Purpose: We aimed to develop matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) for the discovery of AMR-specific biomarkers and biosignatures. Our analyses were applied to species-level identification and characterization of AMR Enterococcus spp. collected from the gastrointestinal tracts of European starlings (Sturnus vulgaris), an invasive avian pest of concentrated animal feeding operations (CAFOs).
Methods: MALDI Biotyping was performed on 718 presumptive Enterococcus spp. isolates using the Bruker microflex LRF mass spectrometer operating with MALDI Biotyper RTC software (Version 3.4), following an ethanol/formic acid extraction of the bacteria. Antimicrobial susceptibilities of these isolates to 13 different antibiotics were then determined via disk diffusion assays. A machine learning approach of data interpretation was then applied to correlate the isolates’ mass spectral signatures with antimicrobial susceptibility fingerprints.
Results: MALDI Biotyping confidently identified (Biotyping score ≥ 1.90) 658 of the 718 isolates as Enterococcus spp. (91.6%), consisting of E. casseliflavus, E. durans, E. faecalis, E. faecium, E. gallinarum, E. hirae, E. mundtii, and E. villorum. Identification of 14 isolates was not possible (1.9%), and the remaining isolates were of the genera Aerococcus, Staphylococcus, and Vagococcus. Importantly, the application of the machine learning strategy enabled the identification of multiple ions and ion pairs that were predictive of specific antimicrobial susceptibility phenotypes.
Significance: This study demonstrates that MALDI-TOF MS is a feasible strategy for the identification of wildlife-associated Enterococcus spp. and for discriminating between AMR phenotypes.