Whole-genome sequencing reveals genotypic resistance in phenotypically susceptible Mycobacterium tuberculosis clinical isolates

Abstract

Background: Whole-genome sequencing (WGS) data of Mycobacterium tuberculosis (MTB) complex strains have revealed insights about genetic variants associated with drug resistance (DR). Rapid genome-based diagnostics are being sought for specific and sensitive identification of DR; however, correct prediction of resistance genotypes requires both informatics tools and understanding of available evidence. We analyzed WGS datasets from phenotypically susceptible MTB strains using MTB resistance identification software. Methods: WGS data for 1526 MTB isolates classified as phenotypically drug susceptible were downloaded from the ReSeqTB database. The TB-Profiler software was used to call Single Nucleotide Variants (SNV) associated with resistance to rifampicin (RIF), isoniazid (INH), ethambutol (EMB), pyrazinamide, fluoroquinolone (FLQ), streptomycin (STR), and aminoglycosides. The SNV were further matched against the 2021 World Health Organization (WHO) catalogue of resistance mutations. Results: Genome analysis of 1526 MTB strains susceptible to first-line drugs revealed 39 SNV associated with DR to be present in across 14 genes in 5.9% (n = 90) isolates. Further interpretation of SNV based on the WHO catalog of mutations revealed resistance that 21 (1.4%) MTB isolates were resistant to first-line (4 to RIF, 14 to INH, 3 to EMB) drugs. While, 36 (2.6%) isolates were resistant to second-line (19 to STR, 14 to FLQ, and three to capreomycin) agents. The most frequent predictive SNV were; rpoB Ser450 Leu for RIF; katG Ser315Thr, inhA Ser94Ala, fabG1-15C >T (for INH); gyrA Asp94Gly for FLQ; embB Met306 Leu for EMB; rpsL Lys43Arg for STR; and tlyA Asn236 Lys for Capreomycin. Conclusions: Our study highlights the value of WGS-based sequence data for identifying resistance in MTB. It also shows how MTB strains may be misclassified simply on phenotypic drug susceptibility testing, and that correct genome interpretation is key for correct interpretation of resistance genotypes that can be used to guide clinical treatment.