Abstracts from AIDS Vaccine 2010 Atlanta,Georgia,U.S.A. 28 September–1 October,2010

1 University of Washington, Seattle, Washington, USA 2 NCI, Frederick, Maryland, USA 3 Irsi-Caixa-HIVACAT, Barcelona, Spain 4 Ragon Institute, Boston, Maryland, USA 5 University of Pennsylvania, Philadelphia, Pennsylvania, USA 6 University of Miami, Miami, Florida, USA 7 Microsoft Research, Los Angeles, Califorina, USA 8 Fundació Lluita contra la Sida, Barcelona, Spain OA01.01 A Global AIDS Vaccine Based on Conserved Elements of the Viral Proteome

Background: We are developing a global, HIV-1 M group vaccine that focuses immune responses on conserved protein elements (CE) essential to the function of the virus while precluding responses against immunodominant decoys, i.e., targets on the virus that can mutate while retaining function. We are testing CE DNA vaccines derived from HIV-1 M group Gag p24 as well as the entire HIV-1-M proteome.

Methods: Two variants of expression-optimized Gag p24 CE DNA constructs (Core 1/2) were generated, with components of p24 joined by linkers chosen to maximize expression and processing in human cells. Each of 7 CE segments differ by one “toggled” amino acid (Core 1 vs Core 2), to address ∼99% of the known variability in HIV-1-M.

Results: Broad recognition of Clade B and A peptides were elicited in Balb/C mice; including CD4 and CD8 responses and Gag antibodies. Proteolytic processing of CE revealed 87% of the known optimal epitopes. Core 1 expressed in human dendritic cells elicited strong CD4 and CD8 responses in human T cells ex vivo. p24 CE are highly immunogenic in HIV-1 infections and include >30 epitopes (some previously unknown) restricted by >40 HLA. There was no difference in the number or magnitude of responses to CE regions between controllers (n = 23, excluding all those with B57, B58 and B27 HLA) and non-controllers (NC, n = 21) assessed by a set of 311 peptides (10mers overlapping by 9 AA) covering p24 and including the Core 1/2 CE. Controllers had higher avidity (p = 0.01) and more cross-reactive responses than NC (p = 0.01) to the entire p24 and CE regions. Also, responses of high functional avidity showed a superior ability to recognize peptide variants than low avidity responses (p = 0.01).

Conclusion: p24 CE are immunogenic in HIV-1 infection and immune responses can be focused on these elements, to the exclusion of decoy epitopes.

1 University of Alabama at Birmingham, Birmingham, Alabama, USA OA01.02 Codon Optimization of HIV-1 Vaccines Decreases the Potential Breadth of CD8⁺ T-Cell Responses

Background: To overcome the enormous diversity of HIV-1, many vaccines being tested in clinical trials were designed to generate broad cytotoxic CD8⁺ T-cell responses to immunogenic proteins. To ensure efficient translation of viral proteins, the primary reading frames can be codon-optimized. While this approach increases gene expression in the primary reading frame, it is still not known how optimization impacts expression of the alternative reading frames (ARF). Our lab recently showed that peptides derived from ARFs (cryptic epitopes, CE) are frequently recognized during HIV-1 infection. We therefore sought to determine the impact of codon optimization on CE by comparing transmitted founder virus (TFV) sequences from 12 acutely Clade B-infected individuals in the U.S. with 4 vaccine sequences from the Thai trial (Virogenetics, RV144), Step trial (Merck, MRKAd5), and 2 current Phase II trials (GeoVax, MVA/HIV62; Vaccine Research Center, VRCrAd5). RV144 and MVA/HIV62 contain natural, non-optimized proteins; both MRKAd5 and VRCrAd5 are codon-optimized.

Methods: CE from ARFs of the Gag protein in the TFV and vaccine sequences were analyzed. Using pairwise alignments, protein sequences from the vaccines were compared with a consensus sequence generated from the TFV sequences. Phylogenetic analyses of individual TFV and vaccine sequences were performed for all six reading frames using the neighbor-joining method to construct unrooted trees with Jukes-Cantor correction.

Results: The ARFs of the non-optimized and optimized sequences had on average 90% and 35% similarity, respectively, with the TFV consensus sequence. The natural sequences had phylogenetic distances smaller than the distances between TFV sequences, whereas the codon-optimized sequences had distances 2.1 to 4.8 (+/−0.5) times greater.

Conclusion: Because cytotoxic T-cells cannot recognize epitopes that differ by more than 33%, these data suggest that translation of ARFs in codon-optimized Gag sequences produces aberrant CE that may negatively impact the potential breadth of CD8⁺ T-cell responses.

1 Drexel University College of Medicine, Philadelphia, Pennsylvania, USA 2 University of Pennsylvania, Philadelphia, Pennsylvania, USA 3 The University of Pennsylania, Philadelphia, Pennsylvania, USA 4 Inovio Biomedical, Blue Bell, Pennsylvania, USA 5 University of Alabama, Birmingham, Alabama, USA 6 Tulane National Primate Research Center, Covington, Lousiana, USA 7 Duke University School of Medicine, Durham, North Carolina, USA OA01.03 Co-delivery of Mucosal Chemokine Plasmids in Systemically Administered DNA Vaccines Elicits Systemic and Mucosal Immune Responses in Rhesus Macaques

Background: Induction of mucosal immunity is a goal for HIV vaccines, as the mucosa is the primary site of HIV transmission/viral replication. DNA vaccines which are non-live/non-proliferating have had limited success in this area. We hypothesized that mucosal immunity could be elicited by systemic DNA vaccination along with mucosal chemokine co-delivery.

Methods: Using the rhesus macaque model, we co-immunized CCL27, CCL25, and CCL28 adjuvants with optimized/consensus macaque SIVpol/sooty mangabey gag/env antigens. Macaques were immunized intramuscularly/electroporation (IMEP) with antigenic plasmids plus/minus each chemokine.

Results: In the periphery, we observed significant IFN-gamma in all groups (∼6,000 SFU each), while intracellular cytokine staining showed a trend toward functional CD8 + T cells at mucosa sites in CCL27 co-immunized macaques. Enhanced antigen-specific IgA in sera, genital and duodenal washes of chemokine-vaccinated macaques was observed. IgA responses included western blot positive responses induced by systemic vaccination approach and development of neutralizing antibodies against SIVsmE660.11 by DNA vaccination. A repeated low dose intravaginal challenge with SmmE660 is in progress to determine whether the functionality and phenotype of vaccine-induced immunity, either at the mucosa or in systemic compartments, is a driving determinant of protection.

Conclusion: Results suggest that IMEP-delivered DNA vaccines encoding mucosal chemokines can induce mucosal immunity. The ability to expand on this study, redirecting functional immunity in vivo has clear significance for HIV vaccine development. This is the first example of the use of mucosal chemokines to influence a DNA vaccine strategy, suggesting a novel approach for manipulation of vaccine-induced immune responses.

This work is supported by grants (NIH/NIAIDS F32AI054152, HIVRADP01AI0739).

1 University of Washington, Seattle, Washington, USA 2 Yerkes National Primate Research Center, Emory University, Atlanta, Georgia, USA 3 Dana-Farber Cancer Institute, Boston, Maryland, USA OA01.05 Prevention of Systemic Infection by a Multigenic Recombinant Protein Vaccine after Heterologous R5 Clade C SHIV Challenge: Correlates of Protection

Background: Data from recent clinical trials imply that effective AIDS vaccines should recruit both the humoral and cellular arms of the adaptive immune system. Vaccine efficacy testing in primates should also use challenge viruses that are heterologous to the virus used for vaccine preparation since no human vaccinee is likely to be exposed to HIV strains exactly matching his/her vaccine.

Methods: We tested a bimodal vaccine strategy designed to induce cellular as well as neutralizing antibody (nAb) responses; efficacy testing employed multiple low-dose, upfront heterologous R5 clade C simian-human immunodeficiency virus (SHIV) challenges by the intrarectal (i.r.) route in rhesus monkeys (RM). Two groups of RM were vaccinated with HIV Tat, multimeric clade C HIV1084i gp160 and Gag-Pol particles of either SIVmac239 or SIVmne origin. All RM were challenged i.r. with SHIV-1157ipEL-p, a clade C virus heterologous to both the vaccine Env and the SIVmne Gag-Pol given to one of the groups. RM remaining virus-free after the 5th low-dose challenge were rechallenged with a single high-dose of SHIV-1157ipEL-p.

Results: Overall, one third of all vaccinees remained virus-free during the low-dose challenge phase; in contrast, 94% of controls were infected. Two of 4 vaccinees remained aviremic after high-dose rechallenge, with sterilizing immunity in one, and one vaccinee had only transient viremia. In contrast, all unvaccinated controls had high peak viral RNA loads. Among vaccinees that became infected after all viral challenges, onset of viremia and peak viremia were significantly delayed, and peak viral RNA loads were significantly lower compared to control RM. At the time of initial virus challenge, both virus-specific nAb and cellular immune responses were inversely correlated with peak viremia levels (p < 0.001).

Conclusion: The results indicate that protein-only vaccination can protect against heterologous virus exposure and support the notion that both humoral and cellular virus-specific immunity should be generated.

1 Oregon Health and Sciences University, Beaverton, Oregon, USA 2 MCB Program, University of Washington, Seattle, Washington, USA 3 Institute for Protein Biochemistry, Naples, Italy 4 Seattle Biomedical Research Institute, Seattle, Washington, USA OA01.06 Sustainable high titer neutralizing antibodies are rapidly induced by novel Virus-Like Particles when combined with DNA vaccines in rabbits

Background: A major challenge for HIV vaccine design has been to identify antigen presentation/delivery systems capable of eliciting strong, persistent immunity that is effective against a broad range of isolates. To focus the immune response on conserved determinants, we have designed virus like particles (VLPs) displaying one of three conserved neutralizing determinants of Env: (1) the third hypervariable region (V3); (2) gp41 extracellular domain (gp41EC); and (3) the gp41 membrane proximal region (MPR). The VLPs are based on the acyltransferase component (E2 protein) of the pyruvate dehydrogenase complex of Geobacillus stearothermophilus. The E2 protein self-assembles into 60mer particles with icosahedral symmetry, displaying 60 epitopes on each VLP.

Methods: We expressed four N-terminal fusion protein constructs: two versions of Env(V3)-E2, Env(gp41EC)-E2 and Env(MPR)-E2. Recombinant E2 monomers were isolated from E.coli inclusion bodies, refolded separately into 60-mer particles, and purified as VLPs by gel filtration.

Results: Immunization of rabbits with gp160 DNA (intradermally, via Gene Gun) or individual E2 fusion proteins (intramuscularly) 2-3 times did not produce significant neutralizing antibodies (NAbs). However, all Env-E2-immunized rabbits developed strong NAbs when co-administered with gp160-encoding plasmid DNA. Remarkably, high levels (10³) of NAbs were generated after only two immunizations within six weeks, levels typically not seen until after 20 + weeks and many more immunizations. These titers were sustained with three additional inoculations for six months, when the experiment was terminated. NAbs generated from rabbits immunized with V3-E2 particles are competed by the autologous V3 peptide, and we are currently exploring the breadth and specificity of responses to MPR and gp41.

Conclusion: Multiple HIV-1 epitopes displayed on E2 VLPs co-administered with DNA rapidly induced high levels of NAbs that focused the immune response on conserved epitopes. Supported by NIH-R01AI AI074379 and the Ruth Kirschstein T32 Viology Training Grant.

1 University of Stellenbosch, Johannesburg, South Africa 2 FIT Biotech, Tampere, Finland 3 National Institute for Communicable Diseases, Johannesburg, South Africa 4 Pharma, Turku, Finland OA01.07 Indicators of therapeutic vaccine effect using GTU®-MultiHIV B clade DNA in treatment-naïve subtype C HIV-1 infected subjects

Background: Therapeutic vaccination is an important adjunct to the management of HIV infected individuals to decrease transmission and slow down progression of disease. The final report using a plasmid DNA comprising complete sequences of Rev, Nef, Tat, p17/p24 proteins, and epitope stretch of previously identified T cell epitopes in pol and env of a subtype B-HIV-1 Han-2 isolate is presented.

Methods: 60 subtype C HIV-1 infected individuals with CD4 350 × 106/L and viral loads 50 copies/ml or more at screening were randomly allocated to vaccine or placebo at 2:1 and received either ID(0.5 mg/dose) or IM(1 mg/dose) immunizations using the Biojector. Vaccinations were administered at 0, 1 and 3 months, with boosts at 19 and 20 months with 1 mg/dose (ID) and 2 mg/dose (IM). Clinical endpoints were viral loads and CD4. Immunogenicity was evaluated using IFN-γ ELISPOT reactivity to B and C consensus peptide pools. HLA typing was performed with sequence-based typing.

Results: IM vaccine delivery resulted in a significant increase in CD4 counts (p = 0.013) and a 0.5 log decrease in viraemia (p = 0.002) at 108 weeks of follow-up. Vaccine effect was emphasized by a decrease in viral load and increase in CD4 after boosting at 64 weeks. HLA typing also supported the vaccine effect after the exclusion of known HLA types which are associated with slower progression. T cell responses recognizing Gag and CTL epitopes in B and C peptide pools were observed after ID vaccination rather than IM. There were no vaccine related serious adverse events and mild to moderate reactogenicity.

Conclusion: There was a significant reduction in viral load and increased CD4 counts after IM vaccination. The concept was further demonstrated by responses after boosting and in data controlled for HLA type. The investigational HIV-1 vaccine, GTU®-MultiHIV B clade is safe, well tolerated and clinically effective.

1 Biomedical Primate Research Centre, Rijswijk, Netherlands 2 Centro Nacional de Biotecnología-CSIC, Madrid, Spain 3 Arizona State University, Tempe, Arizona, USA 4 University of Lausanne, Lausanne, Switzerland 5 University of Cambridge, Cambridge, United Kingdom OA01.08 Improved immunogenicity against HIV-1 clade C antigens by using NYVAC-C with restored replication competence

Background: In order to improve the immunogenicity of currently available non-replicating pox virus HIV vaccine vectors, NYVAC was genetically modified through re-insertion of two host range genes (K1L and C7L), resulting in restored replicative capacity in human cells.

Methods: In the present study these vectors, expressing either a combination of the HIV-1 clade C antigens Env, Gag, Pol, Nef, or a combination of Gal, Pol, Nef were evaluated for safety and immunogenicity in rhesus macaques, which were immunized at weeks 0, 4 and 12 either by scarification (conventional poxvirus route of immunization), intradermal or by intramuscular injection (route used in previous vaccine studies).

Results: Replication competent NYVAC-C-KC vectors induced higher HIV-specific responses, as measured by IFN-γ ELISpot assay, than the replication defective NYVAC-C vectors. Application through scarification only required one immunization to induce maximum HIV-specific immune responses. This method simultaneously induced relatively lower anti-vector responses. In contrast, two to three immunizations were required when the NYVAC-C-KC vectors were given by intradermal or intramuscular injection and this method tended to generate slightly lower responses. Responses were predominantly directed against Env in the animals that received NYVAC-C-KC vectors expressing HIV-1 Env, Gag, Pol, Nef, while Gag responses were dominant in the NYVAC-C-KC HIV-1 Gag, Pol, Nef immunized animals.

Conclusion: The current study demonstrates that NYVAC replication competent vectors were well tolerated and showed increased immunogenicity as compared to replication defective vectors. Further studies are needed to evaluate the most efficient route of immunization and to explore the use of these replication competent NYVAC vectors in prime/boost combination with gp120 protein-based vaccine candidates. This study was performed within the Poxvirus T-cell Vaccine Discovery Consortium (PTVDC) which is part of the CAVD program.

1 Emory University, Atlanta, Georgia, USA 2 Zambia Blood Transfusion Service and ZEHRP, Lusaka, Zambia 3 Microsoft Research, Los Angeles, California, USA 4 University of Alabama at Birmingham, Birmingham, Alabama, USA OA02.01 CTL Escape and Reversion in HIV-1 Gag, Pol and Nef in an African Subtype C cohort

Background: HIV immune escape is not random and follows a predictable mutational path in response to the HLA alleles carried by an individual. Understanding the dynamics between CTL epitope escape and reversion will further our knowledge of the immune pressures that drive viral evolution and diversity at both an individual and population level.

Methods: To address the question of CTL epitope escape and reversion, we examined HLA-linked polymorphisms in Gag, Pol and Nef in a large number (80) of epidemiologically linked transmission pairs from a Zambian subtype C cohort, where the genotype of the transmitted virus is known, for 24 months.

Results: The results of this analysis indicate that a surprising fraction (18% in Gag) of the total potential HLA-linked polymorphisms are transmitted from the donor to their epidemiologically linked recipients. Moreover, a large fraction (30%) could not be attributed to the donor HLA-I alleles (statistically linked HLA-I and cross-reactive HLA-I). In Gag, at 24 months post transmission, 12% of polymorphisms present represented de novo polymorphisms; however, only 41% were statistically linked to the recipient HLA-I, while 29% could have beens due to cross-reactive HLA and 30% were unable to be linked to an HLA-I allele. A similar trend was observed in Nef and Pol. Also, in Gag, the majority of the de novo polymorphisms were selected during the first year, while, in contrast, in Nef and Pol the majority of de novo polymorphisms were not selected during the first 12 months of infection. Finally, at 24 months post-transmission only a small fraction of transmitted polymorphisms (11% in Gag, 23% in Nef and 5% in Pol) reverted to consensus.

Conclusion: These data demonstrate the importance of transmission pairs in accurately assessing HLA linked viral evolution in a newly infected individual and allow the accurate definition of the kinetics of immune driven HIV-1 variation.

1 Duke University Medical Center, Durham, North Carolina, USA 2 Los Alamos National Laboratory, Theoretical Division, Los Alamos, New Mexico, USA 3 University of Oxford, The Jenner Institute, Compton, United Kingdom 4 Weatherall Institute of Molecular Medicine, Oxford University, Oxford, United Kingdom 5 University of Alabama at Birmingham, Birmingham, Alabama, USA 6 University of Pennsylvania, Philadelphia, Pennsylvania, USA 7 Department of HIV Immunology, National Institute for Communicable Diseases, Johannesburg, South Africa 8 Vaccine Research Center/NIH, Bethesda, Maryland, USA OA02.02 High magnitude and multifunctionality epitope-specific CD8 + T-cell responses are associated with selection of escape mutants in acute HIV infection

Background: We sought to evaluate the early profile and impact of multifunctional HIV-1-specific CD8 + T-cell responses on the selection of escape mutations during acute HIV-1 infection.

Methods: We studied the longitudinal functional profiles of 23 epitope-specific CD8 + T-cell responses detectable in 7 HAART-naive patients following acute infection. The epitopes represented Env, Gag, Pol, Nef, and other regulatory gene products. Intracellular cytokine staining was used to characterize the functional profile of the Ag-specific memory CD8 + T-cells, evaluating their ability to degranulate and produce IFN-γ, IL-2, MIP-1β, and TNF-α. Cells were stimulated using individual 18-aa peptides representing the sequence of the autologous transmitted HIV-1 isolates.

Results: The 23 epitopes were defined as early (e-ME; n = 12) and late (l-ME; n = 6) mutating, if analysis of the sequences revealed mutations within 55 or 87 weeks following onset of symptoms (WFOSx), respectively; five were non-mutanting (nME). Overall, a higher magnitude of CD8 + T-cell responses was detected at the earliest time points to e-ME compared to the l-ME and nME (p = 0.03). This association was not observed when the highest response magnitude recorded at any time point was analyzed. Moreover, the order in which epitopes mutated within each individual was significantly associated with the total frequency of multifunctional epitope-specific T-cells (2-5 functions; p = 0.018) but not with the frequency of monofunctional subsets (p = 0.117). The MIP-1β + CD8 + T-cell subsets constituted more than 75% of the total CD8 + response detected within the first 24 WFOSx. We also observed that early appearance of escape mutants was associated with high entropy of the epitope (p = 0.003).

Conclusion: The data reveal an association between the appearance of epitopes mutating during acute/early HIV infection and both the magnitude and multifunctionality of the epitope-specific T-cell response and epitope entropy. The high contribution of MIP-1β-producing CD8 + T-cells to early responses suggests that mechanisms other than cytotoxicity could be exerting immune pressure.

1 Instituto de Higiene e Medicina Tropical, Lisboa, Portugal 2 Instituto Superior de Ciências da Saúde Egas Moniz and URIA/CPM/FFUL, Monte de Caparica and Lisboa, Portugal 3 Swedish Institute for Infectious Disease Control, Karolinska Institute, Stockholm, Sweden 4 Serviço de Doenças Infecciosas, Hospital de Santa Maria, Lisboa, Portugal OA02.03 Escape from neutralization is a frequent event in HIV-2 infection and is strongly associated with X4 tropism

Background: HIV-2 infection induces the production of a broad neutralizing antibody (NAb) response and this has been linked to a better control of viral replication and disease progression. However, little is known about the neutralizing targets and the dynamics of the NAb response and viral escape from neutralization during HIV-2 infection. This study was set out to investigate these issues.

Methods: A cohort of 28 HIV-2 infected patients followed during 4 years was analyzed. Autologous and heterologous neutralizing activities were determined with purified IgG antibodies in a TZM-bl cells-based assay against patient uncloned isolates. Coreceptor usage of the viruses was determined in GHOST cells. The sequence of the C2, V3 and C3 env regions was determined for all patients.

Results: Twenty-four new primary isolates were obtained from 12 patients. Most patients (8 out of 12) were infected with R5-viruses; the remaining 4 patients harbored X4-viruses. Autologous NAbs (median IC₅₀ = 3.91 microgram/ml; range = 0.049-38 microgram/ml) were detected only in 6 patients, all infected with R5-viruses. Most (4/6) patients unable to produce autologous NAbs were infected with X4-viruses. All but one patient produced NAbs targeting heterologous R5-viruses. The heterologous NAb response differed widely in potency between patients (median IC₅₀ = 4.14 microgram/ml; range = 0.049–49.08 microgram/ml). Remarkably, however, none of the patients produced antibodies against X4-viruses. A limited set of amino acids located in the V3 loop was associated with X4 tropism and resistance to neutralization.

Conclusion: Heterologous neutralization in HIV-2 patients is restricted to R5 isolates. Escape from autologous neutralization is a frequent event in HIV-2 infection and is strongly associated with X4 tropism. Viral determinants of neutralization resistance and X4 tropism seem to be located in the V3 region. These results have clear implications for the design of HIV-2 vaccine immunogens able to elicit the production of broadly reactive neutralizing antibodies.

1 Ragon Institute of MGH, MIT and Harvard, Charlestown, Maryland, USA OA02.04 Modulation of cytolytic activity of CD8+ T cells by HIV-1-specific IL21+ CD4+ T cells

Background: CD4+ T cell helper cells are required for the maintenance of long-lived and effective CD8+ T cell responses, but the role of CD4+ help in the control of chronic persistent infections is unknown. Understanding exactly how CD4+ T helper cells coordinate the immune system and respond to immune challenges is likely crucial for the development of an effective HIV vaccine.

Methods: Elite controllers, chronic progressors, treated or untreated subjects with primary HIV-1 infection (PHI) STI were studied. Transwell-Luminex and a modified intracellular cytokine staining was used to detect IL21 cytokine production by CD4+ T cells stimulated with overlapping HIV peptide sets. The effect of IL21 signals on HIV-specific CD8+ T cells was measured by flow cytometry and qPCR. The inhibitory capacity of HIV-1-specific CD8+ T cells in the presence of IL21 secretion by CD4+ T cells was studied in a viral inhibition assay directly ex vivo.

Results: HIV-specific triggering of IL-21 by CD4+ T-cells was significantly enriched in subjects spontaneously controlling HIV-1 infection. Responses were mediated by recognition of discrete epitopes largely in the Gag protein, often in the absence of IFNγ secretion, and were nearly completely absent in subjects with persistent persistent viremia and progressive disease (p = 0.0007). The majority of HIV-1-specific IL21 responses by CD4+ T cells was directed against epitopes in Gag, but was not related to HIV-1-specific IFNγCD4+ T cell responses. Moreover, IL21-secreting CD4+ T cell responses were able to enhance perforin and cytokine production by HIV-1-specific CD8+ T cells from chronic progressors, and HIV-1-specific effector CD8+ T cells showed up to 1000 fold enhanced ability to efficiently inhibit viral replication in-vitro after IL-21 binding.

Conclusion: Our data demonstrate that HIV-1-specific IL-21 + CD4+ T cell responses are critical for the control of viral replication in humans and likely to be of great importance for vaccine design.

1 University of Pennsylvania, Philadelphia, Pennsylvania, USA 2 University of California at San Francisco, San Francisco, California, USA OA02.05 A Potential Role for T-bet in Eliciting the Increased Cytotoxic Potential Demonstrated by HIV-specific CD8+ T-cells in HIV Elite Controllers

Background: Recent data from several groups suggest that perforin-mediated CD8⁺ T-cell effector activity may be an important component to the control of HIV replication in elite controllers (EC). However, the factors regulating perforin expression in human CD8⁺ T-cells remain relatively unclear. Recent findings from murine infection models indicate that one critical element of effector differentiation is the transcription factor T-bet, which is known to promote expression of IFN-γ, perforin, and granzyme B.

Methods: Using polychromatic flow cytometry, we assessed the relationship between T-bet and perforin, granzyme A, granzyme B, and granulysin expression in HIV Gag- and Nef-specific CD8⁺ T-cells among EC (n = 20; undetectable viral load) and chronically-infected progressors (n = 18; viral load > 10,000 copies/mL). All subjects had CD4⁺ T-cell counts above 350 cells/mm³.

Results: As we have recently reported, HIV-specific CD8⁺ T-cells from EC demonstrated a superior ability to express perforin as well as higher levels of granzyme B. However, we detected no differences in the levels of granzyme A or granulysin. Notably, we observed higher levels of T-bet in HIV-specific CD8⁺ T-cells from EC (p = 0.02). Upon proliferation of antigen-specific CD8⁺ T-cells, the kinetics of T-bet upregulation mirrored that of perforin and granzyme B. For both Gag- and Nef-specific CD8⁺ T-cells, there was a positive correlation between the levels of T-bet and either perforin or granzyme B expression (p < 0.05; Spearman test). Finally, lowering viral load with HAART (n = 19) does not itself rescue levels of T-bet, perforin, or granzyme B to those observed in EC.

Conclusion: Rapid and continuous perforin and granzyme B expression following initial antigen encounter allows for the sustained cytotoxic potential of HIV-specific CD8⁺ T-cells in EC. Our results suggest that T-bet may be playing an important role in driving effector function. Collectively, these data imply that strategies to modulate this transcription factor could lead to enhanced HIV-specific CD8⁺ T-cell effector activity.

1 Vanderbilt University, Nashville, Tennessee, USA 2 Division of Adult Infectious Diseases, Department of Medicine, Nashville, Tennessee, USA OA02.06 Dominant clonotypes within HIV-specific T cell responses display an exhaustion phenotype and reduced epitope cross-reactivity

Background: In untreated HIV infection, higher viral loads are associated with an exhaustion phenotype on T cells characterized by high levels of PD-1 and low levels of IL-7Rα expression. HIV-epitope-specific T cell responses are often comprised of clonotypic expansions with distinct functional properties.

Methods: We measured PD-1 and IL-7Rα expression, MHC-I tetramer binding characteristics, and cytokine production profiles on dominant and sub-dominant T cell receptor clonotypes within 35 epitope-specific responses in 22 HIV + individuals to better understand the relationship between the structural composition of the HIV-specific T cell repertoire and T cell exhaustion.

Results: Dominant clonotypes within epitope-specific TCR repertoires were characterized by higher PD-1 expression (p = 0.0003) and lower IL-7Rα expression (p = 0.007) when compared to sub-dominant clonotypes. TCR avidity only weakly correlated with dominance within the clonotypic repertoire (p = 0.09) in a paired comparison. Despite a phenotype consistent with cellular exhaustion, dominant clonotypes were capable of IFN-γ and TNF-α production when stimulated with consensus peptides. In response to variant peptides, however, sub-dominant clonotypes produced higher relative levels of cytokines and displayed greater cross-recognition compared to dominant clonotypes.

Conclusion: Our findings suggest that polyclonal HIV-specific responses may retain sub-populations of T cells capable of highly functional responses and effective suppression of viral variants.

1 INSERM SC10, Villejuif, France 2 Université Paris-Descartes; INSERM CIC BT 505; AP-HP, Hôpital Cochin, Paris, France 3 Hôpital Henri Mondor, Université Paris-Est Créteil, INSERM U955, Créteil, France 4 AP-HP, Hôpital Tenon, Paris, France 5 CHU Hôtel-Dieu, Nantes, France 6 Hôpital Purpan, Toulouse, France 7 CHU Sainte-Marguerite, Immuno-hématologie clinique, Marseille, France 8 ANRS, Paris, France 9 INSERM U912, IRD, Université de la Méditerranée–Aix-Marseille II, Marseille, France OA03.01 Psychosocial consequences of participation in HIV preventive vaccine trials: a cross-sectional study in ANRS COV1-COHVAC cohort

Background: ANRS COV1-COHVAC cohort is a prospective multicenter cohort study of participants in preventive HIV-vaccine phase I and II trials in France, dating back to 1992 to 2008 with various vaccine candidates. The purpose of this study is to examine psychosocial consequences of trial participation including negative experiences as well as changes in risk behaviours in healthy low HIV risk volunteers.

Methods: In this cross-sectional study, an anonymous questionnaire was completed by vaccine volunteers at enrolment in the prospective cohort. The questionnaire dealt with aspects of socioeconomic conditions, coping strategies, personal and social consequences of trial participation and unsafe sexual practices.

Results: Started in December 2008, enrolment comprises 250 vaccine volunteers as of April 2010. Questionnaires were completed by 235 volunteers at six vaccine clinical sites. The median duration between trial entry and the first follow-up visit was 5.2 years (2.2 to 17.8). The median age was 52 years (25 to 78), 51.5% were male. Most of the subjects had graduated from high school (86%), were employed (79%) and lived in couple (70%). A minority of volunteers (13%) experienced trial-related problems, in relationships with spouse (5%), family (4%) or with employment (4%). Four subjects (2%) reported issues with insurance. All but 5 volunteers reported no regret (98%) and 48% no constraints at all for participating. Report of constraints was significantly associated with mucosal vaccine administration route, after adjustment on socio-demographic variables. Nineteen volunteers (8%) had casual partners within the past year; including 9 who reported unprotected intercourse with casual partner. Four subjects reported unsafe sex with HIV-positive or unknown HIV status partners.

Conclusion: Regret and perceived negative consequences were infrequent among HIV vaccine trial participants. Moreover very few subjects experienced unsafe sex several years after their selection as low HIV risk volunteers. HIV vaccine trials are well accepted and objectives understood by most participants.

1 University of Toronto, Toronto, Ontario, Canada OA03.02 “Speaking the dialect”: Understanding public discourse in the aftermath of an early HIV vaccine trial termination

Background: Amidst a history of sometimes contentious HIV chemoprophylaxis trials, scant investigations have addressed the social processes and outcomes of HIV vaccine trials. The purpose of this investigation was to explore community reactions and narratives in the aftermath of the early termination of the Step study HIV vaccine trial in order to inform future trial recruitment and implementation.

Methods: From May 2008 to April 2009, we conducted 9 focus groups in Ontario, Canada with African-Caribbean women, men who have sex with men, female sex workers, injection drug/crack using men and women, and Aboriginal men and women; and 6 key informant interviews with community leaders and healthcare providers. All groups/interviews were digitally recorded, transcribed and analyzed using NVivo software and narrative thematic techniques.

Results: Focus group participants' (n = 72) mean age was 39.5 years. Most (60%; n = 43) were women and half (50%; n = 36) identified as gay, lesbian or bisexual. About one-third (35%) identified as Black/African/Caribbean, one-third (32%) white, one-fifth Aboriginal, 7% Latino, and 5% Asian/mixed. Participants revealed mental models and social interpretations of HIV vaccine trials in striving to make sense of the Step study. We uncovered a series of disjunctures between public and biomedical interpretations of the same phenomena (e.g., trial recruitment, informed consent, risk behaviors and HIV infections among trial participants, dissemination of results) that reveal a coherent, alternate narrative to biomedical discourse.

Conclusion: Public discourse on HIV vaccine trials reveals active interpretation of complex clinical trial processes and outcomes in the context of existing beliefs, conceptions and experiences viz. HIV vaccines, medical research and historical disenfranchisement. Formative research to explore and uncover social meanings and mental models in public discourse on HIV vaccines trials may provide an empirical foundation for knowledge translation and community engagement strategies to support the long-term process of HIV vaccine development.

1 Armed Forces Research Institute of Medical Science (AFRIMS), Bangkok, USA 2 Armed Forces Research Institute of Medical Science (AFRIMS), Bangkok, Thailand 3 Armed Forces Research Institute of Medical Science (AFRIMS), Bangkok, Thailand 4 Forces Research Institute of Medical Science (AFRIMS), Bangkok, Thailand 5 Armed Forces Research Institute of Medical Science (AFRIMS), Bangkok, Thailand 6 Ministry of Public Health-Thai AIDS Vaccine Evaluation Group, Bangkok, Thailand OA03.03 False HIV Seropositive among uninfected HIV vaccine recipients of phase III vaccine trial in Thailand

Background: HIV vaccine induced false positive HIV serologic test result can lead to a former participant being denied employment, international travel, life and health insurance and from donating blood. The double-blind, placebo-controlled Prime-Boost HIV Vaccine Phase III Trial conducted in Thailand enrolled 16,395 HIV non-infected volunteers who received four injections (0, 4, 12, 24 weeks) of recombinant canarypox vector (ALVAC) vaccine and two (12, 24 weeks) recombinant gp120 subunit vaccines (AIDSVAX) or 6 injections of placebo; HIV diagnostic tests were performed at the completion of vaccination series and every six months thereafter for 3 years. This study aims to explore frequencies of false positive HIV serology among trial participants.

Methods: HIV diagnosis was based on ELISA (Vironostika HIV-1 Mocroelisa System) and Western blotting (GS HIV-1 Western Blot) with confirmatory nucleic acid test (Amplicor HIV Monitor version 1.5). Specimens which demonstrated false positive (verified by nucleic acid test) were tested later by Thai FDA-licensed test kits, Genscreen HIV-1/2 V.2., Determine HIV-1/2, Serodia HIV and Immunocomb II HIV1&2.

Results: Of 39,789 HIV non-infected samples from vaccine group 55 (0.14%) were falsely reactive by Vironostika HIV-1 Microelisa system. Of the 55 false reactive samples 19 samples (34.55%) were reactive by Genscreen HIV-1/2 V.2, 3 samples (5.45%) reactive by Immunocomb II HIV 1&2 and only 1 sample each reactive by either (1.82 %) Determine HIV-1/2 or Serodia HIV. No sample tested false positive by all 3 Thai-FDA test kits. The false positive result was highest at 6-month post vaccination and lowest at 36-month post vaccination period.

Conclusion: False positive HIV serology induced by study vaccines was extremely low and declined post vaccination over time.

1 Faculty of Tropical Medicine, Mahidol University, Bangkok, Thailand, Bangkok, Thailand 2 Thai Component–Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand 3 EMMES Corporation, Rockville, Maryland, USA 4 U.S. Army Medical Component–AFRIMS, Bangkok, Thailand 5 U.S. Military HIV Research Program, Rockville, Maryland, USA 6 Department of Disease Control, Ministry of Public Health, Nontaburi, Thailand OA03.04 Exploratory Age-stratified Analysis of Risk-taking Behaviors and Trial Participation Outcomes in the Thai Phase III HIV Vaccine Trial

Background: RV144 Trial revealed possible protection from HIV infection; efficacy was 26.2% (−13.3% to 51.9% 95% CI) in per-protocol; 31.2% (1.1% to 52.1%) in modified-intention-to-treat analysis (mITT) excluding infected participants at randomization. Sample size of 16402 was not powered for subgroup analyses; plus low incidence (132 events) requires cautious interpretation. With this caveat, mITT analyses suggested different protection levels: 49.0% (12.8% to 70.2%) for age 20-25 but −7.1% (−14.3% to 52.7%) in <20, and 18.7% (−49.3% to 55.7%) in > 25. This study explored age-stratified risk-taking behavioral markers and trial participation outcomes.

Methods: The trial used questionnaires to assess risk behaviors including injecting drug use and unsafe sexual behaviors during scheduled visit intervals. Baseline HIV infection risk was classified into 3 levels. Selected adverse events, road accidents were evaluated as markers of risk-taking behavior.

Results: There were 27.7%, 44.8%, 27.5% of participants aged <20, 20–25, >25 respectively. Gender and marital status differed significantly—62%, 65%, 57% were males; 33%, 52%, 65% married. Younger groups had higher education. Each age-group had similar risk levels: 47% low, 28% moderate, 24% high. AEs in each subgroup were similar; 31% had at least one injury, 15% from road accidents. Of 2389 road accident cases: 45% (20-25) and 27% (<20/>25). Similarly, among 160 participants who died: 44% (20-25) and 28% (<20/>25). Dropout rates were 11%, 10%, 7.7%. Infection rates among the low-risk age-groups were 0.47%, 0.56%, 0.81%; among medium-risk, 0.17%, 1.05% 0.84%; and high-risk, 1.01%, 1.15%, 1.40%.

Conclusion: The trial suggested possibility of disparate efficacy by age-groups; however, differences in characteristics and risk taking patterns were not observed. It was not possible in this study to conclude that trial participants were heterogeneous and age was an effect modifier. The participant's risk markers and outcomes explored in this study could help the design and analyses of future trials.

1 HIV Vaccine Trials Network, Seattle, Washington, USA 2 Vaccine Clinical Research Branch, VRP, DAIDS, NIAID, NIH, DHHS, Bethesda, Maryland, USA 3 Merck & Co. Inc, North Wales, Pennsylvania, USA 4 HIV Research Section, San Francisco Department of Public Health, San Francisco, California, USA OA03.05 Analysis of the relative risk of HIV acquisition among Step study participants with extended follow-up

Background: Initial analysis of Step study data indicated that non-efficacy criteria were met; vaccinations were halted and participants unblinded to treatment assignment. In posthoc analyses, more HIV infections were seen among men who received MRK Ad5 gag, pol, nef HIV vaccine vs. placebo recipients. Extended follow-up assessed relative risk of HIV acquisition and whether it waned with time from vaccination.

Methods: Participants were followed for 4 years since first vaccination or until 12/31/2009, whichever came first. Cox proportional hazard models were used to estimate relative risk of HIV adjusted for baseline covariates. Constancy of vaccine effect over time was assessed in Cox models. We used logistic regression modeling to evaluate predictors of incomplete follow-up and sensitivity analysis to assess possible bias due to loss-to-follow-up.

Results: Fifteen infections occurred among 1134 women and 168 infections among 1836 men. This analysis is restricted to males, in whom HIV incidence was 3.63/100 py (95% CI: [2.92, 4.47]) within 18 months after enrollment and 2.88/100 py (95% CI: [2.28, 3.59]) thereafter. Vaccine effect on HIV acquisition did not differ by baseline Ad5 titer but did differ significantly by self-reported circumcision status (p = 0.03, adjusted Cox model). Specifically, there was no significant vaccine effect among circumcised men (hazard ratio [HR, vaccine: placebo] = 1.24, 95% CI [0.83, 1.85]); the vaccine effect diminished over time among uncircumcised men with a treatment effect of HR = 4.04 [1.65, 9.88] during the first 18 months after enrollment and HR = 1.07 [0.5, 2.29]) thereafter. Loss-to-follow-up was greater among placebo recipients (p = 0.07), especially around time of unblinding. However, sensitivity analyses assuming higher or lower hazards among lost participants showed minimal impact on overall inference.

Conclusion: The Step vaccine did not prevent HIV infection. The increased relative risk of infection initially observed in vaccine recipients waned over time, especially among uncircumcised men.

1 Centre Hospitalier Universitaire Henri Mondor, Créteil, France 2 Centre Hospitalier Universitaire Vaudois (CHUV), Lausanne, Vaud, Switzerland 3 MRC Clinical Trials Unit, London, United Kingdom 4 Centre Hospitalier Universitaire Henri Mondor, Créteil, France 5 CIC Vaccinologie Hôpital Cochin, Paris, France 6 University of Regensburg, Regensburg, Germany 7 Imperial College, St. Mary`s Hospital, London, United Kingdom 8 INSERM U593/Clinical Trials Unit, Bordeaux, France OA03.06 Optimal priming of poxvirus vector (NYVAC)-based HIV vaccine regimens requires 3 DNA injections. Results of the EV03/ANRS Vac20 Phase I/II Trial

Background: Two or three DNA primes have been used in previous smaller clinical trials, but the number required for optimal priming of viral vectors has never been assessed in adequately powered clinical trials. The EV03/ANRS Vac20 phase I/II trial investigated this issue using the DNA prime/poxvirus NYVAC boost combination, both expressing a common HIV-1 clade C immunogen consisting of Env and Gag-Pol-Nef polypeptide.

Methods: 147 healthy volunteers were randomly allocated through 8 European centres to either 3xDNA plus 1xNYVAC (weeks 0, 4, 8 plus 24; n = 74) or to 2xDNA plus 2xNYVAC (weeks 0, 4 plus 20, 24; n = 73), stratified by geographical region and sex. T cell responses were quantified using the interferon γ Elispot assay and 8 peptide pools; samples from weeks 0, 26 and 28 (time points for primary immunogenicity endpoint), 48 and 72 were considered for this analysis.

Results: 140 of 147 participants were evaluable at weeks 26 and/or 28. 64/70 (91%) in the 3xDNA arm compared to 56/70 (80%) in the 2xDNA arm developed a T cell response (P = 0.053). 26 (37%) participants of the 3xDNA arm developed a broader T cell response (Env plus at least to one of the Gag, Pol, Nef peptide pools) versus 15 (22%) in the 2xDNA arm (P = 0.047). At week 26, the overall magnitude of responses was also higher in the 3xDNA than in the 2xDNA arm (similar at week 28), with a median of 545 versus 328 SFUs/106 cells at week 26 (P < 0.001). Preliminary overall evaluation showed that participants still developed T-cell response at weeks 48 (78%, n = 67) and 72 (70%, n = 66).

Conclusion: This large clinical trial demonstrates that optimal priming of poxvirus-based vaccine regimens requires 3 DNA regimens and further confirms that the DNA/NYVAC prime boost vaccine combination is highly immunogenic and induced durable T-cell responses.

1 Muhimbili University of Health and Allied Sciences, Dar es Salaam, Tanzania, United Republic of 2 National Institute for Medical Research, Dar es Salaam, Tanzania, United Republic of 3 Muhimbili National Hospital, Dar es Salaam, Tanzania, United Republic of 4 Swedish Institute for Infectious Disease Control (SMI), Solna, Sweden 5 National Institute of Allergy and Infectious Diseases/NIH, Bethesda, Maryland, USA 6 Walter Reed Army Institute for Research (WRAIR), Rockville, Maryland, USA 7 Swedish Institute for Infectious Disease Control/Karolinska Institutet, Stockholm, Sweden 8 Karolinska Institutet/Södersjukhuset, Stockholm, Sweden OA03.07 Broad and strong immune responses in a trial of a heterologous DNA prime MVA boost HIV vaccine among healthy Tanzanian volunteers

Background: A phase I/II HIV vaccine trial using a multiclade, multigene HIV-1 DNA prime with two heterologous HIV-1 MVA boosts has been conducted among healthy adults in Dar es Salaam, Tanzania.

Methods: Sixty HIV-uninfected volunteers randomised to three groups of 20, received DNA plasmids expressing HIV-1 gp160 subtypes A,B,C;revB;p17/p24gagA,B and Rtmut at 1 mg intradermally (i.d.) or 3.8 mg intramuscularly (i.m.) or placebo at months 0, 1 and 3 using a needle-free injection device. They were boosted i.m. with 10⁸ pfu of recombinant MVA expressing HIV-1 env, gag, pol of CRF01_AE or placebo at months 9 and 21. HIV-specific interferon (IFN)-gamma ELISpot, lymphoproliferation assay (LPA) and 4-colour intracellular cytokine staining (ICS) responses were determined.

Results: The vaccines were well tolerated. Two to four weeks after the second HIV-MVA boost, 28 (97%) of 29 vaccinees had positive IFN-gamma ELISpot responses, 27 (93%) to Gag and 23 (79%) to Env peptides. After the first HIV-MVA boost, the i.d. primed vaccinees showed higher immune responses to Env compared to the i.m. primed vaccinees. Two weeks after the second HIV-MVA boost, all of 25 vaccinees (100%) showed strong positive LPA responses to each of the AT-2- treated HIV-1 antigens of clades B, A and A_E (from J Lifson, NCI, USA). Six months after the second HIV-MVA boost, 16 (89%) of 18 vaccinees had positive LPA responses. Testing by ICS for Gag-specific IFN-gamma/IL-2 production 4 weeks after the second HIV-MVA boost showed CD8 T-cell responses in 17 (59%) and CD4 T-cell responses in 16 (55%) of 29 vaccinees. Altogether 25 (86%) of 29 had CD8 or CD4 T-cell responses. All of 29 vaccinees exhibited vaccine-induced seropositivity, 28 (97%) to Gag and 19 (66%) to Env.

Conclusion: This heterologous HIV DNA prime MVA boost vaccine approach demonstrated broad and high immunogenicity in Tanzanian volunteers.

1 Swedish Institute for Infectious Disease Control and Karolinska Institute, Solna, Sweden 2 Laboratory of Viral Diseases, NIAID, NIH, Bethesda, Maryland, USA 3 Karolinska Institute, Stockholm South Hospital, Stockholm, Sweden 4 U.S. Military HIV Research Program, Walter Reed Army Institute of Research, Rockville, Maryland, USA OA03.08 Vaccine-induced seropositivity (VISP) following immunization of healthy individuals with HIV DNA and recombinant modified vaccinia virus Ankara

Background: Forty healthy Swedish participants were recruited to an HIV DNA (HIVIS) prime-recombinant modified vaccinia virus Ankara (MVA-CMDR) boosting vaccine protocol. Here we report the induction and duration of antibody responses.

Methods: Sera from eligible volunteers were evaluated after immunization with HIVIS DNA plasmids encoding Envelope of subtypes A, B and C; Gag A and B, and RT B, followed by two boosts with recombinant MVA-CMDR encoding Envelope E and Gag-Pol A. Serological responses were assayed by Abbott IMX/HIV-1/HIV-2/III, Enzygnost HIV Integral II, Western blot (WB), Elisa and virus neutralization assays.

Results: Following HIV DNA and a first MVA-CMDR boost, 57% of individuals displayed anti-p37Gag antibodies but only 5% anti-HIV-1 gp160 Env antibodies. Routine serological assays were positive in 7 individuals. Neutralizing anti-vector vaccinia antibodies were found in 62% of volunteers. Four weeks following a late (3 years) second boost with MVA-CMDR, 96% developed antibodies against Env antigen and all developed anti-Gag antibodies. In addition, 40% had RT-specific antibodies. The reactivity translated into positive reactions in HIV diagnostic tests: 100% in Abbott IMX and 13/24 (54%) in Enzygnost. All of the latter also had two or more bands in WB, thus meeting the criteria for seropositivity. Importantly, 6 months after the last HIV MVA-CMDR boost, positive IMX (23/24) and Enzygnost (10/24) reactivities were still present. No HIV-1 clade B (SF162) or CRF01_AE (CM235) pseudovirus neutralizing antibodies were detected in 24 volunteers tested. Viral RNA quantification excluded HIV-infection in all volunteers.

Conclusion: The full prime-boost schedule with heterogeneous HIV antigens appears necessary to induce broad and durable humoral immune responses as well as for maintaining broad and strong cellular responses. The antibody responses were directed to most antigens in the vaccine and also detected in standard HIV diagnostic tests. Molecular methods are necessary to separate vaccine-induced reactivity from those of an infection.

1 University of Washington, Seattle,Washington, USA 2 Fred Hutchinson Cancer Research Center, Seattle, Washington, USA OA04.01 HIV-1 acute infection: evidence of selection?

Background: Characterizing HIV-1 populations in acute infection is critical to help devise HIV-1 vaccines.

Methods: We sequenced near-full-length genome (nflg) from 3-12 time points during the first year of HIV-1 infection of 11 individuals.

Results: HIV-1 subtype B evolution was primarily stochastic before signs of adaptive selection emerged about five weeks post-infection. Nonetheless, several findings suggested that selective processes may play a role at transmission or in acute infection. Using additional data, we found that 75% of homosexual transmissions of HIV-1 were established by a single lineage. Examining four transmission pairs, we noted that the restriction to a single founder was not due to a lack of variation in transmitter viruses, suggesting that the mucosal barrier impedes the establishment of infection by multiple variants. Additionally, although one would expect that sequences in the seroconverter would correspond to the most frequent variant in the transmitter, when the transmitter was in chronic infection (three pairs) we saw that rare variants established infection, leaving open the possibility that selection may affect the outgrowth of particular variants. When comparing mean pairwise nucleotide diversity across nflg, we observed a slight dip in nucleotide diversity that temporally overlapped with peak viremia. Interestingly, the dip in diversity paralleled a dip in the mean number of APOBEC-specific mutations per nflg. This suggests that the sequences found at the earliest time points may have been relatively enriched with APOBEC-induced mutations (yet not hypermutated), and that the viral population may have subsequently undergone a clonal purification related to these mutations.

Conclusion: Neither diversifying nor directional selection was detected in acute infection. Thus, it appears unlikely that the founder strain has adapted rapidly to its new host. Yet, perplexing evidence of selective pressure in acute infection warrant further evaluation of the selective forces governing the transition from a localized to a systemic infection.

1 University of Alabama at Birmingham, USA, Birmingham, Alabama, USA 2 University of Tennessee Knoxville, Knoxville, Tennessee, USA 3 NIH NIAID, Bethesda, Maryland, USA 4 Duke University Medical Center, Durham, North Carolina, USA 5 SAIC-Frederick, Frederick, Maryland, USA 6 Los Alamos National Laboratory, Los Alamos, New Mexico, USA OA04.02 Molecular targets and potency of HIV-1 neutralization revealed by dynamic assessment of transmitted/founder virus antibody recognition and escape

Background: To map early Nab epitopes, establish clinically relevant neutralization titers, and enable minimal estimates of the immune pressure afforded by Nab, we performed dynamic molecular analyses of Nab recognition and escape by transmitted/founder viruses and their progeny.

Methods: We identified, cloned and characterized the transmitted/founder viruses in three acutely-infected subjects and then used single genome amplification and parallel allele-specific sequencing to characterize env sequence evolution longitudinally. We performed JC53-bl Env pseudotype entry assays to assess Nab titers against the transmitted/founder viruses and their progeny. We mapped the Nab epitopes utilizing site-directed mutants and adsorption assays with peptides and autologous Env proteins. We then used mathematical modeling to characterize the kinetics, efficacy and fitness cost afforded by epitope-specific Nabs.

Results: In each subject, the transmitted/founder virus was identified and shown to exhibit heterologous neutralization patterns similar to other primary viruses. Infectivity assays with autologous plasma demonstrated Nab titers with IC50s of 1:40-1:2000 developing within 12–24 weeks post-infection. Escape mutations conferring Nab escape evolved simultaneously with, or preceded, the development of these Nab titers, and rapidly expanded to comprise 100% of the viral quasispecies. In each subject, initial Nab epitopes were mapped and shown to be conformational involving at least two areas of gp120. Mathematical modeling of virus escape from Nab provided a minimal replacement rate of the transmitted/founder virus with Nab escape variants of between 8.6 and 13.2%/day.

Conclusion: Early Nabs targeted conformational epitopes involving two or more env regions. Early Nab titers with IC50s of 1:40-1:2000 drove complete virus escape. Estimated rates of early viral Nab escape are slightly slower than those for CTL escape (Goonetilleke 2009). The equivalent kinetics and extent of virus escape from Nabs at titers ranging from 1:40-1:2000 suggest that such titers are sufficient for inhibiting de novo virus infection in vivo, and possibly in the setting of vaccination.

1 University of California, Irvine School of Medicine, Irvine, California, USA 2 University of Alabama at Birmingham, Birmingham, Alabama, USA 3 University of Pennsylvania, Philadelphia, Pennsylvania, USA 4 University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA 5 Duke University Medical Center, Durham, North Carolina, USA OA04.03 Early Appearance of ADCC- and ADCVI-mediating Antibody Responses Against Autologous HIV-1 Transmitted/Founder Virus

Background: Non-neutralizing Env antibodies might prevent or modulate HIV infection. Using autologous and heterlogous transmitted virus, we studied antibody dependent cell-mediated virus inhibition (ADCVI)- and antibody-dependent cellular cytotoxity (ADCC)-mediating Ab responses during acute HIV-1 infection.

Methods: ADCVI and ADCC activities were measured using longitudinal plasma samples from 3 CHAVI 001 acute infection subjects (CH040, CH058, and CH077) as early as 21 up to 588 days following onset of symptoms (DFOSx). Full-length infectious molecular clones representing the transmitted virus from each subject were used to produce target cells for ADCVI assays. Autologous recombinant CH040 gp140- and heterologous WITO gp120-coated and HXB2 chronically infected CEM.NKRCCR5 cells were used as target cells in the ADCC assay.

Results: ADCC-mediating antibody titers of 1:4,182, 1:1,266, and 1:5,445 to the CH040-transmitted gp140 antigen were detected at 21, 28, and 42 DFOSx in subject CH040, CH077, and CH058, respectively. ADCC responses capable of recognizing the WITO gp120 were detectable at the same timepoints. Plasma from CH058 displayed highest ADCC-mediating activity regardless of the target platform. ADCVI Ab responses against the autologous transmitted isolates were detectable at 137 DFOSx in CH040 and within 39 and 57 DFOSx in CH077 and 058, respectively. ADCVI reactivity increased with time from infection in all three subjects. When infected target cells were used in both assays, the ADCC and ADCVI responses appear to peak earlier in CH058, who was subsequently defined as a controller (virus load <2,000 copies/ml).

Conclusion: For both ADCVI and ADCC, autologous and heterologous Ab responses arise at similar times post HIV-1 transmission, unlike the longer delay observed between the appearance of autologous and heterologous neutralizing antibodies. Despite some differences in the timing of appearance, ADCVI and ADCC-mediating Ab targeting heterolgous HIV strains are part of the initial antibody response to HIV-1 infection and may be easy to elicit by vaccination.

1 University of KwaZulu Natal, Durban, KwaZulu Natal, South Africa 2 Ragon Institute of MGH, MIT and Harvard Infectious Disease Unit, Boston, Massachusetts, USA 3 Centre for the AIDS Programme of Research in South Africa (CAPRISA), Durban, KwaZulu Natal, South Africa 4 CAPRISA, Durban, KwaZulu Natal, South Africa 5 NCI, SAIC-Frederick, Frederick, Maryland, USA 6 National Institute of Communicable Diseases (NICD), Sandringham, Johannesburg, South Africa 7 NICD, Sandringham, Johannesburg, South Africa 8 Ragon Institute, Boston, Massachusets, USA OA04.04 Ligand Dependant Increase in KIR Gene Expression in Primary HIV-1 Infection

Background: The interaction between killer immunoglobulin like receptors (KIRs) and human leukocyte antigens (HLA) has been shown to play a key role in HIV-1 antiviral immunity but the exact mechanisms are not fully understood. We hypothesized that increased expression of KIR gene expression during primary HIV-1 infection is dependent upon the presence of the cognate putative HLA ligands and that this increased expression is associated with better control of HIV-1 during this early phase.

Methods: Matched samples from 12 black South African women were available before HIV-1 acquisition, during the acute phase (<3 months post-infection) and early chronic phase (3-12 months post-infection). HLA and KIR typing were performed using PCR-SSP technology. Activating and inhibitory KIR gene expression was measured by quantitative real-time RT-PCR pre-infection and at the various time points post infection and normalized to GAPDH expression. Statistical analyses were performed by fitting a generalized estimating equations model to the data.

Results: Inhibitory KIR genes showed higher relative expression levels compared to activating counterparts at all time points although the differences were not statistically significant. The inhibitory receptor, KIR 3DL1 showed significantly higher expression in the presence (n = 5) of HLA-B Bw40I ligand than in its absence (n = 5) (p = 0.01) post-infection. Furthermore, KIR expression, regardless of whether activating or inhibitory, trended to be higher in the presence of respective HLA class ligands with the differences increasing post infection (p = 0.07). Subjects with an activating KIR present with their respective HLA class I ligand had significantly lower viral loads (half a log less) than participants without an activating KIR/HLA ligand pair post HIV-1 infection (p < 0.0001).

Conclusion: Taken together, our findings support earlier findings of HLA-dependant KIR expansion in acute HIV-1 infection, and suggest a potential role for activating KIR/HLA ligand interaction in viral control during the acute to early chronic phase of HIV-1 infection.

1 Duke University Medical Center, Durham, North Carolina, USA 2 Los Alamos National Laboratory, Los Alamos, New Mexico, USA 3 University of Alabama at Birmingham School of Medicine, Birmingham, Alabama, USA OA04.05 Decrease in the Ratio of Infectious to Total Virions During Acute HIV-1 Infection

Background: A better understanding of the viral dynamics during the early phase of HIV-1 infection could be essential to the design and testing of an effective vaccine. We have studied the kinetics of virus infectivity during the acute phase of infection and characterized the neutralization properties of the “transmitted/founder” uncultured virus as compared to the corresponding PBMC-cultured and Env-pseudotyped viruses.

Methods: Infectious virus in closely spaced plasma samples from 11 HIV-1 infected individuals was quantified in TCID50 assays and compared to plasma viral RNA values. Autologous and heterologous neutralization was assessed for three subjects by using as virus inoculum either the earliest plasma sample, the PBMC-cultured plasma virus or the corresponding pseudotyped virus. All assays were performed in PBMC or TZM-bl cells.

Results: Infectious virions were detected at the earliest time points from all 11 subjects. Plasma TCID50: viral RNA ratios were generally low (peak range 0.0007 to 0.03) and behaved independently (r = 0.29 p = 0.07). A dramatic decline in plasma TCID50 was observed over time in most subjects that was not predicted by plasma viral RNA loads. An autologous neutralizing antibody response was detected in three subjects examined but the response did not coincide with the decline in plasma TCID50. Only minor differences in the autologous and heterologous antibody responses were observed with the different sources of virus.

Conclusion: These observations begin to define the biology and neutralization phenotype of transmitted/founder HIV-1 as it exists in plasma in vivo. The precipitous decline in infectious plasma virus prior to the initial downregulation of plasma viral RNA may offer clues about early immune responses that would benefit vaccines. In addition, the similar neutralization properties of uncultured, cultured, and Env-pseudotyped viruses derived from plasma strengthens the validation of current assay technologies.

1 New York University School of Medicine, New York, New York, USA 2 New York University School of Medicine and VA Medical Center, New York, New York, USA OA04.06 Evolution of HIV-1 Diversity in regions of High Viral diversity and implications for Vaccine design

Background: West-Central Africa is an epicenter of the HIV pandemic; endemic to Cameroon are HIV-1 viruses belonging to all (sub)subtypes and numerous Circulating Recombinant Forms (CRFs). The rural villages of Cameroon harbor many strains of HIV-1, though these areas are not as well monitored as the urban centers.

Methods: Eighty two specimens obtained in 2000 and 2001 from subjects living in the rural villages of the South and West Regions of Cameroon were subtyped in gag, pol, and env and compared to 90 specimens obtained in 2006–2008 in the same regions, in order to analyze HIV-1 evolution in these rural areas. RNA from plasma was amplified by RT-PCR, sequenced, and phylogenetically analyzed to determine subtype. Statistical analysis was performed by Two-way Chi-square test.

Results: It was found that in the South Region, the proportion of unique recombinant forms (URFs) remained constant (∼40%), while the amount of URFs containing fragments of a CRF increased by 25%. (Sub)subtypes A1, F2, H, and K, and CRF09_cpx, identified in 2000 and 2001, were replaced by CRFs 01_AE, 13_cpx, 14_BG, and 18_cpx in 2006-2008. In the West Region, (sub)subtypes A2, C, G, and H, and CRFs 01_AE and 09_cpx, identified in 2000-2001, were replaced by sub-subtype A1 and CRFs 25_cpx and 37_cpx in 2006-2008. The proportion of URFs in the West Region dropped significantly over the time period by 43%. In both Regions, the proportion of CRF02_AG increased at all loci.

Conclusion: These findings demonstrate that the evolution of HIV-1 is distinct for each endemic region, and suggests that the proportion of URFs containing CRF fragments is increasing as the genetic identity of the virus continues to shift dramatically. This highlights the concern that subtype-specific vaccines may not be relevant in Cameroon, and that the distribution of viral diversity in these regions of Cameroon must be carefully monitored.

1 International AIDS Vaccine Initiative (IAVI), New York, New York, USA 2 The Scripps Research Institute, La Jolla, California, USA OA05.01 Structural insights into HIV-1 neutralization by broadly neutralizing antibodies PG9 and PG16

Background: Human monoclonal antibodies PG9 and PG16 have been characterized recently that neutralize nearly 80% of HIV-1 isolates across all clades ¹ . These antibodies are unusually potent with preferential binding to the trimer of Env gp120.

Methods: The crystal structure of Fab PG16 has been determined at 2.5Å resolution as well as the crystal structure of the PG9 light chain at 3.1Å resolution ² . Binding and mutagenesis data have been obtained to determine the specificity of antigen binding ² .

Results: The Fab PG16 structure reveals that the unusually long, 28-residue, CDR H3 forms a hammerhead–like subdomain that protrudes well above the antibody combining site². The PG9 and PG16 CDR H3 loops both contain sulfated tyrosines and a 7-residue region of hypervariability that controls the fine specificity. Homology modeling of PG9 suggests PG16 and PG9 adopt similar three-dimensional structures. An N-linked glycan modification of the variable light chain (V_L) does not appear to affect antigen binding. The binding and mutagenesis data provide further insights into residues in the antigen binding site that are important for neutralization ² .

Conclusion: Structural characterization and biochemical analysis of PG16 and PG9 antibodies have uncovered a novel CDR H3 substructure that appears to confer most of the specificity for interaction with the gp120 trimer. Thus, these human antibodies to the HIV-1 Env appear to have evolved yet another novel mechanism for viral neutralization.

1 Unit of Virus Host Cell Interactions (UVHCI), Grenoble, France 2 Institute for Research in Biomedicine, Bellinzona, Switzerland 3 Pepscan Therapeutics B.V., Lelystad, Netherlands OA05.02 Crystal structure of the neutralizing antibody HK20 in complex with its gp41 antigen

Background: The monoclonal gp41-specific antibody HK20 was isolated from human memory B cells. The epitope was mapped to the triple stranded coiled-coil region formed by HR-1 of gp41. HK20 is broadly neutralizing, particularly as a Fab fragment but reveals strong assay-dependent selectivity in its neutralization activity.

Methods: In order to understand the structural basis of HK20 reactivity we have solved the crystal structure of a Fab fragment in complex with the 5-helix bundle construct of gp41. We used SPR measurements to determine the K_D of both antibodies and Fabs of HK20 as well as D5, an antibody with similar HR-1 specific reactivity.

Results: The structure was refined to a resolution of 2.3 Å. HK20 utilizes its CDR H1-3 and L3 loops to contact a hydrophobic pocket formed by two adjacent HR1 helices. Molecular details show that CDR H2 residues Ile53, Phe54 and Asp55 are ideally positioned to fill the hydrophobic pocket that is occupied by HR2 residues Trp626, Trp631 and Ile635 in the post fusion conformation. HK20 contacts only HR1 residues in comparison to D5, which revealed also few HR2 contacts. The HK20-gp41 interaction is such that each trimer could interact with three antibodies. Although the binding affinities of HK20 and D5 are similar their angle of interaction with gp41 is slightly different.

Conclusion: Although the HR1 epitope is only transiently exposed during the conformational changes that lead to fusion of viral and cellular membranes, the high conservation of this epitope renders it an attractive target for vaccine development. Our structural studies provide the atomic details of interaction and give insight into the requirements of an antigen that could induce HK20-like immune responses.

1 Torrey Pines Institute, San Diego, California, USA 2 Arzeda Corporation, Seattle, Washington, USA 3 Academic Medical Center of the University of Amsterdam, Amsterdam, Netherlands 4 University of Washington, Seattle, Washington, USA OA05.03 Role of Complex Carbohydrates in Human Immunodeficiency Virus Type 1 Infection and Resistance to Antibody Neutralization

Background: Complex N-glycans flank the receptor binding sites of the outer domain of HIV-1 gp120, ostensibly forming a protective 'fence' against antibodies. Here, we investigated the effects of rebuilding this fence with smaller glycoforms by expressing HIV-1 pseudovirions from a primary isolate in a human cell line lacking N-acetyl glucosamine transferase I (GnTI-), the enzyme that initiates the conversion of oligomannose N-glycans into complex N-glycans. Thus, complex glycans, including those that surround the receptor binding sites, are replaced by fully trimmed oligomannose stumps. Conversely, the untrimmed oligomannoses of the silent domain of gp120 are likely to remain unchanged.

Methods: Pseudoviruses were produced in 293T cells and GnTI- cells. These were analyzed for infectivity, neutralization sensitivity and antibody binding in native PAGE. We also determined the effect of endoglycosidase H on infectivity. This enzyme selectively digests only oligomannose glycans.

Results: Both parent and GnTI- viruses exhibited increased sensitivities to V3 loop-specific mAbs and soluble CD4. The N301Q virus was also sensitive to "non-neutralizing" mAbs targeting the primary and secondary receptor binding sites. Endoglycosidase H treatment resulted in the removal of outer domain glycans from the GnTI- but not the parent Env trimers, associated with a rapid and complete loss in infectivity. Nevertheless, the glycan-depleted trimers could still bind to soluble receptor and coreceptor analogs, suggesting a block in post-receptor binding conformational changes in the fusion process.

Conclusion: Collectively, our data show that the antennae of complex N-glycans serve to protect the V3 loop and CD4 binding site, while N-glycan stems regulate native trimer conformation, such that their removal can lead to global changes in neutralization sensitivity and, in extreme cases, an inability to complete the conformational rearrangements necessary for infection.

1 NCI Frederick NIH, Frederick, Maryland, USA 2 Human Vaccine Institute, Durham, North Carolina, USA 3 Vaccine Research Center, Bethesda, Maryland, USA OA05.04 New broadly neutralizing human monoclonal antibodies targeting the 2F5 epitope

Background: Shen et al. previously reported a patient (SC44) with broadly neutralizing serum containing antibodies that target the 2F5 epitope (J. Virology 83, 2009).

Methods: To identify monoclonal antibodies (mAbs) that functionally mimic 2F5, phage and yeast displayed antibody libraries constructed using the patient PBMC mRNA were panned and screened against peptides containing the 2F5 epitope and against gp140s.

Results: An antibody (m66) was selected that bound with high affinity to the screening antigens and was further matured by light chain shuffling. The improved antibody (m66.6) exhibited broad neutralizing activity to clade B isolates and some isolates from clades A and C in a TZM-bl cell line assay. Its potency was dramatically (more than 1,000 times) increased for target cells expressing FcγRI. IgG1 m66.6 potently inhibited infection of PBMCs and macrophages. The potency of Fab m66.6 was greatly diminished suggesting an important role of the Fc. M66, m66.6 as well as serum from SC44 bound to alanine-scanning mutants of peptides containing the sequence ELDKW with profiles similar to those of 2F5 and MPER purified immunoglobulin from SC44 sera. In multiple assays m66.6 exhibited high levels of polyreactivity to self antigens that exceeded the levels of polyreactivity of both 2F5 and 4E10 mAbs. The m66.6 putative VH gene germline predecessor (IGHV5-51) was different than that of 2F5 (IGHV2-5). The extent of somatic hypermutation was also significantly lower than that of 2F5 (8 vs 15 amino acid mutations in the VH gene product). Similarly to 2F5, m66 and m66.6 have long (23 amino acid residues) CDR3 of the heavy chain containing a number of hydrophobic residues.

Conclusion: The newly identified mAbs are the first reported mAbs that are functionally similar to 2F5. These antibodies could be used to further explore the mechanisms of elicitation of broadly neutralizing antibodies targeting the 2F5 epitope.

1 Duke University Medical Center, Durham, North Carolina, USA 2 GlaxoSmithKlein Biologicals, Research and Development, 1330 Rixensart, Belgium OA05.05 Isolation of Human Monoclonal Antibodies from HIV-Uninfected Subjects Immunized with gp120/NefTat/AS01B by Dual-Color Antigen-Specific B Cell Sorting

Background: The current goal of HIV-1 vaccine trials is to induce a rapid and robust secondary neutralizing antibody response that targets the virus envelope. In order to characterize the antibodies induced by current vaccine candidates, we used dual-color antigen-specific flow cytometric sorting of single B cells from subjects immunized with gp120/NefTat/AS01_B.

Methods: Subjects were immunized with gp120_W61D and a NefTat fusion protein formulated with AS01_B Adjuvant System and followed for 2 years. Serum and PBMC were obtained at day 0, two weeks after the second and fourth immunizations, and six months after the final immunization. Reactivity to gp120_W61D was determined by ELISA and subjects with high titers selected for sorting. PBMC were stained using a panel of B cell markers and with gp120_W61D proteins labeled with AlexaFluor 647 and PacificBlue to eliminate background reactivity. Memory B cells that stained positive for both colors were sorted as single cells; antibody genes were rescued by PCR and expressed as recombinant human monoclonal antibodies.

Results: Memory B cells that bound to labeled gp120_W61D were identified in five sorted subjects with frequencies that correlated with immunization status [mean ± SEM % labeled memory B cells by day: 0.013% ± 0.003% on day 0; 0.043% ± 0.002% on day 42; 0.11% ± 0.02% on day 182; 0.012% ± 0.004% on day 672]. Fourteen recombinant antibodies from post-immunization time points were isolated and expressed at ≥ 1μg/mL; all bound to gp120_W61D and three bound to AT-2-inactivated virions.

Conclusion: Memory B cells specific for HIV-1 immunogens can be recovered from vaccine trial volunteers using dual-color antigen-specific sorting. This technique provides a way to functionally characterize specific antibodies induced by vaccine candidates and express antibodies for SHIV passive protection trials in non-human primates.

1 Biomedical Primate Research Centre, Rijswijk, Netherlands 2 CEA, Gif sur Yvette, France 3 University Cambridge, Cambridge, United Kingdom 4 Novartis, Cambridge, Massachusetts, USA 5 Dana-Farber Cancer Institute, Boston, Massachusetts, USA 6 Duke University, Durham, North Carolina, USA OA06.01 HIV-1 envelope-CD4 receptor complexes elicit broad T- and B-cell immune responses as well as cross-reactive neutralizing antibodies in rhesus macaques

Background: A major goal for HIV-vaccine development is the identification of HIV envelope structures that generate broadly cross-reactive neutralizing antibodies (bnAbs). Exposure of hidden and/or conserved epitopes involved in co-receptor binding might be critical for eliciting such responses. Binding of Env to CD4 induces conformational changes that result in exposing the co-receptor binding site. These complexes are shown by us and others to induce bnAbs. However, in order to avoid induction of immune responses against the CD4 receptor itself, the current study has exploited the use of a 3 kD CD4 mimetic (miniCD4), derived from a scorpion toxin scaffold, scyllatoxin, which binds to Env with an affinity similar to CD4 and induces the conformational changes.

Methods: In this study we evaluated the immunogenicity of a stable clade B HIVSF162-miniCD4 complex in rhesus macaques (RM).

Results: After 4 parenteral immunizations, strong T-cell responses (IFNγ, IL4, and to a lesser extent IL2) were induced both in RM given gp140 trimer only and those given the complex immunogen. However, higher levels of antigen-specific, antibody-secreting (AC) B cells were observed in the complex group compared to the protein-only group. Functionally, both groups had similar levels of antibodies against the epitopes in the CD4bs, CD4i and gp120 V3. In addition, strong neutralizing activity was observed against various clade B SHIVs with/without soluble (s)CD4 in RM immunized with trimer alone and with the complex. Limited neutralization was found against the relatively resistant SHIV-89.6p, but only in the presence of sCD4. In addition, cross-neutralization was observed against some R5 clade C SHIVs. Furthermore, we observed neutralizing responses against HIV-2 by pseudovirus assay in the presence of sCD4 in the complex group - confirming that the CD4i epitopes are exposed and functional.

Conclusion: In conclusion, this vaccination strategy has induced strong and broad overall immune responses.

1 Emory University, Atlanta, Georgia, USA 2 Louisiana State University, New Orleans, Louisiana, USA 3 Duke University Medical Center, Durham, North Carolina, USA 4 NIAID, NIH, Bethesda, Maryland, USA OA06.02 CD40L expressed on virus-like particles adjuvants prevention of acquisition mediated by a DNA/MVA vaccine

Background: CD40L can co-stimulate B cells and dendritic cells (DC) to enhance humoral and cellular immunity. Here we test whether CD40L expressed on virus-like particles (VLPs) serves as an adjuvant for immunogenicity and protection against a repeat low dose heterologous SIV challenge.

Methods: We constructed a DNA/SIV-CD40L vaccine that expresses macaque CD40L on SIV VLPs. The expression of SIV Gag, Env and CD40L was assessed by flow cytometry and western blotting. The activity of SIV-CD40L VLPs was confirmed following in vitro stimulation of macaque DC. Groups of rhesus macaques (all negative for B08 and B17) were inoculated intramuscularly at 0 and 8 weeks with 3mg of DNA/SIV or DNA/SIV-CD40L, and boosted with MVA/SIV at 16 and 24 weeks. Both the DNA and MVA immunogens expressed Gag, PR, RT and Env. Twelve weekly low dose intrarectal challenges of SIV E660 (91% related in Gag, 83% related in Env to the SIV239 immunogens) were initiated 20 weeks following the final MVA inoculation. Protection against infection was plotted using the Kaplan-Meier method showing a nonparametric cumulative plot.

Results: CD40L-adjuvanted vaccine did not affect the magnitudes or polyfunctionality of elicited CD4 and CD8 T cells. However, it enhanced the titers of anti-Env binding antibody and the avidity against the native form of SIV239 Env. Impressively, the CD40L-adjuvanted immunizations increased protection against acquisition from 25% in the non-adjuvanted group (6/8 infected) to 60% in the adjuvanted group (5/12 infected). Protection in the adjuvanted group was significantly higher compared to protection in the unvaccinated controls (15/15 infected). Among the infected animals, the CD40L-adjuvanted animals had relatively lower peak viremia compared to unvaccinated animals.

Conclusion: SIV DNA vaccine expressing CD40L on SIV VLPs can provide strong adjuvant activity for prevention of acquisition of a heterologous E660 challenge by a SIV DNA/MVA vaccine.

1 University of Pennsylvania, Philadelphia, Pennsylvania, USA 2 Inovio, Blue Bell, Pennsylvania, USA 3 Merck, Westpoint, Pennsylvania, USA OA06.03 Enhanced magnitude of immune responses induced by SIV antigens delivered by plasmid DNA via electroporation and boosted by recombinant adenovirus 5

Background: Recent enhancements in construct design, delivery and formulation appear to be improving the immune potency of DNA vaccines in animals and humans.

We wanted to compare the immunogenicity of DNA vaccination with the most potent viral vector, recombinant adenovirus serotype 5. We then extended these studies to investigate if boosting with adenovirus vaccines or DNA could further expand the observed immune responses.

Methods: Three groups of animals (n = 5) were immunized with consensus SIVmac gag, env, pol constructs by electroporation (DNA-EP), or with recombinant Ad5SIVgag, pol, nef (Ad5) or saline alone (Naïve). The DNA group received 4 immunizations and the Ad5 group received 3 immunizations. As a follow-on study, five months following the last immunization with DNA-EP, each group was boosted twice with the Ad5 vaccine or vice versa.

Results: Following one immunization, the Ad5 group had a two-fold increase in the number of IFNγ producing cells compared to the DNA-EP group. However, following the second immunization, no significant boost in the IFNγ ELISpot responses in the Ad5 group was observed. In contrast, the DNA-EP group had an eight-fold increase in ELISpot responses and proliferative responses that were a log higher than the Ad5 group. Following boost, we observed increased responses in both the DNA-EP/Ad5 as well as the Ad5/DNA-EP group. The DNA EP prime + Ad5 boost group dramatically expanded the T cell response to over 5% following a single administration of Ad5 vaccine. Furthermore, the proliferative capacity and polyfunctionality of T cells was also substantially improved. Conversely, in the Ad5 primed + DNA EP boost group, additional boosting was observed.

Conclusion: The high levels of T cell immunity induced were encouraging as was the ability of these approaches to boost in either direction. We are now challenging these animals with SIV to observe effects in challenge outcome.

1 Rochester University Research Medical Center, Rochester, New York, USA 2 International AIDS Vaccine Initiative (IAVI), New York, New York, USA 3 EMMES Corporation, Rockville, MD, USA 4 Global BioSolutions, Craigieburn, Australia OA06.04 Preliminary results of safety and immunogenicity of Ad35-GRIN/ENV HIV Vaccine in HIV-uninfected subjects (IAVI B001)

Background: Preexisting immunity to a vector can dampen the immune response to expressed proteins. The low prevalence of adenovirus subtype 35 (Ad35) antibodies (<20%) prompted the development of Ad35-GRIN/ENV, a replication-defective Ad35 vector containing HIV-1 subtype A genes encoding gag, RT, integrase, nef (GRIN) and envelope (ENV).

Methods: A randomized, placebo-controlled, dose-escalation Phase I trial of Ad35-GRIN/ENV was conducted in healthy HIV-uninfected, participants at low-risk for HIV acquisition who were Ad35 antibody negative. The vaccine was administered intramuscularly at 0 and 6 months at either low-dose (Group A, 2x10e9 viral particles, vp), mid-dose (Group B, 2x 10e10) or high-dose (Group C, 2x 10e11). Group D received Ad35-GRIN alone at 1x10e10 vp to assess for Env immunodominance. Each group enrolled 10 vaccine and 4 placebo recipients. All volunteers were followed for local and systemic vaccine reactions for 14 days after each vaccination. IFN-γ ELISPOT responses were performed on frozen peripheral blood mononuclear cells 2 and 4 weeks post 1st vaccination and 1, 2 and 4 weeks post 2nd vaccine.

Results: Five percent of screened volunteers had detectable Ad35 neutralizing antibodies and were excluded from participation. Data presented here are blinded to vaccine or placebo within groups. No serious adverse events have been reported. Local and systemic reactions were mostly mild or. Some high-dose volunteers experienced severe reactions, both local (21.4%) and systemic (35.7%; mostly malaise, myalgia and headache). All reactions were transient and resolved spontaneously. IFN-γ ELISPOT responses 4 weeks post 1st and 2nd vaccinations were detected in 38.5% and 54.5% in Group A, and in 41.7% and 63.6% of volunteers in Group B, respectively, to all proteins. Unblinded immunogenicity results will be presented.

Conclusion: Ad35-GRIN/ENV reactogenicity appears to be dose-dependent. Ad35-GRIN/ENV is immunogenic with responses observed across all proteins.

1 California Department of Public Health, Richmond, California, USA 2 Joint Clinical Research Centre, Kampala, Uganda 3 Makerere University, Kampala, Uganda 4 Sanofi-Pasteur, Lyon, France 5 Johns Hopkins University, Baltimore, Maryland, USA 6 Family Health International, Durham, North Carolina, USA 7 Elizabeth Glaser Pediatric AIDS Foundation, Washington, DC, USA OA06.05 Immunogenicity of ALVAC-HIV vCP1521 in Infants Born to HIV-1 Infected Women in Uganda (HPTN 027)

Background: HPTN 027, the first pediatric HIV vaccine study in Africa, is a Phase I randomized, double blind, placebo-controlled trial of ALVAC-HIV vCP1521 in HIV exposed infants. ALVAC-HIV vCP1521 is a live attenuated recombinant canarypox virus expressing gene products from the HIV-1 clade E Env and clade B Gag.

Methods: 60 HIV exposed infants were enrolled between 10/06- 05/07. Infants were randomized to vaccine or saline placebo (4:1) at birth, 4, 8 and 12 weeks of age. PBMC were isolated for T cell studies at multiple timepoints over 2 years of follow-up. Vaccine-specific T cell responses were evaluated by intracellular IFN-g and/or IL-2 production, and T cell proliferation by dilution of CFSE. Samples were analyzed on a 6-color flow cytometer. Pooled peptides corresponding to the sequences of group M consensus env (gp120), and clade B gag were used at a final concentration of 1ug/ml.

Results: 47 infants completed all 4 immunizations (38 vaccine, 9 placebo). HIV-specific CD4-mediated responses were observed in 4/51 (vaccine: 2/42; placebo: 2/9) and 7/47 (vaccine: 6/39; placebo: 1/8) at the 10 week and 24 month visits, respectively compared to 1/53 (vaccine: 1/42; placebo: 0/11) at baseline as determined by cytokine-released assays. Similarly, HIV-specific CD8-mediated responses were observed in 2/51 (vaccine: 1/42; placebo: 1/9) and 5/47 (vaccine: 5/39; placebo: 0/8) at week 10 and month 24, respectively compared to 0/53 at baseline. Env-specific CD4-mediated T cell proliferation was observed in 9/42 (vaccine: 9/35; placebo: 0/7) infants at week 10 compared to 3/54 (vaccine: 3/43; placebo: 0/11) at baseline. There was no statistically significant difference in immunogenicity observed between the vaccine and placebo control arms at any of the tested timepoints.

Conclusion: Although higher immune response rates were observed in vaccine recipients, the differences were not significant in this small phase I study.

1 Tuberculosis Research Centre, Chennai, India 2 National AIDS Research Institute, Pune, Maharashtra, India 3 International AIDS Vaccine Initiative (IAVI), New Delhi, India 4 YRG Care Centre, Chennai, India 5 EMMES Corporation, Rockville, Maryland, USA 6 Global BioSolutions, Craigieburn, Australia OA06.06 Safety and Immunogenicity of DNA Prime and Modified Vaccinia Ankara Virus HIV Subtype Cboost in Indian Volunteers

Background: A modified vaccinia Ankara (MVA) vaccine (TBC-M4), expressing HIV subtype C env, gag, tat-rev and nef-RT and DNA plasmid vaccine (ADVAX) expressing env and gag and pol and nef/tat genes have previously been tested separately and found to be well tolerated and immunogenic in humans. We evaluated safety and immunogenicity of ADVAX prime and TBC-M4 boost regimen in 32 healthy, HIV-uninfected adult Indian volunteers. Results reported are blinded.

Methods: Volunteers received intra-muscular injections of either ADVAX, 4 mg at 0, 1 and TBC-M4, 5x10⁶ pfu at 3 and 6 months (Group A) or TBC-M4 at 0, 1 and 6 months (Group B) with 12 vaccine and 4 placebo recipients per group at two sites. Reactogenicity was monitored at 3, 7, and 14 days post vaccination. T-cell responses were assessed on fresh cells by IFN-gamma ELISPOT at baseline, 1 and 2 weeks post each vaccination.

Results: Vaccine or placebo was well tolerated. Mild or moderate cumulative reactions were observed in 62.6% and 75% for systemic and 50% and 62.5% of volunteers for local reactions in Groups A and B, respectively. No biological abnormality of clinical significance or serious adverse event was reported. In Group A, 2 weeks post injection, IFN-gamma ELISPOT responses were detected in 20% of post 2nd ADVAX prime and 69.2% post 2nd TBC-M4 boost, while in Group B, in 37.5%, 30.8% and 73.3% post 1st, 2nd and 3rd TBC-M4, respectively. Responses were directed to multiple HIV proteins in most volunteers. Cumulative magnitude of HIV specific responses was 4-8 times higher in Group A than in Group B.

Conclusion: Vaccination with ADVAX prime and TBC-M4 boost or placebo appears to be safe and generally well tolerated. Preliminary data suggest a DNA priming effect on immune response relative to MVA alone, modest in magnitude, and directed to all HIV proteins.

1 Federal Medical Centre, Ido-Ekiti, Nigeria, Ido-Ekiti, Ekiti, Nigeria 2 AIDS Support Organization, Masaka, Uganda 3 Care AIDS International, Ado-Ekiti, Nigeria OA07.01 Preventive HIV vaccine acceptability and behavioural risk compensation among men who have sex with men in Nigeria, sub-Saharan Africa

Background: The modest success of the Thai phase-3 HIV vaccine trial last year raises hope of a future effective vaccine. The effective roll-out and use of a future HIV vaccine depends largely on its acceptability. We assessed acceptability of preventive HIV vaccines and behavioural risk compensation among MSM in Lagos, an urban population in sub-Saharan Africa.

Methods: A structured questionnaire was designed based on formative research among some MSM, HIV physicians and HIV preventive researchers in Nigeria. It was programmed on laptop computers and administered by trained interviewers. Participants were recruited using respondent driven sampling (RDS), a novel methodology designed to access hidden populations. We assessed HIV vaccine acceptability using conjoint analysis and a factorial experimental design, and risk behaviour intentions in response to HIV vaccine use.

Results: Participants were 113 MSM; mean age = 30.4 (SD = 4.7) years. Only 9 (8.7%) engaged in paid sex. Had sex with average of 3.4 male partners in the last 6 months. More than half reported STI diagnosis in the last 1 year. HIV vaccine acceptability ranged from 87.9 (SD = 25.2) and 42.3 (SD = 36.5) on an 80-point scale. Vaccine induced seropositivity has the most impact on acceptability (27.4; p < 0.001), followed by efficacy (22.1; p < 0.001), side effects (11.5; p = 0.002), duration (8.3; p = 0.002, and out-of-pocket cost (6.2; p = 0.002). 60 (58.3%) reported intention to increase sexual risk behaviour after HIV vaccine uptake.

Conclusion: Study confirms and support need for development of interventions to reduce impact of vaccine induced seropositivity, enhance acceptability of future vaccine, and improve understanding of possible side effects which may facilitate HIV vaccine uptake among MSM and other vulnerable groups in Nigeria and West African sub-region. Finally, behavioural interventions to reduce risk compensation remains a critical aspect of a total package prevention.

1 Centers for Disease Control and Prevention (CDC), Atlanta, Georgia, USA OA07.02 Mucosal SHIV exposures and oral PrEP can induce specific T cell responses, and can have a chemo-vaccination effect

Background: PrEP (Pre-exposure prophylaxis) for HIV prevention may be efficacious due to multiple mechanisms of action. One contributing mechanism could be a “chemo-vaccination” effect, i.e. induction of adaptive immune responses to virus exposures during PrEP.

Methods: To study chemo-vaccination under realistic conditions, we used the repeat-low dose macaque model and partially effective PrEP. Nine rhesus macaques were given up to 14 once-weekly, rectal, low-dose SHIVSF162P3 exposures; six were concurrently PrEP-treated with oral, intermittent Tenofovir disoproxil fumarate and Emtricitabine. After 36 weeks of rest, four PrEP-protected macaques received 14 additional SHIV challenges with (n = 2) or without further PrEP (n = 2), followed by 5 weeks of rest. We measured antibody production, and studied SHIV-specific T cell induction, longevity, epitope specificities, and functional maturation throughout, using intracellular cytokine analysis and IFNg-ELISPOT with 14 overlapping 15-mer peptide pools representing gag, env, pol, nef, tat, vif, and vpr.

Results: Four of six PrEP-treated macaques were protected from infection; of these, none produced specific antibodies, but two mounted anti-SHIV CD4 + and CD8 + T cell responses. These were robust (up to 3940 SFU/10^6 PBMCs), predominantly central memory cells, short-lived (22 weeks or less), appeared intermittently, and had changing specificities. When one chemo-vaccinated macaque was re-exposed to virus after T cell responses had waned, it was not protected from infection. Additional chemo-vaccination boosted T cell responses and induced differentiation in the other macaque with previous SHIV-specific T cells.

Conclusion: We document and characterize PrEP-induced T cell chemo-vaccination. Although not protective after subsiding in one macaque, chemo-vaccination-induced T cells warrant more comprehensive analysis during peak responses or after boosting, to study their ability to prevent virus acquisition and role in controlling PrEP breakthrough infections. Our findings reiterate the need to closely monitor these responses in clinical trials and suggest that combination prevention strategies with vaccines and PrEP could potentially have synergistic effects.

1 Thailand MOPH-Centers for Disease Control and Prevention Collaboration, Nonthaburi, Thailand 2 Centers for Disease Control and Prevention, Atlanta, Georgia, USA 3 Ministry of Public Health, Nonthaburi, Thailand OA07.03 High HIV Incidence and HBV Vaccination Uptake in a Preparatory Cohort Study among Men who have Sex with Men in Bangkok, Thailand, 2006–2010

Background: To develop research infrastructure and assess HIV incidence, HBV vaccination uptake and follow-up rates among MSM in Bangkok in preparation for HIV prevention trials.

Methods: Longitudinal cohort study, three years follow-up at 4 monthly intervals. Study recruitment targeted community organizations, MSM bars, clubs and saunas and the Internet. Eligible men were asked to come to the clinic to provide informed consent, complete a behavioral questionnaire using audio-computer-assisted self-interviewing, and undergo medical evaluation and HIV and HBV testing with pre- and post-test counseling. Oral fluid was tested for HIV infection (OraQuick), and if reactive, confirmed with three HIV rapid tests (Determine, Double Check and HIV 1/2 Core). Baseline HBV infection status was determined by serology. Those who were susceptible to HBV infection were offered HBV vaccine, free of charge.

Results: Between April 2006 and January 2008, 1,292 men enrolled (mean age 26 years). Baseline HIV prevalence was 22.4% (n = 290), cumulative HIV incidence through April 08, 2010 was 20.5% (n = 129); 18-22 year-olds (29.9%), 23–29 year-olds (19.8%) and ≥30 years-old (12.1%). The overall HIV incidence remained relatively unchanged over the three years: 6.0%, 6.3%, and 5.7%. At baseline, 598 (46.3%) men had serologic evidence of past HBV infection and 103 (8.0%) of prior HBV vaccination. The HBV status of 19 men could not be determined. Of the 572 men who were eligible for HBV vaccination, 506 (88.5%) received the 1st dose. Of these, 96.4% and 92.8% received the 2nd and the 3rd dose respectively. Follow-up rates in the HIV-negative fraction of the cohort were 85.9% after 12 months and 75.2% after 24 months.

Conclusion: High HIV incidence, HBV vaccination uptake and follow-up rates were found in this preparatory cohort study among MSM in Bangkok. These data suggest that Bangkok MSM may be an appropriate population for the evaluation of biomedical interventions.

1 MRC/UVRI Uganda Research Unit on AIDS, Entebbe, Entebbe, Uganda 2 International AIDS Vaccine Initiative (IAVI), New York, New York, USA OA07.04 HIV incidence and risk factors for acquisition in HIV discordant couples in Masaka, Uganda: A prospective HIV vaccine trial feasibility study

Background: To determine the incidence of and risk factors for HIV acquisition in a cohort of HIV uninfected partners from HIV discordant couples in Masaka, Uganda, and to establish its suitability for HIV vaccine trials.

Methods: HIV-1 uninfected partners aged 18-60 years in HIV discordant couples were enrolled in a cohort study and followed for 2 years. Medical evaluations, HIV counseling and testing, collection of blood for HIV-1 and syphilis serology and storage were performed at quarterly visits. Data on sexual risk behaviour were collected 6-monthly.

Results: A total of 495 participants were enrolled, of whom 34 participants seroconverted over 782.4 person-years of observation (PYO). The overall HIV incidence rate [in brackets 95% confidence intervals] was 4.35 [3.10–6.08]; and 4.29 [2.85–6.46] and 4.46 [2.47–8.06] per 100 PYO in men and women respectively). Risk of seroconversion decreased by 6% for every one year increase in age (aRR = 0.94 [0.89–0.99]) in both men and women. Overall, HIV acquisition was significantly associated with alcohol use (aRR = 3.46 [1.22–9.77], non-use of condoms in the past year (aRR = 2.99 [1.14–7.91], genital discharge in the past year (aRR = 4.67 [1.79-12.17] and current syphilis infection (RR = 3.65 [1.28–10.39]. Only one seroconverter had a partner who was on anti-retroviral therapy. The 2-year retention rate was 80.2%.

Conclusion: There is a high HIV incidence in this cohort with similar rates in both women and men. The 2-year cohort retention was moderately high. Such cohorts may be suitable for future HIV-1 vaccine efficacy clinical trials.

1 Emory University Rollins School of Public Health, Atlanta, Georgia, USA 2 Zambia-Emory HIV Research Project, Lusaka, Zambia 3 Projet San Francisco, Kigali, Rwanda 4 Rwanda Zambia HIV Research Group, Atlanta, Georgia, USA OA07.05 Reduction of HIV transmission risk while prescribed antiretroviral therapy (ARVT): Misclassification of ARVT status as a methodological issue

Background: Use of ARVT may decrease risk of HIV transmission, but HIV is found in semen and cervicovaginal fluid of patients on ARVT with suppressed viral loads. Further, observational cohort studies rely on self-reported ARVT status.

Methods: HIV-discordant couples were followed in Rwanda and Zambia from 2002-2008. HIV-negative partners were tested for HIV 3-monthly. We calculated incidence density (ID) of HIV transmission by ARVT status of the HIV-infected partner (HIP) and rate ratios. We calculated odds of unprotected vaginal sex for couples where the HIP was on ARVT. Tests to identify specific ARVT agents were used in banked specimens to evaluate self-reported ARVT status for seroconverting couples.

Results: 2,993 couples were followed for 5,609 person-years [PY]. By self-reported ARVT status, 4 HIV infections occurred from HIPs on ARVT, and 171 from HIV-infected partners not on ARVT. HIV ID overall was 0.7% on ARVT, and 3.4% off ARVT (rate ratio [RR] 0.21, CI: 0.08, 0.59). The odds of any unprotected vaginal sex for couples on ARVT were 0.87 overall (CI: 0.79,0.96), and were similar in couples where the man (OR 0.92, CI: 0.79, 1.08) or the woman (OR 0.83, CI: 0.73, 0.95) was the HIV-infected partner. We found misclassification of ARVT status in both directions: preliminary results indicate 8/105 (8%) HIV transmitters who reported being off ARVT had evidence of antiretrovirals in their blood, and 3/6 (50%) HIV transmitters who reported being on ARVT had no detectable antiretroviral in their blood.

Conclusion: HIV transmission was significantly lower in couples where the HIV-infected partner was prescribed ARVT. However, our preliminary analyses indicate that misclassification of ARVT status and/or poor adherence occur, which impact the estimation of the impact of ARVT HIV transmission.

1 School of Medicine/Rwanda Zambia HIV Research Group, Atlanta, Georgia, USA 2 University of Zambia//Zambia Emory HIV Research Group, Lusaka, Zambia 3 Rwanda Zambia HIV Research Group, Kigali, Rwanda 4 UNAIDS, Geneva, Switzerland 5 Emory University/Department of Economics, Atlanta, Georgia, USA OA07.06 Cost-per-HIV infection averted in Africa by couples' HIV testing vs antiretroviral treatment: HIV prevention in a 2010 global economy

Background: Since the US President's Emergency Plan for AIDS Relief (PEPFAR) began in 2004, one in four HIV infected Africans have been identified and half of HIV+ Africans with stage III-IV disease have been reached by antiretroviral therapy (ART) programs. Nearly half (48%) of the PEPFAR budget is spent on ART while 23% is allocated to prevention. An overview of the PEPFAR budget, published data on cost-effectiveness of three prevention interventions and models calculating cost-per-HIV infection averted (HIA) for two prevention interventions are presented.

Methods: PEPFAR budgets and allocations and a literature summarizing published data on cost-effectiveness of sexually transmitted infection (STI) treatment, male circumcision, and voluntary HIV counseling and testing (VCT) were reviewed. Models calculating HIA were formulated for couples' VCT (CVCT), and antiretroviral treatment of HIV+ partners in discordant couples for prevention (ART-P). Sensitivity analyses were conducted to illustrate the impact of varying couple HIV serostatus distribution, effect size of the intervention, cost-per-client, and duration of impact.

Results: The median of published costs per HIA is $100-$300 for symptomatic STI treatment and $300–$500 for male circumcision. Reported HIA for VCT ranges from $67-$483 and for CVCT $121-$351 HIA. Cost per HIA in 5 years is 126–168 times higher for ART-P than for CVCT if both programs are applied at a national level.

Conclusion: The cost of ART for 5 million HIV patients who meet current criteria for therapy exceeds available funds. Lifelong antiretroviral treatment-as-prevention for 22 million HIV+ Africans is neither feasible nor cost-effective. Evidence based prevention interventions and vaccine development should be funding priorities. Incorporation of couple testing in existing programs for pregnant women and ART patients, along with linkage of CVCT with circumcision and family planning services would contribute to HIV and unplanned pregnancy prevention. Modification of PEPFAR indicators is needed to allow accurate measures of HIV infections.

1 Cancer Vaccine Center, Dana-Farber Cancer Institute, Boston, MA, USA 2 Tulane National Primate Research Center, Tulane University, Covington, Lousiana, USA 3 University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvanina, USA 4 California National Primate Research Center, Davis, California, USA 5 University of California, San Francisco, California, USA 6 Oregon Health and Science University, Portland, Oregon, USA 7 SAIC Frederick, Inc., National Cancer Institute at Frederick, Frederick, Maryland, USA 8 Ragon Institute of Massachusetts General Hospital, Charlestown, Maryland, USA 9 Division of Vaccine Research, Beth Israel Deaconess Medical Center, Boston, Maryland, USA 10 The Scripps Research Institute, La Jolla, California, USA 11 University of Erlangen-Nuremberg, Erlangen, Germany 12 Duke University Medical Center, Durham, North Carolina, USA 13 Dana-Farber Cancer Institute, Boston, Massachusetts, USA 14 Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA OA08.01 Effect of Fcγ-receptor polymorphisms in rhesus macaques on SIVmac251 setpoint viremia

Background: FcγR polymorphisms have been associated with progression and severity of infectious diseases. Recently in HIV1-infected individuals the relative rapidity of progression to disease was associated with an FcγRIIA polymorphism. Additionally, it was shown that the functional efficacy of monoclonal antibodies (mAbs) in vivo in humans may depend, at least in part, on Fcγ-receptor (FcγR) polymorphisms.

Methods: We investigated polymorphisms of FcγRI, FcγRIIA, FcγRIIB, and FcγRIIIA in a total of 136 Indian origin rhesus macaques (RM). Forty-eight SIVmac251-infected RM (Mamu-A*01-, -B08-, and -B17-) were examined to determine if FcγR polymorphisms were associated with the magnitude of setpoint viremia.

Results: Two polymorphisms were observed in FcγRI and FcγRIIB, and three in FcγRIIIA, resulting in two allelic forms of FcγRIIB and three allelic forms of FcγRI and FcγRIIIA. Surprisingly, extensive polymorphisms in FcγRIIA were identified; 26 allelic variants, four of which were predominant. Protein modeling indicated that several of these polymorphisms clustered in a region that can interact with an antibody.

We found that polymorphisms in FcγRIIA and FcγRIIIA independently contributed to increased setpoint viremia. However, the most dramatic difference in setpoint viremia (1.23 log; P = 0.0031) was observed when RM were stratified to include both FcγRIIA and FcγRIIIA polymorphisms. In contrast, we observed only a marginal difference at peak viremia in RM with these polymorphisms. These observations are consistent with significantly lower levels of SIV-specific antibodies at peak than at setpoint viremia.

We also observed that the same polymorphisms associated with more efficient control of setpoint viremia were associated with more efficient depletion of lymphocytes in RM treated with lymphocyte-depleting mAbs.

Conclusion: These data demonstrate that similar to humans, FcγR polymorphisms in RM play a significant role in the efficacy of mAb-mediated function. Moreover these findings highlight the complexity of humoral immune responses in AIDS virus infection.

1 Duke University Medical Center, Durham, North Carolina, USA 2 University of Pennsylvania, Philadelphia, Pennsylvania, USA 3 University of Alabama at Birmingham, Birmingham, Alabama, USA 4 National Institute for Communicable Diseases, Johannesburg, South Africa OA08.02 High titers of cross-clade ADCC-mediating antibodies are detectable in plasma samples collected from individuals with broadly neutralizing antibodies

Background: Non-neutralizing Abs including those capable of mediating ADCC are likely to be important for the development of efficacious vaccines against HIV. We sought to determine how the presence and magnitude of cross-clade ADCC-mediating Abs compared to neutralizing antibody reactivities.

Methods: Plasma samples were collected from 9 individuals enrolled in the CHAVI 001 and 008 cohorts who represented the top 2% of broadly neutralizing Ab responses. These subjects were defined as elite neutralizers among 430 individuals chronically infected with HIV subtype A, B, and C. ADCC responses were measured with a flow-based assay capable of detecting proteolytically active Granzyme B delivered into target cells by effector cells as a result of Ag-specific Ab-Fc receptor interactions. Target cells consisted of CEM.NKR_CCR5 cells that were either infected with HIV-1 or coated with recombinant gp120 representing subtypes B (WITO), C (CAP45.2), and E (CM243).

Results: Cross-clade ADCC responses were detected in all samples. The potency of the response against subtypes B, C, and E as measured by maximum Granzyme B activity were 34 ± 9%, 32 ± 9% and 25 ± 8%, respectively. The range of titers of ADCC-mediating Abs was 25,103 - 52,074, 50 - > 100,000, and 6,954 – 100,000 against subtypes B, C, and E, respectively. Overall, the titers of ADCC-mediating Ab were higher than those detected in neutralization assays. When the same subtype C targets were used for both assays, the range of ADCC-mediating Ab titers was 2 to 600 fold higher than neutralizing Ab titers. The highest ADCC activity was detected in samples with neutralizing activity that was mainly attributed to gp120-specific Ab.

Conclusion: Potent cross-clade ADCC responses were detected in infected individuals who mount potent neutralizing Ab responses against HIV-1. The magnitude of the ADCC response exceeded the neutralizing Ab response in this group of subjects. Epitopes in gp120 may play a dominant role in cross-clade ADCC reactivity.

1 Ragon Institute of MGH, MIT & Harvard, Charlestown, Massachusetts, USA 2 National Cancer Institute, Frederick, Maryland, USA OA08.03 The role of HIV-1 peptide specificity in modulating KIR+ Natural Killer (NK) cell function

Background: NK cell cytotoxic functions are under the control of a large array of inhibitory and activating receptors. One class of receptors are the MHC class I-specific Killer Immunoglobulin-like receptors (KIRs). Different KIRs recognize different groups of HLA molecules, and binding, crystallographic and functional studies have shown KIR recognition of MHC class I is dependent and influenced by the bound peptide, with sensitivity at position 7 and 8 of the peptide. The inhibitory receptor KIR3DL1 recognizes HLA-B allotypes with the serologically-defined Bw4 motif (HLA-Bw4). Recent genetic and functional studies have implicated that particular KIR3DL1 haplotypes in combination with HLA-Bw4 are associated with slower HIV disease progression. We hypothesized that HIV peptides may contribute to this observed protection through modulating KIR-HLA interactions.

Methods: Fluorescently-labeled HLA-B tetramers refolded with HIV peptides were used to determine binding to Jurkat cells expressing KIR3DL1*001.

Results: In accordance with the HLA-Bw4 allelic specificity of KIR3DL1, HLA-Bw4 (HLA-B*2705, HLA-B*5701 and HLA-A*2404) tetramers, but not HLA-Bw6 tetramers bound to KIR3DL1*001. Additionally, differential binding of KIR3DL1*001 to HLA-Bw4 alleles was observed; HLA-B*2705 tetramers presenting Gag263-272 KRWIILGLNK bound weakly to KIR3DL1*001, whilst strong binding of HLA-A*2404 presenting Nef134-141 RYPLTFGW to KIR3DL1*001 was detected. Furthermore, some HLA-B*5701 tetramers bound to KIR3DL1*001, but this interaction was dramatically affected by the presented peptide; strong binding of HLA-B*5701 presenting Gag240-249 TSTLQEQIAW (TW10) or Gag162-172 KAFSPEVIPMF to KIR3DL1*001 was observed, with no binding detected with HLA-B*5701 presenting RT244-252 IVLPEKDSW. Finally, comparison of mutations in the TW10 epitope at natural positions of variation showed that the interaction between HLA-B*5701: TW10 and KIR3DL1*001 could be abrogated. HLA-B*5701 no longer bound to KIR3DL1*001 when presenting either the TW10 T242N mutant or TW10 A248D mutant.

Conclusion: This data suggests that sequence variations within HIV peptides can affect the KIR-HLA interaction, a mechanism that could potentially place selective pressure on the virus.

1 HIV Vaccine Trials Network, Seattle, Washington, USA 2 Vaccine and Gene Therapy Institute/OHSU, Beaverton, Oregon, USA 3 Louisiana State University, New Orleans, Louisiana, USA 4 Oregon National Primate Research Center, Beaverton, Oregon, USA OA08.04 A DNA plasmid prime and rAd5 boost vaccine regimen applied to the tonsils elicits strong cellular and humoral immune responses in rhesus macaques

Background: Induction of potent mucosal immunity would be desirable for HIV vaccines. However, the effect of mucosal versus parenteral vaccination routes on magnitude, quality, and anatomic distribution of immune responses to vaccine remains unclear. Tonsil immunization, which may elicit genital and GI tract immunity, is particularly interesting for HIV vaccines.

Methods: We assessed innate and adaptive immune responses in Mamu-A*01(+)/B*17(−) rhesus macaques following immunization (priming plus boosting) intramuscularly (IM) or submucosally over the tonsils (IT). Animals (8/group) were immunized with plasmid DNA encoding SIVmac239 Gag, Pol, Env, Vif (1 mg/construct) at months 0, 1, 2 and boosted with rAd5 vectors expressing Gag, Pol, Env, Vif-Vpr-Vpx at 6 months (3x10¹⁰ viral particles/construct). At various time-points following immunization, antigen-specific cellular immune responses in blood, lymph nodes, and multiple mucosal tissues were measured by immunophenotyping, intracellular cytokine staining, and tetramer binding assays. SIV-specific antibody response was measured by ELISA in plasma and secretions.

Results: DNA priming elicited low to undetectable vaccine-specific cellular responses in blood. However, in broncho-alveolar lavage (BAL) high frequency polyfunctional SIV-specific CD4 + and CD8 + T cells were detected by ICS in DNA-primed IT animals, although DNA-primed IM animals showed very low frequency responses. CM9 gag tetramer responses were significantly higher in the IT group (P < 0.005). SIV-specific antibody responses in plasma were similar for IT and IM groups, but showed distinct kinetic profiles. Responses were significantly increased after boosting with rAd5; the IT group generated relatively stronger cellular responses in blood, but the difference in magnitude of responses by group was not statistically significant (%CM9 + CD8 + cells; IT = 3.2%, IM = 2.3%). High-level CD8 + responses were seen at mucosal sites in the IT group (%CM9 + CD8 + cells: BAL = 11.8%, colon = 1.5%).

Conclusion: Tonsillar immunization can elicit systemic responses and mucosal responses at distant sites. Further characterization of mucosal responses, including protection from low-dose rectal challenge with SIVmac239, is in progress.

1 University of Minnesota, Shoreview, MinnesotaUSA 2 Department of Microbiology, Medical School, University of Minnesota, Minneapolis, Minnesota, USA 3 AIDS and Cancer Virus Program, SAIC-Frederick, Inc., NCI, Frederick, Maryland, USA 4 New England Primate Research Center, Harvard Medical School, Southborough, Massachusetts, USA OA08.05 Dissecting the Protective Mechanisms in Cervicovaginal Tissues of SIVmac239Γnef-Vaccinated Macaques against Pathogenic SIV Challenge

Background: The live attenuated vaccine SIVmac239Δnef(SIVΔnef)is one of the best protection modalities against pathogenic SIV challenge, although the mechanisms responsible for this protection remain poorly understood.

Methods: To identify potential protective mechanisms against vaginal transmission, we analyzed virus replication and host responses in cervical-vaginal and lymphatic tissues (LTs) of Indian rhesus macaques following intravenous SIVΔnef infection (vaccine phase), and following vaginal challenge with SIVmac251.

Results: In the first two weeks of the vaccine phase, we found that SIVΔnef replication was largely restricted to a pre-activated follicular B helper CD4+ T cell subset, and that the limited replication did not lead to the massive depletion of CD4+ T cells in the gut and enteropathy seen with higher levels of replication of SIV mac251. Thereafter, plasma and LT levels of SIVΔnef declined to low to undetectable levels at 140 days post-infection (dpi). At this time, we detected antibodies to SIV Env and Gag—mainly IgG, but also IgA—in cervical vaginal tissues that had not been detectable at 35 dpi.

Conclusion: In light of evidence that protection against vaginal challenge induced by SIVΔnef increases from 35 dpi to 140 dpi (Paul Johnson, personal communication), the maturation of the antibody response in cervical tissues raises the possibility that antibodies may contribute to protection against vaginal challenge. We also compared the global gene expression profiles in cervicovaginal tissues of rhesus macaques at 35 dpi versus 140 dpi. We will present the findings and discuss the implications.

1 NIH/NIAID, Bethesda, Maryland, USA 2 SAIC Inc., NCI, Frederick, Maryland, USA 3 Case Western Reserve University, Cleveland, Ohio, USA 4 Harvard University, Ragon Institute, Charlestown, Massachusetts, USA OA08.06 SIV infection results in loss of IL-17-producing NK cells in mucosal tissues, which contributes to damage to the mucosal barrier

Background: Dysfunction of the mucosal immune system and immune activation are hallmarks of HIV pathogenesis. Damage to the structural barrier of the GI tract leads to microbial translocation and immune activation, however, the mechanisms underlying this damage remain unclear. Here we used SIV-infected Asian macaques to study the role of NK cells in mucosal homeostasis and AIDS pathogenesis.

Methods: We used flow cytometry to measure NK cell function and phenotype in blood, lymph nodes and mucosal tissues of macaques prior to and after SIV infection. We measured damage to the GI tract, microbial translocation, and immune activation in tissues by immunohistochemistry.

Results: We found a loss of NKs in blood after SIV infection, which was associated with increased proliferation. In the gut, we found a negative correlation between the frequency of colon NK cells and the amount of damage to the colon epithelium (P = 0.0311, r = −0.7280). Intriguingly, though there was not a general loss of NKs in the mucosa after SIV infection, we observed a selective loss of IL-17-producing NKs in both the colon (P = 0.0127) and mesenteric LN (P = 0.0177). Furthermore, the frequency of IL-17-producing NKs negatively correlated with damage to the colon epithelium (P = 0.0341, r=−0.8289). Moreover, there was a significant positive correlation between the frequency of IL-17-producing NK cells and CD4 + T cells in the colon (P = 0.0081, r = 0.6747).

Conclusion: These data suggest that mucosal NK cells have a specialized role in homeostasis, and loss of these cells after SIV infection may contribute to damage to the GI tract and ensuing microbial translocation and immune activation. Furthermore, the correlation between IL-17-producing NK cells and CD4 + T cells suggests that the mechanism underlying loss of Th17 cells after SIV/HIV infection is not preferential Th17 infection. Better understanding the mechanisms that underlie loss of these homeostatic NK cells may lead to novel therapeutic strategies to treat HIV.

1 Ragon Institute of MGH, MIT, and Harvard, Charlestown, Massachusetts, USA 2 University of California at Davis, Davis, California, USA 3 Massachusetts General Hospital, Boston, Massachusetts, USA OA08.07 HIV-1 elite controllers resist HIV-1 infection via p21 (cip-1/waf-1)

Background: Elite controllers are a small population of HIV-1 infected persons with undetectable viremia in the absence of antiretroviral therapy. These individuals with a “functional cure” of HIV-1 infection may serve as a model for successful host immune activity against HIV-1, but correlates of immune protection in this patient population are ill defined.

Methods: HIV-1 replication steps in CD4 T cell from elite controllers, HIV-1 negative persons, HIV-1 progressors and HAART-treated patients were analyzed after ex vivo infection with a VSV-G pseudotyped HIV-1 vector or primary HIV-1 isolates, using established PCR assays to quantify extra- and intrachromosomal HIV-1 DNA and HIV-1 mRNA. p21 gene expression was analyzed by RT-PCR and western blots in quiescent (HLA-DR-) and activated (HLA-DR+) CD4 T cells from elite controllers, and reference populations. Functional HIV-1 infection assays were performed in the presence or absence of a small molecule inhibitor of p21.

Results: CD4 T cells from elite controllers were significantly less susceptible to HIV-1 infection in comparison to reference patient cohorts (p < 0.01). This resistance was due to less effective reverse transcription and HIV-1 mRNA transcription from proviral DNA. Defective viral replication in elite controllers corresponded to significant (p < 0.001), 10-20 fold higher expression levels of p21 in cells from elite controllers compared to the reference populations. In ex vivo infection assay, blockade of p21 significantly (p = 0.01) enhanced reverse transcription and HIV-1 mRNA transcription in CD4 T cells.

Conclusion: These data indicate that the selective upregulation of p21 represents a natural barrier against HIV-1 reverse transcription and mRNA transcription in CD4 T cells from elite controllers, and significantly contributes to the ability of elite controllers to maintain undetectable viral replication. Manipulation of p21 may provide novel opportunities to increase host resistance to HIV-1 infection.

1 Duke University, Durham, North Carolina, USA 2 The Aaron Diamond AIDS Research Center, New York, New York, USA 3 University of North Carolina at Chapel Hill, Chapel HILL, North Carolina, USA 4 University of Alabama-Birmingham, Birmingham, Alabama, USA OA09.01 Early plasma B cell responses to transmitted HIV-1 are directed to envelope gp41 and originate by the activation of mutated B cell clones

Background: The initial antibody response to HIV-1 is non-neutralizing, targeted to gp41 envelope, and is ineffective in controlling viremia. To understand the origins of the initial HIV-1 antibody response, we studied clonal lineages of gp41 antibodies isolated from sorted plasma cells (PCs) of acute HIV-1 infection (AHI) patients and reverted unmutated antibody ancestors.

Methods: Blood PCs from five AHI subjects obtained approximately 17, 20 and 30 days after HIV-1 transmission and 4 subjects 7 days after influenza vaccination and 6 subjects after influenza infection were sorted into 96-well plates for amplification of VH and VL genes by RT/PCR. Antibody clonal lineage was identified by Ig gene analysis. The isolated VH and VL genes and the reverted unmutated antibody ancestors from the antigen reactive lineages were expressed as recombinant mAbs and specificity was determined by immunoassays.

Results: A total of 36 clonal lineages of antibodies were identified among 1,109 antibodies isolated from AHI, of which 10 clonal lineages were reactive to HIV-1 gp41. Of the 10 HIV-1 gp41 clonal lineages, comprised of 78 antibody members, only 27 antibodies (34.6%) reacted with gp41. In contrast, of 398 antibodies isolated from 4 influenza-vaccinated subjects, 199 mAbs were in 47 influenza-antigen reactive lineages. Of the 199 antibodies in influenza cell clonal lineages, 182 antibodies (91.5%) were reactive with influenza antigens. We found the reverted unmutated gp41 antibody ancestors did not react with HIV-1 envelope gp41 in 4/6 gp41 clones, but instead were polyreactive with host and bacterial antigens. In contrast, reverted unmutated antibody ancestors derived from influenza HA-reactive antibodies in 4/4 HA B cell clones reacted with nM affinities with influenza hemagglutinin.

Conclusion: These data demonstrate that HIV-1 may divert the initial envelope antibody by stimulating mutated B cell clones originally activated by antigens other than HIV-1 envelope, thereby limiting the initial antibody response to non-protective antibodies.

1 University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, USA 2 University of Pennsylvania, Philadelphia, Pennsylvania, USA 3 Zambia Blood Transfusion Service, Lusaka, Zambia 4 Emory University, Atlanta, Georgia, USA 5 Seattle Biomedical Research Institute, Seattle, Washington, USA 6 University of Pennsylvania and Yerkes National Primate Research Center, Philadelphia, Pennsylvania, USA OA09.02 Nonpathogenic SIV infection of Sooty Mangabeys is not associated with high levels of autologous neutralizing antibodies

Background: SIV infection of natural host species, such as sooty mangabeys (SM), is characterized by high viral replication, and low generalized immune activation despite evidence of an adaptive immune response. However, most studies have focused on T cells and innate immune cells, with a significant gap existing in our understanding of whether humoral immunity might also differ between pathogenic and nonpathogenic infections. Here the ability of SIV-infected SM to mount neutralizing antibodies (Nab) against autologous virus was compared to HIV-1 subtype C infected subjects during acute/early and established infection.

Methods: Multiple, biologically functional envelope (Env) glycoproteins were cloned from human subjects and SM during both acute and established infection. Nab activity was evaluated against the autologous Envs using plasma collected from later time points using the Tzm-bl pseudovirus assay. Flow cytometry was performed on peripheral lymphocytes.

Results: While high levels of Nab were observed in early and established HIV-1 infection, samples obtained from comparable time points in SM exhibited significantly lower titers of autologous Nab. During acute/early infection, changes in the B cell compartment were consistent with a transient immune activation that was rapidly resolved. In established infection, SM with higher Nab titers also contained elevated peripheral CD4+ T cell levels, suggesting that synergy is maintained between B cells and CD4 T cells.

Conclusion: Taken together, these results indicate that low autologous Nab activity is a novel and previously unappreciated feature of non-pathogenic SIV infection of SM. The fact that high titer Nab are not necessary to avoid disease progression during SIV-infection of SM is consistent with the notion that the apathogenicity of natural SIV infections is not the result of particularly effective adaptive immune responses against the virus. Understanding the qualititative and quantitative differences in the Nab response during pathogenic vs. nonpathogenic infection could provide critical information regarding protection from AIDS.

1 NIH/NIAID, Bethesda, Maryland, USA 2 The Scripps Research Institute, and International AIDS Vaccine Initiative (IAVI) Neutralizing Antibody Center, La Jolla, California, USA OA09.03 Mechanism of neutralization by monoclonal antibody VRC01, a CD4 binding site directed neutralizing antibody with extensive neutralization breadth

Background: We recently isolated three CD4 binding directed monoclonal antibodies, VRC01, 02, and 03, from the memory B-cells of a donor whose sera displayed potent and broad neutralization activity. VRC01 and VRC02 are a closely related pair of somatic variants that neutralize over 90% of circulating HIV-1 isolates. In this study we sought to further characterize their neutralization epitopes on monomeric gp120, their impact on the architecture of the viral Env functional spikes upon binding, and the elements of Env associated with the unusual cases of neutralization resistance.

Methods: Alanine scanning mutagenesis of the primary isolate JRCSF Env were performed on the native functional virus and on JRCSF gp120.

Results: These data confirmed that key contact residues overlapped with those for CD4 binding. However, VRC01 required contacts in loop D and in V5 on gp120 that are distinct from CD4 binding contact residues. Compared to mAb b12, very few mutations resulted in enhanced binding or neutralization by VRC01, suggesting that steric constraint is not a major factor limiting the neutralization activity of VRC01. Similar to CD4, VRC01 interaction induced a large conformational change in monomeric gp120, leading to enhanced binding of the co-receptor binding site mAb 17b, and enhanced binding of gp120 to CCR5. Unlike CD4, however, VRC01 did not alter the conformation of the native viral spike and it did not trigger gp120 shedding from the Env spike upon binding. Ongoing studies are examining that similarities and differences in interaction of VRC01 and CD4 with the viral Env.

Conclusion: VRC01 achieves extensive neutralization breadth by targeting the CD4bs in a manner that partially mimics that of CD4, but diverges in particular interactions with the functional Env spike.

1 Duke University Medical Center, Durham, North Carolina, USA 2 Torrey Pines Institute, San Diego, California, USA 3 NICD, Johannesburg, South Africa 4 University of Alabama, Birmingham, Alabama, USA 5 University of North Carolina, Chapel Hill, North Carolina, USA 6 Vaccine Research Center, Bethesda, Maryland, USA OA09.04 Epitope Specificities of Elite Neutralizing Sera from HIV-1-Infected Individuals

Background: A small proportion of HIV infected individuals generate a neutralizing antibody (NAb) response of exceptional magnitude and breadth. A detailed analysis of the critical epitopes targeted by broadly NAbs in these elite cases, the top 2%, should help to define optimal targets for vaccine design.

Methods: Elite neutralizers from the chronic infection CHAVI 001 and 008 cohorts were identified by assaying sera from 430 subjects against a multi-subtype panel of 12 ‘tier 2’ viruses (4 each of subtypes A, B and C). Sera and purified IgG were obtained from the top 9 elite neutralizers, comprising clade A, clade B, clade C, and clade A/C infections. Epitope specificities and neutralization targets were characterized by using chimeric and site-mutated viruses, protein and peptide absorptions and peptide microarrays.

Results: MPER-directed neutralizing activity was present in 44-66% of these patients, depending on the assay. One plasma exhibited CD4bs neutralizing specificity and another exhibited N332A glycan-dependent neutralization. Several sera contained broadly NAb activity sensitive to V1/V2 mutations. All subjects exhibited a predominant anti-gp140 IgG1 antibody subclass response and 3 subjects also exhibited an anti-gp140 IgG2 response. One subject with MPER-directed neutralizing activity in serum also exhibited low level neutralizing activity in purified IgG from mucosal samples. Purified mucosal IgG from 3 subjects was further examined for antibody specificities using peptide microarrays covering the entire gp160 of clades A, B, C, and D, group M, CRF 1 and CRF 2.

Conclusion: In subjects with either clade A, C or A/C recombinant infections, MPER NAbs were the most prevalent specificity. However, three subjects with the greatest neutralization breadth had antibodies that targeted gp120 (CD4bs, N332A glycan, V2). These data define optimal epitopes for HIV-1 vaccine design efforts.

1 IIDMM, University of Cape Town, Cape Town, Western Cape, South Africa 2 National Institute of Communicable Diseases, Johannesburg, Gauteng, South Africa 3 Centre for the AIDS Programme of Research in South Africa (CAPRISA), Durban, KwaZulu Natal, South Africa OA09.05 Mapping of residues critical for neutralisation by a broadly cross-neutralising antibody targeting a quaternary epitope on the HIV-1 Envelope

Background: Defining the epitopes targeted by broadly cross-neutralising antibodies is one of the major challenges facing HIV vaccine design. One participant from the CAPRISA 002 cohort (CAP256) was previously identified as an individual who developed potent broadly cross-neutralising activity. It was subsequently shown that one of the antibodies mediating breadth targeted a quaternary epitope on the envelope trimer, and further that V1V2 was a critical determinant. However, conventional techniques used to map quaternary epitopes are labour intensive, requiring the generation of numerous point mutants or chimeras. Here we describe an alternative approach, using available autologous and heterologous envelope sequences to inform the generation of mutants.

Methods: In order to map the fine specificity of this broadly cross-neutralising antibody, positively selected mutations in the autologous virus that correlated temporally with neutralisation escape were identified in longitudinal envelope sequences (n = 66). Positive selection was identified using codon-based maximum-likelihood methods (SLAC, FEL, and REL implemented on www.datamonkey.org) after correcting for recombination. Positively selected sites that were conserved in sensitive but not in resistant envelopes from a heterologous virus panel were considered as potentially important for the formation of the targeted epitope. As breadth was mediated by at least one other antibody that targeted outside of V1V2, only heterologous viruses whose neutralisation by CAP256 serum was determined to be V1V2-dependent were considered “sensitive”.

Results: Using this methodology, 2 residues in V2 (R166 and K169) were identified as potentially forming part of the targeted epitope. Site-directed mutagenesis of either of these residues in a sensitive heterologous virus (CAP45) conferred resistance to neutralisation, reducing titres from an ID₅₀ of > 1:7,000 to approximately 1:160, confirming that both residues are critical determinants.

Conclusion: These results demonstrate the utility of this approach as a simple and more targeted alternative to conventional techniques for mapping the specificity of broadly neutralising antibodies.

1 New York University School of Medicine, New York, New York, USA 2 Molsoft LLC, La Jolla, California, USA 3 Veterans Affairs Medical Center, New York, New York, USA 4 All India Institute of Medical Sciences, New Delhi, India 5 New York University School of Medicine, Veterans Affairs Medical Center, New York, New York, USA 6 New York University School of Medicine, Veterans Affais Medical Center, New York, New York, USA OA09.06 Germline variable genes code for contact residues maintained during affinity maturation of human anti-V3 monoclonal antibodies encoded by VH5-51

Background: Immunogenetic studies have revealed the importance of certain immunoglobulin (Ig) genes encoding monoclonal antibodies (mAbs) against various epitopes in the envelope of HIV-1. We have demonstrated recently that VH5-51 gene segment is preferentially utilized by 18 (35%) of 51 human anti-V3 mAbs.

Methods: In this study, a panel of 18 anti-V3 mAbs were examined which were derived from individuals infected with clade B and non-clade B HIV-1; all were encoded by the VH5-51 gene segment and neutralized Tier 1 and Tier 2 viruses. Immunochemical and crystallographic methods were used to study their function and the structure of the antigen-combining site.

Results: The VH5-51 gene used by all 18 mAbs paired only with lambda light chain genes, mainly with VL1-47 and VL3-1 genes. This restricted pairing of Ig genes resulted in the formation of a conserved antigen combining site, as documented by crystallographic studies of five of these anti-V3 Fabs in complex with V3 peptides. The VH5-51-encoded V3 mAbs recognize slight variations of an epitope which contains conserved residues in the N- and C-terminal β-strands of the V3 crown. This finding was confirmed by the binding of the mAbs to a peptide which mimics this region of V3 but lacks the tip of the V3 loop. Crystallographic analysis of the Fab/peptide complex showed that all contact residues in the CDR domains, except CDR H3 and L3, are germline-encoded and thus determine the conserved character of the binding site. The data indicate that affinity maturation of these mAbs has preserved the germline-encoded interaction with the antigen.

Conclusion: The immunogenetic studies of anti-V3 neutralizing mAbs revealed that certain germline variable genes encode the contact residues which are maintained during antibody evolution; targeting epitopes preferentially recognized by such Ig genes with vaccines should induce broadly reactive neutralizing Ab responses.

1 Division of Virology-University of Cape Town, Cape Town, Western Cape, South Africa 2 NICD, Johannesburg, Gauteng, South Africa 3 University of Kwazulu-Natal, Durban, South Africa 4 University of KwaZulu-Natal, Durban, South Africa 5 University of Kwazulu-natal, Durban, South Africa OA09.07 Genetic characteristics of HIV-1 envelope associated with broadly cross-reactive neutralizing antibody responses

Background: Factors associated with development of broadly cross-reactive neutralizing antibody (BCN) responses in HIV-1 infection are not fully understood. We investigated whether subjects who developed such responses had unusual genetic properties of the envelope.

Methods: Nine anti-retroviral naïve HIV-1 subtype C infected individuals from the CAPRISA 002 cohort in KwaZulu-Natal, South Africa were included in this study. Four individuals had BCN antibodies (BCN group) and five had type-specific neutralizing antibodies (non-BCN group). We generated 194 env sequences by single genome amplification, with an average of 10 per participant from early infection (<2 months post infection) and 3 years post infection i.e after breadth has developed.

Results: Viruses from the BCN group (n = 4) had no significant difference in divergence from founder sequence to 3 years compared to the non-BCN group (n = 5). However, analysis of the V1V5 region showed that the two groups differed significantly in the change in loop lengths. In the BCN group, there was an average increase of 9.25 amino acids in V1V5 over 3 years, compared to an average decrease of 4.1 amino acids in the non-BCN group (p = 0.0317). Changes were largely driven by changes in V1V2 with an average increase of 12 amino acids in the V1V2 loops in the BCN group compared to a shortening of 2.5 amino acids in the same region in non-BCN group (p = 0.0159). Viruses from the BCN group showed a greater increase in glycosylation sites over 3 years compared to non-BCN, although this was not significant.

Conclusion: Our data demonstrated that viruses associated with BCN antibody responses show elongation of V1V2 after 3 years infection compared to viruses from individuals with type-specific responses, whose viruses generally undergo shortening of V1V2 loops. These results implicate V1V2 as playing a major role in neutralization antibody responses, either as a direct target or through shielding of epitopes.

1 Duke Human Vaccine Institute, Durham, North Carolina, USA 2 Division of Hematology/Oncology, University of Alabama at Birmingham, Birmingham, Alabama, USA 3 Center for AIDS Research, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA 4 Center for Infectious Diseases, University of North Carolina, Chapel Hill, North Carolina, USA 5 University of North Carolina Project, Lilongwe, Malawi 6 National Institute of Health, Bethesda, Maryland, USA 7 Laboratory for AIDS Vaccine Research & Developmen, Duke Medical Center, Durham, North Carolina, USA OA10.01 Antigenicity and immunogenicity of transmitted/founder HIV envelope oligomers compared to chronic HIV envelopes

Background: Neutralizing antibodies (NAbs) have been shown to play an important role in blocking HIV-1 infection. One current challenge is to develop criteria for the selection of Env immunogens for human clinical trials. Transmitted/founder (T/F) viruses provide a source of recombinant Env oligomers for antigenicity and immunogenicity studies.

Methods: We compared antigenicity of HIV-1 Envs, including 10 from T/F, 15 from chronic infection viruses, and 7 consensus Envs (24 gp140 oligomers, 8 gp120), by ELISA and SPR binding assays with 10 broadly neutralizing monoclonal antibodies (bNAbs) (1b12, 2F5, 2G12, 4E10, 17b, PG9, PG16, VRC01, VRC02, and VRC03). Immunogenicity of 7 T/F, 10 chronic and 7 consensus Envs was assessed in guinea pigs.

Results: Many Envs were poorly antigenic, reacting with very few bNAbs; T/F Envs were the most antigenic by their reactivity with a greater number of bNAbs. On average, T/F Envs reacted with 12% more bNAbs than consensus Envs, and with 28% (P = 0.012) more bNAbs than chronic Envs (with better reactivity (P = 0.004)). T/F Envs 63521.B and 6240.B reacted with all 10 bNAbs; however, the antigenicity of these latter two Envs did not translate into superior immunogenicity. Thus, the two most antigenic Envs induced primarily NAbs against Tier 1 strains of clade B HIV-1. The most immunogenic was T/F Env 1086.C from Malawi that formed spontaneous recombinant trimers and reacted with 6 bNAbs. 1086.C Sera neutralized 12/12 of Tier 1 A, B and C strains, including two SHIV viruses (Bal, SF162P3), but neutralized only 6/42 Tier 2 HIV-1 isolates.

Conclusion: Antigenicity was poorly predictive of immunogenicity. T/F Env 63521.B expressed epitopes for PG9, PG16 and VRC01 very well and may be an oligomer to screen for these types of antibodies. Env 1086.C is a candidate immunogen for testing with novel adjuvants in non-human primates to improve immunogenicity.

1 New York University School of Medicine, New York, New York, USA 2 Molsoft LLC, La Jolla, California, USA 3 University of Massachusetts Medical School, Worcester, Massachusetts, USA 4 Public Health Research Institute/UMDNJ, Newark, New Jersey, USA 5 Beth Isreal Deaconess Medical Center, Boston, Massachusetts, USA OA10.02 Engineered Immunogen Presenting an Epitope Recognized by a Neutralizing mAb Elicits Mammalian Serum that Recapitulate the mAb's Specificity

Background: To exploit the promising properties of cross-strain neutralizing monoclonal antibodies (mAbs), immunogens should elicit polyclonal Ab responses in mammals that mimic the reactivity of these mAbs. To demonstrate the feasibility of this approach, the epitope recognized by the anti-V3 loop mAb 3074—which is present in approximately 90% of circulating HIV-1 viruses—was use as a template for immunogen design.

Methods: A V3 loop-cholera toxin B fusion protein (V3₃₀₇₄-CTB) was designed to eliminate the epitopes targeted by several anti-V3 loop mAbs while preserving the epitope targeted by mAb 3074. New Zealand White rabbits were primed with codon-optimized clade C gp120 DNA and then boosted with V3₃₀₇₄-CTB. Anti-V3 mAbs and rabbit immune sera were assessed for neutralizing activity against (a) V3 chimeric psVs infecting U87 CD4 + CCR5 + cells, (b) primary isolates infecting TZM-bl cells, (c) Tier 1 and (d) Tier 2 psVs infecting TZM.bl cells.

Results: V33074-CTB bound specifically to mAb 3074 but not to several other anti-V3 mAbs. A psV bearing the V3₃₀₇₄-designed sequence was neutralized only by mAb 3074. Immune sera elicited in rabbits using V3₃₀₇₄-CTB demonstrated 50% neutralization of (a) 7/7 V3 chimeric psVs carrying the V3 loops of clades A1, AG, AE, B, C, F, and H, (b) 3/10 primary isolates from clades A, AG and B, (c) 4/4 Tier 1 viruses, and (d) 4/14 Tier 2 viruses from clades B and C. Little neutralizing activity was seen in the immune sera against a V3 chimeric psV lacking the 3074 epitope, and the only three Tier 2 clade C viruses neutralized by mAb 3074 were also the only three Tier 2 Clade C viruses neutralized by the serum of the best responder.

Conclusion: Our results demonstrate that V3₃₀₇₄-CTB elicited a polyclonal, cross-strain neutralizing Ab response in mammalian serum mirroring the specificity of 3074—the mAb used as a template for immunogen design.

1 International AIDS Vaccine Initiative (IAVI), San Diego, California, USA 2 Vaccine Research Center, NIAID, NIH, Bethesda, Maryland, USA OA10.03 Biophysical, antigenic and immunological analysis of HIV-1 Env constructs designed to focus the B cell response on conserved elements

Background: A major challenge for HIV-1 vaccine design is the extensive diversity of the cell-surface envelope glycoproteins (Env). Targeting conserved sites on Env, such as the functionally conserved CD4 binding site (CD4bs) on the exterior Env, gp120, is one potential means to elicit more broadly neutralizing responses. The CD4bs antibody, b12, neutralizes diverse isolates and there is a growing list of more potent and broad CD4bs antibodies recently reported. Mutations in the gp120 Phe 43 cavity can partially stabilize gp120 or core proteins in the CD4-bound state and increase CD4 affinity. Here, we assess if prime: boosting with stabilized cores and trimers in series would better focus the B cells responses to the conserved CD4bs.

Methods: Stabilized HIV-1 Env gp120 core, gp120 trimers and gp140 trimers were transiently expressed from 293F cells and affinity purified. For biophysical and antigenic analysis, the proteins were characterized by selected methods. The FPLC profile revealed that the variant Envs were purified to homogeneity. The impact of the Phe43 pocket filling mutations on CD4 recognition of the cores and trimers was assessed both surface plasmon resonance and microcalorimetry.

Results: For immunogenicity analysis, rabbits were sequentially immunized with HXBc2gp120 core, YU2gp120-GCN4 trimers, wild-type and stabilized gp140-foldons trimers derived from clades A, B or C. Neutralization activity of the rabbit sera was assessed against a panel of 14 pseudoviruses from clades B and clade C. Substantial neutralization of Tier 1 isolates was observed but only sporadic neutralization of Tier 2 isolates was observed. Homologous inoculation of selected trimers elicited the broadest responses compared to core boosting or priming. We performed competition ELISAs using the neutralizing CDbs mAb, VRC01, and mapped antibody responses to the CD4bs in most sera.

Conclusion: We conclude that the biophysical, antigenic and immunogenicity study of these Env-based immunogens provide relevant and critical information to forward HIV-1 vaccine development.

1 Duke University, Durham, North Carolina, USA 2 Los Alamos National Laboratory, Los Alamos, New Mexico, USA OA10.04 Computationally identified mutations suspected to cause gp120 charge inversion and gp41-mediated allosteric effects impact IgG1b12 susceptibility

Background: A phylogenetically-corrected statistical analysis of IgG1b12 neutralization sensitivies and Env sequences for 251 clonal Env-pseudotyped viruses identified Env genetic signatures associated with b12 susceptibility. Several signatures corresponded to previously identified residues in or near the b12 contact surface of gp120, or at locations where mutations have been shown to affect b12 binding. Interestingly, the analysis identified two signature sites outside the b12 binding domain that have not been previously investigated for their effect on b12 binding or neutralizing activity. One signature, an E to K mutation at HXB2 position 268, causes a rare, substantial positive-to-negative charge reversal that could affect the binding of the highly positively-charged b12 MAb. Another signature, N651I, is located within the C-heptad repeat region of gp41 and may affect exposure of the b12 epitope via an allosteric influence on Env.

Methods: To experimentally validate these signatures, we used site-directed mutagenesis of 10 Env-pseudotyped viruses to generate 16 mutant Envs. The mutated pseudoviruses were tested in a TZM-bl neutralization assay against a panel of monoclonal antibodies, including IgG1b12. Neutralization sensitivities were compared to those of parental Env.

Results: Mutations according to the signature at position 268 resulted in significant changes in b12 sensitivity for 5 of 7 viruses. The signature at position 651 also caused significant changes in b12 sensitivity for 5 of 9 viruses tested. All significant changes in b12 sensitivity matched the computational prediction.

Conclusion: These results strengthen the validation of phylogenetically-corrected computational methods to identify genetic signatures of antibody-mediated HIV-1 neutralization. Moreover, these methods appear capable of providing detailed information about neutralization epitopes that is not easily acquired by other methods. Importantly, these results support the use of similar computational methods for the discovery of improved vaccine immunogens that induce broadly neutralizing antibodies.

1 New York University School of Medicine, New York, New York, USA 2 Veterans Affairs Medical Center, New York, New York, USA OA10.05 Map of broad and narrow neutralization in the V3 loop crown

Background: Sequence variability of the V3 loop crown has often been considered an impediment to eliciting broadly neutralization antibodies targeted to this region. Our work maps contacts between human anti-V3 monoclonal antibodies (mAbs) and the V3 crown in an attempt to make a structural distinction between the sites targeted by broadly as opposed to narrowly neutralizing mAbs.

Methods: The 3D structure of the 16-residue V3 crown bound to anti-V3 mAb 2219 was divided into 4 regions (stem, hydrophilic patch, hydrophobic patch and turn) based on previously published analyses (Almond et. al., 2010). Known mAb/V3 contacts from Jiang et al., 2010 and others were then mapped onto this structure and compared to measurements of neutralization breadth by the respective mAbs in infectivity assays.

Results: mAbs that entirely engage the less sequence variable zones of the V3 loop (the stem, turn and hydrophobic zones) tend to be more broadly neutralizing than mAbs that contact at least one side-chain in the most variable region of V3 (the hydrophilic zone). The contact in the highly variable region appears to narrow the Ab reactivity regardless of how many other contacts are formed with V3 backbone atoms or with side-chains in the conserved regions.

Conclusion: Broadly neutralizing mAbs are targeted to the structurally conserved region, and they avoid contact with the highly sequence-variable zone. More narrowly neutralizing Abs make at least one side-chain contact within the highly-sequence variable zone: these latter mAbs are vulnerable to a single escape mutation at that side-chain/amino acid location, and indeed these side-chains are frequently mutated in circulating viruses. This knowledge could be exploited for HIV vaccine design by designing immunogens that immunologically silence the variable zone in the V3 loop.

1 Northwestern University, Chicago, Illinois, USA 2 Tulane University, Veterinary Pathology, Covington, Louisiana, USA OA11.01 Identification of the first cells infected during vaginal transmission in the Depo-Provera rhesus macaque model

Background: There is currently a significant effort to develop interventions to prevent the transmission of HIV. To facilitate these efforts, a better understanding of how the virus can breach mucosal epithelial barriers is required. We have previously explored the interaction of HIV with the female genital tract using human explants. This analysis revealed that HIV could penetrate both the columnar epithelium of the endocervix and the squamous epithelium of the vagina and ectocervix. Establishing whether the penetrating virus could lead to productive infection however remained to be elucidated.

Methods: To identify the first infected cells, we utilized a SIV based vector system that would express the mCherry fluorescent protein. The vector was pseudotyped with the R5 tropic HIV envelope JRFL and concentrated by ultracentrifugation. Depo-Provera treated female macaques were vaginally exposed to the inoculum. Four days later the animals were sacrificed, their genital tracts excised, and cryosections of the tissue were analyzed by fluorescent microscopy.

Results: Analysis revealed large but rare foci of red fluorescent cells. The expression of mCherry within these cells was confirmed by emission spectrum analysis. These rare foci of transduced cells were found in areas protected by both the columnar and squamous epithelium. The target cells are primarily CD4 + T cells that have infiltrated into the squamous epithelium or reside just below the columnar epithelium.

Conclusion: The clustering of transduced cells suggests that there are small regions within the female genital tract that are susceptible to infection. Understanding the mechanism that leads to this increased local susceptibility could lead to novel treatments that can reduce the male-to-female acquisition of HIV.

1 University of Alabama at Birmingham, Birmingham, Alabama, USA 2 Dartmouth Medical School, Lebanon, New Hampshire, USA OA11.02 Susceptibility of Epithelial Cells from the Human Female Upper Reproductive Tract to Infection by Transmitted/Founder HIV-1

Background: The mechanisms of HIV transmission in the female reproductive tract (FRT) remain poorly understood. In particular, upper FRT sites, especially uterus and fallopian tubes, are only recently being recognized as potentially vulnerable to HIV-1 infection. The endocrine controlled immune functions of the FRT change profoundly throughout the menstrual cycle. This results in a “window of vulnerability” to HIV infection that has not yet been systematically explored.

Methods: We employed infectious molecular clones (IMC) of transmitted HIV-1, which we recently generated, to investigate the susceptibility of uterine epithelial cells (UEC) to infection. Since UECs can express HIV-1 coreceptors, they may represent first target cells for infection. Primary UEC were isolated from hysterectomy tissues and cultured as polarized monolayers. Cells were exposed to Renilla luciferase reporter viruses expressing transmitted or reference strain envelope (Env) genes (Env-IMC-LucR viruses) allowing sensitive detection of viral gene expression.

Results: Using a panel of 15 transmitted clade B Env from female and male subjects, and 4 reference Env, our findings indicate that UEC are susceptible to HIV-1 infection. Remarkably, in primary uterine cell cultures derived from multiple patients, we did not see evidence for infection/reporter gene expression after exposure to Env-IMCs of commonly used strains, BaL and SF162, while cultures challenged with transmitted Env-IMCs repeatedly exhibited low but detectable levels of luciferase expression. Moreover, when polarized cells were tested for route of infection, apical but not basolateral cell surface exposure led to UEC HIV infection.

Conclusion: To our knowledge, this is the first time HIV-1 susceptibility of uterine cells has been addressed with mucosally transmitted HIV-1. Our data suggest that uterine epithelial cells are possible targets of HIV-1 infection and transmission. Since the uterus is readily accessible to semen, the likely exposure of these tissues to HIV-1 is relevant to development of intervention strategies.

1 Northwestern University, Chicago, Illinois, USA 2 Northwestern Memorial Hospital, Department of Obstetrics and Gynecology, Chicago, Illinois, USA OA11.03 HIV ENV antibodies differentially modulate viral transport in cervical mucus

Background: A better understanding of sexual transmission of HIV is essential for the development of viable vaccines and microbicides. Cervical mucus (CM) is one of the first obstacles HIV encounters during vaginal transmission. CM could potentially impede HIV transport by physically sieving the virions or through specific interactions of HIV virions with mucosal antibodies. Furthermore, the existence of repeatedly exposed but uninfected women – as in the case of discordant couples – could, in part, be explained by the protective function of CM and the antibodies it contains.

Methods: We use multi-color fluorescence video microscopy and computer aided particle tracking to follow the movement of individual HIV virions in CM. We add virions with wild-type GP120 envelope (mCherry-tagged) and control virions without GP120 (GFP-tagged) to each mucus sample. We image them simultaneously. We also image the two HIV phenotypes in the presence of various antiGP120, antiGP41, and antiHLA1 antibody isotypes. We characterize and quantify the motion of each virion by their mean squared displacement, and other standard measures of microscopic transport.

Results: We examine the effect of exogenous monoclonal antibodies, and of endogenous antibodies in HIV infected women on viral motion in CM. We find that the binding of antibodies modulates HIV transport in CM and that the strength of the modulation depends on the isotype used.

Conclusion: These studies lay a foundation for quantitatively studying the interaction of HIV with cervical mucus and the antibody response to HIV in infected women.

1 Harvard Medical School, Boston, Massachusetts, USA 2 SAIC-Frederick, National Cancer Institute, Frederick, Maryland, USA 3 University of Alabama at Birmingham, Birmingham, Alabama, USA 4 Blood Systems Research Institute and University of California, San Francisco, California, USA 5 New England Primate Research Center, Harvard Medical School, Southborough, Massachusetts, USA 6 Duke Human Vaccine Institute, Duke University Medical Center, Durham, North Carolina, USA 7 Divisions of Viral Pathogenesis, Beth Israel Deaconess Medical Center, Boston, Massachusetts, USA 8 Division of Vaccine Research, BIDMC, Ragon Institute, MGH MIT &Harvard, Boston, Massachusetts, USA OA11.04 Low Dose Mucosal SIV Infection Restricts Early Replication Kinetics and Transmitted Virus Variants in Rhesus Monkeys

Background: Defining the earliest virologic events following HIV transmission may be critical for the design of vaccine strategies aimed at blocking acquisition of HIV infection. Here we show that the dose of the virus inoculum impacts the length of the eclipse phase and the number of transmitted virus variants following intrarectal SIV infection of rhesus monkeys.

Methods: 24 adult rhesus monkeys (N = 6/group) received a single intrarectal inoculation with virus doses of 109, 108, 107, or 106 SIV RNA copies of SIVmac251. Plasma SIV RNA levels were determined following infection. The number of transmitted/founder virus variants were determined by assessing the phylogenetic patterns of env diversity utilizing SGA and direct amplicon sequencing. Innate immune profiles were evaluated by quantifying serum cytokine levels, and PBMC phenotyping was determined by flow cytometry.

Results: For monkeys infected by 109, 108, 107, or 106 SIV RNA copies of virus, the eclipse phase was a median of 4, 4, 7, and 8.5 days, and the transmitted/founder virus variants were a median of > 10, > 10, 2 and 1, respectively. Compared with the higher dose groups, the lower dose groups also showed trends toward lower and delayed median peak cytokine levels, and less extensive depletion of central memory CD4 T cells.

Conclusion: Low dose SIV infection resulted in a lengthened eclipse phase, fewer transmitted virus variants, and decreased innate immune activation as compared with high dose SIV infection. These data suggest a mechanism by which it may be easier for a vaccine to protect against low risk HIV transmission as compared with high risk HIV transmission. These findings have major implications in the design and interpretation of HIV vaccine efficacy studies in humans.

1 Northwestern University, Chicago, Illinois, USA 2 Tulane National Primate Research Center, Covington, Louisiana, USA OA11.05 Exploring HIV Transmission in the Male Genital Tract Using the Circumcised and Uncircumcised Macaque Model

Background: The model of HIV sexual transmission in the male genital tract has not been well developed and this has hindered understanding of how male circumcision prevents against HIV infection. We aimed to create a more comprehensive model of HIV transmission in men.

Methods: Eight cadaveric penile explants (3 uncircumcised) were inoculated with photoactivatable (PA) GFP-Vpr HIVBal in culture. The genital tracts of seven uncircumcised male rhesus macaques (Macaca mulatta) from Yerkes and Tulane National Primate Research Centers were also inoculated with PA HIVBal in vivo, and penile tissue was removed during necropsy. Tissue cryosections were stained for Langerhans cells, CD4+ T-cells, keratin, and intercellular junctions. Images were obtained with DeltaVision RT systems and analyzed with SoftWorx software. Four uncircumcised male macaques were separately circumcised at Tulane. Trans-epithelial water loss and skin capacitance of the macaque penis before and after circumcision were measured using the Vapometer and MoistureMeter.

Results: HIV particles were predominantly visualized in the keratin of cadaveric penile epithelia, which tends to be thinner in the circumcised penis. 3-8% of virions were seen in deeper epithelial strata, up to 81 microns from the surface. In vivo inoculations of the macaque penis showed similar findings, with penetrating virions seen at depths of up to 60 microns from the surface and occasionally near target cells, thus validating the use of the human explant model.

The stratified squamous epithelium of the macaque penis was histologically very similar to that of the human penis. Non-invasive measurements of penile skin barrier function in the circumcised macaques demonstrated changes over time (correlating to a loss of barrier function) that may contribute to how circumcision affects HIV transmission.

Conclusion: These findings represent potential routes of HIV transmission in the male genital tract. Building a clearer model of transmission will aid in the development of future HIV prevention methods.

1 University of Cape Town, Cape Town, South Africa 2 Kamuzu Central Hospital, Lilongwe, Malawi 3 Duke University, Durham, North Carolina, USA 4 University of North Carolina Chapel Hill, Chapel Hill, North Carolina, USA OA11.06 Differences in Glycosylation Patterns But Not Neutralization Sensitivity of HIV-1 Env in Acute and Chronic Infection From Heterosexual Transmission

Background: Events surrounding the heterosexual transmission of HIV-1 remain an important area of research. Viral genetic variation offers a convenient source of information to explore questions surrounding the transmission event. We have examined patterns of sequence variation in the viral env gene and expressed these sequences to test their biological properties.

Methods: We used env gene sequences derived from subtype C HIV-1 in acutely infected subjects (n = 69) and in contemporaneous chronically infected subjects (n = 62) from Malawi and South Africa. In addition, HIV-1 env clones from a subset of subjects with acute and chronic infections were examined for neutralization sensitivity by generating pseudotyped viruses and testing them with panels of monoclonal antibodies and broadly neutralizing polyclonal sera.

Results: Env proteins from acute infection have fewer glycosylation sites than Env proteins from chronic infection. This difference is largely driven by the virus in acutely infected males (N = 30). Two-thirds of the glycosylation count difference is located in the variable regions. In addition, five potential sites in the conserved regions are under-glycosylated in Env proteins from acute infection. There were no significant differences in neutralization patterns in viruses pseudotyped with Env proteins derived from viruses representing acute or chronic infections.

Conclusion: Our results indicate that heterosexual transmission specifically from females to males on average results in transmission of viruses with Env proteins that have 10% fewer glycosylation sites. We found no difference in the antibody sensitivity of viruses pseudotyped with Env proteins derived from viruses from acute infection compared to viruses derived from chronic infection. These data suggest that in women there is a restriction of virus that reaches or is able to leave the mucosal surface and the restriction is dependent on the extent of glycosylation on the virus. These results are currently being confirmed with an independent subtype C acute infection cohort.

1 USAMC-Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand 2 Department of Medicine, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand 3 Thai Red Cross AIDS Research Centre, Bangkok, Thailand 4 Chulalongkorn University and Thai Red Cross AIDS Research Centre, Bangkok, Thailand 5 National Blood Centre, Thai Red Cross Society, Bangkok, Thailand 6 U.S. HIV Military Research Program, Division of Retrovirology, WRAIR, Rockville, Maryland, USA P01.01 Genetic Diversity of Incident HIV-1 Infection in Thai Blood Donors

Background: HIV-1 circulating recombinant form (CRF) 01_AE is the predominant strain in Thailand. Subtype B (Thai and US) co-circulates across risk populations. Within the past decade, novel CRFs, including CRF15_01B and CRF34_01B have been identified, mostly in drug users. Recombinant strains have also been reported in antenatal and community populations. Blood donors at the National Blood Centre, Thai Red Cross Society (NBC, TRCS), Bangkok are considered a low-risk population based on a self-administered questionnaire prior to donation. Little has been reported about HIV-1 recombinant strains in this group. Here, we describe the molecular characterization of HIV-1 strains and recombinant forms identified among them.

Methods: Plasma samples from 10 incident HIV-1 infected blood donors using the HIV diagnostic algorithm of the NBC, TRCS in Bangkok were identified from September 2007 to March 2008. Samples were collected within 102 days from the previous HIV seronegative donation. Extracted HIV RNA was genotyped by a multi-region hybridization assay specifically for subtypes B, C and CRF01_AE (MHAbce) and further characterized by full genome sequencing.

Results: Subjects included 7 males and 3 females and the estimated median time to seroconversion was 64.5 days (range: 45.5-102). Six CRF01_AE, 2 CRF01_AE/B recombinants and 1 subtype B were identified by MHAbce. One nontypeable sample (viral load 307 copies/ml) contained subtype B in gp120. Full genome sequences and MHAbce results were concordant. The two CRF01_AE/B recombinants had different genome structures. One had subtype B portions in gag, pol, vif, vpr, gp41 and nef regions. The other contained subtype B genetic materials in the pol, gp41 and nef regions only.

Conclusion: CRF01_AE is also predominant in this group as previously reported in other Thai cohorts. The findings that subtype B and a high proportion of CRF01_AE/B recombinants were identified suggest an increasing complexity of HIV epidemic across risk populations in Thailand despite the limited sample size.

1 Royal Thai Army-Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand 2 USA Medical Component-AFRIMS, Bangkok, Thailand 3 Military HIV Research Program (MHRP), Rockville, Maryland, USA P01.02 ECHO study (Early Capture HIV Cohort): Efficient detection of acute HIV-1 infections in Pattaya, Thailand (RV 217)

Background: The acute stages of HIV-1 infection influence disease progression. Understanding the earliest host-virus interactions will be critical to vaccine development. ECHO began in July 2009 to identify volunteers with very early Fiebig stage HIV infection through intensive, bi-weekly molecular-based diagnosis. The feasibility of volunteers to adhere to this intensive visit schedule was assessed.

Methods: Volunteers considered at high risk for HIV infection (FSWs, MSMs, and transgenders) and determined to be at high risk by ACASI screening are enrolling and will be followed for two years. Small blood volumes (SBV) are collected twice a week with testing for HIV-1 RNA by the Aptima Qualitative Assay to detect acute infection. Volunteers are seen every 6 months for larger blood draws to include HIV diagnostics, immunoassays, and host/viral genetics. Volunteers with reactive Aptima results are entered into an intensive one-month diagnostic verification phase to determine if the Aptima result represents true infection. Volunteers determined to have acute HIV infection are followed for an additional five years.

Results: To date, 277 volunteers completed the screening visit. Overall baseline HIV prevalence was 11.8% (32 of 277) stratified by MSMs 17.8% (18 of 101), TGs 14.0% (12 of 86) and, FSWs 2.3% (2 of 90). 91.5 % began SBV collections. Over 6,000 specimens have been tested for HIV RNA. SBV collection adherence has been 92%. Retention is approximately 90%. Seven volunteers have been identified with acute HIV infection, (3 MSMs, 2 FSWs, and 2 TGs) giving a crude incidence of approximately 8.0/100 PYs. Of these, 6 were identified and samples collected in Fiebig stage 1 or 2.

Conclusion: ECHO has demonstrated that efficient detection of individuals with very early acute HIV infection is feasible in Thailand. This will provide an extremely powerful tool to study host-HIV interactions with direct relevance to HIV vaccine development.

1 University of Cape Town, Cape Town, Western Cape, South Africa 2 Mbeya Medical Research Programme, Mbeya, Tanzania, United Republic of 3 Department of Infectious Diseases and Tropical Medicine, Ludwig-Max, Munich, Germany 4 Department of Molecular and Cell Biology, University of Cape Town, Cape Town, South Africa P01.03 The entry efficiency of transmitted HIV-1 variants from Mbeya, Tanzania plays a role in disease progression

Background: Previous studies have shown that Envelope function is an important determinant of HIV replicative fitness. This study aimed to determine whether the entry efficiency of env clones generated from transmitted viruses was associated with higher viral loads at setpoint.

Methods: Plasma samples were obtained from high risk individuals from Mbeya, Tanzania. Functional envelope clones were constructed from single genome amplification (SGA) derived env amplicons from 12 individuals shown to be infected with a single founder virus. The entry efficiency of the functional Env clones was compared using a pseudovirion entry assay with infection of TZM–bl cells using both the pSG3∆env (subtype B) and pBR264F-Mlu∆env (subtype C) backbones. The tropism of each envelope clone was tested using the U87 cell lines expressing CD4 receptor with either CCR5 or CXCR4.

Results: Of the 12 env clones, one was subtype A, 7 subtype C, one subtype D, as well as one AC and two CD recombinants. The use of different backbones (subtype C and B) did not impact Env function as the relative entry efficiency remained the same for all Env clones. There was no relationship between Env entry efficiency and subtype. However there was a significant correlation between viral loads and entry efficiency using either subtype B or C backbones at 3 months (p = 0.0120, r = 0.4838; and p = 0.0118, r = 0.4852 respectively), as well as 12 months (p = 0.005, r = 0.7194; p = 0.0403, r = 0.3566 respectively) post-infection (Spearman correlation). This relationship was not due to tropism as all the env clones were R5 tropic.

Conclusion: High entry efficiency of Env of transmitted variants was associated with high viral loads at set point suggesting that Env function at transmission could affect disease outcome.

1 Harvard School of Public Health, Boston, Massachusetts, USA 2 Botswana–Harvard AIDS Institute, Gaborone, Botswana P01.04 The in Vivo Evolution of Gag Mutations during the Early Phase of HIV-1 Subtype C Infection

Background: A better understanding of viral evolution is critical for the design of preventive and therapeutic interventions. HIV-mediated T cell response can rapidly select escape mutations that can revert upon transmission to a new host.

Methods: The in vivo dynamics of viral mutations was assessed in a cohort of 42 individuals (8 acute and 34 recent cases) infected with HIV-1 subtype C (HIV-1C). Gag quasispecies were analyzed by single-genome amplification from the time of seroconversion to up to 500 days post-seroconversion. The in vivo evolution of viral mutations was studied by their relation to the wild type virus represented by the contemporaneous HIV-1C Gag consensus sequence, and by their fraction in the pool of viral quasispecies over time.

Results: About a third of Gag mutations detected in primary HIV-1C infection, 29.3%, were reverse mutations, suggesting a restoration of viral fitness after viral transmission. The majority of reverse mutations, 72.8%, had reached dominance or completeness in the pool of viral quasispecies. Potential escape mutations comprised 61.3% of Gag mutations detected. In contrast to reverse mutations, the vast majority of escape mutations, 90.9%, were minor or transient. Sequential series of reverse and escape mutations were also evident at the same amino acid residues, highlighting the complexity of viral mutational pathways in Gag. The number but not the type of observed mutations in HIV-1C Gag was directly associated with enrichment for mutations in the APOBEC context. While escape mutations in Gag were not associated with viral set point, a temporal correlation between the number of reverse mutations and viral set point was evident in subjects with increased viremia during the early stage of HIV infection.

Conclusion: The results of the study highlight quantitative and qualitative differences between reverse and escape mutations in Gag during primary HIV-1C infection, and underline the dynamic nature of viral mutations.

1 University of Cape Town, Cape Town, Western Cape, South Africa 2 Weatherhall Institute of Molecular Medicine, Oxford University, United Kingdom 3 AIDS Research Unit, National Institute for Communicable Diseases, Johannesburg, Gauteng, South Africa 4 CAPRISA, Doris Duke Medical Research Institute, UKZN, Durban, Kwazulu Natal, South Africa P01.05 Early selective pressures acting on the HIV-1 subtype C transmitted whole genome in five individuals with differing disease progression profiles

Background: Immune selective pressures acting on transmitted HIV-1 affect diversification and potentially the course of disease. We investigated changes across the virus genome in the first 6 months of infection in two rapid, two intermediate and one slow disease progressor.

Methods: Samples were collected from five participants from the CAPRISA Acute Infection 002 cohort, Durban, South Africa at screening or enrolment, 3 and 6 months post infection. Whole (n = 121) and half genome (n = 9) sequences were generated using single genome amplification and direct sequencing. Putative immune pressure was identified as clustered amino acid changes in 9-mer peptides, known HLA-associated CTL epitopes or N-linked glycosylation sites. Autologous peptides were tested in IFN-g ELISPOT assays. PARRIS was used to identify sites under positive selection.

Results: In the five participants, a total of 63 epitope regions were identified as under putative immune pressure with 51 positively selected amino acid sites. Fifteen epitopes (24%) were recognized by T cells of which 3 have not been previously described; 7 epitopes spanned known class I HLA-associated CTL epitopes but with no matching ELISPOT responses; 18 were identified as under putative antibody pressure; 9 as under putative CTL reversion; and 14 were unclassified. Proven or putative CTL pressure was most frequent in Nef (n = 6) followed by Pol (n = 5), Env (n = 4), Gag and Vif (n = 2), Vpr, Rev and Tat (n = 1). Shuffling or toggling of mutations was observed in 53% of identified CTL epitopes. The earliest CTL escape mutation was observed at two weeks post-infection in Gag. The highest number of positive selection sites (n = 31), putative CTL (n = 7) and antibody (n = 7) escapes were observed in a rapid progressor.

Conclusion: We identified early immune selection in all five participants, with evidence of T cell pressure as early as two weeks post-infection and shuffling of mutations in more than half of identified epitopes.

1 Walter Reed Project-Kenya, Kericho, Kenya 2 Mbeya Medical Research Programme, Mbeya, Tanzania, United Republic of 3 Makerere University-Walter Reed Project, Kampala, Uganda 4 Walter Reed Army Institute of Research, Silver Spring, Maryland, USA 5 Armed Forces Research Institute for Medical Sciences, Bangkok, Thailand 6 U.S. Military HIV Research Program, Rockville, Maryland, USA P01.06 RV 217: The Early Capture HIV Cohort Study (ECHO): A prospective study of acute HIV infection among high risk populations

Background: Events occurring in acute HIV-1 infection critically influence prognosis. Understanding the nature of host control of viral replication will provide crucial insights to guide vaccine development. ECHO is a prospective cohort study among high risk adults designed to define HIV incidence, retention in these populations and acquire samples from acutely infected persons prior to the advent of detectable HIV antibody.

Methods: HIV uninfected, female sex workers and females considered at high risk for HIV infection based upon a screening questionnaire are enrolling and will be followed for two years. Small blood volumes (SBV) are collected twice a week by finger stick with testing for HIV-1 RNA by the APTIMA Qualitative Assay to detect acute infection. Volunteers are seen every 6 months for larger blood draws to accommodate HIV diagnostics, immunoassays, and host/viral genetic studies. Volunteers with reactive APTIMA results are entered into an intensive one-month diagnostic verification phase. Volunteers with acute HIV infection are followed for an additional five years to correlate clinical outcome with acute events.

Results: To date, 817 female volunteers screened and 87% are at high risk by our pre-defined criteria. Baseline HIV prevalence was 33% in Kericho, Kenya (82 of 251), 52% in Mbeya, Tanzania (107 of 207), and 36% in Kampala, Uganda (69 of 193). SBV collection adherence has been 94% through 3–7 months of monitoring. Six volunteers have been identified with acute HIV infection, giving a crude incidence of approximately 7.4/100 PYs (95% CI: 1.5–13.3%). Of these 6, four were identified and samples collected prior to advent of detectable HIV antibody.

Conclusion: ECHO has demonstrated that efficient detection of individuals with very early acute HIV infection is feasible in East Africa. This will provide an extremely powerful tool to study host-HIV interactions with direct relevance to HIV vaccine development.

1 Rwanda Zambia HIV Research Group, Atlanta, Georgia, USA 2 Emory University Department of Pathology and Laboratory Medicine, Atlanta, Georgia, USA P01.07 Seroincidence over time in M to F and F to M transmission in discordant couples referred from government services in Kigali, Rwanda

Background: More than 90% of new infections in Rwanda are acquired from spouses and more than two-thirds of these can be prevented through couples' counseling and testing (CVCT). CVCT is now routine in antenatal clinics in Kigali, the capital city of Rwanda. Cohabiting discordant couples are referred for participation in research studies. The objective of this study was to estimate the incidence of HIV-1 infection among discordant couples in Kigali.

Methods: Discordant couples (M + F- and M-F+) were screened for eligibility and excluded based on age (men <18 or > 65, women <18 or > 45), duration of union (<3 months), and use of ART by the HIV + partner. Eligible discordant couples were enrolled and followed at monthly (HIV-) and quarterly (HIV+) intervals. HIV-uninfected partners were screened with rapid HIV antibody tests and p24 ELISA. When available, the last sample negative for both antibody and p24 in subsequent seroconvertors was tested with PCR. Incidence rates of HIV seroconversion were calculated for HIV-negative partners who had completed at least one follow-up visit.

Results: Between August 2002 and Jun 2009, 1651 discordant couples (773 M-F + and 878 M + F-) were enrolled and had follow-up. There were 113 cases of seroconversions (68 in men and 45 in women). The M to F incidence rate was 3.2/100 Couple-years in the first 24 months of followup and 3.4/100 CY thereafter with no cohort effect; corresponding rates for F to M were 4.9/100CY for the first 24 months and 2.3/100 CY thereafter showing decrease in seroconversion rates over time.

Conclusion: Despite an ongoing counseling and provision of free condoms, the HIV sero-incidence among discordant couples is still high, though lower than that reported in other studies of discordant couples. HIV vaccine trials among discordant couples may yield results that will be generalizable to most at risk populations in Africa.

1 Emory University, Atlanta, Georgia, USA 2 Zambia Blood Transfusion Service, Lusaka, Zambia 3 Tulane University, New Orleans, Louisiana, USA P01.08 Detailed epitope mapping for two V1V2-directed autologous monoclonal antibodies from early HIV-1 subtype C infection

Background: Recently we have shown that potent autologous neutralizing antibodies (Nabs) develop during early subtype C infection, but the virus rapidly escapes. The manner of escape is important to study because vaccine-induced Nabs should ideally prevent this from occurring. We previously recovered two distinct autologous monoclonal antibodies (Mabs) from a newly subtype C infected individual. These antibodies (13.6A & 6.4C) are directed against the envelope (Env) V1V2 region, which is the dominant target of plasma Nabs in this infection. Here we demonstrated the virus used two escape pathways involving residue changes that altered potential glycan motifs.

Methods: Resistance against these two Mabs was mapped using a newly transmitted Env that was mutated to encode two residue changes found in the 8-month Env escape variant. The role of glycosylation was specifically investigated by creating mutations that either did or did not confer putative glycans at these sites. This panel of mutated Envs was assessed for neutralization sensitivity in the Tzm-bl assay.

Results: Mab 13.6A neutralization was principally dependent upon a glycan in V1; however, this activity was modulated by whether an S or T formed the glycan motif at position135 and was abolished by a non-glycan dependent change at R190 in V2. Mab 6.4C principally targeted R190, but this sensitivity was modulated as well by changes at S135 in V1.

Conclusion: These data demonstrate that the antibodies target multiple epitopes in V1V2 containing cooperative sequence and glycan determinants at positions S135 and R190. The results suggest that even though V1V2 was targeted by multiple Nabs, the virus had numerous options for escape. Nevertheless, adding a glycosylation site at R190 was the pathway ultimately chosen in vivo, and therefore, it might confer escape from other Nabs present in the plasma.

1 Tianjin University, China 2 Fudan University, Shanghai, China P02.01 A DNA Delivery Vector Activates Innate Immunity and Improves the Immunogenicity of HIV DNA Vaccine Through Intranasal Administration

Background: PDMAEMA is a synthetic polycation used as DNA transfection reagent by condensing negatively charged DNA at physiological conditions. To minimize the cytotoxicity of PDMAEMA as a DNA delivery vector in vivo, we made PEGylated PDMAEMA and tested it as carrier for DNA vaccine.

Methods: In vitro transfection efficiency was evaluated by delivering with a GFP expression plasmid into 293T cells. The plasmid DNA pSV-CN54gag and recombinant TianTan vaccinia rTTV-CN54gag were used in vaccination. Mice were immunized intranasally with naked DNA, mPEG113-b-PDMAEMA94/DNA polyplexes or PEI/DNA polyplexes at 2-week intervals for three times, and boosted intramuscularly with rTTV-CN54gag. T cell response was evaluated by IFN-gamma based ELISPOT assay and antigen specific antibody was quantified by ELISA. Production of cytokines by RAW264.7 was measured by qPCR and ELISA.

Results: PEGylated PDMAEMA retained the ability to deliver DNA into cells in vitro, althrough PEGylation compromised the transfection efficiency. Complexing with PEGylated PDMAEMA significantly improved the prime effect of HIV DNA vaccine through intranasal administration. PEGylated PDMAEMA induced cytokines production by murine macrophages.

Conclusion: PEG-PDMAEMA is an efficient vector for delivering DNA vaccine through mucosal route which enhances adaptive immune responses with activating innate immunity.

1 The University of Melbourne, Melbourne, Victoria, Australia P02.02 Role of Adjuvants in Enhancing the Immunogenicity of HIV-1 Envelope Glycoprotein based Vaccine Immunogens

Background: The ability of HIV-1 envelope glycoprotein (Env) based antigens to elicit protective immune responses is complicated by factors influencing translation of antigenicity into immunogenicity like structural constraints, choice of adjuvants and vaccination regime. Synthetic adjuvants present as potent and safer choice to enhance the immunogenicity of recombinant Env antigens. Novel synthetic Pam2Cys based lipopeptides - cationic R4Pam2Cys (R4) and anionic E8Pam2Cys (E8) form electrostatically bound complexes with charged surfaces of Env and activate dendritic cells by engagement of Toll-like receptor 2. This leads to activation of antigen specific naïve T cells to induce stronger immune responses.

Methods: AD8gp140 DNA was engineered into stable cell lines to secrete soluble gp140 which were purified by lentil-lectin and size-exclusion chromatography to yield trimers. Pam2Cys adjuvants were synthesized by Fmoc chemistry whereas Freund's and Montanide are commercial preparations. AD8gp140 with different adjuvant combinations - R4, E8, Freund's and Montanide-ISA51 were co-administered in mice using a double protein regime. Antibody titres were measured by ELISA and neutralizing ability of sera was measured using a pseudovirus based CF2³-GFP neutralization assay.

Results: Sequence analysis predicted the overall charge of AD8gp140 to be + 9.32, hence electrostatic bond formation with E8. Post vaccination antibody titres peaked log10^4.5 except for E8 group. The neutralizing ability against MN, SF162 and homologous AD8 pseudoviruses were higher yet moderate in R4 group followed by Montanide group. Though antibody titres reached log10^4.5 for Freund's and log^103.5 for E8 group, both their sera showed poor neutralization. Complex contour and glycosylation of Env might have favored better binding to R4 than the predicted E8 and hence generated stronger quantitative and qualitative immune responses.

Conclusion: Choosing potent adjuvants can enhance the quality of humoral immune response especially promoting broad neutralizing antibody response. Currently novel Env based DNA and proteins immunogens are being assessed combinatorially using Pam2Cys mucosally and Montanide systemically.

1 University of Washington, Seattle, Washington, USA 2 Fred Hutchinson Cancer Research Center, Seattle, Washington, USA 3 Institute for Systems Biology, Seattle, Washington, USA P02.03 Retinoic acid as an adjuvant during systemic vaccination enhances protection from mucosal viral challenge

Background: Retinoic acid can induce the expression of gut-homing receptors on activated T cells. Our previously presented data showed that all-trans retinoic acid administered in vivo during vaccination increased memory T cells at mucosal sites without compromising the functions of memory cells in secondary lymphoid organs. Here we investigated if the retinoic acid enhanced mucosal responses provide greater protection from mucosal viral challenge.

Methods: Naïve C57B/6 mice were treated with all-trans retinoic acid or vehicle control during intramuscular vaccination with recombinant adenovirus expressing the glycoprotein of LCMV as a model antigen. Vaccinated mice were challenged intravaginally with recombinant vaccinia virus or intranasally with recombinant influenza virus expressing the same gp33 immunogen. Vaccine efficacy was evaluated by systemic and mucosal T cell responses, tissue viral loads as assessed by plaque-forming units or absolute viral copy numbers, body weight changes and survival rates.

Results: Mice that received all trans-retinoic acid during vaccination exhibited stronger T cell responses and lower viral loads at mucosal sites following local challenge. Moreover, the retinoic acid-treated vaccinated mice lost less body weight and were better protected from lethal influenza viral challenge than control vaccinated mice.

Conclusion: All trans-retinoic acid appears useful as an adjuvant during systemic vaccination in mouse model to enhance acquisition of protective mucosal responses against viral challenge at mucosal sites. Thus, incorporation of ATRA into vaccine regimens represents a novel methodology that may be employed for the development of more effective T cell-based vaccines against mucosally-transmitted pathogens such as HIV. This work was supported by the Bill & Melinda Gates Foundation.

1 St George's University of London, London, England P02.04 Evaluation of TLR ligands as potential adjuvants for mucosal immunization

Background: Mucosal epithelial cells may play a critical role in maintaining the balance between tolerance and induction of IgA response. TLR ligands are also thought to play an important role in the induction of pro-IgA factors by subepithelial dendritic cells and B cells including TGF-BIL-4 and APRIL. In this study we have evaluated a broad range of TLR ligands, previously shown to induce cytokine responses in vitro, in the mouse model. We analyzed their potential to enhance systemic and local immune responses using different routes of mucosal immunization.

Methods: BALB/c female mice (aged 6-8 weeks) were immunized three times, either intravaginally, intranasally or sublingually. Immunizations were performed every two weeks and animals were treated with medroxyprogesterone (2mg/mouse) prior to intra-vaginal administration. The TLR ligands evaluated as adjuvants were: FSL-1 Poly I:C, MPLA, Pam3CSK4, R848 and CpG B. The antigens used were tetanus toxoid and HIV-1gp140. Each immunization was performed using 10 μg of antigen and 20 μg of adjuvant. Mice were sampled (blood and vaginal wash) two weeks after the third immunization. Specific IgG and IgA production was assessed by ELISA.

Results: TLR ligands including FSL-1 and poly I:C were able to greatly enhance immune responses when administered through the intranasal or sublingual routes and their effect seems to be antigen independent. In contrast, intravaginal administration of gp140, alone or in combination with TLR ligands, did not succeed in eliciting good specific responses. Interestingly, MPLA showed an enhancing effect when used intranasally but dampened specific responses when administered sublingually.

Conclusion: Our data demonstrate that some TLR ligands are potent enhancers of systemic and localized specific immune responses when administered intranasally and sublingually. These two routes are much more effective for the induction of vaginal IgA than the intravaginal route. Further studies will assess whether similar results can be obtained in NHP studies.

1 Walter Reed Army Institute of Research, Rockville, Maryland, USA 2 USMHRP, Division of Retrovirology, Henry M. Jackson Foundation, Rockville, Maryland, USA 3 The Catholic University of America, Washington DC, USA 4 USMHRP, Division of Retrovirolgy, Walter Reed Army Institute of Research, Rockville, Maryland, USA 5 USMHRP, Division of Retrovirolgy Walter Reed Army Institute of Research, Rockville, Maryland, USA P02.05 Adjuvant-Antigen formulations and route of immunization influence the induction of binding and neutralizing antibodies to HIV-1gp41

Background: Induction of neutralizing antibodies against HIV-1 has been a challenging problem in HIV-1 vaccine studies. Although several well-characterized broadly neutralizing monoclonal antibodies exist, the quest for the generation of vaccine-induced neutralizing antibodies has led to the use of different delivery platforms and prime-boost strategies. However, one aspect that has not received much attention is the inclusion of appropriate adjuvants and the routes of delivery of the immunogen formulations.

Methods: In this study, E. coli expressed gp41 [aa 512-856 with the transmembrane region deleted (684-705)] was incorporated in different formulations either as such or as a fusion protein with SOC, a small outer capsid protein of bacteriophage T4. These formulations were administered to rabbits along with liposomal lipid A or with heat liable E. coli enterotoxin as the adjuvant. The immunogens were either administered by the intramuscular route for all three immunizations or alternatively, transcutaneous immunization (TCI) was utilized as one of the delivery routes. Binding and neutralizing antibodies in individual serum samples were analyzed by ELISA and TZM-bl pseudovirus assays, respectively.

Results: Administration of SOC-gp41 consistently generated an 8-fold increase in binding antibody titers (weeks 4-20) compared to the formulation containing gp41. When these same formulations included TCI as the first boost (second immunization), the antibody titers were 3.5-23-fold higher with SOC-gp41-fusion protein. Neutralizing antibody titers were obtained only when the immunogens were delivered as SOC fusion proteins. The antibodies neutralized clades A, B, and C pseudoviruses.

Conclusion: The route of primary immunization appears to be critical for the priming of the immune system to elicit long-lasting binding antibodies. Both the route of primary immunization and inclusion of SOC contribute to the induction of neutralizing antibodies. The SOC-proteins can also be displayed on T4 bacteriophage for future immunization studies.

1 CNRS, Lyon Cedex, France, Metropolitan P02.06 Co-adsorption of TLR3 ligand and HIV-1 antigens increase potency of particulate vaccine vehicles based on chitosan or poly (lactic acid) backbone

Background: Immunization with particulate delivery-systems based on biodegradable polymers has already demonstrated a broad variety of immune abilities for HIV vaccine design. If co-administration of TLR ligands could increase dendritic cells (DCs) activation and recruitment at the injection site, a co-delivery system should be more efficient for co-stimulation and antigen processing in draining lymph nodes.

Methods: We used Poly (D,L-lactic Acid) (PLA) nanoparticles (180 nm) or chitosan/dextran sulphate (CDS) particles (350 nm) as both vehicle and adjuvant for vaccine delivery of HIV gp140 (CN54) and p24 gag. We chose Poly (I:C) as TLR3 ligand, its adjuvant properties being largely assessed, and we evaluated the dose effect of co-adsorption of Poly (I:C) and p24 or gp140 onto each nanoparticle by analyzing phenotypic and functional modifications of human monocyte derived dendritic cells (MDDCs).

Results: PLA and CDS nanoparticles are efficient protein carriers for p24 or gp140 and for p24 complexed with Poly (I:C) without alteration of the colloidal stability. Interestingly enough, CDS nanoparticles permit also the adsorption of Poly (I:C) without protein layer with a 95% adsorption yield. The in vitro maturation potentiality of human monocytes derived DCs (MDDCs) for each formulation, either alone or in combination, was assessed by FACS using the following maturation markers CD25, CD80, CD83. The co-adsorption of Poly (I:C) and p24 onto PLA or CDS nanoparticles induced a synergistic maturation of MDDCs. Moreover, it is particularly worth noting that, the level of MDDCs maturation by Poly (I:C) was increased when CDS nanoparticles were used to carry Poly (I:C).

Conclusion: We show that the co-adsorption of Poly (I:C) and HIV antigens onto various biodegradable nanoparticles is feasible and confers to the vaccine formulation positive in vitro immuno-stimulatory properties as illustrated by increasing DCs maturation abilities. Furthermore, the presence of Poly (I:C) induces the secretion of p24 vaginal IgGs in mice upon subcutaneous administration of both chitosan or PLA formulation.

1 Drexel University, Philadelphia, Pennsylvania, USA P02.07 HIV-1 gag/pol DNA vaccines elicit gut-associated immune responses when combined with highly optimized mucosal chemokine molecular adjuvants

Background: Since their conceptualization, improved immunogenicity of DNA vaccines has been achieved through improvements in the codon/RNA design of the gene constructs, use of the immune adjuvants and more effective delivery approaches. Highly optimized HIV-1 DNA vaccines and adjuvants that target and generate protective anti-HIV immunity at mucosal sites are important due to the specificity of these sites as an entry point for HIV, rapid elimination of mucosal CD4 + T-cell populations, and the role of the mucosa as a reservoir of the vast majority of immune cells. We hypothesized that immunization with codon/RNA optimized chemokine molecular adjuvants and HIV-1gag/pol plasmids will augment anti-HIV-1 T and B cell responses.

Methods: Balb/c mice were immunized with pHIV-1gag/pol in combination with either codon optimized or non-optimized mucosal chemokines, CTACK, MEC and TECK, followed by electroporation. After three immunizations, systemic, secondary lymphoid, and mucosal anti-HIV-1 T cell and antibody responses were determined by IFN-γ ELISpot and flow cytometry. HIV-1 specific humoral immune responses were measured by B cell ELISpot and in sera and fecal extracts by ELISA.

Results: Optimized chemokine adjuvants mediated significant enhancement in the polyfunctionality of HIV-1 specific splenic and gut-associated CD8 + T cell. Both codon-optimized CTACK and MEC resulted in approximately three-fold increase in IFN-γ spot forming units per million splenocytes respectively when compared with antigen alone group. We observed elevated HIV-1-specific sera and fecal levels of sIgA following co-immunization with optimized mucosal chemokines over non-optimized controls. Analysis of mucosal homing markers like α4β7 and CCR9/CCR10 highlights the role of these chemokines in trafficking of lymphocytes to the mucosal tissue.

Conclusion: These data showing an increase in HIV-1 specific secretory IgA and polyfunctional T cell responses support the use of highly optimized molecular adjuvants in our HIV-1 gag/pol vaccine platform for targeting HIV-1specific T and B cells to mucosal sites.

1 University of Miami, Miami, Florida, USA 2 Department of Medicine, University of California, San Diego, La Jolla, California, USA 3 New England Regional Primate Research Center, Harvard Medical School, Southborough, Massachusetts, USA P02.08 Latent Membrane Protein 1 (LMP1): A Novel Adjuvant for HIV and Single Cycle SIV based Vaccines

Background: Latent Membrane Protein 1 (LMP1) from EBV is able to mimic the protein clustering induced by the interaction of CD40 Ligand and the CD40 receptor and induce B cell immune activation. A chimeric protein, LMP1-CD40, has also been shown to act as a strong immunostimulatory molecule in B Cells. We explore the incorporation of these various TRAF-mediated immune activators into single cycle SIV and HIV constructs, which have potential to enhance the current generation of scSIV and HIV vaccine candidates.

Methods: LMP1, LMP1-CD40, or soluble multimeric SP-D-CD40L were cloned in place of the EGFP gene in construct SIVmac239FS-dPR-dIN-dnef-EGFP (for scSIV) and in pNL4-Bal-EGFP-IRES (for HIV). Plasmids were transfected into 293T cells and virus production was determined. Human and Macaque monocyte derived macrophages and DC were infected with various virus supernatants at a set MOI. Infected cells were analyzed for activation surface markers by flow cytometry. Various Inflammatory cytokines were analyzed from the supernatants by cytometric bead array (CBA). Quantitative RT-PCR was done to analyze the expression of various genes from infected Macrophages.

Results: Both HIV and scSIV incorporating LMP1 and LMP1-CD40 significantly increased expression of CD80 and CD40 on human macrophages and dendritic cells as compared to EGFP or parent virus. There was significantly increased secretion of cytokines such as IL-8, IL-1beta, IL-6, TNF-α and IL-12p70. SIV-LMP1 showed a significant increase in SIV Gag-specific IFN-γ ELISPOT responses in a DC:T cell co-culture assay. Adjuvanted SIV constructs also significantly increased IL-8, IL-6 and TNF-α in macaque macrophages and DC.

Conclusion: Overall, HIV and scSIV expressing LMP1 and LMP1-CD40 produced a broad and potent Th1-biased immune response in human as well as macaque macrophages and DC. This viral immune activation is anticipated to increase the immune response to single cycle SIV vaccines. These viruses are currently being tested in a primate vaccination.

1 University of Miami, Miami, Florida, USA 2 Department of Medicine, University of California, San Diego, La Jolla, California, USA P02.09 Immune responses to an HIV-1 gp140 Envelope DNA vaccine are synergistically enhanced by IL-12 combined with Tumor Necrosis Factor superfamily (TNFSF)

Background: Soluble multimeric forms of the TNFSF proteins CD40 ligand or GITR ligand have been shown to adjuvant HIV-1 DNA vaccines in mice. These novel constructs combine the extracellular domain of murine CD40L or GITRL with the body of the spontaneously multimerizing Surfactant Protein-D (SP-D). We studied a DNA vaccine expressing membrane-bound HIV-1 gp140 and adjuvant plasmids expressing IL-12 and various soluble multimeric TNFSF Ligands (SP-D-CD40L,SP-D-CD27L,SP-D-4-1BBL,SP-D-BAFF,SP-D-APRIL, and SP-D-GITRL) in mice models. Membrane-bound gp140 envelope can form stable trimers on the transfected cell membrane, and induce neutralizing antibodies in vaccinated animals.

Methods: Antigen plasmid p96ZM651gp140-CD5-opt encodes membrane-bound HIV-1 gp140 protein (AIDS Research Reagent Program). Various soluble 4-trimer adjuvant constructs of TNFSF ligands were tested for their immune response in combination with plasmid expressing murine IL-12. BALB/c mice were vaccinated i.m. every two weeks X3 with 80μg of antigen plasmid and 20μg pIL-12 plus 20μg of various TNFSF ligand plasmids. Spleen cells and serum were harvested two weeks later for ELISA and ELISPOT.

Results: Combinations of p96ZM651gp140-CD5-opt with pIL-12 and TNFSF adjuvant plasmids induced strong immune responses, determined by ELISPOT and ELISA. Vaccination with pSP-D-CD27L, pSP-D-GITRL (p < 0.01), or pSP-D-APRIL with pIL-12 and gp140 plasmid synergistically enhanced IFN-gamma and IL-2 ELISPOT responses compared to gp140 plasmid plus pIL-12 alone. In addition, pSP-D-CD27L induced high titers of IgG2a and IgG1 anti-gp140 antibodies, and pSP-D-BAFF plus pIL-12 induced significantly higher IgG1 titers (1:500) when compared to gp140 plasmid plus pIL-12 alone (p < 0.05). Overall, anti-gp140 titers of greater than 1:2,000 could routinely be induced in mice vaccinated with pIL-12 plus various TNFSF ligands. These serum samples will be evaluated in HIV neutralization assays.

Conclusion: The inclusion of IL-12 plus TNFSF ligands in a membrane-associated gp140 vaccine can induce both humoral and cellular immune responses against gp140. Suggest a new role for IL-12 in combination with TNFSF ligands as adjuvants for HIV vaccination.

1 Nankai University, Tianjin, China P03.01 Partial protection of SHIV-infected Chinese rhesus monkeys against super-infection with a heterologous or homologous SHIV isolate

Background: Many studies have revealed a protective effect following SIV or SHIV infections with variable results and interpretations. However, whether a protective response can be induced by natural infection with an immunodeficiency virus is still currently debated in the HIV-1 vaccine field. The aim of this study was to evaluate the protection from the SHIV challenge in Chinese macaques vaccinated with different kinds of SHIVs.

Methods: Fifteen adult Chinese rhesus monkeys were inoculated with pathogenic SHIV-KB9, SHIV-1157ipd3N4, or nonpathogenic SHIV-CN97001. All were infected with high viral loads and demonstrated a specific binding Ab response. After 30 weeks, animals were exposed to SHIV-KB9 or SHIV-CN97001, which carried a heterologous or homologous envelope protein relative to the vaccine strain. Infection was monitored by viral load, binding Ab measurement, and virus genome DNA PCR assay.

Results: All control animals were infected, while in test monkeys protection from super-infection was statistically significant. One group of animals SHIV-CN97001 vaccinated and SHIV-KB9 challenged showed evidence of new infection by virus genome SGA PCR, while others showed no evidence of super-infection. Perhaps this protective state correlated with host immune responses, such as neutralizing antibodies against different characteristic viruses.

Conclusion: These findings indicate that different characteristic SHIV infection confers different levels of protection against a second SHIV infection in Chinese monkeys. This protective state may serve for HIV-1 vaccine development, as a similar degree of protection against immunodeficiency virus infections exists in humans.

1 Ragon Institute of MGH, MIT and Harvard, Charlestown, Massachusetts, USA 2 Center for Immunology and Inflammatory Diseases, MGH and HMS, Boston, Massachusetts, USA P03.02 Functional CD8 + T Cell Responses and Viral Escape In a Humanized Mouse Model of HIV Infection

Background: The rhesus macaque model has proven invaluable to the study of HIV pathogenesis and vaccine design. However, the expense of non-human primates, differences between human and macaque genetics and substantial sequences differences between HIV and SIV, impair the ability to model human immune responses in macaques or to study the role of HIV-specific responses due to targeting of different CD8 epitopes. Recent work has illustrated that humanized NOD/SCID-BLT (bone marrow, liver, thymus) mice are capable of supporting robust levels of HIV, with concomitant CD4 + T-cell declines. Little is known, however, about the reproducibility or function of cellular immune responses in these mice.

Methods: To determine if functional CD8 + T-cell responses were present we first examined viral sequence evolution in infected NOD/SCID-BLT mice and then attempted to expand antigen specific T-cell lines.

Results: Sequencing the entire HIV genome in mice from the same tissue reconstitution identified multiple mutations within 6 weeks of infection with JRCSF. Surprisingly, these mutations were predominately located within well-defined CD8 + T-cell epitopes restricted by the donor HLA (p = 0.002). In particular, reproducible mutations arose within two epitopes that are normally immunodominant in HIV-infected subjects, HLA-A01 Env RY9-RRGWEVLKY and Cw03 Nef AL9-AAVDLSHFL. At 8 weeks post-infection the frequency of these mutations continued to increase, while these epitopes remained unchanged in A01- and Cw3-negative control mice. The presence of epitope-specific CD8 + T-cell responses were confirmed by IFN-gamma Elispot and ICS for both epitopes using expanded T-cell lines.

Conclusion: These data illustrate that NOD/SCID-BLT mice are capable of mounting epitope-specific CD8 + T-cell responses that are functional in terms of IFN-gamma production and their ability to exert selection pressure on HIV. Moreover, the observed escape within the same CD8 epitopes that are immunodominantly targeted and escape during acute HIV infection of humans supports the potential of this model for HIV-specific vaccine and pathogenesis studies.

1 Blood Systems Research Institute, San Francisco, California, USA 2 Jenner Institute, Oxford, Oxfordshire, United Kingdom 3 Harvard University, Boston, Massachusetts, USA P03.04 Pathogenic SIV infection induces a striking acute-phase systemic cytokine response similar to that observed in acute HIV-1 infection

Background: The nature and kinetics of cytokine responses during acute and early SIV infections of non-human primates (NHP) were analyzed to give insight into the relationship between the acute-phase cytokine response and disease pathogenesis and to determine their utility as models for studying cytokine responses after vaccination or other interventions.

Methods: Rhesus (RM) and African green monkeys (AGM), pre-treated with IV IgG, were infected with SIVmac251or SIVagm and blood samples were collected up to 320 days post infection. A broad spectrum of cytokine and chemokine responses was measured over time using cytokine bead array and ELISA assays.

Results: In the pathogenic SIVmac251 infection of RM, during peak in viremia, there was a cascade of cytokine responses similar to that in HIV infection in humans including elevations in IFN-α, IL-15 and IL-18. During peak viremia in the AGM, there was an elevation in some of the analytes measured, such as IL-1β, IL-1Ra and IL-8. However, in comparison to the RM, the cytokine elevations in the AGMs were typically of a lower magnitude. In RM, the magnitude of IFN-γ, IL-15, MIP-1β, TGF-α, VEGF and IL-18 elevations during the first three weeks of infection positively correlated with the viral setpoint at day 100 after SIV infection. There was also correlation between IL-18 cytokine elevations, detectable in 2 of 6 AGM with acute SIVagm infection, and viral set point.

Conclusion: These results give insight into the differences in immune responses in pathogenic and non-pathogenic SIV infections. These primate models can be utilized in different ways to assess the effects of immune modulation or manipulation by vaccines at the earliest stages of infection.

1 NIAID/NIH, Bethesda, Maryland, USA 2 Vaccine Branch, NCI/NIH, Bethesda, Maryland, USA 3 AIDS and Cancer Virus Program, SAIC-Frederick, Inc., Frederick, Maryland, USA 4 Department Pathology and Laboratory Medicine, University of Wisconsin, Madison, Wisconsin, USA P03.06 Cytotoxic Capacity of SIV-Specific-CD8 + T Cells from SIV-Infected Rhesus Macaques Measured Against Primary Autologous CD4 + T Cells

Background: Determining the mechanisms of immune mediated control of lentiviral replication in chronic infection and vaccinees is of critical importance to HIV vaccine development. The HIV-specific CD8 + T cells of long-term nonprogressors (LTNP) are clearly distinguished from those of progressors by per-cell cytotoxic capacity against autologous HIV-infected cells. We are evaluating whether similar mechanisms are operative in SIV-infected rhesus macaques (RM).

Methods: Cytotoxicity and IFN-γ production of SIV-specific CD8 + T cells were examined following 6-day incubation with autologous SIVmac239-infected telomerase-transduced CD4 + T cell targets. SIV-specific CD8 + T cell cytotoxic responses were measured at 1 hour by flow cytometric detection of granzyme B (GrB) delivery to live targets and infected CD4 elimination (ICE).

Results: In these preliminary studies, ten SIV-infected RM with different extents of control of viral replication were evaluated. Data collection is ongoing with laboratory personnel blinded to RM virologic status. There is a significant correlation between ICE and viral load (R = − 0.64, p = 0.04). There is a significant correlation between GrB delivery and ICE (R = 0.71, p = 0.03). LTNP monkeys had higher median ICE than progressor monkeys (65.45% vs 7.6%) but the difference was not statistically significant (p = 0.07).

Conclusion: These results suggest measurements of per-cell CD8 + T cell cytotoxic effector function might be associated with immune control in SIV-infected RM. If these functions are clearly associated with immune control of SIV in chronically infected or vaccinated macaques they may provide critical information for the development and optimization of these responses.

1 New York University School of Medicine, New York, New York, USA 2 Virology Unit/Institute of Tropical Medicine, Antwerp, Belgium P04.01 Primary HIV-1 isolates in infected drug naïve individuals that evolve increase neutralization sensitivity to anti-gp120 monoclonal antibodies (mAbs)

Background: Because a potent vaccine against HIV-1 will induce heterologous antibodies against HIV-1, we examine how virus evolution affects neutralization sensitivity to anti-HIV-1 mAbs and determine changes in core epitopes.

Methods: Thirteen mAbs directed at epitopes in V2, V3, CD4bd, and carbohydrate in gp120 were tested using the TZM-bl neutralization assay with sequential primary HIV-1 isolates obtained from five drug naïve individuals. The gp120 sequences from the viruses were analyzed to identify changes in core epitopes.

Results: Primary viruses collected from patients at first visits were resistant to neutralization by all anti-HIV-1 mAbs with the exception of one virus sensitive to IgG1b12. Four of the five patients' viruses evolved increased sensitivity to neutralization by anti-V3 mAbs. Virus collected from a patient obtained 31 months later, evolved increased sensitivity to anti-V2, anti-V3, and anti-CD4bd mAbs. Furthermore, the anti-V2 and anti-CD4bd mAbs also exhibited increased neutralization capacities against virus collected from a patient 12 months later. Of the seven anti-V3 mAbs, five showed increased potency to neutralize the evolved virus from a patient collected at 11 months, and three exhibited increased potency against viruses from two patients collected 12 and 36 months later. Anti-V3 mAbs exhibited the most breadth and potency in neutralizing the evolving viruses. Sequence analysis of the envelope regions revealed amino acid conservation within the V3 loop while most of the changes identified occurred outside the core epitopes, and in particular within the C3 region.

Conclusion: These studies suggest that HIV-1 strains infecting patients can evolve increase neutralization sensitivity to mAbs and that changes that occur outside the core epitope may account for increased sensitivity. The results also reveal that the spectrum of neutralization sensitivities can be broader when studied longitudinally than cross-sectionally. Understanding how viruses evolve increase neutralization sensitivity may facilitate strategies to induce potent protective antibodies.

1 The Scripps Research Institute, San Diego, California, USA 2 International AIDS Vaccine Initiative (IAVI), New York, New York, USA P04.02 A Limited Number of Antibody Specificities Mediate Broad and Potent Serum Neutralization in Selected HIV-1 Infected Individuals

Background: A protective vaccine against HIV-1 will likely require the elicitation of a broadly neutralizing antibody (bNAb) response. Although the development of an immunogen that elicits such antibodies remains elusive, a proportion of HIV-1 infected individuals evolve broadly neutralizing serum responses over time, demonstrating that the human immune system can recognize and generate NAbs to conserved epitopes on the virus. Understanding the specificities that mediate broad neutralization will provide insight into which epitopes should be targeted for immunogen design and aid in the isolation of broadly neutralizing monoclonal antibodies from these donors.

Methods: We have used a number of established and new techniques, such as selective removal of certain antibody specificities using antigen-coated beads, inhibition of neutralizing activity using a panel of mutant viruses, and the use of chimeric viruses displaying specific epitopes, to map the broadly neutralizing antibody specificities in the serum of individuals who were ranked in the top 5% of neutralizers identified in our previous study.

Results: In approximately half of the donors, key N-linked glycans were critical for expression of the epitopes recognized by the bNAb specificities in the sera. NAbs with similarities to the antibodies PG9 and PG16 directed to conserved regions of the V1, V2 and V3 loops were identified in approximately 21% of donors. In one donor, the broad serum neutralization was primarily mediated by NAbs directed against the glycan shield. CD4 binding site (CD4bs)- directed NAbs contributed strongly to the broad responses in approximately 11% of the donors tested.

Conclusion: We show that the broad neutralization in the sera of most of the individual donors that we studied can be associated with single or a small number of specificities. Across the donor panel, broad neutralization appears associated with 4-5 principal specificities.

1 Biomedical Primate Research Centre, Rijswijk, Netherlands 2 Department of Veterinary Medicine, Cambridge, United Kingdom 3 Novartis Vaccines and Diagnostics, Cambridge, Massachusetts, USA P04.03 Cross-neutralization of heterologous isolates by sera from rhesus macaques immunized with candidate HIV-1 vaccines

Background: Under the conditions of natural HIV transmission, it is unlikely that challenge will be homologous to the vaccine. Moreover, the inocula will vary both in their dose and the relative neutralization resistance of their virus. Our aim was to investigate whether sera from macaques immunized with candidate HIV vaccines can neutralize low doses of heterologous virus.

Methods: Sera from macaques immunized with candidate vaccines with high neutralization activity against homologous virus (HIV-1 SF162) were selected. Virus stocks were prepared in PHA-transformed human PBMCs. Virus dilutions and serum (1 in 20) were incubated for four hours at 37°C. Hi5 GHOST cells were then exposed to the mixture for 16 hours, washed free of residual infectious virus and cultured for a total of two days. The number of infectious virus was determined by quantifying the fuorescent cells. Data were analyzed by linear regression.

Results: Plots of fluorescent cells against virus dilution were linear with r² values greater than 0.85. Two patterns of plots were observed with the relatively neutralization resistant subtype B HIV-1 89.6 isolate. First, a statistically significant reduction in the gradient of the plot despite reductions in virus titre being less than 50%. Second, plots where the reduction in the gradient was not statistically significant although there were reductions in infectious virus titre such that the intercept on the fluorescent cell (y-) axis reached statistical significance. There were also differences between the intercepts for the control and immune sera on the virus dilution (x-) axis. At doses of virus between these values, 100% neutralization was observed.

Conclusion: The indications are that HIV vaccine candidates can induce antibodies able to completely inactivate heterologous virus which is relatively resistant to neutralization but only at very low doses. These observations may be validated by titrating SHIV challenge virus in control and immunized macaques.

1 Walter Reed Army Institute of Research, MHRP, Rockville, Maryland, USA 2 U.S. Military HIV Research Program, Henry M. Jackson Foundation, Rockville, Maryland, USA P04.04 Differences in Lipid Binding of Neutralizing and Non-neutralizing Murine Multispecific Monoclonal Antibodies that Bind to the Same Peptide Epitopes

Background: We have recently reported that the murine IgM monoclonal antibodies (mAbs) WR316 and WR320 were multispecific (binding to protein and lipid) and neutralized HIV-1 (Matyas et al., AIDS 2009; 23:2069–77). The peptide binding epitopes of WR316 and WR320 were identified as SLWNWF and LELDKWAL, respectively. In order to investigate the mechanism of neutralization of WR316 and WR320, their lipid binding specificities were compared to those of non-neutralizing murine mAbs that recognized the same peptide epitopes.

Methods: Lipid, liposome and peptide epitope binding was determined by ELISA. HIV-1 neutralization in PBMC was performed with a replication-competent molecular clone of HIV-1 encoding the envelope protein from SF162 and Renilla luciferase as reporter.

Results: WR316 and WR320 neutralized HIV-1, but WR307, WR308, and WR319 did not. Neutralization was not due to antibody contamination by endotoxin, since the addition of lipopolysaccharide from E. coli 055:B5 to the same donor PBMC did not inhibit infection. WR316 and WR320 bound to HIV-1 at low levels in a virus capture assay. The peptide binding epitope of WR308 were identified as SLWNWF and the epitopes of WR307 and WR319 were LELDKWAL. WR308 bound significantly less to phosphatidylinositiol, cardiolipin, and phosphatidylglycerol compared to WR316. WR307 and WR319 had lipid binding specificities very similar to WR320, except that WR320 had reduced binding to phosphatidylserine and phosphatidic acid. WR307, WR308, and WR319 bound strongly to liposomes whereas WR316 and WR320 bound poorly to liposomes.

Conclusion: The difference in neutralizing capability of the mAbs is not related to the peptide binding epitope, but may be due to the reduced lipid binding of WR308 compared to WR316. In addition, the neutralizing antibodies had diminished binding to liposomes compared to the non-neutralizing antibodies. This suggests that the greater binding to the lipid portion of the membrane by the non-neutralizing antibodies may hinder their ability to neutralize.

1 National Center for AIDS/STD Control and Prevention, China CDC, Beijing, China 2 Duke University Medical Center, Durham, North Carolina, USA P04.06 Profiles of neutralizing antibody responses in chronically HIV-1 clade B' infected treatment-naïve former plasma donors in China

Background: Characterizing neutralizing antibody (Nab) responses in individuals chronically infected with locally circulating HIV-1 strain will facilitate the development of HIV-1 vaccine. Here we assess the breadth, magnitude, prevalence of Nab responses in clade B' chronic infection in China, and correlate former two parameters to clinical variables.

Methods: 106 ART- naïve patients were recruited from former plasma donors in central China. The standardized TZM-bl Nab assay was performed against a panel of tier 2-3 pseudoviruses composed of clade B(8), A(4), C(4), CRF07_BC(4) and 3 well-defined CRF01_AE strains and 2 tier1 viruses(SF162.LS and MW965.26). As negative control virus, SVA-MLV positive samples were excluded from analysis.

Results: Overall, all 103 plasma samples (excluding 3 SVA-MLV positive samples) neutralized both tier 1 viruses. 64% of samples (n = 66) neutralized half of the viruses tested. 2% (n = 2) neutralized all the viruses and 5% (n = 5) none of the viruses tested. Strikingly, 29% of samples (n = 30) neutralized more than 80% strains tested. Significant difference of geometric mean ID50 titer (GMT) between intraclade and interclade samples (p < 0.001) was observed. When correlating magnitude or breadth with CD4 + T cell counts or plasma viral load, only the positive association between viral load and GMT(r = 0.218, p = 0.0268)/breadth (r = 0.197, p = 0.0464) was observed. Notably, the positive association between viral load and corresponding GMT was significant (r = 0.658, p < 0.001) in the samples with broadly cross-reactive neutralizing activities (n = 30). Heatmap analysis against top 8% neutralizers demonstrated that each plasma exhibited unique pattern against the spectrum of viruses.

Conclusion: We detected a high prevalence of Nabs and broadly cross-reactive neutralization activities in 29% of samples tested, indicating presence of potential shared neutralization determinants among circulating HIV-1 subtype B' viruses in China. Our data also suggested that subjects showed more Nab against matched-subtype virus strains compared to other subtype/CRF strains.

1 PHRI Center, University of Medicine and Dentistry of New Jersey, Newark, New Jersey, USA 2 New York University Langone School of Medicine, New York, New York, USA 3 Tulane University Medical Center, New Orleans, Louisiana, USA P04.07 Mapping V2 and V3 determinants of related quaternary epitopes recognized by human and macaque mAbs possessing unusually potent neutralizing activities

Background: Novel mAbs specific for quaternary neutralization epitopes (QNEs) that potently neutralize HIV-1 have been recently isolated from both HIV-1-infected humans and monkeys. A common feature of these mAbs is their dependence on specific residues in the V2 and V3 domains of gp120. In this study we compared the V2 and V3 determinants of human mAb 2909 with those of 11 mAbs isolated from macaques infected with an SF162-derived SHIV.

Methods: MAbs were isolated by screening for neutralization of SF162, and were all specific for this isolate. 2909 came from a human subject infected with a heterologous HIV-1 isolate, while the monkey mAbs were from macaques infected with the autologous SHIV-SF162.P4 strain. The sequence dependence of these epitopes was determined by neutralization assays against a series of site-specific mutants in the V2 and V3 regions of SF162 and JR-FL Envs.

Results: The fine specificity of the monkey mAbs for both V2 and V3 regions differed from that of 2909 and, to a lesser extent, from each other. Whereas the limited breadth of 2909 was mainly due to the requirement of a single substitution at position 160 (N160K), the monkey mAbs were also dependent on substitutions at positions167–169, and, in most cases, required R315 at the tip of the V3 loop. Substitutions at positions 167 and 169 were also found to have a significant masking effect on standard epitopes in the V3 and CD4-binding domains.

Conclusion: The different epitope specificities of 2909 and the monkey mAbs suggest that the specificity and consequent breadth of such antibody responses is strongly influenced by the sequence of the infecting virus. The unprecedented neutralization potency of the human and macaque antibodies argues that this class of QNEs represents extremely sensitive neutralization targets, and that related epitopes with more conserved sequences may be important targets for HIV-1 vaccines.

1 Fraunhofer Institut für Biomedizinische Technik, Sulzbach, Germany 2 ICREA and University Pompeu Fabra, Barcelona, Spain 3 CA-VIMC, Duke University Medical Center, Durham, North Carolina, USA P04.08 First implementation of an automated cell culture and transfection system for HIV-1 Env-pseudotyped virus production

Background: Standardized assessments of HIV-specific immune responses are essential for prioritizing candidate vaccines in preclinical and clinical stages of development. With respect to neutralizing antibodies, assays with HIV-1 Env-pseudotyped viruses (HIVpv) are a high priority. To cover the increasing demands of the HIVpvs, a complete cell culture and transfection automation system has been developed.

Methods: The automation system for HIVpv production comprises a modified Tecan-based Cellerity system. It covers an area of 5x3 meters and includes a robot platform, a cell counting machine, a CO2 incubator for cell cultivation and a media refrigerator. The processes for cell handling, transfection and pseudovirus production have been implemented according to manual standard operating procedures and are controlled and scheduled autonomously by the system. The system is housed in a biosafety level 2 hood that guarantees protection of personnel, environment and the product.

Results: HIVpv stocks have been produced in a scale of 140 ml each on the automated system. Parallel manual production of HIVpvs and comparisons (bridging assays) confirmed that the pseudoviruses produced by the automated system were equivalent to those produced manually.

Conclusion: An automated HIVpv production system has been successfully established. It will allow the high quality production of HIVpvs under GCLP conditions. In its present form, the installed module allows the production of 1,000 ml of virus per week. Thus, this novel approach could facilitate standardized large-scale productions of HIVpvs for upcoming HIV vaccine trials.

This project is a collaboration of the Fraunhofer IBMT (Principal Investigator: Hagen von Briesen) and the “Comprehensive Antibody Vaccine Immune Monitoring Consortium” (CA-VIMC, Principal Investigator: David Montefiori) within the “Collaboration for AIDS Vaccine Discovery” (CAVD), funded by the Bill and Melinda Gates Foundation (#OPP38580_01). Automation of cell culture and HIV pseudovirus production was established in exclusive cooperation with TECAN funded as a strategic investment by Fraunhofer, Germany.

1 University College London, London, United Kingdom 2 Ablynx NV, Ghent, Belgium 3 University of Utrecht, Utrecht, Netherlands P04.09 Llama antibody fragments that recognize gp41 of HIV-1 envelope glycoproteins

Background: Llamas and camels make natural, single-chain antibodies from which the N-terminal antigen-recognizing variable domain (VHH) can be cloned and expressed in prokaryotic and eukaryotic cells. VHH are 12-15kDa molecules that can be expressed in large quantities and be selected to be stable at different pH and formulations, and be made suitable for use as microbicides or to identify new vaccine targets.

Methods: Llamas were immunized with envelope glycoprotein gp140 from a HIV-1 subtype B/C chimeric isolate, 97CN54. The immune library repertoire was panned for VHH that can recognize and bind to gp41. These VHH were then tested in neutralization assays and in kinetic studies.

Results: After screening of the selected clones through binding and neutralization assays, most of the selected VHH were found to bind well to gp41 but were weakly neutralizing. Four different VHH were then selected for further characterization. These were able to neutralize subtype C isolates better than subtype B isolates. Three of these VHH belong to the same family of VHH, and compete with each other for the same epitope on gp41. The fourth VHH appear to bind to another epitope on gp41. All the four VHH recognize a conformational dependent epitope on gp41, and an attempt at epitope mapping was performed.

Conclusion: The results indicate that VHH have a possible use as potent HIV-1 entry inhibitors. As they can be easily manipulated to form stable molecules and can be produced at a relatively low cost, llama VHH may be considered for applications such as vaginal microbicides as well as tools for indentifying vaccine targets.

1 Henry M. Jackson Foundation, Rockville, Maryland, USA 2 University of Alabama at Birmingham, Birmingham, Alabama, USA 3 Walter Reed Army Institute of Research, Rockville, Maryland, USA 4 Duke University, Durham, North Carolina, USA P04.11 HIV-1 neutralization using pooled versus individual PBMC donors as target cells

Background: The peripheral blood mononuclear cell (PBMC) HIV-1 neutralization assay is considered to be a physiologic platform for characterizing monoclonal antibodies (mAbs) and sera/plasma from infected or vaccinated subjects. However, this platform has been plagued by variability and a lack of sustainability of PBMC donor source. Here we determine the range in neutralization profiles observed using individual PBMC donors and assess the effects of pooling PBMC in neutralization assays.

Methods: PBMC were stimulated before pooling equal numbers of cells per donor. Pools or individual PBMC from 6 donors were used to assess neutralization of infectious HIV-1 molecular clones expressing Renilla luciferase. The range in neutralization potency using pooled versus individual constituent PBMC was assessed using sCD4, 4E10 and 2G12 mAbs, and two polyclonal reagents.

Results: The wide range (of nearly 2 logs) in neutralization potency observed using individual donors was mitigated by pooling PBMC; range was reduced by one log for most reagents tested. Using pooled PBMC, greater potency was observed for all reagents tested, except sCD4; thus no loss in the ability to measure antibody-mediated neutralization was observed. Interestingly, while donor pooling had little effect on the IC₅₀ observed for sCD4 (Individual donor mean = 2.02 ug/ml; donor pool mean = 2.34 ug/ml), neutralization by 2G12 was significantly greater using the donor pools (p = 0.003; 11-fold decrease in IC₅₀). The average range in neutralization values for 6 individual donors using 4E10, 2G12 and a polyclonal serum pool was 48.5-fold, compared to 5.7-fold for 8 independent pools composed of the 6 individuals.

Conclusion: Pooling PBMC from multiple donors may provide a way to mitigate the considerable variation in HIV-1 neutralization observed using individual donors. With no loss of assay sensitivity, PBMC pooling increases the sustainability of target cells available for testing multiple samples from vaccine trials; this approach should facilitate standardization of the PBMC neutralization assay.

1 The Scripps Research Institute, La Jolla, California, USA P04.12 Very few substitutions are required to initiate significant domain-exchange in 2G12

Background: 2G12 is a broadly neutralizing anti-HIV-1 antibody reactive with a high-mannose glycan cluster on the surface of gp120. A key feature of the antibody is domain-exchange of its heavy chain variable regions creating a multivalent binding surface with favorable avidity. Here, we systematically explored the minimal requirements for domain-exchange to obtain clues as to how this arrangement might be elicited through immunization.

Methods: Amino acid substitutions corresponding to the wild-type 2G12 sequence were introduced in the heavy chain of selected antibodies at specific positions. IgGs were expressed in FS293 cells and purified using affinity chromatography. Domain-exchange was assessed by size exclusion chromatography of Fabs prepared by papain digest.

Results: A germline version of 2G12 was found to behave as a conventional antibody but substitution of only 5-7 residues with those of the wild-type produced a significant fraction of domain-exchanged molecules. Two wild-type residues not previously implicated, A^H14 and E^H75, were shown to be most crucial for the domain-exchange together with I^H19 at the V_H/V_H’ interface and P^H113 in the elbow region between the V_H and C_H1 domains. Considerable domain-exchange could be introduced with the same 5–7 substitutions in one of six other antibodies tested.

Conclusion: The demonstration that domain-exchange can be initiated by a small number of substitutions in a germline 2G12 antibody may have important consequences for the evolution of a domain-exchanged antibody response. Other experiments show the strict maintenance of the conformation and specificity of the antigen binding site in conventional and domain-exchanged forms of the 2G12 antibody. Therefore binding to antigen can be evolved initially in the context of a conventional antibody. Further studies may provide clues what additional factors are required to engineer domain-exchange into antibodies other than 2G12 and how to promoted domain-exchange through immunization.

1 Emory University, Atlanta, Georgia, USA 2 Emory University, Department of Pathology, Atlanta, Georgia, USA 3 Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts, USA P04.13 Rapid depletion of Activated memory B cells and Involvement of the PD-1 pathway in Rapidly Progressing SIV infection

Background: Rapid HIV disease progression is a significant problem especially in developing countries, where the majority of HIV-infected individuals live. The mechanisms underlying rapid disease progression are however not well understood.

Methods: We sought to thoroughly characterize the B cell compartment of RM with different rates of disease progression to determine the role, if any, B cell dysfunction and immune activation may play in rapid disease progression. We also followed the course of non-SIV infections to correlate our immunological findings with clinical changes occurring in the SIV-infected animals. In addition, we compared the B cell compartments of animals with pathogenic (RM) and non-pathogenic (sooty mangabeys, SM) SIV infections to understand the role of B cell defects in pathogenesis and disease progression.

Results: We show that pathogenic SIV infection in rhesus macaques results in a rapid (as early as week 2) depletion of activated memory B (CD21− CD27+) cells, which was strongly associated with rapid disease progression. This depletion was progressive and sustained in rapid progressors, less severe and transient in typical progressors, and was greater in animals with higher percentages of activated memory B cells prior to infection. Consequently, rapid progressors failed to develop SIV-specific antibody responses, showed a significant decline in non-SIV-specific Ab titers and succumbed faster to intestinal bacterial infections. Depletion of activated memory B cells was strongly associated with preferential depletion of Programmed Death-1 (PD-1) + activated memory B cells and in vitro blockade of PD-1 improved their survival. Similar B cell abnormalities were not found in SIV-infected sooty mangabeys, which do not progress to disease despite high viremia.

Conclusion: Our results identify depletion of activated memory B cells as a very early predictor of rapid disease progression in pathogenic SIV infection and suggest that pre-existing high humoral immune activation may contribute to rapid HIV disease progression.

1 Henry M. Jackson Foundation/USMHRP, Rockville, Maryland, USA 2 USMHRP/WRAIR, Rockville, Maryland, USA P04.14 Binding and Accessibility of HIV-1 gp41 Epitopes to Human and Murine Monoclonal Antibodies

Background: A modified ELISA-based virus capture assay was utilized to investigate the accessibility and conservation of gp41 epitopes on 5 different clades of CCR5-tropic HIV-1 pseudoviruses (A: 93RW, KNH; B: US-1 PV, BAL-PV; C: PBL 288, PBL 286; D: 57128, A07412; and CRF01_AE: CM235, NI) in the absence and presence of sCD4.

Methods: Protein G plates were individually coated with three human mAbs (2F5, 4E10, and 2G12), two murine multispecific IgM mAbs (WR316 and WR320) that exhibit both lipid and gp41 epitope specificities, or one murine IgM mAb (WR301) that exhibits lipid-binding specificity. Pseudoviruses were added to the plates alone or with sCD4. Pseudovirus capture and TZM-bl neutralization were evaluated.

Results: The binding of human and murine mAbs to HIV-1 was dependent upon the virus clade with clades A, B (BAL-PV) and D exhibiting moderate to high binding to 4E10 and 2F5, respectively. Studies on the relationship between virus binding and neutralization in a TZM-bl pseudovirus assay indicated that in most cases, mAbs that exhibited neutralization also bound the virus. However, binding per se did not predict neutralization. Monoclonal IgM antibodies bound significantly lower amounts of HIV-1 than the corresponding IgG isotype. Although murine IgM mAbs failed to neutralize HIV-1 in a TZM-bl pseudovirus assay, human IgG mAbs 2F5 and 4E10 did neutralize isolates from several HIV-1 clades.

Conclusion: Soluble CD4 significantly increased the accessibility of gp41 binding epitopes on several isolates that previously showed little or no binding with 2F5 and 4E10 mAbs. The hidden or conformational gp41 epitopes unmasked by sCD4 may provide additional targets for vaccine design.

The views expressed are those of the authors and should not be construed to represent the positions of the Department of the Army or Department of Defense.

1 National Institute for Biological Standards & Control, Potters Bar, Hertfordshire, United Kingdom 2 AIDS Immunopathogenesis and Vaccine Development, Robert Koch-Institute, Berlin, Germany P04.15 Analysis of neutralizing antibody associated with sterilizing immunity against SIV challenge

Background: There is uncertainty as to which approach to measure virus neutralization correlates with antibody mediated protection in vivo. Vaccination of macaques with inactivated SIV provides potent sterilizing immunity against live SIV challenge, which can be passively transferred with immune serum alone. We have characterized the neutralizing antibody responses in vaccinated macaques to identify features of antibodies that are associated with protection in vivo.

Methods: Sera from groups of vaccinated cynomolgus macaques were analyzed by neutralizing antibody assay and for HLA class I and II antibody by Luminex multiplex technology.

Results: Antisera of protected macaques were able to neutralize virus unrelated to that of vaccine or challenge and was dependent on concordance of cell substrate used for vaccine and virus challenge production. This supports the concept that protection is provided by neutralizing antibodies specific for HLA antigens that are virion components. Complement was critical for the correlation of protection with neutralizing activity since SIV specific neutralization could be detected in the absence of complement in all vaccinated animals irrespective of the outcome of virus challenge.

An association of HLA antibody with sterilizing immunity was also observed. Class I and II antibodies were detected against all alleles and suggests specificity for framework regions. The specificity of anti Class II responses was further defined to be restricted to HLA DR. This is distinct from human vaccinees given a HIV inactivated vaccine where there is an allele specific response.

Conclusion: These data have important implications for assessing potent anti-HIV neutralizing antibodies. In particular it indicates that the ability to activate complement is critical for efficacy in vivo.

1 Academic Medical Center, University of Amsterdam, Amsterdam, Netherlands P04.16 Comparison of early HIV specific neutralizing activity in an elite neutralizer and in four patients who develop broadly neutralizing activity

Background: Previously we screened sera obtained at 3 years post-seroconversion from 82 individuals for broadly neutralizing activity (BrNA) and revealed that ∼30% of HIV-1 infected individuals had BrNA, in line with observations by others. It remains to be established how and when BrNA develops and what the role of autologous neutralizing antibodies is in this process.

Methods: Here we studied retrospectively the development of BrNA in serum samples from 5 patients, including one elite neutralizer, who had the highest geometric mean IC50 titer (GMT) on the large multi-clade virus panels used. For the testing of BrNA, sera were obtained at 3 monthly intervals in the first year after seroconversion and at 6 monthly intervals thereafter until death, initiation of HAART or lost to follow-up. Sera were tested against a representative multiclade panel of 6 viruses (Simek et al., JVI 2010). Clonal virus variants were isolated at multiple time-points covering the disease course from seroconversion till death and tested for sensitivity to autologous serum neutralizing activity using sera from the same time-points that were tested for BrNA.

Results: The elite neutralizer developed BrNA already at 12 months post-seroconversion, in contrast to the other four patients who first had BrNA at 30-35 months post-seroconversion. At this same time-point, all five patients showed a peak in neutralizing activity after which the GMT decreased slightly but remained high. Development of BrNA coincided with neutralizing activity against autologous viruses that were isolated < 12 months post-seroconversion in all five patients. Viruses from later time-points had escaped autologous neutralizing activity in all patients.

Conclusion: The very early development of BrNA in the elite neutralizer may suggest that his antibodies required only minimal affinity maturation to become strongly neutralizing as compared to antibodies from the other patients. A better understanding of this process in the elite neutralizer may help vaccine design.

1 National Institute for Communicable Diseases, Johannesburg, Gauteng, South Africa 2 University of Cape Town, Cape Town, South Africa 3 Centre for the AIDS Programme of Research in South Africa, Durban, South Africa P04.18 Cross-neutralizing antibodies in HIV-1 infected individuals recognize novel epitopes on monomeric gp120

Background: Defining the epitopes recognized by broadly cross-neutralizing (BCN) antibodies will assist in identifying targets for an HIV-1 vaccine. Previously, we found BCN activity in 7 of 35 individuals from the CAPRISA 002 cohort infected with HIV-1 subtype C for at least 3 years. Here we describe the study of BCN antibodies directed against monomeric gp120.

Methods: Recombinant gp120s coupled to magnetic beads were used to adsorb antibodies from sera of the 7 individuals with BCN activity. Adsorbed plasmas were tested for neutralization against selected subtype A, B and C viruses. Envelope amplicons were generated at multiple time-points by single genome amplification and some were cloned to generate pseudoviruses for neutralization assays.

Results: In 2 of the 7 individuals tested, the bulk of the neutralizing activity observed at 3 years post-infection was adsorbed using gp120-coated beads. Gp120s with mutations in the CD4 (D368R) or the coreceptor (I420R) binding sites adsorbed the BCN activity equivalently to wildtype gp120 suggesting these sites are not involved. Heterelogous neutralization developed gradually over time from the second year of infection with no subtype-specificity. Adsorptions of longitudinally collected plasmas suggested that the epitope(s) in gp120 was the target of the BCN activity from the initial development of breadth. However the contemporaneous autologous neutralization was not adsorbed by gp120.

Conclusion: Our results indicate that a novel cross-clade neutralizing epitope in gp120 is recognized by the antibodies in two individuals with BCN activity. The failure to adsorb the autologous neutralizing activity suggests that these BCN antibodies constitute a minor proportion of the autologous neutralization.

1 Chinese Center for Disease Control & Prevention, Beijing, China 2 Laboratory for AIDS Research & Development, Duke University Medical Center, Durham, North Carolina, USA P04.19 Genotypic and phenotypic characteristics of CRF01_AE HIV-1 virus env genes from early infected individuals

Background: CRF01_AE subtype HIV-1 strains are dominated in the newly infected man-to-man homosexually transmitted individuals in China. It is valuable for the design of vaccine antigens to analyze the genotype and phenotypes of the functional env genes from early/acutely infected individuals.

Methods: HIV-1 env amplicons were obtained by single genome amplification from 4 early infected individuals, and were subjected to sequencing and cloning. The single cycle pseudovirus infection assay was applied to analyze the autologous and heterologous neutralization response.

Results: All 4 individuals identified as in early/acute infection phase were infected with CRF01_AE strains. The sequences of env genes in each individual were highly homologous. Recombination was common in the env genes from 3 of 4 individuals. In 3 patients, the autologous neutralization responses were detected about 2-3 months postseroconversion. High autologous neutralization responses were found in one patient as early as 15 days postseroconversion, meanwhile, the viral load in this patient decrease rapidly as the neutralization responses increased in the following days. As the peak of autologous neutralization response appeared, the viral load was decrease to 140 copies/ml. Contemporary autologous virus neutralization was observed in this patient and persisted for months. All virus variants obtained at the first time point were sensitive to MAbs 2F5 and 4E10, resistant to b12, only those from the patient with very early neutralization response were sensitive to 2G12. The sites potentially associated with neutralization sensitivity were dispersed through domains of gp120, but focused in the CT region of gp41.

Conclusion: The env genes from early HIV-1 infected individuals are highly homologous and unstable. The early transmitted viruses tend to be sensitive to MAb neutralization. Autologous neutralization responses are common in early CRF01_AE HIV-1 virus infected individuals, but the neutralization responses are strains-specific.