Abstract
Clarithromycin resistance in Mycobacterium abscessus subsp. abscessus, massiliense, and bolletii occurs through induction of erm(41) or mutations in rrl (23S rRNA) genes. Phenotypic detection of clarithromycin resistance is hindered by the need for extended incubation as well as co-occurrence of mixed populations of M. abscessus with different susceptibility profiles. We developed a quantitative EvaGreen-based droplet digital PCR (ddPCR) scheme for rapid detection of full-length or truncated erm(41) and a probe based ddPCR screening assay for assessment of 23S rRNA rrl mutational resistance. We tested 100 M. abscessus strains, synthetic mixes with different susceptibility profiles, and 13 positive MGIT samples. Truncated and full-length erm(41) genes were detected in 27/100 and 73/100 strains and 4/13 and 9/13 MGIT samples, respectively yielding a sensitivity and specificity of 100%. Clarithromycin resistance mutations in rrl were detected in 26/100 isolates, i.e., A2058G (18/100), A2058C (7/100), and A2059G (1/100), and in 3/13 MGIT samples, i.e., A2058G (2/13) and A2059G (1/13). A screening assay of rrl ddPCR (A2058A/A2058G probes) showed 100% sensitivity in detecting the wild type or A2058G mutation as well as identifying samples requiring further testing. Upon inclusion of additional ddPCR assays, we were able to detect A2058C and A2059G clarithromycin resistance-conferring mutations in the rrl gene. Our ddPCR scheme can differentiate between full-length and truncated erm(41) and identify clarithromycin resistance-conferring mutations in the rrl gene from clinical isolates and positive MGIT samples as well as deconvolute and quantitate mixed populations of M. abscessus with different clarithromycin resistance traits.