Role of the alternative sigma factor σB on Staphylococcus aureus resistance to stresses of relevance to food preservation

Authors


Pilar Mañas, Tecnología de los Alimentos, Facultad de Veterinaria, C/Miguel Servet, 177, 50013, Zaragoza, Spain. E-mail: manas@unizar.es

Abstract

Aims:  To examine the role of the alternative general stress sigma factor σB on the resistance of Staphylococcus aureus to stresses of relevance to food preservation, with special emphasis on emerging technologies such as pulsed electric fields (PEF) and high hydrostatic pressure (HHP).

Methods and Results: S. aureus strain Newman and its isogenic ΔsigB mutant were grown to exponential and stationary growth phases and its resistance to various stresses was tested. The absence of the σB factor caused a decrease in the resistance to heat, PEF, HHP, alkali, acid and hydrogen peroxide. In the case of heat, the influence of the σB factor was particularly important, and decreases in decimal reduction time values of ninefold were observed as a result of its deficiency. The increased thermotolerance of the parental strain as compared with the sigB mutant could be attributed to a better capacity to sustain and repair sublethal damages caused by heat.

Conclusions:  σB factor provides S. aureus cells with resistance to multiple stresses, increasing survival to heat, PEF and HHP treatments.

Significance and Impact of the Study:  Results obtained in this work help in understanding the physiological mechanisms behind cell survival and death in food-processing environments.

Introduction

Nowadays, the consumer’s demands for fresher foods require food preservation methods with a very limited impact on the sensorial and nutritional qualities of the products. In turn, this is promoting the development of new preservation techniques, such as pulsed electric fields (PEF) and high hydrostatic pressure (HHP), as well as the improvement of traditional methods such as heat treatments (Mañas and Pagán 2005). In order to improve their efficiency, much research effort is being directed towards the elucidation of the cellular mechanisms underlying microbial resistance to the various agents and processes used.

Within this context, the influence of bacterial general stress responses has received much attention. Among the various strategies that a bacterial cell is able to deploy under adverse conditions, the use of alternative sigma factors is perhaps the most important one (Abee and Wouters 1999). Sigma factors are the recognition units of the RNA polymerase, and the set of genes that is expressed in the cell depends on which sigma factor is bound to the core enzyme. General stress sigma factors include sigma S, also known as rpoS, in some Gram-negative and sigma B (σB) in some Gram-positive bacteria, which are considered by many researchers as functionally homologous (Gertz et al. 2000; Liu et al. 2005). At present, there is some information available about genetic regulation of σB in Bacillus subtilis, Listeria monocytogenes and Staphylococcus aureus (Pané-Farréet al. 2006; Price 2000). However, its physiological role in cell survival, especially with regards to food preservation processes, has been much less investigated. Moreover, both the genetic regulation and the physiological role of σB seem to differ depending on the species (Bischoff and Berger-Bächi 2001; Pané-Farréet al. 2006).

In that regard, the most extensively studied species are Bacillus subtilis and Listeria monocytogenes (Völker et al. 1999; Ferreira et al. 2003; Wemekamp-Kamphuis et al. 2004; Höper et al. 2005). At least 100 σB-dependent stress proteins have been identified in B. subtilis, conferring protection against a variety of environmental stresses including heat, high osmolarity, high ethanol concentration, high and low pH, oxidizing agents, growth and survival under low temperatures, among others (Völker et al. 1999; Höper et al. 2005). Partial overlap between the σB-controlled stress responses of B. subtilis and L. monocytogenes has been demonstrated (Price 2000). However, still many aspects remain unknown, especially in the case of S. aureus. In this micro-organism, σB seems to be an important piece in the complex network that controls the expression of virulence determinants (Bischoff et al. 2004). However, its role in stress resistance has not been clearly established. Furthermore, in a recent DNA array-based study it has been shown that there is only a small overlap (12%) of σB-dependent genes in S. aureus and in B. subtilis (Pané-Farréet al. 2006). The authors suggested that σB may be related to more basic cellular functions in S. aureus, such as metabolism or membrane transport processes, and its function in general stress resistance would be marginal.

In summary, the role of the alternative σB factor in the survival of S. aureus has not been fully clarified. In previous works carried out in our laboratory we suggested that the differences in resistance to heat treatments observed among 21 strains of S. aureus, 15 of them isolated from foods, could be owing to differences in the expression of the alternative σB factor (Rodríguez-Calleja et al. 2006; Cebrián et al. 2007). Nevertheless, further work is required to confirm this hypothesis. The aim of this work was to study the role of the σB factor in the resistance of S. aureus to physical agents relevant for food preservation: heat, PEF and HHP. In addition, resistance to chemical stresses including acid and alkaline pH and hydrogen peroxide was also explored.

Materials and methods

Bacterial culture and media

The S. aureus strains used in this study were Newman and IK184, an isogenic ΔrsbUVW-sigB knockout mutant (Bischoff et al. 2004). They were kindly provided by Brigitte Berger-Bächi from the Institute of Medical Microbiology, University of Zurich. Cultures were maintained frozen at –80°C in cryovials. Stationary-phase cultures were prepared by inoculating 10 ml of tryptone soya broth (Biolife, Milan, Italy) supplemented with 0·6% w/v yeast extract (Biolife; TSB-YE) with a loopful of growth from tryptone soy agar supplemented with 0·6% w/v yeast extract (Biolife; TSA-YE). The resulting culture was incubated at 37°C for 6 h, in a shaking incubator. Exactly 50 μl of this culture was inoculated into 50 ml of fresh TSB-YE and incubated for 24 h under the same conditions, which resulted in a stationary-phase culture containing c. 7 × 108 cells ml–1. Exponential-phase cells were prepared by inoculating 50 μl of the stationary-phase culture into 50 ml of fresh TSB-YE and incubating for 3·0–3·5 h, until the optical density at 600 nm reached 0·8, which corresponded to c. 8 × 107 cells ml–1. Bacterial cultures were centrifuged at 10 000 g for 5 min, washed once with McIlvaine citrate-phosphate buffer of pH 7·0 (Dawson et al. 1974). Cell suspensions were finally diluted in the same buffer to the required concentration. This buffer was also used as medium for heat, PEF and HHP treatments described next, but in the case of PEF its conductivity was adjusted with distilled water to 2 mS cm–1.

Heat treatments

Heat treatments were carried out in a specially designed resistometer (Condón et al. 1993). Briefly, this instrument consists of a 350-ml vessel provided with an electrical heater for thermostation, an agitation device to ensure inoculum distribution and temperature homogeneity, and ports for injecting the microbial suspension and for the extraction of samples. Once treatment temperature (52–62°C) had attained stability (±0·1°C), 0·2 ml of an appropriately diluted cell suspension (10cell ml–1), was injected into the main vessel containing the citrate-phosphate buffer (pH 7·0). After inoculation, 0·2-ml samples were collected at different treatment times and immediately pour plated (Condón et al. 1996).

PEF treatments

PEF treatments were carried out in an exponential waveform pulse equipment (García et al. 2005). High electric field pulses were produced by discharging a set of ten capacitors via a thyristor switch (Behlke, Kronberg, Germany) in a batch treatment chamber. The capacitors were charged using a high-voltage DC power supply (FUG, Rosenhein, Germany), and a function generator (Tektronix, Wilsonville, OR, USA) delivered the on-time signal to the switch. The treatment chamber was made of a cylindrical plastic tube closed with two polished stainless steel electrodes. The gap between the electrodes was 0·25 cm and the electrode area was 2·01 cm2. The actual electric field strength and electrical intensity applied were measured in the treatment chamber with a high voltage probe and a current probe, respectively connected to an oscilloscope (Tektronix). The equipment includes provisions for measuring sample temperature (Raso et al. 2000). In this study, pulses at electrical field strength between 19 and 31 kV cm–1 and a pulse repetition rate of 1 Hz were used. The energy associated to pulses at electric field strengths of 19, 22, 26 28 and 31 kV cm–1 was 2·66, 3·47, 4·25, 5·30 and 6·30 kJ kg–1, respectively. In all experiments the initial temperature of the samples was room temperature and it never exceeded 35°C during the treatment.

The microbial suspension, at a concentration of c. 108 micro-organisms ml–1, was placed into the treatment chamber with a sterile syringe and the treatments were performed. After treatment, appropriate serial dilutions were prepared in sterile TSB-YE and pour plated.

HHP treatments

One millilitre of the bacterial suspension was washed as described before and sealed in a polyethylene bag after exclusion of air bubbles and pressurized in an 8 ml pressure vessel, held at 25°C with an external water circuit (Resato, Roden, The Netherlands). Owing to adiabatic heating during pressure buildup the temperature in the chamber inevitably increased but never exceeded 35°C; hence, heat shock effects could be excluded. After treatment, appropriate serial dilutions were prepared and the samples were pour plated.

Hydrogen peroxide resistance determinations

Resistance to hydrogen peroxide was evaluated by adding to 5 ml of pH 7·0 100 mmol l–1 Tris-HCl buffer tempered at 25°C to the appropriate amount of 30% v/v hydrogen peroxide (Sigma, St Louis, USA) to obtain a final concentration of 100 mmol l–1. This buffer had been previously inoculated with the bacterial suspension to a cell concentration of 107 cells ml–1. The samples were collected after different treatment times ranging from 0·08 to 30 min. After the desired contact time, 0·1 ml samples were removed and appropriate serial dilutions were prepared in TSB-YE and pour-plated. It was previously checked that the possible residual concentration of hydrogen peroxide in recovery agar plates did not exert any inhibitory or lethal action on S. aureus cells.

Acid and alkaline pH resistance determinations

TSB was either acidified or alkalinized to pH 2·0 or 12·5 with HCl or NaOH, respectively. Broth was filter-sterilized, cells were added to a concentration of 107 cells ml–1, and the preparation incubated at 25°C. Exactly 100 μl samples were withdrawn at intervals, and transferred into 900 μl of TSB-YE. Subsequent serial dilutions were prepared and pour-plated for survival counts.

Incubation of treated samples and survival counting

The recovery medium was TSA-YE and, where indicated, NaCl was added to estimate the percentage of sublethally injured cells (Mackey 2000). The lack of tolerance to the presence of NaCl is attributed to damage to the functionality and/or integrity of the cytoplasmic membrane (Mackey 2000). The sodium chloride concentration used was chosen in previous experiments as the maximum noninhibitory concentration for nontreated cells (data not shown), and corresponded to 12% w/v for the parental strain Newman and 10% w/v for the mutant. Plates were incubated for 24 h at 37°C unless NaCl was added to the agar; in such case, incubation times of 48 h were needed. After incubation, CFU were counted. All resistance determinations were performed at least thrice on independent working days. Error bars in the figures correspond to the mean SD.

Resistance parameters

Survival curves were obtained by plotting the logarithm of the fraction of survivors vs treatment time, expressed as the number of pulses in the case of PEF treatments. The lethality of heat treatments was measured by the decimal reduction time values (Dt value), which is defined as the time (min) of treatment required at constant temperature for the number of survivors to drop one log cycle. Decimal reduction times curves (DRTC) were represented by plotting the log of the Dt values vs treatment temperature, and z values (°C increase necessary to reduce Dt value a log cycle) were calculated.

For PEF treatments, nonlinear but concave upwards survival curves were observed. Therefore, a mathematical model based on the Weibull distribution was used to fit the survival curves (Álvarez et al. 2003). This model is described by the following equation (Mafart et al. 2002):

image(1)

where S(t) is the survival fraction; t is the treatment time (number of pulses; τ ≈ 3μs); and δ and ρ are the scale and the shape parameters, respectively. The δ value represents the time needed to reduce the first log cycle of the population, while the ρ parameter indicates the shape of the survival curve, which is related to the distribution of individual resistance of cells in the whole population. ρ values that are equal to 1 correspond to straight survival curves, ρ values >1 to convex profiles and ρ values <1 to concave profiles. To fit the model to the experimental data and to calculate the δ and ρ parameters the GraphPad PRISM® software was used, by applying the least-squares criterion. Similar to the z value from the DRTC lines for heat treatments, the effect of the electric field strength could be described by a zPEF value as already proposed for Salmonella (Álvarez et al. 2003). This value indicates the electric field strength increase necessary for the δ value to drop one log cycle. Dt, z and zPEF values, SD and statistical significance of differences (P < 0·05) were calculated using the statistical software package GraphPad PRISM® (Graphpad Software, Inc., San Diego, CA, USA).

For HHP treatments, as there is no consensus on the model to use, we opted for presenting data in the most common format used in literature: fraction of survivors after a fixed treatment time.

Results

Heat resistance

Figure 1a shows the survival curves of S. aureus strain Newman and its sigB isogenic mutant at 58°C in pH 7·0 citrate-phosphate buffer. As it can be observed in the graph, the parental strain Newman was more thermotolerant than the sigB mutant. The inactivation kinetics was linear for at least 99·9% of the population, and therefore decimal reduction time (Dt) values were calculated. Dt value at 58°C was 0·74 min for parental strain Newman and 0·082 min for the mutant; thus, deletion of the sigB gene provoked a decrease of ninefold in heat resistance at 58°C. To check whether these differences were maintained constant at different temperatures, the heat resistance was studied in the range 52–62°C. Decimal reduction time values have been included in Table 1. z values estimated were 5·2°C (±0·24°C) for the parental and 5·7°C (±0·37°C) for the ΔsigB strain. No statistically significant differences were found between these two z values (P > 0·05), indicating that the differences in thermotolerance were maintained along the range of temperatures studied.

Figure 1.

 Resistance of stationary phase cells of Staphylococcus aureus strain Newman (bsl00001) and strain Newman ΔsigB (□) to physical agents: (a) survival curves to a heat treatment at 58ºC; (b) a pulsed electric field treatment at 26 kV cm–1 and (c) a high hydrostatic pressure treatment at 450 MPa.

Table 1.   Resistance of Staphylococcus aureus Newman and S. aureus Newman ΔsigB: to heat, Dt (minutes) and z value (ºC); to pulsed electric fields (PEF), δ value (number of pulses), ρ value and ZPEF (kV cm–1); and to high hydrostatic pressure (HHP); log Nt/No (after 15 min)
 S. aureus NewmannS. aureus Newmann ΔsigB
HeatDt (SD)Dt (SD)
 52ºC0·771 (0·076)
 54ºC3·336 (0·291)0·349 (0·021)
 56ºC1·391 (0·079)0·124 (0·011)
 58ºC0·741 (0·030)0·082 (0·007)
 60ºC0·248 (0·011)0·028 (0·002)
 62ºC0·081 (0·009)
 64ºC0·047 (0·008)
z (ºC)5·210 (0·239)5·700 (0·375)
PEFδ (SD)ρ (SD)δ (SD)ρ (SD)
 19 kV cm–162·21 (19·87)0·501 (0·180)35·07 (1·060)0·644 (0·028)
 22 kV cm–123·69 (2·927)0·607 (0·031)13·69 (2·086)0·663 (0·104)
 26 kV cm–112·12 (7·881)0·496 (0·096)2·502 (1·669)0·405 (0·079)
 28 kV cm–17·167 (2·930)0·543 (0·073)0·673 (0·526)0·294 (0·062)
 31 kV cm–16·388 (0·946)0·529 (0·045)0·227 (0·077)0·277 (0·020)
zPEF (kV cm–1)12·60 (2·080)5·458 (0·4030)
HHPlog (Nt/No) (SD)log (Nt/No) (SD)
 400 MPa–0·106 (0·104)–0·343 (0·032)
 450 MPa–0·501 (0·014)–0·921 (0·159)
 500 MPa–0·925 (0·345)–2·281 (0·285)
 550 MPa–1·906 (0·576)–3·279 (0·210)
 600 MPa–2·966 (0·494)–4·318 (0·154)

PEF resistance

The profile of survival curves of both strains to PEF treatments is illustrated in Fig. 1b and data about PEF resistance are included in Table 1. Figure 1b shows the survival curves at 26 kV cm–1. Curves followed concave upwards profiles and therefore a nonlinear model based on the Weibull distribution was used to fit the experimental data and to calculate the kinetic parameters described in the previous section. The effect of the electric field strength is shown in Table 1, where δ values, corresponding to the time needed for the first decimal reduction, and ρ values, which indicate the shape of the curves, at different electric field strengths are included. As shown by Fig. 1b and the data in the table, the parental strain Newman was more PEF resistant than the sigB mutant at all the electric field strengths and treatment times tested in this study. zPEF value was also lower for the mutant strain (Table 1). However, the shape of the survival curves of both strains changed differently with the electric field applied. ρ values corresponding to strain Newman varied randomly between 0·607 and 0·496 within the range of electric field strengths studied and no statistical differences were detected among them (P > 0·05). ρ values corresponding to the ΔsigB mutant decreased as the electric field increased indicating that the profile of the survival curves became progressively more concave with increasing electric field strengths.

HHP resistance

Figure 1c includes the survival curves at 450 MPa corresponding to both S. aureus strains, Newman and its ΔsigB isogenic mutant. The effect of the pressure intensity is illustrated in Table 1, where the fractions of survivors after 15 min of pressurization at different pressures have been included. As it can be deduced, the parental strain, sigB+ was more piezotolerant at any of the pressures tested, although the differences tended to disappear after prolonged treatment times (Fig. 1c). The differences in resistance were lower than those previously observed for heat and PEF treatments. In the survival curves represented in Fig. 1c, it can be seen that the most relevant difference between both strains was the occurrence of a pronounced shoulder in the curve corresponding to the parental strain. It is remarkable that the presence of shoulders in HHP survival curves has been reported only occasionally (Klotz et al. 2007). Tailing was observed for both micro-organisms at the same percentage of survival, which caused a diminution in the difference in resistance after prolonged treatments. In any case, after a typical treatment time of 15 min at different pressures, the parental strain always survived better than the ΔsigB strain (Table 1).

Resistance to extreme pH and hydrogen peroxide stress

Figure 2 includes the number of survivors of both strains after exposure to lethal pH extremes of 2·5 and 12·5, and to oxidative stress conferred by 100 mmol l–1 hydrogen peroxide. It can be seen that the parental strain Newman was more resistant to the three agents tested, although differences in acid media were very slight.

Figure 2.

 Resistance of stationary-phase cells of Staphylococcus aureus strain Newman (white bars) and strain Newman ΔsigB (grey bars) to chemical agents: Log cycles of inactivation after 20 min at pH 12·5, 40 min at pH 2·5 and 20 min with 100 mmol l–1 hydrogen peroxide.

Influence of growth phase

The influence of the growth stage (exponentially growing cells vs stationary-phase cells) in the resistance to heat, PEF and HHP for the parental and the ΔsigB strains is shown in Fig. 3a–c, respectively. These figures demonstrate that cells in the exponential phase of growth of both strains showed a very close resistance. It is remarkable that exponentially growing cells were as sensitive to heat as the ΔsigB strain in the stationary phase. However, in the case of HHP treatment at 450 MPa, and for short-time PEF treatments, stationary-phase cells of the ΔsigB strain were still more resistant than the exponential-phase cells.

Figure 3.

 Survival curves of stationary- (bsl00001, •, continuous lines) and exponential-phase cells (□, ○, dotted lines) of Staphylococcus aureus strain Newman (bsl00001, □) and strain Newman ΔsigB (•,○) to: (a) a heat treatment at 58ºC; (b) a pulsed electric field treatment at 26 kV cm–1; and (c) a high hydrostatic pressure treatment at 450 MPa. Stationary phase survival curves correspond to those included in Fig. 1.

Occurrence of sublethal damage

Cells were recovered in the presence of sodium chloride to study the occurrence of cells with damaged membranes and the role of σB factor on cell recovery. Figure 4 includes the survival curves of both strains grown to stationary phase, treated by heat (a), PEF (b) and HHP (c) and recovered in the presence and absence of sodium chloride. As it can be observed in the graphs, survival curves corresponding to nondamaged heat- and PEF-treated cells, i.e. cells capable to outgrow in the presence of sodium chloride, were practically identical for the parental strain and for its ΔsigB mutant as there were no statistical differences between survival counts. For pressure-treated cells, however, differences were still detectable, up to 20 min of treatment, between nondamaged cells corresponding to the parental and mutant strains.

Figure 4.

 Sublethal injury repair capacity of stationary-phase Staphylococcus aureus strain Newman (bsl00001, □) and strain Newman ΔsigB (•,○): survival curves corresponding to cells recovered in triptic soy agar plus 0·6% w/v yeast extract (TSAYE; bsl00001, •, continuous lines) and TSAYE with sodium chloride (□, ○, dotted lines) to: (a) a heat treatment at 58ºC; (b) a pulsed electric field treatment at 26 kV cm–1; and (c) a high hydrostatic pressure treatment at 450 MPa. TSAYE survival curves correspond to those included in Fig. 1.

Discussion

In this work, we examine the loss of stress resistance associated to sigB null mutation, with special emphasis on various technologies for food preservation. The deletion of the sigB gene provoked a decrease in the resistance to heat, HHP, PEF, hydrogen peroxide, acid and alkaline pH. These results indicate that the σB factor is an important determinant of environmental stress resistance in S. aureus.

The protective effect of the σB factor was particularly significant in the case of heat challenge. It is remarkable that cells of the two strains in the exponential phase of growth, where σB concentration and activity is expected to be very low even in the parental strain (Kullik and Giachino 1997), showed the same heat resistance (P > 0·05). This resistance was similar to that corresponding to the ΔsigB strain in stationary phase. It is well known that the alternative sigma factor σB accumulates in the cell as it enters the stationary phase of growth (Kullik and Giachino 1997). According to the results obtained in this investigation, the absence of the σB factor in S. aureus renders stationary-phase cells to behave like exponential cells with respect to heat tolerance. At present, it is difficult to envisage which changes derived from sigB expression could be responsible for an increased heat tolerance. On the one hand, the mechanisms that provoke the inactivation of a bacterial cell by heat are not clear, and on the other hand, changes linked to sigB expression are not fully known either. Heat provokes several changes in the bacterial cell that may be related to cellular damage and in a later step, inactivation. More specifically, it causes ribosome destabilization, membrane permeabilization, enzyme denaturation and DNA damage, among others (Tomlins and Ordal 1976). Recent DNA microarray studies have demonstrated the σB dependence of at least 251 genes in S. aureus (Bischoff et al. 2004). These include a high number of genes related to virulence factors, such as a gene encoding coagulase or fibronectin-binding protein, genes responsible for the synthesis of capsular polysaccharide, carotenoid pigments, membrane transporters, catalase and many others (Pané-Farréet al. 2006). Various σB-controlled genes could in some way influence survival to heat. For instance, it has been reported that overexpression of sigB leads to increased thickness of the cell wall (Morikawa et al. 2001). Carotenoids are known to regulate membrane fluidity (Chamberlain et al. 1991) and to provide cells with defence against oxidative damage (Liu et al. 2005), which may be generated during heating. Catalase could also protect cells from heat by eliminating oxygen reactive species. Further work is needed to characterize the cellular changes linked to sigB expression that are responsible for the survival of S. aureus to heat.

The recovery of heated cells in medium with sodium chloride showed an interesting result. The net amount of cells able to multiply in the presence of salt after heat treatments was the same for the parental strain and mutant. In the case of the ΔsigB strain no cells with damaged membranes were detected, as survival curves corresponding to cells recovered in the presence and absence of sodium chloride were superimposed. In other words, for the sigB mutant only those intact, nondamaged cells would retain the capacity to multiply and survive, whereas for the parental strain a high degree of sublethal damage and subsequent repair under adequate environmental conditions was detected. In conclusion, the increased heat tolerance of the σB factor positive cells could be related to a high capacity to sustain and repair sublethal damages inflicted by heat.

It is generally acknowledged that cell repair against the action of several stressors depends greatly on the expression of proteases and chaperones (Yura et al. 2000), which are triggered in the so-called heat shock response. It has been demonstrated that the ATP-dependent protease ClpP, which belongs to the Hsp104 family, is part of the σB regulon (Price 2000). Information available to date indicates that the heat shock response overlaps partially, but not totally, with the σB general stress response (Price 2000). Work is in progress in our laboratory to further study the relationship between the heat adaptation response and the general stress response triggered by the σB factor.

Absence of the σB factor also provoked a decrease in tolerance to PEF and HHP. For PEF treatments, the magnitude of this decrease depended on the electric field strength and the treatment time. The shape of the survival curves of the sigB mutant strain became progressively more concave as the electric field strength was increased, as indicated by the decreasing ρ values (Table 1). ρ value is considered to be a reflection of the distribution of individual resistance within cells in the whole population (Peleg and Cole 1998). This change in the ρ value would indicate that as the electric field strength was increased, a progressively greater proportion of the population showed higher sensitivity to PEF. In any case, differences in survival between the two strains were never as large as those observed in heat treatments.

As shown in this investigation, the expression of the sigB gene provides S. aureus cells with multiple resistances to physical stresses. However, its relative importance depends on the technology studied. As it can be observed in Fig. 3, cells of the two strains grown to the exponential phase showed the same heat and HHP resistance. Entry into the stationary phase caused a development of heat tolerance only in the sigB+ strain. With regards to HHP, there was an increase in piezotolerance upon entrance into the stationary phase for the two strains, but it was more pronounced for the parental strain, whose survival curve at 450 MPa presented a shoulder before inactivation started. In the case of PEF treatments (Fig. 3b), cells showed an intermediate behaviour: survival curves of cells in the exponential phase were coincident, and were very close to that corresponding to the sigB mutant strain in the stationary phase, but the latter was still more resistant at least after short treatment times. With respect to the involvement of the cellular repair processes in the apparent resistance of the parental strain and the sigB mutant (Fig. 4a–c), similar conclusions can be drawn if the three technologies are compared. For heat treatments, the difference in the ability to sustain and repair sublethal damages could be entirely attributed to σB, at least under the experimental conditions tested. On the contrary, ability to withstand and repair sublethal damage incurred by PEF and HHP treatments did remain present in the sigB mutant. This was especially relevant for HHP-treated cells, where the percentage of sublethally damaged cells was under some experimental conditions higher than 99% of the surviving population (Fig. 4c).

Therefore, the most obvious conclusion from these results is that the σB factor plays an important role in S. aureus HHP and PEF resistance, but there are also other factors responsible for the increased tolerance in stationary phase, especially in the case of HHP. This observation is in accordance with those of previous reports on the RpoS protein in Escherichia coli (Pagán and Mackey 2000; Robey et al. 2001), where also other factors were accountable for the increased piezotolerance of cells in the stationary phase of growth. It is known that some genes and operons related to stress resistance are often subjected to a double regulation, in such a way that their transcription may be initiated from several sigma factors (Price 2000) and may be triggered by various environmental signals. However, the results described here could also indicate that the acquisition of tolerance to high pressure depends partly on cellular functions, molecules and/or structures that are independent from the sigB general stress response. It has to be remarked that as in the case of heat, the final causes that provoke the inactivation of a bacterial cell under the action of HHP have not been established. From the fragmentary information available, it can be summarized that HHP causes pleiotropic cellular effects and the ultimate inactivation seems to result from multitarget damage. The membrane is an important target, as damages including loss of solutes, loss of osmotic responsiveness, increased permeabilization to nonpermeant dyes, among others, have repeatedly been reported in pressurized cells (Pagán and Mackey 2000; Chilton et al. 2001). But also other damaging events such as solute loss during pressurization, protein coagulation, enzyme inactivation and ribosome conformational changes, have been reported (Mañas and Mackey 2004). In the case of PEF treatments, the formation of pores in the membrane is generally accepted as the final cause of death, although still many aspects such as the influence of the structure of the cellular envelopes, for instance the role of the outer membrane in gram-negative cells, remain to be elucidated (García et al. 2007). Our results indicate that the mode of action of each of the technologies studied is different. A common event is the occurrence of sublethal injury prior to cell inactivation, at least for the resistant strain, Newman.

Finally, in an attempt to characterize the role of the σB factor on S. aureus resistance to agents of different nature, survival to hydrogen peroxide and alkaline or acid pH challenges was also studied. Results showed that the absence of the σB factor led to an increased sensitivity to the three agents. These results support the view of the σB factor as a multiple stress resistance determinant in S. aureus.

Concluding remarks

In this work, we report that the general stress alternative sigma factor σB is an important determinant of the resistance of S. aureus strain Newman to heat, HHP and PEF treatments, as well as to chemical agents such as hydrogen peroxide, acid and alkaline pH. Exponential cells of both strains behaved similarly in their sensitivity to each of the three physical agents studied, and the development of tolerance upon entrance into stationary phase was always more pronounced for the parental strain. In the case of heat treatments, the mutant strain did not develop any heat tolerance in the stationary phase as compared with the exponential phase. The increased thermotolerance of the stationary phase parental strain as compared with the mutant is the result of its σB conferred capacity to sustain and repair sublethal cellular injury. Results obtained in this work help in understanding the physiological mechanisms behind cell survival and death in food-processing environments.

Acknowledgements

The authors would like to thank Dr Berger-Bächi for providing the strains used in this study, and the Spanish Ministry for Education and Science for the predoctoral grant of G. Cebrián. A. Aertsen is a postdoctoral fellow of the Research Foundation of Flanders (FWO Vlaanderen). This investigation has been funded by the Spanish Ministry of Science and Education (CICYT, AGL2006-08856) and the European Commission (FP6, 015710-2NOVELQ).

Ancillary