Abstracts Vol. 1 [International Conference on AIDS (7th: 1991: Florence, Italy)]

Page  [unnumbered] VII International Conference on AID Florence 16-2 1 June 1991 C I ABSTRACT BOOK VOLUME 1 MXNDAV-TI TF!%DAV

Page  [unnumbered] Sponsor: ISTITUTO SUPERIORE DI SANITA Rome, Italy The "Istituto Superiore di Saniti" (ISS) is a large, governmental research institute involved in public health. It serves as the technical and scientific arm of the Italian Minister of Health. Its role within the Italian health system is to promote public health through research, controls and analytical tests in the different fields of health sciences: infectious diseases, non-infectious pathology, environment, food, drugs, assessment and planning of health services. The Institute collaborates with the Minister of Health in drawing up and implementing public health programs. It is primarily funded by the Italian Treasury but also carries out studies funded by other national and international organizations such as the Consiglio Nazionale della Ricerca "CNR", the World Health Organization and the European Communities. EXPLANATION OF THE LOGO: the statue of DAVID by Michelangelo (16th century) is the symbol of a bright albeit small man (= mankind) who won the battle against Goliath, the frightening giant. To the left, the great Dome by Brunelleschi (15th century), the Baptistry (11th century) and Giotto's "Campanile" (14th century), three jewels of Florentine architecture.

Page  1 Ae IDS flurnc e 102 1 VII International Conference on AIDS Florence 16-21 June 1991 SCIENCE challenging AIDS Sponsor: ISTITUTO SUPERIORE DI SANITA - Rome, Italy

Page  2 CONTENTS A cknow ledgem ents........................................................................................... W elcom ing M essages........................................................................................... 4 Key to Numbering of Abstracts..........................................................................................7 Conference Com m ittes................................................................................................. 8 Conference Personnel.............................................................................................17 ABSTRACTS: Oral Sessions: Monday 17 June.........1..o....................................................19 Tuesday 18 June.................................................................. 53 Poster Presentation: Monday/Tuesday 17-18 June Track A.................................................................................................... 92 Track B..................................................................................................187 T ra c k B........................................................................................................ 1 8 7 Track C....................................................... 298 Track D.............................................................................. 390 The Conference acknowledges the collaboration of the following international agencies in ensuring maximum representation of delegates from developing countries: Centersfor Disease Control Commission of European Communities (AIDS Task Force) Deutsche Gesellschaftf-ir Technische Zusammenarbeit Family Health International/AIDSTECH Fogarty International Center/National Institutes of Health Office of AIDS Research/National Institutes of Health U.S. Agency for International Development

Page  3 ACKNOWLEDGEMENTS Co-Sponsors: Ministry of Health of Italy World Health Organization (WHO) International AIDS Society (IAS) European Economic Community (EEC) Regione Toscana Also sponsored'in association with: Associazione Nazionaleper la Lotta contro IAIDS (ANLAIDS) Associazioneper la Salute della Donna Comune di Firenze Fondazione Italiana per la Ricerca sul Cancro (FIRC) The major contributions of thefollowing Companies are gratefully acknowledged: Bristol-Myers Squibb Company, Rome Farmitalia Carlo Erba S.p.A., Milan Fidia S.p.A., Abano Terme I.B.M. Semea Prodotti Roche Sp.A., Milan Wellcome Italia Sp.A., Rome The Executive Committee of this Conference also acknowledges the generous assistance of: Datamat Ingegneria dei Sistemi Sp.A. Smith Kline & French S.p.A Abbott S.p.A. Istituto Behring Sp.A. Ortho Diagnostic Systems S.p.A Istituto Merieux Sp.A Bruschettini S.r.l. The Fifth International Conference on AIDS and the International AIDS Society (IAS) have generously contributed to the Developing Countries Fund

Page  4 WELCOMING MESSAGES FRANCESCO A. MANZOLI Director General, Istituto Superiore di Sanitd, Rome, Italy It is with great pride that on behalf of the Istituto Superiore di Sanita, I welcome you to the VII International Conference on AIDS. The epidemic has now entered its second decade - a fact that makes us more and more aware of what scientific research has and has not accomplished. The least one can say about the progress made in the past ten years, is that it has been unprecedented. The etiology of the disease has been unveiled, an effective antiviral chemotherapy is being developed and, most importantly, major prevention programs put in place around the world are yielding beneficial results. We know that much of this dramatic progress depended upon scientific discoveries accumulated in the previous years, particularly with respect to the immune system and molecular biology of recombinant DNA. Despite these major achievements, one is reminded that we have neither a cure nor an effective vaccine, and what is more alarming is that the epidemic is still spreading. A more urgent effort is needed to cope with the economic impact of AIDS, both in the developed and developing worlds. The VII International Conference on AIDS has purposely chosen to emphasize that Science, with its multi-faceted connotations, is the way to overcome the AIDS challenge. Discriminatory precautions originated by social and political policies will not work. Only new scientific knowledge will lead to the formulation of rational public policy. Florence was in the past one of the great capitals of science and, more recently, few cities on earth have witnessed the results of the worldwide collaboration and enthusiastic response that, in 1966, enabled Florence to overcome the damage and perils posed by the Arno flood. I trust that the great cultural heritage of Florence and Italy will be conducive in your effort to pave the way to stop the epidemic. Again, welcome here and enjoy both your science and our hospitality.

Page  5 GIOVANNI B. ROSSI* & LUIGI CHIECO-BIANCHI** Co-Chairs, VII International Conference on AIDS On behalf of the co-sponsors, we welcome you to the VII International Conference on AIDS. In our Second Announcement we anticipated that, being held in Florence, this Conference would be different from the previous ones, as we could not aim at ever larger crowds of participants in view of the space restrictions of the city of Florence imposed by its century-old history. We did not realize that the Conference would differ so greatly in the number of early registrations, as we have almost reached our targeted 8,000 attendees two months before the Conference. We feared that the Gulf War might have endangered the flow of registrations and abstracts, we are happy that it was not so. Since 1985 every international AIDS conference has been very valuable in enhancing the spread of new knowledge and the exchange of ideas aimed at controlling the AIDS epidemic. This Program shows that in June all major challenges posed by AIDS in the nineties will be dealt with. In addition to covering all major aspects of the four thematic tracks, the International Steering Committee of the VII Conference has also decided to include non-abstract derived Parallel Sessions regarding topics mainly concerning social and behavioral science. Looking back to San Francisco, we can all be happy that an ill-advised policy restricting free-movement of HIV-infected individuals has now been annulled. The San Francisco Conference led a world-wide effort to change that policy and we all unite in maintaining that the fight against AIDS must not have borders posed for the HIV-infected persons. We are confident that, in Florence, the program will satisfactorily meet the needs of an heterogeneous group of delegates. When large sections of the population were still living in the Middle Ages, Florence led the world through the Renaissance. We trust that its history will help us to bridge the existing gap between disease and cure, and to preserve the sense of unity, solidarity and common purpose that has enlivened the actions of the Steering Committee in shaping the Program. On behalf of all those who shared with us the heavy workload to bring the Florence Conference into reality, we look forward to seeing you in Florence where your collaboration will be rewarded with a smiling, but rigidly punctual organization. * Director, Laboratory of Virology, Istituto Superiore di Sanita, Rome, Italy. ** Director, Institute of Oncology, University of Padua, Italy.

Page  6 WELCOMING MESSA GE FROM FERDINANDO DIANZANI Chair, Program Committee VII INTERNATIONAL CONFERENCE ON AIDS As Chairman of the Program Committee for the VII International Conference on AIDS, I would like to express my gratitude for the diligent efforts of the individuals that have served on our Program Committees, Local and International, as well as the many reviewers who have evaluated all submitted abstracts. Together, we have shared the burden - and the excitement - of the unprecedented demand posed by planning such a large conference. Choosing the theme "Science Challenging AIDS" for this Conference, our International Steering Committee recalled a well-established principle that governs our struggle against diseases: first understand, second plan, third apply. In other words, the momentum of human efforts against diseases must move from Science to Policy and not vice versa. With AIDS, this principle must be implemented on domestic as well as international levels. Scientists must work closely with policymakers in the on-going struggle against AIDS; the role of the scientist is thus to understand and plan; that of the policy-maker is to plan and apply. Therefore, the main goal of the Scientific Conference is to provide a deeper understanding of the factors which affect the disease process - this is achieved through the presentation of new research data accompanied by comprehensive discussion and exchange of ideas. Among these components, new information is crucial for the whole process. Accordingly, the entire scientific program (including plenary, oral and poster sessions) has been planned on the basis of the scientific merit of the submitted abstracts, established by three independent reviewers (one from Italy, one from Europe, and one from overseas), as well as our track committees. Consequently, among the presentors, you will see both old and new faces. Their presence in the program has been established strictly on the quality of the work they submitted, taking into account originality and potential impact. There are a few exceptions, however, involving the plenary sessions: in fact, the presentation of research data to a general audience not necessarily specialized in the particular topic requires a "state of the art" presentation and a brief introduction by the session moderator; this delicate duty was given to particularly qualified and experienced persons selected by the International Steering Committee. The number of oral sessions given to each track reflects the quality of abstracts received in each of these four areas. We have tried whenever possible to avoid sessions with overlapping topics. We hope that individuals interested in a particular subject will be able to attend every session dealing with that topic. The high standards applied to the oral sessions have also been applied to poster presentations. Thus, of the 4700 abstracts submitted, we have been able to accept approximately 2950 for posters. Unfortunately, more than 1000 valuable abstracts could not be included in the program. Writing this now, a few weeks before the meeting, I feel like a film director before a premiere: everything appears to be in place, but...you never know.... I am certain, however, that Florence will offer an ideal background for our work. In fact, the success of our efforts against AIDS may depend on our ability, as human beings, to maintain the ethical and cultural values of our great civilization as we move forward in implementing scientific technology. I am sure that Florence, with its rich artistic and scientific traditions, will provide a perfect setting to spontaneously and immediately humanize the most exasperated technology with the tops of its belfrys and in the shadow of its churches. I hope you will enjoy both.

Page  7 KEY TO NUMBERING OF ABSTRACTS The presentations (oral and poster) are grouped by day and by thematic track as follows: Volume 1: Abstracts for Monday, 17 June, and Tuesday, 18 June. Volume 2: Abstracts for Wednesday, 19 June, Thursday, 20 June, and Friday, 21 June, as well as complete author, chair, and keyword index. The system used to rumber the abstracts identifies the day, track, and type of presentation. An abstract with a number below 1000 is an oral presentation. One with a number above 1000 is a poster presentation. EXAMPLES M. B. 2057 Monday Track B Poster presentation (numbered above 1000) in Volume 1 (Monday/Tuesday) in Volume 2 (Wednesday/Thursday/Friday) Th. A. Thursday Track A 38 Oral presentation (numbered below 1000) DEFINITION OF SESSION TERMS MORNING SESSIONS AFTERNOON SESSIONS SYMPOSIA THEMATIC SESSIONS WORKSHOPS POSTER SESSIONS Plenary Sessions - Monday-Thursday - 9:00-10:30; 10:30-12:00 Friday - 9:00-10:30 Each plenary session (two per morning) will be devoted to a different thematic track, and will consist of a lecture (30-40 min.) by an expert of international stature presenting state-of-the-art findings in his/her field of interest, followed by four 10 -min. presentations of the best astracts submitted. The Moderator will explain briefly to the interdisciplinary audience the general message carried by the abstracts selected. Lectures only (no discussions). Monday-Thursday, 15:00-16:30; 17:00-18:30 A session comprised of strictly related topics (e.g. Neurologic aspects of AIDS, membrane alterations by HIV, etc.). Six presentations of 10 minutes, each followed by a 5 minute discussion. Presentations selected from submitted abstracts. A session aimed at bringing together multiple specialties in a discussion related to a single theme (e.g. laboratory diagnosis of opportunistic infections, models of care delivery, etc.). Six presentations of 10 minutes, each followed by a 5 minute discussion. Presentations selected from submitted abstracts. Six presentations selected from abstracts, each 10 minutes in length, examining a single topic from a multidisciplinary or international perspective; followed by a 30 minute moderated discussion among the participants. Statements by the moderator at the end of the session, highlighting consensus or differences in opinions. Approximately 1500 posters on a wide variety of subjects will be displayed for two 2-day sessions: Session 1: Monday and Tuesday, 17-18 June 1991 Session 2: Wednesday and Thursday, 19-20 June 1991 Authors will be stationed by their posters from 13:00-14:00 on each day of their poster display to discuss their work.

Page  8 CONFERENCE COMMITTEES EXECUTIVE COMMITTEE INTERNATIONAL STEERING COMMIT TEE Giovanni B. ROSSI, Rome, Conference Chair Luigi CHIECO-BIANCHI, Padua, Conference Co-Chair The Members of the Executive Committee and: World Health Organization Michael MERSON Lars Olof KALLINGS Manuel CARBALLO Interhational AIDS Society Paul A. VOLBERDING Meinrad KOCH Friedrich DEINHARDT Gaetano GIRALDO, Naples Paola VERANI, Rome Elke BETH-GIRALDO, Naples European Economic Community (EEC) Alexandre BERLIN Ferdinando DIANZANI Chair, Program Committee Elio GUZZANTI Chair, Communities Committee Antonio SICCARDI Chair, Communications Committee SCIENTIFIC PROGRAM COMMITTEE LOCAL TRACK SUBCOMMITTEES Ferdinando DIANZANI, Rome, Chair Track A: Basic Science Luigi CHIECO-BIANCHI, Padua, Coordinator Ferdinando AIUTI Andr6 E. BAERT Carlo D. BARONI Mauro BENDINELLI Track R Clinical Science & Trials Carlo ZANUSSI, Milan, Coordinator Licinio CONTU Mauro MORONI Pietro PACI Track C: Epidemiology & Prevention Gaetano M. FARA, Rome, Coordinator Piero CROVARI Donato GRECO Adriano LAZZARIN Track D: Social & Behavioral Science Elio GUZZANTI, Rome, Coordinator Leonardo ANCONA Carlo Lorenzo CAZZULLO Antonio FORNARI Lieve FRANSEN Elke BETH-GIRALDO Gaetano GIRALDO Sergio ROMAGNANI Antonio SICCARDI Paola VERANI Marcello PIAZZA Elio Guido RONDANELLI Alberto TERRAGNA (in memoriam) Stefano VELLA Paolo PASQUINI Giovanni REZZA Umberto TIRELLI Alessandro ZANETTI Giuseppe IPPOLITO Mario RACCO Mario SOSCIA Francesco TARONI COMMUNICATIONS COMMITTEE Antonio SICCARDI, Milan, Chair Catherine DASEN, WHO Fabio MALAVASI Gabriele MILANESI Daniela MINERVA, Hypothesis Giuseppe MONTEMURRO Thomas NETTER, WHO Ragnar NORRBY Roberto SATOLLI Umberto TIRELLI Giovanna VIALE

Page  9 DEVELOPING Elke BETH-GIRALDO, Naples, Chair Gaetano GIRALDO Franco Maria BUONAGURO Ranieri GUERRA COUTRES SPONSORED Andrea CAPRARA Agostino MIOZZO DELEGATES COMMITTEE Giuseppe CASTELLO Marina Lina TORNESELLO Renato CORRADO COMMUNITIES Elio GUZZANTI, Rome, Chair Don Luigi CIOTTI, Community Task Force Giuseppe IPPOLITO Luigi CERINA, Community Task Force COMMIT EE Giovanni REZZA Irinus SERAFIN Carlo VETERE COMMUNITY TASK FORCE HIV+ PEOPLE'S COMMUNITY Coordinamento Nazionale delle Persone Sieropositive: Presidenza: Luigi Cerina Sezione Lazio: Gaetano Porcelli Sezione Liguria: Claudio Cappuccio Sezione Veneto: Gianluigi Manfroi Sezione Toscana: Marco Serventi Sezione Lombardia: Melania Petani Co.R.A.- Positifis: Massimiliano Fiorenzi POSITIFS-Italia: Giovanni Nicola Lecce POSITIFS-France: Jean Rene Grisoni POSITIFS-The Netherlands: Dirk Van Graaf UDAVH Unione detenuti affetti da HIV: Fabrizio Faverato ONG AIDS SERVICE ACT UP: Charles Franchino ARCI GAY: Nazionale: Franco Grillini Donna: Antonella Codice ARCI - ORA D'ARIA: Carmen Bertolazzi A77: Maria Luisa Albera Associazione Antigone: Paola Cecchi Associazione Prometeo: Susanna Ronconi Circolo Culturale Mario Mieli: Andrea Pini C.N.C.A.- Coordinamento Nazionale Comunita di accoglienza: Vinicio Albanesi Comitato Diritti Civili delle Prostitute: Pia Covre Cooperativa T.S.R.: Stefano Magini Coordinamento Lesbiche AIDS: Rosaria Jardino Co.R.A.: Marco Taradash Gruppo Abele: Luigi Ciotti I.C.A.S.O. International Council of Aids Service Organizations: Jim L. Holm INFORMAGAY - Gruppo Solidariett Aids: Enzo Cucco L.I.L.A. - Lega Italiana per la Lotta contro l'Aids Nazionale: Vittorio Agnoletto; Beppe Ramina Bologna: Diego Scudiero; Rino Varrasso Toscana: Vera Lazzari LIBERTE: Renata Taddei

Page  10 INTERNATIONAL PROGRAM COMMITTEE Track A. Basic Science Alberto ALBERTINI, Italy Jean-Pierre ALLAIN, USA Alberto AMADORI, Italy Paolo AMATI, Italy Giuseppe BARBANTI-BRODANO, Italy Frangoise BARRE-SINOUSSI, France Jorge BARRETO, Mozambique Arrigo BENEDETTO, Italy Abdellah BENSLIMANE, Morocco Umberto BERTAZZONI, Italy Jay A. BERZOFSKY, USA Gunnel BIBERFELD, Sweden Peter BIBERFELD, Sweden Robert BIGGAR, USA Margherita BIGNAMI, Italy Edoardo BONCINELLI, Italy Margaret BOTTIGER, Sweden Melchiorre BRAI, Italy Samuel BRODER, USA Arsene BURNY, Belgium Raffaele CALIO, Italy Gabriella CAMPADELLI-FIUME, Italy Piero CAPPUCCINELLI, Italy Enzo CASSAI, Italy Luca CECCHERINI-NELLI, Italy Jean-Claude CHERMANN, France Alfredo CHIARINI, Italy Ermanno CICCONE, Italy Massimo CLEMENTI, Italy Giulio COSTANZI, Italy Franco DAMMACCO, Italy Rodney DANIELS, UK Rafael DE ANDRES MEDINA, Spain Eric DE CLERQ, Belgium Anita DE ROSSI, Italy Guy DE THE', France Friedrich DEINHARDT, Germany Sergio DEL GIACCO, Italy Jan DESMYTER, Belgium Antonina DOLEI, Italy Giuliano D'AGNOLO, Italy J. ECONOMIDOU, Greece Jorg EICHBERG, USA Barbara ENSOLI, USA Volker ERFLE, Germany Alberto FAGGIONI, Italy Antonio FANTONI, Italy Manlio FERRARINI, Italy Robin FOA, Italy Guido FORNI, Italy Genoveffa FRANCHINI, USA Luigi FRATI, Italy Herman FRIEDMAN, USA Donato FUMAROLA, Italy George GALASSO, USA Enrico GARACI, Italy Guglielmo GARGANI, Italy Hans GELDERBLOM, Germany Marc GIRARD, France Jaap GOUDSMIT, The Netherlands John GUARDIOLA, Italy David HO, USA Harvey HOLMES, UK Francesco INDIVERI, Italy Phillis KANKI, USA Ronald KENNEDY, USA David KLATZMANN, France S. KOTTARIDIS, Greece Kai KROHN, Finland Reinhard KURTH, Germany Paolo LA COLLA, Italy Michele LA PLACA, Italy Maria Paola LANDINI, Italy Rainer LAUFS, Germany Jay A. LEVY, USA Jean-Paul LEVY, France J. MACHADO CAETANO, Portugal Fabrizio MANCA, Italy Alberto MANTOVANI, Italy Vittorio MANZARI, Italy Malcolm MARTIN, USA Souleymane M'BOUP, Senegal Andrew McMICHAEL, UK James McDOUGALL, USA Thomas MERIGAN, USA Gabriele MILANESI, Italy Therese MULANDI, Kenya James MULLINS, USA Paul NUNN, Kenya Albert OSTERHAUS, The Netherlands Giorgio PALU, Italy Cesare PESCHLE, Italy Giuseppe PIEDIMONTE, Italy Scott PUTNEY, USA Ruy PROENCA, Portugal Mario RICCI, Italy Paola RICCIARDI-CASTAGNOLI, Italy Douglas D. RICHMAN, USA Gert RIETHMULLER, Germany Nicolo RIZZUTO, Italy Robert W. RYDER, USA Nino ROMANO, Italy Helga RUEBSAMEN-WAIGMANN, Germany Alfredo SALERNO, Italy Angela SANTONI, Italy Giuseppe SATTA, Italy Giuseppe SCALA, Italy Geoffrey C. SCHILD, UK Gianpietro SEMENZATO, Italy Frederick SIEGAL, USA Nicola SIMONETTI, Italy Enrico SOLCIA, Italy A. SOW, Senegal Silvio SPADARI, Italy Bruno TAVOLATO, Italy George TEMBO, Uganda M. TERSMETTE, The Netherlands Antonio TONIOLO, Italy Giuseppe TRIDENTE, Italy Adolfo TURANO, Italy 10

Page  11 PierGiovanni TURILLAZZI, Italy Pasquale URBANO, Italy Antti VAHERI, Finland Guido VAN DER GROEN, Belgium Guido VALESINI, Italy Oliviero E. VARNIER, Italy Giancarlo VECCHIO, Italy Giovanna VIALE, Italy Giuseppe VICARI, Italy Track R- Clinical Science & Trials Roberto ACCOLLA, Italy Francisco ANTUNES, Portugal Henry H. BALFOUR, USA Dante BASSETTI, Italy Anita BELMAN, USA B.H. BELHRADSKY, Germany Carlo BERNASCONI, Italy Giovanni BIASI, Italy Francesco BISTONI, Italy Antonio CALI, Italy Federico CALIGARIS-CAPPIO, Italy Onorio A. CARLESIMO, Italy Giampiero CAROSI, Italy Antonio CASSONE, Italy Decio CERIMELE, Italy Francesco CHIODO, Italy Nicola CIANI, Italy Robert COLEBUNDERS, Belgium Dino COLLAVO, Italy Sven A. DANNER, The Netherlands Gholamreza DARAI, Germany Carlo DE BAC, Italy Franco DE ROSA, Italy Stephane DE WIT, Belgium Salvatore DELIA, Italy Franco DI SILVERIO, Italy Manfred DIETRICH, Germany Gianfranco DONELLI, Italy Karl EIHNAUPL, Germany Peter ERIKI, Switzerland Roberto ESPOSITO, Italy K. FELGENHAUER, Germany P.A. FISCHER, Germany Margaret FISCHL, USA Stig FROLAND, Norway Mitchell H. GAIL, USA Josep M. GATELL, Spain Marcello GIOVANNINI, Italy Giuseppe GIUSTI, Italy Frank D. GOEBEL, Germany Fausto GRIGNANI, Italy Claude GRISCELLI, France Karl-Otto HABERMEHL, Germany Robert HEMMER, Luxembourg H. HIPPIUS, Germany Joachim KALDEN, Germany Christine KATLAMA, France Patrick KENYA, Switzerland John KOSMIDIS, Greece Markus VOGT, Switzerland Klaus VON DER HELM, Germany Herman WAGNER, Germany Mark A. WAINBERG, Canada James H. WALSH, Ireland Hans WIGZELL, Sweden Flossie WONG-STAAL, USA Serafino ZAPPACOSTA, Italy Juan Gonzales LAHOS, Spain Joep M.A. LANGE, The Netherlands Ahmed S. LATIF, Zimbabwe Bernard LEBLEU, France Richard N. LEVINE, USA Vincenzo LONGO, Italy Ruedy LUTHY, Switzerland Mario MAJ, Italy Franco MANDELLI, Italy Piero MARCONI, Italy Mario MIDULLA, Italy Julio MONTANER, Canada Richard A. MORISSET, Canada Cristoforo MOROCUTTI, Italy Jens Ole NIELSEN, Denmark Luigi ORTONA, Italy Antonio PACHI, Italy Giuseppe PAPA, Italy Franco PARADISI, Italy Giuseppe PASTORE, Italy Catherine PECKHAM, UK Pehr Olov PEHRSON, Sweden Carla PETTINELLI, USA Anthony PINCHING, UK Jens PINDBORG, Denmark Paola PIVETTI-PEZZI, Italy Lucio POLLICE, Italy Linda RABIN, USA Linda RABIO, Italy Paul RACZ, Germany Annamari RANKI, Finland Robert REDFIELD, USA Antonio RIBUFFO, Italy E.O. RIECKEN, Germany Giovanni ROCCHI, Italy Ernst ROSCAM ABBING, The Netherlands Willy ROZENBAUM, France Armido RUBINO, Italy Bijan SAFAI, USA Adrien Gerard SAIMOT, France Leonardo SANTI, Italy Wolfgang SCHRAMM, Germany Raffaella SCORZA, Italy Franco SORICE, Italy Dario TEDESCO, Italy Glauco TORLONTANO, Italy Sante TURA, Italy Giuseppe VISCO, Italy Alberto VIERUCCI, Italy 11

Page  12 INTERNATIONAL PROGRAM COMMITTEE Robert M. WACHTER, USA Jonathan WEBER, UK Steve WOLINSKY, USA John ZIEGLER, USA Track C: Epidemiology & Prevention Damiano ABENI, Italy Michael ADLER, UK Gioacchino ANGARANO, Italy Francisco ANTUNES, Portugal Laura AYRES, Portugal Courtnay BARTHOLOMEW, West Indies Jean Baptiste BRUNET, France Fernanda BERGAMINI, Italy William BLATTNER, USA James CHIN, Switzerland Nathan CLUMECK, Belgium Roel COUTINHO, The Netherlands Carlo De BAC, Italy Giovanni FADDA, Italy Luigi FRATI, Italy Michael S. GOTTLIEB, USA Angelos HATZAKIS, Greece Dominique HAUSSER, Switzerland Bila KAPITA, Zaire Hailu KEFENIE, Ethiopia Paolo MANCONI, Italy Pier Mannuccio MANNUCCI, Italy Silvio MONFARDINI, Italy Gary NOBLE, USA Koudou ODEHOURI, Ivory Coast Sergio PADERNI, Italy George PAPAEVANGELOU, Greece Giuseppe PAPA, Italy Catherine PECKHAM, UK Carlo Alberto PERUCCI, Italy John PETRICCIANI, USA Peter PIOT, Belgium Phillip PIZZO, USA Francis PLUMMER, Canada Robert REDFIELD, USA Nicolo RIZZUTO, Italy George RUTHERFORD, USA Irinus SERAFIN, Italy Girolamo SIRCHIA, Italy Bertino SOMAINI, Switzerland Enrico TEMPESTA, Italy Giorgio TONIETTI, Italy Salvatore VENUTA, Italy Arduino VERDECCHIA, Italy John WARD, USA Track D: Social & Behavioral Science Vittorio AGNOLETTO, Italy Rosemary ANCELLE, France E. Maxine ANKRAH, Uganda Grazia ARANGIO-RUIZ, Italy Carlo Maria AVIO, Italy Mauro BARNI, Italy Mario BERTINI, Italy Guido BERTOLASO, Italy G. Battista CASSANO, Italy Paolo CATTORINI, Italy Deborah COTTON, USA Anthony P.M. COXON, UK Farzin DAVACHI, Zaire Giorgio DEL MARE, Italy Mohamed-Ridha GHARBI, Tunisia Alberto GIANNETTI, Italy Noel GILL, UK Franco GRILLINI, Italy Alessandro HANGELDIAN, Italy Danielle HANSEN-KOENIG, Luxembourg J.C. JAGER, The Netherlands Lazare KAPTUE, Cameroon Elly KATABIRA, Uganda Yamil KOURI, USA Jeffrey LEVI, USA Ruthanne MARCUS, USA Warren NAAMARA, Uganda Rafael NAJERA, Spain Thomas NETTER, Switzerland Elizabeth NGUGI, Kenya Antonia NOVELLO, USA D. M. OWILI, Kenya Giorgio PARDI, Italy Jostein RISE, The Netherlands William ROPER, USA Claudio RUGARLI, Italy Leonardo TOTI, Italy Carlo VETERE, Italy Vincenzo VISCO-COMANDINI, Italy Bill WHITTAKER, Australia 12

Page  13 REVIEWERS Track A: Basic Science Jan ALBERT, Sweden Adriana ALBINI, Italy June W. ALMOND, UK Francesca ANDRONICO, Italy Edoardo ASCARI, Italy Birgitta ASJO, Sweden A.M. AUBERTIN, France Gabriella AUGUSTI-TOCCO, Italy Maria Francisca AVILLEZ, Portugal Peter BALFE, UK Filippo BELARDELLI, Italy Ascension BERNAL ZAMORA, Spain F. BEX, Belgium Graham BIRD, UK J.O. BISHOP, UK Mauro BOIOCCHI, Italy Dani P. BOLOGNESI, USA Ferruccio BONINO, Italy Enzo BONMASSAR, Italy L.K. BORYSIEWICZ, UK Manfred BOSH, The Netherlands Leo BRADY, UK Rudiger BRAUN, Germany Pierfrancesco BRAVO, Italy Kristina BROLIDEN, Sweden Giuseppe Roberto BURGIO, Italy Maria Jose CAMARASA, Spain A. CANN, UK Franco CAPSONI, Italy Luis CARRASCO, Spain Antonino CASCINO, Italy Anna CATANIA, Italy Paolo Martino CEREDA, Italy Irvin CHEN, USA Cecilia CHENG-MAYER, USA Francesca CHIODI, Sweden C.B. CHRISTIANSEN, Denmark Vittorio COLIZZI, Italy Paolo M. COMOGLIO, Italy Angus DALGLEISH, UK Graham DARBY, UK Raphael de ANDRES MEDINA, Spain Carlo DE GIULI MORGHEN, Italy Patrice DEBRE, France Ronald DESROSIERS, USA Ulrich DESSELBERGER, UK Christian DEVAUX, France Carlo DI BELLO, Italy Ebbe DICKMEISS, Denmark Ursula DIETRICH, Germany Dino DINA, USA Steven DOUGLAS, USA Lawrence DREW, USA Toby DYNER, USA Edgar ENGLEMAN, USA Myron ESSEX, USA Arturo FALASCHI, Italy Anthony FAUCI, USA Vito Michele FAZIO, Italy Hans FEICHTINGER, Austria T. FEIZI, UK Eva Maria FENYO, Sweden Gaetano FILICE, Italy Massimo FIORILLI, Italy Bernhard FLECKENSTEIN, Germany Thomas FOLKS, USA Domenico FREZZA, Italy Patricia FULTZ, USA Robert C. GALLO, USA Roberto GAMBARI, Italy Murray GARDNER, USA L. GATTEGNO, France Richard GAYNOR, USA V. GEORGOULAS, Greece Felice GIANGASPERO, Italy Enrico GINELLI, Italy Colomba GIORGI, Italy Jean-Claude GLUCKMAN, France Felix M. GONI, Spain Andres GONZALEZ MOLINA, Spain Frances GOTCH, UK Peter GREENAWAY, UK Fulvia GREMO, Italy Pasquale GRIPPO, Italy M. GROENINK, The Netherlands R. GRUTERS, France Alberto GULINO, Italy L. GURTLER, Germany Beatrice H. HAHN, USA S. HAMILTON-DUTOIT, Denmark Shinji HARADA, Japan D.A. HARBOUR, UK William HASELTINE, USA S. HEAPHY, UK Dag HELLAND, Norway Martin HEYWORTH, USA Stefan HOGLUND, Sweden Yun-De HOU, China James HOXIE, USA J.G. HUISMAN, The Netherlands D. HUTCHINSON, UK Martino INTRONA, Italy Anna Maria IORIO, Italy 0. JARRET, UK Claude JASMIN, France D.J. JEFFRIES, UK J. JENDIS, Switzerland Klaus Dieter JENTSCH, Germany Lars Olof KALLINGS, WHO Jonathan KARN, UK A.J. KINGSMAN, UK Susan M. KINGSMAN, UK Peter J. LACHMANN, UK Alan LANDAY, USA D.S. LATCHMAN, UK Jeffrey LAURENCE, USA Pedro LAZO, Spain Andrew LEIGH-BROWN, UK Evelyne LENNETTE, USA Norman LETVIN, USA A.M.L. LEVER, UK Giulio LEVI, Italy Bjarne Orskov LINDHARDT, Denmark Ute LINZ, Germany Dan. R. LITTMAN, USA Cecilio LOPEZ GALINDEZ, Spain Paul LUCIW, USA W. LUKE, Germany Fabio MALAVASI, Italy Enrico MARINELLO, Italy Giuseppe MARTINI, Italy Ian McCONNEL, UK Michael McCUNE, USA Steve McDOUGAL, USA C. McGUIGAN, UK Jane McKEATING, UK Leandro MEDRANO, Spain Giovanni Antonio MELONI, Italy Pier Luigi MERONI, Italy F. MIEDEMA, The Netherlands K. MILLS, UK Edward MOCARSKI, USA Karin MOLLING, Germany Luc MONTAGNIER, France J. MOORE, UK Paolo NICOLA, Italy Claus NIELSEN, Denmark Guido NORBIATO, Italy Alberto Emilio PANERAI, Italy Carlo PARRAVICINI, Italy Ronald PENNY, Australia E. Gloria PEREZ-ROJAS, Venezuela J.F. PEUTHERER, UK Mario PEZZELLA, Italy Pier Franco PIGNATTI, Italy Angela PISTA, Portugal Giovanni PIZZOLO, Italy Raffaele PORTA, Italy K. PSARRA, Greece Orsola PUGLIESE, Italy Ileana QUINTO, Italy Lee RATNER, USA Stanley M. ROBERTS, UK Craig ROSEN, USA Filippo ROSSI, Italy Mose ROSSI, Italy E.J. RUITENBERG, The Netherlands Giuseppe SAGLIO, Italy M.P. SCHMITT, France Thomas SCHULZ, UK Claire SHARROCK, UK Lorenzo SILENGO, Italy J.G.P. SISSONS, UK D. SPANDIDOS, Greece Kathelyn STEIMER, USA 13

Page  14 P.E. STEPHENS, UK Fiorenzo STIRPE, Italy J. STOTT, UK Orjan STRANNEGARD, Sweden Maurizio TOMASI, Italy Maria Caterina TURCO, Italy Stefania UCCINI, Italy Dick W. van BEKKUM, The Netherlands David VOLSKY, USA Hagen von BRIESEN, Germany Britta WAHREN, Sweden Simon WAIN-HOBSON, France Bruce WALKER, USA Christopher WALKER, USA Hans WOLF, Germany Naoki YAMAMOTO, Japan H.W.L. ZIEGLER-HEITBROCK, Germany Susan ZOLLA-PAZNER, USA Track R" Clinical Science & Trials Walter ALMEIDA, Brazil Laura AYRES, Portugal J. ALTES CAPELLA, Spain R. ANDREESEN, Germany Jean-Marie ANDRIEU, France Ann ARVIN, USA Michael BACH, USA Alexandra BECKETT, USA Jeanne E. BELL, UK Constance BENSON, USA Timothy BERGER, USA Edward BERNARD, USA Giancarlo BERTONI, Italy Terry BESWICK, USA John BLACK, USA Alicia BOCCELLARI, USA Paola BONARA, Italy Caroline BRADBEER, UK A.M. BRECKENRIDGE, UK H. Reinhard BRODT, Germany Yvonne BRYSON, USA Pedro CAHN, Argentina Aurelio CAJOZZO, Italy Leonard CALABRESE, USA Nicola CANAL, Italy Luigi CARENZA, Italy C. A. CARNE, UK Giuseppe CASCIO, Italy Ercole CATTANEO, Italy David CHERNOFF, USA Carlo CHEZZI, Italy Maria Jose CILLERUELO, Spain B. CLOTET, Spain David COHN, USA Ann COLLIER, USA J. COSIN OCHAITA, Spain Giulio COSSU, Italy Janet DARBYSHIRE, UK Bonnie DATTEL, USA Francesco DE CATALDO, Italy Pier Giuseppe DE FILIPPI, Italy Maurizio DE MARTINO, Italy Claudio DE SIMONE, Italy James DILLEY, USA Juncal ECHEVARRIA DE LECUONA, Spain Joel EPSTEIN, Canada Margaret ESIRI, UK C. ESTANY, Spain Andrea FACCHINI, Italy Mary FANNING, Canada Judith FEINBERG, USA Francisco FERNANDEZ, USA Pasquale FERRANTE, Italy Celso FERREIRA RAMOS-FILHO, Brazil Alvin FRIEDMAN-KIEN, USA Stuart GARAY, USA Mikel Garcia GARCIA, Spain Felice GAVOSTO, Italy Giuseppe GERNA, Italy Fabio GIANNELLI, Italy Valerio GIANNINI, Italy Carlo GIAQUINTO, Italy Paolo GIOANNINI, Italy Jeffrey GLASSROTH, USA Fred GORDIN, USA Deborah GREENSPAN, USA G.E. GRIFFIN, UK Francesco M. GRITTI, Italy Luigi GUGLIOTTA, Italy W. David HARDY, USA M.J.G. HARRISON, UK Harry HOLLANDER, USA Philip HOPEWELL, USA Jennifer HOY, Australia Mark JACOBSON, USA Clotilde JANNUZZI, Italy Joyce JOHNSON, USA F. D. JOHNSTONE, UK Auguste KADIO, Cote D'Ivoire Kayembe KALAMBAYI, Usa Harold KESSLER, USA J. M. KINDELAN JAQUOTOT, Spain G.R. KINGHORN, UK S. KNIGHT, UK Donald KOTLER, USA Susan KROWN, USA F. LAGUNA, Spain J.Gonzalez LAHOZ, Spain H. Clifford LANE, USA P.L. LANTOS, UK Anna Maria LAVERDA, Italy Lindsay LAWSON, Canada M. LEAL NOVAL, Spain Christine LEE, UK Alexandra LEVINE, USA Josep LLORENS-TEROL, Spain Antonio LOPEZ BRAVO, Spain Sebastian LUCAS, West Arica C.A. LUDLAM, UK Giuseppe LUZI, Italy Bruno MARESCA, Italy Gianni MARONE, Italy Deborah MARRIOTT, Australia Henry MASUR, USA Lars MATHIESEN, Denmark Charles MAYAUD, France Kenneth MAYER, USA Patrizio MAZZA, Italy Aldo MAZZONI, Italy Joseph McCORMICK, USA Alistair McLEOD, Canada Maria Jose MELLADO, Spain Fred MHALU, Tanzania Anne MIJCH, Australia Francesco MILAZZO, Italy Lorenzo MINOLI, Italy David MITCHELL, UK Ronald T. MITSUYASU, USA Jacqueline MOK, UK A. Bruce MONTGOMERY, USA Alberto MUSSO, Italy R. NAVARRO, Spain S. NEWMAN, UK Stuart NICHOLS, USA S. Ragnar NORRBY, Sweden Sandra NUSINOFF LEHRMAN, USA Felix OMENACA TORES, Spain Graziella OREFICI, Italy Roberta PACIFICI, Italy Ivo PANNACCIULLI, Italy David PARENTI, USA Luigi PEGORARO, Italy Tim PETO, UK Hans POHLE, Germany Luciano POLONELLI, Italy William POWDERLY, USA Edoardo POZIO, Italy Richard PRICE, USA Tom QUINN, USA Stanley READ, Canada A. REDONDO CAMP, Spain Peter REICHART, Germany Paolo RICCIO, Italy C. R. RIZZA, UK Richard B. ROBERTS, USA Paolo ROSSI, Italy R. RUBIO GARCIA, Spain Michael RUST, Germany Sharon SAFRIN, USA Merle SANDE, USA J. M. SANTAMARIA, Spain Massimo SCAGLIA, Italy Giorgio SCALISE, Italy Guglielmo SCARLATO, Italy Robert SCHOOLEY, USA G. SCOTT, USA Antonio G. SECCHI, Italy F. SEGURA, Spain Gianpietro SEMENZATO, Italy Yuen SO, USA Rosemary SOAVE, USA E. SORIANO, Spain Stephen A. SPECTOR, USA G. D. STERGIOU, Greece 14

Page  15 Graeme STEWART, Australia E. Richard STIEHM, USA Diane STOVER, USA Roberto STROM, Italy John SULLIVAN, USA Richard SWEET, USA Henry TAELMAN, Rwanda Michael L. TAPPER, USA T. THIRUMOORTHY, Singapore Pietro Emanuele VARALDO, Italy J. VERDEJO ORTES, Spain J.L. VILDE, France M. WANSBROUGH-JONES, UK Hetty WASKIN, USA Hilary WASS, Canada Ian D. WELLER, UK Anne WILLOUGHBY, USA Alec WITTEK, USA Constance WOFSY, USA Stephen K. WOO, Canada Robert YARCHOAN, USA Franco ZACCHELLO, Italy Mario ZINGIRIAN, Italy Kathy ZOON, USA Track C: Epidemiology & Prevention Michel ALARY, Canada Susan ALLEN, Rwanda Roy M. ANDERSON, UK Massimo ARCA, UK Andrew AVINS, USA Peter BACCHETTI, USA Leif BAKKETEIG, Norway David BELL, USA Ruth L. BERKELMAN, USA Roberto BERTOLLINI, Italy Ann-Britt BOHLIN, Sweden Margaret BRANDEAU, USA Raymond BRETTLE, UK Susan BUCHBINDER, USA Jordi CASABONA, Spain Euclides CASTILHO, Brazil Willard CATES, USA Mary Ann CHIASSON, USA Thomas COATES, USA Stefan COSTANTINESCU, USA Piera CUNEO CROVARI, Italy James W. CURRAN, USA Luigi DARDANONI, Italy Luis DE LA FUENTE, Spain Santiago DE TORRES, Spain Giuseppe DEL CORNO, Italy Don DES JARLAIS, USA Roger DETELS, USA Klaus DIETZ, Germany Peter DROTMAN, USA Maria EKSTRAND, USA Patricia EVANS, USA Gino FARCHI, Italy Martin FAVERO, USA Oscar FAY, Argentina Manning FEINLEIB, USA Pierino FERRONI, Italy Antoine FLAHAULT, France Patricia L. FLEMING, USA Francesco FORASTIERE, Italy Silvia FRANCESCHI, Italy Don FRANCIS, USA Samuel R. FRIEDMAN, USA Mindy FULLILOVE, USA Massimo GALLI, Italy John GALLWEY, Switzerland Gabriella GAMBA, Italy Jacob A. GAYLE, USA Alain Jean GEORGES, Central African Rep. Julie GERBERDING, USA David R. GIBSON, USA Peter GILL, Canada D.J. GOLDBERG, UK Sheila GORE, UK M. HATZIVASILIOU, Greece Norman HEARST, USA Alan R. HINMAN, USA Mimmo IANNELLI, Italy Carlo IMPERATO, Italy Jose IZAZOLA, USA Harold JAFFE, USA Anne M. JOHNSON, UK Stephen JONES, USA Richard KASLOW, USA Edward KATONGOLE-MBIDDE, Uganda Renata KIEFER, USA Lawrence A. KINGLSEY, USA Jan KRISTOFFERSEN, Norway Fred KROGER, USA Peter LAMPTEY, USA Philippe LEHMANN, Switzerland Christina LINDAN, USA Miwendaweli MABOSHE, Zambia Per MAGNUS, Norway Luigi MAJORI, Italy Yvonne MALDONADO, USA Barbara MARIN, USA M. Jose MEDRANO ALBERO, Spain Paolo MEZZELANI, Italy James W. MOSLEY, USA Alvaro MUNOZ, USA Kenrad E. NELSON, USA Alfredo NICOLOSI, Italy Gloria ORNELAS HALL, Mexico Dennis OSMOND, USA Jean William PAPE, Haiti Fabio PARAZZINI, Italy Khorshed PAVRI, India Felice PICCININO, Italy Vadim POKROVSKI, USSR Rosa POLO, Spain Gaetano PRIVITERA, Italy Arthur REINGOLD, USA Michael REKART, Canada Robert REMIS, Canada Massimo RICCIO, UK Joanna RINALDI, USA Martha ROGERS, USA A. ROUMELIOTOU, Greece Stefania SALMASO, Italy Michael SAMUEL, USA Martin SCHECHTER, Canada Giuseppe SEGNI, Italy Peter SELWYN, USA Giovanni SERPELLONI, Italy Werasit SITTITRAI, Thailand Zena STEIN, USA Rand STONEBURNER, USA Odorina TELLO, Spain Cecilia TIBALDI, Italy Rob TIELMAN, The Netherlands Jorge TORGAL GARCIA, Portugal Pier Angelo TOVO, Italy Giuseppe TURBESSI, Italy Joan M. TUSELL, Spain Ronald O. VALDISERRI, USA Philippe VAN DE PERRE, Rwanda Anneke van de HOCK, The Netherlands Sten VERMUND, USA Andrew A. VERNON, USA John WATERS, USA Laurie WERMUTH, USA Judy WILBER, USA James WILEY, USA Ronald W. WILSON, USA Warren WINKELSTEIN, USA Albert WU, USA Track D: Social & Behavioral Science Donald ABRAMS, USA Sabino ACQUAVIVA, Italy D. AGRAFIOTIS, Greece A. Carlo ALTAMURA, Italy Dennis ALTMAN, Australia E. ANAPLIOTOU-VASEOU, Greece Dennis P. ANDRULIS, USA Peter ARNO, USA Mark ASKEW, Australia Carlo Maria AVIO, Italy Ronald BAYER, USA Ramon BAYES, Spain G. Giovanni BELLOTTI, Italy Marie BERLINGUET, Canada Mario BERTOLINI, Italy Gabriel BEZ, France Mary BOLAND, USA Lydia BOND, USA Flavia BORTOLOTTI, Italy Mary BOULTON, UK Antonio BRENNA, Italy Kenneth BRIDBORD, USA Lawrence BROWN, USA Ernst BUNING, Holland Maria Luisa CANALS, Spain Cipriano CANOSA, Spain Manuel CARBALLO, Switzerland 15

Page  16 Antonietta CARGNEL, Italy Vittorio CARRERI, Italy Francesco CARRIERI, Italy Phillip CARSWELL, Australia Massimo CASACCHIA, Italy Marcello CESA-BIANCHI, Italy Nicola CIAVARELLA, Italy Antonio CONDINI, Italy Molly COOKE, USA Ellen R. Cooper, USA Mary COTTON, USA Fiorenza CRESPI POZZI, Italy Peter DALGLISH, Canada Herbert DANIEL, Brazil William DARROW, USA P.M. DAVIES, UK Don DE GAGNE, Canada Daniel DEFERT, France Livia DI CAGNO, Italy Pierre DIONNE, Canada Gary DOWSETT, Australia Jean Alain DUBOIS, Switzerland James D'ERAMO, USA Pilar ESTEBANEZ, Spain Rita FAHRNER, USA Bernardino FANTINI, Switzerland Rosa FEIJOO AGESTA, Spain Suzy FLETCHER, USA Lieve FRANSEN, Belgium Luigi FRIGHI, Italy Costanzo GALA, Italy Paolo GENTILI, Italy Barbara GERBERT, USA Juan GERVAS, Spain Raanan GILLON, UK Pierluigi GIORDANO, Italy Getachew GIZAW, Ethiopia David GOSLING, UK Ruth GREENBLATT, USA Adriana GUARESCHI CAZZULLO, Italy Felix GUTZWILLER, Switzerland Mary HAOUR-KNIPE, Switzerland Margaret HARDY, USA G.J. HART, UK Michael HELQUIST, USA Sandra HERNANDEZ, USA Subhash HIRA, Zambia Rainer HORNUNG, Switzerland Antoine HOUEHOUGBE, Cameroun Michel HUBERT, Belgium Michael ISKOWITZ, USA Rabbi Yoel KAHN, USA Noerine KALEEBA, Uganda Joel M.R. KASWARRA, Uganda Carole KAUFFMAN, USA Dargut KEMALI, Italy Edward O. LAUMANN, USA Carol LEVINE, USA Alfonso MANGONI, Italy Lucia MANNETTI, Italy Guglielmo MARIANI, Italy Jean MARTIN, Switzerland Vickie MAYS, USA Amir MEHRYAR, Switzerland Silvio MERLI, Italy Gabriella MORASSO, Italy Debhanom MUANGMAN, Thailnad Tsunetsugu MUNAKATA, Japan Laura MUSETTI, Italy N. McKEGANEY, UK Leon McKUSICK, USA Pilar NAJERA, Spain Giuseppe NAPPI, Italy Ibra NDOYE, Senegal Alvin NOVICK, USA H. M. NTABA, Malawi Dan S. OBIKEZE, Nigeria Angela OKOLO, Nigeria Philista ONYANGO, Kenya David OSTROW, USA Mead OVER, USA Fred PACCAUD, Switzerland Augusto PANA, Italy Richard PARKER, Brazil John V. PEGGE, South Africa Luisa PERRONE, Italy Praphan PHANUPHAK, Thailand Michel POLLAK, France Nicola PRINCIPI, Italy Richard RECTOR, Denmark Michelle ROLAND, USA Maria Antonietta ROSCI, Italy Joan ROVIRA I FORNS, Spain Renee SABATIER, Canada Alessandro SALVINI, Italy Carmen SANCHIS PINOL, Spain Joao SANTOS LUCAS, Portugal Giuseppe SARTORI, Italy Ben SCHATZ, USA Neil SCHRAM, USA David SCHULMAN, J.D., USA William SCHULTZ, USA Jaime SEPULVEDA, Mexico Jim SORENSEN, USA Norbert SPECHT, Germany Ron STALL, USA Gerald V. STIMSON, UK Carol SUSSMAN-OLMSTEAD, USA Fredy SUTER, Italy Francisco SY, USA Alberto Giovanni UGAZIO, Italy Anne USHER, Canada John Paul VADER, Switzerland Alberto VAGLIA, Italy Linda VALLEROY, USA Luigi Amedeo VIGNOLO, Italy Robin WEISS, USA Kaye WELLINGS, UK Alfredo ZAMPIERI, Italy Mario ZANETTI, Italy Graziella ZECCA MANSUETO, Italy Debrework ZEWDIE, Ethiopia Daniel ZULAICA, Spain 16

Page  17 CONFERENCE PERSONNEL Giovanni B. ROSSI, Conference Chair Director, Laboratory of Virology Istituto Superiore di Sanita' Rome, Italy Cristina D'ADDAZIO, Executive Assistant to the Conference Chair and Secretary General Laboratory of Virology Istituto Superiore di Sanita Viale Regina Elena, 299 00161 Rome, Italy Telephone numbers: (396) 4457888; (396) 4462331 Telefax: (396) 4453369 The above numbers will be in operation until June 15 1991. CONFERENCE GENERAL SECRETARIAT Rosella BELLIZZI Brigitte NEHRWEIN Deborah WOOL Catia BUSCHITTARI, Administration Stefania DE MENNA Alessandra MARIANI Daniela MARSILI Support: Roberta BERARDI Veronica BIZZOTTI Flavia FEDELI Angela FRESOLONE Maria Rosaria GALATI Cristiana BOROS, Operations Stefano CREMA Alexandra POINTNER Developing Countries Sponsored Delegates Program: Benjamin JUNGE, Coordinator Dominique HAGL Community Relations Laura THOMAS, Coordinator Computer Consultant Marco PANATTA Security Consultant: Ferdinando Ridi REGISTRATION Catharina RAMEL, Coordinator Tina BIANCO Emilia CIATTI Elena FACCIA Mausi FRIGO Simona GERMONI Rita LICATA Angela LONG Federica MAGNANI Vilma SPAGNUOLO Michela TIDEI PROGRAM Lucy FELICISSIMO, Coordinator Ulla JOHANSSON, Assistant Coordinator Cecilia AVICO Alessia CARATELLI Wilma DURANTE Enzo FRAGOLA Cristina QUATTRINI Bonnie ROSE Pamela WOOL 17

Page  18 LOCAL ORGANIZING SECRETARIAT C.S.S. Centro Servizi Segreteria Via A. Lapini, 1 50136 Florence, Italy Tel. (3955) 670182 - 670369 Telefax (3955) 660236 C.S.S. (a congress organization service based in Florence) was appointed by the VII International Conference on AIDS to organize and manage the local organizing secretariat. The specific tasks of C.S.S. were to plan and organize the use of all the congress areas, furnish all local services necessary for the preparation of the Conference, supply all the personnel needed to assist in the daily, on-site management of the Conference, organize and manage the Commercial and Nonprofit Exhibits, and to schedule and manage local tours and social programs. C.S.S. has worked in continuous contact with all local authorities in Florence and in close collaboration with the Conference General Secretariat in Rome to guarantee the orderly and safe running of the Conference. OFFICIAL TRAVEL ALIWEST TRAVEL S.R.L. AGENT Viale Guidoni, 123/i 50127 Florence, Italy Tel. (3955) 2280593 - 2280600 Telefax (3955) 417165 Tlx. 575659 - 572139 ALIWEST TRAVEL S.R.L. (a travel agency in Florence) was appointed by the VII International Conference on AIDS to provide hotel accomodations for Conference delegates. ALIWEST has been actively working for the past two years with Hotels' Management to guarantee adequate housing to delegates, as June is peak tourist season in Florence. A wide variety of other services are also being offered by ALIWEST. MEDIA SECRETARIAT HYPOTHESIS COMMUNICATIONS Via G.G. Belli, 39 00193 Rome, Italy Tel. (396) 3231153 Telefax: (396) 3200604 Relations with the media are handled by HYPOTHESIS Communications, who will also staff the Media Center at the Conference. The Media Secretariat is responsible for the following services extended to accredited media: press credentialing; preparation of reference materials for the press (Advance Media Information; Press Kit; a complete state-of-the-art report on AIDS research); handling of an Interview Assistance Service to help journalists contact Conference delegates. The Media Secretariat will also be responsible for publishing the Conference newsdaily: "Science vs. AIDS". HOST BROADCASTER RAI, Radiotelevisione Italiana, will be responsible for audio and video coverage of the Opening and Closing Ceremonies, the plenary sessions, the press conferences and some of the afternoon simultaneous sessions. In addition, it will be responsible for organization of the international broadcasting center for radio and television. 18

Page  19 s '~

Page  20 SM.C1 RISK REDUCTION AND STABILIZATION OF HIV SEROPREVMAID E AK3MG DUG INJECTORS IN NEW * YORK CITY AND BANGKOK, THAILAND Des Jarlais, Don C.*; Choopanya, K.**; Wenston, J.***; Vanichseni, S.**; Sotheran, J.L.***; Plangsringann, K.**; *Beth Israel Medical Center, New York, New York, USA, **Health Department Bangkok Metropolitan Ad5inistration (BMA), Bangkok, Thailand, ***Narcotic & Drug Research, Inc., New York, NY, USA. Objective: To compare long term and recent trends in HIV seroprevalence and AIDS risk reduction among drug injectors in New York City (NY) and Bangkok (B). Methods: Data on seroprevalence and risk behavior have been collected by the participating researchers since early in the epidemic in both cities. New subjects were 821 IDUs from NY and 601 IDUs from B were recruited in 1989-90. Recruitment was primarily frnm drug treatment programs, with special subsamples of new-to-treatment IDUs in B (n = 259), and not-in-treatment IDUs in NY (n = 271). A standardized World Health Organization questionnaire was used and a blood sanple collected for HIV testing. Statistics: Chi-square and t-tests were used for bivariate corparpisons. Capture/recapture and dynamic population modeling were used to estimate seroconversion rates. Results: IDUs in NY were older, more likely to be female, and more ethnically diverse than in B (all p <.01). 34% of the IDUs in B were HIV+ and 47% in NYC. Seroprevalence was higher in the regular treatment program samples in both cities (40% in B, 50% in NY) than in the new-to-treatment sample in B (27%) or the not-in-treatment sample in NY (42%), both p <.01. 90% in B and 76% in NY reported changing behavior to reduce AIDS risk. Seroprevalence among IDUs in B has been stable for 1 year and in NYC for 6 years. Seroconversions were estimated to be 4-8/100 person-years in B and 2-4/100 person years in NY; seroconversions in NY were concentrated among new injectors. Conclusions: Despite the many cultural differences, large scale behavior change appears to have produced stable seroprevalence with rmderate seroconversion rates in both cities. New injectors and IDUs not in treatment appear to be critical for long term prevention progranming. NOTES M.C.2 THE IMPACT OF STD CONTROL AND CONDOM PROMOTION ON THE INCIDENCE OF HIV IN KINSHASA PROSTITUTES TuLiza Mulivanda*, Manoka AT*, NziLa N*, Way Way*, St.Louis n*/**, Piot P***, Laga M***. * Projet SIDA, Kinshasa, Zaire, **C.0.C.. Atlanta, US, ***Institute of Tropical Medicine, Antwerp, Belgium Obijetives: To determine the impact of an intervention programme, including condom distribution and treatment for STD, on the incidence of HIV and other STD among female prostitutes in Kinshasa. Methods: Prostitutes were asked to return monthly to a "women's clinic" where they were interviewed about their sexual exposure, examined and treated for STD, and where condoms were freely distributed. Every 3 months blood was drawn for HIV serology. Trends in incidence of HIV and other STD, as well as trends in sexual exposure were monitored over time. Results: Until November 1990, 434 initially HIV(-) women have been followed for a total duration of 22 months. The mean follow up rate per month was 76 %, ranging from 72 % to 89 %. Sixty one (14 %) women seroconverted during the follow up period. The incidence of HIV has been declining significantly (see figure) ranging from 18 I/year at the beginning of the intervention to 2.Z/year in the last 3 months of the study. Concurrently all other STD showed a declining trend in incidence. (Figure: bars represent adjusted yearly incidence of HIV per 3, recruitment, all other points are monthly STD incidence rates; STD 2 monitored: N.gonorrhoeae by culture, C.trachomatis by antigen detection, T.vaginatis by direct microscopic exam and genital ulcer as a clinical diagnosis). Regular use of condoms with clients I= increased initially from 4 i at the start of the study to 55 after Sc 6 months, and remained stable afterwards, ranging from 48-62 i per month. All, except 4, women who seroconverted, admitted having i ' used condoms irregularly. Condom use with stable partners remained a0 6-,2 -. 5M 0- low throuout the study ( ) for all women. The mean number of moniM o0ol1oiw-1p clients per week reported, ranged from 5.5 to 6.8 and their was no -to -=,M -Tn O".1i.n significant trend during the follow-up period. Conclusions: The intervention programme with condom promotion and STD treatment led to a significant decrease of HIV incidence in this population. The incidence of STD declined as well, but is still high. Promotion of safer sex practices should therefore continue and be reinforced among this wmnen and their sex partners. NOTES M.C.3 PERINATAL HIV- I TRANSMISSION IN NAIROBI, KENYA: 5 YEAR FOLLOW-UP Datta Pratibha", Embree J*,'*, Ndinya-Aohola JCf, Kreiss J*,"0, Plummer FA*,". aUniversity of Nairobi, Nairobi, Kenya, *"wUniversity of vianitoba, Winnipeg, Canada, $"University of Washington, Seattle, USA. Objective; To determine the perinatal HIV-1 transmission rate and evaluate the associated child morbidity and mortality. Methods: Prospective cohort study of infants born to HIV-1 seropositive mothers (case children) and those born to HIV-1 seronegative mothers (control children). HIV Infection in a case child was defined by a positive Western blot (WB+) at 15 months. Results: Three hundred and sixty-one case children and 366 controls have been followed for a mean of 15 (range 0.5-60) months. By survivorship analysis, 12% cases (HIV status indeterminate) and 6% controls died during the first 15 months of life (RR=2.5, p-.03). Of the 79 case children followed for over 15 months, 39% (31/79) are WB positive. A perinatal transmission rate of 45% was estimated based upon WB criteria and mortality rates. A 24% mortality rate for the first 5 years of life was estimated for infected children. The infected survivors are at greater risk of tuberculosis (6.5% vs 0%, p=.02), failure to thrive (FTT) (68% vs 48%, OR=2.3, 95%CI=.9-5.5), pneumonia (32% vs 18%, OR=2.1, 95%CI=.8-5.4), and otitis media (OM) (26% vs 11%, OR=2.8, 95%CI=1-7.8, p=.06) than the controls. The noninfected case children are also at greater risk of OM (27% vs 11%, OR=3, 95%C1=1.3-7, p=.01) and FTT (63% vs 48%, OR=1.8, 95%CI=.9-3.6) than the controls. Conclusions: Perinatal transmission rate of HIV-1 in this cohort is between 39%-45%. Seventy-five percent of the infected children are likely to survive for at least 5 years before they develop AIDS or die. In addition, we have observed increased morbidity amongst noninfected children born to HIV-1 positive mothers. M.C.4 Muel0c Masalmo', Saracco A", Nicolosi A', Gervasoni C" and Arid C, Angarano G, Vaccher E. Quirino T, Sinicco A, Turbessi G. Costigliola P, Gafa S, for the Italian Partner Study. 'National Research Council of Italy, ITBA - Department of Epidemiology, Milan, Italy; "Institute of Infectious Diseases, University of Milan, Italy. Objective: This cross-sectional and prospective study of 524 women who were steady partners of HIV-infected men was carried out in order to assess the incidence and risk factors of man-to-woman sexual HIV. Methods: The survey considered women, with no risk factors other than that of exposure to their sexual partner, recruited from 19 Italian HIV clinics between 1988-90. The women were interviewed and tested for HIV antibodies, and those who were seronegative were followed-up. The incidence of seroconversions was calculated by the person-years (PY) method. Odds ratios (OR), adjusted by multiple logistic regression, were used for the cross-sectional analysis, and crude relative risk (RR) in the follow-up analysis (95% confidence intervals (CI) were also calculated). Results: The cross-sectional study revealed that 136 (26%) of the 524 women Involved were HIV-positive. Women using condoms for all intercourse (OR=0.2; Cl 0.0-0.7) and those using oral contraceptives (OR=0.4; Cl 0.3-0.7) had less risk. Increased risk was associated with genital infection (OR-5.5; Cl 2.9-10.4), anal intercourse (OR-1.8; Cl 1.0-3.2) and a partner with AIDS (OR=2.5; Cl 1.0-5.5). In the prospective part of the study, 171 (44%) of the seronegative women were followed-up for an average 13 months. No seroconversions occurred among the 26 women abstaining from sexual Intercourse (N-26), although there were 6 seroconversions among remaining 145; giving an incidence rate of 3.6 per 100 PY. Increased risks were associated with anal intercourse (RR-4.9; CI 1.1-22.6) or an AIDS-affected partner (RR=4.5; CI 0.6-31.6). The use of condoms had a protective effect (RR=0.3; CI 0.1-1.5), and there were no seroconversions among the 19 women using oral contraceptives or the 5 women reporting genital infections. Conclusions: The risk of HIV infection among steady female partners of seropositive men is decreased in users of condoms or oral contraceptives, and greater in those who have genital infections, practise anal intercourse or have partner with AIDS. The direction of risk was the same in both the cross-sectional and the prospective study, although the latter was based on a small number of seroconversions and did not allow detailed study of all risk factors. r0's 0 O C" G? CS

Page  21 M.A.5 SMALL REGIONS OF THE ENVAND TAT GENES CONTROL CELLULAR TROPISM, CYTOPATHOLOGY AND REPLICATIVEPROPERTIES OF HIV-1. Cheng-MaeL. Ccili: Shioda, T. and Levy, J.A. Cancer Research Institute, University of California, School of Medicine, San Francisco, California 94143-0128 Objective: To identify genetic regions of HIV-1 that control the pathogenic properties of the virus. Methods: Recombinant viruses were generated between two sequential isolates (HIV-1SF2, HIV-1SFp3) that are highly related at the genomic level and yet display distinct in vitro biological properties. Results: The sequential isolates (HIV-1sF2, HIV-ISFI3), molecularly cloned and sequenced, were found to be closely related at the genomic level (>97% overall sequence homology). However, when compared to the earlier isolate (HIV-ISF2mc), the later isolate (HIV-1SF13mc) replicated faster and to higher titers in PMC, Hut78, MT-4 and Jurkat cells and was highly cytopathic. It induced plaque in the MT-4 cell line. Moreover, HIV-1SF13mc displayed a wide host range; productively infected primary macrophages, the CEM and U937 cell lines, and CD4-negative fibroblastoid cells. Studies with recombinants showed that a 0.48kb StuI/MstlI fragment, encompassing the V3 domain of gpl20, determines cytopathogenicity in PMC and the Hut78 and MT-4 cell lines. This same region also contains the major determinant of macrophage tropism. A 0.39kb MstlI/StyI fragment, encompassing the V4, V5 and CD4 binding domains of gpl20, contains the major determinant for entry into the Jurkat, CEM and U937 cells. A 0.29kb RI/Hind III fragment, encompassing the first coding exons of tat and rev, determines the rate of virus replication in the Hut78, Jurkat and MT-4 cell lines. A single aa change in this region of tat, however, appears to be responsible for the difference observed between HIV-lSF2mc and HIV-1SF13mc. Discussion: Our studies showed that small genetic regions of HIV-1 control biological properties of the virus that correlate with its virulence in the host. Specifically, a single aa change in tat appears to affect the replicative rate of HIV-1. Studies with other site-directed mutants are in progress to identify specific amino acid changes in the envelope gpl20 that determine the different host range properties of HIV-1. NOTES M.A.6 FACTORS GOVERNING HIV-1 PERSISTENCE IN WVI Stevenson. Mario*#, Bukrinsky, Michael 1.*, Stanwick, Trevor L* and Dempsey, Michael P.* *Department of Pathology and Microbiology, University of Nebraska Medical Center, Omaha, NE USA #National Institute for Medical Research, Mill Hill, London The underlying basis for HIV-1 persistence and latency and the nature of the virus reservoir in the course of AIDS remains unresolved. The large number of cells in infected individuals which harbor proviral DNA is difficult to reconcile with low numbers of productively infected cells suggesting a restricted reservoir where a majority of infected cells are non-permissive to a complete HIV-1 replication cycle, Our previous work in vitro has demonstrated that quiescent T lymphocytes are infectible by HIV-1 with equal efficiency when compared to activated lymphocytes, although the virus life cycle is restricted due to an as yet unidentified block to HIV-1 integration. Importantly, the preintegration complex within quiescent lymphocytes retains full capacity to integrate upon subsequent activation of the infected host cell. These in vitro experiments prompted us to determine whether Infected quiescent lymphocytes represent a reservoir for HIV-1 in vivo and whether this reservoir could explain some of the features of viral persistence In AIDS. We have developed a PCR protocol which allows us to distinguish a single copy of the HIV-1 genome and to determine whether viral DNA is in an integrated or unintegrated state. Application of this method to analysis of lymphocytes from a cohort of HIV-1 infected asymptomatic and AIDS individuals demonstrated that a major portion of viral DNA in asymptomatic individuals exists as full length unintegrated HIV-1 genomes. In contrast, the major DNA form in AIDS patients was an integrated provirus. Importantly, in vitro activation of lymphocytes from asymptomatic individuals resulted in efficient integration of viral DNA demonstrating an Inducible virus reservoir. Several lines of evidence, including ligation mediated PCR analysis of patient lymphocyte subpopulatlons depleted of activated cells, specifically identified quiescent T lymphocytes as the source of extrachromosomal HIV-1 DNA. Our data demonstrates the existence of a novel HIV-1 reservoir in infected Individuals. The nature of this novel quiescent T lymphocyte reservoir indicates that HIV-1 latency and Dersistence is aoverned in Dart by events orior to proviral intecration. NOTES a sC C> C. M.A.7 MUTATION IN T CELL EPITOPES OF HIV GAG PROVIDES A MECHANISM FOR ESCAPE FROM IMMUNE CONTROL rnajxxps, Rodney; Rowland-Jones, S.; Gotch, F.; Nixon, D.F.;, Ogunlesi, A.; Edwards, 3.; McMichael, A. J. Institute of Molecular Medicine, John Radcliffe Hospital, Oxford, U.K. Objectives: In longitudinal studies of HIV positive haemophiliacs, three distinct cytotoxic T cell (CTL) epitopes have been identified in Gag. To investigate the mechanism of fluctuation in CTL recognition of these epitopes and to search for escape mutants, we have determined the nucleotide sequence of HLA-B8 restricted Gag epitopes in serial peripheral blood samples. Synthetic peptides based on mutant sequences were then tested in CTL assays. Results: CTL from seropositive haemophiliacs recognised three distinct epitopes of the conserved Gag gene although the dominant epitope varied with time. There was no evidence of permanent lose of epitope recognition but sequence variation within epitopes defined by peptides 17-3 and 24-13 appeared during serial studies. Variant peptides based on these changes were usually recognised in CTL lysis assays. This study provides clear evidence that mutation occurs within conserved HIV CTL epitopes in vivo; accumulation of these variations could provide a mechanism for immune escape. M.A.8 GENOMIC DIVERSITY AND ANTIGENIC VARIATION OF THE V3 DOMAIN AS DETERMININC FACTORS OF HIV-1 PATHOGENESIS Goudsmit, Jaap*; Wolfs, Tom; Kuiken, Carla*; Zwart, Gabriel*; Tersmette, Thijs** Human Retrovinrs Laboratory, University of Amsterdam, Amsterdam, the Netherlands, ** Central Laboratory of the Netherlands Re Cross Blood Transfusion Service, Amsterdam, the Netherlands. pbjecive: To determine the relationship between the diversity of the viral population, the emergence of HIV-1 variants escaping V3 spedfi tibody, the switch in virus phenotype from non-syncytium inducing (NSI) to syncytium inducing (SI), and progression to AIDS. ethods: Sequential serum samples were used as source of viral RNA. Sequential peripheral bood lymphocytes (PBL) were used as sour f viral DNA and for virus isolation. After establishing the biological phenotype of the virus in primary lymphocyte cultures, these vimr oitive cells were used as source of viral DNA as welL Deduced amino acid sequences were used for the production of peptides as antiger probing reactivity with natural antibody. Results: Before V3specific antibodies appear and during the initial antigen and virus replication peak, the V3 coding region shows no or little diversity in RNA and DNA. Subsequent to the rise of antibody to the initially amplified V3 domain, a progressive accumulation 0,primarily nonsilent) mutants is demonstrated. Coevolution of mutants is observed and the largest sequence change in viral RNA occur uring disease progreion, NSI to SI switch and antigenemia Pptides derived from the V3 RNA in serum bind to antibodies in the srun d amino acid changes H308 ->P308 as well as P308>H308 define mutants escaping V3 antibody binding Our data indicate that thi iersity of the virus population decreases prior to the emergence of Svirus. The V3 master sequence that emerges in the first SI isolate Salready found in an earlier lymphocyte DNA sample. 6 months after the NSI-SI switch, this V3 variant has become the predominant on SPBLs as well. The evolution rate of the V3 domain of isolated NSI virus is close to tbe one of SI virus (10.2 nucleoke/sitCeyear) and y during the short period of the switch an increase inV3 evolution rate is observed. G->A transitk is the most frequent point mutatior bserved but does not explain in any case the antigenic switch at position 308. icussion and conclusions: Upon HIV-1 transmission the diversity of the V3 coding sequences is limited most likely because of the expansion of the virus population most efficiently propagated by a given host. Our results suggest that V3-specific antibodies select antigenk variants during the period that NSI virus is isolated. Prior to the phenotype shift from NSI to SI virus, a decrease in diversity of the virus population is observed. Emergence of SI variants coincides with a sudden genotypical shift.

Page  22 1 I~'1

Page  23 R4, o OWA.....................

Page  24 M.A.9 ROLE OF N-LINKED GLYCANS IN THE FOLDING OF HIV-1 ENVELOPE GLYCOPROTEINS. Emmanuel FENOUILLET and Jean-Claude GLUCKMAN. CERVI, H6pital Piti6-Salp6triere, Paris - France. Objective:Our recent results indicate that carbohydrates (CHO) present on mature env glycoproteins (gp) do not play a major role for HIV-1 infectivity (J. Exp. Med. 1989, March; J. Virol. 1990, June). However, sugar analogs like I deoxynojirmycin (dNM) that interfere with the glycosylation process of nascent gpl60 markedly reduce virus infectivity. To resolve this apparent discrepancy we approached the mechanisms by which glycan processing interplays with gp160 tridimensionnal structure. Methods:We investigated the effect of dNM (4mM) on glycosylation, production, bioactivity and immunoreactivity of gp160 produced by recombinant vaccinia virus-infected BHK21 cells. Glycosylation was studied by endoglycosidase analysis. Binding to CD4+ cells was determined by quantitative indirect immunofluorescence using anti-HIV-1 human antibodies. Because it is thought that exposed V3 loop of gp120 plays an essential role in post CD4-binding events, accessibility of V3 to mAb 110-4 (Genetic Systems) was studied by ELISA capture and by immunoaffinity purification. Non-glycosylated gp120 obtained in E. coli (ec gp120) and mannosylated gp160 from recombinant baculovirus-infected insect cells (bcv gp160) were used as controls. Results: dNM did not affect the amount of rgpl60 recovered nor its secretion from the cells. As already described, dNM effect was incomplete, resulting in the production of molecules the glycosylation of which was heterogeneous with respect to their apparent MW and to sensitivity to Endoglycosidase H. Strongly reduced binding to CD4+ cells was noted with rgpl60 produced in dNM treated cells. bcv gp160 bound to CD4+ cells, but not ec gp120. Similarly, dNM treatment dramatically altered (by at least 10 fold) the accessibility of V3 to 110-4, whereas ec gpl20 and bcv gp160 presented unchanged immunoreactivity. Conclusion:. Proper processing of CHO of nascent gp160 is essential for normal folding. dNM profoundly affects the glycosylation, bioactivity and conformation of rgpl60. This may account for impaired HIV-1 infectivity elicited by dNM, while glycans present on the mature molecule are not of paramount importance. NOTES M.A.10 QUANTITATIVE ANALYSIS OF THE BINDING KINETICS AND THERMODYNAMICS OF GP120-CD4 INTERACTION REVEALS A THREE STEP PROCESS AND A NEW MODE OF ALLOSTERIC INHIBITION OF GPI20 BINDING Repke. Heinrich', Shin, Jaekyoon" and Haseltine, William A.' Division of Human Retrovirology, Dana-Farber Cancer Institute and Harvard Medical School, Boston, USA, Department of Biochemistry, Harvard University, Cambridge, USA Obvective: To understand the conformational pertubations of CD4 and gpl20 during the binding process and to ind alternative ways of inhibiting binding, syncytia formation and infectivity. Methods: gpl20 (nlV-1 and HIV-2), soluble and membrane bound CD4/CD4-mutants and immunoadhesins were usd to study association/dissociation kinetics, satuation/displacement kinetics and thermodynamics of binding. Selective internalization of CD4 and CD4 mutants was triggered by changing the biophysical properties in the outer leaflet of the membrane via fusion with glycolipid vesicles. Results: The first step of gpl20-CD4 interaction is followed by an energy-dependent process resulting in a high ' affinuitybinding (KD = 2.6 e-10 M). This step is affected by the proper glycosylation of gp120. A subsequent slow isomerization of the complex leads to an almost irreversible binding (K0: 1.2e-l and involves gp120 binding to an additional site at CD4 which has been identified by fusion proteins and mutagenesis. Using a series of truncations and mutants of CD4, we examined a new mechanism of CD4 self aggregation resulting in complete and selective endocytosis. Surprisgly, a series of peptide derivatives which are supposed to mimic one of the extracellular sites of CD4 involved in this interaction, proved to alter the conformation of the primary binding site of the gpl2-CMD4 interaction (step I and 2). They bind to CD4 and act as allosteric inhibitors of gp120 binding, syncyt formation and infectivity. Conclusions: The CD4-gpl20 (HIV-1 and HIV-) interaction is a three step "induced fit" process resulting in a much h r affinity than previously assumed and alters the conformations of both gp120 and CD4. Under certain circumstances, C4 can aggregate by virtue of interactions in the VI region. CD4 has a new allosteric binding site. There are ligands interacting with this site which inhibit gpl20 binding, syncytia formation and infectivity in the uM range. NOTES. M.A.11 RECOVERY OF C8166 CELLS FROM HIV INFECTION IS DUE TO REVERSIBILITY OF PROVIRUS INTEGRATION? Benedetto Arrigo? Di Caro A.T Ella G.** Garbuglia A.R.*** and Delfini C!** *)Center Virology, S. Camillo Hosp.; **)Ist. Med. Sper. CNR; ***)Ist. Sup. SanitA, Rome Objective: Investigation of CD4 receptor, citopathic effect, viral expression, and inte gration in C8166 cells and in derived clones, surviving to HIV (HTLV-IIIB) infection. Methods and Results: C8166 cells show maximal susceptibility to HIV, resulting in a near-synchronous infection cycle and cell death, within 4-5 days. Only in a few cases, we found rare cells escaping cell death; cell duplication recovered after at least 20 days. These expanded cultures showed low growth rate, normal morphology (no syncytia) and production of large amounts of infectious virus. After 30-35 days, cells were cloned in soft agar, with a very low cloning efficiency (10-3). The main characteristics of isolated clones were: i) absence of CD4 receptor (FACS analysis); ii) presence of soluble CD4 in medium (ELISA method); iii) absence of HIV provirus LTR, gag and env sequences (PCR). All clones could be reinfected by HIV, with the following features: a) inhibition of infection by anti-CD4 MAb; b) all cells produced low amounts of fully infectious virus; c) infected cells normally duplicated without syncytia formation. Such reinfected clones maintained for at least 2 months the virus production, even if progressively declining. One out of 3 persistently infected cultures stopped producing virus and resulted negative for proviral DNA. This non-producer line was still susceptible to HIV. Discussion and Conclusions: Recovery from HIV infection with provirus disappearance and maintainance of susceptibility to HIV reinfection suggest that, in these clones, the integration status of HIV could be reversible. M.A.12 FUNCTON OF THE HIV-1 ENVELOPE CARBOXYL TERMINUS. Gabuzda, Dana, Lever, A., Haseltine, W., and Sodroski, J., Dana-Farber Cancer Institute and Harvard Medical School, Boston, MA, USA. Objective. The HIV-1 envelope glycoprotein carboxyl terminus contains sequences which are essential for efficient viral replication. To determine the function of this domain in the virus life cycle, the effect of mutations in the carboxy terminal 160 amino acids of gp41 on viral replication, envelope glycoprotein synthesis and processing, syncytium for mutation, and incorporation of envelope glycoproteins into virions was investigated. Methods. An HIV-1 envelope expressor plasmid was used to identify and characterize replication-defective carboxyl terminus mutants in transfected Cos-1 and Jurkat cells. The replicative potential of the mutant envelope glycoproteins was tested using viruses produced in Cos-1 cells in a transient anscomplementadon assay. (Helsetheta(,J. Virol. 64:2416-2420,1990). Radioimunoprecipitation was used to study the synthesis, processing, and virion incorporation of the mutant envelope glycoproteins in transfected cos-l cells. Syncytium forming ability was studied in Cos-I cells and Jurkat-Tat cells. Results. Deletions involving the 42 amino acids at the 3' c-terminus reduced cell-free virus transmission to less than 25% of the wild-type level. The c-terminal 160 amino acids were not required for syncytium formation or for incorporation of the envelope glycoproteins into virions. Conclusions. The HIV-1 envelope glycoprotein C-terminus is not essential for syncytium formation or efficient incorporation of envelope into virions. This region may be involved in an early stage in the virus life cycle following CD4 binding and membrane fusion.

Page  25 M.A.13 METABOLIC AND STRUCTURAL EFFECTS OF THE INTERACTION OF HIV WITH PBL AND LYMPHOBLASTOID CELLS CAN BE MONITORED WITH 1H NMR Federico, M.*, Maggiorella, M.T.*, Sulli, N.*, Guidoni. Laura **, Luciani, A.M.**, Rosi, A.**, Viti, V.**, Verani, P.*, Rossi, G.B.* - **Lab.di Fisica e INFN sez. SanitA *Lab. di Virologia, Istituto Superiore di SanitA, Rome, Italy Objectives: To elucidate changes induced in cell membrane organization and in cellular lipid metabolism following HIV infection of susceptible cells, taking advantage of NMR spectroscopy that can be performed on intact viable cells. Methods: PHA stimulated human peripheral blood lymphocytes (PBL), like other proliferating cells such as tumor cells, display a 1H NMR spectrum characterized by signals from fatty acid chains of lipid molecules, mostly triglycerides, present in fluid domains of the plasma membrane.These molecules can be used as natural probes in monitoring membrane modifications following HIV infection. Choline based metabolites are also visible in these 1H NMR spectra. Results: 1H NMR spectra of PBL and lymphoblastoid cells HUT-78 and CEM-ss show a transient decrease of the lipid methylene signal intensity in the early stages of infection, in concomitance with HIV internalization, monitoring a general rearrangement of the membrane structure. Analogous effects were observed a few days after infection, during HIV release by infected cells, as assayed by high reverse transcriptase activity in cell supernatant. Signals arising from choline based metabolites, transiently depressed in the early stages of HIV infection, are strongly enhanced when HIV is replicating inside cells, indicating a possible slowing down of phospholipid synthesis. Discussion and Conclusions: These phenomena, confirmed by experiments on HIV chronically infected cells, seem to be a general feature of the interaction of HIV with susceptible cells. 1H NMR spectroscopy allows to monitor metabolic and structural events of large entity when very few virus particles enter the cells. NOTES M.A.14 NONCOMPETITIVE INACTIVATION OF HIV-1 BY rCD4 AND CD4-Ig. Berkower, Ira, Murphy, Dano, and Smith, Gale, E., CBER FDA, Bethesda, MD; *MicroGeneSys, Inc., Meriden, CT, USA Obiective: To determine how many recombinant CD4 (rCD4) or CD4-immunoglobulin hybrid molecules must bind each virus to cause inactivation. Methods: HIV-1 inactivation by rCD4 and CD4-Ig were measured in a sensitive plaque-forming assay, which gave a quantitative measure of virus survival as a fraction of input virus. The antiviral effects of both agents were blocked by adding excess recombinant gpl20, allowing us to compare virus survival as a function of receptor occupancy by rCD4 over the full range of competitor gpl20 concentrations. Results: In the presence of 20nM rCD4, virus survival increased progressively from.03 to.73 as the fraction of viral gpl20 receptors occupied by rCD4 decreased from.89 to.04, due to competitor gpl20. For dimeric CD4-Ig, virus survival increased from.02 to.91 as receptor occupancy decreased from.89 to.01. Virus survival followed the Poisson distribution, falling linearly as a semilog function of receptor occupancy by rCD4 and CD4-Ig. At 1/e survival, corresponding to 1 inactivation event per virus, just 1/10th of receptors bound rCD4 and 1/50th bound CD4-Ig. Conclusions: These results indicate a functional valency of at least 10 to 50 for viral inactivation by rCD4 and CD4-Ig and support a noncompetitive mechanism of viral inactivation, triggered by rCD4 binding, but not saturating, receptors and suggest that rCD4 may act before or after virus binds to the cell surface. NOTES M.A.15 HIV ENVELOPE CONTAINS A STRUCTURAL MHC CLASS II HOMOLOGY CONSISTENT WITH AN ALLOEPITOPE. Angus Dalgleish, John Habeshaw, Fabrizio Manca+, Elizabeth Hounsell. MRC Clinical Research Centre Harrow HA1 3UJ UK. +San Martino Hospital Genoa Italy. Objective. 1) To show that gpl20 contains an alloepitope which would be available to interact with the T cell receptor (TCR), and to explore if this interaction is with a specific Va/P region of the TCR. 2) To determine if T cell responses against HIV are T cell restricted at the Va/p level. Methods. Modelling of gpl20 carboxyterminal region incorporating homologies with IgG variable regions, predictions of protein secondary structure (Chou-Fasman and Garnier et al algorithms), molecular dynamics and energy minimisation. Raising and characterisation of T cell clones to gpl20 and gpl60. Results. HIV has a surface accessible alpha helical forming sequence which can exist in close proximity to a region of homology with the minor alpha helix of MHC class I and regions of 0 strand motif of class I and class II. This model represents a conformation for part of gpl20 which mimics the binding domain for the TCR on MHC molecule. In addition it has been shown that in vitro T cell lines against gpl20 and 160 are restricted in their use o TCR genes. Discussion. These results are consistent with HIV acting as an alloepitope which could lead to chronic immune activation and T cell deletion similar to that seen in graft versus host disease. Treatment and vaccines should therefore be directed at GVHD as well as the virus. Specific treatment should be targeted at the specific responsive TCR gene, using well known strategies based on T cell vaccination and idiotypy. This approach may also lead to novel drugs designed to inhibit this interaction M.A.16 MM~EOI R MIM=R BETWEEN 1V 1 GP41 AND HLA CLASS I CORRELATION BETWEEN BTHE G EERATION OF CbSSREACTIVE AN'IBODIES, T CELL F,C AND DISEASE PROGRESSIONN I INFECTED INDIVIDUALS. Goldinq, Hana; Clericai, M.; Mann, D. Shearer, G. M. **; BlAgburn, R.* *Division of Virology, FDA; Experimental Immunology Branch, NCI, NIU' "'Viral Carcinogfnesis Branch, NCI, NIH. Bethesda, MD, USA. two highly hamlogous sequenes were found in onmserved regins of HIV 1 gp41 (aa837-844), and human HA class II beta cdains (aal9-2). Early study demnstrated that 35% of HIV 1 infected individuals produce antibodies against the gp41 sequence which crossreact with HLA class II calls (CRAb). uch CRAb uld block in vitro proliferative responses of CD4+ T cells, nd kill class II cells via Ab dependent cellular cytotoxicity (ADCC). A ve (a) To correlate in asymptamatic patients (#CD4 cells>500/ml) the presence of CRAb ith their T cell function. (b) To determine in a retrospective study of HIV 1-infected aemcphiliacs, if production of anti-class II CRAb affect the rate of progression to AIDS. ethods 1) Patients' sera were tested by ELISA for reactivity against the gp41 and class II leived hanologous peptides. 2) PBL from stage 1+2 AIDS patients, were tested for their ability proliferate and produce IL2 in respcnse to recall antigens (flu/tetanus), allogeneic stimuli, d PHA. 3) The #CD4+ cells and clinical status of patients were followed routinely. ts (a) 50% of asymptomatic patients demonstrated a selective early loss of their CD4 ependent T cell resposes to recall antigens, while still responding to PHA and allogeneic mul Interestingly, the presence of CRAb in patients sera correlated closely (p<o.001) with larck of responses to ftetanus. (b) HIV 1 infected hemqphilias producing high titers of RAb early in the disease, progressed faster to full blown disease. sc The two studies in concert, suggest that the early appearance of anti-class II aitoantibodies as a result of the ~mlecular mimicry between HIV 1 gp41 and self HLA antigens, May be a factor in the prognosis of HIV 1 infected individuals. 0> G\ tJ" ýjl

Page  26 , M.A.17 ROLE OF THE LFA-1/ICAM-1/2 PATHWAY OF CELL ADHESION IN HIV INFECTION SButini Luca, Pantaleo G, Poli G, Dayton Al, Fauci AS LIR, NIAID, NIH, Bethesda, MD, USA Obiective: To determine the role of the LFA-1/ICAM-1/2 pathway of cell adhesion in the process of HIV-mediated cell fusion and HIV spreading. Methods: Jurkat cells were transfected with HXB2, a plasmid that contains a full length, infectious human T lymphotropic virus (HTLV)-IlllB. PHA-activated LFA-1 and LFA-1 lobtained from a leukocyte adhesion deficiency (LAD) patient] peripheral blood mononuclear cells (PBMC) were inoculated with supernatants (SN) containing either HIV-1LAV or HIV-2Roo. Anti-LFA-1 (anti-CD18), anti-ICAM-1 and anti-ICAM-2 mAb and rCD4 were added to the cultures of Jurkat cells or PHA-activated LFA-1' PBMC immediately after transfection or inoculation respectively, and readded every two days. Cell cultures were observed daily for syncytia formation and monitored for viral production by assay of culture SN for reverse transcriptase (RT) activity. Results: High concentrations (10 gg/ml) of anti-LFA-1 mAb significantly inhibited HIV replication in both experimental conditions as measured by RT activity. Inhibition of the peak RT activity was 50 + 3% (mean + standard deviation) as obtained from 4 independent experiments on HIV-1 infected LFA-1' PBMC, whereas it was 45 + 5% (mean + standard deviation) from 3 independent experiments on Jurkat cells transfected with HXB2. Lower concentrations of anti-LFA-1 had no significant antiviral activity in both experimental conditions. Similar results were obtained on PHA-activated LFA-1 PBMC inoculated with HIV-2. In contrast, syncytia formation was completely suppressed, even at the lowest concentration of anti-LFA-1 used (1 pg/ml). Anti-ICAM-1 and antiICAM-2 mAb used alone or in combination did not suppress either virus replication or syncytia formation; furthermore, anti-LFA-1 mAb but not anti-ICAM-1/2 mAb cooperated with rCD4 in suppressing HIV infection in both experimental conditions. In LFA-1' PBMC inoculated either with HIV-1 or HIV-2 efficient viral spreading occurred whereas syncytia formation was never observed. Discussion and Conclusion: These results indicate that: a) LFA-1 is required for HIV-mediated cell fusion; b) the partial inhibitory effect of anti-LFA-1 mAb cannot be explained by inhibition of cell-to-cell spreading of HIV; c) the lack of antiviral activity of ICAM-1/2 mAb suggests the existence of an additional ligand for LFA-1 other than ICAM1/2. NOTES M.A.18 Amadori, A.*; Zamarchi, R.*; Veronese, M.L.*; De Silvestro, G.**; Salmaso, L.**; Barelli, A.***; Borri, A.***; Chieco-Bianchi, L* *Institute of Oncology, Padova; **USSL 21, Padova; ***USSL 36, Mestre, ITALY Premise: Much in vitro data indicate the effects of gpl20/CD4 interactions, but no evidence that this phenomenon occurs in vivo has been provided so far. Methods: PBL were analyzed cytofluorographically with different mAb against CD4; CD4 expression and proliferative response to anti-CD3 was studied before and after in vitro pre-culture; the specificity of Ab recovered in culture supernatants (SN) of purified CD4+ cells was studied by Western blot. Results: PBL from most AIDS patients showed selective masking of the CD4 epitope recognized by Leu-3a mAb, which is closely associated with the gpl20 binding site. This phenomenon was strictly associated with HIV infection; a good correlation with the number of circulating CD4+ cells and p24 antigen was seen. In vitro PBL culture for 24-36 h was associated with re-regulation of CD4 expression, as well as an increase in responsiveness to anti-CD3; Ab directed against env proteins were recovered in culture SN of purified CD4+ cells. Conclusions: These data indicate that circulating CD4+ cells from AIDS patients are covered with gpl20/anti-gpl20 complexes, which down-regulate CD4 expression and cell function. The pathogenetic relevance of this phenomenon is thought-provoking, especially in light of the data on anti-CD4-mediated tolerance in mirinp mndpl.NOTES M.A.19 THE ROLE OF IL-6 AND TNF-ALPHA PRODUCTION BY B LYMPHOCYTES IN HIV INFECTION Rieckmann. Peter; Poli, Guido; Fox, Cecil H.; Kehr, John H.; Fauci, Anthony S. Laboratory of Imnmunoregulation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD, USA Objective: To investigate the role of IL-6 and TNF-alpha produced by activated B cells in HIV Infected individuals for their potential to support B cell activation and HIV expression. Methods: B cells were Isolated from the blood or lymph node, cultured in the presence or absence of polyclonal B cell activators or recombinant gpl20. TNF-alpha, IL-6 and immunoglobulin (Ig) production in the culture supematants was measured using specific ELISAs. The HIV inductive capacity of cytokines released by activated B cells was tested in a direct coculture with chronically HIV infected T cell (ACH-2) or promonocytic (Ut) or autologous infected T cells and virus release was measured by RT or p24 antigen assays. In situ hybridization was performed to detect virus expression in the lymph nodes. Results: Both cytokines were spontaneously secreted by B cells from HIV infected patients and could further be induced with gp120, whereas gp120 had no effect on B cells from seronegative donors in which IL-6 and TNF-alpha production was stimulated only with polyclonal B cell activators. Both cytokines were shown to be involved in the spontaneous in vitro Ig production by B cells from HIV infected patients. The specific effect of gpl20 on B cells is independent of CD4 or Fc receptors. A correlation was found between the expression of virus in germinal centers of lymph nodes and the spontaneous capacity of B cells to secrete IL-6 and TNF-alpha. p24 antigen release in autologous coculture experiments of B and T cells from HIV infected individuals was dependent on the presence of activated B cells. Discussion and Conclusions: Spontaneous and specifically (gp120) induced cytokine release by activated B cells from HIV infected individuals is important for the induction of HIV expression and may contribute to the high Ig levels found in many patients. M.A.20 EFFECT OFIN VITRO INFECTION WITH HIV-1 ON THE FUNCTION OF CD4+ ANTIGEN-SPECIFIC HUMAN T CELL CLONES. Romagnani. S.; Macchia, D.; Parronchi, P.; Piccinni, M-P.; Simonelli, C.; Ravina, A.; Mazzetti, M.; Santoni, F.; Maggi, E. Istituto di Clinica Medica 3, Universith di Firenze (Italia) Objective: To examine the alterations induced by HIV-1 infection on the function of helper T cells with particular regard to their involvement in the polyclonal B cell response. Methods: In vitro infection with HIV-1 of PPD- or tetanus toxoid-specific CD4+ human T cell clones (TCC) derived from uninfected individuals. Assessment of HIV-1 infection of T cells by PCR and by the measurement of RT activity and p24 protein in TCC supernatants. Evaluation of the proliferative response, production of cytokines and helper activity for immunoglobulin (Ig) synthesis of HIV-1-infected TCC. Results: Human TCC infected in vitro with HIV-1 showed reduced ability to proliferate and to produce IL-2, IL-4 and IFN-gamma, but they exhibited enhanced ability to produce TNF-alpha. Furthermore, infection with HIV-1 enabled TCC to stimulate the synthesis of extraordinarily high amounts of IgM, IgG, and IgA in both autologous and allogeneic B cells in the absence of any stimulant. This helper effect was mediated by a physical contact between T and B lymphocytes. Conclusion: These data demonstrate that in vitro infection with HIV-1 enables human CD4+ TCC to produce TNF-alpha and to stimulate Ig synthesis by B cells yi an antigennonspecific, MHC-unrestricted, contact-dependent, mechanism. This may explain, at least in part, hypergammaglobulinemia and the other phenomena indicative of polyclonal B cell activation seen in HIV-l-infected individuals.

Page  27 M.A.21 PREVENTON OF HIV-2 AND SIsm INFECTION BY PASSIVE IMMUNIZATION IN CYNOMOL-US MONKEYS. Putkonen. Per*; Thorstensson, Rigmor; Albert, Jan"; Hild, Kerstin*; Biberfeld, Gunne*; Norrby, Eriing*. "Dept. of Immunology, National Bact. Lab., " Dept. of Virology, Karolinska Institute, Stockholm, Sweden. Objectff: To evaluate the efficacy of passive immunization against HIV-2 and SIVam Infection In cynomolgus monkeys (Macaca Fascicularis).I Metod;- Sera with high antibody titers were heat treated and injected intravenously (iv) Into nalve animals six hours before iv challenge with live homologous cell-free virus. In the first experiment four animals received a low dose (3 ml/kg) and three a high dose (9 ml/kg) of anti-HIV-2 serum (from avaccinated monkey). Passively immunized animals together with 7 controls were challenged with 10 animal 1060 of HIV-2sBL-6eeeH& In the second experiment four animals received g ml/kg of anti-SIVsm serum (from a clinically healthy SIVsm infected monkey) and werechalenged along with 6 controls with 1-10 ID60 of SIVsm. Infection was determined by virus isolation, by antibody response and by PCR analysis. Results In the first experiment at a low dose of anti-HIV-2 antiserum one of fouranimals and at the higher dose two of three monkeys showed no signs of infection after the homologous HIV-2ichallenge. All 7 control animals became infected with HIV-2. In the second experiment three of four monkeys pretreated with SIVsm antiserum were protected against a SIVsm challenge whereas all six controls became infected. Conclusion:These results show that passively transferred antibodies can protect against a low dose lentivirus challenge in a non-human primate. M.A.22 HIV ROD INFECTION OF MACAQUES RESULTED IN U3 DELETION IN VIVO Nicol Isabell*, Boussin F*, Vogt G*, Le Grand R*, Montagnier L & Doront D*~. *CEA, CRSSA, DSV/DPTE/SSA, Fontenay aux Roses, France.'Institut Pasteur, Paris, France. Objectives: AIDS-like clinical manifestations were observed in an HIV2-ROD infected macaque (33215). This isolate was reinoculated into macaques to obtain an animal model for AIDS disease. Here, we report the in vivo HIV2-ROD genomic variations in the monkey 33215 and the results of the second inoculation. Methods; Inoculation- In May 1987, we inoculated 10 macaques with HIV2 ROD. Of these 10 monkeys, 2 presented clinical signs of AIDS: 1 died and 1 was sacrified (33215). Then, in february 1989, the 8 resting animals were IV reinoculated with supernatant of human cord blood lymphocytes coculture with PBL of the 33215 infected monkey. In addition, one naive monkey was IV inoculated with this supematant, as a positive control of infection. Seroloical. Virologicaland Clinical Follow-upD Wstem blot, ELISA, neutralizing antibodies, CD4/CD8 ratio, cocultures with human cord blood lymphocytes, and PCR using primers in env and LTR were performed. Partial restriction ma of the 33215 isolate and seuencin: DNA was extracted from 33215 infected MOLT4 clone 8 cells, then amplified by PCR using primers in U3 region from 8412 to 9579 and digested by restriction cndonucleascs. The sequence was obtained by the didcoxy chain termination procedure. Results: 1) PCR analysis of HIV2/33215 LTR resulted in a fragment of 870 bp, in contrast to HIV2 ROD which showed a fragment of 1167 bp. Numerous cndonucleases digestions demonstrated a U3 deletion in HIV2/33215 which was confirmed by sequencing. This deletion includes partly NEF, "NRE-like", Enhancer sequences, but not the Spl sites. 2) After the 2nd inoculation, 8/9 animals were seropositive. Virus isolation could be performed in 6/9 animals. In addition, all the nine animals are found positive using PCR and their CD4+ cells count decreased since the 2nd inoculation. The presence of lymphadenopathy is often observed and skin rash affects chronically one animal. Morever. another monkey died presenting lymph-nodes abnormalities, massive pulmonary infection and atrophic spleen. The infection control monkey shows a severe cachexia. Conclusion: The obtention of a more "virulent form" of the HIV2 ROD after the in vivo passage in a rhesus macaque may provide a clinical immunodeficiency animal model induced by a human retrovirus. This might be an alternative to the HIVl/chimpanzee model for vaccine trials. The appearence of a deletion in U3 might be of importance in regard of the pathogenic power of HIV2, and of the close relationships between the viral negative regulation genes and the pathogenesis. 01~ I~ NOTES NOTES A2MGENERAn ii, ur HiGm AmuNgiS iOtr-iEsLIivE uDreuIVE iNFECTiuu i BY M.A.23 TRANSCOMPLEMENTATION OF A MUTATED REV WITH HTLV-1 REX. Kai Krohn', Kati Hakkarainen*, Einari Aavik*, Steven Dewhurst", Reza Sadaie*** and James Mullins"" 'University of Tampere, Tampere, Finland, "University of Rochester Medical Center, Rochester, NY, USA, "*LTCB, NIH, Bethesda, MD, USA and *""Stanford University School of Medicine, Stanford, CA, USA Background and Objectives: Genetically engineered replication deficient viruses can theoretically be used as attenuated vaccines. In HIV and SIV, the rev protein is known to be necessary for expression of viral structural proteins and for the production of infectious virus. We have determined the ability of HTLV-I rex protein to complement for a non-expressed rev protein in SIV. Methods: Rev defective infectious clones HIV-l(pHXB2 107) and SIVSMM PBj 1.5 were used to transfect Hela, Hela-rev or Hela-rex cell lines. The supernatants were collected at day three and used to infect MT-4 cells, the latter of which are abortively infected with HTLV-I. Northern and western blotting, p24 antigen determination and infectivity assays were used to monitor expression of viral mRNA and proteins as well as infectious virus. Results: Transfection of Hela cells with the rev defective HIV or SIV clones resulted in no virus production. In contrast, both Hela-rev and Hela-rex cell lines yielded moderate or high amounts, respectively, of infectious virus, when transfected with p107 or PBj 1.5. Infection of MT-4 cells with the cell free supernatants from transfected Hela-rev or Hela-rex, but not from Hela cultures, resulted in extremely high production of replicative defective, infectious virus (>109 particles/ml, 105.106 TCIU/ml). Conclusions: We have shown that HTLV-I rex transcomplements for SIV rev.The action of rev is thought to be mediated through an arginine rich motif that binds to the RRE in ENV mRNA. Similar motifs are found in other RNA binding proteins regulating the expression and transport of mRNA. A promiscuous rev-like S sequence PPRPRPRRBXB (P=polar, B=basic, Xanything), present in rev from HIV-I, HIV-I and SIV as well as in HTLV-I, is proposed. M.A.24 VACCINATION AGAINST SIV IN MACAQUES Mills, Kingston H.G.*, Stott, E.J.*, Chan, W.L.*, Taffs, L.F.*, Cranage, M.**, Greenaway, P.**, Thiriart, C.***, DeWilde, M.***, Bruck, C.***, Page, M.* and Kitchin, P.A.*. *NIBSC, Potters Bar, Herts;, U.K. **CAMR, Porton Down, U.K. ***SmithKline Beecham, Rixensart, Belgium. Objective: To assess the ability of SIV vaccines to generate humoral and cell-mediated immunity and to protect against challenge in an SIV macaque model. Methods: Macaques were immunized with 4 doses of inactivated whole virus or recombinant SIV proteins. Humoral immune responses were tested by ELISA, immunoblots, RIPA and neutralization assays, cellular responses by proliferation, cytotoxicity and lymphokine release assays. Vaccinated and control animals were challenged with 10 HID50 SIVac251 32H isolate and tested for infectivity by virus isolation and PCR analysis. Resultst Formalin inactivated SIVy a 32H in syntex adjuvant formulation (SAF-1) protected 100Z animals (groups of ) against homologous or heterologous (SIVdelta' provided by M. Murphey-Corb) SIV challenge but not HIV-2 (HIV-2 SBL 6669, provided by G. Biberfeld). BPL-inactivated SIVac BK28 in Freunds adjuvant protected 3 of 3 animals against SIV 32H challenge. A recombinant SIV gag protein expressed in yeast as p27:Ty-virus-like-particles (provided by British Biotechnology, Oxford) failed to protect against SIV challenge or to prime for anamnestic anti-env antibody response, in the presence of gag-specific T helper cell responses. An SIV r BK28 env/gag recombinant vaccine, comprising gpl50 expressed in vaccinia virus and p7 expressed in baculovirus, in SAF-1 induced consistent anti-SIV T cell responses and high levels of neutralizing antibodies. The outcome of SIV challenge will be reported., Conclusion: Protection against homologous or heterologous challenge is possible using inactivated SIV in different adjuvants. The immunogenicity and efficacy studies with the inactivated virus and recombinant SIV proteins may help in the elucidation of the mechanism of protective immunity and in HIV vaccine design. K Kk '^ '^

Page  28 o M.A.25 MAPPING OF NEUTRALIZING ANTIBODIES ON THE SIV env GLYCOPROTEIN Collignon, C.*, Thiriart C., Langedijk H.*, Meloen R.* Desrosiers R., De Wilde M.", and Bruck, Claudine*. *SmithKline Beecham Biologicals, Rixensart, Belgium, Centraal Diergeneeskundig Instituut, Lelystad, The Netherlands, ***NERPRC, Southborough, MA, USA. Several monoclonal antibodies were derived from Balb/c mice immunized with vaccinia expressed SIV gpl50. These reagents were characterized for reactivity to different forms of SIV envelope protein and to different SIV isolates using radioimmunoprecipitation and Western Blot (WB). Among these, 2 showed a neutralizing activity. Their reactivity to envelope subframents expressed in E. coli was monitored by enzyme imnuno assays (EIA). One of these reagents was shown to recognize an epitope located in the aminoterminal moiety of SIV gpl20. The other one was unreactive to these subfragments when tested in EIA or WB and did not react to denatured gpl50 in Western Blot, although clearly positive in RIPA. Therefore the target epitope of this neutralizing mAb appears to be conformationnaly determined. The E. coli expressed subfragments of SIV env were furthermore injected into rabbits and one of them encompassing the carboxyterminal half of SIV env gpl20 was able to induce a strong neutralizing response in these animals. These results allow us to define at least 3 neutralizing epitopes on SIV env gpl20. Further mapping of these epitopes using the pepscan system will be presented. NOTES M.A.26 CHANGES IN THE TRANSMEMBRANE PROTEIN ARE ASSOCIATED WITH CYTOPATHIC M.A. EFFECTS OF SIV LaBranche. Celia C.*; Bugelski, P.J."; Hart, T.K.**; Hoxie, J.A.* *Univ.of Pennsylvania, Philadelphia, PA, "Smith-Kline Beecham, King of Prussia, PA, USA. tbiective: To identify the determinants of cytopathogenesis in a cytopathic variant of molecularly-derived SIVMACZ51. lethods: SIVMAC251 has been shown to infect SUP-T1 cells without killing, fusion, or CD4 down-modulation. A ytopathic variant of this virus arose spontaneously during long-term culture. The parental and variant viruses were ompared by time-course analysis of equivalent inocula onto SUP-T1 cells. Modulation of CD4 from the surface of ifected cells was determined by flow cytometry (FACS) using monoclonal antibodies (mAb) to CD4. The presence I gpl 20 (SU) on the surface of infected cells and the ability of SU to bind CD4 were determined by FACS using olyclonal antiserum to SIVMAc and fluoresceinated CD4 (F-CD4), respectively. Viral proteins were analyzed by SDSAGE and western blot or radioimmunoprecipitation of viral and infected cell lysates. esults: In contrast to prototypical strains of HIV-1 and HIV-2, SIVMAc2s derived from the infectious molecular clone K28 (NC-MAC) infected SUP-T1 cells slowly and without cytopathic effect or down-modulation of CD4. However, ie cytopathic variant (CP-MAC) exhibited a cytopathic phenotype in SUP-T1 similar to that of HIV-1 and -2: infectior as rapid and down-modulated CD4 from the surface of infected cells. Western blots using a mAb to the ansmembrane protein (TM) showed that the TM of CP-MAC migrated at 28 kDa, as opposed to 32 kDa for NC-MAC. his mAb also identified a 77kDa species, likely a stable TM dimer, in NC-MAC but not in CP-MAC. F-CD4 bound to U expressed on the surface of CP-MAC-infected cells with an affinity similar to that observed for HIV-1; this is )ntrary to published reports showing a reduced affinity of SIVMA SU for soluble CD4. Electron micrographs ivealed that CP-MAC virions possessed a significantly greater number of envelope spikes as opposed to NC-MAC rions. onclusions: The cytopathic CP-MAC expresses a TM protein that is smaller than that of the NC-MAC. While other ictors may also be involved, changes in TM may profoundly affect processing and stability of the viral envelope id the cytopathicity of the virus. In addition, because these two viruses were derived from a single infectious lolecular clone, sequence comparison will allow us to identify the genetic determinants of cytopathogenesis for P-MAC in SUP-T1 cells and establish their bioloaical effects. NOTES K KJ,^ K1 Cc M.A.27 C-sis/PDGFB IS TRANSCRIBED IN PRIMARY LYMPHOCYTES EXPOSED TO ETLV-I: A POSSIBLE ROLE FOR THE PDGF AUTOCRINE CIRCUIT IN THE CONTROL OF PROLIFERATION IN INFECTED CELLS D'Onofrio, Chiara; Faraoni, Isabella; Macchi, Beatrice; Bonmassar Enzo Dept. Exp. Med. Biochem. Sci., II University of Rome, Rome, Italy. Objective: PDGF has been described not to be espressed in T lymphocytes under physiological conditions. More recently, PDGFB and its receptor have been found to be expressed in T cell lines infected with HTLV-I (D'Onofrio et al., J. IFN Res.9(2s):S82, 1989; Goustin et al., Growth Factors 2:189, 1990). These data suggest that the c-sis protoncogene/PDGFB, a highly mitogenic factor, could play a main role in the control of proliferation of virus-infected cells, possibly contributing to cellular transformation. To elucidate this point, HTLV-I mediated expression of c-sis/PDGFB was studied in primary cord blood-derived momonuclear cells (CBMC) and in CD4+ and CD8+ isolated T lymphocyte subsets. Methods: CBMC were isolated by Ficoll-Paque gradients and further separated, depending on their phenotype markers, by immunomagnetic microbeads conjugated with specific monoclonal antibodies. Amplification of c-sis in the genome of CBMC, cocultured with lethally irradiated MT-2 virus-donor cells, was evaluated by Southern blots. Expression of c-sis was evaluated as mNA transcription during culture time following exposure to HTLV-I (Northern and dot blots) and as indirect immunofluorescence for the PDGFB protein in infected cells (double immunofluorescence staining). Results: No amplification of c-sis/PDGFB was observed in CBMN/MT-2 cocultures, 1 week post infection (p.i). This gene was not transcribed in early hours p.i., whereas mRNA transcripts were present 5 days p.i. Similarly, c-sis/PDGFB was transcribed in cocultured CD4+ and CDo+ subsets, 1-2 weeks p.i. Interferon-p, which has been previously shown to inhibit early transcription of HTLV-I in infected cells, down-regulated the expression of c-sis in cocultured cells. Discussion and Conclusions: c-sis/PDGFB can be expressed also in cells of the lymphocytic lineage, not only in monocytes, platelets and progenitor cells, as previously known. Its expression is associated with exposure of lymphocytes to HTLV-I and is down-regulated by interfer - P, which in turn can inhibit infection of recipient cells. The role of PDGFB in the control of cell cycle progression in infected CBMC is presently under evaluation. Supported by the Italian Miniscery of Public Health, I 0nAJkrn D - - - - 1% sr M.A.28 COMPARATIVE ANALYSIS OF HTLV-I ENVELOPE SEQUENCES Calabro' Maria Luisa, Hoad J.H., Matutes E., Catovsky D., Weiss R.A. n Schulz T.F. Institute of Cancer Research, Chester Beatty Labs., London, England. Objective: Comparative analysis of HTLV-I envelope sequences derived from ATL cases of different countries was perfomed in order to evaluate their genetic distance from prototype sequences and to identify amino acid changes localized in immunogenic regions of the envelope. Methods: PCR-amplified envelope regions of 5 Brazilian, 2 Caribbean and 1 Romanian cases of ATL were cloned in Bluescript and sequenced by dideoxy chain termination method. Nucleotide changes were confirmed by direct sequencing. Results and discussion: The data confirmed that the HTLV-I env sequence is highly conserved among strains from different geographical regions; however, none of the env sequences was completely identical. The sequence from the Romanian patient, in spite of its different geographical origin, was not significantly more divergent than any of the others. Sequence analysis revealed two changes in a region (aa176-209) which has previously been shown to contain a linear B-cell epitope recognized by most patient sera. In order to assess whether these changes would affect antibody binding to this epitope, we expressed aa 157-225 from two sequence variants in E.Coli. We could show that monoclonal antibodies to this epitope exhibited differential reactivity to these sequence variants. This is the first demonstration of envelope epitope variability of this highly conserved human retrovirus. s; Cf c> ^i It

Page  29 M.A.29 COMPARISON OF HTLV-I/II INFECTION BY SEROLOGY & PCR D. Smith*, T. Mann, E. Anderson, Michel Canavaguio & H. Lee; Abbott Laboratories, North Chicago, IL, *Blood Center South East Louisiana, New Orleans, U.S.A. Objective: HTLV-I EIA screening assays and the confirmatory procedures of Western blot (WB) and radioimmunoprecipitation (RIPA) show reactivity with antibodies to HTLV-I and the closely related HTLV-II. Patterns of serologic reactivity were compared to HTLV-I or II PCR results. Methods: 181 serologically confirmed samples were evaluated by PCR using a primer pair in the tax/rex region to discriminate HTLV-I or II infection. To study potential HTLV heterogeneity, a subset of 76 specimens was further tested using multiple primers to amplify either homologous or unique sequences of HTLV-I or II. Results: 62/181 (34%) specimens were HTLV-I, 90/181 (50%) were HTLV-II and 29/181 (16%) were PCR negative. Of the HTLV-I specimens, 95.2% (59/62) had p19 on WB. However, HTLV-II samples showed a more hetergeneous pattern: 43/90 (48%) with p24 only on WB and 47/90 (52%) with both p19 and p24. For most of the samples, HTLV infections were confirmed by amplification of tax/rex, env p21, and gag p19 regions. For 56/62 samples, the PCR reactivity of different primers was generally equivalent. Several samples reacted only with selected primers: 1 HTLV-I and 1 HTLV-II sample had weak tax/rex reactivity only. Three HTLV-II infected DNA's lacked tax/rex reactivity and 1 HTLV-II sample showed only env p21 amplified sequences. Conclusion: Because current serologic methods for HTLV show cross reactivity for HTLV-I and II, discrimination of these closely related viruses should be determined by PCR or peptide specific assay. NOTES M.A.30 Loss OF ANTIBODY REACTIVITY AGAINST CYTOMEGALOVIRUS DURING ASYMPTOMATIC AND LATE PHASES OF HIV-I INFECTION Fehniger. Thomas E. *, **,; Converse, P.*,**; StrannegArd, O.**;Ehrnst, E.**; Maasho, K.*; SbnnerbOrg, A.*; Britton, S.**Karolinska Institute, Dept. of Infectious Diseases,**Central Microbiology Laboratories, Dept. of Virology, Stockholm, SWEDEN Objective: To define the patterns of antibody reactivity against CMV in HIV-1 infected subjects. Methods: The study group of CMV seropositive subjects included HIV+ clinically asymptomatic (asym: n=13), ARC/AIDS (n=7), and HIV- homosexual and non-homosexual adult subjects. IgG antibody reactivity was measured against intracellular and extracellular CMV antigens. Results: Regardless of their clinical status or peripheral T-lymphocyte values HIV+ subjects showed significantly lower reactivity against CMV nucleocapsid antigens and an antigen associated with virion particles of mol. wt. 28 kd. which were recognized by all CMV+ HIV- controls. Equivalent antibody reactivity was observed for all HIV-/HIV+ subjects with CMV infected fibroblast nuclei, however all HIV+ also showed high reactivity with uninfected fibroblast nuclei. The loss of antibody to the CMV nucleocapsids and 28 kd was most pronounced in asym HIV+ nonhomosexual subjects, regardless of their HIV-1 infection clinical status. Conclusions: Selective losses in specific anti-CMV reactivity occur frequently in subjects following HIV-l infection and before general changes in immune status are apparent. Such changes could be indicative of developing lesions in specific immune function and could contribute to the high incidence of clinical CMV infection in HIV+subjects. NOTES M.A.31 INTERACTIONS OF HIV AND HPV IN THE DEVELOPMENT OF PENILE CANCERS Tornesello, Maria Lina*; Beth-Giraldo, E.*; Galloway, D.A.**; S McDougall, J.K.**; Buonaguro, F.M.*; Giraldo, G.*. * Natl. Cancer Institute "Fond. Pascale", Naples, Italy; ** Fred Hutchinson Cancer Research Center, Seattle, USA. Objective: Analysis of the interactions and the role of HIVs and HPVs in the etiopathogenesis o penile cancers from patients of African regions with high prevalence of HIV infections and genital neoplasms. Methods: To test the prevalence of HIVs and HPVs, DNAs extracted from Penile Cancers (PC) biopsies were tested by Dot and Southern blot analysis at high (Tm-28) and low (Tm-50) stringency, and by Polymerase Chain Reaction (PCR) focused on the HIV-1 gag region and the HPVs E6/E7 and L1 regions. Furthermore a genomic library has been constructed. Results: 15 PC samples, tested for the presence of HPV 6, 11, 16, 18, 31, and 33 DNA sequences, are mainly positive at low stringency (Tm-50) with an HPV 16/18 DNA probe, suggesting the presence of HPV strains related to them. The presence of HPVs has been also identified by PCR. The HPV present in PC8 has been further characterized by the construction of a genomic library and a partial recstriction map. HIV sequences have not been identified in these PC biopsies by PCR. Discussion and Conclusions: A higly pathogenic HPV or a cofactor such as immunodeficiency and/or circulating HIV-related regulatory factors could determine the high incidence of genital cancers in these regions. DNA sequence analysis of the E6/E7 region of the cloned HPV, the in vitro transforming activity and the role of HIV regulatory genes (particularly tat) are the current ongoing approaches. The specific results will be discussed. Supported by the ICSC WORLD LABORATORY and ISS-ITALIAN AIDS PROGRAM M.A.32 -6 AS A POTENTIAL COPACTOR IN AIDS: TRANSCRIPTIONAL ACTIVATION OF CD4 S- AND INDUCTION OF SUSCEPTIBILITY TO HIV-1 INFECTION IN CD8+ T LYMPHOCYTES. Lusao. Paolo*, Malnati, M.**, De Maria, A.***, Lori, F.*, Garzino-Demo, A.*, Arya, S.*, Gallo, R.C.* *LTCB, NCI; **LIG, NIAID; ***LIR, NIAID, Betheada, MD. Obiectivei To investigate the possible role of HHV-6 as a cofactor in AIDS and the interactions between HHV-6 and HIV-1. Methods: Limiting dilution cell cloning, magnetic bead separation, fluorocytometry, Northern blot, PCR, p24 Ag capture, PFA inhibition, DEAE-Dextran transfection. Results: Productive HHV-6 infection was observed in both CD3+CD4+ and CD3+CD8+ normal human T-cell populations and clones. HHV-6 significantly upregulated CD4 expression in CD4+ T cells and induced de novo expression of CD4 mRNA and proteins in CD8+ T cells. In contrast, the CD3/TCR complex was transcriptionally downregulated. Immediate early HHV-6 genes were responsible for CD4, but not CD3 regulation, as demonstrated using the viral inhibitor PFA. The HHV-6-induced CD4 enabled HIV-1 to penetrate and replicate in previously non-susceptible, normal CD8+ T cells. Analogous results were obtained on cells of different lineage origin. Data will be presented from current studies aimed to identify and clone the HHV-6 gene(s) responsible for the transcriptional activation of CD4. We are also assessing the percentage of circulating T cells that coexpress CD4 and CD8 in HIV-infected patients and whether these cells harbour HHV-6 and/or HIV-1. Conclusions: These data indicate that mature CD8+ T cells can be induced by a naturally occurring agent to reexpress CD4 in their post-thymic life. Since CD4 is the receptor for HIV-1, HHV-6 may thus broaden the cellular host-range of HIV-1, favoring its spread in coinfected patients. These findings are consistent with the CD8+ T-cell defects documented in AIDS and further support the concept that some HHV-6 strains may enhance the pathogenic effects of HIV-1, thereby accelerating the disease progression. K

Page  30 S M.B.33 IDENIFICATION OF A 90K OR ASSOCIATED ANTIGEN IN HI INFECTION lacobelli S., Natoli C., Tinari Nicola, Ghinelli F.*, Sighinolfi L.*, Bicocchi R.*, Ortona L.**, Antinori A.**,Tamburrini E** Cattedra di Oncologia Medica, Universita' di Chieti, *Divisione Malattie Infettive, Arcispedale S. Anna, Ferrara, and **Istituto Malattie Infettive, Universita' Cattolica, Roma - Italy. A 90.000 molecular weight protein (90K), detected by a monoclonal antibody has been identified in the sera of patients with various malignancies. We recently observed 3 cases of Kaposi's sarcomas who had the highest 90K levels yet recorded in our laboratory. This prompted us to investigate this antigen in HIV infection. We report here the results of a cross-sectional study performed on 339 blood samples from HIV infected individuals. Serum 90K levels, measured by an IRMA procedure, were significantly higher in AIDS (28.5~21.3 U/ml) than in ARC (21.7-13.1 U/ml; p=0.02) and asymptomatic patients (15~11.8 U/ml; p<0.0001). A positive (>11 U/ml) 90K value was observed in 60%, 83%, 88% of asymptomatic, ARC, and AIDS patients, respectively. Serum 90K correlated with CD4+ lymphocyte (r= -0.43; p=0.006) and serum 82 -microglobulin (r= 0.53; p=0.0001). The levels of 90K were significantly (p<0.02) lower in patients treated with Zidovudine. These results have been confirmed in a retrospective study on a group of 17 HIV infected patients which had been followed up for two years. Though preliminary, these data suggest that 90K may represent a further serologic marker for monitorina the course of HIV infection and the response to therapy. NOTES M,.3 4 ACID HYDROLYSIS INCREASES THE SENSITIVITY OF p24 ANTIGEN DETECTION FOR THE EVALUATION OF ANTIVIRAL THERAPY IN ASYMPTOMATIC HIV-1 INFECTED INDIVIDUALS. Bollinger. Robert C*; Kline, R**; Francis, H*; Moss, M*; Bartlett, J.G.*; Quinn, T**. *Johns Hopkins University, Baltimore, MD; **LIR, NIAID, Bethesda, MD. Objective: The sensitivity of p24 antigen detection in asymptomatic HIV infected individuals is decreased by the presence of HIV-specific antibody. We examined the effect of acid pretreatment on the sensitivity of antigen detection in asymptomatic HIV infected patients in order to evaluate the effect of AZT on antigenemia. Methods: Plasma was obtained prospectively from asymptomatic HIV infected individuals who were randomized to receive AZT (n=25) or placebo (n= 17), as part of the ACTG 019 trial. Coulter HIV p24 antigen ELISA was performed at entry and after 4 months of therapy. Results of specimens pretreated with acidification were compared to the standard assay. Results: Acid treatment resulted in a 4-fold increase in specimens considered positive for p24 antigen (9.9% to 48.6%). Of those individuals who were initially p24 positive, 25.0% of the placebo group and 33.3% of the AZT group became antigen negative at 4 months. Although, analysis revealed that this trend toward increased antigen clearance with AZT was not statistically significant due to the small cohort, untreated specimens revealed no such trend. Conclusions: Acidification of specimens greatly increases the sensitivity of the p24 antigen assay in asymptomatic HIV infected individuals and may improve the utility of p24 antigen, as a marker for therapeutic efficacy. T cell subset phenotyping and investigation of the quantitative measurement of p24 antigenemia after acid pretreatment in this cohort is currently underway. NOTES ito G\ M B 5 SERIAL CD4 COUNTS IN RELATION TO THE DECISION TO TREAT ASYMPTOMATIC HIV-1 INFECTION 5 " Transfusion Safety Study (TSS) Group represented by Aledort, Louis M; Mount Sinai Medical Center, New York City, NY and other participating institutions in New York City, NY, Miami, Fl, Detroit, MI, Seattle, WA, and San Francisco and Los Angeles, CA, USA. A rapid, steady decline in CD4 usually precedes the onset of AIDS. This suggests that a large change between visits may indicate imminent progression, and identify the asymptomatic persons most likely to benefit from anti-retroviral therapy. We tested the hypothesis by determining the individual and combined predictive value of 2 CD4 values at 6-month intervals for a third count 6 months after the second. The analysis is based on data for 149 anti-HIV-1(+) mates with treated congenital clotting disorders who had 3 successive CD4 values approximately 6-months apart (CD4s, CD42nd, and C043,d) during 1986-7, with both CD4,ts and CD4a2nd 200 cells/A/L. Counts were determined by standardized methods and monitored by intra- and interlaboratory quality control methods. we found for this cohort that there was an average annual decrease of 9.5% regardless of the outset level above 200. CD418t and CD4,d were individually highly predictive of CD4rd. Taking into account the change between CD41t and CD42 d did not add to the predictive value of either with respect to CD43rd. The table shows the wide variability between counts 6 months apart. CD43rd CD42nd <200 200-299 300-399 400-499 500-699 700+ 200-299 4 8 4 2 0 0 300-399 3 8 8 5 2 0 400-499 1 2 9 11 8 1 500-699 0 5 4 8 13 7 700+ 0 0 1 3 13 19 Short-term changes in CD4 are too variable to be used in predicting rapidity of progression. A proportionately large change during a 6-month period, therefore, is not a good basis for deciding to begin anti-retroviral therapy. (Supported by NHLBI Contracts No. N01-HB-4-7003 and 9-7074.) M.B.36 HUMAN HERPES VIRUS 6 AND MARKERS OF DISEASE PROGRESSION IN HIV - POSITIVE MALES. Fairfax, Marilynn; Schacker. Timothy; Cone, Richard; Collier, Ann; Corey, Lawrence. Departments of Laboratory Medicine and Medicine, University of Washington School of Medicine, Seattle, Washington, U.S.A. Objective: To investigate the natural history of human herpes virus 6 (HHV-6) in HIV-infected males by comparing the amount of HHV-6 DNA in blood and saliva with stage of disease, T cell subsets, anti-viral therapy, and serology Methods: Blood and saliva were collected from 32 HIV-positive males. Semi-quantitative polymerase chain reaction (PCR) was used to detect the amount of HHV-6 in saliva and in peripheral blood mononuclear cells (PBMCs). T cill subsets and HHV-6 serology were determined simultaneously and correlated with stage of disease, antiviral therapy, and nature of opportunistic infections. Results: A subset of 18 patients had HHV-6 DNA detected in saliva. The amount of HHV-6 DNA in PBMCs did not correlate with total lymphocyte count, T cell subsets, stage of disease, HHV-6 antibody titer, or antiviral therapy. HHV-6 was detected in PBMCs of 57.9% of subjects with a CD4 count < 400 and in 100% of subjects with CD4 > 400. HHV-6 DNA in PBMCs was directly correlated with CD4 cell count (R=.74 and P <.001). Linear regression analysis showed the relationship between CD4 cells and the presence of HHV-6 genomes was strong (P< 0.001) while the relationship was not significant when comparing genomes of HHV-6 to total lymphocytes (P -.57). Discussions and Conclusions: The frequency and amount of HHV-6 DNA In PBMCs was highest in persons with CD4 cell counts > 400 and was lowest (even when corrected for the amount of cellular DNA in host cell PBMCs) irl subjects with advanced disease. Prospective studies of the role of this herpes virus in CD4 cell depletion are warranted. btl ~iCI C> C>~ ii C/)

Page  31 M.B.37 NEARLY ALL DEATHS IN PATIENTS (PTS) ON ANTI-HIV THERAPY OCCURRED IN INDIVIDUALS WITH LESS THAN 50 CD4 CELLS/MM3 IN THE ADULT NCI EXPERIENCE: IMPUCATIONS FOR THE DEVELOPMENT OF CD4 AS A MORTALITY RISK INDICATOR Yarchoan. Robert; Venzon, D; Lietzau, J; Pluda, JM; Wyvill, KM; Steinberg, S; Broder, S National Cancer Institute (NCI), Bethesda, MD, USA Objective: Depletion of CD4 cells is central in the pathogenesis of AIDS. However, there is relatively little information available on the relationship between CD4 cells and the immediate hazard of dying. Methods: Records of 55 consecutive pts with AIDS or severe ARC on AZT-based regimens treated between July 1985 and Dec 1990 at the NCI were analyzed. They represent all patients on the 1) phase I trial of AZT, 2) pilot study of AZT/acyclovir, and 3) pilot study of AZT/dideoxycytidine (ddC). Results. 44 pts have expired (CD4 counts proximal to death evaluable in 41), 10 are alive, and 1 was lost to follow-up. For the 51 evaluable pts, who are either alive or whose CD4 count proximal to death is known, all but one death occurred in a pt whose CD4 count had fallen below 50 cells/mm3 (p<0.00000001 using the two-sided exact binomial distribution to analyze the deaths/pt-month). For the patients who died, the geometric mean of the last 3 CD4 counts was 8 cells/mm3 (95% CI 1-63 CD4 cells/mm3). The median survival of patients after their CD4 count fell below 50 cells/mm3 was -9 months. The hazard of dying in patients with <50 CD4 cells/mm3 increased from 0.06 deaths/pt-month during the first year after entry to 0.09 deaths/pt month during the fourth year after entry. Discussion and Conclusions. Nearly all deaths in patients on anti-retroviral therapy occurred in pts. with <50 CD4 cells/mm3. These data provide support for the hypothesis that a CD4 count of 50/mm3 may be used as a mortality risk indicator in clinical studies. However, CD4 may not be the only indicator of benefit. Analyses should be undertaken of other data bases to explore this hypothesis. NOTES M.B.38 A SENSITIVE METHOD FOR MONITORING EFFICACY OF ANTIRETROVIRAL THERAPY IN HIV-INFECTED INDIVIDUALS: A HIGHLY SENSITIVE P24 ANTIGEN ASSAY. Edward E. Winger. Mohan M. Reddy, David Hargrove, Thomas McHugh, George F. McKinley, and Michael H. Grieco, IMMUNODIANOW 1C LABORATORIES, INCSAN FRANSSCO, CALIFORNA, USA, AND AIDS CNICALTRALS UNITST.LUKEROOSEVELT HOSPITAL CENRNEW YORK, USA. Objective: Current methods for monitoring of HIV viral burden include quantitative PCR, quantitative plasma viral culture and HIV p24 antigen assay. The first two techniques are currently impractical while the latter is relatively insensitive detecting antigen in only a fraction of infected individuals. Methods: A highly sensitive p24 antigen antigen-capture ELISA assay was developed by Immunodiagnostic Laboratories with sensitivity at one picogram/ml. To further enhance the rate of antigen, detection, serum is pretreated with HCI to denature immune complexes and is neutralized prior to analysis. Results: HIV p24 antigen was detected in 50 to 60% of symptomatic patients and 15 to 20% of asymptomatic patients. The highly sensitive assay described was able to detect antigen in greater than 95% of these sera. Patients with ARC and AIDS treated with AZT or ddl who were HIV p24 antigen negative by a commercial antigen-capture ELISA could not be monitored for drug efficacy. All of these cases could be monitored using the sensitive assay technique. Conclusion: The use of this highly sensitive HTV p24 antigen assay has permitted the monitoring of the efficacy of anti-retroviral therapy. (Supported by Cooperative agreement Grant No. AI25902, NIAID, AIDS program, Bethesda, MD). NOTES "I pr c00 CJo 1'

Page  32 M.C.39 ESTIMATE OF HIV-1 INFECTION AMONG U. S. HOSPITAL PATIENTS Stafford, Randall S; Janssen, RS; St. Louis, ME; Petersen, LR; Dondero, TJ; the Sentinel Hospital Surveillance Group. Centers for Disease Control, Atlanta, GA, USA. Objective To estimate the annual number of HIV-l-infected patients admitted to U.S. hospitals. Methodrs As a modified part of CDCs Sentinel Hospital Survey, from 7/89 to 6/90 a systematic sample of blood specimens left over after clinical testing were anonymously tested for antibodies to HIV-1. Sera from 52,500 patients in 19 U.S. hospitals were tested, allowing estimates of hospital seroprevalence by age, sex, and admission diagnosis. To generalize this data to all U.S. hospital patients we developed a step-wise, multiple regression model of hospital HIV-1 seroprevalence. The best independent predictors of HIV-1 hospital seroprevalence were the respective county's total AIDS incidence rate and AIDS incidence rate among females ages 25-44 (from CDC's AIDS Reporting System), and the presence of outpatient AIDS services at the hospital (from the American Hospital Association). Using national data on these predictor variables, we estimated HIV-1 seroprevalence in areas without sentinel hospitals. Modeling of seroprevalence by admission diagnosis stratified HIV-1 infected patients into those with AIDS, those with other HIV-1 suggestive diagnoses, and those without HIV-1 suggestive diagnoses. Results: From 7/89 to 6/90, an estimated one of every 115 patients (0.9%) admitted to hospitals was infected with HIV-1, for an estimated total of 250,000 (range: 200,000-300,000) HIV-1 positive patients. Of these patients, 26% were admitted for AIDS and 44% for HIV-1 suggestive diagnoses. The remaining 30% were admitted without HIV-1 suggestive diagnoses. Discussion: The large fraction of patients with potentially unrecognized HIV-1 infection identifies a subpopulation that may benefit from hospital-based HIV-1 counseling and testing. M.C.40 INTERVILLAGE VARIATIONS IN HIV SEROPREVALENCE IN A RURAL UGANDAN COMMUNITY Kengeya-Kayondo J.F., Kamali Anatoli, Nunn A.J., and Mulder D.W. * MRC (U.K.) Programme on AIDS in Uganda ** Uganda Virus Research Institute, P. 0. Box 49, Entebbe, Uganda. Aims: The present study examines between-village variation in HIV seroprevalence and explanatory variables for the observed variation. The study is part of the MRC Programe on AIDS which aims, among other objectives, at elucidating the population dynamics of HIV infection. Study Methods: The population cohort covers a defined geographical area of 15 villages in one sub-county of Masaka District with a total censused population of 10,411. Enrollment of the cohort was done through a baseline ethnodemographic, socio-economic, medical and serological survey. Sera were collected from 8,203 individuals. All sera were tested for antibodies to HIV-1 using HIV-1 ELISA (Cambridge Bioscience Corp.) and Welcozyme IlV recombinant ELISA. Western blotting was done when repeat ELISA results were discordant or weak concordant. Confirmed serostatus data is available for 7798 individuals. Results: Overall, 4.9% (all ages) and 8.5% (13+ years) were HIV antibody positive. Crude seroprevalence rates for all ages combined varied by village from 2.9% to 8.8% (p=0.00065). For adults (13+ years) seroprevalence rates varied from 4.9% to 13.6% (p=0.022). After allowing for age and the sex-age group interaction both of which were significantly associated with serostatus intervillage differences in seroprevalence rates were between 3.2% and 8.5% (p<0.05). The intervillage variation could not be explained by analysis of wealth levels, religion and occupation, as performed by the time of writing this abstract. Conclusion: Intervillage variations in HIV seroprevalence have been demonstrated in a rural Ugandan community. To date they cannot be explained by socio-economic factors. t NOTES NOTES M.C.41 DIRECT MEASUREMENT OF INCIDENCE OF HIV-1 AND HIV-2 IN FEMALE PROSTITUTES IN SENEGAL Kanki, Phyllis*; M'Boup, S*; Marlink, R.*; NDoye, I#,Gueye, A. **, Siby, T"; Thior, 1. ", Dia, M. ", Travers, K.*; Essex, M'. 'Harvard School of Public Health, Boston, MA, USA, # tnstitut Hygiene Sociale, Dakar, Senegal. "University of Dakar,Dakar, Senegal. Objective: We have followed registered female prostitutes in urban centers in Senegal to better understand the epidemiology of HIV-2 infection. The direct measurement of HIV-2 and HIV-1 incidence allows for a better understanding of HIV population dynamics in this endemic area. Methods: Sequential serum samples from prosittues in Dakar and Ziguinchor were obtained for retrovirus (HIV-1, HIV-2 and HTLV) examination semi-annually. HIV serodiagnosis was performed with immunoblot, RIPA, recombinant env peptides of HIV-1 and HIV-2. All seroconversions were verified on standardized antigen. Results: The cumulative seroprevalence in Dakar prostitutes was 9.5% HIV-2, 2.4% HIV-1 and 0.5% HIV-1/2. Through evaluation of sequential samples on 1033 of these women (>3200 samples) 2100 person-years of observation from 1985-90. We determined the annual incidence for HIV-2 to be 8.1 per 1000 per year (17 seroconversions/2099 seronegative person-years) The HIV-2 incidence rate was found to be similar over the 5 year period, with a constant rate over the 5 year period this would predict a doubling time of 11 years. Incidence of HIV-1 in this population was 4.8 per 1000 per year; this rate had doubled over the last 2 year period, indicating a doubling time of less than 5 years. Among Ziguinchor prostitutes the seroprevalence for HIV-2 was 34.0% with an incidence of 40.0 per 1000 per year. From 1987-90, 169 seronegative person-years were observed with 7 seroconversions. There was no incidence of HIV-1 during this period. This region has a significantly higher prevalence of HIV-2 and assuming a constant incidence over time the doubling period would be 8-9 years, comparable to that in Dakar. Results of clinical evaluation of these women continues to indicate a marked difference in disease development between HIV-1 and HIV-2 (Marlink, Thior, and Siby et al). Conclusion: This is the first report of incidence of both HIV-1 and HIV-2 from the same geographic area and study population. Despite higher prevalence of HIV-2 the incidence in the Dakar cohort indicates a lower rate of new infections by HIV-2 as compared to that of HIV-1. Our results support other epidemiologic, clinical and virologic findings demonstrating significant differences between HIV-1 and HIV-2. M.C.42 TRENDS IN HIV-1 AND HIV-2 INFECTIONS IN ABIDJAN, COTE D'IVOIRE, 1987-1990 DOORLY Ronan *; Kadio A **; Brattegaard K *; Gnaore E *; Coulibaly I-M#; De Cock KM *,##. * Projet RETRO-CI Abidjan, ** University of Abidjan, # Centres Antituberculeux, Abidjan, Cote d'Ivoire, ## Centers for Disease Control, Atlanta, USA. OBJECTIVE: To examine trends in HIV-1 and HIV-2 infections in Abidjan between 1987 and 1990. METHODS: Serum specimens collected between 1987-1990 from 18 045 persons in Abidjan were tested for HIV-1 and HIV-2 antibodies (ELISA; Western blot and synthetic peptide suppemental tests). Groups tested included women of reproductive age (n=4395), male STD patients (n=1090), blood donors (n=2327), TB patients (n=4778) and hospitalized patients (5455). Trends over time were examined. RESULTS:In pregnant women, HIV-1 prevalence increased from 5% in 1988 to 9.1% in 1990, but HIV-2 prevalence (2.6% vs 1.9%) and dual reactivity (1.1% vs 1%) changed little. Similar trends were seen in STD patients (1987 HIV-1 7.0%, HIV-2 1.6%; HIV-l/HIV-2 1.4%; 1990 HIV-l 13.9%, HIV-2 2.1%, HIV-l/HIV-2 1.4%). In TB and hospitalized patients marked increases in overall prevalence were observed (Infectious Diseases 1987 38.5%, 1990 59.7%), largely due to increased HIV-1 prevalence (TB 1987 HIV-1 15.6%, HIV-2 4.3%, HIV1/HIV-2 4.8%; TB 1990 HIV-1 31.4%, HIV-2 4.1%, HIV-1/HIV-2 9.3%). CONCLUSIONS: (1) HIV prevalence is increasing in general population; (2) marked increase is occurring in patients presenting to health care structures; (3) increase is largely due to HIV-1; (4) little change has occurred in HIV-2 or HIV-1/HIV-2 prevalence; (5) dynamics of H V-1 and HIV-2 epidemics seem different. 0 Q

Page  33 M.C.43 THE MAGNITUDE OF HIV/AIDS EPIDEMIC IN EASTERN PART OF EUROPE A, Gromyko, World Health Organization. Rtional Office forx Europe. - _1W end of 1990, a total of 41t 3 ease of AIM wre re orted to the World lealth Organisation from 31 guropean countries. Of these soly 1a AIS cases swee reported from the eastern part of Europe, which includes Albania, Bulgaria, Caechoslovakia, Hungary, Poland, Romania, the USSR and Yugoslavia. Although in thes contries the initiation of the epidemic is clearly connected with the introducties f the infection from other continents, at present transmission of the infection is continuing within different propulation groups with clear variation in modes of transmission. Czechoslovakia, Hungary and Poland have a clear majority of AIDS c-es within the homo/bisexual group. The majority of the Bulgarian cases are in the heterosexual community and the USSR and Yugoslavia exhibit a range of transmission patterns. Only Yugoslavia and Poland show AIDS cases among intravenous drug users. The number of HIV infected among intravenous drug users in Poland has dramatically increased during the last year and now reaches 975 of a total of 1373 HIV infected parsons detected so far. During the last two years a new alarming situation aroses the outbreaks of HIV infection in children hospitals in several cities of the USSR where more than 260 children have been infected through nosocomial transmission of the virus, and the occurrence of 1153 AIDS cases among institutionalised children in Romnia. Rather low figures of HIV/AIDS in the eastern part of Europe frequently cause questions with regard to reliability of AIDS reporting systems in these countries but the results of HIV seroprevalence received on the basis of extensive screening of various population groups seems to confirm the reported data. Special efforts are now being made by the World Ikelth Or- saties to swifet thes- - sr in AIDS prventioa and control in view of the recent politial and seiasl haiaan1 to tresmdoUItu incraU of emtaets btm th11 ht -an th __,t M..44 UNITED STATES ARMY: 1985-90 MeNeil. Jhn*; Brundage. J.*; Gardner, L. *; Wann, F.*; Renzullo, P.*; Burke, D.' and the U.S. Military Medical Consortium for Applied Retroviral Research *Walter Reed Anny Institute of Research, Washington DC, USA: *SRA Technologies, Alexandria Virginia, USA. Objective: To describe secular trends and demographic correlates of an evolving HIV infection epidemic among soldiers in the United States Army. Methods: Because soldiers in the US Army are recurrently tested for the presence of HIV antibody. HfV seroconversion rates can be directly measured. From November 1985 through October 1990, seroconversion rates were calculated for 4 time periods: Nov.'85-Oct.'87 (Interval 1:) Nov.'87-Oct '88 (Interval2): Nov.'88-Oct.'89 (Interval3): and Nov.89-Oct.'90 (Interval4), A Poisson regression model was used to assess the demographic and temporal correlates of seroconversion. Results: Between November 1, 1985 and October 31, 1990, 613 HIV seroconversions were diagnosed during 1,774,081 person-years (P'Y) of followup (0.35/1000P*Y. Seroconversion rates declined significantly (P<.001) from 0.46/1000P*Y (Intervall) to 0.32/1000P*Y (Interval2), and stabilized at 0.27/1000P*Y (Interval3) and 0.26/1000P'Y (Interval4). For the combined 5 year period, H1V seroconversion risk was significantly associated with minority race-ethnicity, male gender, and single marital status. Rates trended downward most significantly among 20-30 year old males, but did not change or trended upward among black females, soldiers _35 years of age, and teenagers. Conclusion: The HIV infection epidemic is dynamic. HIV incidence rates and trends have changed in important ways over the past 5 years in the US Army. Rates declined significantly overall and for many demographic groups between November 1985 and October 1990. but remained largely unchanged or even increased among teenagers and black women. Based on these observed trends, the efficacy of prevention efforts can be assessed, and interventions revised and focused in responsive ways. NOTES NOTES M.C.45 RISK OF DEVELOPING AIDS AFTER HIV SEROCONVERSION IN INJECTING DRUG USERS: ANALYSIS OF EARLY MARKERS OF DISEASE EVOLUTION. Rezza. Giovanni: Pezzotti, P; Lazzarin, A; Angarano, G; Sinicco, A; Aiuti, F; Pristera, R; Salassa, B; Ortona, L; Castelli, F; Gafa, S; Zaccarelli, M. IMCSA, AIDS Unit, Istituto Superiore di Sanita-Roma (Italy) Objective: To examine the risk of developing AIDS, the incubation period to full-blown disease, and risk factors for disease progression among a cohort of Italian seropositive injecting drug users (IDUs). Methods: A longitudinal study was conducted among IDUs, enrolled in 15 clinical centres in Italy, who had undergone sequential HIV tests and who had seroconverted: Seroconversion date was estimated as the midpoint between last negative and first positive HIV test. Demographic, clinical and laboratory data were collected by a standard form. Disease progression was analyzed using Kaplan-Meier survival curves in which date of AIDS diagnosis was the end point in the analysis. Differences between curves were assessed by log-rank test. Result: Of the 440 IDUs in the study, 324 (74%) were males. Mean age at seroconversion was 25 years. The median period between the last negative and first positive tests was 8 months, and the median follow-up time was 43 months. Twenty-five developed AIDS during the study period. The estimated cumulative incidence of AIDS was 21% by 6.3 years after seroconversion. Those.25 years at seroconversion progressed more rapidly than those <25 years. A CD4+ count less than 500 at first visit was also associated with faster progression of HIV infection. Discussion and conclusion: The risk among IDUs of developing AIDS following seroconversion is similar to that of other risk groups in Italy. Predictors of disease progression may be identified in the early stage of HIV infection and include older age and low CD4+ count at first visit. These findings suggest that host factors may play a role in the outcome of HIV infection in IDUs. M.C.46 SEX TRADE ACTIVITY AMONG FEMALE INJECTION DRUG USERS IN SAN FRANCISCO Lorvick Jennifer*; Watters, John*; Shade, Startey*; Cheng, Yu-Teh** *University of California, San Francisco, **University of California, Berkeley Obective: To measure the frequency of sex-for-money activity among a sample of female injection drug users (IDUs) and examine its relationship to condom use, history of sexually transmitted diseases and rate of HIV infection. Methods: Using a targeted sampling method, two cross-sections of female IDUs were recruited from three inner-city street locations in San Francisco during late 1989 and early 1990 (N =366). Respondents were interviewed and tested for HIV-1 antibodies. Self-report data were collected regarding the frequency of sexual activity in exchange for money during the six months prior to interview. For purposes of analysis, the women were divided into three groups: a zerofrequency (ZF) group of those never having sex for money in the past six months (n =275); a low frequency (LF) group composed of those having sex for money 1-49 times in the past six months (n= 56); and a high frequency (HF) group of those having sex for money 50 times or more (n =35) In the past six months. Groups were compared by frequency of condom use, history of sexually transmitted diseases, and HIV infection. Results: The ZF group reported that in the past six months, 77.4% never used condoms, 16.1% used condoms some of the time and 6.6% used condoms all of the time. The LF group reported that 35.7% never used condoms, 48.2% used condoms some of the time, and 22.5% used condoms all of the time. The HF group reported that 2.9% never used condoms, 60% used condoms some of time, and 37.1% used condoms all of the time (Pearson's r = 49; p <.000). The rate of HIV infection was 11.9% in the ZF group, 17.3% in the LF group, and 6.1% in the HF group. These differences were not significant. Differences in the incidence of syphillis, genital warts, herpes or chlamydia also were not significant. However, women in the LF group were significantly more likely to report a history of gonorrhea (67.9%) than women in the ZF (29.5%) or HF (42.9%) groups (p<.000). Conclusions: These data suggest a strong correlation between frequency of paid sexual activity and condom use. It suggests that female IDUs who have sex for money, espeeiatty those in the HF group, are more likely to practice safe sex than their counterparts. Thedata also suggest that female IDU'swho have sex for money are no more likelythan their counterparts to contract or spread HIV or sexually transmitted diseases, with the exception of gonorrhea. The higher rate of gonorrhea among the LF group, combined with their lower rate of condom use compared to the high-frequency group, may indicate a higher risk of contracting sexually transmitted diseases and HIV via sex among female IDUs who have sex for money on a less frequent basis. c\

Page  34 SM.C.47 HIV WASTING SYNDROME IN THE UNITED STATES Nahlen, Bernard; Nwanyanwu, O; Chu, S; Berkelman, R. Centers for Disease Control (CDC), Atlanta, Georgia, USA. Objective: To describe the characteristics of AIDS patients reported with HIV wasting syndrome (WS) in the United States. Methods: Adult/adolescent AIDS patients (>13 years) reported with WS to CDC from 9/87 through 12/90 were compared to those without WS (non-WS). Results: Of the 85,846 AIDS patients reported since 9/87, a total of 5.5% had WS as the only AIDS-indicator condition (Al), and 10.5% had WS plus at least one other Al. Overall, there were no differences noted by race, but WS patients were significantly more likely to be female and to be in the 13-19 and 60-69 year age groups. WS patients were also more likely to be intravenous drug users (IVDUs) or sex partners of IVDUs, and were more likely than non-WS patients to have died. The proportion of AIDS patients reported with WS by geographic distribution ranged from a low of 13.3% in the mid-Atlantic states to a high of 50.1% in Puerto Rico. The Als most strongly associated with WS were candidiasis of the lungs (OR 3.3; CI 3.0-3.6), esophageal candidiasis (OR 3.3; Cl 3.1-3.4), HIV encephalopathy (OR 3.0; CI 2.8-3.2), and coccidioidomycosis (OR 2.5; Cl 1.8-3.4); there was no association with mycobacterial disease. Conclusions: Overall, 16.0% of AIDS cases reported since 9/87 had WS as a diagnosis. The association of WS with IV drug use and the gender and geographic differences may be related to differences in diagnostic or reporting practices and/or access to medical care. NOTES M.C.48 PREVALENCE OF HIV INFECTION AMONG INJECTING DRUG USERS IN LONDON Crosier, Adam N., Stimson G.V., Ettorre E.M., Stephens S.F. Centre for Research on Drugs and Health Behaviour, Charing Cross and Westminster Medical School, London, England. Objective: To determine the prevalence of HIV infection and describe associated risk behaviours among injecting drug users (IDUs) in London, England, in collaboration with the WHO Study on 'Drug Injecting and Risk of HIV Infection'. Methods: 534 IDUs were recruited, 117 (22%) from 13 treatment sites and 417 (88%) from other sites including social networks. Saliva samples were taken and tested anonymously for HIV antibodies using GACELISA with HIV positives confirmed by Western Blot. Information on self reported risk behaviours was gathered using a standard questionnaire developed in collaboration with the WHO. Results: HIV prevalence rate was 13%. Levels of reported risk behaviour were similar to the findings of other studies of IDUs in London: 46% reporting having shared needles or syringes in the previous 6 months; 70% and 34% of those who had vaginal intercourse with casual partner(s) and primary partner(s) of the opposite sex respectively, reporting never using condoms. Sample characteristics match the findings of other studies; mean age at interview = 28.4 years, 65.5% of the sample were men and 34.5% women, age of first injection =-19.3 years, mean length of drug use by injection = 9.1 years. The most commonly injected drug = heroin followed by cocaine, heroin and cocaine together and amphetamines. Discussion and Conclusions: Studies of HIV infection among IDUs in London before 1990 found prevalence rates of no higher than 6%. This study indicates that whilst reported risk behaviours remain unchanged, HIV prevalence may be increasing. NOTES cn 4^ c1 M.C.49 BLEACH DISINFECTION OF NEEDLES BY INTRAVENOUS DRUG USERS: ASSOCIATION WITH HIV SEROCONVERSION Vlahov D, Celentano David D, Munoz A, Cohn S, Anthony JC, Nelson KE. ALIVE Study, Johns Hopkins School of Hygiene and Public Health, Baltimore, MD, USA. Objective: To examine association between use of bleach disinfection of needles and HIV seroconversion (SC) among intravenous drug users (IVDUs). Methods: IVDUs initially screened in 1988-89 were scheduled to return for serological rescreens and risk behavior interviews at 6 mos. intervals. The sample for analysis was restricted to blacks who shared needles and excluded homosexual males. Each SC was matched with up to five controls on date of study entry (~ 3 mos.), duration of follow-up (all controls with length of follow-up > case), gender, and recent cocaine injection (yes/no). Data on needle cleaning (none, <all, all the time) were from the first seropositive visit for each case and from the visit closest to the case for the matched controls. Odds ratios (OR) for SC by frequency of disinfection were performed using conditional logistic regression procedures. Results: 22 eligible SC were matched with 95 controls. Compared to no bleach use, the OR for bleach "< all the time" was 0.91 (0.26,3.22) and for bleach "all the time" was 0,77 (0.25,2.33). Conclusions: Bleach disinfection showed a modest but not significant protective effect for seroconversion. Lack of significance may be due to limitation of the size of the sample and/or bias inherent in those who opt for bleach use. Alternatively, the effect of bleach in preventing HIV infection intravenous drug users may be limited. M.C.50 THE RISK THAT HIV-SEROPOSITIVE DRUG USERS (DU0) TRANSMIT THEIR HIV-INFECTION M. B Harters C, Krijnen P, Van den Hoek JAR, Coutinho RA and Van der Pligt J. Municipal Health Service, Amsterdam, The Netherlands. Objective: To determine rates and potential predictors of HIV-transmission risk behavior in HIV-seropositive (HIV+) DUs (n=122) enrolled in a longitudinal Amsterdam AIDS-study. '4ethodsqo Data on current (i.e. in the previous 4-6 months) sexual and injecting behavior, perception of HIV-transmission risk, self- and response efficacy with regard to condom use, attitudes and intentions were recorded in HIV+ DUe between June-Nov.'90. Results: Lending a used needle was reported by 4 (4%) of 95 injectors (78%). 62 DUs (51%) had vaginal sex (VS) with clients and/or private partners (PP). Not always using a condom when having VS (=unsafe sex) with clients was reported by 4 (22%) of 18 prostitutes (15%). Unsafe sex with PP was reported by 25 (45%) of 56 DUs (46%) who had VS with casual/steady PP. In the total group there were 25 DUB (20%) who could have transmitted HIV to someone else by private/commercial unsafe sex (excl. 8 with unsafe sex only with an HIV+ steady partner) and/or by lending a used needle. 47% of the total group thought they had a chance to get syphilis in the future, 44% thought they might infect someone with HIV via sex in the future and 21% thought this might happen through needle sharing (combined: 49% via sex and/or sharing). 48% think condoms are not/not always effective in preventing HIV-transmission. 31% have a negative attitude about condoms: interrupting sex and/or feeling unpleasant. 59% of prostitutes and 71% of those with PP have a low self-efficacy with regard to always using condoms with reap. clients and casual PP. 94% of prostitutes and 88% of those with PP have a strong intention to use condoms with resp. clients and casual PP; 93% of injectors have a strong intention not to lend used needles to others. Conclusions: Most of these HIV+ DUs are current injectors and half are sexually active. Although these HIV+ DUs have good intentions, 20% could have transmitted HIV to someone else in the previous 4-6 mo. and 49% think there is a chance they will transmit HIV to someone in the future, either through sex or through needle sharing. These results suggest that HIV-prevention counseling targeted at HIV+ DUs ought to be intensified. C.> 0 Q^

Page  35 M.D.51 AFRICAN AIDS POLICIES AND LEGAL TRADITIONS Shandera Wayne X.*, *Dept of Medicine, Baylor College of Medicine, Houston, TX USA Objectives. To assess current African AIDs policies and determine if policies correlate with legal traditions of responding countries. Methods. African ministries of health were contacted twice in 1990 regarding HIV/AIDS policies dealing with notification of health authorities and sexual contacts, anti-discrimination, immigration, blood donation, and mandatory testing. National legal traditions were divided into common, civil, mixed common/ civil, and mixed Islamic/ civil laws. Survey results and traditions were correlated with statistical software. Results. The 20/53 responding African nations included 3 - common, 8 - civil, 4 - mixed common/civil, and 5 - Islamic/civil law backgrounds. Countries with civil law traditions were more likely to require that all donated blood be tested for HIV antibodies (X chi square - 2.5), and not require notification of sex partners of cases; only countries with civil law traditions had antidiscrimination policies or policies requiring HIV-infected physicians to notify their patients. Only nations with Islamic/civil backgrounds had laws requiring HIV screening of at-risk groups or border laws. The 50% of respondents with policies preventing HIV-infected from donating blood represented no specific background. Conclusions. African ministries of health are achieving a consensus regarding AIDS policies. A few differences exist by legal background, usually in countries with civil law traditions. African antidiscrimination legislation is unusual; blood donation policies are not uniform. NOTES M.D.52 THE FOOD AND DRUG ADINISTRATION'S COMMUNITY BASED RESEARCH EFFORTS Couig, Mary Pat U.S. Food and Drug Administration, Rockville, MD, USA Objective: To expedite access to new therapies for HIV infection and AIDS-related illnesses by expanding access to investigational therapies and by increasing the size of the available data base. Community-based health professionals receive education, training and technical assistance. Information is provided on the FDA requirements for conducting clinical trials under Investigational New Drug (IND) applications with the goal of producing reliable and accurate data capable of supporting approval of new therapies. Methods: 1) Liaison with the National Institute of Allergy and Infectious Disease's Ccanunity Program for Clinical Research on AIDS (CPCRA), the American Foundation for AIDS Research (AmFAR), and other health professional organizations with a primary interest in HIV infection; 2) Education and Training. Participate in workshops and other training sponsored by FDA, NIAID/CPCRA, and AmFAR. Assemble a manual of FDA regulations, guidelines, and other FDA references that contain essential information for the conduct of studies under INDs; and 3) Provide technical assistance as needed in the development of protocols, clinical investigator responsibilities, or regulatory issues. Results; This effort has fostered close collaboration with NIAID/CPCRA and AmFAR. Although it is too early for results, we hope that these efforts will increase the investigators' understanding of the FDA regulations for conducting clinical trials under IND, will improve the quality of the data subnitted, and will ultimately speed up the approval for therapies for HIV. Discussion and Conclusion: This effort is significant because this is the first time FDA has taken such intensive steps to educate clinical investiaators. NOTES C) I IC M.D.53 BALANCING FEAR AND RESPONSIBILITY: PROFESSIONAL ETHICS IN AIDS CARE Koenig, Barbara A.; Cooke, M.; Beery, N.; Folkman, S. University of California, San Francisco, CA, U.S.A. Objectives: To assess the relationship between fear of AIDS and perceived professional responsibility among U.S. internal medicine residents, and to describe the process of accommodation to AIDS fear while caring for HIV-infected patients. Methods: A self-administered survey was completed by 40% of medical trainees at 41 internal medicine residency programs in the U.S. during 2 calendar years (Year 1, n=1045; Year 2, n=958). Measures consisted of multiple Likert-scaled items, including assessments of: professional responsibility, fear of contagion, and intentions to treat AIDS patients after training. A sub-sample of physicians (n=31) at 3 sites participated in ethnographic interviews during years 1 and 2 of training (data were audio-recorded, transcribed, and analyzed using qualitative content analysis). Programs represent 3 U.S. epidemic patterns: low AIDS incidence; high incidence, >30% IV drug-user patients; and, high incidence, *66% gay patients. Results: T-tests indicated that physicians at high AIDS incidence sites exhibit greater levels of both fear (p<.005) and responsibility to provide care in the future (p<.0001) than did physicians at low incidence sites. Of physicians at highIVDU sites, 22% report needlesticks from HIV+ patients, as opposed to 7% at high-gay sites, and 5% at low incidence sites. Interview data reveal a complex process of adaptation to fear through experience with AIDS patient care. Conclusions: Fear and a positive sense of ethical responsibility coexist among the residents most committed to AIDS care. Educational interventions aimed at increasing the number of physicians willing to provide AIDS care should focus not on over^ coming fear, but on helping care-providers understand its sources and meanings. -L n / INWCRASAIRIG DENTISTS' WILLINMNESST TO TREAT HIV+ PATIEITS - M.D.54 Sadowskv, Donald;* Kunzel, C.* *Albert Dinstein College of Medicine, Bronx, New York, USA Objective: What predicts dentists' willingness to treat HIV+ patients (PHIV+)? A model has been developed to answer the question and is described below. Methods: A disproportionately stratified (high AIDS prevalence urban, other urban, and rural areas) random sample of American general practitioner dentists was drawn from the master list of the American Dental Association. 1351 usable 12 page.questionnaires were returned - response rate 88%. Logistic regression was used to test a variety of models predicting willingness to treat PHIV+: yes/no. Predictor variables were: dentists ethically obligated to treat PHIV+ (ETHICS); patients would leave practice if they knew I treated infected patients (PTSLEAVE); staff does not want to treat PHIV+ (STAFF); I probably have treated PHIV+ (HASTXHIV); I feel I can safely treat PHIV+ (SAFEHIV). Results: DENTISTS' WILLENGNESS TO TREAT I[V+ PATIENTS Independent variables B.E. df Si. Odds Ratio ETHICS.6008.0984 1 <.0001.1744 1.8235 HASTXHIV.9671.2037 1 <.0001.1331 2.6304 PTSLEAVE -.3774.1543 1.0144 -.0586.6856 STAFF -.7000.1632 1 <.0001 -.1189.4966 SAFEHIV 1.5428.1407 1 <.0001.3193 4.6776 Constant -2.4026.7779 1.0020 Model Chi-Square df Sig. 531.53 5 <.0001 Discussion and Conclusion: It appears that dentists' personal feeling of safety is the single most important determinant of their willingness to treat PHIV+. This is substantiated by the fact those who recognize that they have probably already treated PHIV+ are much more willing to treat them in the future. This phenomenon may be analogous to desensitization from a phobia. A sense of ethical obligation to treat PHIV+ is also an influential predictor of willingness to treat PHIV+.

Page  36 M. 5 AIDS, EUTHANASIA AND GRIEF Van den Boom, FMLG*d Mead. hristopher**: Gremmen, T; Roozenbur, H *Netherlands Institute of Mental Health, Utrecht, The Netherlands; ** University of Leiden, Dept. of Psychology, The Netherlands Objective: Gain insight in 1) the occurrence of euthanasia among Aids patients within the context of Dutch eutianasia policy; 2) the grieving process and its complications. Methods: In-depth structured interviews were held with 59 relatives of deceased Aids patients. Results: Euthanasia was discussed by 60% of all patients, mainly because of their fear of dementia and physical deterioration. Euthanasia was performed in 23% of the cases. Abstination of treatment, while at the same time using pain medication, occurred in another 31% of the cases. Problems around euthanasia are the dosage and type of medication used, and the moral, legal and ethical problems physicians are confronted with. The legal dilemma can be resolved, the moral and ethical dilemma remains. The usual reaction to loss is rief, which expresses itself in feelings of loneliness and intense sorrow (68%), sleeping problems 58%), lethargy (27%), eating disorders (17%) psychosomatic complaints (22%), increased use of alcohol (14%) and medication (29%). Half of the respondents did not seek professional support; for them the support from selected friends and family members was sufficient; 40% sought professional help in many cases out of an anxiety not to burden significant others; 10% did not seek help, although they were in need of it. Although the decision for euthanasia is never simple, in this study it does not appear to complicate the grief process. However, the chance that a complicated grief reaction occurs is high for people who initially react with many psychosocial problems and for those who do not receive adequate social support. Three patterns of adaptation to loss can be distinguished. Some people go through the 'classical' pattern; some do not show intense sorrow, and for others the working through process takes much longer than one would expect. Conclusions: I) Euthanasia is a major theme for people with Aids. Health care providers should anticipate this by drawing up eutanasia and abstination protocols. Legal obstacles should be resolved. 2) Many people can rely on a 'healing web' of friends and relatives but 40-50% need professional assistance during the grief process. Therefore, ade.quate services should be provide d NOTES M.D.56 FACTORS EFFECTING THE DECISION AND INTENTION TO SEEK AN HIV Ab TEST Mves. TdJ; Orr, K.**; Locker, D.'; Jackson, E."' SUniversity of Toronto, Toronto, Ont., Canada ** Canadian AIDS Society, Ottawa, Canada, *** AIDS Committee of Toronto, Toronto, Ont., Canada Objective: To identify among gay and bisexual men the socio-demographic and behavioural variables related to a previous decision or future intention to be tested and to reasons for not being tested' MethodA A convenience sample of 1295 men who have sex with men was systematically recruited in Toronto's gay identified bars and bathhouses. Self-completed questionnaires examined socio-demographic characteristics and behaviour (sexual and testing), as wel as. future intention to be tested and reasons for not being tested. Individual relationships of personal, socio-demographic and sexual behaviour variables with a) previous decision to be tested b) expressed intention to be tested and c) reasons for not being tested were analysed using contingency tables and analysis of variance; and multiple relationships were examined using logistic regression. Factor analysis was used to establish groups of reasons for not being tested. Results: Of the sample 53% had been tested for the HIV antibody and 13.6% reported they were HIV+. Perceived risk, type of sexual practice, bisexual activity, monogamy and residence (urban vs rural) were variables associated with previous decision to be tested. Reasons for not being tested formed three factors: desire for anonymity, sell-perceived health and lack of benefits from results. The single most prominent reason, "Do not want name on government list.", was seen as "very important" by 56.3% and "somewhat important" by 20.5% of the respondents. Reasons for not being tested and future intention to be tested varied with age, monogamy and bisexual activity. Conclusions: The proportion of men who had made the decision to be tested was higher than anticipated in view of the fact that testing until recently was discouraged within this gay community. Access to the test as well as personal, lifestyle and behavioural characteristics appear to influence decisions to be tested. For a substantial proportion of the study population concerns about the confidentiality and anonymity and perceived effectiveness of treatment remain as issues and barriers to early identification of HIV infection. NOTES bs 0O M.D.57 TELECONFERENCING: HOW 8000 PHYSICIANS CAN GAIN ACCESS TO THE LATEST MEDICAL EDUCATION ON HIV CARE AND TREATMENT Allerton. Michael; Wolfe, E.; Wibbelsman, C.; McNalty, K; Gordon, N.; Fuller, J.; Casal, T. Kaiser Permanente Medical Care Program (KPMCP), Northern California Region, Oakland CA. USA. Objective: To increase the number of physicians in a large group practice (n=3000) willing and able to provide primary medical care for HIV infected patients and their families in Northern California KPMCP, a Health Maintenance Organization with 2.4 million members ( estimated HIV sero-prevelence of 0.5%) and to make HIV related information available to an additional 5000 KPMCP physicians practicing nationally. Methods: Live bi-monthly broadcasts to 22 sites in the Northern California Region of KPMCP were presented from July 1989 through December 1990 for a total of 10 teleconferences to date. These were used to provide education on the diagnosis and primary management of HIV infection and to allow updates or changes in care and treatment to be disseminated to the physician group in a timely and efficient manner. Video tapes are available of each segment and distributed for use by all 8000 KPMC physicians practicing across the nation. A needs assessment was initially conducted and ongoing surveys are collected to decide the main content of each broadcast. The content is planned well in advance of each session and the "state of the art" of a specific topic is presented through expert panel discussions. Each session has an "Update" section developed immediately prior to the broadcast for "up to the minute" accuracy. The conclusion of each broadcast consists of a phone-in segment that enables physicians at all 22 receiving sites to interact with the panelists. Results: Surveys indicate 25-50% of primary care physicians participate in the program. Ongoing quantitative evaluations indicate significant changes in physician comfort with, willingness to treat, and knowledge of HIV infection. Physicians report feeling better prepared to treat HIV patients (70.3%-92.2% depending on session content.) The number of physicians willing to treat patients has increased 20%. Conclusion: Teleconferencing is an effective, efficient way reach large numbers of physicians practicing in widely dispersed sites. Providing physicians with basic content, frequent updates, and and an opportunity to interact with an expert can achieve the objectives of this program. M OBSERVED AND SELF-REPORTED COMPLIANCE WITH UNIVERSAL PRECAUTIONS AMUNU M. 58 EMERGENCY DEPARTMENT PERSONNEL AT TWO SUBURBAN COMMUNITY HOSPITALS Hcnry..Keith**; Collier, P*; O'Boyle-Williams, CA; Campbell, S* *University of Minnesota Medical School, Minneapolis, MN USA; *St. Paul-Ramsey Medical Center, St. Paul, MN USA; ^Fairview-Southdalc Hospital, Edina, MN USA Objective: To examine the compliance of Emergency Department (ED) personnel with the Universal Precautions (UP) recommendations derived from CDC guidelines. Methods: From 6/90-8/90, ED personnel were observed by 7 RNs at 2 medical centers in suburban Minneapolis. In 8/90, a survey was sent to the same personnel to determine self-reported compliance. Results: 1,822 patient-HCW encounters were observed. When appropriate, gowns were worn 16.3% of the time, (95%CI (9.8%,22.8%)); masks 22.9%, {95%CI (12.5%,33.5%)), goggles 50.8% (95%CI (38.6%, 62.9%)), gloves 72.2% {95%CI (69.5%,74.9%)} excluding IV starts and phlebotomy procedures. Gloves were worn for 32.3% of IV starts and 31.0% of phlebotomy procedures observed. Needles were discarded as recommended 51.2% of the time (95%CI (47.4%,55.0%)}. Needle recapping occurred 32.2% of the time (95%CI (28.7%,35.7%)). When recapping was done, unsafe "2-hand" recapping occurred 78.8% of the time (95% CI (73.4%, 84.2%)). Self-reported rates of appropriate glove use (80.7%, 95%CI (77.1%,84.4%)) and gown use (56.8%, 95%CI (53.1%,60.5%)} were higher than observed rates. Selfreported 2-hand recapping [78.8%, 95%CI (73.4%,84.2%)) was lower than the observed rate. Selfreported mask, goggle, and recap rates were within 1%-15% of observed rates and did not differ significantly. Non-compliance reasons were: no time(68%); pt appearance(59%); reduced dexterity (57%). Conclusions: Previous studies have shown that problems exist with Universal Precautions at urban teaching hospitals. This data indicates that significant problems exist with UP policy at community hospitals. Dexterity issues, time constraints, and patient characteristics influences on HCW compliance with UP. Various strategies to increase compliance with UP should be explored at all types of hospitals throughout the country. 0 ~to C> 0 tC> 1:

Page  37 M.D.59 RISK OF ENDEMIC HIV AND HEPATITIS B VIRUS (HBV) TRANSMISSION TO PATIENTS DURING INVASIVE PROCEDURES Bell David M. Martone WJ, Culver DH, Margolis HS, Shapiro CN, Tokars JI, Marcus R, Furmal LI, Jaffe HW, Curran JW, Hughes JM; Centers for Disease Control, Atlanta, GA, USA Objctive: To estimate 1) the probability (Pr) of endemic HIV and HBV transmission to patients from an infected surgeon or dentist due to percutaneous injury during an invasive procedure; 2) the number of patients infected by this route in the USA between 1980-1990. Methods: Risk was estimated using data from prospective CDC studies, including observation of injuries during surgery and the risk of HIV transmission after exposure to infected blood. The Pr of transmission (PrT) to a patient during a procedure (p)= the Pr of injury to the surgeon (2.5% x the Pr that the sharp object causing the injury recontacts the patient's open wound (32%) x the PrT after this exposure (.03%-.3% for HIV; 3%-30% for HBV). The cumulative PrT to >1 patient during N procedures by an infected surgeon is 1-(l-p)N. The total number of patients infected in 1980-90 was estimated based on 1) number of infected surgeons and dental workers, from AIDS surveillance data (adjusted for HIV infection) and HBV prevalence surveys; 2) number of procedures done by workers while infective; 3) range of PrT per procedure. Results: The PrT from an infected surgeon to a patient is.0002%-.002% for HIV;.024%-.24% for HBV. The PrT to >1 patient during 3500 procedures (7 years, an HIV+ surgeon's estimated remaining working life) is.81%-8.1% for HIV, 57%-100% for HBV (if the surgeon is an HBeAg+ carrier). This model suggests that, from 1980-1990, 12-129 patients may have acquired HIV infection and 478-4773 HBV infect on from infected surgeons and dentists in the USA. Conclusions: This risk assessment and these estimates, based on limited data and not taking outbreaks into account, suggest that prevention strategies may require re-evaluation. NOTES M.D.60 TRANSMISSION OF HIV INFECTION FROM SURGEON TO PATIENT:ESTIMATING THE RISK. Lowenfels, Albert B. MD, Wormser, GP, MD. Departments of Surgery and Division of Infectious Disease. Department of Medicine, NY Medical College, Valhalla NY. 10595. Objective: What is the likelihood of a patient becoming HIV + as a result of a puncture wound to the surgeon during a typical one hour operative procedure? Methods: We have used published reports to calculate the overall probability of such an event which depends upon: 1) the probability of a puncture injury during surgery; 2) the probability that the surgeon is HIV+; 3) the risk of transmission of HIV virus from surgeon to patient. Results: Reports of the frequency of puncture injuries give an average frequency of 0.005 per operative hour. The probability of a surgeon being HIV+ is unknown but probably lies between 0.015 and 0.007 which are estimates derived from US Army volunteers and patients in US sentinel hospitals. The risk of transmission of HIV virus from surgeon to patient would be somewhat less than 0.003, the recently reported occupational risk of seroconversion for health care workers after injury. Conclusions: Multiplying these three estimates gives an overall risk of 1 chance in 3' million per hour of surgery (95% Poisson-based CI = 1 in 1.3 billion to 1 in 3.7 million chances. If the surgeon were HIV positive the risk would still only be 1 chance in 133,000 (95% CI = 1 in 2 million to 1 in 26,000). These estimates are reassuring because even under the most unfavorable assumptions the risk of reverse HIV transmission is low. NOTES 0 Ntj 8~ M.D.61 EFFICACY OF HIV TRANSMISSION AFTER AT-RISK EXPOSURES IN HEALTH CARE SETTINGS: THE ITALIAN MULTICENTRIC STUDY Iuui.L LU. Giuseope*; Puro, Vincenzo* and "Italian Collaborative Study Group on Occupational Risk of HIV Infection" ** Obiective To evaluate efficacy of HIV trasmission after exposures to at/risk body fluids in health care workers (HCW), and to study incidence and type of exposures. Methods:Data were collected prospectively by 30 hospital occupational health services. All exposed HCW were tested for HIV/Ab at the moment of the accident and at intervals of 3, 6, 12 months. Results As of December 1990, 1340 occupational exposures to HIV were reported:19% by physicians, 64% by nurses and 17% by other occupational groups.Needlesticks were 56%, cuts 8%, nonintact skin and mucous contaminations 25% and 11% respectively.Most accidents occurred in patient's room (58%) during routinary assistance; 11% during surgery, 5% during emergency. Most HCW knew the patient's HIV seropositivity before the accident. HCW were followed for a mean time of 10 months (range 3-36).Two seroconversions occurred (0.149%,C.I.95%, Poisson distribution 0.018--0.539). ConclUsiog:The study confirms a low risk of HIV transmission after accidental exposures in HCW. However, the number of occupational at-risk exposures remains common. Sustained effort is required to enhance adherence to universal precautions and to find safer devices and techniques. * Coordinating Center: AIDS Unit at 1. Spallanani Hospital, Rose, Italy. Supported by Italian Kinistry of Healtb, AIDS project, grant 5202.01 ISS. "A' garano,G.; Arici,C.; Bertucct,R.; BoSazziL.;Cadrobbi,P.; Cestrone,A.; Chiodera,A,; Chiriam ni,A,; Colaiacoio,N,; Collina,D.;;orradini,S.; Cristini,.; De Carli,G.; Francavilla,B.; Fracesconi,K.; S linzi,G.; La Torre,.; oibardo,K.; Perna,l.C.; Pietrobon,.; Poapili,s.; Portelli,V.; Raineri,G.; Ruachiao,.; Rafitreoli.t.: SoaeIlla.L.: Suter.F.: Turbessi..: Vaalia.A.: Vlicos.D.: Zullo.G.. M.D.62 LOW RATE OF HIV-1 INFECTION AMONG HEALTH-CARE WORKERS WHO DONATE BLOOD Chamberland. Maryl; Petersen, L'; Munn, V1; White, C1; Busch, M2; Grindon, A3; Kamel, H'; Ness, P5; Zeger, G6; and the HIV Blood Donor Study Group. lCenters for Disease Control, Atlanta, GA, 2Irwin Memorial Blood Centers, San Francisco, CA, 3Atlanta American Red Cross (ARC), Atlanta, GA, 'ARC-Penn/Jersey, Philadelphia, PA,5Chesapeake ARC, Baltimore, MD, 6ARC-Los Angeles, Los Angeles, CA,USA Objective; To estimate the rate of occupationally-acquired HIV-1 infection among health-care workers (HCWs) who are blood donors (BDs). Methods: In an ongoing study, HIV-1 seropositive BDs at 20 U.S. blood centers (BCs) are interviewed using a standard protocol to assess HIV risk factors. In these 20 BCs, 1.7% of blood is collected at hospitals/health facilities where HCWs comprise an estimated 85% of all BDs. To collect more detailed demographic and occupational data which BDs do not routinely provide, 4151 BDs from hospitals/health facilities at 5 of 20 BCs in high/moderate HIV-1 prevalence areas completed anonymous questionnaires, Results: From Jan-Dec 1990, an estimated 33,600 HCWs donated at hospitals/health facilities served by the 20 BCs; 42 (0.1%) were seropositive. At initial interview, 20 (47.6%) reported behavioral risk factors; further follow-up is underway for the remaining 22 BDs. Using data from the 5 special study BCs, we estimated that of these 33,600 HCW-BDs, those with the greatest risk for occupational blood contact included physicians (3.4%), surgeons/obstetricians (0.9%), dentists (1.6%), nurses/nurse aides (21.5%), and lab technicians (5.9%); 82.8% had worked >3years in these occupations. Work locations for physicians, nurses, and lab technicians included medical/surgical floor (49.6%), intensive care unit (17.1X), emergency room (12.3%), and lab (25.7X). Conclusion: BDs offer a unique opportunity for surveillance of uncommon modes of HIV-1 transmission, including occupationally-acquired HIV infection, as persons with known risks for HIV-1 infection are actively discouraged from donating. Among the many HCWBDs studied thus far. HIV infection is uncommon. b G\

Page  38 S M.A.63 INFLUENCE OF CARBOHYDRATE MOIETIES ON THE IMMUNOGENICITY OF HIV-1 RECOMBINANT GP160 (rgp160). Benouad. Abdelazlz*; Gluckman, J.C'; Rochat, H"; Montagnier,L**; Bahraoui, E** *CERVI Pitie-Salp6triere Medical School, Paris, France "*CNRS UA 1179 Marseille, France, "'Unitd d'oncologie virale,lnstitut Pasteur, Paris, France Objective: To investigate the specificity and neutralizing capacity of antibodies raised against native, or partially or totally deglycosylated rgpl60 of HIV-1 Bru. Methods: Rabbits were immunized with rgpl60 or deglycosylated derivatives obtained by either neuraminidase, or mannosidase or Endo F-N glycanase. Specificity of antibodies thus elicited, were analysed by RIA against HIV-1 or HIV-2 radiolabeled envelope glycoproteins, and by ELISA against a panel of synthetic peptides. Anti-sera were further characterised for their ability to block HIV-1 binding to CD4 -positive cells and to inhibit syncytium formation. Results: Immunized rabbits produced antibodies that recognized rgp160 from HIV-1 in RIA or ELISA. Analysis of the reactivity of antibodies raised after carbohydrate removal against a panel of synthetic peptides indicated that serum reactivity to some peptides improved while the reactivity to others decreased or remained unchanged. For example, only antibodies raised against mannosidase-treated rgp160 failed to react with the V3 loop synthetic peptide of gp120. Rabbits immunized with desialylated rgpl60 generated antibodies able to recognize, with high titers, not only rgpl60 from HIV-1 but also rgpl40 from HIV-2. Although all antisera produced against glycosylated or deglycosylated rgpl60 were able to block HIV-1 binding to CD4-positive cells in vitro, only antibodies raised against desialylated rgpl60 totally inhibited syncytium formation between HIV-1 infected cells and non-infected CD4-positive cells. Conclusion: These results indicate that carbohydrate moieties of HIV-1 rgp160 can modulate the specificity of the immune response as demonstrated by the large cross reactivity of desialylated rgpl60 and by the variation of the reactivity to V3 loop of antibodies raised against different forms of deglycosylated rgpl60. Such information can be used to develop an efficient vaccine against HIV. NOTES M.A.64 DEVELOPMENT OF AN AIDS VACCINE CANDIDATE: RECOMBINANT BCG VACCINE VEHICLES INDUCE HUMORAL AND CELL-MEDIATED IMMUNE RESPONSES TO HIV ANTIGENS Aldovini. Anna; Young, Richard Whitehead Institute for Biomedical Research, Department of Biology, Massachusetts Institute of Technology, Cambridge, MA USA We have usr I the potent immunological properties of the human tuberculosis vaccine Mycobacterium bovis BCG trdevelop a live recombinant vaccine vehicle that can produce selected pathogen proteins and stimulate the mammalian immune system to respond to them. BCG has a number of features that make it a particularly attractive live recombinant vaccine vehicle: BCG and other mycobacteria are highly effective adjuvants, BCG has a long record of safe use in man, it is one of the few vaccines that can be given at birth, it engenders long-lived immune responses with only a single dose, and there is a worldwide distribution network with experience in BCG vaccination. Genes encoding HIV1 gag, pol, and env precursor proteins were placed under the control of the mycobacterial hsp70 promoter and ribosome binding site and were introduced into BCG on autonomously replicating plasmids. In the recombinant bacteria, the HIV1 proteins accounted for up to 0.1% of total protein. When injected into mice, recombinant BCG expressing HIV gag stimulated both humoral and cell-mediated immune responses to the viral protein. These results demonstrate that HIV1 structural proteins can be expressed in BCG recombinants and that these foreign polypeptides can be immunogenic. Features of this vehicle may be of particular importance for inducing strong and persistent immune responses against HIV1. NOTES C> C> G\ G\ M.A.65 PROMISCUITY OF CYTOTOXIC T-CELL EPITOPES FROM THE HIV-1 ENVELOPE Berzofsky. Jav A., and Shirai, M. Metabolism Branch, NCI, NIH, Bethesda, MD, USA. Objective: We observed that an immunodominant cytotoxic T lymphocyte (CTL) epitope of HIV-1 gpl60, P18 (residues 315-329), is presented by class I major histocompatibility complex (MHC) molecules to helper T cells as well as by class I MHC molecules to CTL, and that several peptides of gpl60 identified as helper epitopes also sensitized targets for lysis by human CD8+ CTL. We therefore asked in the mouse how broad a range of class I MHC molecules, if any, could present these peptides. Methods: Spleen cells from mice immunized with recombinant vaccinia expressing gpl60 of HIV-1 IIfB were restimulated in culture with helper peptides T (428-443) or HP53 (also known as TH4.1, 834-848), or with P18, plus IL-2. Resulting effector cells were tested for lysis of autologous targets treated with the relevant peptide and transfectants or vaccinia-infected cells expressing the whole gpl60 molecule. Results: P18 was presented by 5 different class I MHC molecules in the mouse. Helper peptide HP53 was also found to be presented by 4 different class I MHC molecules to CTL and TI was presented by 2 class I MHC molecules to CTL. The CTL were CD8+ and CD4-, and killed targets expressing endogenous whole gpl60 or pulsed with peptide. Overlapping and mutant peptides showed remarkable similarity in the core of the peptide required for CTL recognition of a peptide presented by distinct class I molecules, but more subtle differences in fine specificity were found. Conclusions: The 2 helper epitopes tested were also recognized by CD8+ CTL with class I MHC molecules. One of these and the CTL epitope P18 showed a striking degree of promiscuity in the range of class I molecules that could present them. For all class I molecules, the core peptide presented was approximately the same, indicating that these are not merely adjacent or overlapping epitopes contained in the same peptide. These promiscuous epitopes should be useful in a vaccine to induce CTL immunity in a an MHC diverse population. The ability to also elicit T help may give them a dual role in such a vaccine. M.A.66 Induction of anti HIV antibodies by imunization with monoclonal anti-idiotypes. Habib Zaghouani and Constantin Bona Department of Microbiology, Mount Sinai School of Medicine, New York USA Anti-idiotypes which possess the internal image of an antigen can induce protective humoral immunity toward bacterial, viral and parasitic infections. Herein, we demonstrate antigen mimicry by monoclonal anti-idiotypes to a distinct epitope of the HIV envelope protein which is defined by a synthetic peptide. This peptide corresponding to amino acid residues 503-535 of HIV-1 IIIB gpl60 induced antibodies in three mammalian species which interacted with HIV-1 gpl20 and inhibited in vitro syncytia formation caused by HIV-1 IIIB and MN isolates. In addition, antibodies to this epitope were purified from the sera of HIV infected humans using the 503-535 peptide. Three monoclonal anti-idiotypes were generated against rabbit anti-gpl20 antibodies specific for the 503-535 peptide. These anti-idiotypes recognize an interspecies cross-reactive idiotype expressed mouse, chimpanzee, baboon, rabbit and human anti-gpl20 antibodies specific for the 503-535 peptide. The interaction with the cross-reactive idiotypes is inhibited by both synthetic peptide and HIV-1 gpl60. Furthermore, rabbits immunized with the monoclonal anti-idiotypes produced antibodies which also bind both HIV-1 gpl20 and gpl60 and recognized the epitope defined by the 503-535 peptide. CI C>

Page  39 M.A.67 PROGRESS REPORT IV ON AIDS VACCINE IN HUMAN: PHASE I CLINICAL TRIAL IN HIV INFECTED PATIENTS. Zagury, D.1,, Picard, O.1, Halbreich, A.', Bernard, J.2, Carelli, C., Bizinni, B.3, Salaun J.J.4, Lurhuma Z.5, Mbayo K.', Burny A.6, Frottier, J.1 and Imbert J.C.1 xUniversit4 Pierre et Marie Curie, Paris, France, 2Institut J. Godinot, Reims, France, 3Institut Pasteur, Paris, France, 4INRB, Kinshasa, Zaire, 'Universit6s Cliniques, Kinshasa, Zaire, 6U.L.B., Belgique. Vaccine trial phase I performed in AIDS patients, using autologous cells expressing HIV1 antigens was performed at the University Clinic of Kinshasa and at Hospital Saint Antoine. Clinical and biological benefit was observed'(a,), but no repair of specific cellular immunity was apparent as if 1) the benefit was due to restoration of natural immunity, the breakdown of which is responsible of opportunistic infections and 2) a peripheral immune suppression was blocking all specific immune reactions. The latter defect prompted us to evaluate whether antisuppressive immunization including fixed autologous suppressive cells would repair specific cellular immunity against multiple antigens. Preliminary clinical trial phase I done on 6 volunteers under vaccinetherapy showed that adding antisuppressive immunization to the ongoing vaccine cure done every 10-12 weeks (2) provided after over one year, an immune improvement as detected by T4 cell count and some cell mediated immunity restoration. In addition, this modified treatment did generate neither immediate side effect nor clinical complication. "Mbayo K..t.1., V& Conf6rence.urt 1 SIDA.n AfrLque. Kins..h.. 1990: bPicard 0.. et al.. Lancet 336:179, 1990. M.A.68 CONSUTCION OF INNOGENIC SYNTHETIC HIV VACCINE CANDIDATES C1harles Sia. *Matthews, T., Chong, P., Haynes, J., Rovinski, B., Cao, S-X. *Bolognesi, D. and Klein, M. Connaught Centre for Biotechnology Research, 1755 Steeles Avenue West, Willowdale, Ont. M2R 3T4, Canada. *Department of Surgery, Duke University Medical Centre, Durham, NC 27710, USA. Objective: To construct synthetic HIV vaccine candidates using defined T- and B-cell epitopes of HIV structural proteins. Methods: Chimeric oligopeptides containing either a gpl20 or a gp41 B-cell epitope fra the MN and LAV isolates, respectively, linked in tandem to either the N- or C-terminus of a defined p24 T-cell epitope, p24E (residues 292-306) were synthesized. Immunogenicity studies of the synthetic constructs were tested in inbred mice to determine the ability of the T-cell epitope to mediate anti-envelope antibody responses against the individual B-cell epitopes. Anti-B-cell epitope antibody levels were measured by peptide-specific Enzyme Imnunoassays, and the functional properties of the antibodies assessed by "in vitro" neutralization and syncytia formation inhibition assays. Results: The inmunogenicity of the major neutralizing epitope (V3 loop, residues 311 -327) was higher when the peptide was linked to the C-rather than the N-terminus of p24E, whereas the Bi-cell epitope, BE3 (residues 721-751), required to be strictly linked to the C-terminus of the T-cell epitope to elicit an anti-BE3 antibody response Only the antiserum raised against the p24E-V3 but not V3-p24E construct was found to manifest potent neutralizing and syncytia inhibition activites against the homologous (MN) virus. Conclusion: Epitope-polarity critically influences the antibody responses to HIV-1 envelope B-cell epitopes in a synthetic T-B peptide construct. NOTES NOTES M.A.69 CHARACTERIZATION OF HUMAN ANTIBODY BINDING SITES ON THE HIV-2 EXTERNAL ENVELOPE de Wolf Frank*, Meloen R.H.**, Bakker M.*, Barin F.***, Goudsmit J.* *Human Retrovirus Laboratory, University of Amsterdam, The Netherlands. *European Veterinary Laboratory, Lelystad, The Netherlands. ***Laboratoire de Virologie, CHRU, Hopital Bretonneau, Tours, France. Obective: to map linear antibody binding sites on the HIV-2 external glycoprotein (EGP) gpl25 and to study antibody reactivity to the most antigenic epitopes in sera of HIV-1- and HIV-2-infected individuals. Methods: HIV-2 immunoblot positive sera from 8 patients were used for antibody-reactive peptide scanning (PEPSCAN) using overlapping nonapeptides derived from the EGP of HIV-2Rod. Peptides representing PEPSCAN selected regions were synthezised and anitibody-reactivity to these peptides was measured in sera from 26 HIV-2 positive African and European individuals, 64 HIV-1 positive Africans and 55 Dutch homosexual men, as well as 49 HIV-1/-2 negative African and 48 Dutch controls. Results: Fifteen antibody-binding regions were identified, among which 2 amino terminal regions (E3,aal 18 -132; E4,aa125-141) and 4 carboxyl terminal regions (El l,aa303-324; E12,aa340-358; E14,aa436-452; E15,aa486-507) were the most antigenic. E3 and E4 are highly variable domains across SIV, HIV-2 and HIV1. E14 and E15 are highly conserved among HIV-2 strains, and had more aa's in common with SIV than with HIV-1. Sera of HIV-2 infected individuals bound the Ell and E15 peptides best (31%). The sera of HIV-1 infected Africans showed the highest levels of cross-reactivity to the HIV-2 peptides, while this was low in HIV-l-positive European sera. Using peptides covering the E15 epitope of HIV-1 and SIV, African HIV-2 positive sera showed cross-reactivity to E15 of HIV-I and SIV. Conclusions: 1) The immunodominant regions of the HIV-2 EGP (El i and E15) align with those of HIV-1. 2, The frequency with which the African HIV-1 positive sera react to the HIV-1 E15 domain is similar to that for the European sera, but African sera frequently cross-react to the SIV and HIV-2 E15 domains, showing only modest primary aa-conservation. M.A.70 HUMAN MONOCLONAL ANTIBODIES AGAINST THE PUTATIVE CD4 BINDING SITE AND THE V3 LOOP OF HIV gp120 ACT IN CONCERT TO NEUTRALIZE VIRUS Tilley, Shermaine A.*, Honnen, WJ.,* Racho, M.E.*, Hilgartner, M.**, & Pinter, A.* *Public Health Research Institute, New York, NY; **New York Hospital-Cornell Medical Center, New York, NY, USA. Objective: To develop human monoclonal antibodies (HuMAbs) against HIV with potent neutralizing and/or other anti-viral activities. Methods: HuMAbs were derived by EBV transformation of PBMC from asymtomatic, seropositive hemophiliacs. The specificity of HuMAbs was determined by ELISAs, Western blot and RIPA analyses, and immunofluorescence (IF) assays. Neutralization of virus was assessed 24 hours after infection by an IF focus assay. Results: At least 2 of our anti- env HuMAbs have potent neutralizing activity against HIV. One of these HuMAbs, 1125H (yl, 1), is inhibited in its binding to gpl20 by CD4, probably indicating that its epitope is in or near the CD4 binding site. The epitope of 1125H is destroyed by reduction, but not by removal of N-linked sugars under conditions that do not disrupt disulfide bonds. Thus, the epitope is conformationally determined and does not appear to involve carbohydrate. HuMAb 1125H has potent neutralizing activity against all of the HIV-1 strains which we have tested thus far MN, RF, SF-2, and IIIB. Our other neutralizing HuMAb, 4117C (yl, 1), is against the V3 loop and is reactive with V3 peptides from the MN, SF-2, and NY-5 isolates, in increasing order of preference. 4117C has been tested for neutralizing activity against the MN strain and has potent activity against it. Recently, we have compared the neutralizing activity against MN strain of the two HuMAbs combined to that of each HuMAb alone. Preliminary experiments indicate that, whereas approximately 10 tg/ml of each HuMAb alone is required to effect > 90% neutralization of 5 x 104 infectious units of virus added to 5 x 10s H9 cells, only 1.25 Ag/ml of an equimolar mixture of the two HuMAbs is required to obtain this level of neutralization. A fine, analysis of this phenomenon will delineate whether the two HuMAbs are actually acting synergistically. Conclusion: HuMAbs against different neutralizing epitopes of HIV gpl20, specifically the putative CD4 binding site and V3 loop, may act in concert, either synergistically or additively, to neutralize HIV. This indicates that combinations of such HuMAbs may be superior for passive immunotherapy to single HuMAbs and implies that a vaccine ideally should elicit Abs against more than one neutralizing epitope. G\ ~1 C3

Page  40 S M.A.71 HIV-1 LOAD IN THE BLOOD IS INVERSELY RELATED TO THE LEVEL OF REVERSE TRANSCRIPTASE INHIBITING (RTI) ANTIBODY. Imagawa, David T.*; Kobata H.*; Lee M*; Ueda M**; Javier B*; Russo F*; Aldrovandi G*; Detels R***. Harbor-UCLA Medical Center, Torrance, CA.; **Instituto Adolfo Lutz, Sao Paulo, Brazil; ***UCLA School of Public Health, Los Angeles, CA. and L.A. Multicenter AIDS Cohort Study (MACS), U.S.A. Objective: Infectious HIV-1 load in the peripheral blood mononuclear cells (PBMC) and titer of RTI in the corresponding plasma were evaluated in 54 asymptomatic seropositive homosexual men. Methods: Limiting dilutions of PBMC were cocultured with PHA- stimulated donor cells in a microculture procedure to quantitate the infectious HIV-1 load. RTI titers in the plasma were determined by purifying IgG with QAE-Sephadex (Pharmacia), diluting and testing for 50% inhibition of HTLV-IIIb RT activity. Results: There was an inverse relationship between the titer of RTI antibody and the infectious HIV-1 load. In 5 individuals, the HIV-1 level in PBMC was high (>500 TCID50/10~ PBMC) and all had low RTI aptibody of <1:20. In contrast, of 13 men in which HIV-1 was not detected with 2x10 PBMC, 10 had RTI titer of >1:20 (p-<.01). Results on number of samples with HIV-l load and RTI titers are summarized below: RTI ANTIBODY TITER NO. OF MEN % WITH RTI TCID50/106 CELLS <20 20 40 80 160 >160 ANTIBODY >500.0 5 0 50.0 9 2 1 25 5.0 10 3 1 1 33 0.5 3 2 2 2 67 <0.5 3 2 2 1 1 4 77 Conclusion: The above data show that as the RTI antibody titer increases, the viral load decreases. Thus the RTI antibody may have protective effect and needs to be considered in immunotheraov and in develooment of vaccines. NOTES M.A.72 MOUSE MONOCLONALANTIBODY (G3-4) DEFINESA NOVEL CONFORMATION-DEPENDENT EPITDPE ON gp120 IMPORTANT FOR NEUTRALIZATION OF DIVERGENT HIV-1 ISOLATES Ho. David D.*; Fung, M.S.C."; U. X.L.*; Sun, C.*; Chang, N.T."; Chang, T.W."; Sun. N.C." The Aaron Diamond AIDS Research Center and New York University School of Medicine, New York, New York, U.SA.; *Tanox Biosystems, Inc., Houston, Texas, U.S.A. Objective: To define domains in gp120 important for antibody neutralization of divergent HIV-1 isolates. Methods: G3-4 monocional antibody was generated by immunizing mice with purified gpl20 of HIV-1 (llIB), and it has been extensively characterized in radioimmunoprecipitation, neutralization, and gp120-CD4-binding inhibition assays, as well as epitope mapping studies. Renults: G3-4 reacted with the gp120 of many (including 1118 and RF) but not all HIV-1 isolates. It also efficiently neutralized both IlB1 and RF laboratory strains with 90% inhibitory doses (ID90) of 1.0 and 0.5 ug/mf, respectively. In addition, G3-4 neutralized 4 primary isolates with ID90 of approximately 0.5 to 6.0 ug/ml, while 5 other primary isolates were only partially neutralized by up to 20 ug/ml of antibody. In competition immunoassays, G3-4 and soluble CD4 were found to inhibit one another in binding to gp120. However, no competition was observed between G3-4 and mouse monoclonal antibodies directed to the V3 loop or the putative CD4-binding region of gp120. In particular, G3-4 did not compete with our human monoclonal antibody 15e ( Virol 1991;65:489), which identified the first conformational epitope on gp120 involved in group-specific neutralization of HIV-1 and in gpl20-CD4 binding. Epitope mapping studies on G3 -4 using synthetic or unglycosylated recombinant peptides were negative, suggesting that its epitope may be conformation-dependent. Indeed, this was confirmed by showing the loss of G3-4 serologic reactivity when gp120 was first denatured with dithiothreltol. Conlusions: G3-4 has identified the second conformation-dependent epitope on gp120 important for the neutralization of divergent HIV-1 Isolates. Experiments are now underway to examine the spatial relationship of the G3-4 epitope with the V3 loop and the.CD4-binding region. NOTES '] N] M.A.73 v3 EPITOPES CAN MEDIATE ANTIBODY NEUTRALIZATION AND ENHANCE MENT OF HIV-1 INFECTION. Kiks. Srisakul C. Wu, W.W.; and Levy, J.A. University of California, School of Medicine, San Francisco, CA, USA. Objective: To investigate the mechanism of strain-specific antibody neutralization and enhancement of infection by HIV-1 strains. The studies were undertaken because sera from humans and primates immunized with HIV-1 envelope glycoproteins have been found to have dual properties. Methods: Human anti V3 monoclonal antibodies (MAb) were tested for their neutralizing and/or enhancing activity against three HIV-l strains: the North American predominant SF2 strain, and two brain derived isolates, SF162 and SF128A. Neutralization was defined as 67% reduction in virus replication; enhancement was defined by a two-fold or greater increase in virus progeny production in culture. Interviral envelope recombinant virsuses using SF2 and SF128A were also tested for sensitivity to neutralization and enhancement with specific monoclonal antibodies. Results: None of the MAbs showed any effect on the SF162 strain. Some human MAbs neutralized SF2 and some showed enhancement with SF128A. A noteworthy observation was the neutralization (with SF2) and enhancement (with 128A) with the same human MAb. The sequences at the corresponding linear epitopes of V3 in these three HIV-1 strains are different by only 1 or 2 amino acids. Other MAb studies with the recombinant viruses suggested the involvement of regions outside the V3 epitope in determining the viral sensitivities to serum antibodies. Conclusions: These results are the first demonstrating that human MAb to the V3 loop of HIV-1 gp120 can be involved in both viral neutralization and enhancement. The difference in the biologic effect depends on the primary sequence within an epitope and perhaps other regions outside this reactive site. M.A.74 INHIBITION OF HIV-l INFECTION BY IMMUNOLOGICAL REAGENTS. Ugen K.E*., Goedert J.**,, Kieber-Emmons T*., Merva,M.*, Boyer, J.*,. Williams W.V and Wciner. David B** The Wistar Institute*, Department of Medicine***, University of Pennsylvania, Philadelphia, PA., Viral Epidemology Section**, NCI, Bethesda MD. USA. Obiective: Understanding the contribution of the humoral immune response to the functions of the HIV envelope in HIV mediated viral entry processes. Anti-syncytial and group specific neutralization activity are observed most frequently in the asymptomatic patient population. High immune responses to the V3 loop region or other regions as defined with synthetic peptides and the ability of several specific peptides corresponding to defined regions of gpl20 and 41, including the V3 loop region, to induce protective in vitro immune responses in experimental animals in general supports this contention. A correlation exists between specific humoral immune responses in HIV infected pregnant women and a lack of transmission of HIV infection to the fetus. Studies: We have continued to examine the entry related functions of the HIV- (HXB2 isolate) envelope using a library of 110 envelope derived synthetic peptides screened against two selected groups of patients 1) adult patients with high neutralizingand antisyncytial activity, and 2) a maternal sera panel divided into mothers who do and do not transmit HIV infection to their fetuses. Based on our mapping studies we raised peptide based antisera to regions of the HIV-1 viral envelope which we have observed reactivity to in these protective patient seras. A subset of our antisera outside the neutralizing loop appears to disrupt entry events of HIV-1 and may have protective value in immunization strategies. One epitope in particular maps to the carboxy region of gpl20 and clearly elicits very significant neutralizing activity. Data will be presented describing 14 other now clearly defined areas which do not contain similar activity. We will also present data ona large number of other peptide defined regions. Importantly we have reconstructed the group specific protective activity of patient sera through defined combinations of anti-peptide sera. Conclusions: Our site specific serology studies will help in the design of immunological reagents which block HIV entry and offer broadly protective anti-HIV immune responses. QnI

Page  41 M.A.75 PERSISTENT INAPPARENT HIV-1 INFECTION OF HUMAN NEUROBLASTOMA CELLS Vesanen, Mika; Linna, Tuuli; Vaheri, Antti Department of Virology University of Helsinki, Helsinki, Finland We have studied human immunodeficiency virus type 1 (HIV-1) Infection in three different human neuroblastoma cell lines; SK-N-MC, IMR-32 and SH-SYSY. In all of these cell lines the infection became productive. However, the virus expression was different as determined by the p24 antigen capture assays from culture supernatants and immunochemical (APAAP) staining of cells. The medium of SK-N-MC cells contained -300 pg p24 antigen per 106 cells, 0.1-1 % of the cells were P24 -antigen positive and characteristic genomic and subgenomic HIV mRNA species were seen in Northern blotting. In infected IMR-32 and SH-SY5Y cell cultures, the HIV-1 production was below the level of detection. However, infectious virus was found by inoculating cultures of the lymphoid cell 08166 with the cell-free supernatant fluid from the neuroblastoma cultures. The lymphoid cells became positive within one week. Moreover, phytohemagglutinin-stimulated normal human lymphocytes produced virus, if cocultured with any of the three infected neuroblastoma cell lines. Also the production of the virus was upregulated in these cell lines, when stimulated with tumor necrosis factor a (TNFa) based on our mRNA and p24 enzyme immunoassay studies. The infection was persistent and has been followed, using the above techniques, for 4 months in the case of SK-N-MC and IMR-32 cells and 6 months in the case of SH-SY5Y cells. During these periods, no alterations in cell morphology, viability, or proliferative capacity were seen. All three neuroblastoma lines were negative for the CD4 receptor mRNA according to Northern hybridization and RNase protection assays. We conclude that HIV-1 produces persistent and inapparent infection in human neuroblastoma cells, using a CD4-independent mechanism of entry to the cells. NOTES M.A.76 NEUTRALIZATION OF HIV-1 INFECTION BY ANTI-V3 LOOP ANTIBODY IS DEPENDENT UPON THE HOST CELL OF THE VIRAL INOCULUM Ohash, t.*r Popovic, M., Liu, Y.**, Hunter, E.**, Javaherian, K.***, Eddy, G.**, and Gartner.Suzanne**. *Fujisaki Institute, Okayama, Japan; **Henry M. Jackson Foundation, Rockville, MD, USA; ***Repligen Corp., Cambridge, MA, USA. Objective: We sought to determine whether antibodies capable of neutralizing virus infection of T cells can also influence infection of mononuclear phagocytes. Methods: HIV-1Lw-PBM, an isolate from a lab worker infected with HTLV-IIIB was used. To generate viral stocks, early passage virus was transmitted to monocyte/ macrophages (MM), T cells and H9 cells. Goat antisera prepared against peptides corresponding to the V3 loops of HTLV-IIIB, -RF, -WMJ-2, -MN and -SC were tested in comparison with AIDS patients' sera. Viral inocula containing 40 TCID50 were used in microwell neutralization assays, and infection determined by p24 antigen assay. Results: Antibodies against V3 loop peptides as well as AIDS patient's sera were able to neutralize infection of H9 cells regardless of whether the viral inoculum was derived from normal T cells, MM or H9 cells. In contrast, while antibodies to V3 loop peptides were able to neutralize virus infection of normal MM when the viral inoculum was derived from H9 cells, they were unable to do so when the viral inoculum was derived from normal T cells or MM. Antibodies present in patient sera were able to neutralize virus infection on MM regardless of the host cell of origin of the viral inoculum. Conclusions: These results suggest that the host cell of the virus can influence the neutralizing capacity of antibodies against an immunodominant domain of gpl20. They also bring into question the practice of using only neoplastic cells as either a source of viral inoculum or as target cells in neutralization assays. NOTES M.A.77 SYNTHETIC PEPTIDES DERIVED FRM THE PRINCIPAL NEUTRALIZING DOMAIN (PND) OF HMAN IWUNIDEFICENCY VIRUS TYPE 1 (HIV-1) ENHANCE HIV-I INFECTION. De Rossi A., Pasti M., Mamano F., Panozzo M., 0Dettin M., ~Di Bello C., Chieco-Bianchi L. Institute of Oncology, ~ Institute of Industrial Chemistry, University of Padova, Italy. Objective: To investigate biological properties of PIN from different HIV-1 strains. Methods: 24-ner peptides derived from PND (301-324 aminoacid sequence) of different HIV-1 strains were synthesized on solid support using a peptide synthetizer (Applied Biosystems 431A). Target cell (MDLT-3 and H938) infection with MN, IIIB, and F HIV-1 strains was carried out in the presence of scalar amounts of peptides. Virus yield was evaluated by reverse transcriptase assay and CAT assay. Results:Peptide designed from the V3-PND of HIV-1 MN strain enhanced viral infection. The activity of HIV-1 HTLV-IIIB- derived peptide was at least tenfold less efficient, while no enhancing effect was shown by HIV-1 RF specific peptide. This enhancing effect occurred in the early steps of viral infection and was not strain-restricted in that both MN- and IIIB-derived peptides were able to increase heterologous virus expression, including that of the RF strain. Both MN- and IIIB- derived peptides were able to block in a dose-dependent manner the CD4 down-regulation induced by the monosialoganglioside GM1, while peptides without enhancing effect did not. MN-derived peptide induced an increase in CD4 epitopes on the cell mrrbrane, and was able to partially block CD4 internalization induced by the phorbol ester 12-0 -tetradecanoylphorbol 13 acetate. Conclusions: These findings suggest that the viral enhancement induced by the PND-derived peptides occurred through a mechanism involving cellular CD4 viral receptor.. CD4 INDEPENDENT FUSION AND INFECTION BY HIV-2 STRAINS ENHANCEMENT BY SOLUBLE M.A.78 CD4. Clapham Paul, McKnight Aine, Weiss Robin. Chester Beatty Laboratories, Institute of Cancer Research, Fulham Road, London SW3 6JB, UK. Introduction We have characterized HIV-2 fusion and infection of CD4-negative human cell lines. Methods HIV-2 strains used were LAV-2ROD, SBL6669, CBL-20, CBL-21, CBL-22 and CBL-23. Human cells included RD/TE671 rhabdomyosarcoma, U87 glioma, HeLa carcinoma, Raji and Daudi B cell lines. HIV-2 fusion was tested by overnight cocultivation of producer cells with appropriate test cells. Cell-free HIV-2 infectivity was tiltated by the TCID method. Recombinant baculovirus LAV-2ROD env gpll10 was labelled with I using Enzymobeads (BioRad). Results A substrain of LAV-2 (LAV-2/B) efficiently induced fusion of several CD4- human cell lines. SBL6669 and CB-22 also induced syncytia but less efficiently. RD/TE671 and Daudi cells were most sensitive to fusion, which was not blocked by soluble CD4 (sCD4) or by inhibitory CD4 monoclonal antibodies, but was specifically inhibited by HIV-2' human sera. sCD4 either induced or enhanced fusion by all HIV-2 strains tested. in cell-free infection by HIV-2, the 10 TCID of LAV-2/B on CD4 cells was only 10-fold less than that-on the CD4C T-cell line, C866. CBL-20 was only infectious for CD4- cells after treatment with sCD4. HIV-2 envelope glycoprotein (gp110) bound to CD4" RD/TE671 cells. Binding was enhanced by sCD4, but inhibited by HIV-2* sera. Again, binding of gp110.was 10-fold higher on CD4' cells. Discussion and Conclusion HIV-2 strains can fuse and infect CD4- human cell lines. This infection is probably mediated by gp110 binding to an unknown cell surface receptor, and the gp110 domain involved may become more accessible following sCD4 treatment. We do not as yet know whether the alternative receptor is also required for CD4-dependent entry. 0

Page  42 ) M.A.79 MATERNAL-FETAL HIV-TRANSMISSION: PLACENTAL CELLS CD4 EXPRESSION AND PERMISSIVITY TO INFECTION WITH HIV. DAVID ranois J.*, TRAN Hic C.**, AUTRAN Briit**", RAPHAEL Marine****, MENU Elizabeth *,HS1 Bac L. ***,** BARRE-SINOUSSI Franoisc **& CHAOUATO6&rad. *INSERM U 262. in. Univ. Bauddocqe, Paris, t**L oieatb w ddBiogie des d iovkw, PaJur,Pari, s *UA CNRS 196 EcotiIe Cdthaiure, CHU PitiSlpetri,Pari,*' *U 210 INSERM, CRHU doNima, FRANC Ob)te. Toinvestiate the mo lari ehanisms alowing dhe ant pa a of HIV, we studied: A- he placental Fc-Rcptors (E-R), possibly involved in the placetal passage ofIgO-complexed viral particles B- The expression of CD4 and the pcissivity of plactl clls to infection with HIV. Methods: M nbrane antigen expession was detected by imuno-smainng of tcnmplacent tissue sections and synctiorophoblast(SThspecificaatigeasidentifed byPACSanalysis,allowinganassessentof cel purity, of the enriched tpoblast population isolated and maintained in cell cultu. The vi infection with V-brof these cells and of huan trophob derived cho-rc o cells were studied in direct culture as well as by cocultwe with PHA-stinlated PBL. Results: Several c-R werrdctctnd, with distint locatios in the pliacnalisu. After hIg-saturation, the ST-associatcd c-R was lead by ithigher affiniy aonuncmlexd d Id The CD4 molecule was colocalised th specific tophobast nak at th e ST srfac of the vi. Tis onassia wa conobn~ d by FACS, as well as by gpl20 & gpl60 Mg binding and it intion with ani CD4 mAbs1 The init infectio of rophobiast cells with H[V-l'bm' was vified by PCR and confirmed by in hykldsanon. The productive infcon was tested bynvae nrnsripm (RI) ativy andth me ofp24 Agin the cultm e tpnamamaot Both CD4dctecon andIV inf web obtained with c cacion cllies which can be pductivy infected with several HIV-1 sains and with HIV-2. Discu m: Ou dam songly upport the possabity thatSTcan beinectted, which may be an inprtat step for the ntra-pacenal spad of HV. Lasty, the i j FHIV-infeetinofplacnal cellsalso provids a suitable model fo further studies concening placental HIV-tansaission and its preventi NOTES M.A.80 REGULATION OF SILENT AND PRODUCTIVE HIV-1 INFECTION IN PRIMARY MONOCYTES BY TWO DISTINCT GENETIC DETERMINANTS Westerveft. Peter, Trowbridge, D B*, Gendelman, H E**, & Ratner, L* *Departments of Medicine & Molecular Microbiology, Washington University School of Medicine, St. Louis, MO and "Walter Reed Army Research Institute, Rockville, MD OBJECTIVE: Determination of the molecular basis for HIV-1 tropism for primary human monocytes. METHODS: Recombinant proviral clones were constructed by reciprocal DNA fragment exchange between a monocyte-tropic done (ADA) and two full-length molecular clones (HXB2 and NL4-3) incapable of infection or replication in primary monocytes ("no infection" phenotype). Infection and replication were assayed for by serial determinations of reverse transcriptase activity in culture supematants and by virus recovery from infected monocytes following cocultivation with fresh uninfected peripheral blood mononuclear cells (PBMC's). RESULTS: Our data indicate that at least two viral genes, ev and ypL play central roles in HIV-1 infection and replication in monocytes and mediate the expression of three distinct replication phenotypes. The UnY determinant has been fine mapped to amino acids #240-333 of the mature ADA gp120 protein, encoding the entire V3 loop but excluding the CD4 binding domain. Virions derived from recombinant HXB2-based clones containing the env determinant infected primary monocytes and were recovered consistently at 24 days post-infection by co-cultivation with fresh uninfected PBMC's, despite their failure to generate detectable virus replication levels in monocytes ("silent infection"). Incorporation of the cv determinant in recombinant NL4-3-based clones, which unlike HXB2 encode a full-length vp protein, resulted in the generation of virions which replicated to high levels in primary monocytes ("productive infection"). Inactivation of Ypr by mutagenesis converted the productive infection phenotype in monocytes to that of silent infection. CONCLUSIONS: These findings suggest that an env determinant is required for HIV-1 infection of primary monocytes, while subsequent replication levels (silent vs. productive infection) are determined, at least in part, by v=. Further elucidation of these phenotypes in vitro should contribute toward a better understanding of In vivo phenomena such as disease latency and progression to AIDS. NOTES ~ C>I C>j C>I h^ hj rj

Page  43 M.B.81 CLINICAL AND AUTOPTICAL DIAGNOSIS IN AIDS: AUTOPSY WARNS THE CLINICAL APPROACH d'Arminio Monforte A*, Vago L", Lazzarin Adriano', Formenti T*, Rizzardi GPA, Guzzetti S", Musicco M~, Costanzi G**, Moroni M* 'Clin Inf Dis, **V Chair Pathol and Al Chair Clin Med, Univ Milan; *CNR-ITBA Milan, Italy Objective: to relate clinical and autoptical observations in AIDS patients, stratified according to the indicator diseases and the lenght of survival. Patients: 533 patients with AIDS since 1984 to Dec 31th, 1990 were included. 341 (64%) were PDA or exIPDA, 150 (28.1%) homosexual men, 33 (6.2%) heterosexual and 9 (1.7%) were infected by blood transfusions. The Kaplan Meier test was used to generate survival curves. Results: 328 (61.5%) patients died. The overall cumulative survival rate was 45% at 1st year and 20% at 2nd year. Survival was stratified according to the risk behaviour and indicator-disease. Autopsies were performed in 230 cases: 63 cases died within 2 months (group 1),111 within 1 year (group 2) and 56 1 year after from AIDS diagnosis (group 3). In 12% of autoptically examined cases at least 3 opportunistic infections (0.1.) and neoplasms diagnostic of AIDS were found; however bacterial pneumonia, not included in the diagnostic criteria of AIDS, represented the cause of death in 32% of cases (73/230). Among the most frequent 0.1., PCP, diagnosed in 33.5% of cases during life, accounted only for 12% of deaths (281230); moreover nearly half out of autoptically demonstrated PCP (16/28) belonged to group 1. On the contrary, the prevalence of CMV infection and neoplasms increased with the lenght of survival. CMV, whose overall prevalence was 43.5% (100/230) was observed in 29% of the cases in group 1, 48% in group 2 and 52% in group 3. Lymphomas were observed in 49/230 autopsies (21.3%) and Kaposi's sarcoma in 34/230 (14.8%). The prevalence of neoplasms was 27% in group 1, 34% in group 2 and 49% in group 3. The overall percentage of misdiagnosed pathologies was 49% (70% in group 3). CMV infections were misdiagnosed in nearly half out of the cases (46/100), lymphomas in 30% and micotic pulmonary diseases in 16/17 cases. Conclusions; despite the improvement in diagnosis and treatment of PCP, the overall differences between clinical and autoptical findings remain still significant, particularly for the misdiagnosis of some pathologies such as CMV infections, micotic diseases and lymphomas. NOTES M.B.82 BACTEREMIA COMPLICATING HIV-INFECTION; frequency, spectrum and outcome. Phillips, Peter*; Le, A*; Duncan, J**; Oh, H*; Schechter, MT*; Montaner, JSG.* *AIDS Research Programme, St Paul's Hospital, University of British Columbia, Vancouver, B.C., Canada, *University of Aberdeen, Scotland. Objective: To determine the frequency, spectrum, and outcome of bacteremia in HIV-infected individuals admitted to St. Paul's hospital in the 5 year period ending December, 1990. Methods: Retrospective review of all hospital admissions of HIV-infected individuals. Bacteremia was defined as 2 or more separate blood cultures positive for the same organism or a single positive blood culture plus the same organism isolated from another site. Results: 58 episodes of bacteremia in 49 patients were identified out of 1895 eligible hospitalizations during the study period. These represented 1.9, 1.7, 3.5, 3.3, 4.3% of the total yearly HIV-related admissions in 1986 to 1990 respectively. Bacterial isolates were: Staphylococcus aureus (17); Staphylococcus epidermidis (7); Streptococcus pneumoniae (7); other streptococci (5); Listeria monocytogenes (2); Pseudomonas spp. (5); nontyphi Salmonella spp. (4); E. coli (4); Klebsiella spp. (4); and other gram negative bacilli (3). Bacteremic episodes were community acquired, nosocomiat and of unknown origin in 33,16, and 9 respectively. The sources of infection included: intravascular catheters (14); respiratory tract (12); gastrointestinal tract (4); urine (4); wound (3); and unknown (21). Bacteremia contributed to a fatal outcome in 7 (18%) of 38 episodes caused by gram positive organisms compared to 9 (45%) of 20 episodes due to gram negative organisms (p=O.03). Bacteremia contributed to a fatal outcome in 5 (56%) of 9 episodes associated with granulocytopenia (< 1.0 x 109/L) compared to 10 (20%) of 49 among non-granulocytopenic patients (p=0.026). Conclusions: Bacteremic illness in HIV-infected patients is becoming more common. This is associated with significant morbidity and mortality, particularly when due to gram negative bacteria or in the face of granulocytopenia. Supported in part by the National Health and Research Development Programme. Health & Welfare, CANADA. NOTES toj C> C> I C?> ~" TOXOPLASMIC OCULAR INFECTION IN PATIENTS WITH AIDS. M.B.83 Accorinti Massimo, Pivetti-Pezzi P., Tamburi S., Ciapparoni V., D'Offizi G.P.*, Tilia G.**, Catania S.***. Institute of Ophthalmology, *Dept. All. Clin. Immunol., **Dept. of Neurology ***Institute of Infectious Disease, University La Sapienza, Rome, Italy. Objectives: to define the characteristic clinical features, the therapeutical regimens and a possible neurological involvement in ocular toxoplasmosis. Methods: 152 AIDS patients were submitted to an ophthalmological examination. All the subjects with ocular toxoplasmosis underwent also a neurological evaluation including CT or MNR and the study of visual evoked potentials. The treatment was carried-out with sulfadiazine 1 gr + pyramethamine 50 mg. Results: focal toxoplasmic retinochoroiditis and an heavy vitreal gaze were found in 5 patients (12.5% of opportunistic retinal disease, 3.28% of all AIDS cases), a severe anterior uvea involvement in 4/5. Furthermore 2 patients (40%) showed typical CNS lesions. The treatment was effective to control the infection in all the cases but only one patient had a good final visual acuity (6/10). Conclusions: ocular toxoplasmosis represents an outcoming infection preceded only by CMV retinitis; the vitreous and anterior uvea severe involvement are its peculiar findings. The poor visual prognosis is due to the frequent localization at the eye posterior pole so the treatment, even if effective, rarely prevents the damage of visual function. A CNS involvement should be researched in all those patients. M.B.84 MORTALITY IN HIV-1 POSITIVE TUBERCULOSIS (TB) CASES, NAIROBI, KENYA. Nunn P*,**, Odhiambo Joseph*, Brindle R*, Carpenter L**, Gathua S*** Wasunna K*, Were J*, McAdam K**. *Kenya Medical Research Institute, Nairobi, Kenya. **London School of Hygiene and Tropical Medicine, London, UK. '**Infectious Diseases Hospital (IDH), Nairobi, Kenya. Objectives: To compare mortality rates and causes among HIV+ and HIV- TB cases (PTBC and NTBC respectively). Methods: 107 HIV PTBC and 174 HIV NTBC were enrolled in a longitudinal follow up study. Kaplan-Meier survival curves were estimated for both groups. Risk factors for mortality by 6 months were determined by proportional hazards regression models, Results: Mortality rates were 39/77 person years of observation (PYO) and 13/157 PYO in PTBC and NTBC patients respectively. Mortality by 6 months was greater inPTBC than in NTBC when adjusted for age and sex (RR = 3.45; 95% CI = 1.6-7.3). Of 39 deaths in PTBC, TB caused 6 (15%), bacteraemia 6 (15%), various other infections 6 (15%), toxic epidermal necrolysis 3 (8%) and miscellaneous causes in 9 (23%). Data were insufficient in 9 (23%). Of 13 deaths in NTBC, TB caused 7 (54%), brain abscess, oesophogeal cancer and renal failure 1 each and data were inadequate in 3 (23%). Risk factors for death before 6 months in PTBC were age >30 years, male sex, <1 year of education, lymphadenopathy, oral candidiasis and presence of epitrochlear nodes. In NTBC risk factors include age >30 years, (1 year of education, weight <50kgs, miliary disease, extent of pulmonary disease, total white cell count >10x109/l, haemoglobin <11g/dl, Beta2 microglubulin >7.6 mmols/1 and albumin <3g/l. Conclusions: Mortality is significantly higher among PTBC than NTBC. The excess is largely due to treatable non-AIDS defining non-TB infections. TB staff in developing countries must be alerted to this Droblem.

Page  44 SM.B85 ACTG 095: A RANDOMIZED COMPARISON OF FOSCARNET (FOS) VS. VIDARABINE SM.B.85 (ARA-A) FOR TREATMENT OF ACYCLOVIR (ACV)-RESISTANT MUCOCUTANEOUS HERPES SIMPLEX VIRUS (HSV) INFECTION IN PATIENTS WITH AIDS. afrin, Sharon1; Crumpacker C2; Chatis P2; Davis R3; Hafner R4; Rush J'; Kessler H5; Landry B4; Mill '. 1-University of California, San Francisco CA; 2-Beth Israel Hospital Boston MA; 3-Harvard School of ublic Health; 4-National Institute of Allergy and Infectious Diseases; 5-Rush Medical College Chicago IL. Objective: The therapy of ACV-resistant HSV has not been studied in a controlled fashion. We compared FOS to ARA-A in a randomized, multicenter trial. Methods: Patients with HSV lesions unresponsive to therapy with IV ACV for a minimum of 10 d and hich showed in vitro resistance to ACV were randomized to receive FOS (40 mg/kg q8h) or ARA-A (15 g/kg/d). Therapy was administered for 10-42 d, with an option for crossover to the opposite agent at lay 10 or later for therapeutic failure or serious toxicity. Results: The study was stopped prematurely due to excessive toxicity and inferior efficacy of ARA-A& Time o complete healing (p=.01), time to 50% reduction in lesion size (p=.01), time to halving of pain =.004), time to pain resolution (p = 0.12), and time to cessation of viral shedding (p=.006) were less in he 8 patients randomized to FOS than in the 6 assigned to ARA-A. Three of 6 pts had neurotoxicity (confusion, myoclonus) on ARA-A. No patient discontinued FOS for toxicity. Every patient who healed tltimately had a recurrence ofACV-resistant HSV, at a median of 42.5 days after discontinuation of FOS (range, 14-191 days). Conclusion: FOS is more effective and less toxic than ARA-A for therapy ofACV-resistant HSV infection in AIDS patients. Future studies should evaluate optimal agents for prevention of recurrence of ACVresistant HSV once healed. NOTES M.B.86 A RANDOMIZED, ONTROLLED STUDY OF IMXED3IATE VS M.8 DEFERRED GANCICLOVIR THERAPY IN AIDS PATIENTS WITH CYTOMEGALOVIRUS PERIPHERAL RETINITIS Spector, Stephen Aj, Barker, C; Buhles, W**; Feinberg, J***; Montague, P; Weingeist, T; DeArmond, B; and the ACTG/Syntex Ganciclovir Treatment Group. U.C. San Diego, Syntex Research, NIAID, U"Univ. of Iowa, USA. Objectives: To evaluate the efficacy of intravenous ganciclovir (GCV) in treating peripheral CMV retinitis in AIDS patients. Methods: A randomized, multicenter, controlled trial was conducted in 42 subjects (39 males and 3 females) with CMV peripheral retinitis. Patients were randomized to receive GCV 5 mg/kg twice daily for 14 days followed by 5 mg/kg daily for 14 weeks or to deferred treatment. An independent ophthalmologic reading center evaluated retinal photographs for retinitis progression. Results: Of the 36 patients on study for >7 days, 13 received GCV and 23 deferred treatment. The mean + SE CD4T lymphocyte count was 35.9 + 47.4 for the GCV treatment group vs 33.1 + 38.9 for those receiving deferred treatment (p=.87). The median time to retinitis progression was 15.0 days for patients receiving deferred treatment (mean ~ SE = 22.9 + 4.3 days; range: 6-111 days) vs 50.5 days for patients receiving GCV (68.5 ~ 14.3 days; range: 12-143 days) (P=.011). conclusions: Most AIDS patients with CMV peripheral retinitis without therapy have rapidly progressive retinitis. GCV significantly delays progression of CMV peripheral retinitis in AIDS patients. NOTES to,bo Oc J j C> |> C> C>i

Page  45 M.C.87 A SINGLE, RAPID PEPTIDE-BASED IMMUNOASSAY FOR THE SIMULTANEOUS DETECTION OF ANTIBODIES TO HIV-1 AND HIV-2 Montana, J.P; Gosting, L.; Cole, C. A.; Monji, N.; Su, P.; Ferrera, C. and Coleman. P. F. Genetic Systems Corporation, Redmond, WA., USA. Objective: To develop a rapid (10 min.), synthetic peptide-based immunoassay for the separate and simultaneous detection of antibodies to HIV- 1 and HIV-2 for diagnostic purposes. Methods: Synthetic peptides representing highly conserved, immunodominant regions of HIV-1 gp41 and HIV-2 gp36 were adsorbed to latex beads. These beads and procedural control beads were individually spotted onto a nylon membrane in an asymmetrical 4-spot array. The membrane was incorporated into a plastic, molded device containing an absorbent pad andoverlayed with a disposable prefilter. Serum or plasma, diluted -1:11, was added to the GENIE test unit which was then washed with buffer. The immunochemical reactions were visualized by the sequential addition of enzyme-anti-human Ig conjugate and chromogenic substrate. The large antigen surface area, high epitope density and sample flow-through allow for very rapid and efficient immunochemical binding. Results: Summary of preclinical and clinical trial data. M.C.88 ALTERNATVE SCREENING AND SUPPLEMENTAL TESTING STRATEGIES FOR HIV-1 AND HIV-2 INFECTIONS BRATTEGAARD Kari *; Doorly R *; Kouadio J *; Maran M *; George R **; Gayle H ** De Cock KM *,**. * Projet RETRO-CI, Abidjan, Cote d'Ivoire, ** Centers for Disease *Control, Atlanta, USA. OBJECTIVE: (1) To evaluate rapid tests for HIV-1 and HIV-2 antibodies; 2) to evaluate use of synthetic peptide-based tests (SPT) as supplemental tests instead of Western blot (WB). METHODS: 1000 consecutive pregnant women in Abidjan were screened for HIV-1 and HIV-2 by reference strategy (ELISA/WB/SPT). All specimens were also tested by HIV-1/HIV-2 mixed rapid test (Abbott Testpack), and positives tested by SPT (Diagnostics Pasteur Peptilav 1-2). Sensitivity and specificity of rapid test, and of combination rapid test/ST were evaluated against reference diagnostic pathway. Genetic Systems Genie test (HfV-1-and HIV-2-specific rapid test) will also be evaluated. RESULTS: True prevalence was 9.6% HIV-1, 2.3% HIV-2, 1.1% HIV-1/HIV2. Sensitivity of Abbott Testpack was 130/130 (100%), specificity 792/870 (91%). Sensitivity of combination rapid test/SPT for HIV reactivity was 128/130 (98.5%), specificity 869/870 (99.9%), and positive predictive value (ppv) 128/129 (99.2%). Overall concordance between rapid test/SPT and reference strategy was 997/1000 (99.7%). False negative specimens on SPT lacked transmembrane antibody on WB. CONCLUSIONS; (1) Rapid test is adequately sensitive for screening; (2) ppv of rapid test/SPT combination is adequate for counselling when WB unavailable; (3) negative SPT after positive rapid tes needs further testing, ideally by WB. q CI 0 0 r^ O O CI G3 HTV-1: 3314 True Negatives 0 False Negative 288 True Positives 1 False Positive HIV-2: 3390 True Negatives 0 False Negative 112 True Positives 1 False Positive Sensitivities of 100% and 100% and specificities of 99.96% and 99.97% for HIV-1 and HIV-2, respectively, were calculated using immunoblot as the reference. Conclusion: This rapid, easily performed, peptide-based assay readily differentiates HIV-1 and HIV-2 seropositivity and is a useful diagnostic test. NOTES NOTES M.C. 89 he use of Saliva and Urine for HIV surveillance in Children S BURNS Sheila M* Robertson P** Yap P L*** Mok J Q** Parry J**** * Regional Virus Laboratory, ** Regional Infectious Diseases Cepartment, City Hospital **" Scottish National Blood Transfusion Service **4* Public Health Laboratory Service, London. UBJLCTIVE: Children born to HIV infected mothers require follow-up until their HIV status is determined, and this may take up to 16 months. A study was carried out on a group of known infected and noninfected children to examine the feasibility of using saliva and urine samples as an alternative to blood. SUBJECTS: Fourteen children aged between 3 and 7 years and born to HIV antibody positive mothers were studied after informed consent was obtained from their parents, METHODS: Saliva was collected by first obtaining the full cooperation of the child who was then asked to spit into a 3.5 cm diameter plastic pot with a screw lid. Salivettes were not used since it was assumed they were not yet widely available. Blood and urine were collected at the same time. HIV specific antibody was measured in all samples by IgG antibody capture ELISA (GAC ELISA) (Dr Parry, Colindale) for the detection of anti-HIV1 and HIV2 and also by "standard" ELISA (Abbott recombinant HIV1 / HIV2 EIA). There was insufficient neat saliva in some cases to test by both methods. Tests were carried out using the manufacturers protocol. For the GAC ELISA, two sampling volumes of urine were examined, 1U ul and 50 ul. RESULTS: In urine the standard ELISA could detect HIV specific antibody in 2 of B samples positive by SAC ELISA. An increase in sensitivity was obtained in the standard ELISA by using 50 ul rather than 10 ul of urine. Detection of HIV antibody in saliva was in complete correlation with serum antibody using GAC ELISA but not standard ELISA. One saliva sample from a confirmed positive child was negative in standard tLISA. CONCLUSIONS: Using a sensitive test method such as GAC ELISA, saliva samples proved an adequate alternative, noninvasive sample to collect in these children. The inability to detect HIV specific antibody in the urine of known infected children may be related to several factors, including declining levels of p24 antibody in symptomatic children. M.C.90 A WHO INTERNATIONAL QUALITY ASSESSMENT SCHEME FOR HIV ANTIBODY TESTING; RESULTS FROM THE SECOND DISTRIBUTION OF SERA Faucuex, A'; Snell, J.J.S.'*; Supran, E.M.*"; Tamashiro, H." "Global Programme on AIDS, World Health Organization, Geneva, Switzerland, ""Central Public Health Laboratory London, UK. Objective: To evaluate WHO International Quality Assessment Scheme for HIV-1 antibody testing which was established to monitor the quality of laboratory performance in HIV testing in member states. Methods: Twenty serum specimens, 10 of which contained antibody to HIV-1, were sent to 103 laboratories in six WHO Regions. Participants were asked to test the specimens by routine (screeening and confirmatory) methods used in their laboratories. Results: Of the antibody positive specimens, 98.2% were reported as positive and 1.8% as indeterminate. No false negative was reported. Of the antibody negative specimens, 90.3% were reported as negative, 1.3% as positive, and 8.4% as indeterminate. Most of the indeterminate reports were associated with one particular specimen. Conclusions: The WHO International Quality Assessment Scheme for HIV-1 is feasible. Performance of laboratories for diagnosis is presently being evaluated by this same system. The results suggest a number of practical applications for the field operations in national programmes.

Page  46 M.C91 PROPOSED WHO CRITERIA FOR INTERPRETING RESULTS FROM.C.9 WESTERN BLOT ASSAYS FOR HIV-1, HIV-2, HTLV-I/HTLV-II Tamashiro, Hiko; Karam, Marc; Esparza, Jose; WHO HIV-2 Collaborating Group Global Programme on AIDS, World Health Organization, Geneva, Switzerland Objective: To assess the WHO revised Western blot (WB) criteria for HIV-1, HIV-2, and HTLV-I/HTLV-II using sera collected from HIV-1 or HIV-2 endemic countries. Methods: 136 serum samples considered WB-positive for HIV-2 by national criteria were collected from 14 countries with endemic HIV-2 and retested blinded in 2 WHO reference centres to compare the revised criteria to criteria previously recommended. In addition, 4 other frequently used WB criteria were evaluated on 424 serum samples of HIV-1 from 4 groups (AIDS patients, other symptomatic patients, asymptomatic homosexual men, and volunteer blood donors). Results: The revised WHO criteria (i.e., at least 2 env bands) provide increased specificity with a slight loss of sensitivity and assure better discrimination betwen sera that show cross-reactivity to both HTV-1 and HIV-2 by WB. Conclusions: The revised criteria which require the presence of at least 2 env bands vary depending on the lot of WB reagent. In some lots, the 2 high molecular weight bands (gpl60 and gpl20) may not be sufficiently separated to appear on independent bands. In this case the single high molecular weight band must be accompanied by the transmembrane glycoprotein for the blot to be scored as positive. These criteria should be evaluated in the areas with different epidemiological patterns of retrovirus infections. NOTES M.92 RELIABLE DIAGNOSIS OF PEDIATRIC HIV INFECTION WITHIN FOUR MONTHS OF LIFE BY SCOMBINED USE OF POLYMERASE CHAIN REACTION AND VIRUS CULTURE Bfni Jgra'; Kind C'*; Jendis J'; Schipbach J* *Swiss National Center for Retroviruses, Swiss Federal Office of Health and University of Zurich, Zurich, "Swiss Neonatal AIDS Study Group and Department of Neonatology, Kantonsspital, St. Gallen Objective: To assess the value of PCR and virus culture in the early diagnosis of pediatric HIV-infection. Methods: Virus culture (VC) [Ho et al. NEJM 1989;321:1621-5], PCR screening with gag, and PCR confirmation with LTR and env primers (all PCR blind and in duplicate) were done with PBMC of so far 33 children born to HIV-pos mothers in 1990. Results: Pos results by both VC and PCR were obtained in 5114 (36%) children from whom > 2 samples were received. All pos results were obtained within 4 months of life and before clinical disease developed; 3 of these children were VC-pos (2 also PCR-pos) in the cord blood (CB) (all confirmed by follow-up). CB from 19 additional children for whom no follow-up samples were received to date were VC-pos in 1 case (5%). The combined minimal rate of HIV transmission is 6/33 (18%); some of the children of the second group are expected to turn out pos in follow-up. Quality Controls: Specificity of PCR better than 99.6%: No pos reactions with 99 HIV-negative control DNA samples tested along with the children's samples in a total of 278 reactions in 24 different experiments. No pos PCR reactions in 25 samples from children >12m age seroneg for HIV-specific IgG,M,A. Sensitivity of PCR: with plasmid DNA diluted in human DNA: 1 copy/sample; with clinical material 10 copies/lig DNA or better; 18/21 (86%) VC-pos samples were PCR-pos.; 15/16 (94%) PCR-pos samples were VC-pos; concordance of PCR and culture results was 101/108 (93.5%). Discussion and Conclusions: Reliable pos diagnosis of HIV-infection can be made within the first 4 months of life by the combined use of VC and PCR. The sensitivities of these methods are similar. A considerable fraction (4/6) of infected infants were already pos at birth, none of them was false-pos. NOTES ko cE M.C.93 MATERNAL HIV INFECTION AND LOW CD4/CD8 RATIO AS RISK FACTORS FOR PREMATURITY AND STILLBIRTH Temmerman. Marleen; Hawala, D.; Ndinya-Achola, J.O.; Plummer, F.; Plot, P. University of Nairobi, Kenya; University of Manitoba, Winnipeg, Canada; Institute of Tropical Medicine, Antwerp, Belgium. Objectives: To determine whether maternal HIV infection is a risk factor for poor obstetrical outcome (POO) defined as prematurity and stillbirth, and if so, to define risk factors for PO0 in HIV(+) women. Methods: Pregnant women less than 28 weeks pregnant were screened for HIV-antibody and syphilis seroreactivity at a City Commission antenatal clinic in Nairobi, Kenya. HIV seropositive women were referred to a research clinic, together with an age- and parity-matched seronegative control group. Data collection Included sociodemographic features; sexual, medical and obstetrical history; clinical signs and symptoms; and an ultrasound examination to verify the gestational age. Cervical and high vaginal swabs were taken for N, aonorrhoeae, C. trachatis and S. agalactiae. Routine antenatal follow-up was provided and patients were instructed to deliver In the same hospital. Maternal and neonatal data were collected at delivery and the ptacenta was kept In formalin. Mothers and babies were followed up to 6 weeks postpartum. Maternal blood for lymphocyte subsets was examined at 34 weeks pregnancy, at delivery and at 6 weeks postpartum. Results: The maternal HIV-1 seroprevaLence was 7.2 % (280/3900). '5.2 I of the patients had a reactive syphiLis serology (203/3900). 212 HIV(+) and 194 HIV(-) controls have been enroted. The mean birthweight in the HIV(+) group was 2880 g as compared to 3035 g in the controls (p =.005), the mean gestational age at delivery 37.5 weeks and 38.5 weeks, respectively (p=.02). Maternal HIV-1 infection was associated with prematurity (22.5 I v 12 X, p =.02, OR = 2.1) and with a history of infant mortality (25.5 % v 14 %, p -.03, OR - 3.9). 4 X of HIV(+) mothers delivered a stillborn baby as compared to.7 X in the HIV(-) group (p =.1, OR - 5.9). 9.6 % of HIV(+) mothers also had a reactive syphilis serology as compared to 6.2 i in the control group (p *.2, OR - 1.8). N. gonorrhoeae was isolated from the cervix in 11 and 8 %, respectively (p =.4). No association was found between other factors including marital status, blood transfusion, contraceptive use, the number of partners, educational level and occupation. A maternal CD4/CD8 ratio of <.50 was found in 6/15 mothers who delivered a preterm baby as compared to 10/90 women who delivered a term baby (p =.01, OR = 5.3., 1.3-21.4). Conclusions: Maternal HIV infection was associated with prematurity. Seropositive mothers with a CD4/CD8 ratio below.50 were at a higher risk to deliver a preterm infant. M.C.94 MORTALITY AMONG 2419 HIV-POSITIVE U.S. MILITARY SERVICE APPLICANTS: AN UPDATE bannenberg. Andrew*; McNeil J.**; and Brundage J.**. *Johns Hopkins School of Public Health, Baltimore, MD, USA. **Walter Reed Army Institute of Research, Washington, DC, USA. Objective: Long term follow-up is ongoing to assess cause-specific mortality among civilian applicants excluded from military service due to HIV infection compared to applicants excluded due to other unrelated medical conditions. Methods: The 1985-1988 files of the National Death Index (NDI) were searched for 2419 HIV-positive applicants (HPA) and 7254 HIV-negative applicants (HNA) (matched 3:1 on age, race, sex, and date and location of screening) identified by the US Dept of Defense (DOD) between 10/85 and 12/88. HPAs were 93% male, 40% white and 53% black. The age range was 17-55 years (median 24 years). Mean follow-up time was 20 months (range 0 -39 months). Identifiers and HIV status of living applicants were known only by DOD. Results: The NDI identified death records for 68 of 2419 HPAs (28/1000) and for 23 of 7254 HNAs (3/1000)(p<.0001). The death rate for HPAs (but not for HNAs) was higher (p<.0001) than the expected all-cause death rate based on age-, race- and sex-specific 1986 U.S. death rates. Based on death certificates from state vital statistics offices, HPAs died from AIDS (n=52), pneumonia (4), lymphoma (1), suicide (5), homicide (2) and drug abuse (4), while HNAs died from unintentional trauma (9), suicide (4), homicide (5), drug abuse (1), heart disease (2), lupus (1) and hepatitis (1). For the five HPAs who died of suicide, the time from screening to death was 1, 2, 14, 20 and 25 months, respectively. Annual mortality follow-up of this cohort is continuing. Conclusion: While these preliminary data are insufficient to calculate death rates for non-AIDS causes such as suicide, they suggest an increased all-cause short term risk of death in HIV-infected persons oprimarily from AIDS. 3C tn C; ~ I~

Page  47 M.C.95 AIDS WITHIN FIVE YEARS OF SEROCONVERSION: Phair ohn, Detels R, Jacobson L, Rinaldo C, Saah A, Munoz A, for the Multicenter AIDS Cohort Study. Chicago, IL, Baltimore, MD, Pittsburg, PA, Los Angeles, CA. USA. Objective: Describe homosexual men with incident HIV-1 infection who develop AIDS within five years and compare behavioral, clinical and immunologic characteristics of those progressing to AIDS (grp A) with those seroconverters remaining AIDS-free (grp B). Methods: Between entry into the study (4/84 - 3/85) and 7/89, 349 seronegative men developed HIV-1 antibody; as of 3/90, 32 had developed AIDS with known dates of seroconversion (SC) and AIDS diagnosis. Grp B were matched to individuals in grp A based upon time from entry to SC and AIDS-free time post-SC. Participants were seen semiannually. Results: The Kaplan-Meier estimates of the proportion of men with incident infection remaining AIDS-free were: 6 mo 100%, 12 mo 99%, 24 mo 97%, and 48 mo 86%. The AIDS-defining events included: PCP 50%, other 01 22%, KS 19%, and other 9% The two groups were demographically similar. Sexually transmitted disease was reported more frequently pre-SC in grp A than grp B (p = 0.05) and grp A engaged in receptive anal intercourse with more partners before (p = 0.003) and after SC (p < 0.005) than grp B. Grp A reported more symptoms at the visit within 3 mo. of SC than did grp B (42% vs 15%); at 6, 12 and 24 mo, symptoms were equally prevalent in both groups. Significantly greater decreases in number and percent of CD4+ cells at 6 mo (p = 0.023), 12 mo (p = 0.001) and 24 mo (p = 0.0009) post SC were noted in Grp A as compared to grp B. A significant rise in proportion of CDm+ cells post-SC was also noted in grp A; 6 mo (p = 0.048), 12 mo (p = 0.013) and 24 mo (p = 0.006). Conclusion: These findings suggest that sexually-transmitted co-factors, pre or post-SC, play a role in the progression of HIV-1 infection. Alternatively, greater sexual activity could increase the hazard of exposure to a more virulent HIV-1 strain. NOTES M.C.96 PRELIMINARY RESULTS OF THE NATURAL HISTORY OF HIV INFECTION IN UGANDAN INFANTS. GuayLaura*, Ball,P.*, Ndugwa,C.**, Kenya-Mugisha**, Hom,D.*, Kataaha,P.**, GoldfarbJ.*, Vjecha,M.*, Mmiro,F.**, Olness, K*. *Case Western Reserve University, Cleveland, Oh. USA, and **Makerere University, Kampala, Uganda Objectives: To determine the natural history of infants born to HIV infected women. Methods: Selected pregnant women presenting in Kampala, Uganda were enrolled in a prospective study of HITV infected women and their infants. The IHV serostatus of the maternal study group (70% HIV+, 30% HIV-) and their 18 month old infants was ascertained by ELISA with confirmatory Western blots of positives. Investigators were blinded to serostatus. Growth parameters, physical signs, symptoms and health of infants were monitored. Results: Four hundred and fifty infants, age 1 to 24 months, are being followed. Excluding perinatal deaths, 70 other infants have died: 64/364 (17%) born to HIV+ mothers and 6/156 (3.8%) born to HIV- mothers. Twenty of 58 (34%) HIV+ mothers who delivered at least 18 months ago had babies that have died. Two of 28 (7%) HIV- mothers had infants who died by 18 months. The average age of infants at the time of death was 7.5 months (10 days to 16 months). At the time of death, 3 infants met clinical criteria for AIDS and 5 had AIDS related symptoms. Thirty-eight infants born to HIV+ mothers are alive at 18 months; 7 (18%) are IlIV+ and symptomatic. Three of these infants have clinical AIDS and 4 have AIDS related symptoms. Growth failure and malnutrition were the most common findings in both living IIIV infected children and dead infants of HIV infected mothers. The majority of these infants presented with generalized lymphadenopathy and repeated common infections (e.g. otitis media and skin infections). Conclusion: HIV infection in Ugandan mothers is associated with increased mortality in their infants in the first 18 months of life. Severe failure to thrive is the most frequent sign of HIV infection in Ugandan infants. NOTES O C\, sl CERVICAL DISEASE AMONG HIV-INFECTED WOMEN BY IMMUNE STATUS M.C.97 LaGuardia, Katherine*; McGuinness, K*; Hunter, D** *The New York Hospital-Cornell Medical Center, New York, NY, USA **Harvard University, School of Public Health, Cambridge, MA, USA Objectives: To examine the relationship between T-helper cell (CD4) status and the prevalence of cervical intraepithelial neoplasia (CIN) and human papilloma virus (HPV) in HIV-infected women. Methods: In order to more closely examine the relationship between cervical disease and HIV infection, a sample of HIV positive women (n=44) attending a gynecology clinic at The New York Hospital in New York City were studied. We examined cervical cytology, presence of HPV by histology, and their association with CD4 counts. Women with CD4 counts less than 200/mm3 were defined as immuno-depleted. Only 2 of the 44 women studied fulfilled clinical criteria for AIDS. Results: Eleven of 18 (61%) women with CD4 counts less than 200/mm3 had evidence of CIN on cytologic smears compared to 10 of 26 (38.4%) with counts greater than or equal to 200 (p=NS). While CIN I and II were equally represented among the 2 groups, significantly more women with severe helper cell depletion had pap smears showing CIN III than those with counts greater than or equal to 200 (5 vs. 0; p-0.05). There was also a high correlation of HPV infection with CIN in this sample of women. Of the 12 cases of histologically confirmed HPV infection, 11 (91.6%) were associated with CIN. There was no correlation, however, with presence of HPV and immune status. Conclusion: In a sample of largely asymptomatic HIV-infected women there appears to be a relationship between the grade of cervical dysplasia and immune status. The association of high grade dysplasia with severe CD4 depletion merits further study. M.C.98 CERVICAL SQUAMOUS INTRAEPITHELIAL LESIONS (CSIL) AND HIV-1 INFECTION IN MALAWIAN WOMEN. Chiphangwi. John**; Dallabetta, G*; Miotti, P*; Liomba, G**; Wangel, A-M*; Saah, A*; *Johns Hopkins University, Baltimore,U.S.A., **Ministry of Health, Malawi. OBJECTIVE: To assess the association between CSIL and HIV-1 infection in Malawian women with heterosexually acquired HIV-1 infection. METHODS: 159 post-partum women (95 seronegative (SN) and 54 seropositive (SP)) followed in a longitudinal infant transmission study of HIV-1 were given a questionnaire and had a physical examination including an STD screen and a PAP smear. Blood samples were obtained for measurement of T cell subsets using monoclonal antibodies and flow cytometry. RESULTS: 12 of 64 (19%) SP women had CSIL, compared with 7 of 95 (7%) SN women (odds ratio - 2.90, p<0.05). This association persisted even when controlling for presence of other STD, socioeconomic status, multiple partners in the last 3 years and age in a logistic regression model. However, in SP women there was no correlation between risk of CSIL and level of immunosuppression as measured by CD4 count. NUMBER WITH SIL ODDS RATIO p-VALUE HIV-1 NEGATIVE (n-95) 7 ( 7%) 1.00 (reference) HIV-1 POSITIVE, >-400 CD4 CELLS (n-46) 9 (20%) 3.06 0.03 HIV-1 POSITIVE, < 400 CD4 CELLS (n-18) 3 (17%) 2.51 0.21 CONCLUSIONS: These preliminary findings indicate that SP women have a higher rate of CSIL than SN women and SIL rate in SP women is not related to the level of immunosuppression. While this suggests that CSIL predisposes to HIV infection, longitudinal and biological research is needed to resolve this issue. Studies are underway to clarify the relationship between Human Papilloma Virus infection, CSIL and HIV-1 infection in these women. C00 -4

Page  48 00 M.C.99 BISEXUAL MEN WITH AIDS IN THE UNITED STATES: EPIDEMIOLOGY AND THEIR ROLE IN THE AIDS EPIDEMIC IN WOMEN Chu, Susan Y; Petennan T; Doll L; Buehler J.; Centers for Disease Control (CDC), Atlanta, Georgia, USA. Objectives: To compare bisexual and homosexual men with AIDS and to assess the relative importance of bisexual men in the sexual transmission of HIV to women. Methods: AIDS cases reported to the CDC through June 1990 were examined. Results: Among 65,389 men with AIDS who reported having had sex with men, 16,793 (26%) were bisexual. This proportion increased with age after age 30-39 years (30-39 [23%]; 40-49 [26%j; 50-59 [35%]; 60-69 [43%]). Bisexual and homosexual men differed by race/ethnicity (bisexual: 57% white, 29% black, 13% Hispanic; homosexual: 75% white, 14% black, 10% Hispanic). Bisexual men were twice as likely to report intravenous (IV) drug use (20%) compared with homosexual men (9%), regardless of race. Among 3,555 women who acquired AIDS through heterosexual contact, 11% (n=401) reported sex with a bisexual man and no other risk factor. In 1989, the rate for AIDS due to sex with a bisexual man was 3 and 5 times higher among Hispanic and black women, respectively, than among white women; and overall was 7 times lower than the rate for AIDS due to sex with an IV drug user. In states with low HIV seroprevalence in IV drug users, the number of cases in women associated with sex with a bisexual man was comparable to the number associated with sex with an IV drug user. Conclusions: Although often considered together, differences between bisexual and homosexual men should be acknowledged in efforts to prevent both male-to-male and male-to-female HIV transmission. NOTES M.C.101 UNDERSTANDING THE HIGH RATES OF HIV RISK-TAKING AMONG YOUNG GAY AND BISEXUAL MEN: THE YOUNG MEN'S SURVEY Hays. Robert B.'; Kegeles, S*; Coates, T*; *Center for AIDS Prevention Studies, University of California, San Francisco, California USA Objective: To examine factors associated with unsafe sex among young gay and bisexual men and generate ideas for effective prevention approaches with this population. Methods: The Young Men's Survey is a longitudinal survey of 400 gay/bisexual men aged 18 to 27, recruited by peers (through social networks, bars, university and community settings) in three medium-sized communities (Eugene, Santa Cruz, Santa Barbara). Results: 36% of the men engaged in unprotected anal intercourse with men in the past two months. 10% had sex with women; 54% of those had unprotected intercourse. Multivariate analysis of variance and chi-square analyses compared men who engaged in unprotected anal intercourse with men who did not. Unexpectedly, high risk-taking men were found to be well-integrated into the gay community: they reported more gay friends, more comfort being gay, less loneliness and were more likely to have a lover (all ps <.05). However, they also reported less social support for safe sex, more interpersonal and communication barriers to safe sex, greater enjoyment of unsafe sex, and more often engaging in sex while high on drugs/alcohol and in public cruising areas (all ps < 05). Conclusion: Social norms and interpersonal pressures within the young gay men's subculture promote engaging in unsafe sex. Community-level, peer-led interventions that target interpersonal issues may be most effective in reducing young men's risk-taking. M.C.100 HIV-1 INFECTION RATES IN THREE HEPATITIS B (HB) VACCINE COHORTS - AMSTERDAM (AM), NEW YORK (NY) AND SAN FRANCISCO (SF),1978-90 Koblin. Bery*,Van Griensven GJP**,Hessol NA+,Byers RH+ +, Stevens CE*,Coutinho RA**. *The NY Blood Center, NY, USA; **Dept. of Public Health, Amsterdam, The Netherlands;+Dept. of Public Health, San Francisco, CA, USA;++CDC, Atlanta, GA, USA. Objective: To compare baseline characteristics and incidence of HIV infection from 1978-90 among homosexual men in placebo-controlled HB vaccine trials carried out in AM, NY and SF from 1978-82. Methods: Study subjects were all men who participated in HB vaccine trials (N) and subsequently enrolled in AIDS follow-up studies in each city (n): AM(n/N)= 714/714; NY(n/N)=326/1090; SF(n/N)=323/359. Probabilities of HIV infection from 1978-90 were estimated using Kaplan-Meier product-limit method. Results: By Dec. 1982, the probability of being HIV infected in AM was 2.7% (95% CI:1.4,3.9), NY 25.2% (20.3,30.1), and SF 36.9% (31.4,42.4) By Dec. 1989, HIV infection was still significantly different in the 3 cohorts and increased to 17.8% in AM, 36.5% in NY and 48.7% in SF. Among the oldest and youngest birth cohorts, no significant differences between NY and SF were observed. At baseline, the median no. of partners/4-6 mon. and % with history of anal sex was lower for AM (p<.0001) (no.partners: AM=8, NY=10, SF=11; anal sex: AM=74%, NY=98%, SF=99%). HBV attack rates in the placebo recipients were also lower in AM than in NY or SF (p<.0I). Conclusion: Differences in HIV infection rates between the US cties and AM may be explained, in part, by differences in high-risk behaviors at the time the epidemic began. NOTES M.C.02 SEXUALRELATIONSHIPS WITH WOMEN IN HOMOSEXUALLY IDENTIFIEDL MEN. THE PROBABILITY OF M. HV TRANSMISSION TO OTHER POPULATIONS. lzazla JA1 '. Gorlmrnakcr S, Basafiez R1, Scputleda J1. I1.-MEXIO MINISTRY OF HEALTH 2.-AR VARD SCHOOLOF PUBLIC HEALTH. Ojectiv;To discuss the main characteristics ofhomosexually identified men who also have sex with women, in order to describe the risk of HIV transmission to other populations (e.g. women). Alet survey in hoimosexual gathecring places like gay bar and discothdiques was conducted in 6 mexican cities. 715 men were interviewed, aprobabilistic household survey was carried out at the same time, and 255 men from the general population were interviewed (MGP). Men from gay places were classified as exclusively homosexual (EH) if they had never had sex with women, active hisexuals (AB) i thyhad sex with men and women in the previous year, and ifthev ever had sex with women but not in the previous vear, they were Ever'Bisexuals (EB). KABP dela about AIDS and scrology(ELISA with WR) were collected. An Insertive Receptive Bchavioir (IRB)indicator was amnstructed from thecombination of individual questions about the frquency of being anal receptive aid/or anal msertively wre classifiedas ti al insriad recepve (1R), xclusively or mainly rcceptive (R), exclusively or mainly inscrtivc () anil with no ir almost no anadlcoitus NO)., test was used to detect diferences among cathegories, only signiicantlv differences are reported (p. (5), unless spcified differently. Re,;uitL, From the..,g..Lher. i- pli:cs s7lpleB: 47pIQ were classificd as ER, 25c, as EB and 274% as AB. There are dilfercrnccs amitun cities lor i sibution. Thir mnarital stalus ws res pred asort follows 53% oelthe MCGPwas mamced, 10% of A R,4.%, of the EB, and 0.3% of ithe EH. Ihe proportion whoreported college or higher education was: MGP 28%, B A 31 %, B 32% and EH 21%, however lic pro|xrtion ot ihose with eleenttary duatiunort;ss was MGOP 7%, BA 8%, EB 9% and EH 12%. When they were.questioned about haiving changd their scxual behavior: 20% ofMGP, 39% A13,479 Ea and45 e of EH rpried to havc made reatichae i n the r sexual lives in order todLcreiasc risk a11 Viteion. There are no differences of knowledge about mechantislis of transninsson. Tlhere ar difflerencecs between tvMCi iand AtB ER and EH (butt there ate no differences betwen AB, EB, and El) in the presence olf Myths about casual transmission and about the use of condoms for AIDS prevent ion MGP and AR consider less frequently than EB and EH that decreasing the number of partners decreases the risk for HIV/AIIDS (80 lt s 88%). 1Th report l having none, and having nc seaxutl particr in the previots 4 months is re spectivecl: iMG 1 7%and 6r BA 6 and 26%, ER 7 and 41%, ai EH 18 and 32%. The report of using a condom in their last sexual intercourse was MISPG 8%, AB 3,% EB 33% and EH 25%;. The dse of Ctms with their sead ' partnrs ad with prostitutes is rcspectivete: NIGP 14 and, A1 37 ant 2). it 3 an, L- 2[ SanduAr. In theirebehavior with men, thel R by subgroup is in ABR 1 48, R 21, B 9%, NO. o 22; in El.:, 2 1, R % NO atd in EH: 113%, I.R 14%, R 48%,andiNO 24 [ In those who rcpeort eing anal rccptive, there are no differences in rcciving semien in rectumi(76i ). There are no differences among these groups in that ticy nmeet dlcir sexual partners in the street, gay bars, and smunas, however, AB meet their scxual partners also in restaurants and in private parties. There arc no differences in having a previous dtction of HIV infection in those recruited in gay gathering places (32%), and there arc no differences in the IIIV prevalnce in these groups (113%). n There is a subt roup of imen who attend to gaygathering places, who also have sc with womcn and who share the gay lifle tyle and the general cncepls (knowledge) abouL AIDS with gay tien, and tit with imien from the general popu laion. Their scxual practices may irnIpl tdie same risk to acquire HIV since their prevalen t is the same. The fact that AB are only receptive less lfrequently than EB and EH, may nolt implv that they have a lower risk of becoming HIV intccted, and therefore transmitling it t6 others. There is a hiehcr prevalence of condom use among AB an< EB than EH and MPG, however its level of one third is still not quite eftective to prolec l thei and their partners as populations. AB are mor se ually active and acqu re their sexual partners in more places than die other groups, therefore despite ihey arc identified more as homosexuals than bisexuas or heterosexuals, they may constitute the main cpidemniological bride tc twcen populutions. There are more types of btisexual ien however, the importanteffortns tochangc behavior i these "bisexual men" most be cnouraoed becautsc of their direct contact with both populations. CI 0;

Page  49 M.C.103 EFFECT OF HIV POST-TEST COUNSELING ON SEXUALLY TRANSMITTED DISEASE (STD) REINFECTION FOR PATIENTS ATTENDING A STD CLINIC Otten, Mac*; Zaidi, Akbar*; Wroten, Jack**, Witte, John**, Peterman, Tom* *Division of STD/HIV Prevention, Centers for Disease Control (CDC), Atlanta, GA, USA; **Dept of Health and Rehabilitative Services, Tallahasse, Florida, USA Background: STD patients are often tested for HIV infection and counseled about the results. In 1989, over 300,000 STD patients were tested for HIV using public support. Yet, the effect of this intervention is uncertain. Objective: Measure the effect of post-test counseling (PTC) for HIV on acquisition of a new STD Methods: We used computerized records of patients at the main Dade County, Florida STD clinic who were voluntarily tested for HIV from 1-1-88 to 2-28-89 and who had visited the STD clinic more than 6 months before the HIV test. We studied 4 groups based on whether the HIV test was + or - and whether there was or was not any PTC. We compared proportions of patients with +gonorrhea (GC) cultures in the 6 months before and after the HIV test. Percent With+GonorreaCulte Percentage Difference Results: HIV Test Post-test In 6 Mo. Before 2-8 Mo.After After vs Before Result Counseling HIV Test HIV Test (95% Confidence Interval HIV+ Yes 331 6.3 4.5 -29 (-67 to +10) HIV+ No 566 5.7 6.2 +9 (-36 to +54) HIV- Yes 666 2.4 5.0 +106 (-4 to +216) HIV- No 3,958 4.4 5.5 +23 (+2 to +45) Discussion: We used an objective measure to examine the effect of PTC. Although all patients were offered PTC, many did not return for PTC. For HIV+ patients, PTC may have moderately decreased the risk of acquiring GC. For HIV- patients, the risk of GC after PTC was double the risk seen before the test. These findings suggest that we may need to alter our approach for PTC to make it more effective in both HIV- and HIV+ patients. M.C.104 ETHNOGRAPHIC AND QUANTITATIVE RESEARCH TO DESIGN AN INTERVENTION FOR TRUCKERS IN ZIMBABWE Wilson, David*; Lamson. Nanc**; Nyathi, B. ***; Sibanda, A.*; Sibanda, T.* *U. of Zimbabwe, Zimbabwe. **AIDSTECH/FHI, NC, USA. ***Bulawayo City Council, Zimbabwe. Objective: To combine survey and and ethnographic research methods to design an intervention for longdistance truckers in Zimbabwe. Methods: Semi-structured interview of 74 truckers, key informant discussions, focus groups and participant observations by ethnographerwho accompanied truckers onjourneys and lived with themin company hostels. Reults: Of those interviewed, 85% were married, 59% had girlfriends, 64% visited prostitutes when away -of whom 61% had visited one in the last month. 33% had used condoms, mostly with prostitutes. In their last commercial sex encounter, 68% reported that they were drunk. All had heard of AIDS, mostly via radio, but knowledge was uneven: (e.g. only 48% knew condoms afforded protection). Ethnography showed how closely interwoven trucking and prostitution are; prostitutes know truck stops often provide companionship and sex in return for free transportation, and "brothels" are often adjacent to trucker hostels. The nature of transport work underpinned risk of sexual transmission of HIV and other STDs. Truckers were away 5-14 days. Due to long absences, an urban family home was considered an unwarranted expense by half the married truckers who, accordingly, lived alone; their wives remain in rural areas. Truckers engaged prostitutes because of loneliness, monotony, drabness of hostels and primarily male "anti-community" environments. Conclusions: HIV risks are deeply rooted in the nature of trucking work. Intensive condom-oriented programs must be implemented. Reliance on peer educators is essential for programs to be empathetic, credible and successful. NOTES NOTES ~ a

Page  50 0 M.D 05 SUICIDALITY AND PSYCHOLOGICAL OUTLOOK IN LONG TERM SURVIVORS OF AIDS. Remien. Robert H,*; Rabkin, J*; Katoff, L**; Williams, J*; *HIV Center for Clinical & Behavioral Studies-NY State Psychiatric Institute & Columbia University, NY, NY, USA; **Gay Men's Health Crisis, NY, NY, USA Objective: To assess past and current thoughts about death and suicide, history of ýuicide attempts, current mood disorders and feelings of hopelessness in a population bf gay men, unique in having lived with AIDS for at least three years (N=53). Method: Subjects were clients at Gay Men's Health Crisis (GMHC), the largest and oldest AIDS delivery service organization. We excluded those subjects whose AIDS diagnosis was onAy Kaposi's Sarcoma. The measures used included the depressive Oisorders modules of the Structured Clinical Interview for DSM-III-R (SCID), iagnostic Interview Schedule (DIS) queries about thoughts of death and suicide, the iuicide item from the Beck Depression Inventory, and Beck's Hopelessness Scale. esults: Only three men (6%) had a current depressive disorder. Although 23% of the ubjects reported a past suicide attempt (most often in adolescence in relation to coming out"), only three subjects reported having made another attempt after an AIDS Jiagnosis; one was clinically depressed at the time. No suicide attempts were made by nen without a history of such attempts, and none reported current suicidal ideation ilthough several considered it a future option should they become severely impaired. lean score on the Hopelessness Scale was 4.5, which Beck classifies as "mild".,onclusion: Overall, these long term survivors with AIDS are resilient and positive in:erms of their mood and outlook. We found low rates of depressive mood disorders, and early all maintain the conviction that good times lie ahead and that their lives are forthwhile. In the absence of a past suicide attempt, a diagnosis of AIDS alone does iot appear to lead to an increased risk for suicide in this cohort. NOTES M.D.106 HIGH SOCIO-ECONOMIC STATUS (HSES) IS A RISK FACTOR FOR HIV-1 INFECTION BUT NOT FOR SEXUALLY TRANSMITTED DISEASES (STD) IN MALAWIAN WOMEN. jalabetta. Gina*; Odaka, N*; Hoover, D*; Chiphangwi, J**; Liomba, G**; Miotti, P*; Saah, A*; *Johns Hopkins University, Baltimore, USA, **Ministry of Health, Malawi. OBJECTIVE: To examine the relationship between STD and HIV-1 infection in urban pregnant women in Malawi. METHODS: 5376 pregnant women presenting for prenatal care to a large urban hospital in Malawi were interviewed about demographics, medical and sexual history, had a physical/pelvic exam and were tested for syphilis and HIV-1 antibodies. RESULTS: 1220 (22.7%) women were HIV-1 seropositive. Multiple sexual partners and husbands' reported partners were positively associated with both HIV-1 infection and current STD. The strongest predictors of HIV-1 infection were HSES (husband's education > 8 yrs) and any current STD. 56% of HIV-1 positive and 33% of HIV-1 negative women were of high HSES, and 64% of HIV-1 positive and 38% of HIV-1 negative women had a current STD. Logistic regression analyses of risk factors for HIV-1 infection and current STD indicated that HSES was a strong risk factor for HIV-1 infection (p<.001), but a weak protective factor for STD (p<.05). HIV-1 INFECTION CURRENT STD ODDS RATIO (95%CI) ODDS RATIO (95% CI) HSES 2.16 (1.89 - 2.47) HSES 0.89 (0.80 - 0.99) Current STD 2.71 (2.37 - 3.09) HIV-1 Infection 2.71 (2.37 - 3.09) The relationships between HSES and HIV-1 infection and HSES and STD persisted after controlling for multiple sexual partners and husbands' reported partners. CONCLUSIONS: HSES is a strong risk factor for HIV-1 infection but protective for STD. Assuming HSES women have greater access to STD therapy, this suggests that reduction in high risk sexual activity involving multiple partners, and not STD control alone, is essential for HIV-1 infection control. NOTES M.D.107 SOCIODEMOGRAPHIC PREDICTORS OF AIDS RELATED KNOWLEDGE AND PRACTICE IN UGANDA McCombe.,,S*; Rwakaglri, F.**; Bukombi, S.***; Cohn, P.**** *AIDSCOM, University of Pennsylvania, Philadelphia, PA USA, **Federation of Uganda Employers, Kampala, Uganda, ***Experiment in International Living, Kampala, Uganda, ****U.S. Agency for International Development, Kampala, Uganda. Objective: To collect baseline data to plan and evaluate an AIDS in the Workplace Training program that uses a peer educator model. Methods: Anonymous Interviews on knowledge, attitudes, and practices related to AIDS were conducted with 623 randomly selected individuals (457 men, 166 women) in five organizations. A multivariate regression model to predict risk behavior and condom use was constructed. Results: Twenty-seven percent reported more than one partner in the last two months, and overall 7 percent reported some condom use over the same time period. Knowledge of AIDS is high and unrelated to number of partners or condom use. The number of partners with whom condoms are not consistently used was related to male sex, older age, the number of partners reported by co-workers, and the belief that one's friends have more than one partner. Condom use was associated with young age, high educational level, and number of partners. After controlling for these variables, condom availability in the worksite, reported use by co-workers, and believing that co-workers used condoms were significant predictors of individual use. Discussion and Conclusions: The results provide support for the influence of social norms on individual risk behavior. The association can be demonstrated for risk taking and risk reduction, and in terms of perceptions of others' behavior as well as the actual behavior reported by co-workers. M.D.108 A study of the psycho-social determinants of risk-taking and preventive behavior related to HIV and AIDS has been undertaken in Metro-Manila. The sample included: 1. the general population consisting of 711 randomly selected households with 1,600 respondents 15-59 years old, using a panel from the National Statistics Office; 2. 200 hospitality girls randomly drawn from registered hospitality girls in Metro-manila; 3. 200 homosexuals, purposively generated from gay bars and gay association and 4. 200 seamen purposively selected from the Seafarer's Registry from the Philippine Overseas Employment Agency. The interview technique was the method used using the modified WHO questionnaire as the instrument. The study aims to: 1. describe the patterns of sexual behavior in terms of the nature of relationships, number of partners and exposure to HIV infection; 2. describe the patterns of preventive behavior in terms of knowledge, attitudes and use of contraception with particular reference to condoms; 3. identify psycho-social determinants highly associated with preventive and risk-taking behavior such as socio-demographic factors, communication patterns, locus of control, social support/significant others, etc, 4. specify practical implications of the research findings for polic formulation in the caaign against HIV infections and AIDS; for planning, implementing and monitorin intervention programs for the general population and special risk groups identify content, thrusa and direction of educational messages so they will be costeffective. Preliminary results yielded interesting findings on the sexual behavior of the general population in an urban setting, prevalence of the use of condom and misperception about its use, myths entertained about the transmission and causes of HIV/ AIDS, important sources of information about AIDS, extent of knowledge about AIDS, injection and drinking practices and locus of control. Same determinants were studied among the high risk population. (see attachments for more details) C) I~ I~

Page  51 M.D.109 PSYCHOSOCIAL CORRELATES OF HIV RISK TAKING BEHAVIOR.. Lserman. Jane; Perkins DO; Boucval J; Evans DL Department of Psychiatry, University of North Carolina, Chapel Hill, NC, USA Objective: The goals of this study were twofold. We examined the relationship between psychosocial factors and 1) risk taking sexual activity within HIV negative men; and 2) sexual activity that exposes others to risk within HIV positive men. Methods: As part of the Coping in Health and Illness project (CHIP), we studied gay men, 41 HIV negative (education T= 16 years, age x=32), and 51 HIV positive (education = 12 ears, age = 31). For HIV negative men, risk taking behavior was measured with a 3 point scale taking into account condom use, type of sexual activity, and number of sexual partners. For the HIV positive men, exposing others to risk was measured with a 3 point scale taking into account condom use, type of sexual activity, and serostatus of the partner. Results: Among the HIV negative, 61%, 19.5% and 19.5% practiced low, moderate and high risk behavior, respectively. Among the HIV positive, 17.6%, 62.7%, and 19.6% exposed others to low, moderate and high risk. Using multiple regression, we found that risk taking behavior in HIV negatives, was: 1) positively correlated with optimism (p=.0004), and anger (p=.007) and 2) negatively correlated with emotional control (p=.007), gay self-acceptance (p=.0004), self-esteem (p=.01), and loneliness (p=.004). Total model R2=.54. In the HIV positive group, exposing others to risk was negatively correlated with education (p=.05) and positively correlated with mastery (p=.02) and conflict in personal relations (p=.04). Total model R2=.25. Conclusions: Risk taking sexual activities may be modified by improving gay self-esteem, more adaptive use of anger, and altering overly optimistic attitudes. Increasing AIDS awareness, and improving social interaction skills to reduce social conflict and overly assertive behavior, may decrease activities that expose others to risk. Understanding these psychosocial correlates of risk behavior may improve the success of risk reduction interventions and decrease the spread of HIV. NOTES M.D.110 WHO PARTICIPATES IN SURVEYS OF HIV-EXPOSING BEHAVIORS? Robert J. Biagar (NCI, Bethesda, MD, USA) and Mads Melbye (State Serum Institute, Copenhagen, DK) Typically, about 50% of potential participants in surveys of HIV-exposing behaviors will respond to mailed, self-administered questionnaires. The behaviors of those not responding introduce uncertainty about how well the results can be generalized to the whole population. In a random, stratified survey of Danes 18-59 years old, we sought methods to boost responses and toenable us to follow-up non-responders. METHOD: With each questionnaire we sent a small card with the subject's name, instructing subjects to return each in a separate envelope. This allowed us to recontact non-responders while preserving the anonymity of the responses. Answers from 3600 recipients of card-questionnaires were compared to those from 1080 recipients of the same questionnaire sent without cards, the traditional approach. RESULTS: After two recontacts 15 days apart, we boosted the response rate from 51.6% to 72.8%. The method did not bias either the immediate (15 days) response rate or responses even to sensitive questions such as about number of sexual partners, prostitute contact and homosexual acts, but those requiring prompting were less likely t, participate in a hypothetical serosurvey (84.1%) than those who responded immediately (75.7%) (p<0.0001). Persons who answered immediately (45.8 %) had similar lifestyles to those in the additional 27.0% obtained by prompting. 100 persons who failed to respond even after two prompts were telephoned to learn why. 75% of the non-responders gave no indication of concern about HIV-exposure compared to 83.2% of responders. CONCLUSIONS: We could boost response rates by this method but the AIDS-risk profile of the additional persons obtained was similar those that of immediate responders. Within the range of achievable responses, higher response rates would probably reveal similar behavior patterns. NOTES - ct M.D.111 THE MERGING OF PUBLIC AND ACADEMIC INSTITUTIONS; A MODEL OF CARE FOR THE HIV-INFECTED CHILD AND MOTHER Gath, Elizabeth C.*i Weber. Kathleen**; Martin, Annie*; Anderson, Deborah**; Yogev, Ram** sCook County Hospital (CCH), **Children's Memorial Hospital (CMH) Chicago. Illinois, USA OBJECTIVE: The establishment of a pediatric AIDS Clinical Trial Subunit (ACTU) of CMH at CCH, the only public hospital in Chicago primarily serving indigent poor, minority, and substance using populations. This Subunit provides investigational therapeutics to HIV infected pregnant women and their children, currently underrepresented in ACTU trials. METHODS: Clinical Research Teams from both CMH and CCH collaborate in the following manner: (1) Select institutional appropriate research protocols. (2) Identify eligible patient populations, (3) Develop and implement a plan utilizing staff provided by ACTU and Subunit, (4) Evaluation of study patient at CCH while providing comprehensive care. RESULTS: (1) Funds have been awarded by the NIAID for the implementation of the Minority expansion of the ACTG. (2) A collaboration between CMII (academic) and CCH (public) has been developed. (3) HIV-infected pregnant women and children now have access to investigational therapeutics. (4) Research quality and patient compliance is maintained CONCLUSION: A replicable model of clinical research and support services has been developed at a public health facility in cooperation with an academic institution. M.D.112 MONITORING HOSPITAL RESOURCE USE AND COST: A THREE-AND-A-HALF YEAR FOLLOW-UP Dijlkaraaf Marcel G.W.*; Borleffs, J.C.C.*; Poos, M.J.J.C.**; Jager, J.C.**; Schrijvers, A.J.P.*** *Utrecht University Hospital, **National Institute of Public Health and Environmental Protection, Bilthoven, ***Utrecht State University; The Netherlands. Objective: To determine trends in hospital resource use and cost related to treatment of HIV infected patients in a university hospital between January 1987 and June 1990. Methods: Detailed data concerning hospital treatment of 126 HIV infected patients have been recorded prospectively over a three-and-a-half year period and have subsequently been analyzed using nonparametrical as well as multivariate and trend analyses. Costs reflect in part estimated costs, partly charges. Results: The number of admissions in the hospital rose from 5 during the first halfyear of 1987 to 38 during the first halfyear of 1990. The number of days per admission dropped from 24.7 to 14.9. Total costs per halfyear quadrupled ($56,800 --> $250,300) during the years monitored. Analyzing data from 30 deceased patients a curvilinear pattern was found between daily cost and disease progression, with peaks at the onset and late phase of the CDC-IV stage, primarily caused by relatively high numbers of admissions and inpatient and outpatient diagnostic examinations compared to the intermediate period. Several differences with respect to resource use were found among patients in different CDC-IV subcategories. Discussion: Analysis of data supports the 1990 governmental initiative to compensate through financial means the public hospitals that are specialized in treating HIV infected patients and therefore attract most patients from their regions. Cost projections should incorporate cost patterns, that are medically based. b C3 \13 t\r. t~ KJ I-^

Page  52 t1 M.D.113 PATIENTS WITH AIDS AND OTHER HIV INFECTIONS: EXAMINING AND ESTIMATING THE TOTAL BURDEN OF HOSPITAL CARE Andrulis, Dennis; Weslowski, V.; Hintz, E. The National Public Health and Hospital Institute, Washington, D.C., USA Objective: To assess the effect of the total HIV-infected patient caseload (i.e., AIDS and other HIV-related conditions) on hospitals; to document similarities and differences in utilization, demographics, financing and other factors between AIDS patients and individuals with other diagnoses related to HIV infection for hospitals located across the United States. Methods: During 1989, and as part of an ongoing annually conducted national study, 1,158 U.S. community and teaching hospitals from five participating hospital associations received a survey. The instrument, which included questions related to utilization, financing and individuals being treated, requested extensive information on patients that hospitals had classified as suffering from CDC-defined AIDS and those with other HIV-related conditions in 1988. Using primarily t-tests, comparisons between these two patient groups were conducted and hospitals were classified by ownership and location. Results: Seven hundred responded to the survey. Two hundred and eleven provided 1988 information on both patients with AIDS and -IIV-related conditions identifying 13,693 and 9,064 respectively. Differences by HIV status were identified for most variables. For example, public hospitals' average number of admissions per patient per year (AIDS-l.7/hosp.; Other HIV-1.3), and average number of inpatient days per patient per year (AIDS-30.8; Other HIV-21.4) were substantially greater for AIDS patients. However, average length stay was similar for the two groups (AIDS 18.2; Other HIV-17.0). Other HIV demonstrated a larger proportion of IV drug users, a greater percent of self pay/other and greater losses per day. Discussion and Conclusions: Results suggest that other HIV patients are less likely to be admitted for inpatient care but once they are admitted their hospital stays more closely approximate patients with HIV. By risk group and/or condition the HIV patient may also be less likely to have insurance, at least in public hospitals, thereby creating the potential for greater hospital losses already tending to be incurred by many institutions. Policy implications include the need to assess total resource allocation and the potential necessity to provide additional financial support for institutions treating large numbers of HIV patients as well as those diagnosed with AIDS. NOTES M.D.114 PREDICTORS OF AIDS REHOSPITALIZATION IN SAN FRANCISCO (SF) Chang, Sophia W*, Arthur J*, Hirozawa A*, Franks D*, Lemp G*. *Department of Public Health, San Francisco, CA, USA. Objective: To determine the rate and predictors of rehospitalization (rehosp) among persons with AIDS (PWAs) discharged from SF hospitals. Methods: Discharge information from nine San Francisco acute care hospitals was linked with AIDS surveillance data to evaluate hospitalizations (hosp) occuring from 4/1/90 through 6/30/90. The linked data enabled first hosp during the interval to be distinguished from subsequent rehosp; average follow-up period was 49 days. Analysis of predictors of rehosp was performed using logistic regression. Results: A total of 898 hospitalizations were reported for 640 PWAs. 640 hosps were analyzed, 480 (75%) without rehosp and 160 (25%) with rehosp (range 1 to 5 rehosp). The overall rate of rehosp was 3.6 per 100 person-weeks. Average length of stay (LOS) for all initial hosps was 10.5 days, with no significant difference in LOS between the rehosp (9.1 days) and non-rehosp (11.0 days) groups. The analysis controlled for age, gender, race, AIDS risk, AIDS diagnosis (DX), insurance status, time since AIDS diagnosis, LOS, and length of observation/survival time. Predictors of rehosp included a DX of Non-Hodgkins Lymphoma (p<.01), having veteran's benefits (p<.02), and LOS (p<.05), with shorter LOS associated with rehosp. Conclusion: Rehosp, a significant factor in the cost of AIDS care, is associated with shorter lengths of hospital stay, Non-Hodgkin's lymphoma and having veteran's benefits. Rehosp is not associated with race or AIDS risk. NOTES C> M.D.11.5 INFLUENCE OF SEVERITY OF AIDS COMPLICATIONS ON HEALTH CARE MEXPENDITURES. Markson, Leona E*, McKee, L*; Fanning, T**; and Turner, BJ*. *Center for Research in Medical Education and Health Care, Jefferson Medical College, Philadelphia, PA; **New York Department of Social Services, Albany, NY, USA. Objective: To determine whether AIDS severity assessed in the first 3 months (mos) of AIDS onset is predictive of subsequent health care expenditures. Methods: The Severity Index for AIDS Patients (SIAP) is based on a model derived from expert opinion and groups patients into four categories by expected survival according to the severity of the defining diagnosis and complications within 3 mos (Category "A" is least and "D" is most severe). Analysis of variance (ANOVA) was used to examine average Medicaid expenditures by severity category within 3, 6 and 12 mos after AIDS onset for New York State Medicaid AIDS patients diagnosed between 1983-86, with analyses on 1987 forthcoming. Only patients surviving the interval of interdst were examined. The study population included only patients enrolled prior to AIDS onset and the resulting sample sizes were: 1,863 at 3 mos; 1,621 at 6 mos; and 1,212 at 12 mos. Results: Average expenditures for all patients were $14,401, $19,952 and $27,421 at 3, 6 and 12 mos after AIDS onset, respectively. But expenses varied by severity category. Expenses at 3 mos were $5,674 for the least severe group (Category A) as compared to $12,001, $15,082 and $21,242 for B, C, and D, respectively. Similarly at 12 mos, average Medicaid expenses were $18,042 for Category A, $23,497 for B, $29,012 for C and $37,779 for D. Trends at 6 months were the same. Using ANOVA, 31%, 22%, and 13% of the variance (p<,001) of log average Medicaid expenses at the corresponding 3, 6 and 12 month intervals was explained by SLAP. Conclusions: Although SIAP was designed to be predictive of survival of an AIDS patient, we found that it is also highly predictive of different expenditure levels for Medicaid AIDS patients. StAP may be a valuable tool for policymakers in estimating expenditures for AIDS patient care and for clinicians in preparing their patients for the financial demands of the disease. M.D.116 1987 volume-caseload relationship for HIV-related Pneumocystis carinii pneumonia (PCP) in New York City (NYC). Bennett CL (1,2), Adams J (2), Gertler P (2), Gilman S (3), Keesey J (2), Shapiro N (1,2), Brook RH (1,2), and Park RE (2). UCLA Medical School, The RAND Corporation, Santa Monica, Ca and the Veterans Affairs Medical Center, Long Beach,CA. Object: Policymakers are debating whether hospitals with large numbers of patients with AIDS provide better care. Mortality rates decline with caseload for PCP patients in one California study, as well as in many studies of surgical procedures. To extend these results and assess their generalizahility, we analyzed empirical data for PCP patients in New York City, the city with the largest number of cases in the US. Methods: All patients with Ist episode HIV-related PCP (n-3585) who received care at one of 75 NYC hospitals in 1987 were evaluated through review of administrative data. Patients varied in terms of: age (mean 36.7 years); sex (82% male); race (40% white, 33% black, A 27% Hispanic); insurance status (44% Medicaid, 37% private insurance); severity-of illness (using the staging system of B Turner et al); and % receiving care at hospitals with high-caseloads (> 150 pts/yr) (53%). Hospitals varied in terms of bed-size (40% with > 600 beds); teaching affiliation (20% non-teaching); and Ist episode PCP caseloads (16% with high-caseloads). Approximately half of the patients were treated in hospitals that treated fewer than 150 PCP patients/year ("low-caseload hospitals"). Results: Patients at high-caseload hospitals (n=1866) differed from those at low-caseload hospitals (n-1719) in: mortality rate (19% vs 29%), %male (88% vs 77%), % black (24% vs 43%), severity of illness stage, % with prior hospitalizations (22% vs 29%0, and Medicaid (34% vs 49%) (all with p<0.05). A logistic regression model indicated that after controlling for all patient factors, patients at high-caseload hospitals were significantly less likely to die than those at low caseload-hospitals (odds ratio 0.78, 95% ci 0.62-0.98). Conclusions: There are large differences in mortality rates as function of holpital experience with PCP patients in NYC. Clinical data are needed to evaluate these differences. C)3 Con 0 01\ 0 t4 0~ 0~ tcr

Page  53 . '................... Nsk2"N. 7...4t..... S Og (............/.........

Page  54 STU.B.1 EVAWATIO CTHE mWHOCNwICALDNIrmW or LDIATIC AIDmINIr, RWANDA. Mellati Philn m, g L eP.", Dabis F., DequayL, tiMi D.*, Mkmabano B."', hVanW oithem C, amb, aBIra,4 SalaonL RL andVan de Payse P", rSRMUM 0,U dOnversityofBordbu. IlFresanto DtrmentCof P CdistrCarm osptt# de I (CE Rvwandt; "AMDSRwa t LIborntory,.ltional AS ContrProrl aro, i jifrrm. b.K!To vsart thtWortld nrth Oris CmOt clini ce= nitiofpedatuic IS in a larg smple of onscqti lyh d childrea in CUtrAfric. M ria t4-moh piod(Ja-Moyl990,Uchldros e pitdllgntatfepitrtC duL to C ereaegibleforthesa y, hefolloringdtartecoIlltefor eachchildf,sa, maitalu socioCon tatus ofthe mothrl hstoofherpBsm t, tranafIontandprenvlos h t, clnicalminlation f0osingonliki fthe WO linctl denitio ando a dditionnelsig~e nmo fection +rerato. y distressb hrpos sEotrIfkctionparatitis mi nnro-dorlopmental -).san sampl wre srenadfor IV.1 antibodes with a cmmrcial Elis eonfirmdbyl Wemarn Blatn ~ay.Othrdthe mtherr also rundmy testadfor HIV-l ineio,'Thse nsi ifvity,m atydpost predictiv)raP ofesuch sign andsymptoms tudtd vtr estimatedusingthepresm of HIV-1 od t sthtondar Ad. logistic r*rion analyr uwasperformadontherriables inscudiathe WHO clnical case deinton. Reiw;: 908 chtldren(475boysa nd48 irls, ranf: Oto 18yur of&tM) verenincudsdinthe ^str 2D%(179/908)o the children verepositveefor HlV-1 todi.Th HIVprevtalm inthe mothr wes 29%.(62/211). Reaitsfrte childre aged l5moths or oldtr areavailable. Amo these 46 children,the ffVl prealun u wasl5% the sirk andymtvoiththebrtPPVer ortelthmsh(81 ),histormofhepaos2ter eli9d lymphophy5%), loan ei (48%), Tha WH OdiinicuI finiti hadaeslritivrityf 84%, cific of 94% aadPPVofl4e%. Insit rth sivththebsspredichtias omflV- lnfeon1 in childranveregenralisedlyhmpl pyviht a loss, chronic uar ndorlthmsh.Th sumeanlys villbe availeble for childrn lensthan 15 months of g. on ron i This sudyconfnrmthe low smetivtty, low PPV andgtod c yofthe WM alineal demntlon. Thereforemittsurin clnicalpractic ae lin ited.Acomplre wlll how hich sip havethebet prediction ofIV-1 serpositiityin edi at O.A impiscatio afthe dfition wbep TU.B. 2 EFFECT OF COMBINATION THERAPY WITH ZIDOVUDINE (ZDV) AND DIDANOSINE(DDI) ON SURROGATE MARKERS Collier Ann C, Fischi MA#, Kaplan LD, Northfelt D+,Skolnik PR**, Volberding PA+, Coombs RW Davis LG##, Bachus V##, Tartaglione T*, Corey L *University of Washington, Seattle, WA, #University of Miami, Miami, FL, +University of California, San Francisco, CA, **New England Medical Center/Tufts, Boston, MA, ##Burroughs Wellcome Co, Research Triangle, NC, USA. Objective: To evaluate safety, pharmacokinetics (PK), clinical and antMral effects of concurrent therapy with ZDV and ddl in HIV seropositives with CD4 counts <400/mm3 who have had <3 months prior ZDV experience. Methods: On-going Phase I study of 5 different oral dosing regimens of ZDV (given 3 times daily) and ddl (given twice daily) for 6 months. Total daily doses of ZDV/ddl (mg) in each group were: 1)150/90 2)300/334 3)600/334 4)300/500 5)600/500. HIV p24 antigen assays will be batch tested for subjects at 6 months. Results: Fifty-two subjects (51 males; mean age 33, mean CD4 count 218+115) have entered. Group 1 subjects were p24 antigen positive (mean 118 pg/ml). Therapy (mean duration 14 weeks, range 1 to 24) has been well tolerated to date. Too few subjects have had sufficient therapy to determine whether group differences exist. In group 1 subjects (N -5), mean CD4 counts at weeks 0 and 16 were 212~107 and 327+162, (P<0.02). Among 21 subjects who have completed 12 weeks, mean CD4 counts rose from 201+121 to 256~156 (P<0.05). Among 6 who have completed 24 weeks, mean CD4 counts were stable (median increase of 15 cells/mm3). ddl had no effect on PK parameters of ZDV. Two subjects (both group 3) discontinued therapy due to side effects (lethargy and fatigue in 1 each); 4 subjects had dose decreases, due to anemia (1 In group 5), increased liver function tests (1 each groups 1,3) and peripheral neuropathy (1 in group 2). Other side effects have been mild. Plasma virus culture and p24 antigen results on therapy are pending. Discussion and Conclusions: Concurrent ZDV and ddl are well tolerated in the doses studied for up to 24 weeks. CD4 counts increase significantly at the start of therapy in all dose groups, although additional follow-up is required to characterize this further. Long-term virologic, CD4 and clinical effects of various doses will determine the potential optimal combination of ZDV and ddl. NOTES NOTES TU.B. 3 INFLUENCE OF PRE-TREATMENT HIV-1 P24 ANTIGEN LEVEL AND SUBSEQUENT CLEARANCE ON THE RESPONSE TO ZIDOVUDINE THERAPY IN AIDS PATIENTS. Liang MT', Bodsworth NJ12, Swanson C2, Cunningham P3, Imrie A3, Cooper DA2'3. Sydney Sexual Health Centre, Sydney Hospital'; National Centre in HIV Epidemiology & Clinical Research, University of NSW2; Centre for Immunology, St Vincents Hospital, Sydney, Australia3. Objective: To determine the influence of (a) pre-treatment HIV-1 p24ag levels and, (b) subsequent p24ag clearance, on response to zidovudine (ZDV) in AIDS. Methods: A prospective study of 117 homosexual men receiving ZDV therapy for AIDS was undertaken. Kaplan- o.... /.. Meier survival curves from commencement of ZDV to pattern I pattern 2 pattern 3 death were compared on the basis of pattern of p24 n-.i..6 nU response (fig) using the log-rank statistic, sa d1y- s, d,a14 dan Results: Seven patterns of response to ZDV therapy were noted (fig). Subjects without detectable p24ag at the start of treatment (patterns 1-3) survived longer (median 552 o...... days) than p24ag+ve subjects (patterns 4-7; 436 days; pctern 4 pattern 5 pattern 6 pattern 7 P=0.037). Initially antigenaemic patients, who at any time '9, n' 15s "acleared antigen (patterns 4, 5, 7) survived for longer 41 day 4 days 9 dyw 514 daY, (median 481 days) than those who always tested positive for p24ag (pattern 6; 109 days; P<.0001). p24ag clearance and median survival Conclusions: Pretreatment p24 antigenaemia predicts poorer survival of AIDS patients receiving ZDV therapy. Persons who never clear antigen have the worst outcome. TU.B.4 MARKERS PREDICTIVE FOR THE DEVELOPMENT OF OPPORTUNISTIC NON-HODGKIN'S LYMPHOMA (NHL) IN PATIENTS WITH SEVERE HIV INFECTION ON LONG-TERM ANTIRETROVIRAL THERAPY. Pluda, James M; Lietzau. Jill; Tosato, G; Venzon, D; Birx, D; Broder, S; Yarchoan, R. National Cancer Institute (NCI), Bethesda, MD, USA Objective: NHL are a well known complication of HIV infection. We examined a cohort of pts receiving long-term AZT-containing therapy for markers predictive of the development of NHL. Methods: We analyzed 55 consecutive pts with AIDS or severe ARC on AZT-based regimens treated between July 1985 and Dec 1990 at the NCI. They represent pts on the 1) phase I trial of AZT, 2) pilot trial of AZT/acyclovir, and 3) pilot study of alternating weekly AZT/dideoxycytosine (ddC). Results: 8 of 55 pts (14.5%) developed NHL, with an estimated actuarial probability of 31% (95% CI 16-52%) after 36 mos. All 8 pts had <50 CD4 cells/mm3 at diagnosis (median 6 CD4 cells) (p=0.003 for the null hypothesis that CD4 was not a predictive marker). Having <50 CD4 cells was also predictive for the development of NHL independent of the time on anti-HIV therapy (p=0.009). Pts who developed NHL had elevated IL-6 levels as compared to normal, non-HIV infected controls. Quantitative Ig levels and ESR were also elevated compared to non-infected controls, but did not differ from pts in the cohort who did not develop NHL. All pis were EBV seropositive. Discussion and Conclusions: About 1/3 of pts with AIDS or severe ARC who survive for 3 years on anti-HIV therapy are predicted to develop NHL. Pts with <50 CD4 cells/mm3 are at particular risk for developing this complication. B cell hyperactivation, conceivably mediated in part by IL-6, may contribute to its pathogenesis. The high cumulative incidence of NHL in these pts is most likely related to their prolongation of survival in a profoundly immunosuppressed state. C>i -,1 t^Y fa

Page  55 TU.A. 5 HIV-at INCREASES IL-6 PRODUCTION BY AND PROLIFERATION OF AIDS-KS DERIVED CELLS Miles,S; RezaiA; Gaynor,R; Magpantay,L; Kishimoto,T; Martinez-Maza. Otoniel. UCLA AIDS Center. -Objective: AIS-Kaposi sarcoma (AIDS-KS) derived cells produce multiple cytokines with autocrine and paracrine growth activity and proliferate in response to the trans-activating protein of HIV, HIV-t4a. One of these cytokines, interleukin-6, and its receptor is produced in large quantities within AIDS-KS lesions in vivo. In vitro, AIDS-KS derived cells produce and respond to IL-6. To integrate these observations, we examined the effect of HIV-tat on IL-6 production by AIDS-KS derived cells. Methods: Cell lines derived by limiting dilution sub-cloning from the AIDS-KS containing pleural fluid were used. After confirmation of cell lineage, cells were plated at 5 X 105/ml on gelatin coated flasks. After 24 hours of growth in media (IMDM with 10% FBS, 1% Penn/Strep with fungizone (PSA), 30 pg/ml endothelial growth supplement (ECGS) and 100 USP units of heparin), the cells were washed with PBS and subjected to calcium phosphate transfection overnight with either HIV-tal or a mutant HIV-tat vector (0-10pg/ml) in duplicate. After 24 hours, one set of cells were harvested, counted and replated into 96 well plates. The media was changed (IMDM with 1% PSA, ITS+ (10 ml/L), 30 /g/ml ECGS, with/without 100 USP u/ml heparin) after 24 hours. Six hours later 3H-thymidine (lpC/ml) was added. Cells were harvested after 18 hours with trypsin onto glass wool filters and incorporation of labelled thymidine measured. IL-6 levels were determined by an IL-6 ELISA. IL-6 mRNA levels were determined by northern analysis of the duplicate transfections. Results: IL-6 mRNA and secreted IL-6 is increased in AIDS-KS cells after exposure to recombinant HIV-tat or by transfcction of an HIV-tat expression vector. A mutated non-functional HIV-tat expression vector had no effect. The increase in IL-6 production is accompanied by corresponding increases in -H-thymidine incorporation. These increases can be specifically inhibited by antisense but not sense IL-6 oligonucleotide polymers. There was no effect of HIV-tat or IL-6 antisense on human aortic or umbilical vein endothelia and aortic smooth muscle cells. Conclusions: Our data suggests that the effect of HIV-tat on AIDS-KS derived cells is partially mediated by IL-6 and is unique to the transformed AIDS-KS cell phenotype. The proliferative effects of HIV-tat may be mediated through increased expression of IL-6 mRNA. TU.A.6 BIOLOGICAL PROPERTIES OF TAT, THE TRANSACTIVATOR GENE OF HIV-1 Ensoli, Barbara; Buonaguro, L.; Barillari, G.; Gallo, R.C. Laboratory of Tumor Cell Biology, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA. bjective: Analysis of new biological properties of Tat, particularly as growth actor for cells derived from AIDS-Kaposi's sarcoma lesions (AIDS-KS cells) and as.ransactivator of virus and cellular gene expression. Methods: Release, uptake and biological effects of Tat during acute infection, fgL-transfection or by recombinant rotein were analyzed on target systems (mesenchymal cell growth, HIV-1 expression, ellular gene expression),by combining cellular, molecular, and biochemical echniques. Results: After acute infection or transfection, Tat is released before r in absence of cell death, respectively, and correlates with maximal Tat expression. xtracellular Tat induces cell growth (AIDS-KS, activated mesenchymal cells), HIV gene xpression, and can rescue tat-defective proviruses. Low doses of Tat are necessary or cell growth, while transactivation requires higher protein concentrations. lowever, coculture of infected cells with target cells facilitates this Tat activity. ome of Tat effects may be mediated by induction of cellular gene expression, as we ave shown for the TNF B gene. Conclusion: The results suggest that Tat-growth 4ctivity may be relevant in vivo by cell-free protein, while the transactivation may become significant by cell contact or close proximity between infected and target cells. By these activities Tat and/or induced cellular factors, may participate in the pathogenesis of AIDS and associated malignancies. NOTES NOTES TU.A.7 MOLECULAR PREDICTORS OF SURVIVAL IN HIV ASSOCIATED NON-HODGKIN'SLYMPHOMA Kaplan. Lawrence D.; Shiramizu,B.; Kahn,J.O.; Herndier,B.; Meeker,T.; Northfelt,D.; and McGrath,M. University of California, San Francisco; San Francisco, California, USA Objective: To investigate the influence of molecular characteristics on survival in patients with HIV-associated non-Hodgkin's lymphoma. Methods: B-cell lymphomas from 29 HIV-infected individuals were characterized using Southern blot hybridization with JH,EBV and c-myc gene probes. Molecular features were correlated with clinical characteristics and survival. All 29 patients received mulitagent combination chemotherapy. Results: Fifteen were characterized as typical monoclonal non-Hodgkin's lymphoma and 14 as polyclonal. EBV DNA sequences were present in 6 of the monoclona) and one of the polyclonal tumors. C-yc rearrangements were found only in 7 monoclonal tumors. Other than a trend towards higher CD4+ lymphocyte counts in the polyclonal group (238/mm3) compared with the monoclonal group (142/mm3) P=NS, there were no differences in known clinical prognostic features between the polyclonal and monoclonal groups. Median survival time was significantly longer in those whose tumors were polyclonal (13.3 mo) vs monoclonal (3.4 mo) p =.004. No patient with an EBV-positive monoclonal lymphoma survived longer than 3.8 mo (median 2.0 mo) and the longest survival was in the EBV-negative, polyclonal group (median 13.3 mo). Cox analysis indicates that EBV and clonality predict survival independently of each other and that the CD4+ lymphocyte count is predictive of survival only in the groups with either polyclonal or EBV-negative disease. Conclusion: This analysis demonstrates unique characteristics of the polyclonal and EBV-negative groups in which survival is longer and directly related to the CD4+ lymphocyte count. TU.A.8 MALIGNANT LYMPHOMAS IN SIV-INFECTED CYNOMOLGUS MONKEYS - A MODEL FOR EBV - ASSOCIATED LYMPHOMAGENESIS HFeichtinoerf,. SU-Ling Li2, K.Gronewald3, P.Putkonen4, K.Weyrer3, G.Biberfeld4, P.Biberfeld2 1 Dep.of Pathology, Univ.of Innsbruck, Austria (' partly supported by the European Concerted Action on AIDS), 2 Immunopathology Lab., Karolinska Institute, Stockholm, Sweden, 3 Dep.of Internal Medicine, Lab. of Molecular Genetics, Univ.of Innsbruck, 4 Dep. of Immunology, National Bacteriological Laboratory, Stockholm, Sweden Oblective: EBV-associated lymphomagenesis in experimentally SIV-infected cynomolgus monkeys as a model for malignant lymphomas in AIDS Methods: 24 cynomolgus monkeys were infected with simian immunodeficiency virus (SIVsmm3; Drs.P.Fultz and McClure, Yerkes Primate Center; USA). Malignant lymphomas were characterized histopathologically and immunophenotypically. EBV-association was studied by anti complement immunofluorescence (ACI F) with human and monkey sera and by Southern blot analysis. esults: 10/24 of the monkeys developed malignant lymphomas during an observation period between 5 and 15 months after SIV infection. These tumors show remarkable similarities with human HIV-associated lymphomas with respect to clinical behaviour, morphology and immunophenotype. All of them are shown to be of B-cell origin with a variable expression of B-lymphocyte markers like CD10, CD19, CD20, CD22, CD23, CD75 and CD77. In addition adhesion molecules like CD18 and CD54 as well as MHC class II antigens are expressed. Southern blot hybridization revealed the presence of DNA of an EBV-like virus in the tumor tissue of 8/10 lymphomas as well as in cell lines established from 3 different tumors tested so far. Furthermore EBV-associated nuclear antigens (EBNA) were found in the tumors and the cell lines. The clonality of the EBV-associated lymphoproliferations was established by hybridization with a 1.9kb Xhol fragment specific for the fused termini of the EBV-genome and confirmed by the demonstration of clonal immunoglobulin gene rearrangements. Conclusion: The data strongly suggest an involvement of an EBV-like monkey virus in the pathogenesis of malignant lymphomas in SIV-induced immunodeficiency. This experimental model can therefore provide important information on EBV-induced lymphoproliferation and lymphomagenesis in HIV-associated immunodeficiency. bo

Page  56

Page  57 ~SSy~

Page  58 TU.A.9 Natural Antlviral Responses In HIV Infection Jay A. Levy, Carl E. Mackewicz, and Cecilia Cheng-Mayer Department of Medicine and Cancer Research Institute, University of California, School of Medicine, San Francisco, CA, USA During HIV infection, two responses have been described that reflect a natural antiviral activity. First, the infected host's own CD8+ lymphocytes can suppress HIV infection in CD4+ cells. The mechanism does not involve cytotoxicity, but a lymphokine that we have found is not related to any previously identified cellular product. It does not induce 2-5A synthetase activity in cells, and is not neutralized by antibodies to interferon or tumor necrosis factor. The effect of the factor is only observed when cells are exposed to it; after its removal, viral replication quickly resumes. In evaluation of over ten other human cytokines, none has shown the activity of the CD8+ cell factor. T cell clones have been generated that produce relatively high levels of the factor. They have been helpful for characterizing this cytokine for its molecular size, pH, and temperature sensitivities. T cell clones not releasing factor have also been derived and will be useful in the eventual identification of this novel cytokine. Since higher levels of the factor are produced by CD8+ cells from asymptomatic individuals than those progressing to disease, it appears this cytokine is an important antiviral response in infected individuals. The nef protein of HIV is a second mechanism for controlling replication of certain strains of the virus. Isolates obtained from asymptomatic individuals are sensitive to the effect of the nef protein when introduced into T cells using HIV-LTR or SV40 promoters. Their replication is suppressed. Later isolates from the same individual after progressing to AIDS are resistant to the effects of nef. Using Interviral recombinants of an early and late viral strain from the same individual, a viral region responsible for sensitivity to the nef regulatory protein has been identified, It is not associated with the LTR, but encompasses part of the env and rev coded sequences. Thus, an interaction between HIV genes, most likely involving cellular factors, appears to control viral replication and help establish a state of latency. These observations on immune and viral factors controlling HIV infection merit further attention for potential therapeutic approaches. NOTES -T.A.10 HIGH EXPRESSION OF NEF PROTEIN IN HIV-INFECTED BRAIN ASTROCYTES ASSOCIATED WITH RAPIDLY PROGRESSING CNS DISEASE. Banli Annamari*; Ovod, V."; Haltia, M.***; Nyberg, M."***; Aavik, E.**; Krohn, K." *Dept. of Dermatology, and **'Pathology, University of Helsinki, Helsinki, **lnsititute of Biomedical Sciences, University of Tampere, Tampere, *""Aurora Hospital, Helsinki, Finland. Objective: To study the simultaneous expression of HIV proteins, especially net and tat regulatory proteins, and HIV-specific mRNA in postmortem brain sections of HIV-infected individuals. Methods: A double detection method with monoclonal antibodies (MAb) and in situ hybridization was developed for the simultaneous demonstration of HIV proteins and mRNA. A panel of MAbs against HIV nef and tat proteins was produced, and the target cells were identified with Mabs to GFAP (astrocytes), CD68 (activated macrophages) and RCA-1(microglia, endothelial cells). For hybridization, HIV gag/pol-, vifand env specific antisense RNA probes were used. Postmortem formalin-fixed, paraffin-embedded brain samples of six HIV-infected individuals and four controls were studied. Result: HIV mRNA were most frequently seen in microglial cells, and sometimes in endothelial cells. The microglial cells also expressed HIV nef. Multinucleated giant cells were rarely detected. However, in one of the six samples, abundant numbers of astrocytes (GFAP+) were strongly posititive for HIV RNA (gag/pol and env). The astrocytes expressed HIV p24 weakly, did not express HIV tat protein, but were strongly positive for HIV nef. The HIV RNA copy numbers were clearly higher in the astrocytes than in the microglial cells. Histologically, scattered microglial nodules but no giant cells were seen in the cortex, and diffuse astrocytic hyperplasia and hypertrophy was present in the white matter. Clinically, this patient had experienced a rapidly prgessing dementia resulting in mutism, incontinence and akinesia, and death of encephalopathy followed three months after the diagnosis of dementia. Discussion and conclusion: We have shown that astrocytes are in some instances inlcted by HIV in vivo. This condition may clinically represent as a rapidly progressing CNS disease. NOTES TU.A.11 SYNCYTIUM INDUCING CAPACITY AND SEQUENCE ANALYSIS OF HIV-1 ENVELOPE GENES DERIVED FROM CLONAL ISOLATES. Andeweq, Arno C.*; Groenink, Martijn**; Leeflang, Paula*; de Goede, Ruud E.Y.**; Osterhaus, Albert D.M.E.*; Tersmette, Matthljs** and Bosch, Mamix L.*. *Laboratory of Immunobiology, National Institute of Public Health and Environmental Protection (RIVM), Bilthoven, The Netherlands and "Central Laboratory of the Netherlands Red Cross Blood Transfusion Service (CLB) and Laboratory for Experimental and Clinical Immunology, University of Amsterdam, Amsterdam, The Netherlands. Clinical isolates of HIV-1 differ in their biological properties such as replication rate, cytotropism and syncytiumn inducing capacity. Previously it was demonstrated in a longitudinal study that detection of syncytium-inducing (SlI variants in asymptomatic individuals is strongly associated with subsequent rapid decline of CD4+ cell numbers and progression to AIDS, suggesting a direct role for such (SI) variants in CD4+ cell depletion. To study the tusion-/syncytium-formation process we analyzed the envelope genes of a panel of in vitro phenotypically well defined biological HIV-1 clones. To perform functional single gene analysis, the envelope genes of 8 biological HIV-1 clones were cloned in pSC11 to construct recombinant vaccinia viruses. The generated vaccinia/HIV-1-env recombinants were used to determine the syncytium inducing capacity of single envelope genes expressed in different cell types. Our results indicate that different HIV-1 envelope genes display heterogeneous syncytium inducing capacities which in parti correlate with the syncytium inducing capability of the biological HIV-1 clones from which they were derived. To elucidate the genetic basis of the observed phenotypic variation we determined the nucleotide sequence of 8 envelope genes amplified by PCR. The predicted amino acid sequence analysis revealed considerable phenotype associated amino acid variation in a number of regions including the fusion domain. To study the amino acid variation in the fusion domain more extensively, a region around the envelope cleavage site (spanning the fusion peptide) from 20 different biological HIV-1 clones was also amplified by PCR. The predicted amino acid sequence analysis of this region revealed phenotype associated changes in the fusion peptide of 8 out of 20 biological HIV-1 clones; the syncytium inducing capacity correlates with the presence of solely hydrophobic amino acids in the fusion peptide. The contribution of the observed envelope amino acid variation in determining the syncytium inducing capacity will be discussed. TU.A.12 GENETIC MAPPING OF REGIONS IN THE HIV-1 GENOEJT DETERMINING SYNCYTIUM INDUCTION AND CYTOTROPISM M. Groenink, R.A.M. Fouchier, R.C.M. van der Jagt, R.E.Y. de Goede, J.G. Huisman, and M. Tersmette. Central Lab. of the Neth. Red Cross Blood Transf. Service and Lab. for Clin. and Experimental Immunology, University of Amsterdam, Amsterdam, The Netherlands. Previously we and others have demonstrated a relation between the clinical course of HIV-1 infection and biological properties of HIV-1 variants such as replication rate, syncytium-inducing (SI) capacity and cytotropism. For the molecular analysis of this biological variability we generated a panel of phenotypically distinct, yet genetically highly homologous infectious molecular clones. These clones were derived from HIV-1 isolates, mostly recovered by direct clonal isolation, from a single individual in whom over time a transition from non-syncytiuminducing to SI isolates had been identified. The molecular clones exhibited differences in SI capacity and T cell line tropism. T cell line transfection studies revealed that T cell line tropism is not caused by differences in level of HIV-1 expression, but most probably is restricted at the level of virus entry. Experiments with recombinant viruses between four biological divergent HIV-1 clones indicated that the 3'half of the HIV-1 genome is responsible for differences in SI capacity and T cell line tropism. The sequence analysis of this region revealed that the vpr, tat, rev, vpu, and nef genes of these four clones were highly homologous. In the env gene phenotypic specific amino acid changes were observed. These studies indicate that the envelope region determines SI capacity and cytotropism. C> C sj

Page  59 TU.A.13 LYSIS OF HIV-1 INFECTED TARGET CELLS BY HIV-1 SPECIFIC CD8' CYTOTOXIC TLYMPHOCYTES RESULTS IN INFECTION OF THE EFFECTOR CELLS De Maria, Andrea*, Colombini S**; Pantaleo G*, Schnittman SM*, Greenhouse JJ*, Koenig S*, Moretta L"', Fauci, AS* *LIR, NIAID, NIH, Bethesda, MD, USA, **LTCB, NCI, NIH, Bethesda, MD, USA *"I.S.T., Genova, Italy Objective: To study the mechanisms responsible for HIV-1 infection of CD8+ T- lymphocytes in vitro. Methods: An HIV-1 nef-specific CD8+ cytotoxic T lymphocyte (CTL) clone obtained by limiting dilution technique from peripheral blood lymphocytes (PBL) of an HIV-1 seropositive patient and an allospecific CDS* CTL population were used as effector cells in a specific CTL assay. Epstein Barr virus transformed B-lymphoblastoid cell lines (B-LCL) chronically infected with HIV-1,,, were used as target cells. Both the HIV-1 specific CTL clone and the allospecific CTL population were cocultured with either HLA-matched or HLA-mismatched, chronically infected CD20'CD4' B-LCL for 16hrs. The cell populations were then sorted by cytofluorometry into CD8CD20- and CD8-CD20C subsets (. 99% degree of purity) and analyzed for the presence of HIV-1 proviral DNA by PCR amplification. Results: After coculture with HLA-matched HIV-1 infected B-LCL, sorted HIV-1 specific CDS' CTL harbored HIV-1 proviral DNA as determined by PCR. In contrast, no HIV-1 proviral DNA could be detected when the same effector cell populations were cocultured with HLA- mismatched HIV-1 infected B-LCL. Similar results were obtained using the allospecific CD8* CTL population. Therefore, these results indicate that HIV-1 infection of CD8~ CTL was associated with the lytic phase of CD8' effector function. Discussion and Conclusions: We provide evidence that HIV-1 specific CD8+ CTL can be infected with HIV-1 in vitro during the process of killing HIV-1 infected target cells. Therefore, one can speculate that a similar mechanism may be responsible for the infection of HIV-1 specific CD8' CTL in vivo and may explain in part the depletion or functional impairment of HIV-specific CTL that almost invariably parallels the progression of HIV-1 infection. TU.A.14 HIV-1 VIRIONS CARRY VIRAL DNA ASSOCIATED WITH REVERSE TRANSCRIPTASE. POSSIBLE ROLES OF THIS COMPLEX. Lori Frahco*; Veronese, F.**; DeVico, A.**; Lusso, P.*; Reitz, S.M. Jr.*; Gallo, R.C.* * Laboratory of Tumor Cell Biology, NIH/NCI, Bethesda, MD, USA ** Advanced BioSciences Laboratories Inc., Kensington, MD, USA Objective: To study the role of intermediate enzymatic complexes in the replication and latency of HIV-l. Methods: Concentrated HIV-1 was DNAse treated and reverse trancriptase (RT) was purified from viral lysates by HPLC immunopurification. Purified enzyme was analyzed for the presence of viral DNA by in vitro enzymatic assays, activity gel technique, and PCR. Results: Viral DNA was found associated with RT in HIV-1 mature viral particles. This association has been demonstrated to be highly specific. DNA appears to be in an incompletly synthesized form and RT seems to be bound mainly to the central part of the viral genome, close to the polypurine tract which is used as on of the primers for synthesis of plus strand DNA. Conclusions: The presence of specific viral DNA in extracellular particles represents a novel finding in retroviruses. The DNA seems to originate from a reverse transcription step before virus release. This DNA might play a role in the latency of HIV-1 after infection of resting cells. We also propose that reverse transcription may be involved in entirely intracellular cycles of viral replication. HIV-1 replication may thus bear similarities with the replication of RT dependent DNA viruses NOTES NOTES TU.A.15 HIGH LEVEL HIV REPLICATION DEPENDS MORE ON THE TAT GENE THAN THE LTR. Le Guern. Muriel: Levy, J.A.; Cheng Mayer C., Cancer Research Institute, University of California, School of Medicine, San Francisco, California, USA. Objective: To determine whether differences in the replicative capacity of HIV-1 isolates depend on variations in their tat transactivating capacity or on differences in levels of LTR function. Methods: CD4+ Jurkat T cells and the COS-7 cell line were transfected with various LTRs and tat constructs generated from HIV-1 isolates that show different replicative properties. The isolates studied were: HIV-ISF2, a moderately cytopathic T cell tropic strain; HIV-ISFI3, a highly cytopathic and replicative strain recovered from the same patient as HIV-1SF2 when he was at an advanced stage of the disease; and 2 prototype strains of highly cytopathic, fast replicating viruses (HIV-ISF33, HIV-lnlB). LTR transcription and tat mediated transactivation were evaluated by transient CAT expression assays. Tat gene constructs are tatSF2, tatSF13 and tatlIIB under the SV40 or RSV promoters. LTR constructs are from HIV-ISF2 and HIV-1SF33. Results: In COS-7 cells, the level of transactivation was the same for all rat plasmids and was not influenced by the LTR. In contrast, in lymphoid cells (Jurkat), transactivation mediated by the tat plasmid derived from the fast replicative isolates, tatSF13 and tatIInB, were 4 to 8 fold higher than for taSF2, independent of the LTR used. Differences among LTRs were found only in the extent of activation induced by tatsF2 on the LTR derived from fast replicating strains. Three amino acid diferences were found between the tatsF2 and tatSF13. Experiments with site-directed mutants are in progress to determine which a specific amino acid change is responsible for the observed difference in tat activity. Conclusion: The results indicate that the tat protein of a fast replicating virus is a better transactivator of HIV LTR, and suggest that high level virus replication by these strains depends more on the tat gene han on the LTR TU.A.16 ROLE OF TNV, A TRIPARTITE TAT-ENV-REV FUSION PROTEIN, IN HIV-1 REPLICATION -Ottlinger. Heinrich; Dorfman, Tatyana; Sodroski, Joseph; Haseltine, William, Dana-Farber Cancer Institute, Boston, MA 02115, U.S.A. Objective: A tripartite fusion protein, tnv that includes coding sequences of the ~, env and rev genes and displays both tat and rev activity is synthesized by the prototype molecular clone HXBc2. We wished to determine whether the expression of the tnv protein has any influence on the rate of HIV-1 replication. Methods: To selectively prevent the expression of tny without altering any other coding sequences, the splice sites framing the env exon of my were mutated. The AG and GT dinucleotides immediately upstream and downstream of the exon boundaries, which are nearly invariant in higher eukaryotes, were altered by site-directed mutagenesis. Full-length, otherwise isogenic proviruses carrying the mutations were constructed. Results: Alterations of the 5'splice site of the exon resulted in the inability of the provirus to synthesize the mny protein, but did not affect the expression of other viral proteins. Viruses unable to synthesize tv replicated equally well in established T cell lines and primary human cells as did the parental, Myv-expressing provirus. By contrast, mutations of the AG dinucleotide of the 3' splice site markedly affected or abolished virus replication even when no coding sequences other than that of tv were altered by the mutations. This phenotype could be attributed to the occurrence of aberrant splicing, which largely prevented the expression of the essential regulatory proteins tat and rev. Conclusions: The expression of the nv protein had no significant influence on the rate of virus replication in tissue culture. However, alteration of a 3' splice site in the env region, that is used to generate the tnv message, can have substantial effects on virus replication by modulating tat and rev expression. o 'ar i 0\

Page  60 TU.A.17 NEGATIVE REGULATION OF HIV-1 EXPRESSION IN MONOCYTES; INHIBITION OF THE 65+50 KD NF KB HETEROTETRAMER COMPLEX FORMATION Ruscetti, Francis*, Mikovits, J**; Calvert I*, Ghosh, S***, Kung, H-F* and Raziuddin** *Biological Response Modifiers Program, NCI/FCRDC and **PRI/Dyn Corp, NCI/FCRDC Frederick, MD, USA and ***Whitehead Inst., Cambridge, MA, USA Objective: To study the molecular mechanisms involved in restricted HIV expression. Methods: Gel mobility shift and other assays were performed using nuclear extracts from differentially HIV expressing THP-1 monocytoid cells. Results: Although monocytic cells can provide a reservoir for viral production in vivo, the regulation of HIV transcription in these cells can be either latent, restricted or productive. These differences in HIV gene expression have not been molecularly defined. In THP-1 cells with restricted HIV expession, there is an absence of DNA binding complex formation with the HIV-1 promoter-enhancer leading to markedly less production of viral RNA. This absence of binding was localized to the NF KB region of the HIV enhancer with the 65+50 kD NF KB heterotetramer being completely lost. Addition of purified NF KB protein to nuclear extracts from restrictedly infected cells overcomes this lack of binding. In addition, treatment of these nuclear extracts with DOC restored their ability to form the heterotetramer suggesting the presence of a specific inhibitor of the activity of NF KB. Furthermore, nuclear extracts from these cells with restricted expression treated with LPS leads to increased viral production and increased NF KB activity. Conclusion: Both NF KB binding complexes are needed for optimal viral transcription. The binding of 65+50 Kd heterotetramer to the HIV-1 enhancer can be negatively regulated in monocytes providing one mechanism leading to restricted HIV expression. NOTES TU.A.18 SIVMAC-NEF MUTANTS AS CANDIDATE LIVE VIRUS VACCINES Binninger. Dorothea; Bonn, D.; Ennen, J.; Norley, S.G.; Kurth, R. Paul-Ehrlich-lnstitute, Langen, Germany Objective: Site-directed inactivation of single viral HIV/SIV genes will help us answer the questions of what are the mechanisms of pathogenicity and how do lentiviruses escape host immunity. Methods: The aim is the construction of mutants by recombinant DNA technology which replicate acutely after infection and are no longer able to establish a latent stage. Because nef is a candidate gene for the establishment and maintenance of latency, a series of SIVmac-nef mutants have been constructed by partial deletion and insertions in the net gene. Results: Replication behaviour of the net insertion mutants compared to the wildtype revealed unambiguously a positive effect of the inactivated nef gene product on the kinetics of viral replication. Surprisingly, analysis of net deletion mutants suggested the existence of a Negative Regulatory Element (NRE) in the protein coding region of the nef gene. Discussion and conclusion: It is possible to create SIVmac nef mutants which replicate faster than the wildtype. We are analyzing the features of the mutants in cell culture to select single mutants for their use as vaccine viruses in the rhesus monkey. NOTES hr 1~ Eu TU.A.19 SELECTIVE DOWNREGULATION OF RNA AND PROTEIN EXPRESSION OF LYMPHOCYTE SURFACE RECEPTOR CD28 IN T CELL LINES EXPRESSING THE HIV-1 TAT GENE Kashiwamura, S.I.'; Garcia, V."; Agy, M.*; Ledbetter, J.A.'; Katze, M.G.*, and Clark. Edward A.* *Regional Primate Research Center and Department of Microbiology, University of Washington, Seattle, WA USA; "Fred Hutchinson Cancer Center, Seattle WA USA. Objective: To assess the effect of the HIV tat gene product on the expression of key T cell-associated cell surface receptors. Methods: T cell lines including Jurkat, Molt-4, and CEM have been established using murine retroviruses (Garcia et al., in preparation) expressing functional HIV-1 tat protein. Expression of cell surface molecules was monitored by flow cytometry and mRNA levels by Northern analysis. Results: Compared to control lines, T cell lines expressing tat cease to express cell surface CD28, while retaining normal levels of CD2, CD3, CD4, CD5, CD44 and CD45RA. CD28 is a key accessory cell receptor on T cells, which when signalled via its ligand B7/BB1, augments T cell proliferation by stabilizing cytokine mRNAs (June et al., Imm. Todayll11:211, 1990). Compared to control lines, CD28 mRNA levels were also dramatically reduced in T cell lines producing tat. Treatment of CD28 negative lines expressing tat with PMA led to a restoration of optimal CD28 protein and mRNA levels by 12 hours. Whether tat affects CD28 expression by a transcriptional or posttranscriptional mechanism will be discussed. Discussion and Conclusions: Apparently, immunodeficiency viral products can reduce the expression of key accessory receptors, including CD28 by tat and CD4 by nef (Garcia et al., submitted) on T cells, in effect "handcuffing" the T cell from being triggered effectively. Asymptomatic HIV-infected individuals have decreased proliferative responses to specific recall antigens (Miedema et al., Imm. Today 1:293, 1990). The HIV-1 tatprotein is immunosuppressive and inhibits antigen-induced, but not mitogen-induced, T cell proliferation (Viscidi et al., Science 246:1606, 1989). The possibility that the loss of expression or dysfunction of CD28 in immunodeficiency virus- infected individuals may contribute to loss of memory responses is now being investigated. (Supported by NCRR grant RR00166 and NIH grants A107995 and A127291). TU.A.20 INTERFERONS INIIIBIT TIlE REPLICATION OF IIIV-1 IN MACROPHAGES BY )REDUCING BOTH VIRAL DNA AND REGULATORY GENE mRNAs. MEYLAN, Pascal RA. Guatelli J.C., Munis J.R., Kornbluth R.S., Richman D.D. University of California and VAMC, San Diego, CA, USA, Objective: Interferons (IFNs) display a potent antiviral effect on HIV-I in macrophages (MO). A high multiplicity of infection (MOI) system was used to investigate which steps of the virus life cycle are affected by IFNs. Methods: Monocyte-derived MO from scronegative donors were infected at a MOI>3 TCID/cell. Viral replication was assessed by the production of p24 antigen measured by ELISA. Proviral DNA synthesis was assessed by extracting total cellular nucleic acids and PCR amplification for an env sequence. To assess HIV transcript levels, nucleic acids were reverse transcribed and amplified by PCR with a pair of primers bracketing the splice sites of the first intron of HIV-1. This technique selectively delects the steady-state levels of tat, rev and nefmRNAs (Guatelli et al, J. Virol. 1990;64:4093-8). Results: When M,> were pretreated for 18 h with 1000 U/ml of IFN-a or -y, the signal for env DNA was reduced compared to control, suggesting that the proviral DNA synthesis was reduced severalfold compared with controls. In the same treated cells, however, the signal for spliced HIV-1 RNA's was almost completely abolished with at least a one hundredfold reduction, as was HIV antigen production. This suggests that transcript levels were not only reduced secondarily to reduced provirus formation, but primarily due to decreased transcriptional activity and/or transcript stability. To further assess this effect on HIV-1 transcripts, macrophages chronically infected with HIV and expressing high levels of transcripts were treated with IFNs. Over 1 to 2 weeks of IFN treatment, while the signal for proviral DNA remained basically unchanged, a marked reduction of the signal for spliced RNAs was observed. Conclusion: IFNs impede the life cycle of HIV-1 in MO by (1) impairing the formation of viral DNA, but primarily by (2) impairing the expression of HIV-1 genes at the mRNA level. h C> C> ^t a cs t^

Page  61 TU.A. 21 UA 1 ANALYSIS OF THE ANTI-gpl20 ANTIBODY RESPONSE IN HIV-1 INFECTED PATIENTS KOhler. Heinz*; Nara, P.**; Chamat, S.'; Caralli, V.*; Ryskamp, T.*; Haigwood, N."*, Newman, R.*; Kang, C-Y.* *IDEC Pharmaceuticals Corporation, La Jolla, CA, USA, "The National Cancer Institute, Frederick, MD, USA; "*Chiron Corporation, Emeryville, CA, USA To understand the contribution of the anti-gp120 humoral response in host defense against HIV infection, we have fractionated and characterized by neutralization and binding assays polyclonal antigp120 antibodies present in HIV-infected individuals. Total anti-gpl20 antibodies (TAGA) were purified from a pool of 4 sera by affinity chromatography on a gpl20SF2-Sepharose column and tested for their neutralizing activities. The results indicated that TAGA exhibited both type- and group-specific neutralizing activities. To dissect the epitope specificity of the group-specific neutralizing antibodies, CD4 attachment site-specific antibodies (CAGA) were isolated from TAGA using a CD4 blocked gp120SF2 -Sepharose column and were tested for their neutralizing activities. The results demonstrated that the CAGA exhibited group-specific neutralizing activities. Another approach to dissect type- and groupspecific neutralizing activities of TAGA was to separate V3 region specific antibodies (VAGA) from non-V3 region specific antibodies (NVAGA) and test their neutralizing activities. The results indicated that VAGA exhibited type-specific neutralizing activities whereas NVAGA exhibited group-specific neutralizing activities. By comparing the neutralizing activities of VAGA to those of NVAGA, we concluded that VAGA are far more effective than NVAGA in neutralizing a specific HIV isolate. Collectively, this study indicates that group-specific neutralizing anti-gp120 antibodies exist in the sera of HIV infected individuals and that these antibodies are specific for the CD4 attachment site and possibly other unknown epitopes on gp120. NOTES TUA.22 CD4+ T HELPER CELL FUNCTION IS ACTIVELY SUPPRESSED IN HIV INFECTION. Shearer.Gene M.. RoilidesE.*, PizzoPA.*, and Clerici M. Experimental Immunology and 'Pediatric Branches, NCI, NIH, Bethesda, MD, USA. Objetive. To elucidate the mechanism responsible for the selective functional loss of the CD4+, MHC self-restricted pathway of help, which is characteristic of early stages HIV+ patients, and independent of a decline in the number of CD4+ T helper lymphocytes (Th) (J.Clin.InvesL 84:1892, 1989 ). Methods. Two different approaches were employed: 1) monozygotic twins, only one of whom was HIV+ (two diffcrent pairs) and 2) a patient followed longitudinally, from before seroconversion and after development of AIDS. HIV- peripheral bl(x)d ceukocytes (PBL) were stimulated in vitro with Influenza A virus (FLU), a CD4+Th, MHC self restricted antigen, and IHLA alloantigens (ALLO) either alone or in co-culture with HIV+ PBL. HIV+ PBL were either unfractionated, or depleted of accessory cells (AC) and/or depleted of CD4+ or CD8+ T lymphocytes by plastic adherence and panning. Th function was assessed by 3H-thymidine incorporation (3H) and interleukin-2 (DL-2) production. Results. 1) HIV- cell cultures produced IL-2 and proliferated in response to FLU and ALLO; PBL from both the HIV+ twins and the AIDS patient showed a defect in production of IL-2 and 3H when stimulated with FLU; 2) In both systems, Th function of FLU stimulated HIV- PBL was abrogated in co-cultures, by addition of HIV+ PBL; 3) The suppressive activity resided in the CD8+ and not in the CD4+ lymphocyte population; plastic adherent cells also suppressed CD4+ Th function; 4) Addition of anti-tumor growth factor beta (TGF-bcta) antibody (50ug/ml), Zidovudine, dideoxyinosine or soluble CD4 (sCD4) in the co-cultures did not abrogate suppression of CD4+Th function. Conclusions. A suppressive system selective for the CD4+, MHC self-restricted pathway of help is demonstrable in IlIV infection. The suppression is mediated by CD8+ T lymphocytes and, possibly, accessory cells. The suppression is not due to IIIV itself because addition of either dideoxynucleoside analogs or sCD4 in the co-cultures did not prevent suppression. Suppression was also not prevented by anti-TGF-beta antibody. Identification and characterization of the mcchanism(s) responsible for this suppressive phenomenon may be important in preventing and/or reversing early immunologic changes that occur as a result of HIV infection. NOTES TU.A.23 RELATION BETWEEN HIV DISEASE STAGES AND THE ACTIVITIES OF HIV SPECIFIC CYIOTOXIC EFFECTOR CELLS. Rivibre, Y.*; McChesney, M.*; Porrot, F.*; Tanneau, F.*; Marie, C.**; Sansonetti, P.**; Lopez, O.***; Mollereau, G.**; Pialoux, G.**; Kieny, M P.***; Tekaia, F.*; Montagnier, L'. * Institut Pasteur, ** H6pital Pasteur, 75724 Paris Cedex 15; ** CTS Pitie-Salpptriere, 75634 Paris Cedex 13; *"" Transgene 67082 Strasbourg Cedex. Objective: To investigate the HIV specific cytotoxic activities including cytotoxic T lymphocytes (CTL) and antibody dependent cellular cytotoxicity (ADCC) during the progression of the disease in a cohort of 67 HIV-1 seropositive subjects. Methods: Analysis of proteins of HIV-1 recognized by primary cytotoxic effectors cells - mainly env, gag, pol and nef specific CTL was performed within a group of 67 adults HIV1 seropositive subjects. Autologous EBV transformed lymphoblastoid cells were infected with recombinant vaccinia viruses expressing gag, pol, nef or env and used as targets of the cytotoxic activity present in the peripheral blood mononuclear cells (PBMC) isolated from fresh blood. Results and conclusion:In a cross sectionnal study, 630 observations were made: 341 from subjects in class II, 132 in class III and 157 in class IV (CDC classification). No significant relation between stages of the disease and CTL activities were found. However, there was a significant relation between gag or pol specific CTL activities and biological parameters such as absolute T4 counts, IgG levels or tuberculin skin tests. TUA.24 THE SURVIVING CD4 CELLS IN HIV DISEASE ARE NEGATIVE FOR MEMORY AND NAIVE MARKERS Scala Enrico, Paganelli Roberto, Fanales-Belasio Emanuele, Pinter Elena, Vaccaro Donatella, Kalpaklioglu Fusun, Aiuti Fernando. University "La Sapienza", Rome, Italy. Objective: The aim was to investigate the modification of CD4+ subsets (memory: CD29+ and naive: CD45RA+) in HIV disease, in a large series of patients at different CDC stages. Further effort was directed to identify a CD4+ subset resistant to HIV. Methods: 67 HIV+ patients were compared to 20 age-matched healthy donors. Quantitative analysis for two color immunofluorescence (mAbs used were: PEconjugated anti CD29 and anti CD45RA, FITC-conjugated anti CD4) was carried out with a flow cytometer. Results were expressed as mean ~ Standard Deviation, statistical analysis was performed using the Student's two-tailed t-test. Results: A significant (p<0.001) loss of memory cells was found in CDC stage II (36~15% of total CD4+, vs 58~15% in healthy controls). Naive cells were diminished in the other stages. In all stages we observed an increased percentage (30% compared to 8% in healthy donors) of double negative (DN) cells (CD4+ / CD45RA- / CD29-). The absolute number of CD4+ DN cells in all stages and also in AIDS patients was similar to that found in healthy controls. Discussion and Conclusions: Our data suggest that CD4+/CD29+ cells are more susceptible to direct viral infection, but in the terminal stages of disease the naive subset is reduced so that a readjustment of the ratio between the two subsets was found. CD4+ DN cells a subpopulation previously not described, may represent CD4+ cells which are more resistant to HIV-induced damage. K `a K 4 Eo

Page  62 TU.A.25 DECREASED LEVELS IN MEMORY (UCHL1) AND ACTIVATION (HLADR) SUBSETS, T.A OF T-CELLS IN INFANTS MAY INFLUENCE THE RAPID COURSE OF PERINATAL HIV INFECTION. Brown. Christopher*, **; Kasali, M**; Musey, L**; Kamenga, M**; Manzila, T**; Davachi, F***; St. Louis, M**, ***; Lubaki, N**; Quinn, T*. *NIAID, Bethesda, MD; **Projet STDA, Kinshasa, Zaire; **'Mama Yemo Hospital, Kinshasa, Zaire; **"CDC, Atlanta, GA. Objective: To determine whether mean T-cell subset values unique to infants might predispose them to the more rapid course of perinatal HIV infection. Methods: T-cell subsets in infected (+) and uninfected (-) maternal (M), cord (C), and infant (I) blood at 3,6,9 and 12 months were determined by dual color flow cytometry using leu3, leu2, leu4, TCR-I, UCHLI, CD45R, HLA-DR, IL-2, leull+ 19 labelling of whole blood in a mother-child cohort of 350 HIV+ and 200 HIV- mothers. Results: While mean values of %T4 and %T8 did not differ significantly between M(-) and C(-), the mean % of T4 or T8 which expressed memory (T4+UCHL1+, T8+UCHLI +) or activation (T4+HLADR+, T8 + HLADR +) were significantly lower in cord blood and in infants compared to their mothers regardless, of serologic status (*p< 0.001). %T4 T8 %T4+UCHL1 + %T8+UCHL1+ %T4+HLADR+ %T8+HLADR+ M(-) 41 32 55 21 7.2 28 C(-) 39 27 5.2* 3.5* 1.6* 4.8* Conclusions: Memory and activation T4 and T8 subsets considered critical for host defense to HIV are significantly decreased in infants compared to adults. This difference may reflect the infants' level of immunologic immaturity and help explain their predisposition to a more rapid course of HIV disease. Correlative cytokines (TNF, IL-6) and T-cell subset values will be presented for (+) as well as (-) M, C, and TU.A.26 OPTIMING PROLIFERATIVE RESPONSES TO HIV CORE AND ENVELOPE PEFTIDES IN HIV SEROPOSITIVE INDIVIDUALS, Adams. Susan *; Owels, A*; Melville, L**; Stewart, G*. *Immunology Unit, Westmead Hospital. Sydney, Australia, ** School of Business Studies, University of Western Sydney. Sydney, Australia Objective. Four regions from the HIV core protein, p24, aa208-217, 265-286, 287-306 and 330-350 and 1 gp 120 envelope peptide, aa312-329, were selected as candidate T cell epitopes. All 5 peptides were predicited by computer analysis to be amphipathic segments. The detection of HIV epitopes capable of stimulating T cell proliferation however is made difficult by the loss of the proliferative response to specific antigens, using standard assays, as HIV infection progresses. Methods. We therefore assessed means of optimizing this response through the selection of HIV infected individuals with a CD4 count greater than 500. In addition supplementation with exogenous IL2, in vitro priming and enhancement of antigen presenting cell function were assessed. Results. Without the addition of IL2, from 14 individuals, 12% (7/59) of responses were positive (stimulation index (SI) >2.0) after 6 days and 28% (27/98) after 8 and 10 days (p<0.01). Amongst those responding the mean SI's were 6.7. 17.1 and 7.15 respectively at days 6. 8 and 10. The addition of IL2 increased the responses amongst the same 14 individuals at day 6 to 16% (10/63) however the responses at days 8 and 10 remained at 28%. The mean S's for 6. 8 and 10 days respectively, were 7.15, 7.92 and 10.3. Where individuals had a CD4 count between 500-800, 83% responded overall to 1 or more of the peptides without IL2 and 86% responded with IL2. For patients with a CD4 count >800, 86% responded overall to one or more peptides without IL2 and 43% responded with IL2. It would seem that the addition of exogenous IL2 is more effective for patients whose CD4 count is <800 and less effective when a patient has a CD4 count >800. Conclusions, In total 93% of individuals responded to 1 or more of the peptides, with 54% responding to 2 or more and 50% responding to 3 or more. The highest response (71%) of the group was to the core peptide aa 208-217 and the next highest (57%) was to the envelope peptide aa312-329. NOTES NOTES K K S^ K C> > rj t^ gj h^j ~

Page  63 TU.B.27 T B.27 DIRECT IN-VITRO INFECTION OF HUMAN SMALL INTESTINE AND COLON WITH HIV Fleming Simon, Kapembwa M, MacDonald T*, Griffin G. Division of Communicable Diseases, St George's Hospital Medical School, London U.K. and * Dept Paediatric Gastroenterology St Bartholomews Hospital, London U.K. Objective: To determine whether human intestine can be direcly infected with HIV-1 in vitro, and to monitor infection by immunohistochemistry (IHC), in-situ-hybridiation (ISH), and biochemical assay Methods: Human foetal (16-20 weeks) intestinal explants (2 mm ) maintained in culture, tere exposed to 25ug/ml DEAE-dextran for 60 mins, washed and incubated with HIV-1 (strains RF, RUT,IIIB) all0 tissue culture infectious dose / ml for 2 hours. After thorough washing to remove free virus, explants were maintained in culture. Tissue culture fluid (TCF) and explants were harvested and snap frozen on days 4,7, 10, and 14 post infection, at which times culture fluid was replaced with fresh medium. Day 14 explants and TCF were co-cultured with lurkat T-cells to detect infectious virus panicles. Reverse transcriptase (RT) and p24 Ag levels were assayed in TCF. IHC staining (ABC pcroxidase) for p24 Ag and gp41 Ag and ISH for viral RNA were performed on 8um frozen tissue sections. Results: RT and p24 Ag levels in TCF rose between days 7 and14 days post viral inoculation. Jurkat co-cultures confirmed the presence of infectious virus. Cells in the lamina propria resembling macrophages and lymphocytes of both the small intestine and colon stained positively (IHC) for p24 and gp 41. Similar cells in the lamina propria also positively stained for HIV RNA using ISH. Dual label IHC (ABC peroxidase + APAAP) and combined IHC + ISH confirmed the presence of viral protein and RNA in cells bearing CD3, CD4 (leu 3a) or macrophage (CD68) markers. Epithelial cells showed no evidence of HIV-1 infection at any time. Conclusions. Cells of the lamina propria of the human small intestine and colon bearing lymphocyte or macrophage markers can be directly infected by and support the replication of HIV-1 and this maybe important in the pathogenesis of HIV enteropathy. NOTES TU.B.28 LYMPHOID TISSUES FUNCTION AS "RESERVOIRS" FOR HIV INFECTION Graziosi Cecilia", Pantaleo G, Schnittman S', Greenhouse J, Butini L, Kotler D", Fauci AS'. LIR, NIAID, NIH, Bethesda, MD, USA and St. Luke's Hospital, New York, NY, USA ' jective: To determine whheer lymphoid tissues may function as "reservoirs" for HIV Infection. Methods: Comparative analysis of the frequency of HIV-1 infected cells was performed by quantitative polymerase chain reaction (PCR) amplification on unfractionated and sorted CD4* T cell populations isolated from peripheral blood (PB) and lymphoid tissues (lymph nodes, tonsils or adenoids) of the same patients. Unfractionated or sorted CD4* T cells were lysed and cell lysates serially diluted. HIV-1 DNA was amplified either with gag (SK145/101) or LTR (SK29/30) primers [and probed with SK102 (gag) and SK31 (LTR)] and compared to the DNA amplification of tenfold serial dilutions of cell lysates of a chronically infected T cell clone (ACH2) that contains one proviral copy per cell. Results: Seven HIV-seropositive patients (six Group III and one Group IV according to CDC classification) with a CD4t T cells percent in the PB of 25.9 + 12.1 (mean + standard deviation) were studied. By performing quantitative PCR amplification we have demonstrated that the frequency of HIV-1 infected CD4' T lymphocytes was significantly higher (from 1/2 log to 1 and 1/2 log) in the lymphoid tissues,eompared to PB of the same patients. In fact, in the group of patients studied the frequency of HIV-1 inlectef CD4* T cells in the lymphoid tissues ranged between 1/100 to 1/1000 compared to 1/10,000 to 1/50,000 in the PB. In addition, identical results were obtained on cells that were serially diluted before being lysed, thus indicating that the higher HIV burden in the lymphoid tissue was not the result of few infected cells containing multiple copies of viral DNA. Furthermore, by using a modified PCR method for RNA analysis we detected HIV-1 specific RNAs for structural and regulatory proteins of HIV-1 in cells isolated either from PB or from lymphoid tissues. Discussion and Conuson: These results demonstrate that lymphoid tissues are a major "reservoir" for HIV infection in vivo. In addition, they also indicate that the HIV burden in vivo, even in patients not in the advanced stage of disease, is significantly higher compared to what was previously thought on the basis of analysis limited to the PB. NOTES CI Q-) SI EFFECT OF HIV ON ANTIGEN PRESENTATION BY DENDRITIC CELLS ~UNOCY'ES U.B.29 Knight, Stella C; Macatonia, SE*; Patterson, S*; Gompels, M Pinching, AJi* Clinical Research Centre, Harrow, U.K., *St Mary's Hospital Medical School, London, U.K. Objective: To compare antigen presentation by dendritic cells (DC) and monocytes (MO) at different stages of HIV disease or following in vitro HIV infection. Many DC (3-21%) but few MO (<0.2%) are infected in HIV seropositive patients. Methods: DC and MO were isolated from blood of HIV-infected individuals or from uninfected donors and exposed to HIV in vitro. DC and MO were pulsed with influenza virus (INF) to stimulate autologous T cell proliferation, or used untreated to stimulate allogeneic lymphocytes in primary mixed leukocyte reactions (MjL). Results: DC and MO from uninfected individuals stimulated secondary T cell responses to INF but only DC, and not MO induced MLR. In HIV seropositive asymptomatics (CDC stage II), T cell responses to alloantigens and to INF presented by DC were impaired, whilst presentation of INF by MO remained normal. MO and DC exposed to HIV in vitro produced a similar pattern of stimulation. In cells from patients with AIDS (CDC stage IV), both DC and MO failed to induce T cell proliferation. Conclusion: DC normally present antigens to recruit resting T cells into immune responses. Other class II-bearing cells, such as MO, can promote responses to recall antigens in activated T cells. In early HIV infection stimulation of resting T cells by DC may be blocked, but MO will be able to present antigens to activated cells. Loss of activated cells in AIDS without effective recruitment could underlie the progressive decline of immune responses. TU.B.30 NEURONAL LOSS IN THE FRONTAL CORTEX IN HIV INFECTED INDIVIDUALS Everall lan Paul, Luthert PJ, Lantos PL. Department of Neuropathology, Institute of Psychiatry, London SES, England Objective: To clarify the extent of HIV associated neuronal loss. HIV1 causes the AIDS-dementia complex whose precise mechanism remains unknown but has been associated with HIV encephalitis. Method: We quantitatively assessed neuronal populations in the superior frontal gyrus, by a stereological technique called the disector. This requires consistent visual identification of neurons for quantification and is more reliable than computer-assisted image analysis which defines neurons by an arbitrary cell size range. From a cohort of 65 cases, 11 brains were found to have no evidence of opportunistic infections or neoplasms. Six brains had HIV encephalitis and five had minimal pathology (reactive astrocytosis and perivascular cuffing). They were compared to eight age-matched controls. Reults: There was significant and consistent loss of cortical neurons, approximately 38%, in the HIV infected group (p < 0.0001 by Student's test). This neuronal fall-out occurred in the same proportion regardless of the presence or absence of HIV-1 encephalitis. Discussion and Conclusions: This is the first quantitative study which clearly demonstrates that HIV-1 infection, without apparent encephalitis, caused substantial neuronal loss. This finding contributes to the understanding of dementia in AIDS patients and has therapeutic implications since the neuronal loss may be preventable. C)

Page  64 TU.B.31 HIV ASSOCIATED NEPHROPATHY IN TRANSGENIC MICE Kopp JB*; Weeks BS*; Marinos NJ**; Bryant JL**; Dickie P**; Notkins AL**; Klotman, Paul E.*. *Laboratory of Developmental Biology, *Laboratory of Oral Medicine, ***Animal Care Unit, National Institute of Dental Research, NIH, Bethesda, MD, USA. Objective: The objective of these studies was to investigate the role of viral gene products in the pathogenesis of HIV-associated nephropathy (HIVAN). Methods: We established a transgenic mouse model using a non-infectious HIV-1 provirus with a 3 kb deletion overlapping gag and pol genes. This construct contained the LTRs and encoded envelope and the regulatory proteins Tat, Rev, Nef, Vif, Vpr, & Vpu. Results: Southern blots of 2 lines indicated that multiple proviral copies were integrated into different and unique sites. Heterozygous mice developed proteinuria by age 25 and died of azotemia between 60 and 100 days. Light microscopy showed focal segmental glomerulosclerosis, microcystic tubular dilatation, and a sparse monocytic interstitial infiltrate. Indirect immunofluorescence revealed increased glomerular deposition of laminin, collagen IV, fibronectin, and heparan sulfate proteoglycan. Northern analysis of total kidney RNA demonstrated doubly-spliced, singly-spliced, and unspliced proviral transcripts. Kidney immunoblots demonstrated polypeptides corresponding to GP41. Indirect immunofluorescent analysis showed Rev protein in sclerotic glomeruli. Conclusions: In this transgenic model, HIV-1 genes are expressed in the kidney and expression is associated with renal pathology closely resembling that of AIDS patients. These findings implicate HIV-1 proteins in the pathogenesis of HIVAN. NOTES TU.B.32 MOLECULAR BIOLOGICAL DEMONSTRATION OF VIRAL GENOME IN RENAL BIOPSIES OF HIV INFECTED PATIENTS WITH IgA NEPHROPATHY Kimmel. Paul L*; Phillips, TM*; Abraham, AA*; Szallasi, T**; Ferreira-Centeno, A"; Garrett, CT" Departments of Medicine* and Pathology", George Washington University, Washington, DC, USA. Twenty-two HIV infected patients, all male, had renal biopsies at our center. Only 4 were white. Two of the white males had IgA nephropathy clinically characterized by renal insufficiency, proteinuria and hematuria. Both patients had positive hepatitis B virus surface antibody serology, but negative HBSAg. Both had urinary shedding of HIV p24 antigens, detected by dot-blot, and circulating antibodies to HIV p17, p24 and gp41. Patient A had circulating IgG antibodies and patient B IgM antibodies directed against the HIV antigens. Patient A had focal proliferative glomerulonephritis; Patient B had mesangial nephropathy. Both patients had circulating immune complexes (CIC), detected by Clq binding, and polyethylene glycol precipitation and sedimentation, composed of IgA antibodies. Patient A's CIC was an IgA antibody and a protein antigen which was unreactive with anti-HIV monoclonal antibodies. Patient B's CIC was an IgA antibody and IgM antigen, the latter with specific activity against HIV p24 antigen. His renal biopsy tissue was eluted, and the eluate separated by agarose electrophoresis demonstrated IgM, IgA and p24. Patient A had no tissue available for elution. Both patient's renal biopsies evaluated by polymerase chain reaction (PCR) showed the presence of HIV DNA, using primers and probes to gag gene, detected by liquid hybridization and polyacrylamide gel electrophoresis. We conclude IgA nephropathy may be common in HIV infected white males with renal disease. Renal deposition of CIC's, perhaps a result of polyclonal immune dysregulation secondary to HIV infection, may play a pathogenic role in the development of renal disease. Alternatively, the PCR findings suggest an in situ mechanism of immune complex formation may occur in patients with renal cellular incorporation of HIV genome. NOTES T.B33 DIAGNOSIS OFPERINATALHTV-IINFECI TONBYDETECTIONOFVIRAL-SPECIFIC IGA ANTIBODIES IN U.S. AND HAITIAN PEDIATRIC POPULATIONS Quinn, Thomas C*; Kline, R*; Halsey, N**; Ruff, A**; Boulos, R***; Butz, A**; Hutton N**; Modlin, J**. *LIR, NIAID, Bethesda, MD, *Department of Pediatrics, Johns Hopkins University, Baltimore, MD, ***Centers for Development and Health, Port au Prince, Haiti. Objective: The serologic diagnosis of HIV infection in infants is complicated by persistence of maternal IgG antibody to HIV in the infant. Since IgA does not cross the placenta, we evaluated the sensitivity and specificity of a HIV specific IgA assay for the early diagnosis of HIV infection in infants. Methods: Sera were prospectively collected from children between the ages of 0 and 12 months who were born to HIV seropositive women (n= 104) and to HIV seronegative mothers (n= 100). After 3 absorptions of serum IgG with protein G, specimens were assayed by Western blot with use of IgA-alkaline phosphatase followed by BCIP/NBT. A sample was positive for IgA if viral bands p24, gp41, gpl20 or gpl60 were present. Infection status of the infants was determined by IgG seropositivity after they were > 15 months of age, or by PCR and culture. Results: Of 65 children known to be HIV infected, 62 (95.4%) were IgA positive between 3 and 12 months of age. Of 139 children who were IgG seronegative after 12 months of age, IgA antibodies were absent in 137 (98.6%) samples collected between 3 and 12 months of age. One child who was born to a seropositive mother was IgG seronegative at 24 months, but was PCR and IgA positive. Thus, the sensitivity of the IgA assay for diagnosis of HIV infection between 4 and 12 months of age was 95.4% and the specificity was 98.6%. Positive and negative predictive values were 96.9% and 97.9%, respectively. Conclusions: The IgA assay was sensitive and specific for diagnosis of HIV infection in infants as early as 3-6 months of age. Additional testing of sera from children between 0 and 3 months of age is currently underway. The advantages of this assay is that it is relatively inexpensive compared to culture or PCR, and can be used to identify infected infants early so that anti-viral therapy may be initiated. TU.B.34 LABORATORY DIAGNOSIS OF PERINATAL HIV INFECTION: EXPERIENCE WITH A COHORT OF 350 HIV SEROPOSITIVE MOTHER-INFANT PAIRS FROM BIRTH TO AGE ONE YEAR. Lubaki, Ndongala*; Brown, C*,**; Kasali, M*; Behets, F*; Munsey, L*; Manzila, T*; Davachi, F***, Ou, C****; St. Louis, M****; Nelson, A*, ****; Firpo, A*****; Saisun, J*****; Quinn, T*. *Projet SIDA, Kinshasa, Zaire; **NIAID, Bethesda, MD; ***Mama Yemo Hospital, Kinshasa, Zaire; ****CDC, Atlanta, GA; *****AFIP, Washington, DC; ******INRB, Kinshasa, Zaire. Objective: To determine the utility of varied laboratory tests, singly or in combination, to diagnose perinatal HIV infection at the earliest possible age in children born (CB+) to HIV (+) mothers. Methods: In 350 (CB+) and 250 children born (CB-) to HIV- mothers, we examined Western blot (WB), in vitro antibody production (IVAP), viral culture (VC), T-cell subsets, PCR, IgA and anti-p24 reactivity in tissues for evidence of perinatal HIV infection. Results: Percent positive VC, IVAP, WB results are given for (CB-). Percent difference between values for CB- and VC+ are given for various T-cell subsets: T4, TB, T8+HLADR+(8DR) and T4+HLADR+(4DR) and T8+UCHL1+(8U). VC IVAP IgGWD %T4 %T8 %8DR %4DR %8U cord blood - > 95 - 12 <1 61 54 60 6 months 19 24 - 33 44 50 33 53 12 months 17 19 39 42 38 33 66 55 Conclusions: Used alone, no single test appears to be both sensitive and specific. Use of novel T-cell subsets in combination with other tests, may be useful for diagnosis as early as birth. Utility of various combinations will be examined based on PCR and viral culture results. C> C> C> C> Q-) t4

Page  65 TU.B.35 EARLY DIAGNOSIS OF HIV-1 INFECTION IN NEWBORNS: COMPARISON OF PCR-DNA AND VIRAL CULTURE. Burgard. Marianne*; Rouzioux, C.*; Blanche, S.*, Mayaux, M.J.**; Kouznetzoff, A.*; Griscelli, C.*; HIV Infection in Newborn French Collaborative Study Group.* H^pital Necker, Paris; **INSERM U.292, Hopital du Kremlin-Bic&re, FRANCE. Objectives: Semi quantitative PCR-DNA technique has been developed in order to evaluate its diagnostic value in the first month of life in comparison to a sensitive viral culture. Patients and methods: 35 infants born to HIV-I positive mothers have been recruited through the french prospective study. All specimens were collected before the age of 2 months and available to make the two techniques in good conditions. The children are now older than 18 months: 20 of them are HIV- I infected and 15 have become completly seronegative. PCR-DNA technique has been performed on coded samples, using two pairs of primers in Pol region and dilutions of 8E5 T-cell line as standard for the quantitation of the signal. A standard method of viral culture has been performed on fresh samples in optimal conditions using fresh normal cells and a sensitive system for viral recovery. Results: TU.B.36 DYSRHYTHMIAS, UNEXPECTED ARREST AND SUDDEN DEATH IN PEDIATRIC HIV INFECTION. Lipshultz, Steven E.; Luginbuhl, Lynn; McIntosh, Kenneth. Children's Hospital, Boston, MA, USA. 1 Objective: Cardiovascular abnormalities are noted in >90% of HIV+ infants and children and may result in mnorbidity/mortality. We determined 1) pulse & blood pressure(BP) both at rest and with interventions, 2)1 dysrhythmias, 3) unexpected arrest and sudden death, and 4) autonomic function in pediatric HIV+ pts. Methods: All symptomatic (CDC class P-2) HIV+ pts (n=105) with cardiac evaluation over 7 yrs were! included (79% of all P-2 pts). These pts had 277 load-independent ventricular function echos with resting pulse/BPs. ECG, Holter, autonomic evaluation and medical records were reviewed. Results: Tachycardia: was noted in 47% (49/105) and bradycardia in 7% (7/105) of pts. Hypertension was noted in 8% (8/105) and: hypotension in 11% (11/105) of pts. Severe intractable hypertension proved fatal in 1 pt with encephalopathy' and baseline ventricular dysfunction but was observed in other children with acute deterioration requiring intensive care support. Dysrhythmias were common and were related to hyperdynamic ventricular function and reduced afterload, a state noted in pts with increased sympathetic tone. Cardiorespiratory analysis of autonomic function in 1 pt with Torsades de pointes (pentamidine related), QTc prolongation, hyperdynamic function and reduced afterload demonstrated nearly pure sympathetic cardiac inputs. Over 2.5 yrs of followtip this pt maintained heightened sympathetic tone and arrhythmogenic responses to pentamidine while ientricular function deteriorated sharply, consistent with a sustained increased sympathetic state. Marked kinus arrhythmia was found in 15% of pts including 30% of pts with unexpected arrest. High grade entricular ectopy was found in 5% of all pts but these included 30% of pts with unexpected arrest. Fourteen pisodes of unexpected cardiopulmonary arrest (3 fatal) were noted in 12 pts (11%) and related to new rocedures in 2, new medications in 6 and had no known relation in 6 episodes. 83% (10/12) of pts with udden arrest had HIV encephalopathy (48% of all children with encephalopathy [10/21]). Conclusion: lemodynamic abnormalities, dysrhythmias, unexpected arrest and/or sudden death are common in HIV+;hildren, especially with acute deterioration, medical intervention or neurologic involvement. Abnormal Vardiac autonomic regulation is a likely contributor to these problems and should be assessed prospectively. Infected children (20) Viral culture + Seroreverted children (15) Viral culture PCR 3 + 0 15 Conclusion: PCR and viral culture have a very good specificity. The sensitivity is found very similar and not very high, it is 50 % for PCR despite very controlled conditions (regular detection of 5 copies of viral DNA). Results of semi quantitative PCR permit to show that the number of proviral DNA copies in the specimens of infected newborns is statistically and significantly lower than for low positive controls (asymptomatic adults), sustaining the hypothesis of a late viral transmission during pregnancy. NOTES NOTES TU.B37 NEURODEVELOPMENTAL OUTCOME OF PERINATALLY ACQUIRED HIV INFECTION ON THE FIRST 24 MONTHS OF LIFE - Hittelman. Joan*; Willoughby, A.**; Nelson, N.*; Gong,J.*; Mendez, H.*; Holman, S.*; Goedert, J.**; Landesman, S.* *SUNY, Health Science Center, Brooklyn, N.Y., **National Institutes of Health, Bethesda, MD. USA. Objective: To assess the neurodevelopmental outcome of perinatally acquired HIV infection on the developing child. Methoa4s 138 infants born to either HIV infected mothers (n=59) or seronegative controls (n=68) were assessed on the Bayley Scales at 3 month intervals by an examiner blind to serostatus. Using Chi-square, the following groups were compared: HIV infected infants (n=17), seroreverters (n=48) and controls (n=73). Results: Infected infants showed significantly more delays than either seroreverters (p<.025) or controls (p<.025). No differences were found between the other two groups. Infected infants had a developmental disability rate of 56% in contrast to a rate of 22% of seroreverters and 23% of controls. The relationship between disability and serostatus was found at each assessment point throughout the first two years of life (p<.001 to p<.05); only motor disability differentiated the groups at each age (p<.0001 to p<.005). Except at 3 months of age, neither cognitive nor language disability differentiated the groups at any age. conclusion: The prospective data demonstrated a strong and persistent relationship between perinatally-acquired HIV infection and neurodevelopmental disability, especially with respect to motor deficits. TU.B.38 INFECTION WITH CYTOMEGALOVIRUS (CMV) IN CHILDREN OF SEROPOSITIVE MOTHERS Giaquinto.Carlo,Gabriele LTPagliaro A,Zaninotto M,Giacomelli A,Ruga. EjCogol Cozzar-t F,R'lmano rP',1'Elia R.Dpt of Pediatrics,*Inst.of Hygiene,Univ.of Padova, ITALY Objective:To determine the incidence of CMV infection and its effects on the progression of HIV infection in children born to seropositive mothers. Methods:106 children were followed up from birth clinically, immunologically and virologically 3-monthly.At each clinic urine were collected and inoculated in human fibroblasts culture;CMV was identified by its characteristic cytophatic effect.Children weri classified as HIV infected according to CDC criteria.The prevalence of CMV infection al 6 and 12 months was evaluated in HIV infected and uninfected children. We correlated CMV infection with the clinical progression (CDC classification) and the immune status (CD4/CD8, total CD4, HyperIg) in the 3 months after CMV diagnosis. Results:At 6 months 6/23(25%) infected and 8/83(10%) uninfected children were CMVposit: ve (p.ns;x'); at 12 months 7/23 (30%) infected and 15/83(18%) uninfected were positive (p-ns; ~`). The figure shows the correlation between CMV infection and immune abnormalities or clinical progression of HIV disease in 23 infected children. Immune abnorm Clinical progress. There is a significant correlatior + - + - between CMV infection and the im+ 9 2 4 7 mune abnormalities (p<.05;Fischer CMV - 4 8 5 7 test)but no with clinical progres! Conclusions: The possibility that CMV infectionmay affect the course of HIVinfectior in children without causing directly opportunistic infection must be considered. K0

Page  66 TU.C.39 P24 ANTIGENAEMIA, CD4 LYMPHOCYTE COUNTS AND AIDS Phillips AN,Lee C,Elford J,Janossy G,Griffiths P,Kernoff PBA. Royal Free Hospital & School of Medicine, London, UK. Objective: To assess whether the raised risk of AIDS in persons with p24 antigenaemia is mediated by a steeper decline in CD4 lymphocyte count. Methods: A cohort of 111 HIV +ve haemophiliacs followed for up to 11 years. CD4 lymphocyte counts and the presence/absence of p24 antigen and have been regularly assessed. By 30 Nov 1990, 34 patients had AIDS. Results: Thirty patients were p24 +ve Mdln 04 ymphocoun~ t Medan CD4 lymffh un In on at least one occasion. Of these,, p.tiet. p24,.ntaurt poiu prnty p2 -4 19 developed AIDS compared with 15 of i the 81 patients persistently p24 -ve. Fitting p24 antigenaemia as a time- " dependent covariate in a Cox model 0 the relative risk of AIDS was 6.36, p (p=.0001). The Figure shows that patients developing p24 antigenaemia show a steeper decline in CD4 count (p=.006) which appears to exist even before the first p24 +ve test. When eps m W. ** t m *m*WV the CD4 count was added to the Cox model, also fitted as a timedependent covariate, the relative risk for p24 antigenaemia fell to nonsignificance (p=.l, relative risk=1.99). The relative risk for the CD4 count was highly significant (p=.00001). Conclusion: The raised AIDS risk in persons with p24 antigenaemia appears to be mediated by a more rapid decline in CD4 lymphocyte count. TU.C.40 PROGNOSTIC INDICATORS FOR DEVELOPING AIDS WITHIN 24 MONTHS IN A COHORT OF INTRAVENOUS DRUG USERS (IVDUs): PRELIMINARY RESULTS. Vlahov David, Munoz A, Solomon L, Margolick J, Nelson KE, Bareta JC,* Cohn S, Astemborski J. The Johns Hopkins School of Hygiene and Public Health, *Maryland Dept. of Health and Hygiene Baltimore, MD, USA. Objective: To identify laboratory and clinical variables predicting onset of AIDS in a cohort of IVDUs recruited in 1988-1989. Methods: 631 seropositive who were AIDSfree at baseline were followed for 2 yrs. Relative hazards (RH) using Cox regression were estimated to quantify the predictive power of baseline CD4+ number for developing AIDS within 2 years; to this model, age, race, sex, current drug use, serum neopterin and beta-2 microglobulin (B2M), and number of clinical symptoms (sx) at baseline were added one at a time. Results: 42 IVDUs developed AIDS within 2 yrs. Cumulative % with AIDS by # of months CD4 # N AIDS 6mos 12mos 18mos 24mos RH (95% C.I.) <200 32 9 16% 26% 30% - 1.00 201-500 248 23 2% 5% 7% 12% 0.29 (0.13,0.62) >500 351 10 1% 2% 3% - 0.09 (0.04,0.22) In multivariate analysis, both >1 sx and serum neopterin >12 Nmol/L showed significant association with time to AIDS after controlling for CD4, with the RH (95% C.I.) as 2.5 (1.3,5.1) and 2.2 (1.2,4.2), respectively. Conclusion: In these IVDUs, risk of AIDS was strongly associated with depression in CD4 numbers and >1 sx. Among other markers, serum neopterin, but not serum B2M added predictive power. These parameters may be helpful in patient management and design of therapeutic trials. NOTES NOTES TU C 41 IMPROVED SURVIVAL FOR PERSONS WITH AIDS IN SAN FRANCISCO Lemp, George F; Hirozawa AM; Araneta MR; Young K; Nieri G. San Francisco Department of Public Health, San Francisco, CA, U.S.A. Objective: To examine survival trends for persons with AIDS in San Francisco and to determine the effect of therapy on patient survival. Methods: We calculated survival (Kaplan-Meier product-limit) following diagnosis of AIDS for 8947 cases reported in San Francisco between July 1981 and June 15, 1990. Patients were followed prospectively through December 15, 1990. We also examined the relative effect of drug therapy for 2572 of 7047 patients diagnosed between 1986 and 1990. Results: Significantly (p<0.0001) improved survival was observed for patients initially diagnosed with Pneumocystis carinii pneumonia (PCP), or other opportunistic infections (01). Survival for patients diagnosed with Kaposi's sarcoma (KS) improved slightly, but survival for patients diagnosed with lymphoma(s) was unchanged (see Table): Year of PCP only 01 only KS only Lymphoma(s) Diagnosis Median 1-year Median 1-year Median -year Medan -year 1981-1985 10.3 mo 43.0% 8.7 mo 39.2% 18.7 mo 67.4% 10.5 mo 47.2% 1986-1987 17.9 68.1 9.7 43.5 22.7 75.9 6.1 33.5 1988-1990 21.4 76.2 13.3 53.5 22.1 79.4 7.1 33.1?atients receiving drug therapy exhibited significantly (p<0.0001) improved survival then compared with those who did not (see Table): AZT only ddl only Aeropent only AZT+Aeropent AZT+Acyclov No therapy Number 1101 32 50 205 77 478 Median, mo 22.3 ---- 25.1 24.1 22.8 12.1 2-year,% 46.5% 66.6% 51.5% 49.7% 48.8% 27.1% Conclusion: Survival following diagnosis of AIDS continues to improve. TU.C.42 INCIDENCE OF LYMPHOMAS AND OTHER CANCERS IN 1545 HEMOPHILIACS Rabkin. Charles S.; Goedert, J.J.; and the NCI Hemophilia Cohort Study Collaborators. National Cancer Institute, Bethesda, MD, USA. OBJECTIVE: To determine types and rates of cancers associated with HIV-1 infection. METHODS: We examined cancer incidence in 791 HIV-seropositive (HIV+) and 754 HIVseronegative (HIV-) subjects with hemophilia followed for 5688 and 4637 person-years (p yr), respectively. We calculated expected cancers from estimated U.S. age-specific rates, RESULTS: Among the HIV+ subjects, there were 10 cases of cancers known to be HIVassociated: 9 non-Hodgkin lymphoma (NHL) and 1 Kaposi's sarcoma. In addition, there were 8 cases of other cancers: 2 basal cell and 1 each of squamous cell skin, melanoma, palatal, parotid, lung, and Hodgkin disease. Excluding non-melanoma skin cancers (for hich U.S. rates are unavailable), 0.35 case of NHL and 5.8 other cancers were xpected as compared to the 9 NHL and 6 other tumors observed. NHL incidence was 10/10 5 p-yr in adults and 50/105 p-yr in children, representing a 25-fold excess in each age group. CD4+ lymphopenia and longer duration of infection increased NHL risk. In the IV- subjects, 6 cancer cases were observed (basal cell, melanoma, stomach, bladder, ung, and Hodgkin disease) versus 5.5 cancers other than non-melanoma skin expected. CONCLUSION: Cancer incidence in children and adults with hemophilia is similar to that of the general population, except for an excess of NHL with HIV infection. Paradoxically, mprovements in therapy of HIV infection which prolong survival of advanced mmunodeficiency may lead to further increases in the risk of HIV-associated Ivmohoma. CI I1 0~ r^ Cj Qi C` ~C

Page  67 TU.C.43 INFLUENCE OF GENDER, AGE AND TRANSMISSION CATEGORY ON THE PROGRESSION FROM HIV SEROCONVERSION TO AIDS Pezzotti. Patrizio; Rezza, G; Zerboni, R; Lazzarin, A; Angarano, G; Sinicco, A; Aiuti, F; Ricchi, E; Canessa, A; Barbanera, M; Carbonari, P; Lo Caputo, S; Tirelli, U. IMCSA, AIDS Unit, Istituto Superiore di Sanita - Rome, Italy. Objective: To evaluate the risk of developing AIDS in different risk groups, different age classes and in males versus females. Methods: A total of 738 subjects from 15 clinics in Italy who had undergone serial HIV testing and had seroconverted were enrolled in an incident cohort study. Of these, 440 had probably acquired HIV infection through intravenous drug use (IDU), 183 through male-to-male sex (HM), and 115 through hetero-sexual contact (HC). The majority (74%) were males. Seroconversion date was estimated as the midpoint between the dates of the last negative and first positive test. Generally, clinical and laboratory status were evaluated every 6 months. Disease progression from seroconversion to the date of the AIDS diagnosis was calculated according to the Kaplan-Meier product limit method. For the entire cohort and for each sex, age and risk group, statistical differences between groups was evaluated using the twotailed Cox and Gehan-Breslow exact inverse probabilities. Results: The median follow-up times for IDUs, HMs and HCs were 43, 42 and 29 months respectively. The mean age at seroconversion was 25 years for IDUs, 33 years for HMs, and 28 for HCs. The risk of developing AIDS was significantly higher in HMs than in the other two risk groups. Subjects less than 25 years of age had a significantly lower risk of developing AIDS than those aged 25 years or over. We did not found a difference in progression towards AIDS by gender. After adjusting for age, there was no difference in AIDS incidence among different risk groups. Discussion and Conclusions: Age at seroconversion plays an important role in the clinical outcome of HIV infection. The higher mean age of homosexual males explains the faster progression in this group. Age should be taken into account in any modelling of the HIV epidemic. TUCTTc T CELL SUBSETS IN A COHORT OF INTRAVENOUS DRUG USERS (IVDUs): INITIAL LEVELS, RATE OF CHANGE, AND EFFECT OF PATTERN OF DRUG USE. Margolick. Joseph B, Mufloz A, Vlahov D, Astemborski J, Cohn S, Nelson KE, The ALIVE Study, Johns Hopkins School of Hygiene and Public Health, Baltimore, MD, USA. Objective: To evaluate changes in T cell subsets in IVDUs who were HIV-1 seropositive (SP) and seronegative (SN) at enrollment, and in IVDUs with known time of seroconversion (SC). Methods: 859 IVDUs (621 SP, 152 SN, and 86 SC) in the ALIVE cohort were evaluated semiannually. Proportions (%) and absolute numbers (/mm3) of CD3, CD4, and CD8 lymphocytes were determined by flow cytometry and complete blood count with automated differential. Changes in T cell subsets were analyzed by dividing change in cell numbers by total observation period for SP subjects and by period of seropositivity for SC. Results: Median numbers/mm3 of CD4 and CD8 lymphocytes at enrollment were 1061 and 627 for SN and 508 and 894 for SP. For SC, the corresponding levels were 734 and 889 at the first visit (median of 4.5 months) after the estimated time of seroconversion. For SP, median rates of decline in absolute and percent CD4 lymphocytes were 8 cells (0.0%)/6 mo (median followup - 18 mo). For SC, these figures were 49 cells (1.9%)/6 mo post seroconversion (median followup - 12 mo). Among SP classified as either drug nonusers, infrequent users (use in some 6-mo intervals), or persistent users (use in all 6-mo intervals), median CD4 cell numbers at enrollment were 532, 522, and 475 cells/mm3, respectively, with N - 32, 310, and 279, respectively. However, rates of CD4 cell decline did not differ among these groups. Discussion and Conclusions: SP IVDUs in the ALIVE cohort experienced a much more gradual decline in CD4 lymphocytes than did SC. Among SP, self-reported pattern of drug use influenced baseline levels, but not rate of decline, of CD4 lymphocytes. The data suggest that HIV-1 infection in IVDUs resembles that reported for other risk groups, with an early rapid decline in CD4 cells followed by slow progression. NOTES NOTES TU.C.45 INCUBATION TIME FOR VERTICALLY INFECTED AIDS MaWhinney, Samantha; Pagano, M. Harvard School of Public Health, Boston, MA, USA Objective: Estimation of the incubation time for vertically infected HIV. Methods: The pediatric AIDS epidemic is in its infancy and long incubation periods have not been observed. The average time to onset of disease will therefore, be underestimated if based simply on cases already reported. In contrast to adults, the number of HIV vertical infections can be directly estimated by screening newborns for HIV antibodies. Using Newborn Screening Data from New England states, (total positive = 405, diagnosed = 13) with estimates of the conditional incubation distribution and transmission rate, we estimate the actual distribution using maximum likelihood methods. Prior distributions are placed on the estimated parameters in the likelihood function, thereby incorporating their uncertainty into the results. Results: Using a published conditional distribution and an estimate of vertical transmission (mean - 33, std dev =.075), we examine three models that allow the latency assuming-1.0,.95 and.90 as the probability of being HIV positive, given a positive test. The results give median times of 10.4, 9.3 and 8.7 years respectively. The yearly cumulative distribution assuming no false positives is approximately 11% for the first year with an additional 4% per annum for the next nine years. The standard deviation is.05 for the first year and increases to.18 by year ten. Discussion: Because the observed cases of vertically transmitted AIDS have so far been primarily in younger children, it has been reported in the literature that most become symptomatic before one year of age. Our results, however indicate a longer incubation time with a median time to AIDS similar to that of adults and children u infected by contaminated blood products. TU.C.46 HETEROSEXUALLY ACQUIRED HIV-1 INFECTION AND SAFETY OF THE BLOOD SUPPLY: A MULTICENTER STUDY OF SEROPOSITIVE BLOOD DONORS. Petersen, Lyle'; Doll, L'; White, C'; Investigators at 20 U.S. blood centers. Centers for Disease Control; Atlanta, Georgia, USA Objectives: To determine risk factors for HIV-I seropositive blood donors, estimate the proportion who acquired HIV-1 heterosexually, and evaluate possible measures to exclude these persons from donating blood. Methods: From May 1988 - Dec 1990, we interviewed seropositive blood donors at 20 large U.S. blood centers to determine HIV-1 related risk behaviors. Donation and laboratory characteristics were obtained from blood center records. We also interviewed and HIV-tested sex partners of persons with no identified risk (NIR). Results: Of 1,299 (0.018%) HIV-1 positive donors identified from 7,206,000 donations, 771 have been interviewed to date. Of 564 men, 9% had heterosexual contact with a person at risk for HIV (HC)(86% with intravenous drug users [IVDU]) and 29% had NIR. Of 207 women, 49% had HC (67% with IVDU, 15% with bisexual men) and 46% had NIR. Many with HC or NIR had other abnormal test results (HTLV-1, HBcAb, HbsAg, Hepatitis C, RPR)(HC: 43% of men, 16% of women; NIR: 43% of men, 21% of women). A similar proportion reported STDs (syphilis/gonorrhea <3 years previouslyXHC: 40% of men, 11% of women; NIR: 17% of men, 7% of women). Few had numerous sex partners (median since 1978 = HC: 12 for men, 5 for women; NIR: 6 for men, 5 for women). Of those with HC, 24% (25% of women, 20% of men) were unaware of their risk before donation. Conclusion: Although the low HIV-1 prevalence signifies that most high-risk persons are deferred from donating, those with HC may be relatively difficult to identify and defer. Many seropositive donors with HC were unaware of their risk before donation. We estimate that 20-53% of seropositive donors acquired HIV-1 by HC (depending on what proportion of those with NIR actually had HC). Routine serologic testing for other diseases will preclude from transfusion 20-40% of donations from those with HC or NIR. Pre-donation questioning about surrogates of high-risk behavior (e.g. STDs) would eliminate relatively few additional high-risk donors with HC. 4 (J^ 4 0>

Page  68 S TU.C.47 WESTERN BLOT (WB) RESULT, PLACE OF BIRTH AND p21e ELISA ARE PREDICTORS OF INFECTION BY HTLV-I OR BY HTLV-II AMONG BL9OD DONORS. Rlos. Maria, Ferreira, O.C., Sheridan, K., Uehlinger, J., Hemsptead, P., Del Valle, C., Lightbourne, B. and Bianco, C. New York Blood Center, New York, NY, USA. Objective: Confirmatory assays for HTLV-I/II do not discriminate between viral types and generate a large number of inconclusive test results. The present study searched for demographic and serological correlates of viral type in an attempt to predict infection by HTLV-I or by HTLV-II among blood donors. Methods: Sixty donors repeatedly reactive on ELISA screen and indeterminate (IND) or positive (POS) WB were studied by p21 e ELISA and PCR, and were interviewed about risk factors. PCR was performed on DNA from periphera blood leukocytes using the primer pair SK43-44 which amplifies HTLV-! and HTLV-I1 sequences. Sau 96 I, Sau 3A I and Dde I restriction fragments of the PCR products were analyzed by liquid hybridization with 32 P-labelled SK45. Results and Discussion: 26 (43.3%) donors were typed as HTLV-I, 14 (23.3%) as HTLV-II, 3 (5%) had double infection and 17 (28.3%) were negative (NEG) on PCR. These donors were then classified according to their initial WB pattern. POSITIVE W8 - All 34 donors were POS on p21e; 32 (94.1%) were POS on PCR. Among the 19 donors born in the Caribbean, South America, Middle East or South India, 18 (94.7%) were PCR POS, 17 were typed as HTLV-I, and one had double infection. Among 15 individuals born in the U.S. or Europe, 14 were POS on PCR. Of these, 8 (53,3%) were typed as HTLV-I, 4 (26.6%) as HTLV-II, and 2 (13.3%) had double infection. INDETERMINATE WB - Among 26 donors, 11 (42.3%) were PCR POS. Of these, 10 (90.9%) were typed as HTLV-II; 8 of these were also POS for p21e. One, born in Trinidad, was typed as HTLV-1 and was p21 e POS, While helpful in confirmation, p21e lacked sensitivity, missing 2 (20%) of the donors typed as HTLV-II. These donors had p19, p28 and p36 on WB. All 15 WB IND without p24, p36 or gp46 were also p21e NEG and PCR NEG. Conclusions: Birthplace, WB band pattern and p21e EUSA are correlated with HTLV type. Donors with POS WB born in area of high prevalence of HTLV-I are likely to be infected by HTLV-I (94.7%). Those born in the US may be infected by HTLV-1 (66.6%) or HTLV-II (28.6%). Donors born in the US or Europe, with IND WB showing p19, p28 and p36, or with POS p21e EUSA, are likely to be infected by HTLV-II (100%). A lew donors may be infected by both viruses. NOTES TU.C.48 LIMITNG HE RISK OF SSIN ASSOCIATED AIDS: A DECISI ANALYTIC APPFCM Ieymann, Sally Jodv* Brewer, TP* *Harvard University, Cambiridge,, M USA. bjiytie: In many areas of the world, the risk of transfusion associated AIDS remains significant either because screening is unavailable or because the accuracy of screening is limited by the quality of the test, window period before those infected have antibodies, and human error. In this study, we quantitatively address the difficult dileumsa of when to transfuse. MithlO L This study uses decision analysis to compare the survival outcomes of severely anemic patients who are transfused with those who are not transfused. A transfusion thr Ihold is calculated for how high the risk of dying frm anemia needs to be before a patient should be transfused. Transfusion thresholds are given for rates of HIV cCntamination of blood ranging from 0.0000065, ormpatible with the best available screening, to 0.5 per transfusion, above the risk in the most infected regions of the world. Sensitivity analyses arioud published values were performed for all variables. Results: In countries with low rates of non-AIDS fatal transfusion reactions and rhere 1% of the blood transfused is HIV contaminated, those with a risk of dying from anemia greater than 0.7% should be transfused. In countries with fatal transfusion reaction rates colparable to Zaire and where 5% of the blood is HIV contaminated, those with a risk of dying from anemia of 6% or greater should be transfused. aonlusior; Recrmmendations of when to transfuse should be country specific because of the variable risk of HIV transmission, transmission of other infectious diseases, and severe transfusion reactions. ITis study demnstrates how to make quantitative location specific reconmurdations. These tools can also be used to evaluate whether current policies lead to unneooessary deaths from over or under-transfusion. NOTES TU.C.49 RISK OF TRANSFUSION-TRANSMITTED HIV-1 AND HTLV, TYPES I AND II Donahue James G.* Muioz, A.*; Nelson, K*; Ness, P.; McAllister, H. Yawn, D. Reitz, B.; Lee, H. Cohen, N.. Johns Hopkins University, Baltimore *St. Luke's Episcopal, ***Methodist Hospitals, Houston; Abbott Laboratories, N. Chicago; * Texas A&M, College Station, TX. U.S.A. Objective: To determine the effectiveness of HIV-1 screening of the blood supply and the risk of transfusion-transmitted human T-cell lymphotropic virus (HTLV), types I and II, prior to and after HTLV screening. Methods: Sera were collected from patients before and 6 months after cardiac surgery in a prospective study conducted from 1985 to 1990 in Baltimore and Houston. Postoperative sera were tested by enzyme immunoassay, confirmed by western blot for HIV, and by western blot and radioimmunoassay for HTLV. Preoperative sera corresponding to positive postoperative sera were also tested. To distinguish HTLV-I and II, sera were tested using synthetic peptide-coated beads in an immunoassay. Results: Of 10,036 patients transfused with 96,068 units of blood components, 2 seroconverted to HIV-1. The risk of transfusion-transmitted HIV-1 was 0.0021% per screened unit with a 95% upper bound of 0.0065%. One HIV-positive donor was found for each of the seroconvertors who were transfused in mid-1986 and mid-1989. Of the 5,908 patients who received 50,650 units prior to the start of HTLV screening, two seroconverted to HTLV-I and four seroconverted to HTLV-II. The pre-screening risks (and 95% upper bounds) of transfusion-transmitted HTLV-I and HTLV-II were, respectively, 0.0039% per unit (0.0124%), and 0.0079% per unit (0.0180%). Comparison of seroconvertors with controls matched by number of units transfused indicates transfusion of platelets is associated with HTLV transmission (p=0.11). The risk of transmission was significantly (p=0.03) reduced subsequent to HTLV screening; none of the 4,128 post-screening patients (45,418 units) seroconverted to HTLV-I or II. Conclusions: The risk of transfusion-transmitted HIV-1 is small (2/10' units). The risks of HTLV-I and II transmission (4/105, 8/10s units) were significantly reduced by HTLV screening. Furthermore, the prevalence of HTLV-II may be greater than previously thought. TU.C.50 TRANSFUSION-TRANSMITTED HIV INFECTION: RATE OF PROGRESSION TO AIDS Transfusion Safety Study Group (TSS)* represented by Operskalaki, Eva A.** *Participating institutions in New York, Miami, San Francisco, and Los Angeles, **University of Southern California, Los Angeles, CA, USA. Objective: To evaluate whether persons with transfusion-transmitted HIV infection progress to AIDS rapidly as a result of a large inoculum, underlying disease requiring transfusion, and generally older age. Methods: Since 1986, we have enrolled into long-term follow-up 111 persons infected with HIV after receiving a contaminated component in 1984-85. AIDS incidence by time since infection was analyzed using Kaplan-Meier plots. Results: The table compares overall rate of progression to AIDS among TSS recipients with those of earlier studies. Initial no. Time since transfusion (yrs) Study of recipients 2 3 4 5 Giesecke et al (1988) 48 3% 16% 23% 29% Ward et al (1989) 267 5% 8% 18% 33% Present study 111 6% 10% 16% 16% Age-specific rates at 5 years were: under 45, 10%; 45-69, 13%; and 70+, 32%. Conclusion: Prospective identification of HIV-infected recipients does not show the very rapid HIV progression to AIDS previously reported. (Supported y Contracts no.-"NU-llHB3--~--7, 4-U3, and 97 c^ 0a t^ ~ IC o~ 0; 2:

Page  69 T CTT.51 PROSPECTIVE STUDY OF PRIMARY MEDICAL CARE UTILIZATION BY HIV+ AND HIVINTRAVENOUS DRUG USERS(IVDU) IN A NYC METHADONE TREATMENT PROGRAM(MMTP) Selwyn, Peter A, Budner N, Wasserman W, Arno P. Montefiore Medical Center, Albert Einstein College of Medicine, Bronx, New York, USA. Objective. To study primary care utilization in a cohort of HIV+ and HIV- IVDUs in a prospective study of HIV infection at a NYC MMTP with on-site primary medical care. Methods. We used computerized medical encounter data to analyze the frequency of primary care visits and diagnoses (dx) for HIV+ and HIV- cohort subjects receiving onsite MMTP medical services over a 12-month period from 1/11/89-30/10/90. Results. Of 462 cohort patients (pts) in the MMTP,209(45%) were HIV+,253(55%) HIV-;174 (83%) HIV+ and 167(66%) HIV- pts(p<.01) made 2212 primary care visits.HIV+'s made 1544 (70%) visits(mean 8.9. range 1-45),HIV-'s made 668(30%)(mean 4.0, range 1-29)(p<.01). Leading dx's for HIV+'s(% visits with dx) were: ARC(22.8),AIDS(17.0),asymptomatic HIV+ (13.1),psychiatric dx(5.9),pregnancy(5.7),substance abuse(5.2),pneumonia/bronchitis (5.0),upper respiratory infection(URI)(4.8),fatigue(4.8);for HIV-'s,URI(9.7),pregnancy (7.0),substance abuse(7.0),diabetes(6.3),cirrhosis/hepatitis(6.1),psych dx(6.0),musculoskeletal dx(5.5),asthma(4.8). HIV+ pts were also more likely(p(.05) to have(% pts with dx): TB(10.9 vs. 3.6),dermatitis(12.0 vs. 4.2),neurologic dx(12.0 vs. 6.0), and anemia(12.6 vs. 4.8). Vaginitis/cervicitis was dx'd in 30/91(33%) HIV+ and 21/95(22%) HIV- women(p-.10). Among 174 HIV+'s,95(55%) received AZT and/or PCP prophylaxis; mean visits for pts with (200, 200-500,and >500 CD4 cell/mm3 were 12.2, 9.3, and 5.5. Conclusion. On-site primary care services were frequently used by IVDU's in this MMTP, especially by HIV+'s with advanced disease. Data reflect varied HIV-related dx in HIV+'s, other chronic diseases in HIV-'s, and psychiatric, substance abuse,and women's health care dx in both groups. MMTP's are ideal sites to provide primary medical care to IVDU's. but require a wide range of services in addition to HIV-specific care. NOTES TUC 52 SURVEILLANCE OF CD4 LYMPHOCYTE COUNTS OF HUMAN IMUNODEFICIENC VIRUS *.*.5- INFECTED PERSONS IN SOUTH CAROLINA. Luby, Stephen*; Jones J; Horan J; Ball R *CDC, Atlanta, Georgia USA Objectives: Because CD4 lymphocyte counts decrease with the duration of human immunodeficiency virus (HIV) infection, predict opportunistic infection, and predict progression to AIDS, they help to identify more severely affected and more recently infected persons. Using CD4 counts to enhance surveillance of HIV infected persons in a general population has not been previously reported. Methods: From March 1989 to October 1990, we obtained CD4 lymphocyte counts, measured by flow cytometry, from all reported newly diagnosed HIV infected persons who agreed to testing in state public health clinics. Results: Of 1,828 newly diagnosed HIV infected persons statewide, 727 (40%) were CD4 tested through this health department program. Of these, 107 (15%) had CD4 counts <200, the level below which prophylaxis for pneumocystis pneumonia has been demonstrated to prolong survival, and 338 (46%) had CD4 counts <500, the level below which zidovudine (AZT) has been demonstrated to prolong survival. Persons >24 years old were more likely to have C-D4 counts <200. (relative risk [RR] 9.5, 95% confidence interval [CI]) 3.1, 29.4). The highest CD4 counts (>900), which are associated with more recent HIV infection, were more common in females (RR 1.7, CI 1.2, 2.3) especially among those <24 years old (RR 3.1, CI 1.5, 6.3). Conclusions: In South Carolina almost half of newly reported HIV infected persons who agreed to CD4 testing at the health department would benefit from immediate drug therapy. Within this population, among females, especially those <24 years old, there may be an emerging risk group of increasing significance which requires aggressive and specifically directed HIV prevention efforts. NOTES TU.C.53 EFFECTS OF AN AIDS EDUCATIONAL INTERVENTION IN LOW INCOME ADOLESCENTS IN PUERTO RICO. A PILOT PROGRAM DEVELOPED BY MEDICAL STUDENTS. BERNASCHINA, CLAUDIO*; Zorrilla, C.; Rodriguez, M.; Romaguera, J.; Dueno, A.; Olazagasty, J.; Tome, F.; *JPR, School of Medicine. Knowledge about AIDS was significantly increased after an education intervention. Three hundred eighty four summer job adolescents received a Pre-Test, followed by a lecture on AIDS, by medical students. A post-test was administered two weeks later to 77% of the students. Performance was significantly improved at the posttest in all age groups and both sexes. The number of correct answers improved overall by 33% at the post-test. Differences among sexes were found at the PreTest in all age groups and at the Post-Test in the 16-18 years adolescents. The type of questions that elicited the most dramatic increase in knowledge where those dealing with AIDS misconceptions or how AIDS is not transmitted (casual contact, living near a person with AIDS, etc.). Most students knew already how AIDS is transmitted (needle exchange, sexual contact, pregnancy). This sample of students included 67% females and 33% males. Mean age was 15 years. The proportion of adolescents reporting sexual activity was 12%, mostly males. High risk sexual practices (unprotected vaginal or anal intercourse) were prevalent. Self reported substance use was as follows: Cigarette smokers: 12% males, 6% females; alcohol: 19% males, 12% females; marijuana: 8% males, 1% females; cocaine: 8% males. Results show that the general public is not getting the correct information about how the disease is not transmitted. This information is essential to reduce bias and should be in public compaigns. As a final conclusion we can see that using medical students as health educators is a valuable resource, cnpriallv for reachino the adolescent population. TU.C54 DECREASED LEVELS OF SYRINGE SHARING, LOW HIV RATES FOR SYRINGE EXCHANGE CLIENTS AND OTHER U.K. INJECTORS. Stimson, Gerry V., Dolan K A, Donoghoe M C and Jones S G. Centre for Research on Drugs and Health Behaviour, Charing Cross and Westminster Medical School, London, England. OBJECTIVE: To examine differences in HIV prevalence, incidence and risk behaviour between a cohort of drug injectors (IDUs) attending syringe exchange schemes and a comparison cohort. METHOD: All subjects injected in the 4 weeks before interview in 1989 and were quota-sampled by sex. 86 IDUs recruited form syringe exchange schemes and 121 IDUs not attending but from same area. In 1990, 51 (59%) attenders and 60 (n=50%) nonreinterviewed. No significant difference on demographics and drug use between those reinterviewed and not. Saliva samples were collected and tested for HIV antibodies. RESULTS: Attenders' regular access to syringes decreased (96% to 78%) while nonattenders' access increased (63% to 76%). Both groups reduced syringe sharing; attenders, 35% to 16%; non-attenders, 42% to 22%. Both groups increased condom use but non-attenders shirted to non injecting sexual partners. HIV rates below 6% and no seroconversion. CONCLUSIONS: Syringe sharing levels decreased in both groups. Levels of syringe sharing among syringe exchange clients are the lowest recorded in the UK. Injectors attending schemes have lower risk of infection. HIV rates remain low among injectors. Increased syringe availability is associated with risk reduction. C\

Page  70 S TU.C.55 MATHEMATICAL MODELS FOR ASSESSING THE IMPACT OF PREVENTION PROGRAMMES ON 0 C55 HIV INFECTION ArcA Massimo; Spadea T.; Perucci C.A.; Michelozzi P. Regional Epidemiology Unit, Rome, Italy OBJECTIVE: To evaluate the impact of a family of independent prevention programmes on HIV transmission dynamics in Latium, Italy, through model simulations. METHODS: Using a multi-population model of the HIV and AIDS epidemic in Latium, parameter values were modified at different times (5th, 10th, 15th year) to simulate behavioural changes; outcomes were assessed separately for IVDIIs and General Population (GP) at a summarizing end point (40th year). RESULTS: Halving the rates of new needle-sharing partners among IVDUs and a 50% reduction of the probability of transmission per single sharing are the two most effective programmes at the early implementation, preventing more than 7000 cumulative NIDS cases among both IVDUs and GP. The reduction of the probability of transmission per sexual contact, simulating condom use, in all relations involving IVDUs and in "promiscuous" intercourse are mostly effective in the GP, with a 40% reduction of HIV prevalence, even with delayed implementation. On the contrary, lowering the rates of new sexual partners, when associated to a lengthening of partnerships, causes a rise of HIV infection and AIDS. Finally, a longer incubation time, simulating effective therapy, yields a delayed but expanded AIDS epidemic, proving that treatment must not be considered as an alternative to prevention. CONCLUSIONS: These results emphasize the need for multiple-strategy prevention programmes. Although mathematical modelling relies on simplified assumptions, it may, given the well known difficulties of unbiased study designs for assessing prevention effectiveness, help and support programme planning and implementing. TU.C.56 STUDY To REVEW EFFECTVENESS AND COVERAGE OF CURRENT SE ORK INTERVENTIONS IN DEVELOPING COUNTRIES Fereniclc Nina'; Alexander, Priscilla*; Slutkin, Gary*; Lamptey, Peter** * World Health Organization, Global Programme on AIDS, IDS ** Director AIDSTECH, FHI, Durham, North Carolina, USA Objeeive: To determine what has been demonstrated in ongoing sex-work peer education interventions in developing countries and make recommendations for programmes and research. Method: Amatrix was developed to categorize governmental, NGO, university, multi- and bilateral organization projects in the area of sex-work. Projects were reviewed for their coverage, effectiveness and cost, using existing and new data, with 45 projects identified as of end January 1991. Reults: Coverage varies across sites. Individual projects reached anywhere from 25 to 7,768 prostitutes, and up to 19,000 clients, with the proportion reached ranging from less than 5% to up to 85% of the estimated target population. Factors affecting coverage include size, density and distribution of prostitution sites, migration, law-enforcement practices, project organization and attitudes of project staff. Coverage is also affected by the tendency of projects to center on some forms of prostitution (eg. bars, street, females) leaving out significant segments of the sex-work population (eg. massage and escort services, male prostitutes, clients). Almost every project has found an increase in condom use, with up to 81% of the target audience reporting use of condoms some or all of the time. Conclusions: While it is too early to reach final conclusions, initial review indicates that thus far, most projects are small scale. Increases in reported condom use are feasible, but improved methodologies and validation of measurement need to be pursued. Some obstacles to consistent condom use seem to include client resistance, lack of support by managers and inadequate access to condoms. Some projects are already trying to expand both geographically and in terms of coverage to include actual and potential clients. However, expansion of projects from a pilot to multiple sites and sustainability beyond an initial, limited duration continue to pose the main challenges to projects. WHO is reviewing what contributes to effectiveness and coverage, and providing assistance for intervention studies which address these challenges. Ih t^ NOTES NOTES

Page  71 TU.D.57 SOCIO-ECONOMIC IMPACT OF AIDS ON FOOD PRODUCTION IN EAST AFRICA Norse, David. Food and Agriculture Organization of the U.N., Rome, Italy. Objective: Analysis of the impact of AIDS related mortality on household labour supply and on food and non-food production. Methods: First, countries were selected for case studies (Malawi, Rwanda and Tanzania) using 3 criteria: availability and quality of data on farm household labour use; current existence of seasonal labour bottlenecks in agriculture; potentially high rural HIV incidence. Secondly, farm labour use models were developed. Thirdly, projections were made of HIV/AIDS by gender and age cohort using WHO's EPIMOD. Finally, the potential impact of AIDS on household labour supply and food production was assessed. Results: By the year 2000 up to 25% of farm households could be affected depending on the country or region. Some farming systems are particularly sensitive to labour shortages and certain labour intensive crops, Like tobacco and cotton, may decline with a consequent reduction in foreign exchange earnings. Conclusions: Farm incomes and food consumption levels will be negatively affected. Female-headed households are particularly vulnerable. Policies need to give greater support to labour saving actions on farms, and for water and fuelwood collection. NOTES TU.D.58 USE OF EPIDEMIOLOGICAL MODELS FOR SOCIOECONOMIC AIDS IMPACT SCENARIOS ILLUSTRATED BY ANALYSIS OF DATA ON THE NETHERLANDS Jager, Johannes C *; Poos, M.J.J.C.*; Fostma, M.J.*; De Haan, B.J.*; Heisterkamp, S,H,*; Reinking, D.P.**; Van Den Boom, F.M.L.G.** *National Institute of Public Health and Environmental Protection, Bilthoven, The Netherlands,**Netherlands Institute of Mental Health, Utrecht, The Netherlands. Objective: To construct scenarios on the socioeconomic impact of AIDS. Methods: Mathematical models at different levels of sophistication are used to assess the impact of AIDS in The Netherlands. The methods include (i) statistical extrapolation of AIDS incidence after adjustment for reporting delay, (ii) assessment of HIV prevalence by backcalculation and (iii) application of process models. Results: A range of scenarios have been built limited by a pessimistic (AIDS epidemic evolves with growth rate estimated over the period 1988-1990) and an optimistic (no new HIV infections occur after 1 January 1989) course. By 1 January 1991 the number of AIDS cases reported amounts to 1531 and to 1653 after adjustment for reporting delay. The estimated cumulative number of HIV infections by 1 January 1989 is 7000 cases. For the period 1991-1995 projected AIDS cases amount to 2329-3794 and hospital costs to, $ 83-112 million (assumptions: hospital costs $ 19,500 per person-year; 5% discount rate). In 1995 projected potential years of life lost per 1000 population is 1.1-1.9 and hospital bed need is 140-217 halfway the year (assumptions: 51 inpatient days per person-year; 90% occupancy rate). Intermediate explorative scenarios based on process modelling and relevant assumptions will be specified. Discussion and Conclusion: The selected set of methods represents a standardized procedure which extends epidemiological monitoring of the epidemic to its socioeconomic consequences to support public health strategies. NOTES C) t-s 0 Gr TU.D.59 INCIDENCE AND SOCIOECONOMIC CONSEQUENCES OF AIDS ORPHANS IN 422 FAMILIES WITH HIV-1(+) MOTHERS IN KINSHASA, ZAIRE Kamenga, M*; Muniaka, Nkusu*; Batter, V*; DaSilva, M*; Ryder RW**/***; *Projet SIDA, Kinshasa Zaire; **CDC, Atlanta, GA; ***Mt. Sinai School of Medicine, NY, NY. OBJECTIVES: To calculate the yearly incidence, health and social consequences of becoming an AIDS orphan in Kinshasa (population - 4 million; HIV-1 seroprevalence-6%). METHODS: From 1986 to 1990 422 newborn infants with HIV(+) mothers (case children) and their siblings who became AIDS orphans were compared with similarly aged children (control I) and their siblings of HIV(+) mothers who remained alive and age-matched children (control II) and their siblings of living HIV(-) mothers. RESULTS: AIDS orphan incidence for newborn children with HIV(+) mothers was 2.8/100 child years of follow-up (F/U). Each year 515.3 children per million Kinshasa population became AIDS orphans. Case(n=26) Control I Control II Morbidity in uninfected child. during F/U BMD AMD (n-26) (N-26) Acute Diarrhea*/Persistent Diar.* 194/9 229/0 192/22 195/4 Acute Fever*/Pulmonary Infection* 300/9 326/23 336/26 413/5 Purulent Otitis Media*/Yrs. Observation 0/22 40/18 12/60 8/76 BMD-Before Maternal Dean; AMD-After Maternal Death; *No cases/100 Yrs. Observation Socioeconomic status of Case parents (BMD),' Case guardians (AMD) and Control I and II parents was similar. Child care (clothes/shoes, school attendance, purchase of school books, nutritional status, morbidity) was similar during F/U in all 3 groups. CONCLUSION: Because all AIDS orphans (cases and siblings) was adopted by concerned extended families, no differences in the quality of child care and morbidity could be in detected in the 3 groups of children during the 3 years of F/U. Previous studies may have over-estimated the AIDS orphan incidence. TU.D.60 EFFECT OF PERINATAL HIV INFECTION ON INFANT MORTALITY:POLICY IMPLICATIONS Wise Paul H. Pediatric,Adolescent,and Maternal AIDS Branch,NICHD,NIH,Bethesda,MD,USA Objective: In order to examine the interaction of HIV infection and broader issues of maternal and child health, this study estimated the impact of neonatal HIV-seroprevalence on the infant mortality rate (IMR), an important measure of general perinatal and child health. Methods: Regression models were constructed to estimate the impact of neonatal HIV seroprevalence on the IMR using empirical data at four levels: the risk of newborn infection given a positive test at birth; the risk of infant death given actual infection; the influence of seroprevalence levels on the provision of services to noninfected childbearing women and infants; and the risk of infant death given altered levels of services to the noninfected. Results: The models suggest two categories of effect: direct and indirect. At low seroprevalence, direct effects due to HIV-related illness were 0.8 deaths/ 1% increase in HIV seroprevalence. However, at approximately 3.5% prevalence, indirect effects on noninfected populations predominate, increasing the slope of effect to 1.9 deaths/ 1% increase in HIV seroprevalence. Conclusions: As seroprevalence rises, the impact of perinatal HIV infection on the IMR will increasingly be due to indirect effects due to the competitive reduction of traditional, yet effective, maternal and child health services on noninfected populations. The policy implications of perinatal HIV infection must be examined within the broader context of maternal and child health. F21 G\ C3

Page  72 S TU.D.61 Quality of Life Measurement For Persons with HIV Infection, Robert M. Kaplan, Ph.D., John P. Anderson, Ph.D., Thomas L. Patterson. Ph.D.. Allen McCutchan, M.D., James D. Weinrich, Ph.D., Robert K. Heaton, Ph.D., J. Hampton Atkinson, M.D., James Chandler, M.D., Igor Grant, M.D., the HNRC Group, Naval Hospital, San Diego, and University of California, San Diego, La Jolla, CA, USA. Objective This paper offers a method for capturing the diverse outcomes of HIV disease. The Quality of Well-being Scale (QWB) combines preference-weighted measures of symptoms and functioning to provide a numerical point-in-time expression of well-being that ranges from zero (0) for-death to onc(l.0) for asymptomatic optimal functioning -- i.e., higher scores represent better health. In the General Health Policy Model, QWB inputsare integrated with terms for the number of people affected and the duration of time affected to produce the output measure, which is known as the "well-year.'" Mlethods: This study reports data for 86 male adults in three categories: CDC IV, CDC IIII, and gay male control. The QWB as well as a variety of neuropsychological and biochemical measures were administered to all participants. Resullts The CDC IV group scored about.74 on the 0 to 1.0 scale. When QWB was broken down by HIV grouping, the CDC IV group significantly differed from the CDC I/IIl groups and the control groups. The differences between the Class IV and Class Il/Il1 is about.07 units of well being, suggesting that individuals lose 7/100 equivalents of well years of life for each year they are in the AIDS category in comparison to the asymtomatic groups. In comparison to the controls, this would equal a one year of life loss for each 14 infected individuals. The QWB was shown to be significantly associated with CD4+ lymphocytes (p<.01), clinician ratings neuropsychological impairment (p<.04), neurologists ratings of dysfunction (p.03) and several psychiatric variables including POMS scores for vigor (p<.001) and depression (p<001). Multivariate models arc used to adjust for covariation between predictors of QWB. Conclusion: We conclude that the QWB is a significant correlate of biological, neuropsychological, and psychiatric outcomes for male HIV infected patients. In related studies we have shown that the QWB is a strong significant prospective predictor of mortality. Future work will use mathematical modeling techniques to translate observed QWB differences into quality adjusted life units for use in policy analysis. NOTES TU.D.62 EARLY INTERVENTION IN PEDIATRIC HIV DISEASE: THE ECONOMIC IMPACT Amo Peter S,' Green J,1" Mofenson L,'** Bonuck K,* Futterman D,* Shenson D* *Montefiore Medical Center, **NYU Medical Center, *NICHD, NIH The components of an early intervention strategy for treating infants and young children with HIV infection are rapidly evolving particularly with the growing clinical acceptance of PCP prophylaxis and FDA licensure of zidovudine among young children with evidence of immunosuppression. We estimated the economic impact of a national program to provide monitoring and treatment services to children with early HIV infection, The model's parameters include: the expected number of HIV seropositive babies born each year, the percentage in treatment and monitoring each year and the costs associated with outpatient services and pharmaceuticals. The costs of identifying scropositives are excluded. Primary medical care and pharmaceuticals represents 40% and 60% of all treatment costs, respectively. Delivering early intervention services to approximately 5,900 seropositive babies born each year are estimated to cost $12 million for the first year and $20 million during the second year of the program. This represents an annual cost of approximately $2,700 per child - far less than the cost of one pediatric hospitalization. NOTES O Co Q) C) C 1^3

Page  73 TU.A.63 HIGHLY POTENT AND SELECTIVE INHIBITION OF HIV-1 REPLICATION BY A NOVEL SERIES OF 6-SUBSTITOTED ACYCLOURIDINE DERIVATIVES Baba Masanori*; Shigeta, S.*; De Clercq, E.**; Tanaka, H.***; Miyasaka, T.***; Ubasawa, M.****; Umezu, K.****; Walker, R.T.***** *Fukushima Medical College, Fukushima, Japan, **Rega Institute, K.U. Leuven, Leuven, Belgium, ***Showa University, Tokyo, Japan, **Mitsubishi Kasei Corporation, Yokohama, Japan, *****University of Birmingham, Birmingham, UK. Objective: Identification and development of potent and selective anti-HIV agents with novel chemical structure and a new mode of action. Methods, We designed and examined new 6-substituted acyclouridine derivatives for their inhibitory effects on the replication of HIV in a variety of cell systems including MT-4 cells and peripheral blood lymphocytes (PBL). The mode of inhibition was also investigated. Results: Among the derivatives, 5-ethyl-l-ethoxymethyl-6-(3,5-dimethylbenzyl)uracil (E-EBU-dM) proved to be an exquisitely potent and selective inhibitor of HIV-1. Its 50Z antivirally effective concentrations for HIV-1 in HT-4 cells and PBL were 2.2 and 0.45 nM, respectively. These concentrations were more than 100,000 times lower than the 50Z cytotoxic concentrations. E-EBU-dM was equally inhibitory to several clinical isolates of HIV-1, including an AZT-resistant variant. However, E-EBU-dM did not inhibit HIV-2 replication. The compound acts specifically on HIV-1 reverse transcriptase according to a mechanism that is different from that of dideoxynucleosides. Conclusion: E-EBU-dM is a promising candidate for anti-AIDS chemotherapy. NOTES TU.A.64 HOVEL SULFONIC ACID POLYERSS SA NEW CLASS OF POTENT AND HIGHLY SELECTIVE ANTI-HIV-1 AGENTS Mohan, Prem*; Baba, M.** *University of Illinois at Chicago, Chicago, Illinois, USA, **Fukushime Medical College, Fukushima, Japan. Objective: Evaluation of novel sulfonic acid polymers for anti-HIV-3 activity in a variety of assays. Methods: Anti-HIV-1 activity was determined by measuring inhibition of (1) HIV-1 induced cytopathogenicity in both the lab strain (HTLV-IIIg) and a clinical isolate (HE strain). (2) HIV-1 induced giant cell formation. Results: In all assays, potent and highly selective anti-HIV-1 activity was observed in all the tested polymers at doses that were non-toxic to the host cells. In the cytopathogenesis assay involving the lab strain, poly(2-acrylamido-2-methyl-l-propanesulfonic acid) (PAMPS) and poly (vinylsulfonic acid) (PVS) emerged as the most active compounds and displayed EC50 values of 0.28 and 0.31 Bg/ml, and in vitro therapeutic indices of >1785 and >1612 respectively. In this assay, PAMPS is almost as active as dextran sulfate (MW=5000). Sensitivity in the same assay involving the clinical isolate of HIV-1 is less for all compounds. In the giant cell formation assay, PAMPS (IC50=0.38 g/ml) demonstrated the most activity, exhibiting a potency superior to that of dextran sulfate (MW=5000) (IC50=5.2 ug/ml). In this assay, poly(anetholesulfonic acid) (PAS) (IC50o2.0 g/ml) was also more active than dextran sulfate (MW=5000). Conclusion: These sulfonic acid polymers are a new class of potent and selective anti-HIV-1 agents that warrant further investigations. NOTES Con C 'fcq T A.65 POTENT AND SELECTIVE ANTT-HTV ACTTVTTY OF 9-(2-PHOSPHONYL4METHOXYETHYL) ADENINE (PMEA) IN HUMAN MACROPHAGES Ercoii Lucia*, Perno CF*, Balzarini J**, Parisi SG*, Milanese G*, Calio' R*, Rocchi G*. *IT University of Rome, Italy, **Rega Inst. for Medical Research, Leuven, Belgium. Objective: Examination of the antiretroviral activity of 9-(2-phosphonyl-methoxyethyl) adenine (PMEA) in monocyte/macrophages (M/M) under various experimental conditions. Methods: Blood-derived M/M and T4-lymphocytic cell lines were challenged with a monocytotropic strain of HIV in the presence of PMEA. Antiviral effect was assessed by p24 antigen production, syncytia formation and cell cocultivation. Results: Continuous exposure of 0.1 uM PMEA afforded >99% inhibition of p24 antigen and syncytia in M/M (ED50 between 0.01 and 0.1 uM, 100 fold lower than that observed in MT4 and CEM T4-lymphocytic cell lines). 70% inhibition of M/M-to-lymphocyte viral transfer was achieved by 0.2 uM PMEA. No toxicity was observed up to 100 uM. Antiviral activity was markedly reduced when PMEA was removed after viral challenge (EC50 18 uM). similarly, PMEA activity is 10-100 fold reduced by cytokines enhancer of viral replication, with an EC50 between 0.1-1 uM with granulocyte-macrophage colony stimulating factor, and between 0.01 and 0.1 uM with macrophage colony stimulating 'factor. However, even under these experimental conditions, the therapeutic index of PMEA was always 1,000. Conclusions: The high efficacy of PMEA against HIV in lymphocytes and M/M,together with the pronounced antiviral activity described "in vivo", suggest that PMEA is a good candidate drug to initial clinical trials in HIV-infected patients. 1J)0...... TU.A.66 PXIDATIVE STRESS: A MARKER OF PROGRESSION AND A THERAPEUTIC TARGET IN,-IV INFECTION Revillard, Jean-Pierre - Lab. Immunol.- INSERM U80 UCBL Hop E. Herriot, Lyon, France xidative stress is produced by reactive oxygen species (ROS) during infection or inflammation. The ajor antioxidant defence mechanisms are alpha-tocopherol in cell membranes and glutathione in he intracellular compartment. n HIV-infected patients increased production of ROS and/or defective antioxidant mechanisms were emonstrated by elevated levels of malondialdehyde (MDA) a marker of lipid peroxidation. Serum IDA levels were 2.50 ~ 0.20 pmol/L in 30 healthy controls, 2.93 ~ 0.32 pmol/L in 26 CDC stage II patients and 3 32 ~ 0.44 pmol/L in 37 CDC stage IV patients (difference stage II/controls: p < 0.001; stages II/IV: p < 0.001). Increased lipid peroxidation was independently reported in another group of 30 patients at different stages of HIV infection. Other reported evidences of oxidative stress in HIV infection include decreased GSH level in blood mononuclear cells, serum and pulmonary. lining fluid, high serum glutamate and low methionine levels. The occurence of toxic ROS is likely to be associated with the detrimental role of TNFa in the progression of HIV infection. Indeed TNFa induces the production of ROS in target cllls whereas the synthesis of TNFa by monocyte/macrophages is stimulated by oxidants and inhibited by antioxidants. Furthermore oxidative stress can trigger NF-KB translocation and activation of HIV-LTR transcription, a process which can be inhibited by antioxidants in vitro. These data suggest that HIV activation and disease acceleration could be mediated through increased production of ROS. Such a mechanism could thus be slown dawn by an anti-oxydant therapy combined with anti-viral drugs. Diethydithiocarbamate (Ditiocarb Sodium), a sulfur containing compound with both direct and )ndirect antioxidant properties, was shown to prevent immunodeficiency when administered in vivo in animal models of ROS toxicity. In three randomized placebo-controlled clinical trials, ditiocarb was Shown to delay the progression of HIV infection. It is suggested that this therapeutical effect is 4ccounted for by the in vivo antioxidant and immunoprotective activities of this compound. a,b\ hj I

Page  74 TU.A.67 HIV-1 SPECIFIC PYRIDINONE RT INHIBITORS: I. PRECLINICAL BIOLOGICAL CHARACTERIZATION OF TWO INVESTIGATIONAL NEW DRUGS. Goldman, Mark E., O'Brien, J.A., Ruffing, T.L., Stern, A.M., Gaul, S.L., Saari, W.S., Wai, J.S., Hoffman, J., Rooney, C.S., Quintero, J.C., Schleif, W.A., Emini, E.A. and Nunberg, J.H. Merck Sharp & Dohme Research Laboratories, West Point, PA 19486, U.S.A. Obiective: To characterize, pharmacologically and enzymologically, a structurally-novel class of non-nucleoside HIV- RT inhibitors currently undergoing clinical evaluation. Methods: Analogs that possessed potent and selective HIV-1 RT inhibitory activity were tested for antiviral properties in cell culture using multiple HIV-1 isolates and cell types. Results: The selected compounds, L-697,639 and L-697,661, inhibited HIV-1 RT activity in a concentration-dependent manner with IC50 values in the 20-600 nM range, depending upon the template-primer substrate. At 300 /M, however, these compounds did not inhibit ten other enzyme activities. These molecules were reversible dead-end inhibitors which gave mixed-type inhibition patterns with respect to dGTP and TTP as well as to rC-dG and rA-dT. These results are consistent with the formation of enzymetemplate primer-product-inhibitor complexes. Using multiple laboratory and clinical isolates of HIV-1 as well as several human T-lymphoid cell lines and human peripheral blood leucocytes, L-697,639 and L-697,661 inhibited by at least 95% the spread of HIV-1 infection at concentrations of 25-200 nM. Conclusion: These results provide the basis for the ongoing clinical evaluation of L-697,639 and L-697,661 as antiviral agents for the treatment of HIV-1 infection alone or in combination with nucleoside analogs. NOTES TU.A.68 SYNTHESIS AND ANTI-HIV ACTIVITY OF A NEW CLASS OF NUCLEOSIDE ANALOGUES. Verheyden, Julien P.H. Chen, M.S.; Chu, N.; Maag, H.; McRoberts, M-J.; Pettibone, M.; Prisbe, E.J. and Wu, J.C. Syntex Research, Palo Alto, CA, USA. Objective: To investigate the anti-HIV activity of a new class of nucleoside analogues having 4'-substituent. Method: Iodine azide addition to a series of 4',5'-unsaturated nucleosides provides the corresponding 4'-azido-5'-deoxy-5'-iodonucleosides. Displacement of the 5'-iodo function followed by complete deprotection gives the desired 4'-azidonucleosides which were evaluated in a variety of HIV screens. Results: Among this series of 4'-azidonucleosides, the thymidine analogue (ADRT) is a potent and selective inhibitor of HIV replication including that of strains resistant to AZT. In vitro studies comparing the effect of ADRT and AZT on three HIV-1 isolates in four cell types (H9, A301, PBL and MT-2) show that both compounds have comparable inhibitory effect. However, ADRT appears to be 25-40 times more effective than AZT in MT-2 cells infected with two HIV strains resistant to AZT. ADRT is efficiently converted to the corresponding triphosphate, which is a potent inhibitor of HIV-1 reverse transcriptase yet is a poor inhibitor of DNA polymerase a and 3. No significant accumulation of ADRT monophosphate is observed in cells treated with labeled ADRT. Plasma levels and absorption of ADRT and AZT were compared in monkeys. Higher levels of the drug in plasma were observed in animals given 20 mg/kg of ADRT than in those given the same amount of AZT. Bioavailability of ADRT is superior to that of AZT and plasma half-lives are longer following oral or iv administration. The major metabolite of AZT in humans is a glucuronide conjugate. In monkeys, only trace amounts of the glucuronic acid conjugates of ADRT are observed, ADRT being excreted mostly unmetabolized in the urine. This suggests that oral administration of ADRT in humans may result in higher and more persistent levels of circulating drug as compared to AZT. Conclusion: ADRT, a novel nucleoside, has potential advantages for the treatment of AIDS. NOTES NI NI C5 TU.A.69 EVIDENCE FOR NEGATIVE REGULATION OF HIV EXPRESSION IN HUMAN GLIAL CELL LINES PERSISTENTLY INFECTED WITH HIV-1 Kleinschmidt, A.*, Ludvigsen, A.~, Neumann, M.~, Erfle, V~ and Brack-Werner. RI. *Med. Poliklinik d. Universitit Minchen D-8000 Miinchen; ~GSF Abt.Molekulare Zellpathologie, D-8042 Neuherberg, FRG; Objective. Human glioma cell lines persistently infected with HIV-1 show very low production of infectious virus. The nature of virus restriction in these cells was studied by analyzing expression of viral antigens and molecular characterization of the integrated provirus. The activity of various HIV-LTR constructs was determined as an indicator for LTR-dependent negative regulation. Methods. Expression of viral antigens was examined by immunoperoxidase staining (IPS) on single-cell level and by Western blotting analysis. Southern blotting analysis was carried out to characterize the integrated provirus. HIV-LTR activity was determined in transient transfection assays utilizing an HIVBrnLTR-CAT construct and a ANRE-HIVBru-LTR-CAT construct in which the NRE (negative regulatory element; -453 to -156) had been deleted. Results and conclusion. IPS of established human glial cell lines persistently infected with HIV-1 revealed marked synthesis of the viral nef protein. Expression of nef is concomitant with expression of viral structural proteins on a single-cell level. Western blot analysis disclosed much higher levels of nef in the HIV-infected glial cell line TH 4-7-5 containing a single integrated provirus than in HIV-1 producing Tlymphoma cells with multiple integration sites. The NRE of the HIV-LTR has been implicated in negative regulation of virus expression and hence may contribute to restriction of virus production in persistently infected glial cells. Deletion of the NRE-region of HIVBRU LTR led to an average 15-fold increase in HIV-LTR activity in persistently infected, nef-producing glial cell line TH 4-7-5. This effect was about 5-10 times higher than observed in the parental uninfected glial cells (85 HG 66) as well as in productively infected human adherent fibroblastoid cells (LC5/HIV) showing no detectable expression of nef These results point to a prominent negative influence of the NREregion on HIV-LTR activity in HIV-infected glial cells which is possibly mediated by the nefprotein. TU.A.70 FUNCTIONALSTUDIES OFTV.1 NEFPROTEIN. S. Venkatesan: Maitra, R.; Nebreda, A.; Dickie, P4 Ahmad, N.; Holland, S. M.; Santos, E. Laboratory of Molecular Microbiology, National Institute of Allergy and Infectious Diseases. Bethesda. MD. USA Objectives: 1) The difficulty of some laboratories to reproduce the findings that HIV-. NEF protein is a negative regulator of HIV1 LTR led us to re-examine the role of NEF in HIV-1 provirus expression and HIV-1 LTR transcription; 2) to ascertain whether NEF from various HIV isolates possess GTP binding and GTPase activities and 3) to investigate the pathogenic role(s) of HIV-1 Nef using transgenic mouse models. Methods: 1) HeLa and Jurkat cell lines carrying the nef gene linked to the CMV promoter or the HIV-l LTR were isolated by coselection for neomycin resistance and single cell isolates were selected for NEF expression. Provirus expression and HIV-1 LTR linked transcription were assesed in the NEF+ cell lines 2) NEF proteins of three different HIY-i strains were expressed in E. colt using a thermo-inducible expression system and purified for studying their biochemical and biological properties. 3) Three transgenic founder mice bearing integrated copies of an HIV-I LTR linked NEF expression plasmid were generated by pronuclear injection of fertilized embryos. Results: Transcriptional repression of the HIV-1 LTR was directly correlated with NEF expression under transient expression conditions. NEF+ cell lines were five-fold less efficient than NEF- cell lines in transient proviral expression. TAT activated LTR transcription from an HIV-1 LTR linked CAT expression vector was repressed ten fold in the NEF+ HeLa and Jurkat cell lines. When infected with HIV-1. NEF expressing T lymphoid cell lines showed moderate delays in onset and peak of virus production. However, none of these cell lines completely arrested virus replication. NEF from three different HIV- strains was expressed in E. coli and in each case, the bacterially expressed NEF proteins lacked measurable GTP binding or GTPase activities. NEF proteins with a threonine at position 15 had measurable autophosphorylation activity, using any purine ribo-nucleoside triphosphate other than GTP as substrate. NEF also failed to induce oncogenic transformation in NIH 3T3 cells under conditions that led to oncogenic foci using ras oncogenes. Both the NEF transgenic founders and the Fl and F2 progeny exhibited distinctive skin lesions varying from squamous papillomas to severe ulcerations characterized by acanthosis, hyperkeratosis and extraordinary squamous cell proliferation with lymphocytic infiltration of the dernnis. NEF RNA and protein expression were observed only in the diseased areas of the skin. NEF protein staining was confined to the basal cell layer of the epithelum. Discussion and Conclusions: 1) Under the defined conditions of transient expression, HIV.1 NEF has a consistently reproducible repressive effect on the HIV-1 LTR. 2) Although NEF is membrane associated, it is an unlikely candidate for GTPase and GTP binding and does not induce oncogenic transformation. 3) NEF expression in the skin of transgenic mice is associated with proliferative lesions of the epidermis. a ricS C> c> C>

Page  75 TU.A.71 A SPECIFIC INHIBITOR OF CYSTEINE PROTEASES IMPAIRS A VIF DEPENDENT MODIFICATION OF HIV1 ENV PROTEIN Guy, Bruno*; Geist, Michel*; Dott, Karin*; Spehner, Daniele**; Kieny, Marie Paule*; Lecocq, Jcan-Pierre* *Transgtne S.A., Strasbourg Cedex, France, **U74 INSERM, Strasbourg, France. Objective: To determine the role and the biological activity of HIV1 VIF protein. Methods: Vif has been expressed in. coil or by in vitro translation, and in eukaryotic cells using a recombinant vaccinia virus. In addition, targetted point mutations have been introduced in the vif gene, and the native and mutated Vif have been coexpressed with other HIV proteins, particularly env. Results: Vif regulates env processing and/or conformation through a modification of gp41 C-terminus. E-64, a specific inhibitor of cysteine proteases impairs Vif activity on env, as does the mutation of Cysll4 of Vif. Furthermore this amino acid interacts directly with E-64. Discussion and Conclusions: These results indicate that a proteolytic activity is associated with vif, regulating env processing. Interaction of Vif with E-64 as well as structural homologies between Vif and the thiolproteases family suggest that Vif might be a protease. NOTES TU.A.72 CHARACTERIZATION OF HIV SEQUENCES WHICH INHIBIT THE SYNTHESIS OF VIRAL STRUCTURAL PROTEINS Maldpelli. Frank; Martin, M. A.; Strebel, K. Laboratory of Molecular Microbiology, NIAID, NIH, Bethesda, MD, USA. Objective: To identify and characterize cis-acting, inhibitory sequences which supress expression of the HIV-1 gag/pol gene. Methods: Transient expression assays using a novel, feedback-stimulated tat reporter plasmid and the pCMV-CAT reporter plasmid were used to identify sequences within HIV1 gag/pol gene containing inhibitory activity. Quantitative RNA determinations were performed to determine the structure and the intracellular distribution of RNA expressed from reporter plasmids and recombinant derivatives containing the inhibitory sequences. Results: Two regions of the gag/pol gene, 1300-1900 nt in length, were identified which each inhibited reporter gene expression 10-20 fold; subsequences within each of these regions were identified with inhibitory activity. The strongest inhibitory effects were observed only when sequences were present in the foward orientation, suggesting a requirement for a specific RNA secondary structure. The suppressive effect could be eliminated by flanking the sequence between splice donor and acceptor sites, implying that splicing may remove the inhibitory sequence, and permit reporter gene expression. The presence of inhibitory sequences in ci did not alter the relative levels of reporter gene RNA, indicating that the inhibition occurred on a post-transcriptional level. Conclusion: The HIV-1 gag/pol gene contains several elements that inhibit gene expression on a post-transcriptional level. NOTES.A.73 PHENOTYPIC VARIATION IN HIV-I NEF DEMONSTRATED WITH A PANEL OF TU.A.73 EPITOPE SPECIFIC MONOCLONAL ANTIBODIES. Vladimir Ovod*, Anssi Lagerstedt*, Annamari Ranki**, Frank Gombert***, Renate Schaude***, Marja Tahtinen*, Giinther Jung*** and Kai Krohn*. University of Tampere, Tampere, Finland, "University of Helsinki, Helsinki, Finland, **University of Tibingen, Tiibingen, Germany. Background and Objectives: The function of HIV and SIV nef protein is unknown and conflicting results have een obtained. We wanted to study the phenotypic forms and cellular localization of HIV-I nef with a panel of monoclonal antibodies (MAB). ethods: Monoclonal antibodies were produced in BALB/c mice immunized with recombinant nef from HIV/BRU. Epitope specificity was defined with the use of synthetic peptides. Cell lines were infected with 8 ifferent HIV isolates or cloned virus. The size of nef proteins expressed was studied by western blotting and elular localization with immunohistochemistry. esults: Ten MAB clones recognized seven different epitopes, located at the N- and C- terminal parts of nef. ost MABs cross-reacted with several HIV-I isolates but strictly type-specific clones were also obtained. In cutely infected T-cell lines, nef localization was originally cytoplasmic, but later appeared in nucleus, too. n the cytoplasm, nef was most prominently located in the Golgi complex and within the nuclear membrane, orresponding to the localization of ENV glycoproteins. More diffuse localization of nef was sometimes bserved also. Western blotting revealed two species of nef (p24, p27) in HIV-I BRU and IIIB isolates. In all solates, full length nef was produced, even when the nucleotide sequence predicted truncated nef protein:BHX2, MN). Nef was found in purified virus preparations. Conclusions: Several species of nef protein may be produced, differing in length and posttranslational nodification. Some forms of nef, entering nucleus, may have direct regulatory effect on DNA/RNA level. these findings could explain the dual biological effects attributed to nef. TUA.74 MECHANISM OF INCORPORATION OF THE HIV-I VPR PRODUCT INTO VIRUS PARTICLES Cohen, Eric A.*; Haseltine, W.A.** tUniversit6 de Montr6al, Montreal, Canada, **Dana Farber Cancer Institute, Boston, MA, USA. The vpr product of HIV-1 has been shown to accelerate the replication and cytopathic effect of HIV-1 in CD4+ T cells. The vpr protein is the first regulatory protein of HIV-1 to be found associated with virus particles. The mechanism by which vpr becomes associated with the virus particles is not known. Other capsid proteins are made as gag or ag-pol precursor and assembled as a unit into the budding particle. The vpr product is not known to be part of such a larger precursor and must have an independent means of association with the nascent particles. To examine whether the incorporation of vpr into virus particles could occur in trans, Cos-7 cells were cotransfected with a vpr provirus and a vprexpressor plasid. Forty-eight hours post-transfection, the cells were labeled. Immunoprecipitation of the cell-free supernatant with a vpr antiserum revealed the presence of the vpr 15 kd protein indicating that vpr is packaged in trans in the virus particles. The nature of the interaction that allows the incorporation of vpr into the viral particles as well as the critical region of the vpr protein will be discussed. 4 I o!

Page  76 \ TU.A.75 MUTIPLE PROTEINS FROM HIV-1 CAN BE SIMULTANEOUSLY RECOGNIZED BY CYTOTOXIC T LYMPHOCYTES FROM SEROPOSITIVE DONORS. Lamilamedi-CherradiSalma; Culmann,B.; Gomard,E.; Levy, J-P. INSERM U 152, ICGM, Hopital COCHIN, PARIS, FRANCE Objective: In most virus systems, anti-viral cytotoxic T lymphocytes (CTL) issued from one donor recognize preferentially only one or a small number of viral protein. It would be interesting to know if the same observation is true in HIV system, and to determine the degree of immunogenicity of each HIV protein. Methods: Peripheral blood mononuclear cells from a large panel of HIV seropositive asymptomatic donors were stimulated in vitro with irradiated autologous blast cells. The cytolytic activity was tested against autologous B lymphoblastoid cells infected with various recombinant vaccinia virus encoded the Env, Gag, Pol, Nef, Vif, Rev and Tat proteins. Results: CTL issued from one responder donor were able to recognize simultaneously several proteins from HIV-BRU isolate (2 to 5 of the tested proteins). Nevertheless, each BRU protein was recognized with.a different frequency: more than 80% for Gag, about 60% for Nef, about 50% for Env. Vif and Rev, and 0% for Tat. Discussion and Conclusions: Our results evidence several remarkable points: a) Several HIV I proteins could be simultaneously recognized by CTL in most healthy HIV seropositive asymptomatic donors. b) Structural proteins as Env, Gag and Pol were immunogenic for CD8+ cells as good as regulatory proteins like Nef, Vif and Rev. c) The observation that these CTL could recognized with a high frequency proteins from HIV-BRU isolate suggested that relatively conserved epitopes were the targets of these CTL. Altogether, these results are encouraging for vaccinal purposes with CTL stimulation. NOTES TU.A.76 RECRUITMENT OF MAST CELLS AND BASOPHILS INTO THE CYTOTOXIC ELIMINATION OF HIV INFECTED CELLS Krauss, Susanne*; Kufer, P.*; Federle, C. *; Riethmfiller, G.*; Rieber, E.P.* Institute for Immunology, University of Munich, Munich, FRG Objective: Basophils and mast cells are the main effector cells in type I hypersensitivity and are potent stimulators of inflammatory reactions. Recently mouse mast cells have been implicated in cytotoxic reactions and human virus specific IgE has been demonstrated. Therefore, we wondered whether basophils and mast cells could be engaged into the elimination of HIV infected cells. In a first approach we evaluated the cytotoxic capacity of human basophils and designed tools to target basophils to HIV infected cells. Methods: Basophils from healthy blood donors were obtained at > 96% purity by density centrifugation, followed by negative selection using monoclonal CD2 and CD14 antibodies and final purification by anti-IgE resetting. In order to bring basophils into contact with HIV or HIV-infected cells a recombinant IgE/CD4-hybrid molecule was constructed where the VH domain of IgE was replaced by the gpl20-binding first and second domain of the CD4 molecule. Results: Stimulation of basophils with anti-IgE led to release of histamine and TNFa. TNFa mRNA was demonstrated in purified basophils. Release of mediators could be enhanced by IL-3. The IgE/CD4-construct was able to bind both the Fc receptor I on basophils and the low affinity Fc receptor II (CD23) on lymphoblastoid B cell lines. It was also capable to block gpl20 binding to CD4+ cells. When armed with the IgE/CD4-construct basophils could be induced to release histamine by monoclonal CD4 antibody as well as by free HIV and HIV infected H9 cells. Conclusion: Our work provides evidence that human basophils have cytotoxic capacity and can be specifically directed against HIV infected cells. Thus, recruitment of the potent IgE-dependent effector system may supplement therapeutic strategies against HIV infection. NOTES NI a TU.A.77 MULTIPLE AND VARIABLE EPITOPES OF HIV GAG PRESENTED TO CYTOTOXIC T LYMPHOCYTES THROUGH HLA B8. Rowland-Jones Sarah; Phillips, R.E.; Nixon, D.F.; Gotch F.; Ogunlesi, F.; Edwards, J.; McMichael, AJ. Institute of Molecular Medicine, John Radcliffe Hospital, Oxford, U.K. Objectives: To examine the cytotoxic T lymphocyte (CTL) response to HIV gag in seropositive patients with HLA B8 during disease progression to ARC/AIDS and to correlate it with analysis of HIV sequence from those patients; and to assess whether sequence variability in HLA B8 restricted epitopes represents the emergence of escape mutants. Methods: Peripheral blood lymphocyte cultures from three HLA B8 positive seropositive haemophiliacs were established be restimulation with autologous PHA-blasts, and tested for cytotoxiclty against autologous B cell lines infected with vaccinias expressing gag or p17 or incubated with synthetic peptides from the gag sequence. Lymphocyte DNA was analysed by PCR using primers from the gag sequence corresponding to epitopes identified as restricted through B8. Peptides were synthesised to correspond to sequence variation in these epitopes and tested for recognition in the cytotoxicity assay. Results: Three B8 restricted epitopes were defined, to which each donor responded at some time. The pattern of recognition varied with time, and one donor has ceased to make a CTL response altogether. One donor has shown recognition of a synthetic variant peptide whilst failing to recognise the index peptide. Discussion: The CTL response in patients with HLA B8 is of particular Interest in view of the association of the HLA A1-B8-DR3 haplotype with poor outcome in HIV infection. Unusually, we have observed recognition of multiple epitopes presented through B8, together with the subsequent disappearance of the CTL response, which may correspond to the emergence of HIV escape mutants. TU.A.78 PRIMARY RECOGNITION OF A CONSENSUS T EPITOPE ON HIV gpl20 BY HUMAN CD4 T HELPER CELLS DEPENDS ON HYPERVARIABLE FLANKING SEQUENCES Manca, Fabrizio; Dessi, Vitalia; Costantini, Maria; Habeshaw, John;* and Dalgleish, Angus* Dept. Immunology, University of Genoa, Italy and *Retrov. Res. Group, CRC, Harrow UK Objective: Human CD4+ T lines specific for HIV gpl20 can be generated by repeated in vitro stimulation with antigen (gpl20 or peptides) from unprimed individuals. Sequence 235-250 is immunodominant in the individual examined. Comparison of analog sequences 236-251 from different HIV strains shows that residues 246-251 are conserved whereas upstream residues are variable. Thus we asked whether in vitro immunization with different analog peptides 236-251 containing upstream variable sequences (236-245) and C terminal consensus sequences (246-251) was equally effective for induction of T cells responsive to the conserved sequence 246-251. Methods: 4 T cell lines were generated by priming with 4 analog 236-251 peptides with the BRU, MN, SF2 and RF sequences. Response patterns of 12 clones from each line to the analog peptides defined 3 epitopes (236-240, 241-245, 246-251). Results: Recognition of the consensus region 246-251 (NVSTVQ) was dependent on the upstream sequence. In fact 4/12 and 5/12 clones from the BRU and RF peptide induced lines were responsive to all of the analog peptides (i.e. specific for 246-251). No clones from the SF2 peptide induced line recognized the consensus epitope. Conclusions: The sequences that flank a defined T epitope may modulate its recognition by the available T repertoire, possibly by interfering with processing and/or MHC association. Thus induction of a desirable T cell response to conserved regions by vaccination may be influenced by the flanking molecular context. CI C> C> C>,

Page  77 TU.A.79 DEFINITION OF A CONSERVED HIA-A3.1 RESTRICTED CTL EPIOPE ON HIV-1 GP41. Takahashi, Kazuo*, Dai,L.*, Fuerst,T.R.**, Biddison, W.E***, Earl,P.L.*** Moss. B.***, Ennis, F.A.* *Dept of Medicine, University of Massachusetts Medical Center, **Meditmfune Inc., ***National Institutes of Health. USA. Objective: Define human CTL epitopes on HIV-1 gpl60. In this report we mapped the epitope and defined the HLA restriction of an HIV-1 gpl60 specific CD8+ CTL Methods: The PBMC of an HIV-1 infected individual were stimulated with anti-CD3 Mab and allogeneic feeder cells, and were assayed for HIV gpl60 specific CTL activity using EBV-transformed B cells infected with recoabinant HIV-1 gpl60/vaccinia virus. A CTL clone E7.20 was made by limiting dilution. The epitope was originally localized to a.a. 751-850 using vaccinia viruses which expressed truncated genes of gpl60, and then was mapped with synthetic peptides. HIA restriction was defined using allogeneic cells, and cells which had been transfected with HEA genes. RESULTS: The epitope recognized by E7.20 was localized to amino acids 765-778 (IAV env sequence) of the internal domain of gp41. E7.20 killed target cells pulsed with peptides of this region of BRU, WMJ1, JH3, and RP strains, despite a change frcm T to V at a.a.777. However, E7.20 did not recognize this region of MN strain which has a deletion of a.a. 769, and a change from R to H at a.a.768. E7.20 killed allogeneic target cells which shared the HIA-A3 antigen; moreover, they lysed a HLA-A and -B negative human plasma cell lines HMy.ClR after transfection with and expression of the HA-A3. 1 gene. Conclusions: 1. A CD8+ CTL clone E7.20 recognizes a conserved epitope on the internal domain of gp41. 2. This cytotoxicity was restricted by the HIA-A3.1. 3. This study defines a crossreactive CTL epitope on gpl60 of HIV-1 which may be important for vaccine development. NOTES TU.A.80 INHIBITION OF HIV REPLICATION BY HIV-SPECIFIC CTL I.A, Koenig ScotI', Jones G", Boone E", Brewah A, Newell A', Torbett B', Mosier D', Fauci AS' Medlmmune, Inc., Gaithersburg, MD. USA, LIR, NIAID, NIH, Bethesda, MD, USA "'Medical Biology Institute, LaJolla, CA. Obiective: To determine if CTL clones specific for different HIV proteins can inhibit viral replication in vitro and in vivo. Methods: HIV-specifkc, MHC-Class I restricted CTL to net, env, and gag were cloned by limiting dilution from CD8 cells from HIV-seropositive donors. CTL were tested for their ability to lyse cells acutely infected with HIV or to inhibit HIV replication in freshly sorted CD4+ cells activated with anti-CD2 and IL-2 from HIV-seropositive donors. Net-specific CTL were examined for their ability to inhibit an acute HIV-1,,,, infection in SCID mice reconstituted with peripheral blood mononuclear cells (PBL) from MHC-HLA Class I matched and mismatched seronegative donors. Results: The HIV-specific CTL lysed cells acutely infected with HIV and inhibited HIV replication of activated CD4' cells from HIV-infected patients in a MHC-restricted manner in vitro. Some HIV-specific clones, activated with synthetic peptides containing the epitope to which they were restricted, inhibited HIV-replication by diffusion of a factor through a semi-permeable membrane. Suppression of HIV replication in spleen and peritoneal exudate cells was observed in 2 of 5 SCID mice reconstituted with the MHC-matched PBL, 2 weeks after acute infection with HIV after receiving a single transfer (107 cells) of a nef-specific CTL clone. No suppression of viral replication was seen in the hu-PBL-SCID mice receiving the HLA mismatched PBL. Experiments are in progress to evaluate the inhibitory effects of multiple transfers of these CTL to hu-PBL-SCID mice challenged with HIV. Conclusions: In vivo induction or adoptive transfer of CTL to nef, gag, and env sequences may be useful in protecting individuals against HIV infection and this may occur by cytolysis of infected cells or by inhibition of viral replication through soluble mediators. NOTES

Page  78 0 TU.B 81 NON-HODGKIN LmHOMA IN PATIENTS WITH ADVANCED HIV-INFECTION TREATED WITH ZIDOVUDINE. Moore, Richard D,; Kessler, H.; Richman, D.D.;D Chaisson, R.E.; and the ZVD Epidemiology Study Group, Johns Hopkins University School of Medicine, Baltimore, Maryland, USA. Objective: To investigate the relationship of zidovudine therapy and other factors to the development of HIV-related high-grade non-Hodgkin lymphoma (NHL). Methods: Cohort of 1028 patients with AIDS and advanced ARC followed for 2 years, who received zidovudine and had no history of suspected or proven lymphoma. Results: NHL developed in 24/1028 (2.3%) patients who received zidovudine during 1456 person-years of follow-up (rate = 1.6/100 person years of therapy). The relative hazard for development of HIL was stable throughout 2 years of therapy with the risk of developing NHL 0.8% for each additional 6 months of therapy. Factors associated with development of NHL were prior history of Kaposi sarcoma (p < 0.01) and, less strongly, a prior history of herpes simplex virus infection (p < 0.05), a history of cytomegalovirus disease (p = 0.07), and homosexual transmission of HIV (p - 0.07). By Cox proportional hazards analysis, a history of Kaposi sarcoma and of cytomegalovirus disease were most strongly associated with development of NHL (p < 0.01). Patients with NHL were significantly more likely to have a lover average daily dose of zidovudine than those who did not develop NHL (p < 0.04). Discussion and Conclusions: Our study demonstrates a relatively high incidence of NHL in patients with advanced HIV disease on prolonged antiretroviral therapy, and also indicates that development of NHL is not well correlated with the duration of exposure to zidovudine. NOTES TU.B.82 DEFINITIVE EXCLUSSIDN OF HIV INFECTION IN A KAPOSI'S SARCOA BISEXUAL NAN. SU55ESTIONS ON A PATH06ENIC MODEL DF KS Scnano V, Hwlett I, ri edman-Kien.t, Tor J, Huang it, Epstein J. Hosoital aermans Trias i Pjoli, Barcelona, Spain. INYU ledical Center, Newa orn City, USA. OBJECTIVES. Sugestios aout a s ia5l transmittd soet other than HIV in 5 has bee rct E. rntly, this is in contrast;:th lakoratory results which suBgests that cvtokines a0no H-Tat protein, released by *ctvatI and infected cells other than spirdie ceis, in con cert, can orrduce the tuaor. Hoo/bisexual sen who have KS and are HIV seron~Gn t;e L:Ofrise the most interesting group In Whih to investigate the etiroloy o0 F S PATIENT AND NETHODS, A 42-year-3ld ise.ual man was initiay seen in 1n86, is5eminated sin lesions and gastrointestinal involvement oi KS were re~gnise I at that aoment. Serum sa5slie c'!lcteO in a~re than five tiAs sines that "scen't er. tes-e rc. ea`et ' reegative for H!V-A:b EIA nd B) and HIV-Ag. For this reasFo we decied. to investigate the 'resence of HIV infection using PCR (gag an env prisers in,A and?NA n racted frosm PvC anM serl; acllected in April 10, Bcause Tax protein prTduced by, HTLV-!II has sefects similar to H!V-Tat and - has been reco3nlsed in HiV-! inferctd subjects, we investigate the presence of HTLV-iI!! Ab using EIA an i,,ifin; e, e anilyzed; PCR the presence oa specific HIV-Tat sequencrls, because f'ragents of t;E HIV genoE otain Tt mav becomr in:oroorated in some cells. which Say proGucs Tat protein in the absence of expresson f other Iv _?e", RESULTS. Negative P'R resulTs wrer catAind itn both serua and celle Cooking -4A and RNA HIV secuences. Antiscdi:s t -TlV-I/jI w*:lrc negative sv EIA anc N5. FCinall, PCR was negative fur HIV-Tat sesuenceE in PRbC, DISCUSSION. ConsistEnt with these result, an ortory d a t and epindiiological data, we propose a hyptet-ical e-el for the pathogenesis of KS ir whi;h so causal factors ae required. Initiaily, an infectious agent possibly a virus nLhea thar HI'V, ser-1vs as a "trigoer whic causes the differentiation, and possible transformation, f a mrr;iss C eogenitor cell into a KS spfndie-snaped cli, This KS -ell is a cytSoine responsive ceil, KS develop When a secnOd rircumstancE, namely prolonged iseunostiaulation is resent, e.g, HIV infection, malaria so;e;rrcan subjects or ianosprciri sive therapy in organ tranansolant rrecpiens.!n H! infected homosexual men, the ircreased prevalencE c, 6 acoards With thpir likely greater eposure to the putative sexual transaissible KS agent and with the addit-onai pct-rntat g e -ffet:C-5eC oY td e HIV-Tat nrotein, hich may act as a stimulatory cofactor on KS spindle cells. NOTES TU.B.83 KAPOS'S SARCOMA AMONG HIV-SERO-NEGATIVE HIGH RISK POPULATION Safai._BWjan*; Peralta H*; Menzies K*; Tizon H*; Roy P*; Flomenberg N*; Wolinsky S**. *Memorial Sloan-Kettering Cancer Center, New York City, NY, USA; **Northwestern University Medical School, Chicago, IL, USA. Objective: To describe a small number of cases of Kaposi's Sarcoma (KS) among HIV-sero-negative, high risk population, who have been identified and followed for more than 9 years. Methods: Since the beginning of the AIDS epidemic we have seen and followed more than 350 cases of AIDSAssociated KS and identified a subpopulation of 44 KS cases who have lived more than 36 months without developing opportunistic infections (01). Among these long-term survivors we have identified 6 cases who have lived longer than 36 months without 01 and are HIV-sero-negative. Demographic data has been obtained, progression of the tumor has been recorded, complete hematologic, immunologic and virologic profiles have been repeatedly determined. Results: The average age of these cases is higher than HIV-Associated cases and is lower than the Mediterranean KS. Five of the 6 cases are still alive with survival ranging from 38 to 130 months. The disease is active in only one case. Virological data including antibody testing and PCR indicate no evidence of HIV infection Hematologic and immunologic profiles demonstrate normal numbers of WBC, T-cell count and T-cell function. Of interest is the observation of a drop in these parameters in almost all patients at the beginning of their disease onset which is later corrected. Discussion and Conclusions: The presence of KS among high risk population of HIin New York is seen without the presence of HIV infection. No evidence of continuous immune deficiency is detected. One of three following possibilities may e operative: 1) these individuals have the Mediterranean KS and, although high risk, they are free of HIV infection; 2) the necessary HIV genome for induction of KS is incorporated in the DNA of appropriate cell type to produce KS, but the virus is not expressed to be recognized by the immune system; and 3) a separate virus is involved in causing KS and that this virus has been transmitted along side with HIV, but in these very few individuals only the KS virus has been transmitted and not HIV. TU.B.84 SURVIVAL PATTERN AND RESPONSE TO THERAPY IN 104 PATIENTS WITH HIVRELATED KAPOSPS SARCOMA Krickeberg. Holger, Rasokat,H Dept.of Dermatology, Univ.of Cologne, Cologne, Germany. Objective: To describe response to treatment with AZT or alpha-interferon (a-IFN) and parameters of survival in 104 patients with HIV-related Kaposi's sarcoma (KS). Methods: 104 HIV-patients with KS [age: 38 (21-60) years; year of diagnosis: 1983-1990; follow-up: 17.9 (1-77) months] were included in an retrospective analysis. Response to therapy with AZT or a-IFN was analysed in terms of complete remission (CR), partial remission (PR), and stable disease (SD). The influence on survival was tested for the following parameters: age, year of diagnosis of KS, stage of KS, CD4+ -count at diagnosis, treatment with AZT or a-IFN. Results: Significapt tumor remissions were seen with a-IFN but not with AZT alone. In 31 patients treated with a-IFN (10-20x10U/mZ) there were 11 CR and 9 PR, while in 54 patients receiving AZT there were 4 PR and 12 SD. However, AZT stabilized responses to prior a-IFN-therapy. 11/20 patients responding to initial therapy with a-IFN received AZT afterwards: in these patients time to relapse was significantly longer than in 9 patients not given AZT (40.3 and 21.4 months). 63/104 patients died. Survival depended strongly on stage and course of underlying HIV-disease. Mean of survival for all patients was 20 months (CI-95%: 15-22) and differed (p <.0001) with CD4-counts at diagnosis [CD4 < 200/1l: 14 months (N=58); CD4 200-399/41: 37 months (N-.28); CD4 400/pl: 57 months (N=11). There was no correlation of survival and age, year of diagnosis, or stage of KS. However, survival was significantly longer in AZT-trcated patients with initial CD4-counts < 200/4l than in not treated patients (18 and 7 months; p <.0006). Conclusion: AZT seems to prolong survival in patients with KS and prolongs time to relapse in patients responding to prior treatment with a-IFN. C> cz C> toj (?) Q'^

Page  79 TIBB85 PHASE I/II EVALUATION OF AN IL-2 RECEPTOR TARGETED FUSION TOXIN, DAB486IL-2, FOR TREATMENT OF AIDS-ASSOCIATED KAPOSI'S SARCOMA. LeMaistre. Charles, F.*, Craig, F.*, Smith, J.*, Meneghetti, C.*, Banks, P.*, Parker, K.**, Nichols, J.**, Woodworth, T.** *University of Texas Medical Center-San Antonio, TX, USA, **Seragen, Inc., Hopkinton, MA, USA. DAB48eIL-2 is a recombinant fusion protein composed of human interleukin-2 and cytotoxic fragments of diphtheria toxin. This fusion toxin selectively intoxicates cells expressing the high affinity Interleukin 2 (IL-2) receptor. When administered to patients with chemotherapy refractory IL-2 receptor expressing hematologic malignancies, DAB486IL-2 induces significant anti-tumor effects at well tolerated doses. Since certain epidermoid cancers and sarcomas can also express functional IL-2 receptors and further, since DAB486IL-2 has been shown to selectively kill HIV infected lymphocytes and monocytes in vitro, we have begun to evaluate this agent in AIDS associated Kaposi's sarcoma. IL-2 receptors were demonstrated on Kaposi's sarcoma cells using B-D phycoerythrin labelled monoclonal anti-TAC (detects the p55 chain of the IL-2 receptor) and modified immunohistochemical techniques. Two patients with advanced AIDS and disseminated, chemotherapy resistant Kaposi's sarcoma have demonstrated approximately a 30% regression in their skin lesions following one course of 5 doses of DAB4g6IL-2 administered as a daily 90 minute infusion. The selective cytotoxicity for IL-2 receptor expressing malignant cells as well as the possibility for anti-HIV activity underscore the importance of development of DAB4g6IL-2 in HIV associated Kaposi's sarcoma. Additional patients and dose/schedule regimens are currently being evaluated. NOTES TU.B.86 HIV, ANAL CONDYLOMATA AND ANAL INTRAEPITHELIAL NEOPLASIA de Ruiter Annemiek*, Carter P+., Katz D*., Whatrup C+., Hawkins A.*, Northover J.+, Mmidel A.*, *University College & Middlesex School of Medicine & +St Mark's Hospital, London, UK. Obective: To study the possible associations between HIV infection, anal condylomata and anal intraepithelial neoplasia (AIN). Methods: HIV antibody positive and negative homosexual and bisexual men attending a Central London Sexually Transmitted Diseases clinic were assessed for the presence of anal condylomata and AIN using microanoscopy (proctoscopy with an operating microscope) and biopsy. Statistical methods used were log linear modelling and chi squared analysis. Results: Thirty one out of the 60 HIV positive patients had anal condylomata. Fourteen of these 31 had histological evidence of AIN compared to only 3 of the 29 without anal condylomata and this difference was statistically significant (p=0.003). Twenty one out of 28 HIV negative patients had anal condylomata, of which 7 had biopsy proven AIN compared to none of the 7 without anal condylomata, which was not statistically significant (p=0.208). In patients with anal condylomata, there was no significant association between AIN and HIV status (p=0394). Eight out of the 17 HIV positive patients with AIN were CDC stage 2 or 3 and 9 were CDC stage 4. All 4 cases of AIN3 occurred in HIV positive patients. All patients with AIN are being closely followed and of 14 patients (9 HIV positive, 5 HIV negative) that have been followed for a median of 3 months (range 1-6), there has been no apparent progression or regression of their AIN. Conclusions: 1. There is a high prevalence of AIN in homosexual and bisexual men with anal condylomata. 2. In HIV positive patients there is a statistically significant association between anal condylomata and AIN. 3. The more severe form of AIN occurred in HIV positive patients. 4. Prospective studies are required to determine the natural history of AIN in HIV positive and negative patients. NOTES TU.B.87 QUANTITATION OF AZT RESISTANT VIRUSES IN THE HIV COHORT AT THE ROYAL LONDON HOSPITAL Broadhurst. Katrina; Ball, A.; Stein, C.A.; Levantis, P.; Forster, G.E.; Goh, B.T.; Colvin, B.; Jackson G.G. & Oxford, J.S. The London Hospital Medical College, London, England Objectives: To monitor a cohort of 120 HIV patients (41% sexually acquired, 22% haemophiliacs, 6% intravenous drug users and 31% risk group unknown) for clinical parameters and surrogate markers and to characterise their viruses. 45 individuals have been receiving AZT for 1 month to 3 years. Methods: Isolation of AZT-resistant (rHIV) and sensitive(sHIV) virus is by cocultivation with PHA stimulated PBMC's in the presence and absence of 51M AZT respectively. Molecular analysis is by PCR amplification within a 1.7Kb region of pol gene encoding RT followed by M13 or direct sequencing. Results: Forty-nine sHIV's have been isolated from 42 individuals (with sequential isolates from 7 patients) who have never received AZT. No rHIV have been isolated from this group. From patients receiving AZT 24 rHIV's have been isolated from 18 individuals, (sequential from 5), including rHIV's from 2 heterosexuals, 1 haemophiliac and 1 IVDU. Nine individuals have never yielded rHIV despite AZT treatment for up to 2 years. By lowering the screening concentration to 1ipM we have isolated 14 viruses from 10 additional drug-treated individuals and we also have 7 putative rHIV's from 7 individuals who have never been treated with AZT. However, no confirmatory molecular analysis has yet been done on these viruses. Our reference rHIV, from a patient AZT-treated for two years, is resistant to 125pM AZT in culture and has shown cross-resistance with another dideoxy nucleoside. There are mutations from the wild-type RT at aa positions 41, 138, 142, 210, 215, 248, 259, 467, 483 and 490. In other cohort viruses isolated in 5pM AZT changes occur at positions 70 and 211, as well as 210 and 215. Conclusions: Detection of rHIV is dependent on treatment of patients with AZT for at least two months. sHIV continues to be isolated from a proportion of drug treated patients either continuously or:. intermittently with rHIV. TUlJB.88 RAPID GENERATION OF AZT-RESISTANT STRAINS OF HIV-1 AFTER LOW-DOSE THERAPY AND BY IN VITRO SELECTION. Wainberg. Mark A.; Gao, Q.*; Yao, X-J.*; Parniak, M.A.*; Fanning, M."; Montaner, J.S.G."; O'Shaughnessy, 4.*; Tsoukas, C.**; Ruedy, J.* *Jewish General Hospital and McGill AIDS Centre, Aontreal, Canada, and the "Canadian AZT Multicentre Study. Our group has previously described the isolation of AZT-resistant variants of HIV-1 from over 50% of asymptomatic subjects who were treated with high doses of this compound (1000-1200 mg/day) for periods greater than 27 weeks by inclusion of 7.5 pM AZT in the medium of cord blood lymphocytes used in co-culture with the peripheral blood mononuclear cells of infected donors. We now report, by this nethod, the isolation of AZT-resistant virus from 7 of 10 subjects who received this irug at concentrations of 300-400 mg/day for periods of 16-24 weeks. Mean time of therapy to demonstration of AZT-resistant variants from 16 of 24 patients receiving 300 mg AZT/day was 30 weeks, while that for patients on low-dose AZT (300-400 ng/day) was 21 weeks. Each of four patients who were switched to ddl therapy continued to have AZT-resistant viruses in their circulation at 10 weeks therafter, and did not yield any viruses possessing a ddl-resistant phenotype. The use of 3olymerase chain reaction (PCR) on blood samples from patients in our cohort, using primer sequences described by other groups (Larder et al.) revealed concordance in approximately 60% of individuals who yielded drug-resistant isolates after being treated with either high or low dose AZT. We have also demonstrated, using in vitro selection procedures and increasing concentrations of AZT and other nucleosides (0.01 pM - 10 pM), the generation of HIV variants with resistance to AZT, ddI or JdC. Four of 6 such AZT-resistant variants were shown by PCR to have polymerase gene nutations similar to those described bv other groups, while the other two did not, ' bo bo oo (00

Page  80 S TU.B.89 ZIDOVUDINE SENSITIVITY OF HIV DURING THE DEVELOPMENT FROM THE SASYMPTOMATIC STAGE TO AIDS Boucher, Charles*; Goudsmit, Jaap*; O'Sullivan, Eithne"; Ramautarsing, Chitra*; Lange, Joep*; Larder, Brendan** * Human Retrovirus Laboratory, University of Amsterdam, Amsterdam, Netherlands, ** Molecular Sciences Department, Wellcome Foundation Ltd, Beckenham, Kent, England. Objective: To determine the pattern of zidovudine sensitivity of HIV isolates obtained during the development from asymptomatic stage to AIDS. Mehods: 18 asymptomatic persons (HIV core antigenaemic), six progressed to AIDS receiving 1000 mg AZT/24 hrs, two years follow up. Sequential HIV isolates were analyzed with 1) A HeLa-CD4 plaque assay to determine AZT sensitivity and 2) PCR to detect four mutations on reverse transcriptase (codons 67, 70, 215, 219) which confer resistance. Resuts: Zidovudine sensitivity of isolates obtained during the asymptomatic stage and once AIDS developed were partially resistant Genomic analysis showed that the mutation at codon 70, creating a partial resistant phenotype, appeared first (at six months present in 65%). The next mutation to appear was located on codon 215, also causing partial resistance. Results from isolates sampled at short intervals show that once the 215 mutation appeared the population with a mutation at codon 70 disappeared for some time, to reappear later. The 215 mutation occurred earlier in those individuals who eventually progressed. From all individuals after two years of therapy 89% of the isolates were mutant on codon 215, 33% at codon 70 and none at codon 67 or 219 (confirming the observed partial resistance). In one individual with a partial resistant isolate at the time AIDS developed all four mutations could be detected 12 months later. Termination of AZT treatment for 12 months in an asymptomatic individual and eight months in person who had developed AIDS did not result in a change in the partial resistant pheno/geno type. Discussion and concuions: Long term AZT treatment (two years) of asymptomatic individuals results in only partial AZT resistant isolates, which persist after discontinuation of therapy. Progression to AIDS can occur before the emergence of fully resistant isolates. NOTES TU.B.90 EVALUATION OF EFFICACY OF THERAPY AND EMERGENCE OF ZIDOVUDINE (AZT) RESISTANT HIV-1 BY QUANTITATIVE CULTURES AND PCR. Israele, Victor; Chelyapov, Nickolas V.; Ho, David D.; Wittek, Alec E.; Moudgil, Tarsem; Alam, Masud; Brunell, Philip A.: Cedars-Sinai Medical Center, Los Angeles, CA. Objectives: To evaluate the effect of AZT on HIV-1 viral load, to detect the emergence of HIV-1 AZT resistant strains and to assess the impact of HIV-1 resistant strains on viral replication in HIV-1 infected children. Methods: Quantitative end-point-dilution cultures of patients' plasma and peripheral blood mononuclear cells (PBMC) were performed in the presence and absence of AZT. DNA was amplified by PCR and quantitated as previously described. Results: In this ongoing study, serial determinations of viral titers of 8 AZT treated patients revealed a decrease from a pretreatment mean of 2131 tissue-culture infective dose (TCID) per 10 PBMC to 11 TCID per 106 PBMC after a minimum period of four weeks on AZT treatment, and from 179 TCID per ml of plasma to 5 TCID per ml. Mean proviral HIV-1 DNA copy numbers decreased from 4.6 per 103 PBMC to 0.28 per 103 PBMC. Four of the five treated patients tested for AZT resistance had detectable levels of resistant HIV-1 after receiving AZT therapy for a 2-9 month period. In three patients who developed HIV-1 AZT resistant strains, a mean 70, 9 and 6 fold increase in the viral burden in PBMC, plasma and HIV-1 proviral DNA copy number, respectively, was observed 5-9 months after the initial decrease produced by AZT therapy. One perinatally infected child acquired a resistant strain from his mother, and failed to respond to AZT therapy. Conclusions: Zidovudine treatment for 4 weeks or more decreased the viral burden in PBMC, plasma and the proviral HIV-1 DNA copy number. HIV-1 AZT resistant strains were detected 2-9 month afteinitiation of AZT therapy and-tteir appearance was associated with an increase in the level oT viraT-rep on. NOTES ~tl 00 <0 TU.B.91 LONG-TERM FOLLOW UP OF PATIENTS WITH HIV ISOLATES RESISTANT TO ZIDOVUDINE(F (AZT) TREATED WITH DIDEOXYINOSINE (DDI) BACH, Michael C.*; St. Clair, M.**; King, D.**; Vavro, C.** *Maine Medical Center, Portland, Maine, USA, **Wellcome Research Laboratories, Research Triangle Park, North Carolina, USA. Objective: To document the clinical course and virologic susceptibility of six patients with documented AZT-resistant HIV who were clinically failing therapy on AZT treated with DDI for 12-15 months. Methods: Peripheral blood mononuclear cells (PBMC) were cocultured with PBMC's from healthy donors and cell free supernatant harvested at peak reverse transcriptase (RT) expression. Healthy PBMC's infected with HIV exposed to AZT or DDI for 10 days, then cell-free supernatant analyzed for RT and IC50 determined. Results: All six patient isolates were resistant to AZT (6.08-16.1 uM). Five were initially sensitive to DDI (.3-1.68 uM), one had an IC50 of 2.2 uM. Four had dramatic clinical responses to DDI with falls in P24 antigen. Over 12-15 months, repeat DDI sensitivities showed rises of 1-6 uM with corresponding falls in AZT IC50 of 5-13 uM. Four patients developed gradual clinical deterioration and died, three with CMV. Discussion: Clinical and virologic resistance to AZT and DDI occurs. Reversion from AZT- resistant to AZT-sensitive strains occurred on DDI therapy. The mechanisms of DDI-resistance and reversion of resistance to AZT needs further evaluation. The clinical significance of resistance reversion of AZT is at present unknown. TU.B.92 DECREASED DIDEOXYINOSINE (DDI) SENSITIVITY OF HIV ISOLATES OBTAINED FROM LONG TERM RECIPIENTS OF DDI. Reichman, Richard: Lambert J; Strussenberg J; Dolin R. University of Rochester AIDS Clinical Trials Unit (ACTU), Rochester NY USA. Objective: To measure in vitro sensitivity to ddl of HIV isolates obtained from patients prior to and after receiving ddl for symptomatic HIV disease. Methods: HIV was isolated in peripheral blood mononuclear leukocyte (PBML) cultures from patients enrolled in a Phase I study of ddl (NEJM 322:1333-1340. 1990). Sensitivity of isolates to ddl was determined in a PBML assay using reduction of p24 antigen production as an endpoint. Results: Paired isolates from 14 patients who had received ddl for a median of 52 weeks were evaluated for sensitivity to ddl. One patient had AIDS, and 13 had ARC. Twelve had received AZT previously for a median of 8.5 months. Eight patients had measureable levels of serum p24 antigen, and in all 8, antigen content decreased soon after initiation of ddl therapy. However, in 6 patients, p24 levels began to increase after 6 months of treatment. The median minimum inhibitory concentration (MIC) of pre-treatment isolates for ddl was 5 micromolar, and the median MIC of post-treatment isolates was 10 micromolar. Among the post-treatment isolates, 7/14 (50%) had MICs 4-fold or greater than the pre-treatment isolates. Discussion and Conclusions: Some HIV isolates obtained from patients who had received prolonged ddl therapy demonstrated decreased sensitivity to ddl compared to pre-treatment isolates. The clinical significance of this observation is unknown, and requires further investigation. Studies designed to determine the molecular basis of the decreased sensitivity of these isnoltes to ddl a re ongoing Co CI C> I~R I: rY t3

Page  81 TJ.B.93 ZIDOVUDINE THERAPY IN PEDIATRIC PATIENTS: REPORT ON 622 CHILDREN ENROLLED IN A TREATNENT IND Creaqh-Kirk, Terri*; naha,M.*; Yankaskas, B..; Balsley, J.**; Andrews, E.*; Tilson, H.* *Burroughs Wellcome Co., Res. Triangle Park, NC, USA; **NIAID, NIH, Bethesda, ND, USA Objective: Between 1986 and 1990, clinical studies of zidovudine (ZDV) therapy were conducted in children with HIV disease. While data from these studies were being summarized for review by the Food and Drug Administration, a Pediatric Treatment IND was initiated to make zidovudine therapy more widely available to children. ethods: Enrollment criteria included the following: age<12 years, evidence of HIV infection, and HIV-related symptoms and/or evidence of compromised immunity. Informed consent was obtained from a parent or guardian. Initial dose was 180 mg/ma every 6 hours. Data were collected on patient status as well as on serious or previously unreported adverse events. Results: 56% of enrolled patients were male and 79% were of a minority ethnicity. Nedian age was 35 months. 86% of patients acquired HIV perinatally. 5% of patients required a dosage adjustment to 120 mg/am due to ZDV intolerance, and 3% required temporary discontinuation. Some followup was received for 540 pts., among whom median followup was 80 days. 17 deaths due to HIV-related complications were reported among 622 enrolled children during the 7 months in which the program was in effect. The only previously unreported adverse experience was one case of muscle spasm and torticollis in a 6-year-old boy. Conclusion: In a cohort of 622 children treated with zidovudine during a 7-month period, drug tolerance appeared to be at least equivalent to that seen in adults. NOTES TR.B. 4 PROLONGED DIDEOXYINOSINE (ddl) THERAPY IN HIV INFECTED CHILDREN: Butler Karina M: Husson R; Lewis L; Mueller B; Gress J; Jacobsen F; Higham C; Montrela K; Jarosinski P; Balis FM; Poplack DG. and Pizzo PA,. Pediatric Branch, National Cancer Institute, NIH, Bethesda, MD, USA Objective: We reported that HIV-infected children who received 24 weeks of ddl had a significant increases in CD4 counts and decreases in p24 antigen levels (NEJM 1991,323:137). It is important to determine i these changes are sustained and whether more prolonged therapy is associated with an increase in druqtoxicity. Methods: Children enrolled in a Phase 1/11 trial of ddl given at doses ranging from 60 to 540 mg/mz/day have been monitored for up to 97 weeks (median 42) and serially assessed for safety and tolerance of ddl, and changes in clinical, immunological and virological manifestations of infection. Results: 89 HIV-infected children, 0.7-19 yrs (median 6.8), including 43 vertical, 46 blood product acquired infections; 39 previously untreated and 50 who had received prior therapy were evaluated. 57% patients (pts) had a history of thrush, 52%/ of an 01, 19% of LIP, 29% of bacterial infections &/or pneumonia and 34% were encephalopathic at entry. The CD4 count (median:89) was <200/mm3 in 50, 200-499 in 21 and >500 in 18. p24 antigen was detectable in 61% at entry. 59/89 (65%) continue on study: 12 died, 2 continue on ddl off protocol, 1 withdrew day 1, 13 switched to an alterative agent and 2 elected to discontinue all therapy. Significant adverse events included pancrealitis (6 pts receiving ddl at ~ 360 mg/m2/day), asymptomatic peripheral retinal atrophy (3 pts), seizures (2), cardiac arrhythmia (2 pts, each with pre-existing cardiomyopathy), peripheral neuropathy (2 pts, 1 who had stopped ddl 8 days previously and was pre-terminal), diabetes insipidus (1) and neutropenia (1). Paired values for CD4 counts at 0, 24 & 52+ weeks were available on 35 pts. Collectively, values were higher at 24 weeks (median 324 vs 238 cells/mm3, p=0.0069) but returned to baseline by 52 weeks (median 212 cells/mm3, p=0.272). However, on an individual basis the CD4 count remained >10% & 50 cells higher than baseline value at 52+ weeks in 11/33 pts. p24 antigen levels were detectable in 24/38 (63%) at week 0, 12/38 (32%) at 24 weeks and in 13/38 (34%) at 52+weeks. p24 antigen levels available on 38 pts at week 0, 24 and 52+ weeks were significantly lower at 24 (median <31 vs 159 pg/ml, p=0.005) and 52+ weeks (median <31pg/ml, p.0.0002). Conclusions: ddl is well tolerated and antiretroviral activity is sustained in many up to and beyond 1 yr. Toxicities that may prove to be dose or duration related include pancreatitis, peripheral neuropathy and peripheral retinal atrophy NOTES TU.B.95 SIMULTANEOUS COMBINATION THERAPY WITH ZIDOVUDINE (AZT) AND DIDEOXYINOSINE (ddl) IN CHILDREN WITH HIV INFECTION. Husson..1; Tishler, D.2; Kovacs, A.3; Farley, M.1; Woods L.2; Ono, J.3; Butler, K.1; Lewis, L.1; Jarosinski, P.1; Holcenberg, J.2; Avramis, V.2; Pizzo, P.1. National Cancer Institute, Bethesda, MD; 2Los Angeles Children's Hosp., Los Angeles, CA; 3Los Angeles County/USC Medical Center, Los Angeles, CA, USA. Objective: AZT and ddl have antiviral activity in adults and children, but their use is limited by toxicities in many patients, and their activity may not be sustained with long-term treatment. Combination therapy with these agents nay result in less toxicity and/or increased antiviral activity. We are evaluating the safety and toxicity, as well as the antiviral activity, of combination AZT/ddl therapy in children with HIV infection. Methods: Children with CDC Class P-2 symptomatic HIV infection, or those with asymptomatic infection but whose CD4 cell count is less than 500/mm3 are eligible for enrollment. Patients with no prior antiretroviral treatment are enrolled into Arm A, and those with a history of hematologic intolerance on full dose AZT are enrolled into Arm B. Seven dose levels ranging from 360 mg/m2/day of AZT and 120 mg/m2/day of ddl to 720 mg/m2/day of AZT and 270 mg/m2/day of ddl are being evaluated in Arm A. Six dose levels ranging from 240 mg/m2/day of AZT and 180 mg/m2/day of ddl to 360 mg/m2/day of AZT and 360 mg/m2/day of ddl are being evaluated in Arm B. In addition to T cell subsets and p24 antigen, the virus burden in blood is being monitored as a potential marker of antiviral activity. The intracellular pharmacodynamics of AZT and ddl are also being evaluated. Results: As of 1/18/91., nine patients ( 8 in Arm A and 1 in Arm B) have been enrolled at 4 dose levels. The median age at entry was 7.8 years (range 3.3 to 18), the median T4 cell percentage was 23% (4-28) and the median T4 call number was 290/mm3 (24-808). 5 patients were p24 antigen positive at entry. The plasma viral titer in 5 patients at entry ranged from <2 to >250 TCID50/ml and the viral titer in mononuclear cells ranged from 1.8 to 370 TCID50/106 cells. The median time on study is 6 weeks (4-19 weeks). Significant toxicity has not been observed in any patient thus far. p24 antigen declined in 4/4 patients in whom follow-up values are available. Conclusions: The combination of AZT and ddl appears to be well-tolerated in the short term, with no acute toxicities observed at the lower dose levels evaluated thus far. Evaluation of the longer term safety, toxicity and 00 antiviral activity, of these and higher doses, continues and will provide insights into the safety and activity of this I combinalion. TU.B.96 EFFICACY OF INTRAVENOUS IMMUNOGLOBULIN (IVIG) FOR THE PROPHYLAXIS OF SERIOUS BACTERIAL INFECTIONS (SBI) IN SYMPTOMATIC HIV-INFECTED CHILDREN MofensonLyane M *, for the NICHD IVIG Clinical Trial Study Group. * Pediatric, Adolescent & Maternal AIDS Branch, NICHD, Bethesda, MD, USA OBJECTIVE: To determine the efficacy of IVIG for prevention of lab-proven and clinically-diagnosed SBI in symptomatic HIV-infected children. METHODS: Between 3.1.88 and 10.31.90, 372 symptomatic HIV-infected children (12% CDC class PlB, 88% P2; mean entry age 40 months; 91% perinatal acquisition) were enrolled in a multicenter, randomized, double-blind, placebo-controlled clinical trial of IVIG, 400 mg/kg every 28 days, versus albumin placebo (PL). Children were stratified by entry CD4 count and previous infection history; median follow-up was 17 months. At entry 15% of children were receiving PCP prophylaxis, increasing to 49% by 10.31.90; 39% of children had received AZT after entry. RESULTS: As of 10.31.90, 77 PL patients experienced.1 SBI compared to 54 IVIG patients. In children with entry CD4 >_200/mm3, IVIG significantly prolonged SBI-free time (57 PL patients with 21 SBI compared to 38 IVIG patients, p = 0.008 for 24 month infection-free survival); at 24 months, infection-free survival was 48% in the PL group compared to 68% in the IVIG group (p = 0. 005). PL children experienced 1.4 times more bacterial infections of any nature than IVIG children (p = 0.02). For children with entry CD4 >_200/mm3, IVIG treatment was also associated with reduction in number of hospitalizations; approximately 45 excess hospitalizations were observed per 100 patient-years in PL children. For children with CD4 <200/mm3 at entry, no efficacy was seen (.1 SBI in 15 PL patients compared to 14 VIG patients). Mortality was equal in PL and IVIG groups (31 deaths in each). Adverse reactions were observed in 0.6% of PL and 0.5% of IVIG infusions. CONCLUSIONS: IVIGis effective in prolonging SBI-free time in symptomatic HIV-infected children with CD4 counts 200/mm3 or above. In those children with CD4 <200/mm3, alternative interventions may be needed to prevent SBI.

Page  82 S TU.B.97 GRADUATED THERAPY WITH IVIG, PCP-PROPHYLAXIS, ZIDOVUDINE AND DDI IN HIV-INFECTED CHILDREN W. Kreuz, T. GOnglr, M. Funk, I. Kynast, A. Allendorf, R. Linde, S. Ehrenforth, Chr. Lotz, K. Debatin, D. Hofmann, B. Kornhuber Department of Pediatrics, University children's hospital, 6000 Frankfurt am Main, Germany In our clinic we currently care for 75 children (0,1 - 16 ys) exposed to HIV 1. 34 are definitely HIV-infected, 21 by vertical transmission, 13 patients (3-16 ys) by blood products. These children are (CDC)-classified as PI (n =15), P2 (n - 9), and AIDS (n - 5); 5 died. Since the end of 1985 we have conducted a intravenous immunoglobulin (IVIG)-prophylaxis (200 mg/kg b.w./every other week) to prevent bacterial and viral infections. We combined a polyvalent 7 S-immunglobulin (Intraglobin F~, Biotest, Germany) with a hyperimmunoglobulin containing high titers of CMV- and EBV-antibodies (Cytotect~, Biotest, Germany). IVIG-prophylaxis was started when clinical signs and symptoms of B-cell dysfunction were present. Since 1987 we have added oral Zidovudine (5-28 mg/kg b.wJday q.i.d.) in case of HIV-related neurological symptoms, after opportunistic infections or rapid immunological deterioration (CD4-cells < 200-5001/1). Since the end of 1987 all children have received PcP-prophylaxis prior to Zidovudine-therapy, either with oral Co-trimoxazol (5 mg/kg b.wJday, n=4) or inhaled Pentamidine-lsethionate (4 mg/kg b.w. every 4 weeks, n=8). In patients with severe HIV -associated or Zidovudine-induced neutropenia and opportunistic infections we applicate rhu-GM-CSF (100 - 250 pg/qm BSA i.v. or. s.c./day). Dideoxyinosine (DDI) was administered in patients with Zidovudine-intolerance. -toxicity and -refractory disease (2 x 67 mg/day). Results: In the first two ys without PcP-prophylaxis we observed 3 cases of PcP. Since initiation of PcP-prophylaxis no PcP occured. Until now two children with opportunistic infection and neutropenia (< 1000/4d) have been successfully treated with rhu GM-CSF. According to this graduated therapy concept (observation period 1988 - 1991) we have observed 3 CMV- or EBVassociated infections (2 LIP's. 1 CMV-myelo-retinopathy). We noticed 3 children with HIV-encephalopathy; 2 of them responded well to Zidovudine-therapy, the third one died due to progressive HIV-encephalopathy. 4 of 14 children staged PI B in 1988 progressed to P2. 10 of 14 have not shown progression of their disease for over 2 ye. 3 of 6 children staged P2 developed full blown AIDS, 3 did not show a change since 1988. 1 of 3 patients with full blown AIDS remained stable. 2 died after a duration of 2,6 resp. 5 ys of AIDS due to CMV-myelopathy and cardiomyopathy. DDI seems to be of benefit in Zidovudine-intolerance, -toxicity and -refractory disease. y Supported by "Federal Ministry for Youth, Family affairs, Women and Health', Germany (BMJFFG, BMAS) NOTES TU 8 ITALIAN EXPERIENCE ON EFFICACY AND TOLERANCE OF ORAL ZIDOVUDINE (AZT) lU.~ IN SYMPTOMATIC HIV INFECTED CHILDREN Elia L.,Campelli A.Caselli D.,Castelli G.,Palomba E.,Principi N.Terragna A.Timpano C.Zuccotti G.and the italian pediatric Zidovudine trials group OBJECTIVE: A nationwide programme has been set up to investigated the tolerance and the efficacy of oral AZT administration in HIV infected children. METHODS: In 32 centers partecipating to the programme 167 patients (3 months to 8 years old) with symptomatic HIV infection (CDC-P2) have been enrolled in an open phase II trial. RESULTS: Up to now we evaluated 100 children (53 M, 47 F) with vertically trasmitted HIV infection, treated for more than 24 weeks with oral AZT (dosages from 300 to 720 mg/mq). Mean age was 34 months (range 4-87),and mean time on treatment was 44.7 weeks (range 4-116). At entry 45 were AIDS, 22 ARC and 33 P2-A. Out of the 12 patients that died, 11 were AIDS at entry. Levels of IgG and IgA were elevated at enrollment in 75% afd 50% of patients. Significant reduction of concentrations were seen respectively in 80% and 66%. 44 children had CD4+ cells <500 at entry (mean 210), 16/44 (34%) become >500; also 18/29 with persistent lower value showed a transient rise. 41/64 children had p24 antigenemia at entry, in 21/41 (51%) it was no longer detectable and in 2/41 returned positive. 7/11 persistent positive showed marked reduction in concentrations. 7/23 negative children became positive. leran: Possible clinical adverse effects occurred in 33 (33%). 58 % had a decline in Hb to <8 gr/dl requiring either a dose modification or transfusions; neutropenia was observed in 17%. Total number of reported adverse experiences was 77, in 17 was more than one. CONCLUSION: Children treated with AZT showed benefical effects using as markers survival,immunological function and p24 antigen.Matched controls are not available for comparison, but the survival of children appears to be higher than expected. Anemia and neutropenia was the most common laboratory adverse effects noted but the freauencv of those effects was sliqhly lower in children than in adults. NOTES CI ~iCI cr t^

Page  83 TU.C.99 EFFECT ON TIME TO AIDS OF SYMPTOMS, IMMUNE ACTIVATION MARKERS AND PROPHYLAXIS AT DIFFERENT LEVELS OF CD4 Mufoz, Alvaro, Hoover D, He Y, Detels R, Kingsley L, Graham NMH, Vermund SI, Phair J, for the Multicenter AIDS Cohort Study (MACS), Bethesda, MD, USA. Objective: Determine the extent to which thrush, other clinical symptoms(fatigue, fever, weight loss, diarrhea), prophylaxis(Zidovudine and/or Aerosol Pentamidine), neopterin and /2microglobulin modify the effect of CD4 cell count in predicting AIDS-free time. Methods: Maximum likelihood estimates for parameters of life-table regression models were obtained. Procedures were developed to use 12,691 longitudinal measurements collected semiannually on 1,996 IIV positive homosexual men. Robust methods incorporating dependence of within individual data assessed variability of estimates. Results: The Lognormal model explained the distribution of time to AIDS better than did Weibull and Gamma models. Using the lognormal model, regression coefficients and standard errors for the log of the median of the time to AIDS in years were: Constant Main Effects Interactions X1 22 X 3 X4 s 5 5 27 xx 5 a26X 5751X Coefficient 2.37 1.24 -0.49 0.15 1.41 -0.47 -0.33 0.09 -0.12 -0.03 -0.40 Standard Error 0.05 0.06 0.02 0.52 0.10 0.05 0.15 0.09 0.04 0.09 where z,=log(CD4#/500), z2=log(CD8#/800), z3=(hemoglobin-15) in g/dl, z4=( platelets-0.25) in millions, s5=thrush, z6=total number of other clinical symptoms(0 to 4), z7=prophylaxis. Median for the reference (all x=0) was 10.7 years. Conclusions: Prognostic information of thrush was stronger for high levels of CD4. Using the other clinical symptoms individually was as informative as using the total number and their effect was the same for all levels of CD4. In this cohort study, the protective effect of prophylaxis was stronger in lower categories of CD4. Similar analysis found high levels of neopterin and Pzmicroglobulin associated with shorter time to AIDS particularly for low levels of CD4. Neopterin exhibited a greater interaction with CD4, NOTES TU.C.100 THE CORRELATES OF SURVIVAL AFTER DIAGNOSIS OF AIDS. Saah. Alfred; Hoover, I D; Munoz, A; Detels, R; Phair, J; Rinaldo, C; Vermund, SH and the MACS, Bethesda, MD, USA. Objective: Identify correlates of survival after the diagnosis of AIDS in gay men. Methods: Seropositive gay men were studied before and after the diagnosis of AIDS. Variables collected included CD4 number, initial AIDS diagnosis (PCP, KS, other single diagnosis, both KS & PCP) thrush, fever, and AZT therapy. Analysis was performed using the Cox porportional hazards survival model. Pre-AIDS AZT therapy and post-AIDS AZT were considered as separate mutually exclusive variables with post-AIDS AZT being a timedependent covariate, Results: Relative hazard (RH) is a measure of the likelihood of death. Values <1.0 indicate longer survival. The reference group for RH of death after PCP, KS and other single diagnoses is the both PCP & KS group. Other Pre-AIDS Post-AIDS PCP KS Single CD4* Fever Thrush AZT AZT RH.902.884 1.385.838 1.800 1.350 1.168.503 P-Value.402.375.034 <. <.001 *(Per 100 Cells) Conclusion: The risk of death, after controlling for CD4, thrush, fever and AZT therapy, is similar for those with single diagnosis of PCP or KS. Fever, thrush, and CD4 count are still helpful prognostically, even at this late stage of infection. Post-AIDS therapy almost halves the hazard of death after onset of AIDS. Clinical trials have shown that AZT delays the onset of AIDS. In this cohort study, taking AZT prior to AIDS failed to show improved post-AIDS survival. - NOTES C 0 I~ 0 Q) K C3 TU.C.101 DIFFERET RATES OF CD4+ T CEM W IE IC LATMED 1 VI~ VARIANS AND CLIICAL OUITME IN HIV-1 SEROPOSITIVE MEN Schellekens, Peter*,***; Tersmette, Thijs*; RBos, Marijke*; Keet, Rene**; de Wolf, Frank***; Miedema, Frank*. *Central lab. Netherl. Red Cross Blood Transf. Service & lab. of Exp. and Clin. Immunology, University of Amsterdam, **Municipal Health Service, ***Acade55ic Medical Centre, Amsterdam, The Netherlands Objtactive: To estimate rates of decline of CD4 T-cell numbers in non progressors and progressors to AIDS-OI and to correlate these rates with different HIV-1 isolates. MeLthos: The rate of decline was fitted by linear regression analysis. Linear trend analysis related to the different HIV subtypes was calculated in each single individual over the period that a particular virus type was present. Three groups of virus isolates were characterized: 1. Syncytium inducing, high replicating, transmissible to the H9 T-celline (SI) 2. Non syncytium inducing, high replicating, not transmissible (Fast-NSI) 3. Not syncytium inducing, low replicating, not transmissible (Slow-NSI). R ts: In a cohort of 196 HIV positive homosexual men, 53 patients had progressed to AIDS-OI after a mean follow-up of 54 months. In the non progressors the decline of CD4' cell numbers was steady and continuous (0.006x109/1 per month). The decline in the progressors was biphasic: until 18 months before diagnosis with a rate comparable to that in the non progressors, thereafter 3-5 times faster. The decline of CD4+ cell numbers related to HIV subtypes showed that the fast decline was associated with the presence of SI variants and the slow decline with Slow-NSI and Fast-NSI variants. Conclusion: Apparently a critical pathogenic event 18 months before diagnosis determines progression to AIDS. This critical event may be the change within the individual to a more virulent HIV subtype. TU.C.102 A COLLABORATIVE STUDY ON KAPOSI'S SARCOMA AND NON-HODGKIN'S LYMPHOMA AMONG AIDS PATIENTS IN ITALY. Greco Donato*, Tirelli U.**, Zaccarelli M.*, Benedetti P.* Serraino D.**, Errante D.**, Vaccher E.** *AIDS Unit, Istituto Superiore di SanitA, Rome, Italy; **G.I.C.A.T., Aviano, Italy. Objective: To evaluate the association between Kaposi's Sarcoma (KS) and Non-Hodgkin's Lymphoma (NHL) and HIV transmission risk-groups; to describe the prevalence over time of KS and NHL among AIDS patients in Italy. Methods: Statistical analysis was carried out on data from 8,068 cumulative AIDS cases, reported to the Italian National AIDS Registry by November 30, 1990. Odds ratios (OR) and 95% confidence intervals (Cl) were calculated. Results: At AIDS diagnosis, 563 patients (7%/) were found with KS and 320 (4%) with NHL. Homosexual males were at the highest risk of KS (OR=10.1, 95%CI: 8.4-12.1) and homosexual male injecting drugusers (IDUs) had a five-times higher risk of KS (OR=4.4), both compared to heterosexual male IDUs. Homosexuals also had significantly higher risks of KS than did haemophiliacs (OR=7.0) and blood transfusion recipients (OR=7.2). Among heterosexuals, the risk of KS was higher for men (OR=2.8) than for women (OR=l.8). The risk of Burkitt's lymphoma for homosexuals was double than for IDUs (OR=2.0, 95% CI: 1.0-4.1), while in homosexuals the risk of brain lyphoma was significantly lower than in IDUs (OR=0.3, 95% CI: 0.1-0.9). The proportion of AIDS patients with KS decreased from 9.8% in 1985 to 5.3% in 1990, and such a trend was particularly evident in homosexuals. IDUs with NHL decreased from 6.0% to 2.8%, while a slight increase was seen in homosexuals. Discussion and Conclusions: KS is more common in persons who acquired HIV infection through sexual transmission than in those who were infected through parenteral transmission. The frequence of KS is decreasing over time. The study also suggests that Burkitt's lymphoma and brain lymphoma may have different etiologic mechanisms. Further epidemiological research in this field is needed to determine possible cofactors in AIDS-related tumors.

Page  84 ri n c INCUBATION TIME TO SYMPTYMATIC DISEASE (SD) AND AIDS IN WOMEN WITH A S U.-C.103 KNOWN DURATION OF INFECTION Anzala Aggrey O*, Pluimner FA*,***, Wambugu P*, Nagelkerke NJD*, NdinyaAchola JO*, Bwayo J*, Ngugi EN**. *U. of Nairobi, Kenya, **Kenya Medical Research Institute, Kenya, ***U. of Manitoba, Canada. Objective: To determine the rate and risk for development of HIV-1 related disease in a population of women with a known time of seroconversion. Methods: A cohort of 163 women working as prostitutes in Nairobi with a known time of seroconversion have been followed for the occurrence of clinical illness related to HIV-1 infection at 6-month intervals for 6-60 months. Survival analysis was performed using the log rank, Kaplan-Meier and Cox proportional hazards with time dependent covar iates. Results: Of 163 seroconverting women identified, 144 continue in follow-up. Fifty women have developed SD, 34 have developed AIDS and 11 have died. The Kaplan-Meier estimates of the median time to the occurrence of SD and AIDS were 34.2 and 44.6 months, respectively. On univariate analysis, increased risk of development of SD and AIDS was associated with gonococcal (Gc) infection. Condom use was associated with a reduced risk of development of SD and AIDS by univariate analysis. However, on Cox proportional hazards modeling with time dependent co-variates, only Gc infection remained independently associated with development of AIDS (RR=2.12, 95%CI=1.51-3.0, p=.0001, for each unit increase in rate of Gc infection). Conclusions: The incubation time to SD and AIDS is very short in this group of women. This may be related to their sex, the sexual exposure as prostitutes, or unidentified risk factors unique to Africans. Although Go infection is as ociated with increased risk of development of AIDS, the direction of cause and effect cannot be presumed. NOTES S.lon PROSPECTIVE STUDY OF THE NATURAL HISTORY OF HIV-2 TU.MC.1 Marlink. Richard*; Thior, Ibou**; Dia, Mamadou Cire **; Gueye, El Hadj**, Ndoye, Ibra**; Essex, Max*; Mboup, Souleymane**; Kanki, Phyllis* * Harvard School of Public Health, Boston, MA, USA, ** University of Dakar, Dakar, Senegal. Objective: To determine the natural history of HIV-2 infection, especially as compared to HIV-1 infection, by comparing the incidence of clinical and laboratory abnormalities between seropositives and seronegative women. Methods: We have enrolled and followed female prostitutes visiting a nationally supported STD clinic in Dakar, Senegal, since 1985. HIV-2+, HIV-1+ and HIV-1/2+ prostitutes have been stratified (1:2 ratio) with HIV- prostitutes by age, nationality and year of clinic registration. In Dakar, with 9.5% HIV-2 and 1.8% HIV-1 overall seroprevalence, we have collected repeat clinical data 354 on prostitutes. In a satellite STD clinic in Ziguinchor, Senegal, with a 34% seroprevalence for HIV-2, we have attempted to enroll all the registered prostitutes since 1986, now totalling 196 prostitutes. DTH skin testing and lymphocyte subset determinations have been obtained. Results: HIV-2+ HIV-1+ HIV-1/2+ HIV negative Overall Clinical Outcomes N = 166 N = 24 N = 5 N = 355 Person-Years Observed PYO= 399 PYO = 61 PYO = 9 PYO = 926 ARC 1 2 0 0 AIDS 1 2 1 0 By defining "lost to follow-up" as those prostitutes not seen in the past 12 months or those without travel history of leaving the country with a healthy status, we have been able to account for 90% of those prostitutes initially being enrolled in our clinical cohort. The mean T4 values on repeat testing of a subset of HIV-2 seropositives were not significantly different one year later. Discussion and Conclusions: To date, there is a significant difference in disease incidence in this cohort when HIV-2+ versus HIV-1+ individuals are compared. Evaluation of immunologic abnormalities over time appear to parallel these observations. NOTES Cd ro I2 CS 0 '"* ~~t

Page  85 TU.D.105 A PSYCHOLOGICAL STUDY OF LONGTERM SURVIVORS OR AIDS. Katoff, Lewis*; Rabkin,J**;Remien, R**; *Gay Men's Health Crisis, NY, NY, USA; **HIV Center for Clinical and Behavioral studiesColumbia University, NY, NY, USA lObiective: To determine the psychological consequences of prolonged illness and stress; land to identify factors associated with the maintenance of hope and quality of life. Method: Subjects in the study were male clients of an AIDS service CBO. A semistructured interview of approximately 90 minutes duration and several self-rating scales were administered to 53 men who had been diagnosed with an opportunistic infection at least three years before their participation. The interview included questions covering demographic variables, current health status, reactions to illness onset, mental status and mood, current activities, stressors and coping strategies. Standardized scales included measures of social functioning, hopelessness, social support and anhedonia. Results: The average age of study participants was 39 years old; and 75% were not originally from New York City. The most frequently mentioned problem was the restrictions on activities and the great effort required for even simple tasks. Subjects reported frustrations dealing with bureaucracies but a high degree of satisfaction with their current physician. Conversely, respondents reported that their illness led to a greater appreciation of relationships and that they enjoyed things more. Respondents commented on the importance of being well informed. Conclusions: As a group, these individuals were remarkably pragmatic and very realistic about their illness. The absence of denial was very striking; these individuals did not elieve that they would beat the illness, and they were acutely aware that their horizons were foreshortened. These individuals viewed their illness as both a burden and a hallenge; and most reported a sense of accomplishment about their lives and specifically NOTES TU.D.106 THE POTENTIAL FOP THE SEXUAL TRANSMISSION OF HIV: HEROIN AND AMPHETAMINE INJECTOSP COMPARED Klee, Hilary, Manchester Polytechnic, Manchester, United Kingdom Amphetamine sulphate is the stimulant most widely used in the UK and elsewhere in continental Europe. It is cheap, availahle, and injected by an increasing number of the younrg. Like cocane, it is reputeed to facilitate social and sexual interactions, in stark ciontrast to heroin. The aim of this study was to ePxplore differences in the sexual behavior-u of heroin, and amphetamirne usercs and to evaluate their potential fortrarnsmitting HIV to their sexual partners, Methods Research assi-tarint interviewed 100 amphetamine and 142 heroin injectors contiated through other- drug users and drcug ageprcies.The samples were comparable with respect to gender and level of drug use. Data on injecting and sexual behaviour, and attitudes to HIV-related risk were subjected to contingency and regression analyses. Results A marked difference in perceived vu..lnerahility to HIV infection was observed beweer groups. Eighty-one percent of heroin users elieved themselves to be at risk through sPe in contrast to 20% of amphetamine users, despite the significantly greater sexual activity of the latter. Casual sexual ercounters were significantly related to age among amphetamine users (p<-001) the major difference lying in the under 20 age group where 95% reported casual sex> in the 6 months prior to interview compared to 47% of the heroin group. Both groups displayed negative attitudes towards condoms. Discussion and Conclusions Differences in sexual behaviour will be described, and prospects for risk reduction explored. Very few amphetamine users were in contact with drug services and they do not feel they ar a 'risk group'. The potential for the sexutal trarrsmisrnion of HIV is greater among these stimulant users than among heroin users and harm reduction efforts need to be directed specifically at them. NOTES 0 CI Ih C1 TU.D.107 ARE BISEXUALS UNDER-RECOGNIZED AMONG WHITE MEN WITH AIDS IN THE U. S.? Fulton, Robert; Kennedy, R. E., Jr. University of Minnesota, Minneapolis, MN, USA. Objective: This study proposes to show that adult male AIDS cases of "undetermined HIV risk exposure" are disproportionately covert bisexuals, especially in recent years (post 1987), and among AIDS cases reported outside large metropolitan areas. Methods: This research is based on the Centers for Disease Control's AIDS Public Information Data Set of every AIDS case reported through June, 1990. This data base gives the year of diagnosis or reporting of AIDS cases and separates bisexual from homosexual men and metropolitan from rural areas. The criteria used to select subjects were: 1) non-Hispanic white males; 2) twenty years or older at time of first diagnosis; 3) only one HIV exposure risk known (except for bisexual or homosexual IV-drug users; 4) born in the U. S. The 67,480 men with AIDS thus selected constituted 91.7% of all nonHispanic white adult/adolescent AIDS cases reported in the U. S. through June, 1990. The outcome variable was the "HIV exposure" category. Six predictor variables were used for all cases, plus a seventh for men from large metropolitan areas. The "uncertainty coefficient," a nominal-level measure of association, was used to measure the amount of similarity between pairs of groups. Results: Of all HIV risk groups, the non-IV-using bisexual men were most similar to the AIDS cases of "unidentified HIV risk exposure" 1) in large urban areas both before and after January 1, 1988, and 2) outside large metropolitan areas in recent years. Conclusions: Many of the "undetermined HIV risk exposure" AIDS cases are covert bisexuals. Given the current annual +91.2% increase in male "undetermined HIV risk C exposure" cases, we foresee a new stage in the spread of AIDS to non-tv-drug using 00 women. TU.D.108 PHASE I RESULTS OF A MULTI SITE STUDY OF THE HETEROGENEITY OF SEX WORK IN MEXICO, THAILAND, AND ETHIOPIA: IMPLICATIONS FOR TARGETTING BEHAVIORAL INTERVENTIONS. Supulveda J*; Zewdle D**; KaufmanJoan***; Sittitral W****Hemandez M*;de Zaduondo B*****;Bishaw M******;Chen L***;*General Directorate of Epidemiology, Mexico DF, Mexico*;**Natlonal Research Institute of HealthAddis Ababa.Ethlopia,***Harvard School of Public Health, Boston,MA.USA,*"*Program on AIDS of Thai Red Cross Society, Bangkok, Thaliand,*****Johns Hopkins University,Baltimore,MD,USA ******Addis Ababa University, Ethiopia QObectlve: To strengthen efforts to transfer results of sucessful pilot models, a multi site intervention research project on prostitution and HIV transmission (MIRPHT) was begun in Mexico, Thailand and Ethiopia in 1990. We aimed to develop and test an analytic approach to assess and respond to heterogeneity in: a)actors; b)settings; c)practices; d)motivations, expectations and resources of commercial sex workers (CSW) through design and evaluation of targetted interventions. Methods: We use a common conceptual framwork and methodology in all three sites. Geographic mapping to identify work places In each setting has been followed by ethnographic investigation: in depth interviewing with key informants, focus groups with subgroups of sex workers classified by workplace and socioeconomic level, then a baseline KAPB survey. Besults: Phase I results suggest: 1)Heterogeneity in forms of sex work exist both within and accross sites and is important for determining risk status and epidemiological projections; 2)Superficial similarities In workplaces (e.g.street, bar) cross culturally are inadequate indicators of intervention needs; 3)Descriptions of underlying dimensions--such as economic need, autonomy in client selection, relationship to manager, and culture/class specific gender role ideals--are useful for defining subgroups of FCSW and clients who require separate interventions, and for designing content and format of those interventions.

Page  86 TU.D.109 THE RELATIONSHIP OF ATTITUDES AND BEHAVIORS RELATING TO PREGNANCY AND THE 00 USE OF CONDOMS BY HETEROSEXUALS TO PREVENT AIDS AND OTHER STDS O\ Endias, Robert*; Minkoff, Howard*; Fleisher, Jay* * State U. of N.Y. Health Science Ctr. at Brooklyn, Brooklyn, NY, USA Objective: To assess the frequency and determinants (e.g., attitudes and behaviors related to pregnancy) of condom use among women in a community endemic for HIV. Methds: 1,393 sexually active women using reproductive health clinics at 3 central Brooklyn sites were interviewed on demographics, sexual, drug using & contraceptive histories, & attitudes and believes about STDs, pregnancy, contraception & sexuality. Results: Among non-pregnant women who were not trying to get pregnant (n=888), condoms were used during 25.2% (sd, 39.1%) of incidents of vaginal intercourse with main sexual partners (MSP). They were used in 9.5% (sd, 28.6%) of incidents of anal intercourse (p=.001) and in 6.3% of vaginal with MSP among pregnant women (n=363)(p=.0001 pregnant vs. non-pregnant). Women at low pregnancy risk because they were sterilized or using IUDs were significantly less likely to use condoms than women using all other forms of contraception. Condom use was significantly associated with negative attitude toward pregnancy, with women's sense of AIDS seriousness, and with felt susceptibility to AIDS, but not associated with felt susceptibility to STDs in general. Similarly, perception of partner's attitude toward pregnancy, condoms and AIDS was significantly associated with condom use, but perception of partner's felt susceptibility to STDs was not. Conclusions: 1) When fear of pregnancy is less of an issue (e.g. already pregnant, anal intercourse, surgically sterilized) condoms are less likely to be used; 2) Attitude toward pregnancy is an important determinant of condom use; 3) Caution should be exercised before advising dual contraception for women (e.g. condoms for disease prevention, IUD for birth control) lest the contraceptive reduce incentive to use condoms. NOTES TU.D.111 MOBILIZATION OF WOMEN IN THE STRUGGLE AGAINST AIDS IN URBAIN AEREAS: ACTION AMONG YOUNG PEOPLE. BISHAGARA K.* KASALI M.* TULIZA M.* MULANGA K.* GRANDPIERRE J.**; PIRI PIRI L.* * SOCIETY FOR WOMEN AND AIDS IN AFRICA / Section Zaire (SWAA/ZAIRE) ** UNICEF /ZAIRE *** BCC/SIDA ZAIRE OBJECTIVES: To provide pupils with education to life for a better understanding of their sexuality. To widen their knowledge on sexually transmitted diseases (STD) s~ch as AIDS, and teach them preventive methods. To evaluate the impact of our educational intervention among young people. METHODS: Thirty members of SWAA / ZAIRE have been trained to officer pupils of a secondary school (from 2nd to 5 th class secondary school) in KliTshasa. Talks are help informally; each sponsor keepsbusy 20 pupils (i.e. 2 groups of 10 pupils for one hour per week during two months talking about: sexuality, STD, AIDS and diseas prevention. To make a pre-test before the educational intervention and a post test aft~ the intervention. RESULTS: A very big impact. Pupils show a very big interest by spreading the informatip even out of the school frame. Very lively exchanges, positive modifications ofknowledg and attitude towards sexual education in what concerns AIDS. CONCLUSION: The message of young people to then school friends has a particular interep This intervention of SWAA makes it easy the mobilization of pupils who will soon be promotors in the struggle against AIDS in neighbouring communities starting from their own schools. TU.D.110 HIV RISK BEHAVIORS AMONG ADOLESCENTS/YOUNG ADULTS IN THREE COUNTRIES Albrecht. Gary L.*; Wells, J.A.**; Valleron, AJ.***; *University of Illinois at Chicago, U.S.A.; **Project Hope, Washington, D.C., U.S.A.; ***INSERM, Paris, France Objectives: This paper examines HIV risk behaviors in three countries among adolescents and young adults. Using data collected in France, the U.K., and the U.S., we: (1) estimate the levels of high-risk sex and drug-use; (2) compare patterns of risk behaviors; and (3) estimate riskreduction practices. Methods Using nationally representative samples, personal interviews were conducted with 856 adolescents and youths aged 16-24 in France, 634 in the U.K. and 539 in the U.S. in a comparative study of HIV risk-behavior. Logistic regression was used to compare levels of risk across countries and estimate changes in risk practices. Results Approximately 28.5% of the adolescents and young adults perceived themselves to be at risk for AIDS. Despite knowledge about and perceptions of risk, persons aged 16-24 persist in experimenting with risky behaviors such as having sex with multiple partners (29%), high-risk vaginal intercourse (10.5%), anal intercourse (5%), and drug injecting (4.4%). Risk behaviors are least prevalent in the United Kingdom. Behavior change among at-risk adolescents and young adults is reported by reducing partners (42%), adopting condoms (34%), and stopping drug injecting (20%). Discussion and Conclusions: The results indicate that adolescents and young adults continue to experiment with sex and drug-use behaviors that put them at risk for AIDS despite knowledge and perceptions of risk. These findings suggest that intervention programs that merely educate young people about risk behaviors are inadequate in producing behavioral change. Risk-reduction strategies are likely to succeed only if they intervene in the social networks of these risk takers to change norms and behaviors. NOTES TU.D.112 SURVIVAL PATTERNS OF WOMEN AND MEN WITH AIDS: IMPACT OF HEALTH CARE USE PRIOR TO AIDS. Turner, Barbara J*; Markson, LE*; McKee, L*; and Fanning,T**. *Center for Research in Medical Education and Health Care, Jefferson Medical College, Philadelphia, PA, and **New York Department of Social Services, Albany, NY, USA. Objective: The Centers for Disease Control reported higher short term mortality rates for women after AIDS diagnosis, raising concerns that women may have poorer access to care than men. Methods: We examined this issue by determining patterns of survival of AIDS patients depending upon their health care resource use in the months prior to AIDS diagnosis (as a proxy for access to care). Our data include clinical and resource use histories before and after AIDS onset of 1272 men and 632 women aged 13-60 years and enrolled in New York Medicaid in 1983-87; analyses on 1988 data are forthcoming. Proportional hazards models were used to explore the influence on survival of age, sex, risk group (intravenous drug user (IVDU) vs. non-IVDU), severity of AIDS-defining diagnosis (with 3 levels based on a model developed from expert opinion), and total charges for health care (low, moderate, high) within 3 and 6 months prior to AIDS diagnosis. Results: We found that charges before AIDS diagnosis varied by both sex and risk group, with women lower than men for the same risk groups (p<0.001). Average 3-month charges were: $5450 for IVDU men; $4300 for IVDU women; $3570 for non-IVDU men; and $2950 for non-IVDU women. Six-month charge patterns were similar. In multivariate models, the most important factors influencing survival were severity of AIDS diagnosis (95% confidence interval (CI) = 1.6-2.7, high severity vs. low), age (95% CI = 1.1-1.6, 30-39 vs. 20-29 years), and charges 3 months prior to diagnosis (95% CI = 1.2-1.8, high charges vs. low). Conclusions: While women on Medicaid may have lower use of health care resources prior to AIDS, this does not appear to have a major influence on survival. Instead, survival is primarily dependent on severity of the AIDS-defining diagnosis and age. b C3 K^) b fV Q3 O O C) con e 4 I I ~t3

Page  87 TU D 113 Women and AIDS in Mexico: A Socioepidemiological Approach. SValdcspino.L., Dl Rio. Aurora Garcia M.L., Morales R.A., Basafie R.A., Magis C., Scpulvcda J. General Directorate of Epidemiology, Ministry of Health. Mexico. Objctive. To analyse the specific profile of knowledge and attitudes concerning AIDS, the sexual practices and HIV- I scroprevalence among Mexican women, as well as, the characteristics of cases reported among this group in Mexico., Methods. Analysis of dataobtained through 1,984 KABP surveys in 6 cities, 5,367 sentinnel surveys in 18 cities and 837 female cases reported to the National AIDS Register. Surveys include data from women in the general public, College students, health workers and female prostitutes. Differences in knowledgeand attitude scores were identified through ANOVA. Scroepidemiologic surveys and case register data were analysed through epidemiological analysis of rates and risk factors. Rcs.li. Mexico has a female population of 42 million of whom less than 50% are under 25 years old, this way large numbers of women are potentially at risk of HIV-1 infection. Confidence intervals (95%) of the knowledge scores among different groups of women are as follows: adult women from general public (5.7 - 9.8), female prostitutes (6.2 - 9.8) health services workers (7.9 - 8.9) and female College students (7.9 - 9.5). Average scoreon appropriate attitudes ishigher among female prostitutes (39) while lower scores were found among women from the general public (20). Sexual practices among women who had started their sexual life showed the following pattern: 9% of College students reported more than one sxual partner in the four months previous to the survey, while only 5% of the adult women in general public and i % of those who work at health services had 2 or more sexual partners. Naturally, average number of sexual partners is higher among female prostitutes. Condom use rate in this groups' partners is also the highest (47.8%), meanwhile lower rates were observed among parners of College students (21.7%) and the lowest ftor the partners of general public wonmtn (5.4%). Only prostitutes have shown a growing trend in condom use. HIV-l seroprevalences found in womuienarc s follows: 27.5% among saxual partnersof HI V positive men, 6.6 % for women attending information and detection services, 0-5.2 % for prostitutes, 0-4.6 % in tuberculosis female patients,< 1% in prisoners and <0.05% in blood donnors. A total of 837 AIDS cases in women have been reported up to Dcccmbcrl990. In the past 12 months 16% of reported cases have occurred among women; 63% of them arc housewives and more than 50% are between 25 and 44 years old. Regarding transmission ways, 65% of female cases were transmitted through blood or its derivates, 28% heterosexually, 6% through perinatal transmission and < 1% in IVDU. Conclusions. An increase of HIV-1 transmission in women is mainly a result of infected blood transfusions related to obstetric causes, however heterosexual transmission tosexual parmersof scropositivemen(biscxual men and blood recipients is also rclvant).Theren ust be more emphasis on actions directed to Mexican women in AIDS education programs. Most young and adult woman show static levels of attitudes and AIDS preventive behavior however, prostitutes have shown a growing rate of condom use and appropriate attitudes even though they have a low level of knowledge about AIDS. NOTES TU.D.114 PARTICIPATION OF WOMEN IN A MULTICENTER HIV CLINICAL TRIALS PROGRAM IN THE UNITED STATES Cotton, Deborah* **; Feinberg, J***; and Finkelstein, D*. *Statistical and Data Analysis Center, AIDS Clinical Trials Group (ACTG), Harvard University, **Beth Israel Hospital, Boston, MA, USA; and ***Johns Hopkins University, Baltimore, MD, USA. Objective: To examine patterns of accrual of women to ACTG trials. Methods: Demographic trends in accrual were analyzed by ACTG site and ACTG protocol and summarized by descriptive statistics, categorical and regression methods. Results: Women have accounted for 6.7% of 11,898 adult ACTG study entrants. The number of female study entrants (FSE) has increased each year (130 in 1987 to more than 250 in 1990) but is plateauing. The % of FSE who are non-white is less than the % of females reported nationally with AIDS (FAIDS) who are non-white (52% vs. 73% p<.001) as is the % FSE who have used intravenous drugs (23% FSE vs. 51% FAIDS p<.001). Accrual of women varies widely among individual protocols (range 0 -38.5%, median 4.9%) and types of protocols. At 32 adult sites, female enrollment has ranged from 1-37% and only 6 sites have had > 30 FSEs. ConcIluion: Accrual of women to ACTG trials has increased over time but remains low and is plateauing. FSE differ in ethnicity and risk factors from FAIDS. Further analysis of trends in demographic, geographic and protocol variation in female trial participation is needed. Examination of enrollment patterns may be useful in designing strategies to increase trial access for women. NOTES TU.D.115 WOMEN AND AIDS: A HUMAN RIGHTS PERSPECTIVE Hausermann, Julia; Danziger, Renee Rights and Humanity, London, United Kingdom Objective: To illustrate the relationship between a)women's low social status in the family and society, and the lack of protection of their fundamental human rights and b) women's increased risk of HIV infection and the impact of the epidemic on their lives Method: Unstructured interviews in United Kingdom, Uganda and Zambia with HIV-infected women, women health care workers, female partners of infected men, women caring for family members with AIDS and representatives of women's organisations, AIDS service organisations and NACs. Results: Interviews confirm that women's low status and lack of rights increase their vulnerability to the risk of infection, and to the personal, social and economic consequences of the epidemic. Discussion and Concluiopns: Even where women are informed on how to avoid HIV infection their traditionally subordinate role within the family and society, combined with their economic dependence on men, have in many cases prevented them from refusing unwanted, and often risky, sexual intercourse. Additionally, the lack of protection of women's human rights, including social, economic and legal rights, has not only heightened their subordination and dependence on men, but has also served to worsen the personal and social consequences of AIDS on their lives, eg in their role as carers. These factors must be taken into account by AIDS prevention and care programmes if they are to help women andý-nhance their contribution to the prevention of HIV transmission. TU.D.11 ITALIAN SEROSURVEY OF HIV INFECTION IN PARTURIENTS: AN UPDATE Stegagno, M i chele; Ippol ito,G.; Costa,F.; Angeloni,P.; Aebischer, M.L.; Angeloni,U.; Guzzanti,E.- Italian Collaborative Study Group for HIV Prevalence in newborns. - Coordination: AIDS Unit, L. Spallanzani Hospital-Rome, Italy Objective: In Italy the high number of pediatric AIDS cases reflects the rate of HIV infection in females in reproductive age. Therefore it is of primary importance to assess the prevalence of HIV infection among parturients in order to estimate the future rate of pediatric AIDS cases and to monitor the trend of the infection. Methods: Between June '88 and December '90, 144,462 blood samples collected for newborns' metabolic screenings in several hospital nurseries from different Italian Regions were examined for anti-HIV antibodies (HIV-Ab). Blood saturated disks were punched out from the collection papers, without identification and screened for HIV-Ab in ELISA (Pasteur); the positive results were confirmed in Western Blot. Results: Among the 144,462 blood samples examined 198 (0.00137 Confidence Intervals 95%, Poissons Distribution:.00119--.00158) were found positive for HIV-Ab. In 1988: 18 samples among 20,682 tested turned to be positive (0.00087, C.I.:.00052--.00138), in 1989, 91 out of the 55,406 samples examined were positive (0.00164 C.I.:.00132 --.00202), while in 1990, 89 out of 68,374 samples tested resulted reactive (0.00130, C.I.:.00105--.00160). The lower prevalence of HIV infection observed during 1990 is not significant and it is likely due to the enrolment in the study of a higher number of Regions with low incidence of AIDS cases. Conclusion: The serosurvey for HIV-Ab on newborns provides data in the evaluation of HIV prevalence and can represent an useful tool to monitor the trend of HIV infection in females in reproductive age. Work supported by Italian Ministry of Health - AIDS Research Project

Page  88

Page  89 .........

Page  90 C V

Page  91 ~II

Page  92 nJ M.A.1000 IN VITRO HIV-1 INFECTION OF HUMAN BONE MARROW EOSINOPHILS Freedman. Andrew; Gibson, F; Fleming, S; Spry, C; Griffin, G. St. George's Hospital Medical School, London, UK. Objective: To investigate the potential of HIV-1 to infect human eosinophils in vitro. Methods: Bone marrow aspirates from healthy human volunteers were cultured in medium supplemented with recombinant human Interleukin-5 (IL-5) to promote eosinophil differentiation. After 1-3 weeks cultures were inoculated with cell-free HIV-1 supernatants and maintained for a total of 8 weeks. At weekly intervals aliquots of culture supernatant were removed for reverse transcriptase (RT), p24 and Jurkat co-culture assays and cells harvested for cytospin slide preparation. Slides were stained by immuno-alkaline phosphatase for CD4 and HIV-1 p24 and gp41 and by in-situ hybridisation (ISH) for HIV-1 RNA. Results: CD4 expression by eosinophil precursors was maximal in the first 4 weeks with a gradual decline thereafter. At 2-4 weeks after viral inoculation eosinophils were positive for HIV by both immunostaining and ISH. Productive HIV infection was confirmed by co-culture experiments and the demonstration of rising p24 and RT activity in supernatants. Discussion and Conclusion: Human eosinophils, produced from IL-5 stimulated bone marrow cultures, express CD4 early in their maturation and are susceptible to infection by HIV-1. The possibility that eosinophils are a target for HIV infection in vivo has important implications for disease pathogenesis. NOTES M.A.1001 DUAL TROPISM OF HIV-1 IN CHIMPANZEE LYMPHOCYTES AND MONOCYTES KpciL, Zaruhi*; Mannhalter, J.W.*,**; Schaff, S.***; Eder, G.*; Eibl, M.M.** *Immuno AG, Vienna, Austria, **Institute of Immunology, University of Vienna, Vienna, Austria; ***Semmelweis Medical University, Budapest, Hungary Objective: To study the cellular tropism of HIV-1/IIIB in the chimpanzee. Methods: Monocytes were prepared from peripheral blood of chimpanzees by adherence to plastic. Prior to infection with HIV-I strain IIIB the adherent cells were cultured for 7 days in medium containing M-CSF (100 U/ml) and GM-CSF (40 U/ml). Nonadherent cells (lymphocytes) were stimulated with PHA for 3 days and maintained in medium containing IL-2(20 U/ml). Both monocytes (5x104 IU HIV/ml) and lymphocytes (1x103 IU HIV/ml) were infected with HIV-1/IIIB, a lymphocyte-tropic strain of HIV-1. Virus replication was monitored by release of HIV-1 into culture supernatants (assessed by antigen ELISA and RT) as well as by electron microscopy. Results: While release of HIV-1 from infected PHA lymphoblasts was detectable on day 3 and peaked on day 7, only very low levels of HIV-1 were released from infected monocytes. However, if infected monocytes were cocultivated with human PHA blasts, transmission of virus was readily detected. Replication of HIV-1/III B in chimpanzee monocytes was further confirmed by the presence of intracellular virus particles as well as by the detection of HIV-1 budding from the surface of these cells. Discussion and Conclusions: Our studies demonstrate that HIV-1/IIIB, which is considered to be predominantly lymphocyte-tropic in humans, exhibits a dual tropism for lymphocytes and monocytes in the chimpanzee. NOTES C3 l~~ I, c3 I^A ~~ M.A.1002 FcRII ALSO MEDIATES HIV ENTRY INTO C04 NEGATIVE ACROPHAGES Nagy,Kroly, Dobronyi I,Ti mr I. Institute of Isotopes,Hungarian Academy of Sciences and Department of Pathology, Semmelweis Medical UniversityX,Budapest,Hungary Objective: To evaluate the role of FcRII(CD32) of CD4 negative human machrophage cells in receptor mediated enhancement of HIV infection. Methods: Membrane markers characterisation of the machrophes cell line DD were done by immunofluorescence (IF), DD cells infected with HIV directly or complexed with HIVspecific antibodies. Viral replication were monitored by IF, reverse transcriptase (RT) assay and electron microscopy (EM). Results: DO cells express CD14 antigens and FcRII (CD32) receptors onto the cell membrane. Cells are negative for FcRI, Leu M3, HLA-DR and also for CD4, even after receptor induction by vitamine D3. Dot blot hybridisation for C04 mRNA was also negative. DD cells were not permissive for direct infection with HIV-1(BC041) and HTLV3B. Treatment of cells by immunocomplexes resulted in moderate CPE but IF and RT assay were negative. EM, however revealed HIV particles in cytoplasmic vacuoles of DD cells treated with highly diluted ( 24.000) viral:immunocomplex. Conclusion: HIV had been endocytosed in form of immunocomplex mediated by the FcRII receptors of CD4 negative DD cells. Because of the low rate viral replication, particle, accumulated in cytoplasmic vacuoles and have not been expressed into culture medium. Observation suggests that similarly to FcRI, FcRII receptor mediated enhancement represents an alternative infectious pathway independent of CD4. MA ROE OF RnTHE CDR3 REGION OF CD4 IN HIV- 1 IN ECI IUN AND CYTIOPATlHOOENES M.A.1 3 DEFINED BY HUMAN- AND CHIMPANZEE-BASED CD4 SYNTHETIC PEPTIDES. ee E Eiden*. D.M. Rausch*, V.S. Kalyanaraman**, K.M. Hwang, E. A. Berger"" and J.D. Lifson*. *NIMH, Bethesda, Maryland, USA, **Bionetics Research Inst., Rockville, Maryland, USA, #Oenelabs Incorporated, Redwood City, California, USA, "*NIAID, NIH, Bethesda, Maryland, USA. Objective: Peptide derivatives of CD4(81-92), a region which includes a portion of the CDR3 domain important in gp120 binding to CD4, were synthesized, based on either the chimpanzee or human sequence. Differential ability of the two peptides to inhibit gp120 binding to CD4, HIV infection, or cell fusion could help to explain the observed differences in human and chimpanzee CD4-positive cells to support HIV infection and fusion, as well as the pathogenic sequelae of HIV infection in the two species. lethods: The potency of 1,4,5-tribenzyl,10-acetyl-TYICEVEDQKEE or -TYICEVGDQKEE to inhibit HIV- 1HTLVIIIB and HIV-1RF-II infection and cell fusion, and gp 120 binding to CD4 were tested using previously published methods (J. Immunol. 145:4072, 1990). Results: Chimpanzee- and human-based CD4(81-92) peptides were equipotent to inhibit 1251 -gp120 binding to CD4-positive cells and HIV-1 infection of CEM-SS cells (IC5s 12-60 uM). The human-based CD4(81-92) peptide inhibited HIV-1-induced cell fusion at 16-32 JIM, while the chimpanzee-based peptide inhibited fusion only at 1250 lMM, The human-based peptide was also considerably more potent than the chimpanzee-based peptide to induce the release of gpl20 from the sur face of cells expressing the HIV-1 gp41/gp120 envelope complex. Conclusions: The CDR3 region of CD4 is involved in HIV- 1 binding to CD4, in HIV- I infection, and In HIV-1-induced syncytium formation. Subtle differences in the structural requirements for gpl20 binding occuring during HIV-1 infection compared to HIV-1-induced cell fusion have been distinguished in the CDR3 region, and may provide a partial explanation for the differences in the biology of HIV infection and pathooenesis in chimpanzee and man. C> a C> c> C>.

Page  93 M.A.1004 ASSOCIATION OF A KINASE ACTIVITY WITH THE C04 MOLECULE EXPRESSED BY HUMAN HYELOID CELL LINES AND MACROPHAGES, AND MODULATION OF CD4 EXPRESSION AND THE CD4-KINASE ASSOCIATION BY CYTOKINES. oBrton Gioroio, Sorio C., Cassatetla N.A. and Rossi F. Institute of General Pathology, University of Verona, Italy. Objective. Since the significance of CD4 expression by human mononuclear phagocytes is largely unknown, we initiated studies aimed to establish whether CD4 expressed by human ayetoid cell lines and acrophageas ts able to deliver a signal. Furtheremore we have investigated regulation of CD' expression on human myeloid cell lines and macrophages. lo.ds The association of CD. with a tyrosin knase activity has been studied by performing in vitro knase assays on anti-CD4 immunoprecipitates from lysates of U937 and ML3 cells, and monocyte-derived macrophages. C04 expression was studied by cytofluorografic analysis; C4 mRNA expression was studied by Northern blotting analysis. Baauils. In vitro kinase assays performed on anti-CD4O lnunoprecipitates from Lysates of U937 and NL3 cells, and monocyte-derived macrophages (MON) demonstrated that, also in these cells, CD4 is associated with a kinase activity able to phophosphorylate enolase, and a polypeptide of about 56 kDa present in the ianunoprecipitate. Immunoprecipitates done with anti-82 chain of leukocyte integrlns Abs, from lysates of U937 cells, or with anti-CD4 Abs, from neutrophils or eosinophi s, did not display any kinase activity. This observation indicated that, as well established for T cells, also in myeloid cells CD4 is associated to a putative tyrosin kinese able to phosphorylate itself. We are pursuing attempts to identify the aminoacid phosphorylated in the 56 kDa, CD4-associated polypeptide as phosphotyrosine, and to identify this kinase as one of the member of the src family of intraceLtular tyrosine kinases expressed in myeloid cells (fgr,fyn,lyn,hck). In fact, we obtained evidence that the 56 kDa phosphorylated protein is not Ick nor src. The kinase activity associated with CD4 in U937 and ML5 cells correlated with CD4 expression. In fact, down-modulation of CD4 on U937 cells induced by phorbol esters was accompanied by a dereased detectability of the kinase activity in anti-CD4 imunoprecipitates. Furthermore, we observed that differentiation of ML3 cells to monocytes induced by tumor necrosis factor-atfa (TNF-a) and interferon gama IFN-y) was accompanied by an enhanced expression of CD4 and a parallel increase of the kinase activity In anti-CD4 imnunoprecipitates. Also treatment of lDM with cytokines (IFN-y, GM-CSF) caused an enhancement of the kinase activity detected in anti-CD4 imunoprecipitates; this phenmwenon did not depend on an enhanced expression of surface CD4. The observation that cytokines which induce differentiation of ML3 cells also induce surface expression of CD4, prompted studies to reveal the molecular basis of this phenomenon, We obtained evidence that TNF-a and IFN-y induce the C04 nRNA in HL3 cells.Conctusjons. The association of a kinase activity with CD4 in myeloid cells indicates that, also in these cells, interaction of C04 with appropriate ligands can generate a signal possibly implicated in stimulation of selective functions. Furtheremore, the evidence that cytokines can modulate CD4 expression and CD4-kinase association in myeLoid cells suggests that the infection of macrophages with HIV in sites of inflamation and infection can be variable. NOTES M.A.1005 BINDING OF HIV-1 gpl20 PROTEIN TO CD4-NEGATIVE COLON CANCER CELL LINES Heyworth. Martin F; Sullivan, KT; Liu, G-H; Kim, YS VA Medical Center and University of California, San Francisco, CA, USA lymphocytes and colon cancer cell lines. Methods: Recombinant glycosylated gp120 was labeled with biotin, by using sullosuccinimidyl 6-(biotinamido) hexanoate. Binding of biotinylated (b) gp120 was then assessed by flow cytometry of lymphocytes or colon cancer cells incubated with bgp120 followed by streptavidin (SA)-labeled fluorochromes. Presence or absence of CD4 was also assessed, by flow cytometry of cells incubated with fluorochrome-labeled anti-CD4 monoclonal antibody (mAb). Results: Binding of bgpl20 to CD4+ lymphocytes was observed at bgp120 concentrations of O0.5Sig per 106 cells. The bgp120 did not bind to CD4- lymphocytes. In contrast, bgpl20 showed equally strong binding to CD4- and CD4+ colon cancer cells (see table). Colon Cancer Mean fluorescence intensity (arbitrary units) cell line SA-phycoerythrin (PE) bgp120 PE-conjugated alone + SA-PE anti-CD4 mAb HT29 3.38 10.72 3.30 (CD4-) SW620 1.68 11.09 13.76 (CD4+) LSG 2.92 10.69 2.94 (CD4-) s \Y hI t^Y Oo 1-A st Discussion and Conclusions: The results suggest that bgpl20 became bound to the colon cancer cells by a CD4-indeoendent mechanism. NOTES A.1 DUAL TROPISM FOR LYMPHOCYTES AND MACROPHAGES IS A COMMON FEATURE OF ALL HIV ISOLATES M.A.1006 A.Valentinl, R. Fredriksson1, J. Albert2, E.M. Feny61, and B. AsjO1 Depts. of Virology, Karolinska Institute1, and National Bacteriological Laboratory2, Stockholm, Sweden AIM OF THE STUDY The aim of the study was to investigate whether HIV isolates show any difference in the capacity to infect and replicate in peripheral blood leukocytes (PBL) and blood derived macrophages (BDM). MATERIAL AND METHODS; PBL and BOM obtained from HIV seronegative healthy blood donors were infected with 50 HIV-1, 12 HIV-2 primary isolates and 8 molecplar clones (MC) derived from HIV-1 isolates. In addition, the HIV-1 isolates TIIB and BaL, considered as prototype lymphotropic and macrophag.etropic viruses respectively, were included. Viral replication was monitored by reverse transcriptase (RT) activity and presence of p24 antigen in culture supernatants. Occasionally, when viral replication was undetect. able, cocultivation experiments with PHA-stimulated PBL were done. RESULTS; HIV-1 gp120 F) PULL- LEU TH (FL) gp160 CAN BIND TO CELLS BY DIFFERENT M.A.1007 gMECHAISMS. Kelker Hanna C., Ueorgesou R, Valentine FT. NYU Medioal Center, New York, NY 10016. Objectives: To oompare the binding of several recombinant gp120a and gpl60s to CDI in an ELISA and a cellular system. Methods: We have examined the binding of env proteins to: 1) solid phase CDl (Biogen) using an ELISA assay which we have developed, and 2) CD4i cells (flow cytometry). Binding of gp120 produced by Ameriean Biotechnologies, Cellteoh, Chiron or HioroGeneSys (nMS), as well as FL gp160s produced by MGS or Repligen and a gp160 missing much of the transaembrane region (Pasteur Herieux) were assessed utilizing HIV+ timane globulin or a monoclonal antibody to gpo 1 to detect binding. Resultsp gp120 and FL gp160 bind to cells by different mechanisms, even though saturable binding was observed with both molecules. The modified gp160 binds to CD4 in the same manner as gp2gp20 120's bind to solid phase CD4 in ELISA whereas FL gpl60s bind poorly in this assay. Binding of all gp120 preparations to either recombinant CD4 or to CD4+ cells is blocked by OKT4A or preinoubation with rCD4. Binding of modified gp160 also is inhibited by OKT4A. By contrast, binding of FL gp160, measured either by flow cytometry or by ELITS, is not inhibited by OKT4A. Preinoubation of CD4+ cells with a saturating concentration of gp120 fails to block binding of FL gpl60 suggesting that gp120 and FL gp160 bind to CD4+ cells at different sites. gp120 binds to CD4+ cells only, wbhile FL gp160 binds also to some cells that do not express CDI, or have lost CD4. Half saturation of binding of NGS gp120 and FL gp160 to CDI cells occurs at approximately the same concentrations. Half saturation of binding occurs at 5x10" M Celltech gp120. Conclusions: Binding characteristics of gp120 to CD4 are identical in an BLISA assay and a CD4+ cell assay, gp120 and FL gp160 bind to CD4+ cells by different mechanisas, whereas anlfin without some of the transmembrane reilon binds to the name determinants as ro120. Virus Number HIV-1 52 HIV-2 12 MC 8 Virus replication PBL BDM RT P24 52 22 27 12 5 7 8 1 7 Coculture Total 3 52 12 '^ cs a Total 72 72 28 41 3 72 CONCLUSION; All HIV isolates, including HTLVIIIB and HIV-IBaL, are tropic for both T-lymphocytes and blood derived macrophages. However, they show great variations in their capacity to replicate In BDM. IJI

Page  94 S M.A.1008 A DY DEPENDrT EANCaMN OF HIV-1 INFECTION RQUIRES CD4 REC R ANDI) FcR I OR II. Kobayashi, loju*, Takeda, A.*, Fanger, M.W.**; Ennis,F.A.* * Division of Infectious Diseases and Inmunology, University of Massachusetts, Worcester, MA. ** Dartmouth Medical School, Hanover, NH, USA. Objective: We previously reported that the sera of HIV-1 infected patients have antibodies which enhance infection by HIV-1 of U937 cells and human manocytes (Takeda, A., et al, Science, 1988). Antibody dependent enhancement (ADE) required interaction with both CD4 and Fc receptor (FcMR) I (Takeda, A. et al, J.Virol. 1990). This studl was designed to evaluate the role of Fc1R II in ADE of HIV-1 infection. Methods: HIV-1 (HTLVIIIB strain) was incubated with mouse anti-HIV-1 gpl20 monoclonal antibody (gpl.11.2, IgGl), and then used to infect U937 cells or HeLa cells transfected with genes coding for CD4 and/or Fc1RII. Infection was measured in culture supernatants of gpl.11.2Fab/IV. 3Fab (anti-FcIR II) produced by covalently linking the Fab of the mAb to gpl20 to the Fab of the mAb to FclR II. Results: 1)p24 production in U937 cells infected with HIV-1 cxmplexed with the gpI.11.2/IV.3 heteroantibody was significantly (2.1 times) higher than by HIV-1 alone. ADE by the heteroantibody was inhibited by incubation of the cells with aggregated human IgG, indicating ADE was mediated by Fc6R II. 2) Hela cells transfected with CD4 and FcR4 II genes showed increased p24 production when infected with HIV-1 coupled to conventional nAb to gp 120 or ot heteroantibody which also contained Fab to FcgR II. Conclusions: F cR II mediated ADE of HIV-1 infection. CD4 is a required receptor for FcyR II, similar to our results with FcyR I-mediated enhancement of HIV-1 infection. NOTES M.A.1009 ANALYSIS OF HIV INFECTIONS IN CD4 MUTANTS OF HUMAN MACROPHAGE-LIKE CELL LINES. Brs Lea, Verhaegen S., Saman E.", Fransen L.",van der Groen G.', Van Heuverswyn H., and De Baetselier P. Institute of Tropical Medicine - Antwerpen - Belgium, I nnogenetics N.V. - Antwerpen - Belgium, Institute of Molecular Biology, V.U.B. - Brussels - Belgium. Objective: To analyze the role of CD4 in the infectionand productionof HTLVIIIB in human macrophage-likecell lines, CD4' mutants were generated and tested for their susceptibility towards productive HTLVIIIB infections. Methods: U937 and THP-1 cells were treated with mutagcns and selected for the absence of CD4 via panning on antiCD4 coated petri dishes. The selected cell lines were CD4 negative as evaluated by membrane immunofluorescence and Northern blotting analysis. HTLVIIIB infections in these cell lines was monitored by an antigen-captureELISA, coculture with CD4+ T-cell lines, p24 membrane immunofluorescence, reverse transcriptase activity and PCR. Results: Infections of these CD4 cell lines with HTLVIIIB showed two patterns of virus production: (i) a low productive infection (detected by coculture with CD4+ acceptor T-cell lines) that weaned away after a few weeks in culture, (ii) a low productive infection becoming highly productive after a latency period of 60-90 days. Both patterns of virus productioncould be completely blocked or delayed by treatment with CD4-specific antibodies. This observation led to a re-evaluation of the expression of CD4 on these cell lines via PCR analysis. Using a coupled reverse transcriptase/polynerase chain reaction (RT/PCR), a weak CD4 signal could be detected in all the putative CD4 cell lines. Conclusions: These results indicate that an extremely low CD4 signal is still sufficient to promote HIV entry in monocytic cell lines. A productive infection however may be restricted by a low expression of CD4 leading either to abortive or latent viral production. Furthermore the absence of CD4 on putative CD4' cell lines should be considered with caution. NOTES M.A.1010 EXPRESSION AND LOCALIZATION OF CD4 IN HUMAN FETAL NEURONS AND GLIAL CELLS. Torelli S.*,Ennas M.G.*,Sogos V.*,Cocchia D.**,Gremo Fulvia* *School of Medicine, Cagliari; University "Tor Vergata", Rome, ITALY Infants with AIDS present severe neurological symptoms, which could be either a consequence of the general disease or the product of a direct infection of the brain occurring during pregnancy. So far, no direct evidence has been provided about the presence of the cellular receptor for HIV, CD4, in human fetal brain. We have investigated the expression and localization of CD4 gene product in vivo and in vitro in 8-20 week-old human fetal brains, with the uss of polymerase chain reaction (PCH), immunoprecipitation, immunohistochem ical staining at light, scanning and transmission electron microscope. Results showed that CD4 was expressed in fetal neurons since the firsl week in culture, regardless the age of the embryo. The molecule was pref ent both in the cytoplasm and on the surface of cell bodies and processes. CD4 was synthesized by fetal neurons in culture and contained thE gpl20 binding portion, as also assayed with gpl20 binding inhibition experiments. However, CD4 molecular weight was slightly inferior to T lymphocyte CD4 and mRNAs lacked about 240 bases both in vivo and in vitro. We conclude that HIV could indeed target human fetal neurons. M.A.1011 LECTIN-CARBOHYDRATE INTERACTIONS AND INFECTIVITY OF HIV-1. Gattegno, Liliane* ' Ramdani, Abdelhafid*; Jouault, Thierry"; Saffar, Line'; Gluckman, Jean Claude" *Faculte de M6decine, Paris-Nord and * CERVI, HOpital Piti6-Salpltriere, Paris, France. Objective: To examine whether mannosyl-specific lectins intervene in rgpl60 binding to the cell membrane in a CD4-dependent or- independent manner, and to what extent does lectin-mediated enhanced binding of env glycoprotein correspond to modified susceptibility tc HIV-1 infection of monocytic cells as compared with lymphoid cells. Methods: We investigated the effect of mannosyl-specific lectins, ConA, LCA, PSA and VFA i) on 125Irgp160 binding to CEM, U937 cells and to monocyte-derived-macrophages; ii) on the interaction of viral gpl20 and sCD4; and lii) on HIV-1 infectivity for monocytic cells as compared with lymphoid cells.Results: sCD4 did not interact with a ConA-Sepharose affinity matrix and HIV-1 preincubated with buffer or with ConA bound to sCD4 in a similar manner, When preincubated with rgpl60, or the cells, the lectins significantly enhanced rgpl60 binding to the cells In a dose-dependent, carbohydrate specific, and CD4 independent manner. Despite lectin-mediated enhanced binding of env glycoproteins, ConA neutralized HIV-1 infectivity for monocytic as well as for lymphoid cells, Conclusion: These results demonstrate that mannosyl-speficic lectins i) induce CD4-independent bridge formation between env glycoprotein and CD4+ cells, ii) do not inhibit gp120-CD4 interactions - which demonstrates that gp120 mannosyl residues are not involved - iii) neutralize HIV-1 infectivity for monocytic as compared with lymphoid cells. Therefore mannosyl-specific lectins behave like neutralizing antibodies that do not interfere with CD4 binding of gp120 but with post-binding events. a sr c~ c~ h3 c~

Page  95 M.A.1012 HIV INFECTION OF BONE MARROW-DERIVED DENDRITIC CELLS (DC) Patterson, Steven; Gross, J; Bedford, P; Knight, SC Clinical Research Centre, Harrow, Middlesex, UK. Objective: To characterize different populations of peripheral blood DC and investigate their susceptibility to HIV infection. Methods: The presence of DR, DQ, CD4, CD3, CD19, CD16 and CD14 antigens on DC was assessed by immunogold electron microscopy (EM). Susceptibility to in vitro HIV infection was examined by EM and in situ hybridization to detect viral RNA and DNA. The effect of antibody against CD4 (leu 3a) on infection was also studied. Results: Two distinct morphological types of DC (1 and 2) were observed, both expressed DR and DQ antigens, low levels of CD4 and lacked markers specific for T (CD3), B (CD19), NK (CD16) and macrophage (CD14) cells. A few cells similar to afferent lymph veiled cells were also seen (type 3 DC). The presence of cells with morphologies intermediate between these types suggests that they represent a differentiation pathway from immature type 1 DC to mature type 3 (veiled) DC. Types 2 and 3 but not type 1 DC were susceptible to HIV infection. By in situ hybridization a higher percentage of DC were positive for viral RNA and DNA than for RNA alone. Infection was blocked by the leu 3a anti CD4 monoclonal antibody. Discussion & Conclusions: Morphologically distinct DC that may reflect stages of a developmental pathway were identified. Only those DC considered to be more mature by morphological criteria supported HIV growth. The higher numbers of DC positive for HIV DNA than for RNA suggest that some DC may become latently infected or allow only a low level of virus replication. Despite their ability to bind a vast array of antigens, infection of DC by HIV is mainly dependent on attachment to CD4. NOTES M.A.1013 DECREASED HIV-1 INFECTIVITY AFTER IN VITRO DOWN-REGULATION OF CD4 WITH Pl.1I HORBOL 12-MYRISTATE 13-ACETATE (PMA). Sanhadji Kamel, Benard I. & Touraine J.-L. Facult6 de Mddecine A. Carrel, INSERM U.80-CNRS URA 1177, Rue Guillaume Paradin, 69008 Lyon, FRANCE. Objective: To analyze the variation of HIV infectivity of lymphocytes when CD4 expression is modulated by in vitro pre-treatment with PMA. Methods: The MT4 cell line (3x105 cells/ml) was pre-treated with 1.6x10-8 M of PMA (or with control solution) for 0.5, 1, 5 or 20 hrs, then washed and assayed for CD4 expression by cytofluorometry. The cells were then incubated with HIV-1 for 1 hr and immediately cultured. Infection was determined by cytopathogenic effect (CPE), reverse transcriptase (RT) activity and P24 antigen release in supernatant. Results; Over 80% of cells initially expressed the OKT 4A epitope. This expression decreased following PMA treatment and the lowest value was observed at lhr (6%). A partial re-expression was however observed when the incubation period was extended to 20 hrs. The RT activity was decreased by this pre-treatment but it was even lower after 5-20 hrs of incubation than after 1 hr. This observation was confirmed by the CPq appearance and the P24 antigen release in culture supernatant. HIV-1 infectivitj therefore decreased-with the reduced OKT 4A expression resulting from PMA treatment but it continued to decrease when OKT 4A was re-expressed. Conclusion: Although PMA increases production of several cytokines and augments HI\ replication in already infected cells, this agent reduces the infectivity of cells wher it is introduced in the culture before HIV. This effect is very likely to result fron down-regulation of CD4 expression and it lasts even longer than the inhibition o expression of this virus receptor. NOTES M.A.1014 A NON-IMMUNOGLOBULIN, NORMAL SERUM PROTEIN BINDS SPECIFICALLY TO THE GP41 BINDING DOMAIN OF HIV GP120 Pinter, C.; Siccardi, A.G.; Clivio, Alberto Dipartimento di Biologia e Genetica per lc Scienze Mediche, University of Milan, Italy Objective: to search, in normal human sera, HIV binding proteins possibly involved in HIV spreading strategies such as dissemination, alternative tropism or masking of functionally relevant epitopes at the virion surface. Methods: Normal human sera were pooled and fractionated by high resolution anion exchange HPLC (High Performance Liquid Chromatography). The fractions were monitored for HIV binding with ELISA assays on HIV-coated plates using an antiserum raised against total serum proteins as the tracer. The active fractions were labelled and incubated with HIV in solution. Molecules actively involved in HIV binding were recognized using an HPLC retardation assay and SDS-PAGE. ELISA radiobinding assays with free and coupled synthetic peptides were used to assess binding specificity. Results A single active fraction was found by monitoring the anion exchange column elution. This contains a 42kd protein which is responsible for binding to HIV virions in solution. Attempts to demonstrate binding to specific HIV proteins in western blots failed, but the binding to intact virions is strongly and specifically competed by the synthetic peptide SA2 (HEDIISLWDQSLKDC),. a gpl20 domain shown to be involved in the binding of gp41 (McPhee, personal communication). Discussion and Conclusions: A HIV binding protein with unknown function was found in normal human serum pools and was partially characterized. It binds to the intact virion even in the presence of high titres of anti-HIV antibodies and the binding is specific for the gp41-binding domain of native HIV gpl20. Future research should involve 1) investigation on a possible role of this protein (CP-1) in HIV infection and 2) studies on binding of CP-1 to specific cell types and its possible role either in alternative tropism or in virus dissemination throughout the body. HIV-1 gpl20 MIMICKS A HLA CLASS I HIDDEN EPITOPE M.A.1015 Meneveri. Raffaella; Grassi F.; Marozzi A.: Agresti A.; Siccardi A.; Ginelli E. Dipartimento Biologia e Genetica per 11 Scienze Mediche, UniversitA dl Milano, Italy. Objective: To characterize a 45 kD human protein which shares an epitope with HIV-1 gpl20. The protein, present at high concentrations on the surface of activated tmmunocytes, is defined by two Mabs (M38 and L31) one of which (M38) is cross-reactive with the virus. Methods: Immunoscreening of an expression cDNA library from human activated lymphocytes. Sequencing of an isolated recombinant clone. Data bank analysis of derived nucleotide and amino acid sequences. Immunoprecipitation and Western-blot analysis. Results: A clone, producing a L31- and M38-positive recombinant protein, has been isolated. The insert sequence (1549 bp) contains an ORF coding for a protein of 366 aa which is highly homologous (96%) with the allele Cb-1 of the HLA Class I multigene family. Furthermore, immunoprecipitation experiments indicate that the epitopes defined by Mabs L31 and M38 are common and monomorphic determinants of HLA Class I heavy chain. Discussion and Conclusions: HIV-1 gpl20 shares an epitope with HLA Class I heavy chain. Moreover, by comparing the expression patterns of L31/M38-deflned epitopes and of classical HLA antigens, it can be suggested that L31/M38 determinants reside on structurally altered HLA molecules present on the cell surface upon cell activation.

Page  96 S M.A.1016 BINDING OF CD4 TO SYNTHETIC PEPTIDES PATTERNED ON THE PRINCIPAL NEUTRALIZING DOMAIN OF GP120. Autiero,M';Abrescia,P,' Dettin,Monica*;DiBello,C*;Guardiola.CNR,Naples, Italy;*University of Naples,Naples,Italy;*University of Padua,Padua,Italy. Objective:Studies on the binding of gpl20 and gpl20- derived peptides to Sepharose 4B-linked Soluble CD4 were performed. The synthetic peptides corresponding to the PND of gpl20 Caa. 308-331) are able to bind CD4 both as an immobilized recombinant molecule and as a membrane-bound antigen and the interaction is specific. Methods: Peptide were obtained by solid phase synthesis and purified by FPLC and HPLC. The affinity matrix was prepared incubating sCD4 with CNBr-activated Sepharose 4B. The amount of eluted peptide wasketermined spectrophotometrically. Competition experiments were made using mAbs in presence of [125) -I labelled G-E-3 peptide (sequence (307-330) of MN isolate). Results: The specificity of the column was studied by competition experiments using anti -CD4-mAbs. Peptides corresponding to the gpl20 PND domain of different isolates specifically bind to sCD4 Sepharose 4B even if the PND region is not expected to take part in the binding of gpl20 to CD4. Discussion and conclusion: The differences of peptides in the interaction with CD4 bring to these conclusions: the GPGR sequence is probably involved in the binding but the context in which the GPGR motif lies may be relevant; modification of carboxy-terminal function is likely to drastically reduce the propensity of PND peptide to assume the appropriate conformation for binding; peptides binding region is not coincident with enl20-bindine site. NOTES M.A.1017 GROWTH PATTERN OF LABORATORY HIV-1 STRAINS DOES NOT SREFLECT THE BIOLOGICAL BEHAVIOUR OF THE VIRUS IN VIVO Lu Wei, Ferreira N, Israel-Biet D, Andrieu JM. Laboratory of Tumor Immunology, LaEnnec Hospital, Universit6 de Paris V, Paris, France. Objective: Given the importance of a precise knowledge of HIV tropism and growth characteristics for a better understanding of the HIV disease, we have undertaken a study of the infectivity, growth kinetics, and cytopathogenic effect of laboratory adapted HIV-1 strains, and compared them with those of fresh primary HIV-1 isolates recovered directly from patients' cells. Methods: The cell-free HIV-1 freshly isolated from PBL (n=4) or bronchoalveolar lymphocytes (n=3) of 7 seropositive patients, and 5 HIV- isolates maintained i1 laboratory T lymphocytic and monocytic cell lines (2 in CEM, 2 in HUT78, and I in H9), were studied. 10" PHA stimulated PBL from a normal HIV-negative donor were infected by a standardized inoculum (1 ng of HIV core p24 protein) of each single isolate. The growth rate of viral production in each case was assessed by serial measurements of the p24 content of supernatants (Abbott ELISA) during culture over 28 days. The cytopathic changes in infected cells were monitored by trypan blue exclusion dye test and flow cytometry. Results: All the primary isolates from each individual patient exhibited similar infectivity and growth kinetics. The peak of p24 level (day 7, 837~331ng/ml) was associated with 100% death of the CD4+ lymphocytes in the culture. In contrast, only 3 of the 5 laboratory strains productively infected PBL, with various infectivity and growth rates. HIV recovered from laboratory cell lines led to a very low viral production, always < lOng/ml, with a peak delayed after day 11. Strikingly, no cytopathic change was ever observed in these cases, nor CD4 -receptor down regulation, as assessed by flow cytometry. Concluslon: Unlike fresh primary isolates, the laboratory adapted HIV-1 strains infected PBL very poorly and lost their cytopathogenic potential in vitro. Therefore, results of studies using such laboratory cell-line maintained virus do not accurately reflect the situation in vivo and should be interpreted with caution. NOTES M.A.1018 TGF-( OVERCOMES THE LYMPHOCYTOTROPIC RESTRICTION OF HIV-1 Lazdins Janis, Alteri E., Woods-Cook K., Klimkait T., Cox D., Bilbe G., Cerletti N. and McMaster G., CIBA-GEIGY Ltd. Basel CH 4002. Objective: evaluation of the role of TGF-'on HIV-1 replication in blood derived macrophages (M0). Methods: monocytes were isolated by leukapheresis and purified by elutriation. After differentiation M0 were infected with HIV-1 ADA (monocytotropic) IIIB, LAV or Z-84 (lymphocytotropic). Infection was evaluated by supernatant p24 ag., RT activity, morphology and immunoblot analysis. Results: Infection of M0 with ADA shows an enhanced viral replication in the presence of TGF-1I or 2. When Mo were infected with III8, LAV, Z-84 no p24 or RT was detected in cell supernatants, however immunoblots show a p55 band. Wren infection with these isolates is performed in the presence of TGF-r a dose dependent virus production was demonstrated. The virus that emerged retained its original cell tropism. Discussion and Conclusions: TGF-% a factor produced at the site of injury facilitates virus replication in M0, independent of the viral tropism, this effect is not due to apparent changes in cell phenotype. This effect was not seen with other cytokines. Since MO can produce TGF- this molecules could play not only a critical role in the establishment of infection but as well as progression of disease. Furthermore this effect of TGF-tcould play an important role in infection in tissues were M0 are important cell targets ie: brain. M.A.1019 INTERACTION OF HIV-1 WITH HUMAN ENDOTHELIAL CELLS EXPRESSING A TRANSFECTED CD4 GENE. Huritz. Arthur A., Hatch, W.C., Lyman, W.D., Hatcher, V.B., Berman, J.W. Albert Einstein Coll. Med., NY, NY, USA Objective: To test the hypothesis that endothelial cells can be infected by HIV-1. Methods: Early passage human umbilical vein endothelial (HUVE) cells were transfected with a full-length CD4 cDNA and a selectable neomycin-resistance marker (CD4+EC). CD4 EC and nontransfected control HUVE were tested for CD4 expression by nucleic acid hybridization and flow cytometry. CD4 EC and nontransfected HUVE (CD4"EC) were cultured with HIV1lb. Infection of HUVE by HIV-1 was determined by measuring reverse transcriptase (RT) activity and viral antigen (p24) concentration, and by in situ hybridization using a full-length cDNA probe. Results: Transfected HUVE both contained and expressed the CD4 gene as determines by Southern and Northern blotting. Flow cytomet confirmed that approximately 20% of CD4 EC expressed surface CD4. Supernatants of CD4 EC exposed to virus contained more HIV-1 RT and p24 than CD4'EC supernatants. No HIV-1 was detected in any HUVE culture by in situ hybridization. Discussion and Conclusions: Although HIVIb bound to CD4+EC, these cells were not productively infected. This may have been due to viral cytotoxicity for HUVE, an endothelial cell regulatory mechanism, or to the HIV-1 isolate tested. However, these data do not exclude the possibility of a latent infection nor do they eliminate the expectation that HIV infection of endothelial cells may occur via CD4-independent mechanisms n vitro or iLn vivo. The present findings are consistent with the minimal evidence for HIV-1 infection of endothelium in vivo, either in the central nervous system or in the systemic circulation. (Supported by United States Public Health Service grants MH 47667, MH 46815 and DA 05583)

Page  97 M.A.1020 HIV-1 INFECTION OF HUMAN FETAL MACROPHAGESIN VVO AND IN VITRO: KINETICS OF VIRAL GENE EXPRESSION IN CULTURE. Hatch. William C., Burstein, Y*, Calvelli, T.A., Rashbaum, W.K., Lyman, W.D. Albert Einstein Coll. Med., NY, NY, USA and *Tel Aviv Univ. Med. Center, Tel Aviv, Israel. Objective: To compare the kinetics of HIV-1 gene expression in human fetal macrophages (HFM) infected in vivo with the kinetics of infection of normal HFM in vitro using different HIVI isolates. Methods: HFM from fetal livers were obtained from abortuses of HIV-1 seropositive and seronegative women. The cells were cultured using standard conditions and analyzed for phenotype by flow cytometry and non-specific esterase staining. HFM from abortuses of HIV-1 seropositive females were monitored for infection by reverse transcriptase (RT) and p24 assays, immunocytochemistry and nucleic acid hybridization. HFM from abortuses of seronegative women were exposed to the RF (L-cytotropic) and JRFL (M-cytotropic) isolates of HIV-1. These cultures were monitored for HIV-1 infection using the same methods. Results: HFM obtained from HIV-1 seropositive females showed a 3-fold increase in RT activity after 30 days in culture. Similarly, HFM from abortuses of seronegative females incubated with the RF strain showed a 3-fold increase in RT activity after 30 days in culture. In contrast, HFM incubated with the JRFL strain showed a 50-fold increase in RT activity by 14 days in culture. Discussion and Conclusions: The kinetics of HIV-1 expression in cultured HFM infected in vivo were similar to HFM infected in vitro by the L-cytotropic RF isolate of HIV-1. This slow-low expression of HIV-1 may be related to viral latency. In contrast, the expression of viral genes after exposure of HFM to the M-cytotropic isolate (JRFL) demonstrated high-fast kinetics. These data suggest that HIV-1 infection of the human fetus by distinct viral isolates may contribute to the observed differences in post-natal pathology. (Supported by USPHS grants MH 47667, MH 46815 and DA 05583 and the Atran Fund) NOTES M.A.1021 POTENTIAL ROLE FOR COMPLEMENT RECEPTOR 2 (CR2) IN COMPLEMENT (C) MEDIATED ENHANCEMENT OF HIV IN VITRO AND IN VIVO. June, R.; Spear, G.; Stefanlk, K.; Lint, T.; Landay. Alan L. Rush Presbyterian St. Luke's Medical Center, Chicago, Illinois, USA Objective: This study was designed to determine the mechanism of complement mediated enhancement of HIV-1 Infection by monitoring provirus formation and virus binding. It also characterized a population of T lymphocytes in vivo which expressed the receptors required for enhancement. Methods: Supernatants containing HIV/RF were incubated with anti-HIV antibody (Ab) and/or C for 1 hr. at 370C. Cells (MT2 or Peripheral blood lymphocytes IPBL]) were then added and incubated an additional 4 hr. The cells were washed, cultured for 3-7 days and evaluated for HIV-antlgen (Ag), reverse transcriptase (RT) activity, or provirus formation by polymerase chain reaction. To assess virion binding, Ab and C treated virus was incubated with cells on ice for 1 hr, followed by immediate cell lysis to evaluate bound HIV-Ag. Complement receptors on PBL were characterized by two and three color flow cytometry. Results: HIV-Ag production and RT activity of MT2 cells and PBL enriched for CD4+ lymphocytes infected with virus pre-treated with Ab and C was significantly greater than that of cells infected with virus alone, or virus treated with Ab or C separately. This increased HIV-Ag and RT production correlated with provirus formation and HIV binding. Flow cytometric studies of PBL from healthy volunteers demonstrated co-expression of CD4 and CR2. Three color flow cytometric evaluation of this population demonstrated co-expression of CD45RA, CD45RO dim, and CD29 intermediate. The coexpression of CR2 on CD4+ lymphocytes was decreased by 60% in HIV infected individuals. Conclusions: We have directly shown that the increased HIV-Ag production seen in cells infected by Ab plus C treated HIV is due to increased binding of virus to cell surfaces and increased infection. We have also shown that a population of lymphocytes exists in vivo which may be capable of exhibiting complement mediated enhancement. These cells were found to be decreased in AIDS patients. NOTES s CI tl M.A.1022 KINETICS OF HIV-1 FUSION WITH LIPOSOMES AND ERYTHROCYTE GHOSTS. Larsen. Charles E.*; Nir, S.**; Alford, D.R.*; Jennings, M.***; Lee, K-D*; and Diizgiineg, N.*. *University of California, San Francisco, CA, USA; **The Hebrew University of Jerusalem, Rehovot, ISRAEL; and ***University of California, Davis, CA, USA. Objective: We investigated the kinetics and extent of membrane fusion between HIV-1 and phospholipid vesicles and erythrocyte ghost membranes and studied the effects of pH, divalent cation, target membrane composition, temperature and viral trypsinization on this process. Methods: Purified HIV-1 was labeled with octadecyl rhodamine B chloride (R18); fusion was monitored continuously as the dilution of probe into target membranes. Ultrastructural analysis of the virus and reaction products by transmission electron microscopy was also utilized to confirm the results of the RIB assay. Results: HIV-1 fusion is strongly dependent upon target membrane lipid composition. The virus fuses with anionic liposomes dramatically faster than with erythrocyte ghost membranes or pure dioleoylphosphatidylcholine (DOPC) liposomes at neutral pH. Reduction of pH from 7.5 enhances the rate and extent of HIV-1 fusion with vesicles containing anionic lipids. Physiologically relevant concentrations of calcium stimulate HIV-1 fusion with several liposome compositions andwith erythrocyte ghost membranes. Compared to intact virions, the fusion products show a striking reduction in the fusion rate with liposomes. Infection of CD4-expressinT cells was reduced when the fusion products of HIV with CL liposomes were applied instead of the intact vinons. HIV-1 fusion with CL vesicles is temperature dependent and, in the presence of calcium, is greatly reduced following virus trypsinization. Conclusions: HIV-1 fuses with model and biological membranes lacking the known viral receptor, CD4. Like Influenza and Sendai, HIV-1 fusion with membranes containing its own envelope protein(s) is strongly inhibited. Unlike these viruses, HIV-1 fusion is promoted by calcium and appears to be more sensitive to envelope protein surface density. HIV-1 fusion with liposomes is similar to SIVmac fusion (Larsen et a&l. (1990)L. en. V irl. 71, 1947-1955). This work was supported by USPHS Grant AI25534 (ND), U.S.Israel B. S. F. grant 86-00010 (ND/SN) and fellowships (CEL) from the ACS (3394) and NIH (AI08117). M.A.1023 USE OF A RAPID CD4-BASED ELISA TO MEASURE BINDING OF VARIANT GP120 MOLECULES AND TO SCREEN FOR INHIBITORS OF BINDING. Brigido, Luis Fernando de Macedo*; Gilbert M*; Culp,J**; Mills J. *SFGH-University of California,San Francisco, CA & **Smith Kline Beecham Pharmaceuticals, King of Prussia, PA, USA OBJECTIVE To developed a simple, rapid and economical system to measure the binding of variant gpl20 molecules to rsCD4, and to screen for inhibitors of binding potentially suitable as antivirals. METHODS: An ELISA previously described (Gilbert et al J. Clin. Micro 29;142-147;1991) was modified for these purposes. 10 ng/ml of recombinant or native gpl20, with or without preincubation with inhibitors were added to wells coated with 1 ug of rsCD4 and binding was detected with sheep anti-gpl20-biotin-avidinphosphatase system; with results expressed as percent of control wells. RESULTS: Up to 80% of enzymatic deglycosylation with Endo F/Endo H did not modify the binding to CD4 and recognition by the antiserum, although a non-glycosylated gp120 tested, produced in yeast (NIH 388), gave a absorbance 10x less then that obtained with glycosylated material. The presence of gpl20 partially blocked the binding of leu3a, not decreasing the binding of other anti-CD4 monoclonal antibodies (mAb) unrelated to the T4a epitope. On the other hand, leu3a blocked sequential binding of gp (50% inhibition, Ir0 = 250 ng/ml), whereas other mAbs, including OKT4, did not. RsCD4 also inhibited (I50% = 80 ng/ml) binding, and several sulphonated polysaccharides, as dextran sulfate and polyanetholsulfonic acid blocked binding at 5 ug to 500 ng/ml, as did aurintricarboxylic acid (150% = 1.5 ug/ml). At Higher concentrations, 1 mg to 10 ug/ml, these inhibitors were shown to bind to CD4, as preincubation of plaques blocked binding of gp added sequentially. Several lectins blocked binding (ConA, Sweet pea, lentil, human mannose binding protein), presumably through attachment to gpl20 sugar residues. CONCLUSION: This assay system can be used to determine easily and simply the relative binding of various gpl20 molecules to CD4, and to quantify inhibitors of that binding. It may be useful as a screen for inhibitors of binding with potential antiviral activity.

Page  98 0 M.A.1024 COMPLEMENT MEDIATES HIV-1 INFECTION OF HUMAN CELLS IN AN ANTIBODY-INDEPENDENT FASHION. Boyer Vironique,* Trabaud Mary-Anne *, Noraz Nelly *, Fischer Elizabeth +, Kazatchkine Michel +, and Desgranges Claude* *Unitd de recherche sur les H6patites, le SIDA, et les r6trovirus humains. INSERM U271. 151, Cours A. Thomas, Lyon.+Unit6 dImmunopathologie et INSERM U28. HOpital Broussais, Paris. Objective: To investigate the role of complement in HIV-1 infection of different types of cells independentely of the HIV specific antibodies. Methods: To determine the role of complement in HIV-1 infection we incubated the HTLV-LIIB and HTLV-RF strains of HIV-1 with seronegative normal human serum under conditions which allow complement activation. The incubation of the complement-pretreated viruses was done with the human MT2 (CR2+) and U937 (CR1+ or CR3+) cell lines and with peripheral blood T cells.Infection of cells was assessed by measuring reverse transcriptase (RT) activity in supematants at different times of culture. Results: Incubation of different cells with complement and virus resulted in enhancement of infection of the cells when low amounts of virus were used.For MT2 cells, complement activation by viral suspensions occurred through the alternative pathway and infection with suboptimal amounts of serumopsonized HIV-1 was suppressed by blocking the C3dg receptor (CR2, CD21) on MT2 cells with monoclonal anti-CR2 antibody and rabbit Fab'2 anti-mouse Ig antibodies. Blocking of the gpl20-binding site on CD4 under similar experimental conditions had no inhibitory effect on infection of MT2 cells with opsonized-virus. For U937 cells and PBL, the mechanism of complement activation by HIV-1 will be presented as well as for the coculture of PHA-activated and HIV-infected PBL from different donors. Conclusion: The complement in the absence of specific antibodies enhances infection of C3 receptor bearing cells just by the interaction of opsonized virus with the CR1, CR2, or CR3 receptors. The clinical relevances of these findings remains to be determined. NOTES M.A.1025 GENOMIC VARIABILITY OF HTLV-I ACCORDING TO GEOGRAPHICAL AREAS AND ASSOCIATED PATHOLOGY Komurian F1-5, Pelloquin F1, Giordano C2, Gessain A3, Sonoda S4, de-Th G5 1. Mdrieux Institute, Marcy l'Etoile, France 2. Cocody Hospital, Abidjan, Ivory Coast 3. NCI, Bethesda, U.S.A. 4; Kagoshima University, Kagoshima, Japan 5. Institut Pasteur, Paris, France In order to study the geographical and disease-associated variations of the HTLVI genome, the ex-vivo DNA from nine patients were studied by PCR, followed by direct DNA sequencing. For each viral strain, about 2,000 bp were thus sequenced, including LTR, env gene, px II, px Ill and px IV coding frames of the px region. The genomic variations observed in the LTR region were significantly higher than those seen in other viral genes. In the px region, very few mutations were present in px II and px III genes, but px IV exhibited numerous single point mutations. No specific mutation could be linked to ATL or PST/HAM specific pathology. Observed changes among HTLV-I from different geographical areas (Ivory Coast, Caribbean area or Japan) suggest that some Japanese HTLV-I may be close to Caribbean HTLV-I, while others are independent African HTLV-I appear to represent a group apart. NOTES M.A.1026 HIV ENVELOPE CHIMERIC PROTEINS FOR RETROVIRAL TARGETING Levy. David N.. Williams, W.V." and Weiner, D.B.'*, *University of Pennsylvania and "Wistar Institute, Philadelphia, Pennsylvania USA Objective: The cellular tropism of HIV is determined by the envelope proteins. The precursor glycoprotein gpl60 is post-translationally cleaved to yield gp120 and gp41. The external glycoprotein gp120 is noncovalently attached to gp41 and mediates virus binding to the T cell antigen CD4. The transmembrane glycoprotein gp41 mediates virus fusion with the cell membrane following virus attachment. Cells that are infected with HIV express gp120 and gp41 on their surface. Retroviral mediated transfer of genes and/or pharmacologically active agents is a promising method for treatment for various diseases including viral infection. An important ingredient in such a system is the ability to specifically target a retroviral vector to the cells of interest. Method: In order to study determinants of viral tropism and to develop a system for specifically targeting retroviral vectors to HIV infected cells we have constructed chimeric HIV envelope proteins in which the determinants of virus binding have been replaced with determinants of other proteins involved in receptor-ligand interactions. Among these we have created chimeric proteins in which the CD4 binding domain of gp120 is replaced by the gp120 binding domain of CD4. Results: We have expressed these proteins in mammalian cells and examined their expression and function with respect to protein folding, cleavage of gpl60 into gp120 and gp41, transport to the cell surface, binding to native gp120, fusion with HIV envelope expressing cells, and incorporation into viral particles. Discussion: These studies suggest that chimeric proteins can have utility in retroviral targeting. M.A.1027 THE ESTABLISHMENT OF A RAT CELL LINE CONTINUOUSLY PRODUCING HIVI Mizrachi. Yaffa, Sternas L., Volsky D.J. St. Luke's/Roosevelt Hospital Center, Columbia University, New York N.Y. Objectives: To establish a rodent cell line stably carrying HIV-I genome and replicating HIV-I over extended periods of time, as a model system for studying HIV-1 replication in the absence of HIV-I receptors. Methods: A vector carrying an infectious clone of HIV-I and the neomycin resistance gene was electroporated into various non-human cells. The cells were analyzed for the presence of the HIV-1 genome, HIV-1 replication and for transactivation of the HIV-1 LTR in a transient transactivation assay. Results: Three subgroups of cells were identified with respect to HIV-I expression and HIV-1 LTR activation i) cells which were non-permissive to both HIV-I LTR transactivation and HIV-1 expression; ii) cells in which the HIV-1 LTR was transactivated but no HIV-1 replication was detected. iii) cells which permitted both the LTR transactivation and HIV-1 replication. The third group included a normal embryonal rat cell line; these cells were capable of carrying the HIV-I genome for 4 months; at the peak of infection, 5-20% of the infected cells expressed HIV-1 antigens as detected by IF, and HIV-I production reached 30-36ng of p24 per Ix10e cells. HIV-1 expression in these cells declined after 3 months; however, HIV-1 expression could be enhanced 5-7 fold by treatment with sodium butyrate, and to a lesser extent by PMA. Virus released by the persistently infected rat cells was infectious in human T lymphocytes. Conclusions: The model system described here will be useful in studies on HIV-1 expression in CD4- cells, identification and characterization of cellular factors needed for HIV-I infection, and characterization of non-CD4 receptor(s) for HIV-1. s; c^ I;

Page  99 M.A.1028 COMPLEMENT ACTIVATION BY GP160 AND COMPLEMENT-MEDIATED ENHANCEMENT OF HIt INFECTION OF MONOCYTIC CELL LINES Haeffner-Cavaillon, N.*; Thieblemont, N.*; Weiss, L.*; ZieglerHeitbrock, H.W.**; Kazatchkine, M.D.* * Hppital Brousais, Paris, France; ** UniversitAt Miinchen, FRG Objective: To study the role of complement in the facilitation of viral entry into monocyte/macrophages. Methods: Glycosylated recombinant gpl60 of HIV-1; monocytic cell lines Mono Mac 6 (CD4 antigen-negative) and THP1 (low expression of CD4 antigen); HTLV IIIB strain of HIV-1; LAV II strain of HIV-2. Results: Incubation of recombinant gpl60 with normal human serum resulted in dose and time-dependent activation of the classical complement pathway and deposition of neoantigens of cleaved C3 on the glycoprotein. No binding of C3 to gpl60 occurred in C2- and Clq-deficient sera. Binding of human anti-gpl60 IgG to the protein enhanced its ability to activate the classical pathway. Opsonization of HIV-1 and HIV-2 with HIV-seronegative normal human serum enhanced infection of the two monocytic cell lines that were tested, as assessed by the higher amount of p24 measured in supernatants of infected cells and the earlier time point at which p24 was detected following infection of the cells. Discussion: The gpl60 envelope glycoprotein of HIV-1 activates the classical complement pathway in the absence of antibody, resulting in binding of C3b/C3bi to the viral particle. Opsonizatlon with complement of HIV-1 and HIV-2 results in an early and enhanced infection of the CD4-negative Mono Mac6 cell line and of THPI cells. Thus, complement facilitates entry of HIV into monocytes. NOTES M.A.1029 MONOCLONAL ANTIBODIES DIRECTED AGAINST SYNTHETIC CD4 PEPTIDES Maino, Vernon*; Singer, R.*; Buck, D.*; Buchanan, T.*"; 'Becton Dickinson Immunocytometry Systems, San Jose, CA, "University of Washington, Seattle, WA. Objective: Synthetic peptides derived from CD4 sequences of the putative binding domain of gpl20 were employed In an effort to further characterize the gpl20 binding site on CD4. Methods: Monoclonal antibodies (MAb) raised against synthetic peptides derived from sequences of the V1 domain were evaluated fortheir ability to bind native CD4 and to block gpl20 and anti-CD4 MAb binding as determined by flow cytometry analysis. In addition we evaluated a panel of monoclonal antibodies with apparent specificity for the V1 domain of CD4 for their ability to bind to a variety of synthetic peptides. Results: A summary of the MAb reactivities are provided In the table below. These studies indicated that two mabs generated against synthetic peptides from regions 44-58 or 40-55 of CD4 demonstrated specific binding to CD4 as determined by FACS analysis. One of these MAbs, 4E4A (44-58) was capable of binding CD4 at titers comparable to many anti-CD4 mabs raised against native antigen. 4E4A also demonstrated weak but significant blocking activity for both gp120 and the anti-CD4 MAb L83. From a panel of 22 anti-CD4 MAbs capable of blocking gpl20 binding, one of these, L204, demonstrated specific reactivity, as determined by peptide ELISA, with CD4 23-55 a 33 residue synthetic peptide derived from the V1 domain of CD4. MAb CD4pepide reactivity CD4r=eadivity og120 blocking 4E4A 44-58, 23-55 + 39-53, 40-60 + (+) 5H3A 39-53, 40-60 L83 23-55 + L204 23-55 + + Leu 3a n.r.b + + Conclusions: This analysis suggests the amino acid region of 44-58 of CD4 contains sequences capable of binding both gpl20 blocking and non-blocking MAbs and therefore may represent at least a part of the gpl20 binding domain. NOTES p ýtl C) 00 M.A.1030 DEVELOPMENT OF AN HIV-1 INFECTION SYSTEM FOR STUDIES ON VIRAL LIFE.A. CYCLE IN NEURAL CELLS AND ISOLATION OF NEUROTROPIC HIV-1 STRAINS. Volskv. Barbara Sakai, K.; Shahabuddin, M.; Volsky, D. J. Molecular Virology Laboratory, St. Luke's/Roosevelt Hospital Center, Columbia University, New York, N.Y. Objective: HIV-1 infects neural cells but major differences exist compared to its infection of T cells. The route of HIV- entry into neural cells is unknown, the infection is small in extent and non-cytopathic, and virus can persist in a dormant/inducible state. Our objective was to develop an in vitro neural cell system which is susceptible to efficient HIV-I infection, and to employ it for studies on the HIV-1 life cycle in neural cells, its interruption with antiviral compounds, and for isolation of neurotropic HIV-1 strains. Methods: Neural cells were stably transfected with a CD4 expression vector pKS286. CD4+ and CD4" cells were infected with laboratory strains of HIV-I, and a natural isolate which could not be propagated in T cells. The infected cells were tested for: kinetics of expression of viral DNA, RNA, and proteins; relative levels of integrated and unintegrated viral DNA; cellular regulation of viral expression; sensitivity to antiviral compounds. Results: Stable transfection of glial HTB148 cells with the pKS286 vector generated a cell line, HTB 148/286, which expressed high levels of surface CD4 receptors. Three days after infection with HIV-I, 20-30% of HTB148/286 cells expressed HIV-1 antigens and produced high titers of progeny virus. HIV-I expression declined upon prolonged incubation. AZT was equally effective in blocking infection in these cells and T cells, suggesting that the early stages of HIV-1 infection were similar in both cell systems. In contrast, agents active in T cells in inhibition of transcription from DNA were significantly less effective in neural cells, reflecting the presence of distinct neural cell factors mediating HIV-1 transcription. The HTB148/286 cells supported the replication of a natural HIV-1 isolate which initially did not grow in T cells. Conclusions: These data establish the utility of this model system for high expression of HIV-I in neural cells. Cell-specific expression of some HIV-I functions should be considered in the design of new therapeutics. M.A.1031 CORRELATION BETWEEN FUSION, ENTRY AND EXPRESSION OF HIV-1 IN HUMAN **.A.1 1 AND MOUSE CELLS. Zeira. Michael'; Littman, D."; and Volsky, D.J. *Molecular Virology Laboratory. St. Luke's/Roosevelt Hospital Center, Columbia University, New York, N.Y.; **Dept. of Microbiology and Immunology; University of California at San Francisco, San Francisco, CA. Objective: To establish parameters of HIV-1 entry into cells by measuring fusion between viral envelopes and CD4+ T cell lines or freshly isolated PBLs, and to correlate between viral entry and expression in these cells and in mouse cells expressing human CD4 receptors. Methods: Sucrose gradient purified HIV- I-CAT virions were labeled with octadecylrhodamine B-chloride and incubated with target cells. Virus-cell membrane fusion was measured as the increase in the fluorescence signal (dequenching, DQ). Expression of HIV-I immediately after entry was determined by PCR and CAT assays, and by intracellular p24 and HIV-I RNA assays several days later. Target cells included PBLs, human T lymphoid A2.01 cells expressing either CD4 or CD8 receptors, and mouse 3T3 fibrolasts and thymoma 1010 cells, stably expressing human CD4 receptors (conferred by transfection). Results: DQ values observed following incubation of human CD4+ cells with labeled virus were dependent on cell number (max DQ at 2x106 cells) temperature (max DQ at 37~C and none at 40C) and incubation time (50% of DQ within 20 min and 100% DQ within 30 min). Preincubation of purified labeled virus with sCD4 resulted in sCD4 dose dependent inhibition of DQ. This showed that the fusion event, as measured by DQ, depends upon specific gpl20-CD4 receptor binding. No difference of fusion values was observed between resting PBLs and PHA stimulated PBLs. Stable expression of CD4 converted mouse 3T3 fibroblasts and thymoma 1010 cell lines from negative to positive DQ with HIV-1. The extent of fusion observed in CD4+ murine cells was comparable to CD4+ human cells. That DQ correlates with viral entry in these cells is being tested by intracellular CAT activity or by detection of HIV-1 DNA by PCR. Conclusions: The DQ technique enables us to determine quantitatively the viral and cellular parameters reauired for viral entry, and to correlate entry and exoressionin human and animal cells. 00 '^ '^

Page  100 I' 0 0 M.A.1032 THE COMPARATIVE TOXICOKINETICS OF 3'-DEOXY 3' FLOUROTHYMIDINE (FLT) IN LABORTORY ANIMALS Johnson, Dale, Dougherty, W.J., Pool, V., Riley, M.G.I., Yacobi, A. American Cyanamid Company, Pearl River, New York, USA Objectives To investigate interspecies differences in the toxicokinetics of FLT. Methodsi The comparative toxicokinetics of FLT in the rat and dog was evaluated in studies supported by pharmacokinetics, hematology and pathology. In monkeys, a study focusing on hematology and pharmacokinetics was conducted. Results: In rats, the AUC and C, x of FLT increased with increasing oral doses (25 to 1000 mg/kg). The elimination half-life was approximately 2 hours. At 25 mg/kg for 30 days, Cmax of 11 pg/ml was associated with decreases in RBC's. In dogs, the AUC and Cmax of FLT also increased with increasing oral doses (5 mg/kg to 1000 mg/kg). The elimination half-life ranged from 1.8 to 2.8 hours over the dosage range studied. At 5 mg/kg for 30 days, VBC's and platelets were decreased. Their effects were associated with a Cmax of 5.3 pg/ml. In monkeys, Cax and AUC values increased with increasing doses (1 to 100 mg/kg/day). The elimination half-life ranged from 1 to 1.6 hours. A linear, relationship indicated that between percent erythrocytes reduction and log AUC, an AUC of 70 pg hr/ml was associated with a 25X reduction in erythrocytes. Further, in rats and dogs peak plasma concentrations and half-lives were similar for both FLT and AZT when given at equal doses. In the cynomolgus monkey, FLT plasma concentrations and AUC values were 2-3 fold higher, and elimination half-lives were longer (1 to 1.6 hours) than those of AZT (0.6 to 0.7 hours). Conclusions3 An analysis of toxicokinetic data showed a strong relationship between plasma levels of FLT and toxicity in three laboratory animal species. NOTES M.A.1033 VIRAL AND CELLULAR REQUIREMENTS FOR REPLICATION OF HIV IN PRIMARY MONOCYTES Schuitemaker H.,, N.A. Kootstra, M. Groenink, L. Meyaard, R.E.Y. de Goede, H. Huisman, F. Miedema and M. Tersmette Central Lab. Neth. Red Cross Blood Transf. Serv. and Lab. of Exp. and Clin. Immunology of the Univ. of Amsterdam, Amsterdam, The Netherlands The biological phenotype and the prevalence of monocytotropic HIV-1 isolates was studied by cell-free infection and primary isolation studies on sequential blood samples. In cell-free infection studies on primary monocyte derived macrophages (MDM) with 17 nonsyncytium-inducing (NSI) and 14 syncytium-inducing (SI) isolates a correlation between NSI phenotype and tropism for monocytes was demonstrated. These monocytotropic isolates are present at all stages of HIV-1 infection as could be concluded from primary isolation studies. Transient expression in MDM of reporter gene CAT, driven by LTR's derived from distinct HIV-1 variants did not reveal differences. Transfection experiments with full-length infectious, genetically related, molecular clones different for SI capacity and host range indicated restriction at the level of virus entry. Five day treatment with rIL-4, previously described to induce differentiation-like effects in monocytes, prior to infection resulted in the complete absence of HIV-1 expression. IL-4 treatment after infection had no effect. IL-4 did not act by down modulation of CD4 expression nor at the transcriptional level since induction of LTRdriven CAT activity in IL-4 treated monocytes was seen. Concomitant with differentiation, complete inhibition of monocyte proliferation was seen. It appeared that monocyte proliferation is a prerequisite for infection with HIV-1. We conclude that NSI HIV-1 variants are more monocytotropic as compared to SI variants with the level of restriction most likely at entry. Although a certain level of monocyte differentiation seems to be necessary for infection, HIV expression will occur only in macrophages that have been infected before the stage of terminal differentiation. NOTES M.A.1034 A PRE-CLINICAL in vitro TOXICOLOGICAL MODEL USING HUMAN LYMPHOCYTES. Foster, B.C., Wilson, D.L., Tryphonas, H., Whitehouse, L.W. and Khan, Sabih. R. Health & Welfare Canada, Health Protection Branch, Ottawa, Ontario, Canada. Objective: The purpose of this study was to determine whether transformed and peripheral blood leukocytes could be used in a pre-clinical in vitro lymphocyte proliferation assay (3H-thymidine uptake) to determine the potential toxicity of AIDS drugs and whether there would be drug interactions. Methods: Epstein-Barr virus transformed and freshly isolated peripheral blood leukocytes isolated from HIV-1 seronegative donors were incubated in RPMI 1640 media with fetal bovine serum and Conconavalin A in 96 well flat bottom plates. ZDV alone and in combination with diclazuril (DC), fluconazole (FC), itraconazole (IC) and ketoconazole (KC) was added at t-0. 3H-thymidine was added at 72 hr. Cells were harvested at 96 hr and counted for uptake of radiolabel. Statistical significance was assigned using Student's T test. Results: ZDV had a linear dose response on reducing the uptake of 3H-thymidine. IC and KC in combination with ZDV significantly reduced 3H-thymidine uptake compared to that observed with either IC or KC alone. DC and FC had no apparent toxicity and did not alter the toxicity observed with ZDV, whereas IC and KC may contribute to the toxicity of ZDV. Discussion and Conclusions: The results obtained in this study are consistent with clinical findings that IC and KC are highly toxic to leukocytes. This assay may provide an inexpensive, rapid model to determine the pre-clinical toxicity of drugs alone and in combination. M.A. 5 PREFERENTIAL SUSCEPTIBILITIES OF ADHERENT CELLS DERIVED FROM HUMAN B.A.1 5 BRAINS TO AN HIV-1 MUTANT AND HIV-2 AND SIV ISOLATES. Takeuchi.Yasuhiro Hoshino, H., Miura, T.*, Hayami, M.*, Clapham, P.R.** and Weiss, R.A.** Gunma University, Maebashi, Gunma, Japan, *Kyoto University, Kyoto, Japan, **Institute of Cancer Research, London, UK We have been studying a mutant HIV-1, which is named GUN-T1 and can productively infect fibroblastoid cells (BT cells) derived from human brains by a CD4-mediated mechanism. BT cells seem to originate from normal blood vessel cells. Proto-type HIV-1 isolates cannot replicate well in BT cells. The genetic difference in cell tropism between GUN-ly and its parental, proto-type isolate, GUN-1WT, has been shown to be a single base exchange in V3 region in the env gene. In this study, we compared the cell tropism of GUN-1V with those of various immunodeficiency viruses including HIV-2s and SIVs. Methods: HIV-1 isolates (IIIB, LAV-1, ARV, RP, Z84, GUN-1WT and GUN-1V), HIV-2 isolates (GH-1, LAV-2 and SBL6669) and SIV isolates (AGM, MND and MAC) were examined. CD4 -positive HeLa (HeLa-CD4+) and U87 glioma (U87-CD4+) cells, BT cells and M8166, MT-4 and MOLT-4 T lymphocytes were cocultivated with persistently infected cells or inoculated with virus stocks. Virus replication was examined by syncytium formation, reverse transcriptase assay and immunofluorescence. Results and Discussion: All virus isolates infected M8166, MT-4 and MOLT-4 cells. IIIB, LAV-1, RF, GUN-1WT, GUN-1V, LAV-2 and SBL6669 isolates, but not others, efficiently replicated and induced syncytia in HeLa-CD4+ cells. GUN-1V and all of HIV-2 and SIV isolates infected both of BT and U87-CD4+ cells, whereas replication of HIV-1 isolates other than GUN-1V was not detected. These results suggest that brain-derived cells, BT and U87-CD4+ cells, might have similar mechanism(s) of virus reception which is available to GUN-ly, HIV-2s and SIVs, but not to nrntn-tvnP HIV-1 a.

Page  101 M.A.1036 INDUCTION OF CYTOTOXIC T LYMPHOCYTES AGAINST HIV-2 IN MICE IMMUNIZED WITH RECOMBINANT SALMONELLA TYPHIMURIUM Agaarwal, Anita, Sadoff, J.", Markham, P.**, Gallo, R."", and Franchini, 6.* *PRI/DynCorp, NCI-FCRDC, Frederick, MD, USA; "Walter Reed Army Institute of Research, Washington, D.C., USA; **Advanced BioScience Lab, Inc., Kensington, MD, USA; **NCI/ NIH, Bethesda, MD, USA. OBJECTIVE: Salmonella-based oral vaccines have shown to be safe, highly immunogenic, and cost effective. Recombinant Salmonella typhimurium (WR4024) expressing HIV-2 antigens could induce viral-specific cytotoxic T lymphocytes (CTLs) in mice. METHODS: PCR-amplified HIV-2 gag (p17 and p24) and env (amino and carboxy half) gene fragments were cloned into Salmonella. Recombinants were tested for antigen expression by western blot using HIV-2 human serum. Balb/c mice were inoculated orally with three doses of 109 Salmonella recombinants expressing gag and env antigens. Five-eight weeks after the immunization, spleen cells from these mice were stimulated in vitro either with specific peptides or with P815 cells infected with recombinant vaccinia virus expressing HIV-2 env or ga proteins and were processed for CTL by standard methods. RESULTS: The spleen cells from mice immunized with HIV-2 p17 Salmonella recombinants caused 8-10% lysis of P815 cells, and the CTL epitope was mapped to the previously described peptide (HB30). The spleen cells from mice immunized with recombinant Salmonella-carboxy half env demonstrated 20-25% killing of target cells. Antibodies against HIV-2 in the immunized mice were not detected. CONCLUSIONS: Preliminary results show that recombinant Salmonella was able to induce CTLs against HIV-2 in mice suggesting that Salmonella could be an efficient, oral, live vaccine candidate for HIV infection. NOTES M.A.1037 HIV ENVELOPE GLYCOPROTEIN gpl20 INTERACTS WITH ASTROCLIAL BETAADRENERGIC RECEPTORS Levi Giulio* Petrucci T."*, Patrizio M.* and Bernardo A.** Istituto Superiore di SanitA, Pathophysiol.* and Cell Biol." Labs., Roma, Italy. Objective: It has been hypothesized that neural cell damage responsible for the AIDS dementia complex may result from CD4-independent infection, from release of neurotoxic viral-coded products (e.g. gpl20) or from exposure to products (cytokines or others) derived from infected or reactive cells. Since astrocytes, the most abundant neural cells in the CNS, are believed to exert a fine control on neural cell function and are directly or indirectly involved in most neurological diseases, our studies were aimed at testing whether interaction with the HIV envelope protein gpl20 might determine functional alterations in these cells. Results: Using secondary cultures of purified type-1 astrocytes obtained from the neonatal rat cerebral cortex, we obtained the following results: exposure of the cells to 100 pM gpl20 (Genentech, S. Francisco) for 10 min caused a 50-150% increase in the level of cyclic AMP which was counteracted by the p-adrenergic antagonist propanolol. Moreover, 100 pM gpl20 counteracted (by 20-60%) the 20-100 fold increase in cyclic AMP levels induced by the i-adrenergic agonist isoproterenol (1-10 nM). Finally, exposure of the cells to 0.1-10 nM gpl20 for 30 min caused alterations in the phosphorylation pattern of astroglial proteins (analyzed by bidimensional PAGE) qualitatively similar (but quantitatively lower) to those induced by 10 nM isoproterenol. In particular, both agents increased the phosphorylation of the intermediate filament proteins vimentin and glial fibrillary acidic protein, and decreased the phopsphorylation of an unidentified 80 kD protein. Preliminary experiments seem to indicate that, as in the case of cyclic AMP levels, gp 120 partially counteracted the effect of isoproterenol. Conclusions: Our data suggest that gpl20 interacts with astroglial P-adrenergic receptors; interaction with astroglial B-adrenergic receptors may affect the release of neuroactive substances such as taurine, which is modulated by the activation of these receptors. Supported by Project on AIDS of the Italian Ministry of Health. NOTES p \;,CI tz 00 M.A.1038 PROTEIN gpl20 ENHANCES GLUTAMATE TOXICITY IN CULTURES OF RAT CEREBELLAR INTERNEURONS Savio Tiziana and Levi G. Istituto Superiore di SanitA, Laboratory of Pathophysiology, Roma, Italy. Objective: It has been suggested that neural cell damage frequently observed in AIDS patients may be partly related to the neurotoxicity of HIV-coded products, such as the envelope protein gp120. Picomolar concentrations of gp120 cause neuronal death in various non-human neural cell culture systems, and this death has been interpreted as due to competition with neurotrophic factors, to increased calcium influx through voltage-gated calcium channels, or to interaction with the N-methyl-D-aspartate subtype of glutamate receptors. The aim of the present study was to determine whether acute exposure to gpl20 could potentiate the neurotoxic action of glutamate on a defined population of purified neurons. Methods and Results: Cultures highly enriched (95%) in granule neurons were prepared from 8-day postnatal rat cerebella. After 7-8 days of incubation the cells were exposed for 15 min either to a buffered Locke's solution (controls) or to the same solution supplemented with: i) 1 or 100 pM recombinant gpl20 (Genentech, S; Francisco); ii) 0.05 mM glutamate, with or without 1 mM APV (an NMDA receptor antagonist); iii) 0.05 mM glutamate and 1 or 100 pM gpl20. After culturing for another 24 h in their original culture medium, total granule cells and dead granule cells (which bind the fluorochrome propidium iodide) were counted. In controls, dead cells accounted for 4-7% of total. Exposure to gp120 did not alter this proportion, while glutamate caused a 1.5-2 fold increase in the number of dead neurons, which was reversed by APV. In the case of simultaneous exposure to glutamate and gp120 the neuronal cell death was 30-300% greater than that caused by glutamate alone. Conclusions: Since exposure to 1 or 100 pM gpl20 did not induce an increase in the release of endogenous glutamate from the cells, these observations suggest that low concentrations of gpl20 can potentiate the excitotoxic action of glutamate on cultured rat cerebellar granule neurons. Supported by Project on AIDS of the Italian Ministry of Health. M.A1 CYTOTOXIC ACTIVITY AND HIV SENSITIVITY OF INTESTINAL LAMINA M.A PROPRIA LYMPHOCYTES Di Massimo Ann-Marla^, Placido R.*, Bach S.*, MastinoAS, Capobanchi M.$, Fals S. +, Pollone F., Colizzi Y.* Department of Biolog*and of Experimental Medicines, II University of Rome, Italy: "Department of Biotechnology Menarini Ricerche Sud, Rome, Italy; $Institute of Virology, I University of Rome, Italy; +Oastroenterology Unit, University of Regglo Calabria, Italy Objective: The role of mucosal immunity in HIV infection has been analyzed at the level of lamina propria lymphocytes (LPL). Methods: LPL were isolated from the normal colon-rectal mucosa by enzymatic treatment, and phenotypically characterized by flow cytometry. T cell mitogen (PHA, staphylococcus toxin, IL-2) activated LPL and peripheral blood lymphocytes (PBL) were used as effector cells in a 4 hour chromium release cytotaxic assay. Tumor cel lines either uninfected or HIV-infected were used as targets. Results: LPL after activation display high cytotoxicity against a variety of tumor cells, and no preferential kiling was observed when intestinal derived tumor cell lines were used Moreover, LPL display cytotoxic activity against the H9/HIV cell line and this cytotxic activity is mediated by activated T cells. Considering the predominance of CD4+ T cells in the LPL population, their sensitivity to HIV infection has been evaluated using low doses of infectious virus. It has been found, that only activated LPL can be infected by HIV and produce infectous virus. Conclusions: The colon-rectal lamina propria mucosa may represent a first defence barrier to natural HIV infection, and these data show that LPL easily activated in situ might be infected with low doses of HIV, although they are also able to exhert cytotoxic activity. ~j3 S C3 0 a

Page  102 M.A.1040 HUMAN DIPLOID FIBROBLASTS AND EPITHELIAL CELLS PRE-TREATED j )WITH INTERFERON GAMMA PRESENT INCREASED HIV ADSORPTION. A. Dolei, C. Serra, M.V. Area & A. Toniolo, Institute of Microbiology and Virology, University of Sassari, ITALY. Objective: to correlate the permissiveness to HIV of adherent cells to the expression of surface molecules. Methods: HIV-1 (strain HTLV-IIIB) was given to human diploid lung embryo fibroblasts and to other adherent human cell lines, either untreated or pre-treated with recombinant interferon (IFN) gamma. Modulation of membrane molecules was evaluated by radiobinding to living cells of a panel of MAbs directed against membrane markers, or by immunofluorescence. HIV infection was monitored by syncytium assays, RT acativity and p24 ELISA. Results: pre-treatment of cells with IFNgamma increased the surface expression of Class I, 02 -microglobulin, and CD4 molecules, and induced de novo Class II DR and DP antigens; HIV adsorption was increased by a factor of 2-4. Virus yield was reduced only in those cells which completed virus cycle within 2-3 days. Infectivity of adsorbed particles lasted longer in IFN-treated cells, suggesting delayed uncoating. Newly formed p24 and infectious virus were detected 48-72 hrs p.i. in continuous cell lines. Diploid fibroblasts started to produce low levels of HIV only 20-30 days p.i.,which were not influenced by pre-treatment with >FNgamma, whereas physiological doses of TNF enhanced p24 production. No matter of preexposure to IFN, persistent infection with HIV could be established in diploid fibroblasts, with chronic production of HIV antigens and low levels of infectivity. Supported by the 1990 I.S.S. grant N,5206.035 NOTES M.A.1041 IN VITRO BONE MARROW COLONY ASSAYS AND LONG-TERM BONE MARROW CULTURE IN HIV INFECTION. Kaczmarski R S, Sutherland S, McManus T*, Moxham J"*, Moran S. Mufti G J. Departments of Haematological Medicine, Virology, Genito-urinary Medicine*, Thoracic Medicine*", King's College Hospital School of Medicine and Dentistry, London, England. Objectives:The pathogenesis of peripheral blood cytopenias in HIV infection is likely to be multifactorial. Proposed mechanisms include direct infection of haemopoietic cells or progenitors, stromal dysregulation of haemopoiesis, immune-mediated destruction and the myelosuppressive effects of drugs. Bone marrow abnormalities reminiscent of myelodysplastic syndromes suggest the haemopoietic stem cell may be the target of HIV infection however there is conflicting evidence to support this. Methods:To study haemopoiesis in HIV infection we performed in vitro culture of marrow from seropositive patients; marrow was cultured directly on methylcellulose and agar for CFU-Mix, BFU-E and CFU-GM, and long-term liquid cultures. Results.le have studied three HIV patients with peripheral blood cytopenia and three seronegative controls. Preliminary results from direct BM culture show no difference in in vitro colony growth of CFU-Mix, BFU-E or CFU-GM in patients and controls. There are no differences in the appearance of the stroma, haemopoietic foci or fat cells in long-term culture, and CFU-GM from cells harvested from LTC are the same. Conclusions.We are performing in situ hybridization and PCR on harvested colonies and stromal layers to determine whether we can detect genomic integration of HIV in these cells, however these preliminary results suggest that in HIV, ineffective haemopoiesis may be due to some inhibitory factors which are absent in in vitro culture systems. We are conducting studies to investigate this possibility. NOTES C3 cn ~c 03 M.A.1042 A CL.L L aIMs 8uSCEI.L tO A SSLCTPrava N aM0PAor -TaOPTC HVI s5AIN (BaL) LACK OF CYTOn&XZCX-.. CD4 DOWaoUuGLATON AND aRXc ta msurhmuiacd 2salatta._Claudoi*, veroness, 7r.*, LorZi 7,~, pal, R.**, Gallo, R.C.*, and Lusso, P,*. *Laboratory of Tumor CaeL BiolOgy, NCI, Bethesda, MD 208921 **Departamnt of Cell Biology, ABL, Kensington, MD 20899. Obj~ctive; TO establish a unified model'for the study of the biology of divergent HZV isol&tes. Nethodls Syncytia assay, p24 Ag capture, flow cytametry, eCD4 competiton assay. neaslts( PFrom the H2V-1lSZ isolate, we obtained a biologically pure macrophagetropic substrain, HNV-l., which is unable to penetrate PtM-activated PBL, ao well as several CD4+ neopl ati"e T-cell or promonooytic cell lines. We identified a humanr CD4+ EaV-tranaforned B-celL clone, eT62, which is susceptible to infection by both HIV-I.. and HIV-lBlM Vowevert while IV-lAi established a persistent infection 'ths completle t B'ance of synoytia formati'oinfad cellular death, HIV-1 IZ infection was highly cytopathic. No receptor interf~erence was observed between HIV-i5- and HIV-11T, since IT62~~,l cooltt be uperinfected with HIV-1~ raesaulr in a ptiduiotive ooinfiteL with rapid syncytia formation, in addition, the 50% inhibitory concentration (ZIo) of recombinant soluble CD4 for HIV-1 was a pproximately 100-fold higher t-ha for HMV-1IZB,3 suggesting a lower CO4-7l120 binding affinity for )l-1V-A. Disoussion The ZT62 cell line way represent an optima-uAified model for the study of both macrophage.-tropio and_.T-eeLk tropia 1HV Isolates. The lack of interferenoe between nooOfztopathi d w-* tratenic ne V-ll isoatof suggests the possibility of a dirt iftteractiL*4 lA $Lf hebween divergent HIV isolates. M.A.1043 MECHANISMS OF GANGLIOSIDE-INDUCED MODULATION OF CD4 MOLECULE Saggioro Daniela*; Panozzo, M.*; Calabro', M.L.*; Callegaro, L.**; ChiecoBianchi, L.* - * Institute of Oncology, IST Biotechnology Section, University of Padova, Italy, **Fidia Research Laboratories, Abano Terme, Italy. OBJECTIVE: To study the mechanism of monoganglioside GMI induced CD4 down regulation in CD4 positive cells. METHODS: 1) CD4 down-modulation was measured by flow cytometric analysis in T cell lines and in HeLa cells expressing a CD4 molecule defective for the cytoplasmic region; 2) HIV-infection in GM1-treated cells was evaluated by RT analysis and by CAT assay; 3) Northern blot analysis of' specific CD4 mRNA expression in presence of GM1. RESULTS: The complete down-modulation of CD4 antigen requires one hour of GM1 treatment (250 ug/ml) and is promptly reversed when the monoganglioside is removed. However, the comparative analysis of exogenous GMI loss with CD4 re-expression indicates that exogenous GM1 loss from cell membrane is slower than CD4 reexpression. GM1 cells treatment is always accompanied by an inhibition of HIV infection. Contrary to TPA-induced CD4 down-modulation, synthesis of CD4-specific mRNA is not affected by ganglioside treatment. Furthermore, GM1 induces downmodulation of CD4 molecules defective for the cytoplasmic region while TPA treatment has no effect in cyt- CD4 HeLa cells. CONCLUSION: GM1-induction of CD4 down-modulation may act throught mechanisms which do not require the phosphorylation of CD4 cytoplasmic domain.

Page  103 M.A.1044 M.A.1045 HIV INFECTION OF A CELL LINE DERIVED FROM NOVEL GASTROINTESTINAL T-LYMPHOCYTE POPULATION (CD4-, CD8-, TCRap+) Hattori Toshio*, Maeda Y*; Maeda M**; Takatsuki K* *Kumamoto University Medical School, Kumamoto, **Kyoto University, Kyoto, Japan. Objective: Dysfunctions of local immunity are presumed to be prerequisite for systemic immunodeficiency. To obtain clues for the mechanisms of immunodeficient states in HIV infected individuals, we have asked whether a T cell line derived from novel T-lymphocyte population in the gastrointestinal tract could be infected with HIV or not. Methods: IL-2 dependent cell line (43/T) was established from a patient with adult T cell leukemia involving the epithelium of the gastrointestinal tract (Lancet in press). The cell line harbored a unique cell surface markers of the original leukemia cells, namely, CD4-, CD8-, bearing T-cell receptor ap heterodimer. 43/T cells were inoculated with HTLV-III MN followed by the cultivation. Culture supernatants were assayed for p24 antigen, and the infection was also confirmed by the use of mAbs; VAK5 (a-p24) and t39.1 (a-V3 of HTLV-III MN). Results: P24 antigen was detected in the culture supernatants 40 days after infection. Thereafter, the infected cell line continuously produced viral antigen for more than 3 months. Over 96% of the infected cells were positive for both VAK5 and A 39.1 mAbs. Moreover, syncytia formation of infected cells co-cultured with uninfected Molt-4 Clone 8 cells was observed. Discussion and Conclusions: Infection of this novel subset by HIV-1 may explain immunodeficient states of the gastrointestinal tract in HIV infected individuals and would be useful to analyze the mechanisms for CD4-independent HIV-1 infection. NOTES p si tj a cCO sY NOTES f M.A.1046 PRODUCTION OF ADHERENT CELL LINES SUSCEPTIBLE TO LYMPHOTROPIC VIRUSES Toniolo, Antonio*; Serra, C.*; Area, M.V.**; Babudieri, S.**; Dolei, A.** *Dpt. of Biomedicine, University of Pisa & **Inst. of Microbiology and Virology, University of Sassari, Italy. Objective: Adherent cells capable of expressing selected leukocyte markers were produced to analyze the relationships between surface receptor molecules and susceptibility to some lymphotropic agents. Methods: Somatic hybridization of human leukocytes with human epithelial cell lines (together with immunoselection) allowed to obtain 22 different cell lines expressing selectively the CD3, CD4, CD21, or DR membrane receptors. These cells were challenged with HIV, EBV, or HHV-6; virus yield and cytopathic effects were evaluated (virus titration, measurement of viral antigens, detection of infectious foci). Results: As expected, productive infection with HIV-1 occurred in CD4-positive hybrids (although the sensitivity of individual lines varied), and EBV (strain B95-8) replicated in cells carrying the CD21 marker. Certain hybrid lines expressing high levels of the HLA class II DR molecule were capable of producing infectious HHV-6 (Uganda strain] but this occurred also in cells of endothelial and thyroid origin which are devoid of this marker; thus, the receptor specificity of HHV-6 remains unclear. Conclusions: By means of somatic cell hybridization we obtained cell lines producing constitutively leukocyte markers which are relevant to viral susceptibility. This technique should allow to explore virus-receptor relationships even in cases were S genomic constructs specific for the marker under study are not available. M.A.1047 ABILITY OF ANTISERA RAISED TO SYNTHETIC PEPTIDES REPRESENTING THE V3 REGION OF DIFFERENT HIV-1 ISOIATES TO NEUTRALIZE INFECION OF BOTH PRIMARY MONOCYTES/MACROPHAGES AND T-CELL LINES Butto'Stefano, McDanal C.B., Greenwell. T.K.. StreileinA.F., LangloisA.J., BolognesiD.P., Matthews T.J. Department of Surgery. Duke University Medical Center, Durham, NC 27710. The pathogenesis of AIDS remains incompletely understood. Although HIV causes a cytocldal infection of CD-4 bearing cells In vitro, this phenomenon alone is insufficient to fully explain the immunologic dysfunction associated with AIDS. In particular, the role of monocytes/macrophages is poorly understood. Objective. To study the role of the variable V3 region from the viral glycoprotein gp120 in the infection of monocytes/macrophages in vitro. Antibodies directed at this region are known to block infection of T-cell lines with high efficiency. The mechanism of this neutralization process is not clear, although it is thought to occur after the CD4/gp 120 binding. The major objective of this study is to test if V3 antibodies display similar blocking activity for the infection of macrophages. Methods, We synthesized peptldes 20 to 25 residues in length containing sequences from the V3 region of several field isolates as well as the monocyte-tropic HTLV-IIIBa-L strain. Some of these peptldes were used to immunoafflnity purify a number of human HIV-1 positive sera. Results and Discussion. Each of these isolates s able to infect productively both T-lymphocytes and monocytes/macrophages. The role and effect of guinea pig antibodies to the synthetic peptides, as well as immunoaffinity purified human antibodies, in neutralizing the infection by these "dual tropic" isolates on both primary monocytes/macrophages as well as T-cells, will be discussed. 4 'a 4 4 Ns i^

Page  104 I4 M.A.1048 PERSISTENT INFECTION OF A HUMAN NEURONAL CELL LINE BY HIV-1: VIRUSCELL INTERACTION. Orecchia. Anqela; Equestre, N.; Carlini, F.; Novello, F.; Federico, M.; Santoro, R.; Genovese, D.. Lab. Virology, Istituto Superiore di SanitB, Roma, Italy. Objective: To study the mechanism of interaction between HIV-1 and Central Nervous System (CNS) cells. Dtsfunction of the CNS is a prominent feature of the AIDS syndrome. The precise role of the HIV In CNS disfunction remains to be determined. For this purpose an "in vitro" model was established to study the mechanism of the persistance of HIV infection in neuronal cells. Methods: A human neuronal cell line, SK-N-SH, was infected with the HTLVIIIb strain. Viral production was monitored via cocultivation with an HIV-1 susceptible cell line, CEN-SS. Characterization of the persistently infected cell line has been performed by: 1) the release of antigens and viral particles; 2) IFA test; 3) molecular analysis of the viral DNA. Results: An infected cell line was established.After six months, cells are permanently infected. The cocultivation of SK-N-SH HIV-1-infected cells with CEM-SS resulted in positive RT, p24 antigen capture and IFA assays. Moreover, CEM-SS showed cythopatic effect with syncytia formation. Integration of HIV-1 proviral genome was revealed by Southern blotting and PCR analysis. Discussion: We demonstrated the susceptibility to HIV-1 infection of SK-N-SH cell line, resulting in the establishment of a persistant infection. This cell line will be used in an attempt to isolate neurotropic HIV variants from cerebral spinal fluid of patients with neurological AIDS syndromes. NOTES M.A.1049 SOLUBLE CD4 (sCD4) SUPPRESS THE ANTIGEN-DRIVEN PROLIFERATIVE RESPONSE IN CERTAIN BCG SPECIFIC T-CELL CLONES. Magnus Gidlund, Eva Halapi, Hans Wigzell and Rolf Kiessling. Department of immunology, Karolinska Institute, Stockholm,Sweden. We have investigated the effect of soluble recombinant CD4 (sCD4) on the antigen (BCG) specific response of polyclonal T-cell lines or on T-cell clones derived from these. Some of the antigen specific clones could be suppressed in their antigen driven response by the addition of sCD4, while others, including the parental polyclonal T-cell line, were not. The suppression of the specific T-cell response was reversed by the addition of anti-CD3 or IL-2, and was independent of the amount of antigen. A decreased capacity to produce IFN-gamma in response to the antigen by the addition of sCD4 was seen only with those clones that were also inhibited in their specific proliferative response. This model may. be used to further deliniate the interaction between T-cell and the antigen presenting cell, and the finding may limit the possible in vivo use of sCD4 in the therapy of HIV infections. NOTES M.A.1050 HIV 'INECTION OF MONOCYTES SELF-PERPEITUATS HIV ENHANCER ACTIVITY THRODGH NF-kB INDUCTION Alcami, J.; Bachelerie, F.; Arenzana-Seisdedos,F; Virelizier JL. Unite d'Immunologie Virale, Institut Pasteur, Paris. Objective. Tissue macrophages are probably the main reservoir for HIV in vivo. Chronic HIV replication in the macrophage lineage would be expected to result from the constitutive presence of nuclear NF-kB, the transcriptional activator of the HIV enhancer. However, in U937 cells, a myelomonocytic line highly permissive to HIV, levels of NF-kB are barely detectable.we have tested the hypothesis that HIV infection can enhance its own replication in U937 cells by induction of NF-kB. Methods. U937 cells were infected with the LAV-BRU strain of HIV-1. Specific enhancer binding proteins (EBP) were detected by bandshift assay. Binding specificity was assessed by the use of cold oligonucleotides competition, consensus-mutated sequence and antibodies against the p50 moiety of NF-kB. Function of detected EBP was measured by transient transfection of vectors containing the viral enhancer in its active or mutated form using a luciferase reporter gene. Results. Chronic phase of HIV infection was associated with induction of EBP in a TNF independent manner. The nuclear factor induced was functionally active as demonstrated by permanently increased enhancer activity in infected cells and was indistinguishable from NF-kB in their binding and antigenic properties Conclusion. In U937 cells, HIV infection is associated with induction of NF-kB and increased HIV enhancer activity.This phenomenon is likely to participate in the perpetuation of HIV infection in monocytes. M.A.1051, DIFFERENTIAL HIV-1 REPLICATION IN NAIVE AND MEMORY CD4 T-CELLS FROM M.A.1051 HIV-PATIENTS ACCORDING TO ACTIVATION PATHWAY EMPLOYED. Cayota Affonso*; Vuiller, F.*; Scott-Algara, D.*; Feuillie, V.** and Dighero, G.* * Unit6 d'Immunohimatologie et d'lmmunopathologie, Institut Pasteur, Paris, France. ** H6pltal de 'Institut Pasteur, Paris, France To study HIV replication in naive and memory CD4 T-cells, as well as, the role of different stimulatory pathways on viral replication, peripheral blood mononuclear cells (PBMCs) from 8 HIV-l-seropositive carriers (SPC) and 4 AIDS patients were positively selected into CD4+CD45RA+ and CD4+CD45RA- subsets using an "EPICS 752" cell sorter (Coultronics, Haleah, FL).These highly-puriled CD4 T-cell subsets (>98%) were cultured In 96-well plates at 2.104 cells per well with rIL-2 (50 U/ml) to which one of the following was added: PHA (5 ug/ml), PMA (10 ng/ml), TNF-alpha (20 ng/ml), soluble antl-CD3 mAb (0.5 ug/ml) and Immobilized anti-CD3 mAb (0.5 ug/ml). HIV.- replication, as measured by p25 antigen production was examined at 4, 7, 11, 15 and 18 days of culture. Upon PHA stimulation, memory CD4 T-cells preferentially replicated the virus in the SPC group ( p25 was detected at day 4). For AIDS patients, no difference in viral replication was observed in the two subsets. On the other hand, when PMA and TNF-alpha were used, p25 production was exclusively observed in memory subsets of two groups. When soluble antl-CD3 was employed, p25 production was exclusively observed in the naive subset from 2 out of 4 SPC patients. Interestly, in these 2 patients, p25 antigen was exclusively produced by memory cells upon PMA or PHA stimulation. However, in the AIDS group, when soluble anti-CD3 was used, 3 out of 4 patients replicated HIV-1 exclusively in the memory subset and only one patient in the naive subset When anti-CD3 and anti-CD4 were crosslinked, both naive and memory cells produced similar levels of p25 in the AIDS group. These results, suggest that HIV-1 replication may depend upon the activation pathway employed, which could differentially activate HIV-1 enhancers through different cis-acting regulatory elements. This differential HIV-1 replication by CD4 Tcells depends not only on the subpopulation studied but also on the current clinical status of the patient. Studies are now being carried out in our laboratory to elucidate the role of these and other T-cell stimulatory pathways In viral replication by naive and memory CD4 T-cells.

Page  105 M.A.1052 CONTROL OF VIRAL REPLICATION IN MONONUCLEAR CELLS OF HIV-ANTIBODYPOSITIVE AND HIV-ANTIBODY-NEGATIVE HEMOPHILIACS Mihoczy, Elisabeth*, Mannhalter, J.W.*,***, Schramm, W.**, Eibl, M.M.*** *Immuno AG, Vienna, Austria; **Department of Medicine (Innenstadt), University of Munich, Germany; ***Institute of Immunology, University of Vienna, Austria Objective: To compare viral replication in mononuclear cells (MNC) of HIV-antibodypositive (HIV+) and HIV-antibody-negative (HIV-) hemophilia patients (HP) after in vitro infection with HIV-1 IIIB. Methods: Long-term cultures of HIV-1 IIIB-infected PHA blasts were set up in IL-2 -containing medium. Virus released into the culture supernatant was determined with HIV-1 antigen capture ELISA and reverse transcriptase assays. At the end of the incubation period mononuclear cells (MNC) and culture supernatants were checked for the presence of infectious HIV-1. Results: Thirteen HIV+ HP, 9 HIV- HP and 4 healthy controls were studied. In HIV- HP and control subjects MNC were readily infected with all virus doses employed (1x104, 1x103, 1x102 TCID5Q HIV-1 IIIB), viral replication was observed throughout the longterm culture, and infectious HIV-1 was found in both culture supernatants and MNC at the end of the long-term culture (46-60 d). In contrast, MNC of HIV+ HP were much less permissive for in vitro infection with HIV-1. In case of an infection viral replication peaked between d 7 and 14 and decreased to background levels at about d 38. Infected MNC completely lost the expression of CD4; however, at the end of the incubation period (46-60 d) HIV-1 was isolated from MNC of 4 out of 8 HIV+ HP. Conclusion: MNC of HIV+ HP have the capacity to control replication of HIV-1 after in vitro infection. Positive virus isolation at the end of the long-term culture, however, suggests that the virus is not completely eliminated. NOTES M.A.1053 Production of HIV-1 Protein but not Infectious Virus in G, Arrested MT-2 Cells. Kieran. Rosemary., Doherty, R.R., & McPhee, D. Macfarlane Buret Centre for Medical Research, Fairfield Hospital, Victoria, Australia Objective: To investigate the mechanism of inhibition of HIV replication in GI (stationary phase) arrested HTLV-1 transformed MT-2 cells. Method: Cells were arrested in G1 phase using 5mM hydroxyurea, shown by FACS analysis of DNA content & >90% reduction of 3H-TTP uptake. After 18 h pretreatment with drug, MT-2 cells were infected with HIV-1, washed & cultured with drug for a further 48 h. Virus replication was studied by RT activity, MT-4 plaque assay, immunofluorescence (IF), Western Blot (WB), CPE and cell viability. Results: Production of cell-free infectious virus was almost completely inhibited in Gt arrested cells compared with actively dividing cells which released progeny virus 24 h post infection. However, HIV proteins, including env and gag, were detected by IF & WB. Although Gi arrested cells were not fusogenic, cell viability was reduced compared with both infected and uninfected controls. Conclusion: While the replication cycle of HIV was inhibited in G, arrested MT-2 cells since no infectious progeny was released, HIV protein production was observed. Inhibition may occur either at integration of DNA intermediates whereby unintegrated HIV DNA would be transcriptionally active, or packaging of particles at the cell membrane. NOTES 0 NI ~I h, oo t4 M.A.1054 INHIBITION OF HIV-REVERSE TRANSCRIPTASE BY SYNTHETIC OLIGONUCLEOTIDES. Idriss Haitham and Stammers, David, K. Wellcome Research Laboratories, Dept. of Molecular Sciences, South Eden Park Road, Beckenham, Kent, England. Objective: The reverse transcriptase (RT) of the HIV virus is an important target for anti-HIV drugs such as azidothymidine (Zidovudine). In addition to the RNA dependent DNA polymerase activity, RT also catalyse a DNA dependent DNA polymerase activity as a step in the formation of pro-viral DNA. The substrate activity and inhibitory properties of short double stranded DNA oligomers have been investigated. Methods: Activity of RT can be assayed by following the incorporation of 3HdGTP onto a poly(rC).Poly(dG) primer template (rCdG), using a conventional filter assay. Oligonucleotides can compete with this substrate and therefore, it is possible to determine the potency of their inhibition for RT by determining the IC50 values. Three series of partially complementary oligonucleotides were tested for their RT inhibitory effect: (i) 20/18mers where the 18mer is analogus to the priming site binding sequence of t-RNA Lys3. (ii) N/9mers where N= 15-28, the 5' overhang consisting of dATPs. (iii) 'Hairpin' oligomers related to (ii) in which the two oligomers are joined by four thymidine bases. Mode of RT inhibition and IC50 values were determined for oligonucleotides belonging to each of the three sets above. Results: Oligonucleotides of series (i) inhibited RT with IC50 of 3 piM. The mode of inhibition was that of mixed competitive. Series (ii) oligonucleotides inhibited with the IC50 (120-80 pM) decreasing with increasing oligonucleotide length, the mode of inhibition was that of mixed competitive. A parabolic pattern was observed for series (iii) oligonucleotides. The IC50 values being significantly lower(5-20 gM), with the 32mer (n=10), having the lowest IC50 (5 iM). The pattern of inhibition was competitive. Conclusions: All three types of oligonucleotides can bind to RT, show substrate activity and compete with rCdG. Addition of a 'hairpin' to the oligonucleotide seems to strengthen this binding and affect the mode of inhibition. Similar studies should provide useful information on the structure/activity relationship for DNA-dependent DNA polymerase 1 activity of RT, to be of use in studies of the structure of the enzyme. ýl/ CHARACTERIZATION OF A EUCARYOTIC FORM OF THE HIV-1 REVERSE M.A.1055 TRANSCRIPTASE. West. Anthony*; Roberts, T. M.*; Kolodner, R. D.* * Harvard Medical School, Dana-Farber Cancer Institute, Boston, Massachusetts, USA Objective: Examination of the HIV reverse transcriptase with respect to subunit structure, kinetics and processivity. Methods: The baculoviral expression system was used to make large quantities of the products of the pol open reading frame. Monoclonal antibodies were made against the reverse transcriptase and an immunoaffinity purification step was developed that provided high yields of very pure enzyme with the proper structure. Results: Analysis of the subunit structure by sucrose gradient velocity centrifugation indicated that the enzyme exists in several forms in equilibrium. When dNTP was the variable substrate, the enzyme kinetics deviated from standard Michaelis-Menten kinetics in that both substrate inhibition and negative cooperativity occur. Further examination demonstrates that dNTPs may act both as substrate and as noncompetitive inhibitors. The processivity of the baculoviral reverse transcriptase was found to be several fold higher than that observed for the previously described bacterially expressed form of the enzyme. Conclusions: The work presented here calls into question the notion that the enzyme subunits form only a dimer. The enzyme kinetics provide a model that accounts for the substrate inhibition and that is consistent with the negative cooperativity. In this model, the reverse transcriptase is a multisubunit holoenzyme where noncompetitive inhibition is allosterically mediated by one subunit, the 51 kDa, binding nucleotide and down regulating the activity of the polymerizing 64 kDa subunit This model has important implications for drug development. In addition, the processivity analysis indicates that this source of enzyme is ideal for the examination of the polymerase. 'a c3 8a c3

Page  106 M.A.1056 ACTIVATION OF HUMAN IMMUNODEFICIENCY VIRUS-1 (HIV-1) BY DNA DEMETHYM 1 LATION. O'Brien, M.C. and Mitsuva. Hiroaki. The Clinical Oncology Program, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA. Objectives: To study the effect of DNA demethylation on HIV expression in various infected cells. Methods: 5-Azacytidine (AZC) and its analog (dihydro-5-azacytidine, DHAC) were employed as agents that induced DNA demethylation. An HIV-infected T-cell line, ACH2, that produces a low level of HIV-1; a persistently HIV-l-producing cell line, H9; and a subacutely HIV-1-infected normal CD4+ clonal T-cell line, TMI1/HIV; and subacutely infected lectin-activated normal donors' peripheral blood mononuclear cells (PHA-PBM/HIV) were used. Results: Following exposure of ACH2 cells to AZC or DHAC, a profound potentiation of HIV replication was observed as assessed by production of envelope glycoproteins, reverse transcriptase (up to 35 fold), and p24 Gag protein (up to 47 fold). Alterations in the DNA methylation pattern in ACH2 were demonstrated by Southern blot hybridization using the isoschizomer restriction enzymes Msp I and Hpa II and HIV-1-specific probes. Northern blot analyses revealed that without AZC treatment, 4 and 2 kilobase (kb) HIV mRNA species were predominant as compared to the unspliced 9.2 kb species HIV mRNA, while following AZC exposure a significant increase in the amounts of all HIV mRNA species occurred in a dose-response manner. Reinfection of ACH2 cells with newly synthesized HIV virions was not associated with the observed potentiation. HIV viruses produced by phorbol-12-myristate,13-acetate (PMA)-stimulated ACH2 cells could productively infect H9 cells, indicating the HIV-1 genome in ACH2 cells was a functionally complete one. AZC-induced potentiation of viral expression did not occur in any of H9, TMII, or PHA-PBM/HIV cells. Southern blot hybridization demonstrated that proviral DNA in these highly HIV-l-producing cell lines had already been extensively demethylated, and no alteration in DNA demethylation pattern of proviral DNA was noted following AZC treatment. Conclusions: These data suggest that DNA methylation may play a role in regulating HIV expression at least under certain circumstances, although its relevance to in vive events remains to be further investigated. NOTES M.A.1057 Characterization of HIV-1 reverse transcriptase using a set of 23 murlne monoclonal antibodies Restle Tobias*; Sczakiel, G.#; MOiler, B.*, Pawita, M. and Goody, R.S. Max Planck Institut fOr med. Forschung, Abtellung Blophysik, Heidelberg, F.R.G. #Deulsches Krebsforschungszentrum, Heidelberg, F.R.G. Objective: Reverse transcriptases (RT) are interesting targets for chemotherapeutic treatment of retroviral diseases. A detailed knowledge of the correlation between structure and function of this enzyme farrily is the prerequisite for a rational drug design. Due to the lack of suitable crystals for X-ray diffraction analysis, we have used monoclonal antibodies as an alterative to obtain structural information. Methods: Mice were immunized with highly purified recombinant HIV-1 RT and hybridomas were prepared using a standard protocol. Screening was performed by ELISA and Western blot analysis. AN 23 antibodies obtained were purified and tested by appropriate methods for influence on the enzymatic activities of RT. In addition, their effect on dimerization of the heterodimeric protein was investigated, since it is known that polymerase activity of RT correlates strictly with dimer content. Results and Discussion: Two different neutralizing antibodies were identified. One effects the polymerase activity and to a certain degree the RNase H activity, whereas the other inhibits exclusively the latter activity. MAb 28, which blocks the polymerase moiety, interferes with the dNTP binding domain of RT as shown by fluorescence spectroscopy. Some antibodies influence dimer formation of RT in different ways, which allows detailed Investigations on domains involved in subunit interaction within the heterodimer. Interestingly, none of the 23 antibodies recognizes recombinant HIV-2 RT. The results taken together could lead to a better understanding of some important aspects of reverse transcription. Currently, a detailed interpretation of antibody effects regarding the enzyme structure is not possible due to absence of information on the precise antibody epitopes. Epitope mapping Is in progress. NOTES M.A.1058 Expression of HIV-2 reverse transcriptase In E. coll and characterization of the recombinant enzyme. Barbara MQiler, T. Restle, H. KOhnel# & R.S. Goody. Abt. Biophysik, Max-Planck-lnstitut for med. Forschung, Heidelberg, F.R.G.; #Chemotherapeutisches Forschungsinstitut, Georg-Speyer-Haus, Frankfurt/M., F.R.G. Objective: When studying the structure and function of HIV reverse transcriptase (RT), it is of special interest to compare the enzymes from HIV-1 and HIV-2, which are functionally closely related but show considerable differences in amino acid sequence. Besides comparison of structural and biochemical properties, we are interested in the interaction of the enzymes with highly specific inhibitors of HIV-1 RT which do not inhibit HIV-2 replication (TIBO derivatives; Pauwels et al., 1990). Methods: To obtain large quantities of protein for structural and biochemical studies, HIV-2 RT was expressed as recombinant protein In E. coll. The RT coding sequence from HIV-2D194 (KOhnel et al., 1990), which shows 60% identity on amino acid level to the corresponding sequence from HIV-1BH10, was cloned into the expression vector pKK233-2. Plasmids for separate expression of the two RT subunits, p66 and p51, and for co-expression of both subunits were constructed. Recombinant proteins were purified and dimerization properties of the RT subunits were investigated using analytical gel filtration. Results: As demonstrated previously for HIV-1 RT (Restle et al., 1990), both subunits of HIV-2 RT are enzymatically active, and activity is strictly associated with the dimeric forms of the protein. With the exception of p51, specific polymerase activities are comparable to HIV-1 RT. Inhibition of HIV-2 RT by TIBO derivatives was investigated. The finding that HIV-2 RT is insensitive against this type of inhibitor (obtained by Pauwels et al. using disrupted virus particles as enzyme preparation) was confirmed in experiments with purified recombinant HIV-2 RT. Co-expression of HIV-2 p66 with HIV-1 p51 in E. coli did not lead to formation of stable chimeric heterodimers, Indicating that the dimerization domains of the two proteins must be different. Recombinant HIV-2 RT shows weak cross reaction with aHIV-1 antisera, but does not react with a set of 23 monoclonal antibodies prepared against HIV-1 iRT. M.A.1059 RETINOIDS STIMULATE HIV-1 EXPRESSION IN MONOCYTES. Jim A. Turpin and Meltzer, M. S. Walter Reed Army Institute of Research,Washington DC USA Objective: Retinoids are potent mediators of mononuclear phagocyte differentiation. Since vitamin A is a dietary component, the effect of this compound on HIV replication is of interest. Methods: Blood monocytes were purified from PBMC by countercurrent centrifugal elutriation and cultured in medium with 10% human serum and 1000 U/ml MCSF with and without retinoic acid (RA). In certain experiments, human serum was delipidized by chloroform:methanol extraction to remove endogenous serum retinoids. Monocytes were infected at 7 days with HIV-1ADA. Results: Monocytes infected with a 1:10000 HIV show maximal reverse transcriptase activity (RT) of < 5 x 103 cpm/ml by day 23. Addition of RA at 1.0 and 0.1 nM throughout the culture interval decreased the time for appearance and increased levels of RT activity with a peak expression of > 8 x 106 cpm/ml by 18 days. Monocytes cultured in delipidized serum fully supported HIV replication, but did not show the RA-dependent stimulation of virus growth. Conclusions: Retinoids are potent stimulators (> 1000-fold) of HIV replication in monocytes. Their action requires a lipid soluble component in serum and likely reflects changes in monocyte differentiation that enhance virus growth.

Page  107 M.A.1060 DIFFERENTIAL EFFECTS OF L-THYROXINE AND PROSTAGLANDINS ON THE tat-MEDIATED TRANSACTIVATION OF HIV-1 LTR IN CD4 CELLS AND MACROPHAGES. Demirhan, I.; Chandra, A.; Hofmann, D.; Chandra, P. Laboratory of Molecular Biology (ZBC) Frankfurt University Medical School (FRC) Objective: Transactivation of HIV-1 LTR is the most important event in the expression of structural genes. Our aim is to look for endogenous factors which affect this process in target cells (CD4 cells and macrophages) modulating the pathophysiological state of the desease. Methods; Jurkat cells or U937 cells were transfected with plasmids containing HIV-1 LTR-CAT or tat gene sequences (FEBS Lett. 236: 282-286, 1988). Transfected cells were cultivated with or without the test compounds. CAT assays were carried out in triplicate. Results: L-Thyroxine was found to inhibit CAT expression in U937 cells to 45% at 1pM, but no inhibition was observed in Jurkat cells. Prostaglandins Al and B1 showed a 30% inhibition in U937 cells at 5pM, but again no inhibition in Jurkat cells. Prostaglandin D2 inhibited the transactivation in U937 cells to about 45% at 5pM, but showed a stimulation to 191% in Jurkat cells. Prostaglandins El and E2 had a similar effect in both types of cells, showing about 25% inhibition at 5pM. Prostaglandin F1 alpha had no effect in both cell lines. Conclusions: Hypothyroxinemia has been reported in patients in a progressive state of the desease. Our results with L-Thyroxine and prostaglandins may be interesting for pathophysiological investigations. NOTES M A 1061 INVESTIGATION OF THE ROLE OF DNA METHYLATION IN THE MAINTENANCE OF HIV LA* * TENCY IN MONOCrTES Walker. Mary Clare*; Bouchard, Jacques**; Leclerc, Jean-Marie***; varin, Mireille*; Patel, Pravin C*. *Inatitut Armand-Frappier, Laval, Que., Canada, **Bureau of Human Prescription Drugs, Health and Welfare Canada, Vanier, Ont., Canada. ***Hbpital SteJustine, Universit& de Montr6al, Montreal, Que., Canada. Objective: The objective of the study was to examine the possible role of DNA methylation in the maintenance of HIV latency in cells of mononuclear phagocyte lineage. Methods; 1.1 cells, a clone of the human promonocyte cell line 0937, which is chronically infected with the LAV strain of HIV-1, were obtained from T.M. Folks (NIH). The pU3R-III CAT plasmid, containing the reporter gene chloramphenicol acetyl transferase (CAT) directed by the HIV-1 long terminal repeat (LTR) (AIDS Res. Ref. Reagent Prog., AIDS Prog., NIAID, NIH. Contributor: J. Sodroski), was cotransfected into U937 cells (ATCC CRL 1593) by electroporation along with the pSV2Neo plasmid, which contains the neomycin gene. A neomycin resistant transfectant clone, U937-2D4, which constitutively expresses a low level of CAT, was selected. V1.1 and U937-2D4 cells were treated with 5-azacytidine (5-AZAC), a nucleoside analog of cytosine which induces the hypomethylation of newly replicated DNA into which it is incorporated, to determine if the treatment would induce the expression of HIV (01.1 cells) or CAT (U937-2D4 cells). For comparison, the cells were also treated with phorbol myristate acetate (PMA) which activates the HIV LTR through the cellular DNA binding factor NF-xB. Results: 10'5 M 5-AZAC and 10'" M PMA induced the expression of HIV in UI.1 cells as demonstrated by the appearance of reverse transcriptase activity and HIV p24 antigen in the supernatants of treated cultures and augmented the CAT activity of the U937-2D4 cells 2 to 3 fold. Discussion and conclusions: The results are consistent with a role for DNA methylation in the maintenance of HIV latency in monocytes and we are presently determining the methylation status of the CpG sites in the HIV LTR of latently infected cell lines. NOTES M.A.1062 CHARACTERIZATION OF RT INHIBITORY ANTIBODIES IN SERA OF HIV-1 INFECTED INDIVIDUALS DeVico, A.*; Rahman, R.*; Gallo, R.**; Sarngadharan, M.*; di Marzo Veronese, Fulvia* *Dept. of Cell Biology, Advanced BioScience Laboratories, Inc., Kensington, MD, USA; **National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Objective: We became interested in the identification of the antibody binding domain related to the induction of RT inhibitory antibodies and in the characterization of the mechanism involved in this inhibition as a means of identifying RT domains important for enzymatic activity. Methods: RT specific Immunoglobulins were purified from sera of HIV-1 infected subjects by immunoaffinity chromatography on recombinant RT coupled to activated CH-Sepharose 4B. Purified IgGs were assayed for direct inhibition of HIV-1 RT activity and the mechanism of this inhibition was analyzed in detail using steady state kinetics varying the antibody concentrations and either template-primer or substrate concentrations. Results: Independently from the disease stage, anti-RT specific IgGs from all the HIV-1-positive sera we tested were found to inhibit HIV-1 RT activity, although with different titers. Thus the distinction present in the literature between inhibitory and non-inhibitory antibodies is not valid. The double reciprocal plots of enzyme activities against either template-primer or substrate were linear at all antibody concentrations and formed patterns of noncompetitive inhibition. Conclusions: HIV-1 RT inhibitory antibodies appear to directly interfere with -' the polymerase activity of RT by binding to sequences outside the catalytic site O of the enzyme.. 2 1 0TERINANS OF THE HIV-1 TRRB1INEMBRA P N MIj i I FOR VIUS ENTRY M.A.1063 AND SNCIA OR N. Dedera, Douglas; Ratner, Washingto university chaool of Medicine, St. Ioui, ID, U.S.A. Objective: This study was performed to define regions of the transrmmbrane protein (TM) involved in virus entry and syncytia formation. Methods: Site-directed mutations were made which altered the N-terminal fusion domain (F3, F6), conserved potential N-linked glycosylation sites and cysteine residues. Infectivity, replication properties, and syncytia induction were examined for all mutant viruses. Recambinant vaccinia virus vectors were also used to expression the mutant envelope proteins without other HIV proteins. Results: Mutations in the fusion domain resulted in partial (F3) or complete loss (F6) of syncytia formation and attenuation of infectivity for several T cell lines, but retained single cell lytic effects. Additionally, F6 demonstrated increased virus yield campared to the parental HXB2 clone due to the loss of syncytial cytopathicity. Conserved regions for potential N-linked glycosylation were found to be involved in proper attachment of the surface protein (SU) to TM. Mutation of these sites revealed virus with reduced infectivity and syncytia formation capacity. Conserved cysteine residues were found to be necessary for processing the 160 kDa precursor to the SU and TM products. Discussion and Conclusions: Several regions of TM were identified for their involvement in virus entry, syncytia formation, and proper envelope structure and processing. Furthermore, the F6 virus which was observed to be completely deficient in syncytia formation capacity allowed demonstration of two distinct cytopathic mechanisms due to syncytia formation and single cell lysis. These findings may provide important| insights into differences in fusogenicity of viral isolates obtained at different stages of disease, and should assist in analysis of therapeutic drugs and antibodies.'

Page  108 0 C? go) M.A.1064 A NOVEL, DIPYRIDODIAZEPINONE INHIBITOR OF HIV-1 REVERSE TRANSCRIPTASE ACTS THROUGH A NON-SUBSTRATE BINDING SITE. Wu, Joe C.; Warren, T.C.; Adams, J.; Proudfoot, J.; Skiles, J.; Raghavan, P.; Perry, C,; Potocki, I.; Farina, P.R.; Grob, P.M., Boehringer Ingelheim Pharmaceuticals, Inc., Ridgefield, CT, USA Objective: A novel, non-nucleoside inhibitor BI-RG-587 inhibits HIV-1 RT (RT-1) with an IC, = 85 nM. The interaction of this inhibitor with RT-1 was investigated. Methods: A tritiated azido-analogue of BI-RG-587 was synthesized ([HIl-BI-RJ-70) and used to label RT-1 in the presence or absence of dGTP, template-primer, tRNA or thio benzimidazolone (TIBO). The labeled enzyme was subjected to stoichiometric analysis and labeling specificity by gel-filtration, SDS-PAGE, and autoradiography. Results: Photoaffinity label BI-RJ-70 was found to reversibly inhibit RT-1, but not RT-2, in the dark, with Ki = 160 nM. Upon irradiation with UV-light, the inhibitor becomes covalently attached to RT-1 and the inactivation becomes irreversible. In the presence of BI-RG-587 or TIBO, RT-1 was highly protected from photolabeling because BI-RG-587 and TIBO compete with BI-RJ-70 for binding to the same site in RT-1. The binding site is not at the enzyme catalytic site because substrates afford no protection for RT-1 from photolabeling. The tritiated photoaffinity probe was found to stoichiometrically and selectively label p66 such that 1 mol of the probe inactivated 1 mol of RT-1. The photoaffinity labeling was highly specific since in the presence of a 100-fold excess of cytosolic proteins from human peripheral blood leukocytes, p66 was the only protein labeled by [3H]-BI-RJ-70. NOTES M.A.1065 CELLULAR DNA SYNTHESIS AND MITOSIS ARE NOT REQUIRED FOR THE ESTABLISHMENT OF INFECTION OF T CELL LINE WITH HIV-1 Handa, A.; Takeuchi, Y.; Hoshino, Hiroo Gunma University School of Medicine, Maebashi, Gunma, Japan. Objective: Upon de novo infection of cells with retroviruses such as murine leukemia viruses or arian sarcoma virus, production of progeny viruses have been detected after cellular DNA synthesis and mitosis. On the other hand, in humans infected with HIV-1, syntheses of viral mRNA or viral proteins have been observed in cells, such as macrophages, peripheral blood lymphocytes or neurons, whose division will take place quite rarely in vivo. Thus, we examined whether cells in the stationary phase of cell cycle can be infected with HIV-1 in vitro. Methods: MT-4 human T-cell line cells were cultivated in RPMI 1640 medium supplemented with 1% fetal calf serum (FCS) and then infected with HIV-1. Production of viral proteins, reverse transcriptase activity and infectious virus and formation of proviral DNA were examined up to 4 days after infection. Cell cycle of MT-4 cells cultivated in 1% PCS medium was assessed by 3Hthymidine or 5-bromodeoxyurine incorporation assay and mitotic index determination. Results: DNA synthesis and mitosis was not detected in MT-4 cells cultivated in 1% FCS medium for over 48 hours. These stationary MT-4 cells could be infected with HIV-1: Proviral DNA was formed and viral proteins and infectious virus were produced efficiently by these cells. Conclusion: Stationary human T-cells can be infected with HIV-1 in vitro. NOTES G\ c3 G\ M.A.1066' DIFFERENTIATION-INDUCING AGENTS OF LEUKEMIC CELLS ENHANCE INFECTION OF T CELLS WITH HIV-1 IV WO Seld. Jun-ichtr*; Yamamoto, A*; Takeuchi, Y.'; Hoshino, H.* *Gunma University School of Medicine, Department of Hygiene, "Nipponkayaku Co. Ltd., Japan. Objective: Dimethyl sulfoxide(DMSO) has often been used as a solvent of many chemicals to be assayed for antiHIV-1 activities. As we noticed that DMSO enhanced HIV-1 infection in vitro, we examined effects of DMSO and other so called differentlation-Inducing agents on HIV-1 infection of T cells. Methods: MTT assay was used to estimate cyotoxicities of chemicals and cytophathic effects of HIV-1. Indirect immunofluorescence and reverse transcriptase assay were used to examine the expression of viral protein in freshly or chronically infected cultures. Results: MT-4 cells were infected with HIV-1 and MTT assay was done after incubation for five days. Amounts of HIV-1 required to show similar cytopathic effects on MT-4 cells were 25-fold smaller in the presence of 1.0 % DMSO than in its absence. Amounts of HIV-1 produced by MT-4 cells which had been infected with HIV-1 and cultured in the presence of 0.3-1.0 % DMSO for four days were as 3-12 times high as those of control cultures. DMSO increased percentages of HIV-1 antigen-positive cells after infection with HIV-1. Similar enhancing effects on de novo infection with HIV-1 were also found when dimethylformamide, hexamethylenebisasetamide, sodium butyrate or retinoic acid were used. DMSO and these differentiation-inducing agents also augmented production of HIV-1 by chronically infected MT-4 or MOLT-4 cells, whereas neither production of human T-cell leukemia virus type-1 by chronically infected MT-4 or MT-2 cells, nor production of HIV-1 by chronically infected monocytic cells was not enhanced by DMSO. DMSO did not increase numbers of MT-4 cell plaques under agarose medium after infection with HIV-1, suggesting It did not affect an adsorption step of HIV-1 infection. Conclusion: These findings suggested that DMSO and other differentiation-inducing agents enhanced HIV-1 infection of T cells at a step which took place after virus adsorption. M.A.1067 DNA-BINDING-CAPACITY OF THE HIV-1 INTEGRASE EXPRESSED IN THE BACULOVIRUS-SYSTEM Rodner, Birmit; Vinga-Martins, C.; Muller-Lantzsch, N. University of the Saarland, Dept. Virology, Homburg, FRG Objective: Expression and functional characterisation of different forms of the HIV-1 integrase in the Baculovirus-system Methods: The complete gag and pol reading frames and a clone representing the integrase of HIV-1, which is encoded at the 3'part of the pol gene, were expressed in the eucaryotic Baculovirus-system as fusion-proteins. These proteins were used in a SouthWestern-Blot to investigate their DNA-binding-capacity. Results: The protease, encoded at the 5'part of the polymerase reading-frame, cleaved the gag/pol-precursor into the different proteins. In immunoblots and kinetics of the infection we could show the processing of the gag/pol-precursor into the gag-proteins (p17, p24 and p16) and the pol-proteins (RT, RNase H and integrase). This way we obtained a naturally processed integrase. The unprocessed integrase was expressed at a much higher level than the naturally processed protein. We could show unspecific DNA-binding of the unprocessed integrase. Conclusion: The expression of the integrase in the gag/pol-clone implies that ribosomal frameshifting has taken place in the overlap region between gag and pol. The protease was translated and proteolytically processed the gag/pol precursor as observed in vivo in HIV-infected cells. Comparative studies of the DNA-binding-capacity and -specifity of the two forms of recombinant integrase are under investigation. Supported by the BMFT grant No 11-074-88

Page  109 M.A.1068 CHARACTERIZATION OF CELL-TO-CELL SPREAD OF HIV-1 Sato. Hironori and Martin, A. Malcolm Laboratory of Molecular Microbiology, NIAID, NIH, Bethesda, USA. Objective: To compare HIV infections mediated by cell-free virus or cell-to-cell spread. Methods: We have compared the kinetics of virus replication following infections mediated by cell-free virus or virus producing cells, using Southern, Northern, and Western analyses. We also examined the effects of metabolic inhibitors and monoclonal antibodies on both types of HIV infection. Results: The capacity of HIV infected cells or cell-free virus present in supernatant fluid (obtained from the same infected cultures) to initiate a productive virus infection was directly compared. Unintegrated proviral DNA, detected by Southern blot analysis, was present within 4 hours following cellto-cell spread and was rapidly followed by new viral protein synthesis (12 hr) and progeny virion production within the first 24 hr. In cells, simultaneously infected with cell-free virions, proviral DNA was not detected until 48 hr post infection, and progeny particles appeared 48 hr later. The rapid infection, initiated by cell-to-cell spread, required both CD4 and reverse transcription. Conclusion: HIV production following cell-to-cell spread is extremely rapid. The underlying mechanism(s) will be discussed since they may reflect events occurring in vivo. NOTES M.A.1069 A RAPID TEST SYSTEM FOR EVALUATING EFFECTORS OF HIV-1 TRNSMISSIO Gaer, Sharon;Epstein, Jay Food and Drug Administration, Center for Biologics, USA Objective:The purpose of this study was to establish a quantitative in vitro test system for rapidly evaluating the effects of antibodies, drugs and cytokines on transmission of BIV-1. Methods:Indicator T-cell (H938) and monocyte (U38) lines (Felber and Pavlakis) were used to develop an in vitro test system. These cell lines contain copies of stably integrated BIV-LTR-CAT gene. CAT activity increases with production of viral TAT protein (Sci.'88). For virus inocula, H9 cells were similarly infected with one of four strains of HIV-1 (IIIB, MN, RF, Z). Infected cells and culture supernatants were cryopreserved for use as cell-associated and cell-free virus inoculum, respectively. Virus inocula were analyzed by EM and reverse transcriptase activity (RT). The signal strength of each strain of HIV was determined at 3, 7 and 14 days. Dilutions of cellular and cell-free inocula were mixed with 1.4 x 104 H938 and U38 indicator cells in 200ul volume in 96-well microtiter wells and incubated at 37 C. At the times indicated, microtiter plates were centrifuged, supernatant was aspirated and cells were lysed by 3 freeze, thaw cycles. CAT activity (cpm/hr/20ul) was determined for cell lysates by diffusion assay (Neumann). Results:T-cell CAT signals 3 days after infection were proportional to the number of infected cells added (up to the number of indicator cells) and averaged 322-fold over controls. Cell-free inoculum produced about a 200-fold increase in CAT activity. Significant CAT activity occurred in monocytes 14 days after infectionscell-associated virus inoculum increased CAT activity an average 86-fold over controls, while cell-free virus caused a 3-fold increase. CAT activity was highly reproducible (<3%) for a given inoculum and linear over 5 hr counting. T-cell CAT signals correlated with RT activity and virion count of cell-associated virus inoculum, but not cell-free inoculum. Monocyte CAT signals did not reflect RT activity or virion count. ConclusiontThe y vitro assay system described represents a powerful means of rapidly evaluating preferences and efficacy of possible effectors of HIV transmission based on transmission of divergent strains of HIV-1 to T-cells and monocytes across a range of cellular and- cll free viru; inecul-&, NOTES 07 IN HIV-1-INFECTED CELL CLONES THE REPLICATION OF SUPERINFECTING HIV MAY BE M.A.1071 BLOCKED AT THE RETROTRANSCRIPTION STEP Taddeo, Brunella; Federico, M; Titti, F; Orecchia, A; Saggio, I; Carlini, F; Verani, P; Rossi, GB. Lab. Virology, Istituto Superiore di Sanita, Rome, Italy. Objective: To define at which step superinfecting HIV replication is blocked in both producer (DI0) and non producer (F12) HIV-1-infected Hut-78 cell clones exhibiting homologous viral interference (Federico et al., AIDS Res Hum Retr, 1989, 5: 385). Methods: D10 and F12 CD4-down-regulated HIV-1-infected Hut-78 clones were HIV-2 superinfected. The fate of superinfecting virus was then followed by DNA-PCR, to monitor viral RNA retrotranscription after the entry of superinfecting HIV-2. Results: DNA-PCR performed on cell lysates from HIV-2-superinfected D10 or F12 clones shows that superinfecting genomic RNA is retrotranscribed at least partially. In fact, DNA-PCR performed 2-4 hours after HIV-2 superinfection was clearly positive in both DIO and F12 clones using vpx primers, while no signals were detectable when primers mapping shortly downstream of HIV-2 "primer binding site" (PBS) were utilized. The positive signals in vpx PCR disappeared 16-24 hours after superinfection. In addition, PCR performed on high molecular weight DNA from DIO and F12 clones after superinfection was at all times negative, thus confirming the inability of superinfecting HIV to complete its life cycle. Finally, infectious virus was released from non-producer F12 clone after pNL43 (HIV-1 infectious molecular clone) transfection and integration, suggesting that the homologous viral interference is not operating at the post-integration step. Conclusions: These data indicate that the resistance to superinfection of an HIV-1-lnfected clone may be due not only to CD4 down-regulation, but also to the inhibition of full retrotranscription of the superinfecting HIV.! ~I M.A.1070 INTERACTIONS BETWEEN MAB AND SOME EPITOPES OF HLA-CLASSI-82 A. ANT.IGEN COMPLEX PREVENT HIV PRODUCTION IN PBMC AND MT4 CELLS. -Corbeas, Pierre*;.miliani Stphase*; Olive, Daniel**; Favero, Jean**; Dornand, Jacques***; Devaux, Chritian*. *Centre detri des molules anti-HlV, CRBM, CNRS-INSERM, Montpellier, France; **Unit6 INSERM 119, Maseille, France; ***Unit6 INSERM 65, Mcatpeier, France. Objective: We have previously shown (1, 2) that monoclonal antibodies (mAb) specific for 82 microglobulin, can delay HIV cytopathic effect in MT4, an HTLVI immortalized T cell line, and, viral production by Peripheral Blood Mononuclear Cel (PBMC). The main purposeof the present study was to invesigate the effect of anti-HA Clal mAb treatntan the replicativecycle aHIVI andHIV2. Methods: Cells were exposed to virus in the presence or absence of anti-HLA Classl mAb using the previously published infection protocols (2) that allow to precise which stage of the viral cycle could be inhibited by the mAb treatment. The infection was monitored by measurement of reverse transcriptase activity, and numeration of yncytia fomation. Results: Among 6 anti-HLA classl mAb, 2 were able to decrease HIVI and HIV2 cytopathic effect in MT4 cells and to inhibit H IV productn by PBMC. However, no effet was evidenced when target Tcells were tumorallines. In contrast o the results obtained by treatng cels with OKT4A, an ant-CD4 mAb, the treatment with anti-HA Class I mAb neither prevented syncytia fonnatian nor inhibitd HIV infection when incubated with the PBMC before viral exposure. Viral production was reduced only when the mAb was added into the culture media immediaty after PBMC have been exposed to the virus (3). Discussion and Conclusions: These new data emphasize the necessity of studying further the mechanism by which binding of mAb to some regions of HLA Claasl-2m antigen complex may interfere with HIV replication in Tcels. This work is under progres. References (1) Devaux C., Boucraut J., Poirir G., ct al. Res. Immunol. (1990) 141: 357-372; (2) Corbeau P., Devaux C., Kourilsky F. and Chenann J.C. J. Virol. (1990) 64: 1459-1464; (3) Corbeau P., Olive D., and Devaux C. Mr. J. Immunol. (1991) in pres. IO 0

Page  110 0 M.A.1072 COPARATIVE I zNVSTIGATIOS ON HI V-1 RIB aos L FRAMESHIFTING IF..I I VIVOQ AND IN VITRO Reit,. eide, Steinhaus, H., Nloosmyer, D*. and Hauser, N. Gesettschaft fur Biotechnologische Forschung, Braunschweig, FRO * Deutsches Printen Zentrum, G6ttingen, FRG Ribosomal frameshifting is essential for the expression of the catalytic proteins (Pro, RT, Int) of HIV-1. A defined sequence of the gagpol overlapping region has been shown to be necessary for this specific process. We have developed an expression system which allows the quantification of frameshifting in vivo and in vitro. The method is based on the expression of an N-terminally extented firefly luciferase gene which depends on -1 frameshifting at the HIV-1 derived sequence, in order to be translated as a functional enzyme. Our data show a reduced frameshift efficiency in mammalian cells in comparison to frameshifting in vitro. No significant tissue or species specificity could be detected. At present we are investigating whether specific antiviral agents or in vivo parameters such as the status of differentiation or virus infection are able to affect the frameshift efficiency. Using the in vitro system we found that the stem loop structure is dispensable for frameshifting. However, deletion of the stem loop leads to a reduced frameshift efficiency. The molecular mechanism of the ribosomal frameshifting such as the influence of the RNA sequence on translational elongation is under investigation. NOTES M.A.1073 HIV-1 REPLICATION IN HTLV-I TRANSFORMED CELLS. Antonelli, Guido; Turriziani, Ombretta; Capobianchi, Maria R.; Amicucci, Paola; Di Marco, Paola; Guamnu, Dong; Riva, Elisabetta; and Dianzani, Ferdinando. Institute of Virology, University "La Sapienza", Rome, Italy. Objective: To investigate the biological aspects of HIV-1 replication in HTLV-I transformed cell lines. Methods: C8166 or MT4 cells were infected with several HIV-1 isolates at high M01 (1-2 TCIDJ cell). Progeny virus as well as viral DNA, viral proteins, and % of infected cells were examined after one cycle of replication (20-24 hrs for C8166 and 40-48 hrs for MT4 cells). Results: Following infection, only 5-20% of cells became productively infected, as determined by IF, with a very low ratio of progeny infectious virus produced per cell (ranging between 0.5 and 10 TCIDSJcell). The cells are not intrinsically resistant to the infection since after multiple cycles of infection 100% of them became infected. The infectious particles do not attach efficiently to the cells (40-60% of the inoculum was recovered in the supernatants) and this does not seem to be due to interfering soluble gpl20. Furthermore once adsorbed, the virus sloughed off into supernatants even 4 hrs after the infection. Viral DNA and proteins, which are produced in exponential rate during the first hrs of infection (starting from 2 hrs Pl for DNA and 6-8 hrs PI for proteins) do not parallel the production of infectious virus, since accumulation of these products occurs also very late, when most infectious particles are already assembled. Conclusions: The process of replication of HIV-1 in HTLV-I tranformed cell lines does not seem to be very efficient since only a minority of cells became infected under high moi conditions and since only few viral infectious particles were released from each infected cells. Furthermore, many viral particles were seen by EM inside the cells, and levels of infectious virus yield did not correlate with production of viral proteins, thus suggesting that during the HIV-1 replication cycle many defective particles could be produced in both the supernatants and the cytoplasm. Work supported by a grant from Ministero della Sanita' (AIDS project). NOTES M.A.1074 HIV PROTEINASE MUTANT IN CELL CULTURE B.Heindl H.Nltschko and K.von der Helm Max-von-Pettenkofer Institut, University of Munich. Munich, Germany Retroviral maturation requires activity of the virus encoded protease which processes the viral gag and gag-pol protein precursor. The HIV encoded protease might be a suitable target for anti-viral therapy using inhibitors which block selectively the viral encoded protease but do not impair cellular functions. The presence of such an inhibitor in HIV infected cell culture caused the production of viral particles which were non-infectious (1). Using in situ made inactivating protease mutation we found also immature and non-infectious HIV particles produced. This tool is currently used to study the mechanism of how the HIV protease is activated by (auto)-cleavage from the gag-pol zymogen precursor. The protease mutant which is unable to cleave and activate itself autocatalytically could be "trans"-processed by a (wild-type) HIV protease to yield active mutant enzyme in vitro. Upon transfection of this mutant via an infectious HIV-cDNA into cultured cells immature HIV particles were produced in a first phase. In a later replication phase an increasing rate of the particle population appeared to be mature. The results will be discussed in context of the mechanism of protease activation. M.A.1075 INHIBITION OF HIV-INFECTIVITY AND REPLICATION BY a-LIPOIC ACID Baur A', Harrer T2, Peukert M3, Jahn G', Kalden JR2, Fleckenstein B' 'Institut for Klinische und Molekulare Virologie, Universitat Erlangen /NGrnberg, Loschgestr. 7, 8520 Erlangen, FGR. 2Medizinische Klinik III und Institut for Immunologie der Universitat Erlangen/NOrnberg, Krankenhausstr. 12, 8520 Erlangen, FRG. 3ASTA Pharma AG, WeiBmuillerstr. 45, 6000 Frankfurt 1, FRG. Objective: To study the antiviral activity of a-lipoic acid, a naturally occuring disulfide compound, that has been applied for years for the treatment of polyneuropathies and hepatic disorders. Methods: Acutely and persistently infected T-cells were treated with various doses of o-lipoic acid. Reduction of viral expression and activity was determined by plaque titration and RT-activity. Viral markers were assessed in HIV-infected patients before and during treatment with a-lipoic acid. Results: In vitro, o-lipoic acid significantly inhibits CPE, RT-activity and plaque forming units (PFU) in acutely and persistently infected cells. At lower doses of the compound, mainly the PFU were reduced, suggesting that the compound affects the infectivity of virus particles. First virological data from patients treated with a-lipoic acid indicate, that antiviral effects are also seen in vivo. A reduction of infectious viral titers was observed as predicted from the in-vitro experiments. Discussion: Alpha-lipoic acid may be suitable for the treatment of HIV-infection. C>1 a C> C> C>

Page  111 M.A.1076 BIOCHEMICAL CHARACTERIZATION AND MUTAGENESIS OF HIV-1 ENDONUCLEASE Drelich, M., Wilhelm, R. and Mous Jan F.Hoffmann-La Roche Ltd, Basel, Switzerland Objective: The identification of the endonuclease active site of the HIV-1 integration protein (IN) by site-directed utagenesis. Biochemical characterization and structure/function analysis of the IN endonuclease and integrase activities. Evaluation of active site nutations when introduced into infectious proviral clones. Mrthods: The HIV-1,HILV-IIIB IN protein tagged with a hexa-histidine tail has been expressed in E.coli. Hig salt extraction followed by affinity chrcmatography over a Ni-chelate colum resulted in mg quantities of soluble, seni-pure and biologically active enzyme. A convenient assay for endonuclease activity of IN was established using radiolabeled double stranded oligonucleotides that mimic the ends of the linear proviral DNA intenmediates as substrates. For structure/function studies highly conserved residues also found in other retroviral endonucleases were changed by oligonucleotide-directed mutagenesis. Results and Discussion: Recombinant HIV-1 IN was biocheaically characterized in terms of substrate specificity, divalent cation requirement and effect of salt and ATP. Endouclease activity was judged by the renoval of the 3' dinucleotide GT from the substrate. In the presence of Mg2+ (5 nM) this cleavage occurred with marked specificity. In the presence of Ph2+ (2 nM) the endonucleolytic cleavage appeared less specific. On the other hand, under these conditions integrative recombination between labeled oligonucleotides was observed. Endonuclease activity was inhibited by EDTA, EGTA, salt concentrations in excess of 10 iM MFC12, 5 Yn MtC12 and 100 n NaCI, and by ATP above 2 nM. Finally, preliminar experiments indicated that certain anino acid exchanges in a well-conserved TDNG motif, 80 anino acids downstrean of the zinc-binding domain, either abolished activity or dranatically changed the specificity of the cleavage reaction. We are now in the progress of evaluating the effects of these mutations on the integration activity of IN, "in vitro" and in the context of an infectious nroviral INA clone. NOTES M.A.1077 virus IsoLArrTcni FOM LYMHO TES OF ANTI-HIV(+) PERSeS: IENriFICATICW OF A VIRUS WITH UNUSUAL MIRPHIEDGY. Kozlov Andrei P., Glebov A.V., Traugott M.N., Nikolaeva E.V., Skadova V.M., Malykh A.G., Kurbatova T.V. AIDS Reference and Research Lab, Leningrad, 193167, U.S.S.R. Objective: To study viruses infecting anti-HIV(+) persons in Leningrad area. Methods: PBL isolated by Ficoll technique were incubated in presence of PHA and IL-2. HIV presence was shown by antigen detection (ELISA) and electron microscopy. Viral nucleic acids were detected by PCR using primers to the gag and env regions of the viral genome. PCR products were analysed by Southern hybridization. Results: In 70 per cent of cases HIV antigen was detected in cultural medium. In some cases virus with morphology typical for HIV was detected by electron microscopy. PCR analysis has shown fragments typical for known HIV isolates. In case of one patient, where no HIV antigen or virus particles were found, another virus with specific morphology was detected by electron microscopy. The size of the virus is 110-160 nm. It looks similar to but district from known herpesviruses. Molecular characteristics of this virus are currently under study. Conclusions: HIV isolates from cells of infected Leningrad residents do not differ from known HIV isolates by parameters studied. The virus with specific morphology described in the present study may be a novel virus involved in pathogenesis of AIDS. NOTES 0 NN ^I C1 i t^ h, cs 00 M.A.1078 A NOVEL POLYMERASE CHAIN REACTION FOR HIV Bawa Nimrat; Manjunath N; Shankar P; Rajamouli P; Broor S; Seth P All India Institute of Medical Sciences, New Delhi, India. Objective: To develop a simplified non radio-isotopic Polymerase Chain Reaction (PCR) and to use this method on HIV isolates from India. Methods: PCR was done using digoxigenin labeled uridine triphosphate in the reaction mixture, so that it could get incorporated in the amplified product. After 10 cycles, nitrocellulose dipsticks to which the oligonucleotide probe had been immobilized as a dot, were dipped in the reaction tubes followed by 1 cycle of denaturation and annealing. The dipsticks were then removed, washed and successively incubated with enzyme labeled anti-digoxigenin antibody and substrate. H-9 cells were infected with 10 viral isolates from infected Indians (+ve western blot for HIV-1); lymphocytes from 3 infected individuals were co-cultivated with stimulated human lymphocytes. One week later PCR was done on the infected cultured cells. Results: A positive result was seen as a blue dot on the dipstick.This method was found to be specific on known controls.Using this technique we could confirm 11 isolates from India to be HIV-1. Discussion and Conclusions: The dipstick hybridization method described simplifies PCR to the extent that it is accesible to developing countries as it does not need radio-isotopes or an UV transilluminator. M.A.1079 A NEW HIGH-SENSITIVE IMUNO-GOLD-SILVER TEST (IGST) FOR DETECTION OF ANTIBODIES TO HIV Mekushinov Krassimir,Vassilev V. Beshkov D. * Laboratory on AIDS,Military Medical Academy, Bioagrogen,* Central laboratory on AIDS,Bulgarien Medical Academy,Sofia,Bulgaria Ob9ective: To develop a high-sensitive and cheap test for screening of antibodies to HIV-1 -' Methods: We used a nitrooellulose membrane,adsorbed with HIV-1 lysate (DiLPont),ten-fold diluted sera for testing and own protein A-gold conjugate.Our new step is silver enhancing with silver nitrate and chlorides of iron as reductor,which enhanced the visualization of the golden spots 100 to 1000 times.We compared IGST with Westrn blot(DuPont) and EIA(Organon,ABBOTT) on 107 sera.We mesured the titer of antibodies in HIV-positive sera with IGST and EIA. Results: We found the IGST as specific as Western blot and at least asi sensitive as EIA.The mean titers of HIV-antibodies in positive sera were 10 - and 1-2 for IGST and EIA respectively. CoInclusions: This study suggests,that IGST is a simple,easy to read, cheap and high-sensitive techniqe,which could be at least as effective asi EIA for screening of HIV-antibodies. NI NI c3

Page  112 K) M.A.1080 PCR DETECTION OF INTEGRATD HIV-1 DNA IN PERIPHERAL BLOOD LYMPHOCYTES (PBL) FROM INFECTED INDIVIDUALS AT VARIOUS STAGES OF DISEASE USING A NON-ISOTOPIC MICROTITER PLATE ASSAY. Dragon. Elizabeth, Kung, K., Wolfe, L.,, Casareale, D., and Giron, J.* PCR Division, Roche Diagnostic Systems Inc, Fair Lawn, NJ USA 07410 and *Flushing Hospital, New York. OBJECTIVu: TO evaluate a non-isotopic method for the detection of PCR amplified HIV-1 products using DNA isolated from 50 HIV-1 seropositive individuals (at various stages of infection).In addition, different methods of PBL isolation and processing were evaluated. METHODS: PBL, isolated from blood using Ficoll-Hypaque, LeucoPREP, or a red blood cell lysis procedure employing saponin, were extracted using Proteinase K. Target DNA was amplified using either gag or pol region primer pairs and analyzed either by a non-isotopic microtiter plate assay with solid-phase probes specific for each primer pair and an avidin-horseradish peroxidase conjugate, or by a standard isotopic oligomer hybridization (OH) assay. These amplification reactions incorporated the Uracil DNA glycosylase (UNG)-restriction based sterilization to prevent carry-over mediated false positive reactions. REStULTS: The data show direct correlation between PCR results and HIV-1 infection. The microtiter plate assay results show concordance with the OH assay. Standardization of the assay from run to run has been possible with the use of a cloned plasmid control which gives a sensitivity of 10 HIV-1 copies with an intra-assay c.v. less than 15%. Analysis of this plasmid positive control and extracted samples from the negative control population show a signal/noise ratio greater than 10 at the 10 copy level.:ONCLUSION: The non-isotopic microtiter plate assay for the detection of PCR products is a sensitive and specific assay that can be used for the confirmatory liagnosis of HIV-1 infection. NOTES M.A.1081 EVALUATION OF A NEW AUTOMATED TEST (VIDAS HIV 1 + 2) FOR THE DETECTION OF HIV ANTIBODIES SUIPHON Edith *, VERNET. G *, DALBON P.*, BARIN F. ** * bioMerieux S.A, Marcy l'Etoile, France, ** C.H.U, Tours, France Objective: To evaluate a new HIV 1 + 2 screening test developed on the automated immunoenzymatic system VIDAS on a panel of HIV 1 and HIV 2 positive sera. Methods: Plastic tips, used as the solid phase, are coated with 2 synthetic peptides corresponding to transmembranous proteins from HIV 1 and HIV 2. Sera or plasma are placed into a ready-to-use strip which contains all reagents. Bound antibodies are revealed with an anti-human IgG monoclonal antibody coupled to alkaline phosphatase and with a fluorescent reaction. The result is obtained within 35 minutes. Results: The sensitivity was assayed with 2 anti HIV 1 low titer panels (n = 30) from BBI (Boston, USA) and a local panel of early HIV 1 and HIV 2 seroconversions (n = 17). All sera reactive in Western blot (at least P 24 / gp 160) gave a positive signal in the VIDAS test. Furthermore, it detected two sera (one HIV I and one HIV 2) positive for antigen and negative for antibodies in Western blot and ELISA tests. A sensitivity of 100 % was achieved with 203 sera positive for HIV 1 and HIV 2 with various origins and at different stages of infection. None of the 200 negative sera which were tested gave a false positive result. Conclusion: Sensitivity and preliminary specificity studies show that the performance of the VIDAS HIV 1 + 2 screening test is similar or better than that of clinical HIV immunoenzymatic tests and that it allows a rapid detection of HIV antibodies with minimum laboratory handling. NOTES M.A.1082 QUANTITATION OF INFECTIOUS VIRUS FROM UNFRACTIONATED WHOLE BLOOD (WB) IN.A HIV-1 INFECTION Arivoshi Koya, Weber J. St Mary's Hospital Medical School, London, UK. Objectives: To compare the practicability and the implication of WB Tissue-CultureInfectious Dose (TCID) in comparison with plasma TCID and PBMC TCID. Methods: HIV-1 levels were quantified simultaneously in heparinized whole blood, plasma and PBMC by limiting dilution analysis in 15 HIV-1 infected individuals at different stages. The various culture were maintained using cord blood mononuclear cells (CBMC) stimulated with PHA for 3 days and fresh donor cells were added into the culture every week. The supernatant were monitored for p24 Ag by capture ELISA. A culture which produced more than 200 pg/ml of p24 Ag was regarded as positive. Results: WB TCID could be determined in 100% of the blood samples (1-lxl02 TCID/ lii WB) whereas PBMC TCID in 92% (5-5x103 TCID/lx106 PBMC) and plasma TCID in 28% (1-lx10 TCID/lml plasma). The estimated number of lymphocytes in the volume of whole blood which gave WB TCID was well-correlated with the number of PBMC which gave PBMC TCID in the individuals whose plasma TCID was low. Discussion and Conclusions: The use of WB for quantitation of HIV infectious units obviates the need for lymphocyte separation, and is therefore cheaper and easier to perform. We could reliably and repeatedly measure HIV TCID titre in WB, whereas plasma viraemia was only detected in about 30% of the subjects. WB TCID is influenced by the titre of HIV in plasma, PBMC and by the lymphocyte number. However, the reliability of the measurement and the reduction of titre with therapy may lead to its use as the easiest TCID assay. M.A.1083 ECTKN OF HIV PROVIRAL DNA IN THE NEEDLES OF INTRAVENOUS DRUG USERS IN CITY OF NEW HAVEN NEEDLE EXCHANGE PROGRAM Robert Heimer*, Edwin C. Cadman*, and Edward H. Kaplan" *Dept. of Internal Medicine, "School of Organization & Management, Yale University, New Haven, CT. ect- Employment of polymerase chain reaction (PCR) to amplify HIV proviral DNA flushed from s ringes returned by IV drug users participating in New Haven's Needle Exchange Program. Determination of HIV prevalence and incidence among program participants can be calculated and c mpared to estimates. Methods. DNA remaining in syringes was flushed with non-ionic detergents and purified by ptenol/chloroform extraction. A 191 bp region of the gag gene of the HIV DNA was amplified by two rounds of PCR--15 cycles with annealing 4~C below the Tm and 35 cycles with annealing 7~C below the Tr--and identified by Southern blotting with an oligomer probe. As controls,.2 copies of an HIV plasmid or DNA from 3000 WBC from HIV+ patients were reproducibly positive. Needles with <5 IlI residual blood were 93% positive. No false positives were detected using *h or human genomic DNA. Results: Of 228 needles tracked by the program in the first 2 months of its operation, 95 (41.6%) were positive. Needles given to and returned by the same person were 37.5% positive; needles reuurned by someone else were 50% positive. We also tested 80 "street" needles returned by, but not di pensed to, 46 program participants and 48 needles from a "shooting gallery". Of "street" needles, 6.5% were positive; of "shooting gallery" needles, 92% were positive. Co ncusions: Estimates of the prevalence of HIV infection among 59 program participants ranged from 6 % based on the return of at least one positive needle to 37.3% based on return of 50% positive. P evalence was the same as that in needles given to and returned by one person. Estimates will be revised throughout the course of the program, permitting repeated testing and increased accuracy. S quential testing will enable estimates of incidence based on the detection of the first appearance HIV+ bl t The data fronm the first six mnths of the _nroaram will b renrsnted at the mentinn C> a C> a C> C>

Page  113 M.A.1084 REVERSE TRANSCRIPTASE ACTIVITY MEASURED BY ELISA Eberle, Josef*; Seibl, Rudolf** * Max v. Pettenkofer-Institute for Hygiene and Medical Microbiology, Ludwig- Maximilians- University, Munich, Germany; ** Boehringer Mannheim GmbH, Biochemical Research, Penzberg, Germany. Objective: To develop an alternative assay for testing HIV reverse transcriptase (RT) activity in cell culture supernatants, avoiding the disadvantages of the standard procedure (radioactive reagents, difficult handling, expensive equipment). Method: In addition to deoxythymidine-triphosphate (dTTP), digoxigeninand biotin-labeled deoxyuridine-triphosphate is included in the RT reaction mixture. After up to 24 hours reaction, the newly synthesized DNA is trapped on streptavidin-coated ELISA plates, and subsequently detected immmunologically by binding peroxidase-labeled antidigoxigenin antibodies and colour development with ABTS" as substrate. Results: In comparison to a standard RT test (using "H-dTTP), our new assay was 1) sensitive and allows detection of as little as 1 mU of avian myeloblastosis virus-RT, 2) specific and reliable in testing 204 HIV culture supernatants (Pearson correlation coefficient with standard test = 0.943), and 3) easier to perform and cheaper. Conclusions: This assay makes RT testing possible for laboratories not licensed to handle radioactive material, and such testing is cheaper and much easier to perform. This new assay can be used easily for screening drugs for HIV-RT inhibitory effects. NOTES M.A.1085 EARLIE DEI'BC~I TION OF ANIM-IV-1 SEROONVESICN BY THIRD GTGATION AN-TII-1/2 ELISA Reesink Henk W*; Plaisier ADD*; Poel CL van der*; Pietersz m NI*; Lelie PN**. *Red Cross Blood Bank Amsterdam, ** Central Iaboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam, Ihe Netherlands Objective: To establish the sensitivity of 2 new anti-HIV-1/2 ELISA's (Wellcame, Abbott) in a HIV-1 serooonversion panel. Methods: Two anti-HIV-1/2 3rd generation ELISA's eaploy recombinant HIV-1 core, HIV-1 envelope and HIV-2 envelope protein as coating antigen and as conjugate. A panel of sequential serum sanples of 16 individuals at the time of seroconversion for anti-HIV-1 was investigated. Results: In 4 of the 16 individuals who converted to anti-HIV-l, both 3rd generation EELSA's detected one anti-HIV-1 follow-up sample earlier (2-11 days) than a 2nd generation sandwich ELISA (Abbott). Conclusion: Anti-HIV-1/2 3rd generation ELISA's have an improved sensitivity for the detection of early HIV-1 infection. NOTES s 1^ C/j 00 I~; M.A1086 HIV-1 EXPRESSION IN 25 INFBCTED PATIENTS: A COMPARISON OF EA POR, M.A.1 P24 EA IN PLASMA AND IN SITU THYRIDIZATION IN M6ONUCLEAR CELLS. Delord, B."; Ottmann, M.; Schrlve M-H.*; Errecart, M-J.; Ragnaud, J-M.; Seigneurin, J-M.; Fleury, Herve J.A.'. *CHRU de Bordeaux, Bordeaux, France; "" CHRU de Grenoble, Grenoble, France. Objective: To compare the data of in situ hybridization (ISH), RNA polymerase chain reaction (PGR) and p24 enzyme immuno-assay (EIA) for the detection of HIV-1 expression respectively in peripheral blood mononuclear cells (PBMC) and in plasma of infected patients at various CDC stages. Methods: PBMC of 25 patients (mostly of CDC stage II) were obtained from heparinized blood samples, then spread out onto glass-slides with a cytocentrifuge and in situ hybridized with a ["S] labeled single-stranded RNA probe specific for gag-pol of LAV-Bru HIV1 allowing the detection of genomic and/or messenger RNA. The corresponding plasma samples were used for the determination of p24 core antigen by EIA and detection of HIV-1 genomic RNA by DNase/RT/PCR using specific primers in the LTR, gag and env regions. Results: Whereas p24 antigen within the plasma was detected in only 6 out of 25 patients, both ISH and PCR enabled the detection of viral RNA. in more than 50 % of the series; the frequencies of samples with cells containing transcriptionally active provirus and of plasmas with genomic RNA appeared to be increasing with the disease status. NON-ISOTOPIC PCR METHODS FOR THE DETECTION OF HIV-1 IN UGANDAN M.A.1087 OTHERS AND INFANTS Jackson., Brooks; Ndugwa.C"; Mairo,F**; Kataaha P"; Guay,L; Ball,T; Dragon,B; Goldfarb,J; Xu,S; Kabengera,S; Vjecha,M; Robbins F*; Ellner,J; Olness,K*. Case Western Reserve U, Cleveland, OH, USA;**Makerere U, Kampala, Uganda;***Roche Diagnostic Systems, NJ, USA. Objective: To determine the feasibility, sensitivity, and specificity of non-isotopic polymerase chain reaction (PCR) methods for the detection of HIV-1 in Ugandan mothers and infants. Methods: PCR testing was performed on crude cell lysates of 42 Ugandan HIV-1 Western blot positive and 16 seronegative mothers and 55 of their infants (aged 0.5-15 mos). Amplification of HIV-1 sequences was performed in duplicate using a biotinylated primer pair to the gag region (SK 462/431) and nested primer pairs (JA 17-20) to the pol region. Gag sequences were detected using the SK 102 probe by enzyme immunoassay (Roche Diagnostics). Pol sequences were detected on ethidium bromide stained agarose gels. Infants and mothers were considered to be HIV-1 infected if both gag and pol sequences were detected. Result= HIV AB GAG PRIMER POL PRIMER BOTH PRIMERS MOTHERS POS 40/42 (95%) 41/42 (98%) 40/42 (95%) NEG 0/16 (0%) 0/16 (0%) 0/16 (0%) INFANTS POS 10/31 (32%) 11/31 (35%) 10/31 (32%) NEG 0/24 (0%) 0/24 (0%) 0/24 (0%) Discussion and Conclusions: Non-isotopic PCR methods are sensitive, specific, and potentially useful in the early diagnosis of HIV-1 infection in developing countries.

Page  114 I' I' 4^~ M.A.1088 HIV DETECTION IN BLOOD MONOCYTES FROM HIV SERONEGATIVE SUBJECTS. Di Lorenzo Angela*, Bergamini A.**, Perno C.F.**, Rosci M.A.***, Miceli M.*, Mannella E.*, Rocchi G.** *National Blood Trasfusion Center of Italian Red Cross, Rome, **II University of Rome and ***Hospital "L. Spallanzani, Rome, Italy. Objective: To determine wheter viral isolation from monocytes (M/M) of seronegative high risk subjects could be a valuable method for early detection of HIV infection. Method: Blood derived M/M from 18 (6 males and 12 females) eterosexual partners of seropositives subjects previously involved in long term unprotected sexual intercourses were obtained by adherence to plastic and then cultured with or without cytokines enhancer of HIV replication (macrophage colony stimulating factor 1.000 U/ml, or TNF 1 ug/ml). HIV infection was assessed by a p24 antigen ELISA assay. Results: All subjects were seronegative by ELISA and W.B., and p24 antigenemia aas also negative. Viral production has been obtained from cytokine-stimulated A/M of 2 out of 18 seronegative subjects (11%). Such production was low (50-60pg/ml) out confirmed in duplicate and with neutralitation test. No viral production was obtained from unstimulated M/M. Both cases were female. One of them seroconverted (p24 antibodies by western blot analysis) 4 months after the first viral isolation. Discussion and Conclusions: There results, showing silent HIV infection in M/M Ln a similar range than that archieved by others with different methods, suggest that M/M culture could be of importance in evaluating HIV infection in selected -igh risk subjects negative to other tests. NOTES M.A.1089 PROGNOIC VALUE OF PLASMA HIV ISOLATION IN MONOCYTES/MACROPHAGES CELLS Mannella E.~, Armignacco 0.~, Mercurio G~, Marinelli L.~, Di Lorenzo A.~, Bergamini A.*. Pezzella M.**. ~N.B.T.C. Italian Red Cross, Rome, OL. Spallanzani Hospital, Rome, Italy, *II University of Rome, Italy. **Universit& "La Sapienza" Rome Italy. Objective: To assess the prognostic value of HIV isolation from normal monocytes/macrophages (M/M) exposed to plasma of ARC/AIDS subjects. Methods: Blood derived M/M from a normal seronegative donor were obtained by five days adherence and then were cultured with plasma from 36 HIV positive patients (17 asymptomatic, 3 ARC, 16 AIDS). Results: Viral production was achieved in M/M exposed to the plasma of 7/19 ARC AIDS patients but from none of M/M cultures exposed to the plasma of asymptomatic subjects. These 7 positive patients had CD4+ lymphocytes count and antigenemia similar to that of the other 12 ARC/AIDS negative patients. Beta 2 microglobulinemia leveles were between 5-7 ug/ml and the general clinical conditions were also similar. A 12 months follow up pointed that all 12 ARC/AIDS negative patients were alive as well as all 17 asymptomatic subjects, While 5/7 ARC/AIDS positive patients (71%) died (the ARC patient after 11 months, and four AIDS patients after an-average of 4 and 1/2 months). Discussion: These data suggest that the infection of normal M/M with HIV from plasma could be a prognostic factor for ARC/AIDS patients, while no prognostic meaning was shown for similar infection of normal lymphocytes. NOTES M.A.1090 HIV DETECTED WITHIN ALVEOLAR MACROPHAGES IN VIVO BY IMMUNOCYTOCHEMISTRY FROM INDUCED SPUTUM SAMPLES Evans, Mark; Watts, S; Wansbrough-Jones, M. St. George's Hospital Medical School, London, England. Objective: To detect HIV within alveolar macrophages in vivo & to correlate with stage of disease. Methods: Induced sputum samples were obtained from controls (no known risk factors for HIV), asymptomatic HIV+ve individuals & AIDS patients. After mucolytic treatment, alveolar cells were washed, fixed in acetone & labelled using a monoclonal antibody to HIV p24 (efficacy previously confirmed against HIV+ve Jurkatt cells). Macrophages were identified morphologically & the proportion +ve recorded. (At least 100 macrophages counted). Results: No HIV p24 antigen was detected in the control group (n=5). In the HIV+ve asymptomatic group (n=10), one subject had 2% alveolar macrophages +ve for HIV p24, two subjects 1% +ve, one subject <1% & six subjects nil. In the AIDS group (n=6), one subject had 2% +ve & the rest nil. Overall 1/4 of the HIV/AIDS patients had detectable HIV p24 Ag in alveolar macrophages. Conclusion: This study confirms HIV infection within alveolar macrophages in vivo but at a low incidence (<2%). No correlation was seen with stage of disease. Others have shown higher proportions of alveolar macrophages +ve for HIV & the possible reasons for this discrepancy will be discussed. M.A.1091 NASBA'M NUCLEIC ACID AMPLFICATION USED FOR THE DETECTION OF HIV-1 IN CLINICAL SAMPLES. Vanemen. Bob van Stnjp, Dianne; Schukkink, Rianne; Beyer, Ria; Huisman, Han; Koppelman, Marco; Malek, Larry"; and Lens, Peter. Organon Teknika, Boxtel, The Netherlands, Central Laboratory of the Bloodbank, Amsterdam, The Netherlands, Cangene coorporation, Mississauga, Canada. The nucleic acid sequence based amplification (NASBA") technology was developed and tested using a HIV-1 model system and clinical samples. NASBAn is an one step primer dependent isothermal amplification technique for both RNA and DNA templates. Amplification factors range from 10e to 109 depending on primer sets used (gfg and pI region). Specificity and accuracy of NASBAm were confirmed by sequence analysis of amplified fragments. In a model system we were able to detect as little as 5-50 in vitro HIV-1 infected cells in a background of 5x104 non-infected cells, 100 pl blood or 100 il plasma. The capability of NASBA'm to directly amplify RNA sequences makes it possible to test for the presence of HIV-1 virus particles in plasma or blood. Up to 10- dilutions, in 100ui1 plasma or blood, of a virus stock solution (2500 TCID,/ml) could be detected. Nucleic acid isolated from 0.5-1 ml plasma samples of HIV-1 infected individuals was used for the amplification of HIV-1 sequences with primer sets specific for the pol and gag regions. All patients (including asymptomatic) scored positive after NASBA analysis, albeit that one primer set (pg0) was less sensitive than another primer set (gag). Variations in the nucleic acid sequences of different isolates of retroviruses suggest that diagnosis should be based on results obtained with several primer sets. C? a sj c^ t~ Ij

Page  115 M.A.1092 APPLICATION OF PCR TO DISCRIMINATE BETWEEN HIV-1 AND HIV-2 INFECTION. Gorrino MT*, Campelo C*, Sarria L*, Fernandez de Aranguiz A*, Lardelli P*, Suarez MD**, Cisterna Ram6n4***. *Microbiology and Immunology Department. University of Basque Country. Bilbao, Spain. **Microbiology Department. Civil Hospital of Bilbao. Bilbao, Spain. PCR has been proposed as a method to identify infected patients or blood donors before seroconversion is detected or to assess the results when the question of infection in an individual is ambiguous by serological assays. An additional potential application of the method is the discrimination between HIV-1 and HIV-2 infection. In certain individuals, antibody assays may be unable to discriminate which virus type is responsible for a confirmed positive test; however, type-specific nucleic acid sequences can be identified in PCR assays. As a first approach, we have screened in a blinded fashion for HIV-1 and HIV-2 infection a group of 150 patients at different risks for AIDS. SK100 and SK104 oligonucleotides were used as primers to amplify a well conserved region of both HIV-l and HIV-2. SK19 and SK109 bibtin labelled oligonucleotiaes were used for HIV-1 and HIV-2 detection respectively. We found 14 cases of HIV-1 and HIV-2 mixed infection among 56 drug addicts; 7 among 15 sexual partners of persons who were mainly seropositive to HIV-1; 3 out of 9 children born of seropositive mothers and 1 of 3 sanitary workers accidentally infected. Two cases of post-transfussional mixed infection were also detected by this method. The remaining 65 patients screened by PCR have not been classified in a concrete risk population yet. These data demonstrate that HIV-2 is present in our population although further larger studies in the different AIDS groups should be done to evaluate the extent of HIV-2 infection in our population. NOTES M.A.1093 DETECTION OF HIV-1 DNA BY PCR INCORPORATION OF DIGOXIGENIN-dUTP AND HYBRIDIZATION TO IMMOBILIZED PROBES Costa Taveira, N4 Monzh PerelraJ.; Odette FerreiraM. University of UsnFacutyfharmacy, Depo biology,bon, Portugal. OBJECTIVE- To develop a simple, rapid and non-isotopic PCR-based method for the detection of HIV-1 DNA sequences present in the total DNA of PBMCs of HIV-1 infected individuals. METHODS- Polymerase chain reaction was performed as described byLaure, F. et. al. (The Lancet,Sept.3,1988,pg.538-540), using amixture ofdNTPs containingboth dig-dUTP and dTTP.Primers to HIV-1 gag and polconservative regions were used. An aliquot of the products was heat denatured and hybridized to nylon immobilized DNA probes specific to the amplified fragments. Detection was performed by an anti-digoxigenin alkaline phosphatase conjugate and standart phosphatase assay. RESULTS- With this procedure we could detect HIV-1 DNA sequences in all the PBMCs DNA samples from 15 HIV-1 seropositive subjects while the samples from 15 HIV-1 scronegative subjects gave consistent negative results. A limit of 10 copies of HIV-1 DNA could be detected using this method. CONCLUSION- We have developed a sensitive and specific non-isotopic PCR-based assay for the detection of HIV-1 DNA in the total DNA extracted from the PBMCs of HIV-1 infected individuals. This method is safe, simple and of rapid executon:it eliminates gel electrophorese analisys of the reaction products; there is no need for Southern transfer of each amplified DNA prior to hybridization; hybridization can be performed immediately after amplification in a simple tube assay and it involves non-isotopic detection procedure; in a single amplification and hybridization assay it is possible to detect several HIV1 DNA regions. This method can be used in any routine diagnostic facilities for the detection of potentially all amplifiable bacterial and viral pathogens. NOTES O 0() 8\ ^2 M.A.1094 DETECTION OF FREE AND ANTIBODY-CONPLEXED ANTIGEN FOR THE DIAGNOSIS OF PERINATAL HIV INFECTION Palomba, Elvia; Gay, Vincenzo*; Barberis, Ezio*; Ciuti, Elana; Peruginl, Laura*; Nicola, Paolo; Tovo, Pier-Angelo Department of Pediatrics, University of Turin; sOspedale Infantile Regina Margherlta, Turin, Italy Objective: Acid treatment (pH 2.5-3) of serum, dissociating immune-complex, Improves p24 antigen detection in HIV-infected adults. This test may prove useful for the early diagnosis of perinatal infection, when high titers of maternal antibody are still present in the Infant's blood. Methods: p24 antigen was assessed by standard ELISA (Abbott), before and after acid treatment, in serum obtained from 16 HIV+ women, 15 P-2 children and in 14 infants whose infection status was indeterminate (P-0), but was subsequently clarified during the follow-up period (4 P-2 and 10 uninfected children). Results: Free-p24 antigen was detected in 1/16 women, 8/16 P-2 children and In 2/4 P-0 infants who subsequently developed symptoms (P-2). Among subjects without serum free-antigen, positive test after low pH treatment was found in 10/15 (66X) women, in 3/7 (42%) P-2 children and in 2/2 P-0 children who subsequently became symptomatic (P-2), whereas it was negative in all 10 P-O children who later resulted uninfected. Conclusions: HIV antibody passively acquired from the mother or endogenously produced may form immune-complex with p24 antigen in infected infants. This study indicates that acid pretreatment of serum, evidencing complexed antigen, may allow an early diagnosis of perinatal HIV Infection. M.A.1095 INTERNATIONAL COLLABORATIVE STANDARDIZATION OF PCR ASSAY FOR VM.A. DvNA BrCchot. Christian1,2; Chaput Agnes1; Montagnier, Luc1; Osmanov, saladin; Milman, Gregory and the international cooperative group. 1. Institut Pasteur, Paris, France; 2. Necker,Enfants Malades, Paris, France; 3. WHO, Geneva, Switzerland; 4. Division of AIDS, NIAID, Bethesda, MD, USA. This study was made with the sponsorship of EEC (concerted action on HIV quantitation) and WHO global AIDS programm. Obiective: Thirty international laboratories participated in the PCR evaluation of a panel of samples containing various amounts of HIV-1 DNA. Methods: DNA was prepared from the 8E5-LAV cell line containing one stabile HIV-1 DNA copy per cell. Specimens in the panel contained defined copy numbers of the 8E5-LAV DNA included in 1 pg of carrier DNA prepared from PBMC of HIV-negative blood donors. The Coded purified DNA samples were provided in duplicate to the participating laboratories and evaluated for the presence of HIV-1 by each laboratory's current PCR assay. Results and discussion: The results enable the comparison of differences in PCR methodology used by different laboratories for HIV detection. Some of these differences include choice of primers, probes, PCR cycle length, temperature, and other variables. The results emphazize the value of a panel of quality control standards and unknowns. Follow-up studies will include the evaluation of panels of HIV RNA samples.

Page  116 M.A.1096 CONSTRUCTION AND EVALUATION OF A NESTED PRIMER PCR FOR THE DETECTION OF 'HIV-2 IN CLINICAL SAMPLES Grankvist Olov*, Bredberg-RAddn U**, Gustavsson A*, Albert J**, Albino P***, Biberfeld G** and Wadell G*. * Dept of Virology, University of Umea, Sweden, ** Natl Bact Lab, Stockholm, Sweden *** Hospital Simao Mendes, Bissau, Guinea-Bissau. Objective To design and evaluate a nested primer PCR for the detection of HIV-2 DNA in clinical samples. Methods A two-step-PCR with nested primer pairs from 4 regions was compared with a single-step-PCR with primer pairs from 3 regions, with or without hybridization on 17 HIV-2 isolates. 37 samples of blood mononuclear cells (PBMC) from HIV-2 seropositive West African individuals ( 34 asymptomatic and 3 with AIDS related symptoms) were tested with the nested PCR. Results 17/17 isolates were reactive in the single-step-PCR with at least 2 of 3 primers after hybridization. Only 14/17 could be identified without hybridization and showed reactivity with fewer primers. With the nested PCR, 17/17 isolates were reactive and only one of them needed hybridization for detection. 31/37 (84%) PBMC samples were reactive with nested primers without hybridization. Discussion and Conclusions The sensitivity of the nested PCR without hybridization is the same as with the single-step-PCR including hybridization. The negative PCR reactivity in a few seropositive individuals may be due to a very low number of HIV-2 infected cells in these cases. NOTES M.A.1097 PROSPECTIVE DETECTION OF HIV-1 PROVIRAL DNA IN PERIPHERAL BLOOD MONONUCLEAR CELLS OF CHILDREN BORN TO SEROPOSITIVE MOTHERS. S.. 1t. Oari@: De Mara A; Ferazmln A LOy A; Gotta C, TerOraM A I, ClIni Malle Intlm, Un"lWere O Genova, &.G0anlk' institute, * Serv. Noonatologla Cllnica Ostetrica - University of Genm, Italy. (W.l: To determne whether and how an early treatment decison could be based on Information obt lined S I dri r e in reactbon (PCR) npllcatoon of HIV.1 DNA in peripheral blood mononuclar cells (P 3MC) fr~omchlpdren born to HIV serepoallte mothers. ~: Chldrn born to HIV.1 seopose mothers were follUowed up from birth every 2-3 months. Known ed P2patlents we onsered u a positive oonrol roup. At each control chldren undewent a lull c Inical examination. HN1 lnfection was monilored by HIV-1 erology (Western Bot-ELISA p24 Ao ELISA) and virus olonetIhe same tim ontso HIV.1 DNA deteton. T e presence f HIV-1 pW DNA In their F'BMC 1 wa ssesed by P, mpying H.-1 gag and env-peclob primer set and "P4eled probes. lts Twenty ~odre., followed-p for a mean of 26 rmnlh (r ane 13-28). were Included n this study. All w _ HIV-1 eroposlt, e at brth. No HIV-1 NA was detected In the PBMC of 17 children on the tirsi detennnia and al had concordant ne HIV-p24 antigen end virs culture. HIV-I DNA was never detecied in their PBMC during te follow-up determn n Fifteen of these patiuents sub uetly 8eroreverted (PX) and 2 patents are sI aged 8 months and are seroposld (PO). WHV-1 provral DNA sitwere Intlially delucied by PCR in of the 20 study patients. One of them had at least a concomitant positve virus isolation or IVAP test and %bseuently developed HIV-1 related symptom (P2). In the other two patients, no HIV-1 DNA in the PBMC at ater e-points were detected. These two patlents eroreverted and are now cliically well and IV.-1 seronegative. SBased upon this mited expeence on a small group of patients, failure to detect HIV-1 proviral DNA at i exarina in PO children I useful to rule out HIV nfection and the need for therapeutic intervention. In HIV-1 DNA poitive patlents, the preence of WV-1 DNA should be confirmed n at least 2 diflirent determinatons to avoid unnecessary and otentlally toxic treatment of otherwise unected children. NOTES M.A.1098 Simple And Efficient Assay For Gpl60 And Gp120 Of Various Strains Of HIV-1 And HIV-2. N. Mahmood,* A.J. Hay" and C. Hentschel*. *MRC Collaborative Centre, Mill Hill, London NW7 1AD *Natlonal Institute For Medical Research Mill Hill, London NW7 1AA. Assay of the glycoproteins of HIV is useful in assessing HIV infection and in evaluating the antiviral activities of compounds in HIV infected cell cultures. Commercially available ELIZA kits have certain disadvantages in regard to cost and the restricted specificities of antibodies used. We have developed a novel assay using a lectin uniquely specific for high mannose oligosaccharides of Gp120 of different strains of HIV and SIV. The lectin GNA (isolated from the bulbs of Galanthus nivalis), immobilized on microtiter plate wells binds with high specificity envelope glycoproteins in crude cell extracts. These are then detected using human anti-sera and an anti-human IgG antibody conjugated to horse radish peroxidase. The assay is simple and can be adapted for use with specific and cross reaction antisera, It is also eminently suitable for the detection of anti-envelope antibodies in human sera. MA.109 LIDENTIFICATION OF HIV-1 IN PARAFFIN WAX EMBEDDED TISSUES M.A.1099 Pudney, Jeffrey and Anderson, D. Harvard Medical School, Boston, MA USA Oblective - Investigate effects of different fixatives on the ability of commercially available monoclonal antibodies (Mab's) to detect HIV-1 in wax embedded tissues. Methods - Jurkat cells infected with HIV-lb fixed in either unbuffered 10% formaldehyde (F), 2.5% gluteraldehyde (G), or a mixture of 4% paraformaldehyde and 1% gluteraldehyde (P+G) both in O.1M phosphate buffer pH 7.3. Lymph nodes from autopsy of HIV-1 seropositve men fixed in 10%F. Mabs used were anti-gpl20, gp41, p24 (DuPont); anti-gpl20, gp4l, p24A, p24B, pl8A, pl8B, p55, reverse transcriptase (RT) (Olympus) detected by an alkaline phosphatase/anti-alkaline phosphatase kit (Dako). Pellets of HIV+ Jurkat cells were fixed (18 hrs), embedded in wax, and sections incubated with each Mab (1:20) at 4'C for 18 hrs. Sections of lymph nodes were incubated with a cocktail of antibodies (DuPont gpl20, gp41, p24) for 18 hrs at 4'C. Controls were 1) omission of primary Mab and 2) incubation with an irrelevant isotypic Mab. Results - F and G Fixation - good staining for all Mab's, except gpl20 (DuPont) and R.T. (Olympus). P+G Fixation - good staining for gpl20, gp41, p24 (DuPont) and gpl20, p24B, pl8B (Olympus); poor to no staining for gp4l, p24A, pl8A, p55, R.T. (Olympus). Lvmoh Nodes - Detected many HIV-1+ cells with Mab cocktail. Controls - no background or non-specific staining. Conclusions. Many HIV-1 antigens withstand aldehyde fixation and embedding in wax, allowing their detection by different commercially available Mab's. Significance: 1) tissue can be optimally preserved and prepared for identification of HIV-1, compared to frozen sections where antigenicity is often maintained at the expense of morphology; 2) pathological tissue (usually fixed in F and embedded in wax) can be analyzed retrospectively for the presence of HIV-1. a C> C> 5~

Page  117 M.A.1100 HIv-1 ISOLATION F CM WHLE B0LID: SUANDAIZATION AND APPLICATICOS. Fico Cecilia, Angarano G., Fiore J.R., Di Stefano M., Monno L., Fracasso C. Pastore G., Clinic of Infectious Diseases, University of Bari, Italy. OBJECTIVE: To evaluate HIV-1 isolation from small amounts of whole blood(W.B.)in order to develop a technique rapid, effective and easily achievable in the clinical practice. METHODS: We have studied 100 HIV-1 infected subjects (51 asymptcaatics, 49 AIDS).HIV-1 isolation was attempted by using standard PMCs cultures and by co-cultivation of different amounts (500,50,5 ul) of heparin or EMA containing WB with 4 millions of normal PMCs or 1 million of Molt-3 cells. Viral replication was evaluated on culture supernatants by HIV-1 p24 Ag detection, every 4 days over a period of 1 month. RESULTS: 500 ul of WB were effective in detecting HIV-1 replication in all stages of cu sease. Isolation rate in asymptomatic subjects was higher capared with standard PMCs cultures (90% vs 47%). The WB minimal amount which allowed the detection of viral repli cation variated among different subjects (fran 500 to 5 ul) and arised during Zidovudi ne treatment. Viral growth in cell cultures from WB was not affected by the presence of serumn p24 Ag, the use of Molt-3 cells or the type of anticoagulant used. CONCWUSIONS: HIV-1 isolation from small amounts of whole blood constitute a sensible and effective method for HIV-1 detection. M.A.1101 IN VITRO LYMPHOCYTE ACTIVATION ENWHACES THE DETECTION OF HIV-1 SEQUENCES BY PCR IN INFECTED-SERONEGATIVE INDIVIDUALS Chen, Hong; Carbonari. Maurizio: Sbarigia, Daniela; Cherchi, Michela; Pesce, Annamaria; Aiuti, Fernando; Fiorilli, Massimo. University of Rome 'La Sapienza', Rome, Italy. Objectyl!: to determine whether: 1) in vitro activation of lymphocytes by polyclonal stimulation enhances the sensitivity of polymerase chain reaction (PCR) for the detection of HIV-1 sequences in infected-seronegative individuals; 2) in vitro production of HIV-1 specific antibodies occurs in infected-seronegatives. Methods: in a cohort of seronegative subjects at risk for HIV-1 infection, proviral sequences were searched by PCR on lymphocytes, either freshly isolated or cultured for ten days in the presence or absence of pokeweed mitogen (PWU). HIV-1-specific antibodies were also determined in culture supernatants by ELISA and western blotting. Results: in two of 36 subjects examined, PCR (done using SK38/39 and SK68/69 primer pairs) was negative with fresh lymphocytes, but positive with cultured cells. In one case both unstimulated and P1iU-stimulated, and in the other only PUM-stimulated cells yelded a positive PCR. Contaminations were ruled out by testing parallel cultures from six normal subjects. No HIV-1-specific antibodies could be detected in culture supernatants of cells from normal controls and infected-seronegatives. Conclusions; 1) in vitro precultivation of lymphocytes, especially in the presence of PUM, can enhance the detection of HIV-1 sequences in infected seronegatives; 2) our data do not support the claim (Sell K. et al., 6th Int. Conf. AIDS, 1990, Abstr. Book 1, p. 137) that PUM-induced in vitro biosynthesis of HIV-1 antibodies permits the detection of infected-seronegatives. CI ~I 00 i~r NOTES NOTES M.A.1102 DETECTION, QUANTITATION AND SEQUENCING OF HIV-1 VIRUS FRO THE PLASMA OF SEROPOSITIVE INDIVIDUALS AND FRO FACTOR VIII CONCENTRATES Zhang, Lin Qi'; Simmonds, Peter'; Ludlam, Christopher A.**; Leigh Brown, Andrew J.* *Division of Biological Science, University of Edinburgh "Department of Haematology, Royal Infirmary, Edinburgh, U.K. Objective: To develop a sensitive RNA PCR method to detect, quantify and sequence HIVviral RNA directly from the plasma or serum of seropositive individuals and from factor VIII concentrate. Method: HIV-1 viral RNA was extracted and DNAse treated prior to reverse transcription with HIV-specific anti-sense primers. The cDNA was amplified in a double PCR, using nested primers to enhance the sensitivity and specificity of the assay. The amount of RNA in positive samples was quantified by limiting dilution prior to amplification. Results: Ten out of twelve haemophiliacs and two out of eight batches of unheat-treated commericially available factor VIII concentrate was found to contain detectable levels of HIV-1 RNA. The amount of cell-free circulating virus ranged from less than 1 x 10' to 3 x 10' copies of RNA per ml of plasma, with a logarithmic mean value of 1.2 x 10' copies for CDC II stage patients, and 5.5 x 10' copies for CDC IV patients. No correlation was found between p24 antigen concentration, CD4' counts and the amount of HIV-1 RNA in the plasma. The abundance of HIV sequence in factor VIII was very low (<3 copies/ml,. reconstituted volume). Conclusioas: HIV RNA can be detected in the plasma of many patients including some whe are p24 antigen negative. This indicates that viral replication continues throughout the course of infection. There is no viral "latent" period corresponding to clinical latency. MLA 10 H y DI^am OF HIV IPrC~TN IN fNilaCY: A iWIl-1ES EVAIAVTION. M.A.1103 Savita* AhransE.**, Paul M.*, TetaliS.*, WangX.*, Pasieca,R.*, Napolitano,B.*. *North Shore University Hospital-Cornell University Medical College, Manhasset, NY and **Harlem Hospital-Columbia University, NY,NY. USA Objective: To evaluate a battery of tests for diagnosis of HIV infection in infants (CDC class P-O), born to HIV seropositive women. Methods: Methods included HIV culture of peripheral blood mononuclear cells (PBM) and of plasma; direct detection of serum p24 antigen by ELISA; polymerase chain reaction (PM) with 8K38/39 (gag) and SK68/69 (envelope) primer pairs; and in vitro antibody production (IVAP) by culture of PBM. 84 infants were analyzed sequentially at 2-3 month intervals until age 24 months. Thus far, outcome has become known in 41 cases: 21 are now P-2 and 20 have seroreverted to P-3 status. Results: At all ages, viral culture, PR, and p24 antigen tests were 100% specific; overall sensitivities were 78%, 80% and 82%, respectively. Under 3 months, sensitivity was relatively low, but sample size is too small for definitive evaluation. After 3 months, sensitivity increased to 90-100% for PCR and viral culture. Some patients who initially tested negative became positive on subsequent testing. The IVAP test was sensitive and specific only after 6 months of age and the incidence of false positives was high in early infancy. After age 6 months, all assays were equivalent in sensitivity and specificity. The predictive power of the combination of tests exceeded that of any single test. Conclusions: Diagnosis of HIV in infants less than 3 months of age is problematic. HIV diagnosis in early infancy cannot be established based on one single test; a battery of tests may be necessary, especially in infants under 3 months of age.

Page  118 M.A.1104 VIRUS ISOLATION, POLYMERASE CHAIN REACTION AND IN VITRO ANTIBODY PRODUCTION FOR DIAGNOSIS OF PEDIATRIC HIV1 INFECTION Garbarg-Chenon A., Vendrell J.P., Silvain A., Courpotin C., Serre A., Nicolas J.C., Grimprel E., Bricout F. (CHU St Antoine, Paris; INSERM U249, Montpellier-FRANCE) The diagnosis of HIV1 infection in infants bom to HIV seropositive mothers is critical. In this study we have compared the results obtained by virus isolation in cell culture (VC), gene amplification (PCR) and in vitro anti-HIV1 antibody production (IVAP) by peripheral blood mononuclear cells (PBMC). Methods: 69 infants or children born to HIV1 seropositive mothers were followed for 1 to 4 years. Virus was detected in supernatants of PBMC (cultured at least 36 days) by reverse transcriptase activity and p24 antigen. PCR used 2 primers for POL and GAG regions of HIV1 with 40 amplification cycles. IVAP was detected by Westem blot assay (Diagnostics Pasteur) in supernatants of 5.106 PBMC cultured for 6 days. Results: The 3 methods (VC, PCR and IVAP) gave concordant results for 21 negative and 18 positive cases. These results were in agreement with clinical status of the infants. Discrepancy was observed between VC and PCR (6 vC+/PCR-;10 VC-/PCR+), VC and IVAP (4 VC+ /IVAP-;8 VC-/IVAP+), PCR and IVAP (6 PCR+ /IVAP-;7 PCR-/IVAP+). These results are discussed according to the serological and clinical status of the infants. NOTES M.A.1105 A NEW PLASMID CONSTRUCT AS INTERNAL STANDARD FOR HIV PCR A. Telenti, P. Imboden. Institute for Medical Microbiology; University of Berne, Switzerland. The use of internal standards in HIV PCR would fulfill two purposes: 1) Control of the efficiency of amplification for each individual sample, 2) facilitation of quantitative analysis of the results. We have constructed a new plasmid using "gene splicing by overlap extension" PCR techniques. This construct incorporates the HIV gag sequence that is amplified by primers SK38 and SK39 to a pUC-type plasmid. The gag fragment was splitted by the insertion of a heterologous 500 bp DNA fragment. PCR amplification of this plasmid using primers SK38 and SK39 will generate a 611 bp product instead of the 115 bp amplimer obtained with native HIV target. This long product does also hybridize to the SK19 probe. The plasmid (pHIVg ) could be added at a fixed copy number to HIV positive samples to simultaneously generate two distinct amplimers that would be separated by electrophoresis and hybridized after Southern blot. Hybridization signals could then be aOantit.ated and cnmnarRd. NOTES M.A.1106 EVALUATION OF FOUR COMMERCIAL HIV-1 ANTIGEN ASSAYS. Maniez. Michble*; Ferroni, A.**; Dupressoir, M.V.*; Rouzioux, C.**; and the Retrovirus Group of the French National Society of the Blood Transfusion *Centre de Transfusion sanguine, Lille, France; ** H6pital Necker-Enfants Malades, Paris, FRANCE. The Objectives of this study are the evaluation of sensitivity, specificity and reproducibility of four HIV-1 antigens assays. Material and Methods: The four assays were produced by Abbott Laboratories (A), Coulter Electronics Inc. (C), Diagnostics Pasteur (DP) and Du Pont de Nemours Co. (DDN). Eight laboratories have been involved in this study. At least 200 seronegative blood donors (different in each laboratory) have been tested for each kit. A panel of 25 specimens have been established. It included a standard calibrator at different dilutions (composed of native viral proteins and produced by DP), HIV-1 patients sera (positive or negative for antigenemia), and "tricky" sera (FR, hyperlipemic, CMV-IgM). This panel of coded samples have been studied in four different labs for each Ag-kit. Results: We found a good specificity for all assays with less than 1 % of false positive results. The detectability of HIV antigen evaluated with the standard curve was respectively 42.7 pg/ml for A, 16.4 pg/ml for C, 29.7 pg/ml for DP and 31.3 pg/ml for DDN. This good rate of sensitivity was confinned by the analysis of results obtained with HIV-1 positive sera from the panel. The mean of coefficients of variation have been evaluated for the optical densities (OD), the ratios (OD/cut-off) and the titers in pg/ml. They showed a satisfying reproducibility for the 4 assays. Conclusion: This interlaboratory study permitted to show the good performancies of the 4 HIV-1 antigens EIAS and to emphasize the necessity of a standardization between manufacturers for HIV-1 antigen quantitation. M.A.1107 QUANTITATIVE PLASMA VIRAEMIA: COMPARISON OF TWO TECHNIQUES. Groupe "Virimie Quantitative": A.C. 11 - ANRS: Rouzioux, Christine*; Puel, J.**; Agut, H.***; Brun-Vezinet, F.; Ferchal, F.~~; Fleury, H.~~~; Tamalet, C.~~~~ *H6pital Necker, Paris; **CHU Purpan, Toulouse; ***H6pital Pitie-Salpetriere, Paris; ~H6pital ClaudeBernard, Paris; ~~H6pital Saint-Louis, Paris; ~~~CHU Pellegrin, Bordeaux; ~~~ZCHU La Timone, Marseille, FRANCE. The Objectives of this work were the evaluation of two techniques published in N.E.J.M. (Dec 14 1989) and having reported very different results in plasma viraemia quantitation. Patients and methods: Seven labs of the working group were involved in the study. Plasmas and cells from HIV-1 patients at different CD4 levels were selected. The two techniques were strickly followed as published, and have been done for 54 fresh specimens in parallel. The Results obtained by both techniques are very similar and plasmas viraemia titers are not significantly different. They are strongly related to the CD4 cells number: positive results were obtained for 83 % of patients with CD4 < 200, 61 % for those with 200 < CD4 < 500 and 0 % for those with CD4 > 500. Cellular viral cultures are positive for 96 % of the patients. Conclusion: At the moment, plasma viraemia seems the most interesting technique to evaluate antiviral therapies, so there is a need for an available consensus technique. R C> C sj

Page  119 M.A.1108 QUANTITATIVE DETECTION OF HIV IN PERIPHERAL BLOOD MONONUCLEAR CELLS FROM INFECTED PATIENTS BY POLYMERASE CHAIN REACTION. Escaich Sonia*, Ritter J**, Rougier P***, Sepetjan M**, Trepo C*. * INSERM U271; **Laboratoire d'Hygitne, Facultd Rockefeller; '***Htel Dieu, Lyon, France. Objective: To determine the number of proviral HIV-containing PBMC as an index of progression of HIV infection.Methods: HIV proviral sequences were amplified by PCR using gag primers, and P32 labelled dCTP incorporation. For the sake of calibration, known copy numbers of HIV-containing pBT1 plasmid were amplified. Using the standard curve, the number of HIV sequences in PBMC from 16 patients at different stages of infection were computed.Results: The number of HIV copies detected in 106 PBMC ranged from 50 to 10,500: assuming one copy per cell this corresponded to one proviral HIV-containing cell per 95 to 20,000 total cells. In asymptomatic patients the calculated mean number of PBMC per HIV copy was significantly higher than in ARC and AIDS. High HIV copy number in PBMC was associated with low CD4 cell number. In patients who progressed from asymptomatic stage to AIDS, the number of PBMC per HIV DNA copy decreased, indicating an increase of HIV replication. Conclusions: The number of infected PBMC increases during clinical progression. A high percentage of HIV infected PBMC is observed in symptomatic subjects with low CD4 cell counts. NOTES M.A.1109 DETECTION OF HIV PLASMA VIREMIA BY CULTURE AND POLYMERASE CHAIN REACTION (PCR) AT DIFFERENT STAGES OF HIV INFECTION. Escaich S*, Ritter J**, Rougier P***, Lepot D*, Lamelin JP*, Sepetjan M**, Tro Christian*. *INSERM unite 271,** Laboratoire d'hygitne Universit6 Rockefeller,*** Hotel-Dieu Lyon France. Objective: To study HIV replication in infected individuals by assessing cell-free virus production in blood. Methods: Two hundred and twenty eight blood samples from 154 HIV seropositive subjects were investigated for the presence of HIV in plasma, using normal PBMC as the target for replication. PCR was used in parallel in 20 cases to detect HIV RNA in plasma. Results: HIV was recovered from 73.2% of PBMC but from only 19.5% of the plasma samples. Plasma viremia was significantly associated with clinical manifestations of HIV infection, indicating that HIV replication increased as disease progressed. This was confirmed by a correlation between CD4 cell counts and plasma viremia. Plasma viremia was associated with low anti p24 antibody titers. When HIV RNA detection by PCR was compared with plasma viremia, IIIV RNA was detected in plasma in all symptomatic cases and in 53.8% (7/13) of the asymptomatic ones. Conclusions: These results indicate that cell-free virus production is associated with the clinical stage of HIV infection and could serve as a marker for disease progression. Detection of HIV RNA by PCR appears as the most sensitive method to detect viremia. NOTES U.LALITY CONTROL OF PCR ASSAY FOR DETECTION OF IIV DNA IN M.A.1110 FRENCH LABORATORIES French Collaborative Study Group for PCR in HIV infection. Objective: 1) to determine the specificity, the sensitivity and the reproductibility of polymerase chain reaction (PCR) assay to detect HIV DNA in different french laboratories. 2) to reach a consensus on conditions of PCR assay in HIV infection. Methods: Samples of frozen lymphocytes (8 to 10 x 106 cells per sample) collected from 20 individuals were studied by 9 french laboratories using PCR assay for diagnosis of HIV infection. These 20 individuals included symptomless HIV-1 seropositive subjects (used as positive controls), seronegative at-risk individuals, individuals with an isolated and persistent anti-p24 antibody and seronegative individuals at low risk of HIV infection (used as negative controls). The 9 laboratories worked on coded samples and used three primer pairs. Each laboratory used its own primer pairs, probes and specific conditions for PCR assay. The results were to be given at the end of a three month period. Results and Discussion: In this quality control, laboratories worked on lymphocytes (and not on the DNA) to compare the differences possibly linked to the DNA extraction method. The results of this study underlined the necessity for quality control between laboratories using PCR assay in the diagnosis of HIV infection. Discrepancies between laboratories can exist. This multicentric quality control allowed a determination of the best conditions of sensitivity for using PCR assay in the diagnosis of HIV infection. Conclusions: The use of PCR for HIV diagnosis needs entirely dependable methods which require the participation of the laboratories to a quality control. M.A.1111 COMPARISON BETWEEN HIV1 AND HIV2 ISOLATION FROM PBMC BY A STANDARDIZED COCULTURE IN 24-WELL PLATE. Pein Jean Michelea Simon F.*, Bartczak S.", Gamba E.*, Matheron S.", Dazza M.C.*, Brun.Vezinet F.* Laboratoire de Virologie*, Mal.lnfectieuses", Hop. Bichat-Claude Bernard, Paris, France. Objective: To study the frequency of retrovirus isolation from HIV1 and HIV2 seropositive patients' Peripheral Blood Mononuclear Cells (PBMC) in relation to clinical status. Material: Between June and December 1990, 76 consecutive PBMC were collected from HIV seropositlve African and European patients diagnosed by WB and RIPA: 51 HIV1 (CDC I n=3; CDC /IIll n=34; CDC IV n=14), 23 HIV2 West African patients (CDC I11111 n=19; CDC IV n=4) and 2 HIV1+2 West-African patients (CDC I/II1). Methods:After Ficoll gradient, PBMC were cocultivated in duplicate in a 24-well plate (Costar): 2x106 patient's P8MC per well were added to 2x106 PHA stimulated HIV negative donor's fresh PBMC in 1.5 ml of Medium (RPMI 1640, FCS 20%, human 112 5%,glutamine and antibiotics 1%). Culture supernatants were assessed twice a week for RT activity (micro-assay) and for HIV1 p24 Ag (EIA) during 4 weeks. Results: All HIV1 (14/14) and HIV2 (4/4) AIDS and ARC patients as well as the 2 HIV dual reactive patients and the 3 seroconverters had positive coculture with viral replication (p24 Ag) detected before the 10th day. In asymptomatic patients, 24/34 HIV1 (71%) and 5/19 HIV2 (26%) were Isolated (p<0,005). Using the p24 Ag EIA all the 43 HIV1 and HIV1+2 supematants were strongly positive with an Optical Density/Cut-Off ratio > 30. The 9 HIV2 supematants were all cross-reacting on HIV1 p24 Ag but the OD/CO ratio was < 10 in all cases. Conclusions:By using standardized PBMC coculture, we demonstrated a significant difference in the isolation rate of HIV1 and HIV2 from asymptomatic patients. HIV2 was isolated from all AIDS patients, as previously observed in our laboratory.Comparison of the HIV liters in PBMC and plasma specimens from HIV1 and HIV2 infected patients is in progress.

Page  120 M.A.1112 DEVELOPMENT OF OLIGONUCLEOTIDE PRIMERS AND PROBES FOR AMPLIFICATION OF M 1 AFRICAN AND NORTH AMERICAN STRAINS OF HIV-1 GENOME USING POLYMERASE CHAIN REACTION (PCR). Dawood, Magdyl,2; Allan, R.1; Fowke, K.2; Hammond, G.1,2 1Cadham Provincial Laboratory, Winnipeg, Manitoba, Canada; 2Medical Microbiology Department, University of Manitoba, Winnipeg, Manitoba, Canada. Objectives: 1) To develop sensitive and specific oligonucleotide primers and probes for the amplification of HIV-1 genome using PCR; 2) To test the developed oligonucieotides on HIV-1 positive and negative African and North American blood samples; 3) To compare Southern blot (SB) & liquid hybridization (LH) for detection of PCR products. Methods: Five sets of primers and probes representing env, gag, nef, pol and vif HIV1 genes were developed. Sensitivities were determined using a known number (5-100) of 8E5 cells (each has one copy of HIV-1 genome) mixed with lysates of 2.5 x 105 normal peripheral blood lymphocytes (PBLs). Specificities were tested with PBL specimens from 10 seronegative individuals with no risk factors for HIV. PCR was carried out for 30 cycles. One tenth of the PCR reaction mixtures were analyzed with SB and LH. Primers were also used to amplify HIV-1 genome in PBLs from 4 African and 19 North American HIV seropositive patients. Results: All primers and probes detected 5 copies of HIV-1 genome using SB and LH. Liquid hybridization was faster and gave a stronger signal. All negative specimens were negative with all primers and probes. All the primers detected North American strains of HIV. Pol primers did not amplify the African strains of HIV. Conclusion: 1) These primers and probes are sensitive and specific for the amplification of HIV-1 genome. 2) Some primers are better than others for amplification of African strains of HIV-1. 31 LH is oreferred over SB for detection of PCR oroducts. NOTES M.A.1113 POLYMERASE CHAIN REACTION (PCR) ANALYSIS ON COCULTURE SUPERNATANTS FOR DETECTING HIV IN BODY FLUIDS. VedBrat, Sharan-it S.*; Pierce, P.F.**; Hellman, K.B.*** *Braton Biotech Inc., Rockville, Md. USA, **Georgetown Univ. Medical Center, Washington DC, USA, ***CDRH/FDA, Rockville, Md. USA. OBJECTIVE: We have examined the possibility of using PCR analysis of coculture supernatants (Sups) for early HIV detection to permit repeat testing of valuable clinical samples which either have extremely small numbers of cells or run the risk of contamination when analyzed in long term cocultures for HIV p24 antigen by antigen capture (AC) assay. METHODS: Supernatants on day 3 or 7 of cocultures from seropositive blood samples were tested for HIV by PCR and AC assays. This approach of testing coculture supernatants was also used for other body fluids. RESULTS: Out of 37 day3 supernatants from blood cocultures tested, 33 (90%) were positive for HIV DNA by PCR while only 9 (24%) were positive for HIV p24 by AC assay. Following are the data for other body fluids: # of Samples Direct PCR On Cells Sup AC Sup PR Saliva 8 28% 12% 100% Seminal Fluid 7 43% 14% 86% Urine 7 50% 0% 60% Vaginal Fluid 2 100% 0% 100% CONCLUSION: Testing coculture supernatants for HIV by PCR is superior over AC assay or over direct PCR assay of clinical samples. Such efficiency presents a viable approach for clinical testing of HIV in fluids easily available from Datients, but with limited cAll n1 mhper. NOTES _ _A. 1,14 ISOLATION OF HIV FROM CHILDREN IN ZIDOVUDINE THERAPY. BERNAL ASCENSION*, GARCIA-SAIZ A*, MARTIN-FONTELOS P**, ORTIZ M*, PEREZ-JURADOU ML*, PABLOS A*, NAJERA R*. *CNBCR **CIC INSTITUTO DE SALUD CARLOS III, MADRID, SPAIN. OBJECTIVE. To correlate biologic characterization of HIV strains isolates from syntomatic children in prolonged zidovudine therapy with susceptibility to antiviral drugs. METHODS. Isolation of HIV was attempted from heparinized blood, plasma and cerebrospinal fluid from forteen children. Kinetics of replication were stablish acording p-24 antigen detection in culture supernatant. Susceptibility to AZT and ddl was asessed by replication in MT-2 cells and detection of cytopathic effect in the presence of the drug. RESULTS. HIV was isolated from the blood of eleven of the forteen cases, all CSF were negative and P-24 antigen detection in culture supernatant was positive in four plasma culture at low level. Six isolates were characterized as high/rapid (one resistant to AZT and susceptible to ddl) and five as slow/low. Two patients with a high/rapid isolate died. Studies are underway to isolate the virus after one year treatment. DISCUSSION AND CONCLUSIONS. High isolation rates were obtained (73%) even in AZT therapy, but there was not apparently correlation between susceptibility to antiviral dugs and biologic characterization. Further investigations are required to assess if decreased drug sensitivity is accompanied by biological differences. HU.MAN IMNODEFICIENCY VIRUS TYPE 2 (HIV-2) STRAINS: BIOLOGIC M.A.1115 mETE.ROGEmiTY. Buffet-Janvresse, Claudine; Bertin, C; Callens, C; Coron, D; Pons, JL; De Matteis, M. - Ch.Nicolle Hospital and University - Rouen - France Objective: Biological characterisation of HIV-2 strains isolated from peripheral blood mononuclear cells (PBCM) of Senegalese patients from France (Haute-Normandie region). Methods: HIV-2 isolations were recovered from patients PBMC and cocultivated with an equal number of PHA stimulated human PBMC from seronegative donors. Virus containing culture surpernatants with high reverse transcriptase activity were utilised for infection of PBMC, macrophages and established T, B and monocyte cells lines. Cytopathicity and infectivity were monitored by reverse transcriptase assay, indirect immunofluorescence, syncytium and change in cell viability and expression of surface CD4. Results: 7 out of 23 isolates of HIV-2 were excluded from our study because of their very low infectivity. On the other hand, all the other isolates replicated in human macrophages and lymphocytes, HUT 78 and Sup T cell lines. Three isolates replicated in the monocyte cell line U 937.Five isolates which did not produce high levels of virus, manifested no cytopathicity and CD4 down regulation. The CPE was often associated with viral protein expression in the PBMC and T cell lines and with depression of surface CD4 expression. The time needed to reach maximum production of virus in the culture of lymphocytes was a good marker to differentiate between the isolates. conclusion: Our results indicate the heterogeneity of HIV-2 isolates. Clearly, there is a correlation between the cytopathic effect and the expression of viral proteins and the CD4 antigen. Furthermore, some preliminary results suggest that a differentiation step is required for optimal replication of certain HIV-2 isolates. a rj st cN

Page  121 k M A 1116 PLASMA VIREMIA IN A RANDOMLY SELECTED GROUP OF HIV SEROPOSITIVE M.A.1 INDIVIDUALS. Dewar. Robin L. Samiento, M. Lawton, E., Clark, H. Kennedy, P. Shah, A. *Metcalf. L, *Lane, HC. and Salzman, N.P. Georgetown University, Washington, D.C. and *N.I.H, Bethesda, MD. Obiective: To analyze various paranters that affect the successful isolation of H~V from the plasma of sropositive individuals. Meghod: Whole blood samples collected in different anticoagulant were processed at various times after phlebotomy to obtain the plasma. Plasma was then used either fresh, or after a cycle of freezing and thawing to infect pellets of PHA-stimulated, fresh or froen/dthawed uninfected donor PBMC. Cultures were monitored for virums replication using an in-houe reverse transcriptase ssay, as well as a commercial (Coulter Corportion) p24 antigen capturessay. Raslt U Tree anticoagulant (heparin, sodium citrate, and EDTA) wercompared and plasmrsamples obtained from heparinized whole blood yielded the most consistent recovery of virus. Virus was more fequently recovered from plasma samples obtained and used imnediately after phlebotomy. Freezing of plasma significantly reduced the ability to isolate virus from the samples Comparison of fesh and frozen/tawed PHA-stih lated donor PBMC did not reveal any differnce in thirability to support the replication of virun from plasma. The p24 antigen capture asay was found to be uperior to the reverse trascriptae ssay in detecting virus inculturesupemmats. C lusions: Heparin as the anticoagulant, immediatea supationand se of plasma to infect either freh or froan/thawed donorPBMC, followed by p24 antign captre assys to detect virs in cultre supernatantwere found to be the optimal conditions to recowver V from plasm Te level ofplasm via was found to be inversly related to the total CD4 out of theindividml, ad also varied with the type and legthof ani-vial therapy. We are presently investigating conditions which would allow us to stabilize the virus in plam samples in order to facilitate the ease with which plasma can be used to monitor the effectiveness of therapetic protocols. NOTES M.A.1117 MONITORING IGG, IGA, AND IGM RESPONSES TO HIV INFECTIONS BY MODIFIED ABBOTT MATRIXT HIV-1/HIV-2 ASSAY. Chan. Emerson: Schulze,W.; Leuther,M.; Knigge,K, Abbott Laboratories, Abbott Park, Illinois, U.S.A. Objective: To develop tests for HIV-specific IgG, IgA, and IgM for the monitoring of individual immunological responses to HIV infection. Methods: Biotinylated goat antibodies specific for IgG (y), IgA (a), and IgM (4) were used in a semiautomated immunodot blot assay - The Abbott MATRIX HIV-1/HIV-2 assay. Results: The 3 tests were sensitive (detection limit 0.2 - 0.3 ng per dot), specific (no cross-reactivity up to 30 ng per dot), and capable of detecting all the HTV-1 and HIV-2 antigens on the immunodot array. In samples where HIV specific IgG was in vast excess of specific IgM, the IgM signal was partially inhibited by intra-sample IgG competition for binding to the solid phase. For these samples, a preincubation with immobilized protein G was found to effectively reduce this competition and thereby enhance the IgM signal. Studies with 5 HIV-1 seroconversion panels revealed the following general pattern. There was a weak initial response that involved all three types of antibodies lasting 20 to 30 days. These antibodies were directed predominantly to the HIV-1 p41 antigen. Thereafter, the IgM response subsided, but IgA and IgG continued with IgG at a much higher rate. Discussion and Conclusions: Specific IgG, IgA, and IgM MATRIX HIV-I/HIV-2 assays were developed by using appropriate probes. These tests were sensitive and specific, and permitted the independent monitoring of three classes of immunoglobulin during the immune response to HIV infection. They have revealed characteristic patterns of response for each of the antibody types during seroconversion. Studies with these tests may reveal useful details during seroconversion, disease progression, and maternal transmission. NOTES I \= t^ 00 ^r M.A.1118 DIONOSIS OF PERATAL mY-1 IECTION BY MEANS OF DOUBLE POLYMERASE CHAIN REACTION (PCR). Aktb SIMONONi, VAIRAD", LEPAGE P*, HIMANADO*, ONDA-THULL D" nd VAN DE PEIRE F. AIDS RCtneofmLatboly, Natnl AIDSlConorPolPgaiKigall, Ralndia) ); Tdafimn OCet, Ullsity of Lige,Bemmn1;Depamnts of PedMisatcs, CenbtaHospimthrdeo Kgli, RWal(*). Qbi Tdel: dteiinetousuas of doubbiPC for I dignos of pedm HIV-I iectkn.,Msa: Twv cohohl of Ifant btom bn HV-l infechd (Rip21) amid timiafecthd riwost (ni216) Ihve bene prospectvely fovuled ip for 2 yeas i Kigali, Rw N- HIV-I Westm botis perfonnd onchldm serum anmpble every 3 monts. We developed a two-stage PCR (or double PCR) asing 3 set of HIV-1 specific oligonrcleomes prims (flrm gag, envr anl pol gems)smd l t it n psablin or Afican labolatry aand in Elope. A PCR st is comasied posi3ve in pmsen e of adetlebi simnl for et least 2 out l t 3 peilr-pair Peipealaal mad-bkmod mmamcar cels v bTen teted bydoub B PCR i4 groupof chiMan msbc ia ts two cohom: 25 infan born seDrorgaiive molws (gIp 1), 15 irfatmsof motrenr vih pobt-prl smexoc In (group 2), 40 infas bom sero+ mthlr vwho ot 1beir mmalelHIV-1 antibodis (group 3), ald 25 inm s bornto seo+ motlms vbo rtmamed seMD+ ate te age of 15 monts. M:(seetabbte): Ntmberofpositiv irs ithtdoubblePCR 111 A COMPARATIVE STUDY OF THREE DIFFERENT NUCLEIC ACID AMPLIFICATION ASSAYS,M.A.1119 FOR THE DETECTION OF HIV-1 IN INFECTED INDIVIDUALS WITH ANTIVIRAL THERAPY Huisman Han, Bruisten S, Koppelman M, +Van Gemen B, +Lens P, Klaassens R, +Weigel H. Central Laboratory of the Bloodtransfuslon Service, Amsterdam, NL. Organon Teknika, Boxtel, NL. ++ Onze Lieve Vrouwe Gasthuls, Amsterdam, The Netherlands. Objective: We evaluate the correlation of the detection of HIV by different nucleic acid amplification procedures with the detection of HIV-antibody, HIV-antigen and clinical parameters. Of 31 persons at different stages of HIV infection, simultaneousl) whole blood, cell-free plasma, peripheral mononuclear cells (MNC) and clinical parameters were examined. Methods: DNA-PCR was performed on HMW-DNA using nested gag and pol reaction conditions. For RNA-PCR, RNA was extracted, copied into cDNA by RT and amplified by PCR. The nucleic acid sequence based amplification: NASBAtm, a one step primer dependent isothermal amplification was used for direct amplification of RNA using gag and pol primer sets. Results: Of 31 HIV-antibody positive persons of which 18/31 were positive for HIV-Ag, HIV-DNA was detected in 31/31 (100%) by PCR-DNA in MNC. Both RNA-PCR and NASBA were positive in all plasma samples. NASBA detected HIV-RNA in less than 50% of the MNC whereas RNA-PCR in 100% of the cases. DNA-PCR was predominantly negative for HIV-DNA in plasma (1/31). Conclusions: HIV-DNA could be detected in MNC of all seropositive persons by DNA-PCR. HIV-RNA could be detected in cell-free plasma of all persons even in those negative for HIV-Ag by NASBA and RNA-PCR. This might has implication for monitoring the effectiveness antiviral therapy. The use of NASBA and RNA-PCR for this purpose will be studied. Folbo -up at51e Group 1 GroupS eGup3 GOoup4 0125 1115 340 11125 RP 7115 RP RP RP: Resmul Peaning CoIsi sos: Double PCR soul be coAsniead a neffectiv toolforealydiagmsi s of HIVl ictoniinfaWt vhich ca be cerried outa f Refeeace Labotay of a dmbptig cormUy. MOOVBMo, itplDWd 1uOiful tb evalua poteabenlathmr--cbiltM HIV-I urtanmiss ionus suhas postambluwalanimisnfollwing *r&*-* xysO CmOlMonvsion.

Page  122 M.A.1120 DTECTION OFSPECIFIC HIV ANTIBODIES BY THE USE OF CYANOGEN BROMIDE (CNBr)ACTIVATED PAPER DISCS. Ansoteui Ionacio J.. Lombardi VRM., Scarlatti G., Plebani A., Boscherini B., Rossi P. and Wigzell H. Department of Immunology, Karollnska Institute, Stockholm, Sweden. Dept. of Pediatrics Univ. of Rome Tor Vergata and Univ. of Milano, Italy. Objective. To establish a new method for detecting specific antibodies against HIVl-peptides. Methods. Synthetic peptides from different regions of the gpl20 molecule were coupled to cyanogen bromide (CNBr)-activated paper discs. Based on the results previously obtained by HIV1-peptide ELISA,100 positive and 50 negative sera, at different dilutions, were incubated with the coupled discs. As conjugate 12Sl-labelled F(ab')2 Sheep anti-Human Immunoglobulines was used. Results: No. of samples Mean* Minimum* Maximum* Positive sera 100 34104 10602 60931 Negative sera 50 1719 918 3030 Blank (no sera) 10 1318 1128 1655 * Results expressed in cpm using the sera at the dilution of 1:1000. Discussion and Conclusions. At the 1:1000 dilution, the test is much more sensitive that the peptide-ELISA. Besides, the test still shows a good sensitivity at the 1:10000 dilution, while the peptide-ELISA has not detection ability at this dilution. Here we conclude that this is a very sensitive technique for detecting antibodies to HIV1 epitopes and it might be useful in the epitope mapping of sera from infected individuals during the time of seroconvertion. NOTES M.A.1121 DETECTION OF HIV-1 SPECIFIC DNA IN SERUM SAMPLES OF INFECTED PEOPLE BY POLYMERASE CHAIN REACTION S. Ovcak, M. Likar, B. Drinovec. Medical Faculty, Institute of Microbiology, Ljubljana, Yugoslavia. Objective: To detect whether viral DNA could be detected In the serum samples of Infected people by use an in vitro DNA amplification procedure termed the polymerase chain reaction (PCR). Method: 22 sera of anti-HIV-1 positive asymptomatic individuals and 12 sera of patients with ARC or AIDS were evaluated. DNA was isolated from serum samples by phenolchloroform extraction. The PCR amplification was performed using gag and env primer pairs. In adjcition 63 sera of high risk seronegative persons and 10 sera of low risk healthy blood donors were tested by the same method. Results: HIV-1 specific DNA was detected in sera of 67 % patients with ARC or AIDS and in 17 % of infected asymptomatic individuals. HIV-1 DNA was also found in 3 sera of seronegative persons from high risk group. It is likely that the appearance of HIV-1 DNA in serum is related to disease pathogenesis. The release of HIV-1 DNA may be due to virus production and cell lysis subsequent to virus release although this remains to be shown. NOTES. (3 a hr M.A.1122 HUMAN IMMUNODEFICIENCY VIRUS TYPE 1 VIREMIA AND HIV-1 TRANSCRIPTS DETECTED BY RT-PCR IN CLINICAL SAMPLES FROM INFECTED PATIENTS. Bagnarelli,P*; Clementi, M**; Menzo, S*; Mansin, A*; Giacca, M***; Varaldo, PE*. Institute of Microbiology, University of Ancona* and Trieste**; ICGEB/Unido***, Trieste, Italy. Objective To define the diagnostic value of reverse-transcription polymerase chain reaction (RT-PCR) for the direct detection of (i) HIV-1 genomic RNA in plasma samples (HIV-1 vjremia) and (ii.) cellular specific viral transcripts in PBLs from HIV-1 infected subjects. Methods Genomic HIV-1 RNA from plasmas and HIV-1 transcripts from PBLs were detected using two optimized procedures of RT-PCR. HIV-1 infected individuals (N. 45) were studied and grouped on immunological basis: CD4(+) T-lymphocytes higher or lower than 400/mm (group "A" and "B", respectively). Results HIV-1 specific transcripts were detected by RT-PCR in 23/33 (70%) group A and in 12/12 group B patients, whereas HIV-1 viremia was detectable Jn 32/33 (97%) group A and in 12/12 group B subjects. Discussion and conclusions Detection of cellular HIV-1 transcripts identifies in group A a subgroup of patients presenting viral activity before fall of CD4(+) T-lymphocytes (it may be an indication for early antiviral treatment). Otherwise, most of the HIV-1 seropositive subjects from both groups are viremic (thus indicating that the molecular counterpart of clinical progression is probably not an on/off switch from silency to virus expression). Quantitative RT-PCR approach for HIV-1 viremia is necessary. M.A.1123 THE PRESENCE OF BUDDING VIRUS-LIKE PARTICLES IN HUMAN LYMPHOID CELLS USED FOR HIV CULTIVATION Dourmashkin. Robert. R.*; OToole, C.M.*, Bucher, D.+ & Oxford, J.S.* *The London Hospital Medical College, London, England; +New York Medical College, Valhalla, NY, USA Objectives: To characterise cell tropism, morphology and infectivity of virus-like particles (VLP) in HIV infected and uninfected lymphoid cells. Methods: Lymphold cell lines were grown in RPMI 1640 with 10% FCS (or serum-free medium). Cord cells were isolated by Ficoll gradient and grown in the same medium. HIV Illb infected and uninfected cells were used. VLP containing culture media were fixed in 2% glutaraldehyde, clarified and pelleted at 100,000g x30mln, then negatively stained with 2% phosphotungstate. Cells were fixed in sequential glutaraldehyde and OsO4, stained (or not) in 0.5% uranyl acetate and embedded for sectioning. Results: VLP were found in T-cell lines CEM, H9 and C8166; in 2 lines of EBV transformed B-cell lines; and in cultures of primary human lymphoid cells from cord blood, which were either PHA stimulated or not and grown with or without serum and in cord lymphocytes directly after Ficoll separation. The particles were absent from a B-cell line (NYMC) and a human bladder cell line (LHMC). The particles were smaller (50-60nm) than HIV particles and possessed an external fringe of peplomers and a single enveloping membrane. In sections of cells the particles had a very dense core and were budding from the cell membrane. There were no virus-like intracellular structures. VLP/cell was 3000-30 000. VLP: HIV particles was 1:10 to1:100. VLP budding areas were distinct from HIV areas. Discussion; 1) VLP are not bovine viruses associated with serum or other medium. 2) The VLP's are endogenous in human cells. They may be a defective retrovirus, with an atypical morphology. 3) Alternately the VLP's may represent an EM artefact or a subcellular organelle. The absence of VLP in one cell line argued against this. VLP transmission to VLP negative cells has been attempted unsuccessfully. 4) VLP were previously found only in HIV and SIV infected cells. 5) Do VLP affect HIV replication?

Page  123 M.A.1124 NEW ASSAY FOR DETECTION OF REPLICATING HIV BY BIOLUMINESCENCE Honigman, Alexander*; Israel, S*; Lawatti, P**; Vonsover, A***; Ulitzur, S****; Swisa, M****; Bentwich, Z** *Dept. of Molecular Genetics, Hebrew University Medical School, Jerusalem; **Ben-Ari Institute of Clinical Immunology, Kaplan Hospital, Rehovot; ***Central Virology Laboratory, Tel Hashomer; ****Virolume Ltd., Haifa, Israel Objectives: Develop an assay for active replicating HIV by transactivation of luciferase genes through HIV-TAT transacting polypeptide. Methods: The Firefly luciferase gene was fused to the HIV-1 LTR and introduced into human kidney cell 293, using the pZIPNEOSV(x) retrovirus vector. The basal expression of the HIV promoter was lowered by the cloning design. Two stably transfected cell lines were isolated: one, (293/luc) harboring the luciferase constructs, the second (293/luc/CD4), harboring also the CD4 receptor. Results: Pronounced increase in light emission ( x3,000 fold) followed TAT transactivation. This was observed following transfection with a TAT producing plasmid and also by HIV infection. Light was also emitted from the tester cells after addition of extracts of lymphocytes from HIV infected H9 cell line and from1 HIV infected individuals. Using this method and monitoring light emission by polaroid film in a simple, specially devised camera kept in a CO2 incubator, we were able to detect minute amounts of HIV-TAT in the tested samples within a short time. Discussion and Conclusions: This system offers a much needed easy and fast assay for active and infectious HIV particles and thus may shed light on viral load in different phases of HIV infection and facilitate testing, monitoring and screening of antiviral therapies. NOTES M.A.1125 DETECTION AND DIFFERENTIATION OF HIV ISOLATES BY PCR ANALYSIS OF THE VIRAL PROTEASE GENE. Pieniazek Danuta*; Owen MS*; Peralta JM**; Krebs JW*; Ferreira Ramos Filho C; Weniger BG*; Pieniazek NJ*; Schochetman G*; Rayfield MA*. *Centers for Disease Control, Atlanta, GA, USA, and **Hospital Universitario, Rio de Janerio, Brazil. Objective. The viral protease gene contains both conserved and variable regions of sequence that display significant differences between HIV-1 and HIV-2. The conserved elements of the gene are useful in diagnostic detection of proviral DNA and genotyping of HIV-1 and -2 infections, whereas analysis of the variable sequences permits differentiation of distinct isolates within a viral serotype. Methods. Validation of this approach was done through oligomer hybridization, restriction enzymes digestion patterns and sequences analysis of the entire HIV protease gene using a variety of cloned HIV protease genes and cultured HIV-1, -2 and mixed HIV-1/-2 infected cells. Results. Our data indicate that: (1) PCR amplification and detection of HIV protease gene is very specific and permits the distinction of HIV-1 and -2 DNA; (2) sensitivity of detection is equivalent to 1-5 copies of HIV-1 or -2 DNA per 1 ug of cellular DNA; (3) sequence analysis confirmed the distinction of isolates within a given serotype that vary by 4-6% in their protease gene sequence. Conclusions. PCR analysis of viral protease gene provides an efficient method for simultanously detecting heterotypic HIV infections as well as differentiating between HIV isolates within the same viral serotype. NOTES M.A.1126 UnAMBIGUOUS DETECTION OF HIV-I PROVIAL DNA IN NEWBORNS FROM INFECTED MOTHERS BY POLYMERASE CHAIN REACTION Salvi St, Tresoldl E.*, Gaudi S*,. Montermini L., Russo S.*, Siccardi A.i Novati R.**, Lazarin A.**. and Andronico Yrancesca*** *Dipartim. Biol. Genet. Scl. Mad., Milano, Italy. ** Ospedale L. Sacco, Milano, Italy. *** Ist. Bioch. Prot. Eazim., CNR, Napoli, Italy. Objective: We carried out the Polymerase Chain Reaction (PCR) on peripheral blood mononuclear cells DNA of newborns from infected mothers for an early detection of the provirus. The main goals of the work were: 1) to optimize the PCR conditions in order to obtain unambiguous results; 2) to assess thu diagnostic value of the method. Methods: A cohort of 17 newborns have been followed so far: iamunological, clinical and PCR data were collected for all of them at 2-3 months intervals; some of the newborns have been followed already for more than 30 months since birth. For the PCR, we have routinely used three sets of primers: one has been chosen within the fag gene sequence, the other two ustu were designed as "nested primers" and were chosen within the env gene. Resultn; The PCR conditions used allow to unaibiguously detect proviral DNA at the earliest stage of infection In those newborns who later developed clinical signs, and no proviral sequence where detected in the uninfected ones as judged by the other criteria. Discussion; The future perspective of the work will be to characterize the in vivo variants initially present and emerging with time in the patients. SCO-AMPLIFICATION OF SPECIFIC SEQUENCES OF HIV-1 AND HCV GCENOlES BY 7E M.A.1127 POLYMERASE CHAIN REACTION ASSAY: A POTENTIAL TOOL FOR THE SIMULTANEOUS DETECTION OF HIV-1 AND HCV gediar, Sayah; Hewlett, I.; Biswas, R. Lab. of Hepatitis and Lab. of Retrovirology, DTS, CBER,FDA, Bethesda, Maryland 20892, U.S.A. Objective: A rapid and simple method using the polymerase chain reaction (PCR), was devised for the co-amplification and simultaneous detection of human immunodeficiency virus, type 1, (HIV-1) and hepatitis C virus (HCV) specific sequences in the same clinical sample. Methods: Genomic RNA was extracted from 13 blood donor sera that were reactive in ELISA for both anti-HIV-1 and anti-HCV. The extracted RNA was amplified with reverse transcriptase, using primer pairs SK 38; SK 39 for HIV-1 and "SN01 and SN02" from clones 81 and 37b for HCV, simultaneously. PCR products were analyzed by liquid hybridization and Southern blot hybridization with P end labeled oligonucleotide probes from the regions between the two primer pairs, excluding the primer sequences. Resutle: HIV- 1and HCV primer pairs were able to detect as few as 10-20 copies of RNA for HIV-1 and 1-3 copies of RNA for HCV, when compared with 8E5 cells as the single copy standard. HCV-RNA was detected in all 13 samples and RIV-KeA was detected in 11 (85%) samples. The remaining 2 (15%) samples had detectable levels of HIV-1 Genomic DNA sequences. conclusion: The ability to co-amplify specific sequences from two different viral genomes in the same reaction mixture provides a simple assay for the simultaneous detection of multiple viral sequences. Moreover, it offers the possibility of simultaneous diagnosis of two infections with two different viral agents in serum samples of symptomatic and asymptomatic individuals. b '^ ~I t L

Page  124 S%.A.1128 PCR AND VIRUS ISOLATION OF HIV INFECTED INDIVIDUALS: ANALYSIS OF ASYMPTOMATIC SEROPOSITIVE ADULTS AND MOTHER/NEWBORN PAIRS. ROQUES. PIERRE; VASLIN, B.; MARCI, D. AND DORMONT, D. CEA and CRSSA, DSV/DPTE/SSA, Fontenay aux Roses, France. Objective. To compare the efficacy of 3 different methods: polymerase chain reaction (PCR), peripheral blood mononuclear cells (PBMC) culture and coculture in the study of the virological status of asymptomatic HIV infected individuals. Methods. Patient Population: The cohort included 89 adults containing 16 well documented HIV infected patients and 73 pregnant women seropositive for fHV-1 (ELISA and Western blot). 30 childrens aged less than 9 month born from seropositive mothers were also analysed. Clinical status were ranging between asymptomatic and ARC. EK: 3 couples of primers were used in the same reaction mixture (SK01-SK39, P3-P4 and El-E2 in gag, pol and env regions). The amplification products (750 bp, 307 bp, 455 bp) were detected by Southern blot with radioalabeled HIV-1 Bru total genome probe. Control of contamination and sensitivity of 10 HIV copy by million cells were used. Virus isolation: PBMC were isolated from patient and either stimulated with pHA-p or cocultivated with normal human cord lymphocytes (2/1). HIV production was tested using both p24 antigen detection (Abott EIA) and reverse transcriptase assay. Results. 250 cultures and PCR were performed and 138 analysis are run off (PCR and coculture). In the 16 seropositive adults group, 100 % of the PCR were positive, 40 % of the pHA-p culture and 95 % of the coculture. For pregnant women, we obtained 86 % of double positive (coculture, PCR) examination data and 98 % of PCR positive. 7 out of the 30 newborns were infected as determined by PCR, culture and clinical examination, 21 sampling were analysed: 71 % positive cocultures, 85 % positive PCR and 15 % negative PCR with 23 % of divergence between PCR and coculure (4 PCRculture a nd PCR-,culture+). 9 non-infected childrens remained negative for the double test (20 assays). Conclusion. In adult group, PCR sensitivity is the same than the coculture with cord blood lymphocytes, and better than the pHA-p culture (p< o,05); but, for pregnant women and newborn, the culture/PCR analysis seems to be less effective; this might be related to the particular immune status of both pregnant women and newborns. M.A.1129 Seru. HIV Antigen and anti-P24 Antibodies Levels in HIV seropositive patients: CLinicaL evaluation of a new generation's semi-quantitative microeLisa systems. Dominique VIGNON TRANSFUSION CENTER FOCH HOSPITAL FRANCE Because serum Ag levels do correlate with the Level of infectious virus in serum, Ag assays have been indicated for use of monitoring patients' clinical disease progression and response to antiviral therapy. Moreover, progression to AIDS has been associated with a decline in anti-P24 levels, whereas antibodies to enveloppe proteins remain detectable. Objective: To study the course of these viral markers by the follow-up of seropositive patients. To evaluate the reproductibility, sensitivity and specificity of new generation reagents. Methods: Both methods (from Organon Lab.) use wells coated with murine monoclonal antibodies to HIV P24; human (HIV Ag test) or murine anti-HIV P24 (anti-HIV Core test) coupled to H RP serves as the conjugate, with TMB and peroxide as the substrate. 1) "VIRONOSTIKA HIV Ag" is an enzyme immunoassay based on the sandwich principle. Semi-quantitative calibration curve is obtained by plotting the signal/cutoff values of 5 serial dilutions of the positive control, versus their relative P24 Ag concentration (160 to 5 pg/m). All repeatably positive specimens should be confirmed using the "NEUTRALIZATION test" which requires the human neutralizing antibody to HIV of the kit. 2) "VIRONOSTIKA HIV Core" is an elisa based on a competitive sandwich inhibition principle. Competition is obtained for a fixed amount of HIV Ag (HIV/H9 viral lysate). By testing a dilution series of the sample, the titer of anti-HIV P24 can be estimated; the titer is the highest dilution which still reacts positively. Results: 35 Sera from previously known Ag positive patients and from seropositive patients followed for 2 to 5 years were tested in duplicate at Least in two series. The reproductibiLity of both techniques was good; the sensitivity correlated with the clinical course or therapy; comparison with results obtained with other elisa systems was most satisfactory. NOTES NOTES M.A.1130 DEVELOPMENT OF AN AUGMENTED WESTERN BLOT ASSAY FOR CHARACTERIZATION OF SAMPLES REACTIVE BY COMBINED HIV-1/HIV-2 SCREENING ASSAYS Chan, Lily; Yin, May Fong; Tay, Hong Soon; Sum, Yoke Wah; Griffiths, Peter J; Reed, Donna; Diagnostic Biotechnology, Singapore Objective: To develop a supplemental serological assay to characterize samples found reactive by combined HIV-1/HIV-2 screening assays. Methods: A HIV-2 specific envelope peptide and anti-human IgG were slot blotted onto one end of nitrocellulose western blotted with HIV-1 viral lysate. The sensitivity and specificity of this augmented western blot system were assessed with a panel of negative, HIV-1 seroconversion, gag reactive, HIV-1 and HIV-2 seropositive sera. Results: All sera reacted with the anti-human IgG band which controlled for sample addition to minimize the risk of false negatives due to operational errors. HIV-1 seroconversion, gag reactive and HIV-1 seropositive sera had characteristic viral lysate profile bands on the immunoblots but did not react with the HIV-2 peptide band. All HIV-2 seropositive sera tested reacted with the HIV-2 specific peptide band. Conclusion: Because of extensive immunological cross-reactivity of HIV1 and HIV-2, it is often not possible to distinguish a gag reactive sample from a true HIV-2 seropositive sample. Our results show that such an augmented system greatly improved the ability to characterize samples found reactive by combined HIV-1/HIV-2 screening assays. M.A.1131 ENZIMATIC AMPLIFICATION OF HIV-1 DNA IN PERIPHERAL BLOOD MONONUCLEAR CELLS FROM PEDIATRIC PATIENTS Alba,Concepci6n: Delgado R.; Fuertes A.; Ruiz Contreras J.; Gomez Castillo E.; Noriega A.R.; Lopez Archilla D. Hospital 12 de Octubre. Madrid. Spain. Objetive: The purpose of this work was to compare the diagnostic value of PCR and p24 antigen in blood from children born to HIV seropositive mothers. Methods: Patients were classified using CDC criteria for pediatric HIV infection.The group was very heterogeneous in relation to age at the time of the blood sampling and the stage of development of the disease. Anti-HIV antibodies were tested by a commercial ELISA and western blot. The HIV p24 antigen was measured by antigen-capture ELISA. Peripheral blood mononuclear cells were studied to detect the presence of conserved regions of the HIV-1 gag gene, using PCR followed by hybridization with radiolabeled probes. Results: Sixty-seventeen children were studied. They belonged to the following groups: PO (30); P1 (8); P2 (12); the remaining 17 were seronegative at the time of sample collection. Of the 30 in P0, 29 were PCR(-) p24(-); 1 was PCR(+) p24(-) and became symptomatic 1 month later (P2). Of the 8 in P1, 2 were new-born and PCR(-), p24(-),4 were PCR(+),p24(+) and 2 PCR(+),p24(-).All 8 eventualy became PCR(+) but 2 remain p24(). In P2, 9 were p24(+) and all PCR(+). The 17 seronegative patients were repeatedly p24 and PCR(-). Conclusions, PCR is a sensitive and specific method for the diagnosis of HIV pediatric infection, in some cases (6/67) more sensitive than antigen detection. However, in the neonatal period, PCR may fail to detect HIV DNA, probably due to a very low blood viral burden. ~ C> c> a C> C>r~

Page  125 MA1132 MATRIX HIV-I/HIV-II: A NEW ALTERNATIVE TO THE WESTERN BLOT.A.1^ Niedbalski, Joseph S,, Dahlen, S., Braun, B., Knigge, K., Stephens, J., Daluga, C., Whiteside, M. Abbott Laboratories, Abbott Park, IL 60064. A new semi-automated HIV assay (Matrix HIV-I/HIV-II) has been introduced as a Research Use Only product. This new assay system shows sensitivity and specificity to antibodies simultaneously against both HIV-I and HIV-II. Highly purified recombinant proteins (greater than 90%) of HIV-I and HIV-II are immobilized on a solid phase consisting of nitrocellulose. The assay is performed in a 35 degree incubator automated for timed incubations, washing, drying, reflectance reading and sample removal. A digital printout records the intensities as a sample to cutoff value from the color reactions. The assay requires less than 6 hours to perform and less than 30 minutes of user attendance. The assay was 4 to 128 fold more sensitive than the currently available Western Blot assay for the HIV-I antigens gpl20 and gp41 when compared to serial dilution panels. Results from a 49 member seroconversion panel, consisting of 9 donors, indicated that the Matrix HIV-I/HIV-II assay confirmed the presence of HIV-I as much as 7-10 days earlier than the Western Blot, when following the CDC criteria for positive interpretation. Additionally, 300 HIV-I and 200 HIV-II known positive samples confirmed positive by both the Matrix HIV-I/HIV-II and Western Blot assays. over 500 negative and Western blot indeterminate samples were tested and resulted in an indeterminate rate of 6.22%. NOTES. M.A. 1134 PEVWELOPMENT AND EVALUATION OF BIOSENSORS FOR HIV-SEROLOGY S A 1 Aberi, Franz; *Woias, Peter, **Koch, Sabine; I**Klinger, Conrad; ***Drost, Stefan; -Wolf. Hans Institut fu5,Molekulare und Tumorvirologie am Max von Pettenkofer-Institut, Ludwig-*laximilians-Universitat Muinchen; Lehrstuhl ffir Integrierte Schaltungen, Technische Universitit Mflnchen; Fraunhofer-Institut f&r Festkorpertechnologie, Miinchen; Germany, During the last decade a new generation of sensors, the biosensors, has been developed for the quantitative measurement of biological substances. Application and specifity of a biosensor is determined by the recognizing biological/biochemical layer and the type of transducer. According to the transducer used, biosensors can be classified as electrodes, optodes, thermistors, piezoelectric crystals and fieldeffect transistors respectively. The specific interactions between enzymes and their substrates, antibodies and antigens or between receptors and hormones are responsible for the selective binding of biological components in solution. The HIV-system has been chosen as a model system. A peptide antigen from the p24 core protein and murine monoclonal antibodies were used for initial experiments. Experiments under more realistic conditions were performed with a consensus peptide from the hypervariable V3-loop of the fIV-virus and the reactive rabbit immunosera. In scrological diagnosis the rapid and sensitive detection of a vir.l antigen or the induced immunological response is of vital interest. Other fields, such as the utilization in production t~d control processes, extend the spectrum of possible applications for such sensor systems. Presently we are testing ionsensitivc fieldeffect transistors and piezoelectric crystals with respect to their suitability for the detection of antibodies. The immunosensors are evaluated wid optimized with regard to their turn-over periods, sensitivity and cross-selectivity under exactly defined conditions. The use of a sensor system on the basis of a biosensor generally possesses advantages like reduced influence of envircnmental interferences, short response times, and the possibility to develop "intelligent sensors" with the help of integrated signal processing. This contribution will present results from experiments performed with both types of transducer, as well as a discussion of the advantages and draw-backs of the respective systems. M.A.1133 HV ISOLATION FROM PLASMA, LYMPHOCYTES AND WHOLE BLOOD IN SUBJECTS IN DIFFERENT STAGES OF INFECTION AND CHEMOTHERAPEUTIC TREATMENT. Sigarl G1usaeppna Lost E.. Zemignan M., Arenare L, Brignole G.. Wicks R. Cuneo-Crovari P. Institute of Hygiene and Preventive Medicine. Genoa University, Genoa. Italy. Objective. The study was carried out in order to evaluate whether HIV detection In plasma can provide more precise informations on the infection development in comparison with virus detection In lymphocytes IPBMC). with or without Zidovudina treatment. We tried moreover to verify the efficiency and the significance of HIV isolation from whole blood. Methods: HIV isolation was contemporaneously attempted from PBMC and plasma in 65 subjects belonging to the II11 Ill. IVa and Vc group of the CDC classification. The technique that we utilized consisted in cocultivation of PBMC with seronegative blood donors lymphocytes previously stimulated with phytohemagglutinin (48-72 h). Plasma from the same subjects, centrifuged and properly filtered was pooled (1 ml In blood donor lymphocytes prepared as mentioned above. Heparinized whole blood was Inoculated in measure of 100 and 500 pl. HIV replication was evaluated with the core protein p24 detection in the cultural fluid and with the optical microscope observation of the cultural cells. Results: Virus isolation was contemporaneously effected from plasma and PBMC. HIV isolation from plasma. especially In advanced stages of infection can reach high values (10-). AZT treatment did not affect HIV isolation rate, whereas it has reduced viremia titre, even if there was no statistical significance. Fourteen subjects in the II e III group underwent a 1 year follow-up. Those who resulted positive to viral isolation from plasma progressed towards more advanced stages of infection with a higher rate than those who resulted negative. HIV Isolation from small quantities of whole blood was possible. Generally whole blood cultures provided a p24 positivity earlier and with higher Ag titres in comparison with PBMC and plasma cultures. The reasons of such a behaving are not very clear and further tests are under study. Discussion and conclusions: HIV isolation from infected subjects can provide a different and complementary knowledge in comparison with data based on the patient's immunological response. Our research shows that HIV isolation and titration in plasma together with viremia pattern allow a better diagnosis of seropositive subjects; moreover It can represent an earlier prognostic marker in comparison with those available until now. The high positivity rate we can obtain with the cultivation of small quantities of whole blood, in comparison with HIV detection in plasma and PBMC. is of great practical Interest and it deserves further studies. NOTES DIAGNOSIS OF VERTICAL HIV-1 TRANSMISSION USING PCR AND DRIED BLOOD SPOT M.A.1135 (dbs) SPECIMENS. Cassol, S.*; Lapointe, N.; Salas, T.*; Hankins, C.***; Arella, M.**; Fauvel, M.***** and M.V. O'Shaughnessy*. *Federal Centre for AIDS, Health and Welfare; **Hopltal Ste-Justine; **Montreal General Hospital; "**Institut Armand-Frappier; *****Laboratoire de sante publique du Quebec, Canada. Objective: To conduct a blinded, prospective study using PCR to identify HIV DNA in dbs samples collected from infants perinatally exposed to HIV. Methods: A total of 193 serial dbs samples from a cohort of 26 mothers and 44 children were tested for HIV1 specific sequences using primers to env and gag. To determine the efficacy of the dbs method, PCR results were correlated with viral culture, serology, p24 antigen, CD4 counts, IgG levels and clinical status. Results: Twenty-two infected mothers were strongly PCR positive with both primer sets; 2 were indeterminant (positive by only one primer set); 1 showed conversion to both PCR and culture positivity during the course of the study; and 1 remained PCR and culture negative, despite being seropositlve. Within the pediatric population, strong PCR signals were generated In 100% of 32 samples from asymptomatic (P-l) children with positive culture and (or) p24 antigen tests and in 100% of 17 samples from children with both clinical (P-2) and laboratory (positive culture, p24 antigen) indices of HIV infection. Of these samples, ten came from 7 neonates between 14 days and 3 months; 13 were from infants aged 5 to 18 months and 16 were from young children between 19 months and 8 years of age. No HIV-1 DNA was detected in dbs samples from 60 seronegative blood donors or 64 out of 66 (97%) samples from children who lacked all clinical and laboratory evidence of HIV-1 infection (Po). Two initially indeterminant samples were found to be PCR negative on repeat testing. Conclusion: dbs PCR is cost-effective, requires only minute amounts of whole blood (heel-prick) and it presents a useful alternative to viral culture for the early identification of infected and uninfected infants. 'a ~1

Page  126 I, 0\ M.A.1136 SEROLOGICAL AND MOLECULAR ANALYSIS OF HIV-1 AND HIV-2 DOUBLE PROFILES. Leonard Guy ', Courgnaud V **, Sangar6 A., Denis F., BrAchot C. * Virology department, CHU Limoges, France, ** INSERM U75, Necker, Paris, France, *** Pasteur Institute, Abidjan, Ivory Coast Qbiective: To analyse serological HIV-1 and HIV-2 double profiles in different groups of population of Ivory Coast by mean of PCR. Methods: Individuals from groups representative of the general population an( patients with AIDS or AIDS-like diseases were recruited in 1990-1991. Blood sample, were analysed by EIA (Abbott HIV-1 and HIV-2 Recombinant) and then by HIV-1 and HIV2 western blotting (Diagnostics Pasteur). PMBC were isolated and, after DNI purification by classical methods, PCR was performed using SK100 and SK104 primers. The amplification products were then hybridized with specific probe of HIV-1 (SK19) and specific probe of HIV-2 (SK109). Results: Preliminary results were obtained from 35 AIDS patients. Serological analysis showed 18 HIV-1, 1 HIV-2, 9 HIV-1+2 double profiles and 7 seronegative. 01 the 9 HIV-1+2 serological individuals, 3 showed only HIV-1 infection determinated b3 PCR, 3 only HIV-2 and 3 both HIV-1 and HIV-2. Conclusion. These results concerning 9 AIDS patientS with serological HIV-1+2 double profile infection revealed discordant results between serological methods anc molecular analysis. HIV-1 and HIV-2 DNA was recovered in less than half of serological HIV-1 and HIV-2 double profiles. However, to confirm these results, numerous samples will be tested in the next months and complementary results will be presenting at the meeting. NOTES MA 1137 CONTAMINATION AND ITS AVOIDANCE IN LABORATORIES PERFORMING HIV-1 PR..A. -*Dwyer, Dominic E., and Saksena, Nitin K. *Unite d'Oncologie Virale and Unite de Biologie de Retrovirus, Institut Pasteur, Paris France. More laboratories are starting to use the polymerase chain reaction as a diagnostic tool in HIV-1 infection. The major complication of PCR is the detection of false positive reactions, usually due to previously amplified PCR products. To investigate this problem problems in two laboratories that undertake HIV cultures, PCR and other molecular techniques, the following experiment was performed. Empty uncapped eppendorf tubes were left for 6 hours during a working day in different rooms (PCB is normally carried out in only one room) of each laboratory, followed by the addition of 10ul of H20 and centrifugation. PCR mix containing either the SK38,39 or SK68,69 primer pairs (the most commonly used HIV-1 primers in each laboratory) was added, and 30 cycles performed. Samples were run on agarose gels before being transferred and hybridised with appropriate probes. Using this method, material was amplified from 11/14 rooms in one laboratory, and 8/13 rooms in the second. In the 19 'positive' rooms, 8 were positive with only one of the primer pairs, and in 10 rooms products were only seen with hybridisation. Samples collected in two non-HTV laboratories wore negative.To assess the efficacy of autoclaving and UV irradiation (techniques previously recommended to prevent PCH carryover) in decreasing contamination, serial dilutions of Ing of a 115 base pair PCr product (SK38.39 HlV-1 primers) were made. Exposure of these dilutions to autoclaving (120~ for 45 minutes) and UV irradiation, followed by reamplification using the same primer pair reduced reamplification by only 3 dilutions and 1 dilution respectively compared with untreated samples, suggesting that neither technique is reliable in eliminating previously amplified products as a source of false positive reactions.These experiments suggest that PCR contamination can spread easily around a laboratory, presumably carried by laboratory personnel, materials and aerosols, and that rigourous physical separation of the various stages of PCR is the most effective way to minimise PCR 'carryover'. These factors need to be carefully considered by laboratories wanting to use PCI as a routine diagnostic tool. NOTES M.A.1138 POLYMERASE CHAIN REACTION (PCR) SENSITIVITY AND SPECIFICITY IN INFANTS AND CHILDREN. Nelson, Robert; Price, L; Di Nicolo, R; Halsey, A; Day, N; Lockey, R; Good, R. University of South Florida, Department of Pediatrics and All Children's Hospital, St. Petersburg/Tampa, FL, USA Objective: To assess the sensitivity and specificity of PCR for establishing the diagnosis of HIV infection in infants and children. Methods: Serial testing was performed on samples from seropositive and seronegative infants at risk for HIV infection and from seropositive children. PCR was done using the primer pair SK38/SK39 on blinded whole blood samples sent to Specialty Labs, Santa Monica, CA. Samples were obtained every 6wks to 3 months (m) until the diagnosis was established (+ HIV culture, + p24 antigen, + HIV antibody after 15m or an AIDSdefining illness). Two-hundred seventy-five samples from infants/children (n=172) were tested. Results below are stratified according to age at the time of testing. False negative (FN) and false positive (FP) percentages are noted. Eight samples were indeterminate, in that control HLA sequences did not amplify. Results: HIV-infected HIV non-infected Sero +. culture - (P-0) 515m(n-21) >15m(n=32) l15m(n-66) >15m(n-19) <15m(n-34) #PCR 30 37 108 32 60 +PCR 28 34 1 0 1 -PCR 2 3 107 32 59 FN 6.7% 8.8% - - FP - -.9% 0% 1.6% Discussion and Conclusions: PCR, using SK38/SK39, is a useful test to detect HIV in infants and children. False negatives, and to a lesser extent, false positives occur. M.119 HIV ISOLATION AND ANTIGENEMIA IN HEMOPHILIACS AND IN M.A.1139 HOMOSEXUAL MEN van den Akker,Ruud*; Vijge,E*; Meijer,A*; Mauser,E**; Borleffs,J***; van Loon,A*. *RIVM, Lab.of Virology, Bilthoven, the Netherlands, **van Creveldkliniek, Bilthoven, the Netherlands, ***University Hospital, Dep.of Internal Medicine, Utrecht, the Netherlands. Objective: To follow virological events in HIV seropositive hemophiliacs and homosexual men. Methods: 10x106 ficollgradient purified peripheral blood mononuclear cells (PBMC) were incubated with three days-PHA stimulated PBMC from healthy donors. Twice weekly the culture supernatants were examined for viral p24 protein. Results: HIV was isolated from all individuals (table). CDC HIV isolation Group Stage No.positive/ No.positive within No.HIV No.tested 7 days (%) antigenemia (%) hemophiliacs II 38/38 29 (76) 3 ( 8) homosexual men II 32/32 24 (75) 10 (31) homosexual men III 16/16 12 (75) 5 (31) homosexual men IV 39/39 32 (82) 17 (44) In 68 of the 88 (77%) HIV antigenemia negative individuals and in 29 of the 37 (78%) antigenemia positive individuals HIV was isolated within 7 days. Conclusion: - hemophiliacs are actively infected. - significant difference in antigenemia was found between asymptomatic hemophiliacs and homosexual men. - rapidity of HIV isolation was not correlated with the stage of infection. a ^ss 5 C> C> C> I>

Page  127 M cAc R 11MM MUR II1 iiNE OF 72M CM CELL LINE DEWMO TRAE PALPIAL BIDCKADE M.A.1140 1 -flD1IU 11w 1 S OFc THEIR 1W-!iPENDEq PATHWAYS AND S TBCEPTIT HIV 1 INFECMK ý iIa.Lln Kaxthla; Gruber, N. F.; Kaufman, J.? Blackburn, R. * KMnischewitz, J.: Norcross, M.; Golding, Hana. Division of Virology and Division of Cytokine Biology, CBER, FDA, Bethesda, Maryland, USA. The mataan EMS was used to derive mutants from the human cell line CEM. several mutants, eepressing signitiant number of CD4 recpors, and binding gpl20, showed reduced susceptiblity to infecti with multiple strains of HIV 1 as judged by syncytia formation and RT assays. Quantitative PCR assays allowing lAction with 17 1, etecte d imilar amounts of viral DNA in the mutants and the parntal line 6 hrs pot ifection. But after 72 hrs, the amount of viral NA in these mutants was 5-10 fold lower than n the parental or poaltive ctrol mutant lines. e To identify the cellular mhanisms responsible for the reduced susceptibility to HIV 1 infection in the CNN mutants. ethods (a) The mutats were t ted with the PC activator TPA, and were analysed for (1) surface epresion of CD4 CD3, IL2R; (2) phoUphaylatitn of cellular proteins; (3) induction of nudlear NFkB binding proteins. (b) CAT assays were performed using HIV 1 (-139)LTR-CAT constructs with TPA or with TAT-xrpression vectors. (c) Direct PKC enzymatic assays. Results: The deficient mutants showed abnormal responses to TPA treatment: (a) CD4 was phaoporylated and down modulated froa their surface within 2 brs, but unlike the controls, nduction of CD3 and IL2R was not seen een after 48 hr. (b) CAT activity using the LTR-CAT vecaor+TPA was completely negative, and TAT could not transactivate LTR-CAT via the TAR elaneat.(<c TPA-inductick of PFB binding proteins was much reduced. (d) PKC enzymatic assays were normal in these mutants, suggesting that the blockade is distal to the enzyme itself. ia-gsan Further analysis of these mutants would allow identification of cellular proteins which are PKO-dependnt and are required for optimal RV 1 infection and TAT transactivation. NOTES M.A.1141 FINE ANALYSIS OF THE IMMUNODOMINANT REGION FROM HIV TRANSMEMBRANE PROTEIN BY CHEMICAL MODIFICATION OF THE SYNTHETIC PEPTIDES Andreev S.M.,Mescherjakova D.V.,Azmuko A.A.,Khaitov R.M. Institute of Immunology,USSR Ministry of Public Health,Moscow Objective: To define the amino acids involved in IgG reactivity to five HIV-1 gp41 peptides overlapping the sequence 584-624 the method based on the chemical modification of three-functional amino acids,especially Lys residues,was suggested. Methods:The acetylation of Lys E-amino group(pH 8-9) was used.The reactivities of the sera from 75 HIV-infected individuals and gp41-specific human mAb(CBHR/3D6) were studied using peptides and their modified forms in indirect and competitive ELISA. Results:Peptides 584-603,609-624 reacted with 80% of HIV-positive sera;the most diagnostic significance(100%) was found to have 584-612 and 603-624 peptides,all sera reacting with the latter in titers higher than other ones.The acetylation resulted in 10-15% 0ecrease of peptide reactivity.Moreover 50%-inhibition oncentration was 1.5x10" M for nonmodified 584-612 peptide compared with 1.5xlO M for modified one. CD-spectra showed that the modification failed to alterate the conformation of these peptides.A coupling of 584-612 peptide to protein carrier at the pH 6.5-7.0 did not influence immunoreactivity of this peptide.HmAb reacted only with 603-624 peptide. Epitope mapping by using of HmAb resulted in conclusion that critical epitope is arranged within 603-609 gp41 region. Conclusion:The proposed approach is a more simple variant of epitope analysis in contrast to sets of deletion or amino acid substitutions.Lys-608 appear to be exposed and involved directly in the interaction with antibody combining sites.In addition Cys-603 and 609 appear to have a structural role in this immunodominant domain. NOTES s 00 1-I a 00 ty M.A.1142 IMMUNOREACTIVITY OF SELECTED PEPTIDES FROM HIV-1 gpl20 PRINCIPLE NEUTRALIZING DOMAIN(PND) AND CD4-BINDING REGION(CD4-BR) Khaitov R.M.,Mescherjakova D.V.,Andreev S.M.,Sidorova M.V.,Trubcheninova L. Vafina M.G.: Institute of Immunology,USSR Ministry of Public Health,Moscow Objective: To investigate the serum reactivity of HIV-1 infected individuals with the set of the overlapping synthetic peptides derived from PND of the BRU and MN strains (3 peptides) and CD4-BR of the BRU(4 peptides);to evaluate the correlation between the presence of antibody response to PND,CD4-BR and the clinical status.Method; To test 75 positive sera from patients of different AIDS Classes the synthetic peptide-ELISA was employed.The sera reactivity with linear and cyclic peptide forms was compared. Results: 86Z positive sera reacted both with linear form of PND peptide 307-329 and its cyclic analogue,whereas only 53% reacted with peptide 301-319.The different levels of serum recognition were observed with the CD4-BR linear peptides - 36%,68% and 79% for the peptides 434-450,431-450,423-450,respectively.The cyclic peptide 423-450 was found to be the most reactive and to provide 94% recognition.It should be emphasized that the profiles of serum reactivities to PND 307-329 and CD4-BR 423-450 were highly similar(70/75 positive sera).Low titers (1:50,1:200) for the majority of sera reacting with the investigated peptides were detected.15 HIV infected individuals were studied over 1.5-3 years.The reliable differences in anti-PND titers were revealed both among the patients in this group and for each individual during the observed period. Conclusion: A most part of anti-HIV-positive sera has noticeable CD4-BR antibody reactivity however with very low titers.CD4-BR and PND responses at some extent are sinchronic.It is possible that a natural CD4-BR antibodies have low affinity and since is not effective to block high-affinity binding of HIV to cell.There does appear to be clinical correlation in some patients with CD4-BR and PND responses,that may be used as a prognosis markers for the disease development. M.A 1143 INTERACTION OF HIV-1 FUSION PEPTIDES WITH LIPID MEMBRANES I.l _ SlpushkinV.Ai.*,Sidorova M.V.&, Andreev S.M.&, Melikyan G.B.#, Chumakov V.M.*,Manukyan R.A.#,Grigoriev V.B.*,Karamov E.V.* * D.I.Ivanovsky Institute of virology,Moscow, & Institute of immunology,Moscow, # Institute of physiology,Erevan,USSR Obiective: To investigate the influence of the structure of fusion peptides on their interaction with artificial lipid membranes. Methods and Results: Two families of peptides were synthesized.The C-terminus of the peptides from the first family was located at the glycine-532 of gpl60 HIV(BRU str.) and they were assembled in stepwised manner to N-terminus of gp4l(aa 517).Hydrophobic synthetic peptides,containing 10-16 aa,formed channels in planar lipid membranes,and the longest 15-16 aa peptides were able to lyse the liposomes.Peptides of the second family begining from the arginin-538 to Val-510 have several hydrophylic aa residues. 15-22 aa peptides increased the conductivity of planar lipid bilayers and lysed the liposomes as the 15-16 aa peptides from the first family.The degree of liposome's lysis depended on peptide length and concentration. The peptides containing aa from C-terminus of gpl20(23-29 aa) were completely inactive.The effects of the second family peptides on membranes were reduced to a great extent at pH lower than 5.5. 22 aa peptide containing E-protected lysine instead of C-terminal arginine didn't show such pH dependence.This modified peptide couldn't induce liposome fusion,in contrast to parent peptide,at any pH. The conjugation of Lys-peptide with bovine serum albumine decreased its lytic activity. The circular dichroism study of this peptide revealed its alfa-helix configuration in hydrophobic medium.The electron microscopy perfomed in aqueous medium showed the filamentous structures formed by this peptide. Conclusion: The effects of fusion peptides on lipid membranes depends on their length as well as secondary structure. 4 4 (J ~a t>J

Page  128 SM.A.1144 ALLOSTERIC PROPERTIES OF HETERODIMERIC HIV-l REVERSE TRANSCRIPTASE 00 3 * De yser, Zeger; Pau els, R.; Andrils, K.1, Desmyter, J.: Engelborghs, Y.T; Janssen, P.A.J.; De Clercq, E. Rega Institute for Medical Research, K.U.Leuven, Leuven, Belgium, +Janssen Research Foundation, Beerse, Belgium, and tInstituut voor Natuurkunde, K.U.Leuven, Leuven, Belgium Objective: The reverse transcriptase (RT) of HIV-1 is present in virions and infected cells as an heterodimer (p66/p51). A new class of potent and selective HIV-1 inhibitors, the TIBO derivatives, was found to exert its antiviral activity by interacting in a unique fashion with monomeric HIV-1 RT (p66). We have now examined the kinetic properties of the heterodimeric p66/p51 and its mode of inhibition by TIBO compounds. Methods: Kinetic studies with recombinant p66/p51 were analyzed with a non-linear regression software program, and confirmed by equilibrium dialysis binding studies. Results: A positive cooperativity between the subunits of the heterodimeric HIV-1 RT with regard to the substrate (2'-deoxynucleoside 5'-triphosphate) and the template/primer [poly(C).oligo(dG)] was observed. Whereas Triton X-100 behaved as an allosteric activator of the enzymatic activity, TIBO compounds showed the characteristics of allosteric inhibitors. Discussion: Allosteric properties have been described for various DNA polymerases. We present the first evidence of positive cooperativity between the subunits of HIV-1 RT. This may be indicative of enzyme regulation, as illustrated by the finding of an allosteric activator, Triton X-100, and an allosteric inhibitor, TIBO. NOTES M.A.1145 INVESTIGATIONS INTO THE CONTENT OF NUCLEIC ACIDS TRAPPED WITHIN ANTIGEN-SUBSTRUCTURES OF HIV 1 BY THE POLYMERASE CHAIN REACTION (PCR). Dennin, Reinhard H., Beyer, A. Institut fur Medizinische Mikrobiologie, Medizinische Universitit zu Lubeck, Lubeck, Federal Republic of Germany. Objectives: This study was projected to characterize the structures which immunologically behave as 'antigens (AG) of HIV 1 not denatured by detergents' in the cell free supernatant of HIV 1-infected cell cultures as well as in serum/plasma from HIV 1-infected individuals. The special question raised here was to look for their content of HIV 1-specific nucleic acids (NA) - especially DNA - compared with the concentrations of HIV 1 specific AG which can be determined by commercial test kits. Methods: Commercially available beads coated with polyvalent anti-HIV 1 positive human sera were incubated with either cell-free supernatant of HIV 1 infected cell cultures or serum/plasma from HIV 1-infected individuals, adjusted with nuclease inhibitors, but without any detergent in order to get 'native' antigens. Lysates were produced by Proteinase K digestion protocols. PCR was performed with products either after reverse transcription (RT) or without RT directly with aliquots of the lysates. Results: Besides a distinct content of HIV 1-specific RNA we reproducibly were able to ascertain HIV 1-specific DNA-sequences in these lysates. The relative amounts of HIV 1-specific DNA-fragments did not turn out proportionally with respect to the amount of HIV 1-AG determined by the commercial HIV 1-antigen test used here. Conclusions: The concentrations of HIV 1-specific AG in sera of HIV 1-infected individuals do not seem to reflect always the same degree of quality of viral activity during infection. Additional determination of HIV 1-NA can provide better information. NOTES 4 4 A cn ^s h, h, M.A.1146 MUTAGEN-MEDIATED ACTIVATION OF HIV-1 EXPRESSION Ouinto, leana. Mallardo. M. Ruocco M R. Squitieri, B Scala. G Dipartimento di Biochimica e Biotecnologie Mediche, 11 Facolta di Medicina e Chirurgia Universitadegli Studi di Napoli, Naples Italy Objective The aim of the present study was to analyse the molecular mechanisms involved in the activation of HIV-1 expression by mutagens in infected cells We addressed the following questions 1) Is the nature of DNA lesion relevant for triggering the activation of the viral RNA transcription 2) Which ci-acting control elements are required for mutagen-activated transcription in the HIV-1 promoter region (LTR)? Mfethods Human B-lymphocytes or monocytes were transfected with the plasmid pwtcat which carries the LTR of HIV-1 inserted upstream to the cat gene After 24 hours, the transfected cells were treated with methylmethane sulfonate (MMS, 0.25 - 1 mM) or mitomycin C (Mit C, 2 - 20 iM). which are known to produce monoadductions or crosslinks to DNA. respectively The CAT activity was tested 24 hours after mutagen treatment. The same experiments were performed by transfection of the mutant plamids pNFAcat (deleted of the NF-kB binding sites), or pSpAcat (deleted of the Spl sites) Results Two- to three-fold stimulation in CAT activity was observed in B-lymphocytes and monocytes which had been transfected with pwtcat and treated with MMS or Mit C. The effect was related to the dose of mutagens, being maximal at doses leaving about 60% surviving fraction, and was not observed when the cells were transfected with the mutant plasmids pNFAcat or pSpAcat Conclusions Our results indicate that the type of DNA lesion is not relevant for the HIV-1 activation which was produced equally by MMS and Mit C Moreover. the viral activation required the presence of NF-tB and the Spl binding sites, thus suggesting the requirement of both regulatory proteins NF-IBlike and Spl-like This work was supported by Ministero della Sanita-ISS. Progetto AIDS 1990 Rome. Italy. M.A.1147 CYTOGENETIC ANALYSIS OF HUMAN CELLS INFECTED WITH HUMAN IMMUNODEFICIENCY VIRUS IN VITRO Kushch. AlliGluchova. L. A.; Shumay E. P. Mamaeva,S. E. * Nossik, 0. N.; kalni na, L.B; Grabovskaya, I. W.; ITuq zov, Sh. M.; Asjo. B. ** The D.I.Ivanovsky Institute of Vfroloy, Moscow USSR *Institute of Cytology, Leni ngrad, USSR; **Karol inska Insti ute, Stockhol m, Sweden Objective:We studied the effect of HIV infection on the karyotype of two highly sensitive to HIV lymphoblastoid T cell lines. Methods: The cell lines Jurkat-tat and MT-4 have been infected with BRU strain of HIV-l. Metaphase chromosomes were subjected to G-, C-banding and Ag-stained. HIV specific proteins was monitored by cell-ELISA using MAb to pl7, p24, gp41. esl s:Changes in Jurkat-tat cells monitored over 17 days postinfection were biphasic. The number of polyploid cells increased from 8 % before infection to 19 % by day 5 and to 23% by day 15 postinfection. Their ploidy reached to 8-10 n. The number of chromosomal aberrations increased 3- fold by day 5 postinfection. Nonrandom chromosomal breaks, tri-, tetraradials and icentric chromosomes have been detected. In the polyploid cells chromosomes were condensed or aopeared as dots. In contrast, in MT-4 cells there was no significant alteration in the number or structure of chromosomes during 11 days of infection up to cell death. EIA has shown that the accumulation of viral proteins was maximal on day 5 and showed a second peak in Jurkat-tat cells by day 15. Discussions and Conclusions:These data lead to the hypothesis that extensive fusion and polyploidization of Jurkat-tat cells induced by HIV lead to early death of a part of infected cells, followed by a second wave of the lytic infection in surviving cells. The same strain of MIV does not lead to massive fusion or chromosomaT damage in MT-4 cells and their death takes place much earlier. We conclude that mechanisms of HIV effect on these two cel strains are different and reouire further analysis.

Page  129 M.A.1148 VIRUS POPULATION SEQUENCE ANALYSIS OF A SUCCESSFUL SIV VACCINE Slade Andrew, Almond N, Szotyori Z, Jenkins A, Jones S and Kitchin P. AIDS Collaborating Centre, National Institute for Biological Standards and Control, Potters Bar, Herts, UK. Objective: To characterise the sequence repertoire of the SIVmac251 (32H) isolate. This isolate was used to demonstrate protection of cynomolgus macaques against SIV by a fixed infected-cell vaccine (Stott EJ et al, 1990, Lancet 336; 1538-41). Methods: The polymerase chain reaction was used to amplify SIV genes from the virus challenge stock. The products were cloned into M13 to produce a panel of clones from which the sequence heterogeneity of the virus pool could be determined. Results: The majority of nucleotide changes in gag were substitutions. The only insertional mutation was a 12 base pair direct repeat in the carboxyl terminal region of p15 (identical to SIVmac239). The V3 loop equivalent of SIV env gpl20, in contrast to that in HIV-1, was highly conserved. Env gp41 varied little; 1 out of 20 clones contained a premature stop codon. In the rev clones, 4 different splice sites were identified. 25% of the nel clones contained premature stop codons. Discussion: This sequence analysis of the virus pool used to produce a successful SIV vaccine will help the design and immunological analysis of future vaccines. This is highlighted by the observation that the env V3 loop equivalent, regarded as the major antigenic determinant in HIV, is well conserved in slV, NOTES M.A.1149 HIV-1 REGULATORY AND ACCESSORY GENE mRNAs. Deacon. Nicholas., Smith, J.Mc. Macfarlane Burnet Centre for Medical Research, Melbourne, Victoria, Australia. Objectives: As well as the three structural polyproteins HIV-1 expresses at least 8 regulatory and accessory proteins encoded by singly and multiply spliced mRNAs. We have set out to characterise HIV-1 spliced mRNAs, the splice donor and acceptor sites used and the temporal expression of the different species. Method: Cytoplasmic RNA extracted from MT-2 or CEM cells infected with local isolate HIV-1 228200 was converted to cDNA and amplified by polymerase chain reaction (PCR) using primers sited upstream (Primer-5'a) of the major splice donor and downstream of either the third tat exon (Primer-3'a) or the nef coding region (Primer-3'b). Amplified products were probed with splice donor/acceptor combination probes, cloned and sequenced. Results: At the time of peak reverse transcriptase and p24 expression more than 20 species of amplified product have been identified. Of a predicted 42 possible splice junctions we have tested 20 and confirmed the use of 5 donor and 12 acceptor sites. Sequences of cloned amplified products have revealed that rev, nef and tat proteins are each encoded by multiple mRNA species. Conclusion: HIV-1 expresses multiple versions of its regulatory and accessory protein mRNAs differing in the combination of splice junctions used. This degeneracy may reflect temporal differences in expression of each mRNA or result from the need to accommodate the high mutation rate of HIV-1. NOTES I --- o1 00 Ic M.A.1150 CLONING OF CONSERVED AND UNIQUE SEQUENCES OF THE HIV-1 GENOME FOR USE AS NUCLEIC ACID PROBES. Materazzo, M.; McQueen, P. W.; Delanev. Stephen Francis. Department of Biotechnology, University of New South Wales, Kensington, NSW, Australia To develop a hybridisation test to detect the HIV-1 genome and to distinguish between active and latent forms of the virus, conserved and unique sequences of the ga and i.Lregions have been cloned into the transcription vector, pGEM-3Z. pARV/7A, a recombinant plasmid containing the HIV-1 DNA genome cloned into pUC19, was digested with PMvu Iand ligated to the Sma I site of pGEM-3Z. A recombinant plasmid, pOK700, containing a 2.154 bp insert, was selected. The insert spans most of the gag region and the 5' half of Dl (co-ordinates 698-2852 on the HEV-1 sequence of Sanchez-Pescador E (1985) Science, 227, 484-492). pOK700, digested with vu II and San 3A, was ligated between the Sma I and ajn HI sites of pGEM-3Z. A recombinantplasmid, pOK701, with a 952 bp insert spanning the entire p24 coding sequence and the N-teminal half ofpl3 (co-ordinates 698-1650), was selected. pOK701 was digested with Ifnd II, selfligated and a recombinant plasmid, pOK702, containing a 567 bp insert spanning the N-termina half of p24 (co-ordinates 698-1265), was selected. pOK700 was also digested with WII1f I, the ends made bluntended with DNA polymerase (Klenow fragment) and ligated to the S I site of pGEM-3Z. Two recombinant plasmids were recovered pOK704 containing a single inf I insert of 228 bp spanning the central region of Zl (co-ordinates 2624-2852); and pOK705, containing two contiguous HinfI fragments (777 bp, co-ordinates 2075-2852) spanning the 5' half of pM1. In a similar way sequences of the =iv region and the 3' end of the p~Lregion are being subcloned into pGEM-3Z. Once all of the clones have been confirmed by DNA sequencing they will be used to produce RNA probes for the detection of HIV-1. M.A.1151 OJ ',0

Page  130 UO M.A.1152 REGULATIOM OF HI GENE EPRESSION BY CLASS I HLA ATIGEMS: TRANSCRPTION OF kB-DRIVEI GENES IS IMDUCED BY TRtIGERI 6 CLASS I MOLECULES ON T CELLS. Turco. laria Caterina; Lamberti, A.*; Bisogni, R.; De Lorenzo, F.*; Petrella, A.; Venuta, S.*. Dipartimento di Biochiica e Biotecnologie Mediche, II Facolta di hedicina, Napoli, Italy; * Dipartimento di nedicina Speriaentale e Clinica, Facolta di Hedicina di Reggio Calabria (Catanzaro), Italy. Objective: Our research goal is to demonstrate that the triggering of Class I HLA molecules on T cell surface regulates cellular expression of genes controlled by kB elements at their 5'. These include HIV genes. Iethods: Humen primary T lymphocytes were transfected by electroporation with plasmid constructs containing the reporter gene CAT under the control of the herpes virus TK promoter and a kB sequence. Folloving a further 24 hour incubation, the cells were harvested and CAT activity was measured. Results: Unstimulated T lymphocytes did not express significative amounts of CAT activity, while, as expected, a strong signal was detectable in PHAstimulated cells. T cells incubated with anti Class I HLA mAbs, although not proliferating, expressed significative amounts of CAT activity. Conclusions: Our experiments assign for the first time to Class I HLA molecule the ability to regulate the expression of kB-driven genes. These results are particularly intriguing in view of the possibility the HIrV peptides can bind Class I molecule (Choppin, 1990; Frelinger, 1990, Siccardi. personal communication). Therefore HIV aight regulate its own expression by interacting with Class I HLA antigens on infected T lymphocytes. York supported by funds from the Italian tministero della Sanit& - Istituto Superiore di Sanita - AIDS Project 1990 - Romea - Italy. M A 1153 CHARACTERIATION OF LARGE GLYCOSYLATED gag-ENCODED PROTEINS EXPRESSED. ON THE SURFACE OP HIV-INFECTED CELLS WHICH REACT WITH MONOCLONAL ANTIBODIES AGAINST THE HIV MATRIX PROTEIN, p17t' A. Pnter, Wj. Honnon, F. Shang, K Revesz, and R. Her; Laboratory of Retroviral Biology; Public Health Research Institute. 455 First Avenue, New York, New York 10016 Objective: To molecularly characterize cll surface molecules identified by monoclonal antibodies against the matrix protein of HIV-1, p17i. Methods: Monoclonal antibodies against the matrix protein of HIV-I, p17A, were isolated from rats which had been immunized with solubilized HIV-I lysate. The affinity-purified antiiodies were shown to react specifically with an antigen exposed on the surface of HV-infected cells by several different assays, including fluorescence microcopy. FACS analysis, and binding of a radioidinated second antibody. The molecules recognized by these antibodies were characterized by radioimmunoprecipitation assays of infected cells. Results: Radioimmunoprecipitation assays of lysates and supernatant media of infected cells labeled with 'H-glueosamine showed the presence of a complex of large glycoproteins which reacted with these antibodies; these components were also recognized to a lesser extent by other antibodies against p17 and p24. This complex was resolved by SDS-PAGE as a major protein doublet with molecular weights of approximately 150 kDa, and a minor protein doublet of approximately 90 kDa. Cell surface radioiodinaton experiments codnfi ed the location of these mocles on the exterior of the cell and demonstrated that they were he major p7-ontaining molecules xpoed on the cell surface. Discussion & Conclusions: These results suget that the HIV gag gne encodes one or more large membrane-associated glycoproteins in addition to the normal prcursor of the inernal cre proteins. These molecules are efficiently expressed on the cell surface, and are secreted from infected cells. These cell surface components may play a role in the generation of antibodies against p17" which is a characteristic of early stages of HIV infection, and may act as natural targets for the immune system and as potential targets for immunotderapy of HIV infection. NOTES NOTES - I - --- M.A.1154 EXPRESSION AND PURIFICATION OF SOLUBLE HIV-l TAT PROTEIN FROM ECOLI D'SouIza Eric Damian Allen and Cousens, Diane, J. The Wellcome Foundation Ltd., Dept. of Molecular Sciences, South Eden Park Road, Beckenham, Kent, England. ObjectivesStructural studies on HIV Tat have in large been impaired by the inability to obtain sufficient amounts of native purified protein. All procedures reported to date for the purification of HIV Tat have relied on the use of strong denaturants and subsequent refolding of the Tat molecule. We report here the first demonstration of expression and purification of soluble Tat from E.oli under native, non-denaturing conditions. Methods: Both 72aa and 86aa Tat constructs were made under the control of an inducible promoter. Each construct was engineered to contain between one and three additional amino acids at the C-terminus which comprise an antibody epitope to facilitate purification. Expression products were purified using immunoaffinity and anionic exchange columns. The biological activity of Tat was assayed by transactivation of a HIV-LTR cell line and by an in vitro RNA binding assay. Results and Discussion: Expressed Tat protein was soluble and present at 5% total cell protein as verified by polyacrylamide gel electrophoresis (PAGE). Immuno-affinity chromatography of the soluble Tat resulted in a 90% homogenous product as detected by PAGE which also immunoreacted with a Tat-specific monoelonal antibody. Anionic exchange chromatography resulted in further purification to >90% homogeneity. The purified protein showed biological activity in a cellular transactivation assay comparable to previously reported data and in vitro gel shifts with TAR RNA were significantly better than previously seen with synthetic Tat peptide. We are currently evaluating the suitability of this material for structural studies. M.A.1155 MECHANISM OF ACTION OF THE HIV-1 Tat PROTEIN:INTERACTION BETWEEN Tat, CELLULAR FACTORS AND VIRAL DNA AND RNA TARGETS. Kuan-Teh Jeang, Ben Berkhout, Anne Gatignol. Laboratory of Molecular Microbiology, NIAID, NIH, Bethesda, MD, USA. Objective: To study the mechanism of action of HIV-1 Tat. Methods: Standard molecular techniques were used. Results: We will show that upstream promoter elements and the HIV-1 TATAA sequence play an important role in Tat trans-activation. We will present evidence that Tat trans-activation occurs during a kinetically brief time period during the initial moments of LTR-directed transcription. We will describe the cloning and characterization of one cellular TAR-RNA binding protein. Discussion: Tat trans-activation of the HIV-1 LTR involves a complex interplay of viral and cellular proteins and both viral DNA and RNA targets. a Is a

Page  131 MA 11 RACERIZATIONOFV MUTANTS OF HIV-1 AND HIV-2 M.A.116 Peden, Kth Fan, Liju Laboratory of Molecular Microbiology, NIAID, NIH, Bethesda, MD, USA. Objective: To determine the phenotype of mutants of HIV-1 and HIV-2 with defects in the vif gene in order to investigate the role of the Vif protein in the life cycle of these viruses. Methods: Insertion and deletion mutations were introduced into the vif genes of infectious molecular clones of HIV-1 and HIV-2; in some cases, mutations were also made in the vpr gene. Infectious molecular clones derived from the BRU, MAL and ELI isolates of HIV-1 and three clones from the ROD isolate of HIV-2 were used. The biological activitity of Vif mutant viruses was assessed following transfection of HeLa cells to obtain viral stocks and subsequent infection of several CD4-positive cell lines and peripheral blood mononuclear cells (PBMC). Viral infectivity was monitored for the appearance of syncytia and by assaying reverse transcriptase activity in the medium of infected cultures. Results: The infectivity of Vif mutants was found to vary depending on the cell types used. For example, two Vif mutants and a Vif.Vpr double mutant of the BRU isolate were able to replicate in the CEM and SupT1 lymphoid cell lines and also the U937 promonocytic cell line, although the time of appearance of the mutants was delayed and less virus was produced compared to wild type. However, these mutants displayed a severe defect on the H9 cell line and on PBMC. So far it has not been possible to infect H9 cells even by co-culture. Similar cell type and virus specific effects were seen with the HIV-2 Vif mutants. Conclusions: These results demonstrate that the phenotype observed for viruses defective for the production of Vif depends upon the genetic background of the virus and the particular cell type used for the infection. The presence of an additional mutation in the vpr gene did not contribute significantly to the Vif defect. Previous work from others had shown that Vif mutants of HIV-1 were defective for growth on cell lines by cell-free infection but could be transmitted by co-culture of transfected cells with CD4-positive cells. However, our results indicate that the phenotype of Vif mutants depends on the system used, and it remains unclear at which step in the life cycle Vif is functioning. M.A.1157 MOLECULAR CLONING AND COMPLETE SEQUENCE ANALYSIS OF A HIGHLY M.A.11 DIVERGENT AFRICAN HIV ISOLATE. Vanden Haesevelde Marleen, Decourt J.L, Heyndricla L., Vanderborght B., Saman E., DeLeys R., van der Groen G., and Van Heuverswyn H. Innogenetics N.V. - Antwerpen - Belgium. * Institute of Tropical Medicine - Antwerpen - Belgium. Objective: The molecular analysis of the proviral DNA from a divergent HIV isolate (Ant70). The sequence data obtained allow the detailed study of the differences with known HIV strains and the determination of the positionof this new variant within the HIV family. Methods: DNA from MT4 cells persistently infected with the Ant70 virus was used to construct a genomic library in 1 GEM11. Ant70positiveplaqueswere identified by hybridizationwitha cDNA fragment obtainedby reverse transcription of viral RNA. About 2000 bp viral information which was missing in the largest proviral clone obtained was amplified via PCR and cloned. Sequencingwas performed on plasmid subclones using the dideoxy chain termination technique on both strands. Results: The isolation and partial biochemical characterization of this highly divergent HIV isolate Ant70 has been described (DeLeys et al. 1990, J Virol 64:1207-16). Sequence analysis shows that the Ant70 genome contains all open reading frames found in other HIV-1 isolates as far as the localizationin the genome is concerned. The over-all sequence homology with prototype HIV strain HXB2 is 70%, but a similar homology was also found with the more divergent HIVm,, strain and with the recently described HIVcpz. The homology at the amino acid level varies between 47% and 79%, depending on the gene considered. Homology analysis with HIV-2 sequences always yielded a significantly lower score. Phylogenetic tree analysis based on the alignment of gag, pol or env protein sequences localizes the Ant70isolate in a separate branch within the HIV-1 family. Coaclusions:The present analysisof the Ant70 isolate shows that the variabilitywithin the HIV-1 family might be greater than assumed until now and may be similar to the more divergent HIV-2 subgroup. 0 hCo NOTES NOTES 'HUMAN HE'ATUBLASTOMA (HEPGZ) CELLS CONTAIN A TAR-BINDING NUCLEAR FACTOR THAT M.A.1158 TRANSACTIVATES HIV-1 LTR EXPRESSION Pizzella, Teresa; Banerjee, R.* Giusti, G.; Acs, G0 Clinica Malattie Infettive, I Fac. Medicina, Universita di Napoli, Italy; *Department of Biochemistry, Mount Sinai School of Medicine, New York, USA The expression of HIV-1 is regulated by both viral and cellular proteins as well as the trans-activator protein tat, which binds to a cis-acting element TAR, located in the long terminal repeats (LTR). Our group has recently reported that HepG2 cells can support transient HIV-1 LTR-directed CAT gene expression in the absence of cotransfected tat gene constructs (Hepatology 10,1008,1989). In the present study, using band shift assays with probes derived from digestion of the 100 nt TAR (-19 to +80, referred to transcription start), we have identified the specific region of HIV TAR sequence involved in the binding to HepG2 nuclear proteins. The specific binding site was confirmed by copper footprinting as 5'-TCTGGTT-3', spanning between nt+7 and nt+13. Gel retardation assays with synthetic oligonucleotides carrying different mutations in the TAR region and competition.experiments using these oligonucleotides confirmed these data. The influence ol this region-binding HepG2 factor on HIV-1 LTR expression was tested by in vivo CAT competition cotransfection experiments, using -65HIVLTRCAT as reporter plasmid, and a series of mutated plasmids as competitors. The data showed that competition occurs with the plasmid containing the unmodified bindingsequence, but not with the plasmids containing mutations in the binding region. These data demonstrate that HepG2 cells contain a nuclear factor that affects HIV-1 gene expression by interaction with a specific TAR sequence. M.A.1159 HIV-1 LTR promoter is functional in yeast Schizosaccharomyces pombe Dhar, Ravi; Toyama, Reiko Laboratory of Molecular Virology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892. Objective: To study HIV-LTR promoter activity and trans-activation by HIV-1 TAT protein in yeast Schizosaccharomyces pombe. Methods: Yeast replicating vectors containing HIV-1 LTR-CAT and different deletions of the HIV-LTR promoter were transfected into S. pombe cells. In separate experiments, SV40 promoter transcribing the TAT gene was used in a co-transfection. Results: Our results show HIV-LTR promoter is functional in yeast and RNA synthesis initiates from two sites in equimolar amounts. One of the 5'-ends of the RNA corresponds to the 5'-end of HIV-mRNA observed in mammalian cells. Deletion analysis of the promoter suggest NF-kB binding sequences are required for promoter activity. However, TAT gene product failed to trans-activate transcription. Discussion and Conclusions: The promoter of HIV-LTR is functional in yeast S. pombe and tat protein fails to trans-activate transcription. (,)

Page  132 L[ M.A.1160 EXPRESSION OF HOMEOGENES IN HIV INFECTED T CELLS FRat DmSiE*.. Sartni. M.*, Mgliaci. E.', Bach S.*, Boncini E.',Colbi V.* *Dipa~iUwt di Biooia,II Uniaei3 di Romar Verga RoItely 'IsMito nl okna di Gemtia e Biefiica, CNR, Napsl, Ily. Te ami ofdt fpojemt is to nvestiga tS ievolvema of hemn h a ogenb in T bocyt difFea amd vbeter HIV infectom d mondify tbir exprssn camsing furibwal a~ as. Homeoti gems wVe odgia ly defled as genes controllng embryonic developmint in Drooplhk HomoIgo- seqances wee isnkld in vetebats (38 in man in 4 diffemnt clustes) and mha been shovn to be expressed with ~ poral a positonlm speaikitydming embryogeisis, in different~atid g mmoblstma and!mtcaxcimN am cell lime induced by atnoic acid. Six T-cel lines and tvo clrocaly HIV-infected ce lies ( H9, U937) wre amtymd for expiexbro of homegenes by Nortema blot amlyis uin o a panel of 38 homeogem plobes.Pbheotype chlsarate Rlian by flov cyao4 tUy of te CD3, CD4, CD6, ITCR dLP amd f6 exp resin allowved s I defie t v r staes of diffesentiation iT-cenMln. The data obteaid from the leukemic cel lines, in partimu from CEM, H9 and tie chronicaly HIV-infected H9 sge t a spacifl nd higeiby ed ctiattfonp ntf oth oer annT-cellines. T CEM H9 an d HIVinfeced H9 sho a pro11s1ion inte t tatm of lnomgogems. We suptMt thUse differem s could be repreIanwei for V osu4 suNes of diffentlataon of te cell lints o far anmled. Fmtu r stdies of lealemic T-ce lim, stil in mprgres, seem to cofam tis hypotesies.'lT aed expressio pfaten ob8seivr in chronfcaly Infectd H9 couMl be combMerd as HIV l flY ace on celutlar geme expresin or coul replp a sea ion of previouslyalterd can ws rsisant tl e vieytlceffo e Aalysis of th exprmssion of hao eogems in ntmml moninfecl and nfecal T-cesb mlad yrce shall one of tese two possibdlities. NOTES M.A.1161 BINDING OF NUCLEAR FACTORS TO THE HIV-1 GENOME: BIOCHEMICAL CHARACTERIZATION AND INHIBITION BY AROMATIC POLYAMIDINES Ferlotto, Giordana*; Nastruzzi, Claudio*; Volinia, Stefano**; Giacomini, Patrizio***; Gambari, Roberto* *Biochemistry Institute and **Department of Evolutionary Biology, University of Ferrara, ***Immunology Laboratory, Regina Elena Institute, Rome. DNA elements of HIV-1 putatively able to bind nuclear proteins have been identified by computer-assisted analyses using PC programs elsewhere described (1). By means of gel retardation and CAT activity assays we have identified a nuclear protein (X-B3) that specifically binds a HIV-1 tat-IVS region containing a DNA sequence 90% homologous to a box of class II MHC genes required for constitutive and IFN- induced gene expression. Nuclear extracts have been prepared from B-lymphoid, T-lymphoid and IFN-r induced promyelocytic and melanoma cell lines (2). We suggest that binding of nuclear proteins to the HIV-1 genome is not restricted to LTR and could be required for transcriptional activation of the integrated provirus as well as for stability of the unintegrated, reverse transcribed HIV-1 DNA forms. We have studied the effects of DNA-binding drugs on the capability of transcriptional factors (NF-kB, Spl, X-B3) to bind HIV-1 genomic sequences. Among the DNA binding drugs, the aromatic polyamidine tetra-p-amidinophenoxy neo-pentane appears to be of Interest, since it strongly inhibits the electrophoretic mobility of HIV-1 genomic sequences and blocks the binding of nuclear transacting factors to specific HIV-i target regions. These compounds are likely to be active inhibitors of the HIV-1 life cycle [Work supported by Istituto Superiore di SanitA (AIDS-1990)]. 1.Volinia et al.,J.Mol.Biol.203,385,1988;2.Barbieri et al.,FEBS Letters,268,51,1990. NOTES M.A.1162 CLONING OF AN HIV-I NEUTRALIZING V3 SPECIFIC MONOCLONAL ANTIBODY AND EXPRESSION AS A MOUSE-HUMAN CHIMAERIC ANTIGEN BINDING FRAGMENT AND ANTIBODY Marks, J.D.*, Wahren, B.**, Gilljam, G., Hinkula, J.**, Winter, G.* * MRC Aids Directed Programme, Laboratory of Molecular Biology, Cambridge, UK, ** Dept of Virology, NatI Bact. Lab. and Karolinska Institute, Stockholm, Sweden Objective: To clone and sequence the variable domains of an HIV-1 neutralizing V3 specific murine monoclonal antibody (Mab) and use them to construct a mouse-human chimaeric antigen binding fragment (Fab) and mouse-human chimaeric IgG1 isotype antibody. Methods: The most neutralization potent (Mab F58) of a battery of 6 murine Mab against the HIV-1 V3 loop region was selected by peptide serology and neutralization assays. The heavy and light chain variable domains (VH and VK) were amplified using the polymerase chain reaction, cloned and sequenced. A chimaeric Fab construct was created containing the murine VH and VK and the human CH1 and Ck domains and expressed in E. coli. A chimaeric IgG1 antibody was created containing the murine VH and VK and the human IgG1 and k constant domains and expressed in a myeloma cell line. Results: Murine Mab F58 neutralized 9 of 11 primary HIV-1 isolates. The critical amino acids recognized were I--GPGRA in the V3 loop sequence SIRIQRGPGRAFV. Sequence analysis of the F58 VH and VK indicated the presence of a single VH chain (murine family IIb) and a single VK chain (murine family III). The chimaeric Fab and IgG1 antibody bind to HIV-1 gpl20, gpl60 and peptides containing the above mentioned amino acids. Neutralization and antigen dependent cellular cytotoxicity (ADCC) studies are pending. Conclusion: A mouse-human chimaeric Fab and IgG 1 isotype antibody with similar reactivity as the parent murine Mab were constructed. The chimaeras may prove useful for passive immunization where they should be less immunogenic than the murine F58 Mab. In addition, the Fab may be able to penetrate tissues better and the chimaeric IgG I antibody may be able to clear infected cells via ADCC. M.A.1163 C OHRODSOMAL INTEXRATION SITES OF EDOGENCOUS RETROVRAL SEQUECES Taruscio, Domenica*; Manuelidis, L. Yale Medical School New Haven,CT,USA We analyzed the chromosomal integration sites of a human endogenous retrovirus (HERV). HERV is present in 50-100 copies in the genome and is related to MLV. Using high resolution in-situ hybridization, HERV integrationsites were localized with respect to chrcmosomal bands containing high concentrations of Alu repeats. Alu PCR products from unfractionated human DNA labeled with digoxigenin, and biotinylated HERV sequences wer< used as probes, and simultaneously detected with two different fluorochromes. HERV sequences were confined to 29 chromosomal sites, and- 90% of these were unambiguously located inGiemsa-light and Alu rich (presumably early replicating) chromosomal domains. Moreover,,70% of the resolved HERV integration sites corresponded to known breakpoint and/or recombination hotspots. The quantitative data suggested that only 13 or fewer sites could potentially contain full length copies of 8.8kb. Remarkably, chramosame 2 contained- 20% of the total positive signals, where each signal represented> 2kb of HERV sequence as determined by single copy control experiments. The highly selective integration of both complete and truncated HERV sequences into genetically active chromosome domains could indicate a preference for more structurally euchromatic sites. Retroviral sequences may initially propagate along a single chromosome, and later juap to the other active chromosome domains where they contribute to chromosome fragility and recombined with evolutionary pressures may lead to a decidedly different pattern of chromosomal organimation. High resolution mapping of rodent endogenous retroviral sequences present at 800-1,000 copies per gencme, combined with LINE hybridiation and BrdU replication studies supported this concept. In at least one species, the Syrian hamster, endogenous retroviral seguences were entirely confined to Giemsa-dark domains that were late replicating. (Supported by NIH grant CA 15044 and *Istituto Superiore di SanitA, Rcmna, ITALY) a ^I s ~I

Page  133 M.A 11 i64 M.A.1164 ANTIBODIES TO HIV p27 NEF PROTEIN IN 63 ZAIRIAN HIV SEROCONVERTERS pdidi, Bazepeyo*; Behets, F', **; Kambembo, L*; Tshebuye, M*; Kayigamba, K*; Brown, C*, ***; Quinn, T***. *Projet SIDA, Kinshasa, Zaire; **Institute of Tropical Medicine, Antwerp, Belgium; ***National Institutes of Allergy and Infectious Diseases, Bethesda, MD. Objective: To determine if antibodies to p27 nef can be detected before and at the time of seroconversion in a cohort of Zairian prostitutes. Methods: A cohort of 575 initially HIV seronegative prostitutes was enrolled in a counseling and STD treatment program. Blood samples were obtained from each patient every three months. Seroconversion Was monitored by ELISA (Vironostika, Organon) and Western blot (DuPont) and was documented by reactivity to both gag (p24) and oev (gp41, gp120/160) reactivity. The Western blot patterns of all available sequential sera of seroconverters were screened for reactivity in the p27 region. Results: Sera were available from 63 prostitutes who had seroconverted. Reactivity to p27 was observed In 11 (17.5%) of 63 sera at the time seroconversion was detected; 15 (23.8%) additional sera showed weak reactivity. Only 3 (4.3%) of the 69 samples collected up to 1 year before seroconversion demonstrated a weak reactivity in the p27 region. Conclusions: Antibody reactivity for nef protein at 3 month intervals up to 1 year before seroconversion was a rare event in this cohort of sexually transmitted HIV seroconverters. NOTES M.A.1165 DELETION MAPPING OF FUNCTIONAL DOMAINS OF HIV-1 GAG POLYPROTEIN EXPRESSED IN BACULOVIRUS-INFECTED CELLS. Boulanger, Pierre. Royer, M., Gay, B., and Hong, S,5. Laboratoire de Virologie & Pathogenese Moleculaires, Facult6 de M6declne, 34060 Montpellier, France. Objective & Methods: The functional domains of HIV- gag implicated in (i) Its self-assembly, and (Ii) Its cellular localization, were determined using 8 recombinant baculoviruses (AcNPV) harboring various forms of full-length or truncated gag gene expressed under the control of the polyhedrin promoter. Results: (i) Most of the Information for gag-particle assembly seemed to be confined to the p9-NC domain of gag, but the N-myristylated terminus could compensate a p 5-NC deletion and induce the budding of gag-particles. (ii) The two consensus karyophilic signals present in the pl7-MA domain were inefficient for the nuclear transport of non-myristylated, p6-deleted forms of gag. In the presence of p6, or with the polyhedrin karyophilic signal added at its N-terminus, gag relocated in the nucleus. Conclusions: The morphopoietic Information for gag-particle assembly is not entirely located in the p9-NC domain, but Is also present at its N-terminus. p6 plays an important conformational role in the gag polyprotein molecule. NOTES I \I oo Ir M.A.1166 EXPRESSION AND ACTIVITY OF HIV-1 PROTEASE IN RECOMBINANT BACULOVIRUS-INFECTED CELLS. Royer, Moniaue and Boulanger, Pierre. Laboratoire de Virologie & Pathogenese Moleculaires, Faculte de Medecine, 34060 Montpellier, France. Qlectiee. To determine the conditions and parameters for (1) high level expression, (il) efficient self-processing, and (il) proteolytic activity of recombinant HIV protease (PR) expressed in eucaryotic non-mammalian cells. Methods. Six recombinant baculoviruses (AcNPV-PR) expressing various forms of polyhedrln (PH)-fused or non-fused protease gene under the control of the PH promoter were assayed for PR self-processing and proteolytic activity on recombinant gag co-expressed in baculovirus-infected insect cells. Results. (i) High level expression AcNPV vectors failed to produce detectable amounts of full-length, active PR, and only inactive, C-terminally deleted PRs were produced in large yields. (il) Cleavage at Its N-terminal maturation site did not occur when a run of 8-58 amino acids from the polyhedrin sequence was fused witht its N-terminus. (iii) Full-length PR produced by a low level expression vector was active in vivo on its natural substrate gag. Conclusion. AcNPV-PR expression is mainly regulated at the translation level. 117 IOLOGICAL FUNCTION OF HIV-1 NEF M.A.117 Tsunetsugu-Yokota, Yasuko*; Takebe, Yutaka**; Saito, Takashi***; Takemori, Toshitada* *Departint of Cellular Immnology and **Central Virus Diagnostic Laboratory, AIDS Research Center, National Institute of Health, Tokyo, Japan, ***School of Medicine, Univeristy of Chiba, Chiba, Japan. ObjectivL; to study the biological function of the HIV-nef in HIV-infected cells. bethona 1) The vector containing HIV-nef under the constitutive SRB promoter was constructed (SRaef). SRa-nef carrying frm=~-shift termination was served as a control (SRa- nef). These vectors were transfected to Hela-CD4 or A3.01 cells similtaneously with a provirus clone pNL4-3, and virus replication was monitored by the reverse transcriptase activity (RK) and Northern blotting of HIVRNA. 2) These vectors were ligated with a hygrcnycin resistant gene and several nef-expressing clones from Jurkat and THP-1 were obtained. Function of these cells was evaluated either by ELISA or by Northern blotting. Reasuilt 1) Transient HIV-nef expression by SBR-nef in Hela-CD4 cells down-regulated a replication of HIV about 1/10 in rB activity. Replication of HIV in A3.01 was also suppressed by SBa-nef. This suppression of RT activity was corresponded to the suppressed level of the transcription of HIV. 2) After infection of HIVBn to constitutively nef-expressing Jurkat and THP-1 cells, kinetics of HIV replication was delayed and RT activity was about 5-fold less than that of the parental cells. Nef+Jurkat cells stimulated with TPA/PHA appeared to produce less IL-2 than the parental cells, but HIV infection did not induce further change of 11-2 productivity. No down-modulation of CD4 was observed in nef-THP-1 cells and these cells showed no specific change of the expression of cytokine genes (TNF-aX, sTGF-- and I1-6). Discussion and Conclusions: HIV-nef down-regulated the replication of HIV by co-transfection as well as constitutive expression presumably at the initial transcriptional level of HIV. IL-2 production in Jurkat cells appeared to be suppressed in nef transformants. In THP-1 cells nefspecific functional change such as CD4 expression and cytokine production was not observed. 'b G\ cn 0\ ss^ UQ 1-^

Page  134 M.A.1168 EFFICIENT PRODUCTION OF A SOLUBLE HIV-1 REVERSE TRANSCRIPTASE IN INSECT CELLS BY A RECOMBINANT BACULOVIRUS - PURIFICATION AND CHARACTERIZATION OF THE RECOMBINANT ENZYME Moosmayer, Dieter; Jung, Roswltha; Petry, Harald; Jentsch, Klaus-Dieter; LOke, Wodfgang; Hunsmann, Gerhard German Primate Center Gottingen, Germany Objective:The reverse transcriptase (RT) of HIV Is essential for virus replication and thus a potential target for antiviral chemotherapy. Large quantities of HIV RT are necessary to develop and evaluate RT inhibitors. Methods: The large HIV-1 RT subunit encoding region was cloned into the transfer vector p36-C (the complete construct was provided by Dr. Darby). The construct and wild-type baculovlrus DNA was cotransfected into Sf9 insect cells. Recombinant baculovlrus expressing RT was purified by end point dilution and antigen dot blotting. After scaling up, the recombinant RT was isolated by a combination of ionexchange and affinity chromatography. The enzyme was characterized and inhibition studies were performed. Results: As an alternative to prokaryotic systems we expressed the large subunit of HIV-1 RT in insect cells with a recombinant baculovlrus. With this approach we produce about 2-5 mg/10 cells of a soluble recombinant RT. This enzmye was purified to almost homogeneity by three chromatographic steps. The purified RT exhibits the authentic enzymatic activities. The results obtained by inhibition studies with Suramin derivatives were comparable to those reported for the viral enzyme. Discussion: High level expression of HIV-1 RT in prokaryots results frequently In a insoluble recombinant protein forming inclusion bodies. Thus purification and characterization requires denaturation and refolding. We have established an efficient eukaryotic expression system leading to soluble HIV-1 RT which is convintently purlfyed and has the characteristics of the viral RT. This recombinant RT is available in amounts which allow large scale screening of potential RT-inhibitors and further enzymatic characterization. NOTES M.A.1169 MOLECULAR MODELING OF GP120: COMPARISON OF PREDICTED MOLECULAR TOPOGRAPHY WITH FUNCTIONAL PROPERTIES OF THE MOLECULE. William M. Mitchell and Jerome L. Gabriel. Dept. of Pathology, Vanderbilt University School of Medicine, Nashville, TN 37232 and Dept. of Biochemistry, Temple University Medical School, Philadelphia, PA 19140 Obiective: The molecular topography of gp120 is essential to understand the binding of HIV to the CD4 receptor, the association with gp41 and its fusogenic properties, and the role of N-glycosylation on conformation, infectivity, and antigenicity. Molecular modeling using a combination of energy minimization and molecular dynamics was used to generate a molecular model for gpl 20. Methods: The Biograf software from Biodesign was used to generate the three dimensional structure of a truncated gp120 consisting of a 357 residue sdction from a.a. #101 to #457. Assignment of intrachain disutfide bonds were according to Leonard et al. (1990, J. Biol. Chem. 265:10373). Initial secondary structure analysis was determined for a non-glycosylated molecule using the genetic computer group sequence analysis program. Results: A stable tertiary protein structure was determined after 120 psec. of molecular dynamics. Neutralizing antibody domains for the second conserved domain (domain 1), the V3 loop (domain 2), CD4 receptor binding domain (domain 3), and a weakly immunogenic domain 4 are all solvent accessible. All complex and high mannose N-glycosylation sites are on the surface of the protein in contrast to that predicted by hydrophilic/hydrophobic analysis. The CD4 binding domain, using the trpncated model, exists as a distinct elongated surface with minimal molecular dimensions of approximately 23A by 15A. The addition of a N-linked high mannose core to Asn 230 with further dynamic modeling results in the association of this main body carbohydrate element with one surface of the CD4 binding site. Trp 427 and Tyr 435 resident in the CD4 binding site are in close proximity to the high mannose core as is Tyr 191 on the adjacent molecular surface. Discussion and Conclusions: The results to date are consistent with known functional features of gp120 but do not provide formal proof of the validity of the model. The CD4 biding site on gp120 formsa distinct structural element of the dimensions predicted by X-ray crystallographic data for the CD4 molecule (1990, Nature 348:411, 419). The cluster of aromatic a.a. along one face of the CD4 binding site on gp120 suggests a pote"iahlw-w docking interaction between gp120 and the exposed Phe 43 on CD4. All of the non-CD4 antigenic sites which induce neutralizing antibodies are found on the contralateral surface of gp120. The release of CD4 atomic coordinates will allow docking experiments between CD4 and the gp120 model. Successful docking would provide a higher level of confidence in the model until X-ray crystallographic data are available. [Supported by HHS Grants A125272 and A1293981 NOTES M.A.1170 A DYAD SIMMETRY MOTIF IN THE NEGATIVE REGULATORY ELEMENT OF HIV-1 LONG TERMINAL REPEATS DOWNREGULATES TRANSCRIPTION Giacca Maura Gutierrez. Maria Ines; d'Adda di Fagagna. Fabrlzio; Menzo. Stefano: Falaschl, Arturo International Centre for Genetic Engineering and Biotechnology - Trieste (Italy) Objective: In the Negative Regulatory Element (NRE) of the Long Terminal Repeats (LTRs) of HIV-1, a dyad symmetry element CACGTG is located at position -166 upstream of transcription start site. This sequence Is present as transcriptional element upstream of many eukaryotic and viral genes, including the Major Late Promoter of Adenovirus (Ad-MLP). The binding properties and the functional role of the site in the HIV-1 LTR were Investigated. Methods: Proteln-DNA interactions between the LTR and nuclear extracts from human cells were Investigated by gel retardation, ONase I footprinting and South-Western experiments. The functional role of the binding site was ruled out by transient transfection assays.. Results: A nuclear binding activity of HeLa cells protects a 23 bp region (-152 to -174) of HIV-1 LTR, containing the CACGTG motif in its center. At least three protein species recognize the LTR sequence. of 44 KDo (corresponding to transcription factor USF/MLTF), 70 KDa and 110 KDa. CAT assays suggest that the binding site located in the HIV-1 LTR is a negative regulator of transcription. The negative function is exerted both in presence or absence of tat transactivation, in different cell lines, and after PMA stimulation. Discussion: The identified binding site contains the CACPuTG motif, which is present also in the upstream regions of several genes of mammals, birds, amphibians and plants, and in most of these cases constitutive and ubiquitous nuclear binding activities have been recognized, that interact with these sequences. The conservation of this binding domain although In the context of different functions represents an intriguing example of evolutionary tinkering. M.A.1171 EPITOPE MAPPING OF THE LOW-MOLECULAR-MASS SUBUNITS OF REVERSE TRANSCRIPTASE IN HIV-1 BY MONOCLONAL ANTIBODIES. Chandra,P.*; Gerber,T.*; Chandra,A.*; Sarin,P.S.**; Pavlikov,S.***; Sidorovitch,l.***; Khaitov,R.***; lvanov,V.***; Koziach,A.***. * Laboratory of Molecular Biology (ZBC), Frankfurt University Medical School (FRG), ** Laboratory of Tumor Cell Biology (NCI), Bethesda (USA), *** Institute of immunology, Moscow (USSR) Objective: We have shown that the biochemical heterogeneity of RT seen on isoelectric focusing is a unique feature of HIV-1 RT; RT purified from HIV-2, SIVagm, HTLV-I and HTLV-Il did not show this heterogeneity (FEBS-Lett. 197: 84-88, 1986). Our objective is to characterize the two species of HIV-1 RT separated on IEF serologically using monoclonal antibodies. Methods: Monoclonal antibodies to HIV-1 RT were obtained from mouse/mouse hybridomas generated in our laboratory. Peptides of defined sequences from the pol gene were synthesized by the solid phase method on a peptide synthesizer. Results and discussion: Using monoclonal antibodies to HIV-1 RT, low-molecular-mass subunits (p29,p32,p40) were identified in HIV-1 RT purified from HIV (HTLV-IIIB) virions by IEF. Epitope mapping with synthetic polypeptides from various regions of the pol gene suggests, that low-molecular-subunits result from N-terminal cleavage of the p51 subunit. The subunits could be separated only by SDS-PAGE and detected by immunoblotting. They could not be separated on chromatographic columns, suggesting that the subunits are complexed or conformationally arranged in such a way that their separation on the basis of molecular mass is not possible.

Page  135 M.A.1172 QUANTITATION OF HIV DNA BY PCR IN SEROPOSITIVE NAVY PERSONNEL Ferre F., Lewis D., Marchese A., Wallace M., *Beecham J., Duf f C., Gersten M., Burnett K., Jensen F., Carlo D. and *Salk J. Immune Response Corp., U.S. Navy Hospital, "salk Institute, San Diego CA (USA) Objective: To develop an accurate quantitative HIV PCR assay to assess the efficacy of the IRC HIV-1 immunogen in clinical trials. Kethods: Quantitation was based on direct labelling of the PCR reaction and comparison with known amounts of HIV DNA (standard curve). One et of HIV primers was used (SK38/SK39) and one of the primer was P-end labelled to directly quantitate the PCR product afte'r electrophoresis. Results: quantitation was done on 75 HIV infected patients. A linear relation was shown between the HIV copy numbers from the standards and the cpm with detection limit of 10 copies per 4x10 cells. The quantitative PCS correlates very well with the HIV viral culture i.e. high viral burder (100-1000 copies/4xl0 ) are found in samples from which virus were isolated after only 7 days in vitro whereas lower viral burden (<100 copies) are found in samples from which virus were isolated after 14 to 21 days ir vitro. In addition, wen normalized by %CD4 cells, a PCR copy numbex corresponding to <1/10 infected CD4 cells was associated with Walter Reec stage I and II disease and with a %CD4 count >0.25. Conclusion: The results presented here demonstrate that PCR can be usec to reproducibly and accurately quantitate HIV DNA in clinical samples. PCI copy number normalized to %CD4 cells is correlated to several measurements of disease progression, and may serve as a convenient surrogate marker ir trials of immuno- and chemotherapy of HIV infection. M.A.1173 THE EBNA2 GENE OF EPSTEIN BARR VIRUS (EBV) TRANS-ACTIVATES THE HIV LONG TERMINAL REPEAT SEQUENCES (HIV-LTR). Scala, Giuseppe; Mallardo, M; Ruocco, M.R.j Squitieri, B1 Venuta, S*; Quinto, I. II Medical School, University of Naples, Naples, Italy, *Medical School, University of Reggio Calabria, Catanzaro, Italy. Objective: This study investigated the relationships between HIV and EBV, two viruses often clinically associated in AIDS patients. Methods: We have transfected De Few and EW36 cells, two EBV-negative B lymphomas, with the pZip-EBNA2 vector, which carries the EBNA2 gene together with the G418-resistance gene. Stably DeFew-EBNA2 and EW36-EBNA2 cells and control DeFew and EW36-pZipneo cells were obtained after selection (Geneticin, 3 mg/ml). These cells were transiently transfected with plasmids carrying the HIV-LTR sequences inserted upstream to the bacterial cat gene. In particular, the following plasmids were utilized: pUCBenn-cat, carrying the wild type HIV-LTR and the TAR region; pNFA-cat, pSpA-cat, and pTAR-cat, with inactivated NFkB, Spl or TAR region of HIV, respectively. The tat-expressing pAR-tat plasmid was utilized in co-transfection experiments. Results; We observed a consistent trans-activation of the HIV-LTR sequences in EBNA2-transfected cells. This activation was dramatically decreased when the pNFA-cat plasmid was used as a target for the EBNA2 gene, suggesting that EBNA2 gene product acts trough NFkB-like factors. Moreover, a cooperative effect of the EBNA2 gene of EBV and the tat gene of HIV was observed. Other EBV genes, namely EBMA1, EBNA3, EBNA-LP were completely ineffective. Discussion: These data suggest that trans-activation of HIV may easily occur in cells infected by EBV, which can act through its EBNA2 gene. This work was supported by grants from AIDS, AIRC, and CNR projects. 0 NI oo NOTES NOTES M.A.1174 SEQUENCE-SPECIFIC RECOGNITION OF CAR/HRE, THE TARGE I SEQUENCE FOR THE REV AXIS OF HIV-1 Dayton, E.T.', Konings,D.",Powell,D.*'&',Shapiro,B.*, Maizel,J.'" & Dayton. Andrew I.* 'Laboratory of Immunoregulation, National Institute of Allergy & Infectious Disease, N.I.H, Bethesda, MD, U.S.A.; "Laboratory of Mathmatical Biology, Division of Cancer Biology & Diagnosis, National Cancer Institute, N.I.H, Frederick, MD, U.S.A.; **Genetics Department, George Washington University, Washington, D.C. U.S.A. Objective: To determine which bases in CARIRRE are involved in sequence-specific recognition by the elements of the REV axis of HIV autoregulation. Methods: We have devised a novel method of analysis called "modified base-switching" to design RNA target sequence analogues with secondary structures identical to wild type but with sequences that differ from wild type at every position. Unlike anti-sense target sequences these analogues distinguish between sequence-specific information and secondary structure Information. The target sequence analogues were tested for REV-responsiveness in COS cells after insertion into HIV-1 "gag-expression" vectors. Results: Base-switching the entire CAR/RRE resulted in a non-functional target sequence. Exchanging stem/loop domains between wild type and base-switched target sequences produced target sequence analogues with only limited areas of wild type sequence and varying degrees of function. Single point mutations were inserted throughout the "single-stranded" regions shown by these analogues to be exquisitely sequence-specific. Discussion and Conclusions: CAR/RRE has considerable sequence-specificity distributed throughout the molecule. There is limited sequence-specificity in the IV-VI domain and in the stems of the REVbinding domain. Exquisite sequence-specificity lies in the "sinale-stranded" RNA lust 5' and just 3' to stem II in the REV-bindinq domain. DEFINING THE ROLE OF NEF IN THE TRANSITION FROM A QUIESCENT HIV M.A.1175 INFECTION TO AN ACUTE DISEASE STATE Psallidonoulam. Miltiades*; Newell, M*; Natarajan, V*; Lane, H.C.**; Salzman, N.P.*. *Georgetown University, Washington DC, U.S.A.; **NIAID, NIH, Bethesda, Maryland, U.S.A. Objectives: TO determine the function of nef in the HIV-i life cycle and to study changes in nef expression during the transition from a quiescent HIV infection to an acute disease state, associated with a pronounced decline in T4 cell counts. This activation from a latent viral infection to viral productivity is hypothesized to be due to qualitative properties of the nef gene. Methods: DNA was obtained from isolated PBMC of HIV-I seropositive asymptomatic individuals and patients with AIDS. Nef specific sequences were determined from cloned PCR amplified products using primers from highly conserved regions flanking the nef gene. Results: The sequences demonstrated insertions of direct repeats, deletions and point mutations. However, these mutations did not introduce in-frame stop codons and were mostly distributed in the amino and carboxy terminals. The biological significance of these mutations is currently being tested in cell lines expressing the cloned nef genes under the control of inducible promoters. The influence of HIV growth by the induction of nef in these cell lines is evaluated by testing effects on viral entry, changes in viral DNA, RNA and protein molecules as well as the production of infectious virions. Conclusion: The findings of this study will help to delineate domains critical to the nef function and contribute to an understanding of the biological role of nef in the HIV life cycle. b N] N] h,

Page  136 1UQ U3 M.A.1176 STRUCTURE-FUNCTION ANALYSIS OF GLYCAN VARIANTS OF HIV I ENVELOPE OLYCOPROTEINS. ViasevachariL M. B Dewar, Robin L.; Stephens, Ann; Psllidopoulos, M.C; and Salzman, Norman P. Georgetown University, Washington, D.C. QO tive: To study the structre-function relationship of envelope glycoproteins of HIV 1 made in wild-type and glycosylation mutant CHO cell lines. Method: HIV I envelope glycoproteinswere produced in wild-type and lectin resistant CHO cell lines infected with a recombinant human adenovinis which carriesthe HIV 1 gee. The glycan tructurewasanalyzed by diolabeling with 3Hsugars, and czyne digestions followed by imnmoprecipitation and gel electropmoresis. Infected cells wereco-cultured with normal CD4 bearingcells to examine syncytium formation. Relative binding ability of the wild-type and the mutant glycoproteins with respect to CD4 wa studied using binding asys with radiolabeled soluble CD4 and by FACS analysis uing Fl-labeled anti-gpl20 namonoclonal antibodie. Rult: The humanadenovirus recombinant (Ad5env) upon infection of wild-type CHO cells expresses envelope glycoproteins that resemble the glycoproteins produced in HIV 1-infected cell. Infection of mutant cell lines yields glycopmtei with altered glycan structures. The were diffeences in the extent of syncytium fomation among these variant glycoprotens, the most extensive syncytia being formd between CD4-bearing cells and gpl20-expressing lec8 cells.FACS analysis of cell-aurface expression of gpl20 indicated that the lec8 cell line expressed much less gpl20 than did the other cell lines. In contrast this form of gpl20 bound slightly more CD4 than did the forms of gpl20 produced by the other cell lines. neluank: The glycan moieties of gpl20 of HIV 1 have been postalatedto play importat roles in virus infectivity and in mounting an immune response to virus infection. The envelope glycoproteins made in glycosylationutant CHO ell lines had the anticipated structuralalterations in their glycans. Lee glycosyl variant gp120 moleles have a g wreaterapacity to bind to CD4 than do the wild-type CHO or the other variant glycoprotein. An increase in syncytium formation seen with gpl20-expressing lec8 cells is thereult of the lck of sialic acid residues on thecabohydrateside-chais of thesemolecules. NOTES M.A.1177 HIV MODULATION OF CELL SURFACE GLYCOPROTEIN EXPRESSION Mscag,.L*;; Rosenberg, Y.'; Perera, P.*; White, B.*; D'Arcy, L.*; Ritchey, D.* and Burke, D. S.+ *Henry M. Jackson Foundation, Rockville, MD; +Div. of Retrovirology, WRAIR, Washington D.C. Obiective: To examine the correlation between CD4 depletion and HIV expression. Ml:thods: Chronically HIV-1 infected cell lines were established using virus derived from a molecular clone. Cell-free media from ten of these lines, in addition to a series of HIV-1 isolates, were tested on SupT1 cells to modulate the expression of cell surface markers. HIV-1 expression was monitored by in situ hybridization, while CD4 and CD8 cell surface expression was monitored by flow cytometry analysis. Results: SupTi cells treated with HIV containing cell-free media from either chronically infected cell lines or HIV-1 isolates resulted in the loss of cell surface CD4 within 4 to 6 weeks. The down modulation of surface CD4 correlated with a precipitous decline in cell viability. Unexpectedly, media from one chronic line, 190, resulted in the sequential loss of CD4 followed by CD8 from the cells surface. Furthermore, HIV-190 induced CD8 depletion from HIV infected CD4 negative cell cultures. In all cases the resulting cell surface phenotypes are stable in culture, even though only a small proportion of the cells (<1% In some lines) are expressing HIV by in situ hybridization. Conclusion: The discrepancy between the low percentage of cells expressing HIV RNA transcripts and the total depletion of measurable surface CD4 by FACS analysis, can not only be attributed to gp120-CD4 complexing intracellulady. Further exploration in this system may lend insight into CD4 depletion in late-stage HIV infected individuals. NOTES NI NI 'a clz M.A.1178 EXPRESSION OF THE NEF GENE OF HIV-1 BY TRANSFECTION INTO B-CELLS Kienzle. Norbert and MOller-Lantzsch, Nikolaus. University of the Saarland, Dept. Virology, Homburg, FRG Objective: Establishment of B-cell clones stably transfected with the NEF gene for investigation of interactions of NEF with cellular components. Methods: Using a transfection vector which contained the essential elements of EBV for episomal replication and the selectable hygromycin B marker, the NEF gene of HIV-1 was stably transfected by electroporation into EBV immortalized B-cells. The NEF gene transcription was driven by the constitutive early promotors of CMV or SV40. Results: NEF gene constructs flanked by the polyadenylation signals from SV40 and by either the CMV or SV40 promoters expressed NEF protein in Raji cells, whereby constructs carrying the CMV promoter stimulated NEF expression to a much higher amount. The NEF protein was detected by immunoblot or indirect immunofluorescence analysis using a mouse monoclonal anti-nef antibody. No NEF protein could be obtained from these transfection plasmids, if parts of the LTR sequence of the HIV-1 provirus were additionally located downstream to the NEF gene. Experiments to elucidate the inhibitory effects on NEF expression as well as cell compartment localization studies and possible interactions of NEF with cellular components are still in progress. Conclusions: We have generated B-cell dones stably expressing the NEF gene of HIV-1 and are now able to investigate associations of net with cellular factors. rhese studies were supported by BMFT grant No.: 11-074-88. M.A.1179 SIEMULTAIEOUS POST-EMlBEDDING ULTRUSTRUCTURAL DETECTION OF RINA AND PROTEINS OF HIV-I. GrigorievVladimir*; Escaig-Haye,,.**; Pournier,J-Gf*; Sharova,N.*; Tugizov,Sh.*; Klimenko,S.* *The D.I.Ivanovsky Institute of Virology,Acad.Med.Sci..,Iioscow,USSR, **Hopital St.Vincent de Paul I.N.S.E.R.i. U43, Paris, Prance. Objective:The purpose of this study was to detect and localize HIV antigens and its RIIA molecules simultaneously in infected cells during morphogenesis of virus at the ultrustructural level. Ilethods:Jurcat-tat cells chronically infected with HIV-I,strain A-899, were fixed and embedded in Lowicryl K4ti resin with the progressiv low temperature technique.A commercial genomic probe containing HIV-I sequences and labelled with biotine was used to hybridize in situ with virus RNA molecules.To detect viral proteins monoclonal antibodies (MIAb) to p17 and p24 were used.Eilectron microscopic visualization immune and hybridization reactions were carried out by immunogold staining.Post-embedding irmmunocytochemistry was used. Results:In HIV infected cells iAbs yielded strong labelling over the mature,immature and budding viral particles..urthermore, the viral proteins occupied a narrow zone under cytoplasmatic membraneIIIV RPA was accumulated in the same region of the infected cells as products of "gag" gene.The rest part of cytoplasm had a weak level of labelling Conclusion:Simultaneous detection of viral nucleic acid and proteins may provide novel and attractive approach for study of migration and relationshiD of these molecules in the infected cells. a CC> cz rj CC>

Page  137 M.A.1180 MECHANISM OF FUNCTION OF yIlV REV AND HTLV REX Benko D.M.*; Unge T.*; Solomn L.*; Ciminale V.*; Tavemaraks Nektarios**; PavlakUs G.N.*; Felber BX.* * Baic Research Program National Cancer Institute-FCRDC, Frederick, MD, USA., **Institute of Molecular Biology and Biotechnology, Heraklion, Crete, Greece, Objective: To study the mechanism of activation of viral structural gene expression by the regulatory proteins Rev and Rex. Results & Discussion: Rev, mutant Rev, and Rex proteins were expressed in bacteria, purified, and analysed for their ability tobind the -acting Rev responsive element (RRE) and Rex-responsive element (RXRE) in vitro. Cotnofection experiments were also performed to study the functional properties of the proteins In vivo. Results confirmed that Rev protein acts by binding directly to RRE, which is located in the cnv portion of the viral genome. Rev mutants that interfere with Rev function (i.e., transdominant mutants) were also shown to bind to the RRE in vitro, with affinities similar to that of wild-type Rev. Studies of RRE mutants demonstrated that Rev and transdominant Rev mutants bind to the same target within the RRE in vitro and in vivo. However, unlike wild-type Rev, the Rev mutants neither Increase nor decrease the steady-state levels of HIV mRNA. This suggests that the mutants are unable to perform a step that closely follows formation of the protein-mRNA complex. Several lines of evidence indicate that additional cellular factors interacting with Rev are necessary for fnetion. The Rex protein of HTLV1 is functionally similar to Rev and acts through the RXRE, which is located in the long terminal repeat of the viral genome. Rex was shown to act by binding directly to the RXRE. In addition, Rex is able to bind to RRE and activates HIV mRNAs via the same mechanism as Rev. However, the targets of interaction within RRE for Rev and Rex are distinct. In contrast, Rev does not activate HTLV expression and purified Rev does not bind to RXRE in vitro. However, a variant HIV protein containing an altered amino terminus and the carboxy portion of Rev (i.e., Tev) can partially substitute for Rex. Thus, a change in the amino terminal part of Rev alters the biding specificity of this portion. Condcusons: Rev and Rex act by binding to their RNA targets. Although both Rev and Rex act on RRE, their targets are distinct. Additional cellular proteins are necessary for the function of both activators. M.A.1181 EXPRESSION OF IMMUNOLOGICALLY REACTIVE RECOMBINANT TAT, REV AND TEV PROTEINS OF HIV-1 Tavernarakis Nektarios*; Pavlakis George N**; Krmbovitis Elias* * Institute of Molecular Biology and Biotechnology, Heraklion, Crete, Greece, ** Basic Research Program, National Cancer Institute-FCRDC, Frederick, MD, USA Objective: To express the regulatory proteins Tat, Rev and the recently identified Tev in bacteria for the study of patient immune responses to these proteins. Results & Discussion: The proteins encoded by the tat and rev genes are necessary for HIV1 replication and are currently the subject of intensive studies. We used a pUC12-based expression vector to produce the Tat, Rev, and Tev proteins in E. colt. The DNA segments encoding these proteins were engineered to allow efficient expression in E. coli cells and to produce authentic proteins with the same Immunological properties as the native viral proteins produced after infection. This was accomplished in some cases by altering the sequence of the relevant genes to make it more suitable for the protein synthesis machinery of E. coli. The production of immunologically reactive proteins was confirmed by ELISA and western blot analysis using monospecific antibodies raised against the authentic proteins. Preliminary results with HIV patient sera indicated specific antibody reactions. The recombinant proteins were purified using standard chromatographic methods. The antigenicity as well as purity and recovery were monitored with gel electrophoresis and immunoblotting. The purified proteins we describe are used to investigate the antibody response of HIV patients to these proteins, which may allow studies on the disease progression. s t~ I0 h~ NOTES NOTES - I - --- CRITICAL SEQUENCES RESPONSIBLE FOR VIRION FORMATION OF HIV-1 M.A.1182 Hoshikawa, Nariyoshi **"; Masuko, S" ""...; Yasuda, A."'; Takayashlki, E*"; Kojima, A"; Kurata, T"* *Kosel Hospital, Tokyo, Japan, **National Institute of Health, Tokyo, Japan, ***Toray Industries, Inc., Kanagawa, Japan, """Nippon Zeon Co. LTD. Kanagawa, Japan. Objectivm: The gag precursor polyprotein of HIV-1 assembles spontaneously into retrovirus-like particles at the cell membrane. To clarify the critical sequence responsible for viral assembly, we constructed recombinant vaccinia viruses (RVVs) containing a variety of truncated gag genes of HIV-1. Metkods: 01 gonucleotide-directed mutagenesis was carried out to produce truncated gag genes of I1IV encoding polyprotein with predicted molecular mass of 41-, 42-, 43-, 45 -and 48-kDa. RWs were constructed by insertion of the gag genes into vaccinia virus and termed as mOP41, aOP42, OP43, mOP45 and mOP48. The expression of gag proteins by RVVs was analyzed by Western lmunoblot and P24-ELISA. Rabbit RK-13 and human CD4+ Tcell lines infected with the RWs were examined with electron microscopy. Resalts: Imlunoblot analyses of the cell lysates infected with mOP41, a0P42, mOP43, aOP45 and mOP48 revealed a prominent single band of 41-, 42-, 43-, 45-, and 48-kDa proteins, respectively. Culture medium of mOP45 and aOP48 showed a single P24-E.TSA peak at d-1.16 by sucrose density centrifugation. Ultrastructurally, budding structures and release of immature retrovirus-likc particles were detected In the cells infected with mOP45 and mOP48 but not in those infected with aOP41, mOP42 and IOP43. Discassion aad Coaclasions: These results indicate that the entire gag gene is not a prerequisite for viral particle formation. Gag protein contains two Cys-Ilis boxes at the C-terminus. P43 does not contain Cys-His boxes. In contrast, P45 has only a proximal Cys-His box and P48 has both proximal and distal Cys-His boxes. Therefore, critical sequences responsible for viral assembly appear to be located at the region of Cys-His box. ESSENTIAL ROLE OF N-MYRISTOYLATION OF GAG PROTEIN IN BUDDING AND ASSEMBLY M.A.1183 OF lIVKojina, Asato*; Hoshikawa, N"'**; Yasuda, A0**; Takayashiki, E"; Sata, T*; Kurata, T' *National Institute of Health, Tokyo, Japan, "*Kosel Hospital, Tokyo, Japan, """Nippon Zeon Co.LTD. Kanagawa, Japan, Objective: We have previously shown that the HIV gag polyprotein precursor is able to assemble into virus-like particle without other viral proteins. The N-tcrainal glycine of gag protein is myristoylated co-translationally. To investigate the functional significance of the modification, we deleted the glyclne codon from the gag gene and expressed the mutant gene in recombinant vaccinia virus (RW). Reatods: Oligonucleotide-directed mutagenesis was done to delete the glycine codon. The mutated and intact gag-pol genes were introduced into the vaecinia virus genome to construct RVVs termed as mOAGVY and aOGV. RWV-infected cells were metabolically labeled with 3H-myrstate ad nanalysed by Imunoprecipitation and lmmunoblotting. RVVinfected cells were also examined with electron microscopy. Results: Gag precursor proteins expressed by aOAGV lacked myristic acid, but were cleaved and processed into mature p24 and p17 proteins authentically. RT activity in cell extracts and culture medium was much lower with nOAGV than with mOGV. A single peak consisting of both RT activity and the mature gag proteins was observed at Ad1.16 by sucrose density centrifugation of culture medium of mOGV-infected cells. In contrast, no peak was detected when mOAGV was analysed. Budding structures at the cell membrane and IIIV-like free particles were not visible in mOAGV-lnfected cells, but were abundant in mOGV-infected cells. Discmssion and Conclusions: These results indicate that the myristoylation of gag precursor protein is essential for budding and assembly of HIV-1. Furthermore, levels of RT activity appear to be affected by the modification..1

Page  138 00 M.A.1184 FINE MAPPING OF THE IMMUNODOMINANT DOMAINS OF THE HIV-1 and HIV-2 INTEGRASE PROTEINS Vinga-Martins, Cristina, Schneider, T.*,Roenspeck, W., Pauli, G.* and Mueller-Lantzsch, N. University of the Saarland, Dep. Virology, Homburg, FRG *Robert-Koch-Institute, AIDS-Centre, Berlin, FRG *Biochrom-Beteiligungs GmbH & Co., Berlin, FRG ~ Objective: Fine mapping of the major immunodominant epitopes of the HIV-1 and HIV-2 integrase proteins by synthetic oligopeptides and recombinant proteins. Methods: Overlapping decapeptides spanning the entire HIV-1 and HIV-2 integrase proteins were used as synthetic antigens in a pin-based oligopeptide-EUSA to test HIV-1 or HIV-2 positive sera. Three of these peptides, representing essential epitopes of the HIV-1 integrase, were used in an oligopeptide ELSA for screening 153 HIVI positive and 126 HIV-1 negative human sera The reactions of half of the HIV-1 positive sera with synthetic peptides were compared to their reactions with different TrpE-integrase fusion proteins comprising the same epitopes in an immunoblot. Results: Nearly 80% of the HIV-1 positive sera tested reacted with synthetic HIV-1 integrase peptides from the amino- and/or the carboxyterminal part of the protein. In the immunoblot most of the HIV-1 positive sera predominantly reacted with only the carboxyterminal part of the integrase protein. All anti-integrase (p31) positive sera could be detected by either the oligopeptide-ELISA and/or the recombinant immunoblot. For the HIV-2 integrase protein only one immunodominant epitope could be detected by the pin-based oligopeptide ELISA. This epitope is different from the HIV-1 integrase epitopes. Screening of HIV-2 positive sera in an HIV-2 integrase oligopeptide-EUSA and testing of cross-reactivities are under Investigation. Conclusions: Fine mapping of the HIV integrase proteins could define three HIV-1 and one HIV-2 integrase immunodominant epitopes. The reactions of human sera against these epitopes depend on the test system used (synthetic peptide-ELISA or recombinant immunoblot). (Supported by BMFT grant No. 11-074-88), M.A.1185 "CHARACTERIZATION OF VPU FROM HIV-1 EXPRESSED IN E.COLI AND MAMMALIAN CELLS" Schubert, Ulrich1; Morelle, Christine3; Hoffmann, Klaus3: Hauser, Hansjorg3; Schneider, Thomas2; Pauli, Georg2; Porstmann, Tomas1 IDep. of Med. Immunology, Med. School (Charite) Humboldt univ. Berlin (HUB); 2Bundesgesundheitsamt, Robert-Koch-Institut, Berlin West; 3Cell Biology and Genetic Section, GBF, Braunschweig, FRG The 16 KD vpu(out) protein is only found in HIV-1 infected cells at very low concentrations. It is thought to be a phosphorylated membrane protein which has a positive effect on budding and assembling of the virus. We expressed the vpu-gene after isolation by PCR in E.coli as an unfused protein by using a two-cistron system. The induction leads to an arrest of cell growth of the recombinant clones. In E.coli recombinant (r) vpu was found in two forms: as cytoplasmic protein or in association with the membrane fraction. vpu from E.coli is immunologically reactive with both anti-vpu monoclonal antibodies and anti-vpu positive sera from HIV-1 infected individuals. The kinetics of transcription (by Northern blot analysis) and expression (by the APAAP-technique) of vpu in H9 cells after infection with HTLVIIIB was investigated. Maximum of expression was determined 9 days post infection. We have stable expressed vpu in hamster (BHK) and human fibroblast (293) cell lines. The expression of vpu with the strong promotor of the myeloproliferative sarcoma virus (MPSV) leads to a cytotoxic effect, in particular when HIV-1 gag proteins are coexpressed. The phosphorylation of vpu has been investigated in vitro and in vivo. We could show that rvpu is phosphorylated by the casein kinase two (CK-2) in vitro. Phosphorylated rvpu tends to dimerization. 4 N] cn K^ oo NOTES NOTES M.A 1186 OLIGOMERIC STATUS, CARBOHYDRATE CONTENTS, AND MOLECULAR M.A.1 WEIGHTS OF HIV-1 GLYCOPROTEINS. Steven, Alasdair; Trus, Benes; Booy, Frank*. Kaczorek Michel; Wall, Joseph; Hainfeld, James; Eiserling, Fred Thomas, Daniel ' National Institutes of Health, Bethesda, MD, USA, **Pasteur Merieux, Val de Reuil, France, *Brookhaven National Laboratory, Upton, NY, USA, UCLA, Los Angeles, CA, USA, U** niversite de Rennes, France. Obiective: With the aim of resolving enduring ambiguities as to certain basic molecular properties of the env glycoproteins, we have applied quantitative electron microscopy and image processing, including scanning transmission electron microscopy (STEM) of unstained specimens, to purified gpl60 and gpl20. Methods: Recombinant forms of gpl20 and gpl60 ("gpl60s") were purified from a vaccinia virus expression system. Gpl60s had been rendered water-soluble by deletion of a hydrophobic sequence, and resistant to proteolytic processing by alteration of its cleavage site. Samples were examined with the Brookhaven STEM and a Zeiss EM902. Image analysis was performed with the PIC system. Results: Our STEM data show that most gpl60s molecules have masses of 240-250 kDa, with a minor component at -125 kDa, whereas gpl20 averages 89 kDa. Taking its amino-acid sequence into account, it follows that gpl60s is a dimer of 123 kDa subunits, of which - 32 kDa are carbohydrate, and 91 kDa are polypeptide. Gpl20 is a monomer of 89 kDa, and contains virtually all of the carbohydrate. Gp41 is glycosylated only slightly, if at all, and is responsible for the interactions that stabilize the gpl60s dimer. A molecular mass-map of gpl60s derived by image processing depicts an asymmetric dumbell. Its two domains have masses that are consistent with a gpl20 dimer and a gp41 dimer, respectively. Discussion and Conclusions: These results indicate that the nominal molecular weights of gpl60 and gpl20 (160 kDa and 120 kDa, respectively) are substantial overestimates, attributable to the effect on the molecules' electrophoretic mobility of large amounts of covalently attached carbohydrates. Our mass-map will serve as a frame of reference relative to which important functional sites may now be localized. M.A.1187 "IN VITRO" SPERM MEDIATED HIV PROVIRAL DNA TRANSFER TO LYMPHOBLASTOID CELLS. EaziL. Vito M.*, Rinaldi, M.*, Catani, N.V.*, Ravagnan, G.*, Metafora, S.**, Manna, C. ~, Peluso, G. *** and Farace, M.G. ~* *Institute of Experimental Medicine, CNR, Rome, Italy, **I.I.G.B., CNR, Naples, Italy, ***I.B.P.E., CNR, Arco Felice (NA), Italy, 'University of Rome "Tor Vergata", Rome, Italy. OBJECTIVE: Our group has recently shown that spermatozoa can take up exogenous DNA which iS transferred to egg cells at fertilization. Moreover, several reports have demonstrated that human opermatozoa bind and penetrate HLA-DR+ leukocytes. We report here that human spermatozoa can capture in vitro HIV proviral DNA, and that they can transfer HIV DNA to lymphoblastoid cells. METHODS: Ejaculated spermatozoa from fertile, healthy donors were repeatedly washed, incubated either with the entire or replication-defective HIV proviral DNA, and co-coltivated overnight with HLA-DR+ B- or T- lymphoblastoid cells. The same experimental procedure was also used to transfer a plasmid containing the resistance to neomicine. Sperm-penetrated cells can be selected in a medium supplemented with neomicine. RESULTS AND CONCLUSIONS: (1) After overnight co-colture with sperms previously exposed to HIV fluorescent photoaffinity labeled DNA, 10-15% of cells were fluorescent. (2) HIV DNA wa$ detected in cello cultured for three months after co-coltivation with spermatozoa incubated with replication-defective HIV DNA. (3) Using the plasmid carrying the neo-gene we have selected cells resistant to neomicine. (4) We are now testing if cells penetrated by spermatozoa with the entire HIV provirus can produce mature viral particles. These data support results presented by others on sperm penetration of leukocytes and suggest the possibility that HIV positive sperms can transfer HIV DNA and infect somatic cells - a I-r C>

Page  139 M.A.1188 HELICAL POLYMERS OF HIV-1 REV PROTEIN Misra, Manoj; Steven, Alasdair; Booy. Frank*; Kocsis. Eva; Wingfield Paul** National Institute of Arthritis, Musculoskeletal and Skin Diseases and Office of the Director, National Institutes of Health, Bethesda, MD, USA Obiective: In order to understand how REV regulates the expression of the ENV gene by binding to the REV-responsive element of ENV mRNA, thereby inhibiting splicing, we are studying the properties of REV protein expressed in LE coli. We have found that, under appropriate in vitro conditions, purified REV polymerizes into well-ordered helical filaments, and are investigating this phenomenon, with a view to determining the filaments' structure. Methods: REV was cloned into the pL-ner plasmid, expressed in E. cli MM294, and purified by ion exchange chromatography. Samples were negatively stained with uranyl acetate and visualized in a Zeiss EM902 electron microscope. Micrographs were analyzed by image processing, using the PIC software. Results: Upon dialysis into 50mM sodium phosphate pH8, 0.1M NaCI, 10mM K2S04, ImM DTT, at 40C, purified REV gradually polymerizes into long filaments. They are of uniform width (20nm), have an axial channel 5-7nm in diameter, and can range up to many microns in length. Under certain conditions, e.g. elevated protein concentrations, these polymers exhibit a tendency to aggregate laterally. The filaments are somewhat flexible, and generally exhibit random, continuously varying, curvature. Filament images were straightened computationally, and their diffraction patterns calculated. These patterns reveal layer-lines, notably an off-meridional reflection at an axial spacing of (5.3nm)-1, that indicate a regular helical packing of REV subunits. Discussion and Conclusions: Several lines of indirect evidence raise the possibility of a broad analogy between REV and the class of RNA binding proteins that includes the coat protein of tobacco mosaic virus. We intend to clarify the status of this hypothesis by determining the helical structure of REV filaments, and investigating the influence of RNA on REV's capacity to self-assemble. NOTES M A 1189 EFFECTIVE SUPPRESSION OF HIV REPLICATION IN VITRO BY COMBINATION ""A.1 9 TREATMENT WITH AZT AND RECOMBINANT INTERFERON-ALPHA Epstein, JavlJoshi, B.; Lee, S.; Pollock, L.; Mayner, R.; Hewlett, I. Laboratory of Retrovirology, DTS, CBER, FDA, Bethesda, MD USA. Introduction; We have evaluated the efficacy of the antiviral agents, AZT and recombinant interferon-alpha (rIFN-a) both singly and in combination, to inhibit the expression of HIV-1 at the nucleic acid level using PCR. Methods: H9 and 0937 cells were infected with 10 infectious doses per ml of HIV-1 (IIIB strain). After virus adsorption, the cells were transferred to fresh medium or medium containing AZT (50 M) or rIFN-a(50 units/ml) alone or in combination at the same concentrations. The drugs were added concommitantly or 24 hours after infection. At 12 days post-treatment, drugs were removed and cultures maintained drug-free for an additional 14 days. Virus production was measured by reverse tranecriptase(RT) and p24 antigen assays. Viral DNA and RMA were isolated from the cells at various time points and nucleic acid analyzed by (PCR) using primers derived from the gag region. Results: Greater than 95% inhibition of HIV-1 by AZT and the combination of AZT and rIFNa was detected from 3-4 day after infection using the RT and p24 assays in both H9 and U937 cells that were treated with drugs either 1 or 24 hours post-infection.Viral DNA and RNA synthesis were not significantly inhibited by rIFNa. Some inhibition of both DNA and RNA was observed with AZT alone in the early phase of infection followed by a recovery of viral nucleic acid synthesis with time. More than 80% inhibition of viral DNA synthesis was observed with the combination of AZT and rIFNa.Viral RNA was undetectable in both infected H9 and U937 cells in the presence of AZT and rIFNa combined. This effect persisted even after removal of drug. Conclusion; Our results suggest that combination therapy with AZT and interferon may ýe more effective for sustained inhibition of virus replication at all levels. Studies are in progress to understand the molecular mechanism of this inhibitory effect. NOTES 4 4 c^ 00 I1 M.A.1190 INDUCTION OF HIV-1 FROM CHRONICALLY INFECTED PROMONOCYTIC CELLS BY CHEMICAL CARCINOGENS Hewlett, Indira; Joshi, B.; Riordan, G.; Lee, S.r Pollock, L.; Epstein, J. Laboratory of Retrovirology, DTS, FDA, Bethesda, MD, USA Objective: To evaluate the effect of chemical carcinogens on the expression of HIV in chronically infected monocytic cells. Methods: The chronically infected promonocytic cell line Ul is characterised by low constitutive HIV production, which is induced by phorbol esters (TPA) and tumor necrosis factor(TNF-a). Cells were treated with various chemical carcinogens for varying lengths of time and culture supernatants tested for the presence of reverse transcriptase (RT) and p24 antigen. HIV expression was evaluated by quantitative PCR using slot blots. Results: HIV antigen and RNA were induced 20 -fold, 6 days after treatment with benzopyrene (BOP) while a-hexachlorohexane (aHCH) treated cultures showed a 6-7 fold induction after 12 days. In our experiments, induction of antigen and RNA by TPA was comparable to that by BOP although it occurred earlier i.e 2-3 days post-infection. Conclusion: Certain chemical carcinogens such as BOP and aHCH induce the expresion of HIV from chronically infected monocytic cell lines expressing low levels of virus. Induction occurs at the transcriptional level resulting in significant virus production. Since BOP is found in cigarette smoke and charbroiled meats and aHCH is a constituent of certain pesticides, our results may have implications for the effects of such agents on the pathogenesis of HIV infection in man. M.A.1191 HIV-1 INTEGRASE EXPRESSED AS A FUSION PROTEIN IN E.COLI EXHIBITS M.A. DNA BINDING AND ENDONUCLEASE ACTIVITY Marcus-Sekura. Carol. Woerner, A. LDNAV, DV, Center for Biologics Evaluation and Research, FDA, Bethesda, MD USA Objective. To determine whether HIV-1 integrase (IN) (the viral protein required for integration of reverse transcribed virus DNA into the host cell genome), when expressed as a fusion protein in E. coli, exhibits properties expected of a biologically active molecule. Methods. Several IN expressing clones have been constructed using the bacterial expression vector pWSSO. One construct, pKNA01, expresses 13 amino acids of the lambda ell protein fused to the 27 C-terminal amino acids of reverse transcriptase (RT), followed by the entire IN sequence. A second clone, pHIP106, contains 13 amino acids of the lambda clI protein fused directly to the entire IN sequence. Additional constructs containing deletions and amino acid substitutions have also been prepared. Purified recombinant proteins expressed by these clones have been examined for both DNA binding activity using a southwestern procedure and endonuclease activity using bacterial plasmids as substrates. Results. Recombinant proteins expressed by both pKNAlOl and pHIP106 demonstrated both DNA binding capability in the presence of Mn2+; and endonuclease activity, preferring Mn2+ over Mg2+ as the divalent cation. Some mutated proteins expressed differences from controls in these assays. Conclusion. HIV-1 IN expressed as a fusion protein in E.coli exhibits both DNA binding and endonuclease activity. The presence of a free amino terminus is apparently not required for either of these activities. This system should be useful in evaluating requirements for IN activity.

Page  140 0 M.A.1192 HUMAN MONOCYTES/MACROPHAGES (HMM) HIV-1 INFECTION: TNFaIpha PRODUCTION, CELL ACTIVATION AND MICROBICIDE ACTIVITY Le Naour R.0, Raoul H.~, Mabondzo Alcsg, Ripoll L., Barre-Sinoussi EC.*, Caire Y00, and Dormont D.o* * SSA, CEA, Fontenay-aux-Roses.* Inst. Pasteur, Paris, France."o Col. Phys. & Surg., Columbia U., N. Y., USA. Objective: To determined 1) the consequence of HIV infection on the TNFa synthesis and regulation, 2) the relationships beetwccn HIV infection and the different activation stages of the HMM by analysing the expression of the 1-6 fructoscbiphosphatase (1-6 FBPasc) gene, and 3) the capacity of the HIV infected HMM to modify the cellular microbicidal activity through MnSOD mRNA expression quantification. Methods: Fresh human peripheral blood monocytes (PBM) were obtained from healthy HIV-1 seronegative donors. PBM were isolated using the centrifugal elutriation technique after a ficoll-hypaque gradient, and cultured by adherence to culture dishes in RPMI 1640 supplemented with 10% fetal calf serum for 7 days, and infected with HIVI/LAV strain (0,02 TCID50/cell). HMM culture supernatants were recovered from 2 hours to 20 days post-infection (p.i.) for a TNFc cytotoxicity assay. Viral production was monitored every 2 days (P25, RT activity). At the same time-point, cells were scraped, and cytoplasmic RNAs were extracted. Northern blots analysis were performed in order to determine the expression level of the TNFra, MnSOD and 1-6 FBPase genes using appropriated 32P-nick-translated probes. Results: HIV-1 infected HMM did not produce any biologically detectable TNFa during the early time p.i., despite the increase of expression level of the TNFa gene at 2 and 4 hours p.i.. Moreover, an increase in the translation level was detected for the MnSOD and 1-6 FBPase genes 4 hours p.i.. Remarkably, in spite of cell viral infection (P25 positive, Abott EIA), little or no MnSOD and TNFa mRNAs were detectable 18 days p.i., whereas an hypcrexpcssion of the 1-6 FBPasc gene is still observed at the same time. Conclusion: The concomitant hyperexpression of TNFa, MnSOD and 1-6 FBPase genes, within the few hours following HIV-1 infection, may support the hypothesis that HIV infection induces a stronger activation of the HM M. Moreover, the low expression level of TNFa and MnSOD genes 18 days p.i. may be related with the persistence of the virus in HMM during HIV infection. NOTES M.A.1193 ASSEMBLY DOMAIN OF THE HIV-1 ENVELOPE PROTEIN Earl, Patricia L. and Moss, B. Laboratory of Viral Diseases, National Institute of Allergy and Infectious Diseases, Bethesda, MD 20892, USA Objective: To examine the assembly domain of the HIV-1 envelope (env) glycoprotein. Methods: Env protein was produced via infection of cells with a recombinant vaccinia virus expressing the bacteriophage T7 RNA polymerase gene and transfection with plasmids containing secreted forms of the env gene under control of a T7 promoter. All env genes studied contain all of gpl20 and at least the N-terminal 129 amino acids of gp41. Results: We have previously shown that the N-terminal 129 amino acids of gp41 constitute the region primarily responsible for dimer stability. Point mutations in highly conserved elements of this region were made. These include alterations in a leucine zipper motif, substitution of Ser for Cys residues, and elimination of potential N-linked glycosylation sites. In addition, two further C-terminal truncations were made at amino acids 68 and 99 of gp41 respectively. The ability of all mutant env proteins to oligomerize was investigated by sucrose density centrifugation and chemical cross-linking. Conclusion: 1) Amino acids 99-129 of gp41 may be necessary for efficient oligomerization. 2) Amino acids 1-68 play a minor role in oligomerization. 3) Individual N-linked glycosylation sites and Cys residues are not critical for oligomerization. NOTES M.A.1194 HIV-I NONINFECTIOUS VIRUS-LIKE PARTICLES INTERFERE WITH VIRIONS Vzorov, Andrei; Bukrinskaya, Alice Central Institute of Advanced Training of physicians, Moscow, USSR Objective: Noninfectious HIV-I virus-like particles were obtained by infection of CV-I cells with two vaccinia viruses containing gag and env genes respectively. It was important to know whether the particles were able to interact with lymphoid cells and to interfere with infectious virions. Methods; HIV-I produced by H9/3B cells was used at the concentration 10 syncytiumforming units per 100 pl of medium. Cells C-8I66 were infected with the virus myxed with various concentrations of virus-like particles produced by 510 CV-I cells. As a control, virus-like particles containing only Gag protein were used. The interference was manifested by the absence of syncytia. Results: When virus-like particles in the medium diluted 1:10 and 1:20 were used, no syncytia were observed in 72 hours after infection with the myxture of the virus and virus-like particles. The dilutions 1:40 and more did not interfere with syncytia formation. The interference did not occur when virus-like particles containing only Gag protein were used. Conclusion: HIV-I virus- like particles besides Gag protein contain Env protein on their surface which recognise cellular CD 4 receptor and interact with it. As a result, the particles could interfere with infectious virions. So far, the noninfectious lIV-I virus-like particles could be used as a tool for inhibition of natural HIV infection. M.A.1195 PRODUCTION OF HIGHLY IMMUNOGENIC RECOMBINANT HIV-1 INTEGRATION PROTEIN (INT) BY RECOMBINANT VACCINIA VIRUS AND E. COL Tachibana Yoichi*, Kojima A**, Kobayashi T***, Yasuda A*, Sata T**, Kurata T** * NIPPON ZEON Co., Ltd., Kanagawa, Japan; ** Department of Pathology, National Institute of Health, Tokyo, Japan; *** UBE INDUSTRIES, LTD., Yamaguchi, Japan Objective: As a more precise diagnostic alternative to the present analysis of HIV-1 p24, gp41, gpl60, p66, p55, etc., in HIV-infected individuals, and to explore the possibility of the anti-HIV drugs which block the integration stage of the HIV-1 life cycle, we produced recombinant INT with both recombinant vaccinia virus (RVV) and E. coli which expresses INT. Methods: 1) An insertion containing only INT encoded by the 3'-end ofpol gene was isolated from pBH10 of the HTLVIIB isolate of HIV-1 by site-directed mutagenesis. 2) Recombinant Vaccinia Virus (RVV): The expression plasmid pAK-INT was derived from the vector pAK10, which encodes vaccinia promotor p7.5k. 3) Recombinant E. coli (E. coli-INT): The expression plasmid pYT-INT, which expresses INT as the fusion protein with the C terminus of enzyme protein was constructed. The fusion INT was absorbed on an affinity collum, and then partially purified by cleave of the fusion site. 4) The expression of recombinant integration protein was determined by western immunoblot analysis (WB) with anti-HIV-1 positive human sera. Results and Conclusion: RVV-INT was revealed as 34-kDa protein by WB, and E. coli-INT was exposed as degradation product of the full length of protein (34-kDa) on SDS-PAGE and by WB. These recombinant proteins were high reactivity to over 95% of HIV-infected individuals' (AC) sera by WB. We will also discuss anti-INT monoclonal antibodies. We have characterized biological functions of recombinant HIV INT. The reproducibility of evaluation of HIV antigen and anti-HIV antibody by screening with the recombinant products shows promise for the future. C>1 C>

Page  141 M.A.1196 NATURAL HIV-I 3'-ORF ACTS AS A POSITIVE REGULATOR FOR VIRAL REPLICATION IN PRIMARY HUMAN LYMPHOCYTES De Ronde, Anthony; Klaver, Bep; Keulen, Wilco; Goudsmit, Jaap Human Retrovirus Laboratory, University of Amsterdam, Meibergdreef 15, 1105 AZ Amsterdam, The Netherlands. Objective: To elucidate the function of HIV-1 3'-ORF in an environment mimicking as closely as possible its natural one. Methods: Six HIV-1 3'-ORF sequences were isolated from HIV- i infected individuals by nested PCR and cloned in an HXB-2 molecular clone background. Chimeric molecular clones were transfected into T-cell lines and primary human lymphocytes. Replication was assayed by p24 production. Results: Upon transfection in Jurkat and SupTI T-cell lines molecular clones containing HIV-1 3'-ORF sequences directly derived from HIV-1 infected individuals replicate with the same kinetics as the corresponding viruses with an in all reading frames interrupted 3'-ORF. Interestingly after transfection in primary human lymphocytes the onset of replication of viruses with an intact HIV-I 3'-ORF appeared faster than for the corresponding viruses without an intact 3'-ORF. The eventual level of p24 production was also higher for the 3'-ORF+ virus than for its corresponding 3'-ORF- virus. Discussion and Conclusions: Our data indicate that the HIV- 3Y-ORF has a positive regulatory role in primary human lymphocytes, but not in T-cell lines. It is only in primary human lymphocytes that the positive regulatory properties of 3'-ORF are revealed. This stresses the importance of the use of in vitro systems which as closely as possible resemble the natural situation. To mimic the natural situation even more closely, experiments are in progress which use the HIV-1 background of molecular clones isolated from primary human lymphocytes with either an non-syncytium or a syncytium inducing phenotype. NOTES M.A.1197 Heptameric CD4-C4bp fusion proteins efficiently block HIV-1 Pasek, Mark: Burrus, B.; Crimmins, M.; Kirchhausen, T.*; Liu, T.; Miatkowski, K.; Meier, W.; Turner, S.; and Winkler, G. Biogen, Inc, Cambridge, MA, USA; *Harvard Medical School, Boston, MA, USA We have constted heptameric CD4 molecules, CD4(1-187) - C4bp (4SCR), by fusing the N-terminal 187 amino acidsfrom human CD4 to proline 248 at the amino terminal edge of the fourth short consensus repeat of human complement 4 binding protein. We have expressed this molecule both in COS and CHO cells. The polypeptide chain covalently assembles into heptameric form, presumably thrugh the unique carboxy terminal domain of C4bp. Gel filtration on TSK estimates that the molecular weight is approximately 500 kd. This is consistent with assembly resembling that of authentic C4bp, even in the absence of C4bp's short arm. Spider-like molecules were seen after rotary shadowing in the electron microscope. Conditioned culture fluid and purified CD4 heptamer were tested in a HIV-1 microreplication assay. CD4 spider blocks HV-1 (IIB) infection ofC8166 cells at 100 pM; in contrast, sCD4 blocks at 10 nM. Similar results were seen with patient virus. No immunosuppession was observed when tested in (1) a PBL proliferation assay and (2) an assay measuring lymphokine production by alloreactive T cell clones. The assembly machinery of C4bp may be of general utility in the construction of multimerics. NOTES I loo r^ M.A.1198 KYNETICS OF HHV-6 TRANSACTIVATION OF HIV-i LTR AND EFFECT OF HIV-1 TAT ON HHV-6 REPLICATION. Di Luca Dario, Secchiero Paola, Bovenzi Pasqualina and Cassai Enzo. Institute of Microbiology, University of Ferrara, FERRARA, Italy. Objective: This research focused upon: 1) the kynetics of IIHV-6 transactivation of HTV LTR and 2) the interactions between HHV-6 and tat. Methods: The kynetics was studied in the human T cell line H938, containing the reporter gene CAT driven by the HIV LTR. The interactions between HHV-6 and tat were studied in clones of Jurkat cells, expressing constitutively the tat protein. Results: 1) Three distinct phases of HHV-6-induced activation of HIV LTR are present at 6-8, 22-24 and 36-38 hours p.i. At 48 hours p.i. the transactivation reaches a plateau. All the phases occur even in the absence of viral DNA replication, and most likely represent early viral,functions. 2) HHV-6 and tat interact synergically in LTR activation, but tat expression significantly inhibits HIHV-6 replication. Conclusions; Viral functions expressed early in HHV-6 infection transactivate HIV LTR. The data confirm the hypothesis that reactivation of latent HHV-6 could contribute to AIDS development in 1IlV seropositives. HIowever, it could be difficult to detect HHV-6 in these cases, since. aHIV expression inhibits HHV-6 replication. M.A.1199

Page  142 M.A.1200 MOLECULAR CLONING OF THE GENE FOR A CRYPTOSPORIDIUM SPOROZOITE SURFACE PROTEIN Petersen. Carolyn; Gut, J; Leech, JH; Nelson, RG. Parasitology Laboratory, San Francisco General Hospital and Departments of Medicine and Pharmaceutical Chemistry, University of Califoria, San Francisco, CA, USA. Objective: To clone the gene for a surface glycoprotein of Cryptosporidium sporozoites that is the target of a monoclonal antibody (Mab 10C6) that reacts with intracelular merozoites and the surface of sporozoites, causes shedding of the sporozoite surface coat and inhibits sporozoite invasion of Mardin-Darby Canine Kidney (MDCK) cells in vitro. Methods: A lambda gt 11 gDNA library was screened with polyclonal mouse antisera raised to sporozoites and oocysts. Positive clones were plaque purified and used to prepare recombinant eluted antibody (REA) from the polyclonal antisera. Results: REA from four clones reacted with a >300 kD protein on Western blot and exhibited an IFA pattern identical to that of Mab 10C6 on fixed sporozoites. REA from all four clones reacted on Western blot with the > 300 kD protein immunoprecipitated by Mab 10C6. Following N-glycosidase F treatment, REA from the clones recognized multiple Cryptosporidium proteins of Mr 150-190, but Mab 10C6 did not react with any. Conclusions: The gene for a potentially protective antigen on the surface of Cryptosporidium sporozoites has been cloned and encodes a heavily glycosylated protein. The Mab 10C6 epitope requires N-linked glycosylation. NOTES M.A.1201 DIFFERENTIAL IMMUNOLOGICAL FEATURES IN PCR-HIV+ AND PCR-HIV- HEMOPHILIA A PATIENTS.Cdstiana Chelucci, Gringed A*., Macioce G.. Vulcano F., Mariani G., Testa U. Bulgarinl D., Mannucci P.M.*, Hassan H.J., Peschle C. Dept. of Hematology-Oncology, Istituto Superiore di Sanita, Rome, Italy,*A. Bianchi-Bonomi, Hemophilia and Thrombosis Center, University of Milan, Italy. Objective: Polymerase Chain Reaction (PCR) technique was used to analyze the presence of HIV-1 sequences in DNA from two matched groups of 26 seronegative and 18 seropositive hemophilia A patients, equally exposed in the 1981-85 period to not virus-inactivated FVIII concentrates (>75% of the batches coincided in the two groups). Furthermore, we have analyzed a variety of immunological parameters in these hemophiliacs population, as well as in normal controls, in an attempt to identify a possible correlation between these parameters and HIV-1 infection. Methods:1 jg of DNA from lymphocytes and monocytes was subjected to amplification by PCR using primers for two conserved regions of 'gag" and 'env" genes. Hybridization signals were compared to those of a standard curve obtained after serial diluitions of H9 III DNA. Peripheral blood subpopulations were double staired using different MoAbs conjugated with two different fluorochromes and analyzed by a fluorescent activated cell analyzer. Results and Conclusion: No hybridization signals were detected in all seronegative patients after twofold 30 cycles amplification in both lymphocytes and monocytes. In all seropositive ones a clear hybridization signal was observed after the first 30 cycles. Seronegative patients showed a normal number of CD4+ lymphocytesj while a marked reduction was observed in the seropositive hemophiliacs; CD8+ lymphocytes were increased in both groups. Circulating T lymphocytes expressing the HLA-DR antigen are markedly increased in seronegative as well as seropositive patients. A pronounced reduction of circulating CDC56s/CD NK lymphocytes was observed in the seropositive hemophiliacs, while in the seronegative subjects the levels were normal. (NK lymphocytes with CD56+/CD3+ phenotype were at normal level in both seronegative and seropositive patients). A slight increase in the y/1 TCR+ T lymphocytes was observed in both populations of hemophiliacs.The marked and selective reduction of CD56+/CD3- lymphocytes in seropositive hemophiliacs may represent a key pathogenetic factor, which facilitates HIV-1 infection in hemophiliacs and perhaps other individuals exposed to HIV-1 contact. NOTES M.A.1202 THE CLEAVAGE OF THE BAIT REGION OF (2-MACROGLOBULIN (a2M) BY HIV-1 PROTEýSE (PR) Billich, A., Mann, K., Meier, Ute-Christiane*, Schramm, * H.J. ** Sandoz-Forschungsinstitut, Vienna, Austria; Max-Planck-Institut for Biochemie, Martinsried bei Miinchen, Germany. Objective: The specificity of PR is known from the cleavage of the ultiprotein of HIV and from other cleavages. The prevalent opinion is hat only extended inter-domain segments are cleaved. Since OýM ontains an accessible "bait" region we used it to test this asumption. A second question was whether the C-terminal domain of a2M an be cleaved off with PR as it can be done with papain and Lys-C. Mthods; The methylamine form of x2M (185 kd) was incubated with PR at H 5.8 and the resulting 90 kd band sequenced. There were no fragments ower than 90 kd and only traces of fragments with higher mass. Results and conclusions: The bait region of O2M was cleaved in position f.-Y. The same bond is also cleaved by chymosin and cathepsin G. his corresponds to the trait of PR to cleave after Phe or Tyr, but a cissile F-Y bond has not been described, so far. Here, both the S1 and 1' subsites of PR are occupied by aromatic side-chains. Phe removal of the receptor-binding C-terminal domain was not found in spite of the Phe-Pro sequence in the corresponding inter-domain stretch. This rules out that PR released from infected cells might prevent the internalization of O M and interfere with its possible functions as protease trap and carrier of cytokines. SM.A.1203 HIV-1 PROTEASE (PR) CLEAVES CALMODUUN (CaM) AND THE a CHAIN OF PM.A.1 PHOSPI)RYLASE-KINA (PK).. SBilich, A., Daube, Helmut, Iann, K., Schramm, H.J. Sandoz-Forschungsinstitut, Vienna, Austria; Max-Planck-lnstitut fGr Biochemie, Martinsried bei MOnchen, Germany. Objective: The specificity of PR Is known from the cleavage of the pol-gag polyprotein of HIV and from other cleavages found in multi-domain proteins. Since PR may also act on cellular proteins, it is necessary to learn more about the conditions of specific cleavage. PK, a large multi-protein complex was used as substrate. Methods: PK and CaM were incubated with HIV-1 PR at pH 5.8 (1). The cleavage pattern was analyzed by SDS-electrophoresis. The new bands appearing after incubation were sequenced. Results and Conclusions: Only two of the four subunlts of PK are cleaved to a measurable extent: the a subunit (138 kd) and CaM, the integral 8 subunit of PK. The shorter B chain (125 kd), which Is homologous to the a chain, does not seem to be altered, however. This Indicates cleavage in one of the 2 large sequence Insertions in the a chain containing the Phe-Pro sequence and other possibly cleavable bonds. The obtained sequences are given. The y subunit possessing the kinase activity is not cleaved. CaM alone is also rapidly cleaved by PR yielding a doublet of bands near 7 kd. This is not surprising since CaM contains a Phe-Pro bond in position 65,66. CaM is involved in many regulatory mechanisms. The finding, therefore, may have Importance for the pathogenic mechanism of HIV. (1) BillIch, A. et al. (1990) Biol. Chem. Hoppe-Seyler 371, 265-272. c> a c> C> C>

Page  143 M.A.1204 DOES THE HIV NEF PROTEIN BIND GUANINE NUCLEOTIDES? W.A._ _olber, Vera; Rensland, H.; Brandmeier B.; Wittinghofer, F.; Kalbitzer II.-R. Max-Planck-Institute for Medical Research, Heidelberg,F.R.G. Objective: Biochemical and biophysical characterisation of recombinant nef (negative factor) gene product. Methods: The nef gene of HIV-1, isolate LAV1, was expressed in E.coli under the control of the tao promoter. The recombinant protein is found mainly in the soluble fraction of the bacterial lysate, therefore the purification can be performed under nondenaturing conditions. The isolated Nef protein was employed to GTP-, GDP-binding and GTPaseassays and fluorescence experiments, using RAS p21 as positive control. Proton magnetic resonance measurements (NMR) were done at 500 MHz. Results: The NMR experiments show that the purified protein is well folded. Freshly prepared Nef has not bound any nucleotides as proved by HPLC. It shows neither GTP-, GDP-binding nore GTPase activity in our test systems. In contrast to other guanine nucleotide binding proteins fluorescence change was observed when incubating Nef with an fluorescence labeled GTP analogue. Furthermore sequence motifs of Nef postulated to be involved in nucleotide binding cannot easily be accomodated by similar motifs in the threedimensional structure of RAS p21. Conclusion: The results obtained so far show that the nef gene product is most probably not a guanine nucleotide binding protein. Since this is contradictory to earlier results the properties of the protein will be further investigated. NOTES M.A.1205 DIR~BC SQE NG AS A TOOL tR HIV-1 GioG~C P Saldnen, ka Leinikki, P, Ayebunie, S, Johansson, B and Strannegard, O 3 *HIV-UJit, National Public Health Institute, Helsinki, Finlad., *p of Virology, the Central Microbiological Laboratory of Stockolm County Council. Objective: Tb develop a phylogenetic tree analysis that can be applied to epidemiological and pathogenesis studies. Methods: Three primer nested CR was used to amplify regicns of HIV-1 from uncultured hutan white blood oells. The primats were selected frn conserved regions of tihe ag and env genes. Biotin was inrorporated into the amplified NA which was then captured onto avidin coated polystyrene beads. This facilitated purification and seiqueing steps. Results: The sequences were aligned with the previously known sequences of the Los Alanos HIV database. Usirg the phylogeny analysis program pakage PHYLIP a phylogenetic tree of the seque~es was cxnstructed. Geographically distant viruses form distinct genotypes as clusters in the evolutionary tree. Clustering was similar using either ag or ane sequenes. Tree patterns were similar using two methods of phylogeny analysis, te parsimony and maximum a liklihood methods. Discussion and Conclusions: Repeated sequencing from the same batch of cells indicated that at a given tine heterogeneity of viral sequence in lymphocytes is not a prominent feature. Different patients had sequences that varied from oneanother at a varying degree. The origin of the virus could still be traced back even three years after infection. Genotyping of proviral HIV-1 UNA from patient cells bydirect sequening seems suitable for epidemiological and pathogenetic studies. NOTES SIMPAIRED HIV REPLICATION IN CEMss CELLS BEARING A NON-PRODUCER HIV GENOME M.A.1206 ENDOWED WITH HOMOLOGOUS VIRAL INTERFERENCE Federico, Maurizio; Taddeo, B.; Titti, F.; Carlini, F.; Orecchia, A.; Verani, P.; Rossi, G.B. Lab. Virology, Istituto Superiore di Sanita, Rome, Italy. Objective: To determine if a non-producer HIV-1 genome inserted in a retroviral vector and transferred in human CD4+ cells is able to inhibit the replication of a superinfecting HIV-1. Methods: An Hut-78 cell clone (F12) infected by a non-producer HIV-1 variant showed a complete resistance to HIV-1, HIV-2 or SIV superinfection (Federico et al., AIDS Res Hum Retr, 1989, 5: 385). The F12/HTV genome was inserted in the pLj retroviral vector, and transfected in: i) PA317 amphotropic packaging cells to recover recombinant retrovirus able to infect human cells; ii) CD4+ human CEMss cells. Results: In CEMss clones either infected by the recombinant retrovirus carrying the F12/BIV-pLJ construct or directly transfected, similar amounts of F12/HIV specific proteins were produced, even if at a lesser extent than in parental F12 clone. Results about the HIV-1 superinfection of both kind of CEMss clones are superimposable, demonstrating: i) a reduction of 60-80% of the HIV-1 release in CEMss clones harboring the HIV/F12 genome as compared to the CEMss clones carrying only the pLj vector; ii) a significant improvement (>1.5 logs) of HIV yield reduction, when CEMss clones are maintained under the G418 pressure. Conclusions: A role of the non producer HIV-1 variant in the homologous viral interference occurring in Hut-78 cell clone has been demonstrated. To improve the interfering action of the F12/HIV pLJ construct, an enhancement of its transcriptional activity will be attempted, e.g. replacing the MLV-LTRa of the pLj retroviral vector with the transactivable promoters of HIV. SSEQUENCE ANALYSIS OF AN HIV-1 PROVIRAL DNA FROM A NON PRODUCER CHRONICALLY M.A.1207 INFECTED HUT-78 CELLULAR CLONE Carlini, Francesca*; Federico, M.*; Equestre, M.*; Ricci, S.**; Ratti, G.**; Verani, P.*; Rossi, G.B.*. *Lab. Virology, Istituto Superiore di Sanita, Rome; **Centro Ricerche Sclavo, Siena, Italy. Objective: To correlate the peculiar features of a non producer HIV-1-infected cell clone (F12) (Federico at al. AIDS Res Hum Retr, 1989, 5: 385) with possible genomic mutations. F12 clone, in spite of the presence of an integrated full-length HIV-1 provirus, does not release virus particles and shows an altered protein pattern: decreased and poorly cleaved p55 gg precursor, uncleaved gpl60 env precursor and a unique p19. In addition, F12 is fully resistant to HIVs superinfection. Methods: Sequence analysis has been performed on the proviral genome cloned from an SstI genomic library in pUC 19 with the chain termination DNA sequencing method. The results have been analyzed with the Sequence Analysis Software Package of the Genetic Computers Group (Wisconsin University) and compared using the HIV-HXB2 as reference. Results: Most mutations found in F12 clone accumulate mainly in U3-LTR region (where a 42bp repetition sequence has been observed), in the vif gene and RT coding region of the pol gene. Tat, rev and nef genes do not differ from the reference and the gpl60 and p55 cleavage sites are well conserved. Conclusions: Although the repetition sequence in U3-LTR region and the changes in RT aminoaeid sequence do not appear to correlate with any of the major features of the F12 clone, an alterated vif could account for the inhability of this clone to release viral particles. A direct analysis by recombinant exchange of structural and functional genomic domains between F12 clone and a producing strain will be useful to clarify this issue.

Page  144 M.A.1208 CLEAVAGE OF HIV mRNA BY HAMMERHEAD RIBOZYMES Crisell, Paul and James, W. Sir William Dunn School of Pathology, University of Oxford, U.K Objectives: to produce and analyse hammerhead-model ribozymes that cleave HIV mRNA; to establish optimum conditions for cleavage in vitro; to analyse the relationship between flanking sequence length and kinetic properties; to compare in vitro cleavage with anti-HIV activity in ivo. Methods:22 bases correspondingtothe hammerhead catalytic motifwereintroduced into an anlisense RNA-encoding (AR) gene by oligonucleotide-drected mutagenesis, The target site is a 5'GUC3' sequence within the first coding exon of tat (5' to the initiation of translation) and the vpr coding sequence. The ribozyme-containing antisense RNA and the substrate RNA were expressed by T7 RNA polymerase-mediated in vitro transcription from pSPT19-derived plasmids using aCP CTP. Reactions were carried out in 20mM Mg" at pH8.0, 50"C for 1 hour and analysed by electrophoresis through 8% acrylamide under denaturing conditions (7M urea, 25% formamide). AR-ribozyme genes and their sense counterparts were stably introduced into Jurkat cells using retroviral expression vectors and were expressed from the retroviral LTR. The cells were subsequently challenged with HIV and viral replication followed using a p24 antigen ELISA and the polymerase chain reaction. Results: The following parameters of the reaction condition were varied: temperature, 0 - 75~C; time, 30sec - 120min; spermidine concentration, 0 - 20mM; formamide concentration, 0 - 70%; Using rbozyme at 128nm and substrate at 384nm, the reaction was essentially complete within 30min.However the ribozyme did not participate in greater than a stoichiometric number of cleavage reactions because of the strong hydrogen-bonding between ribozyme and products consequently molecules with shorter complementary regions were analysed. The reaction was enhanced by 15% formamide suggesting that partial denaturing conditions enable molecules in stable but inactive conformations to refold into the catalytic conformation, Conclusions: The HIV mRNA-cleaving activity of artificial ribozymes may provide a useful therapeutic tool against HIV. NOTES M.A.1209 Neutralization Loop Vrariability i iHIV-1 Islolates from the Kinshasg Area H.-G.Guo, D. Zagur, D. Waters, L. Jagodzinski, J. Fitzgibbon, R.C. Gallo, and M. Reitz Lofstra d Laboratories, GaithersburgMD, USA, 2Pierre and Marie Curie Institute, Paris, FrancC, PRI, FCRC, Frederick, MD, USA, Bliotech Research Labpratories, Rockville, MD, USA, Robt. Wood Johnson Medical School, Piscataway, NJ, USA, LTCB, NCI, NIH, Bethesda, MD, USA We analyzed a central portion of the jY gene of a number of isolates of HIV-1 collected in 1987 from Kinshasa, Zaire and its surrounding regions. The env gene sequences were generated by PCR and included the PB1 loop, a region which is frequently recognized by type specific neutralizing antibodies in hyperimmune sera, and downstream sequences. 20 of the 32 amino acids within the loop were invariant among seven isolates, compared to 12 and 8 consensus amino acids among six North American and six previous Zairean isolates previously analyzed and collected at various times and locations. The PB1 loop was more conserved in general than the region immediately downstream, suggesting functional constraints on change in this region. The data suggest that in a given area within a limited time span, the PB1 area variability is more limited than apparent from previously available sequence data. NOTES M.A.1210 USE OF SYNTHETIC GENETIC BAR CODES IN POLYMERASE CHAIN REACTION TO FACILITATE THE STUDY OF HIV GENETIC VARIABILITY Ou, Chin-Yih; Luo, C.-C.; Bandea, C.; Villamarzo, Y.; Moore, J. L.; de la Torre, N.; Schochetman, G. Centers for Disease Control, Atlanta, GA, USA Objective: To facilitate sequence determination of HIV. Methods: HIV sequences from the peripheral blood mononuclear cells (PBMCs) of infected persons were first amplified through polymerase chain reactions (PCR). Small aliquots of the amplified products were reamplified using a second primer pair with extended sequences specific for HIV at the 3' end and genetic bar codes (GBCs) and restriction endonuclease sites at the 5' end. Amplified DNA from iseveral patients were combined for cloning and sequencing. Results: The primary PCR products of HIV DNA from patients' PBMCS usually contain many non-HIV sequences which hinder the subsequent cloning process. A short extension of a few HIV-specific nucleotides at the 3' end of 1 of the 2 primers in the secondary PCR is sufficient to yield HIV-specific DNA. The presence of GBCs in the secondary primers provides a means to identify the original patients from whom the HIV sequences were derived. Using this method, we were able to pool amplified DNA from 10 or more persons and correctly identify each of their sources. Conclusions: A novel approach is developed to facilitate the sequence determination from a large number of patients for the study of HIV genetic variability, transmission, and evolution. M.A.1211 A POSSIBLE ROLE OF THE P18 PART OF GAG IN VARIATION OF HIV1 CYTOPATHIC EFFECT de MAREUIL Jean *, SALAUN D.*, KUTINOVA L**, VONKA V.**, CHERMANN J.C *, and HIRSCH 1.* *INSERM Unite 322, Marsellle, France, **INST. Sera and Vaccines, Prague. Czechoslovakia Objectives: To elucidate a role of genetic elements localized In HIV1 NDK fragment BssHII/Spel (nt 255-1042) for HIV1 cytopathogenicity. Methods: Recombinant provirus molecules between HIV1 BRU and HIV1 NDK were constructed using conservative BssHII, Spel, EcoRI and Xhol sites. HIV1 NDK derived fragment BsHlII/Spel was further modified by PCR. Recombinant vaccinia virus expressing complete env gene of HIV1 BRU or HIV1 NDK were constructed. Results: Characteristic ability of HIV1 NDK to form large syncytium in MT4 cells was lost after replacing the BssHIl/Spel segment in its provirus by HIV1 BRU sequence. A minimal, transiently defined, HIV1 NDK sequence sufficient to keep its highly fusogenic phenotype consisted of combination of HIV1 NDK fragments BssHil/Spel and EcoRl/Xhol inserted In HIV1 BRU provirus. HIV1 NDK sequence between nucleotides 721-1042 of this provirus, corresponding to N'-terminal part of p25 Gag, was replaced by sequence of HIV1 BRU. Resulting recombinant virus retained the highly fusogenic phenotype of HIV1 NDK. Interaction of pl8 protein of Gag, representing the most external protein of the nudeocapside, with the transmembrane glycoprotein Gp41 could be therefore responsible for formation of the large syncytlum. To analyse the fusogenic potential of HIV1 envelope glycoprotein Itself, In the absence of genetic elements from BssHII/Spel fragment, vaccinia virus recombinants expressing env gene of HIV1 BRU and HIV1 NDK were used to induce syncytia. Both vaccinia virus recombinants were forming a similar size syncytia. Discussion and Conclusions: Genetic elements localized between nucleotides 255-720 of HIV1 NDK RNA were necessary, in addition to enveloppe glycoprotein, for the full expression of HIV1 NDK fusogenic phenotype. Because a highly conservative RNA leader is present between nucleoldes 255-333, the data suggest that p18 protein of Gag could be involved in variation of HIV1 cytopathogenicity. a cz C> C' C> C>

Page  145 M.A.1212 THE HIV-I EE GENE PRODUCT INHIBITS TRANSLATION I VITRO Poulin. Louise*; Darveau, A.**; Levy, J. A.*** *Centre Hospitalier de l'Universit6 Laval, Qudbec, Qe, Canada, **Canadian Red Crc Society, Quebec Center, Qc, Canada,***University of California, San Francisco, CA, USA. Objective: To analyze the function of HIV-1 nefgene product in the negative regulation of HIV replication. Methods: The HIV-1SF2 nef gene has been subcloned into a transcription vector and used to produce a mRNA. The generated messenger has then been translated into a micrococcal nuclease-treat. reticulocyte lysate. The translational efficiency of the n f message was compared to other mRNAs und standard testing conditions. In addition, translation of globin and CAT mRNAs was further analyzed presence of the nef gene product. Resulta: Preliminary tests have shown that translational efficiency of nef mRNA in reticulocyte lysate very weak. Moreover, time course analysis have demonstrated that 10 minutes after the addition ofL mRNA in the reticulocyte lysate, its translation rate is drastically decreased, independent of its stabilit In addition, translation of either globin or CAT mRNA in the presence of nef translation product h revealed a strong inhibition of the protein synthesis. In contrast, no significant inhibitory effect observed when globin and CAT were cotranslated in similar conditions. Furthermore, experiment usil anti-sens nef mRNA are in progress. Discussion and Conclusions: Our results suggest that the nUf gene product can inhibit the translation Yi1o. In a complementary experiment, this translation inhibition was correlated with the phosphorylatii of a p46 protein of the retic lysate in an in vitro kinase assay. The ability of af to inhibit translation cou have an important physiological significance for the control of viral and host gene expression. NOTES M.A.1213 ABSENCE OF TRUNCATED TRANSMEMBRANE PROTEIN IN TWO MACAQUES AT THE TERMINAL STAGE OF THE DISEASE. Fossati I1* Courgnaud V**, Brechot C*, Sonigo P*. *Institut Cochin de Gnetique mol6culaire,,**U75, Necker, Paris, France Objective: Presence of a stop codon that truncates the transmembrane protein (THP) of HIV and SIV isolates has been described by different authors. The truncated form, in the case of SIV isolate has only been found until now, in vitro, after culture in human PBMC or human CD4 positive cells.. As "rapid high" variants of HIV are selected in symptomatic AIDS patients, we wanted to investigate if "rapid" SIV with truncated TMP could be found in two macaques at the terminal stage of the disease. The animals were infected by an infectious clone of SIVmac 239 (R. Desrosiers et al) Methods: PBMC'DNA was amplified by PCR. The primers were flanking the supposed location of the stop codon. The PCR products (obtained after two different amplifications for each sample) were cloned and sequenced. Results: Fifty clones for each macaque, were sequenced and none of them showed the presence of a stop codon. So the "truncated" genotype is not enough, or not at all, represented in circulating lymphocytes of the two studied animals, to be detected in cur experiments. Discussion and conclusion: Those results agree with what has been already published about the counterselection of the "truncated" form in vivo, in macaques' PBMC. However, this does not exclude that they might exist in vive in other cellular types, as it was suggested by Wcrner et al. Desrosiers R.C.et al (1989). J. Virol 63: 4709-4714. NOTES! ~I ^ 00 ^1Z M.A.1214 SQUENCE COMPARISON OF TWO SIVSMM ISOLATES WITH DIFFERENT BIOLOGICAL PROPERTIES. Courgnaud V, Laure F, Montagnier L Fultz P, Brechot C and Sonigo P. U75 Necker, ICGM Genetique des virus, Diagnostics Pasteur, Institut Pasteur, Paris, France, University of Alabama at Birmingham, Birmingham, USA. Objnctive: To compare the nucleotide sequence of a lethal infectious molecular clone derived from isolate SIV/SMM-Pbjl4 (Dewhurst et al) with the sequence of the original non pathogenic isolate SIV/SMM-9. SIV/SMM-Pbjl4 was generated by mutation of SIV/SMM-9 after only one passage in vivo (Fultz et al). Methods; Entire viral genomes were amplified by multiple independent PCR. The PCR products were cloned and sequenced. Resultat The number of amino acid variations which could be involved in the lethal phenotype of Pbj 14 are: GAG: 2, POL: 12, VPR: 2, TAT: 3, REV: 1, ENV: 11 and NEF: 5. In addition, two insertions/duplications were observed in the lethal infectious clone (one in LTR and one in ENV), that are never observed in the original non pathogenic SIV/SMM-9. Discussion and conclusion: Guided by these results, we have constructed chimaeric or mutagenized infectious clones to localize which determinants are involved in the lethal phenotype. Dewhurst et al. Nature 345, 636-640 (1990) Fultz et al. AIDS Res. Hum. Retroviruses 5, 397-409 (1989). M.A.1215 MUTATIONS IN THE HIV-1 REV NUCLEAR LOCALISATION SIGNAL ABROGATE BINDING TO REV RESPONSIVE ELEMENT (RRE) RNA Berger, Johannes*; Aepinus, Christian; Dubrovnik, Marika*; Kreusel, Jutta *; Fleckenstein, Bernhard*; Hauber, Joachim'; Boehnlein,Ernst* *SANDOZ Resarch Institute, Vienna, Austria * Institut fOr klinische Virologie, University of Erlangen, Erlangen, Germany HIV-1 replication depends on the expression of trans-regulatory genes (Tat, Rev) encoded in the 3'part of the retroviral genome. HIV-1 Tat protein augments retroviral gene expression on a transcriptional and post-transcriptional level. Rev is required for the cytoplasmic translocation of incompletely-spliced retroviral mRNAs and the translational switch from regulatory (Tat, Rev) to structural proteins (Gag, Pol, Env). The activation domain of the HIV-1 Rev protein (RAD) is located around amino acid (aa) 80 and believed to interact with cellular factors necessary for Rev function in vivo. The other functional domain is important for RNA substrate binding and overlaps with the nuclear/nucleolar localisation signal (NLS). Our analysis defined a REV protein domain important for RRE binding in vitro to amino acids 31 -50. Mutations in this domain always resulted in the exclusion from the nucleoli. Furthermore, these mutants could not support Rev-dependent p24 Gag production in vivo. The observec tight correlation between subcellular compartimentalization and RNA binding in vitro indicates that this short stretch of amino acids may support two esssential functions required for HIV-1 replication.

Page  146 I-1 c\ M.A.1216 Investigation of V-I packaging. Geoffrey Harrison. and Andrew Lever. Communicable Diseases Unit, St George's Hospital Medical School, Cranmer Terrace, London. SW17 ORE. U-KI Objective: Determination of the secondary structure of the packaging signal of HIV-1. The full-length unspliced mRNA of retroviruses codes for the gag and pol genes and is also the species packaged as a dimer into the exported virion particles as the genome RNA. Sequences essential for efficient packaging of this RNA in HIV-1 are situated between the 5' long terminal repeat and the gag gene initiation codon and they are highly conserved between HIV isolates. This region is postulated to adopt a stable secondary structure which allows interaction with gag proteins and the formation of the virion core. Methods: Computer modelling and enzymatic or chemical modifications. The region has been examined by computer modelling and a prediction of a potentially stable secondary structure has been determined. RNA from this region has been transcribed in vitro and renatured, then modified, either by partial digestion with ribonucleases which have sequence specific cleavage patterns, (notably cutting at either double or single stranded regions) or by chemical methods. The modified RNA is then analysed by primer extension, looking for reverse transcription stop sites, indicative of secondary structure or strand termination. Discussion: Analysis of the renatured, modified RNA show patterns consistent with particular regions having a stable secondary structure. To a very large extent these results agree with the prediction made by computer modelling. A consensus model of this region is presented. NOTES M.A.1217 Bio:synthesis and processing of the Human ImmunodefiMT.A.1 c VItrus -ývvpe 1 structural proteins in infected ceil: the limits for virus growthv. t Z.ides. L. L 3el mova. S. Yu. Klyushnik, T. V. Vsel-vskaya, L.A. Rikova; The D. 1. Ivanovsky Institute of VI clogy, Moskow, 123098, USSR The svnthesis and metabolism of HIV-'1 structural,orroteins in persistently infected cells were studiedd bV pulse-chase protocol. The primary translational products. pr18O (gag-pol genes fusion protein). pIO60 (env) and p55 (gag) were synthesized at the relative molar rates of approximately 1:50 80, and the rates and extents of their proteolytic conversion into mature products were also different. The gpl20 oro-,eir was observed firstly after 0 min chasing in i mnn prelabelled cell extract and gpl60 cleavage was completed during 30 min of chasing. Thus about O of gpi60 molecules left very rapidly the cleavable c0.cc After cleavage both gpl60 and gpi20 underwent very gfficient (up to 100% of molecules) trimming of oligonar:Cose cores and after that unefficient (<C10) terminal glycosvlation. However only terminally glycosylated torins were seen within virus particles. Both prlSG and,5,5 proteirs were cleaved very efficiently although at.1.-htly different rates (with half-period about oO and 130:nin for prl80 and p55 respectively). Atout?.C% of gag gene and 5X of env gene products were revealed in virus parttcles. These results suggest several levels and events limiting of HIV replication in the system studiedNOTES 0\ 'a l~>k M.A.1218 RIBOZYMES AS ANTI-HIV AGENTS: INSERTION OF CIS AND TRANS RIBOZYMES INTO THE HIV-GENOME. Nina H. Lin, Boro Dropulic, Kenny Graham, Kuan-Teh Jeang. Laboratory of Molecular Microbiology, NIAID, NIH, Bethesda, MD, USA. Objective; To study the potential of ribozymes as anti-HIV agents. Methods; Standard molecular biology techniques were used. Results: Certain RNA molecules, called ribozymes, display enzymatic activities. Their unique secondary structure and sequence enable highly specific cleavage reactions. Two classes of ribozymes were designed and targeted to the human immunodeficiency virus type-1. We designed a cis ribozyme that can undergo self-cleavage and a trans ribozyme that can cleave a distal HIV-1 RNA sequence. These two types of ribozymes were functional in vitro and were placed into the HIV-i genome. Replication of the ribozyme containing viruses was substantially reduced, suggesting that these catalytic RNAs may prove to be useful therapeutic agents against HIV-1. Discussion: We have demonstrated that ribozymes can be used as effective anti-HIV-1 agents. M.A.1219 EXTRACELLULAR INTERACTION OF HHV-6 REPLICATION PRODUCTS ON HIV-1 O.A l ngradi, Joseph+, Ceccherini-Nelli, L.++, Pistello, M.++, Specter, S.+++, Friedman, H.+++, Bendinelli, M.++ +Institute of Microbiology, Semmelweis University of Medicine. Budapest, Hungary, +"Department of Biomedicine, University of Pisa, Pisa, Italy, 'Department of Medical Microbiology, University of South Florida, Tampa, FL, USA A potentiating effect, likely of the early gene products, of heterogeneous coinfections by several viruses, among them human herpesvirus type 6 /HHV-6/ (cytopathic on permissive cells such as CD4+ lymphocytes) on HIV-1 replication is well documented in vitro; leading to enhanced transcription, antigenic expression and cytopathogenicity of HIV. Since the nature of the mechanisms involved is poorly understood and since for HHV-6 it has been reported only the effect of intracellular coinfection,we investigated the possible activating role of soluble mediators released from HHV-6 infected cells. Virus free supernatants, harvested daily from HSB-2 cells acutely infected with HHV-6 until cytopathic effect /CPE/ was evident, were added to CEM-ss cells which were then challenged with HIV-1 and monitored for syncytia formation, reverse transcriptase activity ard p24 antigen production. Only supernatants obtained between 24 and 96 hours i.e. before manifest CPE, induced early and enhanced (up to 30-fold) expression and cytopathogenicity of HIV-1. Such activity was reduced in the supernatants harvested at late stages of HHV-6 infection or when the HHV-6 infected cells were treated with bacterial endotoxin /LPS/: interferon appeared to mediate the activity of LPS. Tumour necrosis factors or interleukin-2 seem not to have any role in these processes. The data show that not only intracellular, but also extracellular interactions between HIV-1 and HHV-6 are possible and damaging. C> C> > rj

Page  147 M.A.1220 LYSOPHAGOSOME FUSION OF MACROPHAGES FROM HtY IIFECTED PATIEITS. Estevez IXarJa lena; Pittis G; lannitelli P; Planes BI; Pirola D*; Sen L Academia lacional de Medicina de Buenos Aires and *Centro Diagnostico de Virus, Buenos Aires, Argentina. Objective: Previously we found that peripheral blood monocytes from HIV infected individuals have an impairaent to lyse candidas with a reduction in the generation of toxic oxygen aetabolites. The aim of the study was to evaluate the monocytic lysophagosome fusion, as a previous step of the oxygen independent microbicidal activity. Naterials and methods: Ye studied 20 HIV-infected subjects and 10 normal controls. The peripheral blood mononuclear cells were isolated and cultured for 48 hours on glass coverslip. O psonized zymozan were incubated with the adhered cells for 30 mln. The lysophagosome fusion was measured using acridine orange and determalning the percentage of fused particles (orange colour) over the total number of phagocytozed particles. Results: The % of fused particles were (x+SD) for HIV infected subjects 59+18 and for normal controls 7119. The difference between both groups studied was statistically significant with a p<0.05. Due to a considerable dispersion of the data in HIV infected subjects, next we investigated the relationship between the CD4/CD8 T-cell ratios and the percentages of lysophagosome fusion. A significant inverse linear correlation was found between both variables with a correlation coefficient of r=0.7579, p(<0,01. Conclusions: These data suggest that the impairment of the lysophagosome fusion of nonocytes in HIV infected subjects becomes evident when the CD4/CD8 T-lymphocyte ratic starts to decline at later stages of HIV chronic infection. NOTES M.A.1221 GENETIC RISK FACTORS FOR CLINICAL MANIFESTATION IN PEDIATRIC HIV INFECTION Stein, Zena;* Abrams E;"* Just J;*** Louis L**; Nicholas S**; King MC"**: *HIV Center for Clinical & Behavioral Studies, NYS Psychiatric Institute & Columbia University, New York, NY, USA; **Harlem Hospital Center & Columbia University, New York, NY, USA; ***University of California, Berkeley, CA., USA. Objective: To search for associations between clinical manifestation of illness and genotype of candidate sequences among infants found with active HIV infection following maternal transmission of the virus. Methods: 40 Black American infants infected through the maternal route, all with established HIV infection, were categorized into 2 clinical groups, corresponding to course of the disease, as moderate (P2A, P2C) and severe (P2B, P2D). For each subject, DNA was extracted from fresh blood and analyzed by Southern blot and PCR, with probe pDCH-1/Taql for 6 alleles of the candidate gene, HLA-DQA1. Results: HLA-DQA1 allelic frequencies detected by Southern Blot and PCR among the 40 subjects were: 0101 0102 0103 0201 0301 0401 N of chromosomes Moderate % 15 40 3 17 17 8 60 Severe % 25 30 10 5 20 10 20 Conclusions and Discussion: In these preliminary data, DQA1.0101 occurs more frequently among infants whose disease manifestations are severe than among those moderately affected. A similar pattern was observed among HIV infected Caucasian men who had more rapid progression of disease, compared to those who had a more indolent pattern. NOTES 0 NI 0 ^ \It h, oo F^4 t^ M.A.1222 PHAGUCYTo SISOF M I YOBALTERIUM-TUEiBE-LL S STIMULATES TRANSCRIPTION OF HUMAN IMMUNODEFICIENCY VIRUS(HIV) TYPES 1 AND 2 IN MACROPHAGES. Robin Shattock; Friedland, JS; Griffin, GE. Division of Communicable Diseases, Department of Cellular and Molecular Sciences, St. George's Hospital Medical School, Cranmer Terrace, London, SW17 ORE, England. Objective: HIV infects human macrophages and may remain latent for years before development of the Acquired Immunodeficiency Syndrome (AIDS). We have investigated the possibility that phagocytosis of a pathogen, in particular M. tuberculois known to be important in clinical AIDS, leads to activation of HIV. Methods: DNA plasmid constructs of the Long Terminal Repeat (LTR) promotor region of the HIV genome linked to the Chloramphenicol Acetyl Transferase (CAT) gene were transfected into 2 human monocyte cell lines (THP-1 and Monomac 6). Constructs included either HIV-lor HIV-2 LTR or a mutated HIV-1 LTR lacking the binding site for the regulatory nuclear trans activating factor KB. 24 hours after transfection cells were exposed to M.tuberculosis or other particulate stimuli for 24 hours prior to measurement of CAT activity and phagocytosis assessed by light microscopy. Experiments were repeated in the presence of monoclonal antibody to Tumour Necrosis Factor (TNF). Results: Phagocytosis resulted in up to a 10-fold stimulation of CAT activity in transfected THP-1 cells. M.tuberculosis increased activity more than zymosan granules or latex beads. Monomac 6 cells, a mature monocyte line, expressed higher basal level of CAT activity when transfected which increased after phagocytosis. Anti-TNF did not affect results and no activity was detected in cells transfected with the mutant LTR. Disscussion and conclusions: Phagocytosis by monocytes activates HIV 1 and 2 transcription by a mechanism involving the iB binding site of LTR and independent of TNF. Such enhancement of HIV transcription by intracellular pathogens may have important implications in clinical progression of HIV infection. M.A.1223 IDENTIFCATION AND CHARACTERIZATION OF HUMAN HERPES VIRUS 6 GENE SEGMENT THAT TRANSACTIVATE HIV Charles Wood; GENG, Y; BALACHANDRAN, N' University of Kansas, Lawrence, KS, USA, 'University of Kansas Medical Center, Kansas City, KS, USA. Objective: Human herpes virus 6 (HHV-6) is a recently isolated lymphotropic herpes virus. HHV-6 can productively co-infect HIV-infected CD4' T cells. It has been demonstrated that co-infection by both viruses can lead to accelerated cytopathic effects when compared to HIV or HHV-6 infection alone. The objective of the current study is to identify and characterize the HHV-6 gene segment(s) that may be responsible for the transactivation. Methods: Our approach was to obtain subgenomic clones of the 170kb HHV-6 genome. These cloned fragments were then tested for their ability to transactivate HIV promoter. Results: We have identified a cloned 22kb HHV-6 viral gene segment which may be responsible for the transactivation of HIV promoter. Transactivation can be demonstrated in different cell types such as human T cells and monkey fibroblast cells. We have mapped the target site for transactivation by the 22kb segment using different HIV promoter deletion constructs, and it is located in the NF-KB region of the HIV promoter. A mutation in the NF-cB site can abolish transactivation by this HHV-6 gene segment. Discussion and Conclusion: It is clear that HHV-6 can encode factors that may directly or indirectly activate HIV. Experiments are now in progress to further identify and characterize the HHV-6 gene(s) in the 22kb segment that is responsible for the interaction with and the transactivation of HIV promoter. K K K K fu fJ

Page  148 . M.A.1224 NEUTRALIZING MONOCLONAL ANTIBODIES TO HHV6 IDENTIFY VIRION 00 ENVELOPE GLYCOPRORTEINS. Campadelli-Fiume. G., A. Boscaro, S. Di Gaeta and L. FoATomasi, The Section of Microbiology and Virology, Department of Experimental Pathology, University of Bologna, Bologna, Italy Obje.cti.ve of this work was to characterize HHV6 glycoproteins and to infer their functions through the production of MAbs. Me-thtod6: BALB/c mice were immunized with lysates of cord blood lymphocytes infected with HHV6(U1102). Antibody secreting hybridomas were identified by ELISA and IFA. The reactive proteins were characterized by immunoprecipitation and Western blot, ReuawLt4 and DLAcu4Lono: Two neutralizing MAbs were produced. The reactive proteins were metabolically labeled with 3H-glucosamine and were designated as gp100 and gp112. Both undergo proteolytic cleavage. One of the two MAbs was neutralizing in the presence of Complement, whereas the other was neutralizing even in the absence of Complement, implying that the two glycoproteins are virion envelope components. The complement-independent MAb inhibited virus attachment to cells, indicating that the reactive glycoprotein mediates the initial events of HHV6 interaction with susceptible lymphocytes. The cell surface receptor(s) for HHV6 can now be identified. NOTES M.A.1225 ASSOCIATION BETWEEN SEMINAL WHITE BLOOD CELLS AND HIV-1 IN SEMEN Politch. Joseph*; Martinez, A.*; Abbott, A.*; Mayer, K.**; Padian, N.***; Seage, G.****; O'Brien, T.*****; Horsburgh, C.*****; and Anderson, D.* *Harvard Univ., Boston, MA; **Brown Univ., Providence, RI; ***Univ. of California, San Francisco, CA;****Boston Univ., Boston, MA; *****CDC, Atlanta, GA Objective - Seminal white blood cells (wbc) have been implicated as vectors in the sexual transmission of HIV-1. The purpose of this study was to characterize wbc in semen from HIV-1 seropositive men, and to determine if wbc types, concentrations and/or activation status are associated with a positive HIV-1 semen culture (HIV-14). Methods - Semen specimens and clinical information were obtained from 96 HIV-1 seropositive men. Seminal wbc were identified and quantified by immunohistology. IIIV-1 was cultured from cellular and cell-free fractions of semen on PHA-activated target cells and detected by ELISA. Results - Nine semen specimens were HIV-1+ and 18 were leukocytospermic (>1 x 106 wbc/ml). There was a higher prevalence of HIV-1 in leukocytospermic samples (Prevalence ratio-5; p<0.05). HIV-1+ samples had significantly higher concentrations of total wbc (7.4 vs. 1.0 x 106, p<0.05), macrophages (2.3 x 106 vs. 3.1 x 105, p<0.05), CD4+ lymphocytes (1.5 x 105 vs. 1.4 x 10', p<0.05) and activated T cells (1.1 x 106 vs. 5.8 x 104, p<0.05). AZT treatment was associated with a significant decrease in seminal wbc concentrations and in the prevalence of HIV-1 in semen. Discussion and Conclusions. HIV-1 in semen is associated with elevated levels of CD4+ lymphocytes, macrophages and activated T cells. The sexual transmission of 1IIV-1 may be enhanced by the presence and/or activation of these cells in the rPnrndiirt-vp tract. NOTES M.A.1226 COMPLEXITY OF GENETIC SUSCEPTIBILITY OF HUMAN CELLS TO HIV INFECTION. Cambiagqi C.:, Ramarli Dunia*, Mantero G.1. Primi D#.,Tridente G.*,and Accolla RS*. 'Istituto di Scienze Immunologiche e Servizio di Imaunopatologia, UniversitA di Verona,Policlinico Borgo Roma, Verona, Italy. #Consorzio n: ]e Biotecnologie, Brescia, Italy. The HIV specitic receptor is the CD4 molecule. However CD4 presence is necessary but not sufficient for virus full infecting capacity as demonstrated by the fact that not all CD4- cells in HIV+ individuals are infected and that murine cells trastected with human CD4 bind the virus but do not get infected. This suggests that other human factors are involved in the susceptibility to HIV infection. This aspect is interesting for understanding how infection is trasmitted and/or mantained in high risk subjects such as babies from affected mothers. In foetuses and babies lymphoid organs such as the thymus are full of CD4+ T cells precursors and yet only a reduced percentage of individuals are HIV-positive at birth. To study the possible existence and nature of these additional factors we produced somatic cell hybrids between human CD4-' normal T colls and murine thymic lymphoma BW5147. In this situation a preferential segregation of human chromosomes takes place. If additional factors are involved and encoded by distinct chromosomes one would expect they would segregate independently from CD4 in the hybrids. Human CD4- hybrids were treated with supernatant from HIV infected H9 cells and analyzed for HIV infection and integration by polymerase chain reaction using two primers of the gag gene. The results indicate that capacity to be infected correlates with the concomitant presence of CD4 and additional factors of human origin which segregate independently from CD4. The genetic location of these additional factors is under study. a1HV6 (HUMAN RPERsrRUS 6) IN LITPHOS AID BIV IMFEICTD IEMOPHILIACS M.A.1227 Ceocherini-Nelli. L.;Toreli, o. Maralca, R.'"Luppi, r.'; Pistello, MI I Ghilarducci, R.'; Batoni, G./; Cecconi N. o Panicucci, F., Bendinelli, M.. "University of Pisa, Pisa, University of Modena, Modena. Italy. Objectivei to assess the degree of association of HHV6 active infection with Hodgkin' (Hg) and non Hodgkin's (nHg) lymphomas and with HIV-infected hemophilic patients: HHV6 has been proposed as a potential cofactor in the onset and/or development of theas lymphoproliferative disorders and of AIDS. Methodst Sera from 94 lymphomas (31Hg and 63 nHg), 171 hemophiliacs well characterize clinically, immunologically and virologically (45 HIV+ at different clinical stages o infection and 126 HIV-), 67 chronic lymphatic leukemias, 689 HIV+ and HIV- I.V. dru abusers from different Italian regions, 86 homosexuals, 110 HIV+ and HIV- Africa inpatients, and 1063 healthy controls (563 military personnel; 500 blood donors) were examined by immunofluorescence using HSB2/HHV6 infected cells. PCR DNA analysis or Bglobin amplifiable samples was performed with pZVH14 HHV6 derived primers ancd contemporarily with HIV1 gag, pol and env primers. Resultst A significant difference (p <.001) in seroprevalence and antibody titres (al:40): was observed only between the groups of Hg lymphoma patients and healthy sera. No serological evidence of HHV6 active infection (reinfection or endogenous reactivation), was detected in HIV+ sera from hemophiliacs during a 6 years retrospective study (1985 -1990) if we exclude two seroconversions. By PCR analysis we found positive for HHV6 DNA sequences 3/25 Hg (all belonging to the same histotogical subgroup: nodular sclero-; sis/lymphocyte depleted) and 1/63 nHg. Two Hg were also positive (indicating active viral: replication) and of identical pattern by Southern blot, both polymorphic with rispect to the HHV6 prototype. Among the HIV+ hemophiliacs, PCR analysis showed 11/35 patients positive for both HIV and HHV6; these patients will be part of a prospective study of clinical evolution. Discussion and conclusions: The study is relevant to; a) the importance of HHV6 as cofactor of AIDS; b) the association and potentially the involvement of HHV6 in the pathogenesis of Hg; c) the pathogenesis of lymphoproliferative diseases in AIDS patients.

Page  149 M.A.1228 HEPATITs b VIRUS (HBV) X PROTEIN ACTS THROUGH NF-kB TRANSCRIPTION FACTORs A POSSIBLE MECHANISM FOR HBV-INDUCED STIMULATION OF A LATENT HIV GENOME. G. Natolit, C. Balsanot, M.L. Avantaggiatit, E.De Marziot, E. Elfassi#, M.Levrero* *Fond. A. Cesaloino - I Clinica Medica - University of Rome - Italy, #Unite' d'Iamunologie Microbienne - Institut Pasteur - Paris - France Objectives: HBl miaht act as a co+actor in the development and/or progression of the acquired immunodeficiency syndrome (AIDS). The aim of this study is to understand the molecular basis of the interaction between HBV and HIV. Methods: we performed several cotransfection experiments in which plasmid vectors expressing the HBV transactivating factor HBx were cotransfected with plassids containing the reporter gene CAT under the control of HIV-1 and HIV-2 complete or tar-deleted LTRs, or of a synthetic promoter containing a typical NFkB site. The modification of NFkB binding activity in HBx-transfected cells was studied by gel retardation assays. All the experiments were performed in HBV. negative human hepatoblastoan cells (HepG2) and HBV positive human hepatocarcinoaa cells (Alexander). Results and conclusions: HBx transcriptionally transactivates both HIV-l and HIV-2 LTRs; the deletion of the TAR sequences, which suppresses TAT transactivation, does not abolishes HBE transactivation. Since the main regulatory sequence left in the TARdeleted plassids is NFkB and since the HBx can activate transcription from a canonical NFkB site, we conclude that this is the main factor involved in the HBx induced HIVLTRs transactivation. Gel retardation assays performed using NFkB labeled oligonucleotides reveals that the X protein influences in a dose dependent manner the binding activity. The mechanism by which HBx activates NFkB is currently under investigation. NOTES M.A.1229 ACTIVATION OF THE HIV PROMOTER BY RADIATION AND DRUGS Zmudzka. Barbara Z.; Beer, JZ; Olvey, KM; Strickland, AG; Lee, W; Jacobs, ME Food and Drug Administration, HFZ-114, Rockville, MD 20857, USA Objective: Several laboratories reported that HIV is activated by ultraviolet radiation (UV). We evaluated the HIV-activating potential of radiation exposures used in dermatology, oncology, in experimental AIDS (treatment, for experimental blood product cleansing, and in cosmetology. Such exposures involve ultraviolet radiation in the A and B ranges (UVA and UVB), combined exposure to psoralens and UVA radiation (PUVA, including extracorporeal PUVA photopheresis), and ionizing radiation. Methods: Induction of the HIV promoter activity was measured in cultures of HeLa cells stably transfected with the HIV promoter/chloramphenicol acetyl transferase (CAT) gene vector. The cells were exposed to selected treatments and 20 h later CAT activity was measured. Evaluation of the HIV-activating potential of different exposures was based on comparing the highest levels of HIV promoter induction. Results: The results confirm that the HIV promoter is strongly activated by UVB or UVC but not by UVA radiation. The most effective fluence of UVB was 750 J/m2; thus, the effective fluences are in the range of real life exposures. UVA in combination with 5-methoxypsoralen, 8-methoxypsoralen, or 4'aminomethyl-4,5',8-trimethylpsoralen also strongly activated the promoter. The most effective fluences were inversely related to drug concentration. X rays had little or no effect in inducing the promoter. Discussion and Conclusions: HIV promoter is strongly activated in vitro by UVB and UVA + psoralens under conditions similar to those occurring in therapeutic and cosmetic, as well as environmental exposures. Further studies are needed to determine how the phenomena observed in vitro with the HIV promoter relate or contribute to the effects produced by similar exposures in the human body infected with the complete virus. NOTES O 0() I~; M.A.1230 HUMAN HEMATOPOIETIC STEM CELL TOXICITY ASSOCIATED WITH ZIDOVUDINE IN VITRO: EFFECTS OF G-CSF AND M--CSF Gallicchio, Vincent; Hughes, Nedda Lucille P. Markey Cancer Center, University of Kentucky Medical Center, Lexington, KY, USA. objective: To investigate the capacity of G-CSF and M--CSF to ameliorate the hematopoietic toxicity associated with zidovudine following co-culture studies with human bont marrow cells In vitro. Methods: Adherent and T--cell depleted normal human bone marrow cells were incubated in vitro either in the presence or absence of G--CSF (human recombinant, AMGEN; 10, 25 and 50 units/ml) or M--CSF (human recombinant, Genzyme; 10, 25, and 50 units/ml) and zidovudine (Burroughs-Wellcome) IDo5S X 10-5M for myeloid and macrophage precursors and 5 X 10-9M in the presence of optimal concentration of erythropoietin (1 unit/ml)(EPO, human recombinant, AMGEN), for erythroid precursor BFU-E. Resul ts: Zidovudine reduced the number of myeloid, macrophage and erythroid precusor stem cells following co-culture in vitro. The addition of G-CSF, M-CSF or EPO alone failed to reduce or protect against zidovudine toxicity on hematopoietic stem cells; however it was demonstrated that the addition of G-CSF, but not M-CSF, cultured in the presence of EPO, did provide protection against zidovudine toxicity on BFU--E. Discussions and Conclusions: These studies identify that G-CSF, M-CSF, and EPO alone failed to provide production and/or reduction against zidovudine toxicity; however G-CSF plus EPO was effective in ameliorating zidovudine hematopoietic toxicity for erythroid precusors. With the increased use of EPO in the management strategy of zidovudine toxicity, the potential clinical significance of this response will require further examination. M.A.1231 HEPATITIS B VIRUS INFECTION OF PERIPHERAL BLOOD LYMPHOCYTES IN S TRANSCRIPTIONALLY ACTIVE HIV-1 INFECTION. Manzin, A*; Bagnarelli, P*; Menzo, S*; Valenza, A*; Pauri, P*; Carloni, G**; Clementi, M***. Institute of Microbiology, University of Ancona* and Trieste*** and Institute of Experimental Medicine**, C.N.R., Rome, Italy. Objective To verify whether presence of hepatitis B virus DNA in PBLs of HIV-1 infected patients correlates with transcriptionally active HIV-1 infection. Methods DNAs and RNAs were extracted from nuclei, and cytoplasmic fractions, respectively, of PBLs from 5 ml of blood of HIV-1 seropositive individuals. HBV DNA sequences of the S and C regions and HIV-1 proviral DNA were detected by polymerase chain reaction (PCR) amplification. HIV-1 specific transcripts were detected by reverse-transcription (RT) -PCR. Results Eight out of 53 HIV-1 seropositive patients resulted also positive for HBV DNA specific sequences in PBLs. HIV-1 specific transcripts were detectable in all of the HBV DNA positive samples and in 27/45 patients negative for HBV DNA sequences.n PBLs, irrespectively of the HBV serology. Discussion and conclusions Molecular evidence indicate that HBV is capable to infect cells other than hepatocytes including PBLs and bone marrow cells. At least one HBV product, the X-protein, has been shown to increase HIV-1 transcription rate in vitro. Data shown here point to a significant proportion (15%) of co-infected patients and to the presence of detectable HIV-1 trascripts in all of the patients with HBV DNA infect- nn of pnIN. Folnow-lup researcrh should c. ari fy the role of HBV/HIV-1 co-infection

Page  150 M.A.1232 IN VITRO INTERACTIONS BETWEEN HIV AND TRICHOMONAS VAGINALIS O A. Dolei, C. Serra, A. Mattana, C. Juliano, M.V. Area and P. Cappuccinelli *Institute of Microbiology and Virology, University of Sassari, Italy. Objective: The aim: of this study was to evaluate the possibility that Trichomonas vaginalis (Tv.), responsible for a worldwide diffused sexually-transmitted disease,may behave as a carrier for heterosexual transmission of HIV infection.To this purpose in vitro experiments were devised in order to cstudy the interactions between HIV and TV isolates with different degrees of virulence. Methods: HIV preparations 'HTLV111B, H9-derived) were incubated with i) living T.v., ii) cells disrupted by sonication, iii) supernatants of T.v, cultures. At serial times residual HIV infectivity and virus antigens were evaluated by SFU on C8166 cells, and p24 ELISA, respectively. Results show that HIV is rapidly cell-associated (from 10 to 70% of recovered virus). All T.v. strains inactivate HIV infectivity, although virulent strains are more effective than the less pathogenic ones (95% and 75% specific inactivation after I h at 37"C, respectively), Moreover, infectious HIV persists associated to the less pathogenic T.v. strains for several honrs-. Even though some inactivation is observed with lysed Trichomonas and conditioned medium, maxinmal effect requires living T.v., suggesting a direct involvement of cell activity. Conclusions: persistence of HIV infectious virions on some low pathogenic T.v. strains may be consi-,tent with an "in vivo" passive transfer of HIV from one individual to another. Supported by the 1990.S.S. grant N.5206.035 M.A.1233 EBV REACTIVATION IN HIV-1 SUPERINFECTED LYMPHOBLASTOID CELL LINES SPONTANEOUSLY ESTABLISHED FROM HIV SEROPOSITIVE INDIVIDUALS OTTMANN. Michele; INNOCENTI, P.; MORAND, P.; SEIGNEURIN, J. M. Laboratoire de Virologie, Facult6 de m6decine, Grenoble, France. Objective: To study HIV and EBV interactions, by investigating both virus replication in in vitro HIV superinfected EBV positive Lymphoblastoid Cell Lines (LCL). Methods: Seven spontaneous LCL from HIV seropositive individuals were infected in vitro with two different strains of HIV-1 (HIVLAv-Bru and a strain isolated in our laboratory: HIVpar). HIV expression was monitored by the p24 antigen capture assay. In 3 of 7 HIV infected LCL, HIV infectivity was assessed by infection of PHA stimulated cord blood cells with filtered supernatants, and kinetics of EBV replication was studied by DNA slot-blot analysis. Results: All the LCL were HIV/PCR negative and EBVIPCR positive before in vitro HIV infection. The seven LCL could be infected by both HIVLAV-Bru and HIVpar as measured by p24 level in supematant. Cytopathic effects with both HIV strains were observed in 3 of the 7 infected LCL. In 3 of 3 HIV infected LCL studied, infectious HIV was released after 3 or 4 days post-infection and induction of EBV replication was observed at a high level during the kinetics of HIV infection. Conclusion: 7 of 7 spontaneous LCL could be infected with two different strains of HIV-1. In vitro HIV superinfection of LCL established from HIV seropositive individuals was associated with EBV reactivation and infectious HIV production. We must add that one LCL, CD4 negative by cytofluorometric analysis, could be easily infected by HIV. This suggests an other HIV entry pathway. y t\Q.u NOTES NOTES M.A.1234 HIV- EMHANCES MIRACELLULAR GROWTH OF MYCOBACTERIUM AVIUM IN HUMAN MACROPHAGES. Hoffner S E*, Killienius G*, Koivula T*, RydgArd K J*, Valentin A**, Asj6 B*, Ljungh C*, Svenson S B1 *National Bacteriological Laboratory, S-105 21 Stockholm, Sweden, **Department of Virology. Karolinska Institute, S-105 21 Stockholm, Sweden. Objective: Disseminated Mycrbacterium avium infections are common in AIDS patients and result in a reduced life expectancy. Human monocytes/macrophages are important target cells both for HIV and M avium. The objective of this study was to evaluate in vitro the interaction of M..avium and HIV-1 in human macrophages. Methods: Human monocytes isolated from peripheral blood of healthy blood donors were infected with HIV-1, Mh avium or both. Intracellular growth of M avium and replication of HIV-1 was followed for five weeks. Results: Intracellular bacterial growth was seen in all M avium infected cell cultures, and was parallelled by increased production of Interleukin-1 (alfa and beta). Preinfection of the macrophages by HIV-1 accelerated the Intracellular growth of M avium. Conclusion: These findings may in part explain the impaired control of mycobacterlal infections in AIDS patients. Human herpesvirus-6 and HIV infection in Africa and in Europe. M.A.1235 S.Ranger*, P.Weinbreck**, A.Sangare***, S.M'Boup****, S.Souquiere*, F.Denis*. *Virology department, CHU Limoges, France. ** Internal medecine, CHU Limoges, France. ***Pasteur Institute, Abidjan, Ivory-Coast. ****University, Dakar, Senegal. HHV-6 was first thought to be a cofactor in disease caused by HIV infection, then it was shown to suppress near totally the replication of HIV-1. Objective: To examine the differences in expression of HHV-6 in HIV-1 or 2 antibody positive individuals with or without clinical manifestation, in Africa and in Europe. Material and methods: Blood samples (641) from individuals (controls and patients) living in three areas (France, Ivory-Coast and Senegal) were analysed for the presence of HIV antibodies by ELISA and western-blot, and for HHV-6 antibodies by indirect immunofluorescence assay. A serum giving an HHV-6 titer 20 was considered positive. Results: FRANCE IVORY- COAST S ENEGAL Anony Blood HIV + HIV + Pregnant HIV-2 Pregnant HIV-1 + mous donors patients symptom women symptom women HIV - free free No of samples 142 95 154 50 50 50 50 50 Geom. mean titer 30 16 33 104 125 153 55 34 Prevalence (%) 59,8 ' 30,5 70,1 86,0 880 78,0 7090 64,0 Conclusions: Prevalences and geometric mean titers varied according to areas: both werd notably higher in Ivory-Coast. French population of anonymous individuals at risk but HIV4 showed a significantly lower prevalence than HIV+ individuals or control group. No significant difference was observed in prevalence and in geometric mean titer between HIV + groups and control groups, nor in HIV-1 and HIV-2 populations comparatively to their respective control group. There was no significant association between higher titers of HHV-6 antibody and HIV C>3 C>

Page  151 M.A.1236 FAILURE OF ANTIVIRAL CHEMOTHERAPY: VIRUS OR CELL RESISTANCE? 3. Cinatl jr., J. Cinatl sen.*, H. Rabenau, H. Geis, N. Mainke*, i. Kornhuber*, fH. W. Doerr 3entre of Hygiene, Department of Medical Virology, *Centre of Pediatry, Departement af Haematology and Oncology, J. W. Goethe University, Frankfurt a. Main:n the course of the last two years we have isolated several acyclovir-resistant nutants of HSV from patients with AIDS. Only recently, we obtained HSV type 1.solates from two patients with AIIS and reourent HSV type 1 infection, manifested ry progression of mucocutaneous ISV lesions despite therapy with acyclovir. nterestingly, both these SV type 1 isolates were sensitive in vitro to acyclovir. l similar case was reported for varicella zoster virus infection in a patient with iIDS. These findings suggest that a mechanism other than virus resistance may account.or failure of antiviral chemotherapy. Cell resistance to drugs including nuclesoide inalogs is the main obstacle in cancer therapy. We supposed that cell resistance ould evolve during the prolonged antiviral chemotherapy. The resistant cell may oxtain subtherapeutic levels of antiviral agents or their active forms resulting in herapeutic failure. Xtlture of monkey kidney cells (NA-104) resistant to acyclovir was established; the esistant cells were continuously grown in medium containing 200 uM acyclovir whereas armal cells died in such medium. The domes of acyclovir requiered to prevent virus eplication for several ISV-1 isolates were by far greater in resistant than in ormal MA-104 cells. Results - howed that antiviral effect of acyclovir may be ignificantly desreased in cells which developed resistance to acyclovir. NOTES M.A.1237 IMMUNE ACTIVATION OF LATENT HIV-1 EXPRESSION IN MONOCYTE/MACROPHAGES Mikovits, Judy A.*, Lohrey, N.**, Schuloff, R***, and Ruscetti, F.** *BCDP, PRI/Dyn Corp, NCI/FCRDC, Frederick, MD, USA, **BRMP, NCI/FCRDC Frederick, MD, USA and ***George Washington Univ., Washington, D.C. USA Objective: To investigate the role of immunological stimuli in the activation of HIV expression from latently infected monocytes/macrophages from HIV infected patients. Methods; A latently infected (no RNA expression) monocytoid cell line, THP-1, and freshly isolated adherent monocytes from asymptomatic seropositive individuals were examined for HIV DNA and RNA by polymerase chain reaction (PCR) using primers for gag and spliced tat products and for p24 antigen by elisa before and after stimulation. Results; THP-1 containing latent HIV, previously shown to be induced by 5-Azacytidine, was activated by incubation with Con A activated normal T cells. Adherent, CD3+ depleted, monocytes cultured from the peripheral blood of twenty-one asymptomatic HIV+ donors for ten days were then assayed for HIV. Three samples had HIV DNA and RNA, six had DNA but no detectable RNA by PCR and twelve contained neither detectable DNA nor RNA. After co-culturing these monocytes with normal Con A activated T cells, the six samples containing DNA but no RNA, were now positive for RNA and p24. Surprisingly, eight of the twelve samples which had no detectable DNA were now positive for RNA and p24 indicating that the samples were HIV positive below the detectable level of PCR analysis. Furthermore, while studies using either transwells to separate monocytes from T cells or purified cytokines did not activate HIV, membranes isolated from Con A activated T cells stimulated HIV expression suggesting that activation involves cell contact which induces factor(s) in monocytes responsible for overcoming HIV latency. Conclusions: HIV expression can be activated in monocytes which lack detectable HIV DNA. Monocytes can be a source of silent infection in asymptomatic HIV+ donors. NOTES I rr 00 t^; M.A.1238 Ooswami, Kamal*; Tedder, Richard*: Hamel, Donald*; Harrison. Michael*: Daniels. Rodney**. * University College and Middlesex School of Medicine. London. UK. *National Institute for Medical Research. Mill Hill. London, UK. Expression of human immunodeficiency virus type 1 in the cerebrospinal fluid detected by polymcrase chain reaction (PCR) and its correlation with CNS disease. Methods: The polymerase chain reaction (PCR) was used to amplify HIV-1 sequences in CSF and serum samples. Briefly, total RNA was extracted from these samples and reverse transcribed using either random primers or the aniscnsc primer. The resulting eDNA was amplified in a first round PCR with outer primers. In "nested" PCR a small aliquot of the amplified products was reamplified using a pair of "nested primers", i.e primers internal to those in the first round. Results: Fifty three specimens of CSF and serum were evaluated. Seven of the 24 patients with neurological disease showed the presence of HIV-1 sequences in their CSF. By using end-point dilution technique of CSF cDNA from both groups of patients, it was shown that patients with neurological disease had much higher titre of virus in their CSF. Nucleotide sequence analysis of the PCR products (env gene) showed some heterogeneity between serum and CSF virus. Conlusions: cDNA synthesis and double PCR is a sensitive method for the detection and rescue of HIV-specific sequences in CSF and serum. This correlates with the stages of the disease. The quantification data suggest an increase in viral load in CSF with the progression of neurological disease. The significance of these results are discussed. M.A.1239 HIV-1 INFECTION ALTERS TYROSINE PHOSPHORYLATION IN CD4* T CELLS Cohen David I'; Tani Y"; Welchlin-Bouadjemi H*; Samelson LE"; and Lane HC* *LIR, NIAID, NIH, Bethesda, MD USA, "CBMB, NICHD, NIH, Bethesda, MD USA Obective: Tyrosine phosphorylation of cellular substrates is a key event regulating T cell growth, and it is abnormal in an animal model (MRL Igr mouse) of dysfunctional T cell development. The mechanisms by which HIV-1 infection leads to CD4 T cell depletion and paralysis of antigen specific helper cell proliferation are incompletely understood. The purpose of the present study is to analyze alterations in cellular tyrosine phosphorylation associated either with HIV-1 infection of CD45 Jurkat cells or peripheral blood lymphocytes (PBLs). or with envelope expression by transfected Jurkat cell lines. Methods: Jurkat cells or PBLs were infected with HIV-1 LAV-BRU; sham-intected cultures served as controls. Jurkat cells were transfected to stably express cell surface LAV-BRU envelope glycoproteins gp41 and gp120, and functional tat and rev. Pattems of tyrosine phosphorylation were analyzed with antiphosphotyrosine antibodies. Results: HIV.1 infected cells demonstrated up to a twenty-five fold Increase in tyrosine phosphorylation of an unusual substrate of 30 kilodalton (pp30), accompanied by syncytium formation and extensive cell death. Tyrosine phosphorylation of cellular substrates of 95 kilodalton (kd) and 135 kd was also demonstrably augmented. In contrast, cell triggering by CD3, CD4, or CD2 receptor cross-linking did not induce pp30 phosphorylation. Transfected Jurkat cell lines stably expressing high levels of envelope and tat glycoproteins were studied for parallel alterations in ceilular tyrosine phosphorylation. Such cells ordinarily grew as readily as parental Jurkat cells without undergoing spontaneous cell fusion. These envelope-expressing cell lines did not demonstrate dramatic relative increases in the tyroslne-phosphorylation of the pp30 substrate. Some of the transfected cells lacking cell surface CD4 expression were induced to undergo syncytium formation and cell death by co-culture with CD4* Jurkat cells. These cells displayed marked increments in pp30 tyrosine phosphorylation similar to the abnormal pattern characteristic of HIV-1 infected cultures. Discussion and Conclusions: These studies demonstrate that HIV-1 infection has the capacity to cause the aberrant tyrosine phosphorylation of T cell substrates, particularly pp30. The induction of abnormal tyrosine phosphorylation may represent an important mechanism both for the helper T cell death and for the antigenspecific paralysis of helper T cells mediated by HIV-1.

Page  152 M.A.1240 HIV- 1 MEDIATED INHIBITION OF ENRICHED HEMATOPOIETIC CELL (CD34+) GROWTH Re Maria Carla*, Zauli 0.*, Visani 6.*, Furlini 0.*, Vignoli M.*, Gugliotta L.**, and La Place M*. *Institute of Microbiology, University of Bologna, Bologna, Italy *"L.A. Seragnoli" Institute of Hematology, University of Bologna, Bologna, Italy Objective: In this work the effect of HIV- I and envelope glycoprotein gp120 on the In vitro growth of enriched (CD34+) hematopoietic progenitors was investigated in liquid and semisolid cultures. Methods: 25 bone marrow samples, obtained from consenting healthy donors, were enriched (up to 95-99%) in CD34+ cells. Enriched CD34+cells were inculated with HIV-I+ H9 cell supernatant, ultracentrifuged HIV-I+ H9 cell supernatant, virtually depleted of viral particles and gp120, at concentrations from 10 pg/ml to I ng/ml. Cell viability, HIV- 1 p24 detection in CD34+cell liquid culture supernatants and clonogentl assay for granulomacrophage progenitors (CFU-GM), were.dayly cocked during all the experimental observation period. In same instances PCR analysis was also performed. Results: a significant reduction of viable cell number in liquid cultures (p<0.05) and of CFU-GM in semisolid cultures (p<0,05) was observed In CD34+cell samples treated with entire viral Inoculum, ultrancentrifuged viral supernatant and gp120, starting from the 2nd day of liquid culture. Although no signs of viral replication were ever observed, gp 120 at the maximum concetration explored (10 lIg/ml) caused only a partial (70%) inhibition of CD34cell growth, whereas in the presence of entire viral Inoculum no viable cells could be observed after 8 day of liquid cultures. Conclusions: these data show that HIV- 1 has a strong Inhibitory effect on CD34+cell growth capacity, that cannot be ascribed to a complete viral replication cycle and it is, at least partially, mediated by envelope clvcooroteln 0a 120. NOTES MA.124A1 QUANTITATION OF HIV IN PERIPHERAL BLOOD: CORRELATION WITH SM.A. DISEASE STAGE O'Shea Siobhan, Rostron T, Singh N B & Banatvala J E, Virology Department, UMDS, St Thomas' Campus, London, United Kingdom. OBJECTIVE: Quantitation of HIV in peripheral blood mononuclear cells (PBMC) and correlation with clinical, virological and immunological status. METHODS: PBMC were obtained from symptomatic and asymptomatic patients and serial dilutions co-cultured with cord blood cells. Viral titres were expressed as a tissue culture infective dose (TCID) per 10 PBMC. p24 antigen was measured by EIA, antibody responses to recombinant NEF, REV, TAT and VIF by slot blot, and B-2 microglobulin by RIA. RESULTS: Mean TCID/10 PBMC was 125 among symptomatics and 85 among asymptomatics representing 1:8,000 and 1:12,000 PBMC harbouring HIV. However, expressing virus titre in terms of the proportion of CD4 PBMC showed that among symptomatics 1:208 CD4+ cells harboured HIV compared to 1:2952 among asymptomatics. p24 antigen was detected more frequently and at higher levels among patients with more advanced disease. NEF and REV antibodies were detected in a lower proportions of symptomatic patients (NEF 26%, REV 56%) than asymptomatics (NEF 53%, REV 73%); responses to TAT and VIF were similar in both groups. B-2 microglobulin levels were higher among symptomatics. DISCUSSION AND CONCLUSIONS: Infectious virus, p24 and A-2 microglobulin levels increased and antibodies to NEF and REV decreased with disease progression. NOTES MA 242 USING POLYMERASE CHAIN REACTION TO DIAGNOS HUMAN HERPES VIRUS TYPE 6 M.A.1242 (HHV6) INFECTON IN HIV DISEASE Stroud. _1l; Salvato, P.**; Thompson, C.***; Morrow, J.****; Ragsdale, D.****; Kotarba. J.**** *Houston Immunological Institute, Houston, Texas, *Park Plaza Hospital, Houston, Texas, **University of Texas, Houston, Texas, ****University of Houston, Houston, Texas, *****Texas Woman's University, Houston, Texas, USA. Objective: To describe the role of polymerase chain reaction (PCR) in the diagnosis of HHV6 infection in HIV disease. Methods: The sample consisted of 100 subjects from a population of 4,000 (50% HIV+) patients in an internal medicine practice. HIV+ (n=38), ARC (n=31) and AIDS (n=31) patients had DNA viral sequences (PCR) and cultures for HHV6. Chi square was used to investigate the relationship between HHV6 infection and HIV disease staging. Results: HIV Disease HIV+ (n=38) ARC (n=31) AIDS (n=31) Detection Method HHV6 culture 5 13 13 42 14 45 HHV6 PCR 36 95 31 100 29 94 The detection of DNA viral sequences by PCR resulted in an 82% increase in the diagnosis of HHV6 infection in HIV+ patients, a 58% increase in the diagnosis of HHV6 infection in ARC patients and a 49% increase in the diagnosis of HHV6 infection in AIDS patients. Conclusions: HHV6 is significantly more likely to be diagnosed by PCR than by blood cultures, particularly in asymptomatic seropositive patients. These results indicate that a functioning immune system can suppress viral expression to a degree not detectable by cultures. Implications are that other more serious viral infections (e.g., cytomegalovirus) can also be detected early in asymptomatic seropositive patients by PCR testing. Also, HIV infection may be detected earlier by PCR than HIV cultures with broad implications for health practices and therapeutic intervention. S1 PERINATAL TPANSMISSION OF HIV-I: REIATIONSHIP BETWEEN PROVIRAL COPY NL IER IN VIIO, VIRA M.A.1243PROPERTIES IN VITRO AN) CLINICAL OUTICYE IN INFECTED CHILDREN. Mtmano, F.,De Rossi A., Pasti M., Oretto L., ~Giaquinto C., Chieco-Bianchi L. Institute of Oncology, oDept. of Pediatrics, University of Padova, Italy. Objective: Study of HIV-1 copy nuiber in PBMC in vivo, and in vitro cell tropism, replication rate and transactivation activity of HIV isolates from perinatally infected children, to correlate virologic properties with disease outcme. Methods: HIV-1 copy nutber was determined by direct PCR in lysed patient cells using env and LTR HIV primers. Amplified products were quantified by densitoneter analysis, and copy number was evaluated on a reference curve obtained with 8E51 cells containing 1 provirus/cell. In vitro properties were studied by coculturing PBMC with normal PHA-stimulated PBMC, and with T-lymphoid H938 and monocytoid U38 cell lines, in which HIV-LTR CAT construct had been Introduced. Transactivation was evaluated by CAT assay. Results: Isolates from children with ARC (7) showed rapid replication rates; tropism and transactivating capacity were seen in both indicator lines.Replication rates of isolates from children with minor (8) or no (10) symptons ranged from rapid to slow or intermediate; tropism and transactivating activity also varied, according to with viral growth pattern. In vitro replication rates correlated with proviral copy rntuer in PBMC in vivo. Conclusion: Isolates from children with frank symptoms show tropism for both indicator lines, high transactivating capacity and rapid growth rates; these features correlated with a relatively high proviral copy number. These characteristics in isolates from subjects with minor or no synptons may constitute a progostic parameter of the course of infection.

Page  153 M.A.1244 BIOLOGICAL AND ANTIGENIC CHARACTERISTICS OF SEQUENTIAL SIMIAN IMMUNODEFICIENCY VIRUS ISOLATES DERIVED FROM EXPERIMENTALLY INOCULATED CYNOMOLGUS MACAQUE MONKEYS. Fenvo. Eva Maria* Putkonen, P.** Albert. J.*'. Ohman, P*, Biberfeld, G**. 'Dept. of Virology, Karolinska Institute and "Depts of Immunology and ***Virology, National Bacteriological Laboratory, Stockholm, Sweden. Objectiv to study the biologic and antigenic properties of simian immunodeficinecy (SIVsm) isolates obtained throughout the entire course of infection, from day 12 postinfection until death from immunodeficiency. Methods: Biologic characterization involved tests for replicative capacity of SIVsm isolates in human blood mononuclear cells and T-lymphoid and monocytoid cell lines, whereas antigenic characterization involved virus neutralization in an autologous system. Results: The biologic properties of virus reisolated at 12 days postinfection were similar to theose of the inoculated SIVsm, both were able to replicate in all cell lines tested. Subsequently, the capacity of reisolated viruses to replicate in cell lines was gradually lost. In two macaques, neutralizing antibodies to the first isolate but not to subsequent isolates, could be detected from seroconversion throughout the asymptomatic period. Qonclusions: SIVsm isolates reisolated from macaque have a restricted replicative capacity compared to the inoculum. Changes occur also in the antigenic properties of the virus over time giving rise to variants resistent to neutralization by the monkeys own serum. M.A.1245 SINGLE STEP GROWTH KINETICS FOR HIV: INDUCTION OF CELL DEATH BY SHUTDOWN OF DNA SYNTHESIS. Kiernan, R., Doherty, R.R., & MPhee. Dale Macfarlane Burnet Centre for Medical Research, Fairfield Hospital, Melbourne, Victoria, Australia. Objective: To achieve single step growth kinetics for HIV in order to understand the mechanism of cell killing. Methods: HTLV-I transformed MT-2 cells were infected at 100 pfu/cell with HIV-I and equivalent inactivated virus, both derived from infection of MT-2 cells: Replication was assessed by RT activity, MT-4 plaque assay, cpe, cell viability and 3H-TTP incorporation. Results: Extensive cell death (85%) without syncytial formation accompanied by peak virus production (10"5pfu/ml) was observed by 48h. DNA metabolism was markedly reduced by 24h and maximal (98.5%) by 48h. Using an equivalent amount of inactivated virus (4pg/ml) cell viability was also reduced (21%) with almost complete shutdown of DNA synthesis (98.5%) by 48h. No syncytia were observed with inactivated virus. In contrast, similar concentrations of uninfected cell supernatant (HTLV-I virus) produced a modest mitogenic effect with no cell death occurring. Conclusion: Virus alone, either viable or inactivated is sufficient to induce significant cell cytotoxicity. The mechanism of cell death appears to involve shutdown of host cell DNA synthesis mediated by one or more viral proteins. The protein(s) responsible and the site of action are currently being investigated. I t^T ~*= CS. ^1 l~h, 00r h, ~o NOTES NOTES - I - --- M.A.1246 CERVICAL SQUAMOUS INTRAEPITHELIAL LESIONS (SIL) IN WOMEN WITH.A. HIV: REIATIONSHIP TO PERSISTENT HUMAN PAPILLOMAVIRUS (HPV) INFECTION OF THE CERVIX. Burk Robert D, Fleming I, Ho GYF, Klein RS. Albert Einstein College of Med./Montefiore Medical Ctr., Bronx, NY, USA. Objective; Previous studies indicate that HIV immunosuppressed women are especially vulnerable to cervicovaginal infection with and neoplastic manifestations related to HPV. To understand the pathologic basis of these events, we have investigated the natural history of cervical HPV infection in a group of women at risk for or infected with HIV. Methods: Forty women had repeat cervicovaginal lavage, Pap smear, and medical assessment over time. Type specific HPV infection was determined by restriction enzyme analysis and Southern blot hybridization. Fourteen women had symptomatic (Sx) HIV infection (AIDS, ARC or persistent generalized lymphadenopathy), 6 had asymptomatic (ASx) HIV infection, and 20 were HIV-. Asx women (HIV+ or HIV-) were combined for analyses. Results: Eleven of 14(79%) HIV+ Sx women vs 7/26(27%) ASx women were HPV+ initially (OR=10, 95% CI 1.8-65). Six of 11(55%) HIV+ Sx vs 1/7(14%) ASx initially HPV+ had persistent type specific HPV infection (RR=4, 0.6-25). SIL was found in 6/7(86%) women with persistent HPV infection compared to 3/33(9%) without persistence (RR=9.4, 2.6-34) or compared to 2/11(18%) initially HPV infected but without persistence (RR=5, 1.1-20). Conclusion: These data indicate that women with persistent HPV cervical infection are at increased risk for SIL. Moreover, the association of zervical SIL and Sx HIV infection may be related to an increased propensity of HIV+ Sx women to have versistent cervical HPV infection. M.A.1247 HV-1 ISOLATES SHOW A DISSOCIATION OF REVERSE TRANSCRIPTASE (p66/p51) AND CORE ANTIGEN (p24) PRODUCTION. Masquelier, Bernard"; Delord. B.*, Sallafranque-Andreola, 4-L.** Tarrago-Litvak, L."': Fleury. H.J.A." * Universite de Bordeaux II. Bordeaux, France. ** IBCN-GNRS, Bordeaux, France, Objective: To characterize the in vitro properties of two HIV-1 isolates from a patient receiving long-term treatment with zidovudine (ZDV). Methods: Two HIV-1 isolates were obtained from the peripheral blood mononuclear cells of a patient treated for more than one year with ZDV. They were then grown in CEM cells (a T-lymphoblastoid tumor cell line) with and without ZDV. Viral production in the supernatants was monitored by determination of reverse transcriptase (RT) activity, titration of core antigen (p24) with an EIA and detection of RT protein (p66/p51) plus core peptide (p24) by western Blotting (WB). A reference HIV-1 strain (LAV-Bru) was used as control under the same conditions. Results: In untreated infected CEM cells the level of p24 production compared to RT activity was high; in infected cells treated with ZDV, RT activity was at background level while p24 production was still significant, thus indicating a dissociation of RT activity and core antigen production; in this latter case, WB showed that there was a release of the p24 peptide and not of the p66/p51 peptides. Conclusion: Under the pressure of ZDV. these two isolates produced p24 antigen but no infectious particles; the significance of seric p24 detection in ZDV-treated patients should be reviewed. 4 4 4^ 'N

Page  154 M.A.1248 HIV INHIBITORY FACTORS IN HUMAN SALIVA. Malamud, Daniel, Davis, C. Roth, E. and Friedman, H. Univ. of Pennsylvania School of Dental Medicine and Medicine, Phila. Pa USA. Previous reports have suggested that incubation of HIV in human whole or submandibular saliva inhibited infection of human lymphocytes. Our studies were designed to confirm and extend the findings of Fox et al to gain insight into the specificity and mechanism of the anti-HIV effect of saliva. Stimulated whole, parotid, or submandibular saliva was collected from normal donors, pooled, dialyzed, and lyophylized. Aliquots of HIV were incubated with saliva, or dilutions, for 2-60 minutes. The saliva/HIV mixtures were then filtered through a.45 u filter and added to Sup-T1 cells. Controls consisted of virus incubated in buffer, and then treated as above. Anti-HIV activity was assessed by quantitation of either syncytial formation or p24 immunoassay. For comparative purposes, similar studies were carried out with herpes simplex or adenovirus. Results showed that exposure to all 3 fluids led to a decrease in HIV infectivity, with submandibular saliva demonstrating the greatest effect. There was a direct correlation between the amount of saliva (dilutions of 1:4 to 1:32) and the extent of inhibition. Likewise, inhibition increased as the time of incubation increased from 2-60 min. In contrast to the results obtained with HIV, there was little or no f fect of saliva on the infectivity of HSV or adenovirus. These latter findings suggested that the inhibitory activity present in human saliva nay be specific for HIV. Supported by NIH grant DE09569. NOTES M.A.1249 IL-1 AND IL-6 CORRECT THE ACCESSORY CELL FUNCTION DEFICIENCY OF HIV-1 INFECTED MONOCYTES Lacroix, France*; Fillon, Lionel G*; Fan, Z*; Zhao, D*, Izaguirre, Carlos** *University of Ottawa, **Federal Centre for AIDS, Ottawa, Canada Obective: To determine the effect of HIV-1 infection on the accessory cell function of monocytes. Methods: Accessory cell function was measured by the ability of monocytes to restore PHA and soluble anti-CD3 induced blastogenesis of highly purified and monocyte depleted (<0.5%) resting T cells. Monocytes were isolated using gelatin-fibronectin coated flasks (< 1% T cells) and after 4 days in culture with GM-CSF they were infected with HIV-1. Accessory cell function was tested 2 and 7 days later. RKsults: We found that HIV-1 infected monocytes lack the ability to permit PHA and anti-CD3 induced T cell blastogenesis. Uninfected monocytes restore the proliferative response of purified T cells. The loss of accessory activity occurred 2 days after infection. Preincubation of monocytes with anti-HIV-1 neutralizing human antibodies did not restore the response. Also monocytes treated with inactivated HIV-I had normal accessory cell function. The addition of IL-1 and IL-6 restored the accessory cell function for PHA stimulation but not for anti-CD3 stimulation. i)iscussion and Conclusions: HIV-I infection of monocytes suppressed their accessory cell function in the T cell blastogenesis assay shortly after infection. The suppressive effect was restored by exogenous IL-1 and IL-6 suggesting that infection down regulated the production of both cytokines by monocytes. AntiCD3 induced blastogenesis was not restored perhaps due to down regulation of Fc receptors. NOTES M.A.1250 ACUTE HIV-1 INFECTION OF PERIPHERAL BLOOD LYMPHOCYTES IS ASSOCIATED WITH ENHANCED INTEGRIN EXPRESSION Weeks BS*; Klotman Mary E.**; Kleinman HK*, Yamada KM*, & Klotman PE*. *Lab of Develop. Biology, NIDR & **Lab of Tumor Cell Biology, NCI, NIH, Bethesda, MD, USA. Objective. Specific tissue targeting of virally infected cells or viral proteins may play a role in HIV-1 pathogenesis. Therefore, we examined the effects of viral infection on the expression of receptors involved in cell adhesion to basement membrane components. Methods. PHA-stimulated peripheral blood lymphocytes (PBLs) were infected with HIV-1 IIIB at a multiplicity of infection of 0.1 or 1.0. After 24 hours, free virus was removed and cell attachment to fibronectin was determined 1, 2, and 5 days following infection. Synthetic peptides as well as antibodies to specific integrins involved in binding to fibronectin were used to block binding. Results. Acutely infected PBLs showed enhanced binding to the extracellular matrix component fibronectin at 1, 2, and five days following infection when compared to mock infected PBLs. Infection of cells was confirmed by reverse transcriptase activity and P24 analysis. Binding was inhibited by the synthetic peptide RGD but not the alternatively spliced peptide CS1 and by antibodies to the RGD-binding a5 and pl integrin subunits. Conclusion. Acute HIV-1 infection of primary PBLs is associated with enhanced binding to the extracellular matrix component fibronectin which is most likely mediated through the a5l1 member of the integrin family of cell surface receptors. This enhanced binding may result in the localization of viral infected cells and/or secreted viral products to tissues. M.A.1251 PATHiOGENESIS OF HIV-2 INFECTION IN MACAQUE SPECIES. Morton, William R*; Schmidt, AM*; McClure, J*; Gale, M*; Kuller, L*; Hu, S-L *,**. Regional Primate Research Center, University of Washington, SJ-50, Seattle, WA, USA; "Oncogen, Seattle, WA, USA HIV-2 has a high degree of homology to the simian immunodeficiency virus (SIV), shown to cause AIDS in several species of nonhuman primates. We have identified 2 subgroups of HIV-2 that have differing degrees of cytopathogenicity to macaque peripheral blood mononuclear cells (PBMCs) in vitro. In this study, we will compare the in vivo pathogenesis of 2 representative strains of these 2 subgroups. Two strains of HIV-2 (ROD and EHO) were used to infect 2 groups of animals: one group consisted of 4 Macaca nemestrina (Mn); the other of 4 Macaca fascicularis (M). Autologous PBMCs were infected with either the ROD or EHO isolate in vitro, then administered intravenously along with cell-free virus. Two animals of each group were given each strain of virus. Infection was confirmed in all animals by virus isolation from PBMCs in coculture or by seroconversion. The Mf are now 30 weeks post-inoculation (PI) and remain in apparent good health. One RODinfected Mn was cuthanized at 30 weeks PI following acute CD4t lymphocyte depletion, gastroenteric dysfunction and shock. Findings at necropsy revealed signs similar to SIV-induced AIDS. The 3 remaining Mn of the group have had transient febrile episodes and lymphadenopathy. Whole blood was transfused from one EHO-infected Mn into 2 Mn, and from the surviving ROD-infected Mn into 2 Mn (primary transfusion). The 2 ROD transfusion recipient Mn have shown no signs of infection. One EHO recipient Mn, currently at 30 weeks PI, is CD4+ lymphocyte-depleted. The other EHO recipient transfusion Mn was euthanized at 20 weeks PI with HIV-2-induced AIDS. Clinical signs included CD4+ lymphocyte depletion, fever and opportunistic infections. Whole blood from this animal was transfused into 2 Mn (secondary transfusion). At 10 weeks PI, one of these animals is demonstrating CD4+ lymphocyte depletion. This study demonstrates that HIV-2 ROD and HIV-2 EHO are infectious in both Mn and M. However, AIDS-like signs have developed only in the Mn. In addition, in vivo passage of the HIV-2 EHO appears to have accelerated the onset of an AIDS-like syndrome in recipient Mn. R Con C> C> sr rj

Page  155 M.A.1252 CYTOKINES TNF-A AND IL-1 STIMULATE HIV-1 PROLIFERATION FROM PERSISTENTLY INFECTED HUMAN FETAL BRAIN CELLS IN CULTURE Carlo S. Tornatore and Eugene O. Major Section on Molecular Virologyaand Genetics, National Institute of Neurological Disorders and Stroke, Bethesda, Md. U.S.A. OBJECTIVE: HIV-1 infection in the brain results in an encephalopathy which affects the function of neural and glial cells. To study the direct affect of HIV-I on glial cells we have established a model of HIV-1 persistently infected human brain cells in culture Experiments are described which identify several cytokines that are able to activate HIV-1 production from persistently infected cells. METHODS: Human fetal brain tissue is established in cultures that results in greater than 95% astrocytes. HIV-1 is introduced into these cultures through transfection of pNL4-3 DNA or infection using NL4-3 virions. Assessment of infection is done by measuring p24 protein in culture medium, immunoblots of infected cell lysates, and multiplication of HIV-1 by coculture with human T cells. Cytokines are measured by an ELISA antigen capture assay. RESULTS; Several weeks following HIV-1 infection, astrocytes no longer produce p24 protein but maintain a persistent infecion. These cells can be reactivated to synthesis p24 when either cocultured with T lymphocytes or in the presence of medium shared with T lymphocytes. Persistently infected human astrocytes can also be reactivated by introducing cytokines such as TNF-alpha or Interleukin-1 into the medium. DISCUSSION: These experiments show that HIV-1 can establish a persistent infection in human fetal astrocytes which is reversible in the presence of specific cytokines and suggests a model for HIV-1 infection in human fetal brain. NOTES M.A.1253 ELECTRON MICROSCOPIC 3D MODELLING OF HIV-INDUCED SYNCITIA AND NON INFECTED CONTROLS Cph, Bron 1, J. Jendis 1, S.Loeliger 1, Th. Blichi 2, J. Schlpbach 1. 1 Swiss National Center for Retroviruses, and2 Central Electron Microscopy Laboratory, Institute for Immunology and Virology, ZtLrich, Switzerland. Objective: To develop techniques using single HIV-infected cells for directed immunocytochemical studies and 3D immunoelectron microscopic quantification. Methods: Computer Aided Volumic Ultra Microscopy (CAVUM) takes advantages of correlative photon to electron microscopy and allows the 3D reconstruction and morphometrical measurements of entire cells with 40 nm spatial resolution. Results: Comparative studies between HIV-induced syncitia and Laser-induced infected and non infected H9 cell fusion, are associated with colloidal gold and/or peroxidase immuno-labeling of p24, gp41 and gpl20. The 3D morphometrical parameters (volume and cell surface, number and volume of cell nucleus) as well as 3D immunoelectron labeling allows for the first time tentative semi-quantification. Studies are underway to evaluate the importance of viral antigens as well as the role of monocytes / macrophages for syncytium formation. Conclusion: These studies pave the way toward the use of single cells as reagent tubes and will be presented by computer-controlled projection. NOTES M.A1254 GENERALIZED GIANT CELL DISEASE IN SIV-INFECTED CYNOMOLGUS MeAs ^ ~ MONKEYS. Su-ling Li*t), Ephata E. Kaaya*#1), Per Putkonen*, Hans Feichtinger*@, Carlo Parravicini***, Marianne Ekman*, Peter Bibcrfcld*. 'Imnmunopathology Lab., Karolinska Institute and *National Bactoriological Laboratory, Stockholm, Sweden, ***Department of Pathology, University of Milano, Italy, #Muhimbili Medical Centre, Dar-es-Salaam, Tanzania, @Department of Pathology, University of Innsbruck, Austria. Qbjjyie: To study the pathology of SIV infection in cynomolgus monkeys as a model of AIDS in man. Methods. 28 monkeys were infected with simian immunodeficiency virus (SIVsmm3, from Dr. P. Fultz and McClure, Yerkes Primate Center, USA). Histopathological including immunohistochcmical changes were studied in biopsies and autopsy material. Viremia, antibody response and immune status were examined in blood and serum samples. ResIL4f: Characteristic changes were found predominantly in the lymphoid tissues and CNS reminiscent of HIV-associated changes in humans. In addition, generalized multinucleated giant cell formation in most tissues was observed in 43% of the monkeys. The morphology of these giant cells (GC) varied considerably but was neither by histology nor by immunohistochemistry related to any demonstrable opportunistic infection. However, SIV antigens were consistently found in these cells by immunostaining. The predominant immunophenotype of these GC's corresponded to that of monocytes/macrophages (CD68+), but some GC also expressed T-lineage antigens (CD2+, 4+). Comparison of monkeys with and without giant cell disease (GCD) showed that GCD+ monkeys had longer infection more often antigenemia, more marked immunodeficiency and an increased rate of malignant lymphomas. Conclusions: Siv infection in cynomolgus monkeys is often associated with generalized formation of GC of macrophage origin which appear to be an important factory and reservoir of SIV. GC formation may reflect virus mediatcd fusogenic events corresponding to syncytia formation in vitro. Cytogenesis and pathogenic importance of GC formation in vivo can be studied in this SIV-cynomolgus-AJDS model. 1) Partly supported by SAREC and WORLD LABORATORY. MA 1255 COMPLEMENT-ENHANCED HIV-1 BINDING TO MONOCYTES IS DEPENDENT ON ALTERNATIVE PATHWAY ACTIVATION, AND MEDIATED BY COMPLEMENT RECEPTORS Bakker, Leendert J; Nottet, HSLM; de Graaf, L; de Vos, NM; Van Strijp, JAG; Visser, MR; Verhoef, J. Eijkman-Winkler Laboratory for Medical Microbiology, University of Utrecht, Utrecht, The Netherlands. Objective and Methods: Infection of mononuclear phagocytes with HIV-1 can be enhanced by complement, as has been shown by several investigators. In a previous report, we demonstrated that HIV-1 binding to human monocytes is increased by complement (Bakker et al, VI Int.AIDS Conf., F.A.311). In the present study the mechanism of complement-mediated HIV-1 binding to monocytes (measured by flow cytometry with direct fluorescein-labeled purified HIV-1) was investigated with the use of sera deficient for complement factor Clq, factor B, or C5, respectively, and mAb directed against complement receptors (CR1 and CR3). Results: Compared to control sera, factor B deficient serum significantly decreased complement-mediated HIV-1 binding to monocytes (p<0.05), whereas Clq or C5 deficient serum did not. This suggests that binding is dependent on activation of complement via the alternative pathway. MAb against CR blocked complement-mediated HIV-1 binding to monocytes (anti-CR3 significantly more than anti-CR1, p=0.01). Conclusions: Our results demonstrated that (1) HIV-1 activates complement via the alternative pathway and (2) the opsonization of HIV-1 with complement results in binding to monocytes via CR. LKLJ\

Page  156 a\ M.A.125 HIV INFECTION OF ASTROGLIAL CELL CULTURES FROM RAT CEREBRAL TISSUE. Romano, Nino; Bonura, F; Russo Alesi, D; Vitale, F; Ferraro, D. Dipartimento di Igiene e Microbiologia, University of Palermo, Italy. Objective: The possibility that primary cultures of astrocytes of rats might be a suitable experimental model to study HIV replication in cells of glial origin was assessed. Methods: Mixed glial cell cultures were prepared from dissociating cerebral cortex of I to 2-d-old rat pups. The astrocytes were separated from the mixed cultures by mechanical sieving techniques. The cell cultures displayed glial cell morphology and expressed glial fibrillar acidic protein. Results: Infection of astroglial cell cultures by HIV resulted in transient expression of gag protein of HIV, with maximal amounts of p24 protein observed at 3 days post infection and declining to zero by 7-10 days post infection. HIV infection of astrocytes appeared to be productive; the infected cells were able to rescue virus by cocultivation with blood mononuclear human cells. HIV proviral DNA was detected by polymerase chain reaction with a gag specific primer pair at 2-3 days post infection in virus infected cells. However, this expression was only observed at early times after infection. Conclusions: The low-level HIV productive infection of astrocytes will facilitate the analysis of viral and cellular factors which control the expression of viral genome in glial cell types. NOTES M.A.1257 BIOLOGICAL CHARACTERISTICS OF HIV 1 ISOLATES FROM BLOOD AND CEREBROSPINAL FLUID OF AIDS PATIENTS. -Russo Alesi, Domenico*; Bonura, F*; Vitale, F*; Romano, N*; Farinella, V**; Bruno, M**. *Dipartimento di Igiene e Microbiologia, University of Palermo; **Ospedale "Guadagna" USL 62, Palermo, Italy. Objective: The possibility that isolates of HIV? from the brain may constitute a special group of AIDS virus was assessed. Methods: Paired cerebrospinal fluid (CSF) and peripheral blood (PB) HIV 1 strains were isolated from 10 AIDS patients in human blood mononuclear cells (HBMC). Primary monocyte cultures, monocyte-limphocyte cocultures and established T-cell lines were used to compare the replicative capabilities of the isolates. Results: All the PB isolates, but one, replicated in HBMC causing extensive cytopathic effect and were easily propagated in continuous T-cell lines. The CSF isolates were relatively unable to infect and to cause cytopathic effect in HBMC, but were easily isolated in monocyte-limphocyte cocultures and multiplied efficiently in primary monocyte cultures. Two of the metabolically labeled paired strains submitted to radioimmune precipitation assay showed a clearcut difference in the size of the env glycoprotein. Moreover the PB isolates were neutralized at higher extent than CSF isolates by sera of patients. Conclusions: Strains isolated from CSF share properties (cell tropism, cytopathology and antigenicity) that distinguish them from the PB isolates. NOTES K Ki M.A.1258 HIV INDUCES BOTH ACTIVATION AND SUPPRESSION OFT LYMPHOCYTES: INDUCTION OF PHOSPHORYLATION OF TYROSINE RESIDUES AND LYMPHOKINE GENE EXPRESSION AS WELL AS SUPPRESSION OF PROLIFERATION. Bo Hofnann, Parunag Nishanian, ThangNguyen, Praphaphone Insixiengmay, John L. Fahey. UCLA School of Medicine, Los Angeles, CA 90024-1747 OBJECTIVE: We have earlier shown that HIV proteins induce a partial suppression of the early steps of lymphocyte activation including inositol phospholipid metabolism leading to reduced cell proliferation (J.Immunol. Hofmann et al 1990). The aim of this study was to further identify steps of lymphocyte activation suppressed or increased by HIV. METHODS: Addition of tHV or PHA to cultures of normal Peripheral Blood Mononuclear Cells(PBMC). Phosphorylation of tyrosine residues was measured by Western Blot of cell lysates with specific monoclonal antibody. mRNA for IL-2 and INF-gamma was measured by hybridization of RNA slot blots with specific oligo-DNA probes. RESULTS: Phosphorylation of tyrosine residues on proteins/receptors involved in cell activation is among the very early steps of cell activation. Addition of HIV proteins to cultures of normal peripheral blood mononuclear cells (PBMC) induced phosphorylation of tyrosine residues. In particular a 45 kD and a 30 kD protein were strongly phosphorylated. In contrast to PHA stimulation, HIV induced phosphorylation of the 45 kD protein remained stable for several hours, whereas the PHA induced phosphorylation declined within one hour. HIV also induced mRNA for cytokines in cytoplasm.. HV stimulation for 16 hours induced levels of mRNA for IL-2 and INF-gamma similar to the levels obtained after PHA stimulation. DISCUSSION AND CONCLUSIONS: The study shows that HIV proteins induce early lymphocyte activation as indicated by phosphorylation of tyrosine residues. Also the generation of mRNA for IL-2 and INF-gamma is seen within 16 hours. The activation, however, does not lead tc proliferation. The prolonged phosphorylation of tyrosine residues indicates that HIV activation differs from PHA stimulation. The phosphorylation changes may reflect the same mechanisms contibuting to HIV-induced suppression of lymphocyte functions. S FUSION OF HIV-1 & HIV-2 INFECTED MONOCYTE/MACROPHAGES (MO) WITH UNINFECTED.A.1259 CD4-BEARING T-LYMPHOBLASTOID CELLS: A TRANSMISSION OF ELECTRON MICROSCOPIC EXAMINATION. J Boothman*, J Marshall+, A Colyin*, S Sonza* and SM Crowe*+. *Macfarlane Burnet Centre for Medical Research and +Virology, Fairfield Hospital, Melbourne, Australia. OBJECTIVE: To define at an electron microscopic level the cell fusion process between HIV-L and HIV-2 infected mo with uninfected CD4-bearing T lymphoblastoid cells (VB). METHODS: MO were isolated from HIV-seronegative donors via density centrifugation and glass adherence and cultured in suspension in RPMI-1640/10% AB human serum. Cells were infected for 48 hrs with either HIV-1 (DV strain) MOI 0.5-1 or HIV-2 106 RT counts per 106 cells. 7x105 infected mo were co-cultured with 5x105 uninfected CD4-bearing T cells (VB) to permit cell fusion. Cultures were pelleted, glutaraldehyde fixed and then processed for transmission electron microscopy (TEM). RESULTS: Fusion between the two cell populations occurred within 48 hrs. By TEM, initial cell contact was followed by the extension of small, occasional, finger-like processes. These processes then widen to form cytoplasmic bridges. A number of bridges along the interfacing cell surfaces fused resulting in obliteration of individual cell membrane integrity. TEM revealed retroviral particles in vesicles within infected monocyte/macrophages. HIV-I and HIV-2 particles were identical. Immature virus particles were seen budding from the mg surface with mature virions visible near the cell surface, CONCLUSIONS: HIV-1 and HIV-2 infected mf fuse with uninfected T cells in vitro. Virus particles are actively released from infected mO from the cell surface. Fusion may provide an additional mechanism for T lymphocyte depletion in vivo. r Con C, C> C> C>

Page  157 M.A.1260 RNA-PROTBIN COMPLEXES INTERFERE WITH INHIBITION OF REVERSE TRANSCRIPTASE Lightfoote, Marilyn M.* and Bailey, J.M.** *Food and Drug Administration, Center for Devices and Radiological Health; and**The George Washington university Medical Center. Depar*ment of Biochemistry and Molecular Biology Objective: To determine the importance of cellular factors on HIV reverse transcriptase activity, and to determine if these factors vary among cell types, influencing the efficiency of the viral enzyme. Methods RNA-protein complexes Isolated from infected or uninfected cells were identified on agarose gels. Further analysis of RNA was carried out using probes from suspected RNA species in dot blot hybridizations. Protein In these complexes were detected using antibodies to normal cell factors, viral factors, and various lymphokines. When identified, proteins in these complexes were included in the reverse transcriptase assay to determine their effect and their ability to reverse the inhibition of enzyme activity by known inhibitors. Results: RNA-protein complexes may bind to reverse transcriptase and modify its function by either enhancing or reducing enzyme activity. Monoclonal antibodies to HIV reverse transcriptase were used to identify RMA-protein complexes and to determine if the protein in these complexes I of cellula or viral origin. The proteins were identified in gel retardations assays. Conclusion: The influence of cellular factors on reverse transcriptase activity may lead to the development of effective inhibitors of viral replication. NOTES M.A.1261 RETROVIRAL EFFECTS ON MUSCLE CELLS: Trujillo, J Roberto; Gomez-Lucia, Esperanza; Lee, Tun-Hou; Essex, Max Harvard University School of Public Health, Boston, MA, USA Objective: Whether or not retroviruses play a direct or indirect role in myopathy is unclear. The objective of this work is to compare and study in vitro the role of HIV-1 and HTLV-I infection on muscle cells. Methods: We used 5 strains of HIV-1 and 5 strains of HTLV-I for infection of two muscle cell lines (A-204, HSM). We also evaluated the role of tumor necrosis factor alpha (TNFa) on non-infected muscle cells. Among the HIV-1 strains we included a mutant of gp120 (HxB2 -427) which lacks the binding site for CD4. After infection with virus alone or cocultivation with infected lymphocytes or infected promonocytes the cultures were evaluated for evidence of infection by presence of cytopathic effect (CPE) and reverse transcriptase and evidence of viral antigens using ELISA, immunofluorescence, and western blotting techniques. The virus-exposed cultures were also tested for levels of TNFa production. Results: No HIV-1 infection or CPE was observed on muscle cells, but CPE was observed when using the rTNFa alone. In the case of HTLV-I, infection was seen and CPE was more prominent with the virus itself than with the rTNFa. Discussion and Conclusions: We found no evidence that HIV-1 infects cultured lines of muscle cells. However, TNFa alone may cause alterations in muscle cells. It is conceivable that HIV-1-infected individuals produce elevated levels of this cytokine, resulting in myopathy. However, in this preliminary study, HTLV-1 showed a direct effect on muscle cells. Further studies are warranted to examine the possibility that myotropism may occur with this virus. NOTES s ~lj 00 tEI MA. 1262 R oI N- E-I--oREPRTE I E'D CEIIS INDEPENDENT OF INFECTIOUS VIRUS PARTICLES. Grunow. Roland*, Valentin, A., Feny6, E.-M., Wahren, B., Jondal, M.W *Karolinska Institute, and **National Bacteriological Institute, Stockholm, Sweden Obective: In the serum of most HIV infected persons a significant increase of viral p24 core protein is detectable with the development of the AIDS. The aim was to study whether the p24 is the result of degradation of virus particles or infected cells, or whether p24 could be released as non-virion proteins from living HIV-1 infected cells. Methods: Primary and chronically HIV-1 infected cells were washed until no infectious virus or p24 were detectable in the supernatants. The cells (T-cells and monocytes) were cultured for different time periods and checked for their viability. The supernatants were divided in whole supernatant (ws), pelleted virus (pv) and supernatant of pelleted virus (spy). In each sample the amounts of p24, reverse transcriptase (RT) and the infectious titer (IT) were determined. Results: During cultivation increasing levels of p24, RT and IT were observed in ws. Increased RT and IT, but not p24 were found in pv. In contrast, increasing amounts of p24 were detected in spy. The numbers of dead cells were few. Conclusion: HIV-1 core protein or part of it (p24) was released from infected cells independently of infectious virus particles. It could be speculated that p24 is a target structure for immune response or a reaulatorv component in HIV replication. HETEROGENEITY IN THE OLIGOSACCHARIDE STRUCTURES OF THE ENVELOPE M.A.1263 GLYCOPROTEIN (GP120) OF HIV-1. Pal, Ranajit, di Marzo Veronese, F., Nair, B., and Sarngadharan, M. Advanced BioScience Laboratories, Inc., Kensington, Maryland, USA. Objective: To study the heterogeneity in the N-linked oligosaccharide structures of gpl20 synthesized in chronically infected Molt3 cells. Methods; Metabolically labeled HIV-1 glycoprotein was immunoprecipitated from chronically infected cells and the immunoprecipitates were digested with endo H to probe mannose-rich oligosaccharides. Results: Gpl20 synthesized in Molt3 cells infected with HIV-1 (Molt3/HTLV-III) contained both endo H sensitive (mannose-rich) and resistant (complex type) forms of oligosaccharides. In contrast, proportions of endo H resistant oligosaccharides were significantly lower in gpl20 synthesized in Molt3 cells infected with the molecularly cloned virus HXB2 (Molt3/HXB2). A single cell clone isolated from Molt3/ HTLV-IIIB cells released glycoprotein predominantly modified with complex type carbohydrates in the medium. Although HIV-1 antibody positive human serum immunoprecipitated gpl20 containing both complex and mannose-rich oligosaccharides, an anti-gpl20 monoclonal antibody (M77) directed to the major neutralizing epitope immunoprecipitated this glycoprotein containing mostly mannose-rich oligosaccharides. In spite of this differential reactivity, virions from both Molt3/HTLV-III and Holt3/HXB2 cells were neutralized by M77, albeit at different concentrations. Conclusion: The oligosaceharide structure of the glycoprotein of HIV-1 is highly divergent and such heterogeneity may influence the immunological properties of the glycoprotein. EV,0\

Page  158 S M.A.1264 PROTECTION FROM VIREMIA AND DISEASE PROGRESSION IN SIVBK28 00 INFECTED Macaca mulatta, CHALLENGED WITH SIVmac251/32H H. Niphuis and J.L. Heeney ITRI-TNO, Dept. of Chronic and Infectious Diseases, Rijswijk, The Netherlands. Various strains of the HIV-2 related SIVsm group of primate lentiviruses have different pathogenic properties in Macaca mulatta. SIVBK28 is a relatively non-pathogenic and non-viremic strain causing persistent infection with a strong persistent immune response. Objective: To determine if prior infection with a minimally virulent strain influences disease progression after subsequent challenge with a virulent strain. Methods: Rhesus monkeys infected with SIVBK28 for over 2 years and challenged with 100 MID of the 8811 stock of SIVmac251/32H. Results: Prior infection with SIVBK28 protected three out of four animals from viremia as detected by antigen capture and four out of four from disease progression as determined by absolute CD4/CD29 counts. The two controls infected with 100 MID of SIVmac251/32H developed viremia and developed terminal disease while unchallenged SIVBK2C infected monkeys remain disease free. These data supports observations that viremia is an important factor in disease progression and secondly demonstrates that prior infection with minimally pathogenic strains may permit aquisition of sufficient protective immunity to limit subsequent infection(s) by more virulent strains. Conclusions: These findings suggest that the pathogenic or virulence properties of the initial inoculum may influence establishment of subsequent HIV strains and greatly influence the course of disease. NOTES M.A.1265 DISEASE PROGRESSION CORRELATES WTH SELECTIVE LOSS OF CD4+/CD29* MEMORY T CELLS IN SIV-INFECTED RHESUS MACAQUES, HIV-INFECTED INDIVIDUALS, BUT NOT IN HIV-INFECTED CHIMPANZEES J.L, Heeneyt, F. Miedema2 and H. Niphuisl; 'ITRI-TNO, Dept. of Chronic and Infectious Diseases, Rijswijk, The Netherlands; 2Dept. of Clinical Viro-immunology, CLB, Amsterdam, The Netherlands Simian immunodeficiency virus (SIV) induces an AIDS like syndrome in rhesus macaques similar to HIV in man. Immunological abnormalities can be demonstrated in macaques infected with SIV similar to man infected with HIV. Chimpanzees infected with HIV-1 show no signs of immunosuppressive disease development to date. In HIV-infected individuals following seroconversion, CD4+/CD29+ memory T-cells are observed to decline while they are still asymptomatic. Objective: To determine if CD4+/CD29+ changes are predictive disease correlates in non-human primates infected with strains of varying virulence. Methods: Rhesus macaques were infected with various types of SIV by either cell free or cell associated challenge. Chimpanzees were infected with either HTLV-IIIB, LAV-1 or blood from an HIV-I infected patient and followed for 5 to 7 years. Results: Absolute CD4+ T-cell numbers, specially the CD4+/CD29+ memory subpopulation were observed to decrease rapidly in rhesus macaques infected with pathogenic SIV strains but not in those infected with nonpathogenic SIV strains. In the asymptomatic post-infection interval, CD4+/CD29+ population decreases occurred in animals which progressed to disease. Analysis of the CD4+/CD29+ subset has not demonstrated any changes between HIV-1 infected and non-infected chimpanzees, in line with the observation that chimpanzees are not susceptible to HIV-related disease. Conclusion: These results indicate a relation between selective loss of CD4+/CD29+memory T-cells and progression to disease only in rhesus macaques infected with pathogenic SIV strains and HIV-infected individuals. NOTES t G\ '4 M.A.1266 DEFINITION OF AN HIV p24 EPITOPE ENDOWED WITH DISTINCTIVE IMMUNOSUPPRESSIVE PROPERTIES. Pugliese,Orsola, Luzzati,A.L.,Giacomini,E.,Giordani,L.,Camponeschi,B., *Chersi,A., *Evangelista,M. Dept. of Immunology, Ist. Superiore di Sanita; *Ist. Regina Elena, Rome, Italy. Objective: We have previously shown that a 20 aminoacids peptide (aa 218-237) of p24 is able to block antigen-specific responses in vitro. Objective of the present study is a better definition of the responsible epitope. Methods: Peptides, corresponding to progressively shorter sequences in the 218-238 region of p24, were added in graded amounts to cultures of normal human blood lymphocytes set up for the induction of: a) a specific primary antibody response to the SRC antigen; b) a proliferative response to the PPD antigen. Cultures were assayed for the presence of specific antibody forming cells (a) or for proliferation (b). Results: The 16aa (223-238) and the llaa (228-238) peptides exerted a strong, dose-dependent inhibitory effect in system a ()90% inhibition), but were much less suppressive in system b (the Ilaa peptide, in one experiment, was practically ineffective). Conclusions: Since culture system a deals with the induction of a primary antibody response, while system b deals with a proliferative response of presumably primed lymphocytes, our results suggest that it is possible to define HIV epitopes able to interfere with precise steps of immunity and not with others. This approach may lead to progress in the study of AIDS pathogenesis and to the definition of appropriate antigenic structures for vaccine development. M.A.1267 souRA ECDInMnM mV l20-MUD MPAME OF SINAL Set Yoshitatsu and Arora Prince K. Laboratory of Neurosclence, NIDDK, National Institutes of Health, Bethesda, MD, U.S.A. Obiective: Impaired transmembrane signalling by HIV proteins may contribute substantially to the alterations in immune responsiveness in HIV-infected individuals. To investigate the mechanisms responsible for immune suppression produced by viral products, the effects of HIV-gpl20 on signal transduction in a variety of human lymhocyte subpopulations and monocytes were examined. Methods: Human peripheral blood lymphocytes and monocytes were treated with HIV- SF2-gp 120 in the presence or absence of antl-gp120 antibody. Subsequently, either OKT3 [an antibody that reacts with T cell receptors (TcR)] or concanavalin A (Con A) -induced elevation of [Ca2+]i was analyzed by flow microfluorometry using fluo-3 dye and phycoerythrinconjugated monoclonal antibodies (Leu3a. Leu2a, Leul2 and LeuM3). Results: The combined treatment with gp 120/anti-gp 120 Ab inhibited both TcR- and Con A- induced elevation of ICa2+]l in CD4+ T cells. HIV gp120 alone had no effect. The HIV gpl20/anti-gpl20 Ab treatment did not alter Con A-induced elevation of [Ca2+]i in either LeuM3+ monocytes or Leul2+ B cells. Neveretheless. CD8+ T cells exhibited significant suppression of Con A-induced elevation of [Ca2+11 following treatment with gpl20/anti-gpl20 Ab. Further analysis of CD8+ subpopulatlons revealed that "dim" CD8+ T cells were affected more than "bright" CD8+ T cells. In addition, soluble CD4 blocked gpl20/anti-gp120 Ab -induced impairment of calcium mobilization in CD8+ T cells. Conclusions: The suppression of calcium mobilization in CD8+ as well as CD4+ T cells by HIV products may explain some of the immune dysfunctions observed in AIDS patients. This phenomenon may have clinical relevance since soluble CD4 is currently being used as a treatment for AIDS. C-> a tCC>, C>3

Page  159 M.A.1268 STIMULATION OF PROLIFERATION RATE DURING HIV INFECTION IN CID4 POSITIVE CELL LINES. Di Rienzo Anna Maria*; Petronini P.O.0; Guetard D.*; Favilla R.^; Borghetti A.F.~; Montagnier L.*; Piedimonte G.0 * Unit6 d'Oncologie Virale, Institut Pasteur, Paris, France. o Istituto di Patologia Generale and ^Dipartimento di Fisica, Universith di Parma, Parma, Italy. Obective: We have observed that the mitotic index of CD4+ cell lines (CEM and U937) is significantly increased when thecells are infected by purified HIV preparations suggesting that the virus alone could be endowed with mitogenic activity. This presumptive virus-induced mitogenic activity and its possible involvement in the pathogenic development of the disease is investigated in this work. Methods: Growth rate quotient has been measured by counting cells at intervals throughout the experiment and cell cycle has been analyzed by flow cytometry. Purification of HIV-1 (Bru) particles has been obtained by filtration, centrifugation and isopycnic sedimentation. Viral production and cytopathic effects were evaluated by p24 and Reverse Transcriptase activity detection, syncitia formation and cell death. ReultO^ Following HIV infection a significant increase of the growth rate in CD4+ cell lines is observed. This increase occurs without any significant alteration in the percentage of cells distributed in the cell cycle phases, and is induced by both infectious and heat-inactivated virus, but not significantly by purified gpl20. Discussion and conclusions: The stimulation of proliferation rate has constantly been observed in tested CD4+ cell lines even when the cells were treated with heat-inactivated virus alone. The phenomenon is therefore independent of viral replication and seems rather linked to early stages of infection. It is possible that the virus-cell interaction modifies or increases a pathway of signal transduction. This interaction, which in cell lines results in a proliferative increase, could also alter the responsiveness of peripheral blood lymphocytes. Our observations underline the necessity to clarify the consequences of early events of infection in order to better define the different aspects of the pathogenesis. NOTES M.A.1269 MODULATION OF PROTEIN GYNTHEIS RATE DURING HIV INFECTION IN A CDIPOSITIVE CELL LINE Petronini, P.G. Di RRienzo, A.M.; Urbani,.; Guetard, D. gorghetti, A.F. # Piedimonte, Biusepp. Istituto dl Patologia Generale, Universit di Parma, Parma, Italy, Uni t d'Oncologie Virale, Institut Pasteur, Paris, France. Objective: The pathogenetic mechanism of lytic viruses usually involves a large domination of host protein synthesis. In contrast, during the lif cycle of a typical retrovirus only little interference with the cell biosynthetic machinery occurs. It has been suggested that the retrovirus HIV takes over the cellular machinery thus behaving as a lytic virus. However, recent results do not support the complete domination by HIV of the host macromolecular synthesis. Our effort has been to investigate whether the late cytopathic effects usually observed during HIV infection are associated with significant alterations of the cell protein synthesis. Methods: CEM cells, a clone of human T-cell leukemia exposinq CD4 receptors, has been used throughout this Investigation. Cell and viral protein synthesis has been analyzed by measuring the initial velocity of amino acid incorporation into proteins and the profile of viral and host polypeptides synthesized during l4V infection. Cytopathic effects have been evaluated by analyming syncytia formation and cell death. Results: During the proliferation of CEM cells in vitro, the rate of protein synthesis is markedly affected by the resulting increase of cell density. The infection by HIV does not significantly alter this affect. Analysis of the specific polypeptide synthesized indicates that this retrovirus uses only -a few percentage of host protein synthesis even at the peak of viral replication. In infected cells, however, a definite increase in the degradation rate of cellular proteins occurs. Discussion and conclusion: In the CD4-positive lymphoid cell line CEM used in this work the lytic retrovirus HIV does not dominate the host cell machinery and a shut-off of protein synthesis does not take place during infection. However, the significant increase of cellular protein degradation observed during infection might be involved in the occurrence of the late cytopathic effects. NOTES s NIt ts; M. ENVELOPE SHEDDING AND HIGH CELL DENSITY: BARRIERS TO VACCINE AND IMMUNOTHERAPY M.A.1270 DEVELOPMENT Layne. Sco'; Merges, M."; Dembo, M.; Spouge, J.*; Nra, P." Los Alamos National Laboratory Los Alamos, NM, USA, National Cancer nstitute, Frederick, MD, USA "National Ubrary of Medicine, Bethesda, MD, USA. Objective: HIV-1 and -2 share two properties that are important for understanding Immunity and therapy. First, HIV preferentially Infects CD4+ cels that have a wide range of densities in lymphoid, reticuloendothelial and nervous tissue. Second, the density of gp120 knobs covering the surace of HIV decreases with time. We have therefore undertaken an investigation of how the blocking of infection with soluble CD4 (sCD4) and monoclonal Immunoglobulins depends on cell density, viral stock age (loss of gp120), and blocker concentraton. Methods: The essential tools for accomplishing our objective have been quantitative infectivity assays and a kinetic model of HIV Infection. The model simulated four phsical-chemical reactions that take place during an assay Infectious contact between virlon and cell, spontaneous shedding of gp120, single-t viral nactivation and formation of gpl20-blocker complexes. By analyzing these reactions, we developed a series of experiments for investigating their Influence on infection. Subsequently, quantitative HIV assays were carried out with different cell densities and blocker concentrations, and with viral stocks that were either fresh or preincubated (aged). Assay results were then used to identify agreements or disagreements with the model, to refine experimental techniques, and to reduce the effects of extraneous variables. Results: After analyzing numerous infectivity assays of HIV-1 and -2, an overall picture has begun to emerge. It Indicates that HIV is In a race against time, in which virions must undergo infectious collision with target cells before spontaneously losing their infectious potential. There are two variables determining whether the race favors infection or inactivation. The first one is target cell density. Lowtarget eel densities are associated with small rates of infectious collisions, which favors inactivation. Conversely, high densities are associated with large rates, which favors infection. The second variable is spontaneous shedding of gp120. Our experiments indicate that HIV requires a critical fraction (or minimal number) of unblocked gp120 glycoproteins for efficient nfection of CD4+ cells. When more than this critical fracton is present on a virion, the rate of nfection s proportionto unbloced gp20. When less than this fraction s present, however, the rate Is below proportional Conclusions: Our results show that the blocking of infection depends strongly on target cell density and virion age (loss of gp120). Thus, urappreated variations In HIV stocks and essay onditions may hinder-comparisons of bockers from laboratory to laboratory, and aged HIV chal enpe stocks may cause significant overestimates of vaccine efficacy. The results also suggests that blocking of viral particles In lymphoid ompartments will require very high sCD4 concentrations, which may explainthe refractory outcomes from sCD4-bsed drug trials n humans. Persistence of infection is commonly observed forviruses that infect lymphocytes and monocytes. The inability of gp120-blnding immunoglobullns to block Infection at high cell densities therefore provides new insights into the mechanisms of such persistence. M.A.1271 Exerimental and theoretical indica'tons for HIV-1 nef neurotoxicitv based on its simianty to scornion neurotoxines. Ruth Brack-Werner*, Thomas Wemer*, Ralph Magert, Torben Sacrmarki, Richard B. Banatit. Stefano Ferronit, Lucill Steinaal, Georg W. Kreuzergt & Volker Erfle* 'GSF-Forschungszentrum for Umwelt und Gesundheit GmbH Ingobltudtr LadstreBe 1, D-8042 Neuherberg, tMax-Plmk-Insitut fir Psychatrie. Am Klopferspitz 18s. D-033 Martinried, University of Kopehaugen, Sigurdsgade 34. DK-2200N Kopenhagcn. Objective:Neuropathological analysis of HIV-I infected brains suggests contribution of general toxic processes. We investigated potential involvement of HTV-1 nef protein in these processes. Next to HIV structural proteins and the regulatory proteins tat and vif, human glial cells persistently infected with HIV-1 show abundant expression of the 27 kD nef protein. Furthermore, immunofluorescence studies indicate expression of nef antigen in astrocytes in brain sections from HIV-1 infected patients. Methods: Alignments were carried out using the program GENALGN (IntelliGenetics Inc., for multiple alignment) and according to the method of Argos (J. Mo. Biol. 193, 385-396, 1987).-Electrophysiological studies were performed using the whole cell configuration of the patch clamp technique (Hamill O.P., et al. Pflug. Arch. 391, 85-100 1981) Results and Conclusions: We detected similarities in amino acid sequence and domain structure between HIV-1 nef protein and scorpion neurotoxins. A short amino terminal peptide and a C-terminal pentapeptide located about 42 amino acids apart in the neurotoxins are similar to two equally spaced regions in the net protein. These peptides can interact to form a domain stabilized by ion pairs as deduced from the crystal structure of one scorpion neurotoxin. Recombinant nef protein as well as nef and scorpion neurotoxin peptides from the homologous region reversibly increase the K+ current after membrane depolarization (patch clamp experiment). This indicates that structural and functional features common to HIV-1 nef and scorpion neurotoxins might be important for the pathogenesis of HIV in the central nervous system. C\ NI I', L V^0

Page  160 O M.A.1272 INTRACELLULAR LOCALIZATION OF HIV-1 NEF IN PERMANENTLY INFECTED GLIAL CELLS Kohleisen Birgit Herrmann R., Krohn K.*, Erfle V. GSF Abt.Molekulare Zellpathologie, Neuherberg, FRG; *University of Tampere, Finland. Objective: Human glioblastoma cells permanently infected with HIV-1 (TH 4-7-5) serve as a model system for virus latency characterized by a high expression of the viral regulatory protein nef. Detailed studies are performed concerning the localization of nef in different cellular compartments of TH 4-7-5 cells. Methods: The expression of HIV-I antigens in TH 4-7-5 cells was analyzed by immunoperoxidase staining and Western Blot with different monoclonal antibodies (mab) directed against defined peptides of the nef molecule. Indirect immunofluorescence (F) was used for localization of nef in TH 4-7-5 cells. Results: Besides the major structural HIV-1 proteins the 27 KD nef molecule is the predominant regulatory protein expressed in TH 4-7-5 cells in comparison to HIV-l infected lymphoma cells (K37-3). In IF studies with several mab, nef was shown to be concentrated in the perinuclear region and associated with the nuclear membrane. An intranuclear localization without staining of the nucleoli was observed with mab 3E6 directed against the amino acid 161-180 of nef. Nef was also detected in membrane and nuclear fractions of TH 4-7-5 cell extracts in Western Blot analysis. Immunogenic epitopes are found in theN- as well as the C-terminal region of nef from TH 4-7-5 cells. Discussion: Nef is found in the cytoplasm and even in the nucleus of HIV-1 infected glial cells, which might be due to basic and hydrophobic amino acids within the epitope recognized by mab 3E6. This region might serve as a signal sequence for transportation to the nucleus. NOTES M.A.1273 Pathogenicity qf,HIV-1 1g the Absence of Envelope Protein Expression. gurao. Corrado * and Gallo,,R.C. C.E.O.S.-CNR, Naples, Italy, Laboratory of Tumor Cell Biology, National Cancer Institute, National Institutes of Health, Bethesda, MD, USA Ob.ective: To look for cytopathic effects of HIV-1 independent of Mnx gene expression. Methods: The initiation codon of the envelope (e) gene of HXB2 and a second initiation codon in frame with the first were mutated into stop codons by primer directed mutagenesis, and the premature stop codon In the nef (n) gene was eliminated. The mutant (HXB2/2e-n+) and the wild type (HXB2/e+n+) viruses were cloned into the Xba site of pSp65gpt vector and respectively transfected into Jurkat, H9, 0881 (EBV transformed chimp B cell line), and into PA317 which contains an amphotropic MuLV packaging retroviral vector. Gpt selection was then applied to obtain permanently transfected lines; stable transfection was obtained with Jurkat and PA317 cells. RsultS: Cell associated particles with the morphology of mature HIV-1 were observed in 0881 cells transfected with the emutant, but not Jurkat and H9 cells. Release of p24 by H9, Jurkat, and 0881 cells transfected with the mutant was monitored during selection and upon establishment of stable transfection (Jurkat). A progressive decline in the amount of soluble p24 observed. The selected Jurkat cells express very low levels of p24; Southern blot analysis showed that the viral DNA is unintegrated-and linked to the plasmid vector. Discussion: The decline of p24 and the emergence of cells expressing very low levels of p24 suggest regulation of expression of viral genes by cell factors. Preliminary studies suggest that mitogen stimulation of e- transfected Jurkat cells leads to re-lease of IL2 followed by cytopathic effects. Ongoing studies focus on the effects of activation on T cells transfected with the mutant to gain insight into mechanisms leading to dysfunction or death of CD4+ T cells in HIV-l infected individuals. NOTES o t\Q. NI NI M.A.1274 IMMUNOHISTOCHEMICAL LOCALIZATION OF LAMININ AND TYPE IV COLLAGEN IN HIV ENCEPHALITIS Cenacchi Giovanna, Giangaspero Felice*, Martinelli Giuseppe N Istituto di Microsco