Hyperphosphorylation of Msn2p and Msn4p in response to heat shock and the diauxic shift is inhibited by cAMP in Saccharomyces cerevisiae

Hervé Garreau1, Rukhsana Nilofer Hasan1, Georges Renault1, Francisco Estruch2, Emmanuelle Boy-Marcotte1 and Michel Jacquet1

Laboratoire Information Génétique et Développement, Institut de Génétique et Microbiologie, UMR CNRS C8621, Université Paris-Sud, Bâtiment 400, 91405 Orsay cedex, France1
Dpto. Bioquímica y Biol. Molecular, Universitat de Valencia, and Dpto. Biotecnologia IATA, CSIC, Apdo Correos 73.46100 Burjassot, Valencia, Spain2

Author for correspondence: Hervé Garreau. Tel: +33 1 69 15 46 30. Fax: +33 1 69 15 72 96. e-mail: garreau{at}igmors.u-psud.fr


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In response to various stresses, as well as during the diauxic transition, the Msn2p and Msn4p transcription factors of Saccharomyces cerevisiae are activated and induce a large set of genes. This activation is inhibited by the Ras/cAMP/PKA (cAMP-dependent protein kinase) pathway. Here we show by immunoblotting experiments that Msn2p and Msn4p are phosphorylated in vivo during growth on glucose, and become hyperphosphorylated at the diauxic transition and upon heat shock. This hyperphosphorylation is correlated with activation of Msn2/4p-dependent transcription. An increased level of cAMP prevents and reverses these hyperphosphorylations, indicating that kinases other than PKA are involved. These results suggest that PKA and stress-activated kinases control Msn2/4p activity by antagonistic phosphorylation. It was also noted that Msn4p is transiently increased at the diauxic transition. Msn2p and Msn4p present different hyperphosphorylation patterns in response to different stresses.

Keywords: Saccharomyces cerevisiae, stress response, transcription factor, cAMP-dependent protein kinase

Abbreviations: PKA, cAMP-dependent protein kinase; STRE, stress-response element


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
All organisms have developed mechanisms of adaptation to environmental challenges. In the yeast Saccharomyces cerevisiae, the cellular responses involve a transcriptional activation of numerous genes, some of which promote a specific adaptation to the new environmental conditions, whereas others appear to be less specific, since they are induced under a wide variety of changes. The so-called stress-response system that involves the STRE (stress-response element) motif CCCCT and two transcriptional activators, Msn2p and Msn4p (Estruch & Carlson, 1993 ; Mager & De Kruijff, 1995 ; Marchler et al., 1993 ; Martinez-Pastor et al., 1996 ; Moskvina et al., 1999 ; Ruis & Schüller, 1995 ; Schmitt & McEntee, 1996 ), is an instance of such a broad-specificity response. As an example, heat shock induces the expression of both the Hsf1p regulon, which controls a specific subset of chaperones involved in the protection of proteins, and the Msn2/4p regulon (Boy-Marcotte et al., 1999 ). Similarly, oxidative stress triggers the activation of the Yap1p and Skn7p transcription factors, which induce genes playing a crucial role in maintaining the redox potential of the cell, as well as the Msn2/4p regulon (Godon et al., 1998 ; Lee et al., 1999 ). The Msn2/4p regulon is also activated in response to nutritional limitation. It is involved in the diauxic shift, the reprogramming of gene expression which occurs upon glucose exhaustion and ethanol-sustained growth (Boy-Marcotte et al., 1998 ). In these cases, the signal transduction pathway(s) from extracellular stress to gene induction is still largely unravelled. Whether or not Msn2p and/or Msn4p receive the same signals as the specific transcription factors (such as Hsf1p, Yap1p and Skn7p) remains unknown. It is also possible that the cAMP–PKA (cAMP-dependent protein kinase) pathway, which controls the nuclear localization of Msn2p and Msn4p, could mediate the general-stress response (Görner et al., 1998 ). Indeed most, if not all, of the genes found to be regulated by Msn2/4p during the diauxic shift, heat shock or oxidative stress are also repressed by cAMP (Boy-Marcotte et al., 1999 , 1998 ). To explain how Msn2/4p could be activated in response to different stresses and to evaluate the role of the cAMP pathway in this control, we have examined the state of phosphorylation of both Msn2p and Msn4p by their differential migration on gel electrophoresis at the diauxic transition and in response to heat shock. We have shown that Msn2/4p become hyperphosphorylated in vivo during the diauxic shift and upon heat shock. Msn2/4p, capable of relaying many different stresses, could be the targets of different signal-transducing kinases. This activating mechanism is antagonized by cAMP. Thus the cAMP–PKA pathway cannot be the transducer of the stress signal to these tranctivators but acts in addition to them, to counteract the adaptive response of the Msn2/4p regulon.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Plasmids and strains
Plasmids.
pGEX-3X-MSN2B contains the EcoRV–EcoRI fragment from the MSN2 gene (Estruch & Carlson, 1993 ) inserted into the SmaI andEcoRI sites in pGEX-3X (Pharmacia). It contains codons 44 to 302 of the MSN2 ORF.

pGEX-3X-MSN4B contains the NciI–NruI fragment from the MSN4 gene (Estruch & Carlson, 1993 ) inserted into the EcoRI site of pGEX-3X, after filling the 5' protruding NciI and EcoRI ends. It contains codons 34 to 256 of the MSN4 ORF.

pEL-32/45 was derived from pEL32 by replacing the PvuII–PvuII fragment (not including the MSN2 gene) with the EcoRI fragment (containing the MSN4 gene) from pEL45 (Estruch & Carlson, 1993 ). It contains both MSN2 and MSN4 genes under the control of their own promoters inserted into the YEp24 vector (Carlson & Botstein, 1982 ).

pAdh1-Msn2-GFP, encoding a Msn2–GFP fusion under the control of the strong constitutive ADH1 promoter, was a generous gift from C. Schüller (Görner et al., 1998 ).

Yeast strains.
The following strains were used: W303-1A (MATa ade2 his3 leu2 trp1 ura3) (Estruch & Carlson, 1993 ) and the diploid OL556 (MATa/MAT{alpha} cdc25-5/cdc25-5 his3/his3 leu2/leu2 trp1/TRP1 rca1/rca1 ura3/ura3) (Boy-Marcotte et al., 1996 ). The construction of the OL556-STRE strain containing the reporter gene lacZ under the control of four STRE motifs has been already described (Boy-Marcotte et al., 1998 ).

Antibodies.
Rabbit anti-Msn2p and anti-Msn4p antibodies were raised against the N-terminal parts of Msn2p and Msn4p (expressed in E. coli from pGEX-3X-MSN2B and pGEX-3X-MSN4B, respectively) and showed no cross-reactivity. They were purified by immunoblotting (Harlow & Lane, 1988 ) with the GST-fused peptides expressed in E. coli from pGEX-3X-Msn2B and pGEX-3X-MSN4B. For immunoblotting, these antibodies were used at 1/20 to 1/50 dilutions.

Yeast growth, protein extracts, immunoblotting and alkaline phosphatase treatment.
For the diauxic transition, yeast cells were grown in YNBS medium with 2% glucose (Boy-Marcotte et al., 1996 ). Glucose concentration in the culture medium was monitored enzymically with glucose oxidase using Sigma Diagnostic Glucose reagent kit no 510-A. When present in the medium, cAMP was at 3 mM. For heat shock, cultures grown at 26 °C in mid-exponential phase (OD710{approx}1) were transferred into a prewarmed water bath at 38 °C. Protein extracts were prepared essentially as described by Volland et al. (1992) , except extraction was performed directly on the cell culture without centrifugation. The precipitated proteins from 4x107 cells were dissolved in 500 µl modified Laemmli sample buffer. After gel electrophoresis (10–30 µl per well) in the Laemmli system (Laemmli, 1970 ), separated polypeptides were transferred onto PVDF sheets by the semi-dry transfer method (Kyhse-Andersen, 1984 ). Immunoreactive bands were identified by anti-Msn2p or anti-Msn4p antibodies, alkaline phosphatase-conjugated anti-rabbit IgG antibody (Promega) and 5-bromo-4-chloro-3-indoxyl-1-phosphate (BCIP)/nitro blue tetrazolium colour development. For dephosphorylation by alkaline phosphatase, Laemmli extracts (50 µl) were diluted 10-fold in alkaline phosphatase buffer (100 mM Tris/HCl, pH 8·5, 1 mM MgCl2, 0·1 mM ZnCl2), concentrated again to the initial volume by ultrafiltration, and incubated in the presence of 3 µl calf intestinal alkaline phosphatase (20 U µl-1; Roche Molecular Biochemicals) for 2 h at 37 °C. The control was incubated under the same conditions without alkaline phosphatase.

RNA preparation and analysis.
Total RNA extraction, gel electrophoresis of glyoxylated RNA and hybridization were performed as described earlier (Boy-Marcotte et al., 1996 ). Fifteen micrograms of total RNA per sample were fractionated on a 1% agarose gel. [{alpha}-32P]dCTP-labelled probes [3000 Ci (111 TBq) mmol-1] were synthesized by random priming with the ‘Ready To Go’ DNA labelling kit (Pharmacia Biotech) according to the manufacturer’s instructions. The HSP12 PCR-generated fragment (310 bp) used as template was amplified from genomic DNA with the following primers specific to the HSP12 ORF sequence: upper primer 5'-CAATGTCTGACGCAGGTA-3' and lower primer 5'-CTTCTTGGTTGGGTCTTC-3'. Hybridized membranes were exposed to phosphor screens, which were scanned in a Molecular Dynamics PhosphorImager.

GFP fluorescence microscopy.
The Msn2p–GFP fusion was detected without fixation essentially as described by Görner et al. (1998) .

ß-Galactosidase assay.
Determination of ß-galactosidase activity was performed as previously described (Boy-Marcotte et al., 1998 ).


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
cAMP prevents hyperphosphorylation of Msn2p and Msn4p at the diauxic transition
We have previously reported that Msn2p and/or Msn4p were essential for full induction of a set of genes that occurs at the diauxic transition, and that cAMP inhibited this transition (Boy-Marcotte et al., 1996 ). Therefore we looked at the phosphorylation state of Msn2p and Msn4p throughout the growth curve, in the presence and in the absence of cAMP (Fig. 1a). We used the OL556 strain, carrying the rca1 mutation in the PDE2 gene, which allows one to increase the intracellular level of cAMP by adding cAMP to the medium (Boy-Marcotte et al., 1996 ; Wilson et al., 1993 ). During the exponential-growth phase, two immunoreactive bands of Msn2p were detected (apparent molecular mass around 100 kDa): their electrophoretic mobilities with or without cAMP was the same (Fig. 1b, +/-cAMP, lanes 1). Msn2p was already phosphorylated, as shown by alkaline phosphatase treatment (Fig. 1c, sample 1). As the glucose concentration decreased, but before complete exhaustion, the fast-mobility band almost completely disappeared and new species with lower mobility were detected (Fig. 1b, -cAMP, lanes 2–4). This mobility shift was due to hyperphosphorylation (Fig. 1c, sample 4). With cAMP in the medium, no mobility shift was observed during the glucose-concentration decrease (Fig. 1b, +cAMP, lanes 1–4). A similar phenomenon was also observed for the closely related Msn4p. Indeed, during the exponential-growth phase, Msn4p was detected as two species having an apparent molecular mass around 95 kDa, the faster mobility form being predominant (Fig. 2b, -cAMP, lane 1). These forms were already phosphorylated, as shown by alkaline phosphatase treatment (Fig. 2c). After alkaline phosphatase treatment, two faster migrating bands were observed for Msn4p, probably as a result of either incomplete dephosphorylation or some proteolysis (Fig. 2c). A partial mobility shift occurred before complete exhaustion of glucose: several poorly resolved species of lower mobility were detected in addition to the faster-mobility forms present in exponential phase (Fig. 2b, -cAMP, lanes 2 and 3). The amount of Msn4p increased before complete exhaustion of glucose and Msn4p became barely detectable when the glucose was completely consumed (Fig. 2b, -cAMP, lanes 4 and 5). This was not due to a lower protein content of the sample, as checked by protein staining in the lower part of the transfer. Nevertheless, the faint band detected in lane 4 has a slower mobility than the major fast-mobility form detected in the exponential-growth phase, indicating hyperphosphorylation. In the presence of cAMP, during exponential growth, one major band was detected that had the same mobility as the low-mobility form detected in the exponential-growth phase in the absence of cAMP. No isoform with lower mobility was observed when glucose became limiting and no significant variation in the amount of Msn4p was observed (Fig. 2b, +cAMP, lanes 1 to 5).



View larger version (38K):
[in this window]
[in a new window]
 
Fig. 1. cAMP prevents hyperphosphorylation of Msn2p at the diauxic transition. (a) Two cultures of the OL556 strain were grown in YNBS at 26 °C, one in the absence (white symbols) and the other one in the presence of 3 mM cAMP (black symbols). At the times indicated by the arrows, samples were withdrawn from the cultures for glucose assay in the medium, immunotransfer and RNA analysis. {circ} and {bullet}, OD710; {triangleup} and {blacktriangleup}, glucose concentration in the culture medium. (b) Samples (40 ml) were withdrawn at the times indicated in (a) and proteins extracted. For each sample, 5 and 10 µl were loaded on polyacrylamide gel (7·5% acrylamide). After electrophoresis and transfer, the upper part of the PVDF sheets were processed for immunodetection of Msn2p using specific antibodies (upper panel), the lower part was used for Coomassie blue staining of bulk proteins (lower panel). (c) Alkaline phosphatase treatment of protein extracts of samples number 1 and number 4 in the absence of cAMP. Aliquots were incubated with or without alkaline phosphatase and Msn2p was detected by immunoblotting. s, starting material; m, mock incubation in alkaline phosphatase buffer; p, incubation in the presence of alkaline phosphatase. (d) Northern analysis of HSP12 mRNA in RNA extracts from samples taken from OL556 cultures with or without cAMP. ACT1 mRNA was analysed as the loading control.

 


View larger version (22K):
[in this window]
[in a new window]
 
Fig. 2. cAMP prevents hyperphosphorylation of Msn4p at the diauxic transition. (a) Strain OL556 was grown in YNBS at 26 °C with or without 3 mM cAMP. At the times indicated by the arrows, samples were withdrawn from the culture for glucose assay in the medium and immunotransfer, as described in Fig. 1. The symbols are the same as in Fig. 1(a). (b) Immunotransfer of Msn4p. The same protocol as for Msn2p was applied. (c) To assess the phosphorylation state of Msn4p during exponential growth, we used strain OL556 carrying the multicopy plasmid pEL-32/45. A sample was withdrawn from a culture, grown in YNBS at 26 °C, at OD710{approx}0·8 and the protein extract was treated with alkaline phosphatase. After electrophoresis and transfer, Msn4p was detected by immunostaining. s, starting material; m, mock incubation in alkaline phosphatase buffer; p, incubation in the presence of alkaline phosphatase.

 
The same type of experiment was performed on strain W303-1A. Here again, as described for the OL556 strain, both Msn2p and Msn4p were already phosphorylated during the exponential-growth phase and became hyperphosphorylated as glucose was consumed. A transient increase in the level of Msn4p was also observed prior to the end of exponential growth, reaching a peak before complete glucose exhaustion (data not shown).

As a reporter gene of the trancriptional activity of Msn2/4p (Martinez-Pastor et al., 1996 ), the accumulation of HSP12 mRNA was followed under the same conditions in strain OL556. The HSP12 mRNA accumulation, which occurred when glucose becomes limiting, was impaired in the presence of cAMP (Fig. 1d).

Msn2p and Msn4p are hyperphosphorylated upon heat shock; cAMP prevents and is able to reverse this effect
For the heat-shock experiments described in this section, we used strain OL556 transformed with the pEL32-45 plasmid, which allows overexpression of both Msn2p and Msn4p, to get strong enough immunostaining. Experiments without overexpression gave the same results although the immunostaining was weak (data not shown).

The phosphorylation state of Msn2p and Msn4p was examined when a heat shock from 26 °C to 38 °C was applied in the absence or in the presence of cAMP (Fig. 3a). At 26 °C, Msn2p was detected as two immunoreactive bands (Fig. 3b, left, lane 1). After heat shock in the absence of cAMP, the faster-mobility band disappeared (Fig. 3b, left, lanes 2 and 3), indicating hyperphosphorylation of Msn2p. In the presence of cAMP at 26 °C, the faster-mobility form of Msn2p was predominant (Fig. 3b, left, lane 4) and the Msn2p pattern was not modified after heat shock (Fig. 3b, left, lanes 5 and 6). Msn4p at 26 °C was detected as one predominant band (Fig. 3b, right, lanes 1 and 4). After the heat shock in the absence of cAMP, slower-migrating bands became apparent and the proportion of the fast-migrating band present at 26 °C decreased (Fig. 3b, right, lanes 2 and 3). It can be noticed that the heat-shock-induced modification in the phosphorylation patterns was different for Msn2p and Msn4p. In the presence of cAMP, no mobility shift was observed (Fig. 3b, right, lanes 5 and 6).



View larger version (25K):
[in this window]
[in a new window]
 
Fig. 3. Hyperphosphorylation of Msn2p and Msn4p upon heat shock and effect of cAMP. (a) Strain OL556 grown at 26 °C in YNBS medium (OD710 0·8) was split into two and 3 mM cAMP was added to one part. After 30 min, both cultures were shifted to 38 °C. Samples (40 ml) were taken at the times indicated (numbered 1 to 6). (b) Protein extracts were prepared from these samples and processed as described in Fig. 1. Top panels, immunostaining of Msn2p and Msn4p; bottom panels, Coomassie blue staining.

 
The hyperphosphorylation promoted by heat shock was reversed by the addition of cAMP. Indeed, when cAMP was added 15 min after heat shock (Fig. 4a), the faster-mobility form of Msn2p again became visible and predominant (Fig. 4b). A similar reversion of the heat-shock-induced mobility shift was also observed with Msn4p (data not shown).



View larger version (65K):
[in this window]
[in a new window]
 
Fig. 4. Hyperphosphorylation of Msn2p induced by heat shock is reversed by cAMP. (a) Strain OL556 grown in YNBS medium (OD710 0·8) was submitted to a heat shock from 26 °C to 38 °C in the absence of cAMP (-). After 15 min heat shock, the culture was split into two and 3 mM cAMP was added to one part (+). Samples were withdrawn from the two cultures at the times indicated by the arrows. (b) Protein extracts were prepared from these samples and processed as described in Fig. 1. Top panel, immunostaining of Msn2p; bottom panel, Coomassie blue staining.

 
The transcriptional activity of Msn2/4p during heat shock in the presence and in the absence of cAMP was assessed by monitoring the expression of ß-galactosidase as a reporter gene under control of Msn2/4 binding sites (STRE motifs). As shown in Fig. 5, specific activity was increased 18-fold after heat shock in the absence of cAMP. In contrast, in the presence of cAMP, the specific activity remained constant.



View larger version (12K):
[in this window]
[in a new window]
 
Fig. 5. Heat shock induction mediated by STRE is inhibited by cAMP. Strain OL556–STRE carrying the lacZ reporter gene under control of four STRE motifs was grown in minimal YNB medium supplemented with 0·5% Casamino acids at 26 °C. In the exponential-growth phase (OD710{approx}2), 3 mM cAMP was added to half of the culture. After 1 h, half of the two cultures +/- cAMP were transferred to 38 °C (white bars) and the remaining parts were maintained at 26 °C (black bars). Samples were withdrawn before heat shock, and 30 and 60 min after heat shock. ß-Galactosidase specific activity was assayed as previously described (Rose et al., 1990 )

 
Different hyperphosphorylation forms are produced upon different stresses
The pattern of heat shock-induced hyperphosphorylated forms of Msn2p (Figs 3 and 4) was different from the one obtained during the diauxic shift (Fig. 1), suggesting that different stresses could lead to different states of phosphorylation. To confirm this hypothesis and further extend this notion, we have compared the pattern of phosphorylation obtained in cells either submitted to heat shock or grown on 2% ethanol. As shown in Fig. 6(a), Msn2p presented different phosphorylation patterns under the two conditions: several low-mobility bands were detected in ethanol-grown cells compared to the unique low-mobility band observed after heat shock, all of them migrating still more slowly than heat-shock-induced species. After alkaline phosphatase treatment of the protein extracts from heat-shocked cells, as described in Methods, the electrophoretic mobility was increased (Fig. 6a, Pase). After control incubation in alkaline phosphatase buffer alone, no mobility shift was observed (data not shown). In both cases the stress led to nuclear localization of a GFP-fused Msn2p (Fig. 6b).



View larger version (62K):
[in this window]
[in a new window]
 
Fig. 6. The profiles of hyperphosphorylation due to heat shock and to ethanol are different. (a) Immunoblotting. Strain OL556 grown in YNBS to mid-exponential phase was heat-shocked for 30 min (38 °C) and proteins were extracted. After SDS-PAGE and transfer, Msn2p was detected using specific antibodies (lane Hs). Protein extracts from heat-shocked cells were treated by alkaline phosphatase and analysed by immunoblotting (lane Pase). Strain OL556 was grown on complete YP/ethanol medium at 26 °C and Msn2p was detected in crude extracts by immunoblotting (lane Et). Loading controls (not shown) did not indicate any significant variation in the amount of protein. (b) Nuclear translocation of Msn2p–GFP upon heat shock and in YP/ethanol-grown cells. OL556 transformed with pAdh1-Msn2p-GFP grown at 26 °C to exponential phase (i) were transferred to 38 °C and examined 15 min later on a fluorescence microscope (ii). The same cells grown to exponential phase in YPE medium are shown in (iii).

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Msn2/4p are hyperphosphorylated and activated by stress-signalling kinases
Msn2p and Msn4p induce transcription of a set of genes in response to various stresses via the STRE motif. The negative regulation of STRE-containing genes by the cAMP–PKA pathway has focused attention on a direct effect of this pathway on Msn2p and Msn4p. This hypothesis was further substantiated by the existence of several sites of PKA phosphorylation in each of these factors. Indeed, Msn2p has been reported to be phosphorylated in vitro by PKA (O’Brien et al., 1995 ; H. Garreau, J. d’Alayet and M. Jacquet, unpublished); mutation of these sites in Msn2p has been shown to abolish cAMP-promoted nuclear export and cytoplasmic localization in non-stressed cells (Görner et al., 1998 ). These results led to the suggestion that inactivation of PKA and/or dephosphorylation of the PKA sites would be sufficient to fully activate Msn2/4p-STRE transcription. To test this hypothesis, we analysed by immunoblotting the in vivo phosphorylation state and how it could be modified under stress or during the diauxic transition. Our data clearly show that under two different conditions that have been previously shown to activate the STRE-Msn2/4p system, heat shock and the diauxic transition (Boy-Marcotte et al., 1999 , 1998 ), hyperphosphorylation rather than dephosphorylation was observed for both of these transcription factors. This definitely eliminates the simplistic model in which the cAMP pathway would have been the unique integration system for stresses. It points to the existence of other protein kinases transducing stress signals by catalysing hyperphosphorylations.

In both the diauxic transition and heat shock, hyperphosphorylation of Msn2/4p was correlated with induced transcriptional activity of Msn2/4p. Moreover, the correlation between activation and hyperphosphorylation could be extended to the subcellular localization of these factors. We have confirmed the results previously obtained (Görner et al., 1998 ): the nuclear translocation of Msn2p induced by heat shock was inhibited and reversed by cAMP (data not shown). Therefore, it appears that hyperphosphorylation could be required for activation of Msn2p and Msn4p. The question therefore arises whether hyperphosphorylation is required for stress-induced nuclear localization or if it occurs only after migration to the nucleus.

cAMP–PKA pathway antagonizes both hyperphosphorylation and activation of Msn2/4p
We observed that high cAMP levels prevent stress-induced hyperphosphorylation of Msn2/4p and induction of HSP12 transcription. Thus, dephosphorylation of PKA sites could be a prerequisite for stress-promoted hyperphosphorylation and activation. In addition, the fact that added cAMP reversed hyperphosphorylation suggests that the PKA-phosphorylated sites have been dephosphorylated during stress-induced activation in order to restore responsiveness to cAMP. As illustrated in this report, Msn2p and Msn4p were already phosphorylated during exponential growth on glucose, when the intracellular cAMP level remains higher than in the absence of glucose (Russell et al., 1993 ). This basal phosphorylation could be due to PKA activity. The decrease in cAMP that occurs upon glucose exhaustion could lead to dephosphorylation of the PKA sites, allowing phosphorylation by stress-activated protein kinases. In this respect, it was recently reported how the TOR pathway leads to Msn2/4p sequestration in the cytoplasm by association with Bmh2p, a member of the 14-3-3 protein family. Inhibition of the TOR signalling pathway by rapamycin led to release of Msn2p from Bmh2p and to nuclear translocation of Msn2p (Beck & Hall, 1999 ). The fact that these 14-3-3 proteins are anchoring partners for serine-phosphorylated proteins (Muslin et al., 1996 ) strongly argues for such phosphorylation of Msn2/4p during exponential growth, to keep them present in the cytoplasm in the absence of stress.

These results suggest the existence of a switch between active and inactive forms controlled by two types of antagonistic phosphorylation. A similar phosphorylation switch has been observed at the diauxic transition on Not5p, a subunit of the Not complex, a negative regulator of TATA-less promoters. Not5p is phosphorylated upon glucose starvation and this phosphorylation is inhibited by high PKA activity (M. Collart, personal communication). Multiple phosphorylations with different effects on the proteins appear to be used by the cell for the fine tuning of regulation such as in the case of Pho4p, which is phosphorylated at several sites, each of them being involved at a different level of control (Komeili & O’Shea, 1999 ). Msn2p and Msn4p are other examples of such complex regulation of transcription factors by multiple phosphorylations.

Different stress signals are transduced by different pathways
In this report we show that different conditions (the diauxic transition, heat shock and growth on ethanol) give different patterns of hyperphosphorylation of Msn2p and Msn4p. That could be due to differences in the antagonistic activity of PKA, higher in glucose medium, even after heat shock, than in ethanol. Indeed, the intracellular cAMP level has been shown to be lower in ethanol-grown cells than in glucose-grown cells (François et al., 1987 ). Another explanation could be the involvement of different protein kinases, according to different stresses. Several stress-signalling pathways have been identified that involve specific kinases (Gustin et al., 1998 ). Among them, the HOG pathway has been shown to activate Msn2/4p/STRE-driven transcription in response to osmotic shock (Inoue et al., 1998 ; Martinez-Pastor et al., 1996 ; Rep et al., 1999 ). Another candidate could be the Snf1p protein kinase, which is activated during the diauxic transition (Wilson et al., 1996 ). Although protein kinase C mediates a response to heat-shock-induced cell-wall damage, this pathway does not seem to be involved in Msn2/4p activating phosphorylation, since STRE/Msn2/4p-driven transcription is independent of this pathway (Kamada et al., 1995 ).

Msn2p and Msn4p seem to be differentially regulated
We have observed some differences between Msn2p and Msn4p. Their cellular content appeared to be differentially regulated during diauxic transition: the amount of Msn4p transiently increased at the end of the exponential-growth phase, whereas Msn2p remained constant. This result is consistent with the data obtained by micro-array analysis (De Risi et al., 1997 ): MSN4 mRNA increases at the diauxic transition and MSN4 expression is repressed by Tup1p, whereas MSN2 mRNA decreases at the diauxic transition. The rapid decrease of Msn4p when growth is arrested due to glucose exhaustion suggests an instability of this protein, which was not observed for Msn2p. The ratios of these two factors could vary between growth conditions. Some subtle differences in the phosphorylation patterns between Msn2p and Msn4p after different stresses were also observed. This suggests differential susceptibility to stress-induced phosphorylation and to PKA. As with many other partially redundant genes, MSN2 and MSN4 could play partially overlapping functions with some biological specificity. The amount of the active forms induced by different stresses could be different for each of the two genes. That could lead to a differential induction of two partially overlapping subsets of genes among the Msn2/4p regulon, each depending on its own transcription factor. Indeed, as reported by Treger et al. (1998) , the relative contribution of each factor varies among different Msn2/4p-regulated genes, indicating some specificity for their functions.


   ACKNOWLEDGEMENTS
 
We are grateful to D. Zickler and F. Jammes for their help and advice in fluorescence microscopy and to W. Heidman for critical reading of the manuscript. We thank C. Schüller and H. Ruis for sending us plasmids encoding GFP–MSN2 fusions and M. Collart for communicating unpublished results. This work was supported by the Centre National de la Recherche Scientifique, the Ministère de l’Education Nationale, de la Recherche et de la Technologie and by a grant from l’Association pour la Recherche sur le Cancer. R.H. was supported by a fellowship from l’Association pour la Recherche sur le Cancer.


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Beck, T. & Hall, M. N. (1999). The TOR signalling pathway controls nuclear localization of nutrient-regulated transcription factors. Nature 402, 689-692.[Medline]

Boy-Marcotte, E., Ikonomi, P. & Jacquet, M. (1996). SDC25, a dispensable Ras guanine nucleotide exchange factor of Saccharomyces cerevisiae, differs from CDC25 by its regulation. Mol Biol Cell 7, 529-539.[Abstract]

Boy-Marcotte, E., Perrot, M., Bussereau, F., Boucherie, H. & Jacquet, M. (1998). Msn2p and Msn4p control a large number of genes induced at the diauxic transition which are repressed by cyclic AMP in Saccharomyces cerevisiae. J Bacteriol 180, 1044-1052.[Abstract/Free Full Text]

Boy-Marcotte, E., Lagniel, G., Perrot, M., Bussereau, F., Boudoscq, A., Jacquet, M. & Labarre, J. (1999). The heat shock response in yeast: differential regulations and contributions of the Msn2p/Msn4p and Hsf1p regulons. Mol Microbiol 33, 274-283.[Medline]

Carlson, M. & Botstein, D. (1982). Two differentially regulated mRNAs with different 5' ends encode secreted and intracellular forms of yeast invertase. Cell 28, 145-154.[Medline]

De Risi, J. L., Iyer, V. R. & Brown, P. O. (1997). Exploring the metabolic and genetic control of gene expression on a genomic scale. Science 278, 680-686.[Abstract/Free Full Text]

Estruch, F. & Carlson, M. (1993). Two homologous zinc finger genes identified by multicopy suppression in SNF1 protein kinase mutant of Saccharomyces cerevisiae. Mol Cell Biol 13, 3872-3881.[Abstract]

François, J. M., Eraso, P. & Gancedo, C. (1987). Changes in the concentration of cAMP, fructose 2,6-bisphosphate and related metabolites and enzymes in Saccharomyces cerevisiae during growth on glucose. Eur J Biochem 164, 369-373.[Abstract]

Godon, C., Lagniel, G., Lee, J., Buhler, J. M., Kieffer, S., Perrot, M., Boucherie, H., Toledano, M. B. & Labarre, J. (1998). The H2O2 stimulon in Saccharomyces cerevisiae. J Biol Chem 273, 22480-22489.[Abstract/Free Full Text]

Görner, W., Durchschlag, E., Martinez-Pastor, M. T., Estruch, F., Ammerer, G., Hamilton, B., Ruis, H. & Schüller, C. (1998). Nuclear localization of the C2H2 zinc finger protein Msn2p is regulated by stress and protein kinase A activity. Genes Dev 12, 586-597.[Abstract/Free Full Text]

Gustin, M. C., Albertyn, J., Alexander, M. & Davenport, K. (1998). MAP kinase pathways in the yeast Saccharomyces cerevisiae. Microbiol Mol Biol Rev 62, 1264-1300.[Abstract/Free Full Text]

Harlow, E. & Lane, B. (1988). Purifying antibodies from immunoblots. In Antibodies, a Laboratory Manual, p. 498. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Inoue, Y., Tsujimoto, Y. & Kimura, A. (1998). Expression of glyoxalase I gene of Saccharomyces cerevisiae is regulated by high osmolarity glycerol mitogen-activated protein kinase pathway in osmotic stress response. J Biol Chem 273, 2977-2983.[Abstract/Free Full Text]

Kamada, Y., Jung, U. S., Piotrowski, J. & Levin, D. E. (1995). The protein kinase C-activated MAP kinase pathway of Saccharomyces cerevisiae mediates a novel aspect of the heat shock response. Genes Dev 9, 1559-1571.[Abstract]

Komeili, A. & O’Shea, E. K. (1999). Role of phosphorylation sites in regulating activity of transcription factor Pho4. Science 284, 977-980.[Abstract/Free Full Text]

Kyhse-Andersen, J. (1984). Electroblotting of multiple gels: a simple apparatus without buffer tank for rapid transfer of proteins from polyacrylamide to nitrocellulose. J Biochem Biophys Methods 10, 203-209.[Medline]

Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 277, 680-682.

Lee, J., Godon, C., Lagniel, G., Spector, D., Garin, J., Labarre, J. & Toledano, M. (1999). Yap1 and Skn7 control two specialized oxidative stress response regulons in yeast. J Biol Chem 274, 16040-16046.[Abstract/Free Full Text]

Mager, W. H. & De Kruijff, A. J. J. (1995). Stress-induced transcriptional activation. Microbiol Rev 59, 506-531.[Abstract]

Marchler, G., Schüller, C., Adam, G. & Ruis, H. (1993). A Saccharomyces cerevisiae UAS element controlled by protein kinase A activates transcription in response to a variety of stress conditions. EMBO J 12, 1997-2003.[Abstract]

Martinez-Pastor, M. T., Marchler, G., Schüller, C., Marchler-Bauer, A., Ruis, H. & Estruch, F. (1996). The Saccharomyces cerevisiae zinc finger proteins Msn2p and Msn4p are required for transcriptional induction through the stress-response element (STRE). EMBO J 15, 2227-2235.[Abstract]

Moskvina, E., Imre, E.-M. & Ruis, H. (1999). Stress factors acting at the level of the plasma membrane induce transcription via the stress response element (STRE) of the yeast Saccharomyces cerevisiae. Mol Microbiol 32, 1263-1272.[Medline]

Muslin, A. J., Tanner, J. W., Allen, P. M. & Shaw, A. S. (1996). Interaction of 14-3-3 with signalling proteins is mediated by the recognition of phosphoserine. Cell 84, 899-897.[Medline]

O’Brien, R. M., Halmi, N., Stromstedt, P., Printz, R. L. & Granner, D. K. (1995). Expression cloning of a zinc-finger cyclic AMP-response-element-binding protein. Biochem J 312, 17-21.[Medline]

Rep, M., Reiser, V., Gartner, U., Thevelein, J., Hohman, S. & Ammerer, G. (1999). Osmotic stress-induced gene expression in Saccharomyces cerevisiae requires Msn1p and a novel factor Hot1p. Mol Cell Biol 19, 5474-5485.[Abstract/Free Full Text]

Rose, M. D., Winston, F. & Hieter, P. (1990). Methods in Yeast Genetics: a Laboratory Course Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Ruis, H. & Schüller, C. (1995). Stress signalling in yeast. Bioessays 17, 959-965.[Medline]

Russell, M., Bradshaw-Rouse, J., Markwardt, D. & Heidman, W. (1993). Changes in gene expression in the Ras/adenylate cyclase system of S. cerevisiae: correlation with cAMP levels and growth arrest. Mol Biol Cell 4, 757-765.[Abstract]

Schmitt, A. P. & McEntee, K. (1996). Msn2p, a zinc finger DNA-binding protein, is the transcriptional activator of the multistress response in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 93, 5777-5782.[Abstract/Free Full Text]

Treger, J. M., Schmitt, A. P., Simon, J. R. & McEntee, K. (1998). Transcriptional factor mutations reveal regulatory complexities of heat shock and newly identified stress genes in Saccharomyces cerevisiae. J Biol Chem 273, 26875-26879.[Abstract/Free Full Text]

Volland, C., Garnier, C. & Haguenauer-Tsapis, R. (1992). In vivo phosphorylation of the yeast uracil permease. J Biol Chem 267, 23767-23771.[Abstract/Free Full Text]

Wilson, R. R., Renault, G., Jacquet, M. & Tatchell, K. (1993). The pde2 gene of Saccharomyces cerevisiae is allelic to rca1 and encodes a phosphodiesterase which protects the cell from extracellular cAMP. FEBS Lett 325, 191-195.[Medline]

Wilson, W. A., Hawley, S. A. & Hardie, D. G. (1996). Glucose repression/derepression in budding yeast: SNF1 protein kinase is activated by phosphorylation under derepressing conditions, and this correlates with a high AMP:ATP ratio. Curr Biol 6, 1426-1434.[Medline]

Received 22 March 2000; revised 12 June 2000; accepted 15 June 2000.