Genetic Evidence for a Morphogenetic Function of the Saccharomyces cerevisiae Pho85 Cyclin-Dependent Kinase

Corresponding author: Erin K. O'Shea, Department of Biochemistry and Biophysics, University of California, 513 Parnassus Ave., Rm. S-960, San Francisco, CA 94143-0448., oshea{at}biochem.ucsf.edu (E-mail)

Communicating editor: M. JOHNSTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The Saccharomyces cerevisiae PHO85 gene encodes a nonessential cyclin-dependent kinase that associates with 10 cyclin subunits. To survey the functions provided by Pho85, we identified mutants that require PHO85 for viability. We identified mutations that define seven Pho Eighty-Five Requiring or Efr loci, six of which are previously identified genes—BEM2 (YER155C), SPT7 (YBR081C), GCR1 (YPL075W), SRB5 (YGR104C), HFI1 (YPL254W), and BCK1 (YJL095W)—with one novel gene (YMR212C). We found that mutations in the EFR genes involved in morphogenesis are specifically inviable when the Pho85-associated G1 cyclins encoded by PCL1 and PCL2 are absent. pcl1 bem2, pcl1 pcl2 cla4, and pcl1 pcl2 cdc42-1 strains are inviable. pcl1 pcl2 mpk1, pcl1 pcl2 bck1, and pcl1 pcl2 cln1 cln2 strains are also inviable, but are rescued by osmotic stabilization with 1 M sorbitol. We propose that the G1 cyclins encoded by PCL1 and PCL2 positively regulate CDC42 or another morphogenesis promoting function.

CYCLIN-DEPENDENT kinases (CDKs) are heterodimeric protein kinases found in all eukaryotes. Although most CDKs associate with multiple cyclin subunits, these different cyclin-CDK complexes appear to have similar biological activities. For example, CDK4 associates with four different D-type cyclins that are subject to cell-type specific transcriptional regulation, but these complexes all promote cell-cycle progression through the restriction point (for review see SHERR 1995 ). The Saccharomyces cerevisiae CDK Cdc28 associates with nine different cyclins and these different Cdc28-containing complexes play essential roles in progression through different phases of the cell cycle, suggesting that cyclin binding might also confer some functional specificity on the CDK subunit.

S. cerevisiae has five CDKs. In addition to Cdc28, there are three CDKs (Srb10, Kin28, and Ctk1) that associate with only 1 or 2 cyclins and seem to be involved in regulating RNA Pol II transcription (LIAO et al. 1995 ; STERNER et al. 1995 ; VALAY et al. 1995 ). The remaining S. cerevisiae CDK is Pho85, a nonessential CDK that associates with 10 cyclins that can be divided into two families based on sequence similarity: the Pho80 cyclin family (Pho80, Pcl6, Pcl7, Pcl8, and Pcl10) and the Pcl1,2 family (Pcl1, Pcl2, Clg1, Pcl5, and Pcl9; for review see ANDREWS and MEASDAY 1998 ). In association with the cyclin Pho80, Pho85 phosphorylates the transcription factor Pho4, thereby inhibiting the expression of phosphate-starvation induced genes (for review see LENBURG and O'SHEA 1996 ). In association with Pcl8 and Pcl10, Pho85 phosphorylates the Gsy2 glycogen synthase, thereby inhibiting glycogen accumulation during fermentation (TIMBLIN et al. 1996 ; WILSON et al. 1999 ). Pho80-Pho85 is a potent Pho4-directed kinase but phosphorylates Gsy2 very poorly in vitro. Pcl10-Pho85 displays the opposite specificity in vitro (WILSON et al. 1999 ). These data and the differing phenotypes of strains with mutations in these cyclins indicate that Pho85 complexes containing different cyclin subunits can have distinct functions.

Transcription of PCL1, PCL2, and PCL9 peaks in G1, implicating PHO85 in cell-cycle regulation. The observation that deletion of PCL1 and PCL2 is lethal in a strain lacking CLN1 and CLN2 (two of the three Cdc28 G1 cyclins), causing growth arrest in G1, supports this hypothesis (ESPINOZA et al. 1994 ; MEASDAY et al. 1994 ). The cyclins in the Pcl1,2 family may participate in a process that affects cell polarity and morphogenesis. Strains that lack these Pho85-containing kinases display phenotypes such as salt sensitivity, reduced endocytosis, and a random budding pattern (LEE et al. 1998 ), phenotypes that are common to mutants defective in cell polarity and morphogenesis. Pho85-Pcl2 can phosphorylate Rvs167, a protein involved in actin polymerization, providing a possible mechanism by which Pho85 participates in establishing cell polarity during G1 (LEE et al. 1998 ). However, cells must lack all members of the Pcl1,2 family to display these cell-polarity related phenotypes (LEE et al. 1998 ; TENNYSON et al. 1998 ). It is possible that the morphogenetic defects of strains lacking these five cyclins are the result of a defect in a single Pho85-regulated process to which all these cyclins contribute or that these phenotypes are a secondary consequence of lesions in several distinct Pho85-regulated processes. As such, PHO85 provides a model system for understanding the situations in which cyclin binding confers functional specificity on the CDK subunit and the mechanisms by which this specificity is achieved.

In this study, we performed a screen for Pho Eighty-Five Requiring (Efr) mutants—mutants that require PHO85 for viability. If the functional specificity of Pho85 is achieved by cyclin association, we reasoned that we might be able to phenocopy the inviability of different efr pho85 strains by deleting the cyclins that associate with Pho85 to perform the particular function required for viability. By classifying different Efr mutants according to which Pho85-associated cyclins they require, we hoped to discover functional differences among the cyclins and something about the functions they perform. Our screen identified seven genes that when mutated make PHO85 essential for viability. Characterization of these mutants and directed tests for synthetic lethality between PHO85 and several other genes provides us with clues about the functions provided by the G1-specific Pho85-associated kinases.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Yeast manipulation:
Yeast were cultured and manipulated according to standard laboratory procedures, which have been described previously (GUTHRIE and FINK 1991 ). Strains and plasmids used in this article can be found in Table 1 and Table 2, respectively. Details of the construction of original plasmids are available on request.

The GAL1-10::PHO85 allele was generated by transforming MY0192 with plasmid MP136 linearized with BglII. The vast majority of transformants displayed a Pho⁺ phenotype on galactose-containing media and a constitutive PHO5 expression, or Pho^c, phenotype on dextrose-containing media. One of these transformants was plated on 5-fluoroorotic acid-containing plates. A Ura^- colony that displayed a galactose-dependent Pho85⁺ phenotype was identified and the presence of the GAL1-10 promoter replacing the PHO85 promoter was confirmed by PCR. This strain was named MY0246.

The transposon-insertion-containing DNA fragments we used as a mutagen in our screen for Efr mutants were derived from eight different pools of a yeast genomic library that had been mutagenized with a Tn3 transposon (BURNS et al. 1994 ). Four alleles of EFR1 were obtained from yeast transformed with fragments from two separate pools. Two alleles of EFR2 were obtained from yeast transformed with fragments from one pool of the library.

To generate new alleles of PHO85, PHO85 was amplified by error-prone PCR using Taq DNA polymerase at standard conditions except the concentration of MgCl₂ was lowered from 1.5 to 1.34 mM and MnCl₂ was added to 160 µM. We screened for temperature-sensitive (ts) alleles of PHO85 by cotransforming a cln1 cln2 GAL1-10::CLN1 pho85 strain (MY0126) with these PCR products and MP056 digested with EcoRI. Transformations were plated on SD at room temperature. Temperature-sensitive alleles of PHO85 were identified by replica plating the transformation plates and screening for colonies that were unable to grow at 37°. Positives from this screen were then tested for temperature-sensitive constitutive Pho5 activity (a ts-Pho^c phenotype) by an acid-phosphatase activity plate assay performed on strains growing in SG medium at room temperature or 37°. We recovered the PHO85 plasmid from three strains with a strong temperature-dependent Pho85 phenotype, and the PHO85 allele from each was used to replace the wild-type PHO85 locus. The allele that we named pho85-9 displays the most dramatic temperature-sensitive phenotype and was used in all experiments. pho85-9 strains display a variety of Pho85 phenotypes at elevated temperature such as an inability to grow in glycerol medium (GILLIQUET and BERBEN 1993 ) and supersensitivity to hydroxyurea (our unpublished observations) though we found that the nonpermissive temperature for different Pho85 phenotypes varied from 30° to 37°. We could not test whether the pho85-9 allele results in a glycogen hyperaccumulation phenotype (HUANG et al. 1996 ; TIMBLIN et al. 1996 ) because we cannot detect this phenotype in our pho85 strains.

High-copy plasmids carrying each of the Pho85-associated cyclins were made as follows. Genomic clones of PCL7, PCL8, PCL9, and PCL10 were identified in a YEp13 genomic library (NASMYTH and TATCHELL 1980 ) by colony hybridization using fragments isolated from pBA949, pBA946a, pBA950, and pBA945a, respectively. Fragments containing the cyclin locus from these plasmids were then subcloned into pRS426 using naturally occurring restriction enzyme sites. Plasmids containing other Pho85-associated cyclins were generated by subcloning fragments containing PCL1 (from pFHE27), PCL2 (from pBA619), PHO80 (from pAC800), PCL5 (from pBA906), and PCL6 (from pRS425-PCL6, a generous gift of Anita Sil) into pRS426. pRS426-CLG1 (pBA904) and the other pBA plasmids were the generous gift of Brenda Andrews.

Strains containing deletions of various Pho85-associated cyclins were constructed as follows. Deletion of PCL1 was performed by transformation of strain EY057 with SphI-SalI-digested EB0149 and selecting for integrants on medium lacking histidine. Integration of HIS3 at PCL1 was confirmed by PCR. Deletion of PCL2 was performed by transformation with KpnI-XbaI-digested EB0226 and selecting for integrants on medium lacking uracil. Integration of URA3 at PCL2 was confirmed by PCR. One such strain was named EY0535. PCL9 was deleted by transforming EY0535 with EB0727 that had been digested with NotI and XhoI and selecting for integrants on medium lacking tryptophan. Integration of TRP1 at PCL9 was confirmed by PCR. One such strain was named EY0552.

A deletion of the EFR3 locus marked with the LEU2 gene was generated by transforming EY0099 with MP0030 cut with SalI and BamHI and selecting for integrants on medium lacking leucine. Integration at the EFR3 locus was confirmed by PCR.

Cloning of transposon insertions:
The genomic position of transposon insertions that give rise to an Efr phenotype was determined by first using T4 DNA ligase to circularize fragments obtained by digesting genomic DNA from the Efr strains with Csp6 I. Primers that hybridize to the sense strand of the transposon just downstream of the 5' end or to the antisense strand just upstream of the transposon's first Csp6 I site were used to perform PCR with the circularized DNA fragments, and the resulting PCR products were sequenced directly. We identified the genomic sequence adjacent to the transposon end by searching the yeast genome sequence database (CHERRY et al. 2000 ).

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identification of mutants that require PHO85 for viability:
To screen for Efr mutants we constructed a strain in which expression of PHO85 is galactose dependent and identified strains that can grow only in the presence of galactose. We mutagenized the GAL1-10::PHO85 strain with transposon-containing fragments from a yeast genomic DNA library (BURNS et al. 1994 ) on galactose-containing plates. These transformants were then replica plated to SD plates. Those colonies that could not grow on dextrose plates were retained for further characterization. Among 25,000 transposon transformants, 26 displayed a galactose-dependent phenotype.

In addition to identifying mutants requiring PHO85, our primary screen would also identify mutants unable to grow on dextrose even in the presence of PHO85 (which we name Dead On Dextrose, or Dod mutants). To differentiate between these two classes of mutants, we crossed candidate Efr mutants to strains with a temperature-sensitive allele of PHO85. Eleven of the 26 mutants had a transposon-linked Efr phenotype: spores containing the transposon were dead or very sick in the presence of the GAL1-10::PHO85 allele on dextrose but grew in the presence of the pho85-9 allele at room temperature. These 11 mutants define seven complementation groups.

A temperature-sensitive PHO85 mutation is suppressed by overexpression of Pho85-associated cyclins:
To investigate which cyclins are required to perform the various functions of PHO85, we tested whether high-copy plasmids containing Pho85-associated cyclins could suppress phenotypes of a PHO85 ts allele. This idea is based on the identification of CLN1 and CLN2 as high-copy suppressors of the cdc28-4 temperature-sensitive G1 arrest (REED et al. 1989 ). We first determined whether high-copy plasmids containing PHO80, PCL1, or PCL2 suppress the ts growth phenotype of a cln1 cln2 pho85-9 strain (Fig 1A) or the ts-Pho^c phenotype of a pho85-9 strain (Fig 1B). High-copy plasmids containing PHO80 but not plasmids containing PCL1 or PCL2 suppress the ts-Pho^c phenotype of a pho85-9 strain. Since both pho85 cln1 cln2 and pcl1 pcl2 cln1 cln2 mutants are inviable, the function of PHO85 that allows cln1 cln2 strains to grow requires Pcl1- and Pcl2-containing Pho85 kinases (ESPINOZA et al. 1994 ; MEASDAY et al. 1994 ). High-copy plasmids containing PCL1, but not PCL2 or PHO80 suppress the ts growth phenotype of a cln1 cln2 pho85-9 strain. Given the ability of plasmids containing PCL1 to suppress the ts growth phenotype of a cln1 cln2 pho85-9 strain, the inability of high-copy plasmids containing PCL2 to suppress the same phenotype is surprising. This PCL2 plasmid is functional as it suppresses the ts phenotype of an mpk1 strain (MADDEN et al. 1997 , data not shown). High-copy plasmids containing each of the other seven Pho85-associated cyclins do not suppress either of these temperature-sensitive pho85-9 phenotypes (data not shown).

View larger version (70K): In this window In a new window Download PPT slide   Figure 1. Suppression of pho85-9 temperature-sensitive phenotypes by high-copy cyclin plasmids. (A) Suppression of the synthetic temperature-sensitive lethality of the pho85-9 cln1 cln2 mutant. Strain MY0205 (bf264-15D pho85-9 cln1 cln2) transformed with ARS/CEN URA3 (EB0009), ARS/CEN PHO85 (EB0327), 2µ PHO80 (MP0115), 2µ PCL1 (MP0120), or 2µ PCL2 (MP0121) was streaked on SD-Ura plates and placed at 30° for 3 days. (B) Suppression of the temperature-sensitive constitutive acid-phosphatase expression phenotype of the pho85-9 mutant. Strain MY0158 (K699 pho85-9; transformed with the same plasmids) was placed on SD-Ura and placed at 37° overnight and then stained for acid-phosphatase activity.

Characterization of efr pho85-9 strains:
We next tested whether high-copy plasmids containing the Pho85-associated cyclins suppress the temperature-sensitive growth phenotypes of different efr pho85-9 strains (Fig 2). High-copy plasmids containing PCL1 suppress the ts growth phenotype of efr1 pho85-9 and efr8 pho85-9 strains, whereas plasmids containing PHO80 suppress the efr3 pho85-9 strain. None of the Pho85-cyclin- containing plasmids suppress the ts phenotype of efr2 pho85-9, efr6 pho85-9, or efr7 pho85-9 strains. The ts phenotype of these strains is, however, pho85-9 dependent as it is complemented by PHO85 (Fig 2 and data not shown). We could not perform this analysis on the efr5 pho85-9 strain as it is very slow growing.

View larger version (105K): In this window In a new window Download PPT slide   Figure 2. Suppression of the synthetic temperature-sensitive lethality of various efr pho85-9 double mutants by high-copy cyclin plasmids. Strains MY0276 (efr1 pho85-9), MY0282 (efr2 pho85-9), MY0176 (efr3 pho85-9), and MY0321 (efr8 pho85-9) transformed with high-copy plasmids containing each of the Pho85-associated cyclins or low-copy plasmids containing PHO85 (EB0327) or URA3 (EB0- 009) were diluted from an OD600 = 0.3 by threefold serial dilutions and plated on SD-Ura plates. Plates were placed at the indicated temperature for 3 days.

The positions of the mutations responsible for the Efr phenotype are indicated in Table 3. The class of mutants that is not suppressed by cyclin overexpression contains mutations in a variety of genes involved in transcriptional regulation: SPT7, HFI1 (ADA1), SRB5, and GCR1. With the exception of GCR1, which plays a specific role in regulating carbohydrate metabolism gene expression, the other genes encode general regulators of RNA polymerase II transcription. The PCL1-suppressible class of Efr mutants is due to mutations in BEM2 and BCK1—genes involved in morphogenesis and cell-wall biosynthesis. Mutations in a previously uncharacterized open reading frame, YMR212C, that we have named EFR3 accounts for the PHO80-suppressible mutant. Plasmids containing EFR3 complement the efr3 mutation. efr3 mutants also grow poorly on media containing acetate as a carbon source and efr3 diploids fail to sporulate (data not shown). High-copy plasmids containing PHO80 suppress the efr3 pho85-9 strain, but an efr3 pho80 strain exhibits no growth defects (data not shown). efr3 pcl1 pcl2 pcl9 strains also grow at a rate similar to that of wild type (data not shown).

Morphogenesis-related Efr mutants:
Since Bck1 is the mitogen-activated protein kinase kinase kinase (MAP-KKK) responsible for activation of Mpk1, and bck1 mutants have an Efr phenotype, we examined synthetic interactions between mpk1 and pho85. mpk1 pho85 strains are inviable (this result was subsequently reported in HUANG et al. 1999 ) and mpk1 pho85-9 strains exhibit a synthetic temperature-sensitive growth phenotype (data not shown).

Since high-copy PCL1 plasmids suppress the ts lethality of bem2 pho85-9 and bck1 pho85-9 strains, we wondered if Pcl1-Pho85 complexes provide the PHO85 function required for viability in these mutant backgrounds. bem2 pcl1 strains are inviable (data not shown). mpk1 pcl1 pcl2 and bck1 pcl1 pcl2 strains are inviable when dissected onto standard medium, but are viable on medium containing 1 M sorbitol (Fig 3). This phenotype is different from the bem2 pcl1 phenotype because mpk1 pcl1 or bck1 pcl1 strains have growth rates similar to those of mpk1 or bck1 strains. Furthermore, sorbitol does not rescue the inviability of bem2 pcl1 strains (data not shown). Removing PCL9, a cyclin highly homologous to PCL2, does not exacerbate the sorbitol-requiring phenotype of mpk1 pcl1 pcl2 or bck1 pcl1 pcl2 strains nor do bck1 pcl1 pcl9 or bck1 pcl2 pcl9 strains exhibit any growth defects (data not shown). We also tested for genetic interactions between spt7, gcr1, srb5, or hfi1 and pcl1 pcl2 pcl9. All of these efr pcl1 pcl2 pcl9 strains grow similarly to the corresponding efr strain (data not shown), suggesting that the cause of the synthetic lethality in these Efr mutants is distinct from that of the PCL1-suppressible Efr mutants.

View larger version (61K): In this window In a new window Download PPT slide   Figure 3. Sorbitol-suppressible synthetic-lethal phenotype of the pcl1 pcl2 mpk1 mutant. Progeny of a cross between MY0263 (mpk1) and EY0535 (pcl1 pcl2) or of a cross between MY0296 (efr8) and EY0552 (pcl1 pcl2 pcl9) were diluted from an OD600 = 0.3 by threefold serial dilutions and plated on SD-complete plates or SD-complete plates that had been supplemented with 1 M sorbitol. Plates were placed at 30° for 2 days.

Suppression of cln1 cln2 pcl1 pcl2 lethality by sorbitol-containing medium:
Our screen identified two types of efr pho85-9 mutants that are suppressed by high-copy PCL1 plasmids. bem2 pcl1 mutants are inviable while bck1 pcl1 pcl2 mutants are sorbitol requiring. Since high-copy PCL1 also suppresses the cln1 cln2 pho85-9 strain (Fig 1), we wondered if the lethality of a cln1 cln2 pcl1 pcl2 strain (ESPINOZA et al. 1994 ; MEASDAY et al. 1994 ) would also be suppressed by sorbitol. We compared growth of a GAL1-10::CLN1 cln2 strain with GAL1-10::CLN1 cln2 cln3, GAL1-10::CLN1 cln2 pcl1 pcl2, and GAL1-10::CLN1 cln2 cln3 pcl1 pcl2 strains on YEPG, YEPD, and YEPD + 1 M sorbitol. While GAL1-10::CLN1 cln2 pcl1 pcl2 strains are unable to grow on YEPD plates, they grow on YEPD plates containing sorbitol (Fig 4A). In contrast, the GAL1-10::CLN1 cln2 cln3 strain is unable to grow on YEPD with or without sorbitol. Like the GAL1-10::CLN1 cln2 strain, our WT strain grows more slowly on YEPD + 1 M sorbitol than on YEPD (data not shown). We had found that the PKC1 R398P gain-of-function allele suppresses the ts phenotypes of mpk1 and swi4 strains (data not shown). This plasmid also suppresses the ts phenotype of a cln1 cln2 pho85-9 strain (Fig 4B). These data indicate that the synthetic lethality of the cln1 cln2 pcl1 pcl2 strain has many similarities with the synthetic lethality of mpk1 pcl1 pcl2 and bck1 pcl1 pcl2 strains.

View larger version (38K): In this window In a new window Download PPT slide   Figure 4. Comparison of the phenotypes of the pcl1 pcl2 cln1 cln2 and the pcl1 pcl2 mpk1 mutants. (A) The lethality of pcl1 pcl2 cln1 cln2 is suppressed by 1 M sorbitol. The following strains were streaked on the indicated plates and placed at 37° for 2 days: cln2 GAL1-10::CLN1 (EY0233); cln2 GAL1-10::CLN1 pcl1 pcl2 (EY0409); cln2 GAL1-10::CLN1 cln3 (MY0261); cln2 GAL1-10::CLN1 cln3 pcl1 pcl2 (MY0260). (B) The temperature-sensitive lethality of the pho85-9 cln1 cln2 mutant is suppressed by a PKC1 gain-of-function allele. Strain MY0205 (cln1 cln2 pho85-9) transformed with plasmids containing PKC-R398P (MP0143), 2µ PCL1 (MP0120), 2µ CLN2 (EB0459), ARS/CEN PHO85 (EB0327), or ARS/CEN URA3 (EB0009) was streaked on SD-Ura plates and placed at 30° for 3 days.

Other morphogenesis-related phenotypes of pho85 mutants:
Because pcl1 and pcl1 pcl2 are synthetically lethal with mutations in genes involved in morphogenesis, and some of these phenotypes can be reversed by osmotic stabilization, we wondered if mutations in PCL1 and PCL2 would show interactions with mutations in other morphogenesis genes. We were particularly interested in looking for interactions between PHO85 and the p21-activated (PAK) kinases Ste20, Cla4, or Skm1 because mammalian CDK5, which is involved in regulating PAK kinase activity in neurons (NIKOLIC et al. 1998 ), can complement a pho85 strain (HUANG et al. 1999 ; NISHIZAWA et al. 1999). pho85 ste20 strains are temperature sensitive (Fig 5). cla4 pcl1 pcl2 strains are inviable and are not cured by osmotic stabilization with sorbitol (data not shown). The cla4 pcl1 pcl2 inviability is therefore dissimilar from the sorbitol-requiring phenotype of strains carrying the mpk1, bck1, or cln1 cln2 mutations in combination with pcl1 pcl2.

View larger version (85K): In this window In a new window Download PPT slide   Figure 5. The ste20 pho85 mutant is synthetically temperature sensitive. Progeny of a cross between MY0363 (ste20) and MY0147 (pho85) were diluted from an OD600 = 0.3 by threefold serial dilutions and plated on YEPD plates that were then placed at the indicated temperature for 2 days.

As sorbitol suppresses the inviability of cln1 cln2 pcl1 pcl2, mpk1 pcl1 pcl2, and bck1 pcl1 pcl2 strains, cln1 cln2, mpk1, and bck1 might have similar defects. Since CLA4 is required for viability in the absence of CLN1 and CLN2 (CVRCKOVA et al. 1995 ), we wondered if mpk1 strains also require CLA4. Both cla4 mpk1 and cla4 bem2 strains are inviable (data not shown). These data demonstrate another similarity between the phenotypes of mpk1 and cln1 cln2 mutants and suggest that despite the differences in synthetic lethality with pcl1 pcl2, the defects of a bem2 mutant are related to those of mpk1 and cln1 cln2 mutants.

Since CDC42 is a GTPase that performs an essential function involved in actin polarization and is also required for the activation of Ste20 and Cla4 (ADAMS et al. 1990 ; PETER et al. 1996 ; BENTON et al. 1997 ), we tested whether mutations in PCL1 or PCL2 would alter the permissive temperature of strains carrying a ts allele of CDC42. cdc42-1 pcl1 pcl2 strains are inviable at all temperatures and are not rescued by sorbitol (data not shown).

The synthetic lethality between cdc42-1 and pcl1 pcl2 could have many causes. Expression of PCL1 from the GPD1 promoter raises the permissive temperature of the cdc42-1 strain (Fig 6A), suggesting that overexpression of PCL1 either promotes a process that substitutes for CDC42 function or promotes Cdc42 activity. In contrast, overexpression of PHO80 lowers the permissive temperature of a cdc42-1 strain, suggesting that PHO80 is not able to perform the function PCL1 provides. It also suggests that increasing Pho80 activity interferes with the Pcl1-mediated functions of Pho85, perhaps by decreasing the amount of free Pho85 available to associate with Pcl1.

View larger version (49K): In this window In a new window Download PPT slide   Figure 6. Other phenotypic consequences of altering Pho85 activity. (A) The effect of Pho85-associated cyclin overexpression of the permissive temperature of cdc42-1 strains. Strain MY361 (MATa cdc42-1) transformed with plasmids containing GPD1::PCL1 (EB0128), GPD1::PHO80 (EB0049), or URA3 (EB0009) were diluted from an OD600 = 0.3 by threefold serial dilutions and plated on YEPD plates and placed at the indicated temperature for 3 days. (B) pho85 strains mate as efficiently as WT strains. Strains EY0140 (MATa pho85), EY057 (MATa WT), MY0277 (MATa bem2), and MY0296 (MATa bck1) were diluted from an OD600 = 0.3 by threefold serial dilutions and plated on a lawn of 2.0 OD600 of EY0478 (MAT far1-c Lys-) that had been spread on YEPD plates and placed at 30° for 6 hr. These plates were then replica plated to SD-Min plates (lacking amino acid supplements) and placed at 30° for 2 days.

Since many of the processes required for morphogenesis during the cell cycle are also required for the specialized morphogenetic processes that occur during mating, many mutants with defects in morphogenesis and cell-wall biosynthesis also have mating defects. To look for subtle mating defects, we compared the ability of WT, pho85, bem2, and bck1 mutants to mate to a far1-c strain (VALTZ et al. 1995 ). WT and pho85 strains mate with similar efficiency whereas bem2 and bck1 mutants show a reduced mating ability (Fig 6B). This suggests that pho85 mutants are unlikely to have a defect in the physical processes of morphogenesis or cell-wall biosynthesis.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Our screen for mutants that require PHO85 for viability was designed to survey the functions Pho85 performs in association with its 10 cyclin subunits. We identified mutations in seven genes. Because we found only one allele of many of these genes, and found additional synthetic-lethal interactions involving PHO85 by directed tests, it is apparent that our screen was not saturating. A similar screen for Efr mutants using ultraviolet light as a mutagen uncovered 18 recessive mutations in 17 complementation groups, suggesting that the actual number of loci that can be mutated to give rise to an Efr phenotype is quite large. Two of these UV-induced mutants contained mutations in EFR3.

Our strategy for identifying the cyclins responsible for different functions of Pho85 involved determining if high-copy plasmids carrying the genes for the Pho85-associated cyclins could suppress various phenotypes of a temperature-sensitive allele of PHO85. This strategy was only partially successful as only PHO80- and PCL1-containing plasmids had activity in this assay, and some efr pho85-9 mutants were not suppressed by either plasmid. Mutations in activators of transcription account for the largest class of non-cyclin-suppressible mutants. It is striking that our screen identified mutations in two different components of the SAGA histone acetylase complex (SPT7 and HFI1/ADA1; EBERHARTER et al. 1998 ). We do not know if these synthetic-lethal interactions indicate a role for PHO85 as a general regulator of transcription—like the yeast CDKs Srb10, Kin28, and Ctk1 (LIAO et al. 1995 ; STERNER et al. 1995 ; VALAY et al. 1995 )—or if they cause a defect in expression of a particular gene with which PHO85 is also synthetically lethal.

Several different classes of Efr mutants require PCL1 and/or PCL2. One is bem2, which is synthetically lethal with pcl1. cla4 and cdc42 represent a type of Efr mutant that is synthetically lethal with pcl1 pcl2 under all conditions tested. bck1, mpk1, and cln1 cln2 form a third class of mutants that are sorbitol requiring when combined with pcl1 pcl2. The inviability of bem2 pcl1 strains is the first example of a pcl1 phenotype that does not depend on also removing PCL2. Several models can account for the requirement for PCL1 vs. PCL1 or PCL2 (Fig 7). PCL1 and PCL2 might perform distinct functions: mutants such as bem2 are sensitive to the loss of PCL1-mediated functions while mutants such as mpk1 are inviable only when both PCL1- and PCL2-mediated processes are perturbed. A similar model is that a subset of PCL1 functions are distinct from those performed by PCL1 and PCL2. Both models explain the inviability of bem2 pcl1 and mutants like mpk1 pcl1 pcl2 as being caused by different defects in the pcl1 as compared to the pcl1 pcl2 mutants. A third possibility is that PCL1 and PCL2 perform identical functions and that genes like BEM2 and MPK1 require different thresholds of this function for viability. We cautiously favor the possibility that PCL1 and PCL2 perform identical functions because it allows us to arrange the various Efr mutants along a phenotypic spectrum from more to less dependence on PCL1/PCL2 and from this hypothesis make a prediction about the site of PCL1/PCL2 action.

View larger version (5K): In this window In a new window Download PPT slide   Figure 7. Models explaining the functional relationships between Pcl1- and Pcl2-associated Pho85 kinase activities. (A) PCL1 and PCL2 perform nonoverlapping functions. (B) PCL1 participates in multiple processes. Some of these are independent of PCL2 (labeled X) and some are shared with PCL2 (labeled Y). (C) The functions of PCL1 and PCL2 overlap.

If PCL1 and PCL2 perform the same function, the requirement of BEM2 mutants for PHO85 is most severe because bem2 mutants are inviable when just the PCL1 fraction of Pho85 function is removed and the other mutants are inviable only when the PCL1- and PCL2-mediated functions are both removed. Similarly, the inviability of mpk1 pcl1 pcl2 strains can be rescued by sorbitol, while the inviability of the cla4 pcl1 pcl2 strain cannot. This could indicate that cla4 strains are more dependent on Pho85 function than are mpk1 strains. If these suppositions are correct we can arrange these Efr mutants along a phenotypic spectrum: bem2 mutants have the strongest Efr phenotype, followed by cla4 and cdc42-1, and then followed by mpk1, bck1, and cln1 cln2.

A model to explain the defect that accounts for this spectrum is that PCL1 and PCL2 function together with the PCL1- and PCL2-requiring EFR genes to positively regulate Cdc42 activity (Fig 8). As the novel elements of this model are based on circumstantial evidence presented in this article, we have summarized our principal findings in Table 4. One aspect of our model proposes that PKC1 positively regulates CDC42 activity. A functional connection between BEM2 and CDC42 is supported by the observation that GIC1 and GIC2 are CDC42 effectors and are also high-copy suppressors of the bem2 mutant (CHEN et al. 1997 ). PKC1 may function in between BEM2 and CDC42 as in vitro studies show Bem2 displays strong GAP activity toward Rho1 (KIM et al. 1994 ; PETERSON et al. 1994 ), a GTPase that activates PKC1 (NONAKA et al. 1995 ; DRGONOVA et al. 1996 ; KAMADA et al. 1996 ). We have drawn a distinction between the functions of the Swi4-dependent genes CLN1, CLN2, PCL1, and PCL2 because, of these, only PCL1 and PCL2 can high-copy suppress the temperature-sensitive phenotype of mpk1 strains (MADDEN et al. 1997 ). We propose that PCL1 and PCL2 high-copy suppress the mpk1 mutant as a result of increasing Cdc42 activity. We also hypothesize that CLN1 and CLN2 are positive regulators of PKC1 function because gain-of-function alleles in PKC1 suppress the temperature-sensitive phenotypes of swi4 and mpk1 strains (foreshadowed by GRAY et al. 1997 ) and also suppress the synthetic temperature sensitivity of a cln1 cln2 pho85-9 strain, suggesting that PKC1 functions downstream of each of these genes. As PKC1 is an activator of the MPK1 MAP kinase pathway at the level of either BCK1 or MKK1 MKK2 (HUANG and SYMINGTON 1995 ), overexpression of CLN1 and CLN2 fails to suppress the mpk1 phenotype because, in this mutant, increased Pkc1 function is unable to increase Swi4-dependent transcription.

View larger version (30K): In this window In a new window Download PPT slide   Figure 8. A model explaining the functional relationships between Pcl1- and Pcl2-associated Pho85 kinase activities and some of the Efr mutants identified in our screen. This model hypothesizes that the underlying cause of many of the synthetic-lethal interactions is a defect in CDC42 activity.

The suppression of the cln1 cln2 pcl1 pcl2 G1 arrest by sorbitol suggests that these strains do not have an absolute defect in G1 progression but rather have a defect in a morphogenesis or cell-wall biogenesis related function that results in a G1 arrest. This phenotype is distinct from the non-sorbitol-suppressible G1 arrest of cln1 cln2 cln3 strains and the unbudded G2-phase terminal phenotype of cln1 cln2 bud2 strains (CVRCKOVA and NASMYTH 1993 ). Perhaps the morphogenesis defect in pcl1 pcl2 cln1 cln2 strains occurs upstream of a checkpoint that inhibits passage through START. Such a checkpoint might operate through stabilization of Sic1, an S-phase inhibitor, as pho85 sic1 cells are temperature sensitive for growth (AERNE et al. 1998 ) and Sic1 is more stable in pho85 strains (NISHIZAWA et al. 1998 ). This suggests that entry into S phase is delayed in the pho85 mutant and that removing this delay has deleterious consequences. Consistent with this hypothesis, while the synthetic lethality of cln1 cln2 cln3 is suppressed by deletion of SIC1 (SCHNEIDER et al. 1996 ), the lethality of pcl1 pcl2 cln1 cln2 is not (M. TYERS and B. ANDREWS, personal communication).

In contrast to mutants with defects in morphogenesis and cell-wall biogenesis, pho85 strains mate efficiently. Furthermore, we have been unable to detect any defects in morphogenesis-related processes in either pcl1 pcl2 or pho85 mutants. Perhaps Pcl1- and Pcl2-associated Pho85 kinases regulate morphogenetic events specifically during G1—or perhaps a specialized G1.

Our data help to clarify the cause of inviability in cln1 cln2 pcl1 pcl2 strains, as they suggest that PCL1 and PCL2 promote Cdc42 activity or promote a process that can substitute for Cdc42 rather than participating in the START step of G1 itself. We report a pcl1 phenotype that does not also depend on removing PCL2. It will be interesting to determine if this reflects the ability of Pcl1 and Pcl2 to phosphorylate different substrates or if instead it reflects differential requirements for phosphorylation of a common substrate. The challenge of understanding how cyclin binding confers functional specificity on Pho85 requires us to identify the in vivo substrates of these different Pho85-containing kinases.

	ACKNOWLEDGMENTS

We thank Andrew Murray, Ira Herskowitz, Sean O'Rourke, and the members of the O'Shea lab for their comments on this manuscript and other useful discussions. We thank Brenda Andrews, Fred Cross, Hernan Espinoza, Sue Biggins, Aaron Straight, Anita Sil, Sean O'Rourke, Linda Huang, and Jennifer Philips for kindly providing us with strains and plasmids. M.E.L. is a Howard Hughes Medical Institute Predoctoral Fellow. This work was supported by a National Science Foundation Presidential Faculty Fellowship (E.K.O.).

Manuscript received April 14, 2000; Accepted for publication September 15, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ADAMS, A. E., D. I. JOHNSON, R. M. LONGNECKER, B. F. SLOAT, and J. R. PRINGLE, 1990  CDC42 and CDC43, two additional genes involved in budding and the establishment of cell polarity in the yeast Saccharomyces cerevisiae.. J. Cell Biol. 111:131-142[Abstract/Free Full Text].

AERNE, B. L., A. L. JOHNSON, J. H. TOYN, and L. H. JOHNSTON, 1998  Swi5 controls a novel wave of cyclin synthesis in late mitosis. Mol. Biol. Cell 9:945-956[Abstract/Free Full Text].

ANDREWS, B. and V. MEASDAY, 1998  The cyclin family of budding yeast: abundant use of a good idea. Trends Genet. 14:66-72[Medline].

BENTON, B. K., A. TINKELENBERG, I. GONZALEZ, and F. R. CROSS, 1997  Cla4p, a Saccharomyces cerevisiae Cdc42p-activated kinase involved in cytokinesis, is activated at mitosis. Mol. Cell. Biol. 17:5067-5076[Abstract].

BURNS, N., B. GRIMWADE, P. B. ROSS-MACDONALD, E. Y. CHOI, and K. FINBERG et al., 1994  Large-scale analysis of gene expression, protein localization, and gene disruption in Saccharomyces cerevisiae.. Genes Dev. 8:1087-1105[Abstract/Free Full Text].

CHEN, G. C., Y. J. KIM, and C. S. CHAN, 1997  The Cdc42 GTPase-associated proteins Gic1 and Gic2 are required for polarized cell growth in Saccharomyces cerevisiae.. Genes Dev. 11:2958-2971[Abstract/Free Full Text].

CHERRY, J. M., C. BALL, K. DOLINSKI, S. DWIGHT, M. HARRIS et al., 2000 Saccharomyces Genome Database (http://genome-www.stanford.edu/Saccharomyces/).

CVRCKOVA, F. and K. NASMYTH, 1993  Yeast G1 cyclins CLN1 and CLN2 and a GAP-like protein have a role in bud formation. EMBO J. 12:5277-5286[Medline].

CVRCKOVA, F., C. DE VIRGILIO, E. MANSER, J. R. PRINGLE, and K. NASMYTH, 1995  Ste20-like protein kinases are required for normal localization of cell growth and for cytokinesis in budding yeast. Genes Dev. 9:1817-1830[Abstract/Free Full Text].

DRGONOVA, J., T. DRGON, K. TANAKA, R. KOLLAR, and G. C. CHEN et al., 1996  Rho1p, a yeast protein at the interface between cell polarization and morphogenesis. Science 272:277-279[Abstract].

EBERHARTER, A., S. JOHN, P. A. GRANT, R. T. UTLEY, and J. L. WORKMAN, 1998  Identification and analysis of yeast nucleosomal histone acetyltransferase complexes. Methods 15:315-321[Medline].

ESPINOZA, F. H., J. OGAS, I. HERSKOWITZ, and D. O. MORGAN, 1994  Cell cycle control by a complex of the cyclin HCS26 (PCL1) and the kinase PHO85.. Science 266:1388-1391[Abstract/Free Full Text].

GILLIQUET, V. and G. BERBEN, 1993  Positive and negative regulators of the Saccharomyces cerevisiae `PHO system' participate in several cell functions. FEMS Microbiol. Lett. 108:333-339[Medline].

GRAY, J. V., J. P. OGAS, Y. KAMADA, M. STONE, and D. E. LEVIN et al., 1997  A role for the Pkc1 MAP kinase pathway of Saccharomyces cerevisiae in bud emergence and identification of a putative upstream regulator. EMBO J. 16:4924-4937[Medline].

GUTHRIE, C., and G. R. FINK (Editors), 1991 Guide to Yeast Genetics and Molecular Biology, Vol. 194. Academic Press, San Diego.

HUANG, D., I. FARKAS, and P. J. ROACH, 1996  Pho85p, a cyclin-dependent protein kinase, and the Snf1p protein kinase act antagonistically to control glycogen accumulation in Saccharomyces cerevisiae.. Mol. Cell. Biol. 16:4357-4365[Abstract].

HUANG, D., G. PATRICK, J. MOFFAT, L. H. TSAI, and B. ANDREWS, 1999  Mammalian Cdk5 is a functional homologue of the budding yeast Pho85 cyclin-dependent protein kinase. Proc. Natl. Acad. Sci. USA 96:14445-14450[Abstract/Free Full Text].

HUANG, K. N. and L. S. SYMINGTON, 1995  Suppressors of a Saccharomyces cerevisiae pkc1 mutation identify alleles of the phosphatase gene PTC1 and of a novel gene encoding a putative basic leucine zipper protein. Genetics 141:1275-1285[Abstract].

JONES, J. S. and L. PRAKASH, 1990  Yeast Saccharomyces cerevisiae selectable markers in pUC18 polylinkers. Yeast 6:363-366[Medline].

KAMADA, Y., H. QADOTA, C. P. PYTHON, Y. ANRAKU, and Y. OHYA et al., 1996  Activation of yeast protein kinase C by Rho1 GTPase. J. Biol. Chem. 271:9193-9196[Abstract/Free Full Text].

KIM, Y. J., L. FRANCISCO, G. C. CHEN, E. MARCOTTE, and C. S. CHAN, 1994  Control of cellular morphogenesis by the Ip12/Bem2 GTPase-activating protein: possible role of protein phosphorylation. J. Cell Biol. 127:1381-1394[Abstract/Free Full Text].

LEE, J., K. COLWILL, V. ANELIUNAS, C. TENNYSON, and L. MOORE et al., 1998  Interaction of yeast Rvs167 and Pho85 cyclin-dependent kinase complexes may link the cell cycle to the actin cytoskeleton. Curr. Biol. 8:1310-1321[Medline].

LENBURG, M. E. and E. K. O'SHEA, 1996  Signaling phosphate starvation. Trends Biochem. Sci. 21:383-387[Medline].

LIAO, S. M., J. ZHANG, D. A. JEFFERY, A. J. KOLESKE, and C. M. THOMPSON et al., 1995  A kinase-cyclin pair in the RNA polymerase II holoenzyme. Nature 374:193-196[Medline].

MADDEN, K., Y. J. SHEU, K. BAETZ, B. ANDREWS, and M. SNYDER, 1997  SBF cell cycle regulator as a target of the yeast PKC-MAP kinase pathway. Science 275:1781-1784[Abstract/Free Full Text].

MEASDAY, V., L. MOORE, J. OGAS, M. TYERS, and B. ANDREWS, 1994  The PCL2 (ORFD)-PHO85 cyclin-dependent kinase complex: a cell cycle regulator in yeast. Science 266:1391-1395[Abstract/Free Full Text].

MEASDAY, V., L. MOORE, R. RETNAKARAN, J. LEE, and M. DONOVIEL et al., 1997  A family of cyclin-like proteins that interact with the Pho85 cyclin-dependent kinase. Mol. Cell. Biol. 17:1212-1223[Abstract].

NASMYTH, K. A. and K. TATCHELL, 1980  The structure of transposable yeast mating type loci. Cell 19:753-764[Medline].

NIKOLIC, M., M. M. CHOU, W. LU, B. J. MAYER, and L. H. TSAI, 1998  The p35/Cdk5 kinase is a neuron-specific Rac effector that inhibits Pak1 activity. Nature 395:194-198[Medline].

NISHIZAWA, M., M. KAWASUMI, M. FUJINO, and A. TOH-E, 1998  Phosphorylation of sic1, a cyclin-dependent kinase (Cdk) inhibitor, by Cdk including Pho85 kinase is required for its prompt degradation. Mol. Biol. Cell 9:2393-2405[Abstract/Free Full Text].

NISHIZAWA, M., Y. KANAYA, and A. TOH-E, 1999  Mouse cyclin-dependent kinase (Cdk) 5 is a functional homologue of a yeast Cdk, pho85 kinase. J. Biol. Chem. 274:33859-33862[Abstract/Free Full Text].

NONAKA, H., K. TANAKA, H. HIRANO, T. FUJIWARA, and H. KOHNO et al., 1995  A downstream target of RHO1 small GTP-binding protein is PKC1, a homolog of protein kinase C, which leads to activation of the MAP kinase cascade in Saccharomyces cerevisiae.. EMBO J. 14:5931-5938[Medline].

OGAS, J., B. J. ANDREWS, and I. HERSKOWITZ, 1991  Transcriptional activation of CLN1, CLN2, and a putative new G1 cyclin (HCS26) by SWI4, a positive regulator of G1-specific transcription. Cell 66:1015-1026[Medline].

PETER, M., A. M. NEIMAN, H. O. PARK, M. VAN LOHUIZEN, and I. HERSKOWITZ, 1996  Functional analysis of the interaction between the small GTP binding protein Cdc42 and the Ste20 protein kinase in yeast. EMBO J. 15:7046-7059[Medline].

PETERSON, J., Y. ZHENG, L. BENDER, A. MYERS, and R. CERIONE et al., 1994  Interactions between the bud emergence proteins Bem1p and Bem2p and Rho-type GTPases in yeast. J. Cell Biol. 127:1395-1406[Abstract/Free Full Text].

REED, S. I., J. A. HADWIGER, H. E. RICHARDSON, and C. WITTENBERG, 1989  Analysis of the Cdc28 protein kinase complex by dosage suppression. J. Cell Sci. Suppl. 12:29-37[Medline].

SCHNEIDER, B. L., Q. H. YANG, and A. B. FUTCHER, 1996  Linkage of replication to start by the Cdk inhibitor Sic1. Mol. Biol. Cell 9:945-956.

SHERR, C. J., 1995  D-type cyclins. Trends Biochem. Sci. 20:187-190[Medline].

SIKORSKI, R. S. and P. HIETER, 1989  A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122:19-27[Abstract/Free Full Text].

STERNER, D. E., J. M. LEE, S. E. HARDIN, and A. L. GREENLEAF, 1995  The yeast carboxyl-terminal repeat domain kinase CTDK-I is a divergent cyclin-cyclin-dependent kinase complex. Mol. Cell. Biol. 15:5716-5724[Abstract].

TENNYSON, C. N., J. LEE, and B. J. ANDREWS, 1998  A role for the Pcl9-Pho85 cyclin-cdk complex at the M/G1 boundary in Saccharomyces cerevisiae.. Mol. Microbiol. 28:69-79[Medline].

TIMBLIN, B. K., K. TATCHELL, and L. W. BERGMAN, 1996  Deletion of the gene encoding the cyclin-dependent protein kinase Pho85 alters glycogen metabolism in Saccharomyces cerevisiae.. Genetics 143:57-66[Abstract].

TOH-E, A. and T. SHIMAUCHI, 1986  Cloning and sequencing of the PHO80 gene and CEN15 of Saccharomyces cerevisiae. Yeast 2:129-139[Medline].

VALAY, J. G., M. SIMON, M. F. DUBOIS, O. BENSAUDE, and C. FACCA et al., 1995  The KIN28 gene is required both for RNA polymerase II mediated transcription and phosphorylation of the Rpb1p CTD. J. Mol. Biol. 249:535-544[Medline].

VALTZ, N., M. PETER, and I. HERSKOWITZ, 1995  FAR1 is required for oriented polarization of yeast cells in response to mating pheromones. J. Cell Biol. 131:863-873[Abstract/Free Full Text].

VOLKERT, F. C. and J. R. BROACH, 1986  Site-specific recombination promotes plasmid amplification in yeast. Cell 46:541-550[Medline].

WILSON, W. A., A. M. MAHRENHOLZ, and P. J. ROACH, 1999  Substrate targeting of the yeast cyclin-dependent kinase Pho85p by the cyclin Pcl10p. Mol. Cell. Biol. 19:7020-7030[Abstract/Free Full Text].

This article has been cited by other articles:

				A. D. Mozdy, E. R. Podell, and T. R. Cech Multiple Yeast Genes, Including Paf1 Complex Genes, Affect Telomere Length via Telomerase RNA Abundance Mol. Cell. Biol., June 15, 2008; 28(12): 4152 - 4161. [Abstract] [Full Text] [PDF]

				S. Castillo-Lluva, I. Alvarez-Tabares, I. Weber, G. Steinberg, and J. Perez-Martin Sustained cell polarity and virulence in the phytopathogenic fungus Ustilago maydis depends on an essential cyclin-dependent kinase from the Cdk5/Pho85 family J. Cell Sci., May 1, 2007; 120(9): 1584 - 1595. [Abstract] [Full Text] [PDF]

				M. Gandhi, B. L. Goode, and C. S. M. Chan Four Novel Suppressors of gic1 gic2 and Their Roles in Cytokinesis and Polarized Cell Growth in Saccharomyces cerevisiae Genetics, October 1, 2006; 174(2): 665 - 678. [Abstract] [Full Text] [PDF]