INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Hexose transporters comprise a family of proteins involved in cellular sugar uptake. They have been well described for a variety of organisms, including bacteria, yeasts, plants, and humans. Regarding sugar metabolism, the fission yeast Schizosaccharomyces pombe shares a number of characteristic properties with the budding yeast Saccharomyces cerevisiae. Both species grow as facultative aerobes and use aerobic alcoholic fermentation in the presence of an excess of sugar (17, 13). Among the utilized carbon sources, distinct differences are present. D-Glucose, D-fructose, glycerol, and maltose are metabolized by both yeast species, with D-glucose being the preferred substrate. S. pombe cells can grow on the monosaccharide D-gluconate (23), whereas S. cerevisiae cells can utilize D-galactose and disaccharides such as sucrose (13, 18). In contrast to S. cerevisiae, S. pombe can use ethanol but only in the presence of glucose (53, 54). The narrow spectrum of carbon sources accepted by S. pombe is attributed to corresponding differences in carbon metabolism. The carbon metabolism of S. pombe does not involve the glyoxylate cycle, and furthermore, some enzymes of ethanol metabolism and gluconeogenesis are not constitutively expressed (53, 13).

Considering transport into the cells as the first step of the utilization of sugar, both yeast species express specific transporters on the basis of related functions. In S. cerevisiae, D-glucose and D-fructose uptake is mediated by the hexose transport (HXT) proteins (30, 5, 47, 31, 46), also known as monosaccharide facilitators. The HXT proteins belong to the superfamily of 12 transmembrane transporters (36, 49). Among the 20 hexose transporter-related proteins in S. cerevisiae, only 6 mediate metabolically relevant uptake, while 2 are thought to function as glucose-sensors and 1 mediates D-galactose uptake. The majority of these proteins have been identified on the basis of sequence similarity (5, 7, 46). In the fission yeast S. pombe, glucose uptake was described to be energy dependent, driven by the plasma membrane ATPase-generated electrochemical gradient (µ_H⁺) (24). Kinetic analysis revealed specific D-gluconate-H⁺ symport activity (23, 10). The first isolated glucose transporter, Ght1, of S. pombe was identified by complementation of the glucose transport-deficient S. pombe strain YGS-5 (37). Functional analysis was performed by heterologous expression of Ght1 in a glucose transport-deficient S. cerevisiae mutant strain lacking HXT1 through HXT7 (34).

Indications for the existence of related sequences in the S. pombe genome coding for putative additional transporters were derived from Southern blot analysis of S. pombe DNA hybridized with a conserved region of Ght1 (34). In the present study we describe the identification and characterization of a family of S. pombe genes, Ght1 through Ght6, which encode monosaccharide transporters. A comparison of the predicted Ght protein topologies and the amino acid alignments with those of the hexose transporters of S. cerevisiae, Kluyveromyces lactis, Candida albicans, and humans suggests that they correspond to a distinct family of transporters. Functional analysis of the Ght transporters identified a substrate specificity for D-glucose for Ght1p, Ght2p, and Ght5p, with Ght5p being the most prominently expressed and active D-glucose transporter and Ght6p transporting D-fructose as a preferred sugar. Following isolation of this hexose transporter family, the conditional phenotype of the S. pombe mutant YGS-5 was characterized as defective in Ght1, Ght5, and Ght6. Two unusual transporters, Ght3 and Ght4, though highly similar to other Ght's, serve a different transport function in S. pombe. The construction of three S. pombe mutants bearing single or double mutations in the Ght3 and Ght4 genes provided evidence that these proteins are involved in D-gluconate transport in S. pombe wild-type cells.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Cloning and sequencing of Ght1, Ght2, Ght3, Ght4, Ght5, and Ght6. The cloning and sequencing of Ght1 have already been described (34, 40). Ght2 was isolated from an S. pombe cDNA library after hybridization with a conserved region of Ght1. The library was obtained from F. Lacroute (Centre National de la Recherche Scientifique, Gif sur Yvette, France) and constructed in the expression vector pFL61 as described by Minet et al. (38), which contains the promoter of the phosphoglycerate kinase gene (29, 8). A total of 25 S. pombe genome equivalents of the cDNA library were transformed into Escherichia coli DH5 cells and streaked on Luria-Bertani agar plates. All E. coli clones were replicated to nylon filters and hybridized with the 0.5-kb KpnI fragment (34) of the Ght1 coding region. The DNA probe was radiolabeled by random priming (Megaprime kit; Amersham, Braunschweig, Germany) using ³²P-labeled dCTP. Hybridization was performed with ³²P-labeled Ght1 DNA at 42°C in hybridization buffer (5× SSC [1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate], 5× Denhardt's solution [0.1% bovine serum albumin, 0.1% Ficoll, 0.1% polyvinylpyrrolidone], 30% formamide, 1% sodium dodecyl sulfate [SDS]) for 24 h. Following hybridization, the filters were washed for 5 min at room temperature in 3× SSC, 5 min at room temperature in 2× SSC containing 0.1% SDS, 10 min at 42°C in 2× SSC containing 0.1% SDS, and 15 min at 42°C in 0.2% SSC containing 0.1% SDS. An X-ray film was exposed to the filters for 36 h. The rescreening was performed with isolated plasmid DNAs of positive clones by using the same hybridization and washing conditions as in the first screen. The NotI cDNA inserts of identified clones were ligated with the NotI-digested vector pBSKII (Stratagene, Heidelberg, Germany). A total of four clones were isolated, and three of them contained overlapping cDNA inserts with different lengths of nontranslated regions. The longest open reading frame contained a single gene termed Ght2, which is capable of encoding a protein of 519 amino acid residues. The fourth clone, termed partial Ght5, represented 1,055 bp of a 5'-end-truncated gene.

Yeast strains, media, plasmids, and general genetic and molecular methods. All yeast cells (Table 1) were grown at 30°C. S. pombe wild-type cells were grown on YEL medium (0.5% Difco yeast extract, 0.2% Difco Casamino Acids, 0.01% adenine sulfate) with 2% D-glucose, and the S. cerevisiae mutant strain RE700A (47) was grown on YEP medium (2% Difco Bacto Peptone, 1% Difco yeast extract) containing 2% maltose. Nutritional requirements appropriate for selection and maintenance of mutants and plasmids in the transformed strains were scored on minimal YNB medium consisting of 0.67% Difco yeast nitrogen base and a 2% carbon source (D-glucose, D-fructose, or maltose), supplemented with the appropriate amino acids (Fluka, Buchs, Switzerland) without uracil. Standard recombinant DNA manipulations were performed according to the procedure reported in reference 50. The S. pombe Ght genes were heterologously expressed in the S. cerevisiae D-glucose uptake-deficient mutant RE700A (47). The Ght1, Ght3, Ght4, Ght5, and Ght6 genes were therefore placed under the control of the inducible copper promoter CUP1 of the expression vector pYEX-BX (AMRAD, Biotech). Gene expression was induced by the addition of 0.5 mM CuSO₄ to the growth medium. Ght2 expression was driven by the phosphoglycerate kinase promoter of plasmid pFL61 (38). Transformation of S. cerevisiae RE700A with the plasmids pYEXGht1, pYEXGht3, pYEXGht4, pYEXGht5, pYEXGht6, and pFLGht2 was followed by selection for uracil protrophy. As a control, the RE700A mutant strain was transformed with the vector plasmids pFL61 and pYEX-BX, yielding strains SHYPFL and SHYPYEX (Table 1), respectively.

ade6-M210 ura4-18 leu1-31

Sugar transport and accumulation assays. S. pombe wild-type cells were grown on YEL containing 2% D-glucose, and the S. cerevisiae mutant strain RE700A was grown on YEP with 2% maltose. Standard YNB medium supplemented with 2% D-glucose or 2% D-fructose was used for the growth of the transformants SHYGht1, SHYGht2, SHYGht5, and SHYGht6. The corresponding vector control strains, containing either pYEX-BX or pFL61 without inserts, were grown in YNB medium with 2% maltose. Consumption of D-glucose and D-fructose from the media was assayed enzymatically as described earlier (3). The initial uptake rates of D-glucose and D-fructose were determined at 5-s intervals according to the zero-trans influx assay described by Özcan et al. (41), as modified by Walsh et al. (55). Kinetic parameters were determined with Eadie-Hofstee plots. To determine the putative effects of the endogenous GAL2p transporter in the S. cerevisiae RE700A background (33), consumption and zero-trans influx assays were performed as competition assays involving 1:3 mixtures of D-glucose and D-galactose. A threefold excess of D-galactose did not influence the D-glucose uptake in any of the hybrid strains, thus confirming that the observed D-glucose transport was entirely mediated by the heterologously expressed S. pombe transporters and not by the endogenously expressed S. cerevisiae GAL2p transporter (data not shown). For all assays, mid-log-phase cells were harvested, washed with media, and incubated at 30°C for 2 h in fresh media containing 0.1% D-glucose or 0.1% D-fructose according to the sugar to be tested in the subsequent experiments. For D-gluconate consumption assays, cells were grown in 2% glycerol-0.2% sodium-acetate, harvested, washed in distilled water, resuspended as a 5% cell suspension in 0.15 M KH₂PO₄ buffer (pH 4.5), and incubated at 30°C. The assay was started by addition of D-gluconate to a final concentration of 2 mM. One-milliliter samples were taken after 15 s and 10, 20, 40, and 80 min and centrifuged to prevent any further uptake. One-hundred-microliter aliquots of the cell-free supernatant were subjected to the enzymatic D-gluconate determination as described previously (39).

RNA experiments. S. pombe wild-type and YGS-5 cells were grown for 18 h at 30°C as 200-ml cultures in YNB medium supplemented with 2% D-gluconate. From these cultures, 20-ml aliquots were transferred to 30 ml of ice-cold water to terminate metabolic activity and used as controls. To initiate the experiment, 2% D-glucose or 2% D-fructose was added to the remaining 180-ml cultures, and 20-ml samples were harvested by centrifugation at 3,000 rpm (Biofuge 22R, Heraeus) after 10, 30, 60, and 150 min of incubation at 30°C. Isolation of total yeast RNA was performed by the acidic phenol method (14). RNA was separated on a 1% agarose gel containing 1% formaldehyde-50 mM boric acid-1 mM sodium citrate-5 mM NaOH (pH 7.5) and transferred to Hybond-N membranes (Amersham) in 10× 1.5 M NaCl-0.15 M sodium citrate (pH 7.0). Hybridization was performed at 68°C in hybridization solution containing 0.5 M NaH₂PO₄, 7% SDS, and 1 mM EDTA using DNA probes spanning the entire coding region of any one of the Ght1, Ght2, Ght3, and Ght4 genes or 1,235 bp of the coding region of Ght5. As a control for the integrity and abundance of the isolated RNA, we used a 389-bp fragment of the Pma1 gene of S. pombe, which encodes the plasma membrane ATPase. The Pma1 probe was derived from PCRs involving S. pombe genomic DNA and specific primers (nucleotides 2557 to 2585, 5' CCTTACCAAGAACAAGTTGTCTCTTGGTG 3', and nucleotides 2917 to 2946, 5' GAACGATAACCACGAGAAGCCAAATCACCG 3'). The sizes of the messages were determined relative to the mobilities of a 9.5- to 0.24-kb ladder from Gibco BRL. All DNA probes were labeled using a High-Prime labeling kit (Boehringer, Mannheim, Germany). Hybridized Northern blots were exposed by using the phosphorimager technology (BAS-1000; Fuji). Reverse transcription-PCR (RT-PCR) experiments were performed by subjecting 15 µg of isolated total RNA to a SuperScript cDNA synthesis kit (Gibco BRL) according to the manufacturer's recommendations, followed by polymerase-directed amplification of specific targets by using 2-µl aliquots of the first-strand reaction mix as the template and the gene-specific primers spanning the entire coding regions of the Ght1 to Ght4 genes, 1,235 bp of the coding region of Ght5, and 1,415 bp of Ght6. The abundance of the amplified Ght genes was determined relative to that in control reaction mixtures containing the Pma1 gene and involving the same primers used for the probe preparation. All RT-PCR experiments comprised 25 PCR cycles.

Nucleotide sequence accession numbers. The sequences of Ght2, Ght3, Ght4, Ght5, and Ght6 have been assigned the GenBank accession numbers AF017180, AF051139, AF051140, AF051141, and AF098076, respectively.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation and sequence analysis of the S. pombe hexose transporters. Previous observations from Southern blot analysis of digested S. pombe DNA hybridized with a conserved Ght1 fragment suggested the presence of additional monosaccharide transporter-related sequences (34). Thus, the experimental rationale for cloning additional putative hexose transporter genes of the S. pombe genome involved low-stringency hybridization screening of an S. pombe cDNA library (38) with the same conserved region of the Ght1 gene. Two different open reading frames, designated Ght2 and Ght5, were identified to have considerable homology to Ght1 and moderate homology to the HXT genes of S. cerevisiae. The Ght2 gene, comprising 1,560 bp, encodes a protein of 519 amino acid residues. The nucleotide sequence displayed 70% identity to the Ght1 gene and 51% identity to the HXT2 gene of S. cerevisiae. Amino acid sequence alignment revealed 71% identity to S. pombe Ght1p and 35% identity to S. cerevisiae HXT proteins. The Ght5 sequence, representing a 5'-end-truncated gene (minus 300 bp), was aligned to the corresponding parts of the Ght1 sequence with identities of 61 and 70% on the nucleotide and the amino acid levels, respectively. This high sequence similarity suggests that Ght2 and Ght5 may represent additional monosaccharide transporter genes in S. pombe.

View larger version (32K): [in this window] [in a new window]   FIG. 1.   Dendrogram of sequence similarities among the human and the yeast hexose transport proteins. The dendrogram was derived from an alignment of some representative amino acid sequences of the hexose transporters of S. pombe (S.p.), S. cerevisiae (S.c.), K. lactis (K.l.), Candida albicans (C.a.), and Homo sapiens (H.s.) glucose transporters by the CLUSTAL program (PCGene; Intelligenetics), which uses the method developed by Higgins and Sharp (21, 22). All aligned transporter proteins of the hexose family belong to the 12-transmembrane sugar porter subfamily of the major facilitator superfamily of proteins. The dendrogram classifies the relationships of the transport proteins based on their sequence similarities. The lengths of the horizontal branches are inversely proportional to the similarity of the sequences at each branch tip. The S. pombe hexose transporters are clustered as a distinct group and are less related to the S. cerevisiae and K. lactis transporters, which comprise another group. The human glucose transporters GLUT1 and GLUT4 are set separately from the yeast transporters.

Functional analysis of the S. pombe Ght transporters in S. cerevisiae. Functional analysis of the S. pombe Ght transporters was performed by heterologous expression of each gene in the S. cerevisiae mutant RE700A, which is deficient in HXT1 through HXT7 (47). Expression of Ght1, Ght2, Ght5, and Ght6 could functionally complement both the glucose uptake- and the growth-deficient phenotype of RE700A on media containing D-glucose or D-fructose as the sole carbon source (Fig. 2). Northern blot analysis confirmed the correct sizes and comparable amounts of the Ght1, Ght2, Ght5, and Ght6 transcripts in the transformed strains (data not shown). Ght3 and Ght4 expression failed to complement growth of the mutant RE700A on media containing D-glucose or D-fructose and displayed the phenotypes of the vector control strains (data not shown).

View larger version (45K): [in this window] [in a new window]   FIG. 2.   Suppression of hxt1-7 conferred the growth defect of the S. cerevisiae strain RE700A after transformation with S. pombe Ght1, Ght2, Ght5, and Ght6. The strains were streaked on a medium with 2% D-glucose, 2% D-fructose, or 2% maltose as the sole carbon source and grown for 3 days at 30°C. The S. pombe strains expressing Ght1, Ght2, Ght5, and Ght6 regained the ability to grow on D-glucose and D-fructose medium, suggesting that each of them is a D-glucose transporter. Expression of Ght3 and Ght4 failed to restore the D-glucose growth defect of the hxt1-7 mutant. The resulting phenotypes were the same as that of the mutant RE700A and the corresponding vector control strains, which grew only on maltose as the carbon source. wt., wild type.

Expression analysis of S. pombe Ght1 through Ght6. Expression of Ght1 through Ght6 was analyzed with S. pombe total RNA extracted from cells grown on different carbon sources by semiquantitative RT-PCR and Northern blot analysis. RT-PCR experiments were performed by subjecting 15 µg of isolated total RNA to cDNA synthesis, followed by Taq polymerase-directed amplification of specific Ght1 through Ght6 targets involving gene-specific primers. The abundance of the amplified Ght genes was compared to that in control reactions of the housekeeping Pma1 gene. The expression pattern is presented in Fig. 3. Ght5 and Ght6 were constitutively expressed in all carbon sources used, with Ght5 being the most prominently expressed transporter, followed by Ght6, which had a less intense signal on 0.2% D-glucose and maltose. Ght2 was also constitutively expressed except in cells grown on glycerol. A high concentration of D-glucose repressed Ght1, Ght3, and Ght4 expression. Ght1 was detected only under derepressed conditions on D-gluconate and on maltose. Ght3 and Ght4 were strongly repressed in cells grown on D-glucose. A low concentration of D-glucose and maltose and glycerol derepressed Ght3 and Ght4 expression. In cells grown on D-gluconate, Ght3 and Ght4 were the predominantly expressed transporter genes.

View larger version (44K): [in this window] [in a new window]   FIG. 3.   Differential expression of Ght1 to Ght6 in S. pombe wild-type cells grown on different carbon sources. Total RNA isolated from cells grown on 2% or 0.2% D-glucose, 2% maltose, 2% D-gluconate, and 2% glycerol was subjected to RT and subsequent PCR involving Ght gene-specific primers. A 389-bp fragment of the S. pombe Pma1 gene served as a control for the abundance and integrity of RNA isolation and the amplified gene fragments. Lanes M, molecular weight markers.

View larger version (79K): [in this window] [in a new window]   FIG. 4.   Time-dependent repression of prominent hexose transporters. Expression of Ght1 to Ght5 was monitored in D-gluconate-grown cultures of the S. pombe wild type and the mutant YGS-5. Aliquots for RNA preparations were taken from the D-gluconate-grown cultures 10, 30, 60, and 150 min after addition of 2% D-glucose or 2% D-fructose. To control for the integrity and abundance of the isolated RNA, a 389-bp fragment of the Pma1 gene of S. pombe, which encodes the plasma membrane ATPase, was used in a control blot. Ght1 to Ght5 were expressed at the highest levels when D-gluconate was used as the sole carbon source. Ght2 mRNA was not decreased by D-glucose or D-fructose, whereas downregulation of the Ght1, Ght3, and Ght4 transcripts indicated sensitivity to D-glucose and D-fructose repression. All Ght gene transcripts were again detectable 2 h after addition of D-glucose or D-fructose, indicating a transient concentration-dependent repression. Ght5 was more highly expressed than all of the other Ght genes, suggesting that this transporter is the most abundant hexose transporter.

S. pombe hexose transporter-deficient mutants. The previously described S. pombe D-glucose transport-deficient mutant YGS-5 (37) was examined for expression of the Ght1 through Ght6 genes. The mutant had been obtained by treatment of wild-type cells (972h) with N-methyl-N'-nitro-N-nitrosoguanidine and selected by growth on D-gluconate medium containing 0.05% 2-deoxy-D-glucose (37). YGS-5 total RNA isolated from D-gluconate-grown cells was subjected to RT-PCR using Ght gene-specific primers. In this S. pombe mutant, Ght2, Ght3, and Ght4 (but not Ght1, Ght5, or Ght6) were expressed (Fig. 5). Northern blot analysis of total RNA preparations obtained from D-gluconate-grown YGS-5 cells following the addition of D-glucose to the culture confirmed Ght2, Ght3, and Ght4 expression in YGS-5 (Fig. 4C). Contrary to what occurred in S. pombe wild-type cells, Ght3 and Ght4 mRNAs were not repressed by D-glucose in YGS-5. Thus, the D-glucose uptake-deficient phenotype of YGS-5 can be attributed to the loss of Ght1, Ght5, and Ght6 functions. The Ght3 and Ght4 genes did not complement the D-glucose uptake-deficient phenotype of the S. cerevisiae mutant RE700A. These genes were found to be expressed in the mutant YGS-5, which does not grow on D-glucose but does grow on D-gluconate. This result implied that at least one of the expressed genes, Ght3 or Ght4, may encode a D-gluconate transporter. Because S. cerevisiae cells do not utilize D-gluconate, a knockout approach with the S. pombe wild-type strain was used for the analysis of the functions of Ght3 and Ght4. The two single mutants (SHYght3 and SHYght4 [Table 1]) as well as the S. pombe ght3 ght4 double mutant (SHYght3 ght4 [Table 1]) were obtained by a disruption of the Ght3 coding region by the S. cerevisiae LEU2 gene and/or a disruption of the Ght4 coding region by the S. pombe Ura4 gene. Successful integration of the respective disruption cassettes was confirmed by PCR analysis with genomic DNA as the template (data not shown). Growth on D-gluconate-containing media of the ght3, ght4, and ght3 ght4 double mutants was compared to that of the S. pombe wild-type cells (data not shown). Both the S. pombe ght3 mutant and the ght3 ght4 double mutant did not grow on D-gluconate, suggesting that the Ght3p transporter function is required for D-gluconate uptake.

View larger version (29K): [in this window] [in a new window]   FIG. 5.   Expression of Ght1 to Ght6 in the S. pombe D-glucose uptake-deficient mutant YGS-5. Total RNA was isolated from cells grown on 2% D-gluconate and subjected to RT and subsequent PCR involving Ght gene-specific primers. A 389-bp fragment of the S. pombe Pma1 gene served as a control for the abundance and integrity of RNA isolation and the amplified gene fragments. In the control experiment all Ght genes were amplified from chromosomal S. pombe DNA. Ght3 and Ght4 were the most highly expressed transporters. Ght2 was expressed weakly, and Ght1, Ght5, and Ght6 were not expressed at all. Lane M, molecular size markers.

View larger version (15K): [in this window] [in a new window]   FIG. 6.   Consumption of D-gluconate in the S. pombe wild-type strain () compared to that of the knockout mutants SHYGht3 (), SHYGht4 (), and SHYGht3Ght4 (). All strains were grown overnight in 2% glycerol-0.2% sodium-acetate medium, harvested, and washed in distilled water. The consumption assay was started by an addition of 2 mM D-gluconate to a 10% cell suspension. Samples for enzymatic determination of D-gluconate concentration were taken after 15 s and 10, 20, 40, and 80 min.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We have isolated six genes designated Ght1, Ght2, Ght3, Ght4, Ght5, and Ght6 in S. pombe which encode monosaccharide transporters. Ght1 was the first gene identified as being involved in glucose transport (34), and Ght2 through Ght6 were identified on the basis of sequence similarity. The representation of all Ght genes in the S. pombe cDNA library confirmed the correct transcription of Ght1 through Ght6 and ruled out the possibility of nontranscribed pseudogenes for Ght3, Ght4, and Ght6. The amount of each amplified Ght gene in the cDNA library, which was obtained from the D-glucose-grown S. pombe wild-type cells, corresponded to the mRNA expression data of S. pombe cells grown on 2% D-glucose (Fig. 3). Because both Ght2 and Ght5 were not repressed in D-glucose-grown S. pombe wild-type cells (Fig. 3 and 4), the Ght1 low-stringency screening proved appropriate for their identification. Ght3 and Ght4 were not isolated by this method because both were repressed in D-glucose (Fig. 3 and 4) and were only 60% homologous to Ght1 (Table 2). Sequence analysis of both the genomic and the cDNA amplification products revealed identical intronless coding regions for all Ght genes.

While we were cloning the S. pombe Ght genes, a communication was published in which a different nomenclature for the Ght genes was used (7). It should be noted that our ascending numbering of the Ght genes involves Ght2, which is not included in the dendrogram of the mentioned publication. The dendrogram of sequence similarities among hexose transporters of humans and different yeast species documents the high similarity of the S. pombe transporter proteins (Fig. 1) and classifies three clusters. Within the yeast monosaccharide transporters, the S. pombe proteins are clustered in a group that is distinct from the other groups comprising the S. cerevisiae and K. lactis transporters and the human glucose transporters, which are set separately from all the yeast genes. Because of the sequence similarity, which correlates well with substrate specificities among transport proteins (48) (Fig. 1), it is reasonable to consider the S. pombe Ght proteins to be members of the sugar porter family within the major facilitator superfamily (49, 44). The most prominent feature of this family, the sugar transport protein recognition motif (1), is highly conserved in all identified S. pombe Ght proteins. Within the S. pombe Ght transporter family, the Ght3p and Ght4p transporters, which are 88% identical to each other, build a subcluster versus Ght1p, Ght2p, Ght5p, and Ght6p. Ght3 and Ght4 are also unusual in that they encode two transporters with significant amino acid exchanges in transmembrane 10, which was suggested to contribute to substrate specificity (Fig. 7). In Ght3p and Ght4p of S. pombe, the most highly conserved phenylalanine residue in the characterized D-glucose transporters is replaced by a tyrosine residue in a way similar to the corresponding residue of GAL2p of S. cerevisiae. The other reported amino acid residue responsible for discrimination between D-galactose and D-glucose in GAL2p of S. cerevisiae is tyrosine (27), which is replaced by tryptophan in GAL2p of S. cerevisiae and also in Ght3p and Ght4p of S. pombe. Because S. pombe does not take up or utilize D-galactose, the alteration of these functional amino acids may indicate a putative alteration of Ght3p and Ght4p substrate specificity to D-gluconate compared to that of Ght1p, Ght2p, Ght5p, and Ght6p.

View larger version (65K): [in this window] [in a new window]   FIG. 7.   Alignment of amino acids constituting transmembrane domain 10 of the common 12-transmembrane stretches in sugar transport proteins. The comparison involves S. pombe (S.p.), S. cerevisiae (S.c.), K. lactis (K.l.), C. albicans (C.a.), and H. sapiens (H.s.) proteins. Numbers on the right refer to the amino acid residues. The phenylalanine residue is the most highly conserved throughout all organisms within this transmembrane segment. Y446 and W455 in S. cerevisiae Gal2p were proposed to be responsible for the discrimination of D-galactose and D-glucose (27, 28), and the same tyrosine and tryptophan residues were found in S. pombe Ght3p and Ght4p (indicated in bold).

Heterologous expression of S. pombe Ght1, Ght2, Ght5, or Ght6 in the mutant RE700A of S. cerevisiae complemented its D-glucose uptake-deficient phenotype (Fig. 2), thus proving a D-glucose transport function for each of them. For Ght1p and Ght2p the kinetic transport analysis revealed moderate affinities to D-glucose, indicating that D-glucose is the preferred transport substrate, though this specificity does not exclude D-fructose from uptake. Discrimination between these two sugars takes place intracellularly: in S. pombe by D-fructose-specific hexokinase Hxk1 (K_m of fructose = 1.5 mM, K_m of glucose = 8.5 mM) (54, 45) and in S. cerevisiae by D-glucose-specific glucokinase GLK1 (18, 6). It was already demonstrated for S. cerevisiae that D-fructose and D-glucose cause different repression effects, probably triggered by different signalling pathways (15). Thus, the specificities of intracellular early metabolic enzymes such as hexokinases, and not of D-glucose- or D-fructose-specific transporters, enable the cells to adapt to the available carbon source. Moreover, the kinases may affect the regulation of transporter expression. By comparing the kinetic constants of D-glucose transport of S. pombe wild-type cells (0.4 mM) with those obtained for the Ght5-expressing S. cerevisiae strain (0.6 mM), Ght5p is characterized as the high-affinity D-glucose transporter. D-Fructose uptake in S. pombe wild-type cells, characterized by a K_T of 5.5 mM, is mediated by Ght6p, which transports D-fructose with the highest affinity (K_T of fructose = 5 mM) in comparison with those of the other Ght transporters. Thus, in S. pombe the discrimination between D-glucose and D-fructose is in addition achieved by different transporters for each sugar. However, comparison of the kinetic parameters should be done with caution. The estimated K_T values of D-glucose transport in S. pombe vary over a wide range (K_T values, 5.3, 3, and 15 mM in references 34, 23, and 37, respectively) and may be influenced by many factors. Obviously, depending on the environmental growth conditions, individually expressed Ght genes contribute to different extents to actual D-glucose transport in S. pombe. The observed high V_max values for Ght1p, Ght2p, Ght5p, and Ght6p compared with that of S. pombe wild-type cells may be caused by overexpression of the individual transporters in the S. cerevisiae background.

Because D-glucose transport properties of the S. pombe wild-type cells should be reflected correspondingly in the expression pattern of the involved genes, RNA expression of all Ght genes was monitored in cells grown on different carbon sources (Fig. 3). In cells grown on high D-glucose concentrations, Ght2, Ght5, and Ght6 were expressed, which is in agreement with their functional characterization by heterologous expression. Ght1, Ght3, and Ght4 were repressed under these conditions, thus resembling the regulation of GAL2 and MAL61 in S. cerevisiae (32). Ght5 expression was increased with low D-glucose concentrations, as was expected for this high-affinity transporter. Low D-glucose concentrations derepressed also Ght3 and Ght4. These two genes were most strongly expressed in cells grown on D-gluconate. The expression pattern of Ght1 through Ght6 in cells grown on maltose was similar to that in cells grown on low D-glucose, suggesting that in S. pombe maltose does not induce any of these genes. Maltose was shown to be taken up as D-glucose molecules following extracellular splitting (24, 51) (thus resembling results with low glucose concentrations). The time-dependent D-glucose repression of Ght transporters was monitored in the D-gluconate-grown cultures of the S. pombe wild-type strain and its mutant YGS-5 following the addition of D-glucose or D-fructose. In the S. pombe wild-type strain a transient repression of Ght3 and Ght4 was observed with high D-glucose and D-fructose concentrations (Fig. 4) when transcripts were again detectable after 60 min. This is consistent with observations by Caspari (10) indicating that at the end of the exponential growth phase or at the beginning of the stationary phase the D-gluconate transport system is expressed regardless of D-gluconate availability. In S. pombe, D-glucose repression is mediated by the cAMP protein kinase A pathway (26) and probably by an S. cerevisiae Snf1-Mig1-homologous pathway (52). Derepression seems to be mediated by the Wis1-Sty1-MAP (mitogen-activated kinase) pathway. The S. pombe gti1⁺ gene was identified as a downstream inducing element for the induction of D-gluconate transport (10).

The previously described S. pombe D-glucose uptake-deficient mutant YGS-5 was also analyzed for Ght transporter expression (Fig. 5). Ght1, Ght5, and Ght6 are not expressed in this mutant. Because it is unlikely that the mutagenic agent conferred mutations in the open reading frames of all three genes, one can speculate that there is a regulatory defect in the S. pombe YGS-5 mutant. This conclusion is also supported by the fact that although Ght2 was constitutively expressed in YGS-5, D-glucose uptake was not detectable in this strain. The putative regulatory role may reside in Ght1, which was previously identified as a suppressor of the YGS-5 phenotype. Because the amino acid alignment of all S. pombe proteins revealed a high (44%) degree of conservation within the predicted transmembrane domains, the putative functional role of Ght1 may be attributed to its C-terminal extension. Within this region, major differences concerning length and amino acid composition were recognized relative to the lengths and amino acid compositions of other Ght proteins. This implies that Ght1p possesses a putative regulatory or signalling feature similar to the D-glucose-sensing roles of SNF3p or RGT2p in S. cerevisiae (33, 42, 43). However, none of the reported consensus sequences mediating signalling in SNF3p or RGT2p were detected in the C-terminal extension of S. pombe Ght1p. Figure 8 gives an amino acid alignment of the Ght proteins' C termini starting with the conserved C-terminal tryptophan. The Ght1p C terminus is the largest of the six aligned protein segments. A BLASTP search revealed that a 22-amino-acid stretch near the end of the Ght1p C terminus (Fig. 8) is significantly homologous to the precursor of the mouse insulin receptor (45% identity, 58% similarity).

View larger version (14K): [in this window] [in a new window]   FIG. 8.   S. pombe Ght protein C termini were aligned by using the CLUSTAL package (21, 22). The alignment included the highly conserved tryptophan. The identical residues are indicated in bold characters. Because of the very different lengths of the cytoplasmic extensions, the putative regulatory elements may reside within these extensions. For the Ght1p C terminus, BLASTP searches identified a 22-amino-acid residue stretch (bold) displaying 45% identity and 58% similarity to the mouse insulin receptor precursor. This motif does not occur in the other proteins. Ght2p, Ght3p, and Ght4p aligned to different subunits of the cyclic-nucleotide-gated cation channel from Bos taurus (Q28279, 28% identity in a 49-amino-acid stretch, and Q28181, 61% identity in an 18-amino-acid stretch [underlined]). The cluster of negatively charged glutamate residues is probably a frequent feature of structural proteins. The Ght5p C-terminal region aligned to transcriptional regulatory proteins containing zinc finger clusters with 24% identity and with 42% similarity in a 61-amino-acid stretch (underlined) to proteins of S. pombe, S. cerevisiae, and Mus musculus. The serine-rich regions in all Ght proteins may be targets for serine/threonine kinases.

YGS-5 cells have been selected and grown on D-gluconate as a sole carbon source (37), suggesting that one of the remaining Ght3 and Ght4 genes is the putative D-gluconate transporter. Three lines of evidence suggest that Ght3 encodes the specific S. pombe D-gluconate transporter. First, the knockout approach of disrupting Ght3 or Ght4 or both in the S. pombe wild-type strain proved that Ght3 is required for D-gluconate transport because neither of the ght3 and ght3 ght4 mutants grew on D-gluconate as the sole carbon source. Second, the expression of Ght3 in S. cerevisiae RE700A did not complement its D-glucose uptake deficiency. Third, Ght3 was constitutively expressed in YGS-5 cells and most strongly expressed in D-gluconate-grown S. pombe wild-type cells; moreover, it was sensitive to D-glucose repression.

In contrast, Ght4 is obviously not a D-gluconate transporter, in spite of its striking homology to Ght3. This conclusion was drawn from the fact that the ght3 strain was unable to grow on D-gluconate even though Ght4 was intact in this strain. The physiological role of the Ght4p transporter has yet to be determined. Disruptions of either Ght3 or Ght4 or both in the S. pombe mutant strain YGS-5 were not viable on D-glucose, thus confirming that Ght2p is not active in YGS-5, probably due to the lack of the Ght1 function.

In summary, we have demonstrated that S. pombe harbors a multimember family of functional hexose transporters. Of these, Ght5 is the prominently expressed one and represents the high-affinity D-glucose transporter of the S. pombe wild-type strain. Ght1 is a putative signalling membrane protein, but its physiological regulatory role remains to be determined. Ght2p was characterized as a D-glucose transporter, with moderate affinity and transport capacities. Ght3 encodes the specific D-gluconate transporter. The transport specificity of Ght4p is the subject of current investigations. Ght6p exhibits a slightly higher affinity to D-fructose than to D-glucose and is the suggested predominant transporter for the D-fructose uptake.