An Interview with Gary Nabel
Immunology and AIDS Vaccines
Gary Nabel, MD, PhD, has long been a leader in the development of AIDS vaccines. Since 1999 he has been Director of the Vaccine Research Center (VRC) at the National Institutes of Health in Bethesda, Maryland. Prior to taking this position, he was director from 1997-1999 of the Center for Gene Therapy and a Howard Hughes Medical Institute investigator at the University of Michigan in Ann Arbor. Nabel also served as the co-director of the University of Michigan Center for Molecular Medicine from 1994-1999. He is a member of the Institute of Medicine of the National Academy of Sciences.
Nabel’s diverse research interests range from applied immunology to fundamental molecular biology. His laboratory has developed gene transfer strategies that are being applied to the development of vaccines for and treatment of AIDS, Ebola virus, cancer, and other diseases. In addition, he has long been involved in studying NF-κB gene regulation, with an emphasis on its relation to cell cycle progression, transformation, and viral infection by HIV. Here, Nabel speaks with IAVI Report Editor Simon Noble about our current understanding of the scientific issues critical to developing an AIDS vaccine.
Originally the VRC had a wide remit to tackle multiple infectious diseases. What are your current broader plans for the VRC and how much emphasis is currently being put on vaccines against potential bioterror pathogens?
Our raison d’être is really to drive the effort to develop an effective HIV vaccine, whether that’s by developing a candidate that we bring along or by helping to assess various candidates and find one that works the best, however we can help. Needless to say, what is often lost in our concerns about bio-defense is the fact that HIV today dwarfs the danger that most other biothreats pose to us.
Having said that, when there are economies of scale and when there are platform technologies that will allow us to either understand emerging pathogens or bioterrorism threats, then we are interested, willing and able to help in those efforts.
What particular pathogens, aside from HIV, are you currently focusing on?
The short list for us is HIV, Ebola, West Nile, and SARS viruses, and more recently we’ve tried to help with the influenza virus problems. Those are the other VRC efforts. We just received word from the FDA that it's okay to proceed with our SARS DNA vaccine. That should start next month, provided the rest of the approvals go according to plan. So we’re moving in several areas.
To switch to HIV more specifically, the VRC is closely involved in the efforts to standardize assays. Which assays and protocols do you see as the best way forward in that effort?
For cellular immunity, our major focus has been on flow-based technologies, flow cytometry, and particularly looking at intracellular cytokines, staining for a variety of different cytokines. We’ve also been performing the ELISPOT assay. Our experience so far is that flow cytometry appears to be more sensitive and a bit more informative about which cell subsets are responding. So that’s our preferred assay. Despite the power of the flow-based technology, we still have a lot of work to do, as a field, to develop even more sophisticated assays that I hope will lead us to an immune correlate.
Do you think some of these assays are practical though for resource-poor settings?
You don’t need to implement flow cytometry, at least of the kind that we would put into place to support licensure of a vaccine product, at every international site. But you do need to have the capability of storing and transporting the cells properly to where the analysis can be done. We have to recognize the importance of defining an immune correlate and taking whatever steps are necessary to do those analyses.
On that point, it's perhaps becoming increasingly evident that interferon (IFN)-γ might not be the only marker to focus on. What other markers do you think might be critical?
Mario Roederer and Rick Koup at the VRC have been looking at a variety of factors in different vaccine models; they have looked at TNF-α, MIP-1α, IFN-γ, IL-2, a variety of cell surface markers, as part of our discovery research. We need to evaluate them all before we know which will be the best. I think RANTES, MIP-1α, the chemokines generally, are of great interest, and a number of cytokines as well.
Do you think there is any functionality strongly linked to any of these markers that makes one perhaps more promising than another?
At this point I don't know. I'm intrigued that there may be innate immune mechanisms that might be involved in protecting against infection, like MIP-1α, RANTES, or other chemokines, products of the natural inflammatory response that block infection. Is it possible to develop an acquired immune response that generates some of those protective chemokines in a more rapid fashion?
Although we're all convinced that cellular immunity is important in protecting against viral replication, we still don't really understand the mechanism by which that occurs. We often say that it’s killer T cells, but is it killer T cells that are killing infected target cells or are they elaborating cytokines or are there other mechanisms? We still have a long way to go when it comes to understanding immune mechanisms and immune correlates of protection.
Which of the current viral vector candidates do you think are most promising, the ones that capture your enthusiasm?
In many ways the candidate platforms are still testing the waters. I guess we've put our money down on the things that we think work well; I'm impressed with how well adenoviral vaccines have worked in a variety of animal and human models, not only for HIV but also for Ebola virus and other pathogens. Compared to other vectors that have been tested, the adenoviruses are much more effective in eliciting cellular immunity and also induce humoral immunity, especially after DNA priming. They are very high on our list.
DNA, particularly as a prime for adenovirus, has done very well in animal models for both HIV and Ebola virus; if we can get the DNA to work well in people, the platform technology is one that has great promise.
Together, those technologies will allow us to make a reasonable first-generation vaccine, and that should give us an opportunity to test the concept that cellular immunity can protect against HIV infection.
Whether it be the Merck adenovirus vaccine or our vaccine, I think they will both be informative, with the difference between Merck's and our's being that we'll be testing DNA-adenovirus in a prime-boost combination and we have also included envelopes. So I think we'll learn very different things from these upcoming trials.
How is your multivalent DNA vaccine candidate proceeding?
We've actually completed Phase I for our first-generation multivalent DNA vaccine, and we’ve completed enrollment for our multiclade adenovirus candidate. We decided to make the multiclade, multivalent vaccine after convening a meeting at the VRC where we invited our international colleagues from India, from different countries in Africa, from China, from the Caribbean; we put our heads together and asked “What’s the best we can do to deal with the global problem in terms of a vaccine candidate?” and the concept of the multiclade ABC candidate arose from these discussions. We’re very dedicated to it; it’s our top priority. We’re trying to move it into efficacy testing in different parts of the world, particularly the developing world, as fast as we can.
How do you see the future of DNA vaccines? Do you think they can make the grade?
We’ve been pleasantly surprised at how well DNA vaccines are performing in people. That’s not to say that DNA alone will be the answer, I think it’s likely to be a prime-boost combination. But having said that, we see pretty consistent cellular immune responses, we’re even seeing antibodies with some consistency, not just for HIV but with other vaccine targets as well.
What we should remember about DNA is that, like all new technologies, it can and will evolve. The use of altered regulatory signals, the CMV enhancer with the HTLV-1 R region, for example, represents a technically simple improvement that enhances the immunogenicity of the vaccine significantly, and there will be other improvements in the future; there are technical aspects in terms of how you purify the DNA. We’re still just scratching the surface of DNA adjuvants. If electroporation can be developed into a safe, clinically-acceptable delivery method, it could have utility. I look at DNA vaccine technology as an opportunity.
Moving from platforms to the actual immunogen, what do you think are the best approaches to rational design of immunogens?
The critical path to a highly effective vaccine runs through the HIV envelope. We are already able to induce cellular immune responses to the internal proteins, and we’d like to broaden that response by generating immunity to conserved regions of the envelope.
But to have a highly effective vaccine, the broadly neutralizing antibody remains the Holy Grail. The jury is still out in terms of whether we can get there. I think that the most logical and productive way to proceed is the structure-based approach. Peter Kwong and Rich Wyatt at the VRC have spent years studying the envelope and the interactions with CD4. We’re trying to modify the protein in such a way that it will be conformationally fixed yet expose determinants that we hope will elicit these antibodies. Peter’s also working on new structures, including the V3 loop, getting the trimeric rather than the monomeric structure. Peter and Rich’s recent paper on the 2F5/4E10 region is another important step forward (see AIDS Vaccine 04).
Steve Harrison has reported the un-liganded form of SIV at meetings and it will be an extremely important contribution to the field. We’re conferring with Steve and trying to incorporate that information into vaccine design as well—there may be HIV structures that we can capture with rational design that are more representative of the un-liganded structure. Also, there are clade differences that we don’t understand yet: Why in clade C do we see so little variability of the V3 loop, and yet there are regions just downstream of V3, and V4 of the clade B’s, which are variable? Understanding the structural basis of these clade differences will be important.
It may also be helpful to re-visit the spectrum of neutralizing antibodies that one sees in nature. Many of the antibodies were originally selected by binding to gp120, as a prescreen for neutralization. If you look at Antonio Lanzavecchia’s work in SARS, he simply screened by neutralization in an infection assay and the monoclonals that he selected were ones that may not have been picked up if he pre-screened by binding to the SARS protein. They’re very weak binders that recognize conformational determinants. So I don’t see why we can’t go back and ask whether the universe of broadly neutralizing antibodies for HIV might be larger than those that we have defined so far.
I’m also intrigued with the idea of using the genetic diversity of the virus in a positive way for vaccine design—there are clues in what the virus does and does not conserve. So both structural and genetic approaches are now our highest priorities.
How do you think immunodominance of epitopes can be addressed? Do you think the immune response should be directed towards subdominant epitopes?
That’s a really interesting question. The more I look at the virus and how it manipulates the immune system the more I’m convinced that the concepts of original antigenic sin and immunodominance are major scientific questions that we must resolve. So yes, it’s a very important area of research. The virus has apparently used this phenomenon for its own gain.
How important do you think improved adjuvant approaches will be?
Adjuvants have always been an essential part of vaccines, and they will remain so. Finding ways to improve mucosal immune responses are important. It would be wonderful if we could design a chemical adjuvant that does what adenovirus does to the immune response, because adenovirus, particularly as a boost after DNA prime, generates a very strong cellular immune response and a potent antibody response. Current adjuvants largely improve antibody responses but don’t do a lot for cellular immunity. We’re understanding more and more about the toll receptor pathways and about the nature of the specific antigen-presenting cells, so I think there will be opportunities for adjuvant improvement.
You mentioned mucosal immunity, it does seem to be increasingly accepted that HIV is essentially an infection of the gut-associated lymphocytes (GALT). In light of some recent direct corroborating evidence in humans that there is massive depletion of CD4 cells in the gut throughout infection, should more attention be paid to mucosal immunity in general and GALT in particular?
I think it is very important to look at protecting the mucosal compartment. Going back to Ron Veazey’s and Ron Desrosier’s original paper and, as you’ve said, the recent papers from Danny Douek and from Marty Markowitz (see Research Briefs), it’s clear that the virus takes off in this compartment. It introduces a whole series of questions: What’s the best way to protect that compartment? Do you provide higher levels of immunity by generating mucosal immunity? If you do, what kind of immunity do you want? Is a good IgA response protective in the mucosa? Do you want mucosal cellular immune responses? Do you need both? Are you better off generating a good IgG response so that the virus never can make it to the mucosal site? It would help to have better tools to stimulate those mucosal immune responses more selectively and parse out the relevant issues.
More generally, do you think CD4 help will be critical in a vaccine-induced immune response? Everyone’s focusing on cytotoxic T lymphocytes (CTLs) at the moment, but increasingly evidence points towards CD4 help in propping up any CTL response.
I’ve been impressed with the importance of CD4 cells, particularly memory cells. You’re quite right to point out that their role in protective immunity is sometimes overlooked. Clearly, one won’t be able to generate strong CTL responses without potent CD4 memory responses, and there are other contributions these cells might make. For example, they may elaborate cytokines, like MIP-1β or RANTES. Getting more specificity on precisely what type of CD4 cells are needed would be even more helpful.
What are our best leads for inducing CD4s as part of a vaccine-induced immune response?
DNA vaccines induce CD4 cells pretty nicely and have some potential. In our VRC004 trial, the responses to the DNA vaccine were more consistent in CD4 cells, and they prime well in animal models for the adenovirus boost, which then brings a more balanced CD4:CD8 ratio. DNA vaccines seem to get around some of the questions that you raised earlier relative to epitope dominance. That’s another reason to really drill into the DNA technology and not abandon it prematurely. And there are adjuvants that will induce a CD4 response; we just need to make sure they’re Th1-like, not Th2-like. Some of the CpG adjuvants are interesting in that regard.
How do you think we can better proceed with the non-human primate (NHP) model, given the lack of successful protection data in that system? Do you think there’s still research value in that model or should it be considered more as a filter when moving towards the clinic?
I wouldn’t use the NHP challenge as a filter because we know now that any vaccine that induces a reasonable cellular immune response seems to protect against SHIVs (simian human immunodeficiency viruses). But I do think, as with DNA technology, there’s enormous potential to understand and model the disease in the monkey. It’s very important to look rigorously at more physiological models, for example the low-dose repetitive challenge model like the one David Watkins described last year. If we can work out those models so that they’re consistent and more representative of human infection we can better use them to test our candidate vaccines. Even if the animal groups need to be larger to do those studies it will be much more expeditious to address scientific questions in those animal models. The only other option is to do proof-of-concept trials [in humans] that take hundreds of millions of dollars to perform and years to complete.
The Merck Ad-5 [adenovirus serotype 5] vaccine trial has been widely touted as proof-of-concept that will test whether cell-mediated immunity can impart some efficacy as part of an AIDS vaccine. What should the field do if Merck Ad-5 doesn't work?
Keep working! There are a number of assumptions in that trial that are very specific, and if it were to fail the results will inform us about those assumptions. But that trial is powered to look primarily at differences in viral set-point and CD4 numbers, and it’s not well-powered to look at acquisition [of infection]. So one possibility is that it might not have an effect on set-point or CD4 counts, and yet it might have an effect on acquisition because they involve very different parts of the infectious life cycle. Another caveat is that there is no envelope component in the Merck vaccine, and envelope offers the opportunity to expand the diversity of the immune response and to generate neutralizing antibody responses.
Another possibility is that we all guessed wrong. Maybe the cellular immune response is helpful but not the response that adenovirus induces. The plan that has been espoused by the HIV Vaccine Enterprise is one that I think we all can agree with conceptually, that we need more human studies, potentially ones that induce alternative immunologic responses. Each trial explores one part of the multi-dimensional immunologic space—if the trial fails, we can exclude that part of the space, but there’s a lot of area to explore. We need to make sure that we cover new ground with each trial.
What should be the priorities for the field in the next five years?
I would say Env structure with an eye towards defining a target for broadly neutralizing antibody; definition of immune correlates, and the judicious use of the animal model to define them; advancement of the most promising candidates into proof-ofconcept efficacy studies.
What are your hopes for the Global HIV Vaccine Enterprise?
Well, I hope that it provides a common ground for people to come together and become more organized. Hopefully it will allow us to proceed with the highest priority items, scientifically and clinically, and help us to address the most critical needs. Part of the challenge that lies before us is practical: How do we mobilize people for trials? How do we mobilize a diverse group of scientists? The Enterprise really provides a mechanism to catalyze an intellectual and practical response, and to bring muchneeded resources to the field.
How can young investigators be best encouraged to enter into HIV research?
Young investigators are critical to the success of the field. If you accept the idea that this disease is not going to disappear overnight, we need to ensure that the right people are involved as the efficacy trials are done, whether it’s in five years or in 20 years. Young investigators should recognize that there is no more important problem in biomedical research than this problem, for several reasons. Number one, the scientific questions that underlie it are fascinating and they will uncover basic biology relevant to immunology, virology, genetics and evolution. Secondly, by working on this problem you are contributing to an effort that will have perhaps the greatest impact on human health on this planet. It’s important for us to get that message out; this is a unique opportunity to address the scientific challenges and to meet a public health imperative. In an age where scientific trainees are reading in the newspapers about the excitement of stem cell, neurobiology, or genetic research, it’s important to understand that the energy and excitement in our field is perhaps even greater. We are in the midst of an incredible renaissance in the field of HIV vaccines. The science and technology that we can now apply to advance the field is unprecedented.