Identification of nontuberculous mycobacteria using multilocous sequence analysis of 16S rRNA, hsp65, and rpoB

BackgroundThe isolation of nontuberculous mycobacteria (NTM) from clinical specimens has increased, and they now are considered significant opportunistic pathogens. The aims of this study were to develop a database and interpretive criteria for identifying individual species. In addition, using clinical isolates, we evaluated the clinical usefulness of 16S rRNA, hsp65, and rpoB as target genes for this method. MethodsThe sequences of NTM for 16S rRNA, hsp65, and rpoB were collected from GenBank and checked by manual inspection. Clinical isolates collected between 2005 and 2010 were used for DNA extraction, polymerase chain reaction, and sequencing of these three genes. We constructed a database for the genes and evaluated the clinical utility of multilocus sequence analysis (MLSA) using 109 clinical isolates. ResultsA total 131, 130, and 122 sequences were collected from GenBank for 16S rRNA, hsp65, and rpoB, respectively. The percent similarities of the three genes ranged from 96.57% to 100% for the 16S rRNA gene, 89.27% to 100% for hsp65, and 92.71% to 100% for rpoB. When we compared the sequences of 109 clinical strains with those of the database, the rates of species-level identification were 71.3%, 86.79%, and 81.55% with 16S rRNA, hsp65, and rpoB, respectively. We could identify 97.25% of the isolates to the species level when we used MLSA. ConclusionThere were significant differences among the utilities of the three genes for species identification. The MLSA technique would be helpful for identification of NTM.