Evaluation of recA sequences for identification of Mycobacterium species

Abstract

16S rRNA sequence data have been used to provide a molecular basis for an accurate system for identification of members of the genus Mycobacterium. Previous studies have shown that Mycobacterium species demonstrate high levels (>94%) of 16S rRNA sequence similarity and that this method cannot differentiate between all species, i.e., M. gastri and M. kansasii. In the present study, we have used the recA gene as an alternative sequencing target in order to complement 16S rRNA sequence-based genetic identification. The recA genes of 30 Mycobacterium species were amplified by PCR, sequenced, and compared with the published recA sequences of M. tuberculosis, M. smegmatis, and M. leprae available from GenBank. By recA sequencing the species showed a lower degree of interspecies similarity than they did by 16S rRNA gene sequence analysis, ranging from 96.2% between M. gastri and M. kansasii to 75.7% between M. aurum and M. leprae. Exceptions to this were members of the M. tuberculosis complex, which were identical. Two strains of each of 27 species were tested, and the intraspecies similarity ranged from 98.7 to 100%. In addition, we identified new Mycobacterium species that contain a protein intron in their recA genes, similar to M. tuberculosis and M. leprae. We propose that recA gene sequencing offers a complementary method to 16S rRNA gene sequencing for the accurate identification of the Mycobacterium species.

FIG. 1

Schematic illustration of the primer pairs and sites used in the amplification of the recA gene. The shaded segments correspond to primers applied for seminested PCR. Primer recG1 was used to determine the true sequence of the gap left between primers recR1 and recF4 when seminested PCR was performed.

FIG. 2

Phylogenetic relationships of 31 Mycobacterium species derived from the sequence of the recA gene (excluding the intein regions). The tree was generated from the alignment obtained with the Megalign tool of Lasergene (DNAstar Inc.) with the Clustal method algorithm.