Non-thermal inactivation of Escherichia coli under conditions relevant to the production of uncooked comminuted fermented meats

Thermal inactivation kinetics of Escherichia coli is an area which has received much attention and is relatively well understood. By comparison, inactivation kinetics and mechanisms of cell death are not well described or understood for nonthermal conditions. Since the occurrence of two significant outbreaks associated with uncooked fermented comminuted meats (UCFM) in 1994 and 1995, considerable research effort has been expended to enhance knowledge of acid-tolerant E. coli in response to conditions encountered during production of UCFM. The well publicised illness and loss of life resulting from consumption of contaminated mettwurst forced a rethink on the safety of these products and had a dramatic effect on the entire Australian meat industry. Results of this study, combined with published literature, contradict previously established concepts (the vitalistic and mechanistic) that attempt to describe deviations from exponential inactivation kinetics. Minor variations in the age of inocula harvested at similar cell densities resulted in considerable differences in inactivation and survival. A resistant subpopulation was successfully removed by dilution, proving that all cells do not have a general similarity of resistance. Whilst these observations support the vitalistic concept, it was shown that this was not the main determinant of inactivation. Cells exposed to lethal pH display different inactivation patterns to those exposed to lethal water activity. The latter typically experience biphasic inactivation consisting of a rapid first phase of death followed by a slower more persistent second phase of "tailing". In comparison, cells exposed to lethal pH display triphasic inactivation, with an additional rapid third phase of death. Viable counts on selective media showed that cells exposed to lethal water activity experienced a considerably higher level of injury than those exposed to lethal pH. The latter displayed a consistent, lower level of injury on selective media, but died more rapidly. It was hypothesised that lethal pH placed a considerable energy burden on the cell, which resulted in the rapid third phase of inactivation. Comparison of uptake of a radiolabelled substance indicated that lethal water activity did not place a large energy burden on the cell. The above knowledge was utilised to determine the effect of variations in the order and timing of lethal stress imposition. Lethal acid stress was found to sensitise cells of E. coli to subsequent exposure to lethal water activity. Cells were more sensitive to subsequent stress the longer they were first exposed to low pH. This finding is relevant to food processing where hurdle technology is employed and may lead to novel strategies for the safe manufacture of minimally processed foods. The requirement for detailed, systematic data on nonthermal inactivation during UCFM processing led to the investigation of the feasibility of modelling death in a broth system. Such a system would allow for reproducible data to be obtained without the need for a complex and expensive meat matrix and smokehouse with appropriate temperature and relative humidity controls. An anaerobic environment using cooked meat medium was created that was able to mimic the effects of fermentation during manufacture of UCFM. Data from this system were used to refine a predictive model that has influenced Australian food standards legislation and assisted manufacturers to gain a better understanding of their processing effectiveness to reduce pathogens. This demonstrated the value of strategic research on microbial physiology and ecology. Detailed research using viable count methods is labour intensive and time consuming. The use of flow cytometry as a rapid method for determining the viability of E. coli after exposure to lethal pH was investigated. This technique proved to be reliable and rapid over the first 2-Log\(_{10}\) reduction with high cell numbers. However, beyond this level of inactivation, the sensitivity of the instrument rapidly decreased and flow cytometry ceased to be an effective method. It is concluded that modem methods and equipment still require considerable effort and design before they become as effective as traditional viable count techniques