Transcriptional Regulation of the Schizosaccharomyces pombe Malic Enzyme Gene, mae2 *

The NAD-dependent malic enzyme fromSchizosaccharomyces pombe catalyzes the oxidative decarboxylation of l-malate to pyruvate and CO2. Transcription of the S. pombe malic enzyme gene, mae2, was studied to elucidate the regulatory mechanisms involved in the expression of the gene. No evidence for substrate-induced expression of mae2 was observed in the presence of 0.2% l-malate. However, transcription ofmae2 was induced when cells were grown in high concentrations of glucose or under anaerobic conditions. The increased levels of malic enzyme may provide additional pyruvate or assist in maintaining the redox potential under fermentative conditions. Deletion and mutation analyses of the 5′-flanking region of the mae2gene revealed the presence of three novel negativecis-acting elements, URS1, URS2, and URS3, that seem to function cooperatively to repress transcription of the mae2gene. URS1 and URS2 are also present in the promoter region of theS. pombe malate transporter gene, suggesting co-regulation of their expression. Furthermore, two positive cis-acting elements in the mae2 promoter, UAS1 and UAS2, show homology with the DNA recognition sites of the cAMP-dependent transcription factors ADR1, AP-2, and ATF (activating transcription factor)/CREB (cAMP response element binding).

The NAD-dependent malic enzyme from Schizosaccharomyces pombe catalyzes the oxidative decarboxylation of L-malate to pyruvate and CO 2 . Transcription of the S. pombe malic enzyme gene, mae2, was studied to elucidate the regulatory mechanisms involved in the expression of the gene. No evidence for substrate-induced expression of mae2 was observed in the presence of 0.2% L-malate. However, transcription of mae2 was induced when cells were grown in high concentrations of glucose or under anaerobic conditions. The increased levels of malic enzyme may provide additional pyruvate or assist in maintaining the redox potential under fermentative conditions. Deletion and mutation analyses of the 5-flanking region of the mae2 gene revealed the presence of three novel negative cis-acting elements, URS1, URS2, and URS3, that seem to function cooperatively to repress transcription of the mae2 gene. URS1 and URS2 are also present in the promoter region of the S. pombe malate transporter gene, suggesting co-regulation of their expression. Furthermore, two positive cisacting elements in the mae2 promoter, UAS1 and UAS2, show homology with the DNA recognition sites of the cAMP-dependent transcription factors ADR1, AP-2, and ATF (activating transcription factor)/CREB (cAMP response element binding).
The fission yeast Schizosaccharomyces pombe efficiently degrades L-malate to CO 2 under aerobic conditions and to ethanol and CO 2 under anaerobic conditions (1). Cells of S. pombe are not able to utilize malate as the sole energy source or incorporate the malate into biomass (2) and therefore require glucose or other carbon sources for the energy-dependent transport and efficient degradation of malic acid (3). Three enzymes are involved in malate degradation in S. pombe, namely the malate transporter, malic enzyme, and malate dehydrogenase (4). The transporter, encoded by the mae1 gene (5), uses an H ϩ -symport system for the active transport of L-malate, and the NAD-dependent malic enzyme (EC 1.1.1.38) catalyzes the oxidative decarboxylation of L-malate to pyruvate and CO 2 . The mitochondrial malate dehydrogenase oxidizes L-malate to oxaloacetate in the tricarboxylic acid cycle and is responsible for 10% of the degradation of malate under aerobic conditions. Molecular analysis of the S. pombe malic mae2 gene showed a high degree of homology with malic enzymes from various organisms (6). Eight highly conserved regions were identified in malic enzymes, including the binding sites for L-malate and the dinucleotide co-factors NAD(P) ϩ (7,8). Although the secondary structure of malic enzymes is highly conserved, the coenzyme specificity (NAD ϩ or NADP ϩ ) and cellular localization (cytosolic or mitochondrial) are strongly linked to their regulation and metabolic function (9). Cytosolic NADP-dependent malic enzymes play an important role in lipid metabolism in higher eukaryotes (10), whereas NAD-dependent malic enzymes provide mitochondrial NADH for electron transport or cytosolic NADH for reductive power for other metabolic pathways (9). The degradation of malate in yeasts such as Candida utilis and Hansenula anomala that can utilize intermediates of the tricarboxylic acid cycle as the only source of carbon was reported to be subject to glucose repression and induction by L-malate (11). However, no substrate induction or glucose repression was observed in species such as S. pombe, Saccharomyces cerevisiae, and Zygosaccharomyces bailii that cannot utilize malate as the only source of carbon. A mitochondrial NADP-dependent malic enzyme was recently cloned from S. cerevisiae (12), and a role in the provision of intramitochondrial NADPH or pyruvate under anaerobic conditions was ascribed to the enzyme. DNA sequence analysis of the S. pombe malic enzyme gene (6) did not indicate the presence of a mitochondrial targeting signal, suggesting that it functions in the cytosol where it catalyzes the decarboxylation of malate to pyruvate. This could serve to regulate the levels of intracellular malate for its function in shuttle systems with oxaloacetate or aspartate or to provide NADH or pyruvate for other metabolic pathways.
Because of the inability of S. pombe cells to utilize L-malate as the only source of carbon and energy, the effective transport and intracellular degradation of L-malate by the yeast is intriguing. Transcription of the malic enzyme gene was therefore studied to elucidate the transcriptional regulation and physiological importance of malate degradation in S. pombe. Our results showed increased levels of mae2 transcription when cells were grown in high concentrations of glucose (8%) or under anaerobic conditions. A detailed analysis of the mae2 promoter enabled us to identify three novel negative cis-acting elements, URS1, URS2, and URS3, as well as two positive cis-acting elements that have sequence similarity to the binding sites of cAMP-dependent transcription factors.

EXPERIMENTAL PROCEDURES
Strains, Plasmids, and Media-All strains and plasmids used in this study are listed in Table I. Yeast strains were cultured in SC Ϫleu medium (0.17% yeast nitrogen base without amino acids and ammonium sulfate (Difco Laboratories, Detroit, MI), 0.5% (NH 4 ) 2 SO 4 , amino acid supplements, and 2% glucose, unless stated otherwise). Media were supplemented with 0.2% L-malic acid (Sigma) when required.
Escherichia coli cells were grown in Luria broth with 200 mg/liter ampicillin for plasmid selection.
DNA Isolation, Transformation, and Analysis-Standard procedures were used for plasmid isolation and transformation of competent cells of E. coli (13), S. pombe leu Ϫ (14), or S. cerevisiae (15). Standard procedures were used for DNA sequencing (13), and computerized analysis (16) was done to identify putative regulatory elements. For primer extention analysis, primers MV2 and MV3 ( Fig. 1 and Table II) were annealed to 50 g of total RNA and treated with Superscript Moloney murine leukemia virus reverse transcriptase in the presence of [␣-32 P]dATP (17). The products were loaded on a 6% acrylamide urea gel with DNA sequence ladders obtained with the respective primers.
Malic Enzyme Activity Assays-Wild type and transformed strains of S. pombe and S. cerevisiae were cultured overnight in 10 ml of SC Ϫleu medium with or without 0.2% L-malate. Crude cell extracts and enzyme assays were done as described by Osothsilp and Subden (18). Enzyme activities are given as mol of NADH produced/min/mg of protein as determined by Bradford assays (Bio-Rad).
RNA Isolation and Northern Analysis-Total RNA was isolated from wild type S. pombe cells grown overnight in SC Ϫleu medium with 1% glucose to A 600 of 0.8. Cells were harvested and resuspended in fresh medium with 2% glucose or 2% glycerol, ethanol, with one set of glucose cultures sealed with mineral oil and incubated stationary so as to mimic anaerobic conditions. Incubation was continued, and total RNA was isolated (19) after 30, 60, and 90 min, respectively. Samples of 15 g of total RNA were subjected to electrophoresis on a 0.8% formaldehyde agarose gel (13) and transferred to a Nylon membrane (MSI, Westboro, MA). The 510-bp 1 BglII-EcoRV fragment of the mae2 gene and 1.2kilobase ClaI fragment of the S. cerevisiae ACT1 gene served as probes.
Cloning of Unidirectional Deletion Fragments and lacZ Fusions-Exonuclease III deletion fragments of the mae2 promoter (6), plasmids pMV1 to pMV18 (Table I), were selected to construct mae2 promoter-lacZ fusions. These plasmids were digested with PstI and ApaI (Fig. 2) to replace the mae2 open reading frame (ORF) and 3Ј-flanking region with a 3.7-kilobase PstI-ApaI fragment containing the lacZ ORF from YEp356, resulting in mae2-promoter lacZ fusions (pMZ1 to pMZ18).
␤-Galactosidase Assays-Yeast cells transformed with the respective pMZ plasmids were inoculated into 10 ml of SC Ϫleu with 0.2% glucose, 2% glucose, 8% glucose, or 2% raffinose as the carbon source. L-Malate was added at 0.2% when required, and anaerobic conditions were obtained with mineral oil and and incubation without shaking. Cultures were grown to A 600 of 0.8 to 0.9, and ␤-galactosidase assays were performed with permeabilized cells (13). Assays were done in duplicate with samples representing 8 l/ml of original cell culture on at least three transformants from different transformation sets.
In pMV54, the TCATTCATTT sequence at nt Ϫ244 was mutated in a similar way with primers MV24 and MV29, and MV28 and MV27 (Fig. 1), and mutation of AAAATTGCGAG at nt Ϫ202 in pMV55 was obtained with primers MV24 and MV30. Plasmid pMV61 carries mutations in both the TCATTCATTT and AAAATTGCGAG elements and was constructed using primers MV24 and MV30 with pMV54 as template. The polymerase chain reaction products were digested with SphI and SalI and cloned into the native SphI, SalI sites of pMV46 (Fig. 2). The mutated TGGGCTAAT sequence at nt Ϫ186 (pMZ86) was introduced with primers MV40 and MV2, with direct cloning into the SalI, PstI sites of pMZ1. Plasmid pMZ89 contains mutations in the TCAT-TCATTT, AAAATTGCGAG, and TGGGCTAAT elements and was constructed by subcloning the SalI, PstI fragment from pMZ86 (contains mutated TGGGCTAAT element) into the SalI, PstI sites of pMZ61 (mutated TCATTCATTT and AAAATTGCGAG elements).
In pMV88, the TCCCCTGGCA sequence at nt Ϫ297 was mutated with primers MV24 and MV42, and MV41 and MV2, and the putative TATA element at nt Ϫ88 in pMV59 with primers MV24 and MV31, and MV32 and MV2. Both were subcloned in the SphI, PstI sites of pMV46. Primers MV24 and MV44 were used for mutation of AGGGGGA at nt Ϫ251 in pMV292 (subcloned in the SphI, DraI sites of pMV46); primers MV43 and MV2 for mutation of TGACGT at nt Ϫ175 in pMZ291 (subcloned in the SalI, PstI sites of pMZ1); and primers MV45 and MV2 for mutation of TGACG at nt Ϫ102 in pMV293 (subcloned in the AvaII, PstI sites of pMV46). For all the pMV constructs, the mae2 ORF was replaced with the lacZ ORF as described above. All deletions and mutations were verified with restriction enzyme digestions and sequence analysis (13).

Malic Enzyme Activity and Levels of mae2 Transcription-
The mae2 gene carried on the episomal plasmid pMV652 complemented the S. pombe mae2 Ϫ mutant in its ability to reduce NAD ϩ to NADH via the malic enzyme reaction (Table III). There was no difference in the growth rate of S. pombe wild type, mae2 Ϫ or transformed strains (data not shown) that would indicate that the malic enzyme is essential for growth. When cells of S. cerevisiae were transformed with pMV652, no increase in malic enzyme activity was observed relative to the untransformed strain. ␤-Galactosidase assays on S. cerevisiae cells transformed with pMZ1 showed only 4% activity relative to S. pombe (data not shown), suggesting that the mae2 promoter is not recognized by S. cerevisiae.
No evidence for an increase in either the malic enzyme activity or mae2 transcription was observed for S. pombe cells grown in the presence of 0.2% L-malate (Fig. 3). However, ␤-galactosidase assays done with pMZ1 (Table IV) showed transcription levels of 122 and 187% for cells grown in 2% glucose and 8% glucose, respectively, relative to the 100% for cells grown in 0.2% glucose and 105% in 2% raffinose. Furthermore, the levels of transcription were increased more than 5-fold when grown under anaerobic conditions in 2% glucose. This was confirmed with Northern analysis (Fig. 4) on the wild type strain; a small increase was observed when cells were shifted to 2% glucose, with a stronger response under anaerobic conditions.
Sequence Analysis of the mae2 Promoter-The DNA sequence of the region upstream of the mae2 ORF is available in the GenBank and EMBL data bases (accession number SPCC794). Computerized analysis (16) of the mae2 promoter indicated a putative TATA element (TTATTTAAAA) at nt Ϫ88 in the mae2 promoter (Fig. 1). A predicted CAP signal (ACAG-TAAT) corresponding to the putative TATA element was identified at nt Ϫ51, and primer extension reactions (data not shown) confirmed that it serves as the transcription initiation site. Comparative analysis of the 5Ј-flanking regions of the S. pombe mae2 and malate transporter (mae1) genes, 2 revealed three conserved elements: GTTGATT at nt Ϫ298 in mae2 (GTTTGATT at nt Ϫ292 in mae1), a direct repeat TCAT-TCATTT at nt Ϫ244 (TCATTCATTT at nt Ϫ227 in mae1), and AAATTGCGAG at nt Ϫ202 (AAATTGCTAG at nt Ϫ152 in mae1).
Putative binding sites for several eukaryotic activators were identified in the mae2 promoter (Fig. 1). The conserved CCAAT sequence that serves as binding site for the CCAAT-binding protein (20) is present in the reverse orientation at nt Ϫ397 and Ϫ294. Recognition sites for the TGGCA-binding protein (21), a member of the ubiquitous eukaryotic family of NF1-like nuclear proteins that are involved in transcriptional regulation (22), are present at nt Ϫ395 and nt Ϫ354.
The putative cis-acting elements that were identified based on homology with the recognition sites of cAMP-dependent transcription factors include the conserved GG(A/G)G core sequence for the yeast transcription factor ADR1 (24) at nt Ϫ251 and inverted at nt Ϫ359. The ADR1 binding site was first identified in the upstream-activating sequence UAS1 of the S. cerevisiae ADH2 gene, where it functions as a cis-acting element involved in glucose repression (24). The ADR1 protein (adr1p) is inactivated through phosphorylation by a cAMP-dependent kinase (25), therefore rendering the activity of adr1p sensitive to the levels of glucose. More recently, adr1p was also reported to be involved in derepression of the S. cerevisiae acetyl-CoA synthetase gene in response to sugar limitation (26).
The putative ADR1 binding site at nt Ϫ359 overlaps with the CCCMNSSS recognition sequence reported for the mammalian transcription factor AP-2 that is involved in cAMP-induced expression of the human metallothionein IIA protein (27). Fur-  1. DNA sequence of the 544 bp upstream of the mae2 gene. Putative regulatory elements indicated by shaded blocks include the recognition sites for the TGGCA-binding protein at nt Ϫ394 and Ϫ354, CCAAT-binding protein at nt Ϫ397 and Ϫ294, ADR1 at nt Ϫ358 and Ϫ250, AP-2 at nt Ϫ358, Sp1 at nt Ϫ186 ATF/ CREB at nt Ϫ175 and Ϫ102, and TATA element at nt Ϫ88. The annealing positions of primers used for mutagenesis of putative elements are indicated with arrows, and the altered sequences are shown above or underneath the nucleotide sequence. Restriction enzyme sites are indicated in bold letters as are transcription initiation (tsp) and translation initiation (ATG) sites.

FIG. 2. Introduction of the mutated CCAAT-like element in pMZ53.
Plasmid pMV46 contains the mae2 gene cloned in the SacI, HindIII sites of pRS315 (SalI-XhoI sites destroyed with restriction digest). Primers MV24 and MV26 were used to yield fragment A (nt Ϫ483 to Ϫ282) and primers MV25 and MV27 for fragment B (nt Ϫ305 to Ϫ180). Both fragments were digested with BstEII and ligated, and the product was amplified with primers MV24 and MV27. The product was digested with SphI and SalI and cloned in the native SphI, SalI sites of pMV46. The mae2 ORF was replaced by the lacZ ORF from YEp356 and flanking regions as a PstI-ApaI fragment. thermore, the TGACGT sequence at nt Ϫ175 and the TGACGA sequence at nt Ϫ102 resemble the binding site for the mammalian activating transcription factor ATF/CREB (cAMP response element binding) (28). ATF/CREB is activated through phosphorylation by the cAMP-dependent kinase A (29) and was reported to bind as a dimer to the GTGACGTACAG consensus sequence (30).
Deletion and Mutation Analysis of Putative Regulatory Elements in mae2 Promoter-Eighteen unidirectional deletion fragments of the mae2 promoter were fused in-frame with the lacZ ORF (pMZ1 to pMZ18 in Table I) to evaluate their ability to support transcription. Transcription from plasmids pMZ1 to pMZ12 were unaffected as determined by ␤-galactosidase assays (only the results for pMZ12 are shown in Fig. 3). Removal of the sequences upstream of nt Ϫ218 decreased the activity by more than 50% (pMZ15), whereas deletion of the adjacent 36 nucleotides in pMZ16 restored the wild type levels of activity. Deletion of the sequences upstream of nt Ϫ73 in pMZ17 removed the putative TATA element at nt Ϫ88 and reduced transcription to only 4%. Smaller deletions within the SphI-AvaII region (nt Ϫ472 to Ϫ122) showed decreased transcription, especially for pMZ48 and pMZ71. In contrast, deletions within the XhoI-AvaII region increased transcription 4-fold for pMZ47, 7-to 8-fold for pMZ77, and 8-to 9-fold for pMZ79. This suggested the presence of positive regulatory sequences between nt Ϫ362 and Ϫ284 and negative regulatory sequences between nt Ϫ218 and Ϫ182.
Site-directed mutations of nine putative cis-acting elements   4. Northern analysis of mae2 expression. Cells were grown overnight in 1% glucose and shifted to fresh 2% glucose, either aerobic or anaerobic, or 2% glycerol, ethanol medium. Total RNA was probed with the 510-bp BglII-EcoRV fragment of the mae2 gene or the 1.2kilobase ClaI fragment of the S. cerevisiae ACT1 gene.
were introduced with polymerase chain reaction primers (Fig.  1). Three of these elements were identified as conserved elements in the mae1 and mae2 promoters, whereas the other elements were identified based on homology with recognition binding sites for eukaryotic regulatory proteins. Mutation of the putative TATA element at nt Ϫ88 (pMZ59) decreased transcription to only 10%, confirming an essential function in transcription of the mae2 gene. Mutation of the CCAAT-like element at nt Ϫ298 (pMZ53), ADR1 at nt Ϫ251 (pMZ292), or ATF/CREB at nt Ϫ102 (pMZ293) did not affect transcription significantly (Fig. 3). However, mutation of TCCCCTGGCA at nt Ϫ359 (homology with the ADR1, AP-2, and TGGCA binding sites) in pMZ88 and TGACGT at nt Ϫ175 (homology with ATF/CREB binding site) in pMZ291 resulted in only 53 and 20% of wild type levels, respectively. These elements therefore seem to serve as upstream activating sequences that regulate the expression of the mae2 gene and are designated UAS1 and UAS2, respectively, in further discussions (Fig. 5).
Mutation of TCATTCATT at nt Ϫ244 (pMZ54), AAATTGC-GAG at nt Ϫ202 (pMZ55), and TGGGCTAAT at nt Ϫ186 (pMZ86) increased transcription to 154, 170, and 224%, respectively (Fig. 3). These elements seem to function as upstream repressor sequences and were designated URS1, URS2, and URS3, respectively (Fig. 5). Furthermore, pMZ61 (mutated URS1 and URS2) and pMZ89 (mutated URS1, URS2, and URS3) showed increased levels of transcription of approximately 3-and 4-fold, respectively, suggesting that URS1, URS2, and URS3 may confer a cooperative repression of mae2 expression. DISCUSSION Transcription of S. pombe class II promoters is typical of higher eukaryotes where TFIID is the general transcription factor that binds to the TATA box and initiates the assembly of the transcription complex (31). However, the promoter structure of S. pombe genes in general differ from that of S. cerevisiae and in some cases more closely resembles that of mammalian promoters (32). Our results showed that mutation of the putative TATA element 37 bp upstream of the transcription initiation site reduced transcription to only 10% (pMZ59, Fig.  3). The distance between the putative TATA box and the initiation start site is typical for S. pombe (32) and within the 35 to 180 nt reported for S. cerevisiae (33). However, the S. pombe mae2 promoter was not functional when expressed in S. cerevisiae as measured by malic enzyme activity and ␤-galactosidase assays. This lack of promoter function could be because of the inability of the S. cerevisiae transcription complex to recognize the mae2 promoter structure or the absence of essential transcription factors.
The S. pombe malic enzyme has a high substrate affinity (K m of 3.2 mM) (34) and decarboxylates L-malate to pyruvate and CO 2 with the concurrent reduction of NAD ϩ to NADH. Under anaerobic conditions, the pyruvate is further metabolized to ethanol and CO 2 with the re-oxidation of NADH to NAD ϩ . The degradation of L-malate in S. pombe was reported not to be subject to substrate induction or glucose repression (35). Our results confirmed that expression of mae2 was not affected by 0.2% L-malate (Fig. 3), but induced levels of mae2 transcription were obtained when grown in 8% glucose or under anaerobic conditions (Table IV). A 2-to 3-fold increase in transcription under anaerobic conditions was reported for the S. cerevisiae malic enzyme gene (12), and a role in the provision of intramitochondrial NADPH or pyruvate under anaerobic conditions was proposed for the NADP-dependent enzyme. Unlike the S. cerevisiae malic enzyme, the DNA sequence of the mae2 gene (6) did not indicate the presence of a mitochondrial signal sequence, suggesting that the enzyme is located in the cytosol.
Little information is available on the regulation of enzymes involved in carbon metabolism in S. pombe cells. Expression of the S. pombe fbp1 gene, encoding fructose 1,6-bisphosphatase, was reported to be repressed in 8% glucose (36) because of regulation by a glucose-induced cAMP signal (37). The increased expression of mae2 observed in high concentrations of glucose (and therefore high cAMP levels) suggested that cAMP induction may be involved in this response. We have found two positive cis-acting elements, UAS1 and UAS2, that show homology with binding sites for cAMP-dependent regulatory proteins. UAS1 contains overlapping recognition sites for the transcriptional activator proteins ADR1, AP-2, and TGGCAbinding protein (Fig. 5), whereas UAS2 has homology with the transcriptional activator ATF/CREB. ADR1 is inactivated by a cAMP-dependent kinase, whereas the activity of AP-2 and ATF/CREB are induced in the presence of cAMP (24 -30). Mutation of UAS1 (pMZ88) and UAS2 (pMZ291) decreased the levels of transcription to 53 and 20%, respectively (Fig. 3). It would be of interest to determine whether these mutations also affect the induced expression of mae2 under fermentative conditions.
Deletion and mutation analysis of the mae2 promoter indicated the presence of three negative-acting elements, URS1, URS2, and URS3. Mutation of URS3 (pMZ86) had a stronger effect than that of URS1 (pMZ54) or URS2 (pMZ55). However, mutation of either URS1 plus URS2 (pMZ61) or all three elements (pMZ89) resulted in higher levels of transcription than for any one of the individual elements, suggesting that they function cooperatively to repress transcription of the mae2 gene (Fig. 5). Homologous copies of the URS1 and URS2 elements are also present in the promoter of the S. pombe malate transporter gene, mae1, indicating possible co-regulation of enzymes involved in malate degradation in S. pombe. Further analysis of these elements and the regulatory proteins that bind them is essential to elucidate the transcriptional regulation of the mae2 gene and perhaps also that of the mae1 gene.
Because cells of S. pombe do not utilize L-malate as the only source of carbon or energy, the physiological importance of the active transport and strong intracellular degradation is intriguing. The oxidative decarboxylation of malate coincides with the reduction of NAD ϩ to NADH, suggesting that the malic enzyme may play an important role in maintaining the redox balance under aerobic conditions. Under anaerobic conditions, however, the pyruvate produced during this reaction is further metabolized to ethanol with the concomitant oxidation of NADH to NAD ϩ . Because there is no net gain or loss in NADH during the conversion of malate to ethanol, the increased expression of the malic enzyme under fermentative conditions should not affect the redox balance.
The S. pombe cells may use the malic enzyme to provide pyruvate for essential anaplerotic reactions under fermentative conditions. Pyruvate plays an essential role in the provision of ␣-ketoglutarate and oxaloacetate for the biosynthesis of amino acids and nucleotides. Both these precursors are synthe- FIG. 5. Proposed model for the regulated expression of the S. pombe malic enzyme gene. UAS1 and UAS2 serve as activator sites and have homology with the binding sites for eukaryotic transcriptionactivating factors as indicated. The novel URS1, URS2, and URS3 elements act cooperatively to repress transcription of the mae2 gene. The putative TATA element indicated is required for wild type levels of expression. sized in the mitochondria and transported to the cytosol for biosynthetic reactions; therefore alternative pathways have to be utilized for the synthesis of these precursors when the mitochondria are not functional (38). These anaplerotic reactions include the carboxylation of pyruvate to oxaloacetate via pyruvate carboxylase, the oxidation of malate to pyruvate via the malic enzyme, and the production of succinate via the glyoxylate cycle. The induced expression of the S. pombe malic enzyme under fermentative conditions may therefore serve as an important auxiliary pathway for the production of pyruvate for other metabolic requirements.
To our knowledge, this is the first report on the glucoseinduced expression of a malic enzyme gene in yeast. Results presented here indicate the differential expression of the S. pombe mae2 gene under fermentative conditions, i.e. 8% glucose and anaerobic conditions. We therefore propose that the NAD-dependent malic enzyme from S. pombe may provide cytosolic pyruvate for anaplerotic pathways under fermentative conditions. A function in the provision of NADH for reductive biosynthesis and maintenance of the redox balance is, however, not excluded. Further research is required to elucidate the physiological role of the malic enzyme in S. pombe and the importance of the various UASs and URSs in the regulated expression of the mae2 gene.