Micro- and macroscopic investigation of the food microstructural influence on microbial dynamics: Case study in/on fish products

Microbiological food safety remains a major challenge in today's society, proving the need for improved strategies to avoid the presence of dangerous numbers of pathogens in our foods. By using predictive microbiology, the effect of processing, distribution, and storage conditions on the microbiological safety of foods can be quantified. Mathematical models which predict the growth or inactivation behaviour of pathogens are developed, in turn leading to a more realistic estimation of food safety risks during the entire food processing and distribution cycle. A significant omission in the effectiveness of predictive microbiology, however, is the limited knowledge concerning the influence of food microstructure on microbial dynamics. This limitation is partly caused by the fact that current predictive models are almost exclusively based on experiments performed in liquid model systems. While models developed according to this approach are useful to predict microbial behaviour in liquid foods, their applicability is limited for food products which exhibit a more complex microstructure such as aqueous gels, emulsions, and gelled emulsions. For this reason, researchers recently started to conduct microbiological experiments using model systems with various microstructures, combining the advantages of artificial food model systems and real food products. However, those model systems often exhibit variations in compositional and physicochemical factors, interfering with the elucidation of the isolated effect of food microstructure on microbial dynamics. The overall objective of this dissertation was to elucidate the effect of food microstructure on microbial dynamics, both concerning growth and thermal inactivation. In order to systematically study the isolated microstructural effect, a set of artificial food model systems with various microstructures was developed. As a specific case study, model system composition was based on processed fish products (e.g., fish soup, fish paté), while typical food microstructural elements of such products were also present (e.g., fat droplets, a viscoelastic matrix). In order to reach all sub-objectives of this dissertation, eleven different model systems were developed, i.e., liquid, xanthan (i.e., a viscous liquid), aqueous gel, emulsion (1, 5, 10, and 20% fat), and gelled emulsion (1, 5, 10, and 20% fat). All systems were demonstrated to be suitable for typical growth and (mild) thermal inactivation experiments involving the foodborne pathogenic bacterium Listeria monocytogenes as model microorganism. The use of these model systems enabled the investigation of the isolated effect of different microstructural aspects on microbial dynamics. In order to elucidate the influence of food microstructure on L. monocytogenes growth dynamics, the growth of the bacterium in/on the different model systems was studied both at the macro- and microscale, focussing on the influence of microbial growth morphology, the nature of the food matrix (i.e., viscous or gelled matrix, rheology), and the presence and concentration of fat droplets inside the food matrix. At the macroscale (i.e., based on viable plate counts), growth experiments were conducted at 4, 7 (only for the emulsions and gelled emulsions with different fat content), and 10°C. The nature of the food matrix influenced the maximum specific growth rate μmax of L. monocytogenes, being significantly higher in the viscous systems than in the gelled systems, while the lag phase duration λ was not significantly affected. The presence of a small amount of fat droplets (i.e., 1% fat content) resulted in a shorter λ and a higher μmax. An increase in fat content to 5% resulted in a further reduction of λ, while a further increase to 10 or 20% fat content did not significantly affect λ. The relationship between the fat content and μmax was more complex, but followed the same trends for emulsions and gelled emulsions. At the microscale, growth experiments were conducted at 10°C, using a Green Fluorescent Protein (GFP) L. monocytogenes strain, and results were assessed by Confocal Laser Scanning Microscopy (CLSM), providing explanations for the macroscopically observed phenomena. L. monocytogenes grew as a combination of single cells, small aggregates and microcolonies of different sizes in all model systems, with the distribution over these categories being dependent on the nature of the food matrix, and the food matrix fat content. In this regard, conlony size increased with increasing food matrix viscosity and fat content. Moreover, while L. monocytogenes mainly grew in the aqueous phase of the model systems, a partial preference of the bacteria to grow at the fat-water interface was also revealed. Both the observed growth morphology and the preferred phase for cell growth significantly influenced microbial growth dynamics. The influence of food microstructure on thermal inactivation dynamics of L. monocytogenes was investigated both at a lab scale and at a pilot scale. For the lab-scale experiments, L. monocytogenes was inactivated in the different model systems in temperature-controlled water baths set at 59, 64, and 69°C. The presence of a gelled food matrix increased the thermotolerance of L. monocytogenes, while the presence of fat droplets only resulted in a higher maximum specific inactivation rate kmax. While the influence of the food matrix fat content on kmax was complexly intertwined with the influence of the thermal conductivity, rheological properties, and the inactivation temperature, the bacterial thermotolerance was not affected in a similar fashion due to the small scale of the model systems at the lab scale. Sublethal injury (SI) was higher in viscous than in gelled systems, higher for cells that were grown inside the matrix, and higher with increasing fat content. Concerning the applicability in the pilot-scale set-ups relevant for this dissertation, the viscous model systems (i.e., liquid, xanthan, and emulsions) were suitable for treatments in the Shaka agitated retort (i.e., involving combined heating and reciprocal agitation of larger samples), while the liquid was the only system suitable for microwave treatments. In the Shaka agitated retort system, log reductions followed the order liquid ≥ emulsion 10% ≥ xanthan ≥ emulsion 20%. Heat transfer dynamics were significantly affected by the nature of the food matrix and the presence/concentration of fat droplets. Both factors exerted an influence on the thermotolerance of L. monocytogenes, additional to the direct influence of those factors. SI percentages, in some cases reaching values of up to 99%, were the highest in the liquid and the lowest in the xanthan systems, while the emulsions exhibited an intermediate behaviour. Overall, the influence of food microstructure on microbial dynamics was further elucidated and/or quantified in this dissertation by demonstrating that isolated food microstructural aspects such as the nature of the food matrix and the presence of fat droplets exert a significant influence on microbial growth and thermal inactivation dynamics. Hence, this research will contribute to the development of more accurate predictive models that will incorporate the food microstructural influence on microbial dynamics.

Jaar van publicatie:2020

Toegankelijkheid:Open