INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The yeast Saccharomyces cerevisiae regulates metabolism, gene expression, and growth in response to carbon source availability. When glucose is abundant, the expression of a large number of genes, including those involved in the utilization of alternative carbon sources, gluconeogenesis, respiration, and peroxisomal functions, is repressed at the level of transcription by a process known as glucose repression. This process has been extensively studied both biochemically and genetically (for recent reviews, see references 4, 16, and 26). These studies show that both the Snf1 (Cat1, Ccr1) protein kinase and the Reg1 (Hex2, Srn1)-Glc7 (Cid1) protein phosphatase play a crucial role in glucose repression.

Snf1 is a serine-threonine protein kinase that regulates transcription both by inhibiting transcription repressors (e.g., Mig1) and by stimulating transcription activators (e.g., Cat8 and Sip4) (see reference 4 for a review). The Snf1 kinase is found in complexes containing the activating subunit Snf4 (Cat3) and members of the Sip1/Sip2/Gal83 family (25). Glucose regulates the activity of the kinase (55, 56) and regulates the interaction between Snf1 and Snf4 within the Snf1 kinase complex (24). In glucose-grown cells, the Snf1 kinase complex exists predominantly in an inactive autoinhibited conformation in which the catalytic domain of Snf1 binds to the regulatory domain of the protein (24). When glucose is limiting, the catalytic domain is released and the activating subunit Snf4 binds to the regulatory domain, which leads to an active, open conformation of the complex (24). Activation requires phosphorylation and an essential, conserved threonine in the activation loop of Snf1, which is phosphorylated during the activation of other kinases (11, 30, 55, 56).

Another major component of the glucose repression pathway is the Reg1-Glc7 protein phosphatase complex. GLC7 is an essential gene that encodes the catalytic subunit of protein phosphatase type 1 (PP1) (48). It is involved in glucose repression and also in the regulation of different processes including glycogen metabolism, translation, sporulation, chromosome segregation, and cell cycle progression (3, 12, 14, 22, 38, 46, 48, 54). As is the case for its mammalian counterpart, the Glc7 protein phosphatase participates in the regulation of these processes via the binding of specific regulatory subunits that target the phosphatase to the corresponding substrates (15, 21, 33, 49, 50, 57). Reg1 is one of these regulatory subunits, which targets Glc7 to substrates involved in the glucose repression pathway and other processes (15, 23, 34, 36, 49, 51). Genetic evidence indicates that Reg1 is required for glucose repression, and Reg1 is physically associated with Glc7, as judged by their strong interaction in the two-hybrid system and their coimmunoprecipitation from cell extracts (49).

The Reg1-Glc7 phosphatase complex has a role in regulating the Snf1 kinase complex. Reg1 (but not Glc7) interacts with the Snf1 kinase in the two-hybrid system (30). This interaction is glucose regulated, increasing when glucose is limiting. Moreover, in a reg1 mutant, the Snf1 complex remains in an open conformation even in the presence of glucose. It was proposed that in response to glucose, Glc7, targeted by Reg1, dephosphorylates Snf1 or another component of the Snf1 complex and thereby facilitates its conformational change from its active state to the autoinhibited form (30). Reg1 interacts very strongly with a kinase-dead Snf1 mutant, Snf1K84R, which has arginine substituted for the invariant lysine in the ATP-binding site; moreover, this interaction is not inhibited by glucose. These findings suggest that Snf1 negatively regulates its own interaction with Reg1.

Another component of the glucose repression pathway is hexokinase PII (Hxk2), a glycolytic enzyme that in addition to phosphorylating glucose is involved in regulating glucose repression (9, 10, 16, 31, 32). This enzyme is located in both the nucleus and the cytoplasm, consistent with dual roles in signaling and catalysis (20, 39), but its actual role in glucose repression is still poorly understood. When glucose is limiting, Hxk2 is a phosphoprotein (27, 52), and the Reg1-Glc7 phosphatase complex dephosphorylates Hxk2 in response to glucose (1, 40).

In this study, we explore the functional and physical relationships between Reg1, Glc7, and Snf1. We show that it is the recruitment of Glc7 by Reg1 that promotes the conformational change of the Snf1 complex in response to glucose. We also show that Reg1 is phosphorylated in a Snf1-dependent manner in response to glucose limitation and that Reg1 is then dephosphorylated, dependent on Glc7, when glucose is added back. Finally, we address the relationship of Hxk2 to the Reg1-Glc7 and Snf1 complexes and show that Hxk2 has a role in regulating the phosphorylation of Reg1.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Strains and genetic methods. S. cerevisiae strains used in this study were FY250 (MAT his3 leu2 trp1 ura3 SUC2; gift from F. Winston, Harvard Medical School, Boston, Mass.), FY250 snf1 (FY250 containing snf110), MCY3278 (FY250 containing reg1::URA3), MCY3000 (MATa lys2 his3 trp1 ura3 leu2 glc7-T152K), CTY10-5d (MATa ade2 his3 leu2 trp1 gal4 gal80 URA3::lexAop-lacZ; gift from R. Sternglanz, State University of New York, Stony Brook), and CTY10-5d hxk2 and CTY10-5d reg1 (CTY10-5d derivatives containing hxk2::URA3 or reg1::URA3). To construct strains FY250 hxk2 and MCY3000 hxk2, an SpeI-XhoI fragment of pSB21 (see below) was used to introduce hxk2::TRP1 by gene disruption (42), and the absence of hexokinase PII was confirmed by Western blotting with an anti-Hxk2 antibody (-Hxk2).

Oligonucleotides. The oligonucleotides used were HXK2-1 (AAATGGATCCATTTAGGTCC), HXK2-2 (GATCATAGAATTCATGTTCAC), HXK2-5 (GCGGGGATCCAGAGCTCCACATTGG), HXK2-6 (TACGGGATCCTTATATAAGCATCTTTTACTAC), REG1-1 (TCAAGAATTCTAGCAAATTACTTCG), REG1-2 (TAATCTCGAGGATAATCCCATGGAATTG), and REG1-7 (CATCCCTCGAGTGTGAAGCTGGATATCG). Restriction sites are underlined.

Plasmids. To construct the disruption plasmid pSB21, in which TRP1 replaces codons 15 to 415 of HXK2, oligonucleotides HXK2-5 and HXK2-6 were used to amplify by PCR the HXK2 sequence from nucleotide 602 to 281 nucleotides after the stop codon, using genomic DNA from strain FY250 as the template. The fragment was digested with BamHI and subcloned into pSK93 (see below) to yield pSK-HXK2(5-6), and then a SmaI/SalI fragment from the latter was subcloned into pRS305 (44), yielding pRS-HXK2(5-6). A SmaI/PstI TRP1 fragment from plasmid YDp-W (2) was used to replace a NcoI/PstI fragment from pRS-HXK2(5-6), generating pSB21.

Invertase and -galactosidase assays. Invertase activity was assayed in whole cells as previously described (24). -Galactosidase activity was assayed in permeabilized cells and expressed in Miller units as in reference 30. Where indicated, -galactosidase was assayed using crude protein extracts (5), and activity was expressed as units per milligram of protein.

-Phosphatase treatment. Protein extracts were prepared essentially as described in reference 5. Protein extracts (2 µg) in -phosphatase buffer containing 2 mM MnCl₂ were treated at 30°C for 30 min with 50 U of -phosphatase (New England BioLabs) in the presence or absence of a mixture of phosphatase inhibitors (50 mM EDTA, 50 mM NaF, 100 mM sodium phosphate buffer [pH 8]). The reactions were stopped by adding 1 volume of sample buffer and boiling for 3 min.

Preparation of cell extracts by the fast boiling method. Cells corresponding to 1 U of A₆₀₀ were collected by rapid centrifugation (14,000 rpm, 1 min), resuspended in 100 µl of Laemmli sample buffer (28), and boiled for 3 min. Glass beads (0.3 g, 450-µm diameter) were added to the suspension, and then cells were vortexed at full speed for 30 s. The suspension was boiled again 3 min and centrifuged at 14,000 rpm 1 min; 10 µl of the supernatant was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting.

Coimmunoprecipitation assays. Preparation of protein extracts and immunoprecipitation procedures were essentially as described previously (5). The extraction buffer was 50 mM HEPES (pH 7.5)-150 mM NaCl-0.5% Triton X-100-1 mM dithiothreitol-10% glycerol and contained 2 mM phenylmethylsulfonyl fluoride and Complete protease inhibitor cocktail (Boehringer Mannheim). -Snf1 polyclonal antibody (1 µl) was used in each immunoprecipitation reaction. Precipitates were analyzed by Western blotting using -HA monoclonal antibodies.

Immunoblot analysis. Crude extracts were separated by SDS-PAGE using 7% acrylamide gels and analyzed by immunoblotting using -Snf1 and -Snf4 polyclonal antibodies and commercial monoclonal -HA (Boehringer Mannheim) or -LexA (Clontech). Antibodies were detected by enhanced chemiluminescence (ECL) with ECL or ECL Plus reagents (Amersham).

Analysis of ³²P-labeled proteins. Cultures were grown in synthetic medium selective for plasmid maintenance and then transferred to 50 ml of YEP-PO₄ containing 2% glucose (1) for two doublings (optical densities at 600 nm [OD₆₀₀] of 0.2 to 0.8). Cells were then collected by centrifugation at 200 × g for 10 min and resuspended into 50 ml of YEP-PO₄ plus 2% glucose at an OD₆₀₀ of 0.2. [³²P]orthophosphate (2 mCi) was added, and cells were grown to an OD₆₀₀ of 0.8. The culture was divided in two; cells were harvested by centrifugation, washed in 10 ml of the appropriate medium (YEP-PO₄ plus 2% glucose or 2% raffinose), and resuspended into 25 ml of the same medium. [³²P]orthophosphate (1 mCi) was added to each culture, which was then grown for 20 min. Cells were harvested by centrifugation and washed. Extracts were prepared by the addition of 0.5 ml of acid-washed glass beads and 1 ml of lysis buffer (50 mM Tris-HCl [pH 8.0], Complete protease inhibitor cocktail, 1 µM microcystin-LR, 150 mM NaCl, 1% Triton X-100). This mixture was vortexed five times for 20 s with 1 min on ice between vortexing. The entire sample was then centrifuged at 200 × g for 2 min, and the supernatant (about 1 ml) was drawn off into a microcentrifuge tube and centrifuged at 12,000 × g for 10 min. For immunoprecipitation, the supernatant was precleared with 20 µl of a slurry of equal parts of lysis buffer and protein G-agarose for 30 min. Immunoprecipitations were done as described previously (19) with 5 µg of -LexA, using preconjugated -LexA-agarose (Santa Cruz).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Reg1 coimmunoprecipitates with the Snf1 protein kinase. To confirm the interaction between Reg1 and the Snf1 protein kinase detected in the two-hybrid assay (30), we introduced plasmids expressing the functional proteins HA-Reg1 and VP16-Snf1 (30) into a snf1 strain. VP16-Snf1 was immunoprecipitated from cell extracts with a polyclonal Snf1 antibody, and the precipitates were analyzed by Western blotting. HA-Reg1 coimmunoprecipitated with VP16-Snf1 and also with the inactive mutant VP16-Snf1K84R (Fig. 1). HA-Reg1 also coimmunoprecipitated with LexA-Snf1, thereby confirming that the VP16 moiety was not responsible.

View larger version (57K): [in this window] [in a new window]   FIG. 1.   Coimmunoprecipitation of HA-Reg1 with VP16-Snf1 and LexA-Snf1. Protein extracts (250 µg) were prepared from FY250 snf1 cells expressing the indicated proteins. VP16 and LexA fusion proteins were immunoprecipitated (IP) with -Snf1 polyclonal antibody, and the precipitated proteins were separated by SDS-PAGE and immunoblotted with -HA monoclonal antibody (Blot) (upper panel). Proteins in the input crude extracts (1 µg) were also immunodetected with either -HA (middle panel) or -Snf1 (lower panel). Size standards are indicated in kilodaltons.

Different sequences of Reg1 interact with Snf1 and the PP1 catalytic subunit. We next fused different regions of Reg1 in frame to the activating domain of Gal4 (GAD) and tested for their interaction with LexA-Snf1 and LexA-Glc7 (Fig. 2). LexA-Snf1 interacted most strongly with GAD-Reg1_424-760 and GAD-Reg1_424-1014. LexA-Glc7 showed a different interaction profile, interacting most strongly with GAD-Reg1_3-760. A consensus motif, (R/K)(V/I)XF, for the recognition of regulatory subunits by the catalytic subunit of mammalian PP1 has been identified, and the similar motif (R/K)X(V/I)XF was found in various PP1-binding proteins in yeast (8). Reg1 contains the sequence RHIHF₄₆₈ (Fig. 2), which is present in all of the GAD-Reg1 proteins that interacted with LexA-Glc7. (GAD-Reg1_424-760 also contains the motif, close to the GAD moiety, but no interaction with Glc7 was detected for unknown reasons.) Mutation of this motif dramatically reduces the interaction of Reg1 with Glc7 (1, 7). In contrast, the F468R mutation (1) did not substantially affect the two-hybrid interaction between Reg1 and Snf1. The combination LexA-Reg1F468R plus VP16-Snf1 gave 20 U of -galactosidase activity after a shift to low glucose, and LexA-Reg1F468R plus VP16-Snf1K84R gave 330 U during growth in 4% glucose (values are averages for four to six transformants of CTY10-5d). These values are the same as those for LexA-Reg1 (see Fig. 7). Thus, distinct Reg1 sequences mediate its interactions with Snf1 and Glc7.

View larger version (24K): [in this window] [in a new window]   FIG. 2.   Two-hybrid interaction of Snf1 and Glc7 with different sequences of Reg1. (A) Reg1 sequences fused to GAD that were used in the two-hybrid analysis. GAD-Reg1 fusions were used because certain regions of Reg1 show self-activating activity. The positions of F468 and serines in potential Snf1 recognition sites (see text) are indicated. (B) CTY10-5d transformants expressing the GAD-Reg1 proteins and either LexA-Snf1 or LexA-Glc7 were grown to mid-log phase in selective SC-4% glucose medium; cells were then washed with water and shifted to SC-0.05% glucose medium for 3 h; values are average -galactosidase activities of four to six transformants, and bars show standard deviations. (C) Western blots of proteins from transformants growing in 4% glucose and expressing LexA-Snf1 and the indicated GAD-Reg1 fusion; 10-µl aliquots of cell extracts prepared by the fast boiling method (see Materials and Methods) were immunoblotted with -HA (top) or -Snf1 (bottom). The production of GAD-Reg1424-1014, GAD-Reg1424-760, and GAD-Reg1760-1014 could not be assessed because these constructs lack the HA epitope tag; the production of GAD-Reg1 fusion proteins in transformants expressing LexA-Glc7 followed the same trend as transformants expressing LexA-Snf1 (data not shown). Size standards are indicated in kilodaltons.

Reg1 regulates the conformation of the Snf1 kinase complex by targeting Glc7 to the complex. The interaction between Snf1 and Snf4 within the kinase complex increases in response to glucose limitation; Snf4 binds to the regulatory domain of Snf1 and the autoinhibition of the Snf1 catalytic domain is relieved, leading to an open, active conformation of the kinase complex (24). Reg1 has a role in regulating the interaction of Snf1 and Snf4, by promoting the autoinhibited conformation of the complex (24, 30). Previously, we inferred that these effects of Reg1 result from the targeting of PP1 activity to the kinase complex. The availability of the Reg1F468R mutant allowed a direct test of this model.

View larger version (34K): [in this window] [in a new window]   FIG. 3.   Reg1 and Glc7 regulate the two-hybrid interaction of Snf1 and Snf4 within the kinase complex in response to glucose. Two-hybrid interaction between LexA-Snf1 and Snf4-GAD was measured in wild-type CTY10-5d and reg1 mutant transformants expressing the indicated HA-Reg1 fusion proteins. Values for cells growing in 4% glucose (R) and cells shifted to 0.05% glucose for 3 h (S) are average -galactosidase activities of four to six transformants, with standard deviations lower than 15% in all cases. A Western blot of wild-type transformants growing in 4% glucose and expressing the indicated proteins is shown; 10-µl aliquots of crude extracts (boiling method) were immunodetected either with -Snf1 (upper panel), -Snf4 (middle panel), or -HA (lower panel). Analysis of protein levels at the end of the 3-h shift did not reveal any dramatic changes, and the production of the different fusion proteins in reg1 transformants was similar to wild type (data not shown). Size standards are indicated in kilodaltons.

Reg1 is phosphorylated in response to glucose limitation in a Snf1-dependent manner. Genetic evidence indicates that the Snf1 catalytic activity negatively regulates the interaction between Snf1 and Reg1: a mutation in the ATP-binding site of Snf1 (Snf1K84R) increases this interaction and relieves inhibition of the interaction by glucose (see Fig. 7 and reference 30). To determine whether the mechanism entails phosphorylation of Reg1, we examined Reg1 for phosphorylation in vivo in response to glucose limitation (Fig. 4A). Cells expressing LexA-Reg1 or LexA-Reg1F468R were labeled with [³²P]orthophosphate during growth in 2% glucose and then transferred to fresh medium containing [³²P]orthophosphate and either 2% glucose or 2% raffinose for 20 min. Extracts were prepared, and the LexA fusion protein was immunoprecipitated, separated by gel electrophoresis, and detected by autoradiography. Phosphorylation of both the wild-type and mutant LexA-Reg1 proteins increased substantially (10- to 20-fold in three different experiments) during growth in raffinose, that is, in the absence of glucose. In contrast, the majority of the phosphoproteins detected by two-dimensional gel electrophoresis and autoradiography are the same in glucose- and raffinose-grown cells (1). This differential phosphorylation of LexA-Reg1 requires the Snf1 kinase because a shift to 2% raffinose did not cause a substantial increase in phosphorylation in a snf1 mutant (Fig. 4B). Similar results were obtained for HA-Reg1 (data not shown).

View larger version (47K): [in this window] [in a new window]   FIG. 4.   Snf1-dependent phosphorylation of Reg1 in response to glucose limitation. (A) Strain MCY3278 expressed no LexA fusion protein (), wild-type (WT) LexA-Reg1, or LexA-Reg1F468R. (B) Strain FY250 snf1 expressed LexA-Reg1. Cultures were grown in 2% glucose, labeled with [32P]orthophosphate, collected by centrifugation, and resuspended in medium containing [32P]orthophosphate and either 2% glucose (Glu) or 2% raffinose (Raf) for another 20 min as described in Materials and Methods. Extracts were prepared, and proteins were immunoprecipitated with preconjugated -LexA-agarose. Precipitates were subjected to SDS-PAGE in 6% acrylamide, and gels were dried for autoradiography. Size standards are indicated in kilodaltons.

View larger version (47K): [in this window] [in a new window]   FIG. 5.   N terminus of Reg1 is phosphorylated in low glucose in a Snf1-dependent manner. (A) FY250 cells expressing different regions of Reg1 fused to LexA were grown in selective SC-4% glucose medium (R). When they reached the mid-log phase, the cells were washed with water and shifted to SC-0.05% glucose medium for 3 h (S). Crude extracts (10 µl) prepared by the fast boiling method were subjected to immunoblot analysis with -LexA. (B) FY250 cells expressing LexA-Reg11-400 were grown in selective SC-4% glucose medium and then were shifted to SC-0.05% glucose medium for 5 min or 20 min; after this time, 2% glucose was added to the medium, and cells were collected after 5 min. Crude extracts prepared by the boiling method (10 µl) were analyzed by immunoblotting with -LexA. (C) FY250 snf1 cells expressing LexA-Reg11-400 were treated and analyzed as for panel B. (D) FY250 cells expressing HA-Reg11-443 were treated as for panel B except that an additional sample was taken 20 min after the addition of glucose; proteins in the crude extracts were immunodetected with -HA. (E) glc7-T152K cells expressing HA-Reg11-443 were grown in selective SC-4% glucose medium and shifted to 0.05% glucose for 20 min; 2-µg aliquots of protein extract were treated with -phosphatase (50 U) in the presence or absence of phosphatase inhibitors (see Materials and Methods); HA-Reg11-443 was immunodetected with -HA. Size standards are indicated in kilodaltons.

Reg1 is dephosphorylated by Glc7 in response to glucose. The phosphorylation of LexA-Reg1_1-400 and HA-Reg1_1-443 was reversible: if glucose was added back to cells after a shift to low glucose, the modified form rapidly disappeared (Fig. 5B and D). To determine whether Glc7 is responsible for this dephosphorylation, we used the mutant allele glc7-T152K, which partially relieves glucose repression but does not interfere with the function of Glc7 in other pathways (48, 49). The mutant was defective in dephosphorylating LexA-Reg1_1-400 after the addition of glucose, and the modified form was still present after 5 min (Fig. 6). However, the modified form was not prominent in cells growing exponentially in 4% glucose, presumably because the mutant phosphatase is still partially active.

View larger version (29K): [in this window] [in a new window]   FIG. 6.   PP1 and hexokinase PII are involved in regulating the phosphorylation state of Reg1. Strains with the indicated genotypes (see Materials and Methods) were transformed with a plasmid expressing LexA-Reg11-400. Cells were grown and proteins were analyzed and immunodetected with -LexA as for Fig. 5B. Size standards are indicated in kilodaltons.

Hexokinase PII has a role in regulating the phosphorylation of Reg1. The hexokinase PII, encoded by HXK2, both phosphorylates glucose and regulates glucose repression (16). The hxk2 and reg1 mutations are similar in several respects: both relieve glucose repression of many of the same genes, including SUC, MAL, GAL, HXT2, and HXT4, but do not relieve glucose repression of gluconeogenic genes (16); both affect glucose induction of HXT1 expression (37); both cause Snf1 and Snf4 to exhibit two-hybrid interaction in cells growing in high glucose (24); and snf1 is epistatic to both (35). In addition, the Reg1-Glc7 phosphatase complex dephosphorylates Hxk2 in response to glucose (1, 40). We therefore considered a role for Hxk2 in regulating the phosphorylation of Reg1.

View larger version (31K): [in this window] [in a new window]   FIG. 7.   Two-hybrid interaction of Reg1 and Snf1 in an hxk2 mutant. Wild-type CTY10-5d cells and an hxk2 mutant derivative of CTY10-5d expressing the indicated fusion proteins were treated as described for Fig. 2. Values for cells growing in 4% glucose (R) and cells shifted to 0.05% glucose for 3 h (S) are average -galactosidase activities of four transformants; standard deviations were lower than 10% of the average values in all cases. A Western blot of wild-type (WT) and hxk2 transformants expressing LexA-Reg1 and VP16-Snf1 is shown; 10-µl aliquots of crude extracts (boiling method) were immunodetected with -LexA (left) or -Snf1 (right). Levels of VP16-Snf1K84R were also similar in both strains (data not shown). n.d., not determined. Size standards are indicated in kilodaltons.

View larger version (34K): [in this window] [in a new window]   FIG. 8.   Two-hybrid interaction of Snf1 and Snf4 in reg1 and hxk2 mutants. Wild-type CTY10-5d cells and reg1 and hxk2 derivatives of CTY10-5d were transformed with plasmids expressing LexA-Snf1 and Snf4-GAD or GAD alone. Cells were grown and treated as described in the legend of Fig. 2. Values for cells growing in 4% glucose (R) and cells shifted to 0.05% glucose for 3 h (S) are average -galactosidase activities of four transformants; standard deviations were lower than 10% of the average values in all cases. A Western blot of the wild-type and reg1 and hxk2 transformants expressing the indicated fusion proteins is shown; 10-µl aliquots of crude extracts (boiling method) were immunodetected with -Snf1 (top) or -Snf4 (bottom). Size standards are indicated in kilodaltons.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We have examined the regulatory interactions between the Reg1-Glc7 form of PP1 and the Snf1 kinase complex. Previous work indicated that phosphorylation and dephosphorylation of Snf1 are important in the regulation of its activity in response to the glucose signal. Phosphorylation is required for activation of the kinase, and the conserved activation-loop threonine is required for Snf1 activity and for the positive regulatory interaction between Snf1 and Snf4 within the kinase complex (11, 30, 55, 56). Conversely, the Reg1-Glc7 protein phosphatase also regulates the kinase complex (24, 30).

We have examined the physical and functional interactions of Reg1 with Glc7 and Snf1. We show that different sequences of Reg1 interact with these two proteins. The binding of Glc7 is mediated by the RHIHF₄₆₈ motif, which matches the consensus motif (R/K)X(V/I)XF found in other regulatory subunits of PP1 (1, 7, 8). The F468R mutation greatly reduces the binding of Reg1 to Glc7 (1, 7) but does not affect its binding to Snf1. We were therefore able to use the Reg1F468R form to demonstrate that the effects of Reg1 on the conformation of the Snf1 kinase complex require the recruitment of the catalytic subunit Glc7. Thus, these effects are mediated by PP1 activity. Glc7 presumably dephosphorylates Snf1, or possibly another component of the kinase complex, to promote the autoinhibited conformation of the complex.

The Snf1 kinase negatively regulates its own interaction with Reg1, as the kinase-dead Snf1K84R protein interacts strongly with Reg1, even in the presence of glucose (30). We present evidence that the mechanism entails phosphorylation of Reg1. The N-terminal region of Reg1 (Reg1_1-400) is rapidly phosphorylated when cells are shifted to low glucose, and this phosphorylation is dependent on Snf1 activity. Moreover, we show that dephosphorylation occurs soon after the addition of glucose to the medium and that a mutation in Glc7 (glc7-T152K) affects this process.

The phosphorylation state of Reg1 is also affected by hexokinase PII, a protein that not only phosphorylates glucose in the glycolytic pathway but also regulates glucose repression by an unknown mechanism (16). We found that LexA-Reg1_1-400 is not phosphorylated in a hxk2 mutant. It is possible that Hxk2 contributes to efficient binding of Reg1 to Snf1 in low glucose, but the two-hybrid interaction of the full-length proteins (LexA-Reg1 and VP16-Snf1) was reduced only two- to threefold in a hxk2 mutant. Hxk2 could also stimulate the phosphorylation of Reg1 by Snf1 by some other mechanism. Alternatively, Hxk2 may interfere with the dephosphorylation of Reg1 by Glc7. Consistent with this view, phosphorylation of LexA-Reg1_1-400 was detected in the double glc7-T152K hxk2 mutant, which has impaired PP1 function. Regardless of the precise mechanism, these data suggest that the effects of Hxk2 on glucose repression are mediated by Reg1-Glc7. Previous genetic evidence supported a functional relationship between Hxk2 and Reg1-Glc7, and we have presented additional evidence, namely, that overexpressing Reg1 suppresses the defect in glucose repression of invertase synthesis of a hxk2 mutant. Finally, we detected a two-hybrid interaction between LexA-Reg1 and GAD-Hxk2 in glucose-limited cells when, in addition to the overexpression of the two fusion proteins, Snf1K84R was also overexpressed.

What is the role of the phosphorylation of Reg1? Analysis of the hxk2 mutant suggests that the phosphorylation of Reg1 stimulates the ability of Reg1-Glc7 to promote the return of the open, activated Snf1 complex to a closed, autoinhibited state. In this mutant, Reg1 is not phosphorylated, and the interaction between Snf1 and Snf4 within the Snf1 complex is constitutive and occurs even in cells growing in high glucose (reference 24 and this study). Phosphorylation is also required for release of Reg1 from the kinase complex, as Reg1 remains tightly bound to the kinase-dead SnfK84R protein in glucose-grown cells (30). Finally, the N-terminal 200 residues of Reg1 play a role in glucose repression of SUC2 (7), and it is possible that the N-terminal phosphorylation of Reg1 modulates this function.

We propose the following model for the role of the Reg1-Glc7 phosphatase complex in regulation of the Snf1 complex (Fig. 9). Reg1-Glc7 binds to Snf1 when the kinase is activated by phosphorylation, which occurs at a much higher rate in glucose-limited cells than in glucose-grown cells. Reg1 is then phosphorylated by Snf1, and Hxk2 either stimulates the binding and/or phosphorylation of Reg1 or inhibits the dephosphorylation of Reg1 by Glc7. In response to a glucose signal, Reg1-Glc7 facilitates the transition back to the autoinhibited state, presumably by dephosphorylating Snf1, and Reg1-Glc7 is released from its association with the kinase complex. The phosphorylation of Reg1 by Snf1 appears to stimulate the activity of Glc7 in promoting closure of the complex, and this phosphorylation is also required for release of Reg1-Glc7 from the kinase complex. Glc7 then dephosphorylates Reg1. The glucose signal most likely inhibits the initial phosphorylation of Snf1 but may also activate Reg1/Glc7 function. The complexity of these regulatory interactions suggests that the Snf1 kinase activity is finely tuned in response to glucose.

View larger version (33K): [in this window] [in a new window]   FIG. 9.   Model for regulation of the Snf1 kinase complex in response to glucose. Cells growing in high glucose maintain the Snf1 complex predominantly in an autoinhibited conformation in which the regulatory domain (RD) binds to the catalytic kinase domain (KD). Low levels of glucose favor phosphorylation of the Snf1 kinase, possibly by an upstream kinase. The catalytic domain is released from autoinhibition and Snf4 binds to the regulatory domain, leading to an open and active conformation of the complex. The Reg1-Glc7 phosphatase complex then binds to Snf1, and Reg1 is phosphorylated by Snf1. Hxk2 either stimulates the binding and/or phosphorylation of Reg1 or inhibits the dephosphorylation of Reg1 by Glc7. Reg1-Glc7 facilitates the transition back to the autoinhibited state, presumably by dephosphorylating Snf1. The phosphorylation of Reg1 appears to stimulate closure of the complex. Reg1-Glc7 is released from its association with the kinase complex, and this release also requires phosphorylation of Reg1 by Snf1. Glc7 then dephosphorylates Reg1. In glucose-grown reg1, glc7-T152K, or hxk2 mutant cells, the Snf1 kinase, once activated, becomes trapped in the activated state. Each kinase complex contains one of the related proteins Sip1, Sip2, and Gal83.