Living History Series: As Antibody Findings Mount, What Comes Next?
Four leading researchers assess the field and what it will take to get from broadly neutralizing antibodies to immunogens
In 2009, the first report of new and more potent HIV-specific broadly neutralizing antibodies kicked off something of an antibody frenzy. Since then, researchers have isolated nearly two dozen broadly neutralizing antibodies from HIV-infected donors. These antibodies are vital clues for HIV vaccine development and are now being used by researchers to try to reverse engineer vaccine immunogens. In this second installment of IAVI Report’s Living History of AIDS Vaccine Research, Managing Editor Kristen Jill Kresge and Science Writer Regina McEnery turned to four experts to provide some perspective on these recent advances in antibody research and to frame the current efforts to design vaccine candidates that can induce antibodies against HIV.
DENNIS BURTON is a professor of immunology and microbial science at The Scripps Research Institute (TSRI) in La Jolla. Burton oversees IAVI's Neutralizing Antibody Center at TSRI, which figured prominently in the discovery of several of the new broadly neutralizing antibodies, including two called PG9 and PG16.
JOHN MASCOLA is the deputy director of the Vaccine Research Center at the US National Institute of Allergy and Infectious Diseases, a division of the US National Institutes of Health. He was part of a team of researchers that isolated a few of the new broadly neutralizing antibodies, including VRC01.
NELSON MICHAEL is the director of the US Military HIV Research Program, a key collaborator in the RV144 efficacy trial in Thailand of a prime-boost vaccine regimen (a canarypox vector based candidate, ALVAC-HIV, followed by a recombinant protein, AIDSVAX gp120 B/E) that demonstrated about 31% efficacy in preventing HIV transmission.
JULIE OVERBAUGH is the associate program head at the Fred Hutchinson Cancer Research Center in Seattle. She studies the biology of HIV and transmission of the virus, particularly vertical transmission from a mother to her infant.
Do you feel more optimistic now about AIDS vaccine development? If so, why?
Burton: I feel much more optimistic about the possibilities for an AIDS vaccine than I have for probably a decade. This is primarily because we’re seeing a lot more of the sorts of antibodies that we’d like to induce through a vaccine than we’ve ever seen before. For 10 to 15 years we’ve been working with a handful of antibodies and we really thought maybe they were a very special beast and we would not be able to induce them via a vaccine. But in the last year or so we’ve identified, between the different groups, something like 20 new such antibodies, and they’re a vital clue to how to make an AIDS vaccine. I think this is a tipping point in a way. The next two, three, four years will tell us a lot about the feasibility of an HIV vaccine.
Mascola: I’m absolutely more optimistic. I think the isolation of numerous monoclonal antibodies, the fact that the immune system does this pretty routinely, and the fact that we have much better tools to measure what we’re doing, all suggest to me that the problem is solvable in due time. In science, you just can’t predict how long that is, but I would say I’m quite optimistic now.
What is special about the new antibodies that have been isolated recently from HIV-infected donors?
Burton: The most significant factor is that these new antibodies are probably 100 times better than the previous broadly neutralizing antibodies. Some of them hit 90% of the world’s [HIV] isolates. They’re also very potent, and the more potent they are, then the less of the antibody we’ll have to induce through vaccination, so we like potency a great deal.
Mascola: What makes VRC01 [an antibody discovered by researchers at the Vaccine Research Center] special is what we call a precise mode of targeting. It finds the spot on the virus that is unable to really mutate or change because it’s the initial part of the virus that binds CD4. Therefore, this site has to be exposed and the antibody is able to access it. The fact that we can define this very precise region of the virus that is attacked by the antibody means we can express that very precise region in a number of ways, and so it gives us the opportunity to try a number of different approaches to vaccine design. While we understand the structure very well, when it comes to knowing what the immune system’s going to do, the science really is in the experimental testing.
So how do you get from these antibodies to vaccine immunogens?
Burton: A number of different strategies are being pursued to try and work backwards from the antibodies to immunogens, a process we call reverse engineering. We are taking pieces of the virus that we see are recognized by the antibodies and using those. We are altering the surface of the virus proteins, or the surface of the virus even, based on what we know about how the antibodies bind to the virus.
The target of all neutralizing antibodies against HIV is a complex of proteins on the surface of the virus known as the trimer. This is quite an instable structure, which is one of the problems in trying to make a vaccine. One way around this is to forget about the trimer altogether and to use pieces of the trimer that are recognized by the broadly neutralizing antibodies. An alternative is to really try and make the trimer. If you do it right, if you can really stabilize the trimer, then all the broadly neutralizing antibodies should bind the trimer and in theory [the trimer should] elicit all of the relevant antibodies. Stablizing the trimer is a great problem. Perhaps one of the best advances in trying to stabilize the trimer would be if we had its structure, but we’re in a sort of vicious circle here. We can’t get its structure because we can’t get it stable enough, and we can’t get it stable enough because we can’t get the structure. So now people are trying to bootstrap their way towards the structure.
How challenging is it to reverse engineer immunogens based on these new broadly neutralizing antibodies?
Mascola: There certainly have been examples of vaccines that have been thoughtfully and rationally designed, but we are in uncharted territory in the level of sophistication, I think, that we have to attain to really rationally approach the types of immunogens we need, and the type of vaccine strategies we need. I think it’s a tougher problem than has been approached before.
Burton: Reverse engineering of this type really started largely out of work with HIV. One of the difficulties with the whole reverse engineering strategy is it seems these antibodies are often quite picky about how they recognize the virus, so it means that the immunogens may have to be designed quite precisely and that’s very challenging.
The hope with the new antibodies is that they are more effective so that might give us a little bit more leeway in how well we design the immunogens. But we don’t have the sites defined well enough at this moment to say whether immunogen design is going to be easier, the same, or harder for the new antibodies. I think that’s one of the issues that’s going to be resolved in the next two years or so. We do know that quite a few different infected individuals have made antibodies to these sites, so they’re not completely intractable. That’s cause for hope but there are many, many unknowns still that we need to work out.
If reverse engineering could be made to work, I think it would be very exciting. It would be tremendous for vaccines generally because it would reduce the problem to simply finding some good antibodies and then working backwards.
What is known about whether any of the broadly neutralizing antibodies can actually protect against HIV?
Burton: We believe that they would protect against HIV because they do so in the monkey model. These antibodies, once individuals are infected, don’t seem to affect the course of disease very much. They work much better if they’re present before the virus.
Overbaugh: There have certainly been a lot of examples in nonhuman primates that you can protect against HIV using HIV-specific monoclonal antibodies in very, very experimental settings. But in humans, the data is pretty sparse. Probably the best data comes from mother-infant transmission studies. There’s only been one study done in infants that I’m aware of, but that’s the closest thing to a vaccine setting. And in that study that we recently completed, which is not yet published, we didn’t see any evidence, no matter how we looked, of a benefit of having those antibodies in the infants. It doesn’t preclude the idea that neutralizing antibodies could work, if they’re the right ones, if they’re broad enough, if they’re targeted to the right viruses, and if there’s enough of them. But I think it says that this bar is actually quite high in terms of eliciting neutralizing antibody responses.
Mascola: What it will take for an antibody to protect is something we need to keep a very open mind about. I think for an optimally effective vaccine—for a vaccine that really does prevent most infections, most of the time, against most viruses—it’s likely to require some level of virus neutralization. But we don’t really know what level of neutralization is going to be required in humans.
So are passive immunization studies with the new antibodies something the field should consider?
Mascola: I think the idea of getting proof of concept in humans that antibodies can protect is critical. It would provide sort of a framework for the field to say we know these types of antibodies protect and therefore we really have to work at eliciting them. We have had a lot of discussions at the Vaccine Research Center with a whole group of colleagues in the field about the potential for making clinical grade antibody and for doing clinical trials. Right now, it is our intention to pursue that course and to work with two groups of clinical cohorts, potentially. One would be high-risk adults and the other is the setting of maternal-to-child transmission of HIV.
We would probably expect this approach to work but if it didn’t, it would maybe make us rethink some of the conventional wisdom about amounts of antibody and mucosal immune responses, for example.
Overbaugh: I think testing those antibodies in adults who are highly exposed or in pregnant mothers is an interesting way to examine the question. I would say that doing that in the setting of mother-to-child transmission is now very complicated because we know that taking antiretroviral drugs, either mother or infant, is highly protective. So that will be the benchmark against which these antibodies would have to be tested, and that might make for a very large trial to be able to see any efficacy. But in high-risk populations it may be interesting to look at those antibodies perhaps in mixtures. I don’t think we really know how much antibody is even likely to be protective.
Is it possible that antibodies explain the partial protection seen with the prime-boost vaccine regimen tested in the RV144 trial in Thailand?
Michael: We certainly have confirmed in RV144 that we don’t generate very much in the way of broadly, cross-reactive neutralizing antibodies, which is generally seen in our field as the Holy Grail. What we did see [in RV144] is a very strong amount of binding antibody and a very high level of what we call ADCC [antibody-dependent cellular cytotoxicity]. ADCC is an important effector mechanism of antibodies. If I had my druthers, I would have much rather seen that plus neutralizing antibodies because I would wager that if we had a vaccine that could do that, we’d all jump on it.
So do the RV144 results indicate that non-neutralizing antibodies can play an important role in protecting against HIV?
Burton: There is some emerging evidence that under certain circumstances, at high concentrations, non-neutralizing antibodies can impart some partial protection. This may be contributing to the results of the Thai trial, which do seem to indicate some very, very modest protection. So there is interest that maybe mechanisms other than neutralization can contribute to protection. As to whether they really will, under normal circumstances, is a topic for investigation.
I think it’s worth investigating the protective activities of non-neutralizing antibodies. It is not a new phenomenon to find non-neutralizing antibodies can act against viruses and afford some protection. Generally, the protection afforded by such antibodies is much less effective than neutralizing antibodies.
Do you think it would be possible to develop a highly effective HIV vaccine without inducing broadly neutralizing antibodies?
Burton: I think many of us feel that without broadly neutralizing antibodies, a vaccine would be difficult, if not outright impossible. It’s for good reason that virologists focus on neutralizing antibodies. But of course if all the problems prove intractable then there’s a case to be made for saying, well, okay, we have failed on the neutralizing, but let’s try and induce a high level of non-neutralizing antibodies.
Overbaugh: I still think we’re going to need multiple components to this. Neutralizing antibodies that are very, very broad and are targeted to the best possible epitopes may be part of it. My gut feeling is that will not be adequate and there will have to be something else that will have to contribute to vaccine efficacy.
Michael: I think it’s going to be difficult for us to understand what it really is going to take, frankly, to make a vaccine that generates broadly, cross-reacting neutralizing antibodies. Some of the best minds in the field are on that. I would say that it’s exciting that maybe we can develop a licensed vaccine that will hold the line for that blessed moment when we’re able to reverse engineer vaccines that will actually make broadly neutralizing antibodies. My view is that, once we’re able to do that, you’re going to be looking at efficacies that are very, very high and it will be that wonderful part of my career when I can think about working on malaria.