Background:  Recently, a xenotropic murine leukemia-related virus (XMRV) was identified by virus-generic microarray and polymerase chain reaction testing in 40% of prostate cancer tissues with the homozygous R462Q (QQ) variant of the antiviral RNase L enzyme. Two subsequent studies confirmed XMRV sequences in prostate cancers at a lower polymerase chain reaction prevalence (1.5% and 6%) independent of R462Q polymorphism. One of the studies also reported higher detection of XMRV proteins by immunohistochemistry in 23% of cancers. To further evaluate the prevalence of XMRV in prostate cancer and better define markers of infection, we developed new molecular and serologic assays and used them to screen tissue and plasma specimens from patients with prostate cancer.

Methods:  We developed both a xenotropic murine leukemia virus-based Western blot assay for antibody detection, and new gag, polymerase (pol), and envelope (env) polymerase chain reaction assays for sequence detection. To detect contamination with mouse DNA, we also designed a mouse-specific mitochondrial DNA (mtDNA) polymerase chain reaction test. Following assay validation, prostate tissue (n = 162) and matching plasma (n = 120) from anonymous US prostate cancer patients were tested. Phylogenetic analysis by neighbor-joining methods was used to infer evolutionary relationships. RNase L R462Q polymorphism in all patients was determined using a commercial test.

Results:   Of 165 prostate tissues, 2 (1.2%) were positive by pol and env polymerase chain reaction and had undetectable mouse mtDNA. Phylogenetic analysis showed distinct sequences that clustered with other XMRV. Plasma from both persons (5956 and 6203) were negative by reverse transcriptase-polymerase chain reaction indicating absence of viremia. Both patients were Western blot-negative. Plasma from 118 additional polymerase chain reaction-negative patients tested Western blot negative. Of the patients, 45.1% were R462Q RR homozygotes, 45.6% RQ, and 9.3 % QQ. Person 5956 was RR homozygous; patient 6203 was heterozygous (RQ). Both had intermediate-grade tumors based on Gleason scoring.

Conclusions:  Our results demonstrate a low prevalence of XMRV sequences in prostate cancer patients that is not likely associated with R462Q RNase L genotype. Infection with XMRV is confirmed by phylogenetic analysis and absence of contaminating mouse DNA. The finding of undetectable antibodies and viremia in 2 patients is noteworthy and may reflect sequestered or cleared infections. More studies are needed to better understand the prevalence and significance of XMRV in prostate cancer or other diseases.