The Kinetics and mechanisms of non-thermal inactivation of L. monocytogenes and E. coli in raw milk cheeses

Listeria monocytogenes and Escherichia coli are food borne pathogens of concern to the Australian dairy industry. Recent consumer and producer requests to allow the manufacture and sale of raw milk cheeses in Australia have caused concern for the industry since raw milk has been identified as a vehicle for the transmission of these, and other, pathogens. Globally L. monocytogenes and E. coli have been associated with severe food-borne illness arising from the consumption of cheeses, demonstrating the ability of these organisms to survive despite the adverse environmental conditions imposed upon them in fermented milk products. An understanding of the ability of pathogens to survive the inimical conditions imposed by fermentation could contribute to the prevention of further outbreaks. This thesis describes the influence of temperature on low pH- and low water activity-induced inactivation of L. monocyto genes and E. coli in cheese made from both raw and pasteurised mil and in an analogous broth system. In addition, proteomic responses of L. monocytogenes to progressive acidification such as occurs during fermentation, were undertaken. The influence of temperature, that is not of itself lethal (i.e. in the range of 4 - 45°C), on the survival of L. monocytogenes under growth preventing pH and water activity conditions was investigated. Bacterial inactivation under these conditions is referred to, in this thesis, as non-thermal inactivation. Inactivation rates for three strains of L. monocytogenes were determined in rich broth adjusted to pH 3.5 and water activity 0.90, to prevent growth, and for temperatures in the range 5 to 45°C. Twenty eight inactivation rates were determined, plotted on Arrhenius coordinates (i.e. 1/[absolute temperature] vs. In [rate]), and lines of best fit determined by simple linear regression. The inactivation rate responses were comparable to those reported elsewhere, namely for i) E. coli inactivation in a model salami product and analogous broth system (McQuestin et al. 2006); ii) E. coli and L. monocytogenes inactivation in broth adjusted to pH 3.5 and water activity 0.90 (Zhang et al. 2010); and iii) E. coli inactivation in a variety of fermented meat products and analogous systems (McQuestin et al. 2009). The results showed that non-lethal temperature is also a strong determinant of the rate of inactivation of L. monocytogenes in environments in which growth is prevented by pH and water activity. To extend these studies, a meta-analysis of 45 independent studies was undertaken to investigate the relative influence of pH, water activity, and temperature on L. monocytogenes survival. Published data were re-evaluated to determine rates of inactivation, providing 1195 data over a pH range from 2.7 to 7.4, a\(_w\) range from 0.793 to 0.99, temperature range from 0-68°C, and in various food products, and for various processes and L. monocytogenes strains. When the data were presented as an Arrhenius model, temperature (0-42°C) accounted for 25% of the variance in /n(inactivation rate) data. The pH or water activity measured in broth or in food accounted for 21% and 6% respectively, of the variability in the data. The inactivation kinetics of L. monocytogenes and E. coli in cheese, when introduced as either contaminants in raw milk, or as post- pasteurisation contaminants in pasteurised milk used for cheese making, were evaluated. Raw and pasteurised semihard Manchego style cow's milk cheeses were prepared with milk inoculated with either L. monocytogenes Scott A or E. coli M23 (an acid tolerant non-pathogenic strain) prepared in both stationary and exponential stages of growth. The final pH of the cheese was 4.5 to 4.7 and final aw 0.91 to 0.92, conditions which, in combination, prevent the growth of both L. monocytogenes and E. coll. Cheeses were prepared and stored at six temperatures in the range 4°C to 25°C. Rates of change of microbial population density were determined and modelled using simple linear regression, where appropriate, and rates of change for both species plotted as a function of temperature using Arrhenius plots. Differences due to the physiological state of the cells (i.e., stationary or exponential growth phase) at the time of cheese making were also analysed. Consistent with the results of broth studies, temperature was shown to have the greatest effect on the rate of inactivation of both species of bacteria however the response was not the same as in broth. A second cheese challenge trial was undertaken to ascertain the robustness to accidental contamination with bacterial pathogens of a raw sheep's milk, Roquefort-style, cheese. Cheeses were prepared under laboratory conditions from raw ewe's milk inoculated with three strains each of both L. monocytogenes and E. coli. Triplicate inoculated cheeses were matured at each of six temperatures in the range 4°C to 20°C. Pathogen inactivation rate was modelled using simple linear regression and rates of changes for both species plotted as a function of temperature as an Arrhenius plot. All trials were followed for the equivalent of the normal commercial maturation time for the cheese so that the effect of pH changes in the cheese (specifically, pH returning to neutrality later in ripening as is typical in mould-ripened cheeses) could be assessed. Fat content, protein content, a\(_w\) and pH were measured to ascertain the potential influence of these factors on the fate of the introduced pathogens in the cheese. The results confirmed that raw ewe's milk Roquefort-style cheese does not support the growth and survival of L .monocytogenes or E. coli and that inactivation of these pathogens occurs when conditions of pH and water activity become inimical for growth. The results show that the inactivation kinetics of E. coil and L .monocytogenes in raw ewe's milk Roquefort-style cheese are consistent with those observed in broth (but not those observed in semi-hard cheese) and support the hypothesis that vegetative bacterial pathogens in foods that prevent their growth are inactivated at a rate that increases with increasing temperature as hypothesised by Ross et al (2008). The results are not consistent, however, with one of the findings of Zhang et al (2010) who concluded that there was no systematic differences between either the non-thermal inactivation rates of E. coli and L. monocytogenes at temperatures in the range 5 to 40°C in inimical environments. To begin to elucidate the physiology of such 'non-thermal' inactivation processes, the proteome changes in L monocytogenes in response to progressive acidification were studied in a model cheese fermentation system. Proteins synthesised by cultures subjected to conditions of pH decreasing from pH 5.0 to 4.6 over time were compared with proteins synthesised by stationary phase cultures grown under neutral pH conditions. In the initial growth stages, when pH was mild (pH 5.0) proteins commonly associated with cell growth were up-regulated. Proteins associated with stress responses (significantly, cold shock proteins, heat shock proteins and the Sigma B operon) were down regulated. As the cell culture continued to grow and the pH continued to decrease proteins associated with the acid-tolerance response and virulence (Sigma B operon, listeriolysin 0, internalin B, FOF1 system, PrfA) were upregulated to varying degrees. The use of proteomic techniques provided an overview of the physiology of L. monocytogenes subjected to low, and eventually inimical, pH and was consistent with current knowledge of acid-tolerance responses. In all, the work presented in this thesis develops the present understanding of the response of L. monocytogenes and E. coli to non-lethal storage temperature and inimical conditions relevant to the manufacture of cheeses from both raw and pasteurised, milk. The results from Chapters 2 and 4 support the hypothesis that nonlethal temperature is a key factor governing the rate of inactivation of vegetative bacteria in foods when other hurdles prevent their growth but indicate that the temperature effect may not be independent of the effect of food type, bacterial species and bacterial strain