Trehalose synthesis is induced upon exposure of Escherichia coli to cold and is essential for viability at low temperatures

Abstract

Trehalose accumulates dramatically in microorganisms during heat shock and osmotic stress and helps protect cells against thermal injury and oxygen radicals. Here we demonstrate an important role of this sugar in cold-adaptation of bacteria. A mutant Escherichia coli strain unable to produce trehalose died much faster than the wild type at 4°C. Transformation of the mutant with the otsA/otsB genes, responsible for trehalose synthesis, restored trehalose content and cell viability at 4°C. After temperature downshift from 37°C to 16°C (“cold shock”), trehalose levels in wild-type cells increased up to 8-fold. Although this accumulation of trehalose did not influence growth at 16°C, it enhanced cell viability when the temperature fell further to 4°C. Before the trehalose build-up, levels of mRNA encoding OtsA/OtsB increased markedly. This induction required the σ factor, RpoS, but was independent of the major cold-shock protein, CspA. otsA/B mRNA was much more stable at 16°C than at 37°C and contained a “downstream box,” characteristic of cold-inducible mRNAs. Thus, otsA/otsB induction and trehalose synthesis are activated during cold shock (as well as during heat shock) and play an important role in resistance of E. coli (and probably other organisms) to low temperatures.

Trehalose is a nonreducing disaccharide (α-d-glucopyranosyl-1,1-α-d-glucopyranoside) present in many prokaryotic and eukaryotic organisms. Bacterial and yeast cells accumulate trehalose to very high levels (up to 500 mM) in response to heat shock (1) and osmotic stress (2), and during stationary phase (3, 4). Initially, trehalose was thought to serve as a reserve metabolizable substrate (3), but recent studies indicate that this sugar instead plays a major role in cell protection against harsh environmental conditions (5, 6). In fact, the build-up of trehalose upon heat shock has been shown to be a more important determinant of thermotolerance than the induction of heat-shock proteins (7). However, our understanding of the molecular mechanisms of trehalose's protective effects is just starting to emerge. In yeast, trehalose was shown to stabilize proteins in their native state and to reduce their heat-induced denaturation and aggregation (8, 9). Recently, trehalose was also demonstrated to decrease oxidative damage to cell proteins by oxygen radicals and thus to increase the tolerance of yeast to reactive oxygen species (10). Consequently, yeast strains deficient in trehalose production are quite sensitive to high temperature and to oxidants (10, 11).

Trehalose is found at high concentrations in many organisms that naturally survive dehydration—for example, baker's yeast, some plants, many bacteria, and certain microscopic animals (12). The ability of many of these organisms to survive in a dry state correlates with their trehalose content (5, 12, 13). Furthermore, although mammalian cells cannot produce trehalose, when trehalose-synthesizing enzymes from Escherichia coli were expressed in human fibroblasts, these cells showed much greater resistance to desiccation than normal fibroblasts (14).

The addition of trehalose to cells has also been found to protect them against freezing. Exogenous trehalose enhances viability of bacteria (15) and yeast (16) during freezing. Also, when genes for trehalose synthesis were introduced into mammalian cells, they could survive much longer in the frozen state (17). The mechanism by which trehalose mediates tolerance to freezing or desiccation is not clear, but presumably involves a stabilization of certain cell proteins and/or lipid membranes (18, 19).

This ability of trehalose to protect against freezing led us to hypothesize that trehalose accumulation in microorganisms may also serve as a natural adaptation to decreased temperatures that helps prevent loss of viability in the cold or upon freezing. To test this hypothesis, we chose E. coli, which like other enterobacteria, must often encounter very rapid, large temperature downshifts in their natural environment (for example, upon defecation by warm-blooded animals at low ambient temperatures). Such organisms have evolved adaptive mechanisms, termed the “cold-shock response,” that alter cell composition to adjust to rapid decreases in ambient temperature. The present studies tested whether trehalose may be involved in cold-adaptation in bacteria. Here we demonstrate that enzymes for trehalose synthesis are induced in wild-type E. coli under cold-shock conditions, and that the resulting accumulation of this sugar increases the cells' viability when temperature falls to near freezing.

Materials and Methods

Bacterial Strains.

Bacterial strain MC4100 was used as a wild-type strain throughout most of this study. The trehalose-deficient otsA⁻ strain (FF4169) was provided by R. Hengge-Aronis (Univ. of Konstanz, Konstanz, Germany); the plasmid carrying otsA and otsB under the regulation of tac promoter (pRHO700) was provided by W. Boos (Univ. of Konstanz). The ΔrpoS strain (ZK1000) and the isogenic wild type (ZK126) were provided by R. Kolter (Harvard Medical School). The ΔcspA strain (WB002) and the isogenic wild type (JM83) were obtained from M. Inouye (University of Medicine and Dentistry of New Jersey, Piscataway). All cells were routinely grown in LB or in M9 medium with 0.4% glycerol as a carbon source.

Trehalose Assay.

Trehalose content was estimated as described previously (7, 10). Trehalose was extracted by boiling cell pellets at 95°C for 20 min. Trehalose in the supernatant was converted to glucose with trehalase (Sigma), which was then measured by a glucose assay kit (Sigma). The preexisting glucose in each sample was determined in a control reaction without trehalase and subtracted from the total glucose.

Cell Viability Assay.

Cells were grown in LB or minimal medium at 37°C until mid-logarithmic phase (OD₆₀₀ about 0.5). The cultures were diluted with the growth medium, and equal amounts of cells were plated on the Petri dishes. The plates were then sealed in plastic bags to prevent drying and stored at 4°C. At different times, the number of colonies that survived were measured by transferring the plates to 37°C.

Preparation of the otsA/B Probe for Northern Blotting.

The otsA/B probe was made by PCR using pRHO700 plasmid as a template. The primers were designed to amplify a fragment of otsA gene located between nucleotides 1978780 and 1979510 of the E. coli genome: primer 1, 5′-GGAGTGGTGAAACAGGGAATGAGG; primer 2, 5′-CCATGATGCTGCGGATATTTTTCC. The rs1 probe was made with E. coli genomic DNA as a template: primer 1, 5′-TCGCGGAAACCCACAAC; primer 2, 5′-GTCGCCAACGCTCAGAAC.

Results

Trehalose Levels Correlate with Cell Resistance to the Cold.

In E. coli, trehalose biosynthesis is catalyzed by two enzymes: trehalose-6-phosphate synthase (OtsA) and trehalose-6-phosphate phosphatase (OtsB) (20). Strains lacking OtsA, although unable to synthesize trehalose, grow at 37°C at rates similar to those of wild-type cells (data not shown). To check whether the cells' ability to withstand cold correlates with their content of trehalose, we used an otsA-mutant strain (kindly provided by R. Hengge-Aronis) and a strain overproducing otsA/B on the plasmid (kindly provided by W. Boos). The wild-type, trehalose-overproducing, and trehalose-deficient strains were grown to mid-logarithmic phase in minimal medium at 37°C, and trehalose content was measured as described in Materials and Methods. This sugar was not detectable in the mutant and was present at very low levels in the wild type during logarithmic growth, but its level increased at least 8-fold when the trehalose-synthesizing enzymes were overexpressed on a plasmid (Fig. 1A).

Figure 1.

Cell viability at 4°C correlates with trehalose content. Wild type (WT; MC4100), wild type carrying pRHO700 plasmid, otsA⁻ strain, and otsA⁻ carrying pRHO700 were grown to the mid-logarithmic phase at 37°C. Isopropyl β-d-thiogalactoside (IPTG; 1 mM) was added to induce trehalose synthesis in trehalose-overproducing strains, and growth continued for 1.5 h. (A) Aliquots were taken from each culture to measure trehalose content. (B) Equal amounts of cells were plated on Petri dishes, sealed in plastic bags, and maintained at 4°C. At the times indicated, a set of plates was transferred to 37°C and surviving colonies were counted. This figure represents a typical experiment (one of four independent experiments).

To test whether trehalose content affects viability at low temperatures that prevent normal growth, logarithmically growing cultures were diluted and plated on Petri dishes, which were then maintained at 4°C. At different times, the number of surviving colonies were measured by transferring the plates to 37°C. All three strains lost viability exponentially at 4°C but at quite different rates. As shown in Fig. 1B, after 5 days at 4°C, about 25% of the wild-type cells and significantly less (10%) of the trehalose-deficient cells survived, whereas much more (over 50%) of the trehalose-overproducing cells were still viable at that time (Fig. 1B). Although the absolute rate of cell loss at 4°C varied between experiments, similar differences between the three strains were seen in four independent experiments. The findings in Fig. 1B represent a time-dependent loss of viability at 4°C and not a time-dependent loss of ability to withstand the large temperature up-shift from 4°C back to 37°C, because a similar decrease in viable colonies was found when the plates were transferred from 4°C to 16°C. Thus, even though trehalose content did not affect growth at 37°C (data not shown), or 16°C (see below), increasing trehalose levels can enhance viability at 4°C, whereas a lack of trehalose correlates with reduced ability to survive at this temperature. It is also noteworthy that these three strains maintained the same trehalose content (as they showed at 37°C) during incubation for up to 5 days at 4°C (data not shown).

Cold-Tolerance Can Be Restored to the Mutant by Inducing Trehalose Synthesis.

Because trehalose was barely detectable in wild-type cells during exponential growth, it appeared questionable whether the cold-sensitivity of the mutant was in fact attributable to the lack of this low amount of trehalose. We therefore tested whether this cold-sensitive phenotype could be reversed by restoring trehalose synthesis. The mutant was transformed with a plasmid carrying the otsA/B genes under the regulation of the tac promoter. In the presence of isopropyl β-d-thiogalactoside (IPTG), trehalose content increased (Fig. 1A), and at the same time, the viability of the cells at 4°C increased above the levels seen in the wild-type cells to those of the trehalose-overproducing strain (Fig. 1B). Thus, viability at 4°C varies with the cell's trehalose content.

Trehalose Content and Cold-Tolerance Increase upon Cold Shock.

To find out if this protective effect of trehalose may be part of a physiologically regulated response, we tested whether trehalose accumulates naturally after a temperature downshift. When logarithmically growing E. coli cells are transferred from 37°C to 16°C, growth stops, and after a lag period of several hours (called the “acclimation period”), growth is resumed at a much slower rate (21). During this acclimation period, a specific set of cold-shock proteins is induced that is believed to serve adaptive functions (21, 22). Because increased trehalose content appeared to be advantageous to the cells at 4°C, we tested whether trehalose levels increase during the acclimation period as a part of the cold-shock response. Wild-type and trehalose-deficient cells were grown at 37°C until mid-logarithmic phase, and then shifted to 16°C. Aliquots were then taken at different times to measure trehalose content. The amount of trehalose in the wild type increased about 3-fold by 2 h at 16°C and 7-fold (a maximal increase) by 5 h. However, after 6 h, trehalose content decreased to a new steady-state level (Fig. 2A), which was then maintained for at least 3 h after cell growth had resumed at 16°C (data not shown). Thus, trehalose accumulates during the cold-shock response, presumably as a part of the cell's adaptation to the fall in temperature, and the time course of its accumulation is very similar to that of known cold-shock proteins (23).

Figure 2.

Accumulation of trehalose at 16°C correlates with increased viability upon further temperature downshift to 4°C. Wild-type (WT; MC4100) and trehalose-deficient (otsA) cells were grown to mid-logarithmic phase at 37°C and then transferred to 16°C. At different times at 16°C, aliquots were taken to measure intracellular trehalose content (A). For viability assay (B), aliquots were taken at the same times as in A, and cells were diluted, plated on Petri dishes, and maintained at 4°C (as in Fig. 1). Viable colonies were counted at day 0 and after 5 days at 4°C.

It is noteworthy that, in contrast to these findings upon shift to 16°C, none of the strains used in our experiments showed any ability to increase trehalose content if shifted directly from 37°C to 4°C (data not shown). However, although induced at 16°C, trehalose is not important for cell viability or growth at this temperature, because the wild type and the trehalose-deficient mutant showed the same length of the lag period upon shift to 16°C, then resumed growth at similar rates, and reached similar saturation density (data not shown). Thus, trehalose accumulation at 16°C provides protection only upon further temperature downshift, e.g., to 4°C, where presumably induction of most genes could not occur.

To test this idea further, cells were maintained at 16°C for different times, and then their trehalose content (Fig. 2A) and their viability at 4°C (Fig. 2B) were analyzed. Cultures were diluted, and equal amounts of cells were plated on Petri dishes. After plating, a set of Petri dishes was left at 37°C to determine initial cell number, and the rest were incubated at 4°C. After 5 days at 4°C, the plates were transferred to 37°C and the surviving colonies were counted. As shown in Fig. 2, the ability to survive at 4°C (Fig. 2B) correlated closely with the cells' initial trehalose content (Fig. 2A). Both reached maximal levels at 5 h at 16°C, and then decreased slightly by 6 h (Fig. 2). Thus, trehalose accumulation during the cold shock (down-shift to 16°C) appears to be a very important factor conferring resistance to the further fall in temperature to 4°C.

Interestingly, although preincubation at 16°C did not increase the content of trehalose in the otsA⁻ cells (Fig. 2A), they still showed some increase in cold-tolerance (Fig. 2B), although much less than that seen in wild-type cells. After 5 h at 16°C, only 20% of the mutant cells survived for 5 days at 4°C, in comparison to 75% of the wild type (Fig. 2B). Therefore, although trehalose accumulation appears to be a primary adaptation leading to the acquisition of cold-tolerance, other components of the cold-shock response also contribute to cell protection at 4°C.

otsA/B mRNA Is Induced During Cold Shock Before the Trehalose Build-Up.

Trehalose accumulation during the acclimation phase at 16°C raises the possibility that enzymes responsible for trehalose biosynthesis are cold-shock proteins. To test this possibility, the levels of otsA/B mRNA were measured at 37°C and at different times after the shift to 16°C. Total cellular RNA was isolated by using Trizol reagent (GIBCO Life Technologies) and analyzed by Northern blotting. Because otsA and otsB constitute an operon transcribed as a single mRNA (24), the same probe could be used to follow the levels of transcription of both genes. The probe was prepared by PCR, using an otsA/B-expressing plasmid as the template (see Materials and Methods). As shown in Fig. 3A, the level of otsA/B mRNA was very low at 37°C, but after the shift to 16°C, its level increased, reaching a maximum in 4 h, and after 6 h dropping to a new steady-state level, which was then maintained for several hours, after bacterial growth had resumed at 16°C (data not shown). This increase in the level of otsA/B transcript during the acclimation phase resembles that of other cold-shock genes (21, 25, 26). Thus OtsA and OtsB are previously unknown members of this group. This increase in otsA/B mRNA before the build-up of trehalose is probably responsible for its accumulation during cold shock and suggests increased transcription of otsA/B.

Figure 3.

Northern blot analysis of otsA/B mRNA in different strains subjected to cold shock. Cells were grown to the mid-logarithmic phase at 37°C and shifted to 16°C. At different times at 16°C, 20-ml aliquots were taken, and total RNA was isolated by using Trizol reagent and analyzed by Northern blotting using otsA probe. (A) otsA/B mRNA is induced during cold-shock response in MC4100. (B) otsA/B mRNA is induced in the wild type but not in the ΔrpoS mutant. (C) There is no difference in the rates of otsA/B induction between the wild type and the ΔcspA mutant.

Induction of otsA/B Transcription at 16°C Requires RpoS σ Factor.

Although there have been no prior reports that specialized σ factors are responsible for induction of cold-shock proteins, several observations led us to test whether the stress-related σ factor RpoS is involved in the induction of otsA/B in the cold: (i) RpoS stimulates OtsA/B expression during osmotic stress and in stationary phase (4, 24); (ii) RpoS is involved in the regulation of transcription under various conditions where growth falls (27); (iii) intracellular levels of RpoS increase during steady-state growth at 20°C (28); and (iv) an rpoS mutant loses viability faster than the wild type at 4°C (24).

To test whether induction of otsA/B upon cold shock may be RpoS-dependent, a wild-type strain and rpoS-deletion mutant (kindly provided by R. Kolter) were grown at 37°C to mid-logarithmic phase and shifted to 16°C. otsA/B mRNA levels were measured before and at different times after the cold shock. In contrast to the wild type, in the rpoS-deletion strain there was no induction of otsA/B mRNA upon temperature downshift (Fig. 3B). Thus, induction of otsA/otsB at 16°C is strictly dependent on RpoS function. This requirement for RpoS further suggests that the mechanism of otsA/B induction upon cold shock includes increased transcription.

CspA Is Not Involved in the Regulation of otsA/B.

The major cold-shock protein, CspA, accounts for more than 10% of total protein synthesis during the acclimation phase (29). CspA has been proposed to function as an RNA chaperone at low temperatures (30) and has been implicated in transcriptional regulation of at least two cold-shock genes, hns (31) and gyrA (32). To test whether CspA function is also required for otsA/B mRNA induction at 16°C, we used a cspA deletion strain, kindly provided by M. Inouye. The cspA mutant and isogenic wild type were shifted from 37°C to 16°C, and levels of otsA/B mRNA were determined before and at different times after the shift. Although in these strains otsA/B mRNA levels increased more slowly than in MC4100, there was no difference in the rates of mRNA accumulation between the wild type and the mutant (Fig. 3C). Thus, the increase in the level of otsA/B mRNA at 16°C is independent of CspA.

otsA/B mRNA Is More Stable at 16°C Than at 37°C.

In addition to transcriptional changes, an important factor in the induction of the cold-shock response is stabilization of the cold-shock mRNAs at low temperatures (33). For example, the half-life of the cspA mRNA increases from less than 1 min at 37°C to about 20 min at 16°C, which contributes to the dramatic increase in cspA mRNA in the cell (25, 33). To test whether otsA/B mRNA is also stabilized upon cold shock, cells were grown at 37°C until mid-logarithmic phase and shifted to 16°C to allow induction of cold-shock mRNAs. After 2 h, half of the culture was left at 16°C, and another half returned to 37°C. In both cultures, transcription was stopped by addition of rifampicin, and aliquots were taken at different times to measure levels of otsA/B mRNA as well as a control mRNA encoding the ribosomal protein S1, whose expression does not change after temperature downshift (21). As shown in Fig. 4, the rate of otsA/B mRNA disappearance was much faster at 37°C than at 16°C. The half-life of otsA/B mRNA was less than 2 min at 37°C, but was about 20 min at 16°C. By contrast, the half-life of rs1 mRNA did not change after the shift from 37°C to 16°C. This 10-fold increase in the stability of otsA/B mRNA must also be contributing to the induction of OtsA/OtsB during the cold shock.

Figure 4.

otsA/B mRNA is more stable at 16°C than at 37°C. MC4100 cells were grown at 37°C until mid-logarithmic phase and shifted to 16°C to allow induction of cold-shock mRNAs. After 2 h, the culture was divided into two parts: half was left at 16°C, and the temperature of the other half was adjusted to 37°C by mixing with preheated LB. To stop transcription, rifampicin was added (to final concentration of 200 μg/ml) to both cultures, and aliquots were taken at different times to measure otsA/B mRNA levels by Northern blotting. As a control, in the same blot the levels of rs1 (ribosomal protein S1) mRNA were also analyzed.

otsA/B mRNA Contains a “Downstream Box” (DB).

Because OtsA and OtsB appeared to be cold-shock proteins, we searched for sequence similarities in the regulatory regions of otsA/B and other cold-shock genes. One such sequence, common to all cold-shock genes, which is believed to be critical for their induction in the cold, is the DB (25). DB is a 14-nucleotide-long sequence downstream from the initiation codon of mRNA complementary to an anti-downstream box (ADB) located at the 3′ terminus of 16S rRNA (34). The level of complementarity between DB and ADB correlates with the rate of the cold-induction, and its deletion resulted in the complete lack of cold-induction (25). DB was suggested to allow efficient association of cold-shock mRNA with the ribosome and initiation of translation in the cold (25). We found that a DB is also present in otsA (Fig. 5). Moreover, the DB of otsA mRNA shows an even higher complementarity with rRNA ADB sequence than that of CspA, which is induced more dramatically than other cold-shock proteins (25). Thus otsA/B mRNA, like mRNAs of other cold-shock proteins, has the classic DB box, which seems to allow preferential translation of cold-shock proteins when overall translation is strongly suppressed.

Figure 5.

otsA mRNA contains a DB sequence common to other cold-shock mRNAs. The number of potential base pairs (shown in uppercase letters) between DBs of cspA, cspB, and cspG and ADB of 16S rRNA was given by Inouye and coworkers (25).

Discussion

Trehalose Accumulation Is of Key Importance for Viability at Low Temperatures.

Production of trehalose by stressed cells appears to be a critical adaptation that protects microorganisms against a wide variety of potentially lethal conditions, including high temperatures, oxygen radicals, and high osmotic strength. We demonstrate here that trehalose also accumulates in E. coli during cold shock (i.e., 16°C). This response, although not affecting growth at 16°C, increased cell viability upon further reduction of temperature to 4°C. Thus the induction of trehalose appears to be a protective response that occurs in anticipation of a further fall in temperature, which constitutes a more severe stress to the cell.

It is noteworthy that in the mutant cells unable to produce trehalose, the cold-shock response provided much less protection against low temperatures than in wild-type cells. Thus trehalose accumulation seems to be of primary importance for viability at 4°C. Although in the absence of trehalose induction of the other cold-shock proteins also provided some protection at 4°C, only one such protein, trigger factor, has actually been shown to be important for viability at low temperatures (35). Presumably, trehalose functions together with the other cold-shock proteins in protecting cell constituents against low temperatures, just as it acts synergistically with heat-shock proteins in reducing damage to cell proteins at high temperatures (7, 9).

OtsA/OtsB Induction at 16°C Requires RpoS-Dependent Transcription.

The accumulation of trehalose upon downshift to 16°C occurs because of induction of the trehalose-synthesizing enzymes, OtsA and OtsB, in response to the cold shock. Their induction requires the specialized σ factor, RpoS, which is critical in the adaptation of E. coli to a number of other stressful conditions, including carbon starvation, high temperature, high osmolarity, and entrance into stationary phase (36). Although we have not proven that RpoS acts directly in the transcription of otsA/otsB in the cold, a direct action seems most likely because of the role RpoS plays in osmotic induction of otsA/otsB (4, 37). RpoS levels are higher during steady-state growth at 20°C than at 37°C (28, 38), which had suggested a special role at lower temperatures (28). Our finding of a RpoS requirement for otsA/otsB induction in cold shock raises the possibility that RpoS is involved in the expression of other cold-shock proteins.

Another important question is what promoter elements are involved in RpoS-dependent induction of otsA/B. There are no sharp differences between the promoter sequences recognized by σ^S and σ⁷⁰, and many genes are recognized by either σ subunit of RNA polymerase (39). For example, upon osmotic stress, induction of otsA/B transcription still occurs in a mutant lacking RpoS, although to a much lower extent than in the wild type (24). In contrast, no increase in otsA/B mRNA was seen during cold shock in the ΔrpoS mutant (Fig. 3B). Although the otsA/B promoter area, as defined previously (20), lacks the characteristic −10 sequence element for RpoS-dependent promoters (CTATACT) (37, 39), such an element is present 10 nucleotides downstream from the reported transcription start point. Possibly, this element (by itself or in concert with others) plays a key role in the cold-induction of otsA/B.

OtsA/OtsB Induction at 16°C Involves Posttranscriptional Adaptations.

Induction of cold-shock proteins upon temperature downshift is tightly regulated at multiple levels, including transcription, translation, and mRNA degradation (25). These three factors appear to play a role in otsA/B induction during cold shock. We observed a dramatic increase in otsA/B mRNA during the acclimation phase before the build-up of trehalose, suggesting an activation of transcription. This conclusion is supported by the requirement for the specialized σ factor, RpoS, for the accumulation of otsA/B mRNA. In addition, the stability of this transcript (but not rs1 mRNA) increased dramatically at 16°C. A similar stabilization of mRNA was shown to be one of the important mechanisms causing induction of CspA during cold shock (25, 33).

In addition to enhanced transcription and mRNA stabilization, another factor probably contributing to OtsA induction at 16°C is the presence in otsA mRNA of the DB sequence, which is critical for selective translation of cold-shock proteins (25). During the acclimation phase, overall protein synthesis decreases dramatically mainly because of the inability of ribosomes to translate most cellular mRNAs, aside from those encoding cold-shock proteins (22). The presence of DB sequence in these mRNAs facilitates their association with ribosomes. The DB sequence was originally proposed to enhance the interaction of these mRNAs with the ribosome by base pairing to ADB sequence in 16S rRNA (25, 40) (see Fig. 5), but Dahlberg and coworkers (41) demonstrated that mRNA⋅rRNA base pairing cannot account for the translational enhancement by the DB element. Although the exact mechanism by which this sequence promotes translation remains uncertain, the presence of DB in otsA mRNA strongly suggests its preferential translation upon temperature downshift.

OtsA and OtsB are both cold-shock and heat-shock proteins and are critical for viability at both high and low temperatures. By contrast, other heat-shock proteins are specifically repressed in the cold, whereas most cold-shock proteins can hardly be detected at 37°C (21, 22), and certainly not during heat shock (42°-45°C). Interestingly, a DB element is also present in the mRNA of one of the key heat-shock proteins, the σ factor RpoH, and is essential for its induction at high temperatures (42). Because the DB sequence can promote translation of certain mRNAs at high as well as low temperatures, this mechanism of translational control may contribute to the increase in OtsA/OtsB expression upon heat shock as well as cold shock.

How May Trehalose Protect Against Low Temperatures?

The present findings raise the fundamental question of how trehalose enhances cell viability at low temperatures, and the converse question of what causes death at 4°C in both the wild type and trehalose-deficient cells. At high temperatures, trehalose can protect cells by acting as a “chemical chaperone” (43), which reduces heat-induced denaturation and aggregation of proteins (8, 9). Because a number of proteins are also known to denature and precipitate in the cold where the hydrophobic effect is relatively weak (44–49), it is possible that at low temperatures trehalose also prevents the denaturation and aggregation of specific proteins. Trehalose also functions as a free radical scavenger in vivo and thus protects proteins and other cell constituents from oxidative damage (10). While defense against oxygen radicals is important at elevated temperatures (50), it is unclear whether damage by free radicals constitutes a greater stress in the cold, although enhanced generation of oxygen free radicals has been reported in mammalian tissues in response to cold stress (51). Another possible protective role for trehalose at 4°C may be in stabilizing cell membranes, whose fluidity decreases during temperature downshift. It is noteworthy that exogenous trehalose helps protect a variety of organisms against freezing (17, 52), and that the maximal protection is seen when trehalose is present on both sides of the cell membrane (16, 17). Possibly trehalose enhances viability at low temperatures by multiple mechanisms, including ones distinct from those proposed above.

An important goal for future studies will be to determine the molecular basis for the ability of trehalose to protect against low temperatures. This question may be of broad biological significance because trehalose and close homologs of otsA/B genes are present in a wide variety of organisms, including yeast, Drosophila, and Caenorhabditis elegans. Trehalose's function and the regulation of these genes in higher organisms have received very little attention. Our related experiments show that in Saccharomyces cerevisiae and Drosophila, mRNAs for genes homologous to otsA/B are also induced in the cold (O.K. and A.L.G., unpublished results). Thus, the protective mechanisms we uncovered in E. coli may not be restricted to bacteria but may also be important for the survival of higher organisms, especially cold-blooded species and plants, in cold environments.

Abbreviations

DB

downstream box

ADB

anti-DB