National Cancer Institute NCI Cancer Bulletin: A Trusted Source for Cancer Research News
June 14, 2011 • Volume 8 / Number 12

A Closer Look

Perfect Match: Unraveling the Origin of XMRV

The team that uncovered the origins of XMRV, Drs. Krista Delviks-Frankenberry, Tobias Paprotka, and Vinay Pathak, head of NCI's Viral Mutation Section. The team that uncovered the origins of XMRV included (from left) Drs. Krista Delviks-Frankenberry, Tobias Paprotka, and Vinay Pathak, head of NCI's Viral Mutation Section.

Dr. Vinay Pathak was tying up loose ends at his NCI-Frederick lab and packing for a trip abroad last December, when he received an e-mail message that began, "GOT IT."

The message, from Dr. John Coffin of Tufts University, described a genetic analysis of two mouse viruses. Both viruses—one discovered by Dr. Coffin's lab and the other by Dr. Pathak's—resembled parts of the retrovirus XMRV, but they had never been compared.

When the viral DNA sequences were lined up for the first time, Dr. Coffin wrote in the e-mail, they were nearly a "perfect match" with XMRV (xenotropic murine leukemia virus-related virus). "How the DNA sequences fit together was really remarkable," he said recently. "They clicked like two magnets pulling themselves together."

This single experiment tied together months of clues the researchers had been gathering on the origins of XMRV, which was discovered in prostate cancer cells in 2006. Based on their evidence, Drs. Coffin and Pathak now believed that XMRV arose when the two mouse viruses came together (a process known as recombination) during a lab experiment in the 1990s.

In the experiment, a lab worker was growing a patient's prostate tumor cells in mice until the tumor cells could be cultured—the first step in making a colony of cells, or cell line, for research. When one of the samples was eventually harvested from a mouse, the researchers said, the sample contained XMRV. Then, as many labs began to use the cell line, the virus spread.

The details of this detective story were published online in Science on June 2. In the article, the authors concluded that XMRV emerged when the two "parent" viruses recombined in a lab mouse that was being used to grow human tumor samples.

"These results really tell us that XMRV, as we know it, has not been circulating in the human population," said Dr. Hung Fan, a retrovirologist at the University of California, Irvine, who was not involved in the research.

You must have flash installed and enabled to view the video.

Video produced and edited by Sarah Curry

Evidence Points to Contamination

Three years after XMRV was discovered in prostate cancer cells, another group reported finding the virus in patients with chronic fatigue syndrome (CFS). Few labs have been able to confirm this initial report by detecting the virus in patient samples, however, and the possible role of XMRV in human disease has been a source of controversy.

All XMRV isolates reported to date are closely related to the viral sequence found in the prostate cancer cell line known as 22Rv1. If the retrovirus had been replicating in humans, these sequences would contain much more variation, several researchers noted.

Furthermore, the chance of the virus arising independently in another mouse or in the wild is "vanishingly small—about one in a trillion," said Dr. Pathak. This virus is unlikely to exist in nature, so the chance that a person would be exposed to a mouse with the virus is also unlikely, he added.

"Admittedly, we have no direct evidence of a patient sample being contaminated," said Dr. Jonathan Stoye of the United Kingdom's National Institute for Medical Research, who was not involved in the research. "But all of the indirect evidence leads one to believe that contamination is the most likely explanation" for reports of XMRV, he continued.

At a recent workshop on XMRV, co-led by Dr. Stoye, some presenters cautioned that contamination could be a reason that some investigators had detected XMRV in patient samples.

"The [current] study documents a new example of the transfer of viruses from mouse to human cells," said Dr. Stephen Goff, a retrovirologist at Columbia University who also had no role in the study. "This kind of transfer has occurred many times before in other experiments.

"When human tumors are grown in mice," he continued, "the tumors can pick up mouse viruses."

For researchers in the field, a take-home message of the current study is that "anyone who publishes a study saying that XMRV is present in a patient sample needs to look very hard at the data," said the first author, Dr. Tobias Paprotka, a member of Dr. Pathak's lab. 

"If it's XMRV, it is almost certainly contamination," added Dr. Pathak.

Testing Prostate Cancer Cells from the Donor

When his investigation began, Dr. Pathak was focused on one question: How had XMRV gotten into the 22Rv1 cell line? As the work progressed, the answer—and the origin of the virus itself—emerged.

Dr. Fan called the study a nice example of "molecular viral archeology," noting that the researchers had tracked down tumor samples in freezers around the country. In California, for instance, Dr. Pathak located a series of tumor samples from the experiment in which the tumor samples had been grown in mice and were later used to create the 22Rv1 cell line.

Dr. John Coffin and Oya Cingöz of Tufts University (Photo by Kelvin Ma) At Tufts University, Dr. John Coffin and Oya Cingöz, a genetics student in his lab, helped to trace the emergence of XMRV. (Photo by Kelvin Ma)

An important clue emerged when his lab tested the earliest samples and found no XMRV. "If the original tumor from a patient had XMRV, then all the samples derived from this tumor should have had it," said Dr. Pathak. "This was not the case."

The researchers discovered XMRV in the later samples, however, indicating that the virus must have emerged during the experiment. Without knowing the risk of contamination, nonvirology labs then handled the prostate cancer cells for a decade.

Once the virus was in the cell line, "it is not hard to imagine how it could have been spread through many labs," said Dr. Coffin, who works part of each month at NCI-Frederick and is a special advisor to the NCI director. In his lab at Tufts, Oya Cingöz, a graduate student in genetics, conducted much of the research.

Although XMRV replicates well in culture and could infect people, it is unlikely that the virus would spread easily from person to person. The reason, Dr. Fan explained in an e-mail, is that humans "have a fairly potent intracellular resistance factor that prevents XMRV infection."

On the Trail of a Mysterious Virus

In the summer of 2009, XMRV was seen as a mysterious and potentially threatening retrovirus. A few months earlier, at the annual Cold Spring Harbor Laboratory meeting on retroviruses, researchers presented a study suggesting that, if the virus were real, millions of people in the United States could be infected.

Because of the prostate cancer link and the possibility that a new retrovirus could be circulating in the population and causing disease, NCI convened a meeting of experts in July 2009.

By September, a plan and funding were in place to develop the tools and resources needed to study XMRV. Dr. Stuart Le Grice of NCI-Frederick had assembled a team of retrovirologists to apply their experience working on HIV to questions about a new retrovirus.

"We looked at XMRV from the perspective that it could be real or it could be contamination," said Dr. Le Grice. "We had no bias going in."

Less than 2 years later, the field has an answer. With the findings published, Dr. Le Grice said, it is important to remember that "there is still no evidence that this virus was ever in patients."

The findings appeared in Science along with a companion report on CFS. In that report, researchers were unable to detect evidence of XMRV in patients whose blood samples had tested positive for the virus in the original 2009 study, also published in Science.

The chance of the virus arising independently in another mouse or in the wild is "vanishingly small—about one in a trillion."

—Dr. Pathak

Since then, at least 10 independent studies have also failed to detect XMRV in patients with CFS.

In an accompanying statement, the editor-in-chief of Science, Dr. Bruce Alberts, "expressed concern" about the 2009 study in Science linking the virus to CFS. The journal, he wrote, awaits the outcomes of two NIH-sponsored studies of XMRV and CFS now under way.

Last week, Dr. Dennis Mangan, who chairs a trans-NIH working group on CFS, said that these studies were expected to continue until completion.

The new reports in Science "are probably not the final word [on XMRV and these diseases], but they are definitive on where we stand at the moment," said Dr. Goff.

It is still possible, he noted, that some individuals may be infected with a virus that contributes to the diseases. Now, however, the burden of proof is on investigators to show that any virus they detect is not the result of contamination and is a contributing factor, he said.

A Cautionary Tale about Contamination

"The XMRV story is a cautionary tale about contamination," said Dr. Pathak. "And as our technology becomes ever more sensitive [for detecting the presence of mouse and viral DNA in patient samples] this will become an even bigger concern."

The XMRV story is also an interesting story of how science happens. Although Drs. Pathak and Coffin were each aware that the other was chasing down a different XMRV-like virus, they never imagined that the viruses would be the two parents of XMRV.

Good fortune played a role as well. As Dr. Le Grice noted, it is conceivable that the two labs might have discovered the same mouse virus rather than both parents, delaying their progress.

On December 17, when Dr. Coffin made the discovery, Dr. Pathak was too busy racing around to even respond to the e-mail until after he had boarded a plane. If Dr. Pathak had received the news a few days earlier, he might have postponed a 2-week vacation to a place without Internet access.

As it happened, he had a wonderful trip. And in the final moments before the flight, he managed to fire off a short reply to Dr. Coffin: "That's awesome!!"

Edward R. Winstead

Further reading: XMRV and Human Disease Association: Questions and Answers