Identification of Pathogenic Bacteria in Mixed Cultures by FTIR Spectroscopy

Abstract

Routine identification of pathogenic microorganisms predominantly based on nutritional and biochemical tests is a time-consuming process. In this study, a method based on FTIR spectroscopy was developed to detect and identify the presence of five possible pathogenic bacteria in ten different microorganism mixtures with each cocktail containing up to three different species at a concentration of 109 CFU/mL. A mathematical approach based on factoring out the common features in the spectral fingerprints was developed and implemented in conjunction with discriminant analysis. FTIR spectra of the mixtures were directly obtained and analyzed using canonical variate analysis based on the discriminant model for the presence of particular species in the pathogenic bacteria mixtures. In nine out of ten mixtures, the predictions were 100% accurate; in one mixture, only one false negative was reported, with no false positives. Results suggest that FTIR spectroscopy combined with a suitable analytical procedure has excellent potential as a fast, powerful, and reliable alternative for identifying a specific foodborne pathogen in a complex and/or competing system.