Michael Palm Basic Science, Vaccines & Prevention Project Weblog

After the Fact: Merck’s Safety & Immunogenicity Data Published in CID

Data from phase I safety & immunogenicity studies of Merck’s adenovirus serotype 5 (Ad5) HIV vaccine has just been published online in Clinical Infectious Diseases. The data has been presented at various conferences but this is the first relatively comprehensive report to appear in the literature. The paper reinforces several of the observations that were made during the development of the vaccine: first and foremost, it represented a great leap forward in terms of the proportion of recipients that actually developed a detectable CD8 T cell response to vaccine antigens. Previous bests were around 20-30% of recipients (with an ALVAC canarypox vector). What seems to have been entirely lost in recent coverage of the Merck vaccine failure is that, before this Ad5 construct came along, no one had the luxury of worrying about the impact of issues like the magnitude, breadth, function & specificity of the CD8 T cell response because, for the most part, there was no CD8 T cell response there to analyze!

Secondly, although the issue is only mentioned in passing in the paper and would benefit from a more detailed presentation of the data, it is clear that the T cell responses induced by the vaccine were heavily biased toward CD8 T cell responses and the importance of a balanced CD4/CD8 T cell response is now better recognized. Merck’s decision to drop the DNA priming immunization – criticized at the time by Jerry Sadoff and Bette Korber during an NIAID AIDS Vaccine Research Working Group meeting – may have exacerbated this problem, as DNA vaccines typically bias toward CD4 T cell responses.

Thirdly, the tendency of for Ad5 vectors to persist and thus maintain an activated effector memory CD8 T cell population – which has been reported in small animal models over the past few years – may have been mirrored in humans: the figures in the paper showing the magnitude of HIV-specific CD8 T cell responses over time do not provide much evidence of a contraction phase (click on the image to see the figure), normally a prerequisite for the development of a “central memory” CD8 T cell population endowed with robust proliferative capacity.

Finally, the paper unfortunately offers little data on the breadth of the vaccine-induced responses and yet the authors suggest that it was “broad” – this seems inconsistent with the data from the STEP trial showing that recipients only responded to an average of one epitope from each vaccine-encoded protein.

Clinical Infectious Diseases 2008;46:000–000
DOI: 10.1086/587993

MAJOR ARTICLE

Safety and Immunogenicity of a Replication-Incompetent Adenovirus Type 5 HIV-1 Clade B gag/pol/nef Vaccine in Healthy Adults

Frances H. Priddy,1 Deborah Brown,2 James Kublin,3 Kathleen Monahan,2 David P. Wright,4 Jacob Lalezari,5 Steven Santiago,7 Michael Marmor,8 Michelle Lally,9 Richard M. Novak,10 Stephen J. Brown,6 Priya Kulkarni,2 Sheri A. Dubey,2 Lisa S. Kierstead,2 Danilo R. Casimiro,2 Robin Mogg,2 Mark J. DiNubile,2 John W. Shiver,2 Randi Y. Leavitt,2 Michael N. Robertson,2 Devan V. Mehrotra,2 and Erin Quirk,2 for the Merck V520-016 Study Groupa

1Department of Medicine, Emory University School of Medicine, Atlanta, Georgia; 2Merck Research Laboratories, West Point, Pennsylvania; 3Fred Hutchinson Cancer Research Center, Seattle, Washington; 4Central Texas Clinical Research Center, Austin; 5Quest Clinical Research Laboratories, San Francisco, and 6AIDS Research Alliance, West Hollywood, California; 7Care Resource, Miami, Florida; 8New York University School of Medicine, New York; 9Warren Alpert Medical School of Brown University, Providence, Rhode Island; and 10University of Illinois at Chicago, Chicago

Background. The safety and immunogenicity of the MRK adenovirus type 5 human immunodeficiency virus type 1 clade B gag/pol/nef vaccine, a replication-incompetent adenovirus type 5–vectored vaccine designed to elicit cell-mediated immunity against conserved human immunodeficiency virus proteins, was assessed in a phase 1 trial.

Methods. Healthy adults not infected with human immunodeficiency virus were enrolled in a multicenter, dose-escalating, blind, placebo-controlled study to evaluate a 3-dose homologous prime-boost regimen of the trivalent MRK adenovirus type 5 human immunodeficiency virus type 1 vaccine containing from to viral particles per 1-mL dose administered on day 1, during week 4 and during week 26. Adverse events were recorded for 29 days after each intradeltoid injection. The primary immunogenicity end point was the proportion of study participants with a positive unfractionated Gag-, Pol-, or Nef-specific interferon-γ enzyme-linked immunosorbent spot response measured 4 weeks after administration of the last dose.

Results. Of 259 randomized individuals, 257 (99%) received 1 dose of vaccine or placebo and were included in the safety analyses. Enzyme-linked immunosorbent spot results were available for 217 study participants (84%) at week 30. No serious vaccine-related adverse events occurred. No study participant discontinued participation because of vaccine-related adverse events. The frequency of injection-site reactions was dose dependent. Vaccine doses of viral particles elicited positive enzyme-linked immunosorbent spot responses to 1 vaccine component in >60% of recipients. High baseline antibody titers against adenovirus type 5 diminished enzyme-linked immunosorbent spot responses at all doses except the viral particle dose.

Conclusions. The vaccine was generally well tolerated and induced cell-mediated immune responses against human immunodeficiency virus type 1 peptides in most healthy adults. Despite these findings, vaccination in a proof-of-concept trial with use of this vaccine was discontinued because of lack of efficacy.

Mannose Y Mannose – Two Attempts to Mimic the Anti-HIV Activity of the 2G12 Antibody

Two new papers in the online first section of J. Virology describe attempts to generate anti-HIV antibodies that can mimic the broad neutralizing activity of the monoclonal antibody 2G12. This antibody is unusual in that it targets the clusters of high-mannose glycans that typically help shield HIV’s envelope from antibody-mediated attack (a glycan is a large carbohydrate molecule composed of many smaller sugar molecules linked together).

Rena Astronomo and colleagues utilized bovine serum albumin (BSA) as the building block for an immunogen designed to induce anti-mannose antibodies that mimic 2G12. Unfortunately the approach proved unsuccessful: immunization of rabbits induced antibodies that failed to even bind HIV’s gp120 protein. The researchers suspect that the molecules of the immunogen were spaced too widely apart and failed to mimic the dense clustering of glycans on gp120; strategies that may overcome the problem – including the use of non-infectious, virus-like particles – are now being explored.

In the second paper, a team headed by Robert Luallen took a different tack. They developed a genetically engineered yeast (Saccharomyces cerevisiae) that encodes proteins containing high density glycan clusters and used heat-killed preparations of the yeast to immunize rabbits. The approach successfully induced antibodies capable of binding gp120 from a clade B and clade C primary isolate as well as SIV 239 and 1A11 and recognition increased after several immunizations. However, when the concentrations of antibodies in the ELISA assay were taken into account, the binding activity was far lower (50-100 fold) than that seen for 2G12 against clade B g120s. The researchers suggest that the lower activity relates to the fact that their approach induced a polyclonal antibody response in which the antibodies mimicking 2G12 are a minority population. Consistent with this possibility, the authors note that preliminary results from neutralization assays have mostly been negative.

Despite the limitations, these data may still represent an important first step toward the design of more effective antibody-based immunogens, as the elicitation of 2G12-like antibodies capable of recognizing HIV and SIV Envs has (to the best of my knowledge) not previously been reported. In discussing their findings, Luallen and colleagues outline a number of strategies they are pursuing in order to induce the closest 2G12 mimics possible, and suggest that the yeast platform is particularly well suited to the task.

JVI Accepts, published online ahead of print on 23 April 20
J. Virol. doi:10.1128/JVI.00412-08

An engineered Saccharomyces cerevisiae strain binds the broadly neutralizing HIV-1 antibody 2G12 and elicits mannose-specific, gp120 binding antibodies

Robert J. Luallen, Jianqiao Lin, Hu Fu, Karen K. Cai, Caroline Agrawal, Innocent Mboudjeka, Fang-Hua Lee, David Montefiori, David F. Smith, Robert W. Doms, and Yu Geng

ProSci Incorporated, 12170 Flint Place, Poway, CA 92064; Department of Microbiology, University of Pennsylvania, 225 Johnson Pavilion, Philadelphia, PA 19104; Department of Surgery, Duke University Medical Center, Box 2926, DUMC, Durham, NC 27710; Department of Biochemistry, Emory University School of Medicine, 1510 Clifton Rd NE, Room 4035, Atlanta, GA 30322

Abstract

The glycan shield of the HIV-1 envelope (Env) protein serves as a barrier to antibody-mediated neutralization and plays a critical role in transmission and infection. One of the few broadly neutralizing HIV-1 antibodies, 2G12, binds to a carbohydrate epitope consisting of an array of high mannose glycans exposed on the surface of the gp120 subunit of the Env protein. To produce proteins with exclusively high mannose carbohydrates, we generated a mutant strain of Saccharomyces cerevisiae by deleting three genes in the N-glycosylation pathway, Och1, Mnn1, and Mnn4. Glycan profiling revealed that N-glycans produced by this mutant were almost exclusively Man8GlcNAc2, and four endogenous glycoproteins were identified that were efficiently recognized by 2G12. These yeast proteins, like HIV-1 gp120, contain a large number and high density of N-linked glycans, with glycosidase digestion abrogating 2G12 cross-reactivity. Immunization of rabbits with whole och1mnn1mnn4 yeast produced sera that recognized a broad range of HIV-1 and SIV Env glycoproteins, despite no HIV/SIV-related proteins being used in the immunization procedure. Analyses of one of these sera on a glycan array showed strong binding to glycans with terminal Man1,2Man residues, and binding to gp120 was abrogated by glycosidase removal of high mannose glycans and terminal Man1,2Man residues, similar to 2G12. The fact that yeast are genetically pliable, and can be grown easily and inexpensively, will make it possible to produce new immunogens that recapitulate the 2G12 epitope, and may make the glycan shield of HIV Env a practical target for vaccine development.

JVI Accepts, published online ahead of print on 23 April 2008
J. Virol. doi:10.1128/JVI.00293-08

A glycoconjugate antigen, based on the recognition motif of a broadly-neutralising HIV antibody, 2G12, is immunogenic but elicits antibodies unable to bind to the ‘self’ glycans of gp120

Rena D. Astronomo, Hing-Ken Lee, Christopher N. Scanlan, Ralph Pantophlet, Cheng-Yuan Huang, Ian A. Wilson, Ola Blixt, Raymond A. Dwek, Chi-Huey Wong, and Dennis R. Burton

Departments of Immunology and Microbial Science, Molecular Biology, The Skaggs Institute for Chemical Biology, Chemistry, and the Glycan Array Synthesis Core D, Consortium for Functional Glycomics, The Scripps Research Institute, La Jolla, California 92037; The Glycobiology Institute, Department of Biochemistry, University of Oxford, Oxford OX1 3QU

Abstract

The glycan shield of HIV-1 gp120 contributes to viral evasion from humoral immune responses. However, the shield is recognized by the HIV-1 broadly-neutralizing antibody, 2G12, at a relatively conserved cluster of oligomannose glycans. The discovery of 2G12 raises the possibility that a carbohydrate immunogen may be developed that could elicit 2G12-like neutralizing antibodies (Abs) and contribute to an AIDS vaccine. We have previously dissected the fine specificity of 2G12 and reported that the synthetic tetramannoside (Man4) that corresponds to the D1 arm of Man9GlcNAc2 inhibits 2G12 binding to gp120 as efficiently as Man9GlcNAc2 itself, indicating the potential use of Man4 as a building block for creating immunogens. Here, we describe the development of neoglycoconjugates displaying variable copy numbers of Man4 on bovine serum albumin (BSA) molecules by conjugation to Lys residues. The increased valency enhances the apparent affinity of 2G12 for Man4 up to a limit which is achieved at 10 copies per BSA, beyond which no further enhancement is observed. Immunization of rabbits with BSA-(Man4)14 elicits significant serum Ab titres to Man4. However, these Abs are unable to bind gp120. Further analysis reveals that the elicited Abs bind a variety of unbranched and, to a lesser extent, branched Man9 derivatives, but not natural N-linked oligomannose containing the chitobiose core. These results suggest that Abs can be readily elicited against the D1 arm; however, potential differences in the presentation of Man4 on neoglycoconjugates, as compared to glycoproteins, poses challenges for eliciting anti-mannose Abs capable of cross-reacting with gp120 and HIV-1.

Transmission of CD8 T Cell Escape Mutants is Associated with Lower Viral Loads in Newly Infected Individuals

A new study by Paul Goepfert and colleagues offers compelling evidence that CD8 T cell responses can pressure HIV into mutating in ways that compromise viral fitness. CD8 T cells target tiny slices of viral proteins called epitopes, which are displayed by infected cells as a sort of alarm signal; CD8 T cells that recognize a particular epitope (via a docking bay-type structure called a T cell receptor or TCR) can mediate destruction of an infected cell by releasing cell-destroying substances such as perforin and granzyme B. HIV mutations that impact the structure of an epitope can abrogate CD8 T cell recognition and this phenomenon is called immune escape (it is loosely analogous to the way mutations can allow the virus to resist the effects of antiretroviral drugs). If a particular epitope-specific CD8 T cell response is effective, viruses with escape mutations in that epitope are at a selective advantage because they can persist despite the presence of the immune response. However, certain parts of HIV can tolerate mutations more easily than others, and studies have shown that the Gag protein is so vital to replication that mutations affecting epitopes in Gag can impair the ability of the virus to replicate in a lab dish (in vitro).

To assess whether these observations are relevant in people, Goepfert et al analyzed data from 114 epidemiologically linked transmission pairs in Zambia (the individuals had been participants in a larger cohort of “discordant” couples in which one partner was HIV-infected; despite counseling and increased condom use in the cohort, transmission still occurred at a rate of approximately 8% per year). The researchers looked for evidence of CD8 T cell escape mutations in Gag and Nef and then analyzed whether the transmission of HIV containing mutations impacted viral load in the newly infected individual. The timepoint for the analysis was 6 months after infection, because 579 of the 610 mutations documented in the transmitting partners were still present in recipients at this time point. The results showed that higher numbers of CD8 T cell escape mutations in the Gag protein of transmitted viruses were significantly associated with lower viral loads in the newly infected individuals (no such effect was seen for Nef). Further analyses revealed that the effect was most consistent for Gag epitopes targeted by HLA-B-restricted CD8 T cells (HLA genes manufacture the CD8 T cell TCR and thus govern the epitope structures that a CD8 T cell can recognize). The researchers hypothesized that the impact on HIV replication of mutations in epitopes targeted by HLA-B-restricted CD8 T cells would be most prominent in individuals lacking the same HLA-B genes themselves, and indeed this turned out to be the case: when the analysis was restricted to only these study participants the associations between more mutations in Gag and lower viral load became much stronger (p=0.0003).

In discussing their results, the study authors note that there was a ~10-fold difference in viral loads when individuals infected with viruses containing less than two escape mutations in Gag were compared to those with more than six such mutations, suggesting that infection with these multiple escape mutants may slow disease progression (although longer term follow up of a subset of participants indicates that viral load is, as is typical, increasing over time). They also state that: “these data imply that for CTL-based HIV vaccines to effectively control viral load, they must simultaneously target multiple Gag epitopes, thereby ensuring that fitness constraints prevent the virus from easily mutating.” In an accompanying news brief, JEM editor Hema Bashyam speculates that the study “might explain why a T cell vaccine that induces immune responses against two Gag epitopes failed in a recent trial.” In fact, recipients of the Merck HIV vaccine developed responses to an average of just one Gag epitope.

The Journal of Experimental Medicine
Published online 21 April 2008
doi:10.1084/jem.20072457

BRIEF DEFINITIVE REPORT

Transmission of HIV-1 Gag immune escape mutations is associated with reduced viral load in linked recipients

Paul A. Goepfert1,2, Wendy Lumm4, Paul Farmer4, Philippa Matthews5, Andrew Prendergast5, Jonathan M. Carlson6,7, Cynthia A. Derdeyn4,8, Jianming Tang1,2, Richard A. Kaslow3, Anju Bansal1, Karina Yusim10, David Heckerman6, Joseph Mulenga11, Susan Allen9, Philip J.R. Goulder5,12,13, and Eric Hunter4,8

1 Department of Medicine, 2 Department of Microbiology, and 3 Department of Epidemiology, University of Alabama at Birmingham, Birmingham, AL 35294
4 Emory Vaccine Center at Yerkes National Primate Research Center, Atlanta, GA 30322
5 Department of Pediatrics, The Peter Medawar Building for Pathogen Research, University of Oxford, Oxford OX1 3SY, England, UK
6 Microsoft Research, Redmond, WA 98052
7 Department of Computer Science and Engineering, University of Washington, Seattle, WA 98195
8 Department of Pathology and Laboratory Medicine and 9 Department of Global Health, Emory University, Atlanta, GA 30322
10 Los Alamos National Laboratory, Los Alamos, NM 87545
11 Zambia-Emory HIV Research Group, Lusaka, Zambia
12 HIV Pathogenesis Program, Doris Duke Medical Research Institute, University of KwaZulu-Natal, Durban 4013, South Africa
13 Partners AIDS Research Center, Massachusetts General Hospital, Charlestown, MA 02129

CORRESPONDENCE Paul A. Goepfert: paulg@uab.edu

In a study of 114 epidemiologically linked Zambian transmission pairs, we evaluated the impact of human leukocyte antigen class I (HLA-I)–associated amino acid polymorphisms, presumed to reflect cytotoxic T lymphocyte (CTL) escape in Gag and Nef of the virus transmitted from the chronically infected donor, on the plasma viral load (VL) in matched recipients 6 mo after infection. CTL escape mutations in Gag and Nef were seen in the donors, which were subsequently transmitted to recipients, largely unchanged soon after infection. We observed a significant correlation between the number of Gag escape mutations targeted by specific HLA-B allele–restricted CTLs and reduced VLs in the recipients. This negative correlation was most evident in newly infected individuals, whose HLA alleles were unable to effectively target Gag and select for CTL escape mutations in this gene. Nef mutations in the donor had no impact on VL in the recipient. Thus, broad Gag-specific CTL responses capable of driving virus escape in the donor may be of clinical benefit to both the donor and recipient. In addition to their direct implications for HIV-1 vaccine design, these data suggest that CTL-induced viral polymorphisms and their associated in vivo viral fitness costs could have a significant impact on HIV-1 pathogenesis.

HLA Class II Associations with Resistance & Susceptibility to HIV Infection

Following a similar theme to a recently posted study indicating that certain class II HLA genes – which impact antigen presentation to CD4 T cells – are associated with elite control of HIV replication in infected individuals, a new paper just published in the journal AIDS adds to the literature showing significant associations between class II HLA genes and resistance/susceptibility to HIV infection. The data comes from a large cohort study of sex workers in the Pumwani district of Nairobi, approximately 10% of whom persistently resist HIV infection despite an estimated ~60 unprotected exposures per year.

AIDS. 22(7):807-816, April 23, 2008.

BASIC SCIENCE

Human leukocyte antigen-DQ alleles and haplotypes and their associations with resistance and susceptibility to HIV-1 infection.

Hardie, Rae-Anne; Luo, Ma; Bruneau, Brigitte; Knight, Erin; Nagelkerke, Nico JD; Kimani, Joshua; Wachihi, Charles; Ngugi, Elizabeth N; Plummer, Francis A

Abstract:

Objectives: To determine the association of DQ antigens with resistance and susceptibility to HIV-1.

Design: Despite repeated exposure to HIV-1, a subset of women in the Pumwani Sex Worker cohort established in Nairobi, Kenya in 1985 have remained HIV-1 negative for at least 3 years and are classified as resistant. Differential susceptibility to HIV-1 infection is associated with HIV-1 specific CD4+ and CD8+ T cell responses. As human leukocyte antigen-DQ antigens present viral peptides to CD4+ cells, we genotyped human leukocyte antigen -DQ alleles for 978 women enrolled in the cohort and performed cross-sectional and longitudinal analyses to identify associations of human leukocyte antigen -DQ with resistance/susceptibility to HIV-1.

Methods: DQA1 and DQB1 were genotyped using taxonomy-based sequence analysis. SPSS 13.0 was used to determine associations of DQ alleles/haplotypes with HIV-1 resistance, susceptibility, and seroconversion rates.

Results: Several DQB1 alleles and DQ haplotypes were associated with resistance to HIV-1 infection. These included DQB1*050301 (P = 0.055, Odds Ratio = 12.77, 95% Confidence Interval = 1.44-112), DQB1*0603 and DQB1*0609 (P = 0.037, Odds Ratio = 3.25, 95% Confidence Interval = 1.12-9.47), and DQA1*010201-DQB1*0603 (P = 0.044, Odds Ratio = 17.33, 95% Confidence Interval = 1.79-168). Conversely, DQB1*0602 (P = 0.048, Odds Ratio = 0.68, 95% Confidence Interval = 0.44-1.05) and DQA1*010201-DQB1*0602 (P = 0.039, Odds Ratio = 0.64, 95% Confidence Interval = 0.41-1.03) were overrepresented in the HIV-1 infected population. DQA1*0504-DQB1*0201, DQA1*010201-DQB1*0201, DQA1*0402-DQB1*0402 and DQA1*0402-DQB1*030101 genotypes were only found in HIV-1 positive subjects (Odds Ratio = 0.30-0.31, 95% Confidence Interval = 0.03-3.70), and these women seroconverted rapidly. The associations of these DQ alleles and haplotypes with resistance and susceptibility to HIV-1 were independent of the previously reported human leukocyte antigen-DRB*01, human leukocyte antigen A2/6802, and human leukocyte antigen-A*2301.

Conclusion: The associations of DQ alleles and haplotypes with resistance and susceptibility to HIV-1 emphasize the importance of human leukocyte antigen-DQ and CD4 in anti-HIV-1 immunity.

Targeting DNA Vaccines to Dendritic Cells

A new study from Ralph Steinmans's laboratory indicates that targeting DNA vaccines to dendritic cells can substantially boost their immunogenicity. The approach involves fusing the antigen of interest (in this study, HIV-1 Gag) with a single chain antibody targeting the dendritic cell receptor DEC205 (also called CD205). Combined with delivery via electroporation, the researchers report that the approach boosted vaccine potency 100-fold in a mouse challenge system. Steinman recently won a Lasker award for his work on dendritic cells; writing about his work in Nature Medicine, he noted sadly that most of the research he conducted (and on which the Lasker award was based) would probably not have been funded if the NIH grant payline had been at the low level it is today:

"The vision of Mary Lasker was to urge the government to sponsor basic research to understand and treat disease. Her vision remains critical to meet ongoing challenges to the scientific enterprise. One challenge is our declining ability to support independent scientists. I probably could not have carried out my research under the current funding environment for basic research, because many of my grant applications only ranked in the top quartile, whereas funding by the US National Institutes of Health is currently only sufficient for less than 10% of basic proposals."

In another recent paper, Jinyan Liu and colleagues show that electroporation alone results in a 77-fold enhancement in the recruitment of dendritic cells to the site of DNA vaccine inoculation, likely accounting for the significant enhancements in immunogenicity that have been reported using this delivery method. The question mark hanging over viral vectors due to the Merck HIV vaccine results is likely to refocus attention on these and other strategies for improving DNA vaccine immunogenicity.

J Clin Invest. 2008 Apr 1;118(4):1427-1436. (free full text access)

The efficacy of DNA vaccination is enhanced in mice by targeting the encoded protein to dendritic cells.

Nchinda G, Kuroiwa J, Oks M, Trumpfheller C, Park CG, Huang Y, Hannaman D, Schlesinger SJ, Mizenina O, Nussenzweig MC, Uberla K, Steinman RM.

Laboratory of Cellular Physiology and Immunology, The Rockefeller University, New York, New York, USA. Department of Molecular and Medical Virology, Ruhr-University Bochum, Bochum, Germany. The Aaron Diamond AIDS Research Center, New York, New York, USA. Ichor Medical Systems, San Diego, California, USA. Laboratory of Molecular Immunology, Chris Browne Center for Immunology, The Rockefeller University, New York, New York, USA.

DNA vaccines promote an immune response by providing antigen-encoding DNA to the recipient, but the efficacy of such vaccines needs improving. Many approaches have considerable potential but currently induce relatively weak immune responses despite multiple high doses of DNA vaccine. Here, we asked whether targeting vaccine antigens to DCs would increase the immunity and protection that result from DNA vaccines. To determine this, we generated a DNA vaccine encoding a fusion protein comprised of the vaccine antigen and a single-chain Fv antibody (scFv) specific for the DC-restricted antigen-uptake receptor DEC205. Following vaccination of mice, the vaccine antigen was expressed selectively by DCs, which were required for the increased efficacy of MHC class I and MHC class II antigen presentation relative to a control scFv DNA vaccine. In addition, a DNA vaccine encoding an HIV gag p41-scFv DEC205 fusion protein induced 10-fold higher antibody levels and increased numbers of IFN-gamma-producing CD4(+) and CD8(+) T cells. After a single i.m. injection of the DNA vaccine encoding an HIV gag p41-scFv DEC205 fusion protein, mice were protected from an airway challenge with a recombinant vaccinia virus expressing the HIV gag p41, even with 1% of the dose of nontargeted DNA vaccine. The efficacy of DNA vaccines therefore may be enhanced by inclusion of sequences such as single-chain antibodies to target the antigen to DCs.

J Virol. 2008 Mar 19 [Epub ahead of print]

Recruitment of Antigen-Presenting Cells to the Site of Inoculation and Augmentation of HIV-1 DNA Vaccine Immunogenicity by In Vivo Electroporation.

Liu J, Kjeken R, Mathiesen I, Barouch DH.

Division of Viral Pathogenesis, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02215; Inovio Biomedical, San Diego, CA 92121.

In vivo electroporation (EP) has been shown to augment the immunogenicity of plasmid DNA vaccines, but its mechanism of action has not been fully characterized. In this study, we show that in vivo EP augmented cellular and humoral immune responses to an HIV-1 Env DNA vaccine in mice and allowed a 10-fold reduction in vaccine dose. This enhancement was durable for over 6 months, and re-exposure to antigen resulted in anamnestic effector and central memory CD8+ T lymphocyte responses. Interestingly, in vivo EP also recruited large mixed cellular inflammatory infiltrates to the site of inoculation. These infiltrates contained 45-fold increased numbers of macrophages and 77-fold increased numbers of dendritic cells as well as 2- to 6-fold increased numbers of B and T lymphocytes as compared with infiltrates following DNA vaccination alone. These data suggest that recruiting inflammatory cells including APCs to the site of antigen production substantially improves the immunogenicity of DNA vaccines. Combining in vivo EP with plasmid chemokine adjuvants that similarly recruited antigen-presenting cells (APCs) to the injection site, however, did not result in synergy.

NIAID Issues "Requests for Information" on Vaccine Research

The National Institute of Allergy and Infectious Diseases (NIAID) has released two requests for information, one to solicit comments to inform the "Development of a Funding Opportunity Announcement on Highly Innovative Tactics to Interrupt Transmission of HIV" and the other to "Solicit Input and Ideas on Priorities in Basic Vaccine Discovery Research."

Hints of Correlations Between Immune Responses and Viral Load Emerge from the Merck HIV Vaccine Trial

The Merck HIV vaccine efficacy trial was conducted in collaboration with the National Institutes of Health-sponsored HIV Vaccine Trials Network (HVTN). At Keystone, the head of the HVTN, Larry Corey, provided an update on the ongoing analyses of data from the study and offered his thoughts on the implications for future research. Most intriguingly, Corey reported that some hints of correlations between vaccine-induced CD8 T cells and viral load setpoints have emerged, suggesting that the magnitude of the response may be important. Specifically, among the participants with anti-Ad5 antibody titers of less than 1:18 units, the magnitude of CD8 T cell responses to Gag (as measured by interferon gamma ELISpot four weeks after the second immunization) showed an inverse correlation with viral load setpoint (p=0.01). Responses to Pol displayed a similar trend (p=0.07) while those targeted against Nef did not (p=0.22). It must be emphasized that Corey cautioned that the numbers are small and the significance is fragile (additional data points may alter the findings), but the data is consistent with prior SIV challenge studies demonstrating inverse correlations between the magnitude of pre-challenge SIV-specific T cell responses and post-challenge viral load levels. However, no such correlation was seen among individuals with anti-Ad5 titers over 1:18 and exactly how pre-existing immunity would impact the relationship between CD8 T cell responses and viral load is unclear. Additional analyses of relationships between pre-infection HIV-specific CD4 T cell responses and post-infection viral load setpoints might shed light on this question, as one speculative possibility is that individuals with pre-existing immunity developed weaker HIV-specific CD4 T cell responses due to interference from vector-specific memory CD4 T cells.

In addition to the issue of magnitude, Corey also pointed out two other potentially significant limitations of the T cell responses induced by the Merck vaccine that could conceivably have contributed to its failure: breadth and functionality. Biometric data indicates that HIV’s Gag protein contains ~150 potential T cell epitopes that should be present in a substantial proportion of circulating clade B HIVs (>15%), but STEP participants only developed a CD8 T cell response to an average of one epitope from the full-length Gag protein encoded by the vaccine (similarly, participants only responded to one epitope in each of the other two vaccine-encoded proteins, Pol & Nef). Corey cited several strategies that are being pursued to increase breadth, including the use of genes from multiple different HIV variants in prime-boost vaccines (heterologous inserts), protein or peptide boosting and prime-boost with differing vaccine vectors (heterologous vectors). In terms of functionality, Corey offered his own “revisionist bit of history,” pointing out that it is now understood that Ad5 vectors given alone bias the CD8 T cell response toward a terminally differentiated phenotype with poor proliferative capacity. In support of this argument, he showed data comparing the phenotype of CD8 T cells induced by priming and boosting with Ad5 alone compared to DNA/Ad5 and also cited unpublished data from Rafi Ahmed’s lab showing that an Ad5 vector offered poor protection against LCMV due to the induction of T cells with limited proliferative capacity. It is currently unclear if this problem is also seen with other adenovirus serotypes (some recent macaque data indicates that there is some serotype-specific variability in the phenotype of vaccine-induced T cell responses).

Corey also described a suite of studies the HVTN is either conducting or planning to conduct in order to better understand the outcomes of both the STEP and Phambili Merck vaccine trials, including looking at the interaction between Ad5 immunization and pre-existing Ad5-specific immune responses at mucosal sites, sequencing the viruses in individuals that became infected, and analyzing the impact of HLA types on both susceptibility to HIV infection and the immune responses to the vaccine.

Looking at the bigger picture, Corey echoed recent statements about the importance of discovery research for developing neutralizing antibody-based or other approaches that may lead to complete protection against infection. But he also stressed that the evidence still strongly supports pursuing the nearer-term goal of trying to increase the chance that an individual who becomes infected will control the virus immunologically, thus reducing the risk of onward transmission and preventing or slowing progression to AIDS.

Keystone Symposia: HIV Vaccines: Progress and Prospects
Banff, Alberta, Canada, March 27-April 1, 2008

Abstract #002

An Overview of the SteP trial

L. Corey
MRK HVTN Study Team, University of Washington, Seattle, WA

Background: A collaborative study between Merck, the HVTN, and NIH was initiated in 2004 to perform a randomized, multi-center, double-blind, placebo-controlled test-of-concept study in 3000 HIV seronegative volunteers at high risk of acquiring HIV. Volunteers were randomized (1:1) to receive 3 injections of either the MRKAd5 HIV-1 gag/pol/nef vaccine (a replication defective Ad5 vector) or placebo. Volunteers were tested ~every 6 months for acquisition of HIV, in volunteers who became HIV infected, viral load and CD4 cell counts were measured at multiple time points post diagnosis. Results: The study was stopped in September 2007 at the first planned interim analysis; 82 endpoints were noted in men and only 1 in women. Among male volunteers with low (<200) baseline Ad5 neutralizing antibody titers (the primary analysis population), there were 24 infections among the 741 vaccinees compared to 21 infections among the 762 placebo recipients (using a modified “intent-to-treat” approach), and among the volunteers who became HIV infected, the geometric mean plasma vRNA level was ~40,000 copies/mL in the vaccine groups compared to ~26,000 copies/mL in the placebo group. Vaccine and placebo recipients were well-matched in demographic and risk characteristics at baseline within Ad5 strata. In ongoing multivariate analyses, the following factors appear to be more strongly associated with HIV-1 acquisition; treatment (vaccine vs. placebo), hazard ratio (HR) 1.9, 95% CI (1.1, 3.2); region (North America vs. other), HR 3.2 (1.5, 6.7); and unprotected receptive anal sex (yes vs. no), HR 4.4 (2.4, 8.7). With interaction terms in the model, the treatment HR (vaccine vs. placebo) was 3.1 (1.5, 6.5) in Ad5 seropositives vs. HR 1.0 (0.5, 2.0) in Ad5 seronegative; the treatment HR was 4.5 (1.8, 11.4) in uncircumcised men vs. 1.0 (0.6, 1.8) in circumcised men. A 25% sample of all recipients and all persons who subsequently acquired HIV was performed at week 8 after 2 doses of vaccine. The frequency of CD8+ T cell responses to HIV genes (76%) was found in persons who subsequently acquired HIV-1; similar to that seen in persons without HIV acquisition. Similarly the GMT of responses was also similar in vaccinees who acquired and did not acquire HIV. Conclusions: There was no evidence thatvaccination prevented infection or lowered viral setpoint; in fact, there were more infections in the vaccine group. In multivariate analysis of baseline risk factors the Merck Ad5 trivalent vaccine appeared to be associated with an increased risk of HIV-1 acquisition in men with pre-existing Ad5 immunity and in uncircumcised men. The mechanism for potential increased acquisition will be discussed.

In Vitro Study of Adenovirus Vector Effects on HIV Replication in CD4 T Cells

Another abstract from the recent Keystone meeting (see below). The data suggests that, as some researchers had hypothesized, Ad5-specific memory CD4 T cells may be rendered particularly susceptible to HIV after Ad5 stimulation. Further research will be needed to try and ascertain if these observations are relevant to the evidence of enhanced susceptibility to HIV acquisition that emerged from the STEP trial. Notably, there was no indication that Ad5-specific immunity enhanced post-infection HIV viral load among STEP participants (people with pre-existing immunity to Ad5 actually had slightly lower viral load setpoints than Ad5 seronegative participants).

In light of the concerns about adenovirus vectors raised by the STEP data, it seems a little odd that a week after the NIAID HIV Vaccine Summit, GenVec announced receipt of a $600,000 NIAID grant to develop their Ad5-based malaria vaccine candidate and Crucell – another company that has invested heavily in the adenovirus vector vaccine platform – announced the initiation of a phase I trial of an Ad26-based HIV vaccine candidate. Ad26 is far less prevalent in nature than Ad5, so few people have antibody responses to the Ad26 hexon protein (the major target for antibody responses to adenoviruses). However, studies have clearly shown that Ad-specific memory CD4 T cell responses are highly cross-reactive against different serotypes, which may call into serious question the safety of any adenovirus-based vaccine candidate in individuals at risk for HIV infection (this may even hold true for vectors based on chimpanzee adenoviruses, see second abstract below). As the data stands currently, if GenVec successfully developed an Ad5-based malaria vaccine, it would be contraindicated in anyone at risk for HIV infection with pre-existing immunity to the vector; this would represent a substantial portion of the population that stands to benefit from immunization against malaria. And for the Crucell construct, it is also difficult to imagine a scenario where the vaccine could be ethically studied in individuals at risk for HIV infection, unless it can be proven that the risk of enhancement seen in STEP related entirely to immune responses specific to Ad5 and not immune responses cross-reactive with other adenovirus serotypes (or some other immune-modulating effect of adenoviruses generally). The comments of Crucell’s Jaap Goudsmit in the company’s press release seem not only premature, but bordering on delusional: “We are excited about the first in man study of this newly developed vector, that could provide a solution to the issues that raised from previous HIV vaccine trials.”

Keystone Symposia: HIV Vaccines: Progress and Prospects
Banff, Alberta, Canada, March 27-April 1, 2008

Abstract# 328

HIV-1 replication in CD4 T cells from Adenovirus experienced and Naïve Donors

A. Gregory Spies, Jason Stucky, John McNevin and Juliana McElrath.
Vaccine and Infectious Disease Research Institute, Fred Hutchinson Cancer Research Center, Seattle, Washington 98105.

Preliminary analysis of the Merck STEP Study suggests that there is a higher rate of HIV infection in vacinees who had positive human adenovirus serotype-5 titers on enrollment compared to those with undetectable Ad5 titers. Therefore, we have utilized a human in vitro T cell priming assay to determine whether prior infection by Ad5 affects replication of HIV-1 in human CD4 T cells following exposure to a replication-defective Ad5 vector. Immature dendritic cells were generated using GM-CSF/IL-4 treatment of CD14 positive cells. Immature DC were exposed to an empty Ad5 vector resulting in increased levels of HLA-Class I, HLA-DR, CCR7, CD40, CD80 and CD83. CD4 T cells were primed with mock or Ad5-stimulated DCs and subsequently infected with the CCR5-tropic HIV-1 strain, YU2. HIV-1 replication was monitored by p24 ELISA of supernatants from DC-primed CD4 T cell cultures. Exposure of DC to Ad5 resulted in increased HIV growth in CD4 T cells from both Ad5-naive and Ad-experienced donors. At day 6, following addition of T cells, the Ad5-experienced CD4 T cells were found to produce twice the p24 as the Ad5-naive CD4 T cells. However FACS analysis of these cells using CD38, Ki67, Bcl-2, CCR5, CCR7 and CD-27 showed no phenotypic difference.

Abstract #223

T-cell responses in healthy adults to replication deficient AdHu5 vector

Natalie Hutnick, Hildegund Ertl, Michael R. Betts.
Microbiology Dept, University of Pennsylvania, Philadelphia PA, USA, 19104

A current strategy for an HIV vaccine employs the use of a replication deficient AdHu5 vector. However, it has been shown that pre-existing antibodies against AdHu5 reduce vaccine efficacy. To avoid the problems associated with pre-existing AdHu5 humoral immunity, vectors based on 2 serologically distinct chimpanzee adenoviruses (AdC6 and AdC7) have been developed.. While this strategy may avoid issues with pre-existing humoral immunity, it remains possible that cross-reactive Ad-specific CD8+ T cells could similarly limit vaccine efficacy. Here, we examined the response to replication defective AdHu5, AdC6 and AdC7 vectors in 10 healthy subjects from North America not expected to have been exposed previously to chimp-derived Ad strains. Eight of the ten subjects exhibited a polyfunctional CD8 T cell response to AdHu5 as measured by IL-2, TNF-a, IFN-g and perforin expression. The responding cells were a mixed population composed of CD27- CD45RO+ effector memory and CD27+CD45RO+ central memory phenotype. The 8 subjects who responded to AdHu5 also exhibited a cross reactive response to AdC6, and three responded to AdC7, though at a lower frequency. We further examined Ad-specific T cell responses in one subject who participated in a phase I AdHu5 HIV vaccine safety trial. A substantial expansion of polyfunctional AdHu5-specific CD8+ (1.06%) and CD4+ (0.57%) T cells were noted approximately 4 months after the final inoculation. These cells also exhibited cross-reactivity with both AdC6 and AdC7. These data indicate that Adenovirus-specific T cell responses are common in humans, and that shared elements between disparate Ad strains can be serve as targets for Ad-specific T cells. Vaccination with replication defective AdHu5 vectors also appears to enhance the frequency of Ad specific T-cells.

HLA Class II Associations with Viral Load Control

Last week’s NIAID vaccine summit featured much handwringing pessimism regarding the potential for adaptive immunity to control HIV replication, leading Guiseppe Pantaleo to remind attendees about the strong and consistent association between the class I HLA allele B*57 and elite control/long term non progression. Class II HLA alleles, which influence CD4 as opposed to CD8 T cell responses, have been less well studied, although some associations with slowed disease progression and resistance to HIV infection have been reported. At the recent Keystone HIV pathogenesis conference in Banff, Rachel Owen from UCSF presented new data indicating that the class II HLA allele DRB1*13 is significantly overrepresented among individuals who maintain viral loads in the absence of therapy, suggesting that CD4 T cell responses contribute to the phenomenon. The findings echo a recently published report on class II alleles and elite control in SIV-infected macaques (see second abstract, below).

As a side note, the Keystone HIV pathogenesis conference occurs in parallel with an HIV vaccine meeting and has done so for many years. However, next year the titles of the two meetings will be changed to “Prevention of HIV/AIDS” and “HIV Immunobiology: From Infection to Immune Control.”

Keystone Symposia: HIV Pathogenesis (X8), March 27 - April 1, 2008

HLA Class II associations in HIV infection: Controllers versus Non-controllers

Rachel E. Owen1, 2, Elizabeth Sinclair3, C. Lorrie Epling3, Jeffrey N. Martin3, Steven G. Deeks3, Philip J. Norris1, 2,3.

1Blood Systems Research Institute, 270 Masonic Avenue, San Francisco, CA 94118, USA, 2Department of Laboratory Medicine and 3Department of Medicine, University of California San Francisco, San Francisco, CA 94143, USA.

Strong CD4+ and CD8+ T cell responses are detected in individuals who control viraemia, suggesting but not proving that these cells may be causally related to virus control. The consistent association between certain class I alleles (e.g., B5701) and virus control provides strong evidence that CD8 T cells are able to effectively exert control. However, the role of CD4+ T cells in controlling HIV infection is not fully understood, and few HLA class II gene associations have been made.

We investigated HLA class II gene associations in chronically infected HIV+ individuals from the UCSF SCOPE cohort and categorized individuals as: (1) “controllers”: defined as plasma HIV RNA levels <10,000 copies/ml (in absence of therapy, n= 38); and (2) “non-controllers”: defined as plasma HIV RNA levels >10,000 copies/ml (n=290).

We found an increased frequency of the HLA-DRB1*13 allele in the controllers compared to the non-controllers (42.1% vs. 24.1%, p= 0.029), consistent with previous studies. Some DRB1*13 haplotypes were more common in controllers than non-controllers (p= <0.05 for each pair-wise comparison): DRB1*13 plus DQB1*03 (24% vs. 10%), DQB1*04 (8% vs. 0.7%), DQB1*06 (37% vs. 18%), DRB3*03 (13% vs. 2%) and DRB4*01 (21% vs. 8%). Controllers had higher CD4+ T cell IFN and IL-2 responses than non-controllers, measured by intracellular cytokine staining, following gag stimulation (mean IFN response 0.7% vs. 0.3%; mean IL-2 response 0.3% vs. 0.1%). No difference in the CD4+ IFN or IL-2 response following env or pol stimulation was measured. A small number of DRB1*13 expressing controllers had a trend towards higher CD4+ IFN and IL-2 responses following gag stimulation, when compared to controllers not expressing the DRB1*13 allele or non-controllers.

The epidemiological link between class II alleles and virus control is consistent with observations made with class I alleles, and argues that antigen-specific CD4+ T cells are playing a role in control of viraemia. These responses would be important to study in the pathogenesis of acute HIV infection and in HIV vaccine trials.

This work was supported by the National Institutes of Health to the UCSF-GIVI Center for AIDS Research, 5P30AI027763, and to the UCSF Clinical and Translational Sciences Institute, U54RR023566.

J Virol. 2008 Jan;82(2):859-70. Epub 2007 Nov 7.

The major histocompatibility complex class II alleles Mamu-DRB1*1003 and -DRB1*0306 are enriched in a cohort of simian immunodeficiency virus-infected rhesus macaque elite controllers.

Giraldo-Vela JP, Rudersdorf R, Chung C, Qi Y, Wallace LT, Bimber B, Borchardt GJ, Fisk DL, Glidden CE, Loffredo JT, Piaskowski SM, Furlott JR, Morales-Martinez JP, Wilson NA, Rehrauer WM, Lifson JD, Carrington M, Watkins DI.

Department of Pathology and Laboratory Medicine, University of Wisconsin-Madison, 555 Science Dr., Madison, WI 53711, USA.

The role of CD4(+) T cells in the control of human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV) replication is not well understood. Even though strong HIV- and SIV-specific CD4(+) T-cell responses have been detected in individuals that control viral replication, major histocompatibility complex class II (MHC-II) molecules have not been definitively linked with slow disease progression. In a cohort of 196 SIVmac239-infected Indian rhesus macaques, a group of macaques controlled viral replication to less than 1,000 viral RNA copies/ml. These elite controllers (ECs) mounted a broad SIV-specific CD4(+) T-cell response. Here, we describe five macaque MHC-II alleles (Mamu-DRB*w606, -DRB*w2104, -DRB1*0306, -DRB1*1003, and -DPB1*06) that restricted six SIV-specific CD4(+) T-cell epitopes in ECs and report the first association between specific MHC-II alleles and elite control. Interestingly, the macaque MHC-II alleles, Mamu-DRB1*1003 and -DRB1*0306, were enriched in this EC group (P values of 0.02 and 0.05, respectively). Additionally, Mamu-B*17-positive SIV-infected rhesus macaques that also expressed these two MHC-II alleles had significantly lower viral loads than Mamu-B*17-positive animals that did not express Mamu-DRB1*1003 and -DRB1*0306 (P value of <0.0001). The study of MHC-II alleles in macaques that control viral replication could improve our understanding of the role of CD4(+) T cells in suppressing HIV/SIV replication and further our understanding of HIV vaccine design.

A Surreal, Deeply Troubling Week for AIDS Vaccine Research

At the prompting of several basic scientists, NIAID held a “vaccine summit” this past Tuesday. The event, which from my perspective had all the rationality and conviviality of a hastily-convened Soviet era show trial, was webcast for posterity (the archived version can be seen online). Essentially, frustration with the dismal, unacceptably low NIAID payline for investigator-initiated grants appears to have caused a number of basic researchers to see the failure of Merck’s HIV vaccine candidate as an appropriate latch on which to hang their argument that money should be directed away from human trials of other experimental HIV vaccine candidates and into basic research and discovery.

There is certainly logic to the argument, despite the clumsy and alienating manner in which it has been advanced. The STEP trial has at the least shown that T cell-mediated control of HIV replication will not easily be achieved, and there is a strong case for rigorously reviewing the current vaccine pipeline and trimming it down to only candidates that might be able to address the potential shortcomings of the Merck construct (particularly in terms of HIV-specific T cell breadth, proliferative capacity, functionality, CD4 help and - of course - vector safety). In the longer term, discovery will also be vitally important to develop neutralizing antibody-based candidates (or other novel approaches) with the potential to completely protect against HIV acquisition.

However, this type of discussion is not what primarily emerged from the summit. Instead, panelists such as virologist and summit co-chair Warner Greene used the opportunity as a bully pulpit to make outrageous and offensive claims about the state of the HIV vaccine field and the work of their colleagues. Even cellular immunologist Rafi Ahmed offered a bizarre revisionist history of his thinking on T cell-based candidates: he claimed he had “cringed” when hearing of Merck’s candidate - an odd assertion given that he is on the immunology advisory committee of the HVTN, which conducted the STEP study – and that vaccines that focus on one arm of the immune system are “doomed to failure.” When I interviewed him a little over a year ago he said he was "optimistic" about the potential for T cell-based approaches to reduce post-infection viral load: “I’m a big believer in memory T cells bringing viral load down because that’s what we keep seeing. That doesn’t mean it will hold up in HIV but at least in all the models we study if you have memory T cells - even without any antibody - the initial burst is less and it’s pulled down much faster. And then the HIV data suggests that if that happens the prognosis is better, so that’s where my optimism comes from." Ahmed also stated at the summit that “immunology 101” says antibodies are required for control of HIV replication, yet studies of elite controllers offer little evidence for this claim.

Predictably, the media coverage surrounding the event amplified these soundbites across the world. Researchers in places such as South Africa, where skepticism about AIDS research runs high and AIDS denial has been politically embraced, will be dealing with the fall-out for years to come. Even elsewhere and here in the US, there has been a growth of anti-research and anti-science movements which will feed off these false, widely-publicized statements that no progress has occurred in AIDS vaccine research.

The misinformation actually started on Friday March 21, when a David Brown article in the Washington Post offered a series of misrepresentations about the Merck vaccine in the guise of facts. The deck quote was from Bob Gallo who, in a truly woeful bout of self-serving hyperbole, compared the STEP trial results to the Space Shuttle Challenger Disaster. Brown went on to misleadingly suggest that the strong trend toward enhancement seen in a subset of the STEP study participants was somehow connected to the T cell-based vaccine strategy; he does so by saying that the vaccine “primed” people to be more susceptible, and then stating that immune activation – one of the factors that has been suggested as a potential explanation for the adverse events – has “never been” associated with increased susceptibility to HIV infection. This is simply false; studies have shown that several “immune–activating infections” (to use Brown’s words) such as CMV and HSV-2 are associated with an increased susceptibility to HIV acquisition. Peripheral blood cells from highly immune-activated individuals also show increased susceptibility to HIV infection. In addition, if the enhancement effect had been related to the HIV-specific T cell responses induced by the vaccine, the effect would have been worst for the participants with no pre-existing immunity to the Ad5 vector (who developed the highest magnitude HIV-specific responses), but the opposite was the case (at the vaccine summit, a CDC researcher noted that the multivariate analyses showed that these vaccinated individuals were actually at a slightly but non-significantly reduced risk of HIV infection compared to the other study participants, with a point estimate of 0.6-0.8). In the same piece, Brown also attempts to indict researchers for not testing the Merck vector in a low-dose SIV challenge model as if this was some kind of oversight, completely glossing over the many complexities and controversies that attend these models and not mentioning that David Watkins did in fact use the Merck vector in one of his repeated low dose SIV challenge experiments.

Brown’s follow-up article from the March 25 summit compounded and expanded upon the misinformation offered in the first, claiming that the event was "tantamount to an admission that almost no progress has been made in the search for an AIDS vaccine in the past 25 years." This must have come as an unwelcome surprise to the researchers and study participants who have been slogging away for years, slowly improving the scientific understanding of both the immunological and immunogenetic underpinnings of long-term non-progression and resistance to HIV infection in highly exposed seronegative individuals, not to mention those involved in trying to iteratively improve the immunogenicity of experimental vaccine constructs. There has certainly been waste, duplication, and a sunny over-simplification of the complexities and unknowns that still face the HIV vaccine field from some quarters but these problems are not best addressed by trashing the entire endeavor and everyone involved in it. The statements quoted in Brown’s piece from Warner Greene and James Hoxie are de facto accusations that researchers who followed the strong scientific evidence that T cells play a key role in controlling HIV replication and used that evidence to inform vaccine design have led the field in the wrong direction and brought us no closer to an AIDS vaccine; in addition to being horrifically hostile and unfair, these claims are simply untrue.

Amidst all the sturm and drang at the summit, few people seemed to pick up on the comments of immunologist Mark Connors. Over the years, Connors has studied HIV-specific immune responses and authored many skeptical articles questioning some of the correlations with control of HIV replication that have been reported in the literature. In 2002, however, he identified HIV-specific CD8 T cell proliferation as a potentially important correlate, a finding several other research groups have now confirmed. At the summit, he cited additional assays his lab is working on, and expressed confidence that robust and broadly applicable correlates of immunological control are within striking distance. Furthermore, he also cited the fact that several Merck vaccine recipients who became infected in the STEP trial and carry the favorable immune response gene HLA B*57 are controlling their viral loads to undetectable levels. Initially, this was assumed to be an HLA B*57 effect but Connors argued that the frequency of the occurrence is far higher than would be anticipated in the absence of immunization, suggesting that it reflects an interaction between the vaccine and the favorable HLA allele akin to that seen in Merck’s now notorious SIV challenge studies, in which only immunized macaques bearing the favorable Mamu A*01 allele showed significant viral load reductions. The implication is that while the Merck vaccine was far from optimal, it may have been able to enhance the HLA B*57 effect. If this finding actually holds up, it may offer additional validation of the SIV/macaque model and it would also strengthen the argument for developing improved immunogens with the potential to achieve this outcome in people lacking HLA B*57.

The AIDS Vaccine Advocacy Coalition and the National Association of People with AIDS both issued concerned statements about the summit and the media coverage it engendered. Exactly where things go from here remains to be seen.

CORRECTION: The first version of this posting incorrectly stated that the class I Mamu A*01 allele in macaques is the equivalent to HLA B*57 in humans; thanks to David Watkins for alerting me to the fact that this is not the case, and that the likely analog is Mamu B*17. Evidence also suggests that Mamu B*08 is equivalent to HLA B*27, another HLA allele associated with control of HIV replication in humans.