Scientist Stories: Michael Oldstone, Pioneer in Viral Immunopathology
The video records a faculty lecture introduction and a comprehensive talk by Dr. Michael Oldstone on viral immunopathology, focusing on how certain viruses establish persistent infections. The introduction highlights Oldstone’s distinguished career, seminal contributions (LCMV immunopathology, molecular mimicry, cytokine control, viral receptors), major honors, and his extensive mentorship legacy. Oldstone then surveys the history of the research institute, emphasizing the roles of its early directors. The scientific portion details the mechanisms by which lymphocytic choriomeningitis virus (LCMV) can either be cleared (Armstrong strain) or become persistent (clone 13) through a single amino‑acid change (position 260) that enhances binding to the cellular receptor alpha‑dystroglycan. He describes how this receptor, together with a specific glycosyltransferase (LARGE), mediates entry into dendritic cells, the primary cellular niche for persistence. Persistent infection leads to T‑cell exhaustion, reduced MHC and co‑stimulatory expression, and the production of immunosuppressive cytokines (IL‑10, IL‑21, TGF‑β) and checkpoint molecules (PD‑1). Experimental evidence from numerous post‑doctoral fellows is presented, including in‑situ hybridizations, reverse‑genetics, ENU mutagenesis, and cell‑specific knock‑outs, establishing dendritic cells as the critical reservoir. The talk extends these findings to human chronic infections (HIV, HCV) and to Lassa fever, proposing that genetic variation in the LARGE gene influences susceptibility. Finally, Oldstone outlines future directions: identifying small‑molecule inhibitors of negative regulators, developing glycopeptide blockers of the receptor, and translating these insights into therapies for persistent viral diseases and cancer.
Chapters
Um, so it's, it's a great honor and a great pleasure to introduce Michael Oldstone today to give the the faculty lecture. My introduction will be very brief. Michael's achievements and contributions over a long and distinguished career are so great that obviously I could do a much longer introduction, but I prefer to cede the time to Michael to tell you about his latest work. So very briefly, Michael's quite clearly one of the most distinguished most published, most scholarly virologists in our community. He's made numerous highly significant contributions. In particular, using a single persistent virus, LCMV, he's defined much of what we know about viral immunopathology. And very briefly, chronologically, he's made seminal contributions to our understanding in several areas. Immune complexes in chronic viral disease, the importance of molecular mimicry in autoimmunity, and probably is the greatest champion and spokesperson for that. Lymphocyte killing of virally infected target cells. I think Michael was the first to actually show that. The control of virus infection through cytokines, and then the identification of receptors for EBV, measles virus, and most recently more recently, arenaviruses. As expected for one who's made so many contributions, important contributions, he's received numerous awards and honors, most particularly election to the Institute of Medicine and of course the National Academy. I was talking to Michael the other day and asked him of which of his many achievements he was most proud, and just like a shot, He said his trainees, and that was honoured recently by the highly prestigious AAI Excellence in Mentoring Award in 2011. When you look at his list of trainees, you can't really believe it. It's absolutely phenomenal, and I thought I knew most of them, but in fact there were many more than I knew. So I'll just give you a very brief flavor. I think many of you will recognize some of these names. Welch at UMass, Buchmeier at UC Irvine, Byron, professor and chair at Brown, Kiesling, Karolinska, Sissons, the Regius Professor of Physics at University of Cambridge, Rafi Ahmed at Emory, Lipkin at Columbia of recent fame as a consultant for their film Contagion. Nelson at Oregon, Mucker, UCSF, Campbell at Sydney, von Herrath at LIAI, and Manchester at UCSD, and so on and so on. It's quite a remarkable set of individuals with massive impact on microbiology and immunology, and I think an ongoing testament to Michael's enormous influence and continued influence.. He's written a very popular book which I often quote from entitled Viruses, Plagues and History, which if you want a brilliant account of viruses and their role in disease over the ages, then you should get it. His own interests are very Catholic, so he can certainly hold his own in discussing Greek history with our learned chairman there. Discussing opera with other members of the faculty, and most importantly, discussing English football, soccer, with me. So it's with great pleasure that I introduce Michael today, with many thanks also for his many contributions to the department, and on a personal note, for the great advice that he's given me over the years. So Michael's title is Anatomy of Viral Persistence. Some pathogenesis entry through the alpha distroglycan gate.
It's a pleasure to give this talk, especially at this time. It's the— for those of you who don't know, it's the 50th anniversary of the formation of this institution. And I thought I would start just with about 3 or 4 minutes of talking to you about the institution because I have a feeling that if you don't know your past, you don't know what your future is going to be. You certainly don't know about the present. Actually, there are two phases to the institution and each phase is reflective of the president or director of the institute. The first 25 years, Frank Dixon was the director and his contribution was that he founded the research institution. And if you think about it, he came here with 4 other colleagues from the University of Pittsburgh. They were pathologists interested in immunology. At that time, he was well known as an immunologist. He was very young. I guess he came here, he was about 36 or 40. He was quite young. He came to La Jolla. They didn't have a building. We now have several buildings. There was no salary for the 5 people who came. They had to raise their own salary on grants. No salary whatsoever. They came with 5 faculty members and 7 postdoctoral fellows. Now the faculty are about 180, and postdoctoral fellows are about 500. I mean, it's quite amazing. And Dixon, through his personality in the early years as one of the great leaders in immunology, attracted a number of postdoctoral fellows to work with him, and I was one of them. As was the next director, Richard Lerner, who came to work with Frank Dixon because of his reputation in immunopathology and immunology. Well, he built the foundation, and above— on that foundation, Richard then became the second director. And of course, Richard's great contributions were the buildings that he put up, and better than the buildings, the colleagues that he brought in from different disciplines. And while with Frank, we really just had chemistry with Frank Hennekins and microbiology with John Spesizen and the— and clinical immunology, which became the Department of Experimental Medicine later on. Richard brought in, as you know, great chemists and melded chemistry with biology and brought in the Department of Cell Biology and really brought in molecular biology and brought in structural biology, did those visionary things before they were the right thing to do or when very few people thought about doing them. And so through those two people, we're where we are now, and it's good to mention that. And the talk I'm going to give today I want to dedicate to the postdoctoral fellows that I've had because they've really kept me going and they've contributed enormously to what I've done. And they've been a lot of fun to work with. So I thought I would talk a little bit about what we've done in the past for the ideas we've had and then go into some unpublished work.. And I would mention the people that contributed as I go along. So what I want to talk about today are how viruses abort the host's immune system, the antiviral response, and that by that means establish a persistent infection. And then knowing the principles of how that occurs, ask the question, how does one restore that immune response to purge the virus and to terminate a persistent infection? Since, as we know, persistent infections are a major problem today. And what I'll talk about, or the background you need, I should say, are that cytotoxic T lymphocytes, or CD8 T cells, are programmed to kill virus-infected cells. These are the factories that produce viral progeny. And the CD4 cells, they make and release inflammatory cytokines like interferon gamma and TNF alpha. And the inflammatory cytokines can participate in removing virus from infecting cells without lysing them. So you have two ways cytotoxic T lymphocytes can control a virus infection by lysing the cell and by releasing cytokines that remove virus from infected cells. And viruses, like a check game— chess game, rather— have— that persist have developed mechanisms to escape antiviral T cell activity by several ways. And these include, for a T cell to recognize an infected cell, it must, through its receptor, recognize an MHC molecule complex with a peptide. And viruses have learned how to disrupt the MHC peptide complex from the beginning of the breaking of the protein to peptides, to its translocation in the Golgi, to its binding and to its migration to its presence on the cell.. There are also regulatory cells that veto or regulate cytotoxic T lymphocytes. And then there's a universal mechanism which I'll spend most of our time on, which is T cell exhaustion or hyporesponsiveness of T cells that occurs uniformly in all the persistent infections that have been looked at, whereas these other mechanisms occur in some but not all. And to give you a flavor of cytotoxic T lymphocytes during an acute virus infection, the dendritic cell presents MHC molecule with a peptide to a T cell, which primes it in the presence of certain cytokines. You get initial expansion, you get increased expansion expansion, and then you later get contraction, and you develop memory cells. Remember that the T cells divide about every 6 hours, so you can imagine the amount of T cells that are there. And to get a handle on how many T cells there are, with a gifted former postdoctoral fellow, Dorian McGovern, who is now a section chief at the NIH, at the NIH, we did an experiment where we took either virus-specific CD8 cells or virus-specific CD4 cells, took them separately. They had a genetic marker on them, green fluorescent protein. We put them into a mouse, let them rest for 2 days or so. These numbers of cells we transferred made up less than 1% of the total lymphocytes that were present. We then added virus and we looked to see the expansion of these T cells. And so those will be in green. And we marked the virus— and here we're looking at whole spleens— we marked the virus in red. So you're looking at viral antigen in red. And you look at T cells.. And you can see they start to come in at day 4, both CD8 and CD4. You see here is the viral antigen. By day 7, they cleared the virus antigen. And it's amazing if you think about it how the whole spleen is a bag of cytotoxic CD8 T cells, as you can see. Now the areas that you're looking at that are blank are the so-called B zones, B cell zones in the spleen. That's why they don't have CD8 cells in there. Now from here, from day 7 and 8, these cells have migrated out to their targets. Many of them have been killed by apoptosis. So if you look at day 8 and now you look at day 10, you can see how many spleens, how many cytotoxic T lymphocytes or CD4 cells are left. Then as you go on to a later time point, CD8 cells, there are very few cells in the spleen and maybe only those in the front row can see the few dots. These are memory cells, so they will respond more vigorously and more quickly than the primed cells. Now knowing that, maybe I should give you a flavor of how these cells work. On a single-cell basis. And to do these experiments, we took a cloned CD8 T cell, we took a virus-infected cell, and we in vitro looked at their activity. So here is the cloned CD8 T cell. Here's a virus-infected cell. And this cell, as it moves through the field, —these are Nomarski optics—as they move through the field, they put out pseudopodia in 360 directions. However, when they come within a distance of the cell that they're going to recognize and kill, they'll put out a pseudopodia almost like a medieval army putting out a bridge to go to the next castle. They'll put out a pseudopodia that you'll see. And then on that pseudopodia you'll see these granules that contain serum esterases move down onto the surface of the infected cell. Let me show you how that looks. Here you can see the bridge forming. Here you can see the granules moving down. Okay, now after that's occurred, there are molecules that make holes in the— that are analogous, not quite molecularly, to the complement proteins that make holes in membranes. And into these holes, the serum esterases have gone, and they will then disrupt the actin-myosin system inside the cell so that there will be blebs that are formed and the cell will eventually die. This, by the way, is taken in real time to give you a an effect. Now obviously I can't show you— obviously I can't show you how cytokines work, but that's an idea of how cytotoxic T lymphocytes kill. Oh, I think I'm going the wrong way. Excuse me. Okay, now this is what happens when you have an acute infection. When you have a persistent infection, you have initial expansion, then you get decreased expansion, and you don't make any memory cells. And let me show you what happens when you have a cell, a T cell that no longer can be functional. And this cell is— there's your T cell. Clone T cell. Here's a virus-infected cell. And amazingly, what happens is that this cell will just— we call it sniffing activity— will just move along the surface of the cell but will not kill it. Okay. So, Now to investigate this in molecular terms, we took the model lymphocytic choriomeningitis virus in its natural host. This is sort of a sweet virus. It's already won Nobel Prizes for 3 people. So it obviously has led to lots of information. A lot of the basic immunology that we know of and virology have come from study of this virus that have been expanded from the mouse to humans and to other pathogens. And the virus that we'll talk about will be the parental virus, which is Armstrong, which when injected into an immunocompetent adult mouse generates cytotoxic T lymphocytes and clears the virus infection. So we call this CTL+ P- because there's no persistence. And its variant that we've cloned which we call clone 13, which when injected into an adult at the same route of administration in the same dose does not generate a robust CTL response and causes a persistent infection. So we call this CTL minus, P plus. And we know we'll ask what is the difference between these two viruses. First let me tell you a little bit about these viruses. They're a member of the adenovirus family. There are Old World Arenaviruses, which consists of lymphocytic choriomeningitis virus and Lassa fever virus, a severe hemorrhagic fever. And New World Arenaviruses, Junin, Machupo, and Guanarito and Sabia. These 4 viruses cause severe hemorrhagic diseases of humans. And on the basis of antibody tests and of sequence, they relate into the Old World and New World viruses. The genome of the virus is rather simple. It has 10.6 kb. It consists of a segmented virus, 2 pieces of RNA, a large RNA which in a positive orientation has a Z gene that encodes a Z protein that's important for assembly and budding. In negative orientation, a polymerase, which is part of the transcription complex with a nuclear protein which is on the small RNA and is also in negative orientation. The glycoproteins are made as a common glycoprotein and then post-translationally cleaved into a spike GP1 and a transmembrane GP2. Here is a negative stain and you can see the spike GP1 spike GP1. This is a picture taken just a section through a tissue with infection. Here's the virus. You see these forms inside the virus are ribosomes. The virus is so sloppy in its budding that it incorporates ribosomes which don't play a role in the virus because the same phenotypes that occur in ribosome-free media or ribosome media the virus has. And this EM I'm rather proud of because it's one that I did, and you can see the spikes coming out of the central holes over here. The spikes are about 5 nanometers, and the difference between each hole is about 27 nanometers. I think that's enough about the virus. Well, the first germane experiment that was done was by Raffia Midwin, who was a postdoctoral fellow. He took 5-3B clone that was cloned by another postdoctoral fellow, Frank Dutko, injected into newborn mice, and then isolated from those newborn mice a virus that he— that has clones either from the brain or from lymphoid cells. Now the virus he put in had generated cytotoxic T lymphocytes. And was cleared, so it was persistent negative. And all the clones he got out of the brain, we looked at about 20, 25, all of them lysed the cells, and that's determined by a chromium-51 assay. When you add the isotope chromium-51, it goes into the cell, it binds to a protein in the cell, and then the leakage is dependent on membrane damage.— so here you're getting about 60% lysis out of a total of 100%. And if you look at the amount of virus that's carried, you can see that when you plaque it, it's undetectable. In contrast, when virus was isolated from lymphoid cells, either dendritic cells or CD4 or CD8 cells, you could see that we looked at about 50— separate clones, you see that about 98% of these clones failed to have a robust CTL response compared to these two. And in all these cases, you can see we had virus persist by plaque formation. Obviously, the two that generated cytotoxic lymphocytes cleared the infection. Now, it wasn't a matter of replication. Amounts, because if you looked at the spleen, we found that the amount of virus in the spleen from either of these things were about 10 to the 7th. So there was a— whether it was a CTL virus or not, we had 10 to the 7th. However, the distribution was quite different. As you can see in the CTL minus virus, looking at nucleic acid hybridization, the virus was in the white pulp, whereas in the CTL+ virus, the virus was not in the white pulp— compare this to this— but was in the red pulp. So on the basis of that, Persephone Burroughs, another postdoctoral fellow who in fact did these in situ hybridizations, did a virus overlay assay where you take different cells that are permissive or non-permissive for the virus, you downsize them, you run them on polyacrylamide gel to separate them, and then you add the virus to see where the virus receptor might be. And you pick up this heterogeneity in a variety of cells that are permissive.— and then using a variety of enzymes that break up sugars or lipids, you could show this was a glycoprotein. Then another fellow, Wei Cao, going through 2 years, I think, of torture cycling the materials through different columns and using lectin columns to get a single band identified a protein by mass spec, only to find out that wasn't the receptor, and continued to do the purification until she got a peptide profile identified as alpha-distroglycan, which turned out to be the receptor. And the evidence for the receptor was threefold. First of all, if you took alpha-distroglycan, purified it, and put it on a membrane, you could bind the virus to it. Second, if you took soluble alpha-distroglycan, you could prevent the infectivity in the binding, see? Whereas another virus like VSV didn't do that. But the piece of the most important evidence was that if you had the alpha-distroglycan present in a cell and you infected at different multiplicities of infection, you can infect different cells as you'd expect. If you knocked out the gene, you didn't infect any cells. And if you now put back the gene, either through an adenoviral vector or just by transfection, you can see that you can restore infectivity. So alpha-dystroglycan is a receptor for both LCMV and for clone 13. And then what I want to talk to you about is the identification and characterization of LCMA and Lassa fever virus cellular receptor alpha-DGE and its preferred anatomical location, which is on dendritic cells. And these, as you will find out, are the cells that are essential for persistent infection. And then we'll show you how we mapped the different variants to find out what amino acids and where the difference between clone 13 and Armstrong is. And then we'd analyze the strategies used by the virus during its interaction with dendritic cells that suppress the antiviral immune response. And lastly, we'll speculate how Lassa fever virus and alpha-distroglycan interactions explains Lassa fever virus pathogenesis. So first, let me tell you something about the receptor. It has two units, an alpha-dystroglycan unit, which is 635 amino acids, which has a globular head at the amino terminus and at the carboxy terminus, and in between has a mucin domain. And the binding site is located at the C-terminus of the N-terminus, in the beginning of the mucin domain. It has non-covalently linked a beta-dystroglycan of 243 amino acids. And this beta-dystroglycan domain, transmembrane domain, is not necessary for virus infectivity, as you can substitute several different molecules and still get efficient virus entry and infectivity. Activity. Now the alpha-dystroglycan serves as the receptor for extracellular matrix proteins, aggretin in the central nervous system and laminin throughout the body. And so the virus, to be effective, it has to first push aside laminin and then bind to the receptor. Now in addition to the— In addition to alpha-distroglycan, we recently found out to have a functional receptor, you need a glycosyltransferase enzyme called LARG, which is quite specific. Most glycosyltransferases are permissive. This is a very specific glycosyltransferase for alpha-distroglycan, and it's necessary to bind—— the N-terminus of the globular domain of the first mucin domain changes conformation, put on an O-linked sugar that becomes phosphorylated, and you need that for a functional receptor. So you need both the alpha distroglycan and the large distroglycan. And this is how it works. Distroglycan binds and changes conformation and puts on this O-linked sugar that then becomes phosphorylated. We've mapped the binding site. I should say Stefan Kuntz, postdoctoral fellow, is now a professor at the University of Lausanne. And that's mapped to 18 amino acids, as I told you, at the carboxy end of the first globular domain and the beginning of the mucin domain. It's 18 amino acids. And it's receptor for LCMV, Lassa fever virus, and laminin. And we know that 3 aneines and 3 317 and 319 are absolutely essential for the glycosyltransferase work. Now when we took a variety of isolates that generated cytotoxic T lymphocytes and cleared the infection and we bound them to alpha-distroglycan, you can see that they bound between 2 and 2.5 logs— this is a log scale— less efficiently than those isolates that failed to generate cytotoxic T lymphocytes and caused a persistent infection. And further, when we sequenced these and we looked at clone 13 and arm 5-3B, we found there are only 6 nucleotide differences that that encoded 3 amino acids, 2 in the spike glycoprotein, which would be for binding and entry, and 1 in the large gene. And then by using a system of reverse genetics pioneered by my colleague Juan Carlos de la Torre, we found that only 1 of the amino acid in the glycoprotein, the amino acid at 260, was responsible for binding to alpha-dystroglycan. And of the 30 or 40 or so different clones we looked at, we found that about 85% had a leucine at position 260, about 14% had an isoleucine, and one had a valine. So they had small aliphatic amino acids at position 260 for binding, and that all the other, isolates that generated CTLs had a phenylalanine in position 260. So on the basis of that amino acid, those two amino acids, you got persistence or acute infection. So how does persistence look in the whole animal? Well, to do these experiments, and they were done— I didn't put this here— by Ian Lipkin, what we did was we took a mouse, infected it either with clone 13 or Armstrong, and then we sectioned the entire mouse and we used a riboprobe to see where the nucleic acids are. So you can see that here's a persistently infected mouse and it's almost a bag of virus. Here you look at the Armstrong, you don't see any virus at the same time. I think these are about day 30 after infection. If you look at plaque-forming units, you have 5 logs in these mice.— you have none over here. And if you look at CTL activity, which of course would be day 7 or 8, you find they don't generate good CTLs, but here they generate a lot of CTLs. So that's how the whole animal looks. And then a postdoc fellow currently in the laboratory, Brian Sullivan, asked using the reverse genetic system that Juan Carlos worked out, asked, which of the two amino acids were important that were found in clone 13, the D or the L. And if you look at this chart, and it's all color-coded, here you have both the asparagine and the leucine in the full clone 13, and here genetically you just put in the 260 of the leucine and you see that the binding to alpha-dystroglycan is about the same. In contrast, when you take on the clone 13 background, you just put in the amino acid 176, the speragine, you can see it's like the Armstrong with the phenylalanine. And if you look now at this cold thing, you see the same is for serum titers. It's 260 that's important, not the 176. If you look at the binding, if you look at infectivity of dendritic cells, here we're looking at viral antigen in dendritic cells. Once again, you see it's the leucine in position 260 that's important and not the mutation in 176. So the single mutation is important for binding and for entry. It's important for looking at persistent infection. You can see over a time scale. And it's important for causing a hyporesponsive of cytotoxic T lymphocytes. This is GP33. GP33 to 41 is an immunodominant CD8 T cell. You can see it's reduced. Immunodominant epitope 396 to 404 that was mapped by Lindsay Whitten when he was a postdoctoral fellow. And you can see that it also is suppressed. Now, in terms of where alpha-disterglycan is, Stephen Kutz then chemically eluted from total spleen or from purified— dendritic cells, B cells, CD8 or CD4 cells— alpha-distroglycan and saw where they were. And you can see that about 99.9% of all the alpha-distroglycan in the central nervous system is located on dendritic cells. And then Andrew Lee, currently in our laboratory, asked what happens when you infect with Armstrong or clone 13. And you can see that if you look in this box over here, the Armstrong infection or noninfected animals do not show a virus in the dendritic cells. But if you now look at clone 13, you can see the expression of viral antigen in large amounts. So it's the dendritic cell that's the main target. Now the first Additional proof for the dendritic cell was given by a current postdoctoral fellow in the laboratory, Daniel Popkin, who working with me and also with Bruce Beutler's laboratory screened over 200— 2,400 ENU mice and looked for mutations. So here's how clone 13 looks. That's the kinetics of the persistent infection. So now you do saturation mutagenesis with ENU in these mice, and you ask, if you give them clone 13, will they respond like an Armstrong, because you've knocked out certain genes that are necessary. And when Dan did this, there was one gene that was found, and this gene was important. It's a serine-1 protease. Important for cleavage, the common glycoprotein to the spike glycoprotein in the transmembrane protein. And when he looked at this mutation, he found that you didn't get a persistent infection with clone 13 infection. And in fact, this affected only the dendritic cell. It was not found in other cells that the virus replicated in, again pointing to the dendritic cell. And now in unpublished data, using a flox-CRE system and specifically deleting dendritic cells, T cells, or B cells. B cells are not shown on this graph but would be like the CD4 cells. You can see that deletion of them still allows persistent infection. But when you delete the dendritic cell with the CRE recombinase, you stop the persistent infection. Again, ensuring that the dendritic cell is the cell that's important. So what I've told you so far is that on the basis of a single amino acid, a leucine in position 260, virus is able to push aside laminin. I didn't tell you that clone 13 binds to alpha-distric glycan, 5-fold stronger than laminin does, so it can push it aside, and then it binds to the receptor, and by the binding and entry is controlled by this 260. In contrast, when Armstrong has a phenylalanine in this position, it does— it cannot displace laminin. It has this— it doesn't displace laminin, and the few that may displace a little bit bind it at about 2.5 logs lower affinity. Now the virus then gets into the dendritic cell. There are 2 amino acids— there's 1 amino acid difference at position 1079 in the dendritic cell. And clone 13 has this glutamine, which is important for increased replication. Transcription and replication, and not the lysine of the Armstrong virus. Once in the dendritic cell, the virus releases IL-10, which we'll talk about later, which is a negative immune regulator. That is, it adjusts the immune system so it negatively regulates it so it can't make a large response. But by infecting the dendritic cell— and by the way, these lymphocytic choriomeningitis virus is a non-lytic virus. Everything you see is caused by the virus or the immune system. It doesn't lyse the cell. What happens, you get diminished MHC molecules and co-stimulatory molecules. As I've told you, this is necessary to present antigen. The virus also prevents the migration of MHC to the DC surface, and it also prevents the differentiation and expansion of new dendritic cells to repopulate cells in lymph nodes and spleen. And this results in a global suppression of T cell responses and antibody and leads to the persistent state. This is the evidence done by Brian Sullivan on cells that were generated by Vishnu Nayak, one of our collaborators. And these are immortalized dendritic cells. And the secret here is to use the reverse genetics of Juan Carlos. So you take the glycoprotein of clone 13 so you can get efficient binding to the dendritic cell. And then you have, for the polymerase, you put in either the polymerase of Armstrong or the polymerase of clone 13, making recombinant viruses. And you can see that if you use the polymerase of clone 13, you get this type of replication in two different lines. And if you use the polymerase of Armstrong, you get about 1.5 to 2 logs poorer replication. So the mechanisms of how the virus interferes is with the differentiation function of dendritic cells. Interferes with its differentiation and migration and in the production of IL-10. Now how does it do that? How does it work with the differentiation? These experiments were done by two postdocs, Doreen McGovern, I told you is at the NIH, and Naomi Sevilla, who is now a tenured faculty member at the University of Madrid. So here you have no LCMB infection. This is the MHC expression on a single cell basis. Here you're looking at dendritic cell with a CD11c marker. And here you're looking at the merge of MHC and CD11. Now when you infect with a virus, this is 5 days of infection, look at the decrease that is seen at the protein level of the MHC molecule.— you can see it. And at 15— at 11 days, look at the comparison. Here's the infection. Look at the comparison. And these cells, as I told you, are alive. Now, this observation of a virus turning off a function of a cell without destroying a cell, which we termed in our ignorance a luxury function— it doesn't become a luxury function when this works as we've shown for growth hormone or for insulin. It comes into a disease, it's really an effect on a differentiated function. Was worked out well, I think, for growth hormone. So it's a common event with this virus. So for example, here is a clone cell, pituitary cell, which makes on a clonal basis in a single cell both growth hormone and prolactin. And in these experiments done by Juan Carlos when he was a postdoctoral fellow shows the effect of the loss of transcription with LCMV infection and growth hormone and not in prolactin on a single-cell basis. Now what the Darwinian selection is that favors the virus for turning off growth hormone rather than prolactin, I'm not sure. In the whole animal, of course, the growth hormone-producing cells and prolactin-producing cells have already migrated and are quite different.. But when you now infect once again the whole animal with lymphocytic choriomeningitis virus, you can see that in the C3H strain that only the cells that make growth hormone are infected in the anterior lobe of the pituitary, while in the posterior lobe of the pituitary there's no infection whatsoever. And the consequence is that you have a large animal and you have a tiny animal. If you now give this tiny animal growth hormone, it can revert to the large animal. So what I'm saying is that I think that a lot of the idiopathic diseases that we don't know the cause of, that have only biochemical markers, may well be due to viruses that persist like this virus, in which the pathology, when you look at the slide, looks perfectly normal. There's no inflammation, there's is no cell destruction. What you have is just a biochemical event in which the virus is turning off the function of the cell to make growth hormone. We've seen this with thyroid, the thyroglobulin growth hormone. We've seen it making immunoglobulins. And we've seen it with the dendritic cell in terms of MHC and co-stimulatory molecules. Well. What about the persistent infection and making IL-10? These negative regulators that are made have been seen by a number of laboratories. Our laboratory did IL-10, started IL-21, and TGF-beta, which is then finished by a postdoctoral fellow now at the University of California as a faculty member, Elena Zuniga, and IL-21 was done by David Brooks, who's now at UCLA as a faculty member. And at the same time, in the same year, Rafi Amidz group found PD-1. Since IL-10 and PD-1 are the best defined, I'll just spend my time with those. And since we did IL-10, I'll just spend my time with IL-10. First of all, we found that IL-10, by these suggestive experiments, experiments done by David Brooks in our original publication of this, was made by Clung13, as you can see, as RNA level, and not by Armstrong. And then if you looked at the cells, it was made by dendritic cells. We really couldn't— very little be made by B cells, macrophages, CD4, CD8 cells, but there was a little bit. And you couldn't really rule that out.— and the cells, the dendritic cells that made it were not the plasmosoid dendritic cells but were the common dendritic cells. Now recently, a postdoctoral fellow in our laboratory, Cherie Neg, has taken mice in which IL-10 are labeled with green fluorescent protein, infected these mice with virus, clone 13 virus, and taken out this— purified the cells of different populations, plated them, and did infectious centers. That is to see which individual cells made infectious virus. As you can see, once again, it's the dendritic cell that predominates, very little with any of the other cells, CD8, CD4, or macrophages. And this is the plot that you see. If you wanted to know which type of dendritic cell was involved, she found that the dendritic cell that was involved predominantly— you can see this is the total population, and here you're looking at the CD8 alpha negative population, or the cell population that's involved, and not the CD8 alpha positive, the so-called cross-presenting cells, are not involved in this process. And then lastly, in a recent experiment that Sherry has done provides the absolute proof of the dendritic cell. You look here at virus titers. She's now looking at a flocks CRE system. And as you can see, if you knock out CD8 and CD4 and B cells are not on this graph, you can see you have the same virus titers tighter as a control. But if you knock out dendritic cells, you can see there's a significant difference. It's about 0.003. Now we expect more mice to come through here. This is just a preliminary data that she has. Interestingly, this one outlier is the outlier that's found out over here when you measure the amounts of IL-10 that's being made. You can see it's the same relationship with IL-10. So for our mind, this is as close as we can to prove that dendritic cells are these cells making IL-10, these cells involved in the persistent infection. Now to show you the role for IL-10, if we knock out IL-10 and knock out mouse, we cannot cause a persistent infection. And This is shown over here where we convert a clone 13 infection into an Armstrong-like infection. So here's Armstrong looking at whole spleen, looking at the expression of viral antigen. So you generate a cytotoxic T lymphocyte response and 9 days after you cleared most of the virus. Now clone 13, you have persistent infection. Here's the viral antigen. You can see it throughout the spleen. If you do an IL-10 knockout with clone 13, clone 13, you can see you make the clone 13 just like the Armstrong infection. And if you look for the restoration of function, you find that— you look at CD4, CD8 function, you find that this is what's made during acute infection. This is what's made during clone 13 infection. We're looking at a variety of different immune mediators.. And now if you knock out clone 13, you see you can restore it. It's the same for all these different factors on CD4 or CD8 cells. Now if one wanted to treat a persistent infection, obviously you're not going to knock out the gene. And so that we then looked at antibodies, the IL-10 receptor, and we found that if you did use antibodies IL-10 receptor, you could reduce persistent infection in 8 out of 10 mice, and 2 that had it had very low titers of virus as compared to no treatment whatsoever. So what I'm saying is that using these negative immune regulators and blocking them with an antibody, or hopefully a small molecule, should be a therapy to restore T cell function that's exhausted and should play a prominent role in treating not only persistent virus infections but treating cancers where you also have CD8 exhaustion and cells don't function. Now this finding with the mouse is also found in HIV, though obviously you can't knock out the gene. So here in experiments done by Bruce Walker, he looked at healthy controls who had no load of HIV and had diminished IL-10 levels. He looked at natural controllers that had the same effect. But he looked at people who had chronic HIV infection and you can see the difference, both virus load is not plotted here but these had high viral loads. High IL-10. Andrew McMichael did this experiment one further in that he took those patients who had high viral load, high IL-10, he treated them with triple therapy, reduced the virus load and showed the IL-10 level was down, and took off the triple therapy. The virus load went back up, the IL-10 went back up, and of course went back with the triple therapy. Now what we see with HIV is a polymorphism with IL-10 promoter that correlates with susceptibility. And I told you it correlates with drug therapy as done by McMichaels. And similar results have been found by multiple investigators looking at another persistent virus infection, HCV. Now one of my famous— one of my favorite One of my favorite philosophers said it's a capital mistake to theorize before one has data. And I'm going to end this talk speculating on Lassa fever virus. Now, John, do you know who the philosopher is? I'm not going to ask you then, John. Now, when you look at Lassa fever virus,— or people who are unfortunate to get Lassa fever virus, they have 3 syndromes. First of all, they're immunosuppressed and have high virus titers. And those that have high titers of 10 to the 9 or higher have over 95% morbidities. In contrast, those that have titers of 10 to the 8th or lower have a 95% incidence of recovery from Lassa fever virus. Now it's a hemorrhagic fever and one gets hemorrhage and vascular disorder. And in those that survive, hearing loss is common. It's a peripheral hearing loss, it's the most common type of hearing loss that's found in West Africa where Lassa fever virus is endemic. Now we think that the hemorrhage and the hearing loss relates— this is a picture of endothelial cells— relates to the fact that the virus puts— pushes off laminin and then either opens up gap junctions to allow hemorrhage to occur or signals through a variety of molecules to allow hemorrhage to occur. We think the same thing happens with hearing, which I'll show you the evidence for that in a bit. And of course, the immunosuppression we feel is the infection in dendritic cells and the failure to arm things. Now, as we went around, went about our business publishing this stuff on distroglycan and large and alpha distroglycan, a group led by Eric Landa— Landau, rather— at MIT and this section headed by Pardis Sabatini were doing genome-wide surveys throughout the world. Pardis was doing the survey in West Africa. And she found that the gene under the greatest selection in West Africa where Lassa fever virus was endemic was LARGE. By far had the highest LOD score. Another gene that was also under positive selection but was not as dramatic, as large, was dystrophin, which is the gene that lines up the alpha— lines up the alpha-dystroglycan receptor and important for its migration and positioning. And so her data was as follows, that if you look at here the LOD p-value score, the red is for is for the mutations in large. You can see this enormously large LOD scores for people who have mutations. So the hypothesis then is as follows. To get the high morbidity, you need to have high replication. If you have a mutation in large, you have less of a functioning receptor. You get less virus binding, you get less virus entry, and you get lower virus replication. So let's look at this. Mutations at large dramatically occur in 35% of the total population among the tribe where Lassa is endemic, 35%. The hypothesis is So this means that you have approximately— of this amount, you have approximately 40% are heterozygotes, and you have about 20% that are homozygote for the mutation. Now, so hypothesize that large is under pressure and is forced in shaping human evolution. That is, those that have the mutation survive, and those that don't have the mutation will die. And as I mentioned to you, the mutations do that. So that's the hypothesis. And to prove that, we started experiments— once again, postdoctoral Brian Sullivan— with a collaborator who's a Nigerian epidemiologist physician who spends half her time in Nigeria and half her time at the Human Institute of Virology in Baltimore. And the plan is that we'll get patients' cells and control cells, and then we'll look for Lassa fever virus infection kinetics, et cetera, binding, and what happens. We're also aware that there may be non-coding mutations, and there may be effects on transcription RNA stability, protein stability, et cetera. So this, if we find this, this will be proof of principle. Now to do these experiments, we have a program with Juan Carlos and with our collaborator Kevin Campbell, and that the skin biopsies that we get will then be made into inducible pluripotential stem cells so we can isolate dendritic cells and endothelial cells from these subjects that have large mutations and to do these studies. We've also, or Kevin Campbell has also made transgenic transgenic mouse lines that express the human large mutation and that we can manipulate and add mutations back into transgenic mice. We also, through Vishnu Nayak, have SB40 immortalized cloned DCs that express human large— that express the human large mutation and can also be used for studies. Now, to do these, of course, Lassa fever virus is a BSL-4 agent, and I guess we wanted more laboratory space. We could work on it on our floor, and we can ensure that people in the building would evacuate the building. But to work with Lassa fever virus, we've made use of the wonderful reverse genetics that Juan Carlos and his postdoctoral fellow Sebastian did in taking a 2-segment virus, making 3 segments, and allowing one to make recombinant viruses. And Andrew Lee, postdoctoral fellow, has made a Lassa fever virus glycoprotein on an LC-MD clone 13 core, which we first tested in BSL-3. Showed it was not harmful for the mice or for us, and then we could use it in BSL-2. And from these pictures taken by Andrew, you can see the virus replicates well in liver vessels, in the kidney, heart, and in the adrenal gland. So we now have a virus that we can use safely in BSL-2. Now what about the stability of the membrane and that hypothesis? Well, what we've done is taken inactivated virus that's also sorlin inactivated as well as UV inactivated, Lassa fever virus and LCMV certified by the CDC as being inactivated. So we worked at it in the molecular biology building, I might add, in BSL-3. Which would displace laminin but not really infect the cell. And we did this using cells and membrane that we put a dye in, FM-143. And on the control, this is the type of leakage you get over time. This is the bottom plot. And now if you do inactivated Lassa fever virus or LCMV, you see you get much greater leakage. Plot over here is highly significant. So once again, we think that pushing laminin to the side opens up— probably opens up the gap junctions and allows this to occur. We haven't proven that, but that's what we think happens. Now, in terms of the hearing loss, we did some experiments with Aura around the cat at the Rockefeller in which we took Schwann cells that we cultured with dorsal root ganglia, and these form myelin. So here is the culture, and here is the axon stained with myelin, stained for myelin basic protein. Myelin's being formed. Here's the stain for laminin. You can see it's the same cells, laminin. And looking by electron microscopy, you can see that myelin is being formed. And this is what we see in a regular culture, or if we lay out Armstrong, which doesn't bind to myelin. Alpha distroglycan, and there's a lot of alpha distroglycan on Schwann cells, very heavy amount. However, if we bind either Lassa fever virus or clone 13, you can see the binding, you can see that we get no myelin formation, we get disordered laminin, and it's very difficult to find any myelination. When you do, it's really an atypical or aberrant myelination. That occurs. Now we think that this is related once again to pushing off of laminin that's necessary to maintain a variety of molecules downstream signals to lead to myelin, and we're investigating this as you'd expect with, with gene arrays and with sRNAs and other techniques at present.— so after all this work, where are we? And this is where we are. And this is what we're digging for. We're decoding molecular basis of signals of how viruses persist. We want to dissect the advantages that the polymerase of clone 13 has. We want to discover small molecules that can block negative regulators. I should say by doing the ENU experiments, experiments, and looking at 2,400 mice which cover about 20% of the genome, we calculate with Bruce that there are about 100 genes that are involved in this pathway. We know that they might act separately. We know that PL1 and IL-10 act separately, and we're looking for small molecules that will effective. We want to know about the large mutations which occur in the Yoruba population, but picked up by the genome-wide screen, whether they have biological relevance or not. And if we can— if that's possible, we then want to look for serological markers that could be used to tell who's going to be susceptible or resistant to Lassa fever virus. And then Campbell wants to develop glycopeptides of large that can block Lassa fever virus alpha-dystroglycan binding, and we want to look at the signaling mechanism for the suppression and the signaling on the endothelial and the Schwann cells that I just told you about. So this is the system that we'll use with gene profiling. This is the system that Vishnu and Hannah condition developed and then Hannah Lowicki stained and you can see that these cells are infected by clone 13 of the last efflux virus recombinant and not by arm. We were just looking at a marker for infection. We have these as all the toll-like receptors or toll-like receptors diminished for a variety of reasons.— and this is the experiment to look for small molecules in which if we take out spleen cells for an animal that's been persistently infected and we add peptides or anything to it, these cells are hypo-responsive and non-responsive. And this is the amount of interferon gamma they make. If we then block this with an antibody to PD-L1 or antibody to IL-10 or both, you can see we can restore function. And the idea is now to screen a variety of libraries and to see if we can restore this function and identify those molecules and then manipulate them and then go back and test them in vivo in our system. And so this is the end of my journey. These are the postdoctoral fellows that I want to thank who are currently here. Brian Cherie, Andrew Daniel, John Simone and Kevin. Andrew is leaving for a faculty position at Southwestern Texas, and Daniel will be leaving for a faculty position at Case Western. Then our collaborators, I've told you about Juan, who's a long-term collaborator, Stefan Kuhns, Kevin Campbell, Vishnu Nayak, Pardis Sabatini, and Alashli Abumukhu. Thank you.
More from Joshua
Scientist Stories: Hans Waterham, The value of phenotypic assays in peroxisomal disorders
Scientist Stories: Joel Zlotogora, Genetic carrier screening programs for reproductive purposes