Molecular genetics of herpes simplex virus: the terminal a sequences of the L and S components are obligatorily identical and constitute a part of a structural gene mapping predominantly in the S component.

In herpes simplex virus 1 (HSV-1) DNA, a small sequence, designated the a sequence, flanks the reiterated sequences at the ends of both the L and S components. The a sequence is the only sequence shared by the termini of all isomeric arrangements of HSV-1 DNA that arise from inversions of the covalently linked L and S components. We report that the a sequence, although present in both components, is a part of a structural gene mapping predominantly in the reiterated sequences of the S component. This conclusion is based on the observations that the mutant HSV-1(13)tsC75 is rescued by transfection of cells with the mutant DNA and any one of the four terminal or four L-S junction fragments of wild-type DNA. Furthermore, in doubly infected cells, this mutant shows little or no recombination or complementation with other ts mutants previously mapped within the reiterated sequences of the S component. Because it is otherwise difficult to explain the isolation of a mutant with several independent, equivalent mutations, the data argue for a mechanism that maintains the identity of the multiple copies of the a sequence.The paradox arising from the two observations that all termini rescue the ts mutant but that in coinfection tests the ts lesion is closely linked to the reiterated sequences of the S component could be accounted for by postulating that either recombination occurs while the DNA is in a circular form-in which case all a sequences would be adjacent to the reiterated sequence of the S component-or recombination can occur while the DNA is in a linear form. In this case the only effective substitution of the a sequence that is perpetuated is the one occurring at the L-S junction or in the S component. In light of the observations that tsC75 and the other mutants tested in this study map in the reiterated sequences and fail to yield appreciable recombinational frequencies, it is unlikely that isomerization of the DNA occurs by intramolecular recombination between reiterated sequences.