Construction of a dynamic model to predict the growth of Staphylococcus aureus and the formation of enterotoxins during Kazak cheese maturation. (June 2023)

• Record Type: Journal Article
• Authors: Cai, Huixue; Pei, Sijie; Zhang, Yan; Liu, Rongrong; Lu, Shiling; Li, Baokun; Dong, Juan; Wang, Qingling; Zhu, Xinrong; Ji, Hua
• Is Part Of: Food microbiology. Volume 112(2023)
• Publication Date: 2023-06
• Subjects: Staphylococcus aureus -- Enterotoxin -- Predictive model -- Kazak cheese; Food Microbiology -- Periodicals; Aliments -- Microbiologie -- Périodiques; Food -- Microbiology; Periodicals; Food -- Microbiology -- Periodicals; Food contamination -- Periodicals; 664.001579
• DOI: 10.1016/j.fm.2023.104234
• Languages: English
• ISSNs: 0740-0020

Abstract

Abstract: Staphylococcus aureus is a common pathogen found in cheese whose Staphylococcal enterotoxins (SE) are the main pathogenic factors that cause food poisoning. The objective of this study was to construct two models to evaluate the safety of Kazak cheese products in terms of composition, changes in S. aureus inoculation amount, Aw, fermentation temperature during processing, and growth of S. aureus in the fermentation stage. A total of 66 experiments comprised of five levels of inoculation amount (2.7–4 log CFU/g), five levels of Aw (0.878–0.961), and six levels of fermentation temperature (32–44 °C) were performed to confirm the growth of S. aureus and the presence of SE limit conditions. Two artificial neural networks (ANN) successfully described the relationship between the assayed conditions and the growth kinetic parameters (maximum growth rates and lag times) of the strain. The good fitting accuracy (R 2 values were 0.918 and 0.976, respectively) showed that the ANN was appropriate. Experimental results showed fermentation temperature had the greatest influence on the maximum growth rate and lag time, followed by the Aw and inoculation amount. Furthermore, a probability model was built to predict the production of SE by logistic regression and neural network under the assayed conditions, which proved to be concordant in 80.8–83.8% of the cases with the observed probabilities. The maximum total number of colonies predicted by the growth model in all combinations detected with SE exceeded 5 log CFU/g. Within the range of variables, the minimum Aw for predicting SE production was 0.938, and the minimum inoculation amount for predicting SE production was 3.22 log CFU/g. Additionally, as competition between S. aureus and lactic acid bacteria (LAB) occurs in the fermentation stage, higher fermentation temperatures are conducive to the growth of LAB, which can reduce the risk of S. aureus producing SE. This study can help manufacturers to make decisions on the most appropriate production parameters for Kazak cheese products and to prevent S. aureus growth and SE production. Highlight: The growth of S. aureus during fermentation processing was developed. Probability model of SE production were developed during cheese maturation. Temperature, Aw, and inoculum size are key factors for controlling SE-production. SE production boundaries were described by logistic regression and neural network.