Report

SBF Cell Cycle Regulator as a Target of the Yeast PKC-MAP Kinase Pathway

See allHide authors and affiliations

Science  21 Mar 1997:
Vol. 275, Issue 5307, pp. 1781-1784
DOI: 10.1126/science.275.5307.1781

Abstract

Protein kinase C (PKC) signaling is highly conserved among eukaryotes and has been implicated in the regulation of cellular processes such as cell proliferation and growth. In the budding yeast, PKC1 functions to activate the SLT2(MPK1) mitogen-activated protein (MAP) kinase cascade, which is required for the maintenance of cell integrity during asymmetric cell growth. Genetic studies, coimmunoprecipitation experiments, and analysis of protein phosphorylation in vivo and in vitro indicate that the SBF transcription factor (composed of Swi4p and Swi6p), an important regulator of gene expression at the G1 to S phase cell cycle transition, is a target of the Slt2p(Mpk1p) MAP kinase. These studies provide evidence for a direct role of the PKC1 pathway in the regulation of the yeast cell cycle and cell growth and indicate that conserved signaling pathways can act to control key regulators of cell division.

The coordination of polarized growth and cell proliferation is a critical process in many cell types. In yeast, the PKC1 MAP kinase cascade functions to maintain the integrity of the cytoskeleton and cell wall (1, 2) and is a candidate regulatory system for coupling cell growth and proliferation. The PKC pathway contains Pkc1p, which activates a MAP kinase cascade that consists of sequentially activated protein kinases: Bck1p(Slk1p), a MAP kinase extracellular signal-regulated kinase (MEK) kinase (MEKK), the redundant Mkk1 and Mkk2 MEKs, and Slt2p(Mpk1p), a MAP kinase (3). Genetic evidence has implicated this pathway in the regulation of cell cycle progression (46). In addition, the Slt2p MAP kinase is specifically activated during bud emergence and mating projection formation, periods of the cell cycle during which cell growth is highly polarized (7, 8).

The SBF transcription factor also regulates both the yeast cell cycle and polarized growth (9). SBF is a heterodimeric complex composed of the Swi4p and Swi6p proteins (10). SBF regulates the transition from G1 into S phase by activating the expression of the G1 cyclin genes CLN1, CLN2, PCL1, and PCL2 (1113). Cln and Pcl proteins associate with the Cdc28 (14) and Pho85 (13) cyclin-dependent kinases, respectively, to promote entry into S phase; Cln-Cdc28p complexes also promote polarized growth (15). Thus, both SBF and PKC function to mediate cell cycle progression and polarized growth.

We used a genetic screen (16, 17) to identify a mutation in the SWI4 gene, swi4-100, that requires SPA2 for growth (18). The Spa2 protein localizes to sites of polarized growth in yeast and is required for mating projection formation and proper bud site selection (19). Mutation of SLT2 or BCK1 is also lethal in the absence of SPA2 (16). Moreover, other similarities are observed between swi4Δ and SLT2 pathway mutants. Deletion of each kinase in the SLT2 cascade results in temperature-sensitive growth, and at 37°C cells die by lysis; this temperature-sensitive growth defect is suppressed by the osmotic stabilizer 1 M sorbitol (3, 16) (Fig. 1A). Haploid swi4Δ mutants are also temperature-sensitive for growth at 37°C in some backgrounds (11), and like SLT2 pathway mutants, this temperature sensitivity is suppressed by the presence of 1 M sorbitol (Fig. 1A). Also, swi4Δ mutants exhibit a defect in projection formation upon exposure to mating pheromone, as do SLT2 pathway mutants (5, 8, 16, 18).

Fig. 1.

Growth phenotypes of SLT2 pathway and swi4Δ mutants, and suppression of the slt2Δ mutant by cell cycle regulatory components. (A) Growth of WT, bck1Δ, slt2Δ, and swi4Δ strains at 37°C on either rich YPD (yeast extract, peptone, dextrose) plates (left) or YPD plates containing 1 M sorbitol (right). (B) Assay for the suppression of the slt2Δ (left) and bck1Δ (right) growth defect at 37°C by either YEp24 vector alone or high-copy number plasmids containing either SLT2, SWI4, PCL1, PCL2, PHO85, or MBP1 (22).

The phenotypic similarities between slt2Δ and swi4Δ mutants indicate that SLT2 and SBF may function in the same pathway. Because Swi4p and Swi6p are phosphoproteins in vivo (20) and the regulation of transcription factors through MAP kinase signaling has been shown in other systems (21), we investigated whether SBF might be activated by Slt2p (22). Plasmids containing either SWI4, SWI6, MBP1, PHO85, or G1 cyclin genes were transformed into both slt2Δ and bck1Δ strains, and transformants were assayed for suppression of the 37°C temperature-sensitive growth defect. Overexpression of SWI4 (23), PCL1, or PCL2 suppressed the 37°C growth defect (Fig. 1B). Overexpression of MBP1, PHO85 (Fig. 1B), CLN1, CLN2, CLN3, or SWI6 failed to rescue the growth defect of bck1Δ or slt2Δ mutants at 37°C. High-copy SLT2 plasmids were also transformed into swi4Δ mutants; these plasmids did not suppress the temperature-sensitive growth of swi4Δ strains at 37°C. Taken together, these observations support the hypothesis that SBF functions downstream of SLT2.

In addition to allowing SLT2 pathway mutants to grow at 37°C, overexpression of SWI4, PCL1, or PCL2 suppressed the cell lysis phenotype of these mutants both during vegetative growth and upon exposure to mating pheromone. After treatment with mating pheromone, wild-type cells arrest in late G1 and undergo a prolonged period of polarized growth to form a mating projection; exposure of slt2Δ and other pathway mutants to mating pheromone (24) caused a fraction (37%) of cells to lyse, and few cells formed wild-type projections (Fig. 2A) (5, 8). Overexpression of SLT2, PCL1, or PCL2 or addition of 1 M sorbitol to the medium suppressed both the lysis and projection formation defects of slt2Δ cells (Fig. 2, B, C, E, and F). Overexpression of SWI4 caused a decrease in cell lysis and a slight increase in the percentage of cells that formed wild-type projections (Fig. 2D). The slt2Δ strains overexpressing SWI4 had an increased percentage of budded cells in the presence of mating pheromone (Fig. 2), probably because of the increased expression of the CLN1 and CLN2 cyclin genes mediated by Swi4p (9, 11); Cln1p and Cln2p inhibit the pheromone response pathway (25). In summary, overexpression of SWI4 and the PCL G1 cyclin genes can suppress multiple phenotypes of SLT2 MAP kinase pathway mutants, but overexpression of SLT2 cannot rescue swi4Δ defects.

Fig. 2.

Suppression of the lysis or projection formation defects (or both) of slt2Δ mutants exposed to mating pheromone by overexpressed SWI4, PCL1, and PCL2 (24). (Top) Haploid MATa slt2Δ cells transformed with either YEp24 vector (A) or high copy plasmids containing SLT2 (B), SWI4 (D), PCL1 (E), or PCL2 (F) were treated with yeast α-factor. slt2Δ cells transformed with YEp24 were also grown in medium containing 1 M sorbitol (C) and exposed to pheromone. (Bottom) Quantitation of the percentage of cells (n ≥ 830 cells) from strains described above that display lysis and projection formation defects (24).

An examination of the various in vivo species of Swi6p further supported the hypothesis that SBF functions downstream of SLT2 and suggested that the SLT2 MAP kinase pathway may regulate the phosphorylation state of Swi6p (Fig. 3A). Immunoblot analysis with affinity-purified antibodies to Swi6p (anti-Swi6p) (26) revealed two major isoforms of Swi6p in lysates from log-phase cells (Fig. 3A). Phosphatase treatment (27) of Swi6p immunoprecipitates resulted in the loss of the slowly migrating species and an increase in the faster migrating isoform (Fig. 3A); the presence of the phosphatase inhibitor sodium molybdate prevented this conversion. Thus, the slowly migrating species is a phosphorylated form of Swi6p. A strain overexpressing SLT2 accumulated the slowly migrating phosphorylated Swi6p isoform, whereas only the faster migrating Swi6p isoform was detected in strains lacking SLT2 (Fig. 3A). Expression of a stable but inactive kinase allele, slt2K54R (7), in a slt2Δ strain did not increase accumulation of the slowly migrating isoform, indicating that the presence of phosphorylated Swi6p requires the kinase activity of Slt2p.

Fig. 3.

In vivo evidence for Slt2p-dependent phosphorylation of Swi6p and coimmunoprecipitation of Slt2p and SBF. (A) Immunoblot analysis of Swi6p. Extracts were prepared from the following strains (7): Left seven lanes: wild-type strain PZY143 (SLT2); PZY181 (2μ SLT2) that overexpresses SLT2; YK193 (slt2K54R), a slt2Δ mutant that overexpresses a kinase-inactive allele of SLT2; YK175 (swi6Δ); and Y783 (slt2Δ). Strains were either grown at 30°C or shifted to 39°C for 1 hour before preparation of extracts (28). Right four lanes: Wild-type strain CMY826 was used for pheromone treatment (α 1h, α 2h) and α-factor block and release experiments (15′ release), as described (7). Approximately 5 μg of total yeast protein was loaded for each sample. CIP treatment experiments (27) and immunoblot analysis (36) were done as described. (B) Coimmunoprecipitation of Swi6p and Swi4p with Slt2p. (Left) Immunoblot analysis of Swi6p IPs from extracts of either a swi6Δ strain containing high-copy plasmid SLT2:HA (7) or from SWI6 strains PZY143 (SLT2) and PZY181 (SLT2:HA) (7). (Middle) Immunoblot analysis of Swi4p IPs from extracts of swi4Δ SLT2:3HA cells, slt2Δ cells containing either high-copy SLT2 or SLT2:3HA plasmids, or an experiment in which no antibody to Swi4p was added. SLT2:3HA contains three copies of sequence encoding the HA epitope at the COOH-terminal coding segment of SLT2 in YEp352. (Right) Immunoblot analysis of total cellular protein (2 μg) from swi4Δ SLT2:3HA cells and slt2Δ cells containing either high-copy SLT2 or SLT2:3HA plasmids; similar results were observed with SLT2:HA cell lysates. Cell extracts were prepared from strains incubated at 39°C for 1 hour. For the Swi6p and Swi4p IP experiments, 2 mg of total yeast protein and either affinity-purified rabbit anti-Swi6p (26) or anti-Swi4p (37) were used. After electrophoresis, proteins were blotted onto Immobilon-P (Millipore), probed with monoclonal antibody 12CA5 (BABCO), and primary antibodies were detected with alkaline phosphatase-conjugated goat anti-mouse IgG and the BCIP-NBT color substrate reaction (35).

Swi6p phosphorylation correlates with the activity of Slt2p in vivo. Slt2p activity is stimulated by growth at increasing temperatures or by treatment of cells with mating pheromone (7, 8, 28). In cells grown at 30°C, the faster migrating form was the major Swi6p isoform detected, whereas incubation of cells at 39°C for 1 hour or exposure to mating pheromone resulted in the accumulation of phosphorylated Swi6p (Fig. 3A). The presence of intermediate species may indicate that several phosphorylation events occur. Cells exposed to a high concentration of pheromone for 2 hours and then washed and diluted into fresh medium for 15 min have low Slt2p activity (7) and also have an increase in the faster migrating isoform of Swi6p (Fig. 3A). Thus, for each of these different growth conditions, phosphorylation of Swi6p correlates with the level of Slt2p activity.

SBF was found to exist in a complex with Slt2p and therefore might be a direct target of Slt2p in vivo. Overexpression of SLT2 tagged with either a single (SLT2:HA) (7) or three copies (SLT2:3HA) of the hemagglutinin (HA) epitope complemented the 37°C growth defect of slt2Δ cells. Immunoblot analysis with a monoclonal antibody to the HA epitope detected Slt2:HAp as a 64-kD protein and Slt2:3HAp as a 68-kD protein (Fig. 3B). We used affinity-purified antibodies to immunoprecipitate Swi6p or Swi4p from lysates of slt2Δ cells that overexpressed SLT2, SLT2:HA, or SLT2:3HA. Immunoblot analysis with an antibody to HA revealed that Slt2:HAp was present in Swi6p immunoprecipitates (IP) (Fig. 3B). A small amount of Slt2:3HAp (and Slt2:HAp) also coimmunoprecipitated with Swi4p (Fig. 3B). The majority of Slt2:HAp or Slt2:3HAp was detected in the supernatant of the immunoprecipitation, indicating that only a small fraction of the kinase is associated with either Swi6p or Swi4p, as might be expected for a kinase-substrate interaction. Association of Swi4p and Swi6p with Slt2:HAp was not detected in lysates from swi4Δ or swi6Δ strains or from strains expressing SLT2 with no HA tag; furthermore, antibodies to the yeast Axl2 protein did not immunoprecipitate Slt2:HAp.

In vitro kinase assays indicated that Slt2p can directly phosphorylate Swi6p and Swi4p. Kinase assays with Slt2:3HAp (and Slt2:HAp) immunoprecipitated from yeast extracts exhibited phosphorylation of the immunoglobulin G (IgG) heavy chain, Slt2:3HAp itself, and a third phosphoprotein of ∼100 kD (Fig. 4A). Immunoprecipitated Slt2:3HAp also phosphorylated myelin basic protein (7, 8, 28). Addition of Swi6p, purified from bacterial extracts (26), to the Slt2:HAp kinase reaction resulted in Swi6p phosphorylation (Fig. 4B); several phosphorylated isoforms were observed. A phosphoprotein of the expected size of Swi6p was not detected in kinase reactions that lack Swi6p or in kinase assays done on the Swi6p extracts alone (Fig. 4B).

Fig. 4.

In vitro phosphorylation of SBF by Slt2p (29). (A) Autoradiograph of gel blot of samples from in vitro phosphorylation reactions containing Swi4p. Kinase reactions were done without immune complexes (lane 1), with immune complexes isolated from yeast extracts containing Slt2:3HAp (lanes 3, 5, and 7), or with immunoprecipitates from extracts containing untagged Slt2p (lanes 4 and 6). Either a Swi4p or Swi6p preparation from insect cell extracts (29) (lanes 1, 5, and 6) or a preparation in which Swi4p was immunodepleted (lanes 2 and 7) was added to the reactions. For lanes 3 and 4, the reactions lacked the Swi4p or Swi6p extracts. (B) Autoradiograph and immunoblot analysis of blots from two-dimensional gels containing Swi6p phosphorylated in vitro by Slt2p (38). (a) Immunoblot analysis with affinity-purified anti-Swi6p of kinase reactions containing Slt2:HA IP and Swi6p purified from bacterial extracts (26, 36). (b) Autoradiograph of the same blot that was probed with anti-Swi6p to generate (a). The left three Swi6p isoforms from (a) and (b) are superimposable. (c and d) Autoradiographs of kinase reactions containing either Slt2:HA IP immune complexes or Swi6p extract alone, respectively. Immunoprecipitations and kinase reactions are as described (29), except that 1.25 mg of total yeast protein from lysates of PZY181 cells that had been shifted to 39°C for 1 hour were used for kinase reactions. Swi6p-containing extracts (1.5 μg) (26) were added per kinase reaction.

Addition of partially purified insect cell extract containing Swi4p and Swi6p (29) to the Slt2:3HAp kinase assay resulted in the phosphorylation of a protein that migrated more slowly than Swi4p, yet appeared to be a phosphorylated form of Swi4p (Fig. 4A) (29); immunodepletion of Swi4p from the extracts before the kinase assay decreased the amount of this phosphoprotein (Fig. 4A). Thus, both Swi6p and Swi4p are substrates of Slt2:HAp in vitro.

To determine if Slt2p functions as a positive regulator of SBF activity, we performed RNA blot analysis (26) of four SBF-dependent genes, CLN1, CLN2, PCL1, and PCL2, with RNAs isolated from wild-type and slt2Δ strains (30). No differences in expression levels were detected for any of these genes in strains grown at either 25° or 30°C. However, when strains were incubated at 37°C for 3 hours, the amounts of PCL1 and PCL2 mRNAs in slt2Δ strains were one-half to two-thirds that of wild-type cells; this reduction is similar to the SLT2-dependent expression of several SBF-regulated genes involved in cell wall biosynthesis (23). No difference was detected in amounts of CLN1 or CLN2 mRNA. Thus, SLT2 apparently functions to activate SBF; this activation results in the increased overall expression of a subset of SBF-dependent genes.

Although our data indicate that Slt2p activates SBF, the observation that swi4 and pkc1 as well as swi4 and slt2 are colethal (18, 23) predicts that SBF and Slt2p may also have independent functions. Consistent with this interpretation, strains lacking SBF activity arrest predominantly in G1 (11, 12), whereas slt2Δ strains grown at 25°, 30°, or shifted to 37°C did not reveal an increase in the proportion of G1 cells (30); this suggests that other mechanisms may contribute to the regulation of SBF target genes. Substrates of Slt2p besides SBF may also exist; one candidate is the potential transcription factor Rlm1p (31).

Our results indicate that the activation of Slt2p results in phosphorylation and activation of the SBF transcription factor, enabling SBF to promote the expression of the PCL1 and PCL2 G1 cyclins and additional genes that regulate cell wall biosynthesis (23). The suppression of the slt2Δ temperature-sensitive growth defect by overexpressing PCL1 or PCL2 suggests a dual role for the Pcl-Pho85p complex in maintaining cell integrity and promoting cell cycle progression (13). In addition, the proper activation of Slt2p requires the Cdc28p kinase (7), indicating that the PKC1 MAP kinase pathway is both activated by and an activator of cyclin-dependent kinase signaling; this regulation should result in a rapid amplification of G1 kinase activity. The identification of SBF as a target of Slt2p offers insight into how PKC signaling regulates the G1 kinase activity required for coordinating entry into S phase and polarized growth events. Because PKC (32) and MAP kinases (21) function to activate transcription factors that control cell proliferation in other organisms, this mode of PKC regulation may be a general mechanism.

REFERENCES AND NOTES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
View Abstract

Navigate This Article