We have determined the sequence of a cloned DNA fragment 1108 base pairs long which corresponds to the 3' end of the Moloney murine leukemia provirus. The clone was obtained as the primary product of reverse transcription and begins with the Moloney "strong stop" sequence, then extends towards the 5' end of the provirus. Our sequence: (i) proves that reverse transcriptase switches templates during minus strand synthesis; (ii) defines the limits of the 515-base-pair repeats which occupy both ends of the integrated provirus; (iii) shows that the structure of the proviral repeats has strong analogy to bacterial insertion sequences, indicating that the Moloney provirus is a transposon; (iv) identifies the putative promotor for genomic transcription within these repeats; (v) (MLV) (1) has long been the focus for biochemical and structural studies as a prototype for the mammalian RNA tumor viruses. Central to understanding the biology of retroviruses is the unambiguous assignment of genes to the RNA genome. In addition, because these viruses represent fragments of "selfish" nucleic acid (2) which can exist either free in a virion or associated with a cellular genome, the ends of the replicating molecule and its mode of replication are of particular interest.After infection, the virus-coded reverse transcriptase enzyme copies the Moloney virus single-stranded RNA genome into double-stranded DNA (3). Studies of reverse transcription in vitro indicate that DNA synthesis is initiated by the covalent elongation of the priming tRNA molecule which is bound near the 5' end of the RNA genome (4). Synthesis proceeds in a 5'-to-3' fashion, polymerizing deoxyribonucleotides complementary to the RNA genome such that the 5' end of the template is soon reached. The nascent single-stranded DNA molecule is then thought to migrate to the 3' end of the RNA template, where it may pair by virtue of complementary nucleotides with the string of approximately 60 bases immediately 5' to the poly(A) tail on that end of the genome (5). The nascent chain is then elongated, presumably continuously, to the 5' end of the template, producing a complete minus strand DNA copy of the virus genome. The mechanism for second-strand DNA synthesis and for producing a repeated end structure, thus We are studying the structure of proviral DNA cloned in bacterial vectors. Our initial attention has been drawn to the 3' end, where much of the molecular interest is. We choose to study the nucleotide sequences of cloned DNA fragments rather than that of whole viral RNA preparations because these RNA viruses have a low ratio of infectious virus to particle and the RNA is quite heterogeneous. Ultimately, the cloned DNA segments may be shown by transfection studies to harbor full biological activity, a feature not possible with RNA populations.We have investigated the molecular architecture of the product of the initial events in reverse transcription and have completely defined at the DNA sequence level the 5' end of the minus strand DNA. The sequence ex...