Organization of the endogenous proviruses of chickens: implications for origin and expression

Eleven of the endogenous proviruses of white leghorn chickens have been mapped with restriction endonucleases and specific nucleic acid hybridization reagents. The restriction maps of these endogenous proviruses have been compared with restriction maps of avian sarcoma virus (ASV) and Rous-associated virus O (RAV-O), an endogenous virus which is spontaneously released by cells from certain lines of chickens. Endogenous proviruses have the same basic structure as proviruses acquired by exogenous infection; the gene order is the same in the provirus as in viral RNA, and the ends of the provirus form a characteristic direct repeat which contains sequences derived from both ends of viral RNA. The endogenous proviruses can thus be described “cell DNA-3′5′-gag-pol-env-3′5′-cell DNA,” where 3′ and 5′ denote sequences homologous to the 3′ and 5′ ends of viral RNA. The endogenous proviruses of chickens are more closely related to RAV-O than to ASV, based on restriction maps and on hybridization with reagents specific for the 3′ ends of RAV-O and ASV. However, all but two of the endogenous proviruses lack at least one of the twoSstI sites in RAV-O DNA (see the preceding paper) and can thus be distinguished from RAV-O by digestion withSstI. One of the two exceptions,ev-2, is found in the DNA of line 7₂ and line 100 chickens and is genetically linked to the production of RAV-O. The only other provirus (which we call B) having both theSstI sites in RAV-O DNA was seen only once in a line 100 sample. Of the nine remaining elements, six had large deletions. Three proviruses (ev-4,ev-6, and an element we call A) are missing the left 3′5′ repeat and have sustained deletions extending varying distances intogag orpol. One of these,ev-6, is associated with the gs^?chf⁺ phenotype; the phenotype can be explained from the structure of the provirus.ev-3 has both terminal repeats intact but has sustained a deletion near thegag-pol boundary. This provirus is associated with the gs⁺chf⁺ phenotype and the structure of the provirus could account for the peculiar RNA and protein associated with this phenotype. We have also found two elements which apparently consist of sequences present only in the 3′5′ terminal repeat unit, with no other associated virus specific sequences. Such structures might arise by homologous recombination between the terminal repeats of a normal provirus.