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

12th World AIDS Conference

Abstracts 11101-11105 3

11101 Generation of recombinant analogues of the
C-C chemokine RANTES with enhanced HIV-suppressive
and reduced pro-inflammatory activity
Paolo Lusso1, Simona Polo2, V. Nardese2, E. Prosperini2, C. De Santis2,
M. Rizzi3, M. Bolognesi3. P2/P3 Laboratories-Dibit-San Raffaele, Scientific
Inst.-via Olgettina 58, Milano; 2San Raffaele Scient. Inst., Milano; 3Univ. of
Genova. Phisics Dpt., Genova, Italy
Objectives: To generate analogues of the HIV-suppressive C-C chemokine
RANTES with reduced pro-inflammatory side-effects and enhanced antiviral activity for potential therapeutic use in HIV-infected patients.
Methods: The human RANTES cDNA was cloned from PHA-activated PBMC
by PCR using primers that amplify the entire coding region of the chemokine.
Different mutations were introduced into the wild-type (WT) gene by a modification of the overlap extension technique. A baculovirus expression system was
used for the in vitro production of mature WT RANTES and analogues in soluble
form. All the molecules were tested for their ability to: a) induce calcium mobilization in CCR5-expressing cells; b) chemoattract primary human lymphocytes
and monocytes; c) inhibit infection in PBMC and macrophages by different laboratory-passaged and primary HIV-1 isolates characterized for viral coreceptor
usage; d) induce CCR5 down-regulation in IL-2-expanded T cells; e) compete
with radiolabeled WT RANTES for CCR5 binding in CCR5-expressing cells.
Results: Two RANTES analogues, R15 and R17, demonstrated a markedly
enhanced antiviral activity, compared to the WT, but failed to trigger CCR5-mediated calcium mobilization. The mechanism(s) responsible for the increased
HIV-blocking effect of these mutants is currently under investigation. Other
analogues showed a dramatic loss of antiviral potency, which generally correlated
with a diminished pro-inflammatory activity.
Conclusions: By virtue of their unique biological characteristics, R15 and R17
represent suitable candidates for chemokine-based therapeutic approaches in
HIV infection.
58*/11102] AOP-RANTES inhibits CCR5 recycling - A novel
inhibitory mechanism of HIV infectivity
Amanda E.I. Proudfoot1, M. Mack2, D. Schlondorff2, P. Clapham3, M. Marsh4,
T.N.C. Wells1, M. Opperman5. 1Serono Pharmaceutical Research Institute 14
Chemin des Aulx, 1228 Plan les Ouates, Geneva, Switzerland; 2Medizinische
Poliklinik, University of Munich, Munich 5Univeritaetskliniken, Goettingen,
Germany; 3Chester Beatty Laboratories, London; 4 University College of London,
London, UK
Coupling of a penta-carbon alkyl chain to the amino terminus of RANTES to form
aminooxypentane-RANTES (AOP-RANTES) produces a potent CC chemokine
receptor antagonist. AOP-RANTES is significantly more effective at inhibiting
infectivity by HIV-1 M-tropic strains than RANTES itself, and is able to prevent
infection of primary macrophages whereas RANTES is a poor inhibitor. Whilst
AOP-RANTES is unable to activate monocytes as demonstrated by its inability
to induce both calcium mobilisation and chemotaxis, it is capable of producing a
robust calcium response as cultured monocytes differentiate into the macrophage
phenotype. Freshly isolated monocytes have no surface expression of CCR5
detectable by flow cytometry using a monoclonal anti-CCR5 antibody, but CCR5
expression becomes readily detectable as the cells become adherent over 48 h in
culture. AOP-RANTES induces greater than 90% downregulation of cell surface
CCR5 of cultured PBMC's, whereas RANTES only achieves 70%. Incubation of
RBL cells transfected with CCR5 with AOP-RANTES causes a 3-5-fold increase
in receptor phosphorylation than when treated with RANTES. Moreover, after removal of RANTES from the medium, CCR5 recycles to the cell surface, whereas
receptor recycling is significantly abrogated after treatment with AOP-RANTES.
The inhibition of infectivity by the M-tropic HIV-1 strain SF-162 using GHOST
34/CCR5 parallels the downregulation of CCR5.
Conclusion: Inhibition of receptor recycling after downregulation is a novel
mechanism of inhibition of HIV infectivity. The implications and mechanisms will
be discussed.
56* / 11103 Modulation of HIV-1 infection by CCR5 antagonists
Donald Mosier1, GR. Picchio2, RJ. Gulizia2, R. Sabbe2, P. Poignard2,
RE. Offord3, J. Wilken2, DA. Thompson2, S. Kent2. 1The Scripps Research
Institute 10550 North Torrey Pines Road La Jolla, Calif 92037; 2Gryphon
Sciences S. San Francisco CA, USA; 3Centre Medical Universitaire Geneva,
Switzerland
Objectives: To determine if N-terminal modification of RANTES can more efficiently block virus entry in vitro and in animal models for HIV-1 infection, and
if co-receptor switch variants emerge under the selective pressure of CCR5
antagonists
Design: N-terminally modified RANTES compounds were either added to primary PBMC cultures or administered to hu-PBL-SCID mice. Cultures or mice were
challenged with a molecular clone of HIV-1 242 that requires minimal mutation to

switch from CCR5 to CXCR4 coreceptor use.
Methods: PBMC or purified CD4 Tcell cultures were treated with AOP-RANTES
or other N-terminal modifications of RANTES at concentrations ranging from 1
pg/ml to 100 ng/ml. Cultures were infected with HIV-1 clones 242, 241, and 230
isolated by Chesebro and Wehrly (J. Virol. 70:9055, 1996). These viruses differ
only in the V3 region, and are R5, R5X4, and X4, respectively. p24 capsid antigen
production was followed for 4-15 days. For in vivo experiments, hu-PBL-SCID

mice were challenged with clone 242 HIV-1 one day after initiation of treatment
with AOP-RANTES or related compounds. Continuous infusion of CCR5 antagonists by Alzet osmotic pumps maintained plasma levels at 1-5 ng/ml. Infection
was monitored by plasma viral RNA levels using the Roche Amplicor assay. Any
recovered virus was sequenced and assayed in vitro for reduced sensitivity to
CCR5 antagonists.
Results: N-terminal modified RANTES compounds were powerful antagonists
of HIV-1 infection by the R5-using 242 isolate, but often led to enhancement of
virus replication by the R5X4 241 virus. This enhancement was often noted at
low doses (<1 ng/ml) of CCR5 antagonists. CCR5 antagonists had little effect
on replication of the X4-using 230 isolate. Treatment of hu-PBL-SCID mice with
CCR5 antagonists reduced plasma viremia to low or undetectable levels during
active treatment, but virus replication resumed when treatment was stopped.
Viruses with altered co-receptor use were not observed in initial experiments.
Conclusion: HIV-1 infection can be blocked effectively by CCR5 antagonists in
vitro, and can be substantially reduced in hu-PBL-SCID mice. CCR5 antagonists
can show paradoxical enhancement of R5X4 viruses in vitro. Rapid selection of
R5 to X4 switch variants did not occur.
S11104 Anti-HIV chemokines: Domain mapping and HIV-2
lentivirus delivery
Suresh Arya, M. Owais, A. Davis-Warren. National Institutes of Health National
Cancer Institutes of Health National Cancer Institutes-Bethesda-MD, USA
HIV has expropriated chemokine receptors to gain entry into the host cell, with Mand T-tropic viruses respectively using CCR5 and CXCR4 as co-receptors. This
also subjects the virus to suppression by the natural ligand of these receptors.
Indeed, C-C or / and C-X-C or a chemokines inhibit HIV infection, presumably by
interfering with the binding of the viral envelop to the co-receptor. The mechanism
of this three component interaction - natural ligand (chemokine), bandit ligand
(viral envelop), and the co-receptor - in virus infection and signal transduction is
not well understood. We envision two models - 'same site' model where the two
ligands compete for the same site on the receptor and distant site model where
competition occurs at distance by way of stearic hindrance or conformational
change. To address these issues and to design a dual tropic minimal chemokine,
we have created a number of mutants of both C-C (prototype: RANTES) and
C-X-C (prototype: SDF-1) chemokines focused on the defined structural elements
(flexible coil, /- sheets and a helix) and expressed them in lentivirus and baculovirus expression systems. Domain mapping studies have revealed that the
amino-terminus (N-loop) as well as carboxy terminus (alpha helix) of RANTES as
well as SDF-1 could be deleted without adverse effect on their antiviral activity.
Deletion of more central /1 domain of RANTES resulted in the loss of its anti-Mtropic but not of its anti-T-tropic activity. Lentivirus vector delivery of recombinant
chemokine to relevant target cell could achieve phenotypic knockout by causing
intercellular trapping and extracellular occupancy of the receptor, thus suppressing virus infection and its spread. Our HIV-2 based lentivirus vector clearly express
recombinant chemokines. They can also be packaged in homologous packaging
system as well as pseudotyped with VSV-G protein.
Domain mapping studies suggest that large segments of both C-C and C-X-C
chemokines are not essential for their antiviral activity, that they are composed
of subdomains determining receptor specificity, and that their antiviral effect and
chemotaxis or signal transduction function employ different pathways. The recombinant chemokines in HIV-2 lentivirus vector will be useful for gene transfer
studies in achieving intracellular immunization and extracellular protection in HIV
infection and AIDS.
54* /11105 Mechanism of transdominant inhibition of
CCR5-mediated HIV-1 infection by ccr5A32
Kuan-Teh Jeang. NIH/NIAID/LMM, Bldg. 4, Rm 306, 9000 Rockville Pk.,
Bethesda, MD, USA
Objective: To understand at the molecular level how heterozygote individuals
with the ccr5A32 mutation is less susceptible to infection with HIV-1.
Design: Intracellular studies of the processing of CCR5 chemokine receptor.
Methods: Intracellular processing of the CCR5 molecule was investigated by
RIPA and by confocal immunomicroscopy. Susceptibility of cells to infection by
HIV-1 was also examined.
Results: Naturally occurring mutation of CCR5, ccr5A32 exists in the Caucasian population. The ccr5A32/ccr5A32 genotype is linked to a phenotype
that is "highly" protected from HIV-1 infection while a CCR5/ccr5A32 genotype
confers "relative" protection from AIDS. PBLs from heterozygous individuals
(CCR5/ccr5A32) support ex vivo HIV-1 replication at a reduced level compared
to CCR5/CCR5 cells; however, the molecular mechanism is unknown. We find
that mature CCR5 can be post-translationally modified by phosphorylation and
multimerization. By contrast, mutant ccr5A32, although retaining the capacity
for multimerization, was incapable of being phosphorylated. ccr5A32 heterocomplexes with CCR5, and this interaction retains CCR5 in the ER resulting in
reduced cell-surface expression. Thus, co-expression in cells of ccr5A32 with

CCR5 produces a trans-inhibition by the former of ability by the latter to support
HIV-1 infection.
Conclusion: CCR5/ccr5A32 dimerization is one plausible molecular explanation for the delayed onset of AIDS in CCR5/ccr5A32 individuals.