LAUSR

The worldwide increase in antimicrobial resistance is due to antibiotic over-use in agriculture and over-prescription in medicine. For appropriate and timely patient support, faster diagnosis of antimicrobial resistance is required. Current methods for bacterial identification rely on genomics and proteomics and use comparisons with databases of known strains, but the diagnostic value of metabolites and lipids has not been explored significantly. Standard mass spectrometry/chromatography methods involve multiple dilutions during sample preparation and separation. To increase the amount of chemical information acquired and the speed of analysis of lipids, multiple reaction monitoring profiling (MRM-Profiling) has been applied. The MRM-Profiling workflow includes a discovery stage and a screening stage. The discovery stage employs precursor ion (PREC) and neutral loss (NL) scans to screen representative pooled samples for functional groups associated with particular lipid classes. The information from the first stage is organized in precursor/product ion pairs, or MRMs, and the screening stage rapidly interrogates individual samples for these MRMs. In this study, we performed MRM-Profiling of lipid extracts from four different strains of Escherichia coli cultured with amoxicillin or amoxicillin/clavulanate, a beta-lactam and beta-lactamase inhibitor, respectively. T-tests, analysis of variance (ANOVA) and receiver operating characteristic (ROC) curves were used to determine the significance of each MRM. Principal component analysis (PCA) was applied to distinguish different strains cultured under conditions that allowed or disallowed development of bacterial resistance. The results demonstrate that MRM-Profiling distinguishes the lipid profiles of resistant and non-resistant E. coli strains.