An X-linked gene affecting mouse cell DNA synthesis also affects production of unintegrated linear and supercoiled...

To identify specific cellular factors which could be required during the synthesis of retroviral DNA, we have studied the replication of murine leukemia virus in mouse cells temperature sensitive for cell DNA synthesis (M. L. Slater and H. L. Ozer, Cell 7:289-295, 1976) and in several of their revertants. This mutation has previously been mapped on the X chromosome. We found that a short incubation of mutant cells at a nonpermissive temperature (39 degrees C) during the early part of the virus cycle (between 0- to 20-h postinfection) greatly inhibited virus production. This effect was not observed in revertant or wild-type cells. Molecular studies by the Southern transfer procedure of the unintegrated viral DNA synthesized in these cells at a permissive (33 degrees C) or nonpermissive temperature revealed that the levels of linear double-stranded viral DNA (8.8 kilobase pairs) were nearly identical in mutant or revertant cells incubated at 33 or 39 degrees C. However, the levels of two species of supercoiled viral DNA (with one or two long terminal repeats) were significantly lower in mutant cells incubated at 39 degrees C than in mutant cells incubated at 33 degrees C or in revertant cells incubated at 39 degrees C. Pulse-chase experiments showed that linear viral DNA made at 39 degrees C could not be converted into supercoiled viral DNA in mutant cells after a shift down to 33 degrees C. In contrast, such conversion was observed in revertant cells. Restriction endonuclease analysis did not detect differences in the structure of linear viral DNA made at 39 degrees C in mutant cells as compared to linear viral DNA isolated from the same cells at 33 degrees C. However, linear viral DNA made at 39 degrees C in mutant cells was poorly infectious in transfection assays. Taken together, these results strongly suggest that this X-linked gene, affecting mouse cell DNA synthesis, is operating in the early phase of murine leukemia virus replication. It seems to affect the level of production of unintegrated linear viral DNA only slightly while greatly reducing the infectivity of these molecules. In contrast, the accumulation of supercoiled viral DNA and subsequent progeny virus production are greatly reduced. Our pulse-chase experiments suggest that the apparent, but not yet identified, defect in linear viral DNA molecules might be responsible for their subsequent impaired circularization.