Bruce Walker, professor of medicine at Harvard Medical School and director of the Ragon Institute of Massachusetts General Hospital, M.I.T. and Harvard, talks about his article in the July issue of Scientific American magazine called "Controlling HIV," about rare individuals who never develop AIDS after being infected by the virus
Let curiosity be your guide, as Scientific American Travel takes you on a journey to faraway ports. For information, go to http://ScientificAmerican.com/travel.
Steve: Welcome to the Scientific American podcast, Science Talk, posted on May 29th, 2012. I'm Steve Mirsky. On this episode:
Walker: He knew that he had been infected in 1978, he felt entirely well, he'd never taken any anti-HIV medications and essentially was asking me, "Why am I still alive?"
Steve: That's Bruce Walker. He's professor of medicine at Harvard Medical School and director of the Ragon Institute, a joint effort of the Massachusetts General Hospital, MIT and Harvard, and he's the author of an article in the July issue of Scientific American magazine called "Controlling HIV", about rare individuals who never developed AIDS after being infected by HIV and what they can teach medical researchers that might benefit the millions of other people who are at risk of AIDS after an HIV infection. We spoke by phone.
Dr. Walker, great to talk to you to today.
Walker: Nice to talk to you.
Steve: So, there's a fascinating article in our July issue about the fact that there are these people who will get infected with HIV and deal with it as if it were any other infection, basically.
Walker: Yeah. I was trained as a physician, and really in the beginning had no interest in doing research. I had done it briefly in college and felt that it was really pretty irrelevant, the things that I was working on. I ended up going to medical school and was quite excited about taking care of patients. When I finished medical school and started my internship at Mass General Hospital, I assumed that the doctors there knew every disease that existed and that it would be just a fantastic place to train. And then something very strange happened. A patient came into the emergency room, who had a disease that none of the doctors there had ever seen before. People thought it was probably some weird genetic variant or something that we would probably never see again. But in essence it was a young man whose immune system seemed to have completely collapsed. So I certainly took note of that case, and then shortly thereafter was meeting with a friend of mine, who was training at another one of the Boston hospitals, and we got to talking and came to the realization that he had seen a patient almost identical to the one that I had seen at Mass General; and that really shocked both of us, and was really the first sense I got that there might actually be a new disease. And it was really then around that time that patients like this started showing up at hospitals around the country, and it became clear that this was a new disease, and it was number of years later that it was identified to be a virus.
Steve: We're talking about what, about 1977–78?
Walker: So, this was in 1980, was when I started my internship, and in 1981, was when I saw this patient. It was really seeing that patient that made me realize that we as physicians on the front lines were going to be the first people to see new diseases. And it made me realize that we didn't do research we were not going to have anything to offer these patients. It was really quite disturbing and depressing at the beginning of the epidemic. We really had nothing to offer these patients. At first we didn't even know what they had; when we finally realized that they were infected with a virus, we had no medications to treat this viral infection.
Steve: You know, you bring up the fact that young people, I mean, if somebody is 20 today, they have lived their entire lives in a different environment; because I remember being in New York City in the '70s, '80s, '90s, I had friends dropping like flies here.
Walker: Yeah. No, I mean it was…
Steve: A different world.
Walker: In those days HIV infection was essentially a death sentence.
Steve: And a fast death sentence.
Walker: Well, interestingly most of the people that we saw—you know, in the beginning we didn't know, we didn't have a diagnostic test for HIV; it was really a clinical diagnosis, somebody whose immune system had collapsed, who had particular risk factors. We knew it was associated with blood transfusions, and it could be transmitted sexually, particularly among groups that seemed to be a particularly high risk, gay men in the U.S. in particular. But we had no idea really what the, you know, what the underlying cause was in order to understand how many people actually had this new disease. We just, we'd see them come in when their immune systems had collapsed, and generally, at that point, the time until death was relatively short. What happened in 1985 was that a blood test became available that allowed us to figure out the number of people who were actually infected but not yet sick. And this is probably an important distinction to make: HIV is a viral infection; AIDS is the advanced stage of HIV infection when the immune system has been largely destroyed, and people end up with things called opportunistic infections, which are infections that normally the body's immune defenses would fight against, but in this situation, where the immune system has been progressively destroyed, things that are normally quite benign, now become lethal. What was happening was that in the hospital, we were getting a skewed representation of what the HIV epidemic was because we were seeing people that were sick enough to come to the hospital. Those that had not yet gotten sick hadn't come yet. When we finally were able to diagnose them, we realized that there was this large population of people that had some evidence of immune dysfunction, but had not yet progressed to AIDS. We still at that point though believed that everybody that was infected would ultimately progress to AIDS, and that really was the general sense that this was a uniformly lethal disease.
Steve: And then in 1995, Bob Massie walks into your office.
Walker: Yes. So, as I said, I'm a physician, I see patients, but I also do research and I see patients in the morning and get blood samples from them and go back to the laboratory in the afternoon and work on those samples to try and understand how the body fights infection, and why it usually loses the battle with HIV. And, pretty much, in 1995 when Bob Massie arrived, I thought everybody ultimately lost their battle. He came in there with a quite different story. His story was that he knew that he had been infected in 1978, he felt entirely well, had never been, he'd never taken any anti-HIV medications, and essentially was asking me, "Why am I still alive?" Every doctor he had seen had told him that you could live with HIV infection essentially as long as he had lived, but you couldn't live longer than that. So he kept expecting that he was going to die; little did he know that—and little did they know—that he was just an extreme outlier in this epidemic. What was really pivotal for me in seeing Bob Massie was that it was the first time I had come face to face with somebody who had clearly documented infection for 17 years, who was entirely well; and the other critical piece of information that we were able to get at that time was a new test had just become available experimentally. Mark Feinberg, who was at MIT, was doing this test which allowed one to quantitate the amount of virus in the bloodstream. And so, you know, initially when I saw Bob I thought, "He probably is not infected and he got misdiagnosis," so I did another HIV antibody test which detects whether the people are infected, and in fact he was infected. And then I sent his blood to Mark Feinberg, who measured the amount of virus in the bloodstream, and it was undetectable. So to me this was just almost unbelievable. Here was somebody, you know, who walked in from a waiting room filled with people with advanced AIDS who was clearly infected, who had been infected for longer than anybody had ever been documented to be infected, and he had no detectable virus in his bloodstream, yet he was clearly infected. And what got me so excited about this was that, you know, the way that the immune system normally takes care of viral infections is, for many of them, you never eradicate the viruses from your body; they remain in your body for the rest of your life but your immune system keeps them in check. So, for example, if you ever have mono or if you've ever had chickenpox, those viruses cause disease when they first attack, but your body mounts a defense against them, your symptoms go away, you then feel entirely well; the virus still exists in sanctuaries in your body, but the immune system keeps it in check so that it can't cause disease. What we knew about those other diseases like chicken pox is that if you become severely immune suppressed, the virus can actually reactivate and cause disease again. So it's very clear that the immune system is actively controlling those viruses. Well, here for the first time for me, I was sitting across the table from a person who was doing this to HIV, something I had never thought possible. And so at that point, you know, even though this was a single patient, it became very clear that we needed to study him, and in fact, that had been his motivation for coming in. He thought that if we were to study him, we might learn something that would be useful. Bob was really an exception. He was doing something that nobody else seemed to be doing. He was a clear outlier, and in medicine it's finding these sorts of outliers that can often give you answers to the people that aren't outliers as to what's going on. I mean, whatever was going on in Bob Massie was holding the virus in check, and it just felt as I was talking to him that the answers were right there in his body as to how the immune system could get the upper hand against HIV.
Steve: And what's also fascinating is he has a hepatitis C infection that's doing what's hepatitis C would normally do. He doesn't have any special abilities to fight off hep C.
Walker: One of the remarkable things that we found is we started to study Bob, was that even though he was controlling HIV exquisitely, he also had become infected with another virus, called hepatitis C virus, which is also transmitted by blood products. Unlike his exquisite control of HIV, he was completely incapable of controlling hepatitis C virus. So, it's not as if he had some, you know, global superpower strength in terms of his immune system, it was something that was in this case very specific to HIV. In fact, what happened in terms of his HCV infection is that he continued to deteriorate and ultimately needed to have a liver transplant because his liver was destroyed by the hepatitis C virus.
Steve: Right. The liver transplant was fortunately successful.
Walker: Well, not only was the liver transplant successful but it did something else that had been a dream of his entire life, which was it got rid of his hemophilia. It turns out that the transplanted liver makes the clotting factors that are missing in somebody with hemophilia, and so he now with his liver transplant, has returned to full health and no longer has hemophilia.
Steve: This guy—the sicker he gets, the better he gets.
Walker: Well, he's doing remarkably, remarkably well. He's just written a book that outlines his journey of resilience against hemophilia, HIV and hepatitis C called A Song in the Night, and I'd recommend it for anybody who's interested in hearing more about the remarkable battles he’s fought and won.
Steve: Amazing. So you want to know how is he doing this?
Walker: Right. So my initial thought when I saw him is that if HIV is following the rules that other viruses follow, other chronic viral infections follow, then he would have to have a vigorous immune defense against HIV that we ought to be able to detect in his body. And we had a simple assay to study that, looking for these cells called killer T cells, which are white cells that are, sort of, like an infantry on a search-and-destroy mission where they find infected cells and kill them. So that would be the case in measles or chicken pox or other infections. So, the question was, did he have these killer cells present? There's no detectable virus in his bloodstream—how's he doing that? Well there must be some evidence of a strong army there, keeping the virus in check. So, we did an assay that was being used, you know, routinely in the lab at that time and found that he had the strongest killer T cell responses of anybody that we'd ever measured. That was pretty exciting, but it still left some questions open. We knew that some people who had AIDS actually also had very strong killer T cell responses, but what they lacked were helper cell responses, which I'll, sort of, refer to as generals. So, if you're going to have a coordinated defense against an enemy like HIV, you need an infantry to actually engage the enemy, and you need generals to help coordinate the attack. The biggest hole in the immune repertoire in HIV infection was the lack of helper cells that are programed to coordinate a defense against HIV. Those cells are generated early in a viral infection, and they're specific for that particular virus infection; it's like on-the-job training for generals. What we had found, the first experiments I'd done 10 years earlier when I first entered the lab, was I was looking for generals programmed to coordinate an attack against HIV. And what I found after studying dozens of patients and I remember these were patients who were coming in to the hospitals, so they all had advanced AIDS. What I found was that they had no detectable generals programed to recognize HIV, and, in fact, that seemed to be the biggest hole in the immune repertoire in infection; which on some level made sense, because we know that HIV preferentially infects activated CD4 cells—which is exactly what these generals are. As they're getting programed, they get activated and that makes them susceptible to infection. So there was a general belief at the time that HIV, because of the nature of the infection, just never elicited these generals. But our thinking was that if Bob's immune response were actually controlling HIV, and if it was due to this really strong infantry we just detected, the only way that it could be effective would be if there were generals present as well. So we dusted off the assay that we'd done 10 years earlier, in which we'd never found any detectable responses, and we used that test on Bob, and found that he not only had these generals programed to coordinate an attack against HIV, he had phenomenal numbers of them. So, this to me was the first time that I actually realized that these cells could even exist, and it all pointed towards his immune system being able to control the virus. But, you know, it also raised really big questions, because, well you know, everybody has an immune system—why isn't everybody controlling? How is that he ends up with this really potent response and is able to control infection? Is it truly due to that? Is it some other factor? What's going on that's really accounting for this and making him different than the vast majority of people who are HIV infected?
Steve: Right. And through a series of advances in the tools that are available, you're able, ultimately, to look for genetics that might explain this situation.
Walker: What we wanted to do was really take an unbiased look and see if by examining people like Bob who control the infection on their own, whether we could find, use the genetic code as something to tell us what was important; to see if there was a genetic, whether it was due to genetic variability that was unique to people who controlled that was accounting for the differences in the ability to control HIV infection.
Steve: There's a great part of the article where you're talking about, one of the challenges as just finding sufficient numbers of these people to do the study with. And you're giving a talk to clinicians and at that point, you just have a couple of these, what we're calling elite HIV controllers, a couple of these people who can really deal with the infection, and you're giving these talk to the clinicians and you raise a question.
Walker: Yeah, so in order to do this kind of genetic testing, it really takes very large groups of people who have the outcome you're trying to examine in order to get an adequate statistical sample. And what we're talking about is generally thousands of patients. And so this really seemed initially not something we could possibly do because we only had a handful of patients like Bob that we'd identified. And there were other people like Steve Deeks at UCSF who had also identified some of these so-called elite controllers. But the idea of finding a thousand or more of these individuals in order to do what's called a genome-wide association study just seemed completely out of reach. And then I had been invited to give a lecture to a bunch of practicing clinicians in New York City, all of whom had big HIV practices, and I just kind of off-handedly, you know, thinking about Bob and telling his story, I said to the audience, “Has any of you ever seen somebody like this, who has an undetectable viral load and has never been treated?" And I was just absolutely shocked when more than half of the hands in the room went up. These physicians in private practice actually were seeing these patients. And it dawned on me that of course that was the case; these guys were not sick, they weren't coming to the hospital, they were still out with their primary care doctors, but they'd been diagnosed with HIV. And as I began then during the break to talk to some of these doctors, they said, "Yeah, the patients would, you know, they keep coming and saying, 'You know, what's going on with me?'" And the doctors would say, "Well, we don't really know, but you're incredibly lucky, and you don't seem to be progressing, and you have an undetectable viral load." So it became immediately clear to me that if we could network among physicians and nurses taking care of patients across the country, that we could probably get to the requisite numbers. Anticipating that as a sample size, if there truly was a strong genetic impact that if we could get a thousand patients who control HIV and compare that to a couple of thousand people who didn't control HIV, then we should be able to get an answer. And so that was really what started to astound the line of a genetic study to try and get an unbiased look. And the idea there, just so that you'll understand how these things are done, it's called the genome-wide association study. There are 3 billion nucleotides in the HIV genome, and it's an incredible task to sequence all of those to look for variability, but it turns out that there are certain areas of the genome where most of the variability occurs. And it can be there're certain nucleotides—building blocks of the genome—that can be often one of two different flavors, and the idea behind the genome wide association study is to look at those, what are called single nucleotide polymorphisms, and you can define most of the variability in the human genome by measuring a million or so of those SNPs. And then to find out are there any SNPs that are consistently one flavor in people that have a certain disease outcome, and the other flavor in people who have the opposite outcome. So, let me give you an example. You want to try and understand why a big dog is big and a small dog is small, so you got a whole bunch of DNA from big dogs and small dogs, you do this genome-wide association study, measuring these SNPs, which is done in an automatic way on a little chip that now measures, you know, easily over a million of these single nucleotide polymorphisms from each dog. And then you ask, Is there any SNP that's consistently one flavor in a big dog and the other flavor in the small dogs? And in fact, what they found when they did that study was there was an SNP that was always one form in a big dog and the opposite form in a small dog, and it lay within the insulin like growth factor gene, which regulates cell growth. So there you have an answer to why big dogs are big and small dogs are small. It ends up that it's a variation energy and that controls cell growth. So what we were trying to do with HIV was to see are there any SNPs that are always one variant in controllers and the opposite variant in progressors.
Steve: Right. So that's the task at hand. Now, so you're able to do that study with your human subjects?
Walker: Yeah, so we recruited, in the end we had about 300 collaborators on this study, and we were able to get, recruit patients from across the country; probably the majority of them came from private practitioners. And we also had helped by recruiting people from Australia and Europe and South America that, through physicians there who had come across these same kinds of patients. We relaxed the definition a little bit for the study. We said anybody with a viral load of below 2000 copies, we would consider a controller and anybody with a viral load of above 10,000 copies, we would consider a progressor; and so that was able to let us stratify people into two groups. It was very easy to get the progressors because there were so many people progressing; that happened rapidly. But we were able then to get blood samples drawn at all these different offices across the country and then shipped to Boston where we processed them and then subjected them to the SNP analysis, and when we did that we got, what I thought was a really remarkable result. We got this incredibly strong signal in one gene, in one chromosome—chromosome 6—and nothing in the T cells. So, you know, we had initially thought, "Well maybe there are a bunch of different reasons why people do well; some do well for this reasons, others for that reason." But here was some genetic information that suggested there's something very similar in all of these people, and it mapped to the chromosome really encodes for a lot of different immune functions. So, to us that was really encouraging, because it suggested that, in fact, it was something about the immune system, genetically determined, that was accounting for better control. And it gave us now a chance to try and figure out what that was. Now we had a few, well 300 SNPs that all showed up as positive, but there's something called linkage disequilibrium which confounds the SNP analysis to some degree. What these SNP do is they tell your variability to a single locus; but because genetic material travels in groups over times, what these SNPs are able to do is they're able to also tag variation at other sites so that, you know, that when you get a positive snip you know it's either that SNP at that nucleotide itself or one of the other nucleotides that always co-varies with the SNP that you're measuring. And so, we knew what these 300 that probably not all of them were independently significant, but probably some of them were tagging the same variant that was actually causal. So, we did what's called step-wise regression analysis and in the end found out that there were only four that were independently associated with control; but we still only knew that of those four, that it was either the SNP that was actually being measured or one of the variants that it tags that was actually causal. So, we needed to go further, and that meant that we needed to sequence the entire genetic code in that region, in order to understand what it was that was truly always different in controllers versus progressors.
Steve: Right, so you go ahead, you're finally are able to do that.
Walker: Well, the 'finally able to do that' has a bit of an interesting twist to it as well, because medical students in the program at Boston often do a year or two of research during medical school. We had an incredibly talented medical student who joined our team, Sherman Jia, and Sherman on his own figured out that he could use existing sequence data and SNP data from lots of other studies that have been done in order to impute the sequence through the entire HLA locus through this entire region of chromosome 6 that were showing up as positive. And so he did that, and the remarkable finding there was that it really came down to, of the 3 billion nucleotides in the human genome, it really came down to just a handful of amino acids that the variability, you know, the different flavors of those amino acids, was what was accounting for more rapid progression in some people and slow progression in others.
Steve: The amino acids in the expressed proteins.
Walker: Well, what he did was, once we got the genetic code, he was able to determine the genetic code and then translate that into proteins, and that's what led us to these specific amino acids that were associated with control. And what was remarkable about this was that it was amino acids that line the binding groove of something called the HLA class I molecule. HLA molecules are on the surface of cells and they basically serve a function. Normally they would contain self-peptides derived from the cytoplasm of the cell in the peptide binding groove, but if a virus comes in and infects the cell, that self-peptide gets replaced with a viral peptide, and the immune system, when it sees a viral peptide in an HLA molecule, it knows that something bad is going on inside that cell, and it basically then seeks out and destroys those cells. So that's what was being pointed to by this observation that the peptides, it was something about how the peptide was sitting in that groove that was making those infected cells highly visible to the immune system of people like Bob Massie.
Steve: So is this seemingly a minor kind of steric effect?
Walker: Yeah, and I want to stress that as with all studies, you know, you learn some things and the studies generate more questions. We know that it's something about how the peptide is sitting there that's making the difference; what exactly that difference is we don't yet know, but that's what we're [trying to figure out]—but we know where to look now and know how to understand that. And, in fact, in other studies that we've just completed, it seems that what happens is that it's the way that the peptide sits in that groove [that] allows it to be seen by the T-cell receptor on these killer cells and allows for very efficient killing.
Steve: So, you know, that's all really fascinating; it's an amazing story. But the bottom line still is how can this knowledge be applied to all the people who are not controlling the HIV?
Walker: Well that's the most important thing for people to realize is that although we've figured out that there's a mechanism by which HIV can be controlled in people, the question is how do you take that information and now turn it into something that can make everybody into an HIV controller? And we don't yet know how to do that. I mean, God knows, we all wish we did, but we know that HIV can be controlled by the immune system. We know the specific interaction that's accounting for that, and we now know where to focus our efforts to try and come up with ways to be able to elicit that kind of immune response in people who don't naturally have it. And I think this is the stepwise nature of scientific discovery. You know, what we're seeing with HIV is that's it slowly revealing its secrets. This is another secret that it's now revealed, and we're marching forward with a lot more focus, now that we know exactly what the controlling mechanisms are. I should also say that there are lots of other things that probably play into this. You know, the immune system works as a system so this is a key part of it. But understanding this in the end as a mechanism that appears to be accounting for control allows us now to think about not just this particular interaction, but the other parts of the immune system that feed into making the system work that can hopefully tweak it in such a way as to improve the efficiency of this in this particular response in everybody.
Steve: Bruce Walker's article, "Controlling HIV," is in the July issue of Scientific American magazine. That's it for this episode. Get your science news at our Web site, http://www.ScientificAmerican.com where you can find info about the upcoming transit of Venus across the sun, coming your way on June 5th or 6th depending on where you are on our little planet. And follow us on Twitter, where you'll get a tweet every time a new article hits the Web site. Our Twitter name is @sciam. For Scientific American's Science Talk, I'm Steve Mirsky. Thanks for clicking on us.