i think there are some people here who haven't been here before, am i right? is there anybody? is this the first time you're here? okay. so jeffrey, what is this a picture of?
>> brooklyn bridge. >> why would we have a picture of the brooklyn bridge in a course called demystifying medicine? >> [inaudible] maybe tying the bridge to science to clinical medicine. >> you got it.
>> that's correct. so believe it or not, this picture was actually taken by my grandfather, and it the brooklyn bridge, and the point is that the purpose of this course is to put us like the two gentlemen on the cat walk, around we're connecting brooklyn and
manhatten. it's important when the bridge was proposed, the people in brooklyn objected. they didn't want these city slickers coming into their rural area. the people in new york said we don't want those country
bumpkins coming across and spoiling things in manhattan they built the bridge and everything flourished. there is a moral to this story. because by connecting the incredible advances in basic biological and engineering scientists with major health
problems, it's not so easy. because everybody speaks a different language. so one of the purposes of this course is to have people communicate, and by the way, if people are using a language you don't understand, technically speaking, speak up.
and ask them to define what they're talking about. of course i think that the lack of this communication is probably one of the most difficult serious problems that we confront. i like to think of science as having progressed in the past 25
years as almost logo rhythmically. around it's so-called application to human health, sometimes even human physiology, you can think of arrhythmatic. so this gap gets bigger and bigger and bigger. just throwing money at it
doesn't solve the problem if people don't speak the same language. so we're about to go between brooken and new york and backwards and forwards. the topic today, hiv, is one that we have had as part of this program here at nih.
i think 11 of the 12 years that we have been doing it. and during that period of time, almost everything has changed. the epidemiology has changed. the public awareness has changed. the science has changed. but the problems still remain.
they remain in the science and they certainly remain in terms of public health and knowledge and so forth. and in the global aspect of it, we now have super imposed the major senders in the world where hiv is rampant, particularly in southern africa.
so the problem is a biological challenge, is as challenging as it was almost since the day that the virus was first discovered, and as a public health challenge, although it's been changed to a treatable disease, with the exception, perhaps, of one individual on earth, no one
has ever been cured of hiv. so we also have a bit of a celebration today. because the first speaker, a friend and colleague, john coffin, this is the 20th year that john has been the eastern airlines professor of microbiological, molecular
biology. what do i meanie that? well, john, we were back with members at tuft's university in boston where john still is, he's an american cancer society professor. twenty years ago, he was recruited by harold varmus to
head an hiv drug resistance program here at the nih. and he has dutifully been commuting back and forth. he's outlived eastern airlines. now he's on national airlines, i think. but this has gone on for 20 years now.
and enormous changes have taken place in which he and his colleagues have been major participants. john is a -- one of the world's experts in retro vairology. certainly one of his major contributions to hiv is the recognition that crewing
resistant forms of the virus are present from the very beginning. he will describe this and discuss other, and i'm sure exciting things that have developed in the past two years. last year he had a vacation and didn't teach in the course. so it's a great pleasure, as
always, to have john coffin take part. john. >> thank you very much. it is a pleasure to come back yet again to give this course, give this lecture. and a great pleasure to do it for the first time with my good
friend and colleague jeffrey lifson. we're both in the same building, rather different employees. we're -- neither of us are a government employee. we're both contractors, so you can take that for what it's worth, i guess.
so i'm not going to talk about drug resistance, although that was how we got started, which we call the hiv dynamic and replication program. that's because drug -- the focus has turned away from drug resistance somewhat, but we had to keep the same acronym.
that was very important to me. so i do want to talk about what it is about hiv. one of the medical miracles of the last part of the 20th century was the discovery of anti-retroviral therapy that could almost completely eliminate the progression of hiv
infection to what had been the inevitable death sentence of aids until that time within about 20 years, virtually every perspective infected with hiv would die of aids if they didn't die of something else before that. and when modern combination
anti-retroviral therapy came along, that actually brought that death rate down to close to 0 as long as it was used properly. however, there is a -- the downside of that is that the therapy is not curative. it, in fact, if you look at --
if you monitor hiv infection, the best way to do that is by monitoring the amount of virus that's been released in the blood stream. the reason that's a good indicator, the virus is an indicator of the virus infecting cells else where, probably
mostly in tissues. but it turns out very, very fast. at any given time the amount of virus you see is a good measure of the amount of virus being produced. that's a good measure of the number of infected cells.
when you start therapy, the -- before you start therapy, the virus remains at a pretty steady state level. usually averaging around 20 or 30,000 copies per milliliter of rna, which you can do using pcr, you have good quantitative rna measures.
when you start therapy, that virus drops very rapidly, and goes down by several logs to a level which you can no longer detect it with a standard clinical assay, 25-50 copies of rna per milliliter. if you use more sensitive assays, the level, you can find
that the level actually of virus remains detectable with the more sensitive assays dropping down to, and then leveling off to a level of maybe 3 copies, on average, with a big variance around that average, in multiple phases that i won't go into the reason for.
and this will remain pretty stable in a patient for many years or decades even. any time that after, this in any-- almost every patient that's been looked at, if you stop therapy, the virus comes right back again. it comes back pretty much to the level where you started before.
so there actually is both a reservoir of some kind of virus that allows the virus to come back, and also, a memory of the host virus relationship so that the level actually returns to very much like what it was before. if you monitor instead of the
rna, if you monitor the number of infected cells, hiv is a retro virus, and replicates by copying its rna into dna and causing that dna to be integrated into the cell, i'll tell you more about integration later on. then you find a rather different
kind of curve. it drops a bit, maybe ten fold or so initially. then levels off at a much higher level than the rna relative to what you started with. and that means that although you have a four, five log decline, 10,000 to 100,000 fold decline
in rna, you only have about a ten fold decline in dna. and that means that of this dna, the fraction of this dna that's devoted to making the virus at this point is much, much less than it is here. that's actually another -- instance of an adthen i like to
use which is in hiv cure research, i think it's also true in hiv vaccine research, is that the better the science, the worse the news move because that means that first place, the viral -- virus infection cannot be cured. it cannot be cured clearly
because there is persistence of dna. but also, and this has been very elganntly shown by bob's lab in baltimore a few years ago. the amount of dna that you see here is not a very good measure of the number of cells that are capable of giving rise to the
virus. because that's only a very, very small fraction of these cells that have dna in them, the virus that will give rise to this what the lab showed was only a very, very small fraction of this dew point is in tact. most of the dna is defective
mutants of various kinds that have found their way into very long live cells. these are mutant cells, cd4 positive t cells. their job is to respond to an antigen, then go into a latent memory state that can last and survive for years and years so
the immune memory comes back. the virus has found its way into cells which can expand rapidly and make lots of virus, and can go into the resting state, and live for very long periods of time rendering it effectively incurable by agents that block virus infection.
such as standard antiretroviral therapy. so what's going on during this time? what is the nature of the reservoir? again, the silicono lab first showed and we confirmed, this shows an extension of that
experiment, that here is another-- this is now a real patient with time on therapy, you can see the drop in the viral -- the level of virus in blood. you can also see the smaller decline in the viral dew point in the same way i showed you before -- viral dna, in the same
way i showed you before. during that time, if you follow the genetics of the virus by making these trees, what you find is that with time, if you look at the earliest samples, the virus that you see in those early samples is very, very diverse.
there is no two sequences alike. however, with time, what you see is two things. one is that the virus does not evolve at all. these branches, evolution would be measured by these branches getting longer. i'll show you a case later on
which is kind of a positive control for this kind of experiment. instead, the virus remains evolutionary -- in evolutionary states. property retro viruses and hiv, as they replicate, they accumulate mutations.
you can use that to tell you whether real replication of the virus is going on. what you do see, however, is not the accumulation of additional mutations above what was already present in the diverse population that was in this patient, but you do see the
appearance of these rakes, these groups of absolutely identical sequences that indicate that something has happened to cause a clopal expansion of one particular virus genotype. and a large part of what i'm going to tell you is to try to convines you that this colonial
expansion is not due -- it could have been due to just a virus breaking through, and then clemly rapidly i don'ting -- outgrowing everything else in the body. the evidence is very strong that i'll show you, it's not due to it's due to the expansion of a
single cell which was infected with virus probably before therapy started, and then that cell has expanded as probably what would ordinarily do as part of the immune system. that's what immune system cells do. if they encounter antigen or
homeostatic signals, they will divide into rather large clones, particularly in the case of a new antigen. so in facts, you could see this, you can see thee rakes of identical sequences. this is four patients showing the same kind of thing.
and these appearance of these identical sequences and the failure of new genetic distance to occur, where these open circles are pre-therapy, and the closed circles are after a long time on therapy. and you can see, again, these plots at the bottom are just a
way of measuring the amount of genetic variation occurred with this time. and if these -- if the virus were replicated -- i don't have a control for this. if the requiris were replicating these knots would look like. this instead they're absolutely
flat. and that's also extending to the virus that comes back. if you stop therapy, in this case, after 7 years, in this case after 5 years, the virus that came back again had no additional accumulation of genetic distance that the virus
has shown by the orange circles, but did have the same kind of colonial outgrowth. what i was showing you before for the bulk rna is also true for the infectious virus, the virus that came back once therapy was stopped. in the examples i just showed
you, we started with a patient wh had been treated in chronic infections so the virus is already very diverse. just to show you one example, which shows that much more clearly, we also did a study with some babies in south africa, you don't get infected
babies in the u.s. anymore. thanks to preventive therapy. and who had he been treated at a very early age, worn infected, and their virus was much less this first example shows a patient who -- where the sample was taken at a couple months of age.
therapy was not started until much later. and then the sequences were examined at the age of 8. and in that case, you started out with very much less -- much less diverse viruses in the colored circles here. and -- but you could see the
appearance of evolution of the virus, that we infer occurred during this time before therapy was started. that serves as a control showing what happens when you do have some ongoing replication. however, if you have good suppression, what you find and
this is a patient now who suppressed, first sample was taken at the time, anti-retroviral therapy was started. very early on. and then treated for 8 years. this shows that the virus -- here -- this is the idea call --
this is probably the virus the patient was originally infected w most patients are infected with a single virus particle. early on, the virus that spreads in the patient is colonial before diversity starts to accumulate. a little bit of diversity
accumulated. when you look at 8 years, after 8 years of treatment, you find the same sequence is present and diversity very much like you started with. there is no evolution whatsoever going on in this patient. is this point because there is
a paper that appeared in the lt. about a year ago, completely wrong on this issue, badly flawed. we're trying to get a rebuttal published as we speak. it's a real up-hill slog. it's nice to careen up the literature.
so there is no evidence we can find for the role of hiv replication in maintaining the reservoir. what does maintain the reservoir, what we visualize is that before you start therapy, you have a mixture, short live cells, cell gets infected with a
the virus replicates in that cell, as part of the replication or part of the immune response or both against the virus that cell gives off a burst of virus and dies within about a day and a half or so on average. but during this time, a very small fraction of cells get
infected, and these cells go into some kind of a latent state. it's probably cells that were activated here and in the process of transitioning back into a memory state. these are immune -- these are cd4 positive t cells by and
large. and these long live cells have pro virus expression very strongly suppressed, and on anti-retroviral therapy these get stripped away. they no longer get infected. these, however, will survive. and thee -- two things can
happen. one, they'll gradually be eroded we natural processes. or maybe because they occasionally make virus and die from that respect. but at the same time, there will be -- others will actually undergo clonal proliferation.
this would give rise to the rates of sequences that you see. may pretty much maintain the numbers of -- total numbers of infected cells in sort of a pseudo homeostatic process. so the numbers of infected cells remain about the same. you can see that there the flat
dna curve i showed but clonal populations appear. we developed or improved on an assay for looking specifically and integration sites. i won't talk too much about integration, except that it's in hiv dna, integrated many millions of possible sites.
we estimate there are about 10 million possible sites for integration in a cell. only small fraction of the dome, but a lot of sites. the probability of it happening twice in any given sample of reasonable size independently is very small.
long life cells sites of integration are determined by initial preferences but they can also we modified by selection or by chance. by the lucky one that integrates into a cell that's going to respond to some antigen, and divide like crazy.
so we have -- to look at this, we've look at integration preferences in different kind of cells. i'll show you an example of white blood cells and hilo cells, very cell line in culture. this shows the distribution of
integration sites across the chromosome. i picked 16 here for reasons that where become clearer later. and you can see that very strong hot and cold regions. the scale is megabases, this chromosome is about 80 megabases, 80 million bases
long. this just shows the accumulated integrations within 1 million base pair chunks. that sequence. if we take a small fraction of the genome, it shows a bit right here, and expand that so now we're opa scale where we can see
individual genes, we can see there is a very clear integration pattern shown actually among others dr. levin, in the front row here. the integration very strongly favors genes in these cells, so you can actually really determine and -- i don't show
you this, but you can determine where genes are by looking at integration patterns, perhaps a hard way to do it. you can also see that hella cells, the pbmc said, this is just in the cd4 positive cells. you can also see the same pattern, smaller numbers but the
same pattern. the same pattern in pbmcs which were whether or not they were stimulating by adding pha to them, activating that stimulates the cells to divide. and in fact, we've looked at a number of different cell types. there is a very, very strong
correlation of integrations. almost all the integrtions sites from one cell line, type of cell to another, and the preferences are very, very similar. you can see one case in hela cells of a gene that's not hit in these others.
we presume it's expressed here, not the others. so hiv infected cells, sites of different preferences and by selection after integration. so what happens, distribution, if we look at patients on long term anti-retroviral therapy. i'm going to spend the rest of
the time talking about just this one patient. i could not bring the patient along because sadly, he died about 3 years ago of a malignancy. probably unrelated to the hiv infection. and he was -- he reported to the
clinical center very sick. he was down to about 10cd4 cells per mickro liter. normal for probably we have about a thousand cd4 cells, so he was really -- he had frank aids. he had probably been infected -- this was 1990, about.
sorry, about 2000. heed been infected since about 1990, from the severity of the disease. he responded very well to antiretroviral therapy with a few blips along the way. he lived in a rather remote region of the u.s., iron the
middle of the virgin islands, so had some breaks in ability to get drugs. and as a consequnce, had some rebounds of virus. and then toward the end, toward the end of his life as it turned out, the virus, while still on therapy, the virus was --
started to creep up. the level of viremia reached no where near the original levels. and so we looked more carefully at the rebound virus. you could see here this is just an enlarged view of what happened. and it was clear, as i'll show
you, that the virus was becoming, in part, resistant to the antibiotic antibiotic so the therapy was -- antiretroviral he was diagnosed with a malignancy right in here. we think that is also related to this increase that i'll show you, and i'll show you the
reason woe think that. at this point, we pulled the virus again, this is out of his and when we sequenced it and put it on this teach you could see 2 different populations of sequence. one, a diverse one indicating ongoing replication of the
and that ongoing replication was made possible because this -- all of these sequences here had drug resistance mutation. this virus was becoming resistant to the therapy that the patient was receiving at this point. this virus was not and also
every sequence of this virus that was season was identical, and was -- so this was one of these clones that we talked about before. when the patient went on to therapy, this resistant virus went away, the therapy was switched to things that the
patient was predicted to be sensitive to. sure enough, it got rid of what we thought was a replicating population, showing that it was a replicating population. this completely drug sensitive population remained behind. that was consistent with the
idea it was not a replicating the virus was being produced by a clone of cells that had expanded greatly in this patient. so when we went back we found that if you looked, if you now put these on the same plot, again, the genomesome 16 map,
instead of this nice pattern of integrations in genes, all the genes in this region, all the integrations were actually focused in one gene. that's a gene called mkl2. if we blow up this pattern here, shows a very dramatic difference.
we infer that in all the patients -- this is what we start with. this is what happens when you infect fresh blood cells essentially. this is what ended up, at least within this region, in this patient. there were other pro viruses in
other regions, this is the one that ended up here in this region. if we blow this up it's even more dramatic. if you look in the infected bpmcs at the integration sites, you see they're scattered all over the place.
the different colors indicate different orientations of the pro virus. pointing this way or that way. and random orientation, they're more or less random, although as dr. levine has shown, there a tendency of the pro viruses to prefer the five prime end of the
genome near the start site of transportation and taper off during the three prime end. these vertical lines are the exon, the coding parts of the gene. this gene is mostly in tron with tiny exons. and when you look at the
integration sites in this patient, instead of being distributed across, they're dramatically focused in this very small region. if we where up just that small region which is only about 7 kilbases, the whole genome is about 3 billion base pairs, this
is about two parts in a million of the total genome. all of these integration sites out of about 1,000 that we saw are there. the ones that are circled were seen more than once, and are -- to have been in cells. we think all these cells were
expanded. these are the ones that we could actually see as having more than one integration at the same site in different cells for reasons i can't go into now for lack of time. we also saw exactly the same kind of thing in another gene,
gene called bak2. it had been shown in a mouse experiment by infecting the mouse with another retro virus. that caused to be -- that was the cause of b cell, again with integrations in exactly the same region of the gene. mostly within this one rather
large intron in this case. so -- and again some of these were amplified, and they were in-- again, in a very small region, a very small fraction of the gene compared to what you saw in pbmc. clearly, there had been selection -- the only explanation, there had been
selection for integrations in this region. and this, if you're an old time retro virologist like me, this smells like what happens when retroviruses cause cancer. you get integrations, randomly, but then because some effect the expression of oncogenes here,
those then cause that cell to grow out as a cancer and clone, basically. so the hiv cannot only be in expanded cells, it can also participate in the colonial expansion of those cells. it doesn't always do so. in many -- i'll show you, there
are many other examples where you get colonial expansion, where you can't pipit on the virus doing anything. you only see it once. those are where the cell has been caused to expand by stimulation or some other means. so it has been thought that --
i'll show you one more pro clonenally expanded hiv inspected cells. we falsified this by the example i'm about to show you. when we looked at integration sites, combining this patient and a couple of others, the paper published a couple years
ago, we found some of the immigration sites were in the genes, some hiding down here. there were a lot of others at other sites, some which were in -- unknown, in sequences that were not known to be near a but the one in particular is what we call amby, ambiguous.
when we looked at the sequence, it was present several times in the genome database. we didn't really care which one because i think that's irrelevant to the point. that sequence by pcr amplication, we can amplify the whole thing starting with a
primer from the cell dew point, and going into the fires, starting with a primer going into the virus. and then we get two overlapping fragments. and this you could cotransfect into cells. when we did, that we found that
those cells would make virus and the virus was identical to the virus that we had seen in plasma, as being the prominent non defective one. non resistant one. so this genome was not effective, and it was infectious if we transfected cells with it.
it was also the prominent virus. this is a quick summary of the patient sequence that i showed you, patient's history that i showed you before, just espanneded. and -- expanded. this pro virus, the sequence of this pro virus was identical to
the sequence of this prominent virus we saw in -- appearing in we could see a sample, the same sequence, from a number of years so this was obviously in a clone, a clone that lasted a long time. it came up approximately -- we have a little more data on this,
it came up about as the tumor that was in this patient was growing, it decreased somewhat as the tumor was treated. and that led us to suspect, okay-- first let me just say that if we take now pbmcs, activate them to make virus in an endpoint way, again, we find that at the endpoint, most of
the virus that was produced we these cells was done by john meler, a colleague at the university of pitts burling. most of them produced owe virus idea idea call to amby. so chemical agents -- was, was also again this amby1 pro virus. there were a couple others that
came up that much rarer, but that we haven't identified the viral dna, the dew point for in these viruses. that is a real problem with trying to study this, in that if you look at the dna in the pro viruses, the amount of dna sequences that correspond to the
virus that you see in plasma are extraordinarily rare. most of the time you can't find them. you can look at hundreds and hundreds of dna sequences. none of those correspond to what's in the mastma. the virus you see is made by
total infected cells. those are the ones, however, that you have to deal with, if you want to think about curing i'm not going to tell you how we're going to do that. i don't know. so this -- this is still true, the only known case where -- we
can identify the cells responsible for infectious infected virus in the blood. it's in some kind of epigenetic we know a little bit about what might be driving clinical expansion. this patient unfortunately died. we were able to get autopsy
samples, quite some time after the -- after his death. unfortunately, the pathologist doesn't get around to the virgin islands very often. but what we found was that the virus was present and, in fact, cells containing this virus were present and, in fact, were
strongly enriched in metastatic lesions of this cancer. you can see cd4 on tumor infiltrating lymphocytes. there were quite a few in this tumor sample. and with a good p value, there was strong enrichment of these we hipized that the -- mike
pencized that the -- that the cells expanded because they were reactive to a tumor antigen, basically tumor infiltrating lymphocytes that you may have heard about. mary, a colleague in frederick, developed a very nice assay for looking not only at the dna
sequences in bpmcs, but also pairing the sequences with a level of rna expression in each of those. i won't go into the details on but what she saw, this is -- this is now a very small fraction of the total tree that was done by her and by
[indiscernible] a student in frederick. this is the clone, and what this shows is that these are the dna sequences, and this pbmc sample that you looked at. these are the rna sequences that were expressed we those cells, however -- by those cells.
however, the dna sequences was a very small sample cared to these. so in this sample, we've actually expanded this to much larger numbers now, of 384 infected cells, only 4 of them had detectable rna and none of them had very high levels.
the number of copies here is more or less corresponding to the number of copies of this rap that rna in these cells. it's expressed at very low levels in cells, only a few percent. that is how we presume the cells survive. most of the time the cells, this
clone, we estimate has expanded to about more than ten to the seventh cells total over the whole patient. done so about the virus becoming expressed enough to cause the cell to ki by virus replication, or to die because of the immune response.
so our hypothesis from this is that the cells get infected, a small fraction -- i remember, we were talking about very small fractions of the total cells. ten to the 9 cells infected every day. we estimate. of those, a very small fraction,
and there is a latent state where the cell can still i don't and espanned, be activated by antigen, so on. a very small fraction of these produce rna spontaneously. probably a much smaller fraction actually produce enough rna at anyone time to make virus.
and so it's -- it's going to be a real challenge to be able to look at specifically these populations of cells, and these are the populations of cells you have to root out if you're going to try to actually consider how to cure hiv infection. it's a huge challenge.
and so to summarize, integration patterns, the last part of this, integration patterns are very they can be readily identified in cells. this actually was a technical -- it's only 1 in a thousand cells on average in these patients that have a provirus, to be able
to identify it in sufficient numbers to make conclusions about expansion and so on is quite remarkable. that's thanks to a talented colleague up at nci frederick. the integration site may, itself, influence persistence uncommonal expansion, therefore,
the rebound can be mapped in individual integration sites. and expanded clones can contain niches pro viruses, this was under debate until recently. amby1:s can be expanded in response to tumor antigen. so what i showed you is a very small fraction.
i feel obliged to show you the whole tree but it has to be done in a certain way to really appreciate it. [music playing] and andramatic responsible for the tree you just saw along with lots of collaborators at nih, south africa, whodid the baby
studies. and many others. thank you. [applause] >> thank you, john. that was brilliant. may the force be with you. we have time for some questions. we urge you to ask anything
that's on your mind. don't feel particularly amongst the students or fellows. here is a student. >> something every day. so can you go into more detail about the tumor cells you saw? did they have an integration event?
>> we don't think so. we tried to look for that. i think, however, the possibility that hiv itself could be oncogenic remains to be investigated. if you talk to clinicians who studied hiv and cancer, they will pew-pew the idea.
i think it needs some investigation. hiv tumors -- t cells lymphomas you think would be the prime candidates. they are not common in hiv infection, i don't think they are anymore common than uninfected patients.
they're very rare anyway. there is something about the biology of t cells compared to b cell. i also argue that the reason you don't see t cell lymphomas, if i get a t cell transformed in some kind of precancerous state, that cell is now activated. a very good targeted cell for
hiv infection, in the hiv that's around at the time will wipe it out. hiv doesn't infect b cells. b cell limboa is the main disease for at least some cases, we have no known viral cause for the increase that you see. >> what about neurological.
>> ten to the -- t cells are infected every day with hiv. tell me that one in a million b cells can't get infects, i'll tell you you're rock, probably. >> there is no association with brain cancers or neurological -- >> not timing. no. -- not to my knowledge, no.
there is other neurological issues. >> john, are the long lived cells the only reservoir for hiv? >> i think so. hour, there are still people that like to talk about macro fags, mon oh sites, other cells
that can be infected at least in the lab with hiv. nobody has yet come up with compelling evidence that these cells form a reservoir. nobody has compelling evidence that they don't, either. but there are some studies which really down play the role of
myeloid linage cells. >> did some of these long lived cells actually reside in tissues? >> sure. most of them do. they're circulating everywhere. only about 2% of cd4 cells are in the blood at anyone time.
and so -- and they're moving around fast. contrary to what some people might say. there are people that like to say lymphoid tissue, for example, is a separate compartment from blood. away don't believe that at all,
and we have studies that show that any given time, at least on patients on therapy, it's exactly the same virus population that's present in limp noticed and -- lymph node and it gut or in blood. cells -- there are some cases where there is
compartmentalization. if you look in the genttle track, for example. that's probably because the infection is quite rare. also the brain sometimes can be kind of seeded:ally. over most of the body, the virus in my upon is very well mixed.
>> do you think there is any relationship to the distribution of the long lived cells and the microbiome? >> i don't know. >> there are studies that show the lymphocyte population -- >> there are. whether the specifics of what's
in the microbiome, that redistribution the gut and the rest of the body actually has an effect on the hiv populations that are there, i think, really remains to be seen. might be studies that are better done in animal models. where you can get more better
sampling. >> what about an update, public health standing? what number of infected people are there throughout the world and how does that compare? >> still, the number of infected people worldwide hovers around 30-35000000.
of which if i remember correctly, somebody can correct me if i'm wrong, there are about 2 million deaths a year. there are somewhat -- there are about 3 million new nexts. probably -- infections, probably still going up a bit. death rate is going down
somewhat, even worldwide because of fairly good access, even in some rather poor countries to anti-retroviral therapy. there are still real issues of getting used effectively. particularly monitoring the course of the infection, for example, the ability to switch
therapies. which you can do in the united states now. nobody who is diagnosed with hiv infection should die of aids. the -- but it still happens, you have been, at a pretty good clip. about 30,000 new infections a
year, a few thousand deaths, i but everybody knows how not to get infected with hiv. shouldn't be necessary to spend billions of dollars to deal with it. but it is. >> thank you, there will be time for more questions afterwards,
so thank you, john. so our second speaker is jeffrey lifson, who graduated from northwestern medical school, went to stanford to take a residency in pathologist. this correlated about the time that hiv was emerging and associated with blood banks,
where he was active as part of his pathology trach. that's how he got interested in hiv in the first place. he, then, spent several years in a bio tech company developing new ways of looking at hiv, of understanding it, and came here to the nih in 1995, where he is
at frederick, and his major interest has been in non human primates as models of hiv. and many other aspects of hiv virology, including some consideration of vaccine development. at present, he's a senior principle scientist and director
of the aids and cancer virus program of the nci at frederick. so thank you very much for being with us. >> thank you. and thank you for the introduction. the more we learn, the better the science gets, the worse the
prognosis gets for solving some of the remaining definitive problems for hiv. and those really are coming up with a way to cure infection. we have, for those that have availability to combination anti-retroviral treatment, very good treatments.
as john indicated, those are not curative. and coming up with more definitive treatment that can eliminate the requirement for daily anti-retroviral drug treatment is a current priority. similarly, despite a lot of work by a lot of very clever people,
we still don't have vaccine that's capable of robustly preventing infection. what i'm going to try to do today is to not necessarily in the way of making excuses for my field, but just to give you, as an non specialist autopsies, some perspectives on why it is
that inspite of all the progress we still haven't achieved keel milestones. so we'll demystify the title slide. i'd like to start with a confession. you heard the first part already, in the 12 years that
this program has been operative, i have been always intrigued by seems like a wonderful lecture series. i love the concept, i love the underlying premise. i love the idea of patient involvement. and somehow, i think partly
because i work up in frederick, this is the first time that i made it here. so that's my confession. as i said, i think the -- i love the concept of having a patient involved in these presentations. i have been involved in some clinical work in my past, and
it's a really compelling part of one of the things that makes my work so interesting is the transrational aspect of it. with that said, these days most of my patients are actually not people, but are non human primates. so i did consult with our vets,
asked if i could bring one of our infected rhesus mick aces to build the building. we'd have to accept a proxy for -- and then finally, i want to start with a disclaimer. i am actively involved in aids evacuation research. it's not the sole or the
prominent focus of the work that i do, but having said that, there are multiple people on this campus that, for a specialized lecture on aids vaccine development, you have a whole building of them over in building 40, the vaccine research center.
if you want to hear barney gram give a great talk, you can log on to the meeting held a couple weeks ago and check out the webcast. larry will be on campus in a couple weeks, heads up the hiv trials network. you can hear larry.
but as i thought about this talk, and initially felt a little bit uncomfortable about presenting opsomething that i don't necessarily consider myself an actual expert on, then i realized this is a very broad audience, that maybe it might be bet to have someone that isn't
going to get all down in the weeds on specific details of things. and so instead, the approach we'll take for this lecture is something that akin to this, where i'm going to try to give you a perspective of someone who works in the general area so
very sophisticated cocktail party conversation, a sense of what the remaining challenges are in aids vaccine development and where the field is today. vaccines depend on the concept of acquired immunity, goes back to the fifth century bc. found in a passage, history of
the war describing the playing of athens. the pathogen responsible is not entirely clear to scholars. what was clear, and very vividly accounted, is in this passage, it was though who had recovered from the disease that the sick and dieing found most
compassion. they knew what it was from experience, now had no fear for themselves for the same man was never attacked twice. this is the first historcal example depiction of what we understand is acquired immunity. obviously, the underlying
scientific concepts in terms of the basis for acquired immunity was not well understood at the time, but this is really the first understanding of it. and vaccines are based on this premise. having subjected you to [indiscernible] and the history
of the war i'll skip the otherwise obligtry cow pox over the will succeeding centuries, we have come up with all kinds of vaccines. there is a long, long list of very successful vaccines. having said that i'll make the case thaten contrast to the
situation for hiv, all the successful vaccines, and this is not in any way to down play or minimize the difficulty in coming up with those vaccines or the contributions of the people that have come up with clever ways to generate the vaccines, but to some extent, all the
successful vaccines represent combination of a triumph of successful empirical testing, combination of luck, and for twoiously targets. i'll convince you, none of those things to apply for hiv. as an example of that, this is a photograph from a very famous
press conference from almost 33 years ago, that's a young bob gallow. then secretary of health and human services, margaret heckler who famously gave a long extended quote. hope to have a vaccine ready for testing in approximately 2
years. that was the spring of 1984. mark gret heckler gets a lot of crap for that quote, sometimes taken outs of context. i think people read that quote and interpret it as we'll have a vaccine that prevents hiv infection in 2 years.
that wasn't what she promised. she said we would have a vaccine ready for testing. and we didn't miss it pie that much. it's just that the vaccine that was ready for testing wasn't very good. and the vaccines that were
initially tested, it's hard to go back to this, and we're going wack 30 plus years. i have been doing this for longer than looking around the room, some of you have been around. but that said, the initial vaccine development efforts was
taken place in an environment where i think the field was flushed with the success of initially purified, then ultimately recombyinant subunits vaccines that were preventing hepatitis b infection. people thought this was terrific, we have this whole new
technology. all we need to do is in a turnkey fashion, apply the same thing that worked so well for hepatitis b to hiv. and that will solve the problem. and long story short, it didn't come anywhere near close to solving the problem.
which turned out to be much more complicated. surprisingly, over the intervening 33 years, we have only been a total of 6ish studies that have been conducted at a large enough scale, phase 2b or later, with the potential to identify potential efficacy
of a vaccine. the first two were just recombinant subunit approach, dismal failures. the next two were related trials using a recombininant adenovirus approach, trying -- cellular immune responses. and then the one sort of partial
success is the r.v144 study. showed a marginal estimated about 31% efficacy. intriguingly was a study that was based on -- that a lot of people thought should never have been done, based on combining two components, neither of which worked very well together into
something that worked just barely detectable level. now there is currently a large study going on in south africa trying to evaluate something that's probably best described as an extension of the underlying approach used in that trial.
there haven't been a lot of efficacy trials in part because the results from earlier stage column trials, and preclinical results haven't been encouraging enough to justify large scale clinical trials. why is that? let's talk about some of the
underlying conceptual premise, what has enabled successful vaccine development for other targets. one thing that's important to realize is that a lot of people think about vaccines and simplist notion, you vaccine someone and that prevents
that isn't always true. they typically control and clear infection, and prevent it from becoming clinically manifest. and that's an important distinction, because for some pathogens that's a perfectly acceptable outcome. if you were paying attention to
john's presentation, you realize that hiv is a sneakier pathogen than many, and one of the things it does is to integrate into the chromosomes of some of the cells and it can establish a long state there that doesn't go away until the cell goes away. so simply controlling the
infection which we can do quite effectively with antiretroviral drugs, a vaccine response that does only that may not be enough. similarly, for most. things foreweigh we have effective vaccines, there are clinical examples of basically naturally
occurring or post convalescent immuneity. individuals that survive, become refactory to reinfection. in many cases we can understand an immune correlate of that either at the level of something that is a causative correlate, to say an interventional study
you can show that a particular response is responsible for that protection. or a surrogate marker of you can sew show it may we that this particular -- measuring this particular antibody response may not be directly involved in protecting against
the pathogen, but correlates well enough that you can use that as a reliable marker for whether or not your vaccine is likely to work. finally, animal models have been critical to development of a lot of vaccines for hiv. i've spent the last 20 years
trying to develop better animal models. hiv itself doesn't readily infect other species. and so what we do is work with simeon, relative -- related viruses that recapitulate many of the key aspects of hiv infection in humans.
we can generate chimeric viruses, that infects monkeys. really with colleagues in new york, we have been generating minimally chimeric hives that have minimal changes, 95% hiv sequences but able to productively effect mechanics and cause disease in those
animals. so there is room for expansion and improvement of the animal models but they're better than available for some diseases, for example, hepatitis c by not as good as some of the models used for other targets. not that we don't have a vaccine
for lack of trying. this is just summarizing a lot of the various different vaccine approaches that have been evaluated. either a classical whole or activated, live attenuated, a subunit, or synthetic peptide approach including synthetic pep
tides packaged in creative ways with nanoparticles or anything else. vectors, dna immunization has been vaulted, people are using -- some may have seen a recent publication regarding rap, zika virus immunization. those approaches are being
evaluated for hiv type vaccines. these have been tried with a lot of reativity about a lot of notable success. so there are a number of factors. in combination, they represent a daunting set of challenges for vaccine development.
we've touched on the fact that one of the paradigms for coming up with a successful vaccine is having examples in nature of where an individual has been able to survive the infection and establish a state of refactory immunity. that's not the case for hiv.
we -- we don't have examples of people that become infected, able to survive and spontaneously clear the virus. we have only limited examples of people who, by virtue prominently of having some particular mhc ileals, that present viral antigens that are
particularly required for viral replication, are able to control the virus very effectively. so-called elite controllers that in the absence of therapy, are able to keep their viral replication levels down quite low. even those individuals are not
the equivalent of being uninfected. john mentioned the replication properties of hiv. so hiv is a retro virus. it replicates turning the rna into dna and integrating that into the host chromosome. it does that, rna to dna step,
with a reverse transcriptase, which is an enzyme that does not have a proof reeding function. that means that it's relatively compared to the [indiscernible], air prone process. so you can think of hiv as having adapted a life strategy of saying whatever environment i
find myself in, i'm going to generate lots and lot of variants. and in whatever environment i find myself, i'm going to come up with some variants that have a selective advantage in that environment, and those will proliferate and expand and
take -- and if the environment shifts it will change again. that generates tremendous diversity. and the capacity to escape immune responses either in the course of natural infection, both antibody and t cell responses, or conventionally
generated vaccine responses. hiv attacks and eliminates cd4 cells, critical for cord nadphing immune responses. also in various way disables other key immune cells. the virus gets ahead of the host and never gives up that advantage.
the immune system chases after the viral replication. the structure of the keyantian for antibody mediated neutralization for hiv is confirmally very complex. and extensably glycosylated. more than half the total combliko protein is
glycosylation, which makes it difficult to generate antibodies to the targets that you like. then as john dwelt on quite a bit in his presentation, once the virus gets in, it technologies and can establish latency and other long lived persistent forms.
these are all challenges. this is mostly for john's nostologic value, and old slide that illustrates the point nicely. we think of influenza as a virus that varies a lot requiring a different vaccine every year. this shows for 1996, the
global -- this is a tree showing the extent of diversity in the global influenza in one year. through that diversification process that john talked about generating variations in a single patient over the course of a few years of infection, you can see that the extent in an
individual is comparable to the global influenza variability, if you look at a relatively small cohort of individuals in one city, it's much bigger. if you look in one country, you can get a sense of the relative scale. so there is tremendous diversity
which is a big challenge for coming up with a vaccine when you're trying to come up with something that can be universally applied. when we think about mechanisms of vaccine protection, we don't really know what it is that we're trying to go after.
in terms of what is a protective we don't have that example of somebody that has come through natural infection, in aitate that resists reinfection, that we can say what is the nature of the responses that allow that? so we can look at what are the potential things to look at?
on the antibody side we look at neutralizing antibody, but because i showed you all that can varsity, we need to work about not only the potency of antibodies but also the breadth, their coverage of these diverse isolates. we can talk about non
neutralizing antibodies that may have activity through fc receptor functional activities like targeting other effecter cells for antibody dependent cell mediated cellular toxicity. we can talk about t cell immunity. also trying to enlist the innate
immune system. when we think about what are some of the approaches that we can use vaccine-wise in terms of the nature of immune yeps, just to rewriterate, most of the regimens that people use are a form of prime boost. we stimulate in a couple
different ways. either using subyou want, currently the best immunegens for envelope glycoproteins seem to be primers that start finally after a couple of decades of intensive work. starting to have immune yeps that more authentically
recapitulate the structure of the antigen, and various live vects. competent, incompetent, nucleic aindividuals. one of the things that people have come up with, approaches to deal with that diversity that we talked about, if you take a
single prototype sequence as the vaccine insert, that won't likely cover the entire range of diversity that's out there in terms of all the different viruses. so people are noting on using either consensus sequences wayed on large databases of sequenced
viruses that have been accumulated, or conserved regions of the genome that are conserved because of functional constraints that are obligatory. or coming up with mow asic vaccine inserts, predicted epitopes that may not exist in the real world but basically,
are generated to provide induction of responses for maximum coverage. those mosaic inserts do turn out to give a few fold increased reactivity across a broad range of diverse itlates. also importantly, the immunization scheme used,
convention alimmunization schemes are unlikely to give successful immune responses for reasons i'll get into in just a minute. i have been doing this for a the field has swung back and forth between approaches that emphasize trying to induce human
ral and cellular immune tee. i think -- as with most questions having to do with hiv, the answer to either or hiv questions is yes, probably both. and that's probably true for a successful vaccine approach here. but let's start by considering
neutralizing antibody directed vaccines. this is an area that i follow, not an area i work in directly. so i'm indebted to colleagues, not only for sharing some prepublication information with me but also some of the slides i'm going to show you.
neutralizing antibody approaches, premised on the fact that the way that hiv infects cells is by the combliko protein shown on the bottom in purple and green. and a chemokine receptor, ccr5, a minority of surprisers can use an alternative chemokine
receptor. these 2 molecules need to interact. this geico protein is a primer of heterodimers. we'll get into that more but what happens, as with most enveloped viruses that have an enveloped glycoprotein, engages
a receptor that triggers a confirmational change in what is a spring loaded confirmally flexible molecular that results in pox posher of a hydro phobic domain, that mediates fusion of the viral envelope membrane allowing entry of the core, and subsequent steps of the
replication leading to integration and infection. so there is a multi step process of engaging the cd4 receptor, the coreceptor and getting the virus into a cell. basic premise is that neutralizing antibody shown in red can bind to the
"glycoprotein, and basically block that interaction and keep it from happening. when building -- we're in building 50. i'll show a slide from my colleague upstairs. and this is using a technique called focused [indiscernible].
and basically what you can do is take a relatively large specimen, fix it, and with a dual beam microscope, one beam is conventional imaging but the other is an i own become that vaporizes a thin layer off the top of your sample, you can take a picture, vaporize a layer of
ten to 25 nanometers, take another picture, so on, so on. then integrate all of those images into a 3 dimensional composite image that allows you to get a 3 dimensional perspective on mixtures of cells or tissues. powerful technique.
what that shows you is that rather than simply having to block the access of free virus to a cell floating in a nice culture flask, the situation is likely to be much more when cells of the immune system, particularly interacting with cells have these extensive
membrane interactions and interdigitations. and if you look at an intersection, you can see focused in red, the virus which is being transferred what's known as a virology synapse. you can see that even things that look like just thin cell
processes here, in the 3d construction, are sheets of membrane. if you think about trying to get access for an antibody to disrupt that, that's harder than if you're trying to block access of a free variant to target cell.
and so there are many challenges where the -- as john alluded to, the evolutionary adaptation of this virus has basically subverted the natural physiology of the immune system, and interactions of cells tin immune system, to its own advantage. there is another challenge.
i told you how hard it is to generate neutralizing antibodies. there are some as we'll get into in just a minute. they work well. one of the ways we have been able to validate this is in non human primate models.
rather than the simplistic notion that i had before we started these experiments, that neutralizing monoclonal antibody will act by neutralizing the virus at the portal of entry, and preventing infection, it turns out that that isn't the case.
in a recent study with [indiscernible], we showed in animals that were treated systemically with the neutralizing antibody pg2 is 21, you can't read this, but this is basically across the way lots of different tissues that were surveyed in a serial sacrifice
study in monkeys on day 1-3, day 7 or ten. the tissue highlighted with the blue backgrounds are tissues outside the vaginal inoclayings, the female gentle track. these are all distal tissues in blue. you can see compared to sham
treated animals, the animals that got the antibody highlighted here actually had more virus at distal tissues, at early time points. but they eventually went on to clear that virus. so it's not a simple case of blocking things at the portal of
entry. it's much more complex than that. it wasn't that we were picking up stray non replication competent nucleic acid signals in these distal issues. we actually were able to transfer infection by transfer of cells from these tissues at
early time points. so the biology of this transmission is much more complex than we've appreciated. so neutralizing antibodies need to target the envelope spike. that's a trimer of heterodimers. we have developed a really extensive knowledge of structure
and functional properties, and that has helped to guide but hasn't led us to solve how to make a vaccine with the desired properties as i'll get into. this shows the crystal structure and a car tune structure where the 3 transmembrane subunits are not coverylently associated with
the surveys envelope glycoprotein subunits. you can see that in a -- in this model here with subunits showing, as we rotate around. if we turn to a space fill model where the green is the gp is 20, yellow highlighting the interaction domain and red is
transmembrane, you can appreciate the structure of when you add in the glycosylation, you can see how obstructed that structure really is. there is tremendous challenge to generate antibodies to the kinds of vulnerabilities that we would
like to be targeting with a vaccine mature we've made a lot of progress. a lot of that progress is by studying naturally infected people that have developed antibodies over time. what happens when someone gets infected, they'll initially
develop antibodies that bind to the envelope glycoprotein, but those antibodies typically don't neutralize, even the virus that they got infected with. by about a year or so, they'll start to develop called strand specific antibodies, that is, antibodies that can neutralize
th initial virus that they were infected with. that was present through that first year. but by the time they developed those antibodies, the virus has moved on and mutated so that antibody no longer neutralizes the virus actively replicating.
that process goes on in a way that reminds me of coyote chasing road runner, never catching up. the virus stays one step ahead of the antibody response that's chasing it. but after multiple years, subset of patients, depending on the
cohort between five and 20% do develop antibodies that capable of broadly neutralizing a range of different hiv isolates. unfortunately, it doesn't do them good. the viruses, again, multiple steps ahead and has mutated. we have learned a tremendous
amount, what are the characteristics of antibodies like that? and that's really been enabled by a technique pioneered by michelle at rockefeller and dennis at scripts, in which they've kind of taken sort of a reverse approach.
they start with the antibody, and then do structural studies of that antibody, inter5:00ing with various model envelope and gens. and then try to glean from that information to help guide the design of new immunegens and test them for immunogenicity.
this has been facilitated by techniques that were labeled bates where you use your gp120ant iran to stain individual b cells, able to sort out the individual b cells and clone the b cell receptors from those and generate through molecular techniques,
neutralizing antibodies. and there has been tremendous progress over the last several years in this area, lots and lots of antibodies have been developed. some prototypes are shown here, the gp120, trimer of heterodimers with
[indiscernible] envelope membrane here. there is one class called apex antibodies, typified by pga9. some interact with the syn. we showed all that glycosylation. that actsis a a impediment to the binding of most antibodies.
there is a subclass that are dependent on the patch for their reactivity. cd4 binding site, action that bind at the interface between those subunits. and things that bind in the so-called membrane proximal external region, which as you
can see, is a tough tight fit. there are antibodies that are able to get in there. this shows a table showing the relative potency of the antibodies for in vitro neutralization, relative to the percentage of a large panel of roughly 100 different diverse
virus isolates that they're able to neutralize. and this kind of composite gives you an idea. these are were some of the initial antibodies that were identified to those didomains -- odidomains. we're making tremendous progress
we have a number of antibodies that reasonably broad and quite potent. ago we've studies those, though, and as you might get based on the process that i described to you of how they arise and how long it takes, what's become career is that most of these
antibodies are not your typical garden variety antibodies. they're unusual. they are -- have up usual characteristics just not typical for most antibodies, including extended lengths of the complementary determining regions for some of them.
they need extra long binding domains to access their target andgen. part of that is associated with a requirement for extensive somatic hyper mutation that comes about in the course of. ation of the antibody responses. so that's part of the
explanation for why you need such a long time to develop these antibodies, because you need to calm lots and lots of mutational changes in order to come up with the properties required. so we have a tremendous problem. we have these wonderful
structures of antigen bound to antibodies isolated through the process i just described to you. we're still kind of stymied in how to translate that information into how to come up with an immune genand generate the antibodies in a practical way.
that's bridging the gap from antigen to immunogen. so the things that people are working on is trying to generate more confirmally athentic primers. we made a lot of progress with one of the challenges because of-- you need the. maturation, it terms out for
many of the linages, the final antibody that you want, the final structure that you're trying to engage, the b cell repertoire doesn't contain something that binds to that. you need to bind to something different than that. and then go through the
sequential mutation process to get to where you need to. and so to deal with challenge, people are starting to do some sequential immunization approaches where they immunize first with one simplified version, the next version, so on.
3 or 4 different things. weather we'll get there i don't know. but it says that just a conventional immupization, the way we're used to for other pathogens won't work here. and finally, as you might figure from how long it takes to
develop the kinds of antibodies in the course of natural infection where you have continuous exposure to large amounts of antigen to drive that process, simply giving a couple of shots at 6 month intervals won't get us there. so understanding the
immunization schemes that are going to be required is going to be important. so as we continue to work on that challenge, one of the things that the field has adopted is to say we don't know how to generate these broadly neutralizing antibodies from
immune gens just yet. but we do have breedly neutralizing antibodies that row potent and broad. can we learn something in terms of proof of principal about the potential once we figure out how to induce them through a vaccine, but also have some
practical impact using passive administration of such there are clinical studies in progress now to test that proof of concept, see whether we can impact transmission in paediatric and adult high range cohorts. it's likely, as with most
things, monotherapy with drugs or antibodies or anything else is on a recipe for generating resistant. we'll likely need multiple antibodies, even for a prophylactic application where the challenge is small, and recombinant proteins like
antibodies can be quite expensive. there has been a lot of progress in engineering antibodies with certain changes that dramatically extend their circulating half life from a few weeks to months. and so there is progress there.
and one other approach is the use of gene therapy in particular, adenoassociated virus, encoding these and mike and ronda in non human primate studies we've collaborated with have very encouraging result on the potential off aav delivered
antibodies to either prevent and or treat infection in non human primate models. i'll touch briefly on non neutralizing antibody targeted this is the -- i allied to the 44 time trail, the conceptual paradigm for the potential importance of these non
neutralizing antibodies comes from that study, which as i said had border line efficacy at 31%. but that efficacy was not associated with either conventional cd8t cell responses or neutralizing antibody responses, instead, associated with cd4 cell responses.
but only in the absence of iga responses to the vaccine which actually seemed to have a negative influence on i'm not a big believer in the utility of these kinds of i think the orange vaccine i'm aware of where there is a clear case that non neutralizing
activity plays a key role is for no.a coccal vaccines but it's being actively pursued in the field. and now you know that. let's shift now to looking at cellular immunity targeted as i said, the pend leon mugesera has one -- pendulum has
one back and forth. it was an adenovirus sero type 5 vector vaccine that really sort of killed the whole field of cellular immune targeted vaccines for a while because as you can see here, and study participants that were at 5 sero positive and had preexisting
responses, the vaccine seemed to produce an activating response to enhanced susceptibility to obviously, that was discontinued, enthusiasm for cellular based vaccines in general suffered as a result. i want to spend my last few minutes talking about a
completely different kind of vaccine. and this is work from my long tim collaborator, lewis, out of the oregon center. lewis is one of the people that developed the technique of intracellular cytokine staining. he used that several years ago
in a technical tour deforest using thousands of pep tides to map the complete immunological repertoire of responses against cytomeglo virus. just represented here, a classic paper. but one of the interesting things about those responses is
that they're very broad. but they also have some other characteristics. in particular, they're skewed toward a memory differentiation state rather than memory. when you do a convectional vaccine, this is showing a hypothetical curve of the time
course of viral load and plasma, as a confluence of viral replication, in a naive individual by the time the cellular immune response in peek, you're multiple weeks into the infection, even in a conventionally vaccinated individual.
in a prime boost vaccine typically what happens is you get a peek response, that dies down to a memory state. then you need proliferation and expansion of that response, trafficking of those cells to the appropriate sites before you get a peak response, even there
you have significant delay. it looks like for hiv, that delay is just too dam long. there is a shorter window of opportunity, somebody asked the question, if cmb induces these effecter memory responses that are already differentiated and ready to go and persistent
because the vector cmb is also persistent, could that have a beneficial effect? so we tested the idea of whether or not these cmb vector responses, are broad cd4 and cd8 responses. lots of diversity of the eptaupes that are recognized,
because it's a persistent vector, gives persistent responses skewed toward memory and a broad tissue contribution, including sites of mucowsal challenge. the question, using it as a vector would give responses to the siv inserts in these 9 human
primate studies that had these properties. if so would that be associated with vaccine protection? around the short answer from many, many studies is yes, of a kind. and really, a unique kind. what you're seeing here is
unvaccinated animals challenged for the data shown here. and you can see the classma viremia occurs for the animals. what we saw is that in this study, this vaccine does not induce appreciable antibody responses to the inserts. just inaccuses these interesting
t cell responses. those responses are not able to prevent infection. but what is striking, in overhalf the animals, they're able to control infection and durably control infection over and what's even more striking is that when we look longitudinally
in animals, you can see that we initially see these plasma blips that decrease and go away. initially. i'm showing you rna and dna levels in bone marrow. in multiple other tissues also. initially, in animals as soon as they clear plasma viremia, we
can show there is still replication compete want cultural virus in these tissues. there is rna, dna, we can adoptively transfer. over time, we can no longer detect these things. so these animals are progressively clearing the virus
from an early established multiple tissue sites. and what we think going as john alluded to in his talk, there seems to be a transition from the short live cells infected to long lived cells. we don't fully understand an tomcally or cell compartments or
whatever, what that trance significance is. if the -- transition is. if the infection is contained before you cross the thresh hold, it appears to decay and to be cleared. that's shown here. if we look at animals, looking
at dna levels, rna levels in these various tissues, in animals, 3-6 weeks after challenge they become ave reamic, and are sacrificed and tissue studied. we can still see readily detectable rna and dna. if we follow animals for 3 years
we can not find anything detectable by laboratory analysis or adoptive transfer. we've seeing progressive clearance of virus from these tissues over time. to make it a little more interesting, i started out my training wanting to be a
cellular immunologist. i got diverted for a couple decades doing aids related stuff. now i'm getting back to studying cellular immunology, a variation. it turns out this vaccine indices a very strange kind of
cellular immune response. [indiscernible] got the nhlbi prize in 1996 for their work in the 70s, demonstrating t cell responses by cd4 cells were restricted by class 2 antigen presentation, and responses by cd8 cells restricted by class 1 antigen presentation.
and that was -- still true, but not the whole truth. because it turns out that the responses induced by these vectors, actually, the cd8 responses rather than being conventionally restricted by mhc class 1, about two-thirds are restricted by class 2, and the
rest are actually restricted by allele called class one e, much better known for being involved in regulating in k cell it's a long story. and some of it is pleasured, some in preparation. it looks like a consequence of the fact that the vector that we
used was passed in fibroblasts for a period of time, and acquired some deletions that resulted in altaltered [indiscernible] cell. in vivo, it has altered it and the antigen priming and presentation is quite different than other vaccines, natural
infection, or even wildtype cmv infection, all of which would be interesting. a side note, except for the fact when we make repaired vectors and put the mutations wack and test the various combinations, it's become clear that these unconventional response razz
what's responsibility for the so we're trying to get to the bottom of that. so, that brings us to the final chapter of our hiv vaccine development for sophisticated highly intelligent but non specialists. and i'll leave you with a couple
take home messages. hopefully i've given you a sense coming up with an effective vaccine for this very wildly and sophisticated target pathogen, is -- represented extraordinary and urn precedented challenges based on both the properties as a target, but also the biology
of the pathogen. i hope you've given you a sense we haven't been wasting our we made a lot of progress, but an efficacious vaccine remains elusive. my conviction is that we tried all the easy and obvious stuff over the last 3 decades.
and that vaccine solution to this will yield but it probably is not going to be based on any kind of a conventional immunegen and or vaccine approach. but with all of that frustrating stuff said, i think the pressing need for vaccine and the potential global impact of even
a partially effective vaccine is so important that it continues to motivate me and others in the field, and i think to justify continued research in this area. i'll stop there and if we have time, try to answer any questions. >> thank you very, very much.
>> that was marvelous. so i trust you have some questions you wish to ask? yes. use the microphone. >> thank you so much for the really nice talk. could you elaborate a little bit more on the possibility of us
really getting prophylactic vaccine verses therapeutical. and in terms of timeline, what do you see? >> yeah. so john alluded to the situation of persistent virus in the face of ain't retroviral therapy. people are pursuing a number of
approaches to come up with more definitive treatment in that setting. some of those involve therapeutic vaccine where the idea is to augment immune around often times, that is premised on a combination of an immune clearance mechanism
joined to something, some kind of a stimulant -- without expression, a virus that persisting virus is basically silent to the immune system, not detected. i think that's a really daunting challenge in part because the amount of virus that is there is
higher than what you need to prevent the initial prophylactic as john alluded to, the molecule data suggests most transmissions involve transmission of a single viral variant, across a mu doesal surface. that in principal is less of a challenge, so on.
i hazard to give you a guess about timelines for vaccines able to be effective in either i think they're a huge -- therer huge challenges. i think some of the work with passive antibody that's going on will demonstrate proof of concept, hopefully, for
particularly neutralizing antibody. there may be some practical applications for that before we have immunization based approaches for proxlaxis, in terms of therapeutic vaccines. it's become clear from clinical work and some of the studies
that we and others have done in non human primate models, that to be effective, you need to have completely effective immune surveillance that is going to control whatever vier isis left -- virus is left when you take the ain't retroviral drugs away, or you need to eradicate
every last replication competent and that's an extraordinarily challenging bar to get over. but like john said, the more sophisticated our knowledge becomes, the more daunting the task seems to get. it's a progress of a kind. >> that was a beautiful talk.
thanks for the -- the background. can you say a little bit more about what could be the explanations for a short lived infected cell investors a latent-- verses a latent cell? does it not have anything to do with selection of the integration site or expression
of the virus? and -- i think that there probably multiple things that can contribute to that. i think integration site may contribute to persistence. the one thing for which there is actually some data that are
relatively clear, is that if you look at the distribution for viral dna in different cd4 memory different asian subsets, in people that go on treatment or animals that go on treatment very early in the course of infection, most of that dna is in shorter live cells.
seems to not get into central memory cells that are longer and can establish latency, just proportionately to the same extent. that's one possiblity. another possibility that we're looking at has to do with the fact that there are a number of
different immune privileged tissue sanctuaries for virus. one of the ones we focused on is actually helper cells inside b cell follicles which, for various reasons, t cells that ra able to control and clear virus from other compartments because the natural biology by which
they're not able to effectively traffic to the -- into the follicles, that remains a viral it could be something where, whether or not you crossed the threshold in terms of virus getting into those sanctuary sites or not may make a difference in terms of
establishment or not establish. >> what typically kills a cell that's infected by hiv? is it just overreplication? or is it -- >> so it's clear that in vitro, replication and viral sito pathology can kill cells including primary cells.
the catch there is that in vivo, to the extent we can quantify viral burst size and so on. , it seems to be much less than under maximal stimulated conditions we use in vivo and so with the so called latency reversing agents that have been used to stimulate cells, both in
vitro and in vivo, the initial hope of the called shock and kill paradigm was that you could stimulate cells, get induced expression, and you get so much viruses expression that the critical would die, but even if it didn't, there would be enough antigen expression that you'd
have immunological clearance. shock and kill has so far turned out to be pickle and hope, because the level of viral induction that you get doesn't seem to be enough to result in viral cytopathology. under normal conditions, the majority e of the cytotoxic t
cells are not able to kill those cells, even if you get a tickle of viral expression. >> i would mention to you that all of the power points and reference materials are all available on our website. all you have to do is google dedemystifying embassy.
you can see materials for all the programs for the past 12 you can follow the evolution of hiv by looking at a website, too. >> you can binge watch demystifying medicine. >> so thank you very much. this has been an extraordinarily
exciting.
If you want to know about herbal product visit IBHIndo Herbal Indonesia. Also best herbal product for diabetes visit Obat Diabetes Alami - Obat Herbal Diabetes Paling Ampuh. Visit Jual Obat for online shoping herbal medicine.