Skip Navigation


Skip Content Marketing
  • Share this:
  • submit to facebook
  • Tweet it
  • submit to reddit
  • submit to StumbleUpon
  • submit to Google +

AIDS Vaccine Research Subcommittee, September 15-16, 2009

Bethesda Marriott Hotel, Bethesda, MD

Meeting Summary

The AIDS Vaccine Research Subcommittee (AVRS) met in public session on September 15-16, 2009, at the Bethesda Marriott Hotel in Bethesda, MD.

AVRS members present: Eric Hunter (chair), James Bradac (executive secretary), Jay Berzofsky (ex officio), Deborah Birx, Larry Corey (ex officio), Kevin Fisher, Nancy Haigwood, Jeffrey Lifson, Nelson Michael (ex officio), Gary Nabel (ex officio), Douglas Nixon, Michel Nussenzweig, Louis Picker, Bali Pulendran, Nina Russell, Jerald Sadoff, George Shaw, Bruce Walker.

Other NIH personnel participating:

  • Carl Dieffenbach, Director, Division of AIDS (DAIDS), NIAID;
  • Peggy Johnston, Director, Vaccine Research Program (VRP), DAIDS, NIAID;
  • Michael Pensiero, Preclinical Research and Development Branch, VRP, DAIDS, NIAID;
  • Geetha Bansal, Vaccine Discovery Branch, VP, DAIDS, NIAID.


  • Dan Barouch, Harvard Medical School;
  • Dennis Burton, Scripps Research Institute;
  • Barton Haynes, Duke University School of Medicine;
  • Jerome Kim, Walter Reed Army Institute of Research;
  • Wayne Koff, International AIDS Vaccine Initiative;
  • Bette Korber, Los Alamos National Laboratory;
  • Norman Letvin, Harvard Medical School;
  • Gary Nabel, Vaccine Research Center, NIH;
  • David Watkins, University of Wisconsin-Madison;
  • Otto Yang, Geffen School of Medicine at UCLA.

Call to Order

Dr. Hunter called the meeting to order at 8:30 a.m. and asked the committee members and observers to introduce themselves.

Programmatic Updates

Basic Vaccine Research Update

Geetha Bensal reported that the Vaccine Discovery Branch (VDB), established in May 2008, has the mission of identifying gaps in the DAIDS vaccine research portfolio and forging partnerships to bridge those gaps through programs that foster AIDS vaccine discovery. Two new funding mechanisms support this mission:

  • The Highly Innovative Tactics to Interrupt Transmission of HIV Program (HIT-IT) provides R01 grants for unconventional, outside-the-box ideas. Four of the nine grants in FY2009 concerned vaccines, and $4.5 million is available for a new round of funding in FY2010, with applications due on November 1.
  • The Basic HIV Vaccine Discovery Program awards R01 grants for research relevant to vaccine design. The 17 grants in FY2009 covered such categories as antibodies, adjuvants, B cell and T cell vaccines, innate immunity, nonhuman primate (NHP) pathogens, and vaccine structure. An additional $10 million will be available for FY2010.

The VDB was also instrumental in awarding several new and supplemental grants in FY09:

  • Vaccine Discovery Supplements were given to 13 of the Centers for AIDS Research (CFARs) for innovative work on AIDS vaccines, totaling $2.1 million.
  • The Early Career Investigator Scholar Award (a collaboration with CHAVI, IAVI, GHVE and NCRR) has funded five early career investigators on projects to develop NHP models in support of HIV vaccine research, for a total of $3.5 million over two years.

New initiatives in FY2010 will include a U19 and R01 grants for research on B cells, systems biology research on early immune response, and mechanisms and prevention of sexual transmission. In response to questions, Bansal said that the total funding for VDB programs in FY2010 will be “substantial,” on the order of $40 million, and that she will have more detailed budget figures for AVRS at its January 2010 meeting. VDB is working with the Center for Scientific Review to organize new special emphasis panels and pre-review teleconferences to ensure that proposals for these innovative programs receive a fair review.

Preclinical Research Update

James Bradac reported that DAIDS, in response to recommendations from the NIAID-sponsored AIDS Vaccine Summit held in Nov, 2008 has shifted preclinical research funds to the basic research end of the spectrum. Vaccine Design and Development to support more basic research, and there are no plans to renew many existing programs beyond FY2010 and FY2011. There will be a new award in FY 2011 for a consortium (approved by ARAC the previous day) that will focus on mucosal responses and early events in HIV infection; this consortium is in part a response to the summit call for better use of NHPs. Staff is already thinking about FY2012 budgets: about one-half of current grants concern neutralization; to increase funding for more basic research, DAIDS will need to create new, solicited research programs.

In response to his call for suggestions for FY2012, subcommittee members noted that there are promising new results in a number of areas that haven’t yet appeared in the literature, and that the AIDS vaccine research portfolio should not shift entirely to basic research. There may, for example, be opportunities to focus on antibody or vector research in 2012, and in any event it is detrimental to the field to make too many rapid swings back and forth from basic to applied research. Other members felt that the mechanisms are in place to fund new ideas in vaccine design and development, and that money will always be found for well-scored proposals. Members also expressed interest in conducting research on which funding mechanism produce the best results. NIAID has already engaged outside consultants to identify where the highest-impact papers come from.

THAI Phase 3 HIV Vaccine Trial Update on Planning for Potential Outcomes

Col. Jerome Kim described post-trial planning for RV144, a Phase 3 trial of ALVAC prime plus AIDSVAX boost that has been conducted by the U.S. Military HIV Research Program (MHRP) in cooperation with the government of Thailand. Begun in September 2003, the study was completed on June 30, 2009, and is currently in the final stages of data analysis. The study was sized to detect true vaccine efficiency (VE) of 50 percent and a 0.4-log decrease in viral load (VL), assuming a null effect on acquisition. The study enrolled 16,395 Thai men and women age 18-30, of whom 13,973 (85 percent) received all four vaccinations, and vaccinees have been followed for from three to six years since their first vaccination.

Kim reported that the FDA does not consider this trial pivotal and have rated it Phase 2B rather than Phase 3. Even the sponsors thought that the trial might be stopped early for futulity, but it was not. Now they are planning for four possible outcomes:

  1. VE ≥ 50 percent (a result that noone really expects).
  2. VE ≤ 50 percent but statistically significant (still a highly important outcome).
  3. VL reduction ≥ 0.6 log (very important from a public health perspective).
  4. Vaccine does not prevent infection or reduce viral load (an outcome that noone wants).

Regardless of outcome, investigators are faced with the chores of operational shutdown and contacting participants. They are working closely with their Thai partners to develop appropriate communications strategies for each scenario.

Preliminary data show that 14 percent of participants are HIV-positive at 6 months post-vaccination, but since the data are still blinded, investigators don’t know what this means. Participants who develop AIDS are transferred to trial RV152 for treatment and long-term followup. Sequencing of breakthrough virus will be conducted by the University of Washington using a methodology similar to that used in the STEP trial. Study results should be announced in late September and will be presented at the AIDS Vaccine meeting on October 20. Kim agreed to return to the AVRS meeting in January 2010 to update the committee on trial results and future MHRP plans.

In the discussion that followed, Jim Bradac suggested that the January meeting might also address innate immunity (Jorge Flores), vaccine-induced changes in memory immune response, and news from the Global HIV Vaccine Enterprise. He also noted that breaking news often trumps advance planning.

AIDS Vaccine Design Based on Mosaic Gene Sequences


Michael Pensiero introduced the session that the field has accumulated a lot of preclinical on mosaic gene sequences, which combine features from several different populations of virus and may thus provide better protection, both against initial infection and against subsequent viral escape. Other investigators are now moving to the creation of next-generation vectors to deliver these sequences as a vaccine against HIV. Consequently, Pensiero posed several questions to frame the presentations and discussion:

  • What are the theoretical limitations should be considered with this approach?
  • Are the current immunogenicity data sufficient to justify the increasing trend to use mosaic inserts in the next generation of HIV vaccines?
    • What additional NHP modeling should be done?
    • What caveats should investigators be aware of when modeling mosaic sequences in NHPs?
    • What influence should SIV studies have in clinical development, given the underlying difference between HIV infections in humans and SIV infections in NHPs?
  • Are the proposed clinical trials sufficient to adequately evaluate the utility of the mosaic approach? If not, what additional trials might be needed?

Design and Theory of Mosaic Antigens for HIV-1 Vaccines

Bette Korber noted that HIV is highly variable at the base-pair level and that investigators are trying several approaches to achieve the desired breadth of immunogenicity. Immunizing to a consensus or ancestral strain does nothing to prevent viral mutation and escape following infection, and antigens based on polyepitope strings have also produce poor results. Another approach based on conserved region design is currently in animal trials, but early results indicate that most conserved regions are not well-processed immunologically. Whole protein cocktails might be able to exploit epitope overlap, but mosaic gene sequences promise to provide greater population protection against infection as well as greater protection against escape mutations after infection. The current approach involves an optimized combination 9-mer epitopes that is computationally designed to match the greatest number of 9-mers in virus from the sera of chronic infections, in part because chronic sera are more common than acute. Investigators use 9-mers because the results are almost as good as 8-mers, better than longer chains, and robust over time.

NHP models are a powerful but limited tool in testing this approach: humans and monkeys have different patterns of peptide recognition and dominance; there are more selective forces on HIV mutation; and as a result NHP results don’t always translate well into human scenarios. Progress in this area would be encouraged by deep sequencing during the early, acute stages of HIV infection. There is a sudden explosion of mutation at day 16 resulting in as many as 57 varieties, but this explosion slows by day 45 as the dominant mutant outgrows its competitors. The result in most cases is a single lineage for each patient.

In response to questions, Korber explained that investigators arrived at their combination of sequences by iterative optimization against all 9-mers, beginning with the most frequent. However, the most conserved regions didn’t always evoke the greatest response, nor did the rarest variants. Korber has not tried to determine ho many of her predicted epitopes show up in the 12 “supertypes” of HIV-1, but she agrees that this would be a good experiment. Participants suggested several other follow-on ideas, and Korber agreed that – not knowing how well the SIV model predicts HIV results – many additional studies will be needed to map the exact differences.

Immunogenicity of Escaped CTL Epitopes

David Watkins reported that, in SIVmac239, most viral mutations outside the envelope are caused by selective pressure from CD8+ cytotoxic T-lymphocytes (CTLs). Similarly, two-thirds of all mutations in chronic HIV patients are attributable to CD8+ T-cell pressure, but another 20 percent of mutations involve substitutions toward a conserved (fitter) epitope, and hence a reversion of the virus to wild type. But when NHPs were challenged with a cloned SIV bearing common escape mutations in three immunodominant CTL epitopes, responses to two of the three altered peptides were barely detectable. In a comparison study, NHPs with SIV bearing eight point mutations produced barely detectable responses to the five most immunodominant CTL responses in NHPs infected with wild type SIV. These results reveal that CD8+ T cells select for “escaped viruses.

In response to questions, Watkins added that about 25 percent of T-cell receptors induce a response, but he doesn’t know why some do and others don’t, nor does he know why some mutations revert to wild type. This could be an important question, since some chronically infected patients go on to become elite controllers. Nor do “wild type” and “escaped” have the same meaning across clades or even across patients, since an escape mutation in one study is a susceptibility in another. Participants suggested that more might be learned through a peptide binding study – if the binding is weak enough, the epitope may not be “recognized.” However, most of the variant peptides tested were recognized in standard ELISpot and ICS assays. The solution is not to go after conserved epitopes, since the mutations aren’t always in conserved regions. Perhaps it’s a vector problem, since vaccinia and adenovirus induce so many responses of their own. And the demands will be different in acute infection than in chronic – during the first 72 to 100 hours the immune system only needs to recognize one sequence, but diversification is increasingly rapid thereafter.

Caveats in Measuring Breadth of CTL Recognition of Epitope Variants

Otto Yang observed that HIV diversity poses at least two problems for a vaccine: (1) matching vaccine sequences with challenge sequences across an entire population, and (2) the ability of HIV to escape via mutation within the individual host. In an individual infection, CTLs are a major determinant of immune containment and thus of subsequent viral load, but the early loss of CD4+ helper cells impairs the system’s ability to “retarget” escaped virus in chronic infection. The solution is to increase the number of fit variants that are recognized by the CTL response (i.e., increase the number and/or promiscuity of T-cell receptors). Increasing the diversity of the vaccine will not necessarily increase the breadth of the immune response, however, due to competition and immunodominance among epitopes. Indeed, the more variants are included, the less likely that this approach would be successful. In fact, current peptide assays (ELISpot, ICS) may overestimate CTL activity against infected cells, just as reactivity against a panel of epitope variants doesn’t necessarily reflect true promiscuity. The solution, it would seem, is to include only the more conserved regions in a vaccine; still to be determined is whether there are enough different conserved regions to provide enough epitope for immune control.

In response to questions, Yang agreed that the virus-killing assay itself is a questionable measure of efficacy, because the functional assay has not been validated with clinical data; a meaningful assay would have to be clinically predictive. At present he is unable to explain the differences among peptides, although he suspects that it has to do with glycolyzation. He agrees that mass spectrometry might be a useful approach for understanding these differences, and he already has a post-doc working on this question.

Mosaic HIV Immunogen-induced CD4+ and CD8+ CTL Responses in Rhesus Monkeys / Human Clinical Trial Plans

Norman Letvin presented the results of a comparison study in which 12 rhesus monkeys were vaccinated with plasmid DNA prime plus vaccinia virus boost containing a single consensus gene insert, while 12 other received DNA prime plus a cocktail of four complementary mosaic gene inserts, and 6 control animals received DNA prime plus empty vector. Breadth of CTL response was measured against ten Gag and ten Nef sequences from clades A, B, C, and G. Results showed no dramatic difference in the vaccine-induced antibody responses, although the mosaic vaccine tended to respond to more epitopes, and more variant epitope sequences, as well as eliciting a slightly stronger CD8+ response. Over all, CD8+ CTL recognition of epitope-specific variants was greater in mosaic- than consensus-immunized monkeys by a factor of 2.5:1.

In response to questions, Letvin added that all monkeys received the same dosages. Some proved to be “high responders,” a phenomenon that might be worth studying if an adequate experiment can be designed. He did not see a correlation between CD4+ response and the magnitude or breadth of CD8+ response, and indeed this may an effect of the vaccinia vector, which is known to have a CD8+ bias. These differences suggest that mosaic vaccines would provide protection against diverse HIV isolates and control of emergent variant viruses – clearly a rational for assessing the mosaic vaccine strategy in humans.

Barton Haynes reported progress in designing a Phase 1 clinical trial to address these questions by assessing CD4+ and CD8+ responses to (1) wild-type transmitted/founder Env from clade B virus, (2) consensus Env from Group M CON-S virus; and (3) trivalent mosaic Env. Funding for year 1 (vaccine production) comes from the Gates Foundation and is administered by NIH; clinical trial support will come from the HIV Vaccine Trial Network (HVTN); DAIDS will provide codon-optimized DNAs under contract. Investigators chose the recombinant NYVAC as the vector because it has known safety and immunogenicity; in addition, it induces greater CD4+ than CD8+ responses. However, it will be important to know how mosaics perform in different vectors. For this reason, the RV144 trial with canarypox vector will be informative. The study will have 35 subjects per arm and is powered for immunogenicity for mapping breadth and depth of both CD$ and CD8 responses.

One participant asked why it was necessary to design a comparative trial, why it is necessary to get both CTL responses when only CD8+ is needed, and generally why such a large, complicated trial when a smaller, simpler trial might answer more (simpler) questions. The answer involves validation: if both in silico and NHP models predict success in humans, then a different design is needed to validate those predictions and have confidence in the numbers. As to why NYVAC would be the vector, the reason is that adenovirus has a known CD4+ bias, the design requires both CD4+ and CD8+, and NYVAC is available.

Immune and Viral Tests of Concept for Mosaic Antigen Vaccines: Animal Models and Clinical Development Plans

Gary Nabel suggested that the consensus and mosaic approaches aren’t mutually exclusive. The general paradigm for structure-assisted vaccine design moves from broadly neutralizing sera to monoclonal antibodies and epitopes, which are defined at the atomic level, followed by immunogen design based on epitope mimicry and immunofocusing, immunization, and analysis of immune response. In the case of HIV, a broad diversity of neutralizing antibodies were isolated from the memory B cells of infected individuals, many of them targeting the gp120 and gp140 proteins. Subsequent isolation and characterization have led to the creation of a series of mosaic immunogens that combine two 22-amino acid sequences that are common to gp120 and gp140 and confer enhanced immunogenicity. A preliminary NHP study is currently underway, with results to come in November that will settle question like 2 vs. 3 mosaics and coinjection vs. separate injection. In the longer run, however, t is uncertain whether the results of SIV mosaics will predict HIV results. Translation of SIV results to HIV vaccines might be advanced by the identification of SIV clades. Many other important questions remain to be answered with regard to manufacturing and cost.

Nabel concluded that mosaics are an elegant model, and the preliminary data are encouraging, but those data are informatic and need to be confirmed with biological data. Doing so will provide a proof of concept and will help address questions about mechanisms and correlates of protection. The goal is to identify the best vectors and inserts for further development. Under the current timetable, optimized immunogens might be expected in early 2010, validation in NHPs and humanized mice in late 2010, and Phase 1 safety and immunogenicity trials in 2011.

Mosaic HIV-1 Vaccines Expand the Breadth and Depth of Cellular Immune Responses in Rhesus Monkeys

Dan Barouch agreed that it is increasingly clear that the next-generation T cell-based HIV-1 vaccine candidate must have a vector that avoids preexisting vector-specific NAbs and antigens that improve the Gag-specific breadth and depth of cellular immune response. However, increased valency in mosaic vaccines has diminishing returns, and in some cases 2-valent may be just as good as 4-valent. In an NHP study, a bivalent mosaic vaccine elicited more epitope-specific responses than M consensus, clades B plus C, or optimized clade C vaccines by a factor of 3.8:1, and a greater CD4+ response than the other three vaccines combined, and with no detectable problems of antigenic competition of immunodominance. The mosaic vaccine also increased the depth of the CD8+ response compared with the other candidates, recognizing more peptide subpools spanning actual Gag proteins from clades A, B and C.

There are limits to what can be learned from NHP studies, however. For example, they can’t assess the ability of mosaic vaccines to elicit human HLA-restricted CTL responses, and they may overestimate the breadth of response because of multiple MHC-B alleles. Nor is there any way to assess protective efficacy of mosaic vaccines in NHPs; clinical studies in humans will be required. For this reason, NIH’s Integrated Preclinical/Clinical AIDS Vaccine Development Program (IPCAVD) is working to develop a mosaic HIV-1 vaccine candidate for evaluation that combines (1) a heterologous rare-serotype recombinant adenovirus prime-boost regimen (Ad5HVR49-Ad26, Ad35-Ad26, Ad48-Ad26), and (2) bivalent mosaic Gag-Pol-Nef antigens. Work is also going forward on next-generation HIV-1 vaccine candidates based on optimized modified vaccinia virus Ankara and Ad26 regimes.

In response to questions, Barouch said that the current trial design does not select for HLA type, but this could be discussed. He acknowledged that the evolution of SIV in rhesus macaques and sooty mangabeys is not driven by CD8+ and is thus not a good model for HIV in humans.


Michael Pensiero repeated the questions he had posed at the beginning of the session and asked for general discussion. On the subject of theoretical limits, several participants pointed to the fact that vaccinating NHPs against SIV may not be an adequate test of the concept. On the other hand, why does a mosaic of human antigens elicit a broad response in humanized mice? Probably because 9-mers represent so much diversity that you’re likely to see a response in almost any species or model. Questions of antigenic interference and immunodominance, on the other hand, can only be answered in humans. It seems reasonable that the most conserved regions will be well represented in any mosaic antigen candidate.

A more practical question arose about who can run the software needed to optimize immunogen designs. While the original designer has moved on, the software itself is freely available at the Los Alamos website, which also provides a tutorial on how to use this an other informatics resources.

There was some debate about how much of the preliminary results are an artifact of the PTE peptide pool. However, Letvin didn’t use PTE and got the same results as Barouch, and Barouch himself got the same results with both PTE and non-PTE sequences. This led to the broader questions of how many mosaic designs the field should be testing. For example, Letvin and Barouch used distinct biological approaches and arrived at distinct mosaics, and the VRC approach is different again. Consensus emerged that the studies are complementary and that it would make no sense to duplicate the same work in different laboratories.

A slightly different question involved the valency of mosaic designs. Why does 4-valent work better than 2-valent? Is it just that the same stimulus is repeated several times over, resulting in a more potent response? If that is the case, why not 6-valent or 12-valent, especially if it increased the amount of antigen being expressed? By the same token, why not abandon Env in favor of Gag or Nef. The latter question could be settled with an assay of functional avidity and would be a relatively simple experiment to do. Ultimately, however, none of these questions can be answered satisfactorily without going into humans.

Eric Hunter offered the following as consensus statements of the AVRS:

  • The preliminary results indicated that mosaic vaccines work better than we might expect, and we’d like to know why. We seem to be moving to something common, rather than something unique, but we don’t understand why.
  • At the same time, there is a limit to how far you can extrapolate from NHP data.
  • CHAVI should complete its Ad-based trials, which are sufficient to test the concept.
  • AVRS would like to see HLA diversity in the Nabel and Barouch trials, at least B57, A01 and other alleles for which escape mechanisms are known. If this is a problem for trial design, it might be possible to do a Phase 1 trial in the United States and South Africa at the same time, in order to increase HLA diversity.
  • It would also be desirable to examine HLA in the Letvin trial, possibly as a substudy. The goal is to have balanced HLA, not selective.
  • In all three trials, we need enough cells to test the depth of response – how many variants the immune system can respond to. Breadth, for current purposes, is gravy. This is easier to do in human studies, since you can get more dells from human volunteers, and additional cells can be obtained from a plasmapheresis center.

The meeting adjourned for the evening at 5:00 p.m. and reconvened the following morning at 8:30 a.m.

Large-Scale Efforts to Address the Neutralizing Antibody Problem

Jim Bradac introduce the session by noting the many efforts already underway in this area and posed the following questions to guide the subcommittee’s discussion:

  • How can NIAID best contribute to accelerating this field in the future?
  • Should we rely solely on unsolicited R01 research?
  • Should we attempt to expand already supported efforts?
  • Should we launch a new effort, and if so, of what scope, size, structure?
  • To what extent should NIAID’s plans in this area be guided by the Enterprise Scientific Strategic Plan, which is under development through a broad consultative process?

Overview of the VRC Program

Gary Nabel reported that the VRC’s NAb immunogen program employs the same structure-assisted discovery program he had described earlier for its mosaic vaccine program. The fact that 15 to 20 percent of sera from chronic AIDS patients in all populations contain neutralizing features is proof that it’s possible to generate NAbs. The difficulty comes in designing immunogens to elicit those NAbs.VRC employs a variety of assays for specificity, avidity and neutralization to screen sera for promising molecules, with the goal of better understanding the Abs that produce these effects. Preliminary results reveal that specificity of binding sites is especially important to neutralization.

VRC is working not only with gp120-binding T cells but also with gp140-binding B cells. The current assay is not optimal for B-cell binding sites, but better tools should emerge from ongoing studies, just as data from cryo-electron microscopy (CEM) should yield better models of molecular structure. The b12 trimer appears promising, although it requires a very precise angle at the binding site – a 12-degree shift in alignment removes the neutralizing capacity. Researchers have also deduced the structure of other binding sites that are not neutralizing, as well as hundreds of “cloaks” that can be used in pairs for high-throughput screening of sera for Abs that directed at specific structures on the virus. These cloaks show less promise as immunogens than as probes, but VRC has also identified several trimers that show promise as immunogens. The goals is to identify immunogens that elicit broadly NAbs, and Nabel characterized the current candidates as “broadly neutralizing, but not very.”

He stressed the importance to the field to generate new candidate immunogens to be tested in Phase 1 trials. Since the lag time is about 12 months, we need to have more candidates in the pipeline, and we may also require new technologies to manufacture these antigens on a large scale. VRC plans to collaborate with established leaders in computational design to accomplish this. In response to questions, Nabel said that there is no evidence that we are within years of a B-cell vaccine, for NHPs or for humans – years of scientific discovery are still needed, and translation of these results will be a challenge in itself. But the field needs to turn some of these early mAbs into immunogens while that basic research continues. Several participants complimented the VRC team on its “beautiful work,” but reminded them that the ultimate test is whether they will work in humans.

Overview of CHAVI Program

Bart Haynes summarized the present status of the CHAVI B-Cell Immunology Program, which focuses on the structure and conformational states of conserved envelope epitopes found on transmitted HIV-1 isolates. Key questions include the universe of specificities of autologous NAbs, how to design an immunogen that elicits enough of those specificities to be relevant for a vaccine, and how to elicit the desired antibodies sooner than the 8 to 10 days in natural infection. New technologies being brought to bear on these questions include single-cell sorting of antigen-specific memory B cells and 454 deep sequencing for clonal analysis.

Investigators are analyzing changes in NAb profiles over time in different chronic patients, hoping to discover why most patients don’t make broadly neutralizing Abs, what’s different about those patients who do eventually make them, and why the latter group doesn’t make them at the time of transmission. They are also looking at Ab responses to viral gp41, which does occur immediately after infection but is not neutralizing, and at differences between protective vaccination and natural infection. Analysis of Ab response to acute HIV infection has shown that it is very different from influenza infection, in that HIV elicits a polyclonal immunoglobulin response, but most of the clones are low-affinity and non-neutralizing. Acute influenza also elicits a polyclonal response, but most of the clones are high-affinity, strongly-reactive, and far more likely to be neutralizing. Further studies are under way to determine Env signatures in acute and chronic infection, in order to design an immunogen that presents optimal structures to B cells that might not otherwise be produced. Clinical collaborations have been vital to the progress of these studies.

In the discussion that followed, participants suggested that it might be useful to select a single feature – e.g., HLA57 – and backtrack to the earliest immune response following transmission, in hopes of discovering what’s different about patients who respond soonest. There was discussion of the usefulness of NHPs in studying the evolution and repertoire of immune responses, especially during the “eclipse” period between transmission and seroconversion. Once again, several observers complimented the “beautiful science” in addressing the gap between antigenicity and immunogenicity. Haynes acknowledged that our knowledge of B cell biology is not yet what it should be, but the field is mature enough for additional NIAID investment. Others noted that the gap between infection and response, in both NHPs and humans, will be hard to close: it’s difficult to speed up the maturation of Ab responses, even to childhood vaccines, and it may not be reasonable to expect a high-affinity NAb response in a short time. In addition, viral escape during acute infection frustrates a broad response – even the best Abs don’t recognize two-thirds of the virus. For this reason, proximal Abs are the key to overcoming escape, and polyvalent approaches might work well.

Overview of the IAVI-Funded Neutralizing Antibody Consortium (NAC)

Wayne Koff posited three characteristics of a safe and effective HIV vaccine:

  1. Elicit durable and broadly neutralizing antibodies, in order to prevent infection;
  2. Elicit durable and broad systemic and mucosal cell-mediated immune response, in order to control infection; and
  3. Be applicable for use in the developing world, where the epidemic is most severe, in order to have the greatest global impact.

IAVI is addressing all three problems, but at present the NAb problem is the most challenging, and the biggest gap is in immunogen design. IAVI supports 18 institutions and numerous contracts in the relevant areas of science, with the goal of developing an immunogen that elicits NAbs to 50 percent of HIV isolates. An important component is Protocol C, which follows that natural history of acute and chronic infection at eight clinical centers to better understand the development of NAbs in multiple populations. Related studies will focus on the evolution and breadth of NAbs, the role of bNAbs in controlling superinfections, and the identification of new targets for vaccine design.

Dennis Burton provided an overview of recent scientific advances by NAC. NHP studies have shown that protection against infection can be achieved at relatively low neutralization titers for some Abs (e.g., 2G12), that interactions between Abs and cellular immunoglobulin receptors (FcR) contribute to protection, and that Abs to the membrane-proximal external region (MPER) are also protective. NAC now plans studies of glycan mimics (2G12), novel Abs (PG9 and PB16), Ab synergy in protection, targeting MPER epitopes (2F5, 4E10, Z13), and protection by non-neutralizing Abs. Several of these NAbs have shown considerable potency and breadth against a large panel of 162 viruses from all eight clades.

NAC has assembled a vertically integrated consortium to rapidly move promising candidates into immunogen design, screening and clinical development. It has also formed a sub-cohort of Protocol C, called Protocol G, of chronic patients and elite controllers who produce broadly neutralizing Abs, which can then be ranked by breadth/depth and then characterized in interaction with Env. Burton expects that the near future will see an explosion of additional bNAbs of high potency and breadth that will facilitate work not only on the structure of the functional trimer but also on the design of immunogens for an effective HIV vaccine. He agrees with his colleagues at CHAVI that important lessons can be learned by following the evolution of AB responses in individuals from the time of infection to the development of broad neutralization. He applauds NIAID for funding many of the most important parts of this research.

Overview of CAVD-Funded Programs

Nina Russell presented an update on the Gates-funded Collaboration for AIDS Vaccine Discovery (CAVD) Antibody Consortia, which include three Vaccine Discovery Centers and five Central Service Facilities. The goal of CAVD is to accelerate HIV vaccine discovery by encouraging collaborative research, designing novel candidate vaccines, improving and standardizing laboratory practices and data analysis and sharing materials in information, and moving promising candidate vaccines to preclinical and Phase 1 trials. This is a unique model for big science, involving 100 institutions in 21 countries with over 400 investigators, linked by web portals, legal agreements, and grant agreements. CAVD is also liked to other HIV vaccine efforts, including CHAVI, IAVI, VRC, and HVTN. Metrics for success include both increased productivity and collaboration (publication, abstracts, web use, and staff exchanges) and the sharing of information and materials. Total funding for CAVD is $408 million, of which about 41 percent is exploratory, 25 percent preclinical and 5 percent clinical, with 23 percent for central services and 7 percent for organizational overhead. The original grants will expire in 2011 and recompetition will begin in 2010 with a “robust external review.” In response to a question, Russell characterized this and other Gates efforts as primarily translational, designed to focus the results of more basic research (funded by NIH) and move them toward the clinic.


Jim Bradac opened the discussion of bNAb programs by repeating the questions he had posed at the beginning of the session:

  • How can NIAID best contribute to accelerating this field in the future?
  • Should we rely solely on unsolicited R01 research?
  • Should we attempt to expand already supported efforts?
  • Should we launch a new effort, and if so, of what scope, size, structure?
  • To what extent should NIAID’s plans in this area be guided by the Enterprise Scientific Strategic Plan, which is under development through a broad consultative process?

Participants noted that it would difficult to achieve a perfect balance of NIAID support, but the current portfolio is more or less on target. We need to learn more about what happens during the first 72 to 100 hours after transmission in both humans and NHPs. There have been big advances from both R01 grants and large interdisciplinary programs, and both forms of support. The need for libraries of serum samples, by itself, justifies the expansion of existing collaborative programs, and the new GHVE scientific plan should provide a valuable endorsement of these activities. NIAID can best contribute by continuing to provide financial resources and by “investing in winners.”

On the scientific side, there have been many (and will be more) successes from focusing on antibodies, but many of those Abs won’t work, and there is a growing gap in immunogen design. Participants applauded the progress being made in mucosal studies and urged continued funding for this promising field. IAVI has had notable success in building cohorts, which provide a lot of “bang for the buck,” but they should if possible include more pregnant women and exposed seronegatives. NIAID should look for ways to support IAVI and CAVD work with cohorts.

There is a continuing need for both R01s (for novel ideas) and consortia (for coordination). There is a risk that some promising questions and avenues will receive inadequate support, including the following:

  • What is the biology of B cell response;
  • How does the body produce Abs in response to immunogens;
  • How do we design immunogens to elicit the Abs we need (“It’s not an Ab problem, it’s an immunogen problem”);
  • How do we make optimal use of the NHP model
  • Is the 2-valent approach to HIV vaccine optimistic, given that flu vaccine is 23-valent;
  • Should we vaccinate once with a “cocktail” of virus or repeatedly with different viruses, particularly under the constraint of a suppressed immune system;
  • Do we need new vectors or inserts to overcome maturation, without which HIV is not infectious;
  • Do we need more work on the engineering of proteins that will elicit desirable Abs;
  • Do we need to develop new cell lines that are capable of producing the desired bNAbs.

The R01 grant remains an important mechanism and is one of the most effective ways to support young investigators. However, the current payline creates a frustrating environment, and perhaps NIAID should focus R01 grants in gap areas, like those above, which would produce needed research while still encouraging young investigators. Another approach might be to require collaborative consortia to include young investigators, but consortia use network review rather than peer review. As a result, this approach might help young investigators survive, but it would leave even less room for evaluation.

The meeting adjourned at 12:30 p.m. The next meeting of AVRS will be on February 2-3, 2010, location to be determined.

Action Items

At the January 2010 meeting:

  •  DAIDS staff will provide more detailed budget figures for FY2010;
  • There will be an overview of the research portfolio of the Division of Allergy, Immunology and Transplantation (DAIT) that is relevant to vaccines and HIV.
  • Staff will present information on the ongoing study of high-impact research papers and the relative effectiveness of funding mechanisms.
  • AVRS will consider adding a new member who is expert in basic immunology.
  • Jerome Kim will update AVRS on the results and future plans of the MHRP Phase 3 trial in Thailand.

back to top

Last Updated June 27, 2011

Last Reviewed April 12, 2010