Control of Mitotic Events by the Cdc42 Gtpase, the Clb2 Cyclin and a Member of the PAK Kinase Family Hendri Tjandra, Jennifer Compton and Douglas Kellogg

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Research Paper 991

Control of mitotic events by the Cdc42 GTPase, the Clb2 cyclin and a member of the PAK kinase family Hendri Tjandra, Jennifer Compton and Douglas Kellogg

Background: Cyclins and cyclin-dependent kinases induce and coordinate the Address: Sinsheimer Laboratories, Department of events of the cell cycle, although the mechanisms by which they do so remain Biology, University of California, Santa Cruz, California 95064, USA. largely unknown. In budding yeast, a pathway used by the Clb2 cyclin to control bud growth during mitosis provides a good model system in which to understand Correspondence: Douglas Kellogg how cyclin-dependent kinases control cell-cycle events. In this pathway, Clb2 E-mail: [email protected] initiates a series of events that lead to the mitosis-specific activation of the Gin4 Received: 14 May 1998 in vivo protein kinase. A protein called Nap1 is required for the activation of Revised: 24 July 1998 Gin4, and is able to bind to both Gin4 and Clb2. We have used a simple genetic Accepted: 7 August 1998 screen to identify additional proteins that function in this pathway. Published: 24 August 1998

Results: We have found that the Cdc42 GTPase and a member of the PAK Current Biology 1998, 8:991–1000 kinase family called Cla4 both function in the pathway used by Clb2 to control http://biomednet.com/elecref/0960982200800991 bud growth during mitosis. Cdc42 and Cla4 interact genetically with Gin4 and Nap1, and both are required in vivo for the mitosis-specific activation of the © Current Biology Ltd ISSN 0960-9822 Gin4 kinase. Furthermore, Cla4 undergoes a dramatic hyperphosphorylation in response to the combined activity of Nap1, the Clb2—Cdc28 kinase complex, and the GTP-bound form of Cdc42. Evidence is presented which suggests that the hyperphosphorylated form of Cla4 is responsible for relaying the signal to activate Gin4.

Conclusions: Previous studies have suggested that cyclin-dependent kinases control the cell cycle by directly phosphorylating proteins involved in specific events, such as nuclear lamins, microtubule-associated proteins and histones. In contrast, our results demonstrate that the Clb2—Cdc28 cyclin-dependent kinase complex controls specific cell-cycle events through a pathway that involves a GTPase and at least two different kinases. This suggests that cyclin- dependent kinases may control many cell-cycle events through GTPase-linked signaling pathways that resemble the intricate signaling pathways known to control many other cellular events.

Background the cyclins function to determine the cell-cycle events The events of cell division are controlled by a family of induced by specific cyclin-dependent kinase complexes. proteins called cyclin-dependent kinases, which are acti- In previous work we reasoned that in order for cyclins to vated at specific stages of the cell cycle when they bind confer this kind of specificity they must bind to specific to members of the cyclin family of proteins [1,2]. proteins involved in the control of cell-cycle events, and Although considerable progress has been made towards we used affinity chromatography to search for such pro- understanding the regulation of cyclin-dependent teins [3]. This led to the identification of a highly con- kinases, we still know little about the molecular pathways served protein called Nap1 which binds specifically to used by these kinases to induce the actual events of cell mitotic cyclins in organisms as divergent as vertebrates division. Simple organisms use a single cyclin-dependent and yeast. In budding yeast, Nap1 binds to the mitotic kinase to induce the events of both interphase and cyclin Clb2 and is required for the ability of Clb2 to mitosis, demonstrating that the same cyclin-dependent induce specific mitotic events. One of these events is a kinase can induce different events when activated at dif- switch in the pattern of bud growth that occurs as cells ferent times during the cell cycle. We do not yet under- enter mitosis [4]. A newly formed bud initially grows only stand the molecular mechanisms that make such at its tip; however, during mitosis the Clb2–Cdc28 kinase specificity possible. complex causes bud growth to occur over the entire bud surface, leading to the formation of a round bud [5,6]. The fact that cyclin-dependent kinases are bound to dif- Without Nap1, the Clb2–Cdc28 kinase complex is unable ferent cyclins at each stage of the cell cycle suggests that to induce this switch in the pattern of bud growth and the 992 Current Biology, Vol 8 No 18

new bud therefore continues to grow at its tip during Figure 1 mitosis, giving rise to highly elongated buds [4].

The pathway used by Clb2 to control bud growth during mitosis is not essential for viability, and cells with defects in this pathway can easily be identified in genetic screens because they give rise to colonies with a peculiar morphology [3,4,7]. This pathway therefore represents an ideal genetic model system in which to understand the mechanisms used by cyclin-dependent kinases to control specific cell-cycle events. By screening for mutations that disrupt the control of bud growth during mitosis, we identified a protein kinase called Gin4 [7]. The phenotype caused by deletion of the GIN4 gene is nearly identical to the phenotype caused by deletion of the NAP1 gene, and Gin4 binds tightly to Nap1. Furthermore, the kinase activity of Gin4 is specifically activated during mitosis in a Nap1-dependent manner [7].

To learn more about the Clb2-dependent pathway that controls bud growth during mitosis, we have characterized The phenotype of the ecm30 mutation identified in a screen for a mutation that causes an elongated bud phenotype that mutations that cause an elongated cell morphology [7]. The mutant closely resembles the phenotype caused by loss of func- strain shown here is in the Clb2-dependent strain background used for the original genetic screen [7]. tion of Gin4 or Nap1 in cells that are dependent upon Clb2 for survival. These studies have demonstrated that the Cdc42 GTPase and a member of the PAK kinase family called Cla4 have important roles in the control of Nap1 and Gin4 to control bud growth during mitosis. To mitotic events by the Clb2 cyclin. further test whether Cla4 functions in this pathway, we looked for genetic and functional interactions between Results Cla4 and the proteins known to function in the pathway. Cla4, a member of the PAK kinase family, functions in a We first tested for genetic interactions between CLA4, Clb2-dependent mitotic control pathway GIN4 and NAP1. To do this, we constructed strains that In previous work we used a simple genetic screen to iden- carry a CLA4 gene deletion in combination with either a tify mutations that cause the formation of elongated buds GIN4 or NAP1 gene deletion. In each case, we found that due to a disruption of the control of polar bud growth cells carrying the double deletion were barely viable when during mitosis. This led to the identification of the Gin4 grown at 37°C, whereas cells carrying the single deletions protein kinase, which has an important role in the control formed colonies at a normal or nearly normal rates of mitotic events by Clb2 [7]. In the same genetic screen (Figure 2a). These results demonstrate that Cla4, Gin4 that identified Gin4, we isolated a mutation that causes an and Nap1 are involved in related functions within the cell. elongated bud phenotype that closely resembles the phenotype caused by deletion of the GIN4 gene or of the We next determined whether Cla4 interacts genetically NAP1 gene (Figure 1). We cloned the gene identified by with the mitotic cyclins. There are four mitotic cyclins in this mutation and found that it encodes a previously iden- budding yeast, called Clb1, Clb2, Clb3 and Clb4. These tified member of the PAK kinase family called Cla4 [8]. cyclins are highly redundant, but cells can be made depen- Members of the PAK kinase family were first identified as dent upon the Clb2 cyclin for viability by deleting the proteins that bind specifically to the GTP-bound form of genes for Clb1, Clb3 and Clb4 [14,15]. This is referred to as the Cdc42 GTPase, and members of the PAK kinase a Clb2-dependent genetic background, and allows one to family have been implicated in such diverse events as study pathways controlled by Clb2 without the complica- actin cytoskeletal rearrangements, AIDS progression, and tion of redundant pathways controlled by other mitotic signaling by the ras oncogene [9–13]. cyclins. We had found previously that deletion of the genes for Gin4 or Nap1 in a Clb2-dependent background results We found that deletion of the CLA4 gene causes an elon- in a much more severe phenotype than in a wild-type back- gated bud phenotype that is identical to the phenotype of ground, suggesting that Gin4 and Nap1 mediate Clb2- the mutation isolated in our screen. The elongated bud dependent functions [3,7]. We tested whether the same is phenotype caused by loss of Cla4 function strongly sug- true for Cla4 by examining the phenotype of a CLA4 dele- gested that Cla4 functions in the pathway used by Clb2, tion in a wild-type and in a Clb2-dependent background. Research Paper Control of mitotic events Tjandra et al. 993

Figure 2

Cla4 interacts genetically with proteins that ∆ function in a Clb2-dependent mitotic control (a) (c) gin4 pathway. (a) CLA4 interacts genetically with cdc42-1 NAP1 and GIN4. Strains carrying deletions of ∆nap1 ∆gin4 the CLA4, GIN4 or NAP1 genes either alone or in combination were grown on a YEPD plate at 37°C. (b) CLA4 interacts genetically with the mitotic cyclins. The indicated strains ∆ were grown in liquid YEPD media to log WT ∆cla4 WT nap1 phase at 30°C and photographed using cdc42-1 Nomarski optics. (c) Deletion of the NAP1 gene suppresses the temperature sensitivity of the cdc42-1 allele. The indicated strains ∆ ∆ were grown on a YEPD plate at 37°C. WT cla4 cla4 ∆gin4 ∆nap1 indicates wild type. cdc42-1

(b)

WT ∆cla4

∆clb1 ∆clb3 ∆clb4 ∆cla4 ∆clb1 ∆clb3 ∆clb4

Current Biology

As with GIN4 and NAP1, we found that the elongated bud can be conveniently detected by assaying for a hyperphos- phenotype of a CLA4 deletion was significantly enhanced phorylation-induced shift in the electrophoretic mobility of in the Clb2-dependent background (Figure 2b). CLA4 is Gin4 [7]. By assaying the mitosis-specific hyperphosphory- unique, however, in that deletion of the CLA4 gene pro- lation of Gin4 in synchronized cultures of ∆cla4 cells, we duced a more severe phenotype in a wild-type background found that Gin4 hyperphosphorylation was completely than does deletion of the genes for either GIN4 or NAP1. dependent upon Cla4 in vivo (Figure 3a). Control experi- This might suggest that Cla4 also has a role in the pathways ments showed that the Clb2 protein appeared at the same used by the other mitotic cyclins to control bud growth. time in wild-type and ∆cla4 cells, demonstrating that the defect in Gin4 activation is not due to failure of ∆cla4 cells Previous work has demonstrated genetic and physical inter- to enter mitosis (Figure 3b). Note that Clb2 protein levels actions between CLA4 and CDC42 [8,16]. We therefore remain high in the ∆cla4 cells, consistent with a prolonged searched for genetic interactions between CDC42 and genes mitotic delay. We have found that the Gin4 protein can be encoding proteins that function in the Clb2-dependent detected among the proteins that elute from a Cla4 affinity mitotic control pathway. We found that deletion of the NAP1 column, consistent with a functional relationship between gene completely suppressed the temperature sensitivity of Cla4 and Gin4 (data not shown). However, Gin4 is not the cdc42-1 allele, whereas deletion of the GIN4 gene did not among the major proteins that are easily visualized by (Figure 2c). These results suggest that Nap1 and Cdc42 are Coomassie blue staining of the eluted fractions, and we likely to function together within the cell, and they suggest therefore suspect that the interaction between Cla4 and that Nap1 may be a negative regulator of Cdc42. Gin4 is either of low affinity or indirect.

Cla4 is required for the mitosis-specific activation of the Clb2 and the Cdc42 GTPase act synergistically to induce Gin4 kinase hyperphosphorylation of Cla4 We next tested for functional interactions between Cla4 To learn more about the role of the Cla4 protein in and Gin4. The Gin4 protein becomes hyperphosphorylated mitosis, we determined whether Cla4 undergoes mitosis- during mitosis, which leads to activation of Gin4 kinase specific regulation. As a first step, we arrested cells in activity [7]. Thus, the mitosis-specific activation of Gin4 interphase or mitosis, and then used western blotting to 994 Current Biology, Vol 8 No 18

Figure 3 seemed possible that the mitotic cyclin Clb2 may somehow have a role in Cla4 hyperphosphorylation. We therefore determined whether overexpression of Clb2 and/or the (a) Time after release from α-factor (min) GTP-bound form of Cdc42 causes the hyperphosphoryla- 30 40 50 60 70 80 90 100 110 120 tion of Cla4. For these experiments, we used the condi- tional GAL10 promoter to express a mutant form of Cdc42 WT that is locked into the GTP-bound form (called Cdc42V12) Anti-Gin4 [17]. Similarly, we used the GAL1 promoter to express a ∆cla4 version of Clb2 that lacks the ‘destruction box’ that targets Clb2 for destruction during interphase (called Clb2∆176), thereby allowing Clb2 to accumulate in cells during inter- ∆ (b) Time after phase [18]. Before inducing expression of Clb2 176 or release from α-factor (min) Cdc42V12, we arrested cells in interphase so that we could 40 50 60 70 80 90 100 110 120 130 study the effects of these proteins under conditions where WT there are no other mitotic cyclins present. We found that Anti-Clb2 expression of Clb2∆176 or Cdc42V12 alone did not induce a ∆cla4 major shift in the electrophoretic mobility of Cla4. Expres- sion of the two proteins together, however, led to a rapid Current Biology and dramatic electrophoretic mobility shift (Figure 4b). Treatment of Cla4 with phosphatase caused the elec- Cla4 is required for the mitosis-specific hyperphosphorylation of the trophoretic mobility shift to disappear, demonstrating that Gin4 kinase. (a) Wild-type (WT) and ∆cla4 cells were synchronized in it is due to hyperphosphorylation (Figure 4c). G1 by arrest with the mating pheromone α-factor and were then released from the arrest and allowed to progress through the cell cycle. At each of the indicated times, samples were taken and the The wild-type version of Cdc42 was not capable of induc- hyperphosphorylation of Gin4 was assayed as previously described ing Cla4 hyperphosphorylation, demonstrating that only the using an anti-Gin4 antibody as a probe [7]. Hyperphosphorylation is GTP-bound form of Cdc42 is able to induce this event indicated by a shift in electrophoretic mobility of the Gin4 protein. (Figure 4b). Similarly, a non-degradable form of a related (b) A cell-cycle time course was carried out as in (a), except that the mitotic cyclin called Clb3∆46 samples were probed for the Clb2 mitotic cyclin with anti-Clb2 was only able to induce weak antibody. This experiment was carried out independently of the hyperphosphorylation of Cla4, indicating that hyperphos- experiment shown in (a), and the cells entered mitosis slightly later. phorylation is a Clb2-specific function rather than a result of arresting cells in mitosis (Figure 4d). Controls demonstrated that the non-degradable form of Clb3 was able to arrest cells determine whether Cla4 undergoes a cell-cycle-specific in mitosis with large buds in the same manner as Clb2 (data electrophoretic mobility shift that would be indicative of not shown). The fact that only a fraction of Cla4 undergoes post-translational modifications. We found that a small hyperphosphorylation during a normal cell cycle suggests fraction of the Cla4 protein was shifted to slower-migrat- either that hyperphosphorylation is a highly transient event, ing forms in cells arrested in mitosis (Figure 4a). A similar or that only a fraction of Cla4 is used during mitosis. fraction of Cla4 was shifted during mitosis when a synchronized population of cells went through a normal cell In previous work, we found that overexpression of a non- cycle. If the Cla4 protein from mitotic cells was treated degradable form of Clb2 in cells arrested in interphase is with phosphatase these slower-migrating forms disap- able to induce hyperphosphorylation of Gin4 (see peared, demonstrating that Cla4 is hyperphosphorylated Figure 11 of Altman and Kellogg [7]). In those studies (data not shown, but see Figure 4c below). however, we found that only a fraction of Gin4 is induced to undergo hyperphosphorylation by expression of Clb2, This result raises two questions. First, why does only a frac- which we could detect on a long exposure of a western tion of Cla4 undergo hyperphosphorylation during mitosis? blot. We suspect that a small fraction of Gin4 is hyper- Also, what factors are responsible for inducing the mitosis- phosphorylated in this situation because a fraction of the specific hyperphosphorylation of Cla4? To begin to answer Cdc42 in the cell is transiently in the GTP-bound form. these questions, we looked more closely at the regulation We only observe rapid and quantitative hyperphosphory- of Cla4 hyperphosphorylation. Cla4 is known to be acti- lation of Gin4 when both Clb2 and the GTP-bound form vated in vitro by binding of the GTP-bound form of Cdc42 of Cdc42 are expressed together. [8,16]. It therefore seemed possible that hyperphosphorylation of Cla4 is the result of autophosphorylation induced by The hyperphosphorylated form of Cla4 is likely to signal binding of the GTP-bound form of Cdc42 to Cla4. In addi- the activation of the Gin4 kinase tion, as Cla4 hyperphosphorylation is a mitosis-specific We next addressed the effect of expression of event and Cla4 shows genetic interactions with Clb2, it also Clb2∆176 and Cdc42V12 on Gin4 activation. As with Cla4 Research Paper Control of mitotic events Tjandra et al. 995

Figure 4

Hyperphosphorylation of Cla4 is induced by (a) (b) Clb2 and the GTP-bound form of Cdc42. (a) Cla4 is hyperphosphorylated in mitosis. Cdc42V12 Cdc42 Cells were arrested in interphase with α-factor + + cla4 Clb2∆176 Cdc42V12 Clb2∆176 Clb2∆176 or in mitosis with benomyl, and samples were Interphase Mitosis ∆ immunoblotted with affinity-purified anti-Cla4 0123 0123 0123 0123 Time after addition antibodies. A sample from a ∆cla4 strain of galactose (h) served as a control for the specificity of the anti-Cla4 antibodies. The arrow points to the hyperphosphorylated form of Cla4 observed (c) (d) in mitotic cells. (b) The combined activity of Cdc42V12 Clb2∆176 and Cdc42V12 induces Cla4 PPase + ∆ hyperphosphorylation. Cells were induced to Clb3 46 – + Time after addition express Clb2∆176, Cdc42V12, Cdc42V12 and 0123 of galactose (h) Clb2∆176, or wild-type Cdc42 and Clb2∆176 together. Samples were collected at 1 h Current Biology intervals after induction of expression and were immunoblotted with anti-Cla4 antibodies. (c) Cla4 is hyperphosphorylated in the immunoprecipitated Cla4 was treated with (d) The mitotic cyclin Clb3 is unable to induce cells that express Clb2∆176 and Cdc42V12. either a buffer control (–) or with full hyperphosphorylation of Cla4. Cells Cla4 was immunoprecipitated from cells λ-phosphatase (PPase; +) followed by expressing Clb3∆46 and Cdc42V12 were expressing both Cdc42V12 and Clb2∆176, and immunoblotting with anti-Cla4 antibodies. assayed for Cla4 hyperphosphorylation.

hyperphosphorylation, we found that Gin4 became hyper- hyperphosphorylation did not occur at the restrictive tem- phosphorylated only in the situation where Clb2∆176 and perature in cells carrying a temperature-sensitive allele of Cdc42V12 were expressed together (Figure 5a). Further- the CDC28 gene, consistent with the idea that Cla4 func- more, we found that expression of Clb2∆176 and Cdc42V12 tions in a pathway that is activated by the Clb2–Cdc28 together in a ∆cla4 strain did not lead to Gin4 hyperphosphorylation (Figure 5b), consistent with our previous results demonstrating that the mitosis-specific activation Figure 5 of Gin4 required Cla4 (see Figure 3a). These results provide further evidence that Cla4 is required for activa- (a) Cdc42V12 Clb2∆176 Cdc42V12 + tion of the Gin4 kinase, and they argue strongly that the Clb2∆176 hyperphosphorylated form of Cla4 is responsible for sig- 0123 0123 0123 Time after addition of galactose (h) naling the activation of Gin4.

Cdc42 is required for the mitosis-specific hyperphosphorylation of Gin4 (b) (c) Cdc42V12 The previous results suggest that Cdc42 should be + required for the mitosis-specific hyperphosphorylation of Clb2∆176 WT cdc42-1 Time after addition 0123 23 37 23 37 Temp (ºC) Gin4. To test whether this is true, cells carrying the cdc42-1 of galactose (h) temperature sensitive mutation were arrested in mitosis by treatment with the microtubule polymerization inhibitor benomyl. We then shifted the cells to the restrictive tem- ∆cla4 Current Biology perature for 2 hours in the continued presence of benomyl, and used western blotting to assay Gin4 hyperphosphorylation. We found that Gin4 hyperphosphorylation was lost at Hyperphosphorylation of Cla4 is needed for the hyperphosphorylation of Gin4. (a) Hyperphosphorylation of Gin4 correlates with Cla4 the restrictive temperature in the cdc42-1 cells, but not in hyperphosphorylation. The same samples used to assay Cla4 wild-type control cells, as expected if Cdc42 is required for hyperphosphorylation in Figure 4b were immunoblotted with anti-Gin4 Gin4 hyperphosphorylation (Figure 5c). antibodies to assay for Gin4 hyperphosphorylation. (b) Cla4 is required for the Cdc42V12- and Clb2∆176-induced activation of Gin4. Samples from a ∆cla4 strain that overexpresses both Cdc42V12 and Nap1 and Cdc28, but not Gin4, are required for Cla4 Clb2∆176 were immunoblotted with anti-Gin4 antibodies. (c) Cdc42 is hyperphosphorylation required for the mitosis-specific hyperphosphorylation of Gin4. Log- To further delineate the requirements for Cla4 hyper- phase wild-type and cdc42-1 cells were pretreated with 30 µg/ml ° ° ° phosphorylation, we assayed the ability of Clb2∆176 and benomyl at 23 C for 1 h before incubation at 23 C or at 37 C for an V12 additional 2 h. Extracts made from the yeast cells were then Cdc42 to induce hyperphosphorylation of Cla4 in immunoblotted with anti-Gin4 antibodies. various mutant backgrounds. We found that Cla4 996 Current Biology, Vol 8 No 18

Figure 6 Figure 7

(a) WT cdc28-4 0123 0123 Time after addition of galactose (h) cla4 ∆ Log phase K594A Uninduced Induced

Myelin basic protein (b) ∆gin4 Current Biology 0123 Time after addition of galactose (h) The in vitro kinase activity of Cla4 is not regulated by hyperphosphorylation. The Cla4 protein was immunoprecipitated from rapidly growing wild-type yeast cells and assayed for kinase activity. As controls for the specificity of the kinase assay, ∆cla4 cells or cells carrying the cla4K594A kinase-dead mutation were used. To test (c) ∆nap1 whether hyperphosphorylation of Cla4 leads to activation of kinase 0123Time after addition activity, kinase assays were carried out either using cells in which Cla4 of galactose (h) hyperphosphorylation was induced by expression of Cdc42V12 and Clb2∆176, or using uninduced control cells.

Current Biology and from cells expressing Clb2∆176 and Cdc42V12, and then Cdc28 and Nap1, but not Gin4, are required for Cla4 assayed kinase activity using myelin basic protein as a sub- hyperphosphorylation. (a) Expression of Cdc42V12 and Clb2∆176 was induced with galactose in wild-type (WT) and cdc28-4 cells at 37°C. strate. As controls we used strains carrying either a dele- Samples were collected at 1 h intervals and assayed for Cla4 tion of the CLA4 gene or a mutant version of CLA4 in hyperphosphorylation. Note that Cla4 hyperphosphorylation does not which an essential lysine in the kinase active site is occur as efficiently at 37°C in the wild-type control cells as it does at changed to an alanine (Cla4K594A). We found that the in 30°C (data not shown). (b) Expression of Cdc42V12 and Clb2∆176 was induced in ∆gin4 cells. Samples were collected at 1 h intervals and vitro kinase activity of Cla4 is not significantly affected by assayed for Cla4 hyperphosphorylation. (c) Expression of Cdc42V12 the hyperphosphorylation of Cla4 (Figure 7, compare and Clb2∆176 was induced in ∆nap1 cells. Samples were collected at uninduced and induced lanes). This result does not rule 1 h intervals and assayed for Cla4 hyperphosphorylation. out the possibility that the kinase activity of Cla4 is regulated by hyperphosphorylation in vivo. kinase complex (Figure 6a). We also found that Cla4 Discussion hyperphosphorylation occurs normally in ∆gin4 cells, con- Cla4 and Cdc42 function in a Clb2-dependent mitotic sistent with the idea that Cla4 functions to activate the control pathway Gin4 kinase (Figure 6b). In contrast, Cla4 hyperphospho- A variety of biochemical and genetic experiments demon- rylation is significantly reduced in ∆nap1 cells (Figure 6c), strate that Cla4 and Cdc42 function in a mitotic control consistent with the genetic interactions between Nap1 pathway that includes Clb2, Cdc28, Nap1 and Gin4. First, and Cla4 (see Figure 2a), and with other results demon- deletion of the CLA4 gene causes an elongated bud phe- strating an involvement of Nap1 in the pathway used by notype that is nearly identical to the phenotype caused by Clb2 and Cdc28 to control bud growth during mitosis loss of function of Clb2, Nap1 or Gin4 in cells that are [4,7]. As Nap1 is also required for the activation of the dependent upon Clb2 for survival. Second, CLA4 interacts Gin4 kinase and we have detected a tight biochemical genetically with GIN4, NAP1 and the mitotic cyclins, and interaction between Gin4 and Nap1, it appears that Nap1 CDC42 interacts genetically with NAP1. Third, Cla4 is may have an integral role in the regulation of both Cla4 required in vivo for the mitosis-specific hyperphosphoryla- and Gin4 kinases [7]. tion of Gin4. Fourth, Cla4 hyperphosphorylation is induced in vivo by the combined activity of Clb2, Cdc28, The in vitro kinase activity of Cla4 is not regulated by Nap1 and the GTP-bound form of Cdc42. Finally, strong hyperphosphorylation hyperphosphorylation of Cla4 can be induced by Clb2, but In previous work, it was found that Cla4 kinase activity not by the related mitotic cyclin Clb3, demonstrating that can be detected throughout the cell cycle, and that kinase Cla4 functions in a Clb2-specific pathway. Figure 8 sum- activity increases approximately sevenfold during mitosis marizes the epistasis relationships that we have observed [16]. We therefore tested whether hyperphosphorylation in this study. It should be stressed, however, that several of Cla4 leads to changes in Cla4 kinase activity. To do results are not consistent with a simple linear signaling this, we immunoprecipitated Cla4 from log-phase cells pathway. For example, cells carrying deletions of the Research Paper Control of mitotic events Tjandra et al. 997

genes for both Cla4 and Gin4 have a more severe pheno- Figure 8 type than either single deletion (see Figure 2), and Gin4 and Nap1 bind directly to each other. It seems likely that Polar signaling events in this pathway are carried out by multi- bud protein complexes, and that there is cross-talk between growth Nap1 intricate and partially redundant pathways, as has been Clb2 Cla4 Gin4 Cdc28 Cdc42 observed for other signaling pathways [19–21]. Mitotic spindle function? Previous studies on Cla4 and a related member of the Current Biology PAK kinase family called Ste20 also provide strong evidence supporting the idea that Cla4 functions during A model summarizing the epistasis relationships observed in this study. The Clb2–Cdc28 kinase complex induces the hyperphosphorylation of mitosis. It was found that cells carrying a ste20 deletion the Cla4 kinase in a Nap1-dependent and Cdc42-dependent manner, and a temperature-sensitive allele of cla4 arrest early in and the hyperphosphorylated form of Cla4 appears to signal the mitosis with a short spindle when shifted to the restrictive activation of the Gin4 kinase. The activated form of Gin4 negatively temperature, consistent with the idea that Cla4 carries out regulates polar bud growth and may play a part in the assembly or function of the mitotic spindle (see Discussion). mitotic functions [8]. The fact that the ∆ste20 cla4ts cells arrest with a short spindle is also consistent with the idea that Cla4 functions with Gin4 and Nap1, as loss of func- As Cla4 is known to interact with the Cdc42 GTPase, we tion of either of these genes can also cause a prolonged reasoned that only a fraction of Cla4 undergoes hyper- delay at the short spindle stage [4,7]. Interestingly, Cla4 phosphorylation during mitosis because hyperphosphory- and Gin4 were both identified in screens for mutations lation is a transient event that can only be induced by the that cause lethality in combination with a double deletion GTP-bound form of Cdc42. Consistent with this idea, we of the G1 cyclin genes CLN1 and CLN2 [8,16,22,23]. found that coexpression of Clb2 and the GTP-bound form Although the basis for the synthetic lethality is unknown, of Cdc42 leads to nearly quantitative hyperphosphoryla- this finding supports the idea that Gin4 and Cla4 are tion of Cla4, even in cells arrested in interphase. Two involved in related cell-cycle functions. lines of evidence argue strongly that Cla4 hyperphosphorylation is a biologically relevant event with an important Several recent studies also support the idea that Cdc42 role in the Clb2-dependent pathway that controls polar carries out mitotic functions. First, newly isolated mutant bud growth during mitosis. First, Clb2, Cdc28 and Nap1 alleles of the CDC42 gene cause the formation of elongated are all required in vivo for Cla4 hyperphosphorylation, and buds, consistent with the idea that Cdc42 function is each of these proteins has previously been shown to func- required during mitosis to induce the switch from polar to tion in the Clb2-dependent pathway. Second, we have so isotropic bud growth (K. Kozminski and D. Drubin, per- far found that Gin4 hyperphosphorylation only occurs sonal communication). The original cdc42-1 allele also pro- when Cla4 becomes hyperphosphorylated, suggesting that duces a fraction of cells with elongated buds when crossed the hyperphosphorylated form of Cla4 is responsible for into the W303 strain background [8]. Second, independent signaling the hyperphosphorylation of Gin4. genetic screens for regulators of Cdc42 and for suppressors of a mutant cdc28 allele both led to the identification of We do not yet understand how hyperphosphorylation reg- related proteins called Zds1 and Zds2. Cells carrying dele- ulates Cla4 activity. Hyperphosphorylation does not tions of the genes for both Zds1 and Zds2 grow very appear to have any significant effect on Cla4 kinase activ- slowly, have elongated buds, and undergo prolonged ity in vitro, although this does not rule out the possibility delays in early mitosis at the short spindle stage [24,25]. that hyperphosphorylation regulates kinase activity in vivo. One possibility is that hyperphosphorylation regu- Cla4 and Ste20 show a high degree of sequence similarity lates the ability of Cla4 to bind to other proteins that func- and share redundant functions in vivo [8,26,27]. However, tion in the pathway. For example, perhaps Cla4 we observed that Gin4 activation fails to occur in ∆cla4 hyperphosphorylation causes assembly of a transient cells that have a wild-type copy of the STE20 gene. In complex that leads to activation of the Gin4 kinase. addition, we have so far been unable to detect any genetic interactions between STE20 and GIN4, NAP1 or CLB2 The in vivo functions of the Cdc42 GTPase (data not shown). Therefore, it is unlikely that Cla4 and Studies carried out in vitro have demonstrated that Ste20 share redundant functions in the pathway or path- members of the PAK kinase family are activated by ways used to control Gin4 activation and polar bud growth. binding of the GTP-bound form of the Cdc42 GTPase, but not the GDP-bound form, suggesting that Cdc42 may Cla4 undergoes mitosis-specific hyperphosphorylation function as an on/off switch for PAK kinases [9,28]. It is We found that a fraction of the total Cla4 protein present interesting to note, however, that the GTP-bound form of in the cell becomes hyperphosphorylated during mitosis. Cdc42 is necessary, but not sufficient, for the induction of 998 Current Biology, Vol 8 No 18

Cla4 hyperphosphorylation. This finding suggests that the We have found that Cla4 and Cdc42 have critical roles in a function of Cdc42 in vivo may be more complex than pathway used by the Clb2–Cdc28 cyclin-dependent simply acting as an on/off switch for the activation of kinase complex to control certain mitotic events. We also protein kinases. Cdc42 has been implicated in a remark- found that Nap1, the Clb2–Cdc28 kinase complex and the ably wide variety of cellular events, and it remains unclear GTP-bound form of Cdc42 work together to induce how the same protein functions in so many different con- hyperphosphorylation of Cla4, and that Cla4 is required texts [29,30]. Indeed, the exact molecular role of Cdc42 in for the mitosis-specific activation of the Gin4 kinase. vivo remains unclear. One interesting possibility is that These results show that the Clb2–Cdc28 kinase complex Cdc42 functions to monitor the specificity of signaling controls specific events of mitosis through a GTPase- reactions by linking multiprotein complex formation linked signaling mechanism. It seems possible that cyclin- and/or phosphorylation events to GTP hydrolysis. dependent kinases may control many other cell-cycle events by activating complex signaling cascades rather The role of Cla4 in the control of cell-cycle events than by directly phosphorylating proteins involved in spe- Cla4 is required in vivo for the mitosis-specific activation cific events. Additional genetic and biochemical studies of the Gin4 kinase, and for the proper control of bud should lead to the identification of more proteins that growth during mitosis. In addition, loss of Cla4 function in function in the Clb2-dependent mitotic control pathway, ∆ste20 cells causes an arrest at the short spindle stage. It is as well as to an understanding of the molecular mecha- unclear at present why loss of function of Cla4, Gin4 or nisms underlying the mitosis-specific regulation of the Nap1 can cause prolonged mitotic delays at the short Cla4 and Gin4 kinases. spindle stage, although there are a number of interesting possibilities. One possible explanation is that these pro- Materials and methods teins are directly involved in aspects of mitotic spindle Strains used in this study assembly, and loss of function causes cells to have diffi- All strains are in the W303 background (leu2-3,112 ura3-52 can1- 100 ade2-1 his3-11 trp1-1 ssd1 ho). DK278 was derived by crossing culty assembling a functional mitotic spindle. Another the cdc42-1 mutant allele into the W303 background three times. possibility is that these proteins are primarily involved in DK96: Matα∆nap1::LEU2. DK186: Mata ∆bar1. DK212: Mata ∆clb1 a pathway that controls bud growth during mitosis, and ∆clb3::TRP1 ∆clb4::HIS3 ∆bar1. DK247: Mata URA3::gal- ∆176∆ K594A defects in this pathway activate a checkpoint that arrests CLB2 bar1. DK278: Mata cdc42-1. DK281: Mata CLA4 ∆bar1. RA4: Mata ∆gin4::LEU2. JC2: Mata ecm30-2 ∆clb3::TRP1 the nuclear cycle at the short spindle stage. A third possi- ∆clb4::HIS3. HT1: Mata ∆cla4::URA3 ∆bar1. HT9: Mata bility is that Cla4, Gin4 and Nap1 are part of a checkpoint ∆cla4::URA3 ∆clb1 ∆clb3::TRP1 ∆clb4::HIS3 ∆bar1. HT15: Mata pathway that normally functions during mitosis to inacti- ∆gin4::LEU2 ∆cla4::URA3. HT16: Mata ∆nap1::LEU2 ∆cla4::URA3. ∆ vate a negative regulator of the cell cycle when mitotic HT31: Mata URA3::gal1-CLB2 176/gal10-CDC42V12 ∆bar1. HT36: Mata cdc28-4 URA3::gal1-CLB2∆176/gal10-CDC42V12. HT37: Mata events are occurring properly. Loss of function of such a URA3::gal10-CDC42V12 ∆bar1. HT39: Mata URA3::gal1- pathway would result in a failure to inactivate the CLB2∆176/gal10-CDC42 ∆bar1. HT42: Mata URA3::gal1- inhibitor, which would lead to a mitotic delay. A final pos- CLB3∆46/gal10-CDC42V12 ∆bar1. HT46: Mata ∆cla4::LEU2 ∆ sibility is that Cla4, Nap1 and Gin4 function in a pathway URA3::gal1-CLB2 176/gal10-CDC42V12 ∆bar1. HT52: Mata ∆ ∆ that triggers the destruction of mitotic cyclins or other gin4::LEU2 cdc42-1. HT53: Mata nap1::LEU2 cdc42-1. HT54: Mata ∆gin4::LEU2 URA3::gal1-CLB2∆176 /gal10-CDC42V12. HT55: proteins that must be destroyed before cells can exit Mata ∆nap1::LEU2 URA3::gal1-CLB2∆176 /gal10-CDC42V12. mitosis. These various possibilities are not mutually exclusive. For example, studies on the DNA damage Cloning and deletion of the CLA4 gene checkpoint have demonstrated that proteins involved in We cloned the CLA4 gene by complementation using a genomic library carried in a CEN plasmid. The cloning was aided by the fact that the carrying out specific cell-cycle events can also be involved original mutant allele identified in our genetic screen is hypersensitive to in checkpoints that detect when cell-cycle events have hydroxyurea. After transforming mutant cells with the plasmid library, we not occurred properly [31,32]. were therefore able to grow cells on plates containing 5 mg/ml hydroxyurea, which strongly inhibited the growth of mutant cells without significantly affecting the growth of wild-type cells. Cells carrying the NAP1 Control of cell-cycle events by cyclin-dependent kinases deletion are also hydroxyurea sensitive (D. Kellogg, unpublished obser- The mechanisms used by cyclin-dependent kinases to vations). The original mutation segregates 2:2 with a deletion of the induce and coordinate the events of the cell cycle are CLA4 gene, demonstrating that the mutation lies in the CLA4 gene. largely unknown. For example, it remains unclear whether cyclin-dependent kinases directly phosphorylate To delete the CLA4 gene, we used PCR to amplify the URA gene from the pRS306 vector with 60 bases of homology to the 5′ and 3′ ends of proteins involved in cell-cycle events, or whether they the CLA4 open reading frame on each end (primers: ATGTCTCTTTC- activate intricate signaling pathways that ultimately regu- AGCTGCAGCGAATAAGATATCTGACAACGATTTCCAAAATATCG late the proteins directly involved in cell-cycle events. GACCGGCCGGCATCAGAGCAGATTGTAC and TCATTCCTT- Given the complexity of cell-cycle events and the fact CCACTCCAACAGTGATGTCAAATCCTTTGGATCACATGCCATATT that they must be precisely controlled, it would not be GAAAAAACCGTCCTTACGCATCTGTGCGGTAT). The resulting PCR product was transformed into strain DK186 or DK212, and cells unexpected to find that they are regulated by intricate carrying the CLA4 deletion were identified by PCR and confirmed by signaling pathways. western blotting. Research Paper Control of mitotic events Tjandra et al. 999

Overexpression of Clb2∆176 and Cdc42V12 then immunoprecipitated and assayed as previously described for Gin4 To create plasmids that express Clb2∆176 and Cdc42V12 we used the kinase assays, except that 2.5 µg myelin basic protein was used as a plasmid pDK20, which carries the bidirectional GAL1/10 promoter substrate for each assay [7]. cloned into the KpnI and XhoI sites of pRS306. A plasmid that expresses Cdc42V12 under the control of the GAL10 promoter was constructed by cloning the DNA encoding Cdc42V12 into the GAL10 Acknowledgements side of the promoter (pHT26), and a plasmid that expresses Clb2∆176 We thank John Sisson, Bill Sullivan, Kristina Yu and members of the Kellogg under the control of the GAL1 promoter was constructed by cloning Lab for discussion and/or critical reading of the manuscript; Kim Nasmyth for the DNA encoding Clb2∆176 into the GAL1 side of the promoter the cla4::LEU2 strain, Adam Rudner for pAR39-1, and Doug Johnson for a Cdc42V12 plasmid. Deeann Guzman-Martinez helped with the construction (pAR39-1, gift of Adam Rudner). To construct a plasmid that expresses of the Cla4 kinase domain mutant strain. This work was supported by grants ∆176 V12 both proteins, Clb2 and Cdc42 were cloned into pDK20 from the American Cancer Society and the National Institutes of Health. together into the same sites described above. Plasmids that express wild-type Cdc42 and Clb2∆176, or Cdc42V12 and Clb3 were constructed in the same manner (pHT28 and pDK59, respectively). All References plasmids were integrated at the URA3 gene. 1. Murray A, Hunt T: The Cell Cycle: An Introduction. New York: Oxford University Press, 1993. To induce expression of proteins, yeast cells were grown overnight at 2. Norbury C, Nurse P: Animal cell cycles and their control. Annu Rev Biochem 1992, 61:441-470. room temperature to an OD600 of 0.5–0.6 in YEP media containing 2% glycerol and 2% ethanol. The cells were arrested in interphase by addi- 3. Kellogg DR, Kikuchi A, Fujii-Nakata T, Turck CW, Murray A: Members of the NAP/SET family of proteins interact specifically with B-type tion of α-factor to 2 µg/ml, followed by incubation at 30°C for 2 h. cyclins. J Cell Biol 1995, 130:661-673. Expression of proteins was then induced by the addition of galactose 4. Kellogg DR, Murray AW: NAP1 acts with Clb2 to perform mitotic to 2%, and 1.6 ml samples are taken at the indicated times. Samples functions and suppress polar bud growth in budding yeast. J Cell were prepared for electrophoresis and western blotting as previously Biol 1995, 130:675-685. described [7]. 5. Lew DJ, Reed SI: Cell cycle control of morphogenesis in budding yeast. Curr Opin Genet Dev 1995, 5:17-23. Cell-cycle arrests and detection of proteins 6. Lew DJ, Reed SI: Morphogenesis in the yeast cell cycle: regulation Cell-cycle arrests and preparation of samples for gel electrophoresis by Cdc28 and cyclins. J Cell Biol 1993, 120:1305-1320. Control of mitotic events by Nap1 and the and western blotting were carried out as previously described [7]. To 7. Altman R, Kellogg DR: Gin4 kinase. J Cell Biol 1997, 138:119-130. resolve differently phosphorylated forms of Cla4, we electrophoresed 8. Cvrckova F, De Virgilio C, Manser E, Pringle JR, Nasmyth K: Ste20- samples through 10% polyacrylamide gels (6.5 by 14.0 cm) for 3 h at like protein kinases are required for normal localization of cell 185 V [33]. To detect Cla4, we used an affinity-purified rabbit poly- growth and for cytokinesis in budding yeast. Genes Dev 1995, clonal antibody raised against a glutathione-S-transferase (GST) fusion 9:1817-1830. protein that includes amino acids 140–511 of the Cla4 protein. Gin4 9. Manser E, Leung T, Salihuddin H, Zhao Z, Lim L: A brain hyperphosphorylation was assayed as previously described [7]. For the serine/threonine protein kinase activated by Cdc42 and Rac1. experiment shown in Figure 5c, cells carrying the cdc42-1 allele were Nature 1994, 367:40-46. incubated in the presence of 30 µg/ml benomyl in YPD media at room 10. Tang Y, Chen Z, Ambrose D, Liu J, Gibbs JB, Chernoff J, Field J: Kinase-deficient Pak1 mutants inhibit Ras transformation of Rat-1 temperature for 1 h before shifting to the restrictive temperature of ° fibroblasts. Mol Cell Biol 1997, 17:4454-4464. 37 C for 2 h. 11. Sawai ET, Khan IH, Montbriand PM, Peterlin BM, Cheng-Mayer C, Luciw PA: Activation of PAK by HIV and SIV Nef: importance for Treatment of Cla4 with phosphatase AIDS in rhesus macaques. Curr Biol 1996, 6:1519-1527. Cla4 was immunoprecipitated from cells expressing Clb2∆176 and 12. Sells MA, Knaus UG, Bagrodia S, Ambrose DM, Bokoch GM, Cdc42V12 and treated with lambda phosphatase (New England Chernoff J: Human p21-activated kinase (Pak1) regulates actin Biolabs) as previously described for Gin4 [7]. organization in mammalian cells. Curr Biol 1997, 7:202-210. 13. Manser E, Huang HY, Loo TH, Chen XQ, Dong JM, Leung T, Lim L: Expression of constitutively active alpha-PAK reveals effects of Construction of a Cla4 kinase domain mutant and Cla4 the kinase on actin and focal complexes. Mol Cell Biol 1997, kinase assays 17:1129-1143. To generate a catalytically inactive mutant version of Cla4, two PCR 14. Fitch I, Dahmann C, Surana U, Amon A, Nasmyth K, Goetsch L, et al.: reactions were carried out to amplify the 5′ and 3′ ends of the open Characterization of four B-type cyclin genes of the budding yeast reading frame, overlapping by 20 base pairs. The first oligo pair was Saccharomyces cerevisiae. Mol Biol Cell 1992, 3:805-818. GCGGAATTCGTATCATCAGACAGTAACTCC (Cla4-33) and 15. Richardson H, Lew DJ, Henze M, Sugimoto K, Reed S: Cyclin B GATAGCAACTTTATCTCCTA (Cla4-38), while the second pair was homologs in Saccharomyces cerevisiae function in S phase and in GAGATAAAGTTGCTATCGCGCAAATGGTCTTATCTAAAC (Cla4- G2. Genes Dev 1992, 6:2021-2034. 16. Benton BK, Tinkelenberg A, Gonzalez I, Cross FR: Cla4p, a 35) and GCGCAAGCTTTCATTCCTTCCACTCCAACAG (Cla4-37). Saccharomyces cerevisiae Cdc42p-activated kinase involved in The Cla4-35 oligo incorporates two mismatched bases that change the cytokinesis, is activated at mitosis. Mol Cell Biol 1997, conserved lysine in subdomain II to an alanine. The products of these 17:5067-5076. two reactions were gel purified, mixed together, and used as a tem- 17. Ziman M, O’Brien JM, Ouellette LA, Church WR, Johnson DI: plate for a PCR reaction with primers Cla4-33 and Cla4-37 to generate Mutational analysis of CDC42Sc, a Saccharomyces cerevisiae a full-length fragment that incorporates the mutation, which was cloned gene that encodes a putative GTP-binding protein involved in the into the vector YIplac211. The resulting vector (pDM2) was digested control of cell polarity. Mol Cell Biol 1991, 11:3537-3544. with AvrII to target integration at the CLA4 gene, and transformants 18. Amon A, Irniger S, Nasmyth K: Closing the cell cycle circle in yeast: were grown in the presence of 5-fluoro-orotic acid to select for recom- G2 cyclin proteolysis initiated at mitosis persists until the activation of G1 cyclins in the next cycle. Cell 1994, 77:1037-1050. bination events between the duplicated CLA4 sequences that loop out 19. Burgering BM, Bos JL: Regulation of Ras-mediated signalling: more the URA3 gene. Recombination events that leave the mutation in the than one way to skin a cat. Trends Biochem Sci 1995, 20:18-20. genomic copy of Cla4 were identified by PCR. This procedure was 20. Denhardt DT: Signal-transducing protein phosphorylation carried out with two different plasmid isolates, both of which gave the cascades mediated by Ras/Rho proteins in the mammalian cell: same results. potential for multiplex signalling. Biochem J 1996, 318:729-747. 21. Duffy JB, Perrimon N: Recent advances in understanding signal To assay Cla4 kinase activity, the cells from 15 ml of culture at an transduction pathways in worms and flies. Curr Opin Cell Biol OD600 of 0.7 were harvested and frozen on liquid nitrogen. Cla4 was 1996, 8:231-238. 1000 Current Biology, Vol 8 No 18

22. Benton BK, Tinkelenberg AH, Jean D, Plump SD, Cross FR: Genetic analysis of Cln/Cdc28 regulation of cell morphogenesis in budding yeast. EMBO J 1993, 12:5267-5275. 23. Cvrckova F, Nasmyth K: Yeast G1 cyclins CLN1 and CLN2 and a GAP-like protein have a role in bud formation. EMBO J 1993, 12:5277-5286. 24. Yu Y, Jiang YW, Wellinger RJ, Carlson K, Roberts JM, Stillman DJ: Mutations in the homologous ZDS1 and ZDS2 genes affect cell cycle progression. Mol Cell Biol 1996, 16:5254-5263. 25. Bi E, Pringle JR: ZDS1 and ZDS2, genes whose products may regulate Cdc42p in Saccharomyces cerevisiae. Mol Cell Biol 1996, 16:5264-5275. 26. Leberer E, Wu C, Leeuw T, Fourest-Lieuvin A, Segall JE, Thomas DY: Functional characterization of the Cdc42p binding domain of yeast Ste20p protein kinase. EMBO J 1997, 16:83-97. 27. Peter M, Neiman AM, Park HO, van Lohuizen M, Herskowitz I: Functional analysis of the interaction between the small GTP binding protein Cdc42 and the Ste20 protein kinase in yeast. EMBO J 1996, 15:7046-7059. 28. Brown JL, Stowers L, Baer M, Trejo J, Coughlin S, Chant J: Human Ste20 homologue hPAK1 links GTPases to the JNK MAP kinase pathway. Curr Biol 1996, 6:598-605. 29. Van Aelst L, D’Souza-Schorey C: Rho GTPases and signaling networks. Genes Dev 1997, 11:2295-2322. 30. Hall A: Rho GTPases and the actin cytoskeleton. Science 1998, 279:509-514. 31. Saka Y, Yanagida M: Fission yeast cut5+, required for S phase onset and M phase restraint, is identical to the radiation-damage repair gene rad4+. Cell 1993, 74:383-393. 32. Navas TA, Zhou Z, Elledge SJ: DNA polymerase epsilon links the DNA replication machinery to the S phase checkpoint. Cell 1995, 80:29-37. 33. Anderson CW, Baum PR, Gesteland RF: Processing of adenovirus 2-induced proteins. J Virol 1973, 12:241-252.

Because Current Biology operates a ‘Continuous Publication System’ for Research Papers, this paper has been published on the internet before being printed. The paper can be accessed from http://biomednet.com/cbiology/cub — for further information, see the explanation on the contents page.