Action of trehalose on the preservation of Lactobacillus delbrueckii ssp. bulgaricus by heat and osmotic dehydration


E. Aníbal Disalvo, Laboratorio de Fisicoquímica de Membranas Lipídicas, Facultad de Farmacia y Bioquímica (UBA), Junín 956 (1113) Buenos Aires, Argentina (e-mail:


Aim: This work determines the efficiency of trehalose on the preservation by heat or osmotic drying of a strain of Lactobacillus delbrueckii ssp. bulgaricus. Cell recovery at different trehalose concentrations during drying correlated with the surface properties and osmotic response of cells after rehydration.

Methods and Results: Bacteria were dried in the presence of glycerol, trehalose, sucrose at 70°C and at 20°C. Trehalose attenuates the loss of viability at 0·25 m. At this concentration, the osmotic response and zeta potential of the bacteria were comparable with the nondried ones.

Conclusions: Trehalose diminishes significantly the damage produced by dehydration both when the bacteria are dried by heating or subjected to osmotic dehydration. This effect appears related to the preservation of the permeability to water and the surface potential of the bacteria.

Significance and Impact of the Study: Dehydration occurring during heating or during osmosis appears to have similar effects. As dehydration-induced damage is in correlation with osmotic response recovery and is hindered or buffered by the presence of trehalose, it may be related to water eliminated from biological structures involved in water permeation.


Dehydration appears to be a convenient method to preserve lactic acid bacteria because the same procedure to produce powdered milk could be used to prepare starters for the milk industry (Teixeira et al. 1995). However, there is a lack of knowledge about the mechanism of dehydration, which results in low survival rates, low stability during storage and difficulties in rehydrating the solid product (Potts 1994).

The cell membrane is the main target for damage during freezing and freeze-drying (Steponkus et al. 1995; Gardiner et al. 2000). The injury caused by freezing usually destroys the barrier for ion permeation and, in consequence, a collapse of the membrane potential is produced. One reason for this damage is that physicochemical properties of the bilayer, such as packing, viscoelasticity and membrane integrity are affected drastically by the temperature shift and by changes in the water/lipid ratio (thermotropic and lyotropic transitions) (Disalvo et al. 2002). These transitions and its implicance on the membrane barrier properties have been extensively studied in model membranes of phosphatidylcholines or phosphatidylethanolamines showing a strong dependence on the unsaturation/saturation ratio of the fatty acid chains, and on the polar head groups (Disalvo 1988). In this context, the mechanism proposed for the cryoprotective action of trehalose is the ability to replace water at the polar head group region and in protein structures (Leslie et al. 1995; Luzardo et al. 2000).

Both unsaturation and trehalose affect the permeability of water, a phenomenon that appears important during dehydration (Alonso-Romanowski et al. 1989; Biondi et al. 1992).

Lactic bacteria membranes are composed of cardiolipin, phosphatidylglycerol and three types of glycolipids (Gómez Zavaglia et al. 2000). These membranes do not show a sharp transition in anisotropy at physiological temperatures. In addition, they present a high permeation to Na+ ions in contrast to phosphatidylcholine bilayers (Fernandez Murga et al. 2000).

The survival of lactic bacteria is related to the increase in the unsaturated/saturated fatty acid ratio after growing in appropriate conditions (Gómez Zavaglia et al. 2000), and to the presence of sugars, specifically trehalose, that are effective cryoprotectants. However, little information is available about the action of trehalose on the preservation of lactic bacteria by heat-drying processes in spite of its potential importance. Moreover, the relationship between cell recovery and water exchange during dehydration by heating or by osmosis and its modification by the presence of protective compounds such as trehalose deserves a special analysis because of the complex composition of lipids.

The questions of relevance as regards preservation of cell structure and function are: how effective are the compounds, frequently used as cryoprotectants, to protect bacteria subjected to dehydration–rehydration; how the cell recovery is related to the bacteria membrane integrity, such as the osmotic response; and how temperature may affect the process of dehydration.

For these reasons, we studied the survival of lactic acid bacteria, such as Lactobacillus delbrueckii ssp. bulgaricus after heat and osmotic dehydration in the presence and absence of protective compounds. The recovery after rehydration correlated with the presence of additives during dehydration, and the osmotic response of the rehydrated bacteria.

As mentioned before, membrane integrity affects ion distribution at both sides of the membrane. The surface potential (i.e. the charge distribution at the bacterial surface) is then directly related to the conditions at which the bacteria are grown (Fernandez Murga et al. 2000). Thus, the surface potential (zeta potential) of individual cells measured at different stages of growth can be taken as a marker of the metabolic state of the bacteria.

In this work this criteria has been employed to emphasize the metabolic state of the dried bacteria after rehydration with and without protectants.

Materials and methods

Bacterial strains and growth conditions

Lactobacillus delbrueckii ssp. bulgaricus CIDCA 333 was isolated from a fermented product (Gómez Zavaglia et al. 1998, 1999, 2000). The strain was maintained frozen at −80°C in 120 g l−1 nonfat milk solids. Cultures were grown at 37°C in MRS (Difco, Detroit, MI, USA).

Dehydration procedure

Cultures in the stationary phase were harvested by centrifugation at 10 000 g for 10 min. Aliquots of 1·2 ml were washed with solutions with different concentrations of polyethyleneglycol (PEG) (PM 10000; Sigma Chemical Co., St Louis, MO, USA), sucrose, glycerol, trehalose or CaCl2 (Merck, Darmstadt, Germany).

Washed cells were resuspended in those washing solutions and dehydrated for 30 min in a vacuum centrifuge (Speed Back System ISS 100; Savant Instrument Inc., Farmerdale, NY, USA) at 70°C and immediately resuspended in distilled water.

The number of viable cells, before and immediately after dehydration–rehydration was determined by means of the most probable number (MPN) method in MRS broth. As the number of viable cells after rehydration can be as low as 102 ml−1, the MPN method was employed in all cases to have consistent results. This method has been used and discussed in previous works (Bibiloni et al. 2001).

Growth kinetics after rehydration was followed by measuring the absorbance at 600 nm (A600) during incubation at 37°C in MRS broth.

Osmotic response measured by stopped flow

The permeability experiments were carried out in a rapid reaction stopped flow device (Applied Photo Physics, Leatherhead, UK) attached to a DU 1000 Hitachi (Hitachi Ltd, Tokyo, Japan) spectrophotometer. Aliquots of the bacterial suspension (BS) obtained after rehydration in water (see previous section) were injected with similar volumes of NaCl 10, 20 and 30% w/v in a mixing chamber (mixing time <100 ms). The variation of the absorbance at 600 nm was read as a function of the time.

From the slopes of the curves at time zero plots of ln (AA0)/(AA0) vs time, where A is the absorbance at any time t, A0 the initial absorbance and A the absorbance at equilibrium, the rate of water outflux can be evaluated.

Electrophoretic mobility and zeta potential

The electrophoretic mobility of bacteria (μ) was determined before drying (control samples) and after rehydrating the bacteria dried in different conditions. The rate of migration of individual cells was determined in the stationary layer of a capillary tube at constant temperature when a constant electric field was applied. The effective electrical distance of the capillary tube was calculated by using KCl solutions of known conductivity at 25°C. The rate of migration was determined by microscopic observation of the rectilinear and uniform displacement of individual cells along a reticular lattice of 1 mm length when a potential of 40 V was applied to the electrodes. Measurements were obtained by changing alternatively the polarity of the electrodes to avoid polarization. At least 10 determinations in each direction were made for each sample in order to calculate statistically significant values. The zeta potential (ξ) was calculated with the equation ξ = 4πημ/ɛ, where η and ɛ are the viscosity and the dielectric constant of the solution, respectively.

The zeta potential, which is a measure of the distribution of charges in the surface of a particle suspended in water, varies along the growth curve of the bacteria with different levels of carbon sources (Fernandez Murga et al. 2000). That is, the charge distribution appears linked to the metabolic state of the bacteria and hence the zeta potential has a characteristic value at each stage of growing. Hence, the measurement of this potential was used as a metabolic marker of the bacteria after each treatment.

Growth in milk after drying

After dehydration–rehydration, 0·1 ml of BS was used to inoculate 10 ml of Ultra-High Temperature (UHT) milk and the kinetics of acidification was determined by measuring the pH decrease.


The recovery of heat-dried L. bulgaricus grown in MRS was measured, after rehydration, by the MPN. The highest recoveries were found in all cases when dehydration was carried out in the presence of trehalose, in comparison with the other protectants assayed (sucrose, glycerol, PEG and Ca2+, data not shown). The recovery of cells is a function of the trehalose concentration in the dehydration media. The optimal recovery was 107 at 0·25 m in comparison the 109 in the control culture previous to dehydration (Fig. 1).

Figure 1.

Effect of trehalose concentration during dehydration on the recovery of Lactobacillus bulgaricus after rehydration. Bacteria grown in MRS (109 MPN ml−1) were washed and suspended in the trehalose concentrations indicated before drying. Dried bacteria were rehydrated in water and quantified by MPN in MRS. The log of the most probable number (MPN) is plotted as a function of the trehalose concentration in the dehydration media

Drying in 0·25 m trehalose reduces the lag time of growing in MRS after rehydration to 10 h in comparison with the lag time of bacteria dried without this protectant (Fig. 2). The long lag times correlate with the decrease in the MPN of recovered bacteria for each trehalose concentration (Table 1).

Figure 2.

Recovery of Lactobacillus bulgaricus grown in MRS and dehydrated in media with and without trehalose. Growth kinetics L. bulgaricus in MRS without dehydration (bsl00001); grown in MRS and dehydrated in the absence of trehalose (×); grown in MRS and dehydrated in 0·25 m trehalose (•)

Table 1.  Lag times and most probable number of heat dried bacteria in the presence of different trehalose concentration, after rehydration
Trehalose (m)MPN/mlLag time (h)
0 (control without drying)   6 × 109 5
01·15 × 10322
0·052·74 × 10614
0·1   6 × 10613·5
0·251·15 × 10711
0·5 1·1 × 10512

The integrity of bacteria after the process of dehydration–rehydration, was measured by its osmotic response against NaCl. The osmotic response of the bacteria dried with and without trehalose was followed, after rehydration, in a stopped flow spectrophotometer as described in ‘Materials and methods’. The decrease in absorbance is much more pronounced when the osmotic shock was performed on bacteria that have been dried in the absence of trehalose in comparison with nondried bacteria (Fig. 3). However, the osmotic response of bacteria dried in the presence of 0·25 m trehalose is comparable with the nondried bacteria.

Figure 3.

Osmotic response to hypertonic shock of bacteria dried in the presence of trehalose after rehydration. Control (nondried) bacteria dispersed in water (bsl00001); control (nondried) bacteria shocked with NaCl (•); bacteria dried without trehalose and shocked with NaCl after the rehydration (□); bacteria dried in 0·25 m trehalose shocked with NaCl after rehydration (bsl00066)

The osmotic response of rehydrated bacteria is directly related to trehalose concentration in which they were dehydrated. The rate of volume change taken from the initial slopes of curves of Fig. 3, is a measure of the rate of water outflux. When the bacteria are dispersed in water, no volume change appears to take place as the absorbance remains unchanged. However, the absorbance decreases when the bacteria is dispersed in NaCl 10%. This may be ascribed to the cell volume decrease because of the water outflux driven by the osmotic gradient. The water outflux appears greater in bacteria dried without protectants and is comparable with control samples when the bacteria were dried in 0·25 m trehalose. This concentration is the one at which the maximum recovery was found as shown in Fig. 1 (Fig. 4).

Figure 4.

Rate of volume change (s−1) of rehydrated bacteria, after dehydration in the presence of different trehalose concentrations. Bacteria dried in increasing concentration of trehalose were resuspended in water and shocked the same NaCl concentrations at 25°C

As shown in Table 1, heat drying produces higher lag times and lower number of viable cells. This varies according to the trehalose concentration present in the media during drying. The long lag times could be due to a great number of dead or injured cells. As the zeta potential is a measure focusing on single bacteria, we measure its value for the different heat drying conditions after rehydration.

The zeta potential of control cells (−0·055 V) shifted to −0·065 V in bacteria dehydrated without trehalose after rehydration. The resulting zeta potential approaches that found in control cells with the increase of trehalose in the drying media. Bacteria dehydrated in 0·25 m trehalose showed the same value of zeta potential than nondehydrated ones. This value corresponds to the shorter lag times found with dehydrated bacteria (Fig. 5).

Figure 5.

Correlation between the zeta potential (•) and the lag time (○) of rehydrated cells with the trehalose concentration present during drying. Bacteria dried in different trehalose concentrations were resuspended in water and the zeta potential determined at 25°C

Trehalose appears to preserve the osmotic properties of the bacteria (Fig. 3). It was of interest to test the action of trehalose in the osmotic dehydration, i.e. when the bacteria were exposed to a hypertonic media at constant temperature. Bacteria grown in MRS were exposed to 25% PEG either during 3 or 6 h at 25°C. In Fig. 6, it is observed that the lag time of bacteria shocked with PEG during 6 h was around the same as that found in dehydration by heating, as shown previously in Fig. 2. Interestingly, the lag time was reduced when the osmotic shock with PEG was carried out in the presence of trehalose 0·25 m. The effect of trehalose was also found when the exposure to the same osmotic shock was shorter.

Figure 6.

Effect of trehalose on the growth kinetics of bacteria subjected to osmotic dehydration with PEG at room temperature. Control (nonshocked) bacteria (bsl00001). Bacteria shocked during 6 h with 25% PEG (○); bacteria shocked during 6 h with 25% PEG in the presence of 0·25 m trehalose (•); bacteria shocked during 3 h with 25% PEG (▵); bacteria shocked during 3 h with 25% PEG in the presence of 0·25 m trehalose (bsl00066). Bacteria grown in MRS were added to PEG 25% or PEG 25% supplemented with 0·25 m trehalose and incubated during 3 or 6 h. After the incubation an aliquot was dispersed in MRS and the kinetics of growing was determined at 37°C


The recovery of bacteria dehydrated by heat or by osmosis increases in the presence of trehalose and appears related to the preservation of the zeta potential and of the osmotic response of the cells.

The fact that the cell recovery of bacteria subjected to dehydration is increased by trehalose, with a recovery of the control values of the osmotic response strongly suggests that trehalose can hinder the damage of cell structures related to water permeation.

This mechanism appears to be independent of the process followed to dehydrate the cell because recovery is also enhanced by trehalose when dehydration is performed by osmosis at 20°C. This result indicates that damage is produced by displacement of water from functionally relevant structures.

The absorbance of the suspension remains unchanged when the bacteria are dispersed in water. The bacterial volume in water reaches a maximum because of the presence of the cell wall (at least during the first 100 s). In these conditions, the number of viable cells remains unaltered. In contrast, the decrease observed when the cells are dispersed in hypertonic NaCl (Fig. 3) denotes that cell volume decreases because of the water outflow imposed by the hypertonic gradient. The different rates of changes in the absorbance are indicative of different rates of water outflow in bacteria subjected to different osmotic gradients.

To understand these responses, it is useful to discuss in more detail the procedure followed in the experiment. The dried cells were rehydrated in water during 30 min. After this time, they were mixed with a fixed concentration of NaCl (usually 10%) and the volume changes followed by the stop-flow measure. As the volume decrease is higher in nonprotected cells than the control cells exposed to the same osmotic gradient, the barrier for water permeation appears to be reduced.

As a comparable osmotic response was obtained in rehydrated bacteria heat-dried in the presence of trehalose, it may be interpreted that the barrier related to water permeability is maintained by this sugar when the cells are dehydrated. A possible effect of trehalose protective action is that it may be incorporated into the cell or to induce the synthesis of metabolites that protect against stress. However, it must be noticed that cells after heat dehydration were resuspended in water. In these conditions, it is unlikely that cells may have metabolic activity to transport trehalose or synthesized protective compounds.

The protective action of trehalose has been ascribed to its ability to replace water in proteins and membrane structures (Leslie et al. 1995). The displacement of water to enhance trehalose interaction is induced by heating or freezing. The present results indicate that trehalose can also replace water when it is displaced by osmotic dehydration performed by PEG in excess of water.

An interesting observation in relation to bacterial response is the correlation of the lag time with the changes in zeta potential (Fig. 4). The zeta potential is an electrical potential arising from the presence of charges on the cell surface. These charges could be constitutive groups fixed on the cell wall or cell membrane. However, the zeta potential also depends on the metabolic state of the bacteria, as bacterial growth can be followed by changes in zeta potential when they are grown with different levels of carbon sources (Fernandez Murga et al. 2000). Thus, a contribution to zeta potential could be due to the ion condensation around the cell because of active transport. Therefore, zeta potential is a marker of the metabolic state. In addition, the zeta potential influences the water permeability in lipid bilayers (Disalvo 1999).

As discussed previously, the zeta potential is a measure on a single cell. Hence, the values of zeta potential indicate the presence of injured cells. These injured cells can also affect the lag time.

The presence of trehalose, which recovers the zeta potential and the osmotic response corresponding to the control cells, hinders the injury and reduces the lag time.

Cells showing a more negative zeta potential were found to have a drastic response to the hypertonic shock. Both the osmotic response and zeta potential value of dried nonprotected cells were restored to those of control cells with trehalose. It might be possible that the electrical barrier affecting water permeation can be preserved by trehalose.

However, more complex mechanisms of protection induced by trehalose, such as the synthesis of shock proteins among them, cannot be discarded and will be the subject of future analysis.

In conclusion, cryoprotectants such as trehalose are equally effective in protecting lactic acid bacteria dried by air desiccation or by osmotic dehydration. The mechanism of the action of trehalose appears related to water permeation as trehalose equally protects cells under hypertonic stress and preserves the osmotic response of the bacteria.


This work was supported with funds from Agencia Nacional de Promoción Científica y Tecnológica, grant PICT 06047.

EAD is a member of the research career of CONICET (National Research Council, Argentina). GLDA is a member of the research career of CIC (Comisión de Investigaciones Científicas de la Provincia de Buenos Aires, Argentina).

AGZ is recipient of a fellowship from CONICET. EET is a student fellow of the University of Buenos Aires