Christopher Cox (Presenter)
Colorado School of Mines
Bio: I received my B.S. in Microbiology and Biochemistry from the University of Oklahoma in 1999. I then studied under Dr. Michael Gilmore, first at the University of Oklahoma Health Sciences Center where I received my Master’s in Microbiology, then at Harvard University where I conducted my dissertation research and completed my Ph.D. in 2006. Following the completion of my Ph.D., I came to the Colorado School of Mines in a post-doctoral capacity to expand my expertise into bacterial and viral biochemistry, and investigate the use of phages and various forms of mass spectrometry and analytical chemistry for molecular biomarker measurement and bacterial identification. As an Assistant Research Professor at CSM, since 2008, my main research interests at have been the design, development, and evaluation of novel diagnostic technologies for bacterial ID and antibiotic resistance determination.
Authorship: Christopher R. Cox (1), Nicholas R. Saichek (1), Nicholas R. Stambach (1), Peter B. Harrington (2) and Kent J. Voorhees (1)
(1) Colorado School of Mines, Department of Chemistry, Golden, CO. USA. (2) Ohio University, Department of Chemistry and Biochemistry, Athens, OH. USA.
MALDI-TOF MS has emerged as a rapid approach for clinical bacterial diagnostics. However, current protein-based methods do not address the increasing demand for antibiotic resistance profiling. As a result, additional testing is required, which adds significant time and expense and further delays treatment. New approaches that combine ID and resistance profiling into a single test would drastically reduce turnaround and improve treatment of resistant infections. We investigated metal oxide laser ionization (MOLI MS) of fatty acids for simultaneous ID and resistance profiling. The energy inherent to the MALDI laser allowed for in situ metal oxide-catalyzed lipid fragmentation into taxonomically useful fatty acids. This resulted in simultaneous strain-level ID and accurate differentiation of a diverse collection of methicillin resistant and susceptible Staphylococcus aureus (MRSA/MSSA).
MALDI-TOF MS has emerged as a rapid approach for clinical bacterial diagnostics. However, current protein-based methods do not address the increasing demand for antibiotic resistance profiling. As a result, additional culture-based testing is typically required, which adds significant time and expense and further delays patient treatment. New approaches that combine ID and resistance profiling into a single test would drastically reduce turnaround and improve treatment of drug resistant infections. We investigated metal oxide laser ionization (MOLI MS) of fatty acids as a means of simultaneous ID and antibiotic resistance profiling. The energy inherent to the MALDI laser was used to drive in situ metal oxide-catalyzed lipid fragmentation into taxonomically viable fatty acids using conventional MALDI instruments already widely in clinical use.1 With a CeO2 catalyst in place of a traditional matrix, we achieved strain-level ID and where able to rapidly and accurately differentiate antibiotic resistance among a diverse collection of methicillin resistant and susceptible Staphylococcus aureus (MRSA/MSSA).2
Lipids from five replicates each of nine MRSA and nine MSSA (90 total samples) were extracted and fatty acid profiles obtained by CeO2-catalyzed metal oxide laser ionization (MOLI) MS. Differentiation of resistant strains was achieved by assembly of a database of spectral profiles followed by principal component analysis and K-Nearest Neighbor pattern recognition. Two classification algorithms were evaluated: a fuzzy rule building expert system (FuRES) and self-optimizing partial least squares discriminant analysis (PLSDA). PLSDA used two Latin partitions and ten bootstraps to calculate average pooled prediction errors. The number of components (i.e., latent variables) that minimized error was selected and used to build a model from the set of training data, which was then used as a prediction set.
Leave-one-out cross validation resulted in 100% correct assignment at the species and strain level and 100% correct differentiation of MRSA/MSSA. Confirmation of these results by FuRES classification and PLSDA consistently achieved 94 and 84% accuracy, respectively. Odd-numbered fatty acids (C15:0, C17:0) were more prevalent in sensitive isolates, while a shift to even-numbered fatty acids (C14:0, C14:1, C16:0, C16:1) was observed in resistant strains. A separate study of 160 samples encompassing 32 strains from 14 Staphylococcus species yielded accuracies of 98% and 96% at the species and strain level, respectively.
MOLI MS in situ lipid fragmentation readily produced unique species and strain-level staphylococcal fatty acid profiles. Results from a 160-sample study of 32 isolates resulted in highly accurate differentiation of resistant isolates at the strain level. Importantly, this allowed for simultaneous ID and resistance determination with a single test without the need for secondary methods or additional culturing.
References & Acknowledgements:
1. Cox, C.R., Jensen, K.R., Saichek, N.R. and Voorhees, K.J. 2015. Strain-level bacterial ID by CeO2-catalyzed MALDI-TOF MS fatty acid analysis and comparison to commercial protein-based methods. Nature Scientific Reports. 5:10470.
2. Saichek, N.R., Cox, C.R., Kim S., Harrington, P.B. and Voorhees, K.J. 2016. Strain-level Staphylococcus differentiation by CeO2-metal oxide laser ionization mass spectrometry fatty acid profiling. BMC Microbiology (In press).
IP Royalty: no
|Planning to mention or discuss specific products or technology of the company(ies) listed above:||