ABSTRACT
The function of two fission yeast genes, SPCC74.03c/ssp2+ and SPAC23H4.02/ppk9+, encoding an Snf1-like protein kinase were investigated. Deletion of ssp2+ caused a partial defect in glucose derepression of inv1+, fbp1+, and gld1+ and in assimilation of sucrose and glycerol, while a mutation in ppk9+ had no apparent effect. Scr1, a transcription factor involved in glucose repression, localized to the nucleus under glucose-rich conditions and to the cytoplasm during glucose starvation in wild-type cells. In contrast, in the ssp2Δ mutant, Scr1 localized to the nucleus in cells grown in glucose-rich medium as well as in glucose-starved cells. Immunoblot analysis showed that Ssp2 is required for the phosphorylation of Scr1 upon glucose deprivation. Mutation of five putative Ssp2 recognition sites in Scr1 prevented glucose derepression of invertase in glucose-starved cells. These results indicate that Ssp2 regulates phosphorylation and subcellular localization of Scr1 in response to glucose.
INTRODUCTION
Most microorganisms, including yeasts, respond to environmental conditions to optimize use of available carbon sources by regulating the expression of many genes. Glucose, the carbon source preferred by most microorganisms, is also recognized to be an important primary messenger molecule that mediates this adaptive response. In the presence of high levels of glucose, genes involved in the utilization of alternative carbon sources are repressed (subject to glucose repression), whereas genes required for glycolysis, glucose transport, etc., are expressed (subject to glucose induction). When glucose levels are limited, on the other hand, expression of glucose-repressed genes is derepressed (reviewed in references 9, 31, 32, and 35).
In the budding yeast Saccharomyces cerevisiae, whose glucose-signaling pathway has been well studied, Snf1 protein kinase-mediated signal transduction plays a central role in glucose derepression of many genes (45). Under glucose-rich conditions, the Snf1 kinase, which forms a heterotrimeric enzyme complex with the γ subunit Snf4 and one of the three related β subunits, Gal83, Sip1, or Sip2, is inactivated via dephosphorylation by protein phosphatase 1 complex (Reg1-Glc7) (5, 20). Once cells are shifted to glucose-limiting conditions, the Snf1 kinase is activated by upstream kinases (Tos3, Sak1, and Elm1) and imported into the nucleus (17, 28, 38). Activated Snf1 kinase phosphorylates multiple transcription factors, causing the derepression/induction of many genes. In the case of derepression of invertase-encoding SUC2, Snf1 kinase phosphorylates Mig1, a DNA-binding repressor that recruits the Ssn6-Tup1 corepressor to the promoters of many glucose-repressed genes, resulting in dissociation of Mig1 from the SUC2 promoter (reviewed in references 3 and 32).
In the fission yeast Schizosaccharomyces pombe, the function of the Snf1 pathway has not been characterized, although several factors involved in the glucose signaling pathway have been reported. We previously reported that a C2H2 Zn finger protein, Scr1, an S. pombe homologue of Mig1, represses the transcription of inv1+ (encoding invertase) and gld1+ (encoding glycerol dehydrogenase) under glucose-rich conditions (25, 41). In the presence of abundant glucose, transcription of fbp1+ encoding fructose-1,6-bisphosphatase is repressed by cooperative binding of Scr1 and Tup1-like Tup11 and Tup12 proteins to a cis-acting element in its promoter (13, 14, 19, 29). However, the possibility that the Snf1 kinase pathway is involved in glucose derepression of these genes has not been determined.
In the present study, we characterized the function of two S. pombe Snf1-like protein kinase genes and demonstrate that the Snf1 pathway plays important roles in glucose derepression.
MATERIALS AND METHODS
Strains and media.The parent S. pombe strain ARC039 (h− leu1-32 ura4-C190T) (8) was provided by Asahi Glass Co., Ltd. (Yokohama, Japan). MM is synthetic minimal medium, and YES is standard rich medium (3% glucose, 0.5% yeast extract, and MM supplements). Both media were used to grow S. pombe as described elsewhere (27). 5′-Fluoroorotic acid (FOA)-supplemented YNB medium (0.7% yeast nitrogen base, 2% glucose, 50 mg/ml uracil, 225 mg/ml each of leucine, adenine, histidine, and cysteine, and 0.1% FOA) was used to select for S. pombe ura4 mutants. YES-Suc medium is YES medium containing sucrose instead of glucose. YES medium containing 8% glucose was used for repressing conditions, and YES medium containing 0.05% glucose and 2% glycerol was used for derepressed conditions. Transformation by electroporation was performed as described previously (37). Escherichia coli XL1-Blue (Stratagene, CA) was used for all cloning procedures.
Plasmid constructions.pAL-Ssp2-GFP and pAL-Ppk9-GFP were constructed as follows. A 2.7-kbp fragment containing the ssp2+ open reading frame (ORF) and the ssp2+ promoter was amplified by PCR using the following primers containing the indicated restriction sites: 5′-GTTTTGGGCCCCAGTTCAGGCTCGTTATATTC-3′ containing ApaI and 5′-GTTTTGTCGACGCAGAAAATAACTTGCACAC-3′ containing SalI. A 2.4-kbp fragment containing the ppk9+ ORF and the ppk9+ promoter was amplified by PCR using the following primers containing the indicated restriction sites: 5′-GTTTTGGGCCCCATTCGCACGGTCTGTAATATGC-3′ containing ApaI and 5′-GTTTTGTCGACGCAATTTTAAGGATCAATTG-3′ containing SalI. These fragments were digested with ApaI and SalI and cloned into the corresponding sites of pAL (40). To construct green fluorescent protein (GFP) fusion proteins, pAL-Ssp2 and pAL-Ppk9 were digested with SalI and NotI to facilitate cloning of SalI- and NotI-digested GFP.
pJK148-Scr1 was constructed as follows. A 3.2-kbp fragment containing the scr1+ ORF and the scr1+ promoter was amplified by PCR using the following primers: 5′-CGGTATCGATAAGCTCCTCGAGTATTTAGCTAACCCACAC-3′ and 5′-GGCGACCGGTGGATCGGGCTTGGTCATAGGAGTTAACGGC-3′. The resultant PCR fragment was cloned into pJK148 (22) digested with HindIII and BamHI using an In-Fusion PCR cloning kit (TaKaRa Bio, Japan). pJK148-Scr1 S235A, pJK148-Scr1 S332A S333A, and pJK148-Scr1 S408A S410A were constructed as described previously using pJK148-Scr1 as a template and the following primers with serine-to-alanine conversions underlined (23): 5′-TTGGCTGCTGCGGCTGCTAATCAATTGGATGCTGC-3′ and 5′-AGCCGCAGCAGCCAACAATTGCATCTCGTTCATGG-3′ for S235A, 5′-TCTAACGCTGCTACTAGCCTTCATTCAATGTATGG-3′ and 5′-AGTAGCAGCGTTAGATTTGCTAGGCAGATAGGGAG-3′ for S332A S333A, and 5′-AGGGCTTTTGCTCCCACCCCAGACGTCACACCTC-3′ and 5′-GGGAGCAAAAGCCCTAGTGGAAAAGGTAGGACTAC-3′ for S408A S410A.
Fluorescence microscopy.Cells were collected by centrifugation and suspended in 50 μl of culture medium, of which 5 μl was placed on a glass slide. Fluorescent images of living cells were taken with a cooled charge-coupled-device camera and stored digitally using MetaMorph software (Universal Imaging, Downingtown, PA). For fixed samples, cells were suspended in 70% ethanol, washed with phosphate-buffered saline, and suspended in 5 μl of phosphate-buffered saline containing Hoechst 33342 dye (0.1 mg/ml).
GFP and 3×FLAG tagging of Scr1.S. pombe strains expressing scr1+-GFP were constructed as follows. A 2.1-kbp fragment containing the scr1+ ORF was amplified from S. pombe genomic DNA by PCR using the following primers: 5′-ATGTCCGAAGCCACCACTGCCACAACAACC-3′ and 5′-CAGTTCAACTTTTTGATAAGATCTGC-3′. The PCR product was then subcloned into pGEM-T Easy (Promega, WI). A 5.1-kbp fragment was amplified by PCR using the following primers and pGEM-T Easy-scr1+ as a template (fragment A): 5′-CACCATGGTGGCGACGGGCTTGGTCATAGGAGTTAAC-3′ and 5′-GATTTACGACATCCGTAACCCAATATTTGCGCTGTGA-3′. A 0.7-kbp fragment containing the GFP gene was amplified using the following primers and pEGFP as a template (fragment B): 5′-GTCGCCACCATGGTGAGCAAGGGCGAG-3′ and 5′-CCTCTACAAATGTGGTATGGCTGATTATG-3′. A 2.3-kbp fragment containing the leu1+ ORF was amplified from S. pombe genomic DNA by PCR using the following primers (fragment C): 5′-CCACATTTGTAGAGGGCCTAGCAGTTTGAAACCTTCC-3′ and 5′-CGGATGTCGTAAATCAATTCCATGCTTTTGC-3′. Fragments A, B, and C were ligated using an In-Fusion PCR cloning kit (pGEM-T Easy-scr1+-GFP-leu1+). A fragment containing scr1+, GFP, and leu1+ gene was amplified using the following primers and pGEM-T Easy-scr1+-GFP-leu1+ as a template: 5′-ATGTCCGAAGCCACCACTGCCACAACAACC-3′ and 5′-CAGTTCAACTTTTTGATAAGATCTGC-3′. S. pombe wild-type, ssp2Δ, ppk9Δ, ssp2Δ ppk9Δ, and cbs2Δ strains were then transformed with the resultant PCR product.
S. pombe strains expressing scr1+-3×FLAG were constructed as follows.
A 5.1-kbp fragment was amplified by PCR using the following primers and pGEM-T Easy-scr1+ as a template and ligated using an In-Fusion PCR cloning kit (pGEM-T Easy-scr1+-3×FLAG): 5′-TGTCATGATCTTTATAATCACCGTCATGGTCTTTGTAGTCGGGCTTGGTCATAGGAGTTAACGGC-3′ and 5′-ATAAAGATCATGACATCGATTACAAGGATGACGATGACAAGTAACCCAATATTTGCGCTGTGAACG-3′. A fragment amplified from pGEM-T Easy-scr1+-3×FLAG using primers 5′-TTCAAACTGCTAGGCTTACTTGTCATCGTCATCCTTG-3′ and 5′-GATTTACGACATCCGCACCTTGGTTGTGTGTTATGC-3′ and leu1+ amplified by PCR using primers 5′-GCCTAGCAGTTTGAAACCTTCCACCATAAC-3′ and 5′-CGGATGTCGTAAATCAATTCCATGCTTTTGC-3′ using S. pombe genomic DNA as the template were ligated using an In-Fusion PCR cloning kit (pGEM-T Easy-scr1+-3×FLAG-leu1+). A fragment containing scr1+-3×FLAG and leu1+ was amplified using the following primers and pGEM-T Easy-scr1+-3×FLAG-leu1+ as the template: 5′-ATGTCCGAAGCCACCACTGCCACAACAACC-3′ and 5′-CAGTTCAACTTTTTGATAAGATCTGC-3′. S. pombe wild-type, ssp2Δ, ppk9Δ, ssp2Δ ppk9Δ, and cbs2Δ strains were transformed with the resultant PCR product.
Gene disruptions.ssp2+, ppk9+, and cbs2+ were disrupted using ura4+ (11) as a selectable marker. To disrupt ssp2+, a 1.9-kbp DNA fragment including part of the ssp2+ ORF was amplified from wild-type S. pombe genomic DNA and subcloned into pGEM-T Easy. A 0.97-kbp HindIII fragment constituting part of the ORF was deleted, and a ura4+ cas-sette was inserted. To disrupt ppk9+, a 1.5-kbp fragment including part of the ppk9+ ORF was amplified from wild-type S. pombe genomic DNA and subcloned into pGEM-T Easy. A 0.45-kbp BglII ORF fragment was deleted and replaced with a ura4+ cassette. To disrupt cbs2+, a 1.2-kbp DNA fragment including part of the cbs2+ ORF was amplified from wild-type S. pombe genomic DNA and subcloned into pGEM-T Easy. The ura4+ marker was inserted into the EcoRI site, about 540 bp downstream from the initiation codon of the cloned cbs2+. Linearized DNA fragments carrying the disrupted genes were used to transform the parent strain ARC039, and stable ura4+ transformants were selected. Disruption of the chromosomal loci was confirmed by PCR. ssp2 mutant cells were subsequently plated on FOA-supplemented YNB medium to select for loss of ura4+ activity. These ura4− cells were then used as parent cells to generate a ssp2 ppk9 double mutant. PCR primer sequences are given below: for ssp2, 5′-GTGACACAACTAAGGATGTATACTATGGC-3′ and 5′-TCCTGCACGTTGTAAAGCTCGATACACAGC-3′; for ppk9, 5′-GGCCAAAGTGAAATGTAAGTACATGAGGTG-3′ and 5′-TCCTCATCACAGGGCCGAAACCGAGCTCC-3′; and for cbs2, 5′-ACAGGCCGAAATAATTTAGTAAGTAAGG-3′ and 5′-CCTTCGAGCTTCAAGTTTTCATCAACCACG-3′.
Invertase assay.Wild-type, ssp2Δ, ppk9Δ, and ssp2Δ ppk9Δ strains were grown to mid-log phase (optical density at 600 nm [OD600] = 1.0 to 2.0) under glucose-repressing conditions (see “Strains and media” above), after which a portion of the culture was collected (time zero), and the remainder was washed with deionized water and transferred to derepressing conditions. Aliquots of culture were harvested after 1, 3, and 6 h of incubation under derepressing conditions, washed with 10 mM NaN3, and assayed for secreted invertase as described previously (10).
Northern blot analysis.Wild-type, ssp2Δ, and ppk9Δ cells were processed for Northern blot analysis as described above (see “Invertase assay”). Total RNA was extracted using hot phenol (34). Total RNA samples (20 μg each) were separated on a formaldehyde-agarose gel and transferred to a nylon membrane. Northern hybridization was performed at 55°C using 0.7- to 1.7-kb DNA probes for inv1+, fbp1+, gld1+, and leu1+. Labeling of probes and hybridization were performed using an AlkPhos Direct system according to the manufacturer's instructions (GE Healthcare Co.). Primers used for amplifying probe fragments are listed below: for inv1+, 5′-TTTTCCAACGACACCATTCCGGAAGAGCGC-3′ and 5′-GAATGAGAGAAGTAAGGTTGGTCATGGGGT-3′; for fbp1+, 5′-TACTGCGATGAAGTCGAACGGATGCTGC-3′ and 5′-TGGGTACCAAATCCAGTTTGCGATCTCC-3′; for gld1+, 5′-GTTTTCATATGATTGGTCCTCGTCTTTGCGC-3′ and 5′-GTTTTGGATCCCTATGGATGAATGTCGGTC-3′; and for leu1+, 5′-GGTTGTTACTTTGGTGAGCGCACTGAGGAC-3′ and 5′-CAGCCTTAGTAATATCAGCGGTAGAAGCC-3′.
Immunoblot analysis.Cells were collected by centrifugation for 2 min and were resuspended in 1× sample buffer containing 62.5 mM Tris-HCl (pH 6.8), 25% glycerol, 2% sodium dodecyl sulfate (SDS), 0.01% bromophenol blue, and 5% β-mercaptoethanol. Cells were boiled for 3 min and homogenized using glass beads. Cell extracts were centrifuged at 20,000 × g for 3 min, separated by SDS-polyacrylamide gel electrophoresis (PAGE), and analyzed by immunoblotting.
RESULTS
Localization of S. pombe Snf1-like protein kinases.A search of the Sanger Centre Fission Yeast Genome Sequencing Project database (http://www.sanger.ac.uk/Projects/S_pombe/) for S. pombe homologues of S. cerevisiae SNF1 led to the discovery of SPCC74.03c/ssp2+ and SPAC23H4.02/ppk9+. Ssp2 and Ppk9 share 63% and 44% identities with S. cerevisiae Snf1, respectively, while the threonine residue essential for activation of the Snf1 kinase (7, 26) is conserved in both proteins (Fig. 1). The ssp2 mutation was initially identified to be a suppressor of ppe1 and sts5 mutations. Ppk9 has been designated a eukaryotic protein kinase catalytic domain-containing protein (1, 24). The S. cerevisiae Snf1 kinase forms a heterotrimeric enzyme complex with the γ subunit Snf4 and one of three related β subunits, Gal83, Sip1, or Sip2 (20). It was reported that S. pombe Cbs2 is similar to the γ subunit Snf4, that SPCC1919.03c is similar to β subunits, and that Cbs2 binds to two AMP-activated protein kinase (AMPK)-like catalytic subunits, Ssp2 and Ppk9 (12). However, it is not clear whether these Snf1-like protein kinases are involved in glucose derepression in S. pombe.
Alignment of S. cerevisiae Snf1, S. pombe Ssp2, and S. pombe Ppk9. Amino acid residues identical to those in Snf1 are shaded. The threonine residue essential for activation of the Snf1 kinase (210T) is indicated with an asterisk.
We constructed strains expressing Ssp2-GFP and Ppk9-GFP and analyzed the subcellular localization of these proteins. We found that Ssp2-GFP mainly localized in the nucleus both in glucose-starved cells and in cells grown in glucose-rich medium (Fig. 2A). On the other hand, Ppk9-GFP localized to the cytoplasm and nucleus both in glucose-starved cells and in cells grown in glucose-rich medium (Fig. 2B).
Localization of Ssp2 and Ppk9. (A) Wild-type cells expressing Ssp2-GFP were cultured on MM containing 8% glucose and were harvested during logarithmic growth (0 min). Cells were then shifted to MM containing 2% glycerol and 0.05% glucose as carbon sources and incubated for 30 or 60 min. (B) Wild-type cells expressing Ppk9-GFP were cultured on MM containing 8% glucose and were harvested during logarithmic growth (0 min). Cells were then shifted to MM containing 2% glycerol and 0.05% glucose as carbon sources and incubated for 30 or 60 min. Nomarski differential interference contrast micrographs (Nomarski), GFP fluorescence (GFP), and Hoechst 33324 staining (Hoechst) are shown.
Disruption of S. pombe Snf1-like protein kinase genes.To characterize the function of these two Snf1-like protein kinase genes, we generated ssp2Δ, ppk9Δ, and ssp2Δ ppk9Δ mutants. Both ssp2Δ and ppk9Δ mutants exhibited a normal growth rate on MM containing 2% glucose (Fig. 3A), indicating that neither gene is essential for vegetative growth. However, deletion of ssp2 prevented assimilation of glycerol (Fig. 3B) and failed to induce expression of gld1+, encoding glycerol dehydrogenase (25), on glycerol medium (Fig. 3C). On the other hand, the ppk9Δ mutant grew at a rate similar to that of the parent and exhibited high expression of gld1+ on glycerol medium. The growth rate and gld1+ expression of the ssp2Δ ppk9Δ double mutant were found to be similar to those of the ssp2Δ single mutant. These results indicate that ssp2+ is important for growth on glycerol and glucose derepression of gld1+ in S. pombe.
Glycerol assimilation and expression of gld1+ in ssp2Δ and ppk9Δ mutants. Wild-type (WT; filled circles), ssp2Δ (open circles), ppk9Δ (open triangles), and ssp2Δ ppk9Δ (open squares) strains were cultured on MM containing 2% glucose (A) and 2% glycerol and 0.1% glucose (B) as carbon sources. (C) Northern blot analysis of total RNA extracted from wild-type, ssp2Δ, ppk9Δ, and ssp2Δ ppk9Δ strains grown in MM containing 2% glucose or 2% glycerol and 0.05% glucose and harvested at an OD600 of 0.8.
Invertase derepression in ssp2Δ and ppk9Δ mutants.To investigate the role of the ssp2+ and ppk9+ genes on glucose derepression of invertase, growth profiles of mutant strains on medium containing sucrose as the sole carbon source were examined (Fig. 4A). Because the S. pombe inv1Δ mutant exhibits a defect in sucrose fermentation only in the presence of antimycin (41), an antibiotic that inhibits cytochrome c reductase within the electron transport chain, the assay was performed in the presence of 10 μg/ml antimycin. The growth rate of the ssp2Δ mutant on sucrose medium was considerably lower than that of the parent strain, while the ppk9Δ mutant grew at a rate similar to that of the parent strain. Furthermore, the growth rate of the ssp2Δ ppk9Δ double mutant was found to be similar to that of the ssp2Δ single mutant (Fig. 4A). Next, we determined the amount of secreted invertase in wild-type and mutant cells. In the parent strain, secreted invertase was not detected when cells were exposed to high levels of glucose (time zero) but began to rise within 1 h after glucose depletion and reached a peak after 3 h (Fig. 4B, filled circles). Elevation of invertase activity after glucose depletion was suppressed in ssp2Δ cells (open squares), while invertase activity in ppk9Δ cells (open triangles) was found to increase to the same level observed in the parent cells. In the ssp2Δ ppk9Δ double mutant, the profile of invertase induction was similar to that observed in the ssp2Δ mutant (open circles). The impaired invertase induction in the ssp2Δ mutant was complemented by overexpression of ssp2+ but not by overexpression of ppk9+ (data not shown). To examine whether inv1+ was derepressed in the ssp2Δ and ppk9Δ mutants by glucose depletion, Northern blot analysis was performed. In addition, the derepression profiles of fbp1+ were also analyzed. In the wild-type strain, expression of the two genes was completely repressed when cells were grown in glucose-rich medium (Fig. 4C, time zero), but expression of the two genes was significantly induced after glucose depletion. Induction of these genes by glucose depletion was partially defective in ssp2Δ cells (Fig. 4C). In the ppk9Δ mutant, the derepression profiles of the two genes were found to be similar to those in the parent strain (Fig. 4C). A more severe defect was not observed in the ssp2Δ ppk9Δ double mutant (data not shown).
Glucose derepression of invertase in ssp2Δ and ppk9Δ mutants. (A) Response of ssp2Δ and ppk9Δ mutants to sucrose availability. Identical volumes of 10-fold serial dilutions of exponentially growing wild-type parent strain, ssp2Δ, ppk9Δ, ssp2Δ ppk9Δ, and inv1Δ cells were spotted onto YES (left) or antimycin (Anti.; 10 μg/ml)-supplemented YES-sucrose medium (right) and incubated for 72 h at 30°C. (B) Glucose-derepression profiles of invertase in ssp2Δ and ppk9Δ mutants. Activity of secreted invertase in each strain was assayed at 0, 1, 3, and 6 h after shifting to medium containing 2% glycerol and 0.05% glucose as carbon sources. One unit of invertase activity is defined as the amount of enzyme which catalyzes production of 1 nmol of glucose per minute. Values are means ± SEMs (n = 3). (C) Northern blot analysis of inv1+ and fbp1+ in ssp2Δ and ppk9Δ mutants. Total RNA was harvested at 0, 3, and 6 h after glucose starvation (2% glycerol and 0.05% glucose) and subjected to analysis. Expression of leu1+ was determined as an internal control.
During the course of these studies, we observed additional phenotypes associated with the ssp2Δ strain, including heat sensitivity and a defect in sporulation, similar to that seen in the S. cerevisiae snf1 mutant (see Fig. S1 in the supplemental material) (4, 33, 42). In S. cerevisiae, it has been demonstrated that the Snf1 kinase controls meiosis via transcriptional regulation of the meiotic regulators Ime1 and Ime2 (18), whereas a function for the Snf1 kinase in the heat-stress response is uncertain. Our results indicate that multiple biological processes, including the heat-stress response and sexual development, are both controlled by the Snf1-like protein kinase Ssp2 in S. pombe.
Localization and phosphorylation of Scr1 in ssp2Δ and ppk9Δ mutants.Fission yeast Scr1, in cooperation with Tup11 and Tup12, which are involved in glucose-dependent transcriptional repression and chromatin alteration, represses transcription of inv1+ (41), fbp1+ (29), and gld1+ (25) in the presence of glucose. Scr1 localizes in the nucleus and binds to the regulatory elements of its target genes under glucose-rich conditions. Under glucose-rich conditions, nuclear Scr1 is quickly exported to the cytoplasm within 5 min of cell exposure to glucose starvation (13). To investigate the role of Ssp2 and Ppk9 in export of Scr1 from the nucleus, we constructed strains expressing Scr1-GFP (Fig. 5A) and analyzed the protein's subcellular localization (Fig. 5B). In the ppk9Δ mutant, as in the wild-type strain, Scr1 localized in the nucleus under glucose-rich conditions and in the cytoplasm during glucose starvation. On the other hand, Scr1 localized to the nucleus both in glucose-rich medium and during glucose starvation in mutants lacking ssp2+. Deletion of cbs2+, which encodes a γ subunit of the kinase complex that binds to Ssp2 and Ppk9, also abolished loss of a large proportion of Scr1 from the nucleus and reduced cell growth on glycerol- and sucrose-containing media (see Fig. S2 in the supplemental material).
Localization and phosphorylation of Scr1. (A) Constructs expressing Scr1-GFP and Scr1-3×FLAG. Scr1-GFP and Scr1-3×FLAG were integrated at the scr1+ locus using leu1+ as a selectable marker. EGFP, enhanced GFP. (B) Localization of Scr1-GFP in wild-type, ssp2Δ, ppk9Δ, ssp2Δ ppk9Δ, and cbs2Δ strains. Scr1-GFP-expressing cells were cultured on MM containing 8% glucose to an OD600 of 0.8 (R). Cells were then shifted to MM containing 2% glycerol and 0.05% glucose as carbon sources and incubated for 1 h (DR). Nomarski differential interference contrast micrographs (Nomarski), GFP fluorescence (GFP), and Hoechst 33324 staining (Hoechst) are shown. (C) Immunoblot analysis of the Scr1-3×FLAG protein. Protein extracts were prepared from wild-type, ssp2Δ, ppk9Δ, ssp2Δ ppk9Δ, and cbs2Δ strains expressing Scr1-3×FLAG grown in MM containing 8% glucose (R) or 2% glycerol and 0.05% glucose (DR) and subjected to SDS-PAGE. (D) Protein extracts prepared from wild-type cells grown in MM medium containing 8% glucose (R) or 2% glycerol and 0.05% glucose (DR) were preincubated with phage lambda protein phosphatase (λ-PPase).
We next examined whether Scr1 is phosphorylated in response to glucose availability. Treitel et al. reported that Mig1, a DNA-binding repressor of S. cerevisiae that represses many glucose-repressed genes, is differentially phosphorylated in response to glucose availability (43). Immunoblot analysis showed that Mig1 is phosphorylated under both repressing and derepressing conditions and that Mig1 from derepressed cells migrates slower in SDS-PAGE than that from repressed cells. Hirota et al. detected no visible changes in Scr1 protein levels or mobility in response to glucose availability (13). However, we found that Scr1 mobility was affected by culture conditions. Although the estimated molecular mass of the Scr1-3×FLAG protein was about 62 kDa, the molecular mass of Scr1 from a wild-type strain grown under repressing conditions was approximately 80 kDa and the protein obtained under depressing conditions was approximately 100 kDa (Fig. 5C). These bands corresponding to Scr1-3×FLAG were shifted to faster-migrating forms, approximately 62 kDa, after phosphatase treatment (Fig. 5D). These results show that Scr1 is phosphorylated both in glucose-starved cells and in cells grown in glucose-rich medium and is hyperphosphorylated upon depletion of glucose. The reason why phosphorylated Scr1 bands were not detected in the earlier study (13) remains uncertain. To test whether the Ssp2 and Ppk9 protein kinases are required for phosphorylation of Scr1, we constructed ssp2Δ, ppk9Δ, ssp2Δ ppk9Δ, and cbs2Δ mutants expressing Scr1-3×FLAG. In the ppk9Δ mutant, Scr1 was modified as in wild-type cells. In the ssp2Δ, ssp2Δ ppk9Δ, and cbs2Δ mutants, Scr1 was phosphorylated to the same level in cells grown in glucose-rich medium but was not hyperphosphorylated during glucose starvation. These results indicate that Ssp2 and Cbs2 are required for increased phosphorylation of Scr1 upon glucose deprivation but are not required for phosphorylation during growth in glucose-rich medium.
Mutation of putative Ssp2 phosphorylation sites in Scr1.Potential phosphorylation sites in Scr1 were analyzed by homology to budding yeast Mig1 on the basis of their similarity to the consensus substrate recognition sequence, which contains an arginine at position −3 and hydrophobic residues at positions −5 and +4 relative to the phosphorylated serine (6, 43). Mig1 possesses the following consensus serine residues: S278, S311, S312 S381, and S383. Treitel et al. reported that these serine residues appear to be phosphorylated in vivo, but mutation of these sites did not reduce phosphorylation of Mig1 as substantially as did mutation of SNF1, suggesting that Mig1 also contains additional serine residues that are substrates for Snf1 (44). S278, S311, S312, S381, and S383 of Mig1 are also conserved in Scr1 and correspond to residues S235, S332, S333, S408, and S410, respectively (Fig. 6A). We changed the phosphorylatable serine residues of Scr1 to alanine by site-directed mutagenesis (mutated sites are designated with upward-pointing arrowheads in Fig. 6A) and assayed for repression of invertase by the mutant proteins in response to glucose limitation. Under glucose-repressing conditions, the invertase activity of scr1Δ cells expressing the mutant scr1 alleles was repressed (data not shown). The derepressed invertase activity of scr1Δ cells expressing alleles harboring single or double mutations was as high as that in wild-type cells, but that of scr1Δ cells expressing the allele harboring five S-to-A substitutions (the quintuple allele) was significantly lower than that in wild-type cells (Fig. 6B). These results indicate that these five substitutions, S235A, S332A, S333A, S408A, and S410A, in Scr1 prevent glucose derepression of invertase in glucose-starved cells. While immunoblot analysis indicated that this mutated protein still displayed a glucose-dependent mobility shift, the shift was less pronounced (Fig. 6C). In addition, fluorescence microscopy revealed that cells expressing the quintuple allele fused to GFP produced a protein that localized to the nucleus both in glucose-rich medium and during glucose starvation. In contrast, Scr1-GFP harboring single or double S-to-A substitutions rapidly translocated to the cytoplasm upon glucose removal (Fig. 6D). These results indicate that potential phosphorylation sites in Scr1, on the basis of similarity to Mig1, appear to be phosphorylated in vivo and are essential for export of Scr1 from the nucleus to the cytoplasm as well as for glucose derepression during glucose starvation. However, mutation of these serine residues did not reduce phosphorylation of Scr1 as substantially as did deletion of ssp2+, suggesting that Scr1 also contains additional sites for Ssp2-dependent phosphorylation.
Effect of point mutations in Scr1 on glucose derepression of invertase, phosphorylation, and localization. (A) Alignment of S. cerevisiae Mig1 and S. pombe Scr1. Amino acid residues identical to those in Mig1 are shaded. Potential phosphorylation sites in Mig1 are indicated with an asterisk, and mutated sites in Scr1 are indicated with an arrowhead. (B) Glucose-derepression profiles of invertase in wild-type, ssp2Δ, and scr1Δ strains carrying a point-mutated scr1 gene. Strains were cultured in MM containing 8% glucose and shifted to MM containing 2% glycerol and 0.05% glucose as carbon sources and incubated for 3 h. Invertase activity of wild-type cells cultured under derepressing conditions was taken as 100%. Values are means ± SEMs (n = 3). (C) Immunoblot analysis of mutant Scr1-3×FLAG proteins. Wild-type cells expressing Scr1-3×FLAG or mutant Scr1-3×FLAG were cultured under glucose-repressing (R; 8% glucose) or glucose-derepressing (DR; 2% glycerol and 0.05% glucose) conditions. Proteins were extracted and subjected to SDS-PAGE. (D) Localization of mutant Scr1-GFP. Wild-type cells expressing Scr1-GFP or mutated Scr1-GFP were grown in MM containing 8% glucose to an OD600 of 0.8 (R). Cells were then shifted to MM containing 2% glycerol and 0.05% glucose as carbon sources and incubated for 1 h (DR). Nomarski differential interference contrast micrographs (Nomarski) and GFP fluorescence are shown.
DISCUSSION
In this report, the function of ssp2+ encoding an Snf1-like protein kinase was characterized. Our results indicate that Ssp2 is involved in derepression of invertase, a glucose-repressible enzyme required for sucrose utilization, and glycerol dehydrogenase, required for glycerol utilization. In contrast, Ppk9 is not involved in glucose derepression, even though Ppk9 also closely resembles budding yeast Snf1 and forms a complex with Cbs2 in S. pombe cells (12).
It has been reported that transcription of fbp1+ is repressed by glucose through activation of a cyclic AMP-dependent protein kinase A (PKA) pathway (2, 15, 16, 21) and is activated by glucose depletion through activation of a stress-activated protein kinase (SAPK) pathway (29, 36, 39). In addition, regulation of fbp1+ involves activities of two mutually antagonistic transcription factors, Scr1 and Rst2, that modulate the function of Tup1-like Tup11 and Tup12 proteins (15). It is reasonable to speculate that Snf1-like protein kinases contribute to the regulation of fbp1+ because the budding yeast counterparts of the above-described transcription factors (Mig1 and Adr1, respectively) are modulated by the Snf1 kinase (reviewed in reference 32). To our knowledge, these are the first observations that indicate that the Snf1-like protein kinase is a component of the transcriptional machinery that regulates fbp1+. It is notable that the glucose-derepression profiles of inv1+, fbp1+, and gld1+ in the ssp2Δ mutant were very similar, suggesting a common regulatory mechanism.
Scr1 is phosphorylated both in glucose-starved cells and in cells grown in glucose-rich medium and is hyperphosphorylated upon glucose removal in an Ssp2-dependent manner. This hyperphosphorylation causes export of Scr1 from the nucleus to the cytoplasm. Analysis of Scr1 on the basis of similarity to Mig1 showed that mutation of serine residues 235, 332, 333, 408, and 410 affects its phosphorylation and reduces Ssp2-dependent glucose derepression. Although the extent of Scr1 phosphorylation during glucose starvation is largely dependent on the Ssp2 kinase, Ssp2 may not be directly responsible for all phosphorylation of Scr1. Scr1 is phosphorylated under glucose-rich conditions independently of the Ssp2 kinase because Scr1 is phosphorylated to the same extent both in the ssp2Δ mutant and in a wild-type strain under glucose-rich conditions. It had been expected that ppk9+ would share functional redundancy with ssp2+ because disruption of ssp2+ failed to completely block the glucose derepression of invertase at either the enzymatic or transcriptional level, nor did it completely block the phosphorylation of Scr1 under glucose-rich conditions. However, we found no experimental evidence that ppk9+ is involved in either the glucose derepression or phosphorylation of Scr1. Our results indicate that the mechanism of derepression of invertase in S. pombe is different from that in S. cerevisiae, where an SNF1 mutation completely blocks the glucose derepression of invertase (4, 30). This raises the possibility that S. pombe may possess an Snf1-like protein kinase-independent signal transduction pathway that upregulates invertase expression.
Collectively, our results indicate that the Snf1-like protein kinase Ssp2 functions in multiple biological processes in S. pombe, including glucose derepression. While the function of Ppk9 is still unclear, it may play a role in important biological processes because the ppk9+ mutant has been reported to be sensitive to several toxins, including cytoskeletal poisons (1). While the current study is the first to our knowledge to describe a mechanism for Snf1-like protein kinase-dependent signaling in glucose repression/derepression in S. pombe, further work is clearly needed to identify additional components of the Snf1-like protein kinase pathway in order to better understand the machinery of glucose-mediated gene regulation in S. pombe.
ACKNOWLEDGMENTS
We thank K. Tanaka (Tokyo University, Japan) for providing S. pombe wild-type strain KJ100-7B, h90 leu1-32 ura4-D18.
This work was partly supported by the Project for Development of a Technological Infrastructure for Industrial Bioprocesses on R&D of New Industrial Science and Technology Frontiers by the Ministry of Economy, Trade & Industry (METI) and was sponsored by the New Energy and Industrial Technology Development Organization (NEDO). T.M. was supported in part by the Japanese Society for the Promotion of Science (JSPS).
FOOTNOTES
- Received 13 October 2011.
- Accepted 22 November 2011.
- Accepted manuscript posted online 2 December 2011.
Supplemental material for this article may be found at http://dx.doi.org/10.1128/EC.05268-11.
- Copyright © 2012, American Society for Microbiology. All Rights Reserved.
REFERENCES
