Influence of growth temperature on thermal tolerance of leading foodborne pathogens

Abstract Accurate prediction of the thermal destruction rate of foodborne pathogens is important for food processors to ensure proper food safety. When bacteria are subjected to thermal stress during storage, sublethal stresses and/or thermal acclimation may lead to differences in their subsequent tolerance to thermal treatment. The aim of the current study was to evaluate the thermal tolerance of Escherichia coli O157:H7, Listeria monocytogenes, Salmonella enterica, and Staphylococcus aureus that are incubated during overnight growth in tryptic soy broth at four temperatures (15, 25, 35, and 45°C). Following incubation, the bacteria were subjected to thermal treatments at 55, 60, and 65°C. At the end of each treatment time, bacterial survival was quantified and further calculated for the thermal death decimal reduction time (D‐value) and thermal destruction temperature (z‐value) using a linear model for thermal treatment time (min) vs. microbial population (Log CFU/ml) and thermal treatment temperature (°C) vs. D‐value, respectively, for each bacterium. Among the four bacterial species, E. coli generally had longer D‐values and lower z‐values than did other bacteria. Increasing patterns of D‐ and z‐values in Listeria were obtained with the increment of incubation temperatures from 15 to 45°C. The z‐values of Staphylococcus (6.19°C), Salmonella (6.73°C), Listeria (7.10°C), and Listeria (7.26°C) were the highest at 15, 25, 35, and 45°C, respectively. Although further research is needed to validate the findings on food matrix, findings in this study clearly affirm that adaptation of bacteria to certain stresses may reduce the effectiveness of preservation hurdles applied during later stages of food processing and storage.

is required to destroy a given bacterium. The D-value is defined as the time required at a constant temperature to destroy 90% of the bacteria present, while a z-value is defined as the change in temperature necessary to bring about the 90% change in D-value, which indicates shortening the duration of time with the increase of heat. Hence, they can be used as predictors for how susceptible a bacteria is to changes in temperature and in duration.
When bacteria are subjected to thermal stress during food processing treatments, sublethal stresses and thermal acclimation can occur that may lead to differences in terms of their tolerance to the heat process that follows. Furthermore, most research documenting the time and temperature relationship necessary for the destruction of bacteria has been conducted using a limited number of different bacterial strains or species in isolation (Patchett, Watson, Fernández, & Kroll, 1996;Redondo-Solano, Burson, & Thippareddi, 2016;Semanchek & Golden, 1998;Sörqvist, 2003). These studies are difficult to compare due to differences in multiple variables, including between strains or species of bacteria, between research laboratories and environmental conditions, and between technologies used for the evaluation of microbial destruction. Therefore, culturing and evaluating bacterial pathogens under concurrent conditions will better elucidate the net effect of growth temperature on inactivation and injury of pathogens due to thermal process and thereby facilitate cross-species comparisons.
Escherichia coli O157:H7, Listeria monocytogenes, Salmonella enterica, and Staphylococcus aureus are the leading bacteria accountable for the vast majority of foodborne illnesses, hospitalizations, and deaths in the United States (CDC 2016). Recognizing the importance of these pathogens associated with various food products, as a preliminary phase, the objective of current study was to evaluate the thermal tolerance (D-and z-values) of different species of bacteria grown in a laboratory culture medium (tryptic soy broth supplemented with 0.6% yeast extracts) at four different temperatures (15, 25, 35, and 45°C). In addition, the term "cool," "ambient," "warm," and "excessive heat" for 15, 25, 35, and 45°C, respectively, defined as in U.S. Pharmacopeia 659 (USP 2017) are used for description purpose of thermal stresses in this article.

| Bacterial strains used
Bacterial species used for the study were obtained from American

| Growth temperature
In order to investigate thermal destruction variability in foodborne pathogens induced by thermal stresses during storage in optimum medium, 0.1 ml of each strain was inoculated into 10 ml TSBYE and incubated for 24 hr at 15, 25, 35, and 45°C. Following incubation, the bacteria were centrifuged for 10 min at 2,000 g and 22 ± 2°C in a centrifuge (Model Heraeus Megafuge 16, Thermo Scientific). The pellets were then suspended in 10 ml of sterile 0.85% saline solution and centrifuged again at 2,000 g for 10 min and resuspended in 10 ml of sterile 0.85% saline solution. Equal volumes of either three or four strains of each bacterial species were combined to make an inoculum containing approximately equal numbers of cells of each species of E. coli O157:H7, L. monocytogenes, S. enterica, or S. aureus. In other words, a cocktail containing either 3 or 4 strains of each bacterial species was used as inocula for thermal destruction rate study. Levels of bacterial counts obtained after 24-hr incubation at each temperature are shown in Table 1 and used as inocula for the thermal destruction study. The inoculum levels shown in Table 1 are the average of three independent replicate trials of each bacterial population prior to being subjected to thermal treatments at 55, 60, and 65°C. Three independent replicate trials were conducted for each thermal treatment at each incubation temperature (thermal stress) of 15, 25, 35, and 45°C.

| Thermal destruction
Each bacterial species of inoculum (2 ml) was introduced into a sterile polyethylene whirl-pak® sample bag (Nasco) and sealed. The bags were 7.5 × 12.5 cm in size with a thickness of 0.057 mm. The sample bags were then completely immersed in a water bath (Lab-Line Water Bath Model 18900 AQ, Thermo Scientific) and held at 55°C for 300, 900, and 2,700 s; 60°C for 30, 90, 270 s; and 65°C for 3, 9, and 27 s.
These ranges of temperatures, which are commonly used in cooking beef up to medium-rare, were chosen in this study for future validation study in mind on food matrices such as ground beef and roasted beef (Line et al., 1991). The sample temperature was monitored using a thermocouple connected to a thermometer (Traceable Infrared Dual-Lasers Thermometer, Model S02273, Control Co.). At the end of each exposure time, the sample bags were removed from the water bath and immediately immersed in ice water (0°C) for 5 min to stop further inactivation due to thermal treatment. Bacterial suspensions in the sample bags were then serially diluted in sterile 0.85% saline solution, surface-plated on standard method agar (SMA), and incubated at 35°C for 48 hr prior to quantification of bacterial survival. The counts were expressed as log colony-forming unit (CFU)/ml.

| Calculation of D-and z-values
The destruction rate curves (R 2 ≥ .89) were constructed by plotting  Figure 1. The thermal destruction temperature (z-values) was also calculated by plotting the temperature against log D-value, and the data were fitted by using linear regression with Excel software (2013, Microsoft). The inverse of the slope was reported as the z-value in °C.

| Statistical analysis
The thermal destruction times (D-values) and temperatures (z-values) for species of E. coli O157:H7, L. monocytogenes, S. enterica, and S. aureus were obtained from three independent replications. Data (log CFU/ml, D-values, and z-values) were subjected to an analysis of variance and Duncan's multiple range test (SAS Institute) to determine the significance of the differences (p < .05) in mean values.

| RE SULTS AND D ISCUSS I ON
The effect of thermal stresses (incubation temperature) on the level of bacteria in TSBYE after 20 ± 2-hr incubation is shown in S. aureus 6.87 ± 0.05Bb 9.14 ± 0.11Aa 9.07 ± 0.00Aa 5.67 ± 0.52Bc Note: Means followed by the same uppercase letters in the same column are not significantly different (p > .05); means followed by the same lowercase letters in the same row are not significantly different (p > .05); data are expressed as means ± standard error (n = 3).
TA B L E 1 The level of bacterial species obtained after 20 ± 2-hr incubation at various temperatures and used as inocula

| E. coli O157:H7
The inoculum level of E. coli (Table 1)  Findings from our study agree with the results presented by Jackson, Hardin, and Acuff (1996) that stationary-phase populations of E. coli O157:H7 grown at 37°C were more resistant to heat treatment than populations grown at 23 and 30°C. Studies have also indicated that heat tolerance was greater when cells were grown at 37 or 40°C rather than 10, 23, 25, or 30°C (Jackson et al., 1996;Kaur, Ledward, Park, & Robson, 1998;Semanchek & Golden, 1998). Katsui, Tsuchido, Takano, and Shibasaki (1981) reported that increased heat tolerance associated with changes in growth temperature were

| L. monocytogenes
The reduction level of Listeria inoculum (  Table 3. A review (Sörqvist, 2003)  Note: Means followed by the same uppercase letters in the same column within the same incubation temperature are not significantly different (p > .05); means followed by the same lowercase letters in the same column within the same thermal treatment temperature are not significantly different (p > .05); z-values followed by the same lowercase letters are not significantly different (p > .05); data are expressed as means ± standard error (n = 3).

| Salmonella
Reduction gradients of the bacteria incubated at 25°C and 45°C were the highest and the lowest, respectively, when they were subsequentially subjected to thermal treatments at 55 and 60°C (Data not shown). However, at 65°C thermal treatment, the reduction gradient for the bacteria incubated at 15°C and 35°C was the highest and the lowest, respectively. Results in Table 4 (Table 1) may have contributed to these results. Therefore, further research is in consideration to manifest the influence of bacterial growth phase on their thermal resistance.
In general, the time required to destroy 90% of Salmonella grown at 45°C was the longest compared with those incubated at 15, 25, and 35°C. The increase in temperature needed to bring about the 90% reduction in the duration of time (D-value) was the highest and lowest for Salmonella grown at 25 (6.73 ± 0.28°C) and 45°C (5.47 ± 0.31°C), respectively. The bacteria grown at lower temperature (25°C) required shorter time but higher temperature increase to be destroyed than the bacteria grown at higher temperature (45°C) indicating that destruction of Salmonella bacteria grown at low and high temperature may be more dependent upon duration of time and increase of temperature, respectively.

| Staphylococcus
The reduction level of Staphylococcus inoculum incubated at 45°C and subsequentially subjected to thermal treatments at 55, 60, and 65°C was the highest among the four tested incubation temperatures (Data not shown). At 55 and 60°C thermal treatments, reduction gradients of the bacteria incubated at 45°C were the highest and the lowest at 35°C. However, at 65°C thermal treatment, the reduction gradient Note: Means followed by the same uppercase letters in the same column within the same incubation temperature are not significantly different (p > .05); means followed by the same lowercase letters in the same column within the same thermal treatment temperature are not significantly different (p > .05); z-values followed by the same lowercase letters are not significantly different (p > .05; data are expressed as means ± standard error (n = 3).
TA B L E 4 D-(min) and z-values (°C) of S. enterica that incubated during overnight growth (20 ± 2 hr) at four temperatures (15, 25, 35, and 45°C) and subsequentially subjected to thermal treatment at 55, 60, and 65°C in 0.85% saline solution was the highest and the lowest at 35°C and 25°C, respectively. This discrepancy of the bacterial response to thermal treatments clearly demonstrates thermal tolerance dissimilarity of the bacteria due to prior thermal stresses as well as thermal treatment temperatures. Table 5 demonstrated that the bacteria incubated at 35°C, which is close to the optimum temperature (37°C, Albrecht, 2017;NZMPI 2001) for Staphylococcus growth, took the longest time (D-value) to reduce its population by 90%. While D-value of the bacteria incubated at 45°C was significantly (p < .05) lower at thermal treatment of 55°C than those incubated at other temperatures, no significant difference of D-values among the bacteria that subjected to other incubation temperatures (15, 25, and 35°C) and thermal treatments (60 and 65°C) was observed. However, the increase in temperature (z-value, 6.77 ± 0.52°C) needed to bring about the 90% reduction in the duration of time (D-value) was the highest for the bacteria incubated at the highest incubation temperature (45°C).

Results in
In general, the D-values (1.57 min at 60°C and 0.22 min at 65°C) obtained from our study on the bacteria incubated at 35°C were similar to those previously reported by Amado et al. (2014) affirms that bacterial adaptation to certain stresses may reduce the effectiveness of preservation hurdles applied during later stages of food processing and storage (Benito, Ventoura, Casadei, Robinson, & Mackey, 1999;Lou & Yousef, 1997;Oh et al., 2009;Wiegand, Ingham, & Ingham, 2009). Additionally, studies (Leenanon & Drake, 2001;Lisle et al., 1998;Swientek, 2017) reported that response to stress may not only enable survival of bacteria under more severe conditions, but also enhance their resistance during subsequent processing conditions and increase pathogenicity.

| CON CLUS ION
In conclusion, findings in this study clearly indicate that storage and holding temperatures similar to those encountered in food service influence the ability of foodborne pathogens to survive subsequent thermal treatments. Therefore, further research on food matrix associated with a variety of food processing related stresses (i.e., acid, fat, protein, starch, sugar, and water) as a bacterial growth and inactivation model in vitro and in situ is needed to validate current findings.
Additional research on the influence of growth temperature at the same physiological stage (i.e., lag, log, and stationary phase) of cells on their sensitivity to sublethal stresses will also manifest the determination of adaptive responses in bacteria. Our findings here with further validation may also assist the food industry with the establishment of critical limits for the safe thermal treatment of food products.

ACK N OWLED G M ENTS
The authors acknowledge the technical advice and/or assistance  6.77a Note: Means followed by the same lowercase letters in the same row within each incubation temperature are not significantly different (p > .05); detail values with standard errors can be found in appropriate tables from 2 to 5.