INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The pathway of meiosis and spore formation of the budding yeast Saccharomyces cerevisiae is initiated in response to nitrogen limitation. Underlying this pathway is a global change in gene expression: within 2 h after starvation, 58 genes are activated over fivefold (8). Many of these genes belong to the set of early meiotic genes, which are required for normal progression through meiosis (20). Expression of early meiotic genes depends on Ume6p and Ime1p, two subunits of a heteromeric transcriptional activator. Ume6p-Ime1p interaction is stimulated by nitrogen limitation (32), but the mechanism through which nitrogen governs interaction is unknown.

Rim11p, a relative of glycogen synthase kinase 3 (GSK3), is required for Ume6p-Ime1p interaction (25, 32). Rim11p phosphorylates both Ime1p and Ume6p in vitro within the interaction region of each protein, and multisite amino acid substitutions that abolish phosphorylation in vitro also disrupt complex formation (4, 24, 25). Thus, regulation of Rim11p-dependent phosphorylation of either substrate would provide a simple mechanism for regulation of Ume6p-Ime1p complex formation. In vitro measurements indicate that Rim11p specific activity is fourfold higher in acetate-grown cells than in glucose-grown cells but have not revealed any response of Rim11p to nitrogen limitation (25).

Rim11p is loosely related to Mck1p (44% identity), a second yeast GSK3 homolog that has diverse functions. Mck1p promotes growth at temperature extremes, diauxic adaptation, centromere segregation, and meiosis (6, 16, 27, 35). In meiosis, Mck1p has two known roles (2, 27). First, it promotes IME1 transcription and is thus required indirectly for Ume6p-Ime1p interaction. Second, it promotes spore maturation, which occurs at the end of the meiotic program. Puziss et al. found that overexpression of Rim11p improves growth of an mck1 mutant at low and high temperatures (30), thus suggesting that Rim11p and Mck1p might have partially overlapping functions. However, mck1 mutant phenotypes are unaffected by a rim11 null mutation, and it has thus seemed equally likely that Rim11p can substitute for Mck1p only when it is overexpressed.

In this study, we have examined the state of Ume6p phosphorylation in vivo. Our findings indicate that Rim11p and Mck1p act together to promote Ume6p phosphorylation and activity. The extent of Ume6p phosphorylation depends on nutrient availability and on the nutritional response protein kinase Rim15p (40), thus indicating that a nutritional control pathway governs phosphorylation by Rim11p and Mck1p. Our results predict a simple genetic interaction between rim11 and mck1 mutations that may provide a generally useful redundancy test for multifunctional protein kinases.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains, genetic methods, and media. S. cerevisiae strains were derived from the SK-1 genetic background (19) or from crosses between SK-1 strains and strain Y190 (10), as indicated in Table 1. Several mutations have been described previously, including rim11::LEU2 (4), ime120 (36), P_GAL1-IME1::TRP1 (38), ime24-lacZ (26), ime2-7::HIS3::LEU2 (37), and mck1::TRP1 (27). The mrk1::URA3-Kl mutation as introduced by PCR product-directed gene disruption using oligonucleotides MRK1.-60+kl.URA3 and MRK1.1860.3'+kl.URA3 (Table 2) and plasmid pWJ716 as a template for the Kluyveromyces lactis URA3 gene (11); PCR with outside primers MRK1.-104 and NotI-kl.URA3.1321-3' confirmed the genotype. Strain construction involved standard methods including mating, meiotic crosses, and transformation (18).

ume6

ume6

ura3 ume6

his3 ume6

Plasmids. (i) UME6 derivatives. Plasmid pKB186 carries the hemagglutinin epitope (HA)-tagged UME6-HA allele (25). A 4-kb SpeI-HindIII fragment from pKB186 containing UME6-HA was inserted into SpeI-HindIII-digested vector pRS406 to produce plasmid pYX148, an integrating UME6-HA plasmid. The plasmid carrying untagged Ume6p, pYX147, was created by the same strategy from plasmid pHY14-2 (5). UME6 plasmids with different point mutation were created by PCR-directed sequence alterations. To generate ume6-T99A (pYX285), first-round PCR products were generated with plasmid pYX148 as template and oligonucleotide UME6.-100 paired with UME6-T99A-3' and oligonucleotide UME6-Asp718 paired with UME6-T99A. Two PCR products were purified and used as template in the second-round PCR with oligonucleotides UME6.-100 and UME6-Asp718. The resulting 0.9-kb DNA fragment was purified, digested with SphI and NheI, and inserted into SphI-NheI-digested plasmid pYX147. An analogous strategy was used to create ume6-T103 (pYX286), ume6-S107A (pYX287), and ume6-Ala3 (pYX281). Presence of desired mutations and absence of secondary mutations were confirmed by sequencing the entire region that had been PCR amplified.

(ii) GBD and GAD fusions. Fusions of the Gal4p DNA binding domain (GBD) to full-length Rim11p and to Ime1p (residues 294 to 360) have been described elsewhere (25).

Two-hybrid interaction assays. Strains AMP1562 and Y190 (10) carrying GBD, GBD-Rim11p, or GBD-Ime1p(294-360) were mated with strain AMP1563 or Y187 (10) carrying GAD or GAD-Ume6p(1-232), and diploids were selected on SC-Trp-Leu plates. For assays, 24-h cultures in SC-Trp-Leu medium were diluted 1/25 into YPAc or SC-Trp-Leu medium and harvested after three doublings. -Galactosidase assays were conducted on permeabilized cells. Determinations were averages from three independent transformants; the range was less than 20% of the mean value.

Phosphatase treatment. Whole-cell extracts were prepared as described previously (25) in a mixture containing 20 mM Tris-HCl (pH 7.4), 5% glycerol, 100 mM NaCl, 100 mM KCl, 1 mM EDTA, 0.05% -mercaptoethanol and protease inhibitors (phenylmethylsulfonyl fluoride [0.15 mg/ml], leupeptin [1 µg/ml], aprotinin [1 µg/ml], and pepstatin [1 µg/ml]). Five milligrams of total protein extract was bound to 5 µg of 12CA5 and 200 µl of 50% protein A-Sepharose beads (Sigma) during a 1-h incubation at 4°C. The beads were washed twice with extraction buffer, washed once with phosphatase buffer (100 mM Tris-HCl [pH 9.6], 2 mM MgCl₂, 0.1 mM ZnCl₂, protease inhibitors), and suspended in 60 µl of phosphatase buffer; 15-µl aliquots of the suspension were mock treated (with buffer alone), phosphatase treated (with 2 U of calf intestinal phosphatase), or phosphatase treated in the presence of phosphatase inhibitors (5 mM sodium fluoride, 5 mM sodium phosphate, 10 mM sodium pyrophosphate, 1 mM NaVO₃, 5 mM EGTA, and 5 mM EDTA). After a 15-min incubation at 37°C, the reaction mixtures were boiled for 5 min with 10 µl of 3× Laemmli buffer.

Immunoblot analysis. Cells were harvested from mid-log phase in YPD or YPAc or after transfer from YPAc to sporulation medium for 3 h. The cell pellets were suspended at 10 units of optical density at 600 nm per ml of 3× Laemmli buffer and boiled for 5 min. We found that preservation of phosphorylation states was best with 5% sodium dodecyl sulfate (SDS) in the 3× Laemmli buffer. After centrifugation, supernatants were transferred to fresh tubes and loaded on SDS-polyacrylamide gels [6% for full-length Ume6p; 10% for GAD-Ume6p(1-232)] for polyacrylamide gel electrophoresis (PAGE). Protein transfer and detection with anti-HA monoclonal antibody 12CA5 (BabCo), goat anti-mouse antibody conjugated to peroxidase (Boehringer), and enhanced chemiluminescence detection reagents (Amersham) followed standard procedures (25).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Phosphorylation of Ume6p in vivo. To determine whether Ume6p is phosphorylated in vivo, we examined the electrophoretic mobility of Ume6-HAp. We identified Ume6-HAp through anti-HA immunoblots of UME6 and UME6-HA strains (Fig. 1A, lanes 1 to 3 and 4 to 6, respectively). Ume6-HAp migrated as a disperse protein, and its apparent size was affected by growth conditions: it was 130 kDa in glucose-grown cells (lane 4), 140 kDa in acetate-grown cells (lane 5), and 150 kDa in nitrogen-limited cells (lane 6). Ume6-HAp mobility was unaffected by cell type or IME1 expression (data not shown). The apparent size of Ume6-HAp was reduced to 120 kDa by treatment with phosphatase; absence of phosphatase or presence of phosphatase plus inhibitors had no effect (Fig. 1B). These results indicate that Ume6p is phosphorylated in vivo and that the extent of Ume6p phosphorylation is elevated in response to acetate medium and nitrogen limitation.

View larger version (66K): [in this window] [in a new window]   FIG. 1.   Phosphorylation of Ume6p in vivo. (A) Mobilities of wild-type and mutant Ume6p derivatives. Strain YX423 carrying integrated UME6 plasmids was grown in rich glucose medium (D; lanes 1, 4, 7, 10, 13, 16, and 19), rich acetate medium (Ac; lanes 2, 5, 8, 11, 14, 17, and 20), or acetate medium lacking nitrogen (sporulation medium [Sp]; lanes 3, 6, 9, 12, 15, 18, and 21). Migration of Ume6-HAp on SDS-PAGE was visualized on an anti-HA immunoblot. The UME6 plasmids specified Ume6p (lanes 1 to 3), wild-type Ume6-HAp (lanes 4 to 6), or mutant Ume6-HAp derivatives with substitutions specified above each set of three lanes (lanes 7 to 21). (B) Phosphatase treatment of Ume6-HAp. Anti-HA immune complexes were prepared from strain YX423 carrying a plasmid specifying Ume6-HAp and grown in rich glucose medium (lanes 1 to 4), rich acetate medium (lanes 5 to 8), or acetate medium lacking nitrogen (lanes 9 to 12). Migration of Ume6-HAp on SDS-PAGE was visualized on an anti-HA immunoblot. Prior to SDS-PAGE, the immune complexes were untreated (lanes 1, 5, and 9) or incubated at 37°C with no addition (lanes 2, 6, and 10), with phosphatase (lanes 3, 7, and 11), or with phosphatase plus phosphatase inhibitors (lanes 4, 8, and 12). (C) Effect of GSK3 homolog defects on Ume6-HAp mobility. SDS-PAGE mobility of Ume6-HAp was examined on an immunoblot of strains YX423 (wild type; lanes 1 to 3), YX425 (rim11; lanes 4 to 6), YX426 (mck1; lanes 7 to 9), YX427 (mrk1; lanes 10 to 12), YX428 (rim11 mck1; lanes 13 to 15), YX429 (mck1 mrk1; lanes 16 to 18), YX431 (rim11 mrk1; lanes 19 to 21), and YX424 (rim11 mck1 mrk1; lanes 22 to 24). Growth conditions and symbols are as in panel A. Sizes are indicated in kilodaltons.

View larger version (34K): [in this window] [in a new window]   FIG. 2.   Phosphorylation of the Ume6p NTR in vivo. (A) Effect of GSK3 homolog defects on mobility of GAD-Ume6p(1-232). Strains carrying a GAD-Ume6p(1-232) plasmid were grown under conditions described in the legend to Fig. 1A, and migration of GAD-Ume6p(1-232) on SDS-PAGE was visualized on an anti-HA immunoblot. The strains had defects in Rim11p, Mck1p, or Mrk1p, as indicated at the top, and are specified in the legend to Fig. 1C. (B) Phosphatase treatment of GAD-Ume6p(1-232). Anti-HA immune complexes were prepared from wild-type strain AMP107 carrying plasmids specifying GAD (lanes 1 to 3), GAD-Ume6p(1 to 232) (lanes 4 to 6), or GAD-Ume6-Ala3p(1-232) (lanes 7 to 9) after growth in acetate medium lacking nitrogen and were visualized after SDS-PAGE on an anti-HA immunoblot. Prior to SDS-PAGE, the samples were incubated at 37°C with no addition (lanes 1, 4, and 7), with phosphatase (lanes 2, 5, and 8), or with phosphatase plus phosphatase inhibitors (lanes 3, 6, and 9). Sizes are indicated in kilodaltons.

Functional activity of Ume6p phosphorylation-negative mutants. Ume6p functions as both a repressor and an activator. In vegetative cells, it binds to Sin3p and Rpd3p to repress many genes, including early meiotic genes (3, 17, 28, 39). In meiotic cells, it binds to Ime1p to activate early meiotic genes (5, 25, 32). To determine whether these functions are governed by NTR phosphorylation, we assayed repression and activation of the meiotic IME2 promoter in strains expressing wild-type and phosphorylation-defective Ume6p derivatives.

ime1

ime1

ime1

ime1

Role of the yeast GSK3 family in phosphorylation. Rim11p immune complexes can phosphorylate the Ume6p NTR in vitro (25). However, we found that a rim11 null mutation had little effect on Ume6-HAp phosphorylation under any growth condition (Fig. 1C, lanes 4 to 6 compared to lanes 1 to 3). Therefore, Rim11p is not required for Ume6p phosphorylation in vivo.

Interaction of Ume6p with Mck1p. Rim11p and Mck1p may phosphorylate the Ume6p NTR directly and may thus interact with the Ume6p NTR. Rim11p-Ume6p NTR interaction is detectable through two-hybrid assays (25), and so we used this approach to test for Mck1p-Ume6p interaction. Mck1p-Ume6p NTR interaction was detectable and above background levels in quantitative assays (Table 4) and on 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-Gal) plates (data not shown). No interaction was detected between Mck1p and the Ime1p C-terminal region (residues 294 to 360), which does interact with Rim11p (4, 24, 32). We also detected no interaction between Mrk1p and the Ume6p NTR (data not shown). These results indicate that both Rim11p and Mck1p are capable of direct or indirect interaction with the Ume6p NTR.

Shared roles of Rim11p and Mck1p in meiosis. Our findings indicate that Rim11p and Mck1p are together responsible for Ume6p NTR phosphorylation in nitrogen-limited cells and that this phosphorylation is required for activation of meiotic genes and sporulation. This model predicts that rim11 mck1 double mutants may have a more severe sporulation defect than either single mutant. The assessment is complicated because rim11 null mutants are completely defective in sporulation, as expected from their defect in Ime1p phosphorylation (4, 24). Therefore, we used strains that express a partially defective Rim11p derivative, Rim11-K68Rp, with an ATP binding site alteration. Rim11-K68Rp causes substantially reduced protein kinase activity in vitro, yet strains that express Rim11-K68Rp are capable of efficient sporulation (4). We reasoned that the low level of Rim11-K68Rp kinase activity may be adequate for sporulation only because of the contribution from other protein kinases to Ume6p phosphorylation. We found that Rim11p and Rim11-K68Rp stimulated sporulation to similar levels in strains that express Mck1p (Table 5, strains YX504 and YX502). However, only Rim11p stimulated sporulation in the absence of Mck1p; Rim11-K68Rp did not (strains YX503 and YX501). Sporulation was dependent on Rim11p protein kinase activity in these assays, because the kinase-inactive Rim11-K68Ap did not permit sporulation in any strain. The third GSK3 homolog, Mrk1p, had little effect on sporulation with any Rim11p derivative (strains YX502 and -501 compared to strains YX504 and -503). Therefore, Mck1p can compensate for a reduction in Rim11p activity to promote sporulation. These results support the idea that Rim11p and Mck1p act interchangeably to promote sporulation.

Dependence of Ume6p phosphorylation on Rim15p. Prior studies indicate that Rim15p stimulates Ume6p-Ime1p interaction (41). To determine whether Rim15p is required for Ume6p phosphorylation, we compared Ume6-HAp forms in wild-type and rim15 mutant strains. The rim15 mutation caused a reduction in accumulation of phosphorylated forms of both Ume6-HAp and GAD-Ume6p(1-232) (Fig. 3). Therefore, Rim15p promotes phosphorylation of the Ume6p NTR.

View larger version (27K): [in this window] [in a new window]   FIG. 3.   Effect of a rim15 mutation on Ume6-HAp and GAD-Ume6p mobility. SDS-PAGE mobility of Ume6-HAp (lanes 1 to 4) and GAD-Ume6p(1-232) (lanes 5 to 8) expressed in strains AMP107 (RIM15; lanes 1, 2, 5, and 6) and AMP1631 (rim15; lanes 3, 4, 7, and 8) was visualized on an immunoblot. Growth conditions and symbols are as for Fig. 1. Sizes are indicated in kilodaltons.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Formation of the Ume6p-Ime1p complex is tightly regulated by nitrogen limitation and by the protein kinase Rim11p (32). Here we have shown that these regulatory signals converge to govern phosphorylation of the Ume6p NTR. Our analysis further indicates that Mck1p and Rim11p both contribute to Ume6p NTR phosphorylation under conditions of nitrogen limitation. The functional significance of Ume6p NTR phosphorylation is supported by the synthetic sporulation defect of the rim11-K68R mck1 double mutant and by the sporulation defects caused by Ume6p substitutions that block phosphorylation. Our results are consistent with a model in which nitrogen limitation promotes Ume6p-Ime1p interaction through increased Ume6p NTR phosphorylation (Fig. 4). Growth in the presence of nitrogen results in low levels of Ume6p NTR phosphorylation, thus precluding Ume6p-Ime1p interaction. Limitation for nitrogen causes elevated Ume6p NTR phosphorylation, thus favoring Ume6p-Ime1p interaction, meiotic gene expression, and sporulation.

View larger version (14K): [in this window] [in a new window]   FIG. 4.   Control of Ume6p NTR phosphorylation. Ume6p exists in one of two states: without NTR phosphorylation or with NTR phosphorylation. NTR phosphorylation is greatest in nitrogen-limited cells and permits formation of an Ume6p-Ime1p complex and activation of early meiotic genes. NTR phosphorylation may be carried out directly by the GSK3 homologs Rim11p and Mck1p; Ime1p phosphorylation is carried out only by Rim11p. NTR phosphorylation also depends on the protein kinase Rim15p, a downstream target of the adenylate cyclase pathway.

Functional roles of Rim11p and Mck1p in meiosis. The possibility that Rim11p and Mck1p have overlapping functional roles was first suggested by Puziss et al., who found that overexpression of RIM11 can suppress a mitotic mck1 mutant defect (30). However, the lack of a synthetic rim11 mck1 growth defect suggested that overexpression may permit Rim11p to substitute adventitiously for Mck1p. We have presented three findings that argue that Rim11p and Mck1p naturally share a function in nitrogen regulation of meiosis. First, rim11 mck1 double-null mutants are defective in accumulation of phosphorylated GAD-Ume6p after nitrogen limitation. Second, Mck1p interacts with the Ume6p NTR, as we found previously for Rim11p (25). Third, sporulation of the partially defective rim11-K68R mutant is abolished by an mck1 mutation. There is no novel sporulation defect of a rim11 mck1 double-null mutant because Rim11p is also required for phosphorylation of Ime1p (4, 24, 32) (Fig. 4). Thus, our results indicate that Rim11p and Mck1p have overlapping roles in promoting Ume6p phosphorylation.

Relationship of Rim15p to Rim11p and Mck1p. Our results argue that Rim15p promotes Ume6p NTR phosphorylation and thus tie Rim15p, Rim11p, and Mck1p to a single biochemical event. Rim15p is inhibited through direct phosphorylation by cyclic AMP (cAMP)-dependent protein kinase, and so our results provide a biochemical connection between the cAMP and meiotic gene expression. However, cAMP levels respond to glucose (40), yet here Rim15p is required for a nitrogen starvation response. Thus, our findings explain one of several mechanisms by which the cAMP pathway governs meiosis but do not explain how nitrogen limitation governs Ume6p NTR phosphorylation.

rim15

rim15

A newly defined nitrogen response module. Nitrogen limitation causes diverse responses in yeast, including expression of nitrogen catabolic genes, changes in permease stability, down-regulation of translational machinery, sporulation, and filamentation. Several individual responses are governed by defined gene products (15, 23, 29), and the TOR signaling cascade may serve as a global regulator of nitrogen response pathways (7, 9, 34, 42). Our finding that Rim11p and Mck1p promote functional, nitrogen-responsive phosphorylation of Ume6p raises two questions. First, does a known nitrogen response pathway govern the activities of Rim11p and Mck1p? And second, do Rim11p and Mck1p govern any responses to nitrogen limitation in addition to activation of meiotic genes? We have observed that rim11 mck1 double mutants are defective in stationary-phase survival (Xiao and Mitchell, unpublished), and so these kinases may together play a role in starvation responses broader than previously thought. Analysis of Rim11p and Mck1p may explain how physiological and differentiation responses are coupled.