By Ben Butkus
In recent months, several independent research groups have found that the continued discovery in human genetic samples of sequences from murine leukemia viruses and related infectious agents is most likely a result of contaminated commercial PCR reagents, and not, as originally thought, a case of the viruses infecting humans.
Now, a group from the US Centers for Disease Control and Prevention has published evidence showing that commercial RT-PCR reagents and even human DNA preparations from several life science vendors contain MLV or mouse DNA, suggesting that the extent of this contamination may be much larger than anyone had previously thought.
As such, the scientists believe that their findings underscore the need for laboratories to carefully pre-screen commercial molecular diagnostic reagents and nucleic acid specimens to ensure they are not contaminated with foreign DNA and to avoid potential false-positive PCR results when testing human clinical specimens.
"If people are testing humans for these kinds of viruses, or any viruses really, they need to pre-screen their reagents first and make sure that they don't have some kind of external source of contamination," William Switzer, a CDC scientists and lead author of the recent study, told PCR Insider this week.
Since 2006, when scientists first discovered a new gammaretrovirus called xenotropic MLV-related virus, or XMRV, and reported the presence of XMRV sequences in samples from patients with chronic fatigue syndrome or prostate cancer, the scientific community has systematically debunked the link by showing that the appearance of the MLV-like sequences was likely due to commercial reverse transcriptases contaminated with the mouse virus DNA (PCR Insider, 12/22/11).
A few of these contrary reports even demonstrated contamination in specific life science research tools such as Platinum Taq from Life Technology's Invitrogen business and nucleic acid extraction columns from Qiagen.
In the meantime, Switzer and colleagues in the laboratory branch of the CDC's HIV/AIDS prevention division — who were part of the Blood XMRV Scientific Research Working Group organized in December 2009 to scrutinize the purported link between XMRV and human disease — frequently detected low levels of MLV and XMRV-like protease sequences in control samples or negative blood donor plasma using a qRT-PCR assay employing the AgPath One Step RT-PCR kit, also sold by Life Technologies.
Suspecting contamination of the RT-PCR kit with trace amounts of residual MLV-like plasmid sequences, the group decided to further test different AgPath kits from different lots, as well as water-only control samples, for the presence of MLV/XMRV sequences. As detailed in their research paper, published last month in PLoS One, they found low levels of the mouse virus sequences in seven of 16 replicates of a new, previously unopened AgPath kit, but found no sequences in the water-only samples.
To further investigate the breadth of this contamination, Switzer and colleagues then tested several other commercial RT enzymes and kits from multiple manufacturers using MLV and XMRV RT-PCR tests developed in their CDC laboratory.
As with the AgPath kit, the researchers discovered low levels of either protease or polymerase sequences from MLV/XMRV in five other commercial kits: the TaqMan 1-step master mix and TaqMan RNA-to-Ct kits from Life Tech's Applied Biosystems business; the Brilliant II QRT-PCR master mix from Agilent; the Transcriptor 1-step RT-PCR kit from Roche; and the Robust I RT-PCR kit from Thermo-Fisher's Finnzymes business.
Even more surprising, according to Switzer, was the fact the reverse transcriptases found in some of the contaminated kits were known to be derived from a completely unrelated virus, avian myeloblastosis virus.
"We found contamination in … enzymes that were prepared from MLV expression vectors, which was not that surprising," Switzer said. "But what was surprising was when we decided to check out avian myeloblastosis virus-derived recombinant RTs, and found the same level of contamination of MLV sequences in some of those. Maybe these companies produce both AMV- and MLV-derived RTs, and do it in large bioreactors … or there are sources of contamination in their facilities, where they grow it in extremely large quantities. It's very hard to get rid of this contamination."
Further, the CDC researchers discovered that several different commercially available human DNAs, sold by Sigma-Aldrich and Biochain and used for negative controls in their PCR assays, were contaminated with mouse DNA — "some with very high levels of mouse DNA," said Switzer.
"We contacted [Biochain] and told them about it, and they independently confirmed those results, and went back to their production lines and changed some things to try and reduce and eliminate the contamination," Switzer said. "We found that they did a good job of that, but there was still trace amounts of contamination in the new DNA preparations that they sent us."
Since publishing their PLoS One paper, the CDC researchers have also discovered commercial reverse transcriptases expressing plasmids that were generated from HIV sequences, Switzer said, raising the specter that this type of unexpected contamination may even go beyond MLV/XMRV and has the potential to cause false positives in HIV studies — an idea that the group is currently examining more closely.
"It's been an eye opener," Switzer said. "When we first read some of these papers [linking XMRV with CFS or prostate cancer] we were thinking either that this is real, or it's contamination on a scale that's never been reported in science before. And the latter scenario seems to be what has happened."
Switzer further hypothesized that "mice are used in a variety of experiments around the world in many laboratories. People work with mouse and human cell lines in the same facilities and laboratories. Apparently there has been so much research done on MLVs, and people have used them to develop other tools like RT-expressing plasmids, that … it's just everywhere."
If this is the case — and evidence continues to mount suggesting that it is — the researchers believe that all laboratories using qRT-PCR to detect the presence of sequences from MLV/XMRV or possibly any virus in human samples may need to rigorously screen all diagnostic reagents and specimens for potential contamination.
"If you get these kinds of unexpected findings like people originally reported … you really have to go back and do all the necessary quality control experiments to make sure you're absolutely right before you stick your neck out there and publish these kinds of results that get refuted, and then end up retracting those papers," Switzer said.
Have topics you'd like to see covered in PCR Insider? Contact the editor at bbutkus [at] genomeweb [.] com.