An Interview with Barton Haynes
A new virtual Center
Bart Haynes, MD, has been interested in a wide spectrum of immunology during his research career, from autoimmunity—especially rheumatoid arthritis—to thymus biology and thymus transplantation as a curative treatment for DiGeorge syndrome, to HIV soon after it was identified as the causative agent of AIDS in the 1980s. He has long been affiliated with Duke University, North Carolina, having completed his residency training after gaining his medical degree at Baylor College of Medicine, Houston. Haynes began his research career under the mentorship of Sheldon Wolff and Anthony Fauci at the National Institute of Allergy and Infectious Diseases (NIAID), part of the National Institutes of Health (NIH), before returning to join the faculty at Duke in 1980. There he has held a number of senior positions, including chief of the Division of Rheumatology, Allergy, and Clinical Immunology (1987-1995), chair of the Department of Medicine (1995-2002), and is currently director of the Duke University Human Vaccine Institute. Haynes has also served as chair of the NIAID AIDS Vaccine Research Working Group that advises the NIH.
Haynes leads a team that in July was awarded the prestigious Center for HIV/AIDS Vaccine Immunology (CHAVI) grant from NIAID. CHAVI is a 'virtual consortium' that brings together researchers in a close collaborative effort that is hoped will speed the progress towards an AIDS vaccine by addressing key immunological questions in HIV infection, and design, develop, and test novel vaccine candidates. It will receive more than US$300 million over seven years. Haynes spoke recently to IAVI Report Editor Simon Noble about his vision for CHAVI.
What are your initial priorities with CHAVI, and what have you been able to achieve so far?
Our first priority is to establish a functional organization and I think we've made great strides so far. One of the driving forces of the Global HIV/AIDS Vaccine Enterprise, when Rick Klausner, Tony Fauci, Gary Nabel, myself, and others wrote the commentary proposing the concept in 2003 (Science 300, 2036, 2003), was that we were all frustrated at the slow tempo of research at the time and the limited progress that had been made towards a practical vaccine. CHAVI is the NIH's component of the Vaccine Enterprise.
The overall goals of CHAVI are: to elucidate early viral and immunological events and host genetic factors associated with HIV-1 transmission, establishment of productive infection, and (partial) containment of virus replication; to determine correlates of SIV immune protection in nonhuman primates; to design, develop, and test novel immunogens and adjuvants that elicit persistent mucosal and/or systemic immune responses in humans and nonhuman primates; and to evaluate HIV vaccine candidates in early phase clinical trials. Our initial Scientific Leadership Group is Joe Sodroski and Norm Letvin from Harvard, George Shaw from University of Alabama at Birmingham (UAB) and Andrew McMichael from Oxford.
Now, there are two ways that you can set up a very large, complicated organization like CHAVI. The traditional way is to set up sequentially, one component at a time. When we learned that we had been awarded the CHAVI grant we decided to do everything in parallel to speed progress. So the first people I hired were an intellectual property (IP) person and a lawyer to incorporate the IP issues into the research plan that we had submitted in application. Then we immediately began writing clinical protocols based upon our research plan, and we started dealing with all of the related QA/QC issues of both clinical trial design and clinical sample assays. We immediately contracted with Family Health International for site management of our clinical programs. We also contracted with SCHARP and Frontier Sciences to develop our patient, sample, and central relational databases that can process not only the samples from the patients and their clinical data, but relate data coming from the host and virus genetic studies, and then all of the functional assays that are key to each patient and each time of blood draw. We should be functional and begin to acquire patient samples in the first quarter of '06.
We've had a series of CHAVI strategic planning sessions to discuss the possibility of mucosal and innate immunity studies, and also for laboratory test standard operating procedure (SOP) development. We've also written four clinical protocols to date on acute HIV infection patients and exposed uninfected subjects and we're in the process of getting the necessary approvals for implementation in early 2006. NIH staff have been wonderfully supportive in moving our clinical protocols through the review process.
It's been quite chaotic but I told everybody from the beginning that out of this controlled chaos will come both creativity and an accelerated way of working. At our organizational meeting in August there were 166 people there. I brought not only the scientists but also all the administrators from our scientific sites and from the NIH, and I told the administrators that they are going to be just as critical to CHAVI's success as the scientists.
The CHAVI RFA (request for applications) emphasized two scientific priorities identified by the Vaccine Enterprise: early immunologic and virologic events in HIV-infected/exposed people, and correlates of immune protection from SIV infection in nonhuman primate models. How do you plan to study early events?
We're going to look at the very earliest stages of acute HIV infection using a new pooling strategy of samples from STD clinics developed by Mike Cohen, Chris Pilcher, and Joe Eron and colleagues at University of North Carolina (UNC). The pooling strategy screens samples with RNA PCR to detect virus positive, antibody negative samples at the very early stages, hopefully even when virus load is still increasing. This is one of the strategies that differentiate CHAVI's programs, to find patients at the very earliest stages of HIV infection.
For the studies looking at the ontogeny of T and B cell responses with changes in viral load we'll need sufficient numbers of patients within the first two to four weeks of infection to really get an idea of what the transmission event is like, what the effects are on the host's innate and adaptive immune responses, and to map out what happens in those people that control the virus versus those that aren’t able to initially control.
For this we need a series of collaborating sites that can establish this screening protocol to identify patients immediately after the transmission event, so Mike Cohen has begun to establish collaborations with investigators in Durban (Salim Karim) and Johannesburg (Helen Rees and Wendy Stevens) in South Africa, Lilongwe (Mike Cohen) and Blantyre (Taha Taha) in Malawi, Entebbe (Pontiano Kaleebu and Heiner Grosskurth) in Uganda, and Moshi (Saidi Kapiga, John Shao, John Crump and John Bartlett) in Tanzania, as well as in London (Sarah Fidler and Phillipa Easterbrook) and the US (UNC—Joe Eron, Chris Pilcher; Duke—Charles Hicks). These collaborations will provide the samples for the work that CHAVI investigators will perform. We're collaborating with existing robust sites that are already set up to do this type of accrual work. As many as possible of the studies will be performed on site. We are also talking to other networks about the possibility of collaboration as well. The less CHAVI has to duplicate, the more time and resources we can spend on vaccine development.
We've also canvassed all of our collaborators and found that we have about 70 retrospective samples that are of the quality that we plan to collect prospectively. We're just beginning now to work on these, to set up assays, and to begin to work out all of the problems of data analysis, standardization, SOPs.
We're collaborating with IAVI on one of our first studies, CHAVI 002, on HIV exposed and uninfected individuals. A pilot group of patient samples will be tested by Andrew McMichael in Oxford, and will be simultaneously studied by Jill Gilmour in the London IAVI laboratory. IAVI is being a great help in developing the laboratory procedures standardization, because they have previous experience. It's a good example of how CHAVI is going to work, collaborate and synergize with folks who are doing things already that are helpful to the field, so that we don't waste either time or resources reinventing the wheel where it's not necessary.
How does CHAVI plan to investigate the correlates of immune protection in the SIV/macaque model?
In our original application this study was not expanded to the degree that we ultimately planned that it would be because of limitations in the original grant. So in the second year of the grant, Norm Letvin with multiple collaborators is now putting together an expanded program that will focus on two areas; one is to determine the correlates of protection in live attenuated SIV infection and the other is to study host-virus interactions in dual infections, and to study these in tandem. When live attenuated infection is protective, what are the immune correlates of that protection? And when superinfection occurs, what has the first infection done to the immune system to allow the second infection? That program is just being worked out, and so we’ll have more details in a couple of months.
So you're pretty pleased with progress so far?
I'm thrilled with progress so far, because everyone has accepted this philosophy and has dealt marvelously with the controlled chaos that has come out of trying to do things differently. I think that's one of the things that CHAVI can contribute to the field, to figure out how to do things in a quicker, more effective way. I told everyone that the secret to success here is that everyone puts this organization on their back and when something comes across their desk that needs to be done for CHAVI, they stop what they're doing and get it off their desk, don't let it languish. Immediately move it. The constant challenge is going to be to maintain that momentum over time, but so far I'm pleased with the early stages of establishing a functional organization.
It's interesting that you identified IP as an initial priority. Has that been a bottleneck on progress in previous collaborations, the willingness and ability of people to share reagents and ideas?
Absolutely, the exchange of reagents and the whole issue of whether one can use certain strategies and what have you is always a traditional bottleneck. We already have a CHAVI materials transfer agreement that we can sign once and then enhance the flow of reagents among all of our investigators. Secondly, our IP manager and legal counsel together are currently canvassing all CHAVI investigators and creating an inventory of the existing CHAVI IP portfolio. Number three, we've begun educating all CHAVI investigators on how to protect future IP, so that we have control over it and can make it available to whoever needs it in order to optimally speed the development of a vaccine, should we be so fortunate to be in that position.
There are currently six Discovery Teams in CHAVI: immunology, vaccine design, adjuvant development, viral biology, structural biology, host genetics and genomics. New Discovery Teams will be set up after the first year—what kinds of subject areas do you see being included?
For year 02, we are considering new teams in mucosal immunity, innate immunity, and computational biology and mathematical modeling. This year, Ray Dolin at Harvard is working to design what our first clinical trials might be in year 03, and David Goldstein at Duke is working to expand the CHAVI host genetics programs. Finally, for year 03 we are also considering a mother-to-child transmission (MTCT) team.
What's the primary goal of the computational biology and mathematical modeling team?
It's three-fold. One is to have a team that can analyze the hugely complex data sets that will come out of our acute HIV infection and exposed uninfected observational studies. Secondly, to be able to model the human immune system with regard to observed responses and to predict how vaccines will interact with the immune system based upon responses seen in those individuals that can control the virus. And then third, to help us with vaccine design, particularly focused on viral genetics. The mathematical modeling group will assist Beatrice Hahn at UAB in viral genetics, and also will work with Joe Sodroski and Steve Harrison at Harvard in their structural analysis of the transmitted HIV envelope trimer.
We're currently in discussion with various groups with regard to who might participate. A group that's already on board that has been a tremendous help to CHAVI investigators is the Los Alamos Mathematical Biology and Computational Immunology team, led by Bette Korber and Tanmoy Bhattacharya. I've been working with Bette for over 15 years on immunogen design using computational biology techniques.
One of the outcomes of the CHAVI acute HIV infection observational studies will be a transmitted virus database of full-length sequences of the virus that gets transmitted across the bottleneck at mucosal sites. I think these kinds of sequence analyses are really the only way to design immunogens that can induce T-cell responses that have a chance of dealing with virus diversity, as well as resolving whether transmitted viruses are different to the viruses that evolve later in HIV infection.
If a MTCT team is added to CHAVI what will be its research priorities?
We believe that it's important to ultimately develop a vaccine that will be practical for children as well as adults. MTCT is quite complicated because there are three points of transmission: in utero, peripartum, and during breast feeding. So the first fundamental questions are: Is the biology of transmission at those three time points the same or different? Exactly what kind of vaccine would need to be given to a mother so that either she will be protected or, if she becomes infected, her child will be protected? A separate issue is developing childhood vaccines that would be effective, given the differences in a child’s versus an adult’s immune system.
In year 02 we'll start a strategic planning process to determine whether CHAVI has the resources to address MTCT, but personally I believe this is a very important area to consider. There are so many things that one could study. Whether it's MTCT, innate immunity, adaptive immunity, one of CHAVI's tasks is to decide what the key questions are in all of these areas. Yes, there are many interesting questions, but are they the most relevant questions for vaccine development?
Which vaccine vectors will CHAVI focus on and why?
One of the charges to CHAVI is to develop vectors that induce long-lasting, protective, mucosal immunity. During our strategic planning process we asked which vectors and inserts had the greatest promise for doing that, and initially we will focus on attenuated recombinant VSV [vesicular stomatitis virus], chimeric adenovirus, and recombinant mycobacteria.
We'll have a team working on mucosal assay development and standardization, which we think is critical to collecting the high quality data needed in order to evaluate mucosal vaccines. We'll be combining mucosal and systemic immunogenicity studies for a vector analysis comparing this group of vectors containing the same insert. Then we will compare a concomitant group of inserts in the same vector, and determine which inserts best induce broad neutralizing-antibody and T-cell responses.
VSV can induce robust T- and B-cell responses, similar to adenovirus, and in some studies can induce mucosal as well as systemic responses. Also, VSV can be administered intranasally and at other mucosal sites, and we’ll be evaluating these. Of course the problem with VSV has been the concern of neurotoxicity but we'll be working with those who have developed some new attenuated strains that may well be less neurotoxic, to evaluate them for their immunogenicity.
We're interested in mycobacteria because of the duration of responses that BCG can induce, because BCG and other mycobacteria have been administered orally for many years and are generally safe, and because millions of children get BCG soon after birth. I've had an NIH grant with Bill Jacobs and Norm Letvin for three years to develop a multivalent mycobacterial-vector vaccine incorporating HIV genes for a combined TB/HIV vaccine. That work will now be coordinated with CHAVI work.
What are the major practical and theoretical hurdles in studying/measuring mucosal immune responses?
The practical problems are getting samples in the first place, and then ensuring that they are informative and in sufficient volumes. Theoretically, the key is understanding what the infectious unit is—cell-associated or cell-free virus—the role of local versus systemic immunity in protection at mucosal sites, the first cell or cells infected at transmission, and the nature of the infecting HIV quasispecies. To name just a few.
There's been a drive for some years now to standardize reagents and assays, to enable comparison of results across experimental systems and research groups. Is that going to be a driving force within CHAVI?
There are folks who are already doing this for existing organizations. David Montefiori is doing this with regard to neutralizing antibodies at the HVTN, and so David is leading this effort for CHAVI. Rick Koup is already doing this with regard to T-cell assays for the VRC, and he is working with Clive Gray, Kent Weinhold, and Guido Ferrari at the HVTN, and they're all now also working with CHAVI. Steve Self at SCHARP is working with the Vaccine Enterprise, HVTN, and now CHAVI for database development and statistical support. So by picking key individuals who are already doing this for other key organizations, we hope to speed the harmonization of what CHAVI does and, again, not reinvent the wheel.
The CHAVI grant application required a detailed research plan. What is the process for expanding or changing the scope of that research plan?
We can go out and seek new collaborations and discovery teams, and other folks that think they have ideas or cohorts to offer can contact us. I think that the community needs to view CHAVI not as a funding agency but as a "company" charged with studying the correlates of protective immunity to HIV and developing new vectors for induction of long lasting mucosal immunity, to speed HIV vaccine development. There are many exciting things scientifically that can be done, but we have been tasked to focus on those things that we believe will speed vaccine development.
Will there be a mechanism to consider ideas from outside of the CHAVI researchers?
Yes, AIDS researchers can contact CHAVI with ideas using a two page letter of interest. CHAVI has a defined mechanism to vet new ideas for new discovery teams culminating in review by our CHAVI Executive Committee that is comprised of NIAID staff and our Scientific Advisory Board.
How much of the CHAVI funding, about $49 million/year, will be available as grants to researchers outside of CHAVI?
We don't yet know the status of our year 02 budget, so it's unknown at present. But any funds expended on new programs or investigators that were not in the original grant will be done in a manner such that the new investigators will become members of the CHAVI team, and work to move the "company's" work forward with the same research agreement and policies that initial CHAVI investigators signed on to.
How can young investigators be attracted to HIV research, and can CHAVI play a role?
Absolutely, CHAVI plans to bend over backwards to help the field, including helping to build enthusiasm and to help train the next generation of AIDS vaccine development researchers. A critical issue is how to bring young people into CHAVI teams and convince them that their work will benefit their careers. We are already working with Duke department chairs and promotions committees to educate them of the value of working in research teams at universities, and of recognizing the contributions of young investigators for the work they do in teams. We have to have young people come to this field to bring in the new ideas that we hope will help solve the AIDS vaccine problem.