Factors determining frequency of plasmid cointegration mediated by insertion sequence IS1.
We demonstrate that mutants with deletions at either end of the insertion sequence IS1 lose the ability to mediate cointegration of two plasmids, whereas mutants with deletions or an insertion within IS1 can mediate cointegration at a reduced frequency. These results, together with the nucleotide sequence analysis of the IS1 mutants, indicate that the two ends of IS1 (insL and insR) and two genes (insA and insB) that are encoded by IS1 are required for cointegration. Using a plasmid carrying two copies of IS1, we found that the individual IS1s mediate cointegration at different characteristic frequencies, and that each of two parts of plasmid DNA segments flanked by the two IS1s is a transposon, mediating plasmid cointegration at a unique frequency. When one IS1 was replaced with a mutant IS1, the remaining wild-type IS1 complemented the cointegration ability of the mutant IS1 as well as a resulting mutant transposon that was then flanked by a wild-type IS1 and a mutant IS1. The efficiency of this complementation reflected the characteristic ability of an individual IS1 present on the plasmid to promote cointegration. The results suggest that the IS1-encoded proteins are produced in different amounts, depending on the location of IS1 in the plasmid, and that these amounts determine the efficiency of complementation of the cointegration ability of a mutant IS1 as well as a mutant transposon. However, the location of an individual IS1 itself can also determine the frequency of cointegration in the presence of a given amount of the IS1 proteins. On the basis of the observation that the cointegration ability of a mutant IS1 is less efficiently complemented than is the ability of a mutant transposon, we also suggest that the IS1-encoded proteins can function in trans, but act preferentially on the IS1 or transposon sequence from which they are produced in promoting cointegration.