Implication of Pkc1p protein kinase C in sustaining Cln2p level and polarized bud growth in response to calcium signaling in Saccharomyces cerevisiae

Cell polarity is essential for morphogenesis and development (Drubin and Nelson, 1996). In the budding yeast Saccharomyces cerevisiae, a number of stage-specific events direct the position and the morphology of the bud. Environmental stimuli activate a bifurcated signal transduction pathway to promote a coordinated response involving a specific pattern of gene expression, G₂/M cell-cycle delay, apically polarized actin distribution and unipolar distal bud-site selection.

As yeast cells begin the G₁/S transition, Cln1p and Cln2p G₁ cyclins accumulate, activate the cyclin-dependent kinase Cdc28p, and induce clustering of actin at the presumptive bud site, thus focusing secretion of new cell-wall components onto the nascent bud (Pruyne and Bretscher, 2000). Continued Cdc28p-Cln1/2p activity maintains this clustered distribution of actin in the bud tip and polarized growth (Lew and Reed, 1993). Cells over-expressing Cln1p or Cln2p form a highly elongated bud because of sustained polarized growth (Lew and Reed, 1995). Polarized growth continues in G₂, as the mitotic cyclin Clb1/2p-Cdc28p activity remains sequestered via inhibitory phosphorylation by Swe1p, a protein kinase that inhibits Cdc28p (Booher et al., 1993; Lew and Reed, 1995). Recently, we reported that Ca²⁺-signaling pathways induced the formation of highly polarized bud and delayed the cell-cycle progression in G₂ (Mizunuma et al., 1998; Mizunuma et al., 2001). The effect of Ca²⁺ on cell-cycle regulation and polarized bud growth are pronounced on a zds1Δ background lacking the negative regulator for SWE1 and CLN2 transcription (Bi and Pringle, 1996; Ma et al., 1996; Yu et al., 1996; Mizunuma et al., 1998; Mizunuma et al., 2001; Mizunuma et al., 2004). These phenotypes are peculiar to calcium, because a similar effect was not observed with sorbitol (300 mM) or MgCl₂ (100 mM; our unpublished data).

In S. cerevisiae, a unique protein kinase C, Pkc1p, functions in a multiplicity of pathways, including those related to cell wall integrity, bud emergence, stretching of the plasma membrane (Mazzoni et al., 1993; Levin and Errede, 1995; Gustin et al., 1998; Heinisch et al., 1999) and organization of the actin cytoskeleton (Heinisch et al., 1999). One well-known pathway for Pkc1p involves the sequentially activated protein kinases Bck1p, the redundant Mkk1p/Mkk2p and Mpk1p (Slt2p), which ultimately activates by phosphorylation of transcription factors, including the Rlm1p and SBF (Swi4p-Swi6p) complex (Dodou and Treisman, 1997; Madden et al., 1997; Baetz et al., 2001). This pathway is essential for cell wall integrity, bud emergence, responses to hypotonic shock, stretching of the plasma membrane, and repolarization of actin upon cell-wall stress (Levin et al., 1990; Levin and Bartlett-Heubusch, 1992; Paravicini et al., 1992; Yoshida et al., 1992; Lew and Reed, 1995; Davenport et al., 1995; Kamada et al., 1995; Igual et al., 1996; Marini et al., 1996; Zarzov et al., 1996; Gray et al., 1997; Gustin et al., 1998; Helliwell et al., 1998; Delley and Hall, 1999). In addition, Pkc1p is involved in controlling the depolarization of actin during adaptation to growth at higher temperatures in a manner independent of the Mpk1p MAP kinase pathway (Delley and Hall, 1999). However, little is known about this Pkc1p effecter branch.

Much of our understanding of Pkc1p functions has come from the phenotypes caused by MPK1 deletion and/or severe temperature-sensitive alleles of PKC1, all of which involve a defect in the cell wall. The identification of a Mpk1p-independent role of Pkc1p will be facilitated by the isolation and characterization of a PKC1 allele specifically defective in the novel function.

Here, we describe the isolation and characterization of a pkc1 mutant allele (pkc1-834) defective in a previously unknown pathway. The pkc1-834 allele was defective in the maintenance of Ca²⁺-induced F-actin polarization in a manner independent of Mpk1p activation. This phenotype appeared to be caused by decreased expression of Cln2p. Pkc1p was required for posttranscriptional upregulation of Cln2p. The Rho1 small G protein molecular switch was suggested to be involved in the novel Pkc1p function.

Yeast strains used in this study are listed in Table 1. The stt1-1 (SYT11-12A) mutant was backcrossed to the W303 strain at least five times to homogenize the background. Yeast cells were grown in YPD medium (1% yeast extract, 2% Bacto-peptone, 2% glucose supplemented with 400 μg/ml adenine and 200 μg/ml uracil).

Sequencing of double strand DNA was done by the cycle sequencing method, using a ALFred DNA sequencer (Amersham Biosciences). For PCR amplification of the genomic DNA, a Thermo Sequenase fluorescently labeled primer cycle sequencing kit with 7-deaza-dGTP (Amersham Biosciences) was used. M13 universal and reverse primers from Cy5™ AutoRead Sequencing Kit (Amersham Biosciences) were also used. The entire open reading frame of PKC1 was divided into four regions. PCR products were synthesized using primers, region I: 5′-AAAGAGCTCATGAGTTTTTCACAAT-3′ (sense strand, the SacI site is underlined) and 5′-TGCACCTGCAGATTCTTTGCTGCTGAAGTAGAC-3′ (antisense strand, the PstI site is underlined). Region II: 5′-AAACTGCAGGTGCAAGGCGAAAATA-3′ (sense strand, the PstI site is underlined), and 5′-AACTTCTAGATCCTCTTGGATTTCC-3′ (antisense strand, the XbaI site is underlined). Region III: 5′-AAATCTAGAAGTTGATCACATTGAT-3′ (sense strand, the XbaI site is underlined), and 5′-GATAGGTCGACTCTTTGGTTTTGAA-3′ (antisense strand, the SalI site is underlined). Region IV: 5′-AAAGTCGACCTATCTGTAAGAAGGG-3′ (sense strand, the SalI site is underlined), and 5′-TTTCTAAGCTTTCATAAATCCAAATCATCT-3′ (antisense strand, the HindIII site is underlined).

To construct plasmids containing the PKC1 gene with single mutations, we employed site-directed mutagenesis. The mutagenic primers for PKC1 (P1102S) were 5′-CCACCCTACATCTCAGAAATTAAATCTCCG-3′ and 5′-CGGAGATTTAATTTCTGAGATGTAGGGTGG-3′; and those for the PKC1 (N834D), 5′-GTTCTTGGTAAAGGTGATTTTGGTAAAG-3′ and 5′-CTTTACCAAAATCACCTTTACCAAGAAC-3′. PCR was performed with PfuTurbo DNA polymerase (Stratagene) and pUC119-PKC1 as a PCR template. Successful mutagenesis was confirmed by DNA sequencing. The low-copy plasmid YCp50-PKC1 (P1102S) or YCp50-PKC1 (N834D) construct was generated as follows. The SalI-PvuII fragment of pUC119-PKC1 containing PKC1 (P1102S) or PKC1 (N834D) was ligated into the SalI-NruI site of YCp50. YCp50-PKC1 (P1102S) or YCp50-PKC1 (N834D) was used for complementation analysis for pkc1Δ or zds1Δ pkc1Δ.

The pkc1Δ strain in the W303 background was constructed by gene replacement. Genomic DNA was isolated from the pkc1Δ::LEU2 (DL376) strain (Levin et al., 1990). The PKC1 locus was amplified by PCR using primers 5′-ATGTTTGTCCCACTCCAGGTTGCAC-3′ and 5′-TTTGAGACGTCATGAACTCTCGCGG-3′. The amplified fragment was used to transform a diploid W303 strain [ZDS1/ZDS1 (W303-1D) or zds1Δ::TRP1/ZDS1(YMM235)]. Replacement was confirmed by PCR. The diploid was sporulated and the tetrads were dissected and analyzed. When the spores were grown at room temperature, all of the tetrads produced two colonies that grew at the wild-type rate and two colonies that grew very poorly even in the presence of 1 M sorbitol. The fast growing colonies all had a Leu^– phenotype, indicating that the very poor growth results from the integration of the knockout construct. Deletion of the PKC1 gene in the Leu⁺ strains was further confirmed by complementation of the very poor growth by a centromere plasmid carrying the PKC1 gene.

The pkc1Δ zds1Δ (pkc1Δ) strain bearing the N834D, P1102S or wild-type PKC1 allele was constructed as follows. The YCp50-PKC1 (P1102S), YCp50-PKC1 (N834D) or YCp50-PKC1 plasmid was transformed into a diploid strain (YMM236 or YMM235). The diploid was sporulated and the tetrads were dissected and analyzed. The spores that were grown minus uracil (Ura), tryptophan (Trp) and leucine (Leu) (which selects for the zds1Δ pkc1Δ strain bearing the N834D, P1102S or wild-type PKC1 allele) or minus Ura and Leu (which selects for the pkc1Δ strain bearing N834D, P1102S or the wild-type PKC1 allele) were isolated and their phenotypes examined.

The zds1Δ strain of the YPH500 background was constructed by gene replacement. Disruption of the ZDS1 gene was performed using a zds1 disruption plasmid pUC119-zds1::URA3. The zds1::URA3 fragment was used to transform the YPH500 background.

Actin was visualized in formaldehyde-fixed cells by using Rhodamine-conjugated phalloidin as described previously (Adams and Pringle, 1991).

Qualitative assessment of cell lysis in colonies was done using the alkaline phosphatase assay described previously (Saka et al., 2001). Colonies were formed on YPD agar plates at 25°C for 2 days, and the plates were then shifted to 35°C and overlaid with an alkaline phosphatase assay solution. Colonies that contain lysed cells stain blue, whereas intact colonies remain white.

Cell culture synchronization was done by the procedure described previously (Mizunuma et al., 1998). Total RNA was isolated by the hot phenol method. The isolated RNA was separated on a 1% agarose gel, transferred to a nylon membrane, and then subjected to northern blot analysis. The SWE1, CLN2, and ACT1 probes were generated by random-primed labeling of a 0.7 kb BglII fragment of SWE1, a 1.3 kb NcoI-XhoI fragment of CLN2 and a 1.1 kb XhoI-KpnI fragment of ACT1, respectively, with [α-³²P]dCTP by use of a multiprime DNA labeling kit (Amersham Biosciences).

Preparation of cell extracts and immunoprecipitations were performed basically as described previously (Mizunuma et al., 2001; Mizunuma et al., 2004). For detection of the HA-tagged proteins, Myc-tagged proteins, Pkc1p, phosphorylated Mpk1, Mpk1p and Cdc28p, monoclonal antibodies 12CA5 (BAbCO) against the HA epitope, 9E10 (BAbCO) against the Myc epitope, polyclonal anti-PKC1 (Santa Cruz Biotech.), anti-phospho-p44/42 MAP kinase (New England Biolabs Inc.), anti-MAPK antibody (Santa Cruz Biotech.), and anti-PSTAIRE (Santa Cruz Biotech.), respectively, were used.

The growth of the zds1Δ strain of yeast in medium containing 300 mM CaCl₂ is severely inhibited, showing G₂ arrest and highly polarized bud growth. In our previous study, suppressor mutants of the calcium-sensitive phenotypes were isolated and classified into 14 complementation groups (Mizunuma et al., 2001). One of these mutants, scz6, is the subject of this report. As shown in Fig. 1A, the inhibition of the growth of zds1Δ cells by exogenous calcium was suppressed by an additional mutation, scz6. Among the physiological effects of calcium on the zds1Δ strain, the formation of a highly polarized bud and G₂ delay were elicited by exogenous CaCl₂ at concentrations (∼100 mM) lower than those (∼300 mM) required for the growth inhibition. We therefore used 100 mM CaCl₂ in most of the subsequent experiments to examine the effect of calcium on the cell cycle and morphology. The polarized bud growth induced by calcium was also suppressed by the scz6 mutation (Fig. 1B), whereas the other calcium phenotype (G₂ delay) was not suppressed by it (Fig. 1C). These results suggested that the suppression of growth inhibition and polarized bud growth, but not of the cell-cycle delay, was the consequence of the scz6 mutation. In addition to these suppressor phenotypes, the zds1Δ scz6 mutant exhibited a growth defect at 37°C (Fig. 1A). We further found that the scz6 mutation in itself exhibited a slow growth phenotype at temperatures between 25°C and 37°C (Fig. 1A). To identify the mutant gene, we introduced a centromeric genomic library into the scz6 strain and recovered a plasmid that suppressed the slow growth at 37°C. The PKC1 gene, encoding a unique protein kinase C, was identified as the gene that complemented the scz6 mutation. To demonstrate that the mutation was situated at the PKC1 locus, we constructed a YIp5-PKC1 plasmid for integration and linearized the plasmid for use in transforming the wild-type strain. The transformants were crossed with the scz6 strain, and the resulting diploid was sporulated and subjected to tetrad analysis. In all 12 tetrads the Ura⁺ or Ura^– phenotype cosegregated with normal or slow growth, respectively, indicating that the mutation was situated at, or closely linked to, the PKC1 locus (Levin et al., 1990; Levin and Bartlett-Heubusch, 1992). As will be shown later, because the scz6 mutation causes a change of amino acid 834 of Pkc1p from asparagine to aspartic acid (Fig. 5), we shall hereafter refer to scz6 mutation as pkc1-834.

Why does the pkc1-834 allele of PKC1 suppress only the calcium-induced hyperpolarized bud growth, but not that of G₂ delay? Previously, we showed that exogenous calcium induces the sustained transcription of the SWE1 and CLN2 genes (Mizunuma et al., 1998; Mizunuma et al., 2001; Mizunuma et al., 2004). We therefore compared the effect of exogenous calcium on SWE1 and CLN2 mRNA levels in synchronized cell cultures of wild-type and pkc1-834 strains by northern blot hybridization (Fig. 2). However, no effect of calcium with respect to either timing or expression levels of these genes was detected.

We next investigated the effect of exogenous calcium on Swe1p and Cln2p levels by western blot analysis of synchronized cell cultures of strains with chromosomally integrated constructs for both Myc-tagged Swe1p and HA-tagged Cln2p (Fig. 3A). The Swe1p and Cln2p levels in the cells released from G₁ arrest were analyzed by western blotting. As previously reported, Swe1p and Cln2p levels in wild-type cells oscillated during the cell cycle, peaking at the time of bud emergence and declining before nuclear division (McMillan et al., 1998; Sia et al., 1998). Exogenous calcium caused a 20-minute lag in the increase in Swe1p and Cln2p levels, reflecting the delays in the elevation of the respective mRNA (Fig. 2); and the increased protein levels were sustained longer by calcium (Fig. 3A). There was no significant difference among the strains when cultivated in YPD or YPD plus CaCl₂ medium with regard to the periodic patterns of the Swe1p level. In the presence of exogenous calcium, Swe1p in the pkc1-834 and zds1Δ pkc1-834 cells accumulated, consistent with the observation that the scz6 mutation failed to suppress the G₂ delay caused by calcium (Fig. 1C and Fig. 3A). The G₂ delay was partially suppressed by the deletion of the SWE1 gene, suggesting that the delay was indeed mediated at least partially by the activation of Swe1p (data not shown). However, Cln2p levels in the pkc1-834 cells were significantly lower than those in the wild-type strain. Moreover, Cln2p levels in the pkc1-834 strain, in contrast to those in the wild-type strain, were not elevated by exogenous calcium in either ZDS1 or zds1Δ backgrounds, suggesting that Pkc1p is important for the elevation and maintenance of Cln2p levels in response to exogenous calcium (Fig. 3A). The defect in Cln2p accumulation in the pkc1-834 cells could be due to a decreased rate of Cln2p synthesis, the destabilization of Cln2p, or both.

To examine the stability of Cln2p, we determined the decay rate of Cln2p in the cells treated with cycloheximide at a restricted temperature (Fig. 3B). In random culture of pkc1-834 cells, it was noted that the Cln2p abundance at time zero was similar to that of wild-type cells, possibly reflecting the prolonged G₂ phase in the pkc1-834 cells in random culture. The abundance of Cln2p in wild-type and stt1-1 cells decreased rapidly, reflecting the rapid turnover of this protein (Lanker et al., 1996). On the other hand, the Cln2p level in pkc1-834 mutant cells was maintained more stably. This could be due to a defect in Cln2p destabilization in the pkc1-834 strain (see Discussion). These results, together with the observation that the pkc1-834 mutation suppressed the Ca²⁺-induced hyper-polarized bud growth, further suggested that the Pkc1p-mediated Cln2p accumulation was important for polarizing bud growth. Consistent with this notion, the over-expression of CLN2 from a GAL1 promoter in the zds1Δ pkc1-834 strain led to hyper-polarized bud growth (Fig. 3D).

Using the same synchronous cell cultures as used in the experiment for Fig. 3A, we determined the cellular DNA content (Fig. 3C). As previously reported (Mizunuma et al., 1998; Mizunuma et al., 2001), a delay in G₁, as well as in G₂, was induced by calcium in all the strains examined (compare YPD and YPD + Ca²⁺ of Fig. 3C). The G₁ delay seemed to be due to the downregulation of CLN2 mRNA (Fig. 2). The pkc1-834 cells showed a G₁ delay after release from G₁ arrest, even in the absence of exogenous calcium, indicating that Pkc1p is required for the G₁-S transition (Fig. 3C).

Both actin polarization and hyper-polarized bud growth is dependent on the G₁ cyclins (Lew and Reed, 1993). Cell polarization can be estimated by the distribution of F-actin. We therefore examined the ability of the pkc1-834 cells to polarize their F-actin cytoskeleton in response to calcium. Using the same synchronous cell cultures described in the above section, we examined the effects of exogenous calcium on bud emergence and F-actin polarization at the bud site (Fig. 4). After a 2-hour treatment of the cells with α-factor, about 80% of wild-type cells yielded a mating projection, whereas in the pkc1-834 cells it was about 30%, although the treatment resulted in a G₁ block in the mutant cells (data not shown and Fig. 3C). However, pkc1-834 cells on the zds1Δ background formed a mating projection at a ratio comparable to that of the wild-type cells. Therefore, we used the zds1Δ mutant background in this study (Fig. 4). Upon release of the α-factor-treated cells to YPD medium, bud emergence of the zds1Δ pkc1-834 cells was delayed by 20 minutes compared with that of the zds1Δ cells. Bud emergence of both zds1Δ pkc1-834 and pkc1-834 cells was delayed by exogenous calcium similarly (Fig. 4B). These results suggested that the suppression of the calcium-induced polarized bud growth by the pkc1-834 mutation is not due to the defect in bud emergence.

In YPD medium, the F-actin organization of the zds1Δ and zds1Δ pkc1-834 cells, as determined by rhodamine-phalloidin staining, was essentially normal (Fig. 4A,B, compare 40∼80-minute samples for the zds1Δ and zds1Δ pkc1-834 cells in YPD medium). Patches of F-actin were restricted to the growing buds, and F-actin cables were oriented toward the bud. Large budded cells were capable of polarizing F-actin to the bud neck (Fig. 4A, 100∼120 minute samples for the zds1Δ and zds1Δ pkc1-834 cells in YPD medium). However, in the presence of calcium, F-actin polarization in the zds1Δ pkc1-834 cells was strikingly different from that in the zds1Δ cells (Fig. 4). In zds1Δ cells, the accumulation of F-actin patches in the bud after release from G₁ arrest was transiently disturbed by exogenous calcium (compare with and without (±) Ca²⁺samples of zds1Δ cells at 40 minutes), and then the patches became rapidly localized to the bud (see zds1Δ cells +Ca²⁺ at 40 and 80 minutes). The patches were maintained in the growing buds during polarized growth (compare ±Ca²⁺ samples of zds1Δ cells at 100 and 120 minutes). The timing of F-actin polarization was coincident with the elevation of Cln2p levels during the cell cycle (Fig. 3A), suggesting that Cln2p elevation is important for the polarization of F-actin patches. In the zds1Δ pkc1-834 cells, F-actin patches were transiently dispersed throughout the cell and then rapidly became localized at the bud, as in the zds1Δ cells (compare ±Ca²⁺ samples of zds1Δ pkc1-834 cells at 40 and 80 minutes). However, the maintenance of the localized F-actin patches in growing buds of the zds1Δ pkc1-834 cells was abolished faster than that in the zds1Δ cell (compare +Ca²⁺ samples of the zds1Δ and zds1Δ pkc1-834 cells at 120 minutes). F-actin patches of these cells varied, being concentrated in the bud or the bud neck or dispersed throughout the cells. The failure of maintenance of F-actin polarization in the zds1Δ pkc1-834 cells might be due to the low level of Cln2p expression (Fig. 3A). These results suggested that Pkc1p is important for the maintenance of F-actin polarization, but not for the de- and repolarization of it, in response to exogenous calcium.

Yeast bearing the stt1-1 mutation, a previously characterized temperature-sensitive PKC1 allele, exhibits a defect in cell wall maintenance and in G₂-M transition (Yoshida et al., 1992). So we examined whether the stt1-1 mutation could suppress the calcium sensitivity phenotypes of the zds1Δ strain. However, the stt1-1 allele only partially suppressed the phenotype, suggesting that the defect of the pkc1-834 allele may be functionally distinct from that of the stt1-1 allele (Fig. 1A,B).

In characterization of the pkc1-834 and stt1-1 mutations, the former mutation was found to involve an asparagine to aspartic acid substitution at amino acid position 834, and the latter, a proline to serine substitution at amino acid position 1102 of Pkc1p (Fig. 5A). The stt1-1 mutation (P1102S) was within the kinase domain in a highly conserved proline residue, whereas the pkc1-834 mutation (N834D) was within the predicted ATP-binding site of Pkc1p in an evolutionarily non-conserved residue. We constructed N834D and P1102S mutant versions of the PKC1 gene on a centromeric plasmid. The N834D and P1102S alleles could suppress any of the phenotypes of the pkc1-834 or stt1-1 mutation, respectively (data not shown). Furthermore, the pkc1Δ zds1Δ strain bearing the N834D or P1102S allele exhibited calcium phenotypes similar to those of the pkc1-834 zds1Δ or stt1-1 zds1Δ strain, respectively, confirming that each mutant allele is responsible for the observed phenotype (Fig. 5B).

It was previously shown that the growth of the pkc1-1 strain (on EG123 background) was dependent on the presence of exogenous calcium in the medium (Levin and Bartlett-Heubusch, 1992). Interestingly, the mutation site of pkc1-834 (N834D) coincided with that of this pkc1-1 mutation (N834K). The pkc1-834 mutant was able to grow on a YPD plate (without added CaCl₂) at a rate slower than that of the wild-type strain, and the growth rate could be restored to an apparently normal level by the addition of CaCl₂, but not MgCl₂ (Fig. 1A and Fig. 5B; data not shown). Thus, the N834 residue seems to be involved somehow in the regulation of Pkc1p by calcium. Western blot analysis of Pkc1p in cell extracts of wild-type, pkc1-834 and stt1-1 strains, using an anti-Pkc1p antibody, demonstrated that the expression level of Pkc1-834p and of Stt1-1p was comparable to that of the wild-type Pkc1p (data not shown).

We further genetically characterized the defects of the two PKC1 mutant alleles. Homo- and hetero-allelic diploids were constructed and examined for their temperature sensitivities and resistance to calcium on the zds1Δ background. In the presence of 100 mM CaCl₂ at 37°C, the homo-allelic diploids pkc1-834/pkc1-834 and stt1-1/stt1-1 grew poorly, whereas the hetero-allelic diploid (pkc1-834/stt1-1) was able to grow (all on the zds1Δ background) showing an intragenic complementation (Fig. 6A). This result indicated that the defects of the stt1-1 and pkc1-834 mutations were functionally different. In addition, various diploid strains, with respect to the PKC1 allele, exhibited a range of calcium resistance: (in decreasing order) pkc1-834/pkc1-834, pkc1-834/stt1-1, stt1-1/stt1-1 and PKC1/PKC1 (all on the zds1Δ background), suggesting that the wild-type Pkc1p inhibits growth in the presence of calcium and that the mutation pkc1-834 reduces the inhibitory effect more than stt1-1 (Fig. 6A).

We further compared the phenotypic differences of the pkc1-834 and stt1-1 alleles. To see if the pkc1-834 mutant exhibited similar defects to the stt1-1 mutant, we examined the cell-lysis phenotype by a simple plate-overlay assay using a plate containing BCIP (Saka et al., 2001). Although the stt1-1 colonies exhibited a cell-lysis phenotype at 35°C, the pkc1-834 colonies remained apparently normal, suggesting that the Pkc1p-Mpk1p pathway functioned adequately in pkc1-834 cells (Fig. 6B). To evaluate the activation of Mpk1p in Pkc1p mutants upon shift of the growth temperature from 25°C to 37°C, we determined the level of phospho-Mpk1p using a monoclonal antibody that specifically recognizes phospho-Mpk1p. Although, Mpk1p phosphorylation in stt1-1 cells was very low, Mpk1p was similarly phosphorylated in pkc1-834 and wild-type cells (Fig. 6C). Moreover, over-expression of MPK1 partially suppressed the temperature sensitivity of the stt1-1 strain, but not that of the pkc1-834 strain (Fig. 6D). Furthermore, both pkc1-834 mpk1Δ and stt1-1 mpk1Δ double mutations were lethal, and only the lethality caused by the pkc1-834 mpk1Δ double mutation was suppressed by 1 M sorbitol, suggesting that the pkc1-834 mutation has additional defect(s) in essential growth function(s) (data not shown). Based on these results, we concluded that pkc1-834 cells have an important defect in their Mpk1p-independent function(s).

The degradation of Hsl1p, a negative regulator of Swe1p, via the ubiquitin-mediated 26S proteasome pathway is triggered by the activation of calcineurin and the Mpk1p-Mck1p pathway (Mizunuma et al., 2001). Pkc1p is known as an upstream regulator of the Mpk1p MAPK pathway in cell wall construction. We expected that calcium-induced Hsl1p degradation, which is dependent on the activation of Mpk1p occurs normally in the pkc1-834 strain, but not in the stt1-1 strain. However, the level of Hsl1p in these strains, in contrast to that in the mpk1Δ strain, decreased more rapidly, in a similar manner as in wild-type strain (Fig. 7). Consistent with this observation, both pkc1-834 and stt1-1 alleles failed to suppress the Ca²⁺-induced G₂ delay of the zds1Δ strain (Fig. 1C), suggesting that this effect is due to the activation of Swe1p. Thus, it was suggested that the Pkc1-834p and Stt1-1p mutant proteins are functional in regulating the Hsl1p degradation. Alternatively, it was also possible that lower activity of Mpk1p in the stt1-1 cells is still sufficient to trigger Hsl1p degradation.

It was previously indicated that the stt1-1 mutation causes a defect in the activation of Mpk1p MAPK by heat shock. Our observation that the stt1-1 mutation failed to suppress the Ca²⁺-induced hyper-polarized bud growth of the zds1Δ strain suggested that Mpk1p MAPK function may not be required for sustaining a high level of Cln2p in the presence of exogenous calcium (Fig. 1A,B). Therefore, we presumed that the stt1-1 mutation does not affect the maintenance of high Cln2p levels in the presence of calcium. As expected, no significant difference was detected in the expression levels of Cln2p between wild-type and stt1-1 strains, using western blot analysis (Fig. 3A). There was also no significant difference in the levels of CLN2 mRNA between stt1-1 and wild-type strains (Fig. 2). These results indicated that Pkc1p is required for the maintenance of the calcium-induced elevation of Cln2p levels in a manner independent of Mpk1p MAPK activation.

In the activation of the Mpk1p MAPK pathway, Pkc1p is a down-stream component of Rho1p, a small G protein molecular switch (Nonaka et al., 1995; Kamada et al., 1996; Helliwell et al., 1998; Delley and Hall, 1999). Because Rho1p is an essential protein, several conditional, lethal (high temperature-sensitive) mutations in the RHO1 gene have been isolated and characterized (Helliwell et al., 1998; Saka et al., 2001). Of these, rho1-2 and rho1-5 strains are defective in the activation of Mpk1p, whereas rho1-3 and rho1-4 mutants are not (Saka et al., 2001). We examined allele specificity of various rho1 mutations on the Ca²⁺-induced polarized bud growth of the zds1Δ strain. Consistent with the distinct role of Pkc1p in Ca²⁺-induced polarized bud growth, the rho1-3 and rho1-4 mutations, but not the rho1-2 and rho1-5 mutations, suppressed the calcium phenotypes of the zds1Δ strain, suggesting that the defect caused by the rho1-3 and rho1-4 alleles, rather than the rho1-2 and rho1-5 alleles, are related in having the novel Pkc1p function (Fig. 8). Consistent with the notion that the functional defects of the rho1-3 and rho1-4 alleles are related with that of the pkc1-834 allele (Fig. 3A), the levels of Cln2p in rho1-3 and rho1-4 strains and the pkc1-834 strain were similarly low compared with those of wild-type and rho1-2 and rho1-5 strains (data not shown). The typical effects of calcium, observed in zds1Δ mutants of W303 strain background, namely Ca²⁺-induced growth arrest, hyper-polarized bud growth and G₂ delay, were also observed in the YPH500 strain background at calcium concentrations higher than those for the W303 strain. The pkc1-834 mutation, but not that of stt1-1, suppressed the Ca²⁺-induced hyper-polarized bud growth on this strain background (data not shown). Taken together, these data implicated the Mpk1p-independent novel pathway of Pkc1p in Ca²⁺-induced polarized bud growth in a manner dependent on Rho1p.

We have demonstrated a novel important role of S. cerevisiae protein kinase C, i.e. a role in the maintenance of polarized bud growth in a manner independent of Mpk1p MAPK activation. Thus, the strain with the pkc1-834 allele, which is defective in this function, is apparently normal in polarity establishment, but defective in the maintenance of polarized growth during budding and formation of the mating projection. The effect of calcium on cell-cycle progression and polarized bud growth are pronounced on a zds1Δ background lacking the negative regulator for SWE1 and CLN2 transcription, leading to the elevation of Cln2p and Swe1p (Mizunuma et al., 1998; Mizunuma et al., 2001; Mizunuma et al., 2004). Pkc1p was important for the elevation and maintenance of the Cln2p level, but not for that of Swe1p, indicating that the regulatory mechanisms for Swe1p and Cln2p are different.

The polarized bud growth and defect in cell proliferation, but not the G₂ delay, caused by high calcium in the zds1Δ strain were abolished by the pkc1-834 mutation (Fig. 1). This suggested the possibility that the growth inhibition of zds1Δ cells by high calcium is due to the hyperpolarization of bud growth induced by high Cln2p levels. However, we consider this possibility unlikely for the following reasons. The growth of the cln2Δ zds1Δ strains was more sensitive to calcium than growth of the zds1Δ strain, although cln2Δ zds1Δ cells failed to show hyperpolarized bud growth in response to calcium (data not shown). Moreover, the rho1-3 and rho1-4 mutant strains (like the pkc1-834 strain) that expressed lower levels of Cln2p than the wild-type strain could not suppress calcium sensitivity of zds1Δ strain (data not shown). Furthermore, double mutant strains having the pkc1-834 mutation in combination with either of the rho1-2, -3, -4 or -5 mutations, exhibited more severe growth defect on YPD medium at a permissive temperature (25°C) than either of the single mutants, suggesting that rho1-2, 3, 4 and 5 mutations may have additional defect besides the defect in the activation of the Pkc1p (data not shown).

We have shown that the regulation of Cln2p expression (and probably Cln1p, as well) is mediated by Pkc1p at the posttranscriptional level (Fig. 3A,B). However, its molecular mechanism still remains to be elucidated. Thus, the level of Cln2p in the pkc1-834 strain, defective in the regulatory mechanism, was lower that in the wild-type strain, and Cln2p was more stable in the pkc1-834 strain (Fig. 3A,B). Since the phosphorylation of G₁ cyclin, which is reported to be due to the Cdc28-dependent autophosphorylation of the Cln subunit, is required for their rapid degradation (Lanker et al., 1996), the stable nature of Cln2p in the pkc1-834 strain may be the result of the lower activity of Cln2p-Cdc28p in this strain. Alternatively, it is also possible that Pkc1p may regulate Cln2p at the posttranscriptional level.

An earlier study suggested that the activation of Cdc28p-Cln1/2p in G₂ leads to hyperpolarization of F-actin and formation of elongated buds (Lew and Reed, 1993). Thus, the effect of the pkc1-834 mutation on the calcium-induced hyperpolarized bud growth of the zds1Δ strain can be explained by the reduced activity of the Cdc28p-Cln1/2p kinase, which is insufficient to fully promote F-actin polarization. During adaptation to higher temperature conditions, Pkc1p is required for both de- and repolarization of the actin cytoskeleton. The former is independent of the Mpk1p MAP kinase pathway, while the latter is dependent on it (Delley and Hall, 1999). By contrast, in the response to calcium, Pkc1p was not required for the de- and repolarization of F-actin (Fig. 4). Instead, Pkc1p was required for the maintenance of F-actin polarization (Fig. 4). We also found that Rho1p Rho GTPase is involved in Ca²⁺-induced polarized bud growth in a manner independent of Mpk1p. Pkc1p interacts with Rho1p in vivo and in vitro (Nonaka et al., 1995; Kamada et al., 1996). Pkc1p mutant protein (Pkc1-834p) of the pkc1-834 strain can physically interact with Rho1p, as revealed by the two-hybrid assay using LexA-Rho1p (Q68L, GTP-bound form) with GAD-Pkc1-834p (data not shown). Co-immunoprecipitation of Pkc1-834p and Rho1p in vivo was also observed (data not shown). The interaction of Rho1p and Pkc1-834p is consistent with the ability of Pkc1-834p to activate the Mpk1p pathway, leading to cell wall construction.

How does Pkc1p regulate the establishment and maintenance of polarized bud growth in zds1Δ strain? Earlier studies showed that the Pkc1p-Mpk1p pathway is activated at the G₁/S transition, concomitant with bud emergence, and that its activation is partly dependent on the activation of the Cdc28p-Cln1/2p complex (Marini et al., 1996; Zarzov et al., 1996; Gray et al., 1997). The pkc1-834 zds1Δ cells were able to initiate bud formation, but were unable to sustain polarized bud growth by exogenous calcium (Fig. 4). Even the reduced levels of Cln1p and Cln2p in pkc1-834 zds1Δ cells appeared to be sufficient to establish polarized bud growth, but insufficient to support Ca²⁺-induced hyper-polarized bud growth. Our data indicated that Pkc1p is important for maintaining the Cln2p levels, placing the Pkc1p function upstream of Cln1p and Cln2p. Our results and those of others indicate that the Pkc1p-Mpk1p pathway is required for the establishment of polarized bud growth, while the Mpk1p-independent Pkc1p pathway is required for its maintenance.

How does Ca²⁺ signaling induce polarized bud growth and G₂ cell-cycle delay? These may be achieved by maintaining a high level of the G₁ cyclins and by triggering the Swe1p-mediated inhibition of Cdc28p-Clb1/2p. Of these, the former is dependent on the novel Pkc1p function, whereas the latter is mediated by a coordinated action of the calcineurin and Mpk1p-Mck1p pathways to activate Swe1p, which is achieved by elevating its abundance and downregulating Hsl1p (a Swe1p negative regulator) leading to the accumulation of the hyperphosphorylated, active form of Swe1p (Mizunuma et al., 2001). However, the possibility that Pkc1p is involved in the downregulation of Hsl1p could not be excluded, because the calcium-induced Hsl1p degradation in pkc1Δ strain could not be determined because the growth of this strain was extremely poor even in the presence of an osmo-stabilizer. Significantly, the swe1Δ pkc1-834 double mutant exhibits a severe growth defect even at the temperatures permissive for the pkc1-834 strain (data not shown), suggesting that Ca²⁺-mediated coordinated regulation of polar bud growth and cell-cycle is important for normal cell growth. It will be of considerable interest to study whether a similar mechanism operates in higher eukaryotes. A model summarizing our results is shown in Fig. 9 

We thank Yoshimi Takai, Kazuma Tanaka, Yoshikazu Ohya, and Tsutomu Kishi for gifts of plasmids or strains, as well as Hisako Nonaka, Keiko Ehara and Koichi Kirizawa for their technical help. 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. (14780546, 16770151) and T.M. (14206012, 17208009), and by a Research Grant of the Naito Foundation to M.M.