Sichuan University Analytical Instrument Research Center | JAFC: A Rapid Online Detection Method for Foodborne Pathogenic Bacteria Based on Ultraviolet Ionization Time of Flight Mass Spectrometer

The following article is translated from "X sensors"

Original link: 特约分享| 四川大学分析仪器研究中心| JAFC：基于紫外光电离飞行时间质谱仪的食源性致病菌快速在线检测方法

A rapid detection method for microbial volatile organic compounds (MVOCs) in foodborne bacteria was established using ultraviolet photoionization time-of-flight mass spectrometry (UVP-TOF-MS) technology. The results showed that there were significant differences in MVOCs among the five bacteria (Figure 1). A feature selection algorithm was used to screen the characteristic MVOCs of each bacteria, and several characteristic volatile organic compounds with significant differences among foodborne pathogens were identified, namely dimethyl disulfide, indole, and dimethyl trisulfide produced by Escherichia coli, acetone and isobutyraldehyde produced by Listeria, cyclopentane produced by Staphylococcus aureus, acetone and dimethyl disulfide produced by Shigella, and acetone produced by Salmonella typhimurium (Figure 2).

Figure 1 shows the results of principal component analysis of bacterial samples in standard culture medium. (a) PCA score plot of bacterial samples. (b) PCA score plot of bacterial samples and culture medium background samples

The MVOCs in Figure 2 are normalized content comparisons. (a)m/z=43. (b)m/z=59. (c)m/z=61. (d)m/z=69. (e)m/z=73. (f)m/z=94.(g)m/z=117. (h)m/z=126.

Online monitoring of MVOCs produced during bacterial growth showed that many volatile organic compounds were metabolized during the logarithmic phase of bacterial growth, and different bacteria had similar and different trends in the metabolism of the same MVOCs (Figure 3).

Figure 3 shows the clustering analysis results of the heat map, with the horizontal axis representing monitoring time (1-12h) and the vertical axis representing MVOCs ions (m/z 40-200). (a) Culture medium background (B.g). (b) Escherichia coli (E.c.). (c) Listeria (L.i.). (d) Staphylococcus aureus (S.a.). (e) Shigella (S.s.). (f) Salmonella typhimurium (S.t.).

Finally, the process of bacteria producing MVOC in different food matrices was investigated. The machine learning models of bacteria cultured in different matrices showed good classification performance for all five types of bacteria, with an accuracy rate exceeding 0.95 (Table 1).

Paper link: https://pubs.acs.org/doi/10.1021/acs.jafc.3c01486

Author: Deng Fulong, Department of Measurement and Control, School of Mechanical Engineering, Analytical Instruments Research Center, Sichuan University

Editor in charge: Luo Zewei, Department of Measurement and Control, School of Mechanical Engineering, Analytical Instruments Research Center, Sichuan University