
Methods: ATR-FTIR spectra were acquired by transferring isolated colonies from MacConkey agar plates onto the ATR sampling surface of a portable ATR-FTIR spectrometer. ATR-FTIR spectral acquisition time was ~1 min. Four spectra were collected for each isolate from different colonies on the same culture plate, yielding a spectral database containing approximately 1,400 ATR-FTIR spectra of Enterobacteriaceae. Spectral data analysis was performed by hierarchical cluster analysis (HCA) and principal component analysis (PCA) in conjunction with the use of a feature selection algorithm.
Results: Approximately 350 clinical isolates of Enterobacteriaceae, including Enterobacter, Klebsiella, Proteus, Citrobacter, Salmonella, and Shigella spp. as well as pathogenic (STEC, EHEC, EPEC and UPEC) and non-pathogenic E. coli, were employed in this study. Genus-level and species-level classification of all the clinical isolates was achieved by multivariate statistical analysis of the FTIR data with an overall rate of correct classification exceeding 98%, including complete differentiation between E. coli and Shigella spp. Further multivariate statistical analysis of the ATR-FTIR spectra of the E. coli isolates yielded complete differentiation between pathogenic and non-pathogenic strains and discrimination among the four E. coli pathotypes included in this study.
Conclusion: ATR-FTIR spectroscopy can provide a simple and rapid method for species identification of Enterobacteriaceae as well as for discrimination between E. coli and Shigella spp. within minutes after initial culture.

P. Lebel,
None
L. Lam, None
H. Kim, None
J. Sedman, None
A. Ismail, None
C. Clark, None
I. Iugovaz, None
See more of: Poster Abstract Session