Given the importance of strain variability to predictive microbiology and risk assessment, this study aimed to quantify the magnitude of strain variability in growth and thermal inactivation kinetics behaviors after acid adaptation. Thirty-three Listeria monocytogenes strains were exposed to acid-adapted tryptic soy broth supplemented with yeast extract (TSBYE; pH 5.5) and non–acid-adapted TSBYE (pH 7.0) for 20 h. Next, the growth parameters of these adapted and nonadapted strains that grew in nonbuffered TSBYE at 25°C were estimated. The tested strains were inactivated at 60°C in nonbuffered broth to obtain the heat resistance parameters. The results revealed that strain variability was present in the growth and thermal inactivation characteristics. The maximum specific growth rate ranged from 0.21 to 0.44 h−1 and from 0.20 to 0.45 h−1 after acid and nonacid adaptation, respectively. The lag times were from 0.69 to 2.56 h and from 0.24 to 3.36 h for acid-adapted and non–acid-adapted cells, respectively. The apparent D-values at 60°C of the pathogen ranged between 0.56 and 3.93 min and between 0.52 and 3.63 min for the presence and absence of acid adaptation condition, respectively. Acid adaptation significantly (P < 0.05) increased the magnitude of strain variability in the thermal inactivation characteristics of L. monocytogenes, with the coefficient of variation increasing to 0.17, whereas acid adaptation did not significantly (P ≥ 0.05) influence the variabilities in the growth parameters of the tested strains. Furthermore, the subsequent growth behaviors of all strains did not exhibit significant (P > 0.05) changes after exposure to acidic broth. However, the thermal resistance of most (25 of 33) of the tested strains increased (P < 0.05) after growing in acid-adapted broth. The relevant data generated in the present study can be used to describe the strain variability in predictive microbiology and to deeply understand the behavioral responses of different strains to acid adaptation.
There was considerable variability in the kinetics characteristics of L. monocytogenes strains.
Acid adaptation increased the magnitude of strain variability in thermal inactivation characteristics of L. monocytogenes.
Behavioral responses of L. monocytogenes to acid adaptation were strain specific.
Listeria monocytogenes is a foodborne pathogen and is responsible for causing listeriosis that can lead to severe diseases, such as meningitis and septicemia, in certain population groups (13). Predictive modeling has been used to quantitatively describe or predict the tendency of growth and decay of specific spoilage microorganisms (33) and estimate and reduce the risk brought by such pathogenic bacteria. The outcome of the growth and inactivation profiles estimated through the predictive microbiology models is influenced by variability. As a function of the system, variability cannot be reduced by changing the experimental method or improving the test protocol (15). Factors such as strain differences, growth condition, cell history, physiological state, and phenotypic variability within a population (population heterogeneity) are the main sources of variability in modeling (17). Hence, it is necessary to quantify the magnitude of the effects induced by these factors on the outcome of the modeling, to ensure that the risk of L. monocytogenes is within a defined range.
Among the above-mentioned factors, the inherent difference among identically treated strains of the same microbial species, defined as “strain variability,” may influence the outcome of the predictive modeling (46). Several studies have reported variability among L. monocytogenes strains (2, 3, 7, 31). For example, Lianou et al. (31) determined the growth rates of 25 L. monocytogenes strains that ranged from 0.40 to 0.49 log CFU/mL/h at 30°C. The D-values at 60°C (D60-values) of 20 L. monocytogenes were from 0.6 to 4.0 min (3). When neglected, such strain differences can result in incomplete and even misleading results in the magnitude of microbial behavior kinetic parameters (17, 28).
Environmental factors, such as growth environment and test conditions, can influence the magnitude of strain variability in the growth and inactivation characteristics of pathogens. However, present studies have seldom involved the exploration of the effect of acid adaptation on the magnitude of strain variability in the growth or inactivation kinetics of foodborne pathogens, except for the studies conducted by Lianou et al. (30) and Haberbeck et al. (25) for Salmonella and E. coli, respectively. Furthermore, data on the extent of such variability remain limited. Mild acidic conditions often occur in food matrices or during food processing. For example, acidic materials, such as lactic acid and acetic acid, are often used to reduce or eliminate bacterial contamination in raw meat or on the meat product surface (14). These materials are washed after a period of time, resulting in a mild acidic meat surface. Furthermore, the pH value of slaughtered cattle is ca. 6.5, and pH remains stable at ca. 5.5 under the action of buffer materials, such as muscle fibrin and water-soluble components (mainly carnosine and inorganic phosphate) (27, 34). Bacteria are able to adapt to moderate conditions that may change the physiological state of cells and in turn alter the subsequent growth or thermal inactivation responses of pathogens at certain conditions (43, 44). This response may have important implications in food safety and food processing optimizations. Thus, it is meaningful to investigate the effect of acid adaptation on the magnitude of strain variability in growth and thermal inactivation kinetics and the behavioral responses of L. monocytogenes to acid adaptation.
Accordingly, the objectives of this study were as follows: (i) to characterize 33 L. monocytogenes strains based on the maximum specific growth rate (μmax), lag times, and decimal reduction times; (ii) to evaluate the effect of acid adaptation on the magnitude of strain variability in growth and thermal inactivation characteristics; and (iii) to assess the adaptive behavioral responses of L. monocytogenes to acid adaptation.
MATERIALS AND METHODS
The 33 L. monocytogenes strains used in this study are listed in Table 1. All of the tested strains were identified. The stock cultures were kept frozen at −80°C in 50% tryptic soy broth with yeast extract (TSBYE; Beijing Land Bridge Technology Co. Ltd., Beijing, China) and 50% glycerol (v/v). For the stock culture, a streak was made on a Trypticase soy agar supplemented with 0.6% yeast extract plate (TSAYE; Beijing Land Bridge Technology Co. Ltd.), and incubated for 24 h at 37°C. Next, a single colony from the plate agar was selected and grown in a test tube with 10 mL of TSBYE and incubated in a shaker (BluePard THZ-100, Yiheng, Inc., Shanghai, China) at 37°C for 16 ± 2 h at 110 rpm. The initial concentration of the activated culture was ca. 109 CFU/mL.
The process of acid adaptation was completed during preculturing. A total of 10 μL of the activated culture was incubated into 10 mL of TSBYE that was buffered with 0.1 mol/L of 2-(N-morpholino)ethanesulfonic acid (Shanghai Maclin Biochemical Co. Ltd., Shanghai, China) at pH 5.5 for the acid adaptation of cells and 0.1 mol/L of morpholinepropanesulfonic acid (Shanghai Maclin Biochemical Co. Ltd.) at pH 7.0 for the nonacid adaptation of cells. The pH value was adjusted with 1 mol/L NaOH and verified using a digital pH meter (Mettler-Toledo Instruments Co. Ltd., Shanghai, China). The culture was incubated at 37°C for 20 h at 110 rpm in a shaking incubator. The pH values of the culture did not change much after bacterial culture (ca. 5.1 to 5.3 and 6.8 to 7.0 for acid and nonacid cultures, respectively). Next, 1 mL of the stationary-phase culture was centrifuged at 5,000 × g for 10 min. Afterward, the cell pellets were washed once with 1 mL of normal saline and then resuspended in 1 mL of untreated TSBYE. Finally, the work cultures were prepared, and the initial concentration was ca. 109 CFU/mL. The adaptation approach entailed the use of biological buffers to gain the acid-adapted and non–acid-adapted cultures for 2-(N-morpholino)ethanesulfonic acid and morpholinepropanesulfonic acid. This approach has many advantages, such as good pH buffer capacity, high water solubility, and nontoxicity to cells. In this manner, the pH values of the broth would almost be as stable as the growth of the bacteria, which is closer to the real conditions of meat products.
Growth and thermal inactivation experiments
The growth curves for the 33 L. monocytogenes strains were determined using the automatic Bioscreen C system (Oy Growth Curves Ab Ltd., Helsinki, Finland). The initial concentration of the inoculum was ca. 105 CFU/mL, which was obtained by decimal dilution. Next, the concentration was five times five-fold diluted to the concentration of 102 to 103 CFU/mL. In total, 200 μL per dilution was added to 100-well microtiter plates. Next, the plates were placed in the Bioscreen system set at 25°C. Optical density (OD) at 600 nm was automatically measured every 10 min until the bacteria reached the stationary phase. Two independent experiments were conducted, and two parallel tests per strain were analyzed at each condition.
Thermal inactivation was accomplished at 60°C by water bath. In total, 25 μL of the work culture in 200 μL of thinly walled PCR tubes was inactivated by complete submersion into water, to prevent the culture from evaporating, leading to an inaccurate concentration. One sample was taken out from the water bath after a period of time and was then immediately immersed in an ice slurry. The survivors were counted by spot plating 20 μL of the serial dilutions on TSAYE. Next, the plates were incubated at 37°C for 24 h. The process for the thermal inactivation experiment of L. monocytogenes ATCC BAA-751 (strain 17, as shown in Table 1) was performed using samples taken at a set of time intervals for both conditions. In total, 13 L. monocytogenes strains were subjected to heat treatment, limiting the duration at 60°C to 3 min. Meanwhile, 18 strains were subjected to heat treatment, increasing the time to 6 min, and the last one was inactivated after 10 min. Two data points were obtained during the thermal inactivation of the 32 L. monocytogenes strains. The thermal treatments were reproduced three times on different days by using three single cell colonies.
Growth and thermal inactivation kinetic parameters
The coefficient of variation (CV) of the estimated μmax, λ, and D-values were calculated using Excel 2019 (Microsoft, Redmond, WA), to characterize and quantify the magnitude of strain variability in the growth and thermal inactivation of pathogens.
Growth and thermal inactivation characteristics of the 33 strains
As presented in Figure 1, the range of the μmax values was almost unchanged after acid adaptation. The μmax values varied from 0.21 to 0.44 h−1 and from 0.20 to 0.45 h−1 for the acid-adapted and non–acid-adapted strains, respectively (Fig. 1A and 1B). For the 33 strains, the μmax value of the majority of the tested strains was ca. 0.3 h−1, and the values for the minority of the tested strains (Table 1, strains 18, 19, and 30) were much larger or smaller than 0.3 h−1. For the λ values of the 33 L. monocytogenes strains, a considerable variability was observed, and the fold-change was greater than that of μmax. Specifically, the period of λ ranged from 0.69 to 2.56 h and from 0.24 to 3.36 h for acid and nonacid adaptation conditions (Fig. 2C and 2D). Interestingly, the growth of some of the tested strains was obviously always greater or slower than that of the others, regardless of whether the strain was preexposed to acid-adapted or non–acid-adapted conditions.
Figure 1E and 1F presents the observed heat resistance of the 33 tested strains. The D60-values for the acid-adapted strains ranged from 0.56 to 3.93 min (Fig. 1E). These values occurred with considerable variations that were nearly equal to that of non–acid-adapted strains, which ranged from 0.53 to 3.63 min (Fig. 1F). Three strains (11, 14, and 32) exhibited higher heat resistance. The D60-values were all greater than 3 min under acid or non–acid-adapted conditions.
Effects of acid adaptation on the magnitude of strain variability for 33 L. monocytogenes strains
As shown in Table 2, acid adaptation had no influence on the strain variability in μmax and λ values of the tested strains. The CV values for the growth parameters after acid or nonacid adaptation conditions were almost the same: 0.12 for μmax and from 0.37 to 0.35 for λ in strains under acid and nonacid cultures, respectively. This finding indicates that after acid adaptation, both the growth characteristics and magnitude of strain variability remained stable. However, different from the effect of the acid adaptation on growth variability, this reduced the scale of strain variability in the heat resistance of L. monocytogenes strains. The CV values significantly decreased from 0.68 to 0.51 for acid and nonacid adaptation conditions.
Behavioral response of L. monocytogenes to acid adaptation
The acid adaptation had no significant (P ≥ 0.05) influence on the growth responses of all strains at 25°C. However, the effect of acid adaptation on the subsequent inactivation responses was strain dependent. The thermal inactivation curves for strain L. monocytogenes (ATCC BAA-751) were further used as the reference strain, to gain insight into the bacterial behavior after acid or nonacid adaptation. The relevant thermal inactivation parameters during inactivation at 60°C for BAA-751 were then determined. The survival curves are presented in Figure 2. The curves drawn demonstrated two different shapes. The curve for acid-adapted cells was log linear with a shoulder. A longer shoulder (or the presence of a shoulder) is correlated to increased thermal resistance in that region of the survivor curve. By comparison, the other curve for the nonadapted cells was only log linear.
The models were overall well fitted to the data, and the RMSE value was 0.07 and 0.06 for acid and nonacid adaptation, respectively (Table 3). Because of the occurrence of a shoulder in the survival curves, three log-reduction times (t3d), including the shoulders for the cells, were appropriate for the estimation of the thermal inactivation parameter of the reference strain. The t3d value was 4.72 and 5.21 min for adapted and nonadapted cells (Table 3).
As presented in Figure 3, acid adaptation resulted in significantly (P < 0.05) enhanced heat survival for 25 of the tested strains. Strains 8, 13, 16, and 29 exhibited similar thermal resistance compared with the nonadapted counterparts (P ≥ 0.05). Strains 9, 14, 17, and 19 had decreased thermal resistance after acid adaptation (P < 0.05). The observed strain differences in behavioral responses to acid adaptation may have no correlation to the change in pH values in the 20-h acid adaptation process, because the value was similar for all tested strains.
The influence of strain-to-strain differences in growth and thermal resistance characteristics constitutes a significant portion of realistic outcome in predictive microbiology modeling. Consistent with the present findings, previous studies have reported a substantial variability among the strains of foodborne pathogens, primarily E. coli O157:H7 (25, 46),L. monocytogenes (1–3, 6, 7, 31), Salmonella enterica (9, 17, 28, 29), and to a lesser extent, Vibrio parahemolyticus (12, 26), in terms of growth or inactivation under different environmental conditions. For example, Díez-García et al. (19) reported that the values of λ ranged from −2.12 ± 2.697 h to 4.17 ± 0.260 h and that the maximum growth rate ranged from 0.03 ± 0.002 ΔOD/h to 0.11 ± 0.011 ΔOD/h. Aryani et al. (3) estimated the D60-values of 20 L. monocytogenes strains from 0.6 to 4.0 min, and the values were almost equal to the values herein. Although relevant studies have analyzed the L. monocytogenes variability under different environmental conditions, the present study adds the important addition of the influence of acid adaptation on strain variability. The kinetic parameters generated expand the data for strain variability in growth and inactivation profiles, simultaneously providing a better view of the wide phenotypic variability among L. monocytogenes strains. A limitation of the present study was that there was no information on the inactivation curve shapes of the 32 L. monocytogenes strains for relying on a linear model.
Studies about the influence of acid adaptation on the responses of pathogens applied different methodologies, such as exposing the bacteria to gradual acidification as a result of glucose fermentation (21, 23, 24), growing the cells in the unbuffered media initially adjusted to low pH (18, 38, 39) or a constant pH (25). We obtained the adapted strains by growing the cells in a constant pH condition. We note that the unmodified TSBYE showed a gradual pH reduction as a result of glucose fermentation during the growth of the bacteria, so there is a possibility that the cells were expressing acid tolerance during the growth due to this pH fluctuation. However, this possible adaptation is equal and present under both conditions in the present study that make it possible to compare between treatments. Edelson-Mammel et al. (20) used TSB without dextrose plus 1% glutamine (TSB−G) and TSB plus 1% glutamine plus 1% glucose (TSB+G) to inoculate for 13 L monocytogenes strains being tested, respectively. The addition of glutamine to the two media allowed TSB−G to support the growth of L. monocytogenes to high levels in the absence of the glucose. Consistent with the present results, the ability of acid adaptation to enhance thermal resistance can vary substantially among L. monocytogenes strains. The results of our study also showed acid adaptation increased the magnitude of strain variability in thermal inactivation, but not the growth of L. monocytogenes. Haberbeck et al. (25) and Lianou et al. (30) also concluded that acid adaptation increases the extent of strain variability in the resistance of foodborne pathogenetic strains. Other factors, such as growth and stress condition, also influenced the size of the strain variability, regardless of the growth or thermal inactivation of microorganisms, and the magnitude of strain variability increased as the environment became more unfavorable (4, 29, 30). The impact of these factors on strain variability should be carefully assessed and characterized to avoid a compromised outcome of predictive modeling and risk assessment.
Previous findings on the growth responses of pathogens to acid adaptation concluded that the process of growing in mild acidic conditions improved the subsequent growth ability of pathogens at certain harsh environments, but not at suitable conditions (8, 36, 40, 42). These findings are consistent with the results herein. Skandamis et al. (40) reported that acid-adapted cultures were able to grow at lower pH levels than nonadapted cultures of E. coli O157:H7, whereas at temperatures below 25°C, the growth ability of nonadapted cultures stopped at a higher water activity value than acid-adapted cultures, for the whole pH range from 3.52 to 7.32. Tiganitas et al. (42) reported the shorter lag time and higher ability to initiate the growth of L. monocytogenes at lower water activity after preadaptation at pH 6.0 or 5.0 than at neutral pH. These studies indicate that stricter sanitation procedures should be implemented in food processing to minimize the risk of change in growth behavior of surviving cells. For the thermal inactivation responses of pathogens, several studies have focused on the effect of acid adaptation on the thermal resistance of multiple L. monocytogens (19, 37, 41, 43). For example, Sharma et al. (37) used two strains and found that the thermal resistance of L. monocytogenes G1091 decreased in watermelon juice after acid adaptation, whereas that of L. monocytogenes G1091 and F8369 was unaffected by the acidic broth. In addition, Topalcengiz and Danyluk (43) reported that only one of the three tested L. monocytogenes strains exhibited a decrease in thermal resistance at 56°C in orange juice after exposure to acidic condition, whereas the others exhibited an unchanged thermal resistance. The above-mentioned studies used a limited number of tested strains, whereas the present study used various strains to assess the variability in the thermal inactivation responses between acid-adapted and nonadapted cultures. The increase in thermal resistance after the acid adaptation of L. monocytogenes may have been associated with the changes in membrane lipid composition (47) and the regulation of stress response regulators, such as the regulation of σB regulon, playing an important role in the resistance to temperature, chemical, and other stresses (45). The different effects of acid adaptation on thermal resistance may be the heterogeneity of the stress-response gene expression.
In conclusion, although a few studies have reported data on microbial kinetics in growth and inactivation, more intensive efforts are still needed to generate relevant data to take into full consideration the strain variability in risk assessment. Furthermore, it is essential to investigate the effects of environmental conditions on strain variability in the kinetic characteristics of microorganisms. Simultaneously, the relationship between phenotypic variability of strains and gene expression among foodborne microorganisms is worth exploring.
The present study revealed the existence of various degrees of strain variability in the growth and thermal inactivation characteristics of 33 L. monocytogenes strains. The variation was especially important to the outcome of the predictive modeling and risk assessment, which should be carefully assessed from the food safety point of view. In addition, the magnitude of thermal inactivation variability deceased after adaptation (pH 5.5), whereas there was no change in growth variability compared with nonacid adaptation (pH 7.0). Furthermore, the thermal inactivation response to acid adaptation was highly variable among the different strains. Therefore, it is expected that these data would provide an in-depth understanding of strain variability to risk assessment.
This study was supported by the National Key Research and Development Program of China (grant 2019YFE0103800).
Present address: Centre of Analysis and Test, School of Chemistry & Molecular Engineering, East China University of Science and Technology, Shanghai, China.