nov

Bacterial whole-genome sequences contain informative features of their evolutionary pathways. Comparison of whole-genome sequences have become the method of choice for classification of prokaryotes, thus allowing the identification of bacteria from an evolutionary perspective, and providing data to resolve some current controversies. Currently, controversy exists about the assignment of members of the Mycobacterium avium complex, as is for the cases of Mycobacterium yongonense and 'Mycobacterium indicus pranii'. These two mycobacteria, closely related to Mycobacterium intracellulare on the basis of standard phenotypic and single gene-sequences comparisons, were not considered a member of such species on the basis on some particular differences displayed by a single strain. Whole-genome sequence comparison procedures, namely the average nucleotide identity and the genome distance, showed that those two mycobacteria should be considered members of the species M. intracellulare. The results were confirmed with other whole-genome comparison supplementary methods. According to the data provided, Mycobacterium yongonense and 'Mycobacterium indicus pranii' should be considered and renamed and included as members of M. intracellulare. This study highlights the problems caused when a novel species is accepted on the basis of a single strain, as was the case for M. yongonense. Based mainly on whole-genome sequence analysis, we conclude that M. yongonense should be reclassified as a subspecies of Mycobacterium intracellulareas Mycobacterium intracellularesubsp. yongonense and 'Mycobacterium indicus pranii' classified in the same subspecies as the type strain of Mycobacterium intracellulare and classified as Mycobacterium intracellularesubsp. intracellulare.

Role of the DNA Mismatch Repair Gene MutS4 in Driving the Evolution of Mycobacterium yongonense Type I via Homologous Recombination

We recently showed that Mycobacterium yongonense could be divided into two genotypes: Type I, in which the rpoB gene has been transferred from Mycobacterium parascrofulaceum, and Type II, in which the rpoB gene has not been transferred. Comparative genome analysis of three M. yongonense Type I, two M. yongonense Type II and M. parascrofulaceum type strains were performed in this study to gain insight into gene transfer from M. parascrofulaceum into M. yongonense Type I strains. We found two genome regions transferred from M. parascrofulaceum: one contained 3 consecutive genes, including the rpoBC operon, and the other contained 57 consecutive genes that had been transferred into M. yongonense Type I genomes via homologous recombination. Further comparison between the M. yongonense Type I and II genomes revealed that Type I, but not Type II has a distinct DNA mismatch repair gene (MutS4 subfamily) that was possibly transferred via non-homologous recombination from other actinomycetes. We hypothesized that it could facilitate homologous recombination from the M. parascrofulaceum to the M. yongonense Type I genomes. We therefore generated recombinant Mycobacterium smegmatis containing a MutS4 operon of M. yongonense. We found that the M. tuberculosis rpoB fragment with a rifampin resistance-conferring mutation was more frequently inserted into recombinant M. smegmatis than the wild type, suggesting that MutS4 is a driving force in the gene transfer from M. parascrofulaceum to M. yongonense Type I strains via homologous recombination. In conclusion, our data indicated that MutS4 in M. yongonense Type I genomes may drive gene transfer from M. parascrofulaceum via homologous recombination, resulting in division of M. yongonense into two genotypes, Type I and II.