Exploring physiological responses of Escherichia coli O157:H7 to abrupt cold and hyperosmotic stress

Shiga-toxin producing Escherichia coli (STEC), notably E. coli O157:H7, are a major cause of food-borne outbreaks, causing diseases such as haemorrhagic colitis. An important vector for the spread of STEC is through contaminated foods of bovine origin. Previous research has suggested that STEC can enter a viable but non-culturable (VBNC) state during carcase processing. VBNC cells cannot be detected using conventional enumeration methods and, as such, contaminated meat may not be detected before consumption, potentially resulting in human illness. It is, therefore, important to understand the behavior of VBNC pathogenic E. coli, so that appropriate preventive measures such as more informed risk assessments or product treatments may be developed. When stress conditions relevant to carcase chilling (abrupt temperature reduction from 35°C to 14°C and water activity reduction from a_w 0.993 to a_w 0.967) are replicated in vitro, three distinct phases of E. coli population behavior can be observed. These phases include an initial phase (‘Phase I’) showing a two-hour transient reduction in culturability, during which approximately 99% of cells are not culturable. In ‘Phase II’ they observed a second transient loss of culturable cells over approximately four days before culturability was restored. The subsequent phase (‘Phase III’) involved an extended lag phase followed by “true” exponential growth. The studies described in this thesis aimed to evaluate the existence of putative VBNC E. coli O157:H7 Sakai cells during dynamic changes in growth kinetics induced by simultaneous and abrupt, but non-lethal, downshifts in temperature and water activity similar to those that occur during carcase chilling, and to understand better the mechanisms by which the VBNC physiological state is induced. The formation of VBNC cells was confirmed by increases in culturable cell numbers using dilution-to-extinction assays, or Most Probable Number (MPN) technique, and application of growth?preventing temperatures (10°C and 6°C). Despite growth preventing conditions, culturable cell counts nonetheless increased. Therefore, recovery or resuscitation, not growth, were responsible for the increase in culturable cell numbers. Mechanisms responsible for transient loss of culturability in Phase I could not, however, be determined by flow cytometry with cells stained with propidium iodide (an indicator of membrane permeability), DiBAC (which highlights changes in membrane potential) or dihydroethidium (which indicates presence of free radicals). The loss of culturability in later phases of the population response curve, however, was correlated with a variety of indices of cell physiological state including changes in cell volume, reduction in membrane integrity and increased cellular free radical levels. Studies also involved the manipulation of recovery media composition and assessed the ability of stressed E. coli cells to form colonies on BHI agar plates. BHI agar media was enriched with: i) NaCl so that the water activity was equivalent to that of the stressed cells, ii) pyruvate content to assess the role of free radicals in cell culturability, or iii) the addition of bile salts to determine the membrane integrity of combined-stressed cells. Bile salts were used at concentrations calculated to not affect growth of a non-stressed control (1g/L and 5g/L). Improved recovery was observed on media with water activity reduced to equal that of the NaCl-stressed cells and on media containing pyruvate. Growth on the now isotonic recovery media was increased compared with non-altered media suggests that both hypo-osmotic and oxidative stress occur during normal enumeration on ‘typical’ recovery media and is a factor contributing to non-culturability. Bile salts had a deleterious effect on recovery of combined-stressed cells, but not on non-stressed controls. This suggests that the membrane of the stressed cells is more susceptible to damage, possibly due to exacerbation of damage caused by the combined stress. Action of bile efflux pumps is discussed further in 5.2.3. Experiments to assess changes in the biochemistry of combined stressed E. coli cells are described, including ATP content and number of carbon sources used by the stressed cells. Changes in membrane phospholipid composition were studied throughout the growth response curve including changes in ratios of double bonds, acyl chain length and phospholipid head group (cardiolipin, phosphatidylethanolamine and phosphatidylglycerol). Changes in morphology were also assessed. ATP content per viable cell decreased after the imposition of the combined stress and did not increase or return to pre-stress levels before or during resumption of exponential growth. No carbon sources were utilized by cells after the combined cold and osmotic stress described above was applied, despite the fact that the non-stressed control showed carbon utilization. Changes in phospholipid species were consistent with known responses to cold stress, i.e., shortening of acyl chains, however these changes were often accompanied by other changes such as saturation of the acyl chains which is contrary to the reported typical cold responses, but changes were observed in acyl chain length. Morphology changed throughout the growth curve, with cells forming both filaments and cocci in later stages. L-forms were not investigated in this study because published research suggests that L-forms are only formed by E. coli if treated with agents that inhibit cell wall formation. The novel observations reported and discussed in this thesis advance understanding of the responses of E. coli under stress conditions experienced during forced-air chilling of carcasses, and that may be useful to enhance their subsequent detection and inactivation. Oxidative stress and water activity imposed by the recovery media and changes in the membrane structure are potentially important factors in the culturability of combined-stressed E. coli cells. These factors could be manipulated to increase or further reduce culturability of combined-stress E. coli. Methods that increase the recovery of VBNC cells could be used to improve detection methods for the detection of VBNC pathogens both pre-harvest and after food processing. It has also been demonstrated that these cells are susceptible to both membrane and oxidative damage, and as such additional food processing treatments could be devised that target these phenomena to promote their inactivation. As shown with modified media, such treatments may be more effective within hours of the initial stress but lose efficacy over time, and as such may be more suitable to an environment in which the timing of treatments can be controlled. This would be particularly relevant to food processing, and, for example, could be used to create more informed risk assessments, or additional carefully timed treatments could be employed to inactivate pathogenic VBNC.