Species-specific proteotypic peptides for characterization of non-tuberculosis mycobacteria

Non-tuberculous mycobacteria are opportunistic bacteria that closely resemble Mycobacterium tuberculosis, causing respiratory infections in humans. While genotyping through genome sequencing is found accurate in detecting mycobacterial species but struggles with distinguishing bacterial co-infection. These challenges lead to delayed therapeutic intervention, drug resistance, and disease complications. Lately, mass spectrometry-based (MALDI-TOF-MS) proteomics has routinely been used in diagnosing mycobacterial species in clinical samples. However, it suffers accurate species detection owing to extensive bacterial cultures and poor specificity in polymicrobial infections. In contrast, due to its sensitivity, LC-MS/MS based proteomics is widely employed for accurate bacterial proteome mapping. In this study, in-depth proteomics of 9 NTM species with proteome database searches of 26 datasets was used. In total 20 million peptide spectrum matches were identified aiding to ≥40% proteome coverage in 7 NTMs with highest in M. abscessus. Further, metaproteomic analysis and rescoring of peptides resulted in high-confidence species-specificity proteotypic peptides in M. smegmatis (2342), M. vaccae (960), M. abscessus (75), M. avium subsp. paratuberculosis (3) and M. fortuitum (1). Finally, database search results were converted to spectral library format for easier future usage in targeted proteomic workflows. This workflow in deriving species-specific peptides with high confidence can be extended in distinguishing closely related bacterial species for enhancing microbial diagnostics.