Exonerating a Virus
It’s not every day that scientific controversies spill out of the pages of peer-reviewed journals and into the mainstream press. But that’s exactly what happened in the wake of a 2009 study linking a virus called XMRV to chronic fatigue syndrome and opening the door to a potential cure.
As subsequent research failed to confirm the finding, the original authors stood by their work, raising the hopes of patients with the disease that has baffled scientists since the illness first came to public attention in the mid-1980s. Now, work by investigators at Tufts University published in the June 2 edition of the journal Science has helped put the controversy to rest—so convincingly that editors of the journal have asked the original authors to retract voluntarily their 2009 study.
Accepted as a legitimate medical condition in 1987, chronic fatigue syndrome (CFS) is a mysterious illness characterized by persistent exhaustion that affects as many as 17 million people worldwide. Researchers at the Whittemore Peterson Institute for Neuro-Immune Disease in Nevada—a private center founded to investigate potential causes and cures for CFS and related illnesses—screened blood samples from 101 of their sickest patients and from 218 healthy individuals in search of XMRV, which had been implicated in prostate cancer. In a paper published in Science in 2009, they reported finding the virus in a whopping two-thirds of the CFS patients, but in less than 4 percent of the healthy people. The association between the debilitating disease and the previously little-known virus seemed crystal clear.
But it wasn’t long before contradictory data began to pile up. When XMRV failed to turn up in chronic fatigue patients in study after study, many biomedical researchers began to suspect that XMRV had been accidentally introduced into the tissue samples used in the Whittemore Institute research.
John M. Coffin, a professor of molecular biology and microbiology at the Sackler School of Graduate Biomedical Sciences, and Oya Cingöz, a genetics student in Coffin’s lab, collaborated with colleagues around the world to get to the bottom of the controversy. Their results not only discredit the purported association between XMRV and chronic fatigue syndrome, but also suggest that prostate cancer researchers inadvertently created XMRV in the early 1990s.
Coffin, who suspects an infectious pathogen of some sort might be at the root of CFS, is well suited to weigh in. He’s an expert in retroviruses, the subset of viruses to which XMRV belongs. Retroviruses can be particularly insidious pathogens because they don’t just infect host cells; they insinuate themselves into the host’s DNA, becoming a permanent feature of the host’s genome. That’s why retroviral infections such as HIV and certain types of leukemia tend to lead to long-term illnesses that are difficult to cure.
XMRV stands for “xenotropic murine leukemia virus-related virus.” Put simply, it’s a human virus that strongly resembles a class of mouse viruses called MLV. Because all mouse cells contain MLV—and because mice are the preferred lab animal for many biomedical researchers—MLV contamination “in any retrovirology laboratory is something of an occupational hazard,” Coffin, Cingöz and their colleagues wrote in the July 2010 issue of the journal Retrovirology.
In that study, the researchers tested DNA from human prostate cancer tissues from the United Kingdom, Thailand and Korea and found that specimens from each country initially appeared to be XMRV-positive because they contained gene sequences known to be present in the XMRV genome.
But the XMRV genome is quite similar to certain mouse viruses, and when the scientists looked at the viral fragments more closely, they detected some gene sequences specific to the mouse viruses that are not found in XMRV. While that didn’t rule out the presence of XMRV in these tissues, it did mean the samples were definitely contaminated by mouse DNA.
In another study, also published in July 2010 in Retrovirology, co-investigator Brigitte T. Huber, a professor of pathology at the School of Medicine, tested blood samples from chronic fatigue syndrome patients taken by collaborators at Harvard and in New York in 2005 and in 2009–10. Again the researchers found that all of the samples that appeared to be XMRV-positive also contained traces of mouse DNA, providing “a strong suspicion that the viral sequences detected in these samples were actually of [mouse] origin,” they wrote in Retrovirology.
The second study also highlighted how the test used to screen for XMRV (called a polymerase chain reaction assay) is capable of detecting vanishingly small amounts of mouse DNA, thus giving false positives for XMRV. “It doesn’t take much at all; one [mouse] cell could contaminate a whole bunch of experiments,” says Cingöz.
A Viral Alibi
Though the Whittemore Institute researchers who had announced the XMRV–CFS link weren’t convinced of the mouse DNA contamination hypothesis, the Tufts team and their colleagues developed another line of evidence that proved even more conclusive.
In the paper published in the June 2 edition of Science, Coffin, Cingöz and their colleagues at the National Cancer Institute and the University of California were able to show that XMRV came into existence just 15 years ago, and therefore could not be responsible for CFS, which the Centers for Disease Control and Prevention had identified in 1987 and was a suspected cause of illnesses that had been described in the medical literature as far back as 1934.
Cingöz found a retrovirus in the mouse genome whose DNA closely resembled one section of the XMRV genome. Serendipitously, a team of researchers led by Vinay Pathak at the National Cancer Institute, where Coffin consults with the HIV Drug Resistance Program, found another mouse retrovirus that resembled the rest of the XMRV genome. “They were exactly complementary,” says Cingöz. The discovery indicated the two separate mouse viruses infected human cells, where they combined to become XMRV.
What’s more, Cingöz and colleagues were able to pinpoint when that happened.
Prostate cancer researchers first identified XMRV in 2006. Human prostate cancer cells are notoriously difficult to grow in the lab for study. Since the 1970s, the standard procedure has been to implant prostate tumor cells into lab mice, let the tumor grow and then excise it for study. To maintain the consistency of their research results, scientists around the world use cells from the same lineages, keeping careful records of their cells’ genealogy.
Working backwards, the National Cancer Institute (NCI) team was able to trace the tissue samples containing XMRV to the original source—a prostate cancer study that began at Case Western Reserve University in Cleveland, Ohio, in 1992. A year later, the tumor cells were negative for XMRV, but by 1996, the samples tested positive for the virus. The two mouse retroviruses found by the Tufts and NCI teams must have infected the human prostate cells, where they combined to become XMRV. Because it is a retrovirus, invading the DNA of its host, XMRV appears in every subsequent generation of cells from this lineage.
“XMRV was generated in the lab,” says Cingöz. “Cell lines becoming contaminated with retroviruses is well documented.”
That view is increasingly being accepted in the scientific community. In an editorial accompanying reports by Coffin and his colleagues, Science editor-in-chief Bruce Alberts noted that the 2009 study “is now seriously in question,” and that the journal is publishing what’s called an “expression of concern” and attaching it to the original 2009 paper in Science.
The Reno scientists as well as many chronic fatigue syndrome patients and their advocates still cling to the XMRV hypothesis, despite the evidence that Coffin and others have presented to refute that. That’s not surprising, given the emotional tenor of the debate. And that’s OK, says Coffin. “Some good will come of this,” he says. “It brought chronic fatigue syndrome back into spotlight, and it will prompt more funding.”
Jacqueline Mitchell can be reached at firstname.lastname@example.org.