Principal component analysis-assisted fluorescent phenylboronic acid probes for discrimination and detection of living pathogenic bacteria

Abstract

A simple analytical platform was constructed based on a phenylboronic acid-functionalized fluorescent probe with machine learning methods. Under the optimized experimental conditions, both fluorescein isothiocyanate-aminophenylboronic acid (FITC-APBA) and rhodamine B isothiocyanate-aminophenylboronic acid (RBITC-APBA) could be used for the discrimination of bacteria, including Escherichia coli O157:H7, Salmonella, Shigella, and Staphylococcus aureus, by using principal component analysis (PCA) and K-means clustering analysis to treat the fluorescence response data. As FITC-APBA showed better discrimination performance than RBITC-APBA, the fluorescent FITC-APBA was further used for identifying the target bacteria in real samples. It is interesting to find that the FITC-APBA-based analytical platform can also discriminate the living bacteria from their dead states, providing a good supplement to the PCR-based bacterial analysis. In this study, a PCA-assisted fluorescent phenylboronic acid-based analytical platform was developed for the discrimination and detection of bacteria in food samples, opening a novel and rapid way for bacterial analysis.