5'-terminal nucleotide noncoding sequences of retroviruses: relatedness of two old world primate type C viruses and avian spleen necrosis virus.

Computer-assisted comparison of the 5'-terminal regions of mammalian type C viruses serves as a useful model of evolutionary divergence of noncoding nucleic acid sequences. It has led to the concept that regions of conserved nucleic acid sequences, the slowly divergent sequences, contain signals of translational, transcriptional, or integrative significance. Interspersed among the conserved regions are rapidly divergent sequences in which base changes, insertions, and deletions are especially prevalent. In the present study, CPC-1, a type C virus isolated from Colobus polykomos, was shown to be related to another Old World type C monkey virus, endogenous stump-tailed monkey virus, MAC-1, by analysis of their 5'-terminal nucleotide sequences. The 5'-terminal regions of CPC-1 and MAC-1 showed a 76% nucleotide correspondence and were of similar lengths, 132 and 127 nucleotides, respectively. Previous strong-stop analyses of other type C viruses have defined two subgroups: (i) Rauscher murine leukemia virus and gibbon ape leukemia virus and (ii) baboon endogenous virus and endogenous cat virus RD114. Based on the present sequence analysis of their 5'-terminal sequences, CPC-1 and MAC-1 formed a third subgroup. Computer-assisted comparison of the 5'-terminal sequences of CPC-1 and MAC-1 to the previously reported sequence of avian spleen necrosis virus (SNV) (Shimotohno et al., Nature [London] 285:550-554, 1980) showed SNV to be a member of that subgroup of mammalian type C viruses. Consistent with the inclusion of SNV in this subgroup of mammalian type C viruses, SNV was distantly related to other mammalian type C viruses. Interestingly, the SNV 5'-terminal sequences showed no significant evolutionary relationship by these criteria to the avian leukemia and sarcoma viruses. CPC-1, MAC-1, and SNV contained conserved regulatory signals in similar positions in their 5'-terminal RNA sequences analogous to those observed in other mammalian type C retroviruses. These sequences included the canonical AAUAAA sequence, a palindrome, a putative ribosome binding site, and an integration site. Some of these highly conserved subsequences were common to 3'- and 5'-terminal noncoding sequences of nonviral eucaryotic mRNA's (Efstratiadis et al., Cell 21:653-668, 1980). Thus, analysis and comparison of 5'-terminal nucleotide sequences have been useful in defining common functional signals and in extending the matrix of relationships among retroviruses.