Heterogeneity of "virus-like" genes encoding retrovirus-associated 30S RNA and their organization within the mouse genome

The mouse genome harbors a set of genes reiterated 20-50 times that normally expresses 30S RNA transcripts. These transcripts can be rescued specifically into virions of helper-independent C-type viruses as pseudotypes and are thought to be defective endogenous retroviruses. We cloned DNA sequences coding "virus-like" 30S RNA from a gene library of the inbred mouse strain BALB/c in bacteriophage lambda. We identified the sequences by their hybridization with an RNA abundant in preparation of MuLV virions produced by mouse cells but sharing no sequence homology with the MuLV genome, and by the ability of the cloned DNAs to hybridize specifically with the poly(A)-containing 30S RNA present in the cytoplasm of uninfected mouse cells. The 30S RNAs are encoded by uninterrupted 5 kb DNA sequences in the genome. Individual copies of virus-like genes were distributed at different loci and were flanked by cellular DNA sequences that shared no apparent sequence homology. Each of the virus-like genes is related but not identical to all other copies. Two levels of genetic heterogeneity were detected: the first results from relatively large, continuous sequences of nonhomology up to 300 bp long; the second results from a pattern of scattered base substitutions suggesting that some DNA sequences might have been conserved more than others.