Development of novel mass spectrometric methods for the characterisation and identification of microorganisms

Abstract

This study describes the development of a novel, REIMS-based characterisation and identification tool for unicellular organisms. A bipolar handheld sampling probe was optimised and characterised on several instrumental platforms before being applied to the acquisition of a large scale spectral database of bacteria and yeasts. Using this database, the specificity of the method was characterised using multivariate statistics and found to yield comparable identification results as MALDI-TOF-MS based techniques for a set of 28 clinically relevant bacterial species. Further tests were performed on sub-species level such as ribotype in case of C. difficile, serotype in case of S. pneumoniae, antibiotic resistance in K. pneumoniae and strain-level differentiation for E. coli. The method could be further applied to yeasts and yielded excellent results for a set of five Candida species. To extend the range of the methodology to cell line cultures, the method was further tested for reproducibility and robustness initially using three cell lines. Subsequently, REIMS profiles were collected for the whole NCI60 cancer cell line panel and investigated for their spectral reproducibility, clustering behaviour with regards to tissue type of origin and comparison with spectra of corresponding bulk cancer tissue specimens. REIMS profiles were additionally correlated with publicly available gene and protein expression data in order to elucidate the sensitivity of this REIMS-based approach. Finally, taxon-specific bacterial biomarkers were derived from a dataset containing 228 bacterial species by finding spectral features that show specificity for a certain phylogenetic group of bacteria. Using these markers, bacteria were detected in tissue sections of both cancerous and healthy colorectal tissue previously acquired using DESI-MSI. Findings were in good agreement with data obtained using 16S rRNA gene sequencing-based analysis and relevant literature.