Barotolerance is inducible by preincubation under hydrostatic pressure, cold-, osmotic- and acid-stress conditions in Lactobacillus sanfranciscensis DSM 20451T


M.A. Ehrmann, Lehrstuhl für Technische Mikrobiologie, Technische Universität München, 85350 Freising, Germany


Aims:  This study addresses the inducibility of barotolerance by preincubation of Lactobacillus sanfranciscensis DSM 20451T under various sublethal stress conditions.

Methods and Results:  Stress conditions which reduce the growth rate of L. sanfranciscensis DSM 20451T to 10% of its maximum were determined. These conditions were met at 43, 12·5°C, a pH value of 3·7, 1·9% NaCl, or 80 MPa respectively. In contrast to heat preincubation, other prestresses, including salt, cold and pressure led to an increase of barotolerance by hydrostatic pressure of 300 MPa for 30 min. Stationary-phase cells also showed an increased barotolerance. Sublethal pressure leads to enhanced heat tolerance.

Conclusions:  Stress response to salt, low temperature and acidic pH as well as starvation overlap with that one to high pressure by inducing barotolerance.

Significance and Impact of the Study:  Inactivation of bacteria by high pressure treatment is influenced by their history which modulates barotolerance. Mechanisms of barotolerance appear different from heat shock defence.


The exploration of bacterial stress responses to adverse environmental conditions is motivated by basic scientific reasons but also by industrial and safety aspects in food microbiology. Stress response is often characterized by the transient induction of specific proteins and physiological changes leading to a general enhancement of the ability to withstand these conditions. It has been demonstrated that bacterial survival is greatly increased by the preceding induction of a stress response (Broadabent and Lin 1999). The quality of the acquired tolerance is not necessarily linked to that of the applied prestress (Hecker 1998). Thus, cross-protection, e.g. against freezing or drying may be achieved by the induction of heat or cold shock response (Komatsu et al. 1990; Willimsky et al. 1992).

Hydrostatic pressure treatment is used for preservation purposes in food technology (Knorr 1999) and represents an exceptional stimulus for most mesophilic bacteria. Pressure was shown to induce multiple changes in physiology (Wouters et al. 1998; Korakli et al. 2002), protein expression (Welch et al. 1993; Drews et al. 2002; Wemekamp-Kamphuis et al. 2002) and also transcription (Sato et al. 1995; Ehrmann et al. 2001). It causes cell injury (Ulmer et al. 2002) and finally cell death (Vogel et al. 2002a).

Objective of this study was to investigate the effect of stressful conditions similar to those met in various foods on gain in barotolerance of bacteria. The ability to acquire barotolerance by cross protection via salt-, temperature, pH- and high pressure-stress conditions would have implications on high pressure treatments. Additional complementary experiments may provide insights in survival of cells during storage of foods after pressure treatment.

Cross protection experiments may provide an insight into the mechanisms of barotolerance. L. sanfranciscensis was used as representative strain for mesophilic lactic acid bacteria being dominant in most sourdough fermentations (Vogel et al. 1999, 2002b).

Materials and methods

Bacterium and growth media

Lactobacillus sanfranciscensis DSM 20451T was grown in modified MRS medium (mMRS) as described by Stolz et al. (1995). To determine the effect of pH and NaCl, the media composition was changed as follows: the pH of mMRS was adjusted to values from 3·6 to 6·6 with 2 mol l−1 NaOH or HCl. To avoid pH-shift during fermentation, 20 mmol l−1 aspartic acid and 20 mmol l−1 citric acid were added. NaCl was added to final concentrations of 0–2·5% (w/v).

Determination of growth under stress conditions

Standardized inocula were prepared by growing overnight cultures in mMRS broth to exponential growth phases (O.D.590 nm 0·3–0·6). The pH and NaCl modified growth media were inoculated to an O.D.590 nm of 0·1 and incubated at 30°C in final volumes of 200 μl. Cultures were overlaid with 50 μl paraffine to achieve anoxic growth conditions and to avoid evaporation. Growth was monitored by measuring the optical density (590 nm) in an automated microtitreplate reader (Spectraflour, Tecan, Austria). The influence of temperature on growth rates was studied in 15 ml tubes incubated at the desired temperature ±0·5°C. Samples were measured within a range of 0·2–0·6 (Novaspec II, Amersham Biosciences, Freiburg, Germany). At higher optical densities, samples were diluted with media. From each growth curve, the maximum growth rate μmax was obtained by fitting the optical density readings to the exponential growth curve (Zwietering et al. 1990).

Determination of the growth after 12 h incubation under stress conditions

As growth rates under high pressure conditions could not be monitored during incubation, growth was determined after 12 h incubation at the respective pressure. After inoculation with an overnight culture to an optical density (O.D.590 nm) of 0·1, cells were filled in a glass tube and sealed with a rubber stopper. They were incubated at 50, 100, 150, 200 and 250 MPa at 30°C in a prototype apparatus as described by Ulmer et al. (2000). Changes of the optical densities were measured after 12 h of incubation.

Determination of the optimal intensity of high pressure pretreatment

Twenty millilitres of mMRS were inoculated to a final O.D.590 nm of 0·1 with exponential growing cells. High pressure preincubations were carried out for 30 min at 30°C in a range of 0·1–120 MPa. After 15 min of regeneration at atmospheric pressure cells were treated with a lethal pressure of 300 MPa for 30 min at 30°C, which was lethal to untreated cells. Cell counts were determined before and after pressure treatment by plating on mMRS. For the determination of the optimal duration of the high pressure pretreatment, cells were pressurized for 0–60 min at 80 MPa and 30°C.

Determination of cross resistances

cLogarithmically growing cells were harvested by centrifugation at 30°C, resuspended in fresh mMRS and incubated for 1 h under different sublethal conditions (1·9% NaCl at 30°C; pH 3·7 at 30°C; 80 MPa at 30°C; 43°C and 12·5°C). For regeneration after this first stress cells were harvested and resuspended in fresh mMRS and were incubated for 15 min at 30°C. Then the cells were treated with a second lethal stress for 30 min [300 MPa at 30°C; 20% NaCl at 30°C; pH 2·0 at 30°C (glycin/HCl-buffer); 50°C and a single freeze/thaw cycle at −20°C]. Cell counts were determined before and after lethal stress incubation.


Determination of stress conditions of L. sanfranciscensis DSM 20451T

For the purpose of comparison of the effects of individual prestresses (acid, heat, NaCl or high pressure) on barotolerance, conditions were determined that reduces the growth rate to 10% of its maximum. These conditions were chosen as most pronounced variation of gene expression has been demonstrated there for lactic acid bacteria (Sanders et al. 1999). These conditions were obtained by incubations at 43°C, 12·5°C, pH 3·7 or 1·9% NaCl, respectively (Fig. 1). Optimal growth rates were achieved at 30°C, pH 5·6 and 0·5% NaCl. As it was not possible to determine the growth rate of L. sanfranciscensis during high pressure treatment, the change of optical density was measured after 12 h incubations under different high pressure conditions. Growth at 50 and 100 MPa reduced the growth to 20 and 2% (Fig. 2). Thus, we assumed a reduced growth rate to 10% at an incubation at 80 MPa.

Figure 1.

Determination of relative growth rates of L. sanfranciscensis DSM 20451T under (a) different pH values (adjusted with HCl), (b) NaCl concentrations, (c) temperatures. Reference growth conditions were 30°C on mMRS (initial pH 5·2) under atmospheric conditions

Figure 2.

Effect of high pressure on growth of L. sanfranciscensis DSM 20451T. Cells were grown at different high pressures. Bars indicate changes of O.D.590 nm after growth for 12 h. Results represent mean ± s.d. of at least two independent experiments

Barotolerance induced by hydrostatic pressure pretreatment

High pressure treatment used in this study (300 MPa at 30°C) resulted in exponential inactivation as usually found for Gram-positive vegetative bacteria. The reduction of cell counts was 99% or greater. Preincubation under pressure in the range of 0·1–120 MPa for 30 min revealed an optimal survival at 80 MPa (Fig. 3). Furthermore, acquired barotolerance was increased with duration of the pretreatment (data not shown). To avoid overlapping effects of the high pressure with pH down shift by metabolic activity, no longer pressure duration than 60 min was used.

Figure 3.

Survival of L. sanfranciscensis DSM 20451T after treatment with a lethal pressure (300 MPa, 30 min, 30°C). X axis indicates high pressure used as prestress (30 min, 30°C). Cell counts were determined before and after lethal pressure treatment by plating on mMRS

Barotolerance by cross resistance

Preincubation of cells at hyperosmotic, acidic and low temperature conditions that reduced the growth rate to 10% resulted in an increased barotolerance of 3·1-, 4·9- and 4·3-fold, respectively (Fig. 4). A significant increase in barotolerance (2·7-fold) was observed for stationary phase cells. Interestingly, high temperature incubation at 43°C reduces barotolerance to more than one half compared with unstressed cells (Fig. 4).

Figure 4.

Determination of cross resistances. Cells of L. sanfranciscensis DSM 20451T were subjected to different conditions (first stress) prior to exposure to different lethal conditions (second stress). First stresses were starvation of cells, exposure to 1·9% NaCl, pH 3·4, 43°C, 12·5°C or pressure (80 MPa). Second stress conditions indicated by different shadings. For detailed conditions see Materials and methods. Results represent mean ± s.d. of at least two independent experiments

High pressure induced heat tolerance

In Fig. 4 the consequences on stress tolerance to different stresses (second stress) after pressure treatment (first stress) are demonstrated.

When sublethal high pressure (80 MPa) was used as first stress, survival at high osmotic conditions (20% NaCl) and resistance to multiple cycles of freezing and thawing were not significantly affected (Fig. 4). But, heat resistance was 3·6-fold increased, whereas tolerance to pH 2·0 was fivefold reduced.


Pressure inactivation of micro-organisms is affected by process parameters but also by product parameters such as temperature, type of food and pH. In this study, we demonstrated that physiological status of the target cells as well as its previous history can influence pressure tolerance.

It is well known that lactic acid bacteria react to metabolic stresses and growth limiting factors with the activation of at least 100 general stress proteins (Van de Guchte et al. 2002). In addition, specific stress responses are in evidence for each individual stress, including low pH, ethanol, heat, osmotic or oxidative stress. Both, general and specific stress responses were known to enhance survival in harsh environments (Gaidenko and Price 1998).

Cross resistances have been frequently demonstrated with different combinations of successional stresses (Boutibonnes et al. 1991; Hartke et al. 1996; Kets et al. 1996; Broadbent et al. 1997; Broadabent and Lin 1999). However, the effect of cross resistance to counterbalance high pressure is not well documented in literature. An increased high pressure sensitivity was previously observed in L. plantarum after preincubation at elevated temperatures (ter Steeg et al. 1999). Anata and Knorr (2003) reported on pressure-induced thermotolerance in L. rhamnosus GG corroborating our results.

In our study all kinds of stress except for heat could induce barotolerance in L. sanfranciscensis (Fig. 4). Therefore, heat shock proteins are not considered to cause barotolerance. As not all stresses induce barotolerance, the general stress response induced by all kinds of stresses including heat is not accountable for the increase in barotolerance. In turn high pressure pretreatment induced tolerance to heat and high pressure only. A possible explanation for that may be found in either one and the same change shared by barotolerance inducing stresses (starvation/stationary cells, osmotic, acidic and cold stress), or different effects activated by each type of stress individually. A possible target that reacts contrarily during heat and cold, but commonly in osmotic, stationary, cold and pH stress seems unlikely.

Basically, achievement of stress tolerance is mediated by a de novo synthesis of certain proteins and/or by changing properties of affected cell components. Protein expression under sublethal high pressure in L. sanfranciscensis was recently investigated by Drews et al. (2002), who detected more than a dozen spots that were more than twofold increased or 50% decreased in their intensity. But their role in barotolerance pale in comparison with 63 protein level changes in consequence of acid stress in L. sanfranciscensis CB1 (De Angelis et al. 2001). Most significantly, it has been reported that the accumulation of trehalose in response to high osmotic pressure is more important for barotolerance than accumulation of heat shock proteins, which act as molecular chaperones (Fujii et al. 1996).

Recent studies indicate a contribution of altered membrane composition, permeability and interference with membrane-bound transport systems to the mechanisms of high pressure mediated inactivation (Perrier-Cornet et al. 1999; Ritz et al. 1999; Ulmer et al. 2000). The composition of lipid membrane is known to be altered in response to variations of pH, external osmolality and low temperature to retain membrane fluidity and functionality (Van den Boom and Cronan 1989; Foster 2000; Guillot et al. 2000; Padan and Krulwich 2000).

The major response to low temperature in bacteria is to lower melting temperature and maintaining fluidity of the membrane by shortening acyl chain lengths and inreasing the degree of unsaturation (Nakayma et al. 1980; Beney and Gervais 2001). Less fluid membranes are known to be more sensitive to high pressure while an increase in membrane fluidity cause the opposite effect (McDonald 1984, 1992; Smelt et al. 1994; Casadei et al. 2002). Thermal control by increasing the proportion of unsaturated fatty acids was shown to take place in B. subtilis after 1 h when 5% of recovered total methyl esters were unsaturated (Grau and de Mendoza 1993).

Other links to membrane-associated barotolerance may be seen by other traits, e.g. the formation of special fatty acids as demonstrated by the formation of cyclopropane fatty acids in L. lactis in response to low pH (Guillot et al. 2000) or membrane-bound transport systems are involved in sodium stress adaptation by activating ion extrusion systems (Padan and Krulwich 2000). Complementary experiments using high pressure as prestress showed that high pressure activates no general stress response leading to enhanced tolerance to multiple stresses. Solely heat sensitivity was significantly reduced by high pressure prestress. Thus, high pressure stimulates specific reactions different from heat shock response but contributes to heat resistance.

Albeit the high complexity and the fact, that underlying physiological details are not clear yet, our results provide vital data that should be taken into account when high pressure processes are designed and pressurized foods are stored.


We thank the Deutsche Forschungsgemeinschaft (DFG) for financial support.