Efflux pump genes variants in MDR TB strains with discrepant phenotype-genotype correlations may further guide drug resistance interpretation

Abstract

Abstract Introduction: Whole genome analysis of Mycobacterium tuberculosis (MTB) is an increasingly important method of identification of multi-drug resistance (MDR) in clinical isolates. It is based on the identification of single nucleotide variants (SNVs) in genes associated with resistance. However, there remain gaps in our understanding of phenotype – genotype correlation between strains. Efflux pumps contribute to drug resistance and here we studied SNVs in key efflux pump genes (EP) to investigate their association with resistance. Methods: Whole genome data of 2221 MTB isolates comprising 1432 susceptible and 789 drug resistant strains were downloaded from ReSeqTB database. MTB lineage and resistance genotyping analysis was performed using an in-house bioinformatics pipeline, MTB-VCF. SNVs in 47 EP genes were categorized according to their SIFT/Polyphen scores. Results: We identified variants unique to EP in DR isolates. SIFT/Polyphen effect analysis determined 38 high impact SNVs across twenty EP genes (EP) to be present in these 789 genomes. SNVs were not associated with MTB lineages. The EPs with SNVs in DR isolates were Rv1819, Rv0194, Rv0507, Rv2333c, Rv3728, Rv3823, Rv1250, Rv1273, Rv1458, Rv1634, Rv1217, Rv1218, Rv0450, Rv0676c, Rv0191, Rv3008, Rv3756, Rv2688, Rv1704 and Rv1877. Examination of 52 isolates with discrepant phenotype-genotypes resistance comprising, MDR, pre-XDR and XDR strains revealed SNVs in EP associated with RIF and INH (Rv0194, Rv1217_1218, Rv1819, Rv0450, Rv1458, Rv0507), and those associated with fluoroquinolone (Rv1634 and Rv2688) resistance. Conclusions: We identified SNVs in efflux pumps which could contribute to resistance in MTB strains. It may be important to consider these as part of MTB genome-based resistance interpretation. Functional studies combined with GWAS and RNA profiling would further confirm these findings.