The effect of pH on the growth rate of Bacillus cereus sensu lato: Quantifying strain variability and modelling the combined effects of temperature and pH
Introduction
The determination of cardinal growth parameters (i.e. the minimum, optimum and maximum values of the most important environmental variables) is one of the essential steps for the characterization of a microbial strain. Their knowledge is a pre-requisite to use user-friendly modelling tools, such as Sym'Previus (www.symprevius.eu), that help manufacturers evaluate food-related microbial risks, determine microbiological shelf life and optimize industrial processes. The differences in microbial responses to identical environmental conditions among different strains of the same species is an important source of variability that has been analysed by several microbiological studies (Lianou and Koutsoumanis, 2011; Whiting and Golden, 2002). This is especially true for the Bacillus cereus group (or B. cereus sensu lato), which exhibits a broad diversity, making quantitative risk assessments more difficult. The phylogenetic structure of the B. cereus group was described by Guinebretière et al., 2008, Guinebretière et al., 2010. Seven different phylogenetic groups were established, showing significant differences in their responses to temperature as well as in their ability to cause food poisoning.
A number of quantitative studies for pathogens such as Salmonella and Listeria monocytogenes include the effects of pH, water activity or lactic acid (e.g. Aryani et al., 2015; Lianou and Koutsoumanis, 2011). For B. cereus, the main focus has been on temperature, as in Afchain et al. (2008), Le Marc et al. (2021) and Ellouze et al. (2021). Fewer studies have considered the effects of pH. They include Biesta-Peters et al. (2010), Baril et al. (2012) and Carlin et al. (2013). However, these have focused either on a single strain or on a limited number of strains (e.g. only 2 strains for each phylogenetic group in Carlin et al., 2013). Although the cardinal pH values of B. cereus do not seem to exhibit the same large variability that has been observed for cardinal temperatures, some differences have yet been highlighted: for instance, Carlin et al. (2013) found that Group VI strains grow poorly at low pH levels, while the other strains grow better. However, there are currently no quantitative studies that would provide robust estimates for the distributions of cardinal pH values for the phylogenetic groups of B. cereus. This is of particular concern for Group III, that comprises a number of emetic strains, for which food poisoning risk is the highest (Guinebretière et al., 2010).
In the Bacillus cereus studies mentioned above, the effect of pH on the maximum specific growth rate (μmax) was described by means of the cardinal pH model (Rosso et al., 1995) which has the advantage of using parameters with biological significance through which it is easier to define the growth boundaries of the studied micro-organism. It was identified by Biesta-Peters et al. (2010) as the best performing model for B. cereus F4810/72, among the 10 studied models. However, the goodness of fit obtained with the cardinal pH model was achieved with an estimation of the pHmax parameter of 19.16, which of course cannot represent the maximum pH for growth. The reason behind this estimation lies in the structure of the cardinal model that assumes a close-to-linear relationship between pH and μmax at suboptimal pH values. This was not justified by experimental observations: instead, the growth rate data from Biesta-Peters et al. (2010) highlight a “plateau” in the μmax vs. pH curve between pH 5.8 and 7.0 (i.e. pH had little effect on μmax for this range of pH), preceded by a rapid acceleration as the pH increased from pHmin to ca. pH 5.5. When fitting the cardinal model, the estimate of pHmax (that, biologically, represents the boundary of growth region and, numerically, acts as a shape parameter) increased artificially toward a high value in order to accommodate this “plateau”. This pHmax value therefore lacks biological meaning and while maximizing the goodness of fit, the pHmax parameter became so high that it could not possibly represent the growth boundary, as only pH values ≤14 have biological meaning. As an alternative, Aryani et al. (2015) used a reparameterization of a more versatile model (Presser, 2001) that had not been tested by Biesta-Peters et al. (2010). It has been successfully used for Listeria monocytogenes and Lactiplantibacillus plantarum by Aryani et al., 2015, Aryani et al., 2016 but, to our knowledge, not evaluated for the B. cereus group.
Besides, some authors identified that the estimate of the pHmin parameter could be influenced by the temperature. Martinez-Rios et al. (2019) determined the pHmin-values of Listeria monocytogenes at different temperatures between 5 and 37 °C. They found that temperature had a marked effect on the pHmin estimates, which decreased from 4.9 at 5 °C to 4.3 at 15–25 °C and then increased to 4.7 at 37 °C. This contrasts with the usual assumption that the lowest minimum pHmin-value is observed at the optimum temperature for growth (Topt), which is close to 37 °C for L. monocytogenes (Augustin et al., 2005; Couvert et al., 2010; Nunes Silva et al., 2020). Experimental data from Tienungoon et al. (2000) and Le Marc et al. (2002) are in line with Martinez-Rios et al. (2019) and suggest that the lowest pHmin is obtained at a temperature (denoted by TR) between 20 and 25 °C, and not 37 °C. Interestingly, the TR-value falls in the same temperature range as the parameter Tmes (the “Temperature of Maximum Enzyme Stability”, i.e. the temperature at which protein denaturation is minimal) that was introduced by Ratkowsky et al. (2005). For Listeria monocytogenes, the parameter Tmes was indeed found to be in the range 23–25 °C (Ratkowsky et al., 2005). To date, the effect of temperature on the estimation of the pHmin parameter has not been investigated for the B. cereus group.
Hurdles, such as low pH and low temperature, are often used to ensure food stability and safety Gamma hypothesis (Zwietering et al., 1992), where the combined effects of these hurdles on μmax can be obtained in a multiplicative manner to produce the overall microbial inhibition. Based on this assumption, an existing temperature model based on a Gamma term for temperature, can easily be extended to incorporate the effect of pH by multiplying the temperature and pH Gamma terms. Hence, the validity of the Gamma hypothesis has often been discussed in the literature. There have been studies showing that ‘interactions’ beyond the multiplicative effects occur, mainly near the growth limits when different hurdles are combined. Some authors proposed to take into account these interactive effects through the inclusion of a new synergy term (e.g. Le Marc et al., 2002; Martinez-Rios et al., 2016; Mejlholm and Dalgaard, 2007, Mejlholm and Dalgaard, 2009). Other studies have been sceptical about the need to model the synergistic effects. For instance, Biesta-Peters et al. (2010) did not find synergistic effects between pH and organic acids, at 30 °C, on the μmax values of B. cereus. So far, the existence of synergy between pH and temperature has not been studied for B. cereus.
The overall objective of this work is to evaluate the effects of pH alone or in combination with storage temperature on the maximum growth rate of B. cereus sensu lato (with a focus on mesophilic strains). The study was conducted in two different phases. Phase 1 aims to evaluate different Gamma terms of pH for experiments run at 30 °C and identify the best performing equation. The strain variability in response to pH was also evaluated. In phase 2, the combined effects of temperature and pH on the growth and growth boundaries of B. cereus were investigated. The effect of temperature on the pHmin-values of B. cereus was examined and the performance of three predictive models for the combined effects of temperature and pH were assessed. Finally, new equations were proposed for the growth/no growth boundaries.
Section snippets
Bacterial strains and strain preparation
In total, 18 strains of B. cereus were used in this study: 16 mesophilic strains (13 strains from Group III and 3 strains from Group IV) and 2 psychrotrophic strains (1 strain from Group II and 1 strain from Group V). These strains were isolated from food products or faeces (Table 1) and were filed in the Nestlé culture collection. To prepare the inoculum for each strain, one cryobead stored at −80 °C was transferred in 10 mL of Brain Heart Infusion (BHI) tube (Thermofisher Diagnostics CM1135B)
Performance of the ‘growth rate vs. pH’ models
Growth was observed at pH levels as low as 4.70 for all strains except for Group V strain B600. This is consistent with results from Carlin et al. (2013) who observed B. cereus growth at pH levels 4.60–4.71 for mesophilic strains and 4.69–4.76 for psychrotrophic strains (Groups II and V). The ‘μmax vs pH’ curves obtained at 30 °C exhibit similar shapes for all the 18 strains. The μmax-values increase rapidly with increasing pH until it reaches a quasi-plateau at pH levels ca 5.7–6.0. Examples
Discussion
The classical cardinal pH model assumes a close-to-linear evolution of the growth rate as a function of suboptimal pH. However, the B. cereus data from this study show a different pattern (a rapid increase in μmax as pH increases from pHmin to ca. pH 5.5 followed by a near-“plateau”) and the cardinal pH model did not perform well. Due to its shape parameter, the Presser-Aryani type models have the flexibility to describe different shapes for the μmax vs. pH curves (as shown in Presser, 2001)
Funding
This work was financially supported by Nestlé Research Centre, Lausanne, Switzerland.
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
Yvan Le Marc thanks Tom Ross for helpful discussions around the mechanistic differences between Tmes and Topt.
References (36)
- et al.
Quantifying strain variability in modeling growth of Listeria monocytogenes
Int. J. Food Microbiol.
(2015) - et al.
Modelling the growth rate of Listeria monocytogenes with a multiplicative type model including interactions between environmental factors
Int. J. Food Microbiol.
(2000) - et al.
Sporulation boundaries and spore formation kinetics of Bacillus spp. as a function of temperature, pH and aw
Food Microbiol.
(2012) - et al.
Variation of cardinal growth parameters and growth limits according to phylogenetic affiliation in the Bacillus cereus group. Consequences for risk assessment
Food Microbiol.
(2013) - et al.
Validation of a stochastic modelling approach for Listeria monocytogenes growth in refrigerated foods
Int. J. Food Microbiol.
(2010) - et al.
Modelling the growth kinetics of Listeria as a function of temperature, pH and organic acid concentration
Int. J. Food Microbiol.
(2002) - et al.
A stochastic approach for modelling the effects of temperature on the growth rate of Bacillus cereus sensu lato
Int. J. Food. Microbiol.
(2021) - et al.
Influence of temperature, pH and NaCl concentration on the maximal growth rate of Brochothrix thermosphacta and a bioprotective bacteria Lactococcus piscium CNCM I-4031
Food Microbiol.
(2012) - et al.
A stochastic approach for integrating strain variability in modeling Salmonella enterica growth as a function of pH and water activity
Int. J. Food Microbiol.
(2011) - et al.
Modelling and predicting growth of psychrotolerant pseudomonads in milk and cottage cheese
Int. J. Food Microbiol.
(2016)
Predictive microbiology theory and application: is it all about rates?
Food Control
Modeling and predicting the growth boundary of Listeria monocytogenes in lightly preserved seafood
J. Food Prot.
Development and validation of an extensive growth and growth boundary model for Listeria monocytogenes in lightly preserved and ready-to-eat shrimp
J. Food Prot.
Effect of pH on Thermoanaerobacterium thermosaccharolyticum DSM 571 growth, spore heat resistance and recovery
Food Microbiol.
Cardinal parameter meta-regression models describing Listeria monocytogenes growth in broth
Food Res. Int.
Predicting the combinatorial effects of water activity, pH and organic acids on Listeria growth in media and complex food matrices
Food Microbiol.
Unifying temperature effects on the growth rate of bacteria and the stability of globular proteins
J. Theor. Biol.
An unexpected correlation between cardinal temperatures of microbial growth highlighted by a new model
J. Theor. Biol.
Cited by (10)
Inter-strain variability on the cardinal parameters (pH and A<inf>w</inf>) of clinical and food isolates of Listeria monocytogenes using turbidimetric measurements
2024, International Journal of Food MicrobiologyMonitoring the effect of pH on the growth of pathogenic bacteria using electrical impedance spectroscopy
2023, Results in EngineeringEffects of carbon dioxide and oxygen on the growth rate of various food spoilage bacteria
2023, Food MicrobiologyProduction of poly-gamma-glutamic acid and spore formation in Bacillus velezensis 83 are affected by acidic growing conditions and glucose uptake rate
2024, Journal of Chemical Technology and Biotechnology