Induction of neutral trehalase Nth1 by heat and osmotic stress is controlled by STRE elements and Msn2/Msn4 transcription factors: variations of PKA effect during stress and growth


  • Harald Zähringer,

    1. Institut für Biochemie und Molekularbiologie, Universität Freiburg, Hermann-Herder-Str. 7, D-79104 Freiburg, Germany.,
    Search for more papers by this author
  • Johan M. Thevelein,

    1. Laboratorium voor Moleculaire Celbiologie, Institute of Botany and Microbiology, Katholieke Universiteit Leuven, Kardinaal Mercierlaan 92, B-3001 Leuven-Heverlee, Flanders, Belgium.,
    Search for more papers by this author
  • Solomon Nwaka

    1. Institut für Biochemie und Molekularbiologie, Universität Freiburg, Hermann-Herder-Str. 7, D-79104 Freiburg, Germany.,
    2. National Institute of Bioscience and Human Technology, Higashi 1-1, Tsukuba, Ibaraki 305, Japan.
    Search for more papers by this author

Solomon Nwaka. QuantaNova Canada Ltd, 15 Chipman Drive, PO Box 818, Kentville, Nova Scotia B4N 4H8, Canada. E-mail; Tel. (+1) 902 679 7590; Fax (+1) 902 678 9440.


Saccharomyces cerevisiae neutral trehalase, encoded by NTH1, controls trehalose hydrolysis in response to multiple stress conditions, including nutrient limitation. The presence of three stress responsive elements (STREs, CCCCT) in the NTH1 promoter suggested that the transcriptional activator proteins Msn2 and Msn4, as well as the cAMP-dependent protein kinase (PKA), control the stress-induced expression of Nth1. Here, we give direct evidence that Msn2/Msn4 and the STREs control the heat-, osmotic stress- and diauxic shift-dependent induction of Nth1. Disruption of MSN2 and MSN4 abolishes or significantly reduces the heat- and NaCl-induced increases in Nth1 activity and transcription. Stress-induced increases in activity of a lacZ reporter gene put under control of the NTH1 promoter is nearly absent in the double mutant. In all instances, basal expression is also reduced by about 50%. The trehalose concentration in the msn2 msn4 double mutant increases less during heat stress and drops more slowly during recovery than in wild-type cells. This shows that Msn2/Msn4-controlled expression of enzymes of trehalose synthesis and hydrolysis help to maintain trehalose concentration during stress. However, the Msn2/Msn4-independent mechanism exists for heat control of trehalose metabolism. Site-directed mutagenesis of the three STREs (CCCCT changed to CATCT) in NTH1 promoter fused to a reporter gene indicates that the relative proximity of STREs to each other is important for the function of NTH1. Elimination of the three STREs abolishes the stress-induced responses and reduces basal expression by 30%. Contrary to most STRE-regulated genes, the PKA effect on the induction of NTH1 by heat and sodium chloride is variable. During diauxic growth, NTH1 promoter-controlled reporter activity strongly increases, as opposed to the previously observed decrease in Nth1 activity, suggesting a tight but opposite control of the enzyme at the transcriptional and post-translational levels. Apparently, inactive trehalase is accumulated concomitant with the accumulation of trehalose. These results might help to elucidate the general connection between control by STREs, Msn2/Msn4 and PKA and, in particular, how these components play a role in control of trehalose metabolism.


In yeast, trehalose is accumulated under adverse conditions as a compatible solute that protects proteins and membranes against damage during a variety of stress conditions (for reviews, see Wiemken, 1990; Piper, 1993, 1998; Nwaka and Holzer, 1998). Trehalose synthesis is catalysed by the trehalose synthase complex encoded by TPS1, TPS2, TPS3 and TSL1 (Bell et al., 1992; De Virgilio et al., 1993; Bell et al., 1998). The hydrolysis of trehalose is catalysed mainly by the neutral trehalase Nth1, encoded by NTH1 (Kopp et al., 1993; Nwaka et al., 1995a; Van Dijck et al., 1995). Deletion of NTH1 results in impaired trehalose hydrolysis, higher trehalose concentrations and impaired recovery of cells after severe heat shock compared with wild-type cells (for a review, see Nwaka and Holzer, 1998). The expression of NTH1 is induced by different stress conditions, such as heat shock, oxidative stress and toxic chemicals, and also under starvation conditions, such as glucose exhaustion in the medium (Nwaka et al., 1995b; Zähringer et al., 1997, 1998). The function of Nth1 during stress recovery is not well understood. Under the stress condition, for instance heat stress, trehalose may function as a chemical chaperone, binding to proteins or critical cellular targets to prevent their denaturation. However, during recovery from heat stress, trehalose apparently must be hydrolysed rapidly by trehalase, liberating the cellular structures from bound trehalose (Singer and Lindquist, 1998). This would make them again accessible to chaperone proteins, enabling renaturation. A second function of trehalose could be to provide energy for the renaturation of cellular structures (Nwaka and Holzer, 1998). This indicates that both the synthesis and hydrolysis of trehalose have to be regulated in a concerted fashion in response to a stress condition and during subsequent recovery.

The regulation of Nth1 activity, in particular during stress and recovery from stress, is not clear. Nth1 activity, as measured in crude extracts, is high in log-phase cells, starts to decrease at the diauxic shift and remains low in stationary-phase cells (San Miguel and Arguelles, 1994). In contrast to Nth1 activity, NTH1 mRNA levels are higher in stationary-phase cells than in log-phase cells (DeRisi et al., 1997; Zähringer et al., 1998). Nth1 activity is regulated during the life cycle of yeast at the post-translational level via a cAMP-dependent protein kinase (PKA)-mediated phosphorylation process (Thevelein, 1984, 1988). Probably, the low Nth1 activity in stationary-phase cells is owing to low activity of PKA. Nth1 activity can be increased several-fold by adding cAMP or PKA to extracts of stationary-phase cells (Thevelein and Beullens, 1985; App and Holzer, 1989; Zähringer et al., 1998). Several putative PKA phosphorylation sites have been identified in the predicted protein sequence of Nth1, but whether one or more of these sites is functional in mediating activation by PKA remains to be elucidated (Kopp et al., 1993, 1994).

Different mechanisms, namely post-translational phosphorylation by PKA (De Virgilio et al., 1991), removal of a trehalase inhibitor (Mesquita et al., 1997) and induction of NTH1 expression coupled with de novo synthesis (Nwaka et al., 1995b; Zähringer et al., 1997, Zähringer et al., 1998), have been proposed to explain the induction of Nth1 by heat and chemical stress. Although these mechanisms remain largely uncharacterized, our previous work has shown that maintenance of Nth1 activity during heat stress depends on PKA activity (Zähringer et al., 1998). This suggests that phosphorylation of Nth1 by PKA might be essential for Nth1 stability at elevated temperatures. In this paper, we have investigated to what extent transcriptional induction of NTH1 plays a role in the control of trehalose levels and trehalase activity during stress and growth.

The promoter of the NTH1 gene contains several so-called STREs (stress responsive elements). These elements were shown to mediate stress induction of the genes CTT1, TPS2, HSP12 and DDR2, possibly through increased binding of the transcription factors Msn2 and Msn4 (Martinez-Pastor et al., 1996; Schmitt and McEntee, 1996). STRE elements (consensus sequence CCCCT or AGGGG) activate transcription in response to a variety of stress conditions, especially heat and osmotic stress, low pH and nutrient starvation (Siderius and Mager, 1997). In accordance with the negative regulation of STRE genes by PKA, the current model argues that PKA controls localization of Msn2/Msn4 to the cytoplasm and prevents their entry to the nucleus (Martinez-Pastor et al., 1996; Görner et al., 1998). The presence of three STREs in the NTH1 promoter suggests that stress-induced expression of NTH1 proceeds via the Msn2/Msn4/STRE pathway (Nwaka and Holzer, 1998). To gain more insight into the regulation of NTH1 under stress conditions, we have studied its induction in the msn2 msn4 double mutant. To elucidate whether the STREs present in the NTH1 promoter are functional, we performed site-directed mutagenesis of the STRE elements. We show that the induction of NTH1 by heat and osmotic stress, and to a lesser extent during the diauxic shift, is STRE and Msn2/Msn4 dependent.


Induction of Nth1 activity by heat and osmotic stress is abolished in a msn2 msn4 mutant

The transcriptional activator proteins Msn2 and Msn4 control the general stress response in yeast. To investigate whether Msn2 and Msn4 are involved in the induction of Nth1 activity during stress, we measured Nth1 activity after heat and osmotic stress in the wild type and in the msn2 msn4 mutant. As shown in Fig. 1A, Nth1 activity does not increase after heat stress in an msn2 msn4 mutant, in contrast to wild-type cells. In addition, basal Nth1 activity is reduced by a factor of about two when compared with wild-type cells.

Figure 1.

. Time-course of Nth1 activity in the wild-type strain (W303-1A) and the msn2 msn4 double mutant during heat and osmotic stress. A. Nth1 activity during heat stress at 40°C. B. Nth1 activity in the presence of 0.5 M NaCl. Stress treatments were performed on cells growing exponentially on glucose at 30°C for the indicated time. Zero time activity is the activity at 30°C before stress treatment. Closed symbols, wild-type strain; open symbols, msn2 msn4 mutant. Data represent means of three independent experiments.

It is known that a severe osmotic stress (1.5 M NaCl) does not enhance Nth1 activity or induce Nth1 synthesis (Zähringer et al., 1997). However, we now show that a moderate osmotic stress (0.5 M NaCl) does enhance Nth1 activity (Fig. 1B). In the msn2 msn4 mutant, the increase of Nth1 activity was significantly reduced compared with the wild type. These data clearly indicate that the increase in Nth1 activity after both heat and osmotic stress depends to a significant extent on Msn2 and Msn4.

Trehalose accumulation is reduced upon heat stress and hydrolysis is delayed during recovery in the msn2 msn4 mutant

Nth1 is responsible for trehalose hydrolysis in Saccharomyces cerevisiae. Deletion of NTH1 or mutations that lower Nth1 activity cause higher trehalose accumulation under most conditions, including heat shock and recovery (Nwaka and Holzer, 1998). We investigated whether the lack of Nth1 induction and the reduced basal activity in an msn4 msn4 deletion mutant affects trehalose levels during heat stress and recovery. As shown in Fig. 2, msn2 msn4 mutants have only about 50% as much trehalose as wild-type cells during exposure to 40°C for 60 min. This result may appear inconsistent with the fact that strains with low or no Nth1 activity show high trehalose accumulation during heat stress (Nwaka and Holzer, 1998; Zähringer et al., 1998). An explanation may be that induction of the catalytic subunits of the trehalose synthase complex TPS1 and TPS2 during heat stress is also dependent on Msn2 and Msn4 (Schmitt and McEntee, 1996; Parrou et al., 1997). During recovery from heat shock, the trehalose level in the wild type decreased within 10–20 min and reached the basal level again after 40 min. In contrast, the msn2 msn4 mutant showed a slower decrease in the trehalose level and reached the basal level after about 80 min of recovery. The trehalose profile in msn2 msn4 mutants during heat stress and recovery indicates the existence of a second Msn2- and Msn4-independent mechanism for trehalose accumulation. The profile further suggests that Msn2- and Msn4-dependent induction of Nth1 is important for rapid hydrolysis of the accumulated trehalose.

Figure 2.

. Trehalose levels in wild type (W303-1A) and msn2 msn4 mutant during heat stress and recovery. Exponentially growing cells of wild type (●) and msn2 msn4 mutant (□) were shifted from 30°C (zero time) to 40°C. After 60 min, the cells were shifted back to 23°C for 90 min. The trehalose concentration was determined at the indicated time points. Data represent means of three independent experiments.

Contrary to heat stress, osmotic stress-induced expression did not lead to accumulation of significant amounts of trehalose, in spite of the increases in Tps1 and Nth1 activities (Lewis et al., 1995; Garcia et al., 1997; Parrou et al., 1997). This suggests that different pathways may be responsible for heat and osmotic stress effects. In agreement with this, deletion of HOG1, which encodes a member of the mitogen-activated protein kinase (MAPK) osmosignalling pathway, markedly diminished osmotic stress induction of Nth1 activity but not heat stress induction (not shown).

Induction of NTH1 mRNA transcription by heat, osmotic and oxidative stress is strongly reduced in an msn2 msn4 mutant

The absence of an increase in Nth1 activity after heat and osmotic stress in the msn2 msn4 mutant prompted us to investigate NTH1 mRNA levels in this mutant after different stress treatments. As shown in 3Fig. 3A, heat (lane 2) and H2O2 (lane 3) induction of NTH1 is reduced in the msn2 msn4 mutant compared with the wild-type strain. This is also true for NaAsO2 (Fig. 3B, lanes 2 and 3) and NaCl (Fig. 3B, lane 4). As reported for other STRE-regulated genes (Martinez-Pastor et al., 1996; Schmitt and McEntee, 1996), basal NTH1 expression is also affected in the msn2 msn4 mutant. In relation to the basal NTH1 expression levels, stress induction of NTH1 in the msn4 msn4 mutant is reduced by at least 50% compared with the wild type, as determined using Phospho-Imager from Molecular Dynamics (data not shown).

Figure 3.

. Northern blot analysis of NTH1 expression in the wild-type strain (W303-1A) and the msn2 msn4 mutant after various stress treatments. Exponentially growing cells were subjected to heat and various chemical stresses. A. Lane 1, control 23°C; lane 2, 40°C for 40 min; lane 3, 0.4 mM H2O2 for 40 min; lane 4, 50 μg ml−1 CHX for 40 min. B. Lane 1, 23°C; lane 2, 1 mM NaAsO2 for 20 min; lane 3, 1 mM NaAsO2 for 40 min; lane 4, 0.5 M NaCl for 60 min. RNA from the respective strains was prepared and 30 μg from each strain was separated on 1% agarose gels, blotted and probed with NTH1. rRNA is used as loading control.

In contrast to heat and the other stress agents, induction of NTH1 expression by cycloheximide (CHX) was not abolished in the msn2 msn4 mutant (Fig. 3A, lane 4). This suggests that Msn2/Msn4 is not responsible for the increase in NTH1 mRNA as well as the increase in Nth1 enzymatic activity reported earlier in the presence of CHX (Zähringer et al., 1997). It cannot be excluded at this stage that these effects are a result of post-transcriptional action of this inhibitor on NTH1 expression or the stabilization effect of CHX on RNA, leading to their accumulation (Herrick et al., 1990). Interestingly, it has been reported that TPS2 mRNA expression is also induced by CHX (Gounalaki and Thireos, 1994).

We also investigated stress induction of NTH1 by making use of a reporter gene construct. The first 850 nucleotides of the NTH1 promoter (containing the three STREs) was fused to the lacZ gene, and the construct was chromosomally integrated into the wild-type strain and msn2 msn4 mutant. As shown in Fig. 4, heat stress induced lacZ expression in the wild-type strain by a factor of about two, but in the msn2 msn4 mutant no induction was observed. It should be mentioned here that the large difference between the fold increase in β-galactose reporter activity after heat stress (Fig. 4) and the NTH1 mRNA (Fig. 3) corroborate what has been previously reported for Nth1 activity and NTH1 mRNA in different wild-type strains after heat stress (Nwaka et al., 1995b; Zähringer et al., 1997, 1998). Possibly, the heat-induced expression of Nth1 is also subject to translational control. Alternatively, the accumulated mRNA may be unstable for efficient translation. Unlike TPS1lacZ induction (Parrou et al., 1997), we did not observe a difference in NTH1lacZ induction when a temperature of either 37°C or 40°C was used to induce expression of the reporter. Moderate osmotic stress also induced expression of the NTH1lacZ reporter in wild-type cells. In contrast, there was only a slight induction in the msn2 msn4 mutant after 60 and 90 min exposure to salt stress. Consistent with the Nth1 activity data, we observed that basal NTH1lacZ expression was also reduced in the msn2 msn4 mutant (compare Figs 4 and 1B). Surprisingly, H2O2 and NaAsO2 failed to induce NTH1–lacZ expression in the wild-type strain (not shown). Although we have no definite explanation for this, the chemicals may block NTH1lacZ reporter expression.

Figure 4.

. β-Gal activity in the wild-type strain (W303-1A) and the msn2 msn4 mutant during heat and osmotic stress. PNTH1lacZ construct (YIp357R–NTH1–LacZ) was chromosomally integrated into the wild-type strain and into the msn2 msn4 mutant as described in Experimental procedures. Closed bars represent β-galactose activity after heat shock at 40°C for 40 min. Striped bars represent activity after osmotic stress (0.5 M NaCl) for 60 and 90 min. Open bars represent activity of exponentially growing cells at 23°C (controls). Data represent means of three independent experiments.

Mutagenesis of all three STREs in the NTH1 promoter abolishes heat and osmotic stress-induced expression of NTH1

STRE elements present in promoters of stress-inducible genes are not always functional (Marquez et al., 1998). To investigate whether the STREs in the NTH1 promoter are functional and to investigate the possible contribution of each STRE in stress induction, we performed site-directed mutagenesis of the STRE elements. Wild-type and the various mutated NTH1 promoters were fused to lacZ carried on the episomal plasmid YEp357R. The plasmids were introduced into wild-type cells and β-galactose activity was monitored after heat and osmotic stress treatment.

As shown in Fig. 5, NTH1 expression is induced by heat and osmotic stress in wild-type cells by a factor of about two. Mutation of STRE1 (position −148 to −152) has only a minor effect, whereas mutation of STRE2 (position −334 to −338) caused a significant reduction in heat and osmotic stress induction of NTH1. Mutation of STRE3 (position −343 to −347) nearly abolished heat and osmotic stress induction. Mutation of STRE2 and STRE3 or of all three STRE elements together largely abolished stress induction of NTH1. Basal expression was reduced by about 30%. These results demonstrate that heat and osmotic stress induction of NTH1, as well as, to a lesser extent, basal NTH1 expression, are STRE dependent.

Figure 5.

. β-Gal activity in the wild-type strain and different STRE mutants of NTH1 during heat and osmotic stress. STREwt (YEp–NTH1wtlacZ) represents a strain carrying the wild type PNTH1lacZ construct in an episomal plasmid. STRE1 (YEp–NTH1STRE1lacZ) represents a strain carrying a mutation in the first STRE. STRE2 (YEp–NTH1STRE2lacZ) represents a strain carrying a mutation in the second STRE. STRE3 (YEp–NTH1STRE3–lacZ) represents a strain carrying a mutation in the third STRE. STRE23 (YEp–NTH1STRE23lacZ) represents a strain carrying a mutation in the last two STREs. STRE123 (YEp–NTH1STRE123lacZ) represents a strain carrying a mutation in all three STREs. Open bars represent controls at 23°C, closed bars represent the heat stress condition (37°C for 60 min), striped bars represent NaCl stress for 90 min. Data represent means of four independent experiments. Error bars indicate standard deviations.

Expression of NTH1lacZ reporter activity is compromised in msn2 msn4 and NTH1STRE123LacZ mutants

NTH1 expression is low in cells growing on glucose, starts to increase during the diauxic shift and remains high thereafter and in stationary phase (Zähringer et al., 1998; Parrou et al., 1999). In contrast, Nth1 activity has been reported to decrease during the diauxic shift and is very low in stationary-phase cells (San Miguel and Arguelles, 1994). To understand whether the Msn2/Msn4/STRE pathway is involved in NTH1 induction at the end of fermentative growth, we followed both NTH1 promoter-driven reporter activity during diauxic growth of wild-type cells and the msn2 msn4 mutant harbouring an NTH1–LacZ reporter integrated into the chromosome. β-Galactose activity increased about eightfold as the wild-type strain progressed into the diauxic shift (5–10 h time point, Fig. 6A). Enzyme activity increased slightly in stationary phase (24 h time point, Fig. 6A). In the msn2 msn4 mutant, the initial β-galactose activity was lower than in the wild type and showed no increase during exponential growth on glucose. After the diauxic shift, the activity was four- to fivefold lower than in the wild-type strain. However, the induction ratio was similar for both strains (about eightfold for wild type and about sixfold for the msn2 msn4 mutant).

Figure 6.

. β-Gal activity of PNTH1LacZ (chromosomally integrated) in wild type and in msn2 msn4 mutant and β-galactose activity of NTH1wtLacZ and NTH1STRE123LacZ (encoded on plasmid) during growth. A. Growth curve of strains and β-galactose activity of PNTH1lacZ in the wild-type strain and in the msn2 msn4 mutant growing on YP glucose. Wild type is represented by closed symbols and mutant by open symbols. Squares represent activity and circles represent growth curve. B. Growth curve of strains and β-galactose activity of NTH1wtlacZ (wild-type NTH1 promoter) and NTH1STRE123lacZ (NTH1 promoter with mutations in all three STREs) on SC-URA medium. Closed squares represent the wild-type strain and open squares represent the mutant. Squares represent β-galactose activity and circles represent the growth curve. Data were reproduced three times.

We next investigated whether the differences in reporter gene expression between the wild type and the msn2 msn4 mutant were related to the STRE elements. We therefore compared NTH1wtLacZ expression with NTH1STRE123LacZ expression during the growth cycle. In this case, the reporter constructs were plasmid encoded and may have caused β-galactose activity to be about fivefold higher than the chromosomally integrated constructs. As can be seen in 6Fig. 6B, NTH1wtLacZ expression during exponential growth on glucose (SC medium) is higher by a factor of about two than NTH1STRE123LacZ expression. Expression of both reporter genes is induced upon entry of the diauxic shift and continues to increase until stationary phase is reached. As in the wild-type strain and the msn2 msn4 mutant, NTH1wtLacZ expression was always higher than NTH1STRE123LacZ expression, but the induction ratio was almost the same. These results indicate the existence of Msn2/Msn4/STRE-independent mechanisms for induction of NTH1 expression upon glucose exhaustion. However, the Msn2/Msn4/STRE pathway is essential for a maximal level of NTH1 expression.


We have provided direct evidence for the involvement of STRE and Msn2/Msn4 in the control of basal expression, as well as in heat and osmotic stress and induction of NTH1 during the diauxic shift. Deletion of MSN2 and MSN4 or mutagenesis of the STREs in the NTH1 promoter reduces basal expression by about 30-50%. Heat- and NaCl-induced expression of NTH1 as well as lacZ reporter activity are strongly reduced. These results add a new dimension to the regulation of Nth1 activity and trehalose levels in yeast.

Inactivation of all three STREs reduced basal Nth1 expression by about 30%. This effect of STREs and MSN2/MSN4 on basal Nth1 expression may be explained by the finding that Msn2 and Msn4 are present in the nucleus in small amounts in non-stressed cells (Görner et al., 1998). Therefore, these factors may be bound to STREs under physiological conditions modulating basal transcription (Schmitt and McEntee, 1996). Our data reveal a hierarchy between the STREs in controlling the heat stress- and osmotic stress-dependent expression of Nth1 in the following order: STRE3 > STRE2 > STRE1. The similar effect of STRE2 and STRE3 on induction could be due to their close proximity, supporting the idea that the frequent occurrence of STREs in clusters might be related to their functionality, as recently proposed by Moskvina et al. (1998).

Our measurements (Fig. 2) of the trehalose level during heat stress and recovery show that biosynthetic as well as hydrolytic enzymes are important for maintaining the desired concentration of trehalose. Interestingly, we observed a decrease of about 50% in the trehalose level during heat stress in the msn2 msn4 double mutant compared with the wild-type strain. During recovery from heat shock, a transiently higher trehalose level, possibly owing to slower hydrolysis, is observed in the msn2 msn4 mutant. The decrease in trehalose concentration in the msn2 msn4 mutant during recovery is faster than that reported for the NTH1 deletion mutant (Van Dijck et al., 1995; Nwaka et al., 1995a; Nwaka and Holzer, 1998). These data show that Msn2- and Msn4-independent control mechanisms exist for trehalose synthesis and hydrolysis.

In contrast to heat treatment, moderate osmotic stress leads to almost no detectable trehalose increase in wild-type cells growing in glucose, in spite of the similar induction of trehalose synthesizing (Parrou et al., 1997) and hydrolysing enzymes (Figs 1 and 4). The reason for this difference is unclear. A rather simple explanation could be that the induced expression of enzymes of trehalose synthesized and hydrolysed under the different stress conditions have different catalytic effects. However, it also suggests that different pathways may modulate the heat and osmotic stress effects on trehalose metabolism. This idea is supported by the observation that the osmotic stress induction of Nth1 activity is dependent on the HOG (High Osmolarity Glycerol response) pathway, in contrast to the heat induction (not shown). In the fission yeast Schizosaccharomyces pombe, a similar (wis1–phh1) pathway controls osmostress induction of trehalase (Degols et al., 1996; Fernandez et al., 1997; Soto et al., 1998). Probably, regulation (by different pathways) of the induction of trehalose metabolism by heat and osmotic stress allows fine-tuning of the trehalose concentration according to the specific survival need of the cell. Although mild heat shock may require a higher trehalose concentration to protect rapidly critical cell targets against denaturation, osmotic stress may require hydrolysis of trehalose to supply glucose for the accumulation of the compatible solute glycerol (Blomberg and Adler, 1992).

Trehalose accumulation, as well as its hydrolysis, has important implications in the yeast life cycle and in stress tolerance. Heat stress induction of trehalase was initially thought to be owing mainly to PKA-mediated phosphorylation of the protein. Our results suggest that the increase in trehalase activity is largely dependent on the transcription factors Msn2/Msn4 and therefore transcriptional induction or MSN2/MSN4-dependent expression of an unknown gene. Northern blot analysis and β-galactosidase reporter activity conclusively show that the heat- and NaCl-induced increase in Nth1 activity is associated with an increase in the expression of the NTH1 gene and that this transcriptional event is dependent on Msn2, Msn4 and the STRE elements in the NTH1 promoter (Figs 3–5). However, participation of other as yet unidentified factor(s) is also indicated by both the residual induction of NTH1 mRNA and by the pattern of trehalose hydrolysis in the msn2 msn4 mutant (Figs 2 and 3).

Increased binding of Msn2 and Msn4 to STRE, as well as expression of most STRE-regulated genes (HSP12, CCT1), is negatively regulated by PKA (Kobayashi and McEntee, 1993; Marchler et al., 1993; Görner et al., 1998). TPK1, TPK2 and TPK3 encode the catalytic subunits of PKA (Toda et al., 1987). We have shown previously that deletion of TPK1 and TPK2, which reduces expression of TPK3 as well as phosphorylation activity of PKA (Mazon et al., 1993), has no discernable effect on heat stress induction of NTH1, although it results in reduction of Nth1 activity during heat stress (Zähringer et al., 1998). We proposed that PKA helps to maintain activity of Nth1 during heat stress. Using the PKA mutants tpk1w1BCY1 (S18-D) and TPK1 bcy1Δ (S13-3A), which exhibit low and high activity of PKA respectively (Nikawa et al., 1987), we studied a possible role of PKA on the transcriptional induction of NTH1. Our results show no clear correlation between NTH1 expression and PKA activity during heat stress when compared with HSP12 expression (data not shown). Furthermore, the pattern of osmostress induction of NTH1 mRNA in the TPK1 bcy1Δ (S13-3A) strain appears to support negative control of PKA over NTH1, however this was not corroborated by the tpk1w1BCY1 (S18-D) strain (data not shown; Zähringer, 1999). This variation of PKA effect on heat- and NaCl-induced NTH1 expression supports the idea that different pathways mediate these effects. For both conditions, the effect of PKA on NTH1 expression appears to be unique compared with most STRE-regulated genes for which transcription control has been documented. Possibly the PKA effect on NTH1 transcription is indirect or depends on factors that are not relevant for control of the other STRE-regulated genes. This may also apply to the expression of TPS1 and TPS2, which also show differences in sensitivity to PKA activity (Winderickx et al., 1996).

Earlier work on the regulation of Nth1 activity during growth on glucose showed a post-translational control of this enzyme by phosphorylation (Thevelein, 1984; App and Holzer, 1989; San Miguel and Arguelles, 1994). In cells growing exponentially on glucose, Nth1 activity is high. When the cells enter the respirofermentative stage (diauxic shift), Nth1 activity decreases and remains low through the growth cycle. However, in cell extracts, the enzyme can be activated strongly by endogenous PKA through cAMP addition or by addition of exogenous PKA (Thevelein and Beullens, 1985; App and Holzer, 1989; Uno et al., 1989; Zähringer et al., 1998). Because a correlation exists between Nth1 activity and the activity or level of many other components controlled by PKA in exponential compared with stationary-phase cells, it is believed that Nth1 activity during growth is controlled by PKA-dependent phosphorylation (Thevelein, 1996). In contrast to the drop in Nth1 activity, NTH1 expression is strongly induced at the diauxic shift (Zähringer et al., 1998; Parrou et al., 1999; this work, Fig. 6). Apparently, the cells synthesize a large amount of inactive Nth1 under conditions in which they accumulate a high level of trehalose. The reason for this reciprocal regulation is not evident. A possible explanation could be that cells prepare for conditions that require rapid hydrolysis of the trehalose, such as the sudden availability of glucose in the medium (Thevelein and Beullens, 1985). Although the Msn2/Msn4/STRE pathway is necessary for maximal NTH1 expression, induction still occurs in msn2 msn4 mutants or in NTH1STRE123LacZ mutants at the diauxic shift (Fig. 6). This clearly indicates that other factors are involved in NTH1 induction under these conditions. A similar Msn2- and Msn4-independent induction at the diauxic shift has also been reported for CTT1, a gene for which stress induction is regulated via the Msn2/Msn4/STRE pathway (Martinez-Pastor et al., 1996). Hence, this difference might indicate a general dichotomy between stress regulation and nutrient regulation for STRE-controlled genes.

Experimental procedures

Yeast strains, growth conditions and stress treatment

The source and genetic background of all S. cerevisiae strains used in this study are listed in Table 1. The cells were grown in a shaker at 23°C or at 30°C in YPD (1% yeast extract, 2% peptone and 2% glucose), plus 2% agar for solid media. Transformants were selected on synthetic complete medium (Sherman, 1991) lacking the relevant auxotrophic requirement. Stress treatments were performed with exponentially growing cells at an OD600 of about 1. The growing cells were divided into aliquots and exposed to heat (37°C or 40°C) or chemical stress agents (H2O2, NaAsO2 or cycloheximide) at concentrations known to induce expression of NTH1 (Zähringer et al., 1997, 1998). Non-lethal concentrations of NaCl used for osmotic stress experiments were in the range of 0.3–0.5 M (Martinez-Pastor et al., 1996; Winderickx et al., 1996). Control cells were maintained at 23°C or 30°C without stress treatment.

Table 1.  . S. cerevisiae strains used.Thumbnail image of

Plasmid construction and site-directed mutagenesis

Site-directed mutagenesis was carried out using a modified overlap extension PCR method (Mikaelian and Sergeant, 1992). The two external primers used to introduce HindIII and BamHI sites (underlined) at positions −850 and +23 of the NTH1 promoter were TGAAAAGCTTGGCTCTGACC and CCGGGGATCCTGGCTTGTATTAAC respectively. Three internal primers, GAAATTTTCCATCTATCGTTTTCCGTAG AG, TAGGACGAAGCCCCTTATCATCTAGTTACC, and TAG GACGAAGCATCTTATCCCCTAGTTACC were designed to introduce two nucleotide substitutions (CCCCT to CATCT) in each of the three STREs present in the NTH1 promoter (Nwaka and Holzer, 1998). The location of the respective STREs with reference to the NTH1 start codon is shown in Fig. 7. Wild type (amplified only with the external primers) and the mutagenized versions of the NTH1 promoter were directionally cloned into plasmids YIp357R or YEp357R (Myers et al., 1986), resulting in in frame NTH1::lacZ reporter plasmids. All basic cloning techniques were performed according to Sambrook et al. (1989). Constructs were checked by sequencing.

Figure 7.

. Promoter region of NTH1 showing the position of STREs.


Transformations were performed using the lithium acetate method of Gietz and Schiestl (1995). Integrative plasmids were linearized by cleavage with StuI before transformation in order to direct integration to the URA3 locus (Rothstein, 1991). The transformants were selected for URA3 prototrophy and lacZ expression.

β-Galactosidase assay

Assays for β-galactosidase activity using o-nitrophenyl-β-d-galactose (ONPG) as substrate were performed as described previously by Rose et al. (1990). Specific activity is expressed as ONPG hydrolysed (nmol min−1 mg−1 protein).

Neutral trehalase Nth1 activity and trehalose concentration

Crude enzyme extracts from stressed and control cells were prepared from equal wet weight amounts of cells using glass beads. Nth1 activity was measured using the glucose oxidase/peroxidase method (Werner et al., 1970), as described previously (Uno et al., 1983; App and Holzer, 1989). One unit of Nth1 is the amount that hydrolyses 1 μmol of trehalose in 1 min. Trehalose was extracted from equal wet weight amounts of cells by boiling and was assayed enzymatically using purified acid trehalase (Kienle et al., 1993). Protein concentrations were measured as described by Bradford (1976).

Extraction of total RNA and mRNA analysis

Total RNA was extracted from exponentially growing or stressed cells as has been described previously (Elder et al., 1983; Nwaka et al., 1995a; Zähringer et al., 1998). RNA was electrophoresed on a formaldehyde gel containing 1% agarose followed by transfer to nylon membrane by capillary force. Random-priming kit, [α-32P]-dCTP and Hybond N+ membrane were purchased from Amersham Buchler. Hybridization was performed in the presence of 50% formamide according to the manufacturer's instructions. The following gene probes were used: a 1.7 kb BamHI–HindIII fragment of NTH1 (Nwaka et al., 1995b), a 0.5 kb EcoRI HSP12 cDNA fragment (Praekelt and Meacock, 1990) and a 0.9kb HindIII–XhoI fragment of ACT1 from pYA301 (Gallwitz and Sures, 1980).


Authors thank Drs F. Estruch, H. Ruis and S. Hohmann for the gift of the strains. We also thank Drs B. Bukau, B. Mechler (Freiburg, Germany), H. Iwahashi (Tsukuba, Japan), M. Skrzypek and R. Dickson (Lexington, USA) for suggestions. This work was supported by a grant from the DFG (HO74/27-1) to Helmut Holzer (deceased). The authors dedicate this work to his memory. J.M.T. acknowledges research grants from the Fund for Scientific Research-Flanders, the Research Fund of the Katholieke Universiteit Leuven (Concerted Research Actions), Interuniversity Attraction Poles Network P4/30 and the Flemish Ministry of Economy through the IWT.