A Member of the Sugar Transporter Family, Stl1p Is the Glycerol/H⁺ Symporter in Saccharomyces cerevisiae

Strains

S. cerevisiae strains used in this study comprise two different genetic backgrounds: BY4742 (Brachmann et al., 1998) and W303-1A (Thomas and Rothstein, 1989). The collection of single deletion mutants in BY4742 was obtained from EUROSCARF. The strains with W303-1A genetic background are listed in Table 1. All stl1 deletion mutants were made by transformation with a PCR-generated fragment containing the KanMX cassette (Wach et al., 1994) and a 200-base pair flanking region amplified from the corresponding BY4742 mutant obtained from EUROSCARF. Deletions were verified by PCR. The construct used to complement the stl1 mutation was made by amplification of the gene and flanking regions from genomic DNA using primers 5′ CCG GCT CGA GTT GCA GAT TAC GAA AGA ATC 3′ and 5′ GGC CTC TAG AGA ATA TCA YGG CAA GAC CGC 3′. The resulting PCR product was cloned between the XhoI and XbaI of pRS316. The construct was verified by DNA sequencing. Schizosaccharomyces pombe EG573 (mat1-M mat2/3-del::leuura4-D18 ade6) was used for heterologous expression of STL1. Escherichia coli strains XL1-Blue and DH5α were used for plasmid selection and propagation.

Yeast Media and Growth Conditions

Batch cultures of yeast were grown aerobically in complex medium (YP: 1% [wt/vol] yeast extract; 2% [wt/vol] peptone) supplemented with 2% (wt/vol) glucose (YPD), ethanol (YPE), or glycerol (YPG) as carbon and energy sources. Incubation was performed at 30°C, 180 rpm, orbital shaking with air/liquid ratio 2.5/1. The inability to utilize glycerol as a carbon and energy source was recognized on glycerol-based defined media supplemented with 0.05% peptone (Rønnow and Kielland-Brandt, 1993). Synthetic complete (SC) and SC drop-out media were made according to Burke and collaborators (2000). Drop tests were performed from cellular suspensions containing ∼1 × 10⁶ cells/ml. Tenfold serial dilutions were made, and 5 μl of each suspension was applied. S. pombe EG573 was cultivated on SC with 2% (wt/vol) glucose in identical physiological conditions as S. cerevisiae.

Transformation and DNA Manipulation

S. cerevisiae cells were transformed with the lithium acetate method (Ausubel et al., 1996). Transformants were selected on selective synthetic medium (SC without uracil) or YPD containing 300 μg/ml G418. E. coli strains were transformed with a CaCl₂/heat shock-based protocol (Ausubel et al., 1996). Transformants were selected on LB agar supplemented with 100 μg/ml ampicillin. S. pombe EG573 was transformed using the method of Bähler et al. (1998). Transformants were selected on SC without uracil, supplemented with 2% (wt/vol) agar and 15 μM thiamine.

DNA isolation and manipulation was performed by standard procedures (Sambrook et al., 1989). DNA fragments obtained by PCR were purified using the QIAquick PCR product Purification Kit (Qiagen, Hilden, Germany). The concentration and purity of DNA was checked spectrophotometrically (Sambrook et al., 1989).

Glycerol Transport Studies

The methods of Lages and Lucas (1995, 1997) and Lages et al. (1999) were used to measure initial rates of radiolabeled glycerol uptake and the ratios of intracellular-to-extracellular glycerol concentrations (in/out accumulation ratios). To measure initial uptake rates of [¹⁴C]glycerol (Amersham, Freiburg, Germany), cells were washed twice and resuspended in ice-cold water to a final concentration of ∼30 mg/ml (dry weight). The yeast suspension (10 μl) was mixed with 100 mM Tris/citrate buffer, pH 5.0 (10 μl), in a conical centrifuge tube and incubated for 2 min at 30°C in a water bath. The reaction was started by addition of 5 μl of a 20 mM aqueous solution of [¹⁴C]glycerol with a specific activity of 2000 dpm/nmol, corresponding to a final concentration of 4 mM. Uptake at this concentration is close to V_max (Lages and Lucas, 1997; Sutherland et al., 1997). After 10 s, the reaction was stopped by dilution with 5 ml of ice-cold water. Cells were immediately collected on Whatman GF/C filters (Whatman, Maidstone, England), washed twice with 5 ml water, and immersed in vials containing 5 ml of scintillation fluid (Optiphase Hisafe2, Wallac Manufactured Products, Turku, Finland). The radio-labeled glycerol taken up by cells was measured in a Packard Tri-Carb 2200 CA liquid scintillation counter (Canberra Packard International, Zurich, Switzerland). The assays were made in triplicate and referenced to a blank made by inverting the sequence of addition of glycerol and water.

To determine the transport-driven ratios of [¹⁴C]glycerol inside versus outside the cells (in/out accumulation ratios), the cell suspension (80 μl) and 110 μl 100 mM Tris/citrate (pH 5.0) were transferred to conical glass centrifuge tubes and incubated at 30°C with magnetic stirring. After 2 min of incubation the experiment was started by addition of 10 μl (200 mM) of [¹⁴C]glycerol (specific activity 300 dpm/nmol). At appropriate time intervals, 10-μl aliquots were taken and filtered through Whatman GF/C filters. The filters were then washed twice with 5 ml ice-cold water and transferred to vials containing 5 ml scintillation fluid.

The protonophores/ionophores carbonyl cyanide m-chlorophenylhydrazone (CCCP) and carbonyl cyanide p-(trifluoromethoxy) phenylhydrazone (FCCP) were used to prevent accumulation of glycerol through glycerol/proton symport activity, as well as to induce efflux of intracellular radiolabeled glycerol, in assay concentration of 50 μM. The S. pombe and S. cerevisiae intracellular volumes used to calculate the intracellular glycerol concentrations were determined previously (Lages et al., 1999).

Subcellular Localization Using Green Fluorescent Protein as Reporter

For expression of the STL1-GFP fusion protein we made use of pUG35 (U. Güldener and J. Hegemann, unpublished results), a centromere-based vector containing the MET25 promoter, followed by a polylinker and a sequence encoding yeast-enhanced green fluorescent protein (GFP). STL1 was amplified by PCR using primers 5′ GGC CTC TAG AAT GAA GGA TTT AAA ATT ATCG3′ and 5′GGC CGA ATT CAC CCT CAA AAT TTG CTT TAT C 3′. The resulting PCR product was cloned between the XbaI and EcoRI sites of pUG35, in frame with GFP, yielding pUG35-STL1. The construct was verified by DNA sequencing. Strains were transformed with pUG35-STL1. Cells were grown overnight in SC medium without uracil, at 30°C, with glycerol as carbon source. Cells were then diluted in the same medium to an OD₆₀₀ ≈0.25 and allowed to reach OD₆₀₀ = 0.6. Subsequently the cells were shifted to medium without uracil and methionine to induce the MET25 promoter. After 1–2 h of induction, cells were visualized by fluorescence microscopy. The remaining culture was then shifted to medium containing 2% glucose and methionine to switch off the promoter. Cells were visualized after 1 h of growth in glucose. A Leica TCS SP2 confocal scanning microscope (Heidelberg, Germany) was used to obtain the images. GFP was excitated at 488 nm and emitted light between 500 and 570 nm.

Western Blotting

A PCR-generated fragment encoding a duplicated IgG binding domain (Z) of protein A, and containing a KanMX cassette, was amplified from plasmid pFZ according to Whyte and Munro (2001) using primers 5′ CAA ACA TCA AAA ATG AAG ATA CAG TGA ACG ATA AAG CAA ATT TTG AGG GTG GAG CAG GGG CGG GTG 3′ and 5′ AAT GCT TTC TTA AGT AAA TTA CAA AAT ATG ATT TGT GAG TTG TGT GTG AAG GTC GAC GGT ATC GAT AAG 3′. This was used to transform yeast to construct the epitope-tagged strains. Correct integration was verified by PCR. For Western blots, cells (0.2 mg dry weight) were collected by centrifugation, and protein extracts were prepared according to Burke et al. (2000), separated by SDS-PAGE (10%), and transferred to nitrocellulose filters (Invitrogen, Carlsbad, CA) according to Kyhse-Andersen (1994). The filters were incubated with a commercial mixture of horseradish peroxidase and antiperoxidase rabbit antibody (PAP, cat. no. Z0113, Dako-Cytomation A/S, Copenhagen, Denmark), and reacting polypeptides were visualized using ECL Plus Western Blotting Detection system (Amersham Biosciences) and a Storm 860 Scanner (Molecular Dynamics, Sunnyvale, CA). GFP-containing proteins were detected using rabbit antibodies raised against rx-YFP (Østergaard et al., 2001) by Pineda Antikörper (Berlin, Germany; a kind gift of Drs. Jakob Winther and Henrik Østergaard).

Expression of STL1 in S. pombe

A chromosomal copy of STL1 was obtained by PCR using the primers 5′GGC CTC GAG ATG AAG GAT TTA AAA TTA TCG 3′ and 5′GGC CCC GGG TCA ACC CTC AAA ATT TGC 3′, including restriction sequences. It was inserted into E. coli/S. pombe shuttle multicopy expression vectors pREP82X and pREP4X between the XhoI and XmaI sites. These plasmids contain a thiamine-repressible promoter, nmt1, and ura4⁺, and Amp^r as selective markers. They differ in promoter strength, weak and strong, respectively, because of increased expression-modulating mutations (Maundrell, 1993). The integrity of the constructs was verified by sequencing. S. pombe EG573 was transformed with pREP4X-STL1 or with pREP82X-STL1. Transformants were selected and cultivated in liquid SC medium without uracil and supplemented with 2 nM thiamine for the lower expression plasmid pREP82X and 15 μM thiamine for the higher promoter strength plasmid pREP4X. Cells were harvested by centrifugation at OD₆₀₀ = 1.2 (exponential growth phase), resuspended in the same medium without added thiamine (Sieiro et al., 2003), and grown for 6 h for cultures from medium with 2 nM thiamine or for 30 h for cultures from medium with 15 μM thiamine. Cells were recollected by centrifugation and resuspended in water to a cell density of 30 mg/ml (dry weight) and kept on ice before glycerol transport assays.

Identification of Mutants Impaired in Glycerol Utilization

We began our screen with the assumptions that a mutant lacking the glycerol transporter would grow less efficiently on glycerol than the parent and that the transporter would be a multispanning membrane protein. On the basis of these assumptions, we screened a subset of the EUROSCARF single-deletion mutant collection, comprising 321 open reading frames with six or more predicted membrane spans, for glycerol utilization. Ten deletion mutants with a growth defect on a defined medium containing glycerol as the sole carbon source were found. Five did not grow (PER1, PCP1, DRS2, PMT5, and RHK1), whereas the remaining five grew poorly (PMR1, GEF1, ERG4, GUP1, and STL1). The strains deleted for PER1 (protein processing in the ER), PCP1 (processing of cytochrome c peroxidase), PMR1 (Ca²⁺/Mn²⁺ P-type ATPase, Golgi), and GEF1 (chloride channel localized to Golgi involved in intracellular iron metabolism) are listed at the SGD database (http://www.yeastgenome.org/) as having growth defects on respiratory substrates. Deletions of DRS2 (integral membrane Ca²⁺-ATPase) and PMT5 (protein O-mannosyl transferase of the ER) are listed as having a growth defect on lactate as a carbon source. Thus deletion of six of the genes identified in the present screen are known to be generally compromised for growth on nonfermentable carbon sources and thus probably not directly involved in glycerol uptake. In addition, RHK1 (involved in synthesis of the dolichol-linked oligosaccharide donor for N-linked glycosylation of proteins) and ERG4 (C24-sterol reductase) are both predicted to have 9 transmembrane domains and to be located in the ER. Interestingly, ergosterol is required for targeting of the tryptophan permease to the membrane (Umebayashi and Nakano, 2003); this may also apply to the glycerol transporter. STL1 (Zhao et al., 1994) and GUP1 (Holst et al., 2000) are the only genes from this group encoding proteins with 12 putative transmembrane domains, and STL1 is the only one with an unknown function. Stl1p belongs to the sugar transporter family in S. cerevisiae (Nelissen et al., 1997).

Most studies on active glycerol uptake have been performed in the W303-1A genetic background. Therefore STL1 was disrupted with the KanMX cassette (Wach et al., 1994) in this strain. Deletion of STL1 in both BY4742 and W303-1A genetic backgrounds resulted in cells that grew poorly on glycerol (Figure 1). A copy of the STL1 gene comprising 2 kbp of the promoter region and 200 base pairs of the terminator region was introduced into the stl1-deleted strain using a centromere-based plasmid. The resulting strain grew like the wild-type strain, demonstrating functional complementation of the stl1 growth defect (Figure 1).

Stl1p Is Essential for Active Glycerol Uptake

The glycerol utilization phenotype prompted us to analyze glycerol transport in the stl1 mutant. Wild-type cells and stl1 cells were grown in glucose-based complex medium containing 1 M NaCl and 15 mM glycerol. Cells were harvested at the diauxic shift, and glycerol transport was measured. These conditions were chosen because they are known to allow detection of glycerol accumulation even in a strain lacking GUP1, GUP2, and GUT1 (Neves et al., 2004a). Deletion of STL1 eliminated active transport-driven glycerol accumulation (Figure 2A). The same occurred in the mutant strains stl1 fps1 and stl1 gut1 (data not in the figure). Interestingly, the gpd1 gpd2 stl1 strain, unlike the gpd1 gpd2 strain, did not grow on salt even in the presence of small amounts of glycerol. Additionally, cells deleted for all three genes previously known to be directly or indirectly involved in glycerol transport in S. cerevisiae, GUP1, GUP2, and FPS1 (Holst et al., 2000; Tamás et al., 1999), accumulated glycerol against a chemical gradient (Figure 2B). This accumulation was sensitive to the depletion of proton motive force by CCCP. Moreover, glycerol symport activity has previously been observed on nonfermentable carbon sources like ethanol (Lages and Lucas, 1997). To avoid interference of metabolism with measurements of transport kinetics (Holst et al., 2000; Oliveira et al., 2003), experiments were also performed in a gut1 background. The deletion of STL1 in this background eliminated the active transport-driven glycerol accumulation (Figure 2C). This was also observed in the mutant strains stl1 gpd1 gpd2 and stl1 fps1 grown on ethanol (data not in the figure). Thus, under all growth conditions where S. cerevisiae is known to take up glycerol by active transport, deletion of STL1 abolished this capacity.

Stl1p and Active Glycerol Transport Are Induced in Response to Osmotic Shock

Transcription of STL1 in glucose-grown cells is induced upon a shift to high salt conditions. The expression is transient and dependent on Hog1p and Hot1p (Rep et al., 2000). Glycerol/H⁺ symport activity is absent in wild-type cells during exponential growth in glucose-based complex medium irrespective of the presence of NaCl (Holst et al., 2000). It becomes measurable during the diauxic shift when glucose is exhausted (Lages and Lucas 1997; Neves et al., 2004a).

To analyze the expression of STL1 in more detail, we integrated DNA encoding a duplicated IgG-binding domain (Z) of the Staphylococcus aureus protein A to the chromosomal copy of STL1. The strain expressing the fusion protein (Stl1p-ZZ) grew on glycerol medium like the wild-type, indicating that the fusion protein is functional. Subsequently, we analyzed the Stl1p level and measured uptake of glycerol under osmotic shock conditions. Cells were grown to exponential phase on glucose and shifted to medium containing 0.7 M NaCl. Samples were taken at time intervals and the V_max of active transport using labeled glycerol at a single concentration was estimated. The same cells were used to determine the maximum in/out concentration ratio, reflecting active transport-driven accumulation (Figure 3A). Osmotic shock led to a sharp increase in glycerol accumulation capability. After 1.5 h of incubation at 30°C, significant amounts of Stl1p were produced in cells experiencing osmotic shock (Figure 3B). Stl1p was not produced in cells that were not osmotically shocked. Importantly, the levels of Stl1p and the glycerol uptake activity directly correlated. The high levels of Stl1p appearing after 1.5 h shift to NaCl gradually disappeared, and the protein was no longer detectable after 4 h (Figure 3B, GPD1 GPD2). These results are in agreement with the transient appearance of the STL1 transcript after salt shock, reported by Rep et al. (2000). They found that the transcript appeared 0.5 h after the shock, but was undetectable after 1.5 h. Synthesis and accumulation of glycerol are major responses of yeast to sudden increases in the osmolarity (Hohmann, 2002). The glycerol synthesis is accomplished by induction of the genes GPD1 and GPP2, encoding a glycerol 3-phosphate dehydrogenase and a glycerol 3-phosphate–specific phosphatase, respectively (Ansell et al., 1997). The closure of the Fps1p channel ensures that glycerol does not exit the cell during osmotic shock (Tamás et al., 1999). Interestingly, in cells that are deleted for the two isogenes GPD1 and GPD2, Stl1p appeared during the osmotic shock, but the down-regulation/degradation was prevented (Figure 3B, gpd1 gpd2), presumably because the osmotic imbalance persisted in the absence of glycerol synthesis. The gpd1 gpd2 strain exhibits a significant lag phase and somewhat slower growth after salt shock compared with the parent, so we also analyzed a strain that is deleted only for GPD1 and that is not affected in growth rate after salt shock. Under these conditions, the presence of Stl1p was also stabilized (Figure 3B, gpd1 GPD2).

If cells were grown continuously in glucose-based medium with a high concentration of salt, Stl1p was detectable only when the cells entered the diauxic shift (Figure 4A, arrow, and B), consistent with active glycerol transport (Holst et al., 2000; Neves et al., 2004a). However, gpd1 and gpd1 gpd2 cells, which produce only residual amounts of glycerol (Oliveira and Lucas, 2004), made significant amounts of Stl1p during exponential growth in 0.7 M NaCl (Figure 4C). This is consistent with the fact that gpd1 gpd2 cells display glycerol uptake during exponential growth in glucose-based medium with NaCl (Holst et al., 2000). This is not surprising, considering the growth defect of the gpd1 gpd2 stl1 strain on media containing 1 M NaCl and 15 mM glycerol. Thus, the absence of glycerol synthesis apparently prevented the down-regulation and degradation of Stl1p in this mutant.

Stl1p Is Localized to the Plasma Membrane and Is Subject to Glucose Inactivation

The subcellular localization of Stl1p was analyzed using the GFP fused to the carboxy-terminus of Stl1p, with the gene under the control of the methionine-repressible MET25 promoter. The fusion protein was functional in uptake of [¹⁴C]glycerol. The majority of the fluorescence was detected at the plasma membrane in cells growing exponentially on glycerol-based medium (Figure 5A). This is consistent with a function of Stl1p as a plasma membrane-localized glycerol transporter. The symporter activity (V_max) decreases progressively after transfer from ethanol- to glucose-containing medium. The activity is reduced to 10% within 30 min after transfer (Lages and Lucas, 1997). In full accordance, when cells were shifted from glycerol to glucose, and the promoter driving the expression of the fusion protein was turned off by the addition of methionine, the fluorescence signal disappeared from the plasma membrane within 30 min. Most fluorescence was detected in the vacuolar lumen 1 h after the shift to glucose (Figure 5A). Western blot analysis using anti-GFP antibodies showed a 90-kDa band, presumably reflecting the full-length fusion protein. In addition, a 38-kDa degradation product was detected. On shift to glucose-containing medium the full-length protein disappeared, and the degradation product accumulated and appeared to be stable for at least 2.5 h (Figure 5B). These observations indicate that the protein is internalized into the vacuole, and that the GFP moiety is detached from Stl1p, releasing it from the membrane. Inactivation of Stl1p-ZZ by growth in glucose was also followed by Western blot analysis. When glucose was added to cells growing exponentially in ethanol-based complex medium, Stl1p progressively disappeared (Figure 5C, WT) and was no longer detected 1.5 h after the addition of glucose. This is consistent with the behavior of the GFP fusion protein. If, however, the fate of Stl1p was followed in an end3 mutant, which is defective in endocytosis (Raths et al., 1993), Stl1p was not degraded within 4 h, and a larger molecular species appeared (Figure 5C, end3). This larger form of Stl1p may represent one or more ubiquitinated species of Stl1p. We conclude that Stl1p is inactivated by growth in glucose, internalized to the vacuole by endocytosis in an END3-dependent manner, where it is subsequently degraded.

The S. cerevisiae STL1 Gene Provides Active Uptake of Glycerol When Expressed in S. pombe

Fission yeast, S. pombe, does not take up glycerol by active transport (Lages et al., 1999). This yeast does not have any genes with high sequence homology to STL1. Therefore, we chose to introduce the S. cerevisiae STL1 gene into S. pombe to test if this gene indeed codes for the glycerol/H⁺ symporter. S. pombe transformants expressing STL1 from two promoters of different strength were obtained. The time needed for optimal expression of STL1 was initially determined by glycerol uptake experiments using a single concentration of radiolabeled glycerol in the transport V_max range. For each construct, a short time interval was found in which glycerol accumulation could be measured after the promoter was turned on. This was 6 and 30 h for the weaker and the stronger promoter, respectively.

We determined glycerol accumulation ratios in a single transformant for each construct (Figure 6A) and compared it with untransformed cells as well as with control transformants harboring plasmids without STL1 insert (Figure 6B). Accumulation exceeded equilibrium and was higher for the transformants expressing the gene with the stronger promoter. The accumulation was sensitive to the dissipation of proton motive force through the action of the protonophores/ionophores CCCP and FCCP (Figure 6A). Furthermore, these drugs elicited glycerol efflux (Figure 6C). We conclude that the expression of the S. cerevisiae STL1 gene in S. pombe introduces uptake of glycerol in this yeast that is dependent on the proton motive force.