If you ask me, or any of my editorial colleagues, what attracted us to this sort of work, one common answer you’d hear is that it exposes one to a very broad range of scientific activity. For at least some of us, this comes as a relief from the narrow view of the scientific enterprise that accompanies graduate and postdoctoral fellowships. On the other hand, sometimes you have to be careful what you wish for. You nurse a batch of papers on a host of different topics through the peer review process, some of them are published, and then the next batch comes along. This is never dull, but it does make it a challenge to have at your fingertips the kind of depth of knowledge in any given area that would be desirable. I often find myself wondering what the latest is in a given line of work that was published in the journal several months or years ago.

Occasionally, then, I’d like to highlight a new paper that amplifies and puts in a new light a previous publication in Nature Genetics. The paper by Anatoly Urisman and colleagues, just published in PLoS Pathogens, describes data that was presented at the February meeting of the American Society of Clinical Oncology, and covered in the press at that time. Urisman et al. report a novel gammaretrovirus in 8 out of 20 prostate cancers from individuals homozygous (QQ) for a common R462Q variant in RNase L, an enzyme that degrades viral and cellular single-stranded RNAs. Only one tumor out of 66 from individuals who did not have the QQ genotype harbored such a virus. Carl Zimmer has more on this paper here.

In 2002, Carpten et al. reported in Nature Genetics that the relevant variant underlying the familial prostate cancer susceptibility locus, HPC1, was the R462Q variant of RNASEL. Microdissected tumors with a germline mutation in RNASEL showed loss of heterozygosity, and had no detectable RNase L protein. Heterozygotes had lower RNase L activity than individuals with two copies of the wild-type allele. A subsequent study published in the journal by Casey et al. found at least one copy of the R462Q variant in 59% of the men studied, and implicated RNASEL in up to 13% of all prostate cancers. While this association has received some support, inevitably there have been failures to replicate the result.

Adding to concerns over RNASEL as a prostate cancer susceptibility gene was the nature of its function. If it has antiviral activity, then where was the relevant virus? Urisman et al. used the so-called ‘Virochip’, which is an array bearing the most conserved sequences of approximately 950 fully sequenced viral genomes, to search for viral sequences in prostate tumors. The virus they identified and fully sequenced, XMRV, probably provides the first documented case of human infection with a xenotropic retrovirus.

Urisman et al. raise a number of key issues in their paper. First, and most important, they emphasize that XMRV may in fact have no role in the etiology of prostate cancer. Second, if it does turn out to be important, it will be of great interest to determine if the R462Q variant affects acquisition of infection, or subsequent clearance. Third, XMRV infection seems to be targeted to stromal cells, implying an indirect (paracrine) effect (if any) on tumor growth. Finally, if XMRV is causally associated with prostate cancer, then a non-uniform prevalence of infection in different populations could explain the variability in the RNASEL genetic association studies. Regardless, the identification of XMRV is an exciting discovery, and should refocus efforts to show conclusively that the RNASEL R462Q variant is a risk factor for prostate cancer.