Antibiotic resistance detection and concomitant species identification of ESKAPE pathogens by proteomics

Antimicrobial resistance (AMR) is an increasing challenge for therapy of bacterial infections. Currently, patient treatment is guided by antimicrobial susceptibility testing (AST) using phenotypic assays and species identification by MALDI-ToF biotyping. Bacterial phenotype prediction using omics technologies could offer several advantages over current diagnostic methods. It would allow species identification and AST to be combined in a single measurement, it would eliminate the need for secondary cultivation and could enable the prediction of phenotypes beyond AMR, such as virulence. In this study, the potential of proteomics for clinical microbiology was evaluated in an analysis of 126 clinical isolates covering 16 species including all ESKAPE genera and 29 of the most common AMR determinants. For this purpose, a flexible workflow was developed, which enables to report the AMR phenotype and the species of primary cultures within 2h. Proteomics provided high specificity (99.9%) and sensitivity (94.4 %) for AMR detection, while allowing species identification from very large sequence databases with high accuracy. The results show, that proteomics is well suited for phenotyping clinical bacterial isolates and has the potential to become a valuable diagnostic tool for clinical microbiology in the future.