Introduction

The endoplasmic reticulum (ER) is responsible for folding and modification of nascent secretory and transmembrane proteins. When ER functions are perturbed by increased influx of newly synthesized polypeptides or exposure to stressors causing defects of glycosylation and disulfide bond formation, aberrant proteins accumulate in the ER lumen and membrane. This condition is referred to as ER stress. To restore ER homeostasis, cells under ER stress conditions activate transcription of a variety of genes, including genes encoding ER-resident chaperones and proteins functioning in the secretory pathway or ER-associated degradation1,2. In the budding yeast Saccharomyces cerevisiae, the unfolded protein response (UPR) signaling pathway composed of Ire1 and Hac1 plays a principal role in broad transcriptional response to ER stress1,2.

Protein phosphorylation and dephosphorylation catalyzed by protein kinases and phosphatases, respectively, are fundamental mechanisms by which cells respond to changes of the extracellular environment. Particularly, regulation of phosphorylation/dephosphorylation is essential for cells to initiate adaptive responses to environmental stresses. Previous studies using the budding yeast revealed that several kinases, including the stress responsive MAP kinases (MAPKs), such as Mpk1 and Hog1, and the Snf1 AMP-activated protein kinase (AMPK), are involved in ER stress response3,4,5,6,7,8,9,10. Furthermore, it has been shown that the calcineurin phosphatase participates in protection of yeast cells from ER stress11. It is well-known that many protein kinases can be activated through binding of their activator proteins and phosphorylation mediated by their upstream kinases. Indeed, such mechanisms operate in the budding yeast ER stress response3,4,5,6,7,8,9. In contrast, modulation of the expression levels is frequently observed in the situations where the phosphatase activity is regulated12,13. Furthermore, expression control of protein phosphatases is involved in feedback regulation of the signal and crosstalk between the signaling pathways13. In Saccharomyces cerevisiae, there are approximately thirty genes encoding known or putative protein phosphatases14,15. Previous genome-wide microarray analyses suggested that the mRNA levels of a subset of yeast protein phosphatases increase upon exposure to ER stress16,17. However, none of them were categorized as the target of the UPR pathway. Thus, little is known about regulation of the expression level of protein phosphatases during the budding yeast ER stress response and, furthermore, its physiological significance.

In this study, we analyzed mRNA expression patterns of 35 genes encoding yeast protein phosphatases in ER stress response. We found that the mRNA levels of 8 genes were more than two-fold increased by ER stress. Of these, PTP2 and CMP2 were involved in cellular sensitivity to ER stress. PTP2 and CMP2 encode a tyrosine phosphatase of MAPKs and a catalytic subunit of calcineurin phosphatase, respectively18,19,20,21. Our data presented here suggest that the PTP2 and CMP2 promoters are activated in response to ER stress and that promoter activation is required for the PTP2 and CMP2 genes to effectively fulfill their roles in ER stress response. Thus, regulation of the expression levels of protein phosphatases is critically involved in adaptive responses to ER stress.

Results

The mRNA levels of PTP2 and CMP2 are upregulated by ER stress

In Saccharomyces cerevisiae, 35 genes encode the known or putative catalytic components of protein phosphatases14,15 (Table 1). We first monitored their mRNA levels during ER stress response. Wild-type yeast cells were treated with tunicamycin, which causes ER stress by inhibiting N-linked glycosylation, and then harvested up to 7.5 hr every 1.5 hr; a quantitative real-time RT-PCR (qRT-PCR) analysis was carried out to quantitate the mRNA levels. We found that 10 mRNAs (CMP2, MIH1, PPZ2, PTC5, PTC6, PTP1, PTP2, SAL6, SDP1 and YCH1) increased more than two-fold within 7.5 hr after tunicamycin treatment (Fig. 1A and Supplementary Figs 15). We performed similar experiments using dithiothreitol (DTT), which causes ER stress by inhibition of the disulfide bond formation. Exposure to DTT increased the mRNA levels of 12 genes, including PPZ1 and PTP3 in addition to 10 genes whose expression was increased after exposure to tunicamycin (Fig. 1B and Supplementary Figs 610). We next attempted to examine the mRNA levels in cells which were ER-stressed by genetic alterations. Previous reports showed that activation of the UPR pathway, which is interpreted as an indication of ER stress conditions, was caused by the block of transport form the ER to the Golgi22,23,24. The SEC12 gene is essential for the initiation of coat protein complex II (COPII)-coated vesicle formation in ER-to-Golgi transport25, and its temperature-sensitive allele, sec12-4, causes the UPR activation23 (data not shown). Therefore, we analyzed expression of protein phosphatases in the sec12-4 mutant: Wild-type and sec12-4 mutant cells were incubated for 4 hr at 37 °C; their total RNAs were prepared and subjected to qRT-PCR analysis. The changes in gene expression patterns caused by sec12-4 mutation closely resembled those induced by treatment with chemical ER stressors, with some exceptions such that MIH1 expression was strongly induced by sec12-4 mutation (Fig. 1C and Supplementary Figs 1115). Similar results were obtained using the sec13-1 and sec16-2 mutants (Supplementary Figs 1115), which are also defective in vesicle formation from the ER23. Remarkably, among 10 genes whose mRNAs were increased by both tunicamycin and DTT, expression of 8 genes (CMP2, MIH1, PPZ2, PTC6, PTP1, PTP2, SAL6, and SDP1) was upregulated more than two-fold in the sec12-4

Table 1 Saccharomyces cerevisiae genes encoding the protein phosphatase.

mutant cells compared to wild-type cells. Thus, approximately 25% of mRNAs for protein phosphatases are upregulated under ER stress conditions.

Figure 1
figure 1

CMP2 and PTP2 are induced by ER stress and involved in cellular response to ER stress. (A,B) The changes of the mRNA levels of 35 genes encoding protein phosphatases after exposure to 2 μg/ml tunicamycin (TM) (A) and 4 mM dithiothreitol (DTT). (B) Maximum fold changes within 7.5 hr after ER stressor treatment relative to the untreated state are extracted from Supplementary Figs 110 and shown. (C) The changes of the mRNA levels of 35 genes encoding protein phosphatases in the sec12-4 mutant cells. Wild-type and sec12-4 mutant strains were grown at 25 °C until exponential phase and incubated for 4 hr at 37 °C. Fold changes in the sec12-4 mutant cells compared to wild-type cells are extracted from Supplementary Figs 1115 and shown. (D) Wild-type (WT) and mih1Δ, sdp1Δ, ych1Δ, ptc5Δ, ptc6Δ, cmp2Δ, ppz2Δ, sal6Δ, ptp1Δ, and ptp2Δ mutant strains were spotted onto YPD medium lacking or containing 1.0 or 1.5 μg/ml tunicamycin (TM) and incubated at 25 °C.

Next, we extracted protein phosphatases that are physiologically important in ER stress response from 10 protein phosphatases that were induced by both tunicamycin and DTT. We examined ER stress sensitivity of cells deleted for each gene. Cells were plated on medium containing tunicamycin as an inducer of ER stress, and their growth was monitored. Among them, the cmp2 mutant displayed hypersensitivity to tunicamycin, but the ptp2 mutant was resistant to tunicamycin (Fig. 1D). The CMP2 gene encodes a catalytic subunit of calcineurin phosphatase18,19. Previous studies using an inhibitor of the calcineurin phosphatase activity revealed that the budding yeast calcineurin acts to confer ER stress tolerance11. Furthermore, it has been reported that ptp2 deletion results in increased resistance to ER stress6. Thus, our observations are consistent with previous findings.

PTP2 expression is induced by the Mpk1-Rlm1 signaling axis during ER stress response

We attempted to identify the regulator of PTP2 expression. In the budding yeast ER stress response, the UPR pathway composed of Ire1 and Hac1 activates a broad transcriptional program1,2. Therefore, we examined whether the UPR pathway induces expression of the PTP2 gene. However, the mRNA levels of PTP2 were apparently unaffected by hac1 mutation (Supplementary Fig. 16A). Previous studies demonstrated that several signaling pathways, including Mpk1, Hog1 and Snf1, become activated in budding yeast ER stress response3,4,5,6,7,8,9,10. Therefore, we tested their involvement in regulation of PTP2 expression. We found that induction of PTP2 mRNA upon ER stress was impaired in the mpk1 mutant, but not in hog1 or snf1 mutant cells (Fig. 2A, and Supplementary Fig. 16B,C). Previous studies revealed that Mpk1 phosphorylates and activates Rlm1, a transcription factor belonging to the MADS (MCM1, Agamous, Deficiens, and SRF) family26. Furthermore, it has been reported that Rlm1 acts downstream of Mpk1 in ER stress response7. We therefore examined whether Rlm1 is involved in PTP2 upregulation during ER stress response. Similar to mpk1 mutation, rlm1 mutation significantly reduced the mRNA levels of PTP2 (Fig. 2A and Supplementary Fig. 16D). We next examined the protein levels of Ptp2 during ER stress response. We used the strain expressing the carboxyl-terminally Myc-tagged Ptp2 for western blot analysis. As shown in our previous report9, Ptp2 was increased following exposure to ER stressors (Fig. 2B,C). However, Ptp2 induction was diminished by mpk1 and rlm1 mutations (Fig. 2B,C, and Supplementary Fig. 16E). These results indicate that ER stress-activated Mpk1-Rlm1 pathway upregulates the expression level of the PTP2 gene.

Figure 2
figure 2

The Mpk1-Rlm1 signaling axis induces Ptp2 expression in ER stress response. (A) The mRNA levels of PTP2. Wild-type (WT) and indicated mutant strains were grown at 25 °C until exponential phase and treated with 4 mM dithiothreitol (DTT) for the indicated time, and total RNAs were prepared from each strain. The mRNA levels were quantified by qRT-PCR analysis, and relative mRNA levels were calculated using ACT1 mRNA. The data show mean ± SEM (n = 4). *P < 0.05 as determined by Tukey’s test. (B,C) The protein levels of Ptp2. Wild-type (WT) and rlm1 mutant strains harboring Myc-tagged PTP2 were grown at 25 °C until exponential phase and treated with 4 mM dithiothreitol (DTT) (B) or 2 μg/ml tunicamycin (TM) (C) for the indicated time. Extracts prepared from each cell were immunoblotted with anti-Myc and anti-Mcm2 antibodies. Original blots are presented in Supplementary Figs 17 and 18. The intensities of Ptp2-Myc were measured and normalized to the Mcm2 level. The values are plotted as the fold change from wild-type cells at the time of ER stressor addition. The data show mean ± SEM (n = 3). *P < 0.05 and **P < 0.01 as determined by Tukey’s test. NS, not statistically significant (P > 0.05). (D) Schematic representation of the structure of PPTP2-GFP. (E) Two putative Rlm1-binding motifs in the PTP2 promoter region. (F) Effects of ER stress on expression of the PPTP2-GFP reporter. Wild-type (WT) and rlm1 mutant strains harboring the integration which expresses GFP under the control of the PTP2 promoter were grown at 25 °C until exponential phase and treated with 4 mM dithiothreitol (DTT) for the indicated time. The GFP mRNA levels were quantified by qRT-PCR analysis, and relative mRNA levels were calculated using ACT1 mRNA. The data show mean ± SEM (n = 4). *P < 0.05 as determined by Tukey’s test. (G) Effects of deletion mutations in putative Rlm1-binding motifs on expression of the PPTP2-GFP reporter. Wild-type (WT) cells harboring the integration which expresses GFP under the control of wild-type or mutated PTP2 promoter were grown at 25 °C until exponential phase and treated with 4 mM dithiothreitol (DTT) for the indicated time. The GFP mRNA levels were quantified by qRT-PCR analysis, and relative mRNA levels were calculated using ACT1 mRNA. The data show mean ± SEM (n = 4). **P < 0.01 as determined by Tukey’s test.

Previous reports showed that PTP2 expression is induced by heat shock in a manner dependent on Mpk1 and Rlm127,28. Hence, it has been assumed that Rlm1 transcriptionally activates the PTP2 gene in response to heat shock; however, it remains unclear whether the PTP2 promoter is really activated. Therefore, we investigated whether ER stress increases PTP2 promoter activity. To test this, we constructed the PPTP2-GFP reporter (Fig. 2D), consisting of the 5′ upstream region of the PTP2 gene to drive GFP expression, and monitored the mRNA levels of GFP by qRT-PCR (Fig. 2F). GFP expression from the PPTP2-GFP reporter was increased after treatment with ER stressors (Fig. 2F and Supplementary Fig. 16F). However, GFP induction was impaired by rlm1 mutation (Fig. 2F). This suggests that the PTP2 promoter is activated by ER stress in a manner dependent on Rlm1. We further explored the mechanism by which Rlm1 controls PTP2 promoter activity. Previous analysis revealed that Rlm1 binds the consensus sequence, CTA(T/A)(T/A)(T/A)(T/A)TAG29. Our sequence analysis utilizing JASPAR, a database of transcription factor binding profiles (http://jaspar.genereg.net/), showed that two putative binding motifs for Rlm1 exist in the 5′ upstream region of the PTP2 gene (Fig. 2E). We designated them as the Rlm1-binding motifs (RBMs). To examine whether the RBMs are really important for PTP2 induction during ER stress response, we deleted each RBM or both in the PPTP2-GFP reporter. Deletion of RBM1, the sequence spanning from −282 to −273, clearly suppressed GFP induction (Fig. 2G). On the other hand, deletion of RBM2, the sequence spanning from −42 to −33, alone failed to change GFP expression pattern, but inhibited GFP upregulation in combination to RBM1 deletion (Fig. 2G and Supplementary Fig. 16F). This result supports the model whereby Rlm1 transcriptionally activates the PTP2 gene during ER stress response.

Ptp2 induction is required for effective downregulation of Hog1 activity in ER stress response

To investigate whether ER stress-induced upregulation of Ptp2 is critical for its role in ER stress response, we examined ER stress sensitivity of cells in which PTP2 induction is impaired. The recent study revealed that Mpk1 is involved in the ER stress surveillance (ERSU) pathway, which prevents transmission of stressed ER into daughter cells, and the defect in ERSU causes hypersensitivity to ER stress7. Rlm1 has been reported to act downstream of Mpk1 in ER stress response4,7. Indeed, hypersensitivity to ER stress was observed in the rlm1 mutant4 (Supplementary Fig. 16G). Thus, we considered that the mpk1 and rlm1 mutants are not suitable to be used as cells with diminished PTP2 induction. Therefore, we generated a series of PPTP2-PTP2 constructs, which express PTP2 under the control of wild-type or mutated PTP2 promoter, and compared their ability to complement the ER stress resistant phenotype caused by ptp2 mutation (Fig. 3A). The ptp2 mutant was resistant to ER stress. When harboring wild-type PPTP2-PTP2 integration, ER stress sensitivity of the ptp2 mutants was comparable to wild-type cells. Deletion of each RBM had little effect on the ability of PPTP2-PTP2 integration to complement the ptp2 mutant phenotype. However, the PPTP2-PTP2 integration in which both RBMs were deleted only partly complement ER stress resistance caused by ptp2 mutation. These results suggest that ER stress-induced Ptp2 upregulation is important for its role in ER stress response.

Figure 3
figure 3

Ptp2 induction is required for effective downregulation of Hog1 activity. (A) Effects of the PPTP2-PTP2 integrations on ER stress sensitivity. Wild-type (WT) and ptp2 mutant strains harboring the PPTP2-PTP2 integrations were spotted onto YPD medium lacking or containing 1.5 μg/ml tunicamycin (TM) and incubated at 25 °C. (B) Effects of ptp2Δ mutation on ER stress-induced Hog1 and Mpk1 activation. Wild-type (WT) and indicated mutant strains were grown at 25 °C until exponential phase and treated with 2 μg/ml tunicamycin (TM) for the indicated time. Extracts prepared from each cell were immunoblotted with anti-phospho-p38 MAPK (P-Hog1), anti-Hog1, anti-phospho-p44/42 MAPK (P-Mpk1), anti-Mpk1 antibodies. Original blots are presented in Supplementary Fig. 19. (C) Effects of rlm1Δ mutation on ER stress-induced Hog1 activation. Wild-type (WT) and the rlm1Δ mutant strains were grown at 25 °C until exponential phase and treated with 2 μg/ml tunicamycin (TM) for the indicated time. Extracts prepared from each cell were immunoblotted with anti-phospho-p38 MAPK (P-Hog1) and anti-Hog1 antibodies. Original blots are presented in Supplementary Fig. 20. The intensities of phosphorylated Hog1 were measured and normalized to total Hog1 level. The values are plotted as the fold change from wild-type cells at 6 hr after tunicamycin addition. The data show mean ± SEM (n = 3). *P < 0.05 as determined by Student’s t-test. (D) Effects of the PPTP2-PTP2 integrations on ER stress-induced Hog1 activation. Wild-type (WT) and ptp2 mutant strains harboring the PPTP2-PTP2 integrations were grown at 25 °C until exponential phase and treated with 2 μg/ml tunicamycin (TM) for the indicated time. Extracts prepared from each cell were immunoblotted with anti-phospho-p38 MAPK (P-Hog1) and anti-Hog1 antibodies. Original blots are presented in Supplementary Fig. 21. The intensities of phosphorylated Hog1 were measured and normalized to total Hog1 level. The values are plotted as the fold change from the ptp2 mutant cells harboring wild-type PPTP2-PTP2 integration at 6 hr after tunicamycin addition. The data show mean ± SEM (n = 3). **P < 0.01 as determined by Student’s t-test.

Ptp2 is known to dephosphorylate and inactivate Hog1 MAPK20,21. A previous report showed that loss of Ptp2 causes enhanced ER stress resistance in a HOG1-dependent manner6. This phenomenon was seen in our strain background (data not shown). On the other hand, it has also been proposed that Ptp2 functions in dephosphorylation and inactivation of Mpk127,28. Therefore, we examined the effects of ptp2 mutation on the activities of Hog1 and Mpk1 during ER stress response. It is well-known that antibodies against the phosphorylated form of mammalian p38 and p42/p44 MAPKs can be utilized to detect the phosphorylated form of the budding yeast Hog1 and Mpk1, respectively. We therefore performed western blot analysis using anti-phospho-p38 antibodies to monitor Hog1 activity. As observed previously5,9, tunicamycin treatment increased Hog1 activity in wild-type cells (Fig. 3B). In the ptp2 mutant cells, tunicamycin-induced Hog1 activation was significantly enhanced (Fig. 3B), indicating that Ptp2 dephosphorylates and inactivates Hog1 in ER stress response. We next determined Mpk1 activity by western blot analysis using anti-phospho-p42/p44 antibodies. Consistent with previous observations4,7, Mpk1 became activated by tunicamycin exposure (Fig. 3B). The activated Mpk1 level in the ptp2 mutant was comparable to that in wild-type cells (Fig. 3B). Western blot analysis also showed that the amount of Mpk1 was increased after ER stress treatment (Fig. 3B). This increase has been reported to result from Rlm1-mediated transcriptional activation of the MPK1 gene7. Similar to wild-type cells, Mpk1 upregulation was observed in the ptp2 mutants, suggesting that Ptp2 is not involved in regulation of the Mpk1-Rlm1 pathway in ER stress response. Thus, Ptp2 acts as a negative regulator of Hog1 in ER stress response.

We next examined how Rlm1-mediated transcriptional activation of the PTP2 gene influences Hog1 activity in ER stress response. Hog1 activation upon tunicamycin exposure was upregulated in the rlm1 mutant (Fig. 3C). This result suggests that Rlm1 is involved in downregulation of Hog1 activity during ER stress response. To test whether that ER stress-induced activation of the PTP2 promoter is important for Hog1 downregulation, we compared the Hog1 activity in the ptp2 mutant cells harboring PPTP2-PTP2 integrations (Fig. 3D). When carrying wild-type PPTP2-PTP2 integration, the ptp2 mutant displayed Hog1 activity at comparable levels to wild-type cells. However, Hog1 activation was upregulated by deleting both RBMs from the PPTP2-PTP2 integration. Taken together, these results indicate that Rlm1-mediated transcriptional activation of the PTP2 gene is critically involved in Hog1 downregulation during ER stress response.

The CMP2 promoter is important for ER stress response, but not for salt stress response

The CMP2 gene encodes a catalytic subunit of calcineurin phosphatase18,19. Calcineurin is a heterodimer composed of the catalytic and regulatory subunits30. The budding yeast contains another gene encoding a catalytic subunit of calcineurin, which is termed CNA118,19. The calcineurin regulatory subunit is encoded by the CNB1 gene31. As shown above, cells lacking Cmp2 exhibited hypersensitivity to ER stress (Fig. 1D). To investigate whether Cna1 is also involved in ER stress response, we examined tunicamycin sensitivity of the cna1 mutant cells, together with the cnb1 mutant cells (Fig. 4A). The cna1 mutation alone did not lead to ER stress hypersensitivity, but significantly enhanced ER stress sensitive phenotype caused by cmp2 deletion. ER stress hypersensitivity of the cmp2 cna1 double mutant cells was indistinguishable from that of the cnb1 mutant cells. These results indicate that Cmp2 and Cna1 act as the major and minor catalytic subunit of calcineurin in ER stress response, respectively.

Figure 4
figure 4

Cmp2 induction mediated by the promoter activation is important for ER stress response. (A) ER stress sensitivity of the mutants in calcineurin subunits. Wild-type (WT) and cmp2Δ, cna1Δ, cmp2Δ cna1Δ, and cna1Δ mutant strains were spotted onto YPD medium lacking or containing 1.0 μg/ml tunicamycin (TM) and incubated at 25 °C. (B) The protein levels of Cmp2 and Cna1. Wild-type (WT) cells harboring Myc-tagged CMP2 or CNA1 were grown at 25 °C until exponential phase and treated with 2 μg/ml tunicamycin (TM) for the indicated time. Extracts prepared from each cell were immunoblotted with anti-Myc and anti-Mcm2 antibodies. Original blots are presented in Supplementary Fig. 22. The intensities of Cmp2-Myc and Cna1-Myc were measured and normalized to the Mcm2 level. The values are plotted as the fold change from the level of Cmp2 at the time of tunicamycin addition. The data show mean ± SEM (n = 3). *P < 0.05 as determined by Tukey’s test. NS, not statistically significant (P > 0.05). (C) Schematic representation of the structure of PCMP2-GFP and PCNA1-GFP. (D) Effects of ER stress on expression of the PCMP2-GFP and PCNA1-GFP reporters. Wild-type (WT) cells harboring the integration which expresses GFP under the control of the CMP2 or CNA1 promoter were grown at 25 °C until exponential phase and treated with 2 μg/ml tunicamycin (TM) for the indicated time. The GFP mRNA levels were quantified by qRT-PCR analysis, and relative mRNA levels were calculated using ACT1 mRNA. The data show mean ± SEM (n = 4). *P < 0.05 as determined by Tukey’s test. NS, not statistically significant (P > 0.05). (E) Effects of PCMP2-CMP2, PCNA1-CMP2, PCNA1-CNA1, or PCMP2-CNA1 integrations on ER stress sensitivity. Wild-type (WT) and cmp2 mutant strains harboring the integration of PCMP2-CMP2, PCNA1-CMP2, PCNA1-CNA1, or PCMP2-CNA1 were spotted onto YPD medium lacking or containing 1.0 μg/ml tunicamycin (TM) and incubated at 25 °C. (F) Salt sensitivity of the mutants in calcineurin subunits. Wild-type (WT) and cmp2Δ, cna1Δ, cmp2Δ cna1Δ, and cna1Δ mutant strains were spotted onto YPD medium lacking or containing 0.8 M sodium chloride (NaCl) and incubated at 25 °C. (G,H) The mRNA levels of CMP2 (G) and CNA1. (H) Wild-type cells were grown at 25 °C until exponential phase and treated with 1 M sodium chloride (NaCl) for the indicated time, and total RNAs were prepared. The mRNA levels were quantified by qRT-PCR analysis, and relative mRNA levels were calculated using ACT1 mRNA. The data show mean ± SEM (n = 3). NS, not statistically significant (P > 0.05), by Tukey’s test. (I) Effects of PCMP2-CMP2 or PCNA1-CMP2 integrations on salt sensitivity. Wild-type (WT) and cmp2 mutant strains harboring the integration of PCMP2-CMP2 or PCNA1-CMP2 were spotted onto YPD medium lacking or containing 1.0 M sodium chloride (NaCl) and incubated at 25 °C.

We next asked why CMP2 is more important in ER stress response than CNA1. We compared the expression levels of Cmp2 and Cna1 using yeast strains carrying the carboxyl-terminally Myc-tagged genes. Consistent with an increase in CMP2 mRNA levels after ER stress treatment, Cmp2 expression was upregulated by ER stress (Fig. 4B). In contrast, Cna1 expression remained relatively unchanged following exposure to ER stress (Fig. 4B). Under ER stress conditions, Cmp2 was expressed at significantly higher levels than Cna1 (Fig. 4B). Next, we examined whether the difference in the protein levels between Cmp2 and Cna1 is due to the difference in the promoter activity between the CMP2 and CNA1 genes. To address this, we generated two reporter constructs, PCMP2-GFP and PCNA1-GFP, which express GFP under the control of the CMP2 and CNA1 promoters, respectively (Fig. 4C). GFP expression from the PCMP2-GFP reporter, but not the PCNA1-GFP reporter, was increased after ER stress treatment (Fig. 4D). The CMP2 promoter was more active than the CNA1 promoter under ER stress conditions (Fig. 4D). Thus, the difference in the promoter activity between CMP2 and CNA1 is reflected in their protein levels.

It may be possible that higher functional significance of Cmp2 attributes to higher activity of the CMP2 promoter. To explore this possibility, we generated two integrations: one expresses CMP2 from its own promoter; the other expresses CMP2 from the CNA1 promoter. We compared their ability to rescue ER stress hypersensitivity associated with cmp2 deletion (Fig. 4E). The PCMP2-CMP2 integration could fully rescue ER stress sensitive phenotype of the cmp2 mutant. However, the PCNA1-CMP2 integration failed to restore ER stress hypersensitivity caused by cmp2 mutation, suggesting that the CMP2 promoter is critical for the role of the CMP2 gene in ER stress response. We next examined whether CNA1 driven by the CMP2 promoter can compensate for loss of CMP2. However, the cmp2 mutant cells carrying the PCMP2-CNA1 integration exhibited hypersensitivity to ER stress. This result suggests that not only the promoter but also the sequence including the coding and 3′ untranslated regions are important for the role of the CMP2 gene in ER stress response.

We next compared the functional significance of Cmp2 and Cna1 in other stress response except for ER stress. Previously, it has been reported that the yeast calcineurin functions in adaptation to high salt conditions32,33. Consistent with the previous report32, hypersensitivity to sodium ions was observed in the cmp2 single mutant, but not in the cna1 single mutant (Fig. 4F). The cmp2 cna1 double mutant cells exhibited hypersensitivity to sodium ions indistinguishable from that of the cnb1 mutant cells (Fig. 4F). These results indicate that similar to ER stress response, Cmp2 is more important than Cna1 in salt stress response. However, unlike ER stress response, not only CNA1 but also CMP2 mRNA levels were largely unchanged in salt stress response (Fig. 4G,H). We next asked whether the CMP2 promoter is critical for the role of the CMP2 gene in salt stress response. The PCMP2-CMP2 integration completely rescued salt stress sensitive phenotype seen in cmp2 mutant cells (Fig. 4I). We found that even when CMP2 was expressed from the CNA1 promoter, salt stress hypersensitivity associated with cmp2 mutation could be effectively restored (Fig. 4I). Thus, the CMP2 promoter is less important for salt stress response compared with ER stress response.

Discussion

Reversible protein phosphorylation exerted by the antagonistic activity of protein kinases and phosphatases is one of the most well-analyzed post-translational modifications. Numerous studies have demonstrated that proper modulation of phosphorylation is required for eukaryotic cells to adapt to the environmental stress. In controlling the function of protein phosphatases, regulation of their expression levels is frequently utilized. Under ER stress conditions, yeast cells facilitate the expression of a variety of genes functioning to restore ER homeostasis. However, it remains unclear whether the expression levels of protein phosphatases are regulated in yeast ER stress response.

In this study, we examined the mRNA expression patterns of the budding yeast protein phosphatases during ER stress response. Of 35 mRNAs encoding known or putative protein phosphatases, 8 mRNAs (CMP2, MIH1, PPZ2, PTC6, PTP1, PTP2, SAL6, and SDP1) were upregulated by both chemical and genetic ER stressors; 2 mRNAs (PTC5 and YCH1) were increased by chemical ER stressors, but not by the genetic ER stressor. To investigate the physiological role of ER stress-induced protein phosphatases in ER stress response, we examined ER stress sensitivity associated with each deletion mutant. Contrary to our expectation, only two mutations, ptp2 and cmp2, altered cellular sensitivity to ER stress. In the previous study34, double mutants for 30 genes encoding non-essential yeast protein phosphatases were constructed in 435 combinations, and 4 double mutant cells displayed synthetic growth defects under specific conditions. In 10 protein phosphatases that were identified here as those induced by chemical ER stressors, for instance, two mitochondrial protein phosphatases, Ptc5 and Ptc6, are included35,36,37. Additionally, the budding yeast possesses Ptc7, the third protein phosphatase residing within mitochondria35,36. Therefore, failure of single mutation to alter ER stress sensitivity may be due to the functional redundancy between more than two protein phosphatases.

Here, we revealed that the Mpk1-Rlm1 axis promotes PTP2 expression in ER stress response. Previous reports showed that PTP2 expression is induced by environmental stresses, such as heat shock and osmotic stresses21,27,28,38. Upon heat shock, PTP2 is upregulated in a manner dependent on Mpk1 and Rlm127,28. Thus, it is likely that a similar mechanism operates in regulation of PTP2 expression in various stress responses. However, the roles of PTP2 in heat shock and ER stress responses seem to differ from each other: Ptp2 acts to downregulate Mpk1 in heat shock response, whereas Ptp2 functions in Hog1 inactivation during ER stress response6,27,28. To date, the physiological necessity of PTP2 upregulation in heat shock response remains to be elucidated. We showed here that the ER stress resistant phenotype of the ptp2 mutants was only partially rescued by PTP2 expression from mutated PTP2 promoter in which two putative Rlm1-binding motifs were deleted. This result indicates that PTP2 induction is required for its full function in ER stress response. Taken together, our results clearly demonstrated that the mechanism by which Hog1 is downregulated by the Mpk1-Rlm1 axis through Ptp2 induction exists in the budding yeast ER stress response (Supplementary Fig. 16H). Previous studies have revealed that constitutive activation of Hog1 leads to cell lethality39,40. However, under our laboratory conditions used in this study, Hog1 upregulation caused by ptp2 mutation advantageously functions in adaptation to ER stress, since the ptp2 mutant cells were resistant to ER stress. Furthermore, PTP2 overexpression failed to suppress ER stress hypersensitivity caused by rlm1 mutation, but rather enhanced it (Supplementary Fig. 16G). Nevertheless, why do yeast cells possess the mechanism by which Hog1 is negatively regulated through Ptp2 induction in response to ER stress? Previous studies suggested that ER stress sensitivity of the budding yeast is altered by the extracellular environments, including the composition of the culture medium8,41. Furthermore, it is well-known that the activity of Hog1 and Mpk1 are dynamically modulated by various environmental conditions42,43. Accordingly, the mechanism of Hog1 downregulation whereby the Mpk1-Rlm1 axis activates PTP2 expression may be required for yeast cells under certain conditions (for example, the combined stress conditions) to facilitate adaptation to ER stress.

Previously, it has been shown that the budding yeast calcineurin phosphatase confers resistance to ER stress3,11. However, it remained unclear how two alternative calcineurin catalytic subunits, Cna1 and Cmp2, operate in ER stress response, because the previous results were mainly based on the experiment using an inhibitor of calcineurin. Here, we found that Cmp2 makes a greater contribution to ER stress response than Cna1. Similar relative contribution was seen in salt stress response32 (Fig. 4F). Furthermore, it has been reported that Cmp2 is more important for recovery from the mating factor-induced growth arrest than Cna144. Thus, Cmp2 and Cna1 act as the major and minor catalytic subunits in distinct biological processes, respectively. However, the determinant of the relative functional significance appears to be different between, at least, salt and ER stress responses. Our reporter analysis revealed that the CNA1 promoter is less active than the CMP2 promoter. Nevertheless, Cmp2 expression from the CNA1 promoter could almost completely rescue the salt sensitive phenotype associated with cmp2 mutation. This result suggests that higher promoter activity is not required for the CMP2 gene to confer salt tolerance. In contrast, the promoter activity is likely to be a key determinant of the functional significance in ER stress response, since replacement of the CMP2 promoter with the CNA1 promoter considerably reduced the function of CMP2 in ER stress response. Additionally, the CMP2 gene may possess another element that is functionally important for ER stress response. This idea is derived from our observation that Cna1 expression from the CMP2 promoter hardly rescued cmp2 hypersensitivity to ER stress. Our results presented here suggest that the adaptive response to ER stress requires a higher amount of Cmp2 than salt stress. One plausible explanation for this phenomenon is as follows: the Cmp2 targets are different between ER and salt stress responses; more abundant Cmp2 is required for interacting physically with its target in ER stress response. In the budding yeast, a major target of the calcineurin is Crz1 transcription factor45,46. Crz1 acts downstream of the calcineurin in salt stress response, whereas Crz1 is unlikely to transduce a signal from the calcineurin in ER stress response41,45,46,47. The previous study suggests that under ER-stressed conditions, the calcineurin dephosphorylates Cch1, a subunit of the high affinity calcium channel3. However, it remains obscure whether Cch1 dephosphorylation is related to the physiological role of the calcineurin in ER stress response. Thus, identification of the calcineurin target in ER stress response should be needed to elucidate why higher promoter activity is required for the function of CMP2 in ER stress response. Additionally, since CMP2 induction occurred in the hac1, rlm1, hog1 and snf1 mutant cells (Mizuno et al., unpublished data), identification of the components involved in CMP2 expression should be needed for the understanding of the regulatory mechanisms of the calcineurin in ER stress response.

Materials and Methods

Plasmids

The PTP2, CMP2 and CNA1 genes were amplified from the Saccharomyces cerevisiae W303 derivative48 by PCR with the following primers: 5′-CTCTAGAGGATCCCCGGGGGACACTCGTTTAATTTAGC-3′ and 5′-TCGAGCTCGGTACCCGGGTATGGGTACTGACATCTCTG-3′ for PTP2; 5′-CTCTAGAGGATCCCCGGGCGTCCCAAAAAGGAAATAGC-3′ and 5′-TCGAGCTCGGTACCCGGGTCTCTGAGTCAGACAGTGTC-3′ for CMP2; 5′-CTCTAGAGGATCCCCGGGGATTTTGAAGATACTAGTGC-3′ and 5′- TCGAGCTCGGTACCCGGGTTGGTCGCACAAGGTGTCTC-3′ for CNA1. The amplified PTP2, CMP2 and CNA1 DNA fragments were inserted into the YCplac33 vector49 by In-Fusion cloning kits (Takara), yielding the YCplac33-PPTP2-PTP2, YCplac33-PCMP2-CMP2 and YCplac33-PCNA1-CNA1 plasmids, respectively. To make the PPTP2-GFP, PCMP2-GFP and PCNA1-GFP constructs, 1000-bp, 1000-bp and 997-bp genomic fragments containing the 5′ upstream sequences of the PTP2, CMP2 and CNA1 genes, respectively, were amplified by PCR with the following primers: 5′-CTCTAGAGGATCCCCGGGGGACACTCGTTTAATTTAGC-3′ and 5′-TAACCCGGGGATCCGATCCATCAATAGCAACGTCGATC-3′ for PPTP2-GFP; 5′-CTCTAGAGGATCCCCGGGCGTCCCAAAAAGGAAATAGC-3′ and 5′-TAACCCGGGGATCCGAGACATTGCGGGTTCAAGAAG-3′ for PCMP2-GFP; 5′-CTCTAGAGGATCCCCGGGGATTTTGAAGATACTAGTGC-3′ and 5′-TAACCCGGGGATCCGCGACATTGGCGTTGAGAGTG-3′ for PCNA1-GFP. The DNA fragment encoding GFP followed by the ADH1 terminator (TADH1) was amplified from the pFA6a-GFP-HIS3MX6 vector by PCR with the following primers: 5′-CGGATCCCCGGGTTAATTAAC-3′ and 5′-TCGAGCTCGGTACCCGGGAGATCTATATTACCCTGTTATCC-3′. The amplified 5′ upstream sequences of the PTP2, CMP2 and CNA1 genes, together with the GFP-TADH1 DNA fragment, were fused to the YCplac33 vector by In-Fusion cloning kits (Takara), yielding the YCplac33-PPTP2-GFP, YCplac33-PCMP2-GFP and YCplac33-PCNA1-GFP plasmids, respectively. Deletions of RBMs in the PTP2 promoter were generated by oligonucleotide-directed PCR using the YCplac33-PPTP2-GFP plasmid as a template. The primers used to delete RBMs are: 5′-CTACACATAAAGTTCCATAAAGCAG-3′ and 5′-GAACTTTATGTGTAGTACACCTAAC-3′ for RBM1 deletion; 5′-TGTAAACACTGGGGATCGGACCTAG-3′ and 5′-TCCCCAGTGTTTACAATAAAATAGG-3′ for RBM2 deletion. The PACT1-PTP2 construct was generated as follows. A 728-bp genomic fragment containing the ACT1 promoter was amplified from the Saccharomyces cerevisiae W303 derivative by PCR with the following primers: 5′- CTCTAGAGGATCCCCGGGAAGGGAACGTCAACCTGAAG-3′ and 5′-TGCTATGCGATCCATTGTTAATTCAGTAAATTTTCG-3′. The coding region of the PTP2 gene together with the 3′ downstream sequence was amplified by PCR with the following primers: 5′- ATGGATCGCATAGCACAGCAATATCG-3′ and 5′-TCGAGCTCGGTACCCGGGTATGGGTACTGACATCTCTG-3′. The amplified PACT1 and PTP2 DNA fragments were fused with the YCplac33 vector by In-Fusion cloning kits (Takara), yielding the YCplac33-PACT1-PTP2 plasmid. The PCNA1-CMP2 construct was generated as follows. The CNA1 promoter was amplified by PCR with the following primers: 5′-CTCTAGAGGATCCCCGGGGATTTTGAAGATACTAGTGC-3′ and 5′-AGCGTCTGAAGACATTGGCGTTGAGAGTGTTTTATGG-3′. The coding region of the CMP2 gene together with the 3′ downstream sequence was amplified by PCR with the following primers: 5′-ATGTCTTCAGACGCTATAAGAAATAC-3′ and 5′-TCGAGCTCGGTACCCGGGTCTCTGAGTCAGACAGTGTC-3′. The amplified PCNA1 and CMP2 DNA fragments were fused with the YCplac33 vector by In-Fusion cloning kits (Takara), yielding the YCplac33-PCNA1-CMP2 plasmid. Similarly, the YCplac33-PCMP2-CNA1 plasmid was constructed. The primers used to amplify PCMP2 and CNA1 are: 5′-CTCTAGAGGATCCCCGGGCGTCCCAAAAAGGAAATAGC-3′ and 5′-CAAGTCTTTCGACATTGCGGGTTCAAGAAGAAG-3′ for PCMP2; 5′-ATGTCGAAAGACTTGAATTCTTCACG-3′ and 5′-TCGAGCTCGGTACCCGGGTTGGTCGCACAAGGTGTCTC-3′ for CNA1. To generate the integrations, the inserts in the YCplac33 plasmids were subcloned into the pRS306 vector50. Plasmids used in this study are described in Supplementary Table 1.

Strains

Standard procedures were followed for yeast manipulations51. Yeast strains harboring the complete gene deletions (MIH1, SDP1, YCH1, PTC5, PTC6, CMP2, PPZ2, SAL6, PTP1, MPK1, RLM1, CNA1 and CNB1) and carboxyl-terminally Myc-tagged genes (CMP2 and CNA1) were generated by a PCR-based method as described previously52,53. Primer sets were designed such that 46 bases at the 5′ end of primers were complementary to those at the corresponding region of the target gene, and 20 bases at their 3′ end were complementary to the pFA6a sequence, 5′-TGCAGTACTCTGCGGGTGTATACAG-3′ or 5′- ATTTGACTGTATTACCAATGTCAGC-3′. All strains produced by a PCR-based method were verified by colony PCR amplification to confirm that replacement had occurred at the expected locus. Strains carrying the integrations, PPTP2-GFP, PCMP2-GFP, PCNA1-GFP, PPTP2-PTP2, PACT1-PTP2, PCMP2-CMP2, PCNA1-CNA1, PCNA1-CMP2, and PCMP2-CNA1, were constructed by integrating the linearized plasmids, pRS306-PPTP2-GFP, pRS306-PCMP2-GFP, pRS306-PCNA1-GFP, pRS306-PPTP2-PTP2, pRS306-PACT1-PTP2, pRS306-PCMP2-CMP2, pRS306-PCNA1-CNA1, pRS306-PCNA1-CMP2, and pRS306-PCMP2-CNA1, respectively. The sec12-4, sec13-1 and sec16-2 strains of the YPH499 derivative were kind gifts from Dr. Akihiko Nakano (Riken). Strains used in this study are listed in Supplementary Table 2.

RNA isolation and RT–PCR

Preparation of total RNA and generation of cDNA were performed as described previously9. The cDNAs were quantitated by a quantitative real-time RT-PCR (qRT-PCR) method using a 7500 fast real-time RT-PCR system (Applied Biosystems) with SYBR Premix Ex Taq (Takara), and levels of gene expression were normalized to ACT1 expression. Primers used to analyze the mRNA level are described in Supplementary Table 3.

Protein extraction, western blot analysis and antibodies

Preparation of protein extracts and Western blot analysis were performed as described previously9. Anti-phospho-p38 MAPK monoclonal antibody D3F9 (Cell Signaling), anti-Hog1 polyclonal antibody y-215 (Santa Cruz), anti-phospho-p44/42 MAPK polyclonal antibody (Cell Signaling), anti-Mpk1 polyclonal antibody yN-19 (Santa Cruz), anti-Myc monoclonal antibody 9E10 (Santa Cruz) and anti-Mcm2 polyclonal antibody N-19 (Santa Cruz) were used. Detection was carried out by using a LAS-4000 (Fuji Film) with Immobilon Western (Merck Millipore) or the Odyssey Imaging Systems (LI-COR Biosciences). Signal intensities were quantified by the Odyssey Imaging Systems, and statistical analysis was performed with Excel (Microsoft).

Stress sensitivity

Assays for tunicamycin and sodium chloride toxicity were carried out as follows. Cells were grown to exponential phase, and cultures were adjusted to an optical density of 0.5. Cell cultures were then serially diluted 5-fold, spotted onto normal plates or plates containing the indicated concentrations of tunicamycin, followed by incubation at 25 °C for 3 days (for plates lacking stressors) or more than 5 days (for plates containing stressors).