RESEARCH

Dr. Coffin's Testimony on XMRV at the Federal Advisory Committee on CFS (CFSAC)(Oct 2009)

Dr. Coffin is an acclaimed retrovirologist and the co-author of the leading text in the field. His testimony was arguably the most important of the CFSAC session on XMRV. Thanks to Garcia for transcribing it.

Introduction   -  Dr Coffin: I should probably introduce myself to you a little bit. I’ve been working in retroviruses for a great many years. More than 40 years actually. And my interests have been in basic retrovirology, understanding particularly the co-evolution of viruses and hosts. Retroviruses have some very special properties that allow us to look at evolution in ways that we can’t with any other virus or infectious disease for that matter. More recently within the last 10 or 12 years, I became interested in the various aspects of the HIV epidemic and understanding the HIV-host interaction, also evolutionary issues which occur with that virus. I was the founding director of the HIV drug-resistance program for the NCI at Frederick about 50 miles from here. And I held that position for about 10 years and I’m now a consultant to that and in fact I’m en route in between my two sites of work as it were.

So what I want to do is give a first my reaction to this remarkable finding. And reiterate a number of the salient points that Dr Peterson just made. I think it’s important to re-emphasize them. And then talk a little bit about what it means to be xenotropic, because many of you may actually be curious about that particular point. And then discuss at the end a little bit about what we still need to know about this virus.

XMRV and Prostate Cancer   - XMRV was first reported in prostate cancer samples, particularly ones with a mutation as was pointed out in RNASE-L in 2005. In a recent paper this virus was reported as being present using an immunological assay in about nearly a quarter of randomly chosen prostate-cancer biopsies and in a much smaller fraction of non-malignant prostate biopsies. And as of course we all know has been reported in the published literature in the paper that appeared last week in 67% of CFS cases and Dr Peterson just presented some data suggesting that that is maybe a gross understatement. There may well be much closer to 100% association and it would be very interesting to see how that plays out of course.

The virus can be detected & cloned from prostate cancers. I don’t believe infectious virus has ever been isolated from prostate cancers or patients, at least in the published literature. And that’s in large part because almost all of those studies were based on amplifying DNA out of paraffin blocks. And there’s some debate in the 2 papers that have been published on this as to what cells the virus is in, whether it’s in stromal cells and tumor cells. That remains to be worked out.

Also the association with RNASE-L as we’ve already heard actually remains to be worked out. It may be that the RNASE-L mutation is not related necessarily to the infection with the virus, but the presence of the mutation does make it easier to detect in patients. That’s certainly a hypothesis we can entertain at this point.

Chronic Fatigue Syndrome (ME/CFS) In the case of CFS the association is in many ways much clearer than in the prostate cancer because we can get live virus. But they can get live virus out of PBMCs or even out of plasma which is actually pretty remarkable. It’s in fact quite hard to isolate live virus from plasma of HIV infected patients. It tends to be coated with antibody, virions aren’t in very good shape and so on. It almost looks like it’s easier to do that with CFS (although the virologists who are actually working on this may not like the characterization of being easy). And at least in activated PBMCs it is present in a very high fraction.

It was in a really quite impressive fraction of cells that are isolated from blood, where after they’re activated. It’s unclear to me at this point, if you don’t activate the cells, what the fraction of actual infected cells is in blood. Which clearly, because the activation itself may cause the virus, may allow the virus to spread though the blood cells that you’re looking at. It almost certainly does I would say. The virus that you isolate is infectious for a number of human cell lines. It’s particularly infectious for the prostate cancer cell lines as was mentioned and also for b-cell and t-cell derived lines and for fresh activated PBMCs (Peripheral Blood Mononuclear cells) which are basically a mixture of b-cells and t-cells.

Ed - After the WPI 'activated' the immune cells in their samples and tested them for XMRV they found that a surprisingly large fraction of them were infected.  My understanding is that immune cells  generally troll the body looking for pathogens; after they find one they undergo a metabolically expensive transition into killing machines.  Most of the time most of our immune cells are not activated. Since the WPI only looked at activated immune cells Dr. Coffin is wondering here how many immune cells under normal conditions  have XMRV infection. Dr. Klimas has stated that a large percentage of immune cells in chronic fatigue syndrome patients  usually are activated.

One of the things that struck me the most was this apparent detectability in 4% of the control samples, although it’s important to point out that right now these are not unbiased samples in either case. In the case of the prostate cancer the samples are from prostate biopsies that were non-malignant but were hyperplasic and that may buy us an ability to see the virus. In the case of CFS the controls were not familial or health-worker or anything but they were local and so if the distribution of infection is very uneven, that might be a local bias.

Ed - Could there be pockets of XMRV infection around the world? Three things suggest that this could be a possibility. A German prostate cancer study failed to find any association with XMRV, Dr. Coffin stated he believes XMRV could be a 'recent' escape from a wild mouse and the fact that the virus shows little genetic variability from person-to-person. Its unclear what Dr. Coffin means by 'recent'  (decades? hundreds of years?) but a 'recent' escape would have a localized distribution - it takes time for viruses - particularly for retroviruses, which don't tend to be very transmissible, to spread.

All this needs to be worked out of course and this is an extremely critical thing to do, to get a handle on what the actual prevalence in the US and worldwide is. There is some suggestion, there is one as yet unpublished report which was presented in the meeting earlier this year of a few percent incidence in the blood samples that were obtained by the Japanese red cross and a somewhat higher prevalence in samples from prostate cancer patients in Japan. And so there is some evidence for some worldwide association. You’re probably aware that there was a recent paper that did not find any virus at all, any evidence of infection in I believe six hundred and some prostate cancer patients from Germany . So this all remains to be worked out.

Little Genetic Variability - One of the most very striking things is the very close relationship (in the viruses genetic makeup from person-to-person), and Dr Peterson showed you this and I’ve also got that on the slide in a minute, of how close these viruses are to one another. And the most distant pair actually, both of which were from prostate cancer, differs by only 0.3%. Now a patient who's been infected with HIV for one or two weeks has a greater diversity in his virus population than does virus isolated in different years, different parts of the country, different diseases of this virus.

It’s really remarkable how close that is and that’s a very important implication. It’s not that this virus has a lower mutation rate than HIV. These viruses probably have all about the same mutation rate. But its suggestive in fact that there are very few cycles of replication that separate the viruses that’s in one person from the virus that’s in another.

Vaccine Possibilities - Good, Treatment - Not So Good   -  And in some ways the implications of that are both good and bad news. The bad news is that it suggests the virus is not actively undergoing ongoing replication during the course of infection of a single individual and that would not be good news if one were trying to use antiviral therapy. It doesn’t mean it shouldn’t be tested but it reduces optimism that that kind of treatment regimen (the same kind of thing we do for HIV) might work for this. But it would be good news if one were developing a vaccine because it would mean that the genetic variation issue would not probably be a significant roadblock the way it is expected to be with HIV. And it’s important to note that there are marketed vaccines against a very closely related virus, and that’s feline leukemia virus, which has been approved for use and has not 100% efficacy but has some efficacy.

So the other striking thing is the very close relationship of this virus to xenotropic MLV an endogenous virus of inbred mice. As was noted the pathogenicity of xenotropic MLV has not been studied in mice, for an obvious reason, and that is this virus does not infect mice. But closely related viruses have been very important pathogenic models in virology for cancer, for immunodeficiency disease and for a variety of neurological diseases. So the potential pathogenicity of this virus at least in a host that it can infect should not be underestimated.

XMRV's Close Cousins - So Xenotropic MLV is inherited as an endogenous provirus meaning it’s in the germline, about 10-20 copies, in all inbred mice, and in a much greater copy number in some wild mice. Most of the proviruses that are inherited this way are defective, they won’t give rise to complete virus, but some of these are in fact capable of replicating. There’s one called BXV1 which is intact and infectious if you pull it from out of the mouse DNA you can get infectious virus out of it. And these viruses can infect virtually all mammals, and I think even some birds except for some species of MUS, and what’s happened in these mice is that after the virus became endogenous (became integrated into the germline of the mouse) the mouse lost the receptors for it. Probably by selective pressure that was put on by the virus itself.

As I said these are not directly pathogenic in mice, but if you isolate some pathogenic cancer causing viruses from mice, they will have a sequence, a very important sequence which is the Long Terminal Repeat which contains all of the signals for gene activation for directing transcription of this and for turning on oncogenes which causes the cancer and that LTR comes from the xenotropic virus and its very similar (not quite identical few little differences) but its very similar in sequence to the virus in XMRV.

One of the big worries in this kind of study over the years is these viruses are actually common contaminants of human cell lines.  They infect prostate cancer cell lines very nicely, and these are often isolated by passage through nude mice. Once the nude mice have this infectious provirus, this

 But closely related viruses have been very important pathogenic models in virology for cancer, for immunodeficiency disease and for a variety of neurological diseases. So the potential pathogenicity of this virus at least in a host that it can infect should not be underestimated.

virus can then infect the tumor. There are repeated reports over and over again of human tumor cell lines that have picked up these viruses in this way and so when one first sees this it’s a major concern. The phylogenetic relationship in fact as Dr Peterson pointed out very strongly argues against that kind of contamination from laboratory mice. And I already made this last point, related Murine Leukemia viruses can cause a wide variety of diseases.

Ed. - XMRV's closest cousin has inserted itself into the DNA of inbred  laboratory mice.  These types of viruses  - called endogenous retroviruses - are commonly found in our DNA. Almost all are benign but researchers believe that under some conditions a few may be able to activate themselves and either start pumping out virus or pieces of DNA that could contribute to cancer or other diseases. XMRV has a genetic sequence in it that is similar to one that is known to contribute to cancer.

XMRV's Origin   - This is a little cartoon that I published in the perspective that I wrote that accompanied Judy Mikovits’s paper describing the virus in CFS, and simply showing that the virus came originally from an exogenous virus, an infectious virus infecting mouse ancestors that then happened to infect the germ line so it became fixed as a fossil in the germ line. Subsequent to that the receptor got mutated and so the virus can still infect, but it can’t infect mice because they don’t have the receptor and somehow it got transmitted to humans.  

The clusters and the very close relationship have implied that this transmission may have occurred quite recently and in fact we don’t know but it’s not impossible that it happens all the time, that there is some mouse out in the wild that has this as an endogenous virus and my lab has just begun to look to see if that’s true. And that the virus can be both transmitted from mouse to human and from human to human, it just all remains to be seen, but I think it’s a very important area of study to get aware of where this virus is coming from, when it got into humans and how of course it’s being transmitted among people.

Ed. At some time in the past XMRV infected wild mice. Over time it became a permanent part of the mouse's DNA.  At some point 'the receptor' the virus used to gain entry into mice cells changed making it impossible for XMRV to further infect mice. Instead the virus lived on as a benign 'fossil' in the mice's DNA.  If the original virus no longer infects mice, though, that would suggest it must have jumped to humans from another source (?). Dr. Coffin suggests, though, that there is a wild population of mice which can still get infected by XMRV in which may have transmitted the virus to humans at some point.

"it’s not necessarily easy to establish causality. It took quite a while with HIV despite the real smoking guns you had... whether the virus is the cause of the disease or a passenger, or just a geographic coincidence of infection with the disease, all of this remains to be established."

On the right I show the same kind of phylogenetic relationship and just want to point out again the very striking similarity of the viruses isolated from chronic fatigue and prostate cancer. There’s no obvious division of viruses that are isolated from one disease or another. The two most distant viruses on there are actually both from prostate cancer. And what is really a very close relationship between these endogenous proviruses and infectious xenotropic viruses isolated from mice and then the more distant relationship to other well known mice viruses. If we were to put HIV on there I think it would have to go all the way over to that wall and come back again to get the relationship.

These viruses are very, very distant. They’re all retroviruses, but it’s very hard to even see the relationship when you try to align the sequences up, you just have to look in certain very, very specific areas. So it’s important to keep in mind it’s a very different group of viruses. A very widespread group of viruses, particularly in mammals, and even a few bird isolates. But never before, except in the case of some gene-therapy applications, have these viruses been known to infect humans. And there was some effort actually put in to looking at that particular issue in the case of Feline Leukemia virus, back in the ‘70s when the virus was first discovered, there was a fair amount of concern that cats might be spreading this virus into their owners.

And this is actually my last slide and I just want to emphasize what we don’t know about this virus. This is very important. It’s very, very early days in this. We’ve got one paper on associated Chronic Fatigue Syndrome, two papers on the association with Prostate Cancer, which don’t completely agree with one another. The [CFS] paper in my opinion is as good as you get for the first paper on a subject, but it’s still the first paper. There’s much that needs to be done.

The National Cancer Institutes Efforts - First is of course to establish what the real role of the virus is in this disease. And it’s not necessarily easy to establish causality. It took quite a while with HIV despite the real smoking guns you had. It’s fairly easy to establish in some kinds of cancers because you can actually see integration site specificity that would point to it, but in the case of other diseases whether the virus is the cause of the disease or a passenger, or just a geographic coincidence of infection with the disease, all of this remains to be established. NCI in particular is really mounting a big effort, intramurally at least, to develop this. We don’t know the incidence and prevalence in the human population.

"It’s very important that we have standardized, uniform, well validated and reliable assays. This is really the critical issue here. And immunological assays are particularly difficult in this case. Even with HIV the well established HIV-assays still have a certain false positive rate"

The Critical Issue   - I want to emphasize that in order to do these studies it’s very important that we have standardized, uniform, well validated and reliable assays. This is really the critical issue here. And immunological assays are particularly difficult in this case. Even with HIV the well established HIV-assays still have a certain false positive rate and so it’s really critical that we know exactly what the assay is, it’s been well validated with well standardized specimens and everybody is on the same page, ideally using the same assay for these studies.

Ed. The biggest problem with PCR tests for pathogens is not false negatives but  false positives; i.e. tests which falsely indicate a certain pathogen is there. PCR tests work by identifying small sequences of genes that are, hopefully, unique to a particular organism - not an easy task given the multitude of DNA present.

It’s not necessarily easy to establish causality. It took quite a while with HIV despite the real smoking guns you had. It’s fairly easy to establish in some kinds of cancers because you can actually see integration site specificity that would point to it, but in the case of other diseases whether the virus is the cause of the disease or a passenger, or just a geographic coincidence of infection with the disease, all of this remains to be established.

And this is one of the things we are actually working on in Frederick quite intently right now is to try to develop some assays that will fill this, but other people are as well. We may win out, they may win out but whatever happens before we can really get a real handle on this, these assays are critically important. We need to know the distribution in the human population of course, are there possible clusters of infection or is it just generally widespread?

We have no idea what the mode of transmission is, despite some things you may have read

Transmission   - We have no idea what the mode of transmission is, despite some things you may have read and in my opinion we just don’t know. The ability to easily isolate the virus from blood cells and from blood certainly implies that transmission could be blood borne. Beyond that I haven’t a guess as to what it’s being transmitted by. Bodily fluids? It’s probably not transmitted in aerosols. These viruses are not very transmissible that way in general, but we don’t know that for sure for this virus. And we don’t know the origin of this virus.

The close relationship of this virus means it almost has to have come from mice, I can’t really visualize another scenario. But is that happening today? Does it happen all the time? Is it like a hanta virus outbreak or something like that? Or did it happen once a long, long time ago and it is then transmitted at a level in the human population and of course if it is causal for the disease we don’t know what the attack rate is going to be, we don’t know what fraction of the people infected with this virus. Everybody infected with HIV eventually dies, almost everybody will die, 99 point something percent will eventually die of AIDS, if untreated. However everybody infected with another human retrovirus HTLV1, only a small fraction will actually get disease so we don’t know that for these viruses. I mean this list [of questions] could go on and on.

Thanks again to Garcia for transcribing this talk.