EOT-SERS Portable Detector and High-Throughput Metabolomics Methods for Intelligent Sensing of Foodborne Pathogens

Goals and ObjectivesThe goal of the proposed work is to provide a basic understanding of the distribution, diversity and potential food hazards of representative foodborne pathogens that have frequently occurred in recent years, so as to clarify and reduce the risk of foodborne disease outbreaks. The basic premise of our project is that if we adopt a comprehensive approach that can achieve the following objectives, the result will lead to more effective identification and prediction of foodborne disease outbreaks that may be caused by latent pathogens. The ultimate goal is to develop a portable sensor system that can quickly detect potential foodborne pathogens and establish high-throughput chromatography and tandem mass spectrometry methods based on pathogen-related metabonomic fingerprints. To achieve this goal, specific research objectives are established as follows.Objective 1: Develop a fiber-optic matrix sensing system that integrates multiple principles of lab-on-fiber sensors, including nano-pH and temperature sensing, extraordinary optical transmission (EOT), surface-enhanced Raman spectroscopy (SERS), and surface plasmon resonance (SPR). The matrix sensor array will be integrated into a portable detector to facilitate on-site real-time detection. The sensor units will be activated by specific VOC molecular signals related to foodborne pathogens presented in the sample. The focus will be given to develop and calibrate the fiber-optic matrix sensor array so that it can immediately assess and report the potential risks by displaying simple LED lights to indicate no risk (green), potential risk (yellow), or high risk (red).Objective 2: Use the ultra-high performance liquid chromatography and tandem mass spectrometry (UHPLC-MS/MS) and gas chromatography-mass spectrometry (GC-MS) to establish rapid and high-throughput metabolomics detection methods in PI's Advanced Bioanalytical Laboratory (ABL) at Lincoln University of Missouri. The developed methods will help determine and confirm the foodborne pathogens that may be contained within suspected food samples identified during the on-site testing. The use of the metabolomics method can lower the cost and avoid the time- and labor-consuming conventional methods that involve agar plate culture, PCR amplification, etc. Results that used to take days or even weeks can now be obtained within hours. Successful establishment of GC-MS and LC-MS methods will also simplify the instrumentation requirement among industrial inspection labs and lower their cost.Objective 3: Based on the datasets collected in Objective 1 and the quantitative analysis conducted in Objective 2, Objective 3 aims to develop and implement an ML-based decision-making framework to assist the portable detector in rapid identification and reliable prediction of foodborne pathogens. Different ML models will be tested, and the most suitable one with the highest efficiency and accuracy will be determined. The influence of a series of instrumental parameters will also be evaluated, such as the flow rate and volume, particle content in the aerosol, and so on. The purpose of doing so is to train and evaluate the best ML model's performance with sufficiently complex datasets. Well-trained ML models will then be used to improve the performance of the fiber optic sensors. For example, the proposed gold nanohole structures can be further modified to enhance the interaction with the target molecules in the evanescent field, thus, provide more information about the chemical/physical specificity of the target molecules.