The kelch repeat protein, Tea1, is a potential substrate target of the p21-activated kinase, Shk1, in the fission yeast, Schizosaccharomyces pombe.

The p21-activated kinase (PAK) homolog, Shk1, is a critical component of a multifunctional Ras/Cdc42/PAK complex required for viability, polarized growth and cell shape, and sexual differentiation in the fission yeast, Schizosaccharomyces pombe. Substrate targets of the Shk1 kinase have not previously been described. Here we show that the S. pombe cell polarity factor, Tea1, is directly phosphorylated by Shk1 in vitro. We demonstrate further that Tea1 is phosphorylated in S. pombe cells and that its level of phosphorylation is significantly reduced in cells defective in Shk1 function. Consistent with a role for Tea1 as a potential downstream effector of Shk1, we show that a tea1 null mutation rescues the Shk1 hyperactivity-induced lethal phenotype caused by loss of function of the essential Shk1 inhibitor, Skb15. All phenotypes associated with Skb15 loss, including defects in actin cytoskeletal organization, chromosome segregation, and cytokinesis, are suppressed by tea1 Delta, suggesting that Tea1 is a potential mediator of multiple Shk1 functions. S. pombe cells carrying a weak hypomorphic allele of shk1 together with a tea1 Delta mutation exhibit a cytokinesis defective phenotype that is significantly more severe than that observed in the respective single mutants, providing evidence that Shk1 and Tea1 cooperate to regulate cytokinesis. In addition, we show that S. pombe cells carrying the orb2-34 allele of shk1 exhibit a pattern of monopolar growth similar to that observed in tea1 Delta cells, suggesting that Shk1 and Tea1 may regulate one or more common processes involved in the regulation of polarized cell growth. Taken together, our results strongly implicate Tea1 as a potential substrate-effector of the Shk1 kinase.

In addition to Cdc42, Shk1 function is positively modulated by at least 3 other nonessential proteins, Scd2 (26,27), Skb1 (17,29), and Skb5 (30). Scd2 binds directly to Scd1, Cdc42, and Shk1, and appears to function as a scaffold that positively modulates protein-protein interactions between Scd1 and Cdc42 and between Cdc42 and Shk1 (10,26). Like ras1⌬ and scd1⌬ mutants, scd2⌬ cells are viable but ovoid in shape and mating defective. Skb1 and Skb5 appear to be specialized modulators of Shk1 function in that they are required for proper maintenance of cell polarity under hyperosmotic conditions but not under normal growth conditions (30,31). Shk1 is negatively regulated by a highly conserved WD repeat protein, Skb15 (7). Loss of Skb15 function is lethal and results in severe morphological, cytoskeletal, mitotic, and cytokinesis defects. Skb15 loss results in hyperactivation of the Shk1 kinase and can be suppressed by mutations that partially inactivate Shk1 (7). Thus, phenotypes resulting from loss of Skb15 function can be attributed to Shk1 hyperactivity. A mammalian homolog of Skb15 can substitute for its counterpart in S. pombe cells and inhibits PAK activity in mammalian cell cultures, indicating that Skb15-related protein functions have been substantially conserved through evolution (7,32).
While substantial progress has been made toward elucidating mechanisms of Shk1 regulation in S. pombe, downstream targets of the Shk1 kinase have not yet been described. Previous studies have demonstrated that Shk1 is a regulator of actin cytoskeletal organization in S. pombe (3,18,33). Newly divided S. pombe cells grow initially in a monopolar fashion from their "old ends." Subsequently, via a process referred to as new end takeoff (NETO), they initiate growth from the new end created by the cell division and continue to grow in a bipolar fashion until the onset of mitosis (34). Cortical F-actin patches are localized to the growing ends of S. pombe cells, initially to the old ends, then to both ends after NETO (35). Loss of Shk1 function, either by deletion of the shk1 gene or by overexpression of a kinase defective Shk1 mutant protein, results in the formation of spheroidal cells that exhibit a largely random distribution of cortical F-actin patches (3, 10). 2 By contrast, S. pombe cells rendered hyperactive in Shk1 function as a result of loss of function of the essential Shk1 inhibitor, Skb15, are often branched and/or hyperelongated, and exhibit hyperaccumulations of F-actin in the cytoplasmic space, septumforming region, and/or cell tips (7). An interesting hypomorphic allele of shk1, orb2-34, was identified by Verde et al. (18,33). S. pombe cells carrying the orb2-34 mutation are viable but monopolar for growth and likewise localize cortical F-actin patches only to the single growing cell tip (36). Experiments analyzing the localization of cortical cell markers in orb2-34 cells suggest that Shk1 may play a role in the process whereby fission yeast cells recognize their ends as sites for growth (36). In addition to regulating F-actin cytoskeletal organization, Shk1 is also required for proper organization of both the interphase microtubule cytoskeleton and the mitotic microtubule spindle in S. pombe cells (8). As microtubules play an important role in the establishment of cell polarity in S. pombe (37,38), it is possible that the regulation of polarized growth by Shk1 involves both actin and microtubule-dependent processes.
Consistent with its roles as a regulator of cytoskeletal organization and cell polarity, the Shk1 protein has been shown to localize to the cell ends, septum-forming region, and microtubules in S. pombe cells (8). This pattern of subcellular localization overlaps with that of other known regulators of polarized growth in S. pombe, including the kelch repeat protein, Tea1 (37). While not deleterious to cell growth, loss of Tea1 results in defects in actin and microtubule cytoskeletal organization and cell morphology (33,37). Similar to S. pombe mutants carrying the orb2-34 allele of shk1, cells carrying a null mutation in the tea1 gene exhibit a monopolar growth defect (37). We and other investigators recently showed that a null mutation in the tea1 gene causes growth inhibitory phenotypes when combined with loss of function mutations in genes encoding components of the Ras/Cdc42/Shk1 complex in S. pombe, including ras1, scd1, and shk1 (8,39). These findings raised the possibility that the Ras/Cdc42/Shk1 pathway might interact functionally with Tea1. In this article, we provide evidence that Tea1 is a potential downstream substrate target of the Shk1 kinase.
Plasmids-The oligonucleotide primers 5Ј-AAACCCTCGAGTCTG-CAGATGTCC-CCTATACTAGGT and 5Ј-AGGAAGATCTGTCGACG-GATCCGATTTTGGAGGAT were used to amplify the glutathione Stransferase (GST) protein coding sequence using the plasmid pRP259 (42) as a template. The PCR product was digested with XhoI and BglII then ligated to XhoI-BamHI digested pREP4X (43), producing pREP4XGST, which was used for expressing GST fusion proteins from the thiamine repressible nmt1 promoter (43) in S. pombe cells. pREP3XTea1, which was used for overexpressing tea1 from the nmt1 promoter, was a gift from Eric Chang (Baylor College of Medicine). The oligonucleotide primers 5Ј-TGGTGGATCCTCTCTTTAAAGGAGTAAT-GTCT and 5Ј-TATAGAGCTCATCGTCGAATATTTACACTATGT were used to amplify the tea1 protein coding sequence by PCR using a pREP3XTea1 plasmid as a template. The tea1 PCR product was digested with BamHI and SacI for cloning into the corresponding sites of pREP4XGST and the BglII and SacI sites of pREP4XHA (31), thus producing pREP4XGST-Tea1 and pREP4XHA-Tea1, respectively, and by BamHI and Ecl136II for cloning into the BamHI and SmaI sites of pRP259 to generate the plasmid pGST-Tea1. pREP4XGST-Tea1 and pREP4XHA-Tea1 were used for expressing GST-Tea1 and triple hemagglutinin (HA)-Tea1, respectively, from the nmt1 promoter in S. pombe, while pGST-Tea1 was used to express GST-Tea1 in bacteria. pTrcHisShk1 has been described and was used to express polyhistidinetagged Shk1 (His 6 -Shk1) in bacterial cells (30).
Preparation of Yeast Cell Lysates, Immunoprecipitations, and in Vitro Kinase Assays-S. pombe cells transformed with pREP4XHA or pREP4XHA-Tea1 were grown in EMM containing 50 M thiamine (EMMϩthi) to mid-log phase, washed twice with EMM lacking thiamine (EMM-thi), then subcultured into EMM-thi and grown for 24 h prior to the preparation of cell lysates as described (23). HA immune complexes were isolated by incubating yeast lysate (1 mg of total protein) with 5 l of anti-HA monoclonal antibody 12CA5 ascites (44) and 25 l of packed protein A agarose beads (Roche Applied Science), which had been preequilibrated with YLB, on an orbital rotator for 2 h at 4°C. Immune complexes were washed three times with YLB and once with kinase buffer (50 mM HEPES, pH 7.4, 10 mM MgCl 2 , 2 mM MnCl 2 , 1 mM dithiothreitol). Kinase reactions were performed by resuspending 12.5 l of packed immune complex beads with 25 l of kinase buffer containing 20 M ATP, 0.4 Ci/l of [␥-32 P]ATP (6000 Ci/mmol)), and ϳ1 ng/l His 6 -Shk1 (in control reactions, His 6 -Shk1 was replaced by an equal volume of His 6 -Shk1 storage buffer), and incubating for 20 min at 30°C. Reactions were stopped by adding SDS-PAGE sample buffer and boiling for 3 min. Kinase reactions were resolved by SDS-PAGE and autoradiography. HA immune complexes were also resolved by SDS-PAGE, transferred to nitrocellulose, and subjected to immunoblotting analysis for the detection of HA-Tea1 protein. GST and GST-Tea1 proteins were expressed and purified from E. coli strain BL-21 using glutathione agarose beads as recommended by the manufacturer (Amersham Biosciences). His 6 -Shk1 was purified from BL-21 using nickel agarose beads according to the manufacturer's specifications (Invitrogen). In vitro kinase assays of purified bacterially expressed proteins were performed as previously described (30).
In Vivo 32 P Labeling Experiments-In vivo 32 P labeling experiments for the analysis of Tea1 phosphorylation in S. pombe cells were per-2 P. Yang and S. Marcus, unpublished results.

Tea1 Is a Potential PAK Substrate-Effector in Fission Yeast
formed essentially as described (45). Briefly, S. pombe wild-type and nmt1-shk1K415R cells transformed with either pREP4XGST or pREP4XGST-Tea1 were grown in EMM containing 50 M thiamine to mid-log phase, then subcultured into phosphate-free EMM supplemented with 1 mM phosphate and grown for about 3.5 generations. Cells were then resuspended at 2 ϫ 10 6 cells/ml in fresh EMM containing 50 mM phosphate. 5 ml of each cell suspension was incubated with 1 mCi of [ 32 P]orthophosphate (10 4 Ci/mmol) (NEX054, PerkinElmer Life Sciences), labeled for 4 h, and then mixed with 50 ml of unlabeled carrier cells. Radiolabeled cells were lysed and GST fusion proteins purified as previously described (17). GST proteins were resolved by SDS-PAGE followed by Coomassie Brilliant Blue staining and subsequent autoradiography of the gels.

Shk1
Directly Phosphorylates Tea1 in Vitro-Previous studies have shown that Tea1 is associated with high molecular weight protein complexes in S. pombe cells (37,46). To investigate whether Tea1 or Tea1-associated proteins are phosphorylated by the Shk1 kinase in vitro, we constructed the plasmid pREP4XHA-Tea1 for expressing HA-tagged Tea1 protein (HA-Tea1) in S. pombe cells. HA-Tea1 complexes were immunoprecipitated from S. pombe cell lysates (Fig. 1A) and the immune complexes were incubated in kinase reactions with recombinant His 6 -Shk1 protein, which was purified from bacterial cells. As shown in Fig. 1B, we observed that HA-Tea1 was more heavily phosphorylated in kinase reactions containing His 6 -Shk1 than in reactions lacking His 6 -Shk1. These results suggest that Tea1 is either phosphorylated directly by Shk1 or by a Tea1-associated protein kinase that is stimulated by Shk1. Shk1 kinase activity, as measured by autophosphorylation, was also potentially stimulated slightly by the Tea1 complex, although it is possible that partially degraded Tea1 phosphoproteins comigrated with Shk1 (see HA immunoblot in right panel). Additional phosphoprotein bands detected in the kinase reactions may likewise have been phospho-Tea1 degradation products.
We next performed kinase assays using purified bacterially expressed proteins to determine whether Tea1 is directly phosphorylated by Shk1. The plasmid pGST-Tea1 was constructed for expressing GST-Tea1 in bacterial cells. GST-Tea1 protein was purified from bacterial cell lysates and incubated in in vitro kinase reactions with bacterially expressed His 6 -Shk1. We observed that GST-Tea1, but not GST, was phosphorylated in kinase reactions containing His 6 -Shk1 but not in reactions lacking His 6 -Shk1 (Fig. 1B). The level of Shk1 kinase activity, as measured by autophosphorylation, was not markedly affected by GST-Tea1 in these experiments (Fig. 1B). These results demonstrate that Tea1 is a direct in vitro substrate of the Shk1 kinase. Tea1 Is Phosphorylated in S. pombe Cells in a Shk1-dependent Fashion-Experiments were performed to determine whether Tea1 is phosphorylated in S. pombe cells and, if so, whether its phosphorylation is dependent on Shk1. The plasmid pREP4XGST-Tea1 was constructed for expressing a GST-Tea1 fusion protein from the nmt1 promoter (43). Wild-type S. pombe cells expressing either GST or GST-Tea1 were metabolically labeled with [ 32 P]orthophosphate. 32 P-labeled cells were lysed, and GST fusion proteins were purified from the cell lysates and subjected to SDS-PAGE and autoradiography. As shown in Fig. 2A, GST-Tea1 but not GST, was labeled by 32 P in vivo, thus demonstrating for the first time that Tea1 is phosphorylated in S. pombe cells.
To determine whether Tea1 phosphorylation in vivo is dependent on Shk1, we expressed GST-Tea1 in an S. pombe mutant in which the endogenous shk1 gene is replaced by a sequence from which a kinase deficient form of Shk1, Shk1K415R, is expressed from a weak allele of the nmt1 promoter (8,27,43). Wild-type and shk1K415R cells expressing GST-Tea1 were labeled with 32 P and the degree of 32 P incorporation into GST-Tea1 in each strain was determined. As shown in Fig. 2B, the degree of 32 P labeling observed for GST-Tea1 was significantly greater in wild-type S. pombe cells than FIG. 1. Shk1 directly phosphorylates Tea1 in vitro. A, S. pombe wildtype cells transformed with pREP4XHA or pREP4XHA-Tea1 were grown in EMMϩthi, then subcultured into EMMthi and grown for 24 h at 30°C prior to lysing and isolating HA immune complexes. HA immune complexes were resolved by SDS-PAGE and subsequent immunoblotting to resolve the HA-Tea1 protein, which exhibited a molecular mass close to its predicted size of 133 kDa. B, HA control (Control) and HA-Tea1 immune complexes were incubated in kinase reactions either with or without His 6 -Shk1 protein, as indicated, then resolved by SDS-PAGE followed by autoradiography. HA-Tea1 was more heavily phosphorylated in kinase reactions containing His 6 -Shk1 than in reactions lacking His 6 -Shk1. C, GST (Ϸ10 g), GST-Tea1 (Ϸ1 g), and His 6 -Shk1 (Ϸ25 ng) fusion proteins purified from bacterial cells were incubated alone or together in in vitro kinase reactions as indicated and resolved by SDS-PAGE (6 or 12% gels) followed by autoradiography. GST-Tea1 but not GST was directly phosphorylated by Shk1. Shk1 autophosphorylation was not markedly affected by Tea1.

Tea1 Is a Potential PAK Substrate-Effector in Fission Yeast
in shk1K415R cells. These results demonstrate that Shk1 is required for Tea1 phosphorylation in vivo and, when considered together with the above results showing that Tea1 is directly phosphorylated by Shk1 in vitro, implicate Tea1 as a potential Shk1 substrate in S. pombe.
Genetic Evidence Implicating Tea1 as an Effector of Shk1 Function in S. pombe-The above molecular data implicate Tea1 as a potential downstream target of the Shk1 kinase. We reasoned that if Tea1 is a mediator of Shk1 function, then loss of Tea1 function might suppress phenotypic defects resulting from Shk1 hyperactivity. We therefore determined whether phenotypes associated with loss of the essential Shk1 inhibitor, Skb15, are affected by a null mutation in the tea1 gene. For our analyses, we utilized an S. pombe strain, nmt1-skb15, in which the skb15 gene is conditionally expressed from the thiaminerepressible nmt1 promoter (7). The nmt1-skb15 strain is indistinguishable from wild-type S. pombe cells on media lacking thiamine. However, on thiamine containing media, the nmt1-skb15 strain is inviable and recapitulates phenotypes of the skb15 null mutant (7). We constructed an nmt1-skb15 tea1⌬ mutant and compared its growth characteristics to that of wild-type, nmt1-skb15, and tea1⌬ strains in the presence and absence of thiamine. Strikingly, we found that at 35°C, the nmt1-skb15 tea1⌬ mutant grew about as well as wild-type S. pombe cells on thiamine-containing media (Fig. 3A). In contrast, the nmt1-skb15 mutant was inviable under the same conditions (Fig. 3A). nmt1-skb15 tea1⌬ cells did not grow on thiamine-containing media at 25 or 30°C, indicating that the tea1⌬ mutation is a temperature-dependent suppressor of Skb15 loss (data not shown). The temperature dependence of the suppression of loss of skb15 by tea1⌬ is consistent with our previous observation that tea1⌬ causes a temperature sensitive growth inhibitory phenotype when combined with a mutation that partially inactivates Shk1 (8). Thus, the effects of the tea1⌬ mutation on both loss and gain of Shk1 function are temperature-dependent.
When grown in thiamine-containing medium, nmt1-skb15 cultures display a variety of morphological defects, including a significant frequency of multiseptated, highly branched, severely deformed, and greatly enlarged and/or elongated cells (Fig. 3B, panel ii) (7). The frequency of enlarged cells, multiseptated cells, and cells with elongated branches was markedly lower in nmt1-skb15 tea1⌬ cultures than in nmt1-skb15 cul-tures (Fig. 3B, panel iii). However, the nmt1-skb15 tea1⌬ mutant still exhibited a fairly significant frequency "T"-shaped and otherwise morphologically aberrant cells similar to that observed in tea1⌬ cultures (Fig. 3B, panel iv), but which were not observed in cultures of wild-type S. pombe cells (Fig. 3B,  panel i). Indeed, nmt1-skb15 tea1⌬ cells were for the most part indistinguishable from tea1⌬ cells from a morphological standpoint.

FIG. 2. Tea1 phosphorylation is dependent on Shk1 function in S. pombe cells.
A, wild-type S. pombe cells expressing either GST or GST-Tea1 were labeled with [ 32 P]orthophosphate and then lysed as described ("Experimental Procedures"). Lysates were incubated with glutathione-agarose beads to purify GST fusion proteins, which were resolved by SDS-PAGE. GST-Tea1 samples were resolved on 6% SDS-PAGE gels, while GST samples were resolved on 12% gels. The gels were stained with Coomassie Brilliant Blue to visualize GST and GST-Tea1 (left panels) and exposed to film for autoradiography (right panels). 32 P was incorporated into GST-Tea1 but not into GST. B, wild-type and nmt1-shk1K415R cells expressing GST-Tea1, as well as untransformed wild-type cells (Control) were labeled with 32 P and analyzed for GST-Tea1 phosphorylation as described in A. The degree of 32 P incorporation into GST-Tea1 was markedly greater in wild-type cells than in nmt1-shk1K415R cells.

Tea1 Is a Potential PAK Substrate-Effector in Fission Yeast
Taken together, the above results demonstrate that cell growth and morphological defects caused by loss of Skb15 function are dependent on the presence of Tea1. As Skb15 is an inhibitor of Shk1 and genetic and molecular data suggest that phenotypes associated with loss of Skb15 function are attributable in toto to hyperactivation of Shk1, these results implicate Tea1 as a potential downstream effector of the Shk1 kinase.
Tea1 Potentially Mediates Multiple Shk1 Functions-Cytological analyses were performed to assess the degree to which the tea1⌬ mutation suppresses cytoskeletal, mitotic, and cytokinesis defects resulting from loss of Skb15 function. Thiamineinduced repression of skb15 expression in nmt1-skb15 mutant cultures results in severe abnormalities in F-actin organization (7). Most notably, these include cells exhibiting dense accumulations of F-actin throughout much or most of the cytoplasmic space, the septum-forming region, and/or the cell tips (Fig. 4B). In addition, a small percentage of cells (about 1%) exhibit dense cable-like and/or circular F-actin structures (see circled cell in panel B and cells in small panels to right of panel B). We found that the tea1⌬ mutation strongly suppressed F-actin defects of the nmt1-skb15 mutant at 35°C (Fig. 4C). Indeed, the nmt1-skb15 tea1⌬ mutant was phenotypically similar to the tea1⌬ mutant in that interphase cells of both strains exhibited a primarily monopolar pattern of cortical F-actin localization (Fig. 4, C and D).
In addition to defects in F-actin organization, nmt1-skb15 cultures exhibit a high frequency of cells (Ͼ40%) with significant hyperaccumulations of cell wall material (chitin-rich deposits), particularly at the septum-forming region (Fig. 4, E and F) (7). As in the case of actin cytoskeletal defects, we found that the tea1⌬ mutation strongly suppressed cell wall/septum hypersynthesis defects of the nmt1-skb15 mutant. The degree of suppression was not as complete as was observed for suppression of F-actin defects. However, in this regard, we observed that the tea1⌬ mutant also exhibited a low frequency of cell wall defects similar to those observed in the nmt1-skb15 mutant (Fig. 4F). These results implicate Tea1 as a mediator of Shk1 functions involved in regulating septum and cell wall biogenesis in S. pombe cells.
nmt1-skb15 cells also exhibit severe mitotic defects in thiamine containing media (7). In comparison to wild-type S. pombe cells (Fig. 5A), mitotic spindles in nmt1-skb15 cells are often hyperelongated and/or attached to multiple fragments of chromosomal material (Fig. 5B) (7), the latter phenotype being indicative of defective nuclear segregation. These defects were not observed in nmt1-skb15 tea1⌬ cultures, as judged by microscopic examination of several thousand cells stained to simultaneously visualize microtubules and DNA by fluorescence microscopy (data not shown). Ultimately, nmt1-skb15 cultures reach a high state of aneuploidy (Fig. 5, C and D). As with other phenotypic defects associated with the loss of Skb15 function, we found that the tea1⌬ mutation strongly suppressed the aneuploid phenotype of the nmt1-skb15 mutant (Fig. 5D). These data suggest that Tea1 may also share a role with Shk1 in the regulation of mitotic processes in S. pombe cells.
Evidence That Shk1 and Tea1 Cooperate to Regulate Cytokinesis in S. pombe-In our microscopic analyses of tea1⌬ cultures, we noticed a low frequency of cells (Ͻ2%) with multiple septa (Fig. 6, A and B). A similar low frequency of multiseptated cells was also observed in nmt1-shk1K415R cultures (Fig.  6B), but not in shk1 mutant strains that are less severely impaired for Shk1 function, such as orb2-34 (data not shown). A role for Shk1 in cytokinesis was previously suggested from studies by Ottilie et al. (3). These investigators showed that shk1⌬ cells that conditionally express an N-terminal truncated Shk1 protein (Shk1⌬N276) arrest as round cells containing a septum and two nuclei when Shk1⌬N276 expression is repressed, thus indicating a potential cell cycle block in a late stage of cytokinesis. To investigate whether Shk1 and Tea1 might cooperate to regulate cytokinesis in S. pombe, we determined the consequences of combining a tea1⌬ mutation with the nmt1-shk1K415R mutation. As was previously observed for an adh1-shk1K415 tea1⌬ mutant (shk1K415R is expressed from the strong adh1 promoter) (8), the nmt1-shk1K415R tea1⌬ mutant was temperature sensitive for growth (data not shown). However, unlike the adh1-shk1K415 tea1⌬ strain, we found that the nmt1-shk1K415R tea1⌬ mutant exhibited a high frequency of cells with multiple septa, in comparison to the relatively low frequencies of multiseptated cells observed in nmt1shk1K415R and tea1⌬ cultures (Fig. 6, B and C). These results provide genetic evidence that Shk1 and Tea1 cooperate to regulate one or more processes necessary for the completion cytokinesis in S. pombe cells. . nmt1-skb15 cultures exhibit a high frequency of cells with hyperaccumulation of cell wall material in the cytoplasm, septum-forming region, and/or cell ends. These defects were not observed in wild-type cultures and found at only a low frequency in tea1⌬ and nmt1-skb15 tea1⌬ cultures (see panel F). F, quantitation of the percentage of cells exhibiting hyperaccumulation of cell wall material in wild-type, nmt1-skb15, tea1⌬, and nmt1-skb15 tea1⌬ strains.

Tea1 Is a Potential PAK Substrate-Effector in Fission Yeast
In light of the above findings, we decided to reexamine shk1⌬ cells for cytokinesis defective phenotypes. As shk1 is an essential gene, we generated an S. pombe culture enriched for shk1⌬ cells by inducing a shk1 ϩ /shk1::ura4 diploid culture to undergo sporulation. We then transferred the sporulated culture to fresh minimal medium lacking uracil to induce spore germination and to select for the growth of shk1::ura4 cells (shk1⌬ spores germinate as round cells that typically undergo from one to a few rounds of cell division before arresting growth and eventually lysing, Ref. 10). Microscopic analysis of DAPIstained cells revealed that nearly 60% of shk1⌬ cells contained at least one septum and two nuclei, in contrast to wild-type S. pombe cultures, in which only about 20% of cells were septated (Fig. 6D). The percentage of septated shk1⌬ cells may have actually been higher, as in some cells, the septum may have been parallel to the viewing plane and, therefore, not detectable. Significantly, we observed that more than 20% of septated shk1⌬ cells actually had multiple septa (Fig. 6D), which is a clear indication of a cytokinesis defective phenotype. These results are consistent with the observations of Ottilie et al. (3), which suggested that Shk1 might be required for a late stage(s) of the cytokinesis program in S. pombe.
Analysis of the Monopolar Growth Defect of S. pombe Cells Carrying the orb2-34 Mutant Allele of shk1-S. pombe mutants carrying the orb2-34 allele of shk1 are unable to activate a bipolar phase of growth after cell division, as is the case for cells carrying a deletion of the tea1 gene. tea1⌬ cells have been shown to exhibit an interesting pattern of monopolar growth in which one daughter cell typically initiates growth from its old end after cell division, while the other daughter cell initiates growth from the new end created by the cell division (46). As the growth characteristics of the orb2-34 mutant have not been described in detail, we performed time-lapsed photomicroscopy to determine whether this mutant exhibits a pattern of monopolar growth similar to that of tea1⌬ cells. Whereas in the case of wild-type S. pombe cells, both daughter cells invariably initiated growth from the old cell end after cell division, for the orb2-34 mutant, we observed in the vast majority of cases (97%) that one daughter cell initiated growth from the old cell end, while the other daughter cell initiated growth from its new end (Fig. 7, A-C).
In the case of wild-type cells, the majority of daughter cells (89%) proceeded to activate NETO and grew bipolarly until the next cell division, whereas orb2-34 cells never initiated a bipolar phase of growth (Fig. 7C). These observations demonstrate that the orb2-34 mutant exhibits a pattern of mo-FIG. 5. Suppression of mitotic defects in nmt1-skb15 cells by the tea1⌬ mutation. Mitotic spindle and associated DNA in a wildtype S. pombe cell (A) and an nmt1-skb15 mutant cell (B) from cultures grown in YEAU at 35°C for 40 h. The spindle (red) was visualized by indirect immunofluorescence of microtubules, while DNA (blue) was visualized by direct fluorescence using DAPI. Spindles were frequently found to be hyperelongated in mitotic nmt1-skb15 cells and often attached to multiple fragments of chromosomal material. These defects were not observed in wild-type, tea1⌬, or nmt1-skb15 tea1⌬ cultures (data not shown). C, fluorescence photomicrograph of nmt1-skb15 cells stained with DAPI to visualize DNA. A high degree of aneuploidy was apparent in nmt1-skb15 cultures but not in wild-type (inset panel and panel D) or tea1⌬ cultures (panel D). A low frequency of aneuploidy was observed in nmt1-skb15 tea1⌬ cultures (panel D). D, quantitation of the percentage of aneuploid cells in wild-type, nmt1-skb15, tea1⌬, and nmt1-skb15 tea1⌬ cultures.

Tea1 Is a Potential PAK Substrate-Effector in Fission Yeast
nopolar growth that strongly resembles that observed in tea1⌬ mutants, suggesting that the Shk1 and Tea1 proteins may participate in the regulation of a common process(es) involved in the regulation of cell polarity in S. pombe.
Overexpression of tea1 Is Inhibitory to the Growth of shk1deficient S. pombe Cells-We next determined the consequences of overexpressing tea1 in shk1-defective S. pombe cells. Wild-type and nmt1-shk1K415R cells were transformed with a tea1 overexpression plasmid, pREP3XTea1, in which the Tea1 protein coding sequence is expressed from the nmt1 promoter, or with the control plasmid, pREP3X. As shown in Fig. 8, tea1 overexpression had only a modest inhibitory effect on the growth of wild-type S. pombe cells. In contrast, it was highly inhibitory to the growth of nmt1-shk1K415R cells. Microscopic analyses revealed a high incidence of multiseptated cells in tea1 overexpressing nmt1-shk1K415R cultures similar to that detected in nmt1-shk1K415R tea1 (data not shown). Thus, overproduction of Tea1 has cytotoxic effects on the nmt1-shk1K415R mutant similar to those caused by the tea1 null mutation. A possible explanation for these results is that the high levels of non-phosphorylated Tea1 protein that are produced when tea1 is overexpressed in shk1 defective cells result in dominant inhibitory effects on processes regulated by the Shk1 and Tea1 proteins. These results provide additional genetic evidence for functional interaction between the Shk1 and Tea1 proteins. DISCUSSION An important goal in understanding the function of Ras/ Cdc42/PAK complex in fission yeast is to identify the pertinent substrates of the PAK kinase. In this article, we have provided evidence that the kelch repeat protein, Tea1, is a potential substrate target of the PAK kinase, Shk1. Both proteins regulate cell polarity and are localized to the cell ends, septum, and microtubules. Tea1 was found to exist as a phosphoprotein in S. pombe cells and its phosphorylation was substantially dependent on the presence of Shk1. Tea1 was also directly phosphorylated by Shk1 in vitro, suggesting that it is likely to be a direct substrate of Shk1 in S. pombe cells. Genetic interactions support this molecular interaction and place Tea1 downstream of Shk1. First, a null mutation in the tea1 gene strongly suppresses phenotypes of S. pombe cells rendered hyperactive for Shk1 function as a consequence of loss of the essential Shk1 inhibitor, Skb15. Second, a strain carrying the tea1 null mutation together with a mutation that severely attenuates Shk1 function causes a synthetic interaction. Additional evidence suggesting that Shk1 and Tea1 may regulate one or more common processes involved the regulation of cell polarity comes from our observations that S. pombe cells carrying a non-lethal hypomorphic allele of shk1, orb2-34, exhibit a pattern of monopolar growth that strongly resembles that resulting from the tea1⌬ mutation. Although our attempts to coprecipitate Shk1 and Tea1 proteins from S. pombe cells have not been successful, it is possible that these proteins associate only transiently, as might be expected for a protein kinase-substrate interaction. Nevertheless, our molecular and genetic data provide strong  8. Overexpression of tea1 is highly inhibitory to the  growth of nmt1-shk1K415R cells. Wild-type and nmt1-shk1K415R cells transformed with pREP3X or pREP3XTea1 were grown on EMMϩthi, subcultured onto EMM-thi, then streaked onto EMM-thi and incubated for 4 days prior to scanning the plates. tea1 overexpression was highly inhibitory to the growth of nmt1-shk1K415R cells, but only modestly inhibitory to wild-type cell growth. evidence that Tea1 is downstream target of Shk1.
While our findings provide compelling evidence implicating Tea1 as a biological substrate and effector of Shk1, the mechanistic contribution of Shk1 phosphorylation to Tea1 function remains to be determined. Roles for Shk1 phosphorylation in regulating the subcellular localization and/or protein-protein interactions of Tea1 can be envisaged. Indeed, in a particular shk1 mutant strain, orb2-34, Tea1 is sometimes unevenly distributed at the cell poles, suggesting a potential role for Shk1 in regulating Tea1 localization (37). Tea1 has been shown to associate with the actin binding protein, Bud6, in a large protein complex (46). Like Shk1 and Tea1, Bud6 is localized to the cell ends and septum-forming region of S. pombe cells (46). It is reasonable to speculate that Tea1, through its interaction with Bud6, mediates actin cytoskeletal organizing functions of Shk1. It will be interesting to determine whether Shk1 phosphorylation defective Tea1 mutant proteins exhibit defects in subcellular localization, complex formation with Bud6, or the ability to properly regulate cytoskeletal organization and cell polarity. It is possible that Shk1 and Tea1 have a dynamic interaction in which both proteins contribute to function(s) of the other. For example, Shk1 localizes to illicit growth tips produced by tea1 mutant cells, suggesting that Tea1 may contribute to the regulation of Shk1 localization (8). Although mechanistic details remain to be elucidated, our findings provide an important insight into what are likely to be highly complex mechanisms by which the Ras/Cdc42/Shk1 complex regulates cytoskeletal organization, cell growth polarization, and cytokinesis in fission yeast.
The findings of this study suggest that Tea1 is likely to have more diverse functions than previously suspected. Consistent with a role for Tea1 in cytokinesis, tea1 mutants were observed to have a mild cytokinesis defective phenotype (e.g. multiseptated cells), and tea1 shk1 mutants (this study), as well as tea1 ras1 and tea1 scd1 double mutants (39), exhibit very strong cytokinesis defective phenotypes. A role for Shk1 in cytokinesis was originally suggested from studies by Ottilie et al. (3) and has been substantiated by the findings of this study. The mechanistic contribution(s) of Shk1 and Tea1 to cytokinesis remain to be determined. Our results suggest shk1⌬ and nmt1-shk1K415R tea1⌬ cells are defective for a later stage of cytokinesis after the initiation of septum synthesis. It is possible that these mutants do not produce fully functional septa and/or that they are defective in the final cell separation process, which occurs after the termination of septum synthesis (47). Our observation that the tea1 null mutation strongly suppresses mitotic defects caused by loss of Skb15 suggest that Tea1 may also function in mitosis.
Shk1 has multiple cellular functions and is essential for the viability of S. pombe cells. Because it is a nonessential protein, if Tea1 is indeed a Shk1 effector, as our findings suggest, then it cannot be the only Shk1 effector in S. pombe. Nevertheless, our results showing that loss of Tea1 function causes a growth inhibitory phenotype when combined with mutations that partially inactivate Shk1 suggest that Tea1 contributes to an essential function(s) of Shk1. If this is the case, then Tea1 is likely to share overlapping essential function(s) with at least one other Shk1 effector. Alternatively, Shk1 and Tea1 may function as part of a dynamic, multiprotein complex, which as a whole is essential for cell viability, but which contains at least some individual components, such as Tea1, that are dispensable under normal growth conditions. Interestingly, a second Tea1-related kelch repeat protein, Tea3, has recently been identified in S. pombe (48). Like Tea1, Tea3 is a nonessential protein that contributes to the regulation of both cell polarity and cytokinesis. It will be interesting to determine whether Tea1 and Tea3 share overlapping and perhaps essential functions that will account for the important multiple functions of the Ras/Cdc42/PAK complex in fission yeast.
In animal cells, kelch repeat proteins have been found at focal adhesions (49) as well as ring canals, which are actinbased structures derived from arrested cleavage furrows (50). PAKs have likewise been shown to associate with focal complexes (51) and, more recently, to the cleavage furrow in animal cells (6). Furthermore, recent work in the amphibian, Xenopus laevis, has provided evidence that PAKs, in addition to their established roles as actin cytoskeletal regulators, also contribute to the regulation of microtubule dynamics in animal cells, as is the case in fission yeast (6). It will be of great interest to determine whether Tea1-like kelch repeat proteins mediate functions of PAKs involved in regulating actin and/or microtubule-dependent processes in higher eukaryotes.