In vitro strand transfer from broken RNAs results in mismatch but not frameshift mutations

An in vitro system to compare the fidelity of strand transfers from truncated vs full-length RNAs was constructed. A donor RNA, on which reverse transcriptase (RT)-directed DNA synthesis was initiated, shared homology with an acceptor RNA, to which DNAs initiated on the donor could transfer. All RNAs were derived from the N-terminal portion of the alpha-lac gene. On full-length donors, transfers occurred when DNAs migrated to the acceptor prior to being completed on the donor. On donors that were truncated, most transfers occurred after DNAs reached the end of the donor. Transfer products were amplified by PCR and used to replace the corresponding region in a vector containing the alpha-lac gene. Transformed Escherichia coli were screened for alpha-complementation by blue-white phenotype analysis, with white colonies scored as those with errors in alpha-lac. These errors were derived from RT synthesis and strand transfer. The mutant colony frequency approximately doubled for transfer products derived from truncated donors (0.026+/-0.005 vs. 0.053+/-0.011 (three experiments +/- SD), for full-length vs. truncated, respectively). The increases resulted from additional non-template-directed bases (mostly thymidines) added to the DNAs before transfer. Sequence analysis of DNAs synthesized on truncated donors showed that about 60% had additions (20/34); however, those without additions transferred at a much higher rate than those with. Transfer of the DNAs with additions always resulted in substitutions; no frameshifts were observed. Results are consistent with RT adding nontemplated nucleotides at template termini. Transfer and subsequent extension of these products is severely inhibited relative to products without additions. The potential relevance of these findings to retrovirus replication is discussed.