Written by Kati
This is a rare chance to get a view of how the XMRV/CFS connection is viewed by virologists outside of the ME/CFS field
Part I: Transcribed by Kim
TWiV 76: XMRV with Professor Stephen Goff
by Vincent Racaniello on April 4, 2010
VR: I wanted to talk about XMRV with you.
VR: We could probably talk about all we know in an hour, right?
SG: Yup, that’s probably true (laughs).
VR: There’s not a lot of papers on it, of yet, but it’s a new field. Why don’t you tell us where this virus originally came from. It was picked up in a prostate tumor, right?
SG: Right. So, the ultimate origin of the virus is a topic of discussion, still. But the history of its discovery, is, as you say, was found in prostate cancers. It was one of the first new viruses found in humans using Virochips, where people were looking for new pathogens. And, Ganem and DeRisi and the people at UCSF (with Bob Silverman) were looking, especially in prostate cancer, because of a genetic history, actually. So humans have a familial tendency for prostate cancer that they tracked to a gene in innate immunity. And so there was the plausibility that this disease had a component that could be infectious.
VR: So that’s why they were looking for that?
SG: That’s why they looked. Now that hasn’t held up well. Uh, actually (laughs)…anymore – as the basis for the disease. But it was the reason they looked and coincidental, or lucky or not, they found this virus.
VR: This was the only virus they found in prostate tumors?
SG: This…I don’t know…this was the strikingly correlated one. In their initial look, was present in a fair number of the familial prostate cancer people. And not in most of the controls. So this was the virus that correlated most strongly with the disease.
VR: So I remember when this was picked up and you began to work with it in your lab shortly afterwards.
VR: Which I know because you’re right across the hall from me. And, I don’t remember anyone discussing its origin at that time. That seemed to have come much later when Chronic Fatigue entered the picture. But when it was first picked up by DeRisi and Ganem, did they realize that it was probably of mouse origin?
SG: Yeah, I mean that was certainly the thought. I don’t think it was the first topic people addressed, but the sequence similarity is so great to the mouse xeno-viruses that I think everybody assumed it was going to have had arisen from a mouse.
VR: There’s about 30 bases different from a mouse virus?
SG: Yes. Incredibly close, now I forget the number…98-point-something percent identical to mouse xeno-virus.
VR: Could it have been a contaminant?
SG: So that was also a concern. And it still is in a lot of labs. You know, the levels were low, people were always using PCR to detect it. And when you’re in a lab that’s working with large amounts of any virus, PCR contamination is always a problem. So people do have to be very careful – even paranoid – to make sure they’re avoiding it. But I think at this point, it’s not likely to be a contaminant in most of the assays that we’re seeing because, for one thing, it’s not identical. So John Coffin and the crowd have looked in a very large number of mouse species.
VR: Lab species and wild or just lab?
SG: Both, but mostly lab. But all mice have about 10 or 20 copies of these xeno-viruses in their genomes and all of the ones that have been looked at, in all the species of mice, none of them are identical to the viruses that are being said as circulating in humans. They’re close, but not identical. So, contamination is not very likely.
VR: So even a lab mouse here in our building in gonna have this virus in its genome. It’s an endogenous virus.
SG: Absolutely. Yup.
VR: Do they produce infectious retroviruses?
SG: Some do. So there are known…the great thing about this – people were excited to find the virus because they were so familiar with this family of viruses. So all mice have endogenous proviruses in their genomes. They’re of a variety of families. There are a lot of ways to group them. One of the most useful ways is by the receptor that they use. So there are the so-called ecotropic family, which you use one virus receptor. And the xenotropics, the one we are talking about here, which is a different receptor, which is known as XPR1. There are polytropic and others, but the xenotropic are well known and they’ve been studied lately, since the seventies. They were actually initially picked up – these true mouse viruses – because whenever passed a human cell line through a mouse, for example, you would do for a tumor in a nude mouse, or a human line passed through a mouse for any reason, just to amplify the…or study the line. It would acquire these viruses. The human cells have the receptor for the virus, the mouse does not. So, the mouse would carry the endogenous virus, even sometimes have replication competent versions of the endogenous viruses, but the mouse would never amplify them or replicate them because they don’t have the receptor. But if you put a human cell line into the mouse, it immediately picks up this virus. And does replicate it. And when you take the tumor back out of the mouse, or the human line back out of the mouse, it would always, almost, acquire that virus. So people were very excited about that in the seventies. They thought it was a human virus that had been induced, but no, it was simply the mouse virus that had been picked up by the human line.
VR: Do we know what the receptor is for XMRV?
SG: Yes, so it’s this – we don’t know much about it – it’s a membrance surface protein called XPR1, function largely unknown.
VR: It seems similar to many of the mouse receptors for other mouse viruses.
SG: Yeah. They’re quite diverse, really, the various murine leukemia virus receptors, but it’s not an unusual surface molecule.
VR: So, they’re called murine leukemia viruses. Do they cause Leukemia?
VR: So the endogenous copies that are producing virus in many mice, is that causing Leukemia?
SG: So the ecotropics, yes. Those that can actually replicate in mice will cause and do cause Leukemia. And they can be…and that can occur in at least two ways. You can actually just infect the mice at birth with one of these viruses as an exogenous infection. The mice will become viremic and they will get Leukemia and they will die. So the Maloney virus that we work with is just such a virus. If you give almost any strain of mouse at birth Maloney, it will replicate and the mice will virtually all be dead in a few months of Leukemia.
VR: If you infect an adult mouse, what would happen?
SG: Well, they tend to get control of the infection by the immune system. They don’t get as viremic, they recognize the virus clearly as foreign, and they suppress it most often enough that they don’t get Leukemia, because Leukemia requires the infection of a very large number of cells. I don’t think we appreciate that the long lag that is because of the need for the virus to, by chance, insertionally activate some oncogene.
VR: So could XMRV infect a mouse?
SG: No, not per se, because they don’t have the receptor. There are rare mice that do, wild mice, and rats do. And we know that the difference in the critical amino acids of the receptor are the basis for that. But given the similarity to the ecotropic viruses, which do cause Leukemia in mice, and everything we know about the xenotropics. Even in the seventies, we were certainly aware of the possibility that they would cause, could cause, Leukemia – in man. And we worked with the virus pretty casually, I have to say, back in the seventies. But the history was, even then, that people who had been working with these viruses for years never showed any infections. More rec…only now have people started putting them into primates. There is now…some work being done in companies are seeing infections in primates and following the viremia. Nobody’s yet seen disease, but they have certainly seen virus replication in a variety of tissues, out many months/years in primates. So, you know I think the potential that they could cause disease is out there.
VR: How has the prostate connection held up since 2006, I guess it was?
SG: Yah, so we’re now seeing it in a small number of prostate samples. The range of prevalence depends on who you ask. It’s from very low – I guess in some subsets of people, zero, in some labs hands – to percentages as high as 20 percent and up in the case of our friend Ila Singh, now in Utah.
Transcription Part 2 by Robin
-41:34 to – 29:18
(note, my recording times didn’t quite line up with Kim’s — maybe we have a different browser or something? — so I had to hunt around to find where she left off. So, there might be a discprepancy in my time and the next person’s.)
VR: Are these all American studies? There’s some European studies —
SG: There are…
VR: There’s a German study —
SG: Yep. And I would say the American patients are being seen to have higher prevalence, probably than the European. Don’t know if that’s real or whether it’s just experimental. You know one of the issues —
VR: It is found in European patients so it’s not geographically restricted…?
SG: I can’t remember if all…if the European studies had any? I think yes; I think they were rare cases —
VR: Cause I have German studies, zero out of 589 —
VR: Zero out of 105, so there were two separate German studies. And there was one in Ireland too, zero of 139.
SG: The European ones are very low.
VR: So any — you went to CROI lately, recently, this meeting in San Francisco did you hear anything more about that?
SG: Yeah I mean there was…there were…some of those studies were reported…
VR: Nothing positive, though?
SG: Ummm…there were several low levels in prostate cancer in America, certainly. But the numbers were low.
VR: So this — does it make sense that this could cause prostate cancer?
SG: Well it’s possible. You wouldn’t ordinarily think of the virus as being specific for any tissue. The leukemia viruses are largely leukemia viruses because they grow in cells that are rapidly dividing, and we presume that’s why the murine viruses cause leukemia as opposed to a sarcoma. But, because otherwise insertional activation could happen in any tissue.
So could that be the case here? Yes, and what would make it tissue specific? Well, it could have been for example the distrubution of the receptor levels — we don’t think that’s the case, XPR is not high, particularly, in prostate tissue. We’ve focused on my lab on LTR, that is the promoter for the transcription from the virus, and that is interestingly very hot in prostate cell lines. So certainly that could account for the tendency of the virus to replicate more vigorously in prostate tissue.
We don’t know how, we haven’t done a complete survey (I don’t think anybody has yet) of all the tissues in which the promoter might be very hot, but it’s certainly very hot in a handful of human prostate cancer cell lines, surprisingly hot to us. It is as active in those cell lines as the Moloney mouse virus in the best mouse cells lines.
So it’s a real virus, and it can replicate very vigorously in culture, more so than we ever thought. I think nobody studied it much since the 70s because of the general feeling that it was an incredibly weak virus, that it was going to be difficult to study because it barely replicated in most cell lines. And that’s true. But now having found that it does replicate really really well in some cell lines, the studies will be much easier.
VR: So has any looked at any prostate tumors in the integration site?
SG: Yeah. So that’s being done now, as we speak. Nobody has found it — and this is important — in a negative result so nobody has found it yet clonally integrated next to an oncogene in a tumor. You know the early studies are still such that we don’t even know the cell types that it lives in. People are just doing sections, and Ila and others like that are seeing it in more cell types than originally thought. There were initially concepts that it was only in stromal areas, not in the tumor itself. Seems to be in both now.
But today nobody’s seen the kinds of things you see with leukemia, which is an insertional clonal insertion, next to an oncogene like myc that was clearly a clone triggering a tumor. So we don’t have any reason that we feel confident that it will be the case.
Prostate cancer is interesting — it may be not consistent with that anyway, in that it’s a disease where there’s a long phase of sort of chronic immune stimulation or irritation that might be consistent with some other mechanism than insertional activation. Maybe the tumor is the result of inflammation or expansion of cell lines in a non clonal way, maybe triggered eventually into a clonal tumor but the virus may not be the direct, immediate cause the way it is in leukemia.
VR: Like Hep B where you don’t have any oncogene activation but there’s a long term inflammation that may be involved.
VR: So, the one thing I always wondered about was you have a very high homology with this virus with mouse viruses, so, can you learn from that how recently it went from mice into people?
SG: You can — yeah — and there’s some issues there for this particular virus. Normally that’s right and the similarity is so great that people who really think about this deeply like John Coffin say, “this had to be relevantly recent, this had to be a fairly recent transmission”
VR: [unclear] ten years, twenty?
SG: Oh, more. But you couldn’t narrow it down that much, but, you know but thousands not millions of years ago.
SG: Now the problem is the various isolates that have all been sequenced from humans are unbelievably similar to each other. Very very similar, even more similar than the close similarity to the mouse viruses —
VR: But not identical, right?
SG: Not identical but very similar, and that’s very odd because viruses that we’re used to like HIV are very prone to mutation and they very quickly evolve and diversify so in the case of an HIV infected individual, we think of the person as infected with a swarm of virus, not with a single virus really because when you pull individual clones and sequence them, they’re different. This is not the case, apparently, with XMRV. And so there’s something funny about its lack of evolution across a person or across different people.
VR: So that would be consistent with it just being integrated in the genome, and not rep —
SG: Right. So that would be one thought, that it’s infecting a person and then replication is slow or not happening, and therefore you just retain the original copy, basically, that you were infected with for a long long period of time. But it’s not clear that’s the answer to this slow evolution. HTLV-1 is a little like this; HTLV-1 also is not as diverse as many other viruses and perhaps is replication is also restricted in some way in humans.
VR: So this is a biological issue. It’s not troubling in terms of the origin of the virus?
SG: Not — no, not really —
VR: It’s just something we need to understand…
SG: It might affect things like how often is transmission happening? And how many independent transmission from mice, let’s say, to people, have happened? In contrast, how often is person to person transmission going on? And we don’t know anything about the routes of transmission except by analogy to what we know about other viruses where you would say well you need blood sharing, or you need sex —
VR: Could it be sexually transmitted?
SG: Could be!
VR: Have people looked in semen for it yet?
SG: Yes, and people have also seen stimulation of infection by proteins founds in semen —
VR: That’s right there’s [unclear] fragments, right?
SG: Exactly. And that’s not unique to XMRV. That is true of a lot of retroviruses but consistent, again, with the possibility of sexual transmission.
VR: So what do you think — what has to be done to show that this is the causative agent of prostate cancer?
SG: Well, the smoking gun would be to find the virus integrated next to an oncogene in some subset of the tumors. That may not happen. That would be compelling if we saw that. Otherwise, I think we just need better epidemiology. We need — one of the things that’s really critical is we need reliable, reproducible assays for the virus. By PCR, probably. We need good assays for sero-conversion, because there are claims that there are people that are seropositive against the virus, and we need those to be done widely to get a sense of what is the real rate of infection in the population at large, in people with prostate cancer, and other diseases because we don’t really know the range which of diseases in which it may be found and that brings us eventually to the chronic fatigue issue,
VR: Sure. If it were clear that this virus was involved in prostate cancer it’s not going to be causing all prostate cancers, right?
SG: Oh, no way!
VR; Just a subset.
SG: Yep. Even initially it was only seen in that subset that are familial due to this inherited gene…
VR: That’s no longer the case.
SG: And now it looks like it’s being seen outside that subset. But the numbers remain small in terms of percentages of people.
VR: All right. Then at some point the people out in Nevada, what is it the institute — the Whittemore —
SG: I can’t remember the name…
VR: The WPI.
SG: Yeah (laughs) OK.
VR: They reported in Science that the same virus, essentially, but not exactly the same, but very similar —
SG: But very close.
VR: Yeah, they found in many chronic fatigue syndrome patients.
VR: And now the whole field has exploded.
SG: Right. They’re —
VR: Are there more people with CFS than prostate cancer in the world?
SG: I don’t know the numbers. I mean — the chronic fatigue diagnosis is a very difficult one —
VR: It’s hard to do.
SG: Yeah. It’s almost by exclusion. But, there are now criteria that are used to…score it. But, first of all one has to eliminate all the other causes of potential fatigue and what you’re left with is pretty much chronic fatigue syndrome. It’s been an issue for many decades whether it even existed as a disease. I think it’s well accepted now that it is. And then there was another long phase were many different viruses, bacteria even, were ascribed as the cause, EBV was a popular one, for a while, and I think all of those one by one have been ruled out.
VR: How do you rule them out? By just showing they’re not prevalent in most cases?
SG: Yeah, there’s no correlation. The viruses were in people with and without chronic fatigue equally and therefore there is no link, essentially.
VR: So the first study that was published in Science was a huge number positive of chronic fatigue patients. The first time that those kinds of numbers had been found, right?
SG: Well certainly the first time things looked as good correlating with a given virus. So this looked very striking and unbelievably exciting if you were in the field.
VR: And they also recovered virus from some of the patients.
SG: Right. Frank Ruscetti said it’s in plasma, and he was able to recover live replicating virus out of some of these people. So all the signs look good and today still do look good as far as one can see in this paper. But it’s also true that similar studies of even larger numbers of CF patients have been looked at in Europe and elsewhere and they’re not seeing it. So, either there’s geographic issues, there may be strain issues, there might be PCR primers that people are using that don’t work for some strains of virus. One base pair is all you need in the primer to cause you to miss that.
So at this point I think the CF connection is pretty much up in the air. I don’t feel confident that I would vote one way or the other yet. And I just think we have to get more data.
Transcribed by Catch
VR: ‘Course the people who originally found it are very vehement that it’s definitely the cause and that everyone else doesn’t know how to do PCR
SG: Right well I mean I think that they’re at least adamant that it’s, that the link is real. And I think they’re pretty strongly coming out about cause. And I think that’s a very tough call to make at this point. There’s no way I would feel comfortable claiming that this was the cause of either prostate cancer or any other disease at the moment.
VR: I mean they claim to have data from all over the world implicating the virus in CFS but we haven’t seen that in the literature so until we do-
VR: Were they at the meeting CROI?
SG: Ah not much, no, that crowd was not at CROI
VR: They should have been
SG: They should have been, probably. There are more workshops happening all the time. There’s one coming up in the next month at the NIH. So I think all this will come out.
VR: So you’re basically, you need more data.
VR: That’s what it comes down to.
VR: It could be geographically limited, it could be a detection problem…
VR: I have a review here by Joaquin?? Denner?? who says that in cases of low proviral load antibody can indicate infection even when PCR doesn’t. Is that true?
SG: I’m not sure what example virus he’s talking about-
VR: HIV I think
SG: But I think certainly you could – because PCR can be done at any level of sensitivity, if you include that – you can certainly imagine a virus having been cleared (HIV not being cleared but a non-retrovirus being cleared), completely cleared so there truly is no signal that you could possibly detect by PCR, and the person would still retain seropositivity for decades, so, certainly that is a possibility. And I think it’s also very compelling if you do have seropositivity because it’s hard to, it’s not contamination any more, as an issue. You’ve got a person who’s got an antibody response, he probably had to be infected.