This project developed generic models to predict the survival and subsequent growth of microorganisms that negatively impact food safety and quality. The models facilitate the development of strategies to ensure the safety and stability of products and the reputations of producers.
The growth characteristics of 20 different Listeria monocytogenes and 20 Lactobacillus plantarum strains were assessed under conditions of different temperature, pH, water activity and lactic-acid concentrations. This allowed for establishing the phenotypic variability using growth models. Experimental, biological and strain variation was quantified. The impact of strain and biological variability on the growth kinetics were comparable, and both were higher than experimental variability. The cardinal parameters and their variability were used as inputs for the gamma model to predict growth in food matrices. In addition, heat inactivation of the strains was assessed and D-values were determined and compared with data available in the literature. For thermal inactivation kinetics, the impact of strain variability was much higher than biological variability. Genomes of all twenty L. monocytogenes strains were sequenced, and gene-trait-matching was analysed. Furthermore, the model was validated in milk and ham to quantify the impact of food-product composition. The comparison provided knowledge of the factors influencing variability.
Stable acid-resistant variants of Listeria monocytogenes were isolated and their resistance to a range of other types of stresses was tested. To identify a potential genetic basis for certain resistance phenotypes, a cluster was made based on these phenotypes, showing a large group (11 variants) that clustered together. A common gene trait in this cluster of variants was an SNP mutation (located in rpsU). A ΔrpsU deletion mutant was constructed to confirm the role of rpsU in stress resistance. To determine whether stable resistant variants can also be found in other organisms, other L. monocytogenes strains and mutants of B. subtilis and L. plantarum were investigated. All data were used for simulation of population dynamics (sensitive and resistant population fraction) along a model food chain. An extra deliverable, in cooperation with the Biofilms project, has been set up in which the performance of L. monocytogenes acid-resistant variants in mixed biofilms with Lactobacillus plantarum has been evaluated.
For spore-forming bacteria, the variability in growth and heat inactivation of spores was also assessed for 20 strains of Bacillus subtilis, Bacillus cereus and Geobacillus stearothermophilus. Experimental work on heat inactivation and growth of B. subtilis and B. cereus has been completed. For Geobacillus stearothermophilus strains, optimal media have been established to ensure recovery after heat treatments, and spores have been harvested for 20 strains in 2 independent experiments. The data structure for fitting of the inactivation and growth model has been designed, and prediction of variability of B. subtilis spore inactivation was completed, showing significant strain variability. The experimental data was used to predict variability in outgrowth of spore-formers and heat inactivation of their spores.
Other activities involving spore-forming bacteria included the generation of an overview of all genes involved in sporulation and germination of B. subtilis. This overview has been generated and captured in a database made accessible through SporeWeb. This was extended by adding gene-expression data acquired through the sporulation of wild type B. subtilis strains. In addition, the contribution of specific factors to spore-germination efficiency, in response to various heat treatments, was studied. The resulting data were used as input for a generic model describing the contribution of specific nutrient-receptor complexes in germination responses as a function of heat.