Lisa Purk

The aim of the current study is to develop and characterise novel complex multi-phase in vitro 3D models, for advanced microbiological studies. More specifically, we enriched our previously developed bi-phasic polysaccharide (Xanthan Gum)/protein (Whey Protein) 3D model with a fat phase (Sunflower Oil) at various concentrations, i.e., 10%, 20%, 40% and 60% (v/v), for better mimicry of the structural and biochemical composition of real food products. Rheological, textural, and physicochemical analysis as well as advanced microscopy imaging (including spatial mapping of the fat droplet distribution) of the new tri-phasic 3D models revealed their similarity to industrial food products (especially cheese products). Furthermore, microbial growth experiments of foodborne bacteria, i.e., Listeria monocytogenes, Escherichia coli, Pseudomonas aeruginosa and Lactococcus lactis on the surface of the 3D models revealed very interesting results, regarding the growth dynamics and distribution of cells at colony level. More specifically, the size of the colonies formed on the surface of the 3D models, increased substantially for increasing fat concentrations, especially in mid- and late-exponential growth phases. Furthermore, colonies formed in proximity to fat were substantially larger as compared to the ones that were located far from the fat phase of the models. In terms of growth location, the majority of colonies were located on the protein/polysaccharide phase of the 3D models. All those differences at microscopic level, that can directly affect the bacterial response to decontamination treatments, were not captured by the macroscopic kinetics (growth dynamics), which were unaffected from changes in fat concentration. Our findings demonstrate the importance of developing structurally and biochemically complex 3D in vitro models (for closer proximity to industrial products), as well as the necessity of conducting multi-level microbial analyses, to better understand and predict the bacterial behaviour in relation to their biochemical and structural environment. Such studies in advanced 3D environments can assist a better/more accurate design of industrial antimicrobial processes, ultimately, improving food safety.

Understanding and predicting bacterial behaviour in foods is vital for food safety. Although in the past most food microbiological research was conducted in liquids, it is now known that bacterial behaviour changes fundamentally when grown in structured environments. Furthermore, the bacterial behaviour is also affected by the natural microflora of foods and possible cross-contaminants. These can interact synergistically but can also be used as an antagonistic tool for food safety. The aim of this work is to perform a systematic study on the impact of fat concentrations (FC) on bacterial kinetics, their inter-species interactions, and their sensitivity towards the mild preservation technology of cold atmospheric plasma (CAP) in complex triphasic 3D systems. Building on our biphasic protein/polysaccharide food model (FM), a third fat phase was incorporated into the system (10-60%). Single-cultures of Listeria monocytogenes, Escherichia coli, Pseudomonas aeruginosa and Lactococcus lactis were grown on the surface of the FMs, as well as listerial co-cultures with each of the listed bacterium. A multiscale analysis took place macroscopically (plate count) and microscopically (confocal-laser-scanning-microscopy). Subsequently, the single- and co-cultures grown on the FMs surfaces were treated with CAP. Overall, the macroscopic analysis revealed increased growth of single cultures in comparison to co-cultures, but no significant impact in respect to the tested FCs. However, on the microscopic scale, generally, differences between the FCs were observed. More specifically, the bacterial colony sizes and biofilm formation were increased with increasing FC, more significant in co-cultures than in single-cultures. Due to these microscopic differences, a different level of cell-to-cell and colony-to-colony interaction takes place. This was further demonstrated by the susceptibility and resistance of the single- and co-cultures to the CAP treatment. In conclusion, our results indicate the importance of accounting for the microflora complexity and their interactions in food systems when predicting microbial behaviour and designing food decontamination treatments to ensure food safety.

The demand for products that are minimally processed and produced in a sustainable way, without the use of chemical preservatives or antibiotics have increased over the last years. Novel non-thermal technologies such as cold atmospheric plasma (CAP) and natural antimicrobials such as grape seed extract (GSE) are attractive alternatives to conventional food decontamination methods as they can meet the above demands. The aim of this study was to investigate the microbial inactivation potential of GSE, CAP (in this case, a remote air plasma with an ozone-dominated RONS output) and their combination against L. monocytogenes on five different 3D in vitro models of varying rheological, structural, and biochemical composition. More specifically, we studied the microbial dynamics, as affected by 1 % (w/v) GSE, CAP or their combination, in three monophasic Xanthan Gum (XG) based 3D models of relatively low viscosity (1.5 %, 2.5 % and 5 % w/v XG) and in a biphasic XG/Whey Protein (WPI) and a triphasic XG/WPI/fat model. A significant microbial inactivation (comparable to liquid broth) was achieved in presence of GSE on the surface of all monophasic models regardless of their viscosity. In contrast, the GSE antimicrobial effect was diminished in the multiphasic systems, resulting to only a slight disturbance of the microbial growth. In contrast, CAP showed better antimicrobial potential on the surface of the complex multiphasic models as compared to the monophasic models. When combined, in a hurdle approach, GSE/CAP showed promising microbial inactivation potential in all our 3D models, but less microbial inactivation in the structurally and biochemically complex multiphasic models, with respect to the monophasic models. The level of inactivation also depended on the duration of the exposure to GSE. Our results contribute towards understanding the antimicrobial efficacy of GSE, CAP and their combination as affected by robustly controlled changes of rheological and structural properties and of the biochemical composition of the environment in which bacteria grow. Therefore, our results contribute to the development of sustainable food safety strategies.