Bridging the Gap: Conference Record [Abstract book, International Conference on AIDS (12th: 1998: Geneva, Switzerland)]

12th World AIDS Conference

Abstracts 31121-31124 519

stimulated in vitro with 15 HIV subtype E CTL epitopes identified in previous studies of HLA All, A2 positive or A24 positive subtype E, HIV-infected individuals
from Thailand. CTL assays were performed at effector-to-target ratios of 60:1,
30:1, or 10:1 using individual peptides or peptide pools. A positive response was
defined as lysis of 51Cr-labelled peptide-pulsed autologous B cell lines at a level
of >10% over that of the non-pulsed targets. Presence or absence of the A32
CCR5 polymorphism was determined by PCR.
Results: Five of 8 EU FSWs had positive CTL responses to peptides from
HIV-1 subtype E. Some reactivity was to epitopes conserved between E and B
subtypes. In 4 of 6 HLA-A11 positive persons, reactivity was most often seen to
nef (3 persons), pol (2 persons), and env (2 persons) and less frequently to gag
(1 person). One of 2 non-HLA-A11 positive EU FSWs recognized an HLA-A24
gag epitope. No CTL reactivity was observed in 9 unexposed controls, and all EU
FSWs and unexposed controls lacked the A32 CCR5 polymorphism.
Conclusions: These data suggest that, in the absence of HIV receptor abnormalities which prevent cellular infection, exposed individuals may remain
uninfected or may clear HIV-infected cells through the effects of HIV- specific
CTL. These data also emphasize the importance of CTL induction as a desirable
component of a protective HIV-vaccine induced response.
109*/31121 Initiation
109*/31121 Initiation of antiretroviral therapy during primary
HIV infection is associated with rapid stabilization
of the TCR V/3 repertoire and reduction of CD8+
T cell oligoclonality
Hugo Soudeyns', G. Campi2, A. Lazzarin2, A.S. Fauci3, L. Corey4, G. Casorati2,
G. Pantaleo'. 'Laboratory of AIDS Immunopathogenesis, CHUV-Hbpital de
Beaumont, 29 Avenue de Beaumont, Lausanne, Switzerland; 2San Raffaele
Institute, Milan, Italy; 3Laboratory of Immunoregulation, NIAID, NIH, Bethesda,
MD; 4University of Washington, Seattle, WA, USA
Objectives: TCR V/I repertoire perturbations are frequently observed in patients
during primary HIV infection, and have been associated with transient oligoclonal
expansions of HIV-specific CD8+ CTLs. The aim of this project was to determine
the impact of initiation of antiretroviral therapy (ART) during primary HIV infection
on the dynamics of T cell-mediated immune responses.
Methods: The V/I repertoire was analyzed by qPCR in serial blood samples
obtained from 11 untreated and 13 patients treated with ART (AZT, n = 1; ddC,
n = 1; AZT/ddl, n = 2; AZT/3TC/saquinavir, n = 3; AZT/3TC/indinavir, n = 6). V/I
repertoire variations were evaluated longitudinally using delta scores (A), which
reflect global repertoire differences between time points. Thus, a negative slope
of A vs time indicates repertoire stabilization, whereas a positive slope indicates
destabilization.
Results: 12 of 13 treated patients showed negative A slopes, indicating the
occurrence of repertoire stabilization. Negative slopes of A vs time were observed
in 7 of 11 untreated patients. Furthermore, A slopes were significantly smaller
in treated than untreated patients (p = 0.030, t-test), indicating that repertoire
stabilization occurred more rapidly in treated patients. In order to test whether
this stabilization was associated with increased complexity of T cell populations,
TCR heteroduplex mobility shift assay (HMA) was performed on sequential samples from a patient treated with HAART. Densitometric analysis of HMA profiles
showed a reduction in the number of recognizable TCR clonotypes in 19 of
24 VP families, and a significant decrease in the total number of recognizable
clonotypes (230 vs 181, p < 0.01, t-test) following 6 months of therapy. As well,
appearance of a more homogenous smear further indicated that the repertoire
became polyclonal.
Conclusion: These results suggest that ART may significantly modulate the
pattern of TCR repertoire mobilization during primary infection, and reduce global
CD8+ T cell oligoclonality. The rapid stabilization of the TCR repertoire observed
in treated subjects may prevent clonal exhaustion of HIV-specific CTL clones
observed in untreated subjects.
601*/31122 Cross-clade envelope gp160, gag p55 and pol
specific CD8+ CTL responses early following HIV-1
clade-B infection
Susan Wilson1, S.L. Pedersen1, J.C. Kunich1, G. Mazzara2, J. Tartaglia3,
C.L. Celum4, H.W. Sheppard1. 1California Dept. of Health Services, VRDL,
Berkeley, CA; 2Therion Biologhics, Cambrigde, MA; 3 Virogenetics Corporation,
Troy NY; 4University of Washington, Seattle, WA, USA
Objective: To evaluate the breadth of HIV-1 intra- and interclade env gp 160
(BLAI VS. BMN, A and C), gag p 55 and pol (A and BLAI) CTL reactivity in recently
HIV-1 clade B-infected individuals.
Methods: PBMC were isolated from recently HIV-1 clade B-infected individuals (10 weeks to 24 months post seroconversion) and stimulated in vitro using
autologous PBMC infected WM a vaccinia vectors expressing either a multigene
cocktail comprised of HIV-1 env gp 160LAI, gag P55,LAI, pOILAI, (vAbT-408-6-1)
and nefLAI, (vAbT-23-5-1), or a multigene complex containing env gp 120MN linked
to the transmembrane portion of gp41LAi, gag p55LAI and proteaseLAi, antigens

(vP1291). At day 9-12 CTL reactivity against a panel of targets was assessed
using a standard 4-h chromium release assay. The target panel was comprised of
autologous EBV-BCL infected with vaccinia vectors expressing env gp 160 (BLAI,
BMN, A92UG037 and C92BR025), gag p55 BLAI, A92UG037 and E91CAR402, or pol BLAI
and A92UG037.
Results: All 13 of the 19 recent seroconverters who elicited positive gp 160 BLAI
CTL responses also showed comparable CTL reactivity against gp160 C92BR025,

nine of these individuals also demonstrated CTL cross-recognition of gp 160
A92UG037. In addition, 18/22 (82%) donors elicited positive gag p55 BLAI CTL
reactivity, with 16/22 (73%) and 10/13 (77%) showing CTL cross-recognition of
gag p55 A92UG037 and gag p55 E91CAR402 respectively. Interestingly, four donors
examined showed significantly greater CTL responses against clade A92UG037
and/or E91CAR402 gag p55 infected targets than BLA|. Of a total of 18 donors tested
for pol-specific CTL reactivity, 16/17 (94%) responded to pol BLAI with only 9/17
(53%) showing pol A92UG037 CTL cross-reactivity. All CTL reactivity was found to
be mediated by CD8+ effectors.
Conclusion: This study indicates that there is a considerable degree of CTL
cross-recognition of the highly divergent HIV-1 env gp160 subtypes during early
phases of HIV-1 infection, as well as the more conserved gag p55 and pol
proteins. Such findings suggest that a HIV-1 vaccine based on a single isolate
that can induce extensive cross-clade immunity could potentially demonstrate
significant universal utility.
600*/31123 HIV-1 specific cytotoxic T-lymphocyte responses in
uninfected Kenyan prostitutes: Behavioural and
immunologic correlates
Rupert Kaul1, K. Fowke2, E. Njagi', J. Ruthetford2, K.S. MacDonald3,
F.A. Plummer2. 'Dept. of Medical Microbiology, University of Nairobi, Nairobi,
Kenya; 2University of Manitoba, Winnipeg MB; 3University of Toronto, Toronto
ON, Canada
Background and Objectives: Resistance to HIV-1 has been described in a
cohort of Kenyan prostitutes, most of whom demonstrate HIV-1 env specific cytotoxic T-lymphocyte (CTL) responses. We expand these observations to include
uninfected prostitutes newly enrolled in the cohort, who do not meet the definition
of HIV-1 resistance, and seronegative women with no history of sex work. We
also examine behavioural and immunologic correlates of HIV-1 specific CTLs.
Methods: HIV-specific CTL responses were studied in seronegative prostitutes
enrolled in the cohort for <3 years (new negatives, N = 25); prostitutes enrolled
for >3 years, who met the definition of HIV-1 resistance (resistants, N = 32); and
seronegative women enrolled from a nearby family planning clinic (controls, N =
6). After stimulation with autologous HIV-infected, inactivated PHA-blasts, effector
cells were incubated with autologous B-cell targets infected with vaccinia virus
and vaccinia/HIV-env constructs. CTL activity was determined using a standard
Cr51 release assay.
Results: HIV-1 specific CTL responses were associated with duration of
seronegative follow-up in the cohort of >3 years: responses were present in 17
(53%) resistant prostitutes; 5 (20%) new-negatives; and no control women (P =
0.02). Current HIV risk behaviours, CD4/8 counts and contraceptive use were not
associated with HIV-1 specific CTLs, and nor was an AHIV exposure index which
estimates cumulative lifetime exposure to HIV-1. Several serologic MHC alleles
were associated with CTL: HLA-A24 (P = 0.002), HLA-A9 (P = 0.01), HLA-B18
(P = 0.04), and HLA-DQB604 (P = 0.01).
Conclusion: HIV-1 specific CTL activity was found in HIV-1 exposed but uninfected Kenyan prostitutes, but not in lower risk control women. Among prostitutes,
the probability of demonstrating HIV-1 specific CTL activity increased with a duration of seronegative follow-up >3 years, suggesting a role for CTL responses
in protection against incident HIV-1 infection. HIV-1 specific CTL activity was also
correlated with certain MHC alleles.
1112*/31124 Prolonged suppression of HIV-1 viremia results
in profound but incomplete enhancement of
anti-HIV-1 CTLm responses in advanced
infection
Charles Rinaldo, X.L. Huang, Z. Fan, J. Liebmann, D. McMahon, S. Riddler,
J. Mellors. Univ. of Pittsburgh A427 crabtree Hall 130 Desoto Street; University
of Pittsburgh, Pittsburgh, USA
Objectives: To evaluate memory anti-HIV-1 cytotoxic T lymphocyte (CTLm) responses in patients treated for over 2 yr with combination drug therapy in the
Merck 035 clinical trial.
Methods: We studied CTLm responses to Gag, Pol and Env in a doubleblind, randomized, controlled trial in zidovudine (ZDV)-experienced patients with
50-400 CD4~ T cells/mm3 and >220,000 copies HIV-1 RNA/ml serum. The patients were treated with either a combination of protease inhibitor (indinavir, IDV)
plus 2 nucleoside RT inhibitors (ZDV, 3TC), or ZDV plus 3TC or IDV alone for
24-42 weeks before switching to open label, triple drug therapy.
Results: Patients who first received combination therapy with IDV/ZDV/3TC
had viremia suppressed to undetectable levels (<500 HIV-1 RNA copies/ml) and
increases in CD4+ T cell numbers throughout >2 yr of follow-up. Anti-HIV-1
CTLm bulk lysis and precursor frequencies in these patients were enhanced
within weeks of initiating therapy to HIV-1 Pol, less so to Env but not to Gag.
This enhancement has persisted through >2 yr of followup. Levels of IFNy and
RANTES also were increased in these CTLm cultures. Much less antiviral effect
was noted in patients first receiving ZDV/3TC or IDV treatment. Only transient

increases in CTLm responses to Pol and Env, and not to Gag, occurred in these
patients after switching to the triple drug therapy. Current studies are assessing
T cell repertoire diversity as a possible basis for the lack of increase in anti-Gag
CTLm responses.
Conclusions: IDV/ZDV/3TC therapy has extraordinary yet incomplete enhancing effects on anti-HIV-1 CD8+ CTLm responses in patients with advanced
infection. Prior treatment with single or dual drugs can limit the immunopoten