During grape storage, they are susceptible to fungal contamination, leading to economic losses and food safety concerns. Current methods for detecting grape mold include PCR, ELISA, electronic noses, and hyperspectral imaging. However, these methods are plagued by expensive equipment, personnel training costs, sample preparation, and insufficient sensitivity and specificity. Therefore, there is an urgent need for a low-cost, non-destructive, and highly sensitive detection method to enable early monitoring and warning of mold occurrence in grapes during storage.

To address this issue, researchers from Institute of Food Science and Technology (IFST), Chinese Academy of Agricultural Sciences (CAAS) constructed a set of Escherichia coli reporter strains by fusing 14 stress-responsive promoters with the luciferase reporter gene. They combined various data preprocessing algorithms and an optimized machine learning model to successfully achieve accurate non-destructive monitoring of mold contamination in grapes by Aspergillus niger, A. westerdijkiae, and Botrytis cinerea. The research results indicate that these engineered bacterial reporter strains are highly sensitive to volatile organic compounds released by infected grapes and can clearly distinguish between grapes in the asymptomatic stage (1 day after infection) and the symptomatic stage (2 days after infection). The machine learning prediction model can classify the degree of infection by three fungi on grapes, with classification accuracy reaching 100% for black mold, 92% for red mold, and 92% for gray mold, demonstrating exceptionally high predictive performance.

This novel biosensing technology, based on genetically engineered bacteria, enables rapid, non-destructive, and cost-effective monitoring of fungal spoilage in grapes. It can be applied not only to grapes but also to the microbiological spoilage detection of other agricultural products, making it a promising innovation. The next step will focus on enhancing the sensor's specificity and operational lifespan, with the goal of achieving widespread application in the storage and transportation of agricultural products.

The research results have been published online in the internationally renowned academic journal "Postharvest Biology and Technology" (JCR Q1, Top 5%, IF=7.0) in the field of postharvest technology for biological and horticultural products. Dr. Junning Ma, a postdoctoral researcher from the mycotoxin prevention and control in agro-products innovation team, is the first author of the paper, and Researcher Fuguo Xing is the corresponding author. This research was funded by National Key R&D Program of China (2022YFD0400104) and Agricultural Science and Technology Innovation Program of Institute of Food Science and Technology, Chinese Academy of Agricultural Sciences (CAAS-ASTIP-G2022-IFST-01).