Transformation of naturally competent Streptococcus mutans with replicative and non-replicative Tn916-containing plasmids: implications for a mechanism of transposition.

Based on the observations reported here and what is known concerning transformation of naturally competent strains of S. mutans and other streptococcal species such as S. gordonii, we propose the model shown in Figure 2. The Tn916-intermediate transforms S. mutans as originally proposed for B. subtilis by Scott and coworkers [8]. It is not clear in either system (B. subtilis or S. mutans) whether the Tn916 intermediate enters the cell as ds-DNA or ss-DNA. Because it is likely that transformation of B. subtilis via formation of protoplasts involves a mechanism quite different from natural transformation in S. mutans, it would be unwise to extrapolate findings from their studies. If Tn916 enters S. mutans in a manner similar to plasmid or chromosomal DNA, we would assume that Tn916 binds to a cell receptor and as one strand enters, the other is degraded [9]. This leaves open the question of whether Tn916 recircularizes as ds-DNA before it inserts into the chromosome or whether it remains as ss-DNA, if, indeed, it enters as ss-DNA. The transformation efficiency for the Tn916 intermediate (approximately 10(-7) precluded kinetics studies such as those performed with pAM118. Poyart-Salmeron and coworkers [11] however, described a model in which Tn1545 inserts into the target site as a ds-DNA circular molecule, similar to that seen with lambda phage. Perhaps the most interesting finding presented here is that the predominant mechanism of insertion of Tn916 into the chromosome of the recipient occurs after Tn916 enters the cell. The replicative plasmid pAM118 evidently forms by two-hit kinetics followed by intracellular excision and transposition of Tn916. The helper-rescue experiment shows that in this system, the formation of Tcr transformants, and hence the integration of Tn916, was a function of the transformation efficacy of plasmid pAM118. Since intracellular excision of Tn916 probably follows the re-formation of the plasmid pAM118, the rate-limiting step in this system would be the formation of the transient, intracellular plasmid pAM118. (The transient white colony phenotype probably denotes the slower growth rate of transformants that acquire the large replicative plasmid pAM118). Our findings demonstrate that a practical way of promoting Tn916 insertions into chromosomal DNA for the purpose of obtaining mutations is to use a helper-rescue system. Our model supports the concept that the majority of Tn916 inserts arise from a mechanism similar to 'zygotic induction' as proposed for S. sanguis (gordonii) [1]. However, the frequencies for the co-establishment of the replicative plasmid (Emr) and Tn916 inserts (Tcr) in their paper (10(-6)) differ from our observations for S. mutans. We found nearly 100% of Tcr white colonies to be Emr whereas in S. gordonii, only a fraction (approximately 1%) exhibited the TcrEmr phenotype. If both phenotypes arose independently, the frequency of the TcrEmr phenotype would be 10(-8) rather than 10(-6) survivors/recipients as observed. It was surmised that both Tcr and Emr transformants arise dependently [1] where both the Tn916 intermediate and pAM118 contribute to the formation of Tn916 inserts. We conclude from their data, however, that most Tcr arose from the Tn916-intermediate formed in the donor, in agreement with the explanation of these data by Scott [12]. Support for the contention that the Tcr arose in S. gordonii from transformation by the Tn916-intermediate as in S. mutans (yellows), and not from intracellular excision from pAM118, comes from the observation that so few Tcr are Ems and because pAM150 (rep-) yields Tcr at the same frequency as the rep+ pAM118 in their experiments. In summary, the Tn916 intermediate is capable of transforming S. mutans. In contrast to the hypothesis of Scott [12], however, the Tn916 intermediate is not the only form involved in the transformat++t