The Ca2+‐activated pathways in Saccharomyces cerevisiae induce a delay in the onset of mitosis through the activation of Swe1p, a negative regulatory kinase that inhibits the Cdc28p/Clb complex. We isolated the YAP1 gene as a multicopy suppressor of calcium sensitivity owing to the loss of ZDS1, a negative regulator of SWE1 and CLN2 gene expression. YAP1 deletion on a zds1Δ background exacerbated the Ca2+‐related phenotype. Yap1p was degraded in a calcineurin‐dependent manner when cells were exposed to calcium. In yap1Δ cells, the expression level of the RPN4 gene encoding a transcription factor for the subunits of the ubiquitin–proteasome system was diminished. The deletion of YAP1 gene or RPN4 gene led to the accumulation of Swe1p and Cln2p. Yap1p was a substrate of calcineurin in vivo and in vitro. The calcineurin‐mediated Yap1p degradation seems to be a long adaptive response that assures a G2 delay in response to a stress that causes the activation of the calcium signalling pathways.
Calcium‐triggered signalling mechanisms are used by virtually all eukaryotic organisms to regulate a wide variety of cellular processes, including gene expression. Transient increases in cytosolic Ca2+ result in the activation of diverse enzymes, including calcineurin, a highly conserved Ca2+/calmodulin‐dependent serine/threonine protein phosphatase.
In the yeast Saccharomyces cerevisiae, calcineurin has been implicated in stress‐induced gene expression, ion homeostasis and viability after exposure to mating pheromone (for reviews, see Aramburu et al, 2000; Cyert, 2001). More recently, we demonstrated that calcineurin and the Mpk1p pathway, in a coordinated fashion, activate Swe1p, a negative regulatory kinase of the Cdc28p/Clb complex (Booher et al, 1993), leading the cell to a cell‐cycle delay in the G2 phase (Mizunuma et al, 2001). The activation of these pathways is thought to occur in response to membrane stretching (Mizunuma et al, 1998).
The effect of calcium on cell‐cycle regulation is conspicuous on a zds1Δ background lacking the negative regulator for SWE1 and CLN2 transcription (Ma et al, 1996; Mizunuma et al, 2004). Previously, we showed that the cell cycle of zds1Δ strain cultivated in medium containing a high concentration of CaCl2 was delayed in the G2 phase and that this strain showed polarized bud growth owing to the activation of the cellular calcium signalling pathways (Mizunuma et al, 1998). To identify the proteins that bear Zds1p‐like functions in the calcium‐induced G2 cell‐cycle regulation, we screened for genes the overexpression of which could suppress the calcium phenotypes of the zds1Δ strain. By this screening, the YAP1 gene was obtained as a suppressor. Here, we show that the Swe1p and Cln2p levels were negatively regulated indirectly by Yap1p. On exposure of cells to exogenous calcium, Yap1p was degraded in a calcineurin‐dependent manner, which led to transcriptional downregulation of the RPN4 gene encoding a transcription factor that modulates the expression levels of the genes involved in the ubiquitin–proteasome system. The Yap1p degradation, which eventually leads to Swe1p and Cln2p upregulation, was suggested as a mechanism to assure a G2 delay in response to a stress that causes the activation of calcineurin.
YAP1 gene is a high‐copy suppressor of zds1Δ strain
To investigate the detailed mechanism of the calcium‐induced cell‐cycle regulation, we carried out a screening for genes the overexpression of which could suppress the calcium sensitivity of the zds1Δ strain. The genes contained in the suppressor plasmids were classified into 19 groups, including ZDS1 and its homologue ZDS2 (Bi & Pringle, 1996; Yu et al, 1996). Excluding these two genes, a representative plasmid was chosen from each of the other groups, and they were designated as pSUZ1∼17 (suppressor of zds1). By sequencing of the inserts, plasmids containing the genes, such as PMR1, PMC1 and MIH1, were identified, showing the validity of this screening. By subcloning of the pSUZ14 plasmid, the YAP1 gene, which encodes a mammalian AP‐1‐like protein (Moye‐Rowley et al, 1989), was identified as the gene responsible for suppression. As shown in Fig 1, the growth of the zds1Δ strain in the CaCl2‐containing medium was severely inhibited, exhibiting a G2 delay and highly polarized bud growth. Overexpression of the YAP1 gene partially suppressed all of these calcium‐induced phenotypes. So we further investigated the mechanism by which Yap1p modulated the calcium‐dependent cell‐cycle regulation.
Deletion of YAP1 gene exacerbates Ca2+ sensitivity
To verify the contribution of Yap1p in suppressing the various calcium‐induced phenotypes of the zds1Δ strain, we examined the effect of a YAP1 deletion mutation on these phenotypes (Fig 2A–C). The yap1Δ mutation by itself did not lead to significant increases in the calcium sensitivity. However, the growth defect of the zds1Δ strain on calcium plates was exacerbated by further yap1Δ mutations (Fig 2A). Morphologically, the yap1Δ zds1Δ double deletion cells, in comparison with the zds1Δ cells, showed a slightly more elongated shape in YPD medium (without added CaCl2), and showed a more pronounced elongated morphology than the zds1Δ cells in the presence of CaCl2 (Fig 2B). Fluorescence‐activated cell sorting analysis of the cellular DNA content demonstrated that the yap1Δ zds1Δ strain, in comparison with the respective single deletion strains, suffered from a severe defect in G2 progression in the presence of CaCl2 (Fig 2C). These results indicated that Yap1p and Zds1p may share a common biological function in parallel pathways in calcium signalling.
Yap1p degradation is calcineurin dependent
The results of the genetic experiments indicate that Yap1p is involved in the regulation of the calcium‐induced G2 delay and polarized bud growth. We examined whether the levels of YAP1 messenger RNA and Yap1p were altered by calcium. For this purpose, we used the strain carrying a chromosomally integrated construct for Myc‐epitope‐tagged Yap1p at the carboxyl terminus under the control of its own promoter. The Myc‐tagged Yap1p suppressed the elongated bud growth of the zds1Δ yap1Δ strain in the presence of 100 mM CaCl2, suggesting that the construct is functional in vivo (data not shown). We previously noted that the calcium‐induced, proteasome‐dependent degradation of Hsl1p proceeded more rapidly at 37°C than at 28°C (Mizunuma et al, 2001). Thus, we determined the changes in the levels of YAP1 mRNA and Yap1p at 37°C following the shift of wild‐type cells to the medium containing CaCl2. Although the YAP1 mRNA level was not altered, the Yap1p–Myc level of the wild‐type strain was remarkably diminished by the presence of calcium (Fig 3A). At 28°C, the calcium‐induced decrease in the Yap1p–Myc level was observed at a slower rate (Fig 3B). To distinguish whether Yap1p was downregulated at the translational or post‐translational level, we examined the effect of calcium on the stability of Yap1p–Myc in the presence of cycloheximide. The rate of the decrease in the intensity of Yap1p–Myc in the presence of the protein synthesis inhibitor was still accelerated by calcium (Fig 3C,D). Under similar conditions, the Yap1p–Myc level in cells lacking calcineurin (cnb1Δ) was not altered by calcium. These results indicate that the degradation of Yap1p–Myc was induced by calcium in a calcineurin‐dependent manner (Fig 3C,D).
Phospho‐Yap1p is dephosphorylated by calcineurin
The calcineurin dependence of Yap1p downregulation raises the possibility that Yap1p might be a substrate for calcineurin. We first examined, by immunoprecipitation experiments, whether calcineurin and Yap1p physically interacted in vivo. As shown in Fig 4A, Myc‐tagged Cnb1p and haemagglutinin (HA)‐tagged Yap1p were co‐precipitated in a reciprocal manner. Co‐precipitation was observed by incubation in the presence of EDTA, suggesting that the association of Yap1p and calcineurin is not induced by calcium (data not shown).
As Yap1p phosphorylation has not been well characterized previously, we examined whether Yap1p can be phosphorylated in vivo. When the Yap1p–Myc immunoprecipitate was treated with calf intestinal alkaline phosphatase (CIP), a band that migrated faster than that similarly treated in the presence of a phosphatase inhibitor was observed, showing that Yap1p had been phosphorylated (Fig 4B). We further examined the possibility that Yap1p is a substrate of calcineurin in vivo. When cell extracts prepared from wild‐type cells grown in calcium medium were analysed by electrophoresis, the intensity of Yap1p–Myc bands was diminished with progression of incubation time, reflecting its calcineurin‐dependent degradation. The change in Yap1p intensity was accompanied by the preferential loss of the slower migrating component from the broad band seen at time 0. By contrast, no significant changes in the intensity and mobility of Yap1p–Myc were observed in the cnb1Δ cells. These results indicate that the faster migrating component was generated in a calcineurin‐dependent manner. The CIP‐treated Yap1p–Myc migrated faster than Yap1p–Myc generated in wild‐type cells grown in calcium medium, suggesting that Yap1p contained other phosphorylated residues that were not removable by calcineurin. Moreover, we found that Yap1p was a direct substrate of calcineurin in vitro (Fig 4C). These results show that Yap1p was a direct substrate of calcineurin.
Yap1p degradation leads to a decrease of RPN4 mRNA
Because the effects of the overexpression and deletion of the YAP1 gene in calcium signalling resembled those of the ZDS1 gene, we determined whether the YAP1 gene was involved in the regulation of SWE1 and CLN2 transcription by northern blot analysis. However, the effect of YAP1 deletion on the expression levels of these genes was negligible (Fig 5A).
To examine whether Yap1p functioned at the post‐transcriptional level, we compared the levels of Swe1p and Cln2p in wild type and in yap1Δ strains carrying chromosomally integrated constructs for Myc‐tagged Swe1p and HA‐tagged Cln2p (Fig 5B). The levels of these proteins in the yap1Δ strain were significantly elevated compared with those in the wild‐type strain. Contrary to the effect of YAP1 deletion, overexpression of the YAP1 gene led to decreased levels of these proteins (Fig 5C).
The Yap1p degradation led to the upregulation of Swe1p and Cln2p. What might be the mechanism underlying this upregulation? Recently, Yap1p, together with the Pdr1p and Pdr3p transcription factors, was implicated in the transcriptional activation of the RPN4 gene encoding a transcription factor that modulates the expression levels of the genes for the ubiquitin–proteasome system during the oxidative stress response (Owsianik et al, 2002). We reasoned that the Yap1p degradation may cause downregulation of Rpn4p expression and that the decreased activity of the proteasome system may lead to Swe1p upregulation, which is normally degraded in the G2 phase by the ubiquitin–proteasome system (Kaiser et al, 1998). As expected, the RPN4 mRNA level in the yap1Δ cells, in comparison with that in wild‐type cells, was diminished (Fig 6A).
To investigate whether the downregulation of the RPN4 mRNA level indeed leads to the decreased activity of the proteasome, we compared the activities of the ubiquitin–proteasome of wild‐type and yap1Δ cells by measuring the levels of the ubiquitin–β‐galactosidase conjugate protein, a well‐characterized proteasome substrate (Johnson et al, 1995). Of the doublet bands observed, the slower migrating band was shown to correspond to Ub‐P‐βgal (Saeki et al, 2004). As shown in Fig 6B, wild‐type cells treated with exogenous calcium for 3 h showed a diminished activity of degrading the substrate protein, and its degradation was Yap1p dependent. No significant additive effect of yap1Δ mutation on the rate of protein degradation by the rpn4Δ strain was observed, suggesting a Yap1p–Rpn4p linear regulatory pathway (Fig 6B,C). The degradation of the substrate protein in the cnb1Δ strain was still weakly but reproducibly stimulated by calcium, suggesting that other calcium‐dependent mechanisms may be involved in the downregulation of the proteasome (Fig 6B). Consistent with the idea that the degradation of Swe1p and Cln2p is dependent on the proteasome‐mediated proteolysis, the Swe1p–Myc and Cln2p–HA levels of the rpn4Δ strains were higher than those of the wild‐type strain (Fig 6C). These results indicate that the downregulation of RPN4 expression caused by the calcium‐dependent Yap1p degradation may have contributed to the promotion of the G2 delay and polarized bud growth through the upregulation of Swe1p and Cln2p levels.
Here, we discovered a novel function of Yap1p in the calcineurin‐induced G2 cell‐cycle regulation (Fig 7). Yap1p has been well characterized as a central regulator of responses to oxidative stress (Balzi & Goffeau, 1994; Hirata et al, 1994; Kuge & Jones, 1994; Wemmie et al, 1994; Wu & Moye‐Rowley, 1994; Kuge et al, 1997). We also showed a new Yap1p function in calcium signalling. The calcium sensitivity phenotypes of the zds1Δ strain were partially suppressed by YAP1 overexpression, and, conversely, exacerbated by yap1 deletion. Our genetic data indicated Yap1p to be a negative regulator of the calcium signalling that is linked to the Swe1p‐ and Cln2p‐mediated regulation of cell cycle and morphology (Figs 1, 2). The Yap1p degradation caused the accumulation of Swe1p and Cln2p through the downregulation of the ubiquitin–proteasome system, in a manner mediated by the inhibition of the RPN4 expression. Supporting this possibility, the activity of the proteasome system, as determined by the level of the ubiqutin‐conjugated β‐galactosidase, was decreased by exogenous calcium in a manner dependent on the Yap1p degradation.
Previously, it was demonstrated that calcineurin interacts with Skn7p, a protein that also interacts with Yap1p, by virtue of an intermediate protein Crz1p in calcium signalling (Williams & Cyert, 2001). However, the Yap1p–Myc degradation was induced by calcium on the skn7Δ and crz1Δ backgrounds, similar to the wild‐type background, suggesting that Skn7p and Crz1p do not participate in this mechanism (H.Y. et al, unpublished results).
We showed that Yap1p was phosphorylated by an as yet unidentified kinase(s) and dephosphorylated by calcineurin, thus identifying Yap1p as a substrate of calcineurin (Fig 4). The Yap1p degradation was required for downregulating the expression levels of the Yap1p‐dependent genes, including the RPN4 gene. In the 5′ non‐coding region of the RPN4 gene, a Yap1p response element (YRE; TTACTAA) is located at position −373 relative to the ATG translation initiation codon (Wu & Moye‐Rowley, 1994). It was previously shown that the effect of the YAP1 deletion on the activation of the ubiquitin–proteasome system was seen only in the absence of the Pdr1p transcription factor (Owsianik et al, 2002). Nevertheless, a decrease in the levels of RPN4 mRNA and proteasome activity was seen on the yap1Δ PDR1 background (Fig 6A,B). These different observations may be due to the difference in the strain background. Alternatively, it may be possible that Pdr1p is also degraded similarly as Yap1p.
In the Ca2+‐induced G2 cell‐cycle regulation, calcineurin upregulates Swe1p through two distinct but cooperative pathways: one involving the activation of SWE1 transcription (Mizunuma et al, 1998) and the other the destabilization of Hsl1p, a negative regulator of the Swe1p kinase (Mizunuma et al, 2001). In the present study, we showed another role of calcineurin in Swe1p upregulation, that is, promotion of Swe1p accumulation through the calcineurin‐mediated degradation of the Yap1p transcription factor. Thus, the Swe1p upregulation by calcium signalling seems to be regulated multilaterally to ensure the Swe1p activation by at least three distinct calcineurin‐mediated mechanisms that operate on Swe1p at different levels: that is, transcriptional, post‐translational and degradation. The present mechanism seems to be part of a long adaptive response as it requires degradation of the Yap1p transcription factor that leads to regulation of another transcription factor, which per se regulated the machinery involved in the degradation of Swe1p and Cln2p. The calcineurin‐mediated Yap1p degradation seems to be a mechanism that assures a G2 delay in response to a stress that causes the activation of the calcium signalling pathways.
Isolation of multicopy suppressors of zds1Δ strain. For screening, the zds1Δ strain was transformed with a yeast genomic DNA library constructed in the high‐copy number plasmid YEp24, and the transformed cells were spread on SD minus uracil plates (8.4 × 104 cells per plate). The transformants were picked up and inoculated onto YPD plates supplemented with 300 mM CaCl2, a concentration that virtually inhibited the growth of the zds1Δ strain. From each of 145 calcium‐resistant transformants, plasmids were recovered after transformation into Escherichia coli. Of these, 122 plasmids reproducibly suppressed the calcium‐sensitive phenotype when they were re‐introduced into the zds1Δ strain.
General methods are available in the supplementary information online.
Supplementary information is available at EMBO reports online (http://www.emboreports.org).
We thank Dr M. Longtine for the gift of the pFA6a‐13Myc‐kanMX6 plasmid and Dr A. Toh‐e for the Ub‐P‐βgal plasmid, as well as T. Machida for isolation of the multicopy suppressors. This work was supported in part by Grant‐in‐Aid for Scientific Research from the Ministry of Education, Science, Culture, Arts and Sports of Japan to M.M. (16770151) and T.M. (14206012, 17208009), and by Research Grants from the Naito Foundation to M.M. and from the Noda Foundation and Asahi Glass Foundation to T.M.
- Copyright © 2006 European Molecular Biology Organization