Amplification of Raman spectra by gold nanorods combined with chemometrics for rapid classification of four Pseudomonas

Abstract

This paper demonstrates the application of surface-enhanced Raman scattering (SERS) using positive charged gold nanorods (Au NRs) as an enhancement substrate to classify Pseudomonas spp. coupled with multivariate methods. Four species of Pseudomonas as dominant spoilage bacteria of food were isolated from rotten chicken, namely, Pseudomonas gessardii (P9), Pseudomonas psychrophila (P8), Pseudomonas psychrophila (S2) and Pseudomonas fluorescens (T3). Au NRs were synthesized with positive charge by seed-mediated growth method which can be adsorbed onto the surface of the bacteria by electrostatic adsorption. SERS spectra were collected individually for four types of Pseudomonas and pretreated by mean centering (MC), then principal component analysis (PCA) and hierarchical clustering analysis (LDA) were used to achieve data dimensionality reduction and visualize the result of differentiation for the species of Pseudomonas. Particularly, the classification accuracy of LDA was reached to 100%. Following we applied hierarchical clustering analysis (HCA) to cluster each species of Pseudomonas and the results of HCA consistent with the results of 16S rRNA. This study has shown that SERS combined with LDA and HCA can be used as a reliable method to classify Pseudomonas.

Keywords: Classification; Multivariate calibration; Pseudomonas spp.; Surface-enhanced Raman scattering (SERS).