Integration of Transcriptional and Posttranslational Regulation in a Glucose Signal Transduction Pathway in Saccharomyces cerevisiae

Yeast strains and plasmids.The yeast strains used in this study are listed in Table 1. Cells were grown on either YP (2% Bacto peptone, 1% yeast extract) or YNB [0.67% yeast nitrogen base plus 0.5% (NH₄)₂SO₄ lacking the appropriate amino acids] medium, supplemented with the appropriate carbon sources. Genes were disrupted by homologous recombination using HisG-URA3-HisG (1) or KanMX (39) cassettes. Sequences of the primers are available on request. To construct pBM4747 (MET25 promoter-green fluorescent protein [GFP]-STD1), pBM4748 (MET25 promoter-GFP-MTH1), and pBM4749 (MET25 promoter-GFP-Htr1-23), coding sequences of the genes were amplified by PCR and the resulting PCR products were cloned into the BamHI and SalI sites of pUG34 (pBM3842 ARSH4/CEN-HIS3-MET25 promoter-yeGFP-polylinker-CYC1 terminator; GenBank accession number AF298784). Plasmid pBM4750 was derived from pBM4542, in which the Mig1 site was mutated (CCCCAG to CAACAG [mutated nucleotides are underlined]). To make Std1-9KR (positions K207, K287, K312, K337, K344, K347, K354, K381, and K411), Std1-7KR (positions K287, K312, K344, K347, K354, K381, and K411), and Std1-5KR (positions K312, K344, K347, K354, and K411), the lysine residues in Std1 were converted to arginine by using a gap repair technique (39).

β-Galactosidase assays.β-Galactosidase activity assays were performed using the yeast β-galactosidase assay kit (Pierce, Rockford, IL) according to the manufacturer's instructions. Results are presented in Miller units [(1,000 × OD₄₂₀)/(T × V × OD₆₀₀), where OD₄₂₀ is the optical density at 420 nm, T is the incubation time in minutes, and V is the volume of cells in milliliters]. Mean activities are averages from four to seven assays of at least eight independent transformants. Precultures were grown in selective media plus 2% galactose and 2.5% glycerol. Cells were then transferred to YP medium containing the appropriate carbon source and were incubated at 30°C for 4 h before β-galactosidase activity was measured.

Immunoprecipitation and ChIP.Yeast cell extracts were prepared by vortexing the cell pellets with glass beads in NP-40 buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% NP-40) at 4°C for 10 min. Crude protein (3 mg) was incubated with antibody-conjugated agarose beads (Santa Cruz) at 4°C for 3 h. The beads were then washed with high-salt buffer (NP-40 buffer containing 1 M NaCl) and boiled in sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer at 90°C for 5 min to elute proteins. For Western blot analysis, the eluted proteins were resolved by SDS-PAGE and transferred to a polyvinylidene difluoride membrane (Millipore). Monoclonal antibodies against LexA, GFP, and Myc (9E10) (Santa Cruz) and the ECL system (Pierce) were used for detection of proteins in the membranes as previously described (16). Binding of protein to the HXT3 promoter in vivo was assayed by chromatin immunoprecipitation (ChIP) as described previously (16).

Mth1, Std1, and Rgt1 are required for repression of HXT gene expression.To verify the roles of Mth1, Std1, and Rgt1 in glucose signaling, we analyzed the effect of loss of these genes on expression of HXT1, which is induced by high levels of glucose (2%), and on expression of HXT3, which is induced by low levels of glucose (0.2% [a condition mimicked by 2% raffinose]), as well as by high glucose levels (28) (Table 2, genotype 1). As expected, removal of the Rgt1 repressor substantially relieved repression of the HXT genes (Table 2, genotype 2, Gal). Deletion of MTH1 relieved repression of the low-glucose-induced HXT3 gene but had little effect on expression of the high-glucose-induced HXT1 gene (Table 2, genotype 3, Gal). Further deletion of STD1 in an mth1 mutant relieved repression of HXT1 expression (Table 2, genotype 5). Deletion of STD1 alone had little effect on expression of HXT1 and HXT3 (Table 2, genotype 4). Thus, Mth1 seems to be responsible for repression of the low-glucose-induced HXT3 gene (and for that of HXT4, another low-glucose-induced gene [data not shown]), but either Std1 or Mth1 is sufficient for repression of the high-glucose-induced HXT1 gene in the absence of glucose. These results reinforce previously reported findings (4, 20, 32) and suggest that Mth1 and Std1 work together with Rgt1 to repress expression of the high-glucose-induced HXT1 gene but that Mth1 acts alone to regulate expression of the low-glucose-induced HXT3 and HXT4 genes.

Mth1 and Std1 inhibit the ability of Rgt1 to activate transcription.Rgt1 is a transcriptional activator in cells grown on high levels of glucose but not in cells grown in the absence of glucose (on galactose) (29) (Table 3, genotype 1). Deletion of both MTH1 and STD1 causes Rgt1 to activate transcription in cells grown on galactose (Table 3, genotype 4), suggesting that Mth1 and Std1 play a role in inhibiting transcriptional activation, in addition to their roles in promoting transcriptional repression in the absence of glucose. In mth1 and std1 single mutants, Rgt1 was unable to activate significant transcription in cells grown on galactose, indicating that Mth1 and Std1 are redundant inhibitors of the transcriptional activation function of Rgt1 when glucose levels are low. A form of Mth1 that is resistant to glucose-induced degradation (due to the dominant MTH1-23 [also known as HTR1-23] mutation) (6, 26, 33, 34) prevents Rgt1 from activating transcription in the presence of high levels of glucose. Thus, in addition to promoting transcriptional repression by Rgt1 when glucose is absent (4, 30, 32), Mth1 and Std1 seem to inhibit transcriptional activation by Rgt1 when glucose levels are low. We believe that Mth1 and Std1 regulate Rgt1 function directly, because they interact with Rgt1 (20, 30, 38). In addition, Mth1 and Std1 are associated with HXT promoters (Fig. 1).

• Open in new tab
• Download powerpoint

FIG. 1.

Mth1 and Std1 are associated with Rgt1 at the HXT3 promoter. Chromatin was prepared from cells expressing Mth1 (pBM4560) or Std1 (pBM4544) tagged with 9 copies of the Myc epitope grown under repressing (2% galactose) or inducing (2% glucose) conditions. Chromatin was precipitated with the anti-Myc antibody and used to detect the HXT3 promoter by PCR amplification (ChIP). As a control, the ACT1 gene was amplified. The proteins expressed were visualized by Western blotting. Essentially identical results were obtained when the immunoprecipitates were tested for enrichment of the HXT1 promoter (data not shown).

Glucose regulation of MTH1 and STD1 expression contributes to glucose signal transduction.It seems clear that Mth1 is degraded upon exposure of yeast cells to glucose, but there are conflicting reports regarding Std1 degradation in response to glucose (4, 23). Indeed, in our hands, degradation of Mth1 was reproducibly observed but degradation of Std1 in response to glucose was variable. We suspected that this was due to the different regulation of STD1 and MTH1 expression by glucose: STD1 expression is induced by glucose via the Rgt2/Snf3-Rgt1 pathway, while MTH1 expression is repressed by glucose via the Snf1-Mig1 pathway (15). Induction of STD1 expression by glucose would be expected to counteract glucose-induced degradation of Std1 and obscure its disappearance. Conversely, repression of MTH1 expression by glucose should reinforce the glucose-induced degradation of Mth1, thereby enhancing its disappearance upon addition of glucose to cells.

We interrupted glucose regulation of STD1 and MTH1 by replacing their promoters with the promoter of MET25, which is not regulated by glucose. Expressing STD1 at the basal level of this promoter (by including methionine in the medium) makes degradation of Std1 in glucose-grown cells obvious (Fig. 2A, center panels) and significantly accelerates the rate of loss of Std1 after addition of glucose to cells (Fig. 2B). This suggests that induction of STD1 expression by glucose attenuates the glucose signal to Rgt1 by slowing the disappearance of Std1. By contrast, when repression of MTH1 expression by glucose is interrupted either by expressing MTH1 at the basal level of the MET25 promoter, by deleting MIG1 and MIG2, or by removing the Mig1/Mig2-binding sites from the MTH1 promoter, the extent (Fig. 2A) and rate (Fig. 2C) of degradation of MTH1 are reduced.

• Open in new tab
• Download powerpoint

FIG. 2.

Glucose induction of STD1 expression dampens degradation of Std1. (A) Cells expressed the Std1 and Mth1 proteins from their own promoters (Std1-Myc₉ [pBM4544] or Mth1-Myc₉ [pBM4560]) or from the MET25 promoter (GFP-Std1 [pBM4747] or GFP-Mth1 [pBM4748]). (Both versions of both proteins—with 9 copies of the Myc epitope at their C termini or with GFP at their N termini—appear to be functional, because they complement an std1 mth1 mutant for repression of HXT expression in galactose-grown cells.) Cells were grown to mid-log phase in a medium selective for the plasmids containing 2% galactose. Aliquots were then transferred to a medium containing 2% galactose (Gal) or 2% glucose (Glu), incubated at 30°C for 60 min, and lysed. Cells (pdr5Δ) deleted for the gene encoding a drug efflux pump were also treated with MG132 (50 μg/ml) for 30 min, and the proteins were analyzed by Western blotting with anti-Myc and anti-GFP antibodies (Santa Cruz). (B and C) Time course analysis of the degradation of Std1 and Mth1. (B) Std1 was expressed from its own promoter (pSTD1-STD1-Myc₉) and from the MET25 promoter (pMET25-GFP-STD1). (C) Cells (wild type [FM393] or mig1Δ mig2Δ [YM6682]) expressing Mth1-Myc₉ (pBM4560) from its own promoter were tested for Mth1 degradation. Mig1-DBSD (pBM4750) indicates that the MTH1 promoter carries a mutation of the Mig1 binding site (25). Cells grown on galactose medium (2%) were transferred to glucose medium (2%), and the same aliquots were harvested and analyzed at the time indicated above each lane.

The effects of transcriptional regulation of MTH1 and STD1 on the rate of loss of Mth1 and Std1 are expected to be translated into effects on the rates of induction and repression of HXT expression. We surmised that the glucose repression of MTH1 expression and the resulting acceleration of its disappearance from the cell after addition of glucose might serve to ensure speedy induction of HXT expression. Indeed, in cells in which MTH1 expression is not repressed by glucose (due to deletion of the genes encoding the Mig1 and Mig2 glucose repressors), induction of HXT3 expression by glucose is delayed relative to that in wild-type cells (Fig. 3A). Conversely, we speculated that glucose induction of STD1 expression might serve to replenish Std1 after its initial glucose-induced degradation so as to enable prompt establishment of repression of HXT expression when glucose is exhausted in the culture. Indeed, in cells in which STD1 expression is not induced by glucose, repression of HXT1 expression is established more slowly than in wild-type cells after addition of galactose (Fig. 3B). Thus, transcriptional regulation of MTH1 and STD1 significantly affects the course of induction and repression of HXT gene expression.

• Open in new tab
• Download powerpoint

FIG. 3.

Relieving transcriptional regulation of MTH1 and STD1 results in delayed induction and delayed repression, respectively, of HXT gene expression. (A) FM393 (wild type) (solid line) and YM6682 (mig1Δ mig2Δ) (dashed line) carrying HXT3::lacZ (pBM2819) were grown on 2% galactose. At time zero, 2% glucose was added to induce expression of HXT3. β-Galactosidase was assayed at the times indicated. (B) Cells (YM6292) carrying HXT1::lacZ (pBM2636) and expressing STD1 from its own promoter (pBM4540) or from the MET25 promoter (pBM4747) were grown in glucose. At time zero, the cells were pelleted, washed with water, and resuspended in 2% galactose to induce repression of HXT1 expression. Aliquots of the culture were assayed for β-galactosidase activity at the times indicated. During this time the cells approximately doubled in number.

Degradation of Std1 and Mth1 requires a glucose signal.Degradation of Std1 and Mth1 requires the glucose sensors Rgt2 and Snf3 (Fig. 4), as well as two components of the SCF^Grr1 ubiquitin-protein ligase, Grr1 and Cdc34 (Fig. 5). Glucose addition does not cause Std1 and Mth1 to disappear in a temperature-sensitive pre2 mutant defective in a chymotrypsin-like activity of the proteasome (8-11) or in the presence of the proteasome inhibitor MG132 (Fig. 2). These results support the view that glucose binding to the Rgt2 and Snf3 glucose sensors causes them to initiate proteasome-mediated degradation of Mth1 and Std1 by targeting them for ubiquitination by the SCF^Grr1 ubiquitin-protein ligase.

• Open in new tab
• Download powerpoint

FIG. 4.

Rgt2 and Snf3 promote glucose-induced degradation of Std1 and Mth1. (A) GFP-Std1 (pBM4747) or GFP-Mth1 (pBM4748) expressed in wild-type (FM391) or rgt2 snf3 (YM6212) cells was detected by Western blotting. Control lanes (Ctl.) were loaded with extracts of cells containing the empty vector (GFP alone). (B) The dominant HTR1-23 mutation in MTH1 is resistant to degradation. GFP-Mth1 with the HTR1-23 mutation (pBM4749) was expressed in FM391 (wild type) and was detected by Western blotting.

• Open in new tab
• Download powerpoint

FIG. 5.

Degradation of Mth1 and Std1 requires the SCF^Grr1 ubiquitin-protein ligase complex and the 26S proteasome. For Western blotting, cell extracts were prepared from yeast cells expressing GFP-Std1 (pBM4747) or GFP-Mth1 (pBM4748) and treated as described for Fig. 2A. Strains used were YM4127 (wild type), YM6542 (grr1Δ), FM542 (cdc34^ts), and FM524 (pre2-2^ts). Temperature-sensitive mutant strains were grown at 30°C overnight, then shifted to a medium containing 4% glucose, and incubated for 1 h at 30°C or 37°C. The GFP-Std1 and GFP-Mth1 proteins were then detected by Western blotting.

Potential ubiquitin attachment sites in Std1 are required for Std1 degradation.The SCF^Grr1 ubiquitin-protein ligase catalyzes the covalent attachment of ubiquitin to lysine residues of the target protein (17, 22, 35, 36). Evidence has been presented that suggests that Mth1 is ubiquitinated (37), but similar evidence that Std1 is also modified in this way is lacking. Indeed, our attempts to demonstrate this modification of Std1 have so far proven unsuccessful. We noticed 10 lysine residues in Std1 that are conserved in its orthologues in other yeasts (positions 207, 282, 287, 312, 337, 344, 347, 354, 381, and 411). Conversion of 9 of these lysines to arginine (9KR) prevented glucose-induced degradation of Std1 (Fig. 6B) and severely reduced derepression of HXT1 expression (Fig. 6A). Changing fewer than 9 of these lysine residues (7KR and 5KR) had smaller effects on induction of HXT1 expression, suggesting that ubiquitination at only a few sites of Std1 is required to target the protein for degradation. These results provide indirect evidence supporting the idea that Std1 is targeted for degradation by ubiquitination.

• Open in new tab
• Download powerpoint

FIG. 6.

Changing evolutionarily conserved lysine residues in Std1 prevents induction of HXT1 expression and protects Std1 from degradation. (A) HXT1::lacZ expression was measured in yeast cells (YM6292) carrying Std1-5KR, Std1-7KR, and Std1-9KR, respectively. Black bars, cells grown in 2% galactose (HXT-repressing conditions); gray bars, cells grown in 4% glucose (HXT-inducing conditions). Vec., vector. (B) Extracts prepared from cells expressing GFP-Std1 (pBM4747) (wild type [WT]) or GFP-Std1-9KR (pBM4754) under the control of the MET25 promoter were used for Western blotting.