Gammaretroviruses have a broad host range, infecting organisms as diverse as mammals, birds, and reptiles (9
). Perhaps the best-studied retroviruses within the gammaretrovirus genus are murine leukemia viruses (MLVs), which have both endogenous and exogenous counterparts. Mouse genomes contain a large number of endogenous MLVs, and different Mus
subspecies are highly variable in the number of MLV insertions that they harbor.
Endogenous proviruses are carried as part of the genome of the host species and are subject to the same forces of evolution as their host, resulting in the slow accumulation of mutations over time that can eventually render them inactive. However, in some species there are intact endogenous proviruses that can be activated to produce infectious virus, through external stimuli or through recombination with other endogenous or exogenous viruses (31
XMRV (xenotropic murine leukemia virus-related virus) was first identified as a possible human pathogen in 2006 in a prostate cancer (PC) patient cohort (58
). It was later reported in a high percentage of chronic fatigue syndrome (CFS) patients, as well as some healthy subjects, although no causal link between XMRV and any human disease has ever been established (35
). In contrast, there are a large number of studies from many laboratories in which either no virus or no association with disease was found in a variety of human populations, despite the use of multiple highly sensitive detection methods (2
). Strikingly, two recent replication studies by Knox et al. and Shin et al. showed that patients who were previously reported as XMRV positive were, in fact, XMRV negative (29
). Furthermore, viral sequences detected in human samples do not show the extent of diversity that would be expected from a human retrovirus that replicates within a population, suggesting that the observed clinical isolates could be explained by laboratory contamination from a single source (23
). Finally, XMRV is highly susceptible to the antiviral effects of human cellular restriction factors, making the virus an unlikely candidate to overcome the blocks against its replication in the context of human infection (7
). Taken together, these findings challenge the notion that XMRV is a genuine human virus.
cell line was derived from a human prostate tumor (CWR22) passaged repeatedly as xenografts in nude mice (50
cells contain multiple XMRV insertions (28
) and produce the virus at very high titers (~109
RNA copies/ml; data not shown). The virus produced by these cells differs by only 1 nucleotide (nt) from the consensus XMRV sequence reported in patient samples (28
). In a recent study, analysis of genetic material from early and late xenograft samples of the tumor line that was used to generate 22Rv1
cells revealed the absence of XMRV from early samples but its presence in the late samples (39
). Moreover, the discovery of two endogenous MLV proviruses, PreXMRV-1 and PreXMRV-2, present in the mouse tissues used to passage the xenografts was reported. A heterozygous progeny virion derived from these proviruses most likely generated XMRV by a unique recombination event (39
). Here we describe multiple screening approaches that led to the identification of PreXMRV-2, the integration sites of the parental proviruses in the mouse genome, and their distribution among 94 different mouse strains.