Revertants of mouse cells transformed by murine sarcoma virus. III. Metastable expression of virus functions in revertants retransformed by murine sarcoma virus.

3T3FL cells transformed by the Moloney murine sarcoma virus (MSV) (sarcoma positive, leukemia negative, S+L- cells) segregate spontaneous flat revertants. These S-L- revertants do not release MSV upon superinfection with MuLV but are susceptible to a second cycle of single-hit MSV transformation. The behavior and nature of second-cycle MSV transformed revertants were followed by successive clonal generations. The second transformation by MSV resulted in a number of metastable states. The simplest of these was a tranformed S+L- like state, which in contrast to parental S+L- cells, was also inducible by halogenated pyrimidines. A second type consisted of a transformed S+L- clone which could give rise to a secondary flat S-L- clone, which when followed over several generations could turn again into a stable S+L- type cell. Thus an MSV genome can persist in a nontransformed cell in a nonrescuable state. A third condition has been found in which a previously S+L- type MSV retransformed revertant began to produce infectious MSV with or without an atypical murine leukemia virus (MuLV). Other clonal sublines from a second cycle MSV transformed cell resulted in both a stable S+L- type line and a spontaneous MSV releaser line. The S+L- type retransformed cell resembled parental single cycle MSV transformed S+L- cells in that it contained MuLV group specific antigen(s), and C type particles. Some second cycle MSV transformed S+L- cells produced infectious MSV on chemical induction. No second-cycle reversion was ever observed, if both the first and the second infecting genomes were Moloney-MSV. The nature of the atypical MuLV was examined in terms of host range, growth potential, and antigenicity. This MuLV was at times released with an apparent excess of MSV, yet if eleven weekly blind passages of such a virus stock were made in S+L- cells, replicating MuLV was found in titers similar to the MSV content. This atypical virus replicated in mouse cells, less in rat cells, but not at all in human or cat cells. Neutralization revealed that it was closely related to the Moloney type of MuLV. The above observations are compatible with the interpretation that the second cycle of MSV transformation destabilizes cellular controls over the transforming genome.