Dusty Miller's New XMRV Study

[Mindy Kitei]

Dr. Miller gave a statement to CFS Central about his study, which will be published tomorrow in the Journal of Virology.

Mindy Kitei
CFS Central
http://www.cfscentral.com

 

[Jemal]

Another negative study then.
Thanks for posting Mindy.

 

[eric_s]

So he cancelled his initital XMRV study, because of the reaction of patients or whatever, and then went on to try to "disprove" XMRV?
And at the same time one of the "members" of his lab has done research that a couple of weeks ago surfaced here that concluded XMRV could explain the symptoms of ME/CFS, but then posts had to be removed again, because it was not published yet.
More confusion...

In many labs they have sequenced what they have found. And they say it's XMRV. How could this be explained by an endogenous mouse virus? Could it, i'm no scientist, so i'd like to know...
And then we have the antibody findings and mouse mtDNA tests and even IAP tests coming up negative.
Or Ruscetti who reported, if i remember correctly, that they find more XMRV in prostate cancer patients after some sort of stimulation.
It just doesn't really add up, it seems.

I don't know if they are doing it on purpose, but i really feel like we are being bull*ed a bit here... by whoever.

 

[Jemal]

This is a quote from Dusty Miller:

Unfortunately, we did not find an intact copy or the right half of XMRV in any of the mouse cell lines or tissue that we analyzed, but clearly we did not look hard enough. Paprotka et al. firmly established the origin of XMRV from nude mice. We were pursuing the same hypothesis, but could not get early samples of the cell lines and tissues from which the XMRV-carrying 22Rv1 cells were derived.

I think it is significant that they did not find the right half.

 

[eric_s]

I'm not really sure about the point they're trying to make. It seems they are saying they've found an endogenous mouse virus that can lead to false PCR positives. Why would they then also look for an intact copy or the right half of XMRV? Coffin said it's a very rare event that there's a recombination that would lead to XMRV, right? So if that's what happened in this one mouse where 22Rv1 came from, why would then XMRV also be in the mouse Miller was looking at?

 

[Jemal]

I'm not really sure about the point they're trying to make. It seems they are saying they've found an endogenous mouse virus that can lead to false PCR positives. Why would they then also look for an intact copy or the right half of XMRV? Coffin said it's a very rare event that there's a recombination that would lead to XMRV, right? So if that's what happened in this one mouse where 22Rv1 came from, why would then XMRV also be in the mouse Miller was looking at?

I think they can only say for sure that they found an endogenous mouse virus, when they have both halves.
I don't think this study brings anything new to the table... but I could be wrong. The study hasn't been published yet.

 

[Esther12]

Is it that they thought the different component parts of XMRV were likely to be common, even though the possibility of them combining to form XMRV is so unlikely that its would have been a one off event?

This study doesn't seem to add much new, but I think it was going on at the same time as the other, similar, work.

And at the same time one of the "members" of his lab has done research that a couple of weeks ago surfaced here that concluded XMRV could explain the symptoms of ME/CFS, but then posts had to be removed again, because it was not published yet.

I missed that!

One of the haunting things about XMRV is that there are things about it which, if you ignore all of the negative/contamination studies, seem to fit unusually well with CFS (the prevalence rates make it plausible that XMRV could require some genetic predisposition/trigger to result in CFS, it's apparent reaction to EBV, etc) - but I'm still not expecting it to work out.

It's a strange story, with bizarre coincidences needed to explain XMRV as either contamination or genuine infection, and it's not over yet, but I'd be surprised if we got another twist in the tale to bring XMRV back. (Gandalf lives!)

 

[Jemal]

I missed that!

Many have I think, but it happened. I was the one making the post on this forum and later asked to have it removed, because the information was still under embargo.

Ah Gandalf... someone says after he seems to have died: "He has fallen into shadow". ME/CFS is a bit like that.

 

[alex3619]

I'm not really sure about the point they're trying to make. It seems they are saying they've found an endogenous mouse virus that can lead to false PCR positives. Why would they then also look for an intact copy or the right half of XMRV? Coffin said it's a very rare event that there's a recombination that would lead to XMRV, right? So if that's what happened in this one mouse where 22Rv1 came from, why would then XMRV also be in the mouse Miller was looking at?

Hi eric_s, there are several points I would like to make. The first is that Miller did not find XMRV but one of the endogenous viruses that could give rise to XMRV, from what I recall - if I am wrong about that, please let me know.

Now, Coffin and recombination: recombination is common, typically occurring multiple times in a replication cycle. The argument rests on a claim that it can only occur in one by then to the twelfth power of replications. However, multiple the number of animals in the world, by how many infected cells, by how many replication cycles and you can get a very big number. Using moderate (not extreme) figures I got ten to the power eighteen replication events in the last century alone. So, presuming my calculations were not wildly off, that means XMRV could have arise a million times last century (ten to the sixth). Rare - only if you look at a single replication event, the odds are very very low for one event.

Given the paper at the Belgian conference showing that other murine leukemia viruses might recombine to make XMRV, and that recombination is common, its almost a certainty that any population with PMLVs and XMLVs will lead to XMRV. It is therefore almost a certainty that if Lo and Alter are right, so is the WPI.

That does not mean that contamination is not an issue, but there are established controls and procedures to deal with MLV contamination. Contamination with wild XMRV is the main risk.

However, if XMRV is wild, and can infect humans, then some will have it. This is the background rate. This conclusion then invalidates all the tests used in the zero zero studies.

The only hope for the contamination theorists is that they can show sources of lab culture derived XMRV infecting every lab that has had positive XMRV studies. I think this is extremely improbable.

As for the unpublished study for which the comments were deleted, it ties in with my own model of ME/CFS, circa 2002, which was never published or discussed in public. I may have more to say on this once the paper is published. I will refrain from discussing it until then, however much it is eating at me. :mask:

Bye
Alex

 

[Bob]

So he cancelled his initital XMRV study, because of the reaction of patients or whatever, and then went on to try to "disprove" XMRV?
And at the same time one of the "members" of his lab has done research that a couple of weeks ago surfaced here that concluded XMRV could explain the symptoms of ME/CFS, but then posts had to be removed again, because it was not published yet.
More confusion...

In many labs they have sequenced what they have found. And they say it's XMRV. How could this be explained by an endogenous mouse virus? Could it, i'm no scientist, so i'd like to know...
And then we have the antibody findings and mouse mtDNA tests and even IAP tests coming up negative.
Or Ruscetti who reported, if i remember correctly, that they find more XMRV in prostate cancer patients after some sort of stimulation.
It just doesn't really add up, it seems.

I don't know if they are doing it on purpose, but i really feel like we are being bull*ed a bit here... by whoever.

Hi eric_s, there are several points I would like to make. The first is that Miller did not find XMRV but one of the endogenous viruses that could give rise to XMRV, from what I recall - if I am wrong about that, please let me know.

Now, Coffin and recombination: recombination is common, typically occurring multiple times in a replication cycle. The argument rests on a claim that it can only occur in one by then to the twelfth power of replications. However, multiple the number of animals in the world, by how many infected cells, by how many replication cycles and you can get a very big number. Using moderate (not extreme) figures I got ten to the power eighteen replication events in the last century alone. So, presuming my calculations were not wildly off, that means XMRV could have arise a million times last century (ten to the sixth). Rare - only if you look at a single replication event, the odds are very very low for one event.

Given the paper at the Belgian conference showing that other murine leukemia viruses might recombine to make XMRV, and that recombination is common, its almost a certainty that any population with PMLVs and XMLVs will lead to XMRV. It is therefore almost a certainty that if Lo and Alter are right, so is the WPI.

That does not mean that contamination is not an issue, but there are established controls and procedures to deal with MLV contamination. Contamination with wild XMRV is the main risk.

However, if XMRV is wild, and can infect humans, then some will have it. This is the background rate. This conclusion then invalidates all the tests used in the zero zero studies.

The only hope for the contamination theorists is that they can show sources of lab culture derived XMRV infecting every lab that has had positive XMRV studies. I think this is extremely improbable.

As for the unpublished study for which the comments were deleted, it ties in with my own model of ME/CFS, circa 2002, which was never published or discussed in public. I may have more to say on this once the paper is published. I will refrain from discussing it until then, however much it is eating at me. :mask:

Bye
Alex

Interesting discussion.

Singh detected XMRV in 4% (or 6%?) of healthy controls when testing prostate tissue.
http://www.pnas.org/content/106/38/16351.full
But when she tested blood (in the CFS study), she found 0% in healthy controls.
So, either her prostate results are wrong, or her blood results are wrong.

This is a significant discrepancy which she hasn't highlighted, but when asked about it, she said that she didn't know why there was a difference:
"Not entirely sure, but there were different assays (e.g. immunohistochemistry) and different sample types (blood vs prostate tissue)."
http://www.cfscentral.com/2011/05/dr-ila-singh-we-are-now-convinced-that.html

At least she admits that the differences might be due to her methodologies, but this does question the basis for her conclusions:
"I'd urge people to move on rather than to keep their hopes hanging on the link between XMRV and CFS,"
http://news.sciencemag.org/scienceinsider/2011/05/more-bad-news-for-chronic-fatigue.html

Singh says about finding XMRV in prostate cancer patients:
"This increases the population at risk for XMRV infection from only those homozygous for the RNASEL variant, to all individuals. Our goal is to further understand pathogenesis by XMRV, to determine if this virus causes prostate cancer, and delineate possible mechanisms for oncogenesis. We will also carry out epidemiological studies to study the prevalence and tropism of this new and important virus."
http://projectreporter.nih.gov/project_info_description.cfm?aid=7866444&icde=4780366

So all individuals are at risk from XMRV infection, except ME patients!?! How logical.

 

[eric_s]

Singh detected XMRV in 4% (or 6%?) of healthy controls when testing prostate tissue.
http://www.pnas.org/content/106/38/16351.full
But when she tested blood (in the CFS study), she found 0% in healthy controls.
So, either her prostate results are wrong, or her blood results are wrong.

This is a significant discrepancy which she hasn't highlighted, but when asked about it, she said that she didn't know why there was a difference:
"Not entirely sure, but there were different assays (e.g. immunohistochemistry) and different sample types (blood vs prostate tissue)."
http://www.cfscentral.com/2011/05/dr-ila-singh-we-are-now-convinced-that.html

To be fair, probably one would have to say that both her results could be right, if XMRV was absolutely impossible to detect by PCR in the blood. Because her other work didn't look for it in the blood.

But not finding something and then concluding from that that it can't be found and so the ones who have reported finding must be wrong seems a bit much for me. I feel that the conclusion that one group for whatever reason is better at detecting the virus is more likely. But that's of course not how you do science, i don't know who will be right in the end. I just want everybody to get their fair chance.

 

[eric_s]

Hi eric_s, there are several points I would like to make. The first is that Miller did not find XMRV but one of the endogenous viruses that could give rise to XMRV, from what I recall - if I am wrong about that, please let me know.

Hi Alex. In Mindy Kitei's article Miller was quoted as having said

Unfortunately, we did not find an intact copy or the right half of XMRV in any of the mouse cell lines or tissue that we analyzed, but clearly we did not look hard enough.

That confused me a bit. Because XMRV is thought to be xenotropic (i think it's not 100% xenotropic, i think i once read it can infect some mice) and also Coffin said such a recombination leading to XMRV is extremely rare. So why would they go look for XMRV in mice? Coffin's hypothesis says XMRV is only in the mouse or tumor tissue where that one event took place and then in the cell lines developped from there. Or not? So to some degree he seemed to me to contradict Coffin's hypothesis, even though he argues for contamination himself.

 

[Bob]

Singh detected XMRV in 4% (or 6%?) of healthy controls when testing prostate tissue.
http://www.pnas.org/content/106/38/16351.full
But when she tested blood (in the CFS study), she found 0% in healthy controls.
So, either her prostate results are wrong, or her blood results are wrong.

To be fair, probably one would have to say that both her results could be right, if XMRV was absolutely impossible to detect by PCR in the blood. Because her other work didn't look for it in the blood.

Yes, you're right of course eric... The point that I was trying to make was that the two different sets of results indicate that either the conclusions that Singh made from the prostate study are wrong, or the conclusions that she made from the blood study are wrong... Unless there is some other explanation that we aren't aware of.

 

[alex3619]

Hi Alex. In Mindy Kitei's article Miller was quoted as having said That confused me a bit. Because XMRV is thought to be xenotropic (i think it's not 100% xenotropic, i think i once read it can infect some mice) and also Coffin said such a recombination leading to XMRV is extremely rare. So why would they go look for XMRV in mice? Coffin's hypothesis says XMRV is only in the mouse or tumor tissue where that one event took place and then in the cell lines developped from there. Or not? So to some degree he seemed to me to contradict Coffin's hypothesis, even though he argues for contamination himself.

Hi eric_s, the same section on CFS Central, Miller is quoted: Essentially, we found an endogenous retrovirus (mERV-XL) in NIH/3T3 cells, a commonly used mouse cell line, that is virtually identical to what Paprotka et al. are calling PreXMRV-2.

What they found was pre-XMRV, but only one of the two Coffin found. When the discussion moves on to XMRV, my take it that this is about the pre-XMRV endogenous viruses. These mice are not "infected", this is DNA.

However, these pre-XMRV sequences will show up on PCR as bits of XMRV due to sections that are nearly identical to actual XMRV.

Bye
Alex

 

[eric_s]

However, these pre-XMRV sequences will show up on PCR as bits of XMRV due to sections that are nearly identical to actual XMRV.

Bye
Alex

Yes, but if they sequence what they find, i guess they should be able to distinct XMRV from mERV-XL, or what do you think? And they sequenced XMRV many times, in different labs. So that sort of false positives can't explain the positive studies, i guess. Not to speak of the antibody findings.

 

[biophile]

Given the paper at the Belgian conference showing that other murine leukemia viruses might recombine to make XMRV, and that recombination is common, its almost a certainty that any population with PMLVs and XMLVs will lead to XMRV. It is therefore almost a certainty that if Lo and Alter are right, so is the WPI.

Would it also be possible for a virus in patients to combine with lab contamination?

 

[justinreilly]

My comment:

Dr. Miller:

I am a layperson. Why do you say 'obviously you didn't look hard enough', that the Coffin paper 'proved' preX1 and 2 recombined to make XMRV? He has not been confirmed yet by other labs. Would you say that all those who didn't find XMRV, unlike Lombardi et al., obviously didn't look hard enough because Lombardi proved HGRVs are associated with ME?

If XMRV is everywhere "contaminating"/infecting everything, why can't you and these other labs find it? You can find mERV-XL, but not XMRV or the other 'parent virus'? It seems to this layperson that either your lab is incompetent and shouldn't be publishing on these data or Coffin is wrong or both. Please disillusion me.

 

[alex3619]

Would it also be possible for a virus in patients to combine with lab contamination?

Hi biophile, if the lab contamination is a replication competant MLV, such as lab derived XMRV, then it could do so in a living patient, or a tissue culture. It would not do so in a test tube - the probability is too low. However, the chance of a small culture dish giving rise to XMRV due to contamination with something that was not XMRV might be about one chance in 10000000000. That is just a guess. In a living host the chance would be much higher, over time. For contamination arguments to hold, I think it requires that XMRV itself is the contaminant.

Bye
Alex

 

[alex3619]

My comment: If XMRV is everywhere "contaminating"/infecting everything, why can't you and these other labs find it? You can find mERV-XL, but not XMRV or the other 'parent virus'?

Hi justinreilly, this is indeed a major flaw in the contamination arguments, as is the fact that mouse contamination is routinely screened for and found - and results in equal contamination of controls and patient samples. However, it could be the stain of mouse tested is XMRV immune, and I have not gone back to check that (too much work to do it for such a minor point).
Bye, Alex