Suboptimal Bacillus licheniformis and Bacillus weihenstephanensis Spore Incubation Conditions Increase Heterogeneity of Spore Outgrowth Time

Sporulation and incubation conditions have an impact on the numbers of spores able to recover after exposure to sublethal heat treatment. Using flow cytometry, we were able to follow at a single-cell level the changes in the physiological states of heat-stressed spores of Bacillus spp. and to discriminate between dormant spores, germinated spores, and outgrowing vegetative cells. We developed original mathematical models that describe (i) the changes with time of the proportion of cells in their different states during germination and outgrowth and (ii) the influence of temperature and pH on the kinetics of spore recovery using the growth limits of the tested strains as model parameters. We think that these models better predict spore recovery after a sublethal heat treatment, a common situation in food processing and a concern for food preservation and safety.

ABSTRACT

Changes with time of a population of Bacillus weihenstephanensis KBAB4 and Bacillus licheniformis AD978 dormant spores into germinated spores and vegetative cells were followed by flow cytometry, at pH ranges of 4.7 to 7.4 and temperatures of 10°C to 37°C for B. weihenstephanensis and 18°C to 59°C for B. licheniformis. Incubation conditions lower than optimal temperatures or pH led to lower proportions of dormant spores able to germinate and extended time of germination, a lower proportion of germinated spores able to outgrow, an extension of their times of outgrowth, and an increase of the heterogeneity of spore outgrowth time. A model based on the strain growth limits was proposed to quantify the impact of incubation temperature and pH on the passage through each physiological stage. The heat treatment temperature or time acted independently on spore recovery. Indeed, a treatment at 85°C for 12 min or at 95°C for 2 min did not have the same impact on spore germination and outgrowth kinetics of B. weihenstephanensis despite the fact that they both led to a 10-fold reduction of the population. Moreover, acidic sporulation pH increased the time of outgrowth 1.2-fold and lowered the proportion of spores able to germinate and outgrow 1.4-fold. Interestingly, we showed by proteomic analysis that some proteins involved in germination and outgrowth were detected at a lower abundance in spores produced at pH 5.5 than in those produced at pH 7.0, maybe at the origin of germination and outgrowth behavior of spores produced at suboptimal pH.

IMPORTANCE Sporulation and incubation conditions have an impact on the numbers of spores able to recover after exposure to sublethal heat treatment. Using flow cytometry, we were able to follow at a single-cell level the changes in the physiological states of heat-stressed spores of Bacillus spp. and to discriminate between dormant spores, germinated spores, and outgrowing vegetative cells. We developed original mathematical models that describe (i) the changes with time of the proportion of cells in their different states during germination and outgrowth and (ii) the influence of temperature and pH on the kinetics of spore recovery using the growth limits of the tested strains as model parameters. We think that these models better predict spore recovery after a sublethal heat treatment, a common situation in food processing and a concern for food preservation and safety.

INTRODUCTION

Spores of Bacillus are metabolically dormant but keep the capacity to reinitiate active life after a sequential process that includes germination, outgrowth, and cell multiplication (1, 2). Cell multiplication of many species of Bacillus is a major cause of poisoning and spoilage in food production and distribution (3). Growth control of spore-forming bacteria is a major issue, and prediction of their behavior in food environments a major challenge. The proportion of spores able to germinate, to outgrow, and to multiply and the duration of the germination and outgrowth process preceding cell multiplication depend on multiple factors. Among these are previous stress applied to spores, such as heat inactivation common in food industry, concentration of molecules (germinant) triggering germination, and, more generally, the physical and chemical environment to which spores are exposed (temperature, pH, water activity [a_w], etc.) (4–6). The combination of these sequential events determines the spore recovery, i.e., the ability of stressed spores to form a colony on nutrient agar or to multiply under suboptimal growth conditions in foods. We recently proposed quantitative estimations and mathematical models describing the recovery process of different species of spore-forming bacteria subjected to various stresses, such as suboptimal growth temperature or pH (6–9). These models will gain to take into account the large heterogeneity of phenotypes within a spore population (10), and thus it is necessary to examine the germination and outgrowth process at the single-cell level. Among possible tools, flow cytometry offers the possibility of marking cells at different physiological states using fluorescent probes. Flow cytometry has specifically been applied, for instance, to follow the evolution of a Bacillus cereus ATCC 14579 and Bacillus cereus NCTC 7464 spore population containing dormant spores, germinated spores, and cells outgrowing under adverse conditions (11, 12).

In the present study, we followed with time the proportions of dormant spores, germinated spores, and outgrowing cells and their variations as consequence of heat treatment and of suboptimal sporulation or recovery conditions. A mathematical model defined with parameters of biological significance and describing the process is proposed. The tested strains, Bacillus weihenstephanensis KBAB4 and Bacillus licheniformis Ad978, were selected for their differences in thermotype and resistance to heat, as well as for the availability of phenotypic data, used as input parameters of the mathematical model.

RESULTS

Changes with time of counts of dormant spores, germinated cells, and vegetative cells.

Changes with time in counts of dormant spores, germinated spores, and vegetative cells within spore populations were monitored using flow cytometry (Fig. 1 and Table 1). Cell populations were analyzed until first cell divisions, revealed by an increase in total cell number. Each of the 45 kinetics curves of B. weihenstephanensis spores and of the 27 of Bacillus licheniformis spores was drawn with at least 10 cell population samplings and determination of counts of dormant spores, germinated spores, and vegetative cells. Equations 1 to 3 describe the cell numbers over time. Representative curves showing the changes in dormant spore population, germinated spore population, and vegetative cell population are shown in Fig. 2. The germination and outgrowth kinetics were composed of two major phases: the number of dormant spores decreased as the number of germinated spores increased (from 0 min to 15 min in Fig. 2A and from 0 min to 30 min in Fig. 2B), and then the number of germinated spores decreased as the number of vegetative cells increased (from 45 min to 130 min in Fig. 2A and from 60 to 110 min in Fig. 2B). The evolutions of dormant spores, germinated spores, and vegetative cells were fitted with equations 4, 5, and 6, respectively (Fig. 2), which use the mathematical function of a Weibull distribution. These equations allowed a satisfactory description of germination and outgrowth kinetics for both strains. No change in the scattering of germination time (S_germ) was observed according to sporulation, heat treatment, or incubation conditions, and an S_germ of 1.0 gave a satisfactory description of dormant spore and germinated spore evolution (likelihood ratio test, α = 0.05) (data not shown).

FIG 1.

Flow cytometry dot plots for identification of the populations successively observed during germination and growth recovery of Bacillus weihenstephanensis spores under optimal conditions. These results were obtained with one culture, at different times of incubation. (A) At the initial time, the population was made up of >95% dormant spores; spores appeared small, hardly fluorescent, and dormant (A1), as confirmed with phase-contrast microscopy (phase-bright dormant spores) (Aa) and epifluorescence microscopy (a green fluorescent halo was hardly observed) (Ab). (B) After 10 min of incubation at 30°C, the germinated spores had the same size as dormant spores but were highly fluorescent (B2). Phase-contrast microscopy and epifluorescence microscopy confirmed this observation: phase-dark spores (Ba) were highly fluorescent (Bb). To verify that the population was appropriately gated, the same experiment was performed with passing 10-fold-diluted suspensions (data not shown). (C) After 120 min of incubation, the vegetative cells had a larger size (higher FSC values) and a higher fluorescence than germinated spores. Phase-contrast microscopy and epifluorescence microscopy confirmed this observation: vegetative cells (Ca) and highly fluorescent (Cb) are visible. To verify that the population was properly gated, the same experiment was performed with passing 10-fold-diluted suspensions (data not shown).

TABLE 1.

Germination and outgrowth parameters and cardinal values for Bacillus weihenstephanensis KBAB4 and Bacillus licheniformis Ad978 estimated by the models presented in Materials and Methods

Organism	Incubation condition	Parameter	Optimal value (95% CIa )	Tmin (95% CI)	Topt (95% CI)	Tmax (95% CI)	pHmin (95% CI)	pHopt (95% CI)
Bacillus weihenstephanensis	Temp	tgerm		−1.66 (−1.85, 1.52)
		τgerm	0.94 (0.90, 0.98)	0.58 (−6.33, 7.49)	31.08 (29.35, 32.8)	44.55 (38.67, 50.44)
		toutgr	100.8 (95.51, 114.11)	11.79 (3.48, 20.9)	29.80 (26.5, 33.1)	43.64 (37.41, 49.87)
		Soutgr	6.10 (5.22, 6.98)	5.48 (4.75, 5.76)	24.85 (21.56, 28.15)	44.23 (38.05, 50.41)
		τoutgr	0.84 (0.77, 0.91)	2.98 (−6.53, 12.48)	31.73 (29.36, 32.80)	51.82 (38.39, 65.24)
	pH	tgerm	6.67 (5.66, 7.74)				4.79 (4.53, 5.07)	6.78 (6.39, 7.17)
		τgerm	0.96 (0.93, 0.99)				4.91 (4.82, 5.01)	7.98 (6.81, 8.96)
		toutgr	73.83 (64.47, 83.20)				4.76 (4.57, 4.95)	6.50 (6.36, 6.66)
		Soutgr	6.46 (5.44, 7.48)				4.95 (4.83, 5.08)	6.40 (5.26, 7.54)
		τoutgr	0.88 (0.83, 0.94)				4.95 (4.89, 5.00)	6.82 (6.54, 7.09)
Bacillus licheniformis	Temp	tgerm		13.86 (12.18, 14.50)
		τgerm	0.92 (0.83, 1.00)	3.25 (−2.01, 4.49)	48.07 (44.40, 51.75)	73.86 (56.65, 91.09)
		toutgr	57.48 (51.40, 65.18)	10.31 (1.19, 19.44)	54.78 (53.31, 56.25)	59.81 (58.85, 60.77)
		Soutgr	6.94 (6.09, 7.81)	12.90 (8.20, 16.21)	41.09 (36.14, 46.03)	69.29 (61.34, 77.23)
		τoutgr	0.91 (0.85, 0.96)	12.72 (10.90, 17.45)	47.72 (45.14, 50.30)	82.56 (71.59, 93.52)
	pH	tgerm	9.75 (8.16, 12.11)				4.48 (4.23, 4.73)	6.47 (6.13, 6.82)
		τgerm	0.90 (0.82, 0.97)				4.01 (3.19, 4.81)	7.16 (5.70, 8.62)
		toutgr	69.54 (63.06, 77.51)				4.59 (4.47, 4.71)	6.64 (6.39, 6.89)
		Soutgr	7.10 (7.08, 7.12)				4.48 (4.23, 4.73)	6.48 (6.14, 6.82)
		τoutgr	0.83 (0.67, 1.00)				4.50 (4.22, 4.78)	8.16 (6.06, 10.24)

^aCI, confidence interval.

FIG 2.

Representative kinetics curves of the changes in the numbers (N) of dormant spores (○), of germinated spores (●), and of vegetative cells (◆) and total cell counts (spores and vegetative cells) (□) over time of incubation for Bacillus weihenstephanensis KBAB4 at 30°C (A) and for Bacillus licheniformis at 15°C (B). The line represents the fitting of data to the kinetic model of spore germination and outgrowth (see Materials and Methods for details).

The spore germination and outgrowth were optimal under conditions near the optimal growth conditions. Indeed, the optimal value of germination time (t_germ), proportion of germinated spores (τ_germ), outgrowth time (t_outgr), scattering of outgrowth time (S_outgr), and proportion of outgrowing cells (τ_outgr) were obtained near the optimal growth pH (pH 7.7 ± 0.2 and pH 8.2 ± 0.6 for B. weihenstephanensis and B. licheniformis, respectively) (Fig. 3; see also Fig. S1 in the supplemental material). Similarly, the optimal values of all parameters were obtained near the optimal growth temperature (30°C and 45°C for B. weihenstephanensis and B. licheniformis, respectively), with a noticeable exception for the time of germination, which continuously decreased with increasing temperature (Fig. 4).

FIG 3.

Impact of pH on time to germination (t_germ) (A), proportion of germinated spores τ_germ (B), time of outgrowth (t_outgr) (C), the scattering of outgrowth time (S_outgr) (D), proportion of outgrowing cells (τ_outgr) (E), and vegetative cell proportion (corresponding to the proportion of dormant spores able to achieve germination and outgrowth) (F) for Bacillus weihenstephanensis KBAB4 spores produced under optimal conditions (pH 7.0 and 30°C) (●), for spores produced at pH 5.5 and 30°C (▵), for spores produced at pH 7.0 at 30°C and treated at 85°C (□), and for spores produced at pH 7.0 at 30°C and treated at 95°C (○). The full lines represent the fitting of the data with equations 7 to 10.

FIG 4.

Impact of temperature on t_germ (A), τ_germ (B), t_outgr (C), S_outgr (D), τ_outgr (E), and vegetative cell proportion (corresponding to the proportion of dormant spores able to achieve germination and outgrowth) (F) for Bacillus weihenstephanensis KBAB4 spores. The full lines represent the fitting of the data to the model developed in the present work (see equations 7 to 10).

Description of the impact of incubation temperature and pH on germination and outgrowth using cardinal growth parameters.

The impact of incubation temperature and pH on t_germ, τ_germ, t_outgr, S_outgr, and τ_outgr was determined by the secondary model derived from a set of mathematical equations (equations 7, 8, 9, and 10). The model was fitted on 26 values of t_germ, τ_germ, t_outgr, S_outgr, and τ_outgr for B. weihenstephanensis KBAB4 (13 values to test model the temperature effect and 13 the pH effect) and 24 values for B. licheniformis Ad978 (13 values to test model the temperature effect and 11 the pH effect) (Table 1). Predetermined cardinal temperatures and pHs were used as input parameters. The fitting of the secondary models was as satisfactory as with estimated cardinal pH values and cardinal temperatures on the full range of tested pHs and on the range of temperatures between the minimal and optimal growth temperatures (T_min and T_opt) (likelihood ratio test, α = 0.05). In contrast, as the evolution of germination and outgrowth parameters did not follow the same pattern as growth parameters for temperatures higher than T_opt, a proper maximum growth temperature (T_max) value was estimated.

Impact of heat treatment on spore germination and outgrowth for B. weihenstephanensis KBAB4.

Spores of B. weihenstephanensis, produced at pH 7.0 at 30°C, were heat treated at 85°C or 95°C to achieve a 90% ± 2% reduction of the initial population. Surviving spores were then incubated at pH 7.4. Neither heat treatment at 85°C nor treatment at 95°C significantly impacted the t_germ compared to that of untreated spores (Fig. 3A). Germination rates remained high, 0.84 ± 0.03 and 0.80 ± 0.06, for spores treated at 85°C and 95°C, respectively, but significantly lower than germination rate of untreated spores (0.94 ± 0.04) (analysis of variance [ANOVA], α = 0.05) (Fig. 3B). S_outgr was significantly lower, 3.73 ± 0.26 min at pH 7.4 and 3.85 ± 0.16 min at pH 7.4, for spores treated at 85°C and 95°C, respectively, than S_outgr of untreated spores (6.10 ± 0.88 min at pH 7.4), meaning that heterogeneity of outgrowth times was higher for heat-treated spores (Fig. 3D). In addition, both treatments led to a significantly lower τ_outgr, 0.59% ± 0.05% for pH 7.4 and 0.32% ± 0.11%, than in untreated spores, 0.84% ± 0.07% (Fig. 3E). Interestingly, the heat treatment of spores at 95°C resulted in a proportion of outgrowing cells significantly lower than at 85°C (Fig. 3E). Spores heated at 85°C also showed an outgrowth time (171 min ± 7.7 min) significantly longer than for untreated spores (100.8 min ± 5.3 min). Outgrowth times of untreated pores and of spores treated at 95°C were not significantly different. In addition, both heat treatments did not have a significantly higher impact on spores incubated at pH 6.0 than for spores incubated at pH 7.4 (Fig. 3). Under the tested conditions, the incubation pH did not have a significant additional impact on B. weihenstephanensis KBAB4 spore germination and outgrowth.

Impact of sporulation pH on spore germination and outgrowth for B. weihenstephanensis KBAB4.

Spores produced at pH 5.5 and incubated at pH 7.4 presented a τ_germ of 0.69 ± 0.06, significantly lower than that of spores produced under optimal conditions, 0.94 ± 0.04. Time of outgrowth of spores produced at pH 5.5 was slightly but significantly longer (ANOVA, α = 0.05) (117 min ± 11.8 min) than that of spores produced under optimal conditions (100.8 min ± 5.3 min). The t_germ and the S_outgr were not significantly impacted by the pH of sporulation for B. weihenstephanensis KBAB4 (Fig. 3). The same pattern was observe for spores incubated at pH 6.0.

Impact of sporulation pH on the proteome of B. weihenstephanensis spores.

Proteomic analysis, comprising 788 proteins, revealed significant differences in detection of 187 proteins between spores produced at 30°C and at pH of sporulation (pH_spo) 7.0 and spores produced at 30°C and at pH_spo 5.5. Among those, 166 proteins were detected in larger amounts in spores formed at pH_spo 7.0 than at pH_spo 5.5 (Table S1). Interestingly, proteins linked to germination were overdetected in spores formed at pH_spo 7.0 compared to pH_spo 5.5. In particular, the proteins SpoVAD, SpoVAE, having an important role in intake and release of dipicolinic acid (DPA), and SleB, a cortex lytic enzyme (CLE) playing a major role in germination, were in significantly higher abundance or more extractible in spores formed at pH_spo 7.0 than in spores formed at pH_spo 5.5 (13, 14) (Table 2). Finally, 21 proteins were overdetected in spores formed at pH_spo 5.5 compared to pH_spo 7.0, such as a peroxiredoxin involved in oxidative stress response.

TABLE 2.

Identification of proteins involved in germination process for Bacillus weihenstephanensis KBAB4 spores produced at 30°C and pH_spo 7.00 and spores produced at 30°C and pH_spo 5.50

Codea	Protein	Description	No. of detected spectra
			pHspo 7.00	pHspo 5.50
A9VFH1	SpoVAD	Localizes to the spore inner membrane and appears to interact with the GerAB and GerAC nutrient germinant receptor proteins	67.7b	10.0
A9VUD3	SpoVAE	Transketolase central region, canal protein located in spore inner membrane	70.7b	29.7
A9VSI2	SleB	Hydrolase, cortex peptidoglycan hydrolysis	14.3b	4.3
A9VUD2	GerA (predicted protein)	Germination protein, germinant receptor	43.0b	14.0
A9VRK1	Peroxiredoxin	Peroxidase activity, oxidative stress response	27.7	41.3b

^aFrom database UniprotKB_Bacillus_weihenstephanensisKBAB4-dec2014.

^bSignificant difference in protein detection.

DISCUSSION

The model developed in the present study described the evolution of dormant spores, germinated spores, and vegetative cells of B. weihenstephanensis KBAB4 and B. licheniformis Ad978, formed or grown under optimal and suboptimal incubation conditions. The time of germination (i.e., time to 90% germination within in spore population) and its scattering, which indicates the heterogeneity of time of germination, were used as parameters of a Weibull cumulative distribution. The distribution of germination and outgrowth times of Clostridium botulinum were previously described with a log-normal distribution (15). We also applied the fitting to a log-normal distribution in the present study (data not shown), but the Weibull model provided a better description of the whole germination and outgrowth process, as in previous work dealing with B. subtilis germination (16, 17). The Weibull model takes into account the differences of probability of individual spores to germinate and/or to outgrow, in accordance with our flow cytometry observations. The proposed model assesses the composition of the population at any time of incubation. The model is based only on parameters associated with incubation time and proportions of spores in their successive physiological states during germination and outgrowth.

The impact of incubation temperature and pH on the parameter describing the structure and the evolution with time of the populations of spores germinating and outgrowing has been quantified. First, the estimated S_germ parameter was not impacted by the incubation temperature or pH. Taking into account these observations and in order to simplify the model, the S_germ value was fixed at 1.00 min. No significant difference was pointed out between the fitting with estimated S_germ for each kinetic curve or an S_germ of 1 min for all kinetic curves (likelihood ratio test, α = 0.05). This may be caused by the rapid germination of the majority of B. weihenstephanensis and B. licheniformis spores under the tested conditions and therefore a poor recording of events in early germination (t < 5 min, corresponding to the first experimental observable time). For both strains, times of outgrowth were minimal and probabilities of germination and outgrowth maximal for temperatures close to the optimal growth temperature. The times of outgrowth increased and probabilities of germination and outgrowth decreased as temperature strayed from optimum. In contrast, no optimal temperature for the germination time was observed for either strain. The first step of germination is triggered by germinant molecules access to the inner membrane, binding to the germinant receptors and leading to a partial hydration of the spore core (5, 18). A hypothesis may be that molecule agitation increases with temperature and may promote germinant diffusion to their receptors through spore surface layers. This may facilitate germination initiation and decrease germination time since the diffusion time of germinant in spore outer layers to reach the germination receptors (GRs) is the main determinant of germination time (1). The parameters describing outgrowth, namely, t_outgr and S_outgr, were impacted by incubation temperature in the same way as growth rate (19). Steps following early germination involve active metabolism, as in Bacillus anthracis rapidly showing esterase activity after germination (20). Moreover, the second step of germination involves the activity of cortex lytic enzymes (CLEs) (21). Metabolic activity is influenced by incubation temperature, with impacts on both outgrowth duration and heterogeneity. With regard to pH, t_germ and t_outgr progressively increased when the incubation pH came close to the germination and outgrowth pH limit, and t_germ and t_outgr were minimal at an optimal pH, between pH 6.0 and pH 7.4. Interestingly, the pH had a less progressive impact on the proportion of spores able to germinate than on other parameters (Fig. 3).

The proposed models describing the impact of incubation conditions on spore germination and outgrowth dynamics were based on physiological parameters of growth. Minimal and optimal temperatures of outgrowth were close to predetermined minimal (T_min) and optimal (T_opt) growth temperatures (20), but the estimated maximal temperature (T_max) of outgrowth was higher than the maximal temperature of growth (20). The estimated maximal temperature for germination rate was significantly higher than the maximal growth temperature, as found in a previous work (19). Nevertheless, no significant difference could be pointed out between the fittings of the data with estimated minimal, maximal, and optimal germination temperature and the fittings with predetermined growth T_min and T_opt (likelihood test, α = 0.05). Thus, the evolution of each parameter could be described between the minimal and optimal germination temperature with the model we developed that uses predetermined growth cardinal values. Maximal temperature of germination and outgrowth higher than maximal growth temperature may be linked to the capacity of spores to germinate beyond the growth limits, as some (rare) previous studies showed (22, 23). This observation is consistent with the passive nature of mechanisms involved in the initiation of the germination process. In other words, some temperatures detrimental for growth have a low impact on germination. As with temperature, the predetermined minimal and optimal growth pHs (pH_min and pH_opt) were used as input parameters in the models (24). No significant differences were pointed out between the fittings with estimated minimal and optimal pH of germination and with the fittings with predetermined growth pH_min and pH_opt (likelihood ratio test, α = 0.05), meaning that the model we developed allowed the description of the impact of incubation pH on germination and outgrowth using predetermined growth pH_min and pH_opt.

Spores surviving heat treatment at 85°C for 12 min and 95°C for 2 min showed different germination and outgrowth kinetics, although the inactivation rates achieved with the two treatments were similar (i.e., a 90% reduction of the initial population). In the present study, heat treatment did not significantly impact t_germ (ANOVA, α = 0.05), in contrast with previous studies. For instance, the germination time of B. subtilis spores increased by only 10% (4) and by 200% for Geobacillus stearothermophilus and Clostridium botulinum spores. The impact of heat treatment on germination time seems to depend on the bacterial genera or species as well as heat treatment intensity (6, 25, 26). Surprisingly, the heat treatment at 85°C did not significantly impact the proportion of spores able to germinate (τ_germ) at pH 7.4 and pH 6.0, compared to the heat treatment at 95°C, which resulted in a lower τ_germ value. This could be explained by a different impact of heat treatment on germinant receptors or calcium-dipicolinic acid (CaDPA) canals, as suggested by a previous work on the damage caused by heat on B. subtilis spores (27). A proportion of spores could be blocked after an intense heat treatment at the dormant spore stage, as previously suggested by other studies (28, 29). However, spores remaining dormant after an intense heat treatment may also be superdormant spores, revealed, for instance, by heat treatments on B. subtilis spores (30). Outgrowth times of spores surviving heat treatment at 85°C were prolonged, as was observed for heat-treated C. botulinum spores (6, 25). This extension of outgrowth time may be due to the impact of the treatment on enzymes or functional proteins within the spores (27). The heat treatment at 95°C led also to a lower proportion of spores able to outgrow, τ_outgr, but without any significant impact on time to outgrowth. The damage due to heat treatment may be repaired before or during the outgrowth of spores (31). This hypothesis is supported, at least partly, by restored metabolic activity, when spores are rehydrated, allowing repair of damage mainly on proteins and enzymes caused by heat treatment (27, 28, 32). After both heat treatments, a higher heterogeneity within the spore population was observed during outgrowth, as already observed for heat-treated spore-forming bacteria such as C. botulinum (6, 25) and B. subtilis (4).

Spores of B. weihenstephanensis formed at 30°C and pH_spo 5.5 presented some significant differences in recovery kinetics compared to spores produced at 30°C and pH_spo 7.0. The proportion of dormant spores able to germinate was significantly lower for spores produced at pH 5.5 (ANOVA, α = 0.05). This observation may come from a lower number of germinant receptors set during sporulation at low pH, as the sporulation conditions affect germinant synthesis (12). Proteins of the SpoV family involved in the germination process were detected in smaller amounts in B. weihenstephanensis spores formed at pH 5.5. SpoV proteins are involved in the release of CaDPA during germination (5). Spores lacking these proteins may present a dysfunctional release of CaDPA and thus a lower ability to germinate. The time of outgrowth was significantly impacted by the sporulation pH. It can be assumed that this observation was due to a lower metabolic activity or due to a smaller amount of necessary molecules stored in the spore, leading to an extension of the time of metabolic activity restoration, as spores produced at low pH presented a significantly lower level of general metabolism proteins. Recent work showed a spore phenotypic memory linking the entry and exit of dormancy, hence supporting this hypothesis (33). Our results showed a lower abundance of cortex lytic enzyme SleB for spores produced at pH 5.5, which is one of the essential enzymes for cortex hydrolysis occurring during germination (34–36). We can hypothesize that the hydrolysis of cortex could be less efficient than in spores produced at pH 7.0. The time for total spore hydration and then swell up and outgrowth may be longer because of a smaller amount of enzymes, like SleB. These observations on protein detection changes regarding different sporulation pH, and the possible correlations with the changes in spore germination and outgrowth for the same tested pHs of sporulation, remain putative, and further experiments are necessary to assert these assumptions.

In conclusion, B. weihenstephanensis and B. licheniformis spore recovery observed at the single-cell level is highly heterogeneous. This recovery implies successive physiological steps, including germination and outgrowth. Sporulation conditions, heat treatment intensity, and incubation conditions impact each physiological step in spore recovery. This complexity can be described and assessed by the proposed mathematical model based on their ecological conditions in terms of temperature and pH.

MATERIALS AND METHODS

Bacterial strains and culture.

Bacillus weihenstephanensis strain KBAB4, isolated from forest soil (INRA, Avignon, France) (37), and Bacillus licheniformis strain Ad978 (ADRIA Développement, Quimper, France), isolated from dairy ingredients, were used. Both strains were stored at –80°C in brain heart infusion (BHI; Biokar Diagnostics, Beauvais, France) mixed with 50% (vol/vol) glycerol. A 100-ml volume of BHI was inoculated with a 1-ml aliquot fraction of the stock suspensions and incubated for 8 h at optimal growth temperature (30°C for B. weihenstephanensis KBAB4 and 45°C for B. licheniformis Ad978), and then 1 ml was transferred into 100 ml of BHI for a 16-h incubation at the same temperatures. Finally, 0.1 ml of B. weihenstephanensis suspension and 0.01 ml of B. licheniformis suspension were added to 100 ml of BHI and were incubated for 6 h. For both strains, the final cell concentration in the preculture was approximately 10⁸ CFU ml⁻¹. The number of spores estimated by the number of cells surviving a 5-min 70°C heat treatment was lower than 100 ml⁻¹.

Spore preparation.

Spores were produced through a two-step sporulation process (38). Volumes of 100 ml of the previously described preculture were centrifuged (6, 000 × g, 10 min, and 12°C) and suspended in 100 ml of sporulation mineral buffer at pH 7.0 made of K₂HPO₄ at 4.5 g liter⁻¹, KH₂PO₄ at 1.8 g liter⁻¹, CaCl₂·H₂O at 8.0 mg liter⁻¹, and MnSO₄·H₂O at 1.5 mg liter⁻¹ and at pH 5.5 made of K₂HPO₄ at 4.5 g liter⁻¹, KH₂PO₄ at 1.8 g liter⁻¹, CaCl₂·H₂O at 8.0 mg liter⁻¹, and MnSO₄·H₂O at 1.5 mg liter⁻¹, filter sterilized using 0.22-μm-pore-size filters (38). These suspensions were incubated with shaking at 30°C and pH 7.0 and pH 5.5 for B. weihenstephanensis KBAB4 and 45°C for B. licheniformis Ad978. Spores suspended in sporulation mineral buffer were harvested when free spores represented more than 95% of cells observed under a magnification of ×1,000 by phase-contrast microscopy (Olympus BX50; Olympus Optical Co., Ltd., Hamburg, Germany), i.e., for both strains after 1 to 2 days at optimal growth pH and up to 10 days at suboptimal pH. The spore suspensions were centrifuged (6,000 × g, 10 min, and 12°C). Spore pellets were suspended in 5 ml of sterile distilled water. The 5-ml suspensions were divided into 1-ml aliquots and stored for 1 month at 4°C before use. Spore heat resistance did not change for at least a 6-month storage time (data not shown). The final concentrations of the stock suspensions were 10⁸ spores ml⁻¹ for B. weihenstephanensis and 10⁹ spores ml⁻¹ for B. licheniformis.

Heat treatment of Bacillus weihenstephanensis.

Heat treatments were all standardized for inactivation, with a reduction of 90% ± 2% of the initial spore counts. Spores of B. weihenstephanensis KBAB4 were diluted in phosphate-buffered saline (PBS; K₂HPO₄ at 4.5 g liter⁻¹, KH₂PO₄ at 1.8 g liter⁻¹, and NaCl at 8 g liter⁻¹, filter sterilized through a 0.22-μm filter) to a final concentration of approximately 10⁷ spores ml⁻¹. Capillary tubes of 200 μl were filled with 100 μl of spore suspension, sealed, and then immersed in a water-glycerol bath maintained at 85°C for 12 min or at 95°C for 2 min in order to obtain the inactivation of 90% ± 2% (9, 38). Capillary tubes were removed from the heat treatment bath (12 min at 85°C or 2 min at 95°C) and immediately cooled in a water-ice bath for 30 s for the enumeration of survivors. The tips were broken and the heat-treated spore suspensions were diluted in tryptone salt broth (Biokar Diagnostics, Beauvais, France). Volumes of 1 ml of the appropriate decimal dilutions of heat-treated spores were mixed into molten brain heart agar (BHA; Biokar Diagnostics, Beauvais) at pH 7.4 and pH 6.0 and incubated at 30°C. Spore concentration at the initial time (t₀) was estimated by treating spore suspensions in capillary tubes a water bath at 70°C for 5 min, in order to inactivate the remaining vegetative cells, and counting surviving cells.

Detection and counting of dormant spores, germinated spores, and outgrowing cells by flow cytometry.

Flasks of 100 ml of BHI were inoculated at a final concentration of 10⁶ spores ml⁻¹. Flasks of BHI adjusted at pH 7.4 inoculated with B. weihenstephanensis were incubated at six temperatures (10°C, 15°C, 18°C, 25°C, 30°C, and 37°C). Flasks of BHI adjusted at pH 7.4 inoculated with B. licheniformis were incubated at 18°C, 30°C, 37°C, 45°C, 50°C, and 58°C. Flasks of BHI adjusted at seven pHs (5.0, 5.1, 5.2, 5.8, 6.0, 6.5, and 7.4) and inoculated with B. weihenstephanensis were incubated at 30°C. Flasks of BHI adjusted at pHs 4.7, 4.8, 5.0, 6.0, 6.5, and 7.4 and inoculated with B. licheniformis were incubated at 45°C. At regular incubation times, 0.1-ml volumes of the suspensions were sampled and diluted to approximately 10⁵ spores or cells ml⁻¹ in filtered PBS and stained with Syto9 (Molecular Probes, Life Technologies, Saint-Aubin, France) at a final concentration of 1.5 μM. Changes with time in counts of dormant spores, germinated spores, and/or vegetative cells were quantified using a flow cytometer (Cyflow Space; Sysmex Europe Gmbh, Norderstedt, Germany) equipped with an excitement laser light source at 488 nm and three detectors: forward scatter (FSC), side scatter (SSC) detecting light emission at 488 nm, and a fluorescence light detector with a 536-/40-nm filter (FL1). The software used to collect and analyze the flow cytometry data was Flomax 2.3. Spores and/or cells were analyzed at 1 μl s⁻¹ with the following voltages: 109 mV (FSC), 310 mV (SSC), 251 mV (FL1). The spores and/or cells were quantified in a 200-μl volume of suspension using Sysmex true absolute volumetric counting (TVAC). Cell populations were analyzed until first cell divisions, revealed by an increase in total cell number. Each kinetic curve was replicated three times with independently prepared spore suspensions.

Heat-treated B. weihenstephanensis spores were incubated at 30°C in BHI adjusted at pH 7.4 and pH 6.0, and flow cytometry assays were performed as previously described and replicated three times with independently prepared spore suspensions. The control for a dormant-spore suspension was a suspension of spores stored at 4°C in sterile distilled water, and the control for germinated spores was a suspension of spores germinated for 15 min in BHI at 30°C or 45°C, respectively, for B. weihenstephanensis or B. licheniformis. A suspension of spores incubated for 2 h at 30°C or 45°C, respectively, for B. weihenstephanensis or B. licheniformis was taken as a control for a vegetative-cell suspension. The phase-bright state of the dormant spores, the phase-dark state of germinated spores, and the characteristic shape of vegetative cells were verified by phase-contrast microscopy under a magnification of ×1,000. The fluorescence of each population (i.e., dormant spores, germinated spores, and vegetative cells) stained with 1.5 μM Syto9 at the final dilution was verified by epifluorescence microscopy. Each control was analyzed by flow cytometry. Statistics on the FSC parameter that correlates to spore or cell size, on the SSC parameter that correlates to cell granularity, and on the FL1 parameter that correlates to green fluorescence intensity were determined. The flow cytometry parameter set for each population is given in Table 3.

TABLE 3.

Parameters of the gates associated with dormant spores, germinated spores, and vegetative cells of populations of Bacillus weihenstephanensis KBAB4 and Bacillus licheniformis Ad978 cells analyzed by flow cytometry^a

FC parameter	Value for test strain at indicated physiological stage
	Bacillus weihenstephanensis KBAB4			Bacillus licheniformis Ad978
	Dormant spores	Germinated spores	Vegetative cells	Dormant spores	Germinated spores	Vegetative cells
FSC	2.65 ± 0.06	1.72 ± 0.21	12.66 ± 2.12	1.67 ± 0.07	1.33 ± 0.17	7.42 ± 1.23
FL1	0.65 ± 0.14	8.91 ± 0.15	55.46 ± 4.56	0.88 ± 0.11	4.81 ± 0.13	82.39 ± 5.64

^aValues are expressed as arbitrary units (means ± SDs). See Fig. 1 for flow cytometry dot plots. FSC, forward scatter; FL1, fluorescence 1.

The gating between two populations (e.g., dormant spores and germinated spores) was defined on FSC and fluorescence parameters. The regions were confirmed by analyzing mixtures of dormant spores and germinated spores, dormant spores and vegetative cells, germinated spores and vegetative cells, and dormant spores germinated spores and vegetative cells (Fig. 1). With this optimized setup, the error on counting of each population was less than 5%.

Kinetic model of spore germination and outgrowth.

During germination and outgrowth, the cell population is composed of dormant spores, germinated spores, and vegetative cells. The total cell number over time can be described as

$$N_T\left(t\right)=N_{\mathit{RS}}\left(\text{t}\right)+N_{\mathit{GS}}\left(\text{t}\right)+N_{\mathit{VC}}\left(t\right)$$ (1)

where $N_T$ is the total number of bacterial cells, $N_{\mathit{RS}}$ the number of (refractive) dormant spores, $N_{\mathit{GS}}$ the number of germinated spores, $N_{\mathit{VC}}$ the number of vegetative cells, and t the time of incubation. Dormant spores become germinated spores and germinated spores become vegetative cells. Each test was performed with a constant number of cells (equation 2):

$$\forall t,N_T\left(t\right)=N_0$$ (2)

where N₀ is the initial number of cells in the suspension.

The evolution of each subpopulation was described using mathematical models described below. The evolution curves of germinated spores were fitted using the model inspired from Weibull cumulative distribution.

$$P_i\left(t\right)=1-{10}^{{\left(\frac{t}{t_i}\right)}^{S_i}}$$ (3)

where P_i corresponds to the Weibull distribution of time, t_i is the time to reach 90% of the maximal number of cells in each population i (either dormant spores, germinated spores, or vegetative cells), and S_i expresses the scattering of t_i within the population (the higher the S_i is, the lower the scattering is). The evolution of dormant spore population was fitted using the model presented in equation 4:

$$N_{\mathit{RS}}(t)=N_{\mathit{RS}}{}_0\times \left\{\left[\text{1}-{\tau }_{\mathit{germ}}\times P_{\mathit{germ}}(t)\right]\right\}$$ (4)

where $N_{\mathit{RS}}(t)$ is the dormant spore number, $N_{\mathit{RS}}{}_0$ is the initial number of dormant spores, τ_germ is the proportion of dormant spores able to germinate, and $P_{\mathit{germ}}(t)$ is the cumulative Weibull distribution as described previously (equation 3).

Using the same methodology, the evolution of germinated spore population was fitted using the model presented in equation 5:

$$N_{\mathit{GS}}(t)=\left[N_{\mathit{RS}}{}_0-N_{\mathit{RS}}\left(t\right)\right]\times \left[\text{1}-{\tau }_{\mathit{outgr}}\times P_{\mathit{outgr}}(t)\right]$$ (5)

where N_GS is the number of germinated spore, $N_{\mathit{RS}}{}_0$ is the number of dormant spores at the initial time, N_RS is the number of dormant spores, and τ_outgr is the proportion of germinated spores able to outgrow. As described for germinated spores (equation 3), $P_{\mathit{outgr}}(t)$ is the cumulative distribution of Weibull, depending on the incubation time t, parameterized by t_outgr, a scale parameter corresponding to the time to reach 90% of the number of germinated spores able to outgrow, and S_outgr, the Weibull shape parameter describing the scattering of t_outgr.

Finally, the evolution of vegetative cells was fitted using the model presented in equation 6:

$$N_{\mathit{VC}}(t)=N_{V{C}_{\mathit{max}}}\times \left[\text{1}-{\tau }_{\mathit{outgr}}\times P_{\mathit{outgr}}(t)\right]$$ (6)

where N_VC is the number of vegetative cells, $N_{V{C}_{\mathit{max}}}$ is the maximal number of vegetative cells, and τ_outgr is the proportion of germinated spores able to outgrow.

Modeling of the impact of temperature and pH on germination and outgrowth kinetics.

The developed models describing the evolution of each parameter regarding temperature or pH were derived from the Gamma concept (39) (equation 7):

$$\theta ={\theta }^{*}\times {\displaystyle \prod {\gamma }_{X^{(X_i)}}}$$ (7)

where $\theta$ stands for each of the primary parameters $(\frac{1}{t_{\mathit{germ}}},\hspace{0.17em}\frac{1}{t_{\mathit{outgr}}},\hspace{0.17em}{S}_{\mathrm{germ}},\hspace{0.17em}{S}_{\mathrm{outgr}},\hspace{0.17em}{\tau }_{\mathrm{germ}},\hspace{0.17em}\text{and}\hspace{0.17em}{\tau }_{\mathrm{outgr}})$ describing the evolution of each subpopulation, $\theta *$ stands for each temperature or pH estimated for reference conditions X_i, and ${\gamma }_X(X_i)$ is a function describing the impact of temperature or pH, as described below.

The evolution of $\frac{1}{t_{\mathit{outgr}}}$, $S_{\mathit{outgr}}$, τ_germ, and τ_outgr regarding incubation temperature (T_i) is described by the function presented in equation 8:

$${\gamma }_T\left(T_i\right)=\{\begin{array}{cc}0& \hspace{0.17em}\text{for}\hspace{0.17em}{T}_i\le T_{\mathit{min}}\\ \frac{\left(T_i-T_{\mathit{max}}\right){(T_i-T_{\mathit{min}})}^2}{{(T_{\mathit{opt}}-T_{\mathit{min}})}^1\left[\left(T_{\mathit{opt}}-T_{\mathit{min}}\right)\left(\mathit{Ti}-T_{\mathit{opt}}\right)-\left(T_{\mathit{opt}}-T_{\mathit{max}}\right)(T_{\mathit{opt}}+T_{\mathit{min}}-2T_i)\right]}& \hspace{0.17em}\text{for}\hspace{0.17em}{T}_{\mathit{min}}<T_i<T_{\mathit{max}}\\ 0& {\hspace{0.17em}\text{for}\hspace{0.17em}T}_i\ge T_{\mathit{max}}\end{array}$$ (8)

where T_min, T_opt, and T_max are the minimal, optimal, and maximal conditions of temperature of germination and outgrowth.

The evolution of $\frac{1}{t_{\mathit{germ}}}$ regarding incubation temperature was described by the function presented in equation 9:

$${\gamma }_T\left(T_i\right)=\{\begin{array}{c}\begin{array}{cc}0& \text{if}\hspace{0.17em}{T}_i<T_{\mathit{min}}\end{array}\\ \begin{array}{cc}{\begin{array}{c}\left(\frac{T_i^{}-T_{\mathit{min}}}{T_{\mathit{opt}}-T_{\mathit{min}}}\right)\end{array}}^2& \text{if}\hspace{0.17em}{T}_i\ge T_{\mathit{min}}\end{array}\end{array}$$ (9)

The evolution of all germination and outgrowth parameters regarding the incubation pH (pH_i) was described by the function presented in equation 10:

$${\gamma }_{\mathit{pH}}\left({\text{pH}}_{\text{i}}\right)=\begin{array}{cc}0& {\text{for pH}}_{\text{i}}\le {\text{pH}}_{\text{min}}\\ \frac{\left({\text{pH}}_{\text{i}}-{\text{pH}}_{\text{max}}\right){\left({\text{pH}}_{\text{i}}-{\text{pH}}_{\text{min}}\right)}^n}{{\left({\text{pH}}_{\text{opt}}-{\text{pH}}_{\text{min}}\right)}^{n-1}\left[\left({\text{pH}}_{\text{opt}}-{\text{pH}}_{\text{min}}\right)\left({\text{pH}}_{\text{i}}-{\text{pH}}_{\text{opt}}\right)-\left({\text{pH}}_{\text{opt}}-{\text{pH}}_{\text{max}}\right)\left(\left(n-1\right){\text{pH}}_{\text{opt}}+{\text{pH}}_{\text{min}}-n\times {\text{pH}}_{\text{i}}\right)\right]}& \hspace{0.17em}\text{for}{\hspace{0.17em}\text{pH}\hspace{0.17em}}_{\text{min}}<{\text{pH}}_{\text{i}}<{\text{pH}}_{\text{max}}\\ 0& {\text{for pH}}_{\text{i}}\ge {\text{pH}}_{\text{max}}\end{array}$$ (10)

where pH_min, pH_opt, and pH_max are the estimated minimal, optimal, and maximal conditions of pH of germination and outgrowth and n was equal to 1 to describe the evolution of $\frac{1}{t_{\mathit{germ}}}$, $\frac{1}{t_{\mathit{outgr}}}$, $S_{\mathit{outgr}}$, and τ_outgr or equal to 0.2 to describe the evolution of τ_germ.

Statistics.

The models were fitted on the observations by minimizing the sum of squared errors (SSE) using the lsqcurvefit function from MatlabR2012b (MathWorks, Natick, MA). The goodness of fit of the models was checked by the root mean square error (RMSE) (40, 41). The smaller the RMSEs were, the better the model fitted the data. The 95% confidence intervals were calculated using the nlparci function from MatlabR2012b. The fitting performance of each model was statistically evaluated by the F test, comparing the mean square error of the model to the mean square error of the data. The computed f value was compared to the F table value (0.05 significance level). If the f value was lower than the F value from the table, the F test was accepted, indicating that the model fitting was statistically acceptable. Moreover, variance comparison tests were performed using the anova function from MatlabR2012b.

The fitting performance of models was compared with a likelihood ratio test and test statistic S_L, computed as follows (equation 11):

$$S_L=n\hspace{0.17em}\text{log}\hspace{0.17em}\mathit{RS}{S}_C-n\hspace{0.17em}\text{log}\hspace{0.17em}\mathit{RS}{S}_U$$ (11)

where n is the number of data, $\mathit{RS}{S}_C$ is the residual sum of square for the constrained model, and $\mathit{RS}{S}_U$ is the residual sum of square for the unconstrained model. In this work, the constrained model was the model using a predetermined cardinal temperature and pH for each strain as input parameters. The unconstrained model was the model using estimated recovery cardinal values. When n tends toward infinity, the limiting distribution of S_L is χ² distributed with p_u − p_c degrees of freedom, where p_u is the number of parameters in the unconstrained model and p_c the number of parameters in the constrained model. If S_L is lower than χ² (α = 0.05), the difference in the fitting of both models was considered not significant.

Proteomic analysis.

Volumes of 1 ml of spore suspension at 5 × 10⁸ spores ml⁻¹ from the stock of spores produced at 30°C and pH 7.0 and at 30°C and pH 5.5 were centrifuged at 6,000 × g and 12°C for 10 min. The pellet was resuspended in 300 μl of grinding buffer (Tris HCl [pH 7.5] at 50 mM, EDTA at 5 mM). At the very end of incubation, 1× protease inhibitor cocktail was added and the suspension was homogenized. The 300-μl suspension was transferred into a FastPrep tube containing 0.3 g of zirconium beads. The suspension was subjected to 2 min of agitation at 1,500 rpm (1600 MiniG; SPEX Sample Prep, Metuchen, NJ). This agitation was repeated at least 5 times, in order to obtain more than 95% broken spores. The breakage efficiency was checked by phase-contrast microscopy where more than 95% of the spores appeared phase dark and damaged. Then the suspension containing beads was centrifuged at 4,000 × g and 12°C for 10 min. The supernatant was transferred into 1.5-ml tubes. The protein extract was concentrated on a 10-kDa ultrafiltration membrane at 14,000 × g and 4°C for 15 min, to 200 μl (42, 43). The protein concentration was evaluated by the Bradford method (44).

Iodoacetamide was added at a final concentration of 15 mM to the protein extracts, which were incubated in the dark at room temperature for 1 h. Then the enzyme Lys-C (Wako, Richmond, VA) was added at an enzyme/protein ratio of 1:50 (wt/wt), and the extract was incubated for 3 h at room temperature. Trypsin (Promega, Madison, WI) was added at an enzyme/protein ratio of 1:100 (wt/wt), and the extract was incubated at 37°C for 3 h. Finally, trypsin was again added at the same ratio and the extract was incubated at 37°C for 16 h. The sample was acidified to pH 2.0 with 1% trifluoroacetic acid (TFA). The volume was reduced to 200 to 300 μl with a vacuum concentrator (45, 46).

A Strata-x 33u polymeric reversed-phase 30-mg/1-ml column (Phenomenex, Torrance, CA) was activated with 500 μl of pure acetonitrile. Then the column was rinsed with buffer A (3% acetonitrile, 0.06% glacial acetic acid, in water) three times. The protein extract sample was adjust to 500 μl with buffer A. The diluted sample was charged on the column, and the column was rinsed three times with 500 μl of buffer A (flowthrough). The column was eluted with 300 μl of buffer B (40% acetonitrile, 0.06% of glacial acetic acid, in water) two times and the total volume (600 μl) was collected in a 1.5-ml tube. The sample was totally dried using a vacuum concentrator and stored at –20°C. The extracts were prepared in order to obtain 100 μg of protein in each sample (47).

The dried samples were resuspended in 400 μl of loading buffer (0.05% of TFA, 0.05% of formic acid, in water), and 4 μl was injected on a 50-cm Qexatcive column (Thermo Scientific, Waltham, MA) coupled with Eksigent nano-high-performance liquid chromatography-tandem mass spectrometry (nano-HPLC-MS/MS; AB-sciex, Framingham, MA). Runs were 4 h long for each sample. The assays were performed in triplicate for both conditions, with three independent spore batches.

The generated spectra were merged and searched by X!tandem 3.3.3 (http://www.thegpm.org/tandem/) using X!TandemPipeline (v3.3.3), developed by PAPPSO (http://pappso.inra.fr/bioinfo/) (9 January 2013). The database used to compare the results was B.W.KBAB4 from UniprotKB (http://www.uniprot.org/ updated in November 2014), containing 5,717 entries. The results were also compared to a contaminant database to ensure the protein identification.

The LC-MS data alignments and peptide quantification by peak area integration were performed with the MasschroQ 0.20 R package (R 3.1.1) (48). In addition, MasschroQ 0.20 was used to investigate the impact of the sporulation conditions on the relative detection of peptides within B. weihenstephanensis KBAB4 spores produced at pH 7.0 or at pH 5.5 (ANOVA test, α = 0.05).