to BioTechniques free email alert service to receive content updates.
Walking on eggshells: PCR assays for XMRV detection

Andrew S. Wiecek

Are positive XMRV studies the result of contamination? Or is PCR just not sensitive enough to detect XMRV? Andrew Wiecek investigates the controversy surrounding PCR detection of XMRV.

Bookmark and Share

In September, over 200 virologists gathered at the NIH campus in Bethesda, MD for an international workshop to hash out the issues surrounding the detection of xenotropic murine leukemia virus–related retrovirus (XMRV), a virus that may be associated to several human diseases. Several attendees presented evidence associating the gamma retrovirus with prostate cancer and chronic fatigue syndrome (CFS). Others presented studies that failed to detect XMRV in patient samples. Still others argued that positive XMRV results were the result of contamination.

In this section from a human prostate cancer, XMRV proteins are expressed in infected cancer cells. Cells showing brown, granular staining are malignant prostate cells that express viral proteins. Source: Ila R. Singh, M.D., Ph.D., the University of Utah

“There was no explanation as to why the data was so conflicting, why some groups can detect the virus, but others using the same PCR approach do not,” says Norbert Bannert, the head of the Center for Retrovirology at the Robert Koch Institute in Berlin, Germany. “A real solution was not found as to why the results were so different at the workshop.”

Now, the dirty work of rectifying these conflicting reports continues. A working group has been assembled to standardize protocols for PCR, contamination screening, and sample preparation. Some believe that the group’s results will show how easy samples can be contaminated. Others believe the results will highlight PCR’s detection limits, demonstrating that other detection methods are required. In the end, everyone is just hoping to understand two questions: whether the positive XMRV PCR results were the result of contamination, and whether PCR is sensitive enough to detect XMRV in patient blood samples

Highly sensitive issue

Judy Mikovits, a researcher at the Whittemore Peterson Institute for Neuro-Immune Disease (WPI) in Reno, NV, believes that XMRV may explain why a number of cancers and neurological diseases—not only prostate cancer and CFS, but also lymphoma, breast cancer, and prostate cancer—have been on the rise in young populations over recent decades. “It’s scary how many young people under 30 get cancer,” says Mikovits. “It could be environmental, and, of course, the virus would be an environmental agent.”

XMRV was first identified in 2006 in prostate cancer patients by Anatoly Urisman and colleagues at the University of California, San Francisco; the Cleveland Clinic; and Cleveland State University (1). Using real-time PCR, the group detected the virus in 8 out of 20 prostate cancer patient samples tested. The virus is closely related to murine leukemia viruses (MLVs), which can lead to disease in mice. In 2009, Mikovits’ group published a paper in Science that first identified XMRV in CFS patient samples. Using a variety of techniques, including a nested PCR of culture samples, her team found XMRV in 67% of CFS patient samples, compared with about 4% of health controls (2). The presence of XMRV in prostate cancer and CFS patients was corroborated by several other studies(3-6).

For millions of patients suffering from prostate cancer or CFS, these studies provided a hopeful starting point: if they tested positive for XMRV, then perhaps treatments targeting the retrovirus would afford results. WPI licensed their XMRV detection method to Viral Immune Pathways Diagnostic Laboratories (Reno, NV) to offer CFS patients a XMRV diagnostic test. Thousands of patients paid hundreds of dollars to have their samples tested for XMRV.

But then came the negative studies. Several independent research teams from around the world—including Europe and the United States—failed to detect XMRV using PCR in prostate cancer and CFS patient samples (7-18). Bannert’s group at the Robert Koch-Institute in Berlin, Germany was one of these teams. “We did two studies that looked into prostate cancer samples from Germany, mostly in the Berlin area, close to 600 samples. We did [nested] PCR and did not find any positives,” he says. Recently, Bannert’s team published a report on their study of XMRV in CFS patients from Germany. Again, all the samples were negative.

No sooner had an XMRV-related answer seem to materialize than it began to disappear. Why were some labs detecting XMRV in prostate cancer and CFS patients while others were not? Research shifted focus to evaluating the protocols, the possibility of contamination, and the sample preparation of each study.

According to the WPI group, nested PCR of cultured samples provided the best results for XMRV detection. Capable of amplifying single molecules within a large sample, it is used when targeted DNA templates are in extremely low concentrations. So-called for the way it “nests” two runs using two sets of primers, the product of the first PCR run becomes a template for the second. The technique’s only limitation is the amount of template DNA in the sample.

To increase the sensitivity of their nested PCR protocol, Mikovits turned to Max Pfost, a graduate student in her lab. Pfost researched PCR optimization and began modifying the protocol’s magnesium concentrations and annealing conditions, and choose primers for increased sensitivity rather than specificity. Mikovits allegorizes from what she calls “the HIV days,” when it was first discovered that multiple low–copy number HIV strains existed. “If you are looking for a low–copy number family member, you really have to optimize the magnesium and base everything on the annealing temperatures." Otherwise, she says, researchers using nested PCR would very likely overlook XMRV in their samples.

“A lot of people are just not taking the time to optimize their PCR,” says Pfost.

With great sensitivity comes great contamination risk

Using a variety of techniques, including a nested PCR of culture samples, Judy Mikovits' team at the WPI found XMRV in 67% of CFS patient samples, compared with about 4% of health controls. Source: University of Florida Health Science Center

Other retroviruses—such as HIV and HTLV-1—can be detected by less-sensitive PCR assays. “To detect a gamma retrovirus in animals, you usually don’t need such a high sensitivity,” says Bannert. “Of course, as high as the sensitivity is, the easier it is to catch contaminants. That’s for sure.” And that’s exactly what researchers at the NIH workshop—followed by four research papers published in the journal Retrovirology in December 2010—presented evidence of in nested PCR protocols for XMRV detection (19-22).

Indeed, contamination may be behind XMRV’s capriciousness. Because of its close relationship to MLVs, XMRV’s false-positives may be the result of mouse genomic contamination, some say. The mouse genome contains two sets of mouse-specific genes, but it has a higher frequency of integrated retroviral genomic parts, at more than a thousand copies per genome. Standard protocols to screen for mouse genetic contamination target the mouse-specific genes, but if the contamination is not the complete mouse genome—maybe just a small part of it, a few molecules—the higher-frequency sections would be more likely to contaminate a sample.

To control mouse genomic contamination, the WPI lab checked their PCR results against mouse mitochondrial DNA, using a probe developed by Bill Switzer from the CDC. They never had a positive mitochondrial DNA sample. They used separate rooms, separate buildings, separate hoods for the two experimental rounds of their study. They cross-linked amplicons, tubes, and water supplies. They ran negative plasmas in every culture to prove that no steps were contaminated. “We take every possible precaution for PCR contamination,” says Mikovits.

Greg Towers, a researcher at the center for Medical Molecular Virology at the University College London, says “You may just have parts of the genome which contain these high–copy viruses, but not the gene that you are using for your control. The contamination situation is worse than normal.” Source: Greg Towers

But, says Greg Towers, a researcher at the center for Medical Molecular Virology at the University College London, “you may just have parts of the genome which contain these high–copy viruses, but not the gene that you are using for your control. The contamination situation is worse than normal.” Brigitte T. Huber, a professor of pathology at Tufts Sackler School of Graduate Biomedical Sciences, agrees. “We’re just not used to working at that level of sensitivity, so we are not used to looking at contaminants at such a low level,” she says.

After the Mikovits 2009 Science paper, Huber was inspired to start targeting XMRV in her CFS research. When she got positive results in her nested PCR assays, Huber got even more excited. Then she decided to screen the samples for mouse contamination, using two different mouse probes, the Switzer-CDC probe that targeted mouse mitochondrial genes and a new probe—developed by her colleague John Coffin at Tufts—that targeted the high-frequency integrated retroviral elements in the mouse genome.

In every sample that Huber’s team detected XMRV, they also detected contamination. “We started to see contamination at these extremely low levels,” says Huber.

“It was a brilliant idea,” says Towers of Coffin’s probe. “Several studies have taken their samples, screened them for mouse DNA using John Coffin’s mouse DNA assay, and all of the studies found that all the samples that are XMRV-positive are also mouse DNA–positive.”

Even though the false-positives could be the result of mouse genomic molecules contamination, no one can pinpoint the source of this contamination. The WPI lab does not work with mice or mouse cells. “We do not culture mouse cell lines,” says Mikovits. “We never have.”

But in biomedical research, mouse cells and genomic molecules are almost everywhere, according to Huber. Human cells lines have historically been passaged through mice—specifically nude mice—to increase their numbers for in vitro applications. The MLVs can easily infect human cells. Many human cell lines are suspected of being contaminated by XMRV, producing the virus at high levels. “It’s extremely easy for labs that have any of these cell lines to get contamination with an infectious virus,” says Huber. Specifically, researchers have identified XMRV contamination in the prostate cancer cell line 22Rv1, along with several other tumor cell lines.

Commercial PCR enzymes could also be a source of contamination. One of the Retrovirology papers detected mouse DNA in commonly used Taq polymerase reagents (21). “I could say that I never use mouse samples in my lab,” suggests Towers, “but I used this Taq polymerase that has a mouse antibody, so I am putting mouse DNA into my PCR reactions.” Paul Kellam, Tower’s University College London colleague, believes that this might be applicable to many other reagents in the lab as well. “Maybe a lab uses monoclonal antibodies, they’ve all been created in murine systems one way or another.”

The WPI lab, though, does not use and has not used any of the human cell lines or PCR reagents shown to be contaminated. And if contamination was the issue, it would have affected all the blinded samples at WPI, not just the CFS patient samples. “The head of our Immunology Department congratulated me on being able to not contaminate the controls and only contaminate the patient samples in literally the thousands that we’ve done now,” says Mikovits.

These questions remain unanswered. But researchers are calling for increased PCR contamination management for XMRV detection. “There are plenty of rules and regulations and lab practices on how you can get around or control PCR contamination,” says Kellam. “Perhaps this is another set of rules and thought processes that you need to do if you are dealing with murine viruses.”

PCR’s Achilles’ heel

Contamination controversy is nothing new to Mikovits. Her doctorate thesis posited that HIV latency monocytes could become active and infect T cells. Researchers argued that Mikovits contaminated the nucleus with cytoplasm or forgot about important sample preparation steps. “I got a lot of criticism from my Ph.D.,” says Mikovits. “It turned out that they were not contaminated, and monocytes could be infected by HIV.”

As for her XMRV work, Mikovits says that researchers have focused on her PCR technique because it is the easiest element to replicate. “A lot of the negative studies assume that in this age of molecular technology, there’s not a detection limit for PCR,” she says. “We could detect one copy in one milliliter of blood. But there is a limit of detection to PCR.”

When she was studying HIV latency monocytes in the 1980s, biologists did not rely upon molecular-based detection methods in virology. They relied on culturing—which, for a retrovirus, is a complicated task even for expert biologists. “The first paper on the isolation of HIV was in 1982,” says Mikovits. “We didn’t have a single-copy assay of HIV by PCR until 1991.”

It’s only been about a year since XMRV arrived on the scene, and Mikovits argues that since opposing studies focused only on plasma and did not culture the virus, there could be discrepancies. “If you simply look for proviral DNA and look for very low–copy number virus, you have to make the cells divide.”

“Some of these newer technologies are just not cutting it. You have to go back to culturing the virus,” says Pfost. ““It’s hard to find the needle in the haystack if you’re not looking in a big enough haystack.” To this end, since the 2009 Science paper, Mikovits has traveled around the world to teach other labs how to culture and detect the virus with positive results. “No one to my knowledge has cultured the virus,” she says.

But in a paper published in PLoS One in December, Bannert’s group reported their findings after doing the culturing suggested by Mikovits. The German researchers made the cells divide from CFS patient samples divide and cultured them for seven days before DNA isolation and nested PCR. The team found no evidence for the virus (17).

Also, the more researchers work with a culture, the chance of contamination also increases. “People can culture these cells for weeks and weeks, and it’s likely that during that time there was some contamination occurring,” says Huber. “We have absolutely no evidence that there’s any live virus in the blood cells of humans.”

If PCR cannot detect XMRV in plasma without culturing the virus, that leaves two possibilities: either XMRV is not present in live human blood cells, or PCR—even a single-molecule nested protocol—is not sensitive enough for XMRV detection.

PCR-based techniques are used to detect all other infectious diseases. US human blood banks rely upon FDA-approved PCR-based tests to screen donations for HIV and hepatitis because these tests provide quick, reliable results. “These PCR-based techniques are used in all other infectious diseases and used absolutely perfectly, without any controversy,” says Kellam. “You have to rule out an entire area of modern molecular biology in order to support the notion that XMRV is a transmissible human virus.”

XMRV research cannot move forward without a reliable detection method. Researchers cannot study a virus that they cannot detect. Drug developers cannot monitor the success of their treatments. While a PCR method would provide the quickest method, there are other options that could be explored as well. “What we need is a highly evaluated detection method and positive human samples, so that reference labs can really have agreed positive materials,” says Bannert.

And the XMRV Blood Scientific Research Working Group is doing just that. Headed by Graham Simmons, associate investigator at the Blood Systems Research Institute, the group is distributing blinded positive samples collected from the WPI lab to several labs across the US—including the CDC, NCI, and right back to the WPI—to develop and evaluate standards for XMRV diagnostic assays. “I’m really curious and looking forward to see the results and what will come out of these assays,” says Bannert.

But so far, the working group has only produced more mixed results. In early December, the working group announced its Phase II results. In Phase IIa, two out of three labs detected XMRV in all four samples positive; the NCI lab detected no XMRV at all. In Phase IIb, only one lab—the WPI lab—detected XMRV in any of the samples.

“We were all scratching our heads in the blood working group,” says Mikovits. “Because we’re talking about the detection limits of PCR.”


  1. Urisman, A., R.J. Molinaro, N. Fischer, S.J. Plummer, G. Casey, E.A. Klein, K. Malathi, C. Magi-Galluzzi, R.R. Tubbs, D. Ganem, R.H. Silverman, and J.L. DeRisi. 2006. Identification of Novel gammaretrovirus in prostate tumors of patients homozygous for R462Q RNASEL variant. PLoS Pathogens. 2:e25.
  2. Lombardi, V.C., F.W. Ruscetti, J.D. Gupta, M.A. Pfost, K.S. Hagen, D.L. Peterson, S.K. Ruscetti, R.K. Bagni, C. Petrow-Sadowski, B. Gold, M. Dean, R.H. Silverman and J.A. Mikovits. 2009. Detection of an Infectious Retrovirus, XMRV, in Blood Cells of Patients with Chronic Fatigue Syndrome. Science 326:585-589.
  3. Schlaberg, R., D.J. Choe, K.R. Brown, H.M. Thaker, and I.R. Singh. 2009. XMRV is present in malignant prostatic epithelium and is associated with prostate cancer, especially high-grade tumors. PNAS 106:16351-16356.
  4. Arnold, R.S., N.V. Makarova, A.O. Osunkoya, S. Suppiah, T.A. Scott, N.A. Johnson, S.M. Bhosle, D. Liotta, E. Hunter, F.F. Marshall, et al. 2010. XMRV infection in patients with prostate cancer: novel serologic assay and correlation with PCR and FISH. Urology 75:755-761.
  5. Danielson, B.P., G.E. Ayala, and J.T. Kimata. 2010. Detection of xenotropic murine leukemia virus-related virus in normal and tumor tissue of patients from the southern United States with prostate cancer is dependent on specific polymerase chain reaction conditions. J Infect Dis 202:1470-1477.
  6. Fischer, N., C. Schulz, K. Stieler, O. Hohn, C. Lange, C. Drosten, and M. Aepfelbacher. 2010. Xenotropic murine leukemia virus-related gammaretrovirus in respiratory tract. Emerg Infect Dis 16:1000-1002.
  7. Aloia, A.L., K.S. Sfanos, W.B. Isaacs, Q. Zheng, F. Maldarelli, A.M. De Marzo, A. Rein A. 2010. XMRV: A New Virus in Prostate Cancer? Cancer Res 70:1028.
  8. Barnes, E., P. Flanagan, A. Brown, N. Robinson, H. Brown, M. McClure, A. Oxenius, J. Collier, J. Weber, H.F. Gunthard, et al. 2010. Failure to detect xenotropic murine leukemia virus-related virus in blood of individuals at high risk of blood-borne viral infections. J Infect Dis 202:1482-1485.
  9. Cornelissen, M., Zorgdrager F, Blom P, Jurriaans S, Repping S, van Leeuwen E, Bakker M, Berkhout B, van der Kuyl AC: Lack of detection of XMRV in seminal plasma from HIV-1 infected men in The Netherlands. PLoS One 2010, 5:e12040.
  10. Erlwein O, Kaye S, McClure MO, Weber J, Wills G, Collier D, Wessely S, Cleare A: Failure to detect the novel retrovirus XMRV in chronic fatigue syndrome. PLoS One 2010, 5:e8519.
  11. Groom HC, Boucherit VC, Makinson K, Randal E, Baptista S, Hagan S, Gow JW, Mattes FM, Breuer J, Kerr JR, et al.: Absence of xenotropic murine leukaemia virus-related virus in UK patients with chronic fatigue syndrome. Retrovirology 2010, 7:10.
  12. Hohn, O., H. Krause, P. Barbarotto, L. Niederstadt, N. Beimforde, J. Denner, K. Miller, R. Kurth, and N. Bannert. 2009. Lack of evidence for xenotropic murine leukemia virus-related virus(XMRV) in German prostate cancer patients. Retrovirology 6:92.
  13. Hong, P., J. Li, and Y. Li. 2010. Failure to detect Xenotropic murine leukaemia virus-related virus in Chinese patients with chronic fatigue syndrome. Virol J 7:224.
  14. Jeziorski, E., V. Foulongne, C. Ludwig, D. Louhaem, G. Chiocchia, M. Segondy, M. Rodiere, M. Sitbon, and V. Courgnaud. 2010. No evidence for XMRV association in pediatric idiopathic diseases in France. Retrovirology 7:63.
  15. Switzer, W.M., H. Jia, O. Hohn, H. Zheng, S. Tang, A. Shankar, N. Bannert, G. Simmons, R.M. Hendry, V.R. Falkenberg, et al. 2010. Absence of evidence of xenotropic murine leukemia virus-related virus infection in persons with chronic fatigue syndrome and healthy controls in the United States. Retrovirology 7:57.
  16. Luczkowiak, J., O. Sierra, J.J. Gonzalez-Martin, G. Herrero-Beaumont, R. Delgado. 2011. No Xenotropic Murine Leukemia Virus-related Virus Detected in Fibromyalgia Patients. Emerg Infect Dis. 17:314-5.
  17. Hohn, O., K. Strohschein, A.U. Brandt, S. Seeher, S. Klein, R. Kurth, F. Paul, Meisel C, Scheibenbogen C, Bannert N. PLoS One. 2010 Dec 22;5(12):e15632. No evidence for XMRV in German CFS and MS patients with fatigue despite the ability of the virus to infect human blood cells in vitro.
  18. van Kuppeveldet, F.J.M., A.S. de Jong, K.H. Lanke, G.W. Verhaegh, W.J. Melchers, C.M. Swanink, G. Bleijenberg, M.G. Netea, J.M. Galama, J.W. van der Meer et al. 2010. Prevalence of xenotropic murine leukaemia virus-related virus in patients with chronic fatigue syndrome in the Netherlands: Retrospective analysis of samples from an established cohort. BMJ 340: c1018.
  19. Robinson, M.J., O.W. Erlwein, Kaye S, Weber J, Cingoz O, Patel A, Walker MM, Kim WJ, Uiprasertkul M, Coffin JM, McClure MO. 2010. Mouse DNA contamination in human tissue tested for XMRV. Retrovirology. 7:108.
  20. Oakes, B., A.K. Tai, O. Cingöz, M.H. Henefield, S. Levine, J.M. Coffin, N.T. Huber. 2010. Contamination of human DNA samples with mouse DNA can lead to false detection of XMRV-like sequences. Retrovirology. 7:109.
  21. Sato E, Furuta RA, Miyazawa T. Retrovirology. 2010 Dec 20;7:110. An endogenous murine leukemia viral genome contaminant in a commercial RT-PCR kit is amplified using standard primers for XMRV.
  22. Hué S, Gray ER, Gall A, Katzourakis A, Tan CP, Houldcroft CJ, McLaren S, Pillay D, Futreal A, Garson JA, Pybus OG, Kellam P, Towers GJ. Retrovirology. 2010 Dec 20;7(1):111. Disease-associated XMRV sequences are consistent with laboratory contamination.

Keywords:  PCR