| || |
| || |
| Edward Berger, PhD |
Interview with Edward Berger, PhD Berger is chief of the Molecular Structure Section in the laboratory of viral diseases in NIAID and NIH. Berger's labidentified the first HIV co-receptor, a molecule that Dr. Berger and his colleagues dubbed "fusin." They showed that fusin must be present on the surface of CD4+ T cells in order for HIV to enter and infect these cells. Soon thereafter, Berger's group and others showed that other HIV strains use different co-receptors to gain entry into target cells. Many of these molecules ordinarily function as receptors for chemokines, proteins that help orchestrate immune responses.
How does the HIV virus get into a cell?
The way HIV virus gets into cells is that the membrane on the virus touches the membrane of the cell and then the two membranes fuse together, and the virus introduces its genetic material into the cell. The way the virus fuses with the membrane of the target cell is because it has a protein on its surface called the "envelope protein" whose sole purpose is to get that virus to fuse with the cell.
We've been studying the mechanism, the biochemistry of how that fusion reaction works, how the envelope protein works, and what has to be on the cell in order for the envelope protein to be able to get into that cell because the HIV virus doesn't go into every type of cell in the body. It goes preferentially into a certain class of cells that we call "lymphocytes."
Does the envelope change shapes?
What I used to say back in '96 when we first came out with these discoveries is that it had been known for a long time that the virus has to interact with a particular molecule on the surface of the target cell. If you can imagine a molecule sticking out, it has a certain shape, and the envelope protein of the virus recognizes that shape and somehow that triggers the envelope protein to do its fusion reaction.
But it turns out to be more complicated than that in the sense that the virus needs not only the first receptor, but an additional receptor that we call a "coreceptor" in order to be able to carry out its fusion reaction. The way HIV has chosen to do that is that it hides the part of its envelope that needs to bind to the second receptor so this is what that part is and it's hidden here, and then when the virus finds a cell that has CD4, the first receptor, it binds, it changes its shape, and it now exposes this part. And now this part can bind to that coreceptor.
The reason it's really important is because the HIV virus has to survive in the body at a time when the immune system is throwing everything it can at it, and one of the things that the immune system is trying to make to protect against the virus are antibodies which are proteins that recognize specific shapes on other proteins.
It turns out that this region is very critical and the virus can't afford to change because it has to bind CD4. The reason antibodies can't get in there is because this space is too narrow and the antibody just can't fit in; i.e., the antibodies don't work against this is because this is hidden. It only becomes exposed when the virus comes along and finds the cell that has CD4, changes its shape and now the antibody has to rush in there and bind before this binds to the coreceptor and once it bonds to the coreceptor it triggers the fusion reaction and it's too late [for the antibody].
Can you provide an analogy to this process?
We can think of the virus as a small soap bubble that has the genetic material inside it that will tell the cell to make more HIV. So in order to be able to do this it has to be able to get into the cell.
Now the cell is also a big bubble that's surrounded by a membrane, so in order for the HIV virus to get into that cell, the membrane of the virus has to fuse with the membrane of the cell. If you can imagine a big soap bubble here, coming into contact with a little cell bubble, those membranes have to fuse so they now basically become one and now the genetic material from the virus can go inside the cell.
Is the CD4+ the thing that attracts the HIV virus?
CD4 is a protein. There are certain cells of the immune system, called "CD4 lymphocytes," that play a very important function in the immune system. They're critical players in the immune system and the HIV virus preferentially attacks the CD4-positive T cells. Over time, in the person who's infected with HIV, the CD4 cell counts decline and they get so low that the immune system doesn't function well anymore and can't fight off the virus.
It turns out that the virus actually uses the CD4 protein as the first receptor to which it binds in order to get entry into the cell, but CD4 alone is not enough for the virus to get in. There has to be an additional protein on the surface of the cell in order for the virus to be able to fuse and enter the cell.
Why are we looking at different ways to fight this virus?
In the past five or six years the field of HIV therapy has been radically transformed by the development of inhibitors of the virus replication cycle. These target certain processes of certain steps in the virus lifecycle that are absolutely critical. Many people who take these drugs see dramatic improvements in their health: their immune systems get restored to significant function. Their CD4 cells come back and in many cases they can resume fairly normal lives, but they have to keep taking these drugs.
There are several problems associated with this particular regimen. One is the fact that you have to take multiple drugs and you have to take them multiple times a day, so it's a difficult regimen to adhere to. And if the virus in the body is constantly randomly changing, and if it randomly changes in such a way that it is no longer inhibited by that drug, then that virus is now going to have an advantage and is going to replicate and take over.
You always need to keep the concentrations of drugs very high so that you completely suppress the virus replicating, because if you don't do that then the virus will start replicating and it will start throwing off variants and you're likely to get a drug resisting variant that will take over.
So the practical problem is that for a person who's infected with HIV who is responding well to these drugs, you have to adhere to them and it's a very difficult regimen. The other critical issue is that the drugs now deep in use for several years are starting to show side effect profiles that are very unpleasant and unacceptable for many people. So that's a second problem.
The third problem is that the drugs are very, very costly, so in the Western world, with the insurance that we have and the economies that we have, many infected people can get access to these drugs. But in the world at large, in the large populations where HIV is decimating populations, the prices of drugs are simply too high.
The other scientific problem is that as effective as these drugs are at stopping the replication of the virus, it now turns out we know that there are cells in the body that might have gotten infected before the person started taking these drugs. Some of those cells can live in the body for a very, very long period of time and they have the potential at any moment to get turned on and start producing a virus.
So for a person who takes these drugs and responds well to these drugs, the virus load will go way down in the body-in some cases you can't even detect any virus in the body-and that person might want to come off drugs. There have been clinical trials to examine what happens when a person comes off this antiretroviral therapy. And in every case what happens is that the virus load comes roaring back, and the person is back basically where he or she started before the treatment began. Fortunately in many or most people, they'll respond again to reinitiating the drug treatment.
Because these [virus] reservoirs are so long lived, a person would have to take these drugs for an extremely long time before those reservoirs disappear, and in fact that might end up being for the life of the person. So even though the person is not suffering from the direct affects of the virus, the person will have to keep taking these drugs and with the side effects and the cost that becomes a very daunting idea.
What about drug resistance? Is resistance an issue if you target the molecules?
The drugs that are currently being used effectively to treat HIV infection are drugs that target the virus. One drug is called a "reverse transcriptase inhibitor"; another drug is called a "protease inhibitor," and each one of those targets a different component of the virus that the virus absolutely needs in order to be able to replicate.
Because the HIV virus mutates randomly at a tremendously high rate, the viruses constantly just by chance are making a variant that happens to be resistant to a particular inhibitor. So, if the replication is not being fully suppressed then those resistant viruses are going to take over.
You have the other opportunity of drugs that would target the cell instead of the virus. The good thing about that is that the molecule on the cell is not going to be changing. That's pretty much set by the genes that each person carries. There is a big advantage of targeting a component of the cell that the virus needs in order to replicate.
A big disadvantage is that whatever that molecule is on the cell that you're targeting and trying to block, it's there for a reason. It wasn't put there just to allow the HIV virus to get in. So there is always the concern that by blocking that molecule and preventing HIV from replicating, you're also going to block the normal function of that molecule and that might cause some very unacceptable side effects that would not be predictable in advance.
How did you start figuring this out? How did you get to the first step of identifying the CD4 molecule?
I started working in the field in 1987 and at that time it was known that CD4 is an essential receptor on the cell that the virus uses to get in. But it was also becoming clear that CD4 was not sufficient. There appeared to be a requirement for another molecule on the surface of the target cell.
Is this where you started to put the CD4 molecule on the mouse cell?
Well, actually other people did that first-this notion that CD4 was not enough came into being. The first studies were experiments that were performed at Columbia University. Basically they showed that if you take a cell that normally doesn't have CD4 on it and you use genetic technology to express CD4 on the surface of that cell, that cell would now become permissive to allow the virus in and to replicate.
But oddly enough that works if you put CD4 on a wide variety of different human cells that don't normally have it. But it doesn't work if you put same human CD4 on a cell from a different species like a mouse. So even though that you can show that the CD4 is there and the envelope of the protein can bind that CD4 the virus still doesn't get in.
What were the questions you were asking yourself?
This poses a very simple question. Why is the virus able to get into a human cell that has CD4, but not into a mouse cell that has CD4? And simplistically speaking there are two possibilities. The one is that the mouse cell, even though it has CD4, has something else that's inhibiting the virus from getting in. The other possibility is that maybe there's an additional thing that's required and the human cell has it and the mouse cell doesn't.
So we and others, in the early '90s, did some experiments to distinguish between those possibilities and what we did was lay a hybrid between human cell and a mouse cell and we put CD4 on the hybrid. If the inhibitor molecule-the inhibitor theory-is the explanation, then you would think that hybrid would still have the inhibitor and the virus would not be able to get in.
Whereas if the additional molecule explanation is the real one then you would think the hybrid would have the additional molecule and the virus would be able to get into the hybrid. What we found-we and other people working on this-is that the virus did get into the hybrid. That led to the notion that there had to be another molecule on the human cell in addition to CD4 that was required for the virus to get in and we call that a "coreceptor."
What does the hybrid cell consist of?
Well, different people did it in different ways. We did it by using our fusion system to get a mouse cell to fuse with the human cell but not by using HIV molecules to promote the fusion. We used measles virus fusion proteins-we knew the receptor of the measles virus and we put the measles virus envelope on one cell and the receptor from measles virus on the other. So one went on the human cell; the receptor went on the mouse cell and for that virus that receptor works on mouse cells. Then we made the hybrid.
So the hybrid was the human cell and the mouse cell?
Right. Because things were fluid in the membrane [of the hybrid cell], basically the mouse and human content gets uniformly disseminated on the surface of that hybrid cell. So if there was an inhibitor there that inhibitor should be disseminated all over the hybrid cell. Whereas if there is an additional required molecule that came from the human cell it also should be uniformly distributed on the cell.
You had the hybrid mouse and human cell, and then you had just the mouse cell?
We have three different kinds of cells. We have the mouse cell that we put CD4 on, we have the human cell that we put CD4 on, and a virus will get into the human cell but not into the mouse cell. So then we make a hybrid between a mouse cell and a human cell and we put CD4 on the hybrid. And the answer was that the virus was able to get in.
Did you know at that point what you were looking for?
We decided to take an unbiased approach where we started out not making any assumptions ahead of time about what kind of molecule we were looking for as a coreceptor. But rather we took advantage of the fact that the human cell appears to have it and the mouse cell appears to lack it.
The genetic information that's being expressed in the human cell codes for, among lots of other things, the coreceptor. Whereas the mouse cell doesn't code for a coreceptor.
So by putting genes from the human cell into the mouse cell and finding out which ones render the mouse cell permissive for letting the virus in, then we'd have a way of identifying the coreceptor.
Is this where the gene library comes in, the fusion assay?
We've been working on this whole mechanism of how the virus fuses with the target cell and as I mentioned before, the way the virus fuses with the target cell is that there's a protein on the surface of the virus called the "envelope protein." The envelope protein has the sole function of binding to the receptors on the target cell and catalyzing a fusion reaction so that the membrane of the virus fuses with the membrane of the target cell and the virus can introduce its genetic information into the target cell.
So instead of working with real HIV virus which has some real technical problems and difficulties, as well as bio safety problems, we decided to use a very simplistic system where the only HIV component that we had was the envelope protein. Instead of working with the whole HIV virus which has about a dozen genes, [we can] just to take the gene from the envelope protein. There are technologies now where you can very easily put that gene into a cell and now that cell will express the envelope protein on the surface of the cell.
We developed what we call the "cell fusion assay," and it's extremely valuable. It's very high throughput and very quantitative. It's very fast and very sensitive.
One of the things that gets lost even on the scientists is that one of the best things about it is that it's only measuring the envelope protein function. If you had an assay using a real virus and your final readout was ten days later that it produced more virus, you know there could be lots of steps in-between that are either working or not working that would confuse the issue. But because in this case the only HIV molecule in the whole system was the envelope protein we know that that's what we're looking at.
So we had a very simplistic way of looking at the function of the envelope protein and we reproduced the HIV infection experiment by showing that when you do the cell fusion assay that a cell is expressing the envelope protein will fuse with a cell that has CD4: but it will fuse with a human cell that has CD4 but not with a mouse cell that has CD4.
How do you show that precisely?
The way we used to look at the cell fusion is simply looking under the microscope: the cells are bigger and you can actually count them. But you can imagine that to be a very tedious thing, sitting with a microscope a looking at thousands of cells trying to decide which ones are bigger than others.
When you look under a microscope and they're not fusing, are they just cells?
Yes, you can just see the individual cells sitting there on the bottom. When the cells fuse, you can see bigger cells sitting there on the bottom and sometimes you have to make a judgment. Is that a bigger cell or not? So your biases enter into it-we needed something that was more objective and easier to quantitate as well as faster and more sensitive.
So we used a technology called the "reporter gene." Reporter gene is simply a device that you can rig up (because of the technology that we have available today) so that when the cells fuse they turn on a gene that we can easily measure.
So when the cells are sitting there and not fusing there won't be any of this gene expressed, but when the cells fuse it would be expressed and we can measure that.
The particular reporter that we used can be measured in a few different ways, which were valuable to us. If you mix a bunch of cells that have the envelope and cells that have the right receptors, when the cells fuse, they'll turn on this gene. So we can measure how much of that gene has been turned on in two ways. In one way we just stain the cells that are sitting in the bottom of the dish and the stain works in such a way that if they encounter the reporter gene the cells turn blue. We look in the microscope and we actually see blue cells.
The other way we do it which in fact is faster and a bit more quantitative and easier is to mix the two cells together (the envelope cells and the target cells with the receptor), let them fuse, and then stop the reaction by just pressing the cells open. You just add some detergent and the cells will just break open and now that reporter gene protein is now floating around in this mixture and we can measure the amount that's there in a very simple reaction very accurately and very quickly.
Does the color have something to do with it?
Depending on the particular experiment you're doing you may get a few cells that fuse or you may get a lot of cells that fuse. The more fusion that goes on, the more of this reporter gene you're going to make. When you try to quantitate how much reporter gene is there, you going to see a difference in the color.
So in the mixture of cells, let's say the target cell was lacking one of the receptors that's required so you wouldn't get fusion. When you added the substrate, there's no enzyme there to change its color and so it will stay yellow. If the reporter gene has been turned on, then that enzyme will be there and it will start turning the substrate red. The more enzyme that's there, the faster it's going to turn red.
Is this where the needle in the haystack comes at this point? What do you do to narrow it down?
We decided in searching for this coreceptor that we were not going to start with any biases about what kind of molecule it is. The only thing we were going to rely on is the fact that the human cell seems to have a gene that codes for this receptor whereas the mouse held doesn't.
What we did was to isolate the genetic material of all the genes that are expressed by that human cell: the collection of that genetic material [was called] our library. We don't know it's in the library, but we are assuming that there is something in there, some piece of genetic information that codes for a receptor. So when we then introduce that library into the mouse cells if that gene gets into a mouse cell that mouse cell is now going to become permissive for fusion when we mix it with an envelope cell. And that's what we did.
We took this library and mixed it in with the mouse cells and added the envelope cells and asked: Do we see the reporter gene getting turned on? The easier way to do that-because there was such a low level of it-was by doing the microscopic assay, because ten blue cells wouldn't give you enough of a signal to really see it here, but you can easily see it under the microscope.
We got a positive result when we put the genetic information from the human cell into this plate that had a million mouse cells in it. When we then stained it, some of the cells turned blue and we interpreted that that the library did in fact contain some gene that coded for the coreceptor, and therefore the mouse cell that happened to pick up that gene became permissive for fusion, it turned on the reporter gene and when we stained it, it turned blue.
So back to this library. You know there's something in this and how are we going to find it? Well, we simply took the library and divided it up into ten portions. [We used] a bunch of different tubes, and one tube got these genes, one tube got these genes, one tube got these genes and so on.
One or maybe several of those tubes will have the gene for the coreceptor in it and many of them won't. Then we did the same thing. You actually amplify the genetic material from each one of those collections and for the collection that never got the gene for the coreceptor, when you then do the experiment again you won't see any fusion. But for the collection that happened to get the gene for the coreceptor, when you amplify that and now put that on the mouse cells you will seecells that are fused.
So we start out with a whole library it might have 10,000 different genes and when you divided it up-let's say into a hundred different tubes-you will have a hundred different genes per tube. [After the experiment] you have now narrowed the search to one in a hundred. So now you take that tube, you expand it up again and divide it up and [conduct the experiment again.] And now you're going to have one in ten. You do that enough times and eventually you get one.
So eventually you found what you were looking for.
What exactly did you find?
We identified a gene and with the technology now, you can very quickly determine the sequence of that gene. And from the genetic code, we can tell the exact sequence of the protein that it codes for. With modern computer technology, there are databases of all the known proteins that people have identified over the past many years, so you can access those on a computer. You can put in the sequence of the gene that we identified and ask if it looks like anything else that other people have found. It turned out that it looked the most like one other gene that happened to have a certain kind of function that people were studying. The molecule we isolated looked like a receptor for something called a "chemokine."
What is a chemokine?
Chemokines are small proteins that cells make and release in order to attract other cells to them so they play major roles in the immune system and also in other parts of the body. They are a way that cells use to attract other kinds of cells in order to fight infection.
What is the role of the chemokine with regards to fighting infection?
In the immune system, cells that might be infected with a certain infectious agent might start releasing chemokines which will then attract other cells of the immune system that have the receptors for those chemokines to move towards the site of infection. In other words, it's like calling in the troops to fight the infection. So this is one of the very many complex ways that the immune system has to fight off infections.
Does the chemokine come in and fill the receptor?
Let's say there's an immune system cell that you would like to bring in to fight off this infectious agent. If it has the receptor and the infected cell starts releasing the chemokine, the chemokine will bind to it. This cell not only can detect that the chemokine is there, it can also read the concentration gradient and it can tell when it's moving towards higher amounts vs. lower amounts, so it knows how to go towards the cell that's actually producing the chemokine.
That's what chemokines normally do. And it turns out that HIV has chosen as its coreceptor a molecule protein that looked like it might be a chemokine receptor although there was no data in the literature to prove that it actually was a chemokine receptor.
Was this research going on at the same time as CCR5?
We were doing this independently based on our studies of how the HIV virus gets into a cell. It was the end of July 1995 when we actually sequenced the gene for the coreceptor and we put it into the computer and found out that it was closely related to this chemokine receptor.
What did you call it?
At that point since it looked like a chemokine receptor but neither we nor anybody else had identified a chemokine that actually binds to this, we really didn't know what its normal function is. The only function that we had for it was that it enabled the HIV virus to fuse with a cell that was expressing CD4 so we gave the protein the name "Fusin."
The evidence suggested that there had to be another receptor in addition to CD4, a so-called coreceptor.
Here we've identified a coreceptor for HIV that looks like a chemokine receptor. We knew that there were probably two different coreceptors to explain the biology of different HIV strains. We knew that some strains should be able to use those coreceptors and other strains shouldn't.
There's an additional important complexity to the story and that is the fact that not all HIV 1-we're only talking about HIV 1-not all HIV 1 strains are the same. The virus mutates at an enormously high frequency, so every HIV 1 virus is a little bit different from another HIV 1 virus. And it's actually the envelope protein that is varying the most, because it's sitting on the outside of the virus and the immune system is throwing antibodies at it all the time. It has to keep changing a little bit to escape the antibodies.
It's been known for years now that different isolates of HIV 1 have very distinct properties for infecting different types of human CD4 positive cells. So people in the literature use a terminology, which I think is very inappropriate, very misleading. They might say "T-tropic" or "T-cell tropic".
One of the tools that you have in a laboratory to study a virus are cell lines that you can grow ad nauseum-they'll grow forever-and the virus can get in and replicate in those. So there are T-cell lines that have been adapted to grow in the laboratory and some HIV strains will grow very, very well in those-others won't.
There are other types of cells that you actually have to purify from blood, so we call them "primary cells" because they're not permanently adapted to the laboratory. You can purify macrophages from the blood. So some viruses can enter and replicate a macrophage but they can't enter into T-cell lines.
The reason I'm making this distinction between T-cell lines and T-cells is because if you take "real" T-cells straight out of the body, they can be infected by both types and that is confusing to a lot of scientists in the field. That's why I always use the terminology T-cell line tropic whereas a lot of other people say T-tropic.
Where does the "tropic" come in?
Tropic means having the ability to replicate in that type of cell to get into and to replicate in that type of cell. So, from the same individual you can isolate one isolative HIV 1 that is T-cell line tropic. That means it can get into a T-cell line and replicate in that T-cell line. But it can't get into a macrophage even though the macrophage has CD4. It can't get into the cell and it can't replicate.
So different strains can come from the same person?
One of the complexities that started being appreciated in the late '80s and early '90s is that not all HIV isolates are the same. In fact, it's possible to isolate two different isolates from the same person that have very different properties in the laboratory.
In particular some isolates of HIV can replicate in what we call "CD4-positive T-cell lines" but they can't replicate in another type of cell, a "CD4-positive macrophage". However, from the same person you can isolate another HIV that has the opposite property. It can get into the macrophage and replicate but it can't get into the T-cell line. So just having CD4 on two different human cells is still not the whole answer.
So we began thinking maybe there's a different coreceptor on a T-cell line compared to a macrophage and maybe one isolate can work with CD4 in this coreceptor and the other isolate can work with CD4 and this other coreceptor.
We actually did the same experiment. We asked the question that for the virus that can replicate in the T-cell lines but not macrophages is that because the macrophage has an inhibitor or is it because the T-cell line has something else that's required. So we did the hybrid experiment between a T-cell and a macrophage and the hybrid fuse. [We found] that the permissive cell line had something else that was required and the non-permissive cell lacked it.
This gave rise to the notion that there must be at least two distinct coreceptors, one that is expressed on T-cell lines and it works for the T-cell line tropic viruses and another that's expressed on macrophages and it works for the macrophage tropic isolates.
Was this going on in different labs?
Yes, that work was going on in different labs. Mostly people were doing it by taking real HIV and doing real infection experiments and they came out with these striking differences.
We took it another step. We expressed the envelope proteins of those different viruses and did a cell fusion assay and we found that the envelope proteins from the T-cell line tropic viruses fused very well with T-cell lines, but not with macrophages; whereas the envelope proteins from macrophage tropic viruses fused very well with macrophages but not with T-cell lines.
Which strain binds?
So when we first decided to do this library screening, we arbitrarily decided to go for the T-cell line tropic factor coreceptor and that's for technical reasons. It's a lot easier to grow up these T-cell lines so you can make genetic material much more easily than you can from macrophages which you have to purify from the blood.
So we isolated a gene which when put into a mouse cell expressing human CD4 would now enable a cell expressing the envelope protein to fuse with it. Because we didn't know of any normal function for that gene the only function we knew about it is that it enabled HIV envelope to fuse, we called it "Fusin". We recognized that at some point the real purpose of this gene and the normal function of this gene would be identified and it might acquire a new name.
What is CCR5 and does the cell need this in order for the virus to be transmitted?
So we identified this coreceptor and we proved that it was a coreceptor by doing the experiment of expressing it in a mouse cell that had CD4 and showing that it now allowed entry of HIV. But based on what I said earlier about the fact that different isolates have a strong preference for entering either a T-cell line or a macrophage, the library that we chose for our initial screening was from a cell that supported the T-cell line tropic viruses but not the macrophage tropic viruses.
So based on that we predicted that the coreceptor that we identified would function for the T-cell line tropic viruses but not for the macrophage tropic viruses. We did those experiments and that's the result that we got. So that convinced us that we had the real McCoy and that left wide open the question of what is the other receptor that the macrophage tropic viruses use.
What we should have done-we could have done it-is the obvious thing which is recognizing the fact that this thing looks like a chemokine receptor and there's a whole field of scientists who work on chemokine receptors and a whole bunch of them that have been cloned. One of the best people in that field works right across the street-his name is Phil Murphy. What I should have done back in July of 1995 was to call Phil and say, "Hey, this is what we found. Bring over all your chemokine separate genes and we'll test every one of them and see if one of them works for the macrophage tropic envelopes."
But being hardheaded as I am and possessive as I am, we started doing the same library screen using genetic material for macrophages and that was plodding along.
Then in December of '95 Powell's paper came out reporting that they had identified the Jay Levy factor-of course Levy never believed it-they had identified some molecules released by CD8 cells that inhibit HIV infection and it turned out they were chemokines. Of course when I saw that I said, "Hmm, does that have any connection with what we're studying?" And they actually showed a figure in their paper that the chemokines inhibited a whole bunch of macrophage tropic viruses that they tested, but it didn't inhibit the one T-cell and tropic virus that they tested. They didn't even make a point of it in the paper.
Then I gave a talk at a meeting and I presented the Fusin stuff and I also very stupidly and very honestly actually gave them the sequence of the protein. I told them it was like a chemokine receptor. I sort of said we have this gene and here is what it does but I'm not going tell you what it is.
By the next week I was getting just overwhelmed with phone calls from the press and all that kind of stuff and I was being very naive about it. I had no idea of the furor that this was going engender.
It was the hot new field.
We knew there had to be another receptor that supported the entry of the macrophage tropic isolates. Now by a remarkable coincidence there happened to be going on totally independently, a very, very different line of research focused on a very different problem in HIV biology. That is the fact that certain cells release soluble chemicals that inhibit HIV infection and replication and that discovery was initially made by Jay Levy's group in the late 1980s. The identity of what those released molecules are became a very, very hot topic. And the first group to identify specific candidate molecules was the group of Gallo and Luso.
Robert Gallo's group, at the National Cancer Institute had identified three specific chemicals released by cells that inhibit HIV infection and they happen to be chemokines. So [one could] put our discovery of Fusin, which looked like it might be a chemokine receptor, together with the Luso-Gallo discovery that certain chemokines inhibit HIV infection.
In their paper they showed, although they didn't make a point of it, in the figure where they tested their chemokines against a bunch of different HIV isolates, it inhibited macrophage tropic isolates very well, but it didn't inhibit the one T-cell line tropic isolate that they examined.
So at that point we got the idea that maybe this is a clue to the other chemokine receptor mainly maybe the macrophage tropic coreceptor is a chemokine receptor that binds those three chemokines. At that point I contacted Dr. Phillip Murphy at the National Institute of Allergy and Infectious Diseases, (my colleague right across the street). He had recently cloned a gene for a receptor that bound those three chemokines and that gene was called CCR5.
And so we tested CCR5 to determine whether it would support cell fusion by an envelope from a macrophage tropic virus and it works perfectly, and it didn't work at all for the envelope protein that from the T-cell line tropic virus.
Now we had two distinct molecules, one of which acted as a coreceptor for the T-cell line tropic viruses and the other of which acted as a coreceptor for the macrophage tropic isolates. One of them looked like a chemokine receptor and the other one we knew was a chemokine receptor. So that's where the story stood, ah, in June of 1996.
What were the implications of this?
Why is this significant? Well it turns out that there's tremendous significance for the real world biology of HIV, how HIV gets transmitted from one person to another, how HIV causes disease in people, what controls its replication in the body, how we might find new ways to treat people who are infected with HIV, and how we might find new ways to prevent transmission of HIV.
Before the coreceptors were identified, it was clear that some viruses had this property of replicating in T-cell lines but not macrophages and others had the opposite property. Several groups showed that if you isolate HIV from a person who has just become infected-during the early course of the disease which can be the first couple of years-these viruses had the properties of being able to infect macrophages but not T-cell lines.
We now know it's because the envelope protein of those viruses use CCR5, but they can't use CXCR4, so we have a molecular understanding for that. We still don't know why the viruses can that only use CCR5 are preferentially transmitted but we know that it is its use of CCR5 that determines what virus is going to take over in the body.
Then as the person is living with HIV and can be quite healthy and maybe not even know about the infection, over time the virus starts gaining the upper hand and the virus replication starts going up and the CD4 cells, which are key players in the immune system start declining, that's when the person is at risk for developing all the consequences of HIV infection.
Does CXCR5 contribute to that decline?
It's during that time when the person goes from the asymptomatic phase to the symptomatic phase that you can now start isolating viruses that use CXCR4. So that's what's happening in the body that the viruses that are present during the asymptomatic phase are preferentially using CCR5 and it's only during the transmission to the symptomatic phase that you start isolating viruses that can use CXCR4.
In parallel with those experiments showing what's going on in the body you can do experiments in the laboratory which generally show that viruses that can only use CCR5 are much less effective at killing CD4-positive T-cells than are viruses that use CXCR4. We don't really know why that is. There are lots of good ideas out there and the mechanisms are probably complicated. But the fact is that the transition of the virus population as the virus is changing in the body starting with the virus that is specific for CCR5 and changing over time and eventually acquiring the ability to use CXCR4 that that is in some way causally related to the progression of the disease.
So does the surface of the cell have both coreceptors?
At least two of the obvious things that can be changing would be the expression of the chemokine receptors on T-cells and also the expression of the chemokines that have the potential to block HIV's use of those chemokine receptors as entry coreceptors. So you could imagine that production of chemokines as well as changes in the levels of the receptors could influence the progression of the disease and there's a lot of work going on for that and it's very complicated.
What question were you asking about people who are repeatedly exposed to the virus and don't develop AIDS with regards to CCR5?
It became clear then that the viruses that are initially taking hold in the newly infected person are viruses that are specific for CCR5 and that began raising questions about a very interesting and important phenomenon: namely that there are some individuals who despite getting repeatedly exposed to HIV do not get infected. So various investigators began wondering if there was any connection between the chemokine receptors as coreceptors and this ability of some people to resist being infected by HIV. It turns out that there is an allele, a mutant version of CCR5 that's very widespread in the population. [This] variant of CCR5 codes for a totally nonfunctional CCR5. [This is] one of the major explanations for why some people are not infected by HIV despite repeated exposure. They are lacking the coreceptor that's required for the initial virus that takes hold in the body.
That's not the only explanation for why some individuals are repeatedly exposed and aren't infected but it's the one for which we have the best understanding.
The other remarkable thing is that in people who have no functional CCR5, the only medical consequence we know about it is that they are resistant to HIV. They don't seem to have any other health impairments. That's a remarkable thing because most of the genes that we have in our bodies are there for some reason and if you're lacking them typically you're going to experience some medical consequence.
But it turns out that perhaps because the chemokine receptor system has a lot of built-in overlap and interplay, that in a person who doesn't have a functional CCR5, other systems can take over and perform the functions that CCR5 would normally perform. [This] raises very important implications for drug development because now we can think about the possibility of blocking HIV replication by blocking entry with a drug that binds CCR5 and prevents the HIV envelope protein from being able to use it and so it prevents entry of the virus.
A particularly attractive aspect of that is that CCR5 is a chemokine receptor and chemokine receptors are actually members of an even much larger family of receptors called "G protein couplet receptors." These receptors encode for all kinds of things. They encode for receptors for neurotransmitters, they encode for receptors of hormones, and they encode for receptors for light, for taste, for odor.
If you go to your pharmacy and you get a drug for any particular ailment, there's a large likelihood that the way that drug is acting is that it's binding to a G protein couplet receptor. That whole family of G protein couplet receptors has been very successfully targeted by the pharmaceutical industry. So now we have a critical molecule on the cell that's required for HIV to enter and it is the type of molecule that the pharmaceutical industry has been very successful at developing drugs for. We know that people can live without a functional CCR5 so why not make a drug that targets CCR5?
What is the downside of targeting CCR5?
The idea of using a drug that targets CCR5 is based on the notion that during the early stages of infection-and that can include the first few years of infection-the predominant viruses in the body are specific for CCR5. So if you took a drug that blocks CCR5, you would greatly slow down the replication of the viruses in the body and that ought to have a very beneficial effect and slow down progression of disease.
There is one down side that a lot of people have been concerned about: during the normal history of the virus as it's replicating in the body, eventually a variance arises that can use CXCR4 and those seem to have a greater capacity for killing CD4 positive T-cells.
The concern is that if a person takes a drug that blocks CCR5, you might now give a greater advantage to those viruses that can use CXCR4, and they may take over faster than they would in the absence of that drug. Whether that in fact is going to be a real problem, we won't know until these drugs have been in clinical practice for some time. But in all likelihood, no HIV drug is used in isolation because of the potential for resistance.
Resistance to a CCR5-blocking drug is not going to arise by the CCR5 changing because that is a gene that's encoded by our bodies, not by the virus. It is possible that the virus could mutate in such a way that it can use CCR5 even when the drug is sitting on there-in fact there are experiments that have been published that show that you can force the virus to do that.
Why not target both coreceptors at the same time? What is the downside of targeting CXCR4? Is it because it plays a bigger role in our bodies than CCR5 does?
Well, it does in the sense that it appears to be more associated with CD4 depletion and disease progression.
The appearance of CXCR4-using viruses is associated with a more rapid progression and those viruses are typically found only when the person has gone to to the symptomatic stage. In the laboratory, viruses that use CXCR5 seem to be much more capable of killing CD4 T-cells than viruses that use CCR5. What I didn't say is that there are lots of cases where people do go on to develop AIDS even though you don't recover CXCR4-using viruses. So CXCR4-using viruses might accelerate the development of AIDS, but they might not be absolutely necessary.
Why not also block CXCR4 and not just CCR5? There are major efforts going on to develop CXCR4-blocking agents for possible use in HIV therapy, but one of the real major concerns is the fact that unlike CCR5, it appears that people cannot live without CXCR4. This is based on experiments that are done in animal models in particular with "knockout mice" where you can actually remove the gene for CXCR4 and that mouse cannot make CXCR4. Those animals die during embryogenesis and it's now believed that CXCR4-in addition to playing an important role on cells of the immune system-is probably playing an important role in other systems of the body. So these animals that die during embryogenesis have defects in heart development as well as other kinds of defects.
While it's possible that once a person is born and gets past that embryonic stage maybe CXCR4 isn't necessary and maybe you can get away with blocking CXCR4 to prevent HIV from using that coreceptor. There is as major concern that all through life maybe we need CXCR4 to be functioning.
What's the most promising strategy in dealing with these coreceptors and the pharmaceutical ramifications?
With the recognition that we might be able to develop a new therapeutic strategy based on blocking the CCR5, there's a lot of thought going to what kinds of blocking agents might we use. One of the most obvious things are the natural chemokines that normally bind to the chemokine receptors.
There are several potential problems with that. One is that the chemokines, if introduced in the body, are going to be doing the things that they normally do. They're going to be triggering cells in ways that might be harmful for the person. The other problem is that a chemokine is a large molecule and it wouldn't be able to be taken orally: the most effective drugs are drugs that are taken orally.
There's another kind of agent that a lot of attention has been given to and that would be an antibody that targets the chemokine receptor and binds to it and prevents the envelope protein from using it. Again, you have the problem that antibodies are large proteins and they wouldn't be orally bio-available.
There are also genetic engineering ideas that have been worked on and that is to modify a person's cells so they don't express CCR5. There are a lot of potential ways to do that, but you're really talking about gene therapy: modifying the person's genes outside the body and then reintroducing them back into the person. [This could have] all kinds of complications and that technology is just not ready yet, but it could be very promising down the road.
So the really most promising kinds of CCR5 directed therapies would be small molecular weight chemicals that bind to CCR5 in such a way that they prevent the HIV from using it as a coreceptor and there's a tremendous effort going on in the pharmaceutical industry for that.
How far along is the industry in making these low weight molecular compounds for CCR5?
Several companies now are placing major efforts into developing small molecular weight molecules that will block CCR5 and prevent HIV from using them. Some of them have entered clinical trials. There's some hint of promise and with drugs of this type what we're going to see are first generation, second generation, third generation variances that have improved potency with fewer side effects. These are going to have to go through a long period of development including clinical trials to test safety and efficacy. But there's a real hope that this would provide another tool or another weapon in the armamentarium against HIV.
With a virus like HIV that can change so fast, you want to be able to throw different kinds of things at it. So the idea of having a new kind of target different from the ones that are already being successfully used like the transcriptase inhibitors and the protease inhibitors, has become very, very exciting. This idea of an entry inhibitor that blocks the coreceptors is very exciting. There are others that are also being very actively developed. There are other agents that actually bind to the envelope protein instead of to the chemokine receptor and prevent the envelope protein from functioning. So some of those are in clinical trials as well so we'll be hearing lots more about those in the coming years.
What are microbicides?
With the worldwide pandemic of HIV, there is a tremendous need to slow the transmission of HIV around the world. Mostly the efforts have gone into vaccine development and we've seen a lot of successes in that area as well as frustrations. Whether or not we're going to have a truly preventative vaccine anytime soon? I'd be hesitant to put my bets on and I think most scientists would also feel that way.
So given that, there's more interest in a very different kind of strategy to prevent transmission of HIV from the infected person to the unaffected person and that's what we call a "topical microbicide." A topical microbicide is simply an agent that would either block HIV or kill HIV at the time and place where it's about to infect a new person. For heterosexual transmission that's typically in intercourse, so that would be in the vaginal tract.
What is your strategy?
So the strategy that we've taken is to try to develop a very, very specific inhibitor of HIV that could be applied topically at the site of transmission. We've developed a very potent protein that neutralizes HIV infection. The way it works is based on this mechanism which we now understand about how HIV sequentially has to bind first to CD4 and then to the chemokine receptor.
So we designed a protein based on the knowledge that we've learned about how HIV has hit the cells and how HIV uses these two receptors CD4 and the chemokine receptor. We know now that the envelope is structured in such a way that it combines CD4. This is CD4 in the target cell and this is the HIV envelope protein. The HIV envelope protein can bind here.
So we now know that part of the HIV envelope protein binds to CD4 and part of it binds to the chemokine receptor. The part of it that binds to the chemokine receptor is actually hidden-it's buried-within the structure of the envelope protein and it only becomes exposed when the envelope protein binds to CD4: it can now bind to the chemokine receptor.
So you can imagine that if you had an agent that bound to that part of the envelope protein that recognizes the coreceptor, it might be a good way to neutralize the virus. People have identified antibodies that bind to that newly exposed site that's involved in binding to the coreceptor. The problem is they don't neutralize the virus because the virus is floating around with this site hidden so the antibody can't find it. When the virus binds to CD4 on the cell and now opens up and exposes this new site, it's probably going to bind it really fast before the antibody ever has a chance to get there.
So even though this part of the envelope protein which is very, very highly conserved between different strains of HIV all over the world and even though it's a potential neutralizing target we don't know how to get at it. The only way we know how to get at it is by CD4 binding.
That inspired us to come up with a new idea of how to create a new neutralizing agent. That is to create a hybrid protein or a chimeric protein in which we have a soluble part of CD4, but constructed in such a way that it's not attached to the cell. That soluble CD4 is now linked to an antibody that binds to this highly conserved site that's induced on CD4 exposure and is required for binding to the chemokine receptor. The soluble CD4 portion of this chimeric protein binds to the envelope protein on the virus. It causes it to open up and expose this part of the envelope protein that binds to the coreceptor. Now the antibody part of the chimeric protein can bind and that would neutralize the virus. We've done experiments in the laboratory and shown that this is in fact a very potent neutralizing protein.
How long before this can be used?
We have a very potent neutralizing agent and there are other potent neutralizing HIV agents as well. How might we consider actually using them as a topical microbicide? Well, one way would be simply in some kind of gel form or capsule form that could be applied by the woman at the time of exposure. That's a very important development that's going on now and there are companies that are producing these proteins for this purpose. The problem is the proteins are very expensive to produce. However, there are technologies out there to produce antibodies and proteins like the one we made very, very cheap and the hope is that this would be able to be available in an economically and practically feasible way around the world.
Is it a separate strategy where the lactobacillus are genetically engineered to express the molecules?
That's one way of considering using a protein as a topical macrobecide. But we're actually exploring a very intriguing novel way of doing this and that's based on the fact that the mucosal services of the body-all the mucosal services in all parts of our body-are naturally colonized by bacteria that actually create a healthy environment.
In the vaginal tract, the major protectant bacteria is lactobacillus. Loss of the normal biofilm of lactobacillus from the vaginal tract is often associated with recurrent urinary tract infections and yeast infections, so there are efforts underway now by other investigators to try to treat recurrent urinary tract infections by repopulating with the natural lactobacillus that normally lives there.
So we're taking that idea a step further by engineering lactobacillus to make proteins that would actually neutralize HIV. The idea is that the vaginal tract could be colonized with these lactobacillus that are normal except for the fact that they are making this protective protein and that when the virus encounters the vaginal tract, it will be neutralized by this protein.
The advantage of that over the gel and the cream is that because the lactobacillus is an organism that normally lives in the vaginal tract, it might persist there for long enough periods of time and provide durable protection. We have no idea whether this is actually going to work, but we think it's a very exciting area.
Why is this important for animal models?
One of the major opportunities that the coreceptor discoveries have provided us is the potential for creating transgenic animal models for HIV infection. HIV 1, the main virus that infects people does not infect small animals.
One of the major problems that has impeded all of HIV research is that we don't have a small animal model that can be infected with HIV. One reason for that is that small animals lack the appropriate receptors that allow HIV to get in. So it's possible to engineer mice or rats or rabbits to produce CD4 on the surface and also to produce the chemokine receptors and that work has been going on for several years now.
Progress is being made, but there are other issues because in order to have a really good animal model, not only must the animal cells be permissive for letting the virus in but they also have to be permissive for all the subsequent steps in the virus replication cycle and those are complex.
We believe that any animal model is going to have to have the chemokine receptors in order to be susceptible to a HIV infection in addition to solving these other post entry problems.
What are some of the unanswered questions in this field?
So there are major questions that have been opened up by the discovery of the coreceptors. From the standpoint of understanding the mechanism of how the virus gets into the cell, we need to understand exactly how the HIV envelope protein physically interacts with the chemokine receptor. There is work going on using very sophisticated techniques of structural biochemistry to try to elucidate the binding interactions between the HIV envelope and its receptors.
We have some good information now based on some very important studies done a few years ago on how the HIV envelope binds with CD4, but we as yet have no insight into how the envelope protein binds to the chemokine receptor-the biochemical efforts to try to tease that out have yielded by complex results.
What we might get from that understanding is by actually seeing how they interact, we might be able to use methods of rational drug design to develop drugs that would fit in the right place and block the ability of the envelope protein to interact with the chemokine receptor. So that's a basic question focused on the mechanism of virus entry into cells.
When we extend it to the so-called real life issues of HIV, infection of people, we don't know why the viruses that are initially transmitted from one person to another are specific for CCR5. That's a major, major challenge and there's some very important biology going on there. Perhaps if we understood that we might have some better ideas about how to make a protective vaccine or a protective topical microbicide.
Why are the viruses that use CXCR4 more pathogenic? We really don't understand that. There are some very good experiments out there that suggest certain mechanisms but we still don't have a full answer on that.
Why is it that the virus persists in the body for years only using CD4 and in many individuals, after some time the virus starts evolving into forms that can use CXCR4? What are the selection pressures in the body that prevent the early emergence of the CXCR4-using viruses and allow their subsequent emergence at a later date? Answers to those questions will go a long way towards not only increasing our basic understanding of the HIV disease process but also into devising new ways to block it, to treat it, and to prevent it.