Yeast Mutants of Glucose Metabolism with Defects in the Coordinate Regulation of Carbon Assimilation

Archives of Biochemistry and Biophysics Vol. 365, No. 2, May 15, pp. 279 –288, 1999 Article ID abbi.1999.1163, available online at http://www.idealibrary.com on

Yeast Mutants of Glucose Metabolism with Defects in the Coordinate Regulation of Carbon Assimilation Richard A. Dennis, Mark Rhodey, and Mark T. McCammon 1 Department of Biochemistry and Molecular Biology, University of Arkansas for Medical Sciences, Little Rock, Arkansas 72205

Received December 1, 1998, and in revised form January 11, 1999

The enzymes of the glyoxylate cycle and gluconeogenesis are tightly regulated by transcriptional, posttranscriptional, and posttranslational mechanisms in Saccharomyces cerevisiae. We have previously identified four genes, ACN8, ACN9, ACN17, and ACN18, whose mutant phenotype includes two- to fourfold elevated levels of enzymes of the glyoxylate cycle, gluconeogenesis, and acetyl-CoA metabolism. The affected enzymes are elevated on nonfermentable carbon sources but are still fully repressed by glucose. Catabolite inactivation of the cytosolic malate dehydrogenase is not affected in the mutants. Instead, the phenotype appeared to be manifested primarily at the level of transcription. The ACN8, ACN17, and ACN18 genes were isolated by functional complementation of the respective mutant’s inability to utilize acetate as a carbon and energy source, and these genes were shown to encode subunits of metabolic enzymes. ACN8 was identical to FBP1, which encodes the gluconeogenic enzyme, fructose 1,6-bisphosphatase, while ACN17 and ACN18 were identical to the SDH2 and SDH4 genes, respectively, that encode subunits of the respiratory chain and tricarboxylic acid cycle enzyme, succinate dehydrogenase. Mutants defective in other glyoxylate cycle and gluconeogenic enzymes also display the elevated enzyme phenotype, indicating that the enzyme superinduction is a general property of gluconeogenic dysfunction. Glucose 6-phosphate levels were diminished in the mutants, suggesting that endogenous glucose synthesis can regulate the expression of gluconeogenic enzymes. © 1999 Academic Press Key Words: gluconeogenesis; glyoxylate cycle; tricarboxylic acid cycle; glucose regulation; yeast.

1 To whom corespondence should be addressed at the Department of Biochemistry and Molecular Biology, University of Arkansas for Medical Sciences, 4301 W. Markham St., Slot 516, Little Rock, AR 72205. Fax: (501) 686-8169. E-mail: [email protected].

0003-9861/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

Considerable progress has been made in defining the signaling pathways regulating carbon metabolism in Saccharomyces cerevisiae (15, 21, 24, 44, 53). Many of these mechanisms apply to the utilization of alternative carbohydrates, such as galactose, sucrose, or maltose, as well as to the metabolism of nonfermentable carbon sources, such as ethanol and acetate, that require oxidative metabolism and assimilation through the tricarboxylic acid (TCA) 2 and glyoxylate cycles and gluconeogenesis (14). While regulatory proteins exist that are specific for certain subsets of carbon source utilization pathways, many proteins function in a more global fashion. The regulation of glyoxylate cycle and gluconeogenic enzymes may be the most complex subset of glucose-regulated proteins. While much of the regulation is exerted at the level of gene expression, sophisticated mechanisms for posttranscriptional, allosteric, and posttranslational regulation also exist. The enzyme fructose 1,6-bisphosphatase (FBPase) is inhibited allosterically by the accumulation of AMP and fructose 2,6-bisphosphatase (27). The cytosolic enzymes are subject to inactivation by phosphorylation during glucose-induced adaptation to fermentation (14) that is mediated by a RAS-dependent cAMP-mediated signaling pathway (53). Phosphorylation is followed by ubiquitination (50) and rapid proteolytic degradation of these proteins in response to plentiful glucose supplies. While the mitochondrial succinate dehydrogenase proteins are not turned over in response to glucose, the mRNAs for the SDH1 and SDH2 2

Abbreviations used: TCA, tricarboxylic acid; PEPCK, phosphoenolpyruvate carboxykinase; MDH, malate dehydrogenase; CAT, carnitine acetyltransferase; ICL, isocitrate lyase ; CS, citrate synthase; b-gal, b-galactosidase; SDH, succinate dehydrogenase; FBPase, fructose 1,6-bisphosphatase; UAS, upstream activating sequences; CSRE, carbon source response element; PMSF, phenylmethylsulfonyl fluoride; Pipes, 1,4-piperazinebis(ethanesulfonic acid); DEPC, diethylpyrocarbonate; PEP, phosphoenolpyruvate. 279

280

DENNIS, RHODEY, AND MCCAMMON

genes are rapidly degraded in the presence of glucose (4, 5, 26). Succinate dehydrogenase plays a critical role in oxidative carbon assimilation since it oxidizes succinate that is generated by isocitrate lyase of the glyoxylate pathway. Much of the regulation of oxidative carbon assimilation occurs at the transcriptional level. These genes are extremely sensitive to glucose concentration (33), and rapid repression can occur by external addition of as little of 0.005% glucose (56). Glucose must be phosphorylated to glucose 6-phosphate for most glucose repression signaling to occur (28, 56), but the actual signaling molecules have not been defined. The promoters of these genes contain multiple negative and positive regulatory sites and have been best defined for the FBP1, PCK1, and ICL1 genes that encode FBPase, phosphoenolpyruvate (PEP) carboxykinase, and isocitrate lyase, respectively (32, 37, 39–41, 49, 55). Many of these regulatory sites are responsive to different concentrations of glucose (24, 33, 56). The upstream activating sequences (UAS) elements, sometimes defined as the carbon source response element (CSRE) (49), activate transcription only in the absence of glucose. Activation requires the function of the Snf1 (Cat1) kinase complex. One of the targets of Snf1 is the Cat8 protein that is specifically required for activation of oxidative carbon assimilation genes via the UAS/CSRE in association with other proteins (18, 42). A 27-kDa protein was recently identified as a protein capable of binding the UAS/CSRE of the ICL1 promoter (39). However, the relationship between this protein and Cat8 has not been defined. The Hap2/Hap3/Hap4/Hap5 complex also plays an important role in transcriptional induction under nonfermentable growth conditions. HAP binding motifs have been identified primarily in genes encoding proteins of oxidative phosphorylation, the TCA and glyoxylate cycles, and gluconeogenesis (21, 24, 43). Glucose repression of these genes also involves the Cyc8-Tup1 multisubunit complex that acts as a general negative regulator of a variety of other cell processes, including flocculation, mating, sporulation, and minichromosome maintenance (21). This complex functions by facilitating the action of various DNA-binding repressor proteins (23). One protein that interacts with Cyc8-Tup1 is Mig1, a C 2H 2 zinc finger protein. Mig1 directs glucose repression by binding to elements within their promoters. Mig1 is phosphorylated by the Snf1 kinase, which releases the promoters from glucose repression. As an approach to studying carbon metabolism in yeast, we previously isolated a collection of respiratorycompetent mutants that were unable to utilize acetate as a carbon source. These mutants have been termed the Acn 2 mutants (30). While acetate is a poor carbon source, it can be activated to acetyl-CoA for oxidative catabolism and for assimilation though the glyoxylate cycle-SDH-gluconeogenesis pathway. The original

TABLE I

Genotypes of S. cerevisiae Strains Used in This Study Strain

Genotype

Source

MMYO11

MATa ade2-1 can1-100 his3-11,15 leu2-3,112 trp1-1 ura3-1 Ole 1 MATa [rho 2] acn8-1(fbp1) MATaleu2 trp1 Dfbp1::LEU2 MATa ade2-1 can1-100 his3-11,15 leu2-3,112 trp1-1 ura3-1 fbp1::LEU2 acn9-1 acn9-1 acn17-1(sdh2) acn17-2 acn17-2 acn18-1(sdh4) acn18-1(sdh4) icl1 mdh2 pck1 cit1-1 Dsdh2::URA3 Dsdh4::TRP1 MATa ura3 his3 leu2 sdh2::URA3 MATa sdh4 leu2 ura3 his3

31

30p G23-2 a DFY457 FBP1KO G24-1 a G241-2A a G35-5 a A19-3 a A193-8A a A10-3 a A103-1C a G37-5 a G38-5 a N34-1 a C11L a DSDH2L a DSDH4L a pJL2 KC122 a

30 30 51 H-L. Chiang

30 G24-1 3 30p 30 30 A19-3 3 30 p 30 A10-3 3 30p 30 30 30 13 MMYO11 MMYO11 I. E. Scheffler 6

MMYO11 background; only changes are listed.

Acn 2 collection consisted of more than 100 strains that harbored mutations in as many as 43 different genes. Many of the genes identified to date encode metabolic enzymes in the TCA and glyoxylate cycles and gluconeogenesis. In addition, several mutants with regulatory defects were also observed. Five complementation groups, ACN8, ACN9, ACN17, ACN18, and ACN42, were grouped together with a common mutant phenotype: elevated levels of enzymes of the glyoxylate cycle, gluconeogenesis, and acetyl-CoA metabolism. Enzymes of glycolysis, b-oxidation, TCA cycle, invertase, and cytochrome c oxidase were not significantly or consistently elevated. A negative regulatory defect was proposed since enzyme levels increased when these ACN genes were mutated. In this report the regulatory defects of these genes have been dissected, and three of the genes have been defined as subunits of metabolic enzymes involved in oxidative carbon assimilation. MATERIALS AND METHODS Strains and culture conditions. The strains of S. cerevisiae used in this study are listed in Table I. Strains were grown on three basic types of media with a variety of carbon sources. Rich medium consisted of 1% yeast extract and 2% peptone supplemented with 2% glucose (YPD), 2% ethanol (YPE), or 3% glycerol (YPG). Synthetic medium consisted of 0.7% yeast nitrogen base (Difco) plus 2% glucose (SD) supplemented with amino acids and other nutrients to meet the auxotrophic requirements of the strains or to maintain selection of plasmid-containing transformants (16). Semisynthetic medium contained the same ingredients as synthetic plus 0.05% yeast extract;

YEAST MUTANTS OF GLUCOSE METABOLISM unless otherwise specified, the carbon sources used were 1% potassium acetate (SSAce) or 2% ethanol (SSE). Agar (2%, Difco) was added for plates. Genetic analysis of strains by tetrad analysis followed standard procedures (16). For determination of enzyme activities and for mRNA isolation, strains were typically precultured overnight in YPD. The cells were collected by centrifugation and resuspended in sterile water for inoculation into rich medium at an initial optical density of 0.05 OD 600/ml for wild type and 0.1 OD 600/ml for the mutant strains. Cultures were harvested at an OD 600/ml of 0.8 –1.5 for most experiments or at specific times. Cell pellets were collected by centrifugation, frozen in liquid nitrogen, and stored at 280°C for later use. Biochemical assays. Enzymatic analysis was performed on whole-cell extracts prepared by glass bead lysis (30). The lysis buffer usually consisted of 100 mM Tris–Cl, pH 7.4, 1 mM EDTA, 0.1% Triton X-100, and 1 mM phenylmethylsulfonyl fluoride (PMSF). Enzyme activities are presented as micromoles of substrate or product converted per milligram of total protein (2). All measurements of enzyme activities followed published procedures: phosphoenolpyruvate carboxykinase (PEPCK) (17), malate dehydrogenase (MDH) (29), carnitine acetyltransferase (CAT) (1), isocitrate lyase (ICL) (11), citrate synthase (CS) (46), and b-galactosidase (b-gal) (34). Measurement of FBPase (27) required a cell lysis buffer containing 50 mM imidazole, pH 7.5, 20 mM Tris–Cl, pH 7.5, 1 mM EDTA, 1 mM b-mercaptoethanol, and 2 mM PMSF. Succinate dehydrogenase (SDH) (54) was assayed on lysed mitochondria (30). For extraction of glucose 6-phosphate (10), a 25-ml culture was swirled into 150 ml of 60% methanol at 220°C. All steps were performed in an ethanol– dry ice bath with cold (220°C) reagents. Cells were pelleted by centrifugation in a precooled Beckman JA-14 rotor for 15 min. The pellet was resuspended in 4 ml of a 2 mM Pipes, 1 mM EDTA buffer extracted with 5 ml of cold 100% methanol and 10 ml cold chloroform. The mixture was cooled for 90 min in an ethanol– dry ice bath with frequent vortexing to prevent phase separation and freezing. The phases were separated by centrifugation, and the top layer was collected. The chloroform phase was backextracted with 2 ml methanol plus 2 ml Pipes–EDTA buffer. The combined aqueous phase was extracted with 35 ml diethyl ether before vacuum drying overnight. The extract was resuspended in 1.6 ml of 50 mM imidazole buffer, pH 6.6, frozen in liquid nitrogen, and stored at 280°C. Upon thawing, samples were microcentrifuged at 4°C for 15 min before analysis. Glucose 6-phosphate was quantitated biochemically (25). Identification of the ACN genes. The ACN8(FBP1), ACN17(SDH2), and ACN18(SDH4) genes were isolated by functional complementation of the acetate-negative growth phenotype of strains G23-2, A19-3, and A10-3, respectively. These strains were transformed using a lithium acetate procedure (48) with a yeast genomic library in the plasmid YpC50 (CEN ARS, URA3, AMP R) (45). Transformants on SD plates were replicated to SSAce plates. The plasmids from acetate 1 colonies were extracted (20) and transformed into E. coli strain TG1 (supE, hsdD5 thi D(lac-proAB) F9 [traD36 proAB1 lacIq lacZDM154]), made competent by a CaCl 2 procedure (47). Typically, the identity of the genomic insert was determined by standard dideoxy DNA sequencing using primers that flanked the BamHI site of YpC50. DNA analysis was performed using the Wisconsin Genetics Computer Group package. Individual open reading frames within the insert were subcloned into shuttle vectors (56) and assayed for their ability to complement the acetate phenotype of the yeast mutant strain under analysis. The identified gene was further confirmed as a wild-type copy of the defective ACN gene in two ways. First, yeast strains in which a given gene in question had been deleted were obtained (Table I), and these strains failed to complement the ACN strains in genetic crosses. Second, the enzyme activity encoded by the gene in question was assayed and was observed to be defective in the appropriate mutant strains.

281

LacZ studies. pICL1-LacZ and pCIT1-LacZ reporter plasmids were constructed for the indirect measurement of transcriptional initiation for each gene. The 59 regulatory regions of ICL1 (903 bp) and CIT1 (635 bp) were amplified from template plasmids containing the genes by polymerase chain reaction (PCR) and were subcloned in front of the LacZ gene in the plasmid, pSEYC102 (Amp R, CEN ARS, URA3), such that ICL1 and CIT1 were fused in-frame with the LacZ gene at codons 22 and 15, respectively. Primer pairs for the amplification of the ICL1 promoter region were (59-GTTCTCAAGGAGCACTACAG-39) and (59-ATTTCGGCAGGATCCGCATCTAGTTTTGC-39) and for the CIT1 promoter region were (59-GTTCAGGTACCCGCGTTAAGGGCG-39) and (59-TGTCTTGTGGA(G)TCCCTTGATAAGAAACT-39). These primers contained 100% homology to the templates except for the mutagenic underlined nucleotides and the extra nucleotide in parentheses which created a BamHI site on the 39 end of the fragments. PCR reactions were performed using 1 U of Vent Polymerase (New England Biolabs), 300 mM dNTPs, 0.5 mM each primer, and 200 ng plasmid DNA. The CIT1 reaction was supplemented with 3 mM MgSO 4. The ICL1 reaction was carried out at 94°C/1.5 min, 57°C/1 min, and 72°C/1 min. The CIT1 reaction required 96°C/1.5 min, 43°C/1 min, and 72°C/0.75 min. These parameters were repeated for 30 cycles and then a final extension at 72°C/5 min was added. Fragments were then subcloned into pSEYC102 at BglII/BamHI and KpnI/BamHI sites to create the plasmids pICL1-LacZ and pCIT1-LacZ, respectively. mRNA quantitation. Procedures for isolation of total RNA were modified from published procedures (3). Frozen pellets containing 15 OD 600 of cells were resuspended in 150 ml of LET buffer (25 mM Tris–Cl, 100 mM LiCl, and 100 mM EDTA). LET-equilibrated phenol (150 ml) was added along with 0.5 g of acid-washed, baked, glass beads. Cells were homogenized at 4°C for 2 3 30 s with 2-min rests in between. A 1:1 phenol:chloroform mixture (250 ml) and 250 ml of diethylpyrocarbonate (DEPC)-treated water was added before a final 30-s homogenization. The phases were next separated in a microfuge. The aqueous phase was again extracted with phenol:chloroform and then with chloroform. RNA was precipitated by adding 1/25 vol of 5 M LiCl and 2.5 vol ethanol and stored at 270°C. RNA was recovered by centrifugation, washed, air-dried, and resuspended in 100 ml DEPC-treated water. Steady-state levels of ICL1 and ACT1 mRNA were measured by an RNase protection assay using a commercially available kit (RPAII, Ambion). ClaI DNA fragments (ICL1, 771 bp; ACT1, 563 bp) from the coding regions of the genes were subcloned into the plasmid, pRS314, in the 39–59 orientation relative to the T3 promoter. Plasmids were amplified in the E. coli strain, DH5a (supE44 DlacU169 [f80 lacZDM15] hsdR17 recA1 endA1 gyrA96 thi-1 relA1). The template plasmids, pICL1-Cla and pACT1-Cla, were linearized with NdeI and BglII, respectively. Antisense RNA probes were labeled using CTP as the limiting nucleotide (2.5 mM [a- 32P]CTP and 22.5 mM cold CTP) during in vitro transcription with 20 units of T3 RNA polymerase (Ambion). Each gel-purified probe (8000 cpm) was precipitated with 2 mg of total yeast RNA and 10 mg of bacterial tRNA, resuspended in 20 ml of hybridization solution, and hybridized at 42°C overnight. Samples were then digested for 30 min at 37°C with a mixture containing 1 mg of RNase A and 25 ng of RNase T 1. Protected fragments of 368 and 282 bp for the ICL1 and ACT1 mRNAs, respectively, were electrophoretically separated on a 5% native polyacrylamide gel. The results were visualized using a Molecular Dynamics PhosphoImager 425B and quantitated using ImagQuant Version 3.3 software.

RESULTS

ACN mutants that overexpress gluconoegenic enzymes. Several complementation groups, ACN8, ACN9, ACN17, and ACN18, were previously defined with similar mutant phenotypes (30). First, these mutants were

282

DENNIS, RHODEY, AND MCCAMMON

FIG. 1. Glyoxylate cycle and gluconeogenic enzymes are elevated in the ACN mutants. Enzymes were assayed in whole cell lysates from strains grown on YPE: isocitrate lyase (ICL); malate dehydrogenase, including isozymes MDH-1 of the TCA cycle and cytosolic MDH-2 (MDH), phosphoenolpyruvate carboxykinase (PEPCK), and carnitine acetyltransferase (CAT). Activities and standard deviations were calculated based upon the average of two independent cultures that were each assayed in triplicate. Genotype (strain): WT (MMYO11), acn8 (G23-2), acn9 (G24-1), acn17 (G35-5), and acn18 (A10-3).

members of a collection of respiratory-competent strains that were unable to utilize acetate as a carbon source. Second, enzymes of several metabolic pathways were similarly affected in the mutant strains. Affected enzymes were elevated in the mutants compared to the wild-type parental strain. Figure 1 demonstrates this phenotype for the glyoxylate cycle enzyme isocitrate lyase, the gluconeogenic enzymes PEP carboxykinase and malate dehydrogenase MDH-2 and for the acetatemobilizing enzyme, carnitine acetyltransferase. The malate dehydrogenase activity shown in Fig. 1 includes both the TCA cycle MDH-1 and gluconeogenic MDH-2 isozymes, however, using isozyme-specific antisera, only the MDH-2 protein was elevated (data not shown). Strains with defects in the ACN8, ACN9, ACN17, and ACN18 genes were investigated further in order to define their biochemical defects. A fifth gene, ACN42, that was phenotypically grouped with these strains (30) was not included because the elevated enzyme phenotype was less evident and was not consistently observed. Isocitrate lyase was chosen as a model-af-

fected enzyme since it could readily be monitored in whole-cell lysates and lacked competing isozymes that might complicate analysis (19). Although an enzyme of the glyoxylate pathway that is normally considered to be in peroxisomes, isocitrate lyase is present in the cytosol of S. cerevisiae (7, 31). It appears to be regulated at the transcriptional and posttranslational levels in a manner similar to the cytosolic gluconeogenic enzymes (12, 38, 39, 49). These mutants were first analyzed to determine whether isocitrate lyase was still regulated by carbon source. Isocitrate lyase was repressed when the strains were cultured on glucose medium (Fig. 2A), indicating that it was still subject to carbon catabolite repression. The enzyme was elevated when the mutants were cultured on nonfermentable carbon sources, such as glycerol or ethanol. Maximal induction of the enzyme was observed when the strains were cultured on ethanol and acetate. Most subsequent experiments monitored activity from ethanol-grown cells since the mutants were able to grow with ethanol as a carbon source but not with acetate (30). The induction of isocitrate lyase was monitored during a shift from glucose repression to induction by ethanol (Fig. 2B). Isocitrate lyase induction followed accelerated kinetics in the acn8 mutant strain. Induction of the enzyme was apparent by 6 –7 h, or 2–3 h before induction in the wild type became apparent. The rate of induction was also slightly higher in the mutant strain. Both strains reached their fully induced levels of isocitrate lyase by approximately 16 h. The induction profiles of isocitrate lyase and PEP carboxykinase followed similar accelerated kinetics in the acn9, acn17, and acn18 mutant strains (data not shown). Catabolite inactivation of MDH-2 protein. Experiments were performed in order to determine whether the elevated levels of enzymes observed in these mutants were due to transcriptional or posttranscriptional defects. The cytosolic MDH-2 isozyme, as well as PEP carboxykinase, FBPase, and isocitrate lyase, is subject to regulated proteolytic turnover, termed catabolite inactivation, during the shift from nonfermentable metabolism back to fermentation (21). When these mutant strains were shifted from 1% acetate to 5% glucose, MDH-2 protein was rapidly degraded with a half-life of approx 50 min, confirming previous reports (35, 36). The half-life of this enzyme was longer in the ACN mutants. However, this appeared to be due to the elevated protein and not due to a defect in the turnover of the protein. When an extra copy of the MDH2 gene was integrated into the wild-type strain, similar levels of MDH-2 protein were observed compared to the ACN mutants. A longer half-life for MDH-2 during catabolite inactivation was observed in the strain with the extra copy of the MDH2 gene, and

YEAST MUTANTS OF GLUCOSE METABOLISM

283

FIG. 2. Glucose repression and induction of isocitrate lyase in the ACN mutants. (A) Isocitrate lyase activity (ICL) was measured from whole cell lysates prepared from strains grown on YPD (glucose), YPG (glycerol), or YPE (ethanol). Each assay was performed in triplicate. The strains are the same as those described in the legend to Fig. 1. (B) Strains MMYO11 and G23-2 were precultured on YPD, washed, and transferred into YPE. Aliquots of each strain were harvested at the times indicated, whole cell lysates were made, and isocitrate lyase was assayed.

this half-life was now comparable to the half-life observed in the ACN mutants (data not shown). The effect of MDH-2 protein level on the half-life of this protein was observed previously (35, 36). These results suggest that the regulated proteolytic degradation of the gluconeogenic enzymes was not defective in the ACN mutants. Isocitrate lyase activity is due to increased transcription of the ICL1 gene. Defects at the level of gene transcription were investigated in the ACN mutants. An ICL1 promoter-LacZ plasmid reporter was constructed such that b-galactosidase activity could be monitored as an indication of transcriptional activity of the ICL1 gene (12). A similar construct using the promoter of CIT1 was used as a control. After induction by ethanol, isocitrate lyase and ICL1-mediated b-galactosidase were similarly elevated for each ACN mutant relative to the wild type (Fig. 3). In contrast, citrate synthase and CIT1-mediated b-galactosidase activities were not significantly different between the wild-type and the mutant strains. While citrate synthase activity was greatly decreased in a cit1 mutant strain, b-galactosidase from CIT1-LacZ was normal. These results suggested that isocitrate lyase expression was affected at the level of transcription in all four ACN mutants. Transcription of the ICL1 gene was subsequently measured in the ACN mutants using RNase protection assays. As expected, isocitrate lyase activity was elevated in the four ACN mutants but not in a cit1 mutant (Fig. 4A). When normalized against the ACT1 mRNA, ICL1 transcript levels were also elevated in the ACN mutants (Fig. 4B). A strong correlation existed between the level of ICL1 mRNA and isocitrate lyase

FIG. 3. Isocitrate lyase and ICL1-b-galactosidase activities are similarly elevated in the ACN mutants. Wild-type and mutant strains were transformed with either pICL1-LacZ or pCIT1-LacZ plasmids. Cells from SD medium were transferred into YPE medium and harvested after approximately three cell divisions. Both isocitrate lyase (A) and b-galactosidase activities (B) were measured in lysates from transformants containing pICL1-LacZ, while citrate synthase (C) and b-galactosidase (D) were measured in transformants containing pCIT1-LacZ. b-Galactosidase activity was adjusted to compensate for plasmid loss during growth on YPE. Strains are the same as those described in the legend to Fig. 1 with the addition of cit1 (C11L).

284

DENNIS, RHODEY, AND MCCAMMON

FIG. 4. Isocitrate lyase and ICL1 mRNA are both elevated in the ACN mutants. (A) A comparison of isocitrate lyase activity and ICL1 mRNA levels in wild-type and mutant strains. Isocitrate lyase activity and ICL1 mRNA quantitation represent the average of triplicate assays from two independent YPE cultures. (B) Phosphoimage of protected mRNA fragments for ICL1 and ACT1 as detected from RNase protection assays. For quantitation in A, ICL1 densitometer readings were normalized against ACT1. Genotype (strain): WT (MMYO11), acn8 (G23-2), acn9 (G241-2A), acn17 (A193-8A), acn18 (A103-1C), and cit1 (C11L).

activity (Fig. 4A). These results confirm that the elevated isocitrate lyase activity observed in the ACN mutants is caused primarily by altered ICL1 transcription. The ACN8, ACN17, and ACN18 genes encode metabolic enzymes. The functional ACN8, ACN17, and ACN18 genes were isolated by complementation of the inability of the corresponding mutants to grow on acetate medium after the mutants had been transformed

with a yeast genomic plasmid library. The ends of the genomic inserts were sequenced and compared to the Saccharomyces genome database. This gave the size of the genomic insert as well as the number of genes that resided within the insert. These genes were tested for their ability to complement the corresponding mutation after subcloning fragments which contained individual open reading frames. This analysis revealed that the ACN8 gene was identical to FBP1, which encodes the gluconeogenic enzyme, FBPase. These results were confirmed in several ways. First, the acn8 mutations failed to complement Dfbp1 deletion mutations in genetic crosses. Second, the growth phenotypes of acn8 strains on various nonfermentable carbon sources were compared to Dfbp1 strain. Both the Dfbp1 and acn8-1 strains grew poorly on glycerol (YPG), and neither strain was able to grow with acetate as a carbon and energy source (Fig. 5A). Finally, FBPase activity was measured in four strains harboring different acn8 alleles. FBPase activity was not detectable in an Dfbp1 strain and ranged from 7 to 14% of wild type in the acn8 strains. As reported previously (30) and shown in this report, other gluconeogenic, glyoxylate cycle, or TCA cycle enzymes were not deficient in the acn8 mutant strains. These results confirmed that ACN8 was the structural gene for FBPase. Similarly, the ACN17 and ACN18 genes were shown to be identical to the SDH2 and SDH4 genes, respectively. The SDH2 and SDH4 genes encode the iron– sulfur and membrane anchor subunits of succinate dehydrogenase, an inner mitochondrial membrane protein of the TCA cycle and a component of the respiratory complexes (complex II). Strains harboring acn17 alleles failed to complement Dsdh2 deletion strains, while acn18 mutants failed to complement Dsdh4 deletion mutants. Growth phenotypes were compared between Dsdh2 and Dsdh4 mutants and the acn17 and acn18 mutants. None of the mutants grew on a semisynthetic acetate medium (Fig. 5B). However, acn17 and acn18 strains were able to grow on glycerol medium, while the deletion mutants could not. Succinate dehydrogenase activity was 30% of the wild-type levels in an acn17 strain but was not detectable in an Dsdh2 strain. Deletion of the SDH2 gene resulted in an increase in isocitrate lyase activity (Fig. 6), and the elevated levels were comparable to those observed from an acn17 strain. Many other TCA cycle, glyoxylate cycle, and gluconeogenic enzymes were not deficient in the acn17 and acn18 strains. Taken together, these results indicate that the acn17 and acn18 alleles encode succinate dehydrogenase subunits that retain partial activity. The observation that the ACN8, ACN17, and ACN18 genes encode subunits of metabolic enzymes prompted an inquiry to determine whether mutations in genes encoding other related metabolic enzymes also re-

YEAST MUTANTS OF GLUCOSE METABOLISM

285

FIG. 5. Growth phenotypes of ACN mutants and corresponding deletion mutants. Strains were grown on YPD medium. 2 OD 600 units were pelleted and resuspended in 1 ml of water. The culture was serially diluted in 10-fold steps, and 10 ml of each dilution was spotted onto YPG and SSAce plates. Plates were photographed after incubation at 30°C for 4 days. (A) The acn8(fbp1) and Dfbp1 mutants grow on YPG but not on SSAce medium. Genotype (strains): WT (MMYO11), acn8 (G23-2), and Dfbp1(FBP1KO). (B) The growth phenotypes of the acn17(sdh2) and acn18(sdh4) mutants are different from the corresponding deletion mutants. Genotype (strains): WT (MMYO11), acn17 (A193-8A), Dsdh2 (DSDH2L), acn18 (A103-1C), and Dsdh4 (DSDH4L).

sulted in elevated enzyme levels. Mutants with defects in the ICL1, PCK1, and MDH2 genes (30) were cultured on ethanol, and isocitrate lyase, PEPCK, malate dehydrogenase, and carnitine acetyltransferase activities were assayed in lysates. These metabolic defects resulted in elevated levels of other glyoxylate cycle and gluconeogenic enzymes (Fig. 7). In contrast, elevated isocitrate lyase activity was not observed in a cit1 mutant defective in the TCA cycle citrate synthase (Figs. 3 and 4), suggesting that the elevation of the glyoxylate cycle and gluconeogenic enzymes is not a general property of TCA cycle gene defects. Conversely, TCA cycle enzymes did not appear to be elevated in the ACN mutants. Levels of malate dehydrogenase, citrate synthase, aconitase, fumarase, succinate dehydrogenase, NAD 1-dependent isocitrate dehydrogenase, and the non-TCA cycle enzyme NADP 1-dependent isocitrate dehydrogenase were not elevated in mitochondria

from acn8 or acn17 strains (8). These results confirm that the TCA cycle enzymes are not affected by the metabolic mutants of oxidative carbon assimilation (30). Glucose 6-phosphate concentrations are decreased in the ACN mutants. Glucose 6-phosphate levels were measured to test if gluconeogenic activity was defective in these strains. Glucose 6-phosphate was chosen for two main reasons. First, it is a major end product of gluconeogenesis. Second, since glucose must be phosphorylated to glucose 6-phosphate in order to signal for glucose repression, levels of this metabolite can be taken as an estimate of glucose metabolites available for glucose repression. Glucose 6-phosphate concentrations were highest when the wild-type strain was cultured on glucose medium and dropped approximately 75% when cultured on ethanol (Table II). Glucose

286

DENNIS, RHODEY, AND MCCAMMON TABLE II

Glucose 6-Phosphate Levels in Mutants of Oxidative Carbon Assimilation Genotype of strain

Carbon source

Glucose 6-phosphate (pmol/mg protein)

WT WT cit1 fbp1(acn8) acn9 sdh2(acn17) sdh4(acn18)

Glucose Ethanol Ethanol Ethanol Ethanol Ethanol Ethanol

9.08 6 2.44 2.56 6 0.15 1.83 6 0.22 1.13 6 0.37 0.70 6 0.33 ,0.31 ,0.31

Note. Strains were cultured on YPD (glucose) or YPE (ethanol) in duplicate. Glucose 6-phosphate was extracted and measured in triplicate as described under Materials and Methods. Genotype (strain): same as in Fig. 1 except for cit1 (C11L). FIG. 6. An acn17(sdh2) mutant is defective in succinate dehydrogenase activity. Enzymes were measured in triplicate from YPEgrown strains: Genotype (strains): WT (MMYO11), acn17 (G35-5), and Dsdh2 (DSDH2L). n.d. not detectable.

6-phosphate levels decreased further in the ACN mutants. Metabolite levels were approximately 44% of wildtype levels in an acn8(fbp1) strain, 27% in an acn9 strain,

and 12% for both the acn17(sdh2) and acn18(sdh4) strains. By comparison, glucose 6-phosphate levels decreased about 30% in a cit1 strain. These results confirmed that gluconeogenic activity was impaired in these strains and suggested a mechanism for the increased transcription of the affected genes. Gluconeogenic activity may be self-regulating through the endogenous synthesis of glucose metabolites that may signal transcriptional repression, especially during growth on nonfermentable carbon sources. DISCUSSION

FIG. 7. Enzymes of carbon assimilation are elevated in glyoxylate cycle and gluconeogenesis mutants. Isocitrate lyase (ICL), malate dehydrogenase (MDH), PEP carboxykinase (PEPCK), and carnitine acetyltransferase (CAT) were measured in triplicate in YPE-grown strains. Genotype (strains): WT (MMYO11), icl1 (G37-5), mdh2 (G38-5), pck1 (N34-1), and acn8 (G23-2).

We report a previously undescribed type of glucose regulation controlling the enzymes of the glyoxylate cycle, glucoenogenesis, and acetyl-CoA mobilization. Mutations in several genes encoding enzymes of carbon assimilation resulted in the hyperinduction of glyoxylate cycle and gluconeogenic enzymes. This phenotype was traced to aberrant and elevated transcription of the affected enzymes. The enzyme defects in these mutants resulted in a decrease in cellular glucose 6-phosphate, which is an end product of gluconeogenesis and is also a metabolite required to signal for glucose repression. We believe that the diminished level of glucose 6-phosphate results in the elevated enzyme phenotype. The gluconeogenic genes may be the most sensitive of the glucose-regulated genes to repression by very low concentrations of externally added glucose (33, 56), and it appears that endogenous synthesis of glucose can also exert a significant effect on transcription. The gluconeogenic mutations affect their own gene expression in this way, but they do not severely affect other glucose-regulated enzymes, such as invertase, cytochrome oxidase, b-oxidation, and TCA cycle enzymes (8, 30). These results add new insights into the regulation of gluconeogenic enzymes and suggest that an active monitoring system is capa-

YEAST MUTANTS OF GLUCOSE METABOLISM

ble of responding to changes in gluconeogenic activity by altering gene expression of carbon assimilation enzymes. Another interesting finding from these studies was the significant contribution of succinate dehydrogenase to gluconeogenic activity. Succinate dehydrogenase is situated at the intersection of three metabolic pathways. As part of the TCA cycle, succinate dehydrogenase is involved in the catabolism of acetyl-CoA, which produces reducing equivalents to drive oxidative phosphorylation. Succinate dehydrogenase is also one of the respiratory complexes (complex II), where it functions to transfer electrons from FADH 2 to ubiquinone. Succinate dehydrogenase is also intimately involved in carbon assimilation. The glyoxylate cycle enzyme, isocitrate lyase, generates succinate, which must be metabolized by succinate dehydrogenase for gluconeogenesis. Succinate dehydrogenase regulation shares many properties with gluconeogenic enzymes that are distinct from other TCA cycle or respiratory complex enzymes. As reported here, succinate dehydrogenase mutants display elevated gluconeogenic enzymes, while TCA cycle mutants, such as cit1, contain normal levels of these enzymes. In many experiments a cit1 strain was used as a negative control in which the elevated enzymes were not observed. In sharp contrast, respiration mutations prevented induction of gluconeogenic enzymes. Mutations in ubiquinone biosynthesis, cytochrome bc 1, and ATP synthase were unable to induce isocitrate lyase and PEP carboxykinase during growth on ethanol medium (22, 41). In fact, one mutant in ubiquinone biosynthesis, coq7(cat5), was selected and characterized for its inability to induce gluconeogenic enzymes (22, 41). Another aspect of succinate dehydrogenase regulation that groups it with gluconeogenic enzymes and distinguishes it from TCA cycle or respiratory complexes is the effect of sudden glucose abundance. This results in glucose repression of many oxidative genes and a gradual dilution of the affected enzymes as the fermenting cells divide. In contrast, glucose addition results in the rapid degradation of the mRNAs for the SDH1 and SDH2 genes. Message half-lives decrease from more than 30 min on glycerol medium to less than 5 min in the presence of glucose (4, 26). The cytosolic gluconeogenic enzymes also decrease rapidly in the presence of abundant glucose levels. However, this is due to inactivation by phosphorylation and ubiquitin-dependent proteolysis of these enzymes (21, 50, 53). The two processes of mRNA turnover and proteolytic degradation may be coordinated to achieve the same effect of rapidly decreasing the amount of enzymes involved in de novo glucose synthesis when alternative sources of glucose are present. However, it is not known how or if these processes are regulated.

287

Transcriptional repression of gluconeogenic enzymes by very small concentrations of glucose requires hexose phosphorylation since it is defective in mutants devoid of hexokinase activity (56). The level of glucose 6-phosphate decreased by 75% during gluconeogenic growth on ethanol compared to glucose-dependent growth (Table II), and this metabolite was diminished even further in succinate dehydrogenase or FBPase mutants. If glucose 6-phosphate is responsible for modulating the expression of these enzymes, then the critical cellular threshold concentration must be very close to the levels observed in normal cells growing on ethanol. The elevated enzyme phenotype is not observed when cellular glucose 6-phosphate is 70% of wild type in a cit1 strain but does become apparent when metabolite levels are less than 50% of wild type in an fbp1(acn8) mutant. This suggests that the cell is capable of responding to a relatively small change in glucose metabolites by altering expression of genes involved in endogenous glucose synthesis. While this report has provided new insight into the role of endogenous glucose synthesis in glucose repression, many other details of the glucose signaling pathways remain obscure, despite long and intense efforts by a number of investigators (15, 21, 24, 44, 53). It is not clear whether glucose 6-phosphate is the actual regulatory metabolite signaling glucose repression or what the signaling factors are that monitor glucose abundance. The use of metabolic mutants defective in oxidative carbon assimilation should be useful in the analysis of glucose signaling pathways. By eliminating endogenously synthesized glucose metabolites, the fully derepressed state within the cell can be established. This may allow for a more sensitive evaluation of both cis- and trans-acting factors involved in glucosemediated gene expression. The fourth complementation group analyzed in this report was ACN9. This gene has also been cloned by functional complementation. The ACN9 gene encodes a protein of unknown function (9). However, since it shares several mutant phenotypes with genes encoding enzymes of oxidative carbon assimilation, the ACN9 gene product may also play an important role in glucoenogenesis. ACKNOWLEDGMENTS The authors thank Dr. Lee McAlister-Henn for the MDH-2 antiserum and Drs. Hui-Ling Chiang, Dan G. Fraenkel, and Immo E. Scheffler for yeast strains used in this report. This work was funded by the National Science Foundation (MCB-9604225), the Arkansas Science and Technology Authority, and the University of Arkansas for Medical Sciences Graduate Student Research Fund.

REFERENCES 1. Bieber, L. L., and Markwell, M. A. K. (1981) Methods Enzymol. 71, 351–358.

288

DENNIS, RHODEY, AND MCCAMMON

2. Bradford, M. (1976) Anal. Biochem. 72, 248 –254. 3. Caponigro, G., Muhlrad, D., and Parker, R. (1993) Mol. Cell. Biol. 13, 5141–5148. 4. Cereghino, G. P., Atencio, D. P., Saghbini, M., Beiner, J., and Scheffler, I. E. (1995) Mol. Biol. Cell 6, 1125–1143. 5. Cereghino, G. P., and Scheffler, I. E. (1996) EMBO J. 15, 363– 374. 6. Chapman, K. B., Solomon, S. D., and Boeke, J. D. (1992) Gene 118, 131–136. 7. Chaves, R. S., Herrero, P., Ordiz, I., Delbrio, M. A., and Moreno, F. (1997) Gene 198, 165–169. 8. Dennis, R. A. (1998) Ph.D. dissertation, University of Arkansas for Medical Sciences. 9. Dennis, R. A., and McCammon, M. T. (1999) Eur. J. Biochem., in press. 10. de Koning, W., and van Dam, K. (1992) Anal. Biochem. 204, 118 –123. 11. Dixon, G. H., and Kornberg, H. L. (1959) Biochem. J. 72, 3. 12. Fernandez, E., Fernandez, M., Moreno, F., and Rodicio, R. (1993) FEBS Lett. 333, 238 –242. 13. Gadde, D. M., and McCammon, M. T. (1997) Arch. Biochem. Biophys. 344, 139 –149. 14. Gancedo, C., and Serrano, R. (1989) in The Yeasts (Rose, A. H., and Harrison, J. S., Eds.), 2nd ed., Vol. 3, pp. 205–209, Academic Press, San Diego, CA. 15. Gancedo, J. M. (1992) Eur. J. Biochem. 206, 297–313. 16. Guthrie, C., and Fink, G. R. (1991) Methods Enzymol. 194, 1–933. 17. Hansen, R. J., Hinze, H., and Holzer, H. (1976) Anal. Biochem. 74, 576 –584. 18. Hedges, D., Proft, M., and Entian, K-D. (1995) Mol. Cell. Biol. 15, 1915–1922. 19. Heinisch, J. J., Valdes, E., Alvarez, J., and Rodicio, R. (1996) Yeast 13, 1285–1296. 20. Hoffman, C. S., and Winston, F. (1987) Gene 57, 267–272. 21. Johnston, M., and Carlson, M. (1992) in The Molecular and Cellular Biology of the Yeast Saccharomyces: Gene Expression (Broach, J. R., Pringle, J. R., and Jones, E. W., Eds.), pp. 193– 281, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 22. Jonassen, T., Proft, M., Randez-Gil, F., Schultz, J. R., Marbois, B. N., Entian, K-D., and Clarke, C. F. (1998) J. Biol. Chem. 273, 3351–3357. 23. Keleher, C. A., Redd, M. J., Schultz, J., Carlson, M., and Johnson, A. D. (1992) Cell 68, 709 –719. 24. Klein, C. J. L., Olsson, J. L., and Nielsen, J. (1998) Microbiology 144, 13–24. 25. Lang, G., and Michal, G. (1983) in Methods of Enzymatic Analysis (Bergmeyer, H. A., Ed.) pp. 1238 –1242, Verlag Chemie, Weinheim. 26. Lombardo, A., Cereghino, G. P., and Scheffler, I. E. (1992) Mol. Cell. Biol. 12, 2941–2948. 27. Marcus, F., Rittenhouse, J., Moberly, L., Edelstein, I., Hiller, E., and Rogers, D. T. (1988) J. Biol. Chem. 263, 6058 – 6063. 28. Ma, H., Bloom, L. M., Walsh, C. T., and Botstein, D. (1989) Mol. Cell. Biol. 9, 5643–5649.

29. McAlister-Henn, L., and Thompson, L. M. (1987) J. Bacteriol. 169, 5157–5166. 30. McCammon, M. T. (1996) Genetics 144, 57– 69. 31. McCammon, M. T., Veenhuis, M., Trapp, S. B., and Goodman, J. M. (1990) J. Bacteriol. 172, 5816 –5827. 32. Mercado, J. J., and Gancedo, J. M. (1992) FEBS Lett. 311, 110 –114. 33. Mercado, J. J., Smith, R., Sagliocco, F. A., Brown, A. J. P., and Gancedo, J. M. (1994) Eur. J. Biochem. 224, 473– 481. 34. Miller, J. H. (1972) Experiments in Molecular Biology, pp. 352– 355, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 35. Minard, K. I., and McAlister-Henn, L. (1992) J. Biol. Chem. 267, 17458 –17464. 36. Minard, K. I., and McAlister-Henn, L. (1994) Arch. Biochem Biophys. 315, 302–309. 37. Niederacher, D., Schuller, H-J., Grzesitza, D., Gutlich, H., Hauser, H. P., Wagner T., and Entian, K. D. (1992) Curr. Genet. 22, 363–370. 38. Ordiz, I., Herrero, P., Rodicio, R., and Moreno, F. (1996) FEBS Lett. 385, 43– 46. 39. Ordiz, I., Herrero, P, Rodicio, R., Gancedo, J. M., and Moreno, F. (1998) Biochem. J. 329, 383–388. 40. Proft, M., Kotter, P., Hedges, D., Bojunga, N., and Entian, K-D. (1995) Mol. Gen. Genet. 246, 367–373. 41. Proft, M., Kotter, P., Hedges, D., Bojunga, N., and Entian, K-D. (1995) EMBO J. 14, 6116 – 6126. 42. Rahner, A., Scholer, A., Martens, E., Gollwitzer, B., and Schuller, H. J. (1996) Nucleic Acids Res. 24, 2331–2337. 43. Randez-Gil, F., Bojunga, N., Proft, M., and Entian, K-D. (1997) Mol. Cell. Biol. 17, 2502–2510. 44. Ronne, H. (1995) Trends Genet. 11, 12–17. 45. Rose, M. D., and Broach, J. R. (1991) Methods Enzymol. 194, 195–230. 46. Rosenkrantz, M., Alam, T., Kim, K-S., Clark, B. J., Srere, P. A., and Guarente, L. P. (1986) Mol. Cell. Biol. 6, 4509 – 4515. 47. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd ed., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 48. Schiestl, R. H., and Gietz, R. D. (1989) Curr. Genet. 16, 339 –346. 49. Scholer, A., and Schuler, H. (1994) Mol. Cell. Biol. 14, 3613– 3622. 50. Schork, S. M., Thumm, M., and Wolf, D. H. (1996) J. Biol. Chem. 270, 26446 –26450. 51. Sedivy, J. M., and Fraenkel, D. G. (1985) J. Mol. Biol. 186, 307–319. 52. Sikorski, R. S., and Hieter, P. (1989) Genetics 122, 19 –27. 53. Thevelein, J. M. (1991) Mol. Microbiol. 5, 1301–1307. 54. Veeger, C., Dervartanian, D. V., and Zeylemaker, W. P. (1969) Methods Enzymol. 13, 81–90. 55. Vincent, O., and Gancedo, J. M. (1995) J. Biol. Chem. 270, 12832–12838. 56. Yin, Z., Smith, R. J., and Brown, A. J. P. (1996) Mol. Microbiol. 20, 751–764.