Universal antimicrobial resistance detection from clinical bacterial isolates using proteomics

Antibiotic resistance in pathogens is an increasing challenge for therapy and management of infections. Detection of antimicrobial resistances are often done by phenotypic assays which take time, due to secondary cultivation. In addition the detection of the bacterial species is indispensable for a suitable treatment. Currently no MS-based method is available for routine diagnostics of patient samples, which addresses both needs simultaneously. New systems are tested on the bases of MALDI-ToF mass spectrometry potentially solving that need. But those systems still require secondary cultivation with antibiotics and lacking further analysis depth. Here we present a universal workflow for clinical LC-MS/MS to detect species and antimicrobial resistance related proteins in the absence of secondary antibiotic cultivation in fewer than 4 h from an already grown primary culture. Using DIA MS-data and neural network analysis tools, the entity of every sample is analyzed, even for future studies. We have developed an universal workflow for rapid, sensitive and accurate detection, covering 7 bacterial species and 11 AMR determinants represented by 13 protein isoforms. This prove of concept LC-MS/MS approach could be the next game changer in the microbial clinical diagnostic.