Rapid molecular methods for identification of pathogens and characterization of resistance mechanisms are essential to clinical infectious disease diagnostics. Our work aims to develop novel diagnostic applications of mass spectrometry for these problems in the clinical laboratory. Our approach involves the identification of optimal bacterial protein/peptide markers and the development of assays to detect these markers using LC-MS/MS, Orbitrap, QTOF, and other mass spectrometry methods. Approaches are being pursued for (1) rapid resistance protein identification in bacterial isolates (2) strain-level classification of bacterial pathogens, and (3) culture-free identification of pathogenic bacteria in primary specimens. Foundational work done during the 2016 - 2018 fiscal years with collaborators in the NIH Clinical Center included the development of a genoproteomic approach for identifying strain-specific peptide markers based on LC-MS/MS profiling of tryptic peptides, using clinical Acinetobacter baumannii isolates as a model (Wang et al, Clinical Chemistry, 2016). This approach is based on in silico whole genome translation followed by computational construction of pan- and species-specific peptidomes to guide selection of genome-specific tryptic peptides from LC-MS/MS peptide profiles. Optimal peptides from these sets are selected experimentally based on mass spectral characteristics. An extension of these methods to the design of assays for the culture-free identification of bacteria in primary specimens was developed (Wang et al, Clinical Chemistry, 2017). This proof-of-concept study demonstrated that rapid (90 minute) diagnostic identification of bacterial pathogens in primary specimens may be possible for a variety of specimen types without culture. These approaches were further refined and applied to validate an approach for the direct rapid identification of the KPC carbapenemase in clinical isolates using LC-MS/MS (Wang et al, Scientific Reports, 2017). Work completed and published in the 2019 fiscal year included the development and validation of rapid LC-MS/MS assays to detect OXA-48 (Stritch et al, JCM, 2019) and NDM (Wang et al, AAC, 2019) families of carbapenemases in clinical bacterial isolates. These proteins mediate resistance to carbapenems, a last line antibiotic class, in clinically important bacterial pathogens. Rapid detection of the presence of OXA-48 and NDM carbapenemases is critical to guide appropriate antibiotic selection to treat these infections, particularly with the introduction of new beta-lactam beta-lactamase inhibitor combination drugs that are carbapenemase-specific. Quantitative measurements of cellular NDM carbapenemase concentrations (Wang et al, AAC, 2019) revealed variation over four orders of magnitude in a set of clinical isolates, demonstrating the sensitivity of this approach to variations in resistance protein expression, with implications for precise quantification of proteins mediating antimicrobial resistance. A separate project completed and published in 2019 focused on a rapid LC-MS/MS approach to detect the recently discovered colistin resistance protein, MCR-1, in clinical isolates (Wang et al, Clinical Proteomics 2019). The mcr-1 gene has been mobilized by a composite transposon (Snesrud et al, AAC, 2016), resulting in replicative transposition and spread in a diverse array of bacterial plasmids. Mobilized colistin resistance is of significant concern and rapid methods to detect the MCR-1 protein are needed. This work demonstrated that rapid and accurate detection of the MCR-1 protein in clinical isolates can be achieved using as few as a two peptides.
