[SW4] Discontinuation between genome analysis and phenotypes in Mycobacteria
Mycobacterium tuberculosis (Mtb) has no plasmid or transposon which transfer drug resistance. Then, we used to believe that whole genome analysis will reveal all drug resistant mechanisms of Mtb. However, not as planned, we’re still in the deep forest of unknown orientation. The mechanisms of drug resistances are not so simple. Mutations and indels are most common/simple causes of drug target modification and affinity changes, and rifamycin is typical drug showing high specificity to this resistant mechanism. The sensitivity and specificity of clinical drug resistance estimation with rpoB, encoding beta subunit of DNA dependent RNA polymerase as the target of rifamycin, they are over 95% in general. Even in such highly accurate mechanism, there are several exceptions. Some mutations render relatively low elevation of MIC, then the phenotypic drug susceptibility testing (DST) shows susceptible results to some extent. However, the clinical outcomes are similar to highly resistant cases. The mutations are disputed. Vise versa, in Mycobacterium abscessus, which is macrolide resistant in general due to induced methylation, shows macrolide susceptible variant in approximately 10% in Japan. The major reason is T28C sequence variant, but some isolates are identified macrolide susceptible without this sequence variation. In such case, phenotypic DST makes much sense. We may need to solve the discontinuation between genome analysis and phenotypes.