Effect of pH and temperature on Shiga toxin-producing Escherichia coli O157:H7 isolated from bovine reservoir

Abstract

The aim of this study was to evaluate the individual and combined effect of pH (between 4.0 and 7.0) and storage temperature (between 6 and 35°C) on two wild-type <italic>E. coli</italic> O157:H7 (STEC 902 and STEC 2004) strains. The experiments were performed in laboratory culture medium, employing Central Composite Rotational Design (CCRD). For comparison, Enteropathogenic <italic>E. coli</italic> (EPEC) O55:H7 and non-pathogenic <italic>E. coli</italic> (npEC) strains were used. To validate the model, seven additional tests not included in the experimental design were performed. The model showed goodness-of-fit. The CCRD results indicated that the temperature significantly affected the microbial population of <italic>E. coli</italic> strains studied. With best development at temperatures between 20 and 35°C, while refrigeration temperatures (between 6 and 10°C) are unfavorable for the development of STEC O157:H7, independent of pH. The wild-type STEC 902 and STEC 2004 strains showed similar behavior under pH and temperature conditions in laboratory culture medium. On the other hand, control strains showed different behavior than STEC strains, being significantly affected by pH. The data indicate that the response to stress resulting from exposure to different pH and temperature levels varies according to the <italic>E. coli</italic> strain evaluated.

Key words
microbial growth; microbial survival; acid tolerance; predictive microbiology; Central Composite Rotational Design; response surface methodology

INTRODUCTION

Shiga toxin-producing Escherichia coli (STEC) belongs to a significant group of E. coli, with certain serotypes recognized as food-borne pathogens. Pathogenic STEC serotypes are capable of causing bloody diarrhea, hemorrhagic colitis (HC) and hemolytic-uremic syndrome (HUS) in humans, which, in severe cases, can result in death (Croxen et al. 2013, Amin et al. 2022). The prevalence of HC and HUS has become a public health problem in several countries, due to the large number of people infected and the high rate of morbidity and mortality, which increases the costs related to the treatment and prevention of these diseases (Newell & La Ragione 2018, Kim et al. 2020). STEC O157:H7 is the most virulent serotype and is frequently associated with serious human disease (Croxen et al. 2013, Gonzalez & Cerqueira 2019). Outbreaks of STEC O157:H7 infection have been described in several countries (Luna 2018, Hassan et al. 2019, Coulombe et al. 2020). In Brazil, STEC O157:H7 has been described in sporadic cases of human disease (Gonzalez & Cerqueira 2019, Irino et al. 2002, Souza et al. 2011). Cattle are the main reservoir of STEC O157:H7 (Kim et al. 2020, Gonzalez & Cerqueira 2019, Gonzalez et al. 2016). The transmission of STEC occurs mainly by the ingestion of foods of bovine origin (meat and milk). However, contaminated vegetables and water are also involved in STEC outbreaks (Kim et al. 2020). The STEC O157:H7 isolates from foods of bovine origin have been described in Brazil (Gonzalez & Cerqueira 2019).

Several factors act as hurdler, controlling microbial development in food. In this sense, the quantity, intensity and combination of factors can promote the microbiological food stability and safety (Leistner 1992). pH and storage temperature are hurdles that are, often, sufficient to control bacterial development in food and guarantee its safety (Leistner 2000). These hurdles (pH and storage temperature) can control the development of STEC O157:H7 in food (Juneja et al. 2020, Balamurugan et al. 2014, Akkermans et al. 2017). Some authors have described the effect of pH and storage temperature on the development of STEC O157:H7 in food matrices (Arican & Andic 2011, Biswas & Pandey 2017, Kim et al. 2018) and in culture media (Kim et al. 2018, Valero et al. 2010, Stachelska & Charmas 2018). However, Koutsoumanis and Lianou (Koutsoumanis & Sofos 2004) highlighted the influence of variability in the behavior of individual cells in the assessment of microbiological risks. Each microorganism has unique characteristics, and a strain may show variations over time (Cepl et al. 2016). This raises questions about the variability in the behavior of strains of the same species or of the same serotype. In this context, assessing the development of wild-type STEC O157:H7 strains can provide important information that can be used to estimate microbiological risk of this potential pathogen. Although foods inherently contain a range of variables that can influence microbial development, including its composition and microbiota (Wang & OH 2012, Silva et al. 2023), laboratory culture media allow for more precise control of individual variables. Therefore, this study aimed to evaluate the individual and combined effect of pH and temperature on the development of two wild-type STEC O157:H7 strains, isolated from bovine feces, in laboratory culture medium, using the central composite rotational design (CCRD). Additionally, Enteropathogenic E. coli (EPEC) O55:H7 and non-pathogenic E. coli (npEC) strains were evaluated for comparison.

MATERIALS AND METHODS

Bacterial strains and inoculums preparation

Four E. coli strains were used, two wild-type STEC O157:H7 strains (STEC 902 and STEC 2004), isolated from bovine feces and carriers of a set of important virulence marker genes (stx2c, eaeγ, tir, espAγ, espBγ, iha, E-hlyA, astA, espP) (Gonzalez et al. 2016), an EPEC O55:H7 strain (O157:H7 precursor serotype) (Zhou et al. 2010, Kyle et al. 2012), and a npEC strain, indicative of fecal contamination, previously isolated from Minas frescal cheese.

Bacterial cultures stored at -20°C were activated in Trypticase Soy Broth (TSB; Himedia) incubated at 35°C for 18-20 h. Bacterial inoculums were standardized through the Mac Farland scale 0.5, obtaining approximately 1.5 x 108 cells/mL. A total of 200 μL of each standardized bacterial suspension was seeded individually in 5 mL of modified pH-TSB (MpH-TSB), under the conditions of pH and temperature indicated in Table I, and a fixed value of activity water (Aw) of 0.98, obtaining an inoculum of about 4 x 106 cells/mL (6.6 log CFU/mL).

Table I
Independent factors (pH and temperature) and their coded and not coded levels according to the central composite rotational design (CCRD).

Factor	Level
	-1.41	-1	0	+1	+1.41
Temperature (°C)	6	10	20	30	35
pH	4.0	4.5	5.5	6.5	7.0

The pH was adjusted with 1 mol/L of hydrochloric acid (HCl) or sodium hydroxide (NaOH) and measured in an electronic pH meter (model mPA-210), directly on the MpH-TSB. The assays at 6, 10 and 20°C were carried out in B.O.D. (MS Culture - Eletrolux, model RE120) and at 30 and 35°C in calibrated incubators (Fanem LTDA®, model 002 CB). Aw was measured on the activity water apparatus (Decagon, Pawkit-Aqualab).

Design of Experiments (DOE)

Two independent variables that control bacterial development, pH and temperature, were included in the DOE (Table I). A CCRD was used in (22 + (2 x 2) + 3) experiment, with two level (-1 and +1), two axial points (-1.41 and +1.41) and three central points (0), totaling 11 tests (Table II), performed in three independent replicates. Statistica7® software (StatSoft, OK, USA) was used.

Table II
Predicted and observed values (log CFU/mL) of STEC O157:H7 (STEC 2004 and STEC 902), EPEC O55:H7 and non-pathogenic E. coli (npEC) strains, under the effect of pH and temperature, according to central composite rotational design. T°C: temperature in Celsius degree; N0: initial cells (Log CFU/ ml). *Means and standard deviation of three analytical replicate.

Test	pH	T°C	STEC 902		STEC 2004		EPEC O55:H7		npEC
			Predicted	Observed*	Predicted	Observed*	Predicted	Observed*	Predicted	Observed*
1	4.5	10	4.9	5.4 ±0.06	5.1	5.5±0.08	5.0	5.1±0.95	5.4	5.2±0.15
2	4.5	30	7.0	7.1 ±0.07	7.3	7.3±0.11	6.7	7.1±1.32	7.2	7.2±0.10
3	6.5	10	4.5	4.5 ±0.05	4.6	4.6±0.07	5.6	5.1±0.96	5.8	5.1±0.09
4	6.5	30	7.6	7.2 ±0.02	7.8	7.5±0.11	7.7	7.5±1.41	7.7	7.4±0.06
5	4.0	20	6.7	6.3 ±0.13	7.2	6.9±0.10	5.7	5.4±1.01	6.8	6.8±0.11
6	7.0	20	7.0	7.2 ±0.13	7.1	7.3±0.11	6.8	7.2±1.35	7.4	8.0±0.00
7	5.5	6	3.4	3.0 ±0.00	3.3	3.0±0.00	4.8	5.1±0.95	4.6	5.2±0.11
8	5.5	35	7.0	7.2 ±0.13	7.1	7.3±0.11	7.6	7.4±1.39	7.2	7.2±0.07
9	5.5	20	7.1	7.0 ±0.13	7.3	7.2±0.10	7.4	7.5±0.13	7.4	7.4±0.02
10	5.5	20	7.1	7.1 ±0.13	7.3	7.4±0.11	7.4	7.5±0.13	7.4	7.4±0.02
11	5.5	20	7.1	7.3 ±0.13	7.3	7.3±0.11	7.4	7.3±0.13	7.4	7.4±0.02

• T°C: temperature in Celsius degree; N0: initial cells (Log CFU/ ml). *Means and standard deviation of three analytical replicate.

Microbiological analysis

After 26 h of incubation, wild-type STEC O157:H7 strains (STEC 902 and STEC 2004) and control strains (EPEC O55:H7 and npEC) in MpH-TSB, from each assay, were diluted in phosphate-buffered saline (PBS) and, 100 μL were sown on the surface of Trypticase Soy Agar (TSA; Biomerieux), incubated at 35 +1 °C for 18h. Colonies were counted and the results were expressed as log CFU/mL. A 500 μL aliquot of MpH-TSB with no turbidity was seeded in 5 mL of TSB, incubated at 35 +1 °C for 18-20 h, to recover viable cells.

Mathematical model

The polynomial mathematical model (Eq.1) was obtained by multiple regression analysis of the independent variables (pH and temperature) related to the E. coli count in log CFU/mL, using the Statistica 7® software. From the analysis of variance (ANOVA), factors (independent variables) with significance level of 5% (p<0.05) were selected to compose the model that describes the individual and combined effect of pH and temperature on E. coli in laboratory culture (Eq.1).

​$$Y={\alpha }_0+\sum _{i=1}^2{\alpha }_i{X}_i+\sum _{i=1}^2{\alpha }_{ii}{X}_i^2+\sum _{i<j}^2{\alpha }_{ij}{X}_i{X}_{ij}$$ (Eq.1)

where, Y is the cellular concentration (log CFU/mL), α0 is the intercept coefficient, αi (i = X1, pH and X2, temperature) is the coefficient of the first order model, αii is the coefficient of the quadratic model), and αij is the coefficient of multiplicative effect of the variables.

Model performance

The normality of the residual data of the mathematical model was verified by the Shapiro-Wilk test. The repetitions of the central point allowed to evaluate the lack-of-fit of the model, where values of p > 0.05 indicate the fit of the mathematical model. The model was represented graphically by the response surface. The performance and reliability of the model were measured by the coefficient of determination (R2). R2 values equal to or above 0.7 indicate the fit of the model to the data.

The performance of the mathematical model was observed by evaluating the adequacy indicators, as the accuracy factor (Af) (Eq.2), bias factor (Bf) (Eq.3), percent discrepancy (%D) (Eq.4) and percent bias (%B) (Eq.5), thus guaranteeing the confidence of the results. Values of Af and Bf equivalent to 1 represent perfect correlation between the values predicted by the mathematical model and observed (Ross 1996, Baranyi et al. 1999, Lee et al. 2014). Bf > 1 indicates fail-dangerous, where the predicted values are smaller than the observed ones. While Bf < 1 indicates fail-safe, where the predicted values are greater than the observed ones (Ross 1996). %D determines the discrepancy between model and observed values and %B determines how much the predicted values can describe the observed values. The perfection of the model is observed when %D and %B are equal to 0, which is, predicted values equal to the observed values (Baranyi et al. 1999).

​​$$A_f=\exp \left(\sqrt{\frac{\sum _{k=1}^m{(\ln f(x^{(k)})-\ln {\mu }^{(k)})}^2}{m}}\right)$$​ ​ (Eq.2)

$$B_f=\exp \left(\frac{\sum _{k=1}^m(\ln f(x^{(k)})-\ln {\mu }^{(k)})}{m}\right)$$​​(Eq.3)

where, Ln f(x) are the values predicted by the model, Ln μ the experimentally observed values, m is the number of tests.

​$$\mathrm{\%}D=(A_f-1)\cdot 100\mathrm{\%}$$ (Eq.4)

$$\mathrm{\%}B=\mathrm{s}\mathrm{g}\mathrm{n}(\ln B_f)\cdot (\exp |\ln B_f|-1)\cdot 100\mathrm{\%}$$ (Eq.5)

The relative error (RE) is the difference between the observed and predicted values in relation to the predicted values and was calculated to assess the acceptable prediction zone of the model (Eq.6). The acceptable prediction zone is between the minimum limit of -0.3 and maximum of 0.15. ER < 0 means that the prediction of the model response is greater than the observed values (fail-safe). On the other hand, ER > 0 means that the prediction model is smaller than the observed values (fail-dangerous) (Oscar 2005). Models with proportion of RE (pRE; number of assays with RE in the acceptable prediction zone/total number of assays) > 0.70 (70%) are considered accepted (Oscar 2005).

$$\mathrm{E}\mathrm{R}=\frac{V_{\text{observed}}-V_{\text{predicted}}}{V_{\text{predicted}}}$$​(Eq.6)

where, V observed are the values obtained experimentally by the CFU count (log CFU/mL) and V predicted are the values predicted by the model.

Validation of the mathematical model

Validation of the mathematical model was performed under seven additional conditions, within the minimum and maximum range of pH and temperature not determined by the CCRD (Table IV), in three independent repetitions. Additional tests were not used for the construction of the model.

Table IV
Predict and observed values of STEC O157:H7 (STEC 2004 strain), under the effect of pH and temperature, in the additional experimental conditions for validation of the mathematical model. T°C: temperature in Celsius degree.

Test	pH	T°C	Log CFU/mL
			Predict	Observed
1	5.0	30	8.0	8.0±0.23
2	5.0	15	5.9	5.9±0.17
3	5.0	25	8.5	8.1±0.25
4	6.0	20	8.3	7.8±0.24
5	6.0	25	8.6	8.3±0.25
6	6.0	30	8.1	8.2±0.24
7	6.0	35	8.4	7.8±0.24

• T°C: temperature in Celsius degree.

RESULTS AND DISCUSSION

Mathematical model

The Shapiro-Wilk test demonstrated the normal distribution of residuals data, with p = 0.7865 (EHEC 902), p = 0.3933 (EHEC 2004), p = 0.8107 (EPEC O55:H7) and p = 0.2755 (npEC). Then, the significant individual and combined effects (p<0.05) of the hurdles (pH and temperature) on the development of wild-type STEC O157:H7 and control strains in MpH-TSB were described by the models (STEC 902, Eq.7; STEC 2004, Eq.8; EPEC O55:H7; Eq.9; and npEC, Eq.10) obtained through multiple regression analysis (p <0.05). Also, the response surfaces plot that illustrate the equations are shown in Fig. 1.

Figure 1
Table SI.

$$\log {\mathrm{C}\mathrm{F}\mathrm{U}}_{\text{STEC902}}/\mathrm{m}\mathrm{L}=-1.82+0.38T-0.01T^2$$​(Eq.7)

$$\log {\mathrm{C}\mathrm{F}\mathrm{U}}_{\text{STEC2004}}/\mathrm{m}\mathrm{L}=1.45+0.40T-0.01T^2+0.03pH\cdot T$$ (Eq.8)

$$\log {\mathrm{C}\mathrm{F}\mathrm{U}}_{\text{EPEC055:H7}}/\mathrm{m}\mathrm{L}=-15.34+6.49pH-0.57p{H}^2+0.28T-0.01T^2$$ (Eq.9)

$$\log \mathrm{C}\mathrm{F}\mathrm{U}/{\mathrm{m}\mathrm{L}}_{\text{npEC}}=-1.96+1.56pH-0.13p{H}^2+0.35T-0.01T^2$$ (Eq.10)

where, log CFU/mL is the cellular concentration of each bacterial strain per mL of MpH-TSB; pH is the hydrogen potential; and T is the temperature (°C).

Model performance

The lack-of-fit of the model was evaluated and non-significant p-values (p> 0.05) indicated that the mathematical model adequately describes the functional relationship between STEC 902 (p= 0.051), STEC 2004 (p= 0.061) and EPEC O55:H7 (p= 0.069) strains and the independent variables (pH and temperature) (Table III; Table SI - Supplementary Material). The mathematical models of the STEC 902, STEC 2004 and EPEC O55:H7 strains presented R2 of 0.95, 0.98 and 0.94, respectively (Table III), indicating their reliability, that is, the model explains more than 90% of variability in response (log CFU/mL). For the npEC strain, the lack of fit of the model was significant (p = 0.001) (Table III; Table SI), however, the R² value of 0.88 (Table III) indicates the adequacy of the model, with an adjustment of 88% of the data. Furthermore, as will be described below, the ER also corroborates to indicate the adequacy of the npEC model. CCRD evaluates only the mathematical factors and does not consider the biological factor, which may limit the evaluation of the npEC strain.

Table III
Model performance indicators. R²: coefficient of determination; Af: accuracy factor; Bf: bias fac_tor (Bf); %D: percent discrepancy; %B: percent bias.

Indicator	STEC 902	STEC 2004	EPEC O55:H7	npEC
R²	0.95	0.98	0.94	0.88
Af	1.35	1.24	1.31	1.42
Bf	0.97	1.02	0.99	1.01
%D	34.89	24.31	30.91	41.91
%B	0.09	0.02	0.01	0.00
Lack of fit	0.051	0.061	0.069	0.001

• R²: coefficient of determination; Af: accuracy factor; Bf: bias fac_tor (Bf); %D: percent discrepancy; %B: percent bias.

Therefore, it is important to use more than one parameter to evaluate the performance of mathematical models. In addition, the graphical representation demonstrates the equivalence between predicted and observed values for all E. coli strains (Fig. 3). The evaluation of the graphical representation of the predicted and observed values considers the systematic deviations between these values (Mellefont et al. 2003).

Figure 3
Pareto diagram representing the individual effect of independent variables and their interaction on the behavior of (a) EHEC 902; (b) EHEC 2004; (c) EPEC O55: H7 and (d) non-pathogenic E. coli according to the Central Composite Rotational Design (CCRD). (1) pH (L): linear effect of pH. pH (Q): quadratic effect of pH. (2) T (L): linear effect of temperature. T (Q): quadratic effect of temperature. 1Lby2L: interaction between the linear effects of pH and temperature.

The Af values ranged from 1.24 to 1.42 (Table III), indicating that the observed values are distributed close to the prediction. The Bf values were very close to 1, with a variation between 0.97 and 1.02 (Table III), indicating almost perfect agreement between the observed and predicted values. The %D of the models ranged between 24.31 and 41.91 (Table III). %D shows the percentage of discrepancy between the model and observations. The %B ranged between 0 and 0.09 (Table III), indicating a positive global bias (b> 0). The adequacy indicators demonstrated that the predicted values could describe the observed values.

The performance of the mathematical model was also measured by quantifying the proportion of ER in the acceptable prediction zone between –0.3 and 0.15 (Oscar 2005). The ER of all assays performed with E. coli strains were within the prediction zone (Fig. 4), with pRE = 1.0. This result indicates that the predictions are accurate, and the biases are acceptable (Oscar 2005). Therefore, the evaluated performance indices indicate the adequacy of the model in describing the individual and combined effect of pH and temperature on wild-type (STEC 902 and STEC 2004) and control (EPEC O55:H7 and npEC) strains in the MpH-TSB.

Figure 4
Predicted versus observed log CFU/mL values of (a) EHEC 902, (b) EHEC 2004, (c) EPEC O55: H7 and (d) non-pathogenic E. coli, under the effect of pH and temperature (°C). Points above the line indicate fail-safe prediction values, points below the line indicate fail-dangerous prediction values.

Model validation

Although not identical, the development of STEC 902 and STEC 2004 at MpH-TSB was similar. Therefore, seven additional tests were conducted with the STEC 2004 strain under conditions not established by the CCRD (Table IV). Performance indices, Af and Bf, were used to assess how reliable the model can be for decision-making (Ross 1996, Baranyi et al. 1999, Lee et al. 2014, Jagannath & Tsuchido 2003). The performance of the model, including additional tests with the STEC 2004 (Table IV), resulted in an Af value of 1.43, indicating that the predicted values are close to the observed values (Ross 1996, Baranyi et al. 1999, Lee et al. 2014). The Bf of 0.9 indicated that the model is “fail-safe” (Ross 1996) and has an almost perfect goodness-of-fit, with predicted values slightly lower than observed values (Baranyi et al. 1999). In this sense, it can be confirmed that the mathematical model adequately describes the effect of pH and temperature on STEC 2004 in MpH-TSB.

Influence of pH and temperature on E. coli strains

The result of each of the 11 assays determined by the CCRD, corresponding to the count, in log CFU/mL (dependent variable), of the wild-type STEC (STEC 902, STEC 2004), EPEC O55:H7 and npEC strains, under the individual and combined effect of pH and temperature (independent variables) in the MpH-TSB culture medium, are described in Table II. From these observed results, the response surface graph was generated (Fig. 1), which illustrates the model equation for the STEC 902 (Eq.7), STEC 2004 (Eq.8), EPEC O55:H7 (Eq.9) and npEC (Eq.10) strains. The points corresponding to the lowest counts in log CFU/ml of the response variables (STEC 902, STEC 2004, EPEC O55:H7 and npEC) are found in the dark green zone of the response surface graph, while the points corresponding to the highest counts are found in the dark red zone (Fig. 1).

pH and temperature are stress-inducing factors that can affect bacterial development (Misiou et al. 2023). In this study, temperature showed both linear and quadratic effect on the development of STEC 902 (p<0.001), STEC 2004 (p<0.001 and p=0.04, respectively), EPEC O55:H7 (p<0.001 and p=0.01, respectively) and npEC (p<0.001) strains in MpH-TSB (Fig. 2). These results emphasize the critical role of temperature as a stressor affecting microbial behavior. Previous studies have similarly highlighted temperature as a primary environmental variable influencing the physiology and survival of STEC O157:H7, in laboratory culture media (Balamurugan et al. 2014) and food matrices such as spinach (Wang et al. 2015). The effect of temperature on E. coli cells is directly related to the conformation of the protein structure, the functioning of enzymes and the synthesis of RNA and DNA (Shapiro & Cowen 2012). Temperature is the main factor affecting microbial development (Misiou et al. 2023) and was an important stress factor, capable of affecting the behavior of different E. coli strains in laboratory culture medium (Fig. 1, 2). The increase in the number of cells of the E. coli strains, in relation to the initial inoculum (6.6 log CFU/ml), in MpH-TSB, can be observed at temperatures above 20°C (Fig. 1). E. coli is a mesophilic bacterium, with optimum development temperature around 37°C, its survival is impaired at temperatures below 10°C and above 45°C (37). Ukuku et al. (2013), described the reduction in the final number of E. coli K-12 (ATCC 23716) artificially inoculated into corn product and whey protein stored at 5°C. However, Frozi et al. (2015), observed a psychrotrophic behavior among STEC O157:H7 strains, which were able to multiply in Minas frescal cheese (pH 6.19, A_w 0.96) stored at 8°C. Hsin-Yi & Chou (2001) reported that the survival of E. coli in acidic juices (pH 3.2 to 3.6) was better at 7°C than at 25°C. The different substrates may influence the mesophilic or psychrotrophic behavior of STEC O157:H7 strains. Although the referenced studies provide valuable context, it is important to note that the temperature and pH ranges evaluated differ from those investigated in this study. Consequently, any comparisons must be interpreted with caution and within the scope of the experimental conditions validated here.

Figure 2
Surface response graphs of the behavior of (a) STEC 902, (b) STEC 2004, (c) EPEC O55:H7 and (d) non-pathogenic E. coli, under the effect of pH and temperature (°C).

However, the STEC 902 and STEC 2004 strains responded differently to the interaction of stress-inducing factors, pH and temperature. The STEC 2004 strain also suffered a significant effect from the pH and temperature interaction (p=0.04) (Fig. 2b). Although the interaction between pH and temperature does not have a significant effect on STEC 902 (Fig. 2a), it can be observed that the optimal conditions for the development of this pathogen (Fig. 1b) are similar to those of STEC 2004 (Fig. 1a), in a temperature range between 20 and 35°C, regardless of the pH value.

Unlike wild-type E. coli O157:H7 (STEC 902 and STEC 2004) (Fig. 2a, b), the control strains, EPEC O55:H7 (precursor of serogroup O157:H7) and npEC (non-pathogenic E. coli) also suffered a significant linear (p=0.01 and p<0.001, respectively) and quadratic (p=0.01) effect of pH in the MpH-TSB culture medium (Fig. 2c, d). pH exerts a greater influence on the behavior of non-STEC than in STEC O157:H7 (Fig. 2). In acid environments the pH inside the bacterial cell is reduced, leading to DNA denaturation or enzymes degradation. STEC O157:H7 is able to adapt to acid environments, and this characteristic was identified as a virulence determinant of this pathogen (Elsas et al. 2011, Foster 2004). Some phenomena, such as rpoS gene expression and the use of the arginine and glutamate dependent system may explain the ability of STEC O157:H7 to adapt in acidic environments (Foster 2004, Bae & Lee 2017). Beyond those mechanisms, acid resistance could be induced for pyruvate in O157:H7 (Wu et al. 2014). Several studies have reported the survival of STEC O157:H7 in acid environments (Bjornsdottir et al. 2006, Stopforth et al. 2007, Skandamis & Mychas 2000) being described its involvement in outbreaks with acid food (Besser et al. 1993, Weagant et al. 1994, McCarthy 1996, Cody et al. 1999). Glass et al. (1992), observed growth of STEC O157:H7 in TSB with pH between 4.0 and 4.5, adjusted with HCl.

In this study HCl was used for the medium acidification. The type of acid used in the solution (Kim et al. 2015, Samelis et al. 2002) strongly influences the behavior of STEC O157:H7. Bae & Lee (2017) observed that salt associated with acetic acid increases the acid resistance of STEC O157:H7. Shayanfar et al. (2018) observed the formation of metabolites that induced acid resistance in O26:H11 subjected to pH 3.6 modified with HCl.

Despite being the precursor serotype of STEC O157:H7 (Kyle et al. 2012), the behavior of the EPEC O55:H7 strain in the MpH-TSB culture medium (Fig. 1c), subjected to different combinations of pH and temperature, was closer to the behavior observed in the npEC strain (Fig. 1d) than that observed in the STEC O157:H7 strains, under the same conditions (Fig. 1a, b).

The STEC O157:H7 (STEC 902 and STEC 2004 strains) showed a different response to stress induced by the interaction of pH and temperature (Fig. 2a, b). This fact may be related to the expression or not of proteins at different temperature and pH levels. Considering the independent variables pH and temperature, and with due caution regarding the interpretation of ranges beyond those validated in our study, it is worth noting that Balamurugan et al. (2014), also observed the significant effect of the interaction between pH and temperature on the behavior of STEC O157:H7 in TSB incubated at 5, 20 and 30°C, in the pH ranges of 4.8, 5.4 and 6.0, adjusted with lactic acid. Similarly, Clavero & Beuchat (1996), observed a significant effect on the interaction of different levels of temperature (5, 20 and 30°C), pH (6.0, 5.4 and 4.8) and Aw (0.99, 0. 95 and 0.90) on five STEC O157:H7 strains in TSB. The ability of STEC O157:H7 to tolerate a wide range of pH and temperature further emphasizes its adaptive capacity as a foodborne pathogen. This adaptability is critical to understanding its behavior in various food matrices and under different storage and processing conditions.

CONCLUSIONS

The mathematical model showed goodness-of-fit, being able to adequately describe the individual and combined effect of pH and temperature on the behavior of STEC O157:H7, EPEC O55:H7 and npEC strains in MpH-TSB. The evaluation indexes indicated that the mathematical model tends to fail-safe predictions. The application and analysis of the model allows us to conclude that the individual and combined effect of pH and temperature on the wild-type STEC strains in the laboratory culture medium differed from the effect on the control E. coli strains. Temperature was the stress-inducing factor that most influenced the development of STEC O157:H7 strains in MpH-TSB. The development of STEC O157:H7 strains in laboratory culture medium is favored at temperatures above 20°C, with a gradual reduction in the number of cells at temperatures below 10°C, regardless of the pH value. Non-STEC are more susceptible to the effect of pH than STEC. pH is a barrier that can be used by the food industry to control microbial development. However, this barrier is not sufficient to control the development of STEC O157:H7. The results presented encourage future studies to evaluate the effect of pH and temperature on other STEC serogroups and thus be able to optimize the use of these barriers, widely used in food preservation, thus ensuring food safety.

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

SUPPLEMENTARY MATERIAL

Table SI.

Acknowledgements

This study was funded by the Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ, Brasil), finance code E-26/210.068/2021 and E-26/210.915/2021; Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, Brasil); and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES, Brasil), finance code 001.

References

• AKKERMANS S, FERNANDEZ EN, LOGIST F & VAN IMPE JF. 2017. Introducing a novel interaction model structure for the combined effect of temperature and pH on the microbial growth rate. Int J Food Microbiol 240: 85-96. https://doi.org/10.1016/j.ijfoodmicro.2016.06.011.
  » https://doi.org/10.1016/j.ijfoodmicro.2016.06.011
• AMIN MA, HASHEM HR, EL-MAHALLAWY HS, ABDELRAHMAN AA, ZAKI HM & AZAB MM. 2022. Characterization of enterohemorrhagic Escherichia coli from diarrhoeic patients with particular reference to production of Shiga-like toxin. Microb Pathog 166: 105538. https://doi.org/10.1016/j.micpath.2022.105538.
  » https://doi.org/10.1016/j.micpath.2022.105538
• ARICAN A & ANDIC S. 2011. Survival of E. coli O157:H7 in yogurt incubated until two different pH values and stored at 4°C. Kafkas Univ Vet Fak Derg 17: 537-542.
• BAE YM & LEE SY. 2017. Effect of salt addition on acid resistance response of Escherichia coli O157: H7 against acetic acid. Food Microbiol 65: 74-82. https://doi.org/10.1016/j.fm.2016.12.02.
  » https://doi.org/10.1016/j.fm.2016.12.02
• BALAMURUGAN S, AHMED R & GAO A. 2014. Survival of Shiga toxin-producing Escherichia coli in broth as influenced by pH, Aw and temperature. Lett Appl Microbiol 60: 342-346. https://doi.org/10.1111/lam.12375.
  » https://doi.org/10.1111/lam.12375
• BARANYI J, PIN C & ROSS T. 1999. Validation and comparing predictive models. Int J Food Microbiol 48: 159-166. https://doi.org/10.1016/s0168-1605(99)00035-5.
  » https://doi.org/10.1016/s0168-1605(99)00035-5
• BESSER RE, LETT SM, WEBER JT, DOYLE MP, BARRE TJ, WELLS JG & GRIFFIN PM. 1993. An outbreak of diarrhea and uremic syndrome from Escherichia coli O157:H7 in fresh-pressed apple cider. J Am Med Assoc 269: 2217-2220. https://doi.org/10.1001/jama.1993.03500170047032.
  » https://doi.org/10.1001/jama.1993.03500170047032
• BISWAS S & PANDEY P. 2017. The relative efficacy of thermal and acidification stresses on the survival of E. coli O157:H7, Salmonella, and Listeria monocytogenes in ground beef. J Food Res 6: 74-84. https://doi.org/10.5539/jfr.v6n3p74.
  » https://doi.org/10.5539/jfr.v6n3p74
• BJORNSDOTTIR K, BREIDT-JUNIOR F & MCFEETERS RF. 2006. Protective effects of organic acids on survival of Escherichia coli O157:H7 in acidic environments. Appl Environ Microbiol 72: 660-664. https://doi.org/10.1128/AEM.72.1.660-664.2006.
  » https://doi.org/10.1128/AEM.72.1.660-664.2006
• CEPL J, BLAHUSKOVA A, NEUBAUER Z & MARKOS A. 2016. Variations and heredity in bacterial colonies. Commun Integr Biol 9: e1261228. http://dx.doi.org/10.1080/19420889.2016.1261228.
  » https://doi.org/10.1080/19420889.2016.1261228
• CLAVERO MRS & BEUCHAT LR. 1996. Survival of Escherichia coli O157:H7 in broth and processed salami as influenced by pH, water activity, and temperature and suitability of media for its Recovery. Appl Environ Microbiol 62: 2735-2740. https://doi.org/10.1128/aem.62.8.2735-2740.1996.
  » https://doi.org/10.1128/aem.62.8.2735-2740.1996
• CODY SH ET AL. 1999. An outbreak of Escherichia coli O157:H7 infection from unpasteurized commercial apple juice. Ann Intern Med 130: 202-209. https://doi.org/10.7326/0003-4819-130-3-199902020-00005.
  » https://doi.org/10.7326/0003-4819-130-3-199902020-00005
• COULOMBE G, CATFORD A, MARTINEZ-PEREZ A & BUENAVENTURA E. 2020. Outbreaks of Escherichia coli O157: H7 infections linked to Romaine lettuce in Canada from 2008 to 2018: an analysis of food safety context. J Food Protect 83(8): 1444-1462. https://doi.org/10.4315/JFP-20-029.
  » https://doi.org/10.4315/JFP-20-029
• CROXEN MA, LAW RJ, SCHOLZ R, KEENEY KM, WLODARSKA M & FINLAY BB. 2013. Recent advances in understanding enteropathogenic Escherichia coli Clin Microbiol Rev 26: 822-880. https://doi.org/10.1128/CMR.00022-13.
  » https://doi.org/10.1128/CMR.00022-13
• ELSAS JD, SEMENOV AV, COSTA R & TREVORS JT. 2011. Survival of Escherichia coli in the environment: fundamental and public health aspects. ISME J 5: 173-183. https://doi.org/10.1038/ismej.2010.80.
  » https://doi.org/10.1038/ismej.2010.80
• FOSTER JW. 2004. Escherichia coli acid resistance: tales of an amateur acidophile. Nature Ver Microbiol 2: 898-907. https://doi.org/10.1038/nrmicro1021.
  » https://doi.org/10.1038/nrmicro1021
• FROZI JB, DOMINGUES JR, ESPER LMR, ROSA JMC, SILVA ALSC & GONZALEZ AGM. 2015. Survival of Shiga toxin-producing Escherichia coli O157:H7 in minas frescal cheese. Food Sci Technol 35: 108-114. https://dx.doi.org/10.1590/1678-457X.6528.
  » https://doi.org/10.1590/1678-457X.6528
• GLASS KA, LOEFFELHOLZ JM, FORD JP & DOYLE MP. 1992. Fate of Escherichia coli O157:H7 as affected by pH or sodium chloride and in fermented, dry sausage. Appl Environ Microbiol 58: 2513-2516. https://doi.org/10.1128/aem.58.8.2513-2516.1992.
  » https://doi.org/10.1128/aem.58.8.2513-2516.1992
• GONZALEZ AGM & CERQUEIRA AMF. 2019. Shiga toxin-producing Escherichia coli in the animal reservoir and food in Brazil. J Appl Microbiol 128: 1568-1582. https://doi.org/10.1111/jam.14500.
  » https://doi.org/10.1111/jam.14500
• GONZALEZ AGM, CERQUEIRA AMF, GUTH BEC, COUTINHO CA, LIBERAL MHT, SOUZA RM & ANDRADE JRC. 2016. Serotypes, virulence markers and cell invasion ability of Shiga toxin-producing Escherichia coli strains isolated from healthy dairy cattle. J Appl Microbiol 121: 1130-1143. https://doi.org/10.1111/jam.13230.
  » https://doi.org/10.1111/jam.13230
• HASSAN R ET AL. 2019. A multistate outbreak of E. coli O157: H7 infections linked to soy nut butter. Pediatrics 144(4): e20183978. https://doi.org/10.1542/peds.2018-3978.
  » https://doi.org/10.1542/peds.2018-3978
• HSIN-YI C & CHOU C. 2001. Acid adaptation and temperature effect on the survival of E. coli O157:H7 in acidic fruit juice and lactic fermented milk product. Int J Food Microbiol 70: 189-195. https://doi.org/10.1016/s0168-1605(01)00538-4.
  » https://doi.org/10.1016/s0168-1605(01)00538-4
• IRINO K, VAZ TMI, KATO MAMF, NAVES ZVF, LARA RR, MARCO MEC, ROCHA MMM, MOREIRA TP, GOMES TAT & GUTH BEC. 2002. O157:H7 Shiga toxin-producing Escherichia coli strains associated with sporadic cases of diarrhea in São Paulo, Brazil. Emerg Infect Dis 8: 446-447. https://doi.org/10.3201/eid0804.010490.
  » https://doi.org/10.3201/eid0804.010490
• JAGANNATH A & TSUCHIDO T. 2003. Validation of a polynomial regression model: the thermal inactivation of Bacillus subtilis spores in milk. Lett Appl Microbiol 37: 399-404. https://doi.org/10.1046/j.1472-765x.2003.01416.x.
  » https://doi.org/10.1046/j.1472-765x.2003.01416.x
• KIM JS, LEE MS & KIM JH. 2020. Recent Updates on Outbreaks of Shiga Toxin-Producing Escherichia coli and Its Potential Reservoirs. Front Cell Infect Microbiol 4: 273. https://doi.org/10.3389/fcimb.2020.00273.
  » https://doi.org/10.3389/fcimb.2020.00273
• KIM C, WILKINS K, BOWERS M, WYNN C & NDEGWA E. 2018. Influence of pH and temperature on growth characteristics of leading foodborne pathogens in a laboratory medium and select food beverages. Austin Food Sci 3: 1031.
• KIM G-H, BREIDT F, FRATAMICO P & OH D-H. 2015. Acid resistance and molecular characterization of Escherichia coli O157:H7 and different non-O157 Shiga toxin-producing E. coli serogroups. J Food Sci 80: 2257-2264. https://doi.org/10.1111/1750-3841.12996.
  » https://doi.org/10.1111/1750-3841.12996
• KIM MS, LIM SH, KIM JH, HA IS, CHEONG HI & KANG HG. 2020. Rapid Resolution of Atypical Hemolytic Uremic Syndrome by Eculizumab Treatment. Childhood Kidney Dis 24(2): 138-142.
• KOUTSOUMANIS KP & SOFOS JN. 2004. Comparative acid stress response of Listeria monocytogenes, Escherichia coli O157:H7 and Salmonella Typhimurium after habituation at different pH conditions. Lett Appl Microbiol 38: 321-321. https://doi.org/10.1111/j.1472-765x.2004.01491.x.
  » https://doi.org/10.1111/j.1472-765x.2004.01491.x
• KYLE JL ET AL. 2012. Escherichia coli serotype O55:H7 diversity supports parallel acquisition of bacteriophage at Shiga toxin phage insertion sites during evolution of the O157:H7 Lineage. J Bacteriol 194: 1885-1896. https://doi.org/10.1128/JB.00120-12.
  » https://doi.org/10.1128/JB.00120-12
• LEE YJ, JUNG BS, YOON HJ, KIM K-T, PAIK H-D & LEE J-Y. 2014. Predictive model for the growth kinetics of Listeria monocytogenes in raw pork meat as a function of temperature. Food Contr 44: 16-21. https://doi.org/10.1016/j.foodcont.2014.03.024.
  » https://doi.org/10.1016/j.foodcont.2014.03.024
• LEISTNER L. 1992. Food preservation by combined methods. Food Res Int 25: 151-158. https://doi.org/10.1016/0963-9969(92)90158-2.
  » https://doi.org/10.1016/0963-9969(92)90158-2
• LEISTNER L. 2000. Basic aspects of food preservation by hurdle technology. Int J Food Microbiol 55: 181-186. https://doi.org/10.1016/S0168-1605(00)00161-6.
  » https://doi.org/10.1016/S0168-1605(00)00161-6
• LUNA S. 2018. Outbreak of E. coli O157: H7 infections associated with exposure to animal manure in a rural community—Arizona and Utah, June–July 2017. Morb Mortal Wkly Rep 67: 659-662. https://doi.org/10.15585/mmwr.mm6723a2.
  » https://doi.org/10.15585/mmwr.mm6723a2
• MCCARTHY M. 1996. E. coli O157:H7 outbreak in USA traced to apple juice. Lancet 348: 1299-1299. https://doi.org/10.1016/S0140-6736(05)65758-3.
  » https://doi.org/10.1016/S0140-6736(05)65758-3
• MELLEFONT LA, MCMEEKIN TA & ROSS T. 2003. Performance evaluation of a model describing the effects of temperature, water activity, pH and lactic acid concentration on the growth of Escherichia coli Int J Food Microbiol 82: 45-58. https://doi.org/10.1016/S0168-1605(02)00253-2.
  » https://doi.org/10.1016/S0168-1605(02)00253-2
• MISIOU O, ELLOUZE M & KOUTSOUMANIS K. 2023. Cardinal models to describe the effect of temperature and pH on the growth of Anoxybacillus flavithermus & Bacillus licheniformis. Food Microbiol 112: 104230. https://doi.org/10.1016/j.fm.2023.104230.
  » https://doi.org/10.1016/j.fm.2023.104230
• NEWELL DG & LA RAGIONE RM. 2018. Enterohaemorrhagic and other Shiga toxin-producing Escherichia coli (STEC): Where are we now regarding diagnostics and control strategies? Transbound Emerg Dis 65: 49-71. https://doi.org/10.1111/tbed.12789.
  » https://doi.org/10.1111/tbed.12789
• OSCAR TP. 2005. Validation of lag time and growth rate models for Salmonella typhimurium: acceptable prediction zone method. J Food Sci 70: 129-136. https://doi.org/10.1111/j.1365-2621.2005.tb07103.x.
  » https://doi.org/10.1111/j.1365-2621.2005.tb07103.x
• ROSS T. 1996. Indices for performance evaluation of predictive models in food microbiology. J Appl Bacteriol 81: 501-508.
• SAMELIS J, SOFOS JN, KENDALL PA & SMITH GC. 2002. Effect of acid adaptation on survival of Escherichia coli O157:H7 in meat decontamination washing fluids and potential effects of organic acid interventions on the microbial ecology of the meat plant environment. J Food Protect 65: 33-40. https://doi.org/10.4315/0362-028x-65.1.33.
  » https://doi.org/10.4315/0362-028x-65.1.33
• SHAPIRO RS & COWEN LE. 2012. Thermal control of microbial development and virulence: molecular mechanisms of microbial temperature sensing. mBio 3: e00238-12. https://doi.org/10.1128/mBio.00238-12.
  » https://doi.org/10.1128/mBio.00238-12
• SHAYANFAR S, BROUMAND A & PILLAI SD. 2018. Acid stress induces differential accumulation of metabolites in Escherichia coli O26: H11. J Appl Microbiol 125(6): 1911-1919. https://doi.org/10.1111/jam.14081.
  » https://doi.org/10.1111/jam.14081
• SILVA CB, FERREIRA LM, LIMA AR, ARAÚJO KG, SOUZA RM, FONSECA ABM & GONZALEZ AG. 2023. Microbiological quality and cultivable bacterial community of fresh and ripened Minas cheeses made from raw and pasteurised milk. Int Dairy J 143: 105662. https://doi.org/10.1016/j.idairyj.2023.105662.
  » https://doi.org/10.1016/j.idairyj.2023.105662
• SKANDAMIS PN & MYCHAS G-J. 2000. Development and evaluation of a model predicting the survival of Escherichia coli O157:H7 NCTC 12900 in homemade eggplant salad at various temperatures, pHs, and oregano essential oil concentrations. Appl Environ Microbiol 6: 1646-1653. https://doi.org/10.1128/aem.66.4.1646-1653.2000.
  » https://doi.org/10.1128/aem.66.4.1646-1653.2000
• SOUZA RL, CARVALHAES JTA, NISHIMURA LS, ANDRADE MC & GUTH BEC. 2011. Hemolytic uremic syndrome in pediatric intensive care units in São Paulo, Brazil. Open Microbiol J 5: 76-82. https://doi.org/10.2174/1874285801105010076.
  » https://doi.org/10.2174/1874285801105010076
• STACHELSKA MA & CHARMAS R. 2018. The prediction of Escherichia coli O157:H7 decline in tryptic soy broth with yeast extract treatment with different temperatures and acidity values using a pathogen modeling program. Polish J Appl Sci 4: 51-54. http://doi.org/10.34668/PJAS.2018.4.2.02.
  » https://doi.org/10.34668/PJAS.2018.4.2.02
• STOPFORTH JD, SKANDAMIS PN, GEORNARAS I & SOFOS JN. 2007. Acid tolerance of acid-adapted and nonacid-adapted Escherichia coli O157:H7 strains in beef decontamination runoff fluids or on beef tissue. Food Microbiol 24: 530-538. https://doi.org/10.1016/j.fm.2006.05.007.
  » https://doi.org/10.1016/j.fm.2006.05.007
• UKUKU DO, MUKHOPADHYAY S & ONWULATA C. 2013. Effect of storage temperature on survival and recovery of thermal and extrusion injured Escherichia coli K-12 in whey protein concentrate and corn meal. Foodborne Pathog Dis 10: 62-68. https://doi.org/10.1089/fpd.2012.1269.
  » https://doi.org/10.1089/fpd.2012.1269
• VALERO A, RODRÍGUEZ M, CARRASCO E, PÉREZ-RODRÍGUEZ F, GARCÍA-GIMENO RM & ZURERA G. 2010. Studying the growth boundary and subsequent time to growth of pathogenic Escherichia coli serotypes by turbidity measurements. Food Microbiol 27: 819-828. https://doi.org/10.1016/j.fm.2010.04.016.
  » https://doi.org/10.1016/j.fm.2010.04.016
• WANG J & OH D-H. 2012. Effect of temperature and relative humidity on growth behaviour of Escherichia coli O157:H7 on spinach using response surface methodology. J Food Saf 32: 296-304. https://doi.org/10.1111/j.1745-4565.2012.00380.x.
  » https://doi.org/10.1111/j.1745-4565.2012.00380.x
• WANG X, EVY L, BOECK E, DEVLIEGHERE F, GEERAERD A & UYTTENDAELE M. 2015. Growth and inactivation of Salmonella enterica and Listeria monocytogenes in broth and validation in ground pork meat during simulated home storage abusive temperature and home pan-frying. Front Microbiol 6: 1161. https://doi.org/10.3389/fmicb.2015.01161.
  » https://doi.org/10.3389/fmicb.2015.01161
• WEAGANT SD, BRYAN JL & BARK DH. 1994. Survival of Escherichia coli O157:H7 in mayonnaise and mayonnaise-based sauces at room and refrigerated temperatures. J Food Protect 57: 629-631. https://doi.org/ 10.4315/0362-028X-57.7.629.
  » https://doi.org/10.4315/0362-028X-57.7.629
• WU J, LI Y, CAI Z & JIN Y. 2014. Pyruvate-Associated Acid Resistance in Bacteria. Appl Environ Microbiol 80: 4108-4113. htpps://doi.org/10.1128/AEM.01001-14.
  » https://doi.org/10.1128/AEM.01001-14
• ZHOU Z ET AL. 2010. Derivation of Escherichia coli O157:H7 from its O55:H7 precursor. PLoS ONE5: e8700. https://doi.org/10.1371/journal.pone.0008700
  » https://doi.org/10.1371/journal.pone.0008700

Publication Dates

• Publication in this collection
  19 May 2025
• Date of issue
  2025