Detection and discrimination of enteric bacteria using gas sensor arrays.

摘要

Bacteria may be distinguished from one another by the metabolic products formed. Some of these products are volatile and present distinctive odors. An array of cross-selective gas sensors that generate a pattern of signals associated with the sample vapors can be used to provide the specificity needed to discriminate between samples that are chemically different. This work focused on the application of gas sensor arrays to the detection and discrimination of enteric bacteria as a demonstration of the utility of these arrays to biological testing applications. Multiple species of enteric bacteria were compared to Escherichia coli and blank broth. The headspace vapors from sealed, pure bacterial cultures of varied concentrations were evaluated using three gas sensor arrays including: quartz microbalance (QMB), metal oxide (MOX), and electrochemical gas sensors. The data generated were analyzed by Principal Component Analysis (PCA) and K-nearest neighbor (KNN) methods. The ability to detect all of the species was shown and eight of nine species were readily discriminated from E. coli cultures containing approximately 109 cells/mL. Consistent sampling methods were necessary. A heterogeneous array of gas sensors was more selective than any single type of gas sensor. Generally, the mass sensitive QMB sensors were useful to detect culture growth and the reactive MOX and electrochemical sensors were needed for discrimination. The sensitivity of the sensors to moisture, the order the arrays were used, and the stability of the sensors were evaluated. Industrial applications that can use stationary systems and automated sampling controls are in more immediate reach for analysis by gas sensor arrays than applications with more variable samples. With further refinements in sensor performance and data analysis methods, focused, remote testing applications will benefit from the operational simplicity of gas sensor array systems.