Science
March 10, 2011
More Negative Data for Link Between Mouse Virus and Human Disease
By Jon Cohen
BOSTONA new finding presented at a conference here last week throws cold water on the impassioned debate about the link between a novel mouse retrovirus and prostate cancer and chronic fatigue syndrome in humans. Yet few believe it will end the controversy, which began in 2006.
In an extensive sleuthing expedition that looked back nearly 20 years, two collaborating research teams contend that they have evidence that xenotropic murine leukemia virusrelated virus (XMRV) resulted from the chance recombination of pieces of two mouse viruses in lab experiments and that the connections to human disease are spurious. That nails it, said retrovirologist Nathaniel Landau of New York University. Everyone working on this thing has this virus contaminating their stuff. It's been a tremendous waste of time and money. But even if XMRV is not a threat to human health, the fact that a retrovirus that can readily infect human cells was apparently generated by chance in the lab raises some interesting and potentially troubling issues.
Vinay Pathak, a retrovirologist who works at the HIV Drug Resistance Program run by the U.S. National Cancer Institute (NCI) in Frederick, Maryland, presented the new data at the 18th Conference on Retro viruses and Opportunistic Infections, which focuses mainly on another retrovirus, HIV. (For additional coverage of the meeting, see pages 1248 and 1249.) The fact that the XMRV work garnered so much attention here reflects the high stakes. The possibility that XMRV causes human disease has raised both hope and fear among patients and public health officials (Science, 2 July 2010, p. 18). For people who have prostate cancer or the baffling chronic fatigue syndrome, XMRV offered not only an explanation but also a treatment: The virus is susceptible to some anti-HIV drugs. Blood banks, on the other hand, have worried mightily that, as happened when the AIDS epidemic began, they were unwittingly helping to spread a dangerous retrovirus.
The unusual life cycle of retroviruses explains how such a recombination could occur, as Pathak described. Retroviruses contain RNA that, in addition to coding for viral proteins, carries instructions to make the enzyme reverse transcriptase. After a retro virus infects a cell, reverse transcriptase converts the viral RNA into DNA, which is necessary for the virus to integrate with the host chromosomes. This is the stage in which recombination between retroviral DNA from different genomes can happen.
Pathak explained how skepticism has steadily built about the link between XMRV and these diseases as several labs examined patient samples and could not find the virus or antibodies to it (Science, 17 September 2010, p. 1454). One 2009 study particularly piqued Pathak's interest, as it showed how a human prostate cancer cell line produced high levels of the virus.
The cell line was established at Case Western Reserve University in Cleveland, Ohio, from a human tumor called CWR22. Prostate cancer tumors are difficult to grow in lab experiments, but in 1993, researchers there reported that they had success by injecting tissue from CWR22 into mice, growing tumors, injecting tissue from those xenografts into new mice, and repeating that passaging process until they could reliably grow large enough xenografts for study. In 1999, the same lab described a permanent cell line, 22Rv1, it had made from a CWR22 xenograft. Because it was one of very few cell lines available to study prostate cancer, it was widely used. Pathak's group tracked down samples from different passages of CWR22, different versions of both 22Rv1 and a second cell line made later from CWR22. Before 1996, no CWR22 samples contained XMRV DNA.
Pathak's lab found that some of the early samples of xenografts did have a stretch of DNA that was nearly identical to about half of the XMRV genome. A group led by John Coffin, who works at both NCI and Tufts University here, made a similar discovery with different samples of xenografts. When the teams compared notes, they saw that the two sequences perfectly overlapped to form XMRV. It was an amazing moment, the kind that happens once or twice in a career, Coffin says. It was like seeing a puzzle come together.
As Pathak emphasized in his talk, the DNA sequences in what they dubbed preXMRV-1 and preXMRV-2 are nearly identical to the XMRV sequences reportedly found in humans but suspected to be a lab contaminant by some groups. Pathak's and Coffin's teams both also found preXMRVs in some mice strains used in the experiments. But XMRV itself cannot infect mouse cells, which means the preXMRVs could have recombined only after the mice received prostate tumor transplants that contained human cells. Specifically, RNA from both preXMRV genomes must have been packaged in a newly formed viral particle, or virion. When that virion infected a human cell derived from the prostate tumor, the reverse transcriptase enzyme accidentally mashed up the preXMRVs and created XMRV. It's a very elegant study, says phylogeneticist Stphane Hu of University College London. This is the birth date of the virus.
Hammering the nail in further, Oya Cingz in Coffin's lab looked for XMRV in dozens of inbred and wild mice and reported that she found no evidence that the virus naturally exists.
Coffin believed earlier that studies linking XMRV to human disease deserved serious attention. He co-authored an article in the 23 October 2009 issue of Science, which included the first report of XMRV in patients who had chronic fatigue syndrome. Led by Vincent Lombardi of the Whittemore Peterson Institute in Reno, Nevada, and NCI's Francis Ruscetti, the study provided more evidence that, as Coffin's piece stated, transmission happened in the outside world and was not a laboratory contaminant. Now, Coffin has changed his thinking. It's all contamination, he says. At this point, Coffin questions whether any human has been infected with the virus. It remains a distant possibility, he says.
Hu, who works with Greg Towers in London, presented complementary 22Rv1 data at the conference that they published 20 December 2010 in Retrovirology (Science, 7 January, p. 17). They showed that XMRVs isolated from different 22Rv1 cell lines were more genetically diverse than sequences reportedly found in chronic fatigue and prostate cancer patients. If XMRV infected humans, copied itself, and spread to others, Hu says he would expect to see more diversity in the patients as it evolved to escape immune defenses. I don't think XMRV is a human pathogen, Hu says. It's as simple as that. Like Coffin, he doubts that XMRV has even infected a human but adds that one can never say that something doesn't exist.
The evidence coming out at this meeting is incredibly impressive, and the weight of evidence is indicating that this is not a major human virus in terms of pathogenesis, says Michael Busch, who heads the Blood Systems Research Institute in San Francisco, California, and is part of a working group convened by the U.S. Department of Health and Human Services to examine whether XMRV poses a threat to the country's blood supply. But Busch said that before he concludes XMRV is simply a contaminant, he wants to see the results of studies they are coordinating between several labs with samples from agreed-upon patient and negative controls, as well as blood donors. If most of these fail to find the virus, Busch says, it's going to eliminate concerns that XMRV has caused these diseases.
Even if XMRV has not harmed humans, Busch says we got lucky. This is the first accidental generation of a retrovirus that can infect human cells. It's a warning shot, Busch says. We've created a highly infectious virus that may transmit to humans.
