A phenotypic screening bioassay for Escherichia coli stress and antibiotic responses based on Fourier-transform infrared (FTIR) spectroscopy and multivariate analysis.

Abstract

To develop and optimize a Fourier-transform infrared spectroscopy (FTIRS) phenotypic screening bioassay for stress responses, regarding the effect of nutrient content, bacterial growth phase and stress agent exposure time. A high-throughput FTIRS bioassay was developed to distinguish the stress responses of Escherichia coli to sodium hydroxide, hydrochloric acid, sodium chloride, sodium hypochlorite and ethanol. Principal component analysis and hierarchical clustering were used to quantify the effect of each parameter on bioassay performance, namely its reproducibility and metabolic resolution. Bioassay performance varied greatly, ranging from poor to very good. Spectra were partitioned into biologically relevant regions to evaluate their contributions to bioassay performance, but further improvements were not observed. Bioassay optimization was validated against empirical parameters, which confirmed a closer representation of known mechanisms on the antibiotic-induced stress responses. The optimized bioassay used standard nutrient content, cells in the late-stationary growth phase and a one-shift exposure duration. Only the optimized bioassay adequately and reproducibly distinguished the E. coli stress and antibiotic responses. The absence of performance improvements using partitioned spectra indicated that stress responses are imprinted on the whole-spectra metabolic signature. Highly optimized FTIRS bioassay parameters are vital in capturing whole-spectra metabolic signatures that can be used for satisfactory and reproducible phenotypic screening of stress and antibiotic responses.