INTRODUCTION

Sphingolipids are important components of eukaryotic cell membranes. Animals develop different diseases due to genetically or environmentally altered sphingolipid metabolism (1, 2). Sphingolipids have structural functions in maintaining cell membrane integrity, and they act as anchors to proteins (3). In addition, metabolites of sphingolipids such as ceramide, sphingosine, and sphingosine 1-phosphate (S-1-P)¹ have been demonstrated to be involved as bioeffector molecules and second messengers in key events including cell growth, differentiation, cell senescence, apoptosis, and stress responses (1-5). S-1-P has a proliferative effect on certain quiescent cells and has been shown to trigger intracellular calcium mobilization (5). Platelet-derived growth factor induces an elevation in cellular levels of sphingosine and S-1-P and activates sphingosine kinase in quiescent fibroblasts. These studies imply that S-1-P participates in the mitogenic action of this and other growth factors (6).

In the yeast Saccharomyces cerevisiae, sphingolipids have been demonstrated to be essential for cell viability, based on studies of long chain sphingoid base auxotrophs that are defective in serine palmitoyltransferase, the first enzyme in the de novo pathway of sphingolipid synthesis (7). Importantly, at least some sphingolipid-mediated cell events are conserved between mammalian cells and S. cerevisiae (8). For example, ceramide induces S. cerevisiae growth suppression and cell cycle arrest, and the mammalian counterpart of ceramide-activated protein phosphatase has been identified in S. cerevisiae (9).

In this report, we have identified a novel enzyme in S. cerevisiae as a dihydrosphingosine-1-phosphate (DHS-1-P) phosphatase by characterizing sphingosine resistance genes (YSRs). Our studies show that dihydrosphingosine is phosphorylated to DHS-1-P once it enters cells. Dephosphorylation of DHS-1-P regulates synthesis of sphingolipids and glycerolipids in response to exogenous sphingolipids.

EXPERIMENTAL PROCEDURES

Yeast strains listed in Table I were maintained on YPD medium, synthetic minimal medium (SC), or SC medium without uracil (SC-ura) as described previously (10). To determine minimum inhibitory concentrations of sphingosine to the JS16 strain, 20 µl of 3 × 10⁷ exponential phase cells/ml was spotted on SC agar plates with serial dilutions of sphingosine. The lowest concentration with which cell growth was completely inhibited was designated as the minimum inhibitory concentration. To test sphingosine resistance, cells were streaked on suitable media containing 35 µM sphingosine and 0.005% Nonidet P-40. Bacterial Epicurian coli XL1-Blue competent cells were purchased from Stratagene.

Table I. Yeast strains used in these studies Strain Genotype Phenotype Source SGP(3) MATa leu2-3,112 trp1 ura3 his3 ade8 ras1::HIS3 Wild type, BST1+ Ref. 21 JS16 MATa leu2-3,112 trp1 ura3 his3 ade8 ras1::HIS3 bst::NEO BST1 deletion, sphingolipid hypersensitive Ref. 21 JK9-3d MATa/ trp1 leu2-3 his4 ura3 ade2rme1 YSR2+, YSR2-1+ Ref. 26 JK9-3d MAT trp1 leu2-3 his4 ura3 ade2rme1 Wild type, YSR2+, YSR2-1+ Ref. 26 YSR2 MAT trp1 leu2-3 his4 ura3 ade2 ysr2::NEO YSR2 deletion This study YSR2-1 MAT trp1 leu2-3 his4 ura3 ade2 ysr2-1::NEO YSR2-1 deletion This study

Sphingosine, dihydrosphingosine, ceramide, phosphatidic acid, phosphatidylcholine (PC), phosphatidylethanolamine (PE), and phosphatidylinositol (PI) were purchased from Sigma, Biomol, Avanti, and Calbiochem. NBD-PA and NBD-C₆-ceramide were purchased from Molecular Probes. D-erythro-[4,5-³H]Dihydrosphingosine and D-erythro-[4,5-³H]dihydrosphingosine 1-phosphate were purchased from American Radiolabeled Chemicals Inc. NBD-C₆-ceramide 1-phosphate was synthesized enzymatically from NBD-C₆-ceramide by bacterial diacylglycerol kinase in the presence of excess ATP as described (10). Synthesized ceramide 1-phosphate was separated on preparative Silica Gel 60 TLC plates and purified, and its concentration was measured by determination of phosphate (11).

Vector pYES2 was purchased from Invitrogen. AmpliTaq DNA polymerase was purchased from Perkin-Elmer, and the DNA sequencing kit (Sequenase Version 2.0) was purchased from U. S. Biochemical Corp. Most restriction nucleases were purchased from Boehringer Mannheim and New England Biolabs. The DNA extraction kit and plasmid preparation kits were purchased from Qiagen. Monoclonal antibody M2 (against Flag peptide) and its affinity columns were purchased from Kodak. All oligonucleotides were synthesized by IDT, Inc.

Yeast cells were labeled with either D-erythro-[4,5-³H]dihydrosphingosine (60 Ci/mmol) or myo-[2-³H]inositol (15-20 Ci/mmol), and radiolabeled lipids were extracted and resolved on Silica Gel 60 TLC plates as described (12). Labeled lipids were detected and quantitated by PhosphorImager (Storm, Molecular Dynamics) as recommended by the manufacturer.

Total lipids were extracted from exponential phase cells as described (13). Sphingoid bases were measured using high pressure liquid chromatography as described (14). Ceramide and phytoceramide were measured using the diacylglycerol kinase method (15).

Yeast genes were disrupted essentially as described by Wach (16). Using the primer pairs 1 and 2 listed in Table II, the gene disruption cassettes were amplified from the plasmid pFA6 (16) as the template by PCR. PCR reaction conditions were: 1 cycle of 2 min at 94 °C, 30 cycles of 1 min at 94 °C, 45 s at 55 °C, and 1 min at 72 °C followed by 1 cycle of 10 min at 72 °C. The cassettes with the kanMX module flanked by portions of sequence (nucleotides 38-77 at the 5 end and nucleotides 1174-1205 at the 3 end) of the ORF YJL134w (gene YSR2) or portions of sequence (nucleotides 12-48 at the 5 end and nucleotides 1128-1167) of YKR053c (gene YSR2-1) were transformed into wild type diploid strain JK9-3d a/ using a standard method (10), and G418-resistant clones were selected on YPD plates with 220 µg/ml G418. Gene disruptions were confirmed by PCR and Southern blotting analysis (data not shown). The diploid cells carrying the ysr2 or ysr2-1 allele were sporulated, and tetrads were dissected as described (17, 18) to isolate the haploid strains bearing ysr2 allele (YSR2) and ysr2-1 allele (YSR2-1).

Table II. Primers used for PCR in these studies Primer pair Oligonucleotide sequence Purpose 1 F: 5-GAGCCAGGACTCTCTCTAACCCCAATGACTTTCAA GAGCCCAGCTGAAGCTTCGTACGC-3 YSR2 knockout R: 5-AGGACGGGGCTGCACATTACAACGGTGAATGGGCATAGGCCACTAGTGGATCTG-3 2 F: 5-TCAGACGGTTACTGAATTGGGTGTTACCGAGGACACC CAGCTGAAGCTTCGTACGC-3 YSR2-1 knockout R: 5-GGTTGTAGGTATACCTGCGTATATGAGAAACCTTCCGACGGCATAGGCCACTAGTGGATCTG-3 3 F: 5-CGGGGTACCATGGACTACAAGGACGACGATGATAAGGTAGATGGACTGAATACCTCGAACATTAGG-3 YSR2 Flag tagging R: 5-CGGGAATTCTTATGCTATATTTAGAGGGAAAATAGGACGGGGC-3 4 F: 5-CGGGGTACCATGGACTACAAGGACGACGATGATAAGACCATTATTCAGACGGTTACTGAATTGGGTGTTACC-3 YSR2-1 Flag tagging R: 5-CGGGAATTCCTACCTTAAGTTTGTCCAAGTGAAAAAAACTGGGCATAGC-3

The open reading frames of YSR2 and its homologue YSR2-1 were tagged with Flag peptide sequence by PCR under the same conditions as described under "Yeast Gene Disruption" using the primer pairs 3 and 4 listed in Table II. They were then subcloned into pYES2 in EcoRI and KpnI sites to create overexpression constructs pYSR2 and pYSR2-1. Constructs were verified by DNA sequencing to ensure against occurrence of mutations. The expression constructs were introduced into strains JK9-3d , YSR2, and JS16 using a standard method (10). Expression of tagged YSR2 protein and YSR2-1 protein was detected by Western blotting using the anti-Flag peptide monoclonal antibody M2, and the fusion proteins were purified by the M2 immunoaffinity column as recommended by the manufacturer.

Cells overexpressing Flag fusion proteins were suspended in cold 20 mM Tris-HCl (pH 7.4) containing 5 mM EDTA, 0.1 mM PMSF, 20 µg/ml protease inhibitor mixture CLAP (chymostatin, leupeptin, pepstatin, aprotinin), and 5 mM dithiothreitol. Cells were disrupted with glass beads as described (12). Glass beads and cell debris were removed by centrifugation at 2000 rpm for 5 min at 4 °C. The microsome-containing supernatants were transferred to new Eppendorf tubes, pelleted by centrifugation at 100,000 × g for 40 min at 4 °C, resuspended in the extraction buffer, recentrifuged, and resuspended in the extraction buffer without EDTA. To purify the epitope-tagged protein from the microsomes, total membrane proteins were solubilized in lysis buffer (20 mM Tris-HCl, pH 7.4, containing 150 mM NaCl, 1% Tween 20, 0.1 mM PMSF, 1 mM EDTA, and 1 mM EGTA). After removal of membrane debris, the agarose beads coupled with the anti-Flag monoclonal antibody M2 were added to the above protein solution and incubated overnight at 4 °C. The beads were loaded onto the affinity column coupled with M2 antibody and washed with TBS buffer (50 mM Tris-HCl, pH 7.4, containing 150 mM NaCl). The fusion proteins were eluted with 0.1 M glycine (pH 3.5). The eluants were neutralized immediately by adding 0.05 volume of 1 M Tris-HCl (pH 7.4).

Phosphatase activity was measured as described (19). Briefly, D-erythro-[4,5-³H]dihydrosphingosine 1-phosphate, NBD-C₆-ceramide 1-phosphate, or NBD-PA in methanol was dried down under N₂ and resuspended in the assay buffer (100 mM Tris-HCl, pH 7.4, containing 10 mM EDTA, 0.1 mM PMSF, 20 µg/ml CLAP, and 0.3% fatty acid-free bovine serum albumin) to a final concentration of 20 µM and a specific activity of 1 µCi/mmol for D-erythro-[4,5-³H]dihydrosphingosine. In 50 µl of reaction, 25 µl of microsomes (~20-30 µg of protein), whole cell extract, or the affinity-purified protein (~1 ng) were added to 25 µl of the above substrates to initiate the enzymatic reaction at 30 °C for 20 min. Reactions were stopped by adding 1.5 ml of chloroform:methanol (2:1) and 0.35 ml of water. Lipids were extracted and dried. The products were separated from substrates on TLC plates, which were developed in a solvent system (chloroform, methanol, 4.2 N ammonium hydroxide (9:7:2)). Radioactivity of [³H]dihydrosphingosine as well as fluorescent intensity of NBD-C₆-ceramide and NBD-DAG on TLC plates were quantitated using PhosphorImager (Storm, Molecular Dynamics). The dephosphorylation by enzymes was calculated based on formation of D-erythro-[4,5-³H]dihydrosphingosine, NBD-C₆-ceramide, or NBD-DAG. Higher concentrations of protein (>0.1 µg/µl) were determined by the Bradford method (20) using the protein assay kit from Bio-Rad whereas lower concentrations were determined by silver staining (using a silver staining kit, Bio-Rad) of proteins on SDS-polyacrylamide gel with bovine serum albumin as standard.

RESULTS AND DISCUSSION

In an ongoing effort aimed at identifying genes that impart resistance to sphingolipids in the yeast S. cerevisiae, we have identified several genes that impart resistance to sphingosine. One of these genes, designated YSR2 (yeast sphingolipid resistance), has an ORF indicated as YJL134w in the Saccharomyces genomic data base. It encodes a protein with 409 amino acids and is predicted to be a membrane protein with several transmembrane domains. Upon sequence determination, a homologous sequence was found in Saccharomyces genomic data base (the ORF is YKR053c) with 53% identity at the protein level. This gene encodes a protein with 404 amino acids and is here designated as the YSR2-1. Both ORFs were subcloned into the Saccharomyces expression vector pYES2 to form expression constructs, pYSR2 and pYSR2-1, respectively. pYSR2 and pYSR2-1 imparted resistance to sphingosine when overexpressed into the S. cerevisiae mutant strain (JS16) deficient in sphingosine-1-phosphate lyase. This latter mutant has been demonstrated by Saba et al. (21) to be hypersensitive to sphingosine, such that a concentration of 25 µM sphingosine was completely inhibitory to growth. This is probably due to the accumulation of S-1-P (21). Overexpression of both YSR2 and YSR2-1 in JS16 driven by the Gal1 promoter conferred sphingosine resistance up to 35 µM (Fig. 1). These results suggested that the products of the YSR2 and YSR2-1 genes function either in sphingolipid metabolism or in regulating responses of yeast to sphingoid bases. During our analysis of these genes, Stukey and Carman (22) identified a novel phosphatase motif among several lipid phosphatases and hypothetical proteins. Both YSR2 and YSR2-1 proteins encoded by YJL134w and YKR053c were identified to have these phosphatase motifs. We therefore investigated whether these proteins may be DHS-1-P phosphatases.

[View Larger Version of this Image (49K GIF file)]

To examine the effects of these proteins on metabolism of sphingoid bases, JS16 cells overexpressing the proteins were labeled with D-erythro-[4,5-³H]dihydrosphingosine as described under "Experimental Procedures." These cells accumulate DHS-1-P due to the absence of the catabolic enzyme DHS-1-P/S-1-P lyase. Overexpression of either YSR2 or YSR2-1 in this strain prevented accumulation of DHS-1-P (Fig. 2A). This failure to accumulate DHS-1-P is consistent with the function of these gene products as DHS-1-P phosphatase or as regulators of enzymes of sphingolipid metabolism. Moreover, JS16 cells become resistant to sphingosine, probably due to elimination of S-1-P accumulation by overexpression of these two proteins.

ysr2-1

[View Larger Version of this Image (67K GIF file)]

To further investigate the role of these genes in sphingolipid metabolism, the ysr2 mutant and ysr2-1 mutant were established as described under "Experimental Procedures." Twelve asci from the diploid strain bearing ysr2 allele or ysr2-1 were dissected; four ascospores from each of 12 asci were viable. According to this tetrad analysis, both mutants were viable. To investigate if the ysr2 mutant and the ysr2-1 mutant accumulate DHS-1-P, cells were labeled with D-erythro-[4,5-³H]dihydrosphingosine, and total lipids were extracted and analyzed by TLC resolution. We found that the ysr2 mutant accumulated DHS-1-P at the earlier time points, unlike the wild type strains (Fig. 2B) but similar to the JS16 strain (Fig. 2A). At later time points (more than 2 h), the ysr2 mutant strain had the same DHS-1-P level as the wild type strain, but more glycerolipids (PI, PE, and PC) were labeled in the mutant than in the wild type (Fig. 2C). Thus, excess accumulated DHS-1-P may be converted to glycerolipids, probably through the DHS-1-P/S-1-P lyase pathway. The ysr2-1 mutant also accumulated DHS-1-P at the early time points (less than 15 min) compared with the wild type JK9-3d  (Fig. 2D). We conclude that the proteins themselves degrade DHS-1-P in a reaction different from S-1-P lyase, making them strong candidates to be DHS-1-P phosphatases.

To verify if the protein encoded by YSR2 has phosphatase activity, a system for in vitro assay of the phosphatase activity was established. Whole cell extracts from the YSR2 overexpression strain and its wild type strain were obtained. Because the protein is predicted to be a membrane protein, microsomes from both strains were isolated. Using DHS-1-P as a substrate, the enzymatic reaction was carried out using similar conditions as were used for mammalian DHS-1-P phosphatase (19). Using the same amount of proteins, the microsomes from the overexpression strain had 6-fold more DHS-1-P phosphatase activity (0.19 nmol of dihydrosphingosine/min/mg of protein) compared with the wild type strain (0.03 nmol of dihydrosphingosine/min/mg of protein).

To eliminate the possibility that the increase in activity was caused by activation of a regulatory protein or by a nonspecific phosphatase, we tagged the proteins (YSR2 and YSR2-1) with Flag peptide and purified the resultant fusion proteins. To prevent the substrate DHS-1-P from being degraded by S-1-P lyase, we took advantage of the lyase deletion mutant. Therefore, the epitope-tagged proteins (YSR2-Flag and YSR2-1-Flag) were expressed in the JS16 strain using the Saccharomyces expression vector pYES2. The expression was under the control of the Gal1 promoter, which is induced in the presence of galactose. Cells that expressed the fusion proteins were disrupted by vortexing in the presence of glass beads, and microsomes were prepared as described under "Experimental Procedures." The expression of the fusion proteins was detected by Western blot analysis. The fusion proteins were predominantly located in the membrane pellet fraction (Fig. 3A). The fusion proteins were then purified by an affinity column coupled with the monoclonal antibody against Flag peptide. Partially purified fusion proteins were obtained as shown on SDS-polyacrylamide gel electrophoresis with silver staining (Fig. 3B). The activity assay using D-erythro-[4,5-³H]dihydrosphingosine as a substrate demonstrated that the specific activity of the partially purified fusion protein YSR2 increased 250-fold as compared with the starting crude membrane extract, whereas the specific activity of YSR2-1 increased approximately 200-fold (Table III). Thus, both proteins are enzymes and have an activity to dephosphorylate DHS-1-P. Using the partially purified proteins, we next tested their substrate specificity. No phosphatase activity toward ceramide 1-phosphate and phosphatidic acid was detected in the partially purified proteins (Table III). Thus, we conclude that the YSR2 and YSR2-1 genes encode lipid phosphatases, and they specifically dephosphorylate DHS-1-P.

[View Larger Version of this Image (26K GIF file)]

Table III. Purification and substrate specificity of Flag fusion proteins The Flag-tagged YSR2 and YSR2-1 were partially purified as described under "Experimental Procedures." Phosphatase activities toward DHS-1-P, NBD-C6-ceramide 1-phosphate (Cer-1-P), and NBD-phosphatidic acid (PA) were measured as described under "Experimental Procedures." The results indicate the specific activity (dihydrosphingosine, ceramide, and DAG (nmol/min/mg of protein). Substrate Enzyme activity JK9-3d  curde membrane extract YSR2 crude membrane extract YSR2-1 crude membrane extract YSR2 affinity pure proteins YSR2-1 affinity pure proteins DHS-1-P 0.05 0.18 0.16 49.2 32.8 Cer-1-P 0.17 0.11 0.16 0.0 0.13 PA 0.29 0.22 0.26 0.06 0.11

The cells of the ysr2 deletion strain (YSR2) and the wild type strain (JK9-3d ) were labeled with [³H]dihydrosphingosine for 2 h. Unlike in the JK9-3d  strain, none of the major sphingolipids such as inositol phosphorylceramide, mannosylinositol phosphorylceramide, and mannosyldiinositol phosphorylceramide were labeled in the YSR2 strain. Failure of sphingolipid labeling in the mutant strain was restored by re-introducing the YSR2 gene back into the YSR2 strain (Fig. 2C). Importantly, this failure to label sphingolipids was not due to a generalized defect in sphingolipid biosynthesis. First, the YSR2 strain had normal levels of sphingolipids when analyzed by mass measurements of sphingoid bases or ceramide levels (232.0 ± 16 pmol of sphingoids including dihydrosphingosine and phytosphingosine, and 78.0 ± 12 pmol of phytoceramide per 5 × 10⁸ wild type (JK9-3d ) cells; 266.0 ± 9 pmol of sphingoids and 66.0 ± 14 pmol of phytoceramide per 5 × 10⁸ ysr2 cells). Second, when the mutant cells and the wild type cells were labeled with [³H]inositol, all sphingolipids were labeled in both types of cells (Fig. 4A). These data show that the failure of the deletion mutant to incorporate dihydrosphingosine into sphingolipids is due to the inability to dephosphorylate DHS-1-P, suggesting a role for phosphorylation/dephosphorylation of exogenous dihydrosphingosine in incorporation into sphingolipids.

[View Larger Version of this Image (87K GIF file)]

When the JK9-3d  strain and the YSR2 strain were pretreated with 200 µM fumonisin B1 and then labeled with [³H]dihydrosphingosine, phytosphingosine accumulated in wild type cells but not in the ysr2 mutant cells (Fig. 4B). Thus exogenously supplied dihydrosphingosine was not converted to phytosphingosine in the mutant. These data also suggest that dihydrosphingosine is not directly converted to phytosphingosine without prior phosphorylation followed by dephosphorylation.

In contrast to labeling of sphingolipids, we observed that glycerolipids including PI, PE, and PC were labeled 5-10 times more in the DHS-1-P phosphatase deletion strain (YSR2) than in the wild type strain (JK9-3d ) (Fig. 2C). On the other hand, in the DHS-1-P/S-1-P lyase-deficient mutant (JS 16), glycerolipids such as PI, PE, and PC were hardly labeled when cells were fed [³H]dihydrosphingosine (Fig. 2A). Therefore, dihydrosphingosine was probably incorporated into these glycerolipids through phosphorylation followed by cleavage by DHS-1-P/S-1-P lyase. Taken together, these results suggest that the fate of exogenous dihydrosphingosine appears to be a result of the distinct action of the DHS-1-P phosphatase and the DHS-1-P/S-1-P lyase with the latter enzyme being responsible for the shift of sphingolipid metabolites into glycerolipids. These data are in agreement with previous studies that have suggested a relationship between sphingolipid and glycerolipid metabolism (1, 2). For example, treatment of yeast or mammalian cells with the mycotoxin fumonisin B1, which blocks sphingolipid biosynthesis, resulted in an elevation in the concentrations of free sphingoid bases such as dihydrosphingosine, phytosphingosine, and dihydrosphingosine 1-phosphate as well as an increase in the concentrations of glycerolipids (23, 24). Therefore, these data also suggest that the sphingolipid pathway interacts with the glycerolipid pathway.

While this work was in progress, Qie et al. (25) identified a gene they named LCB3 with an identical sequence to YSR2. They demonstrated that LCB3 was necessary for incorporation of exogenous long chain bases into sphingolipids. They speculated that YSR2/LCB3 transports or facilitates uptake of long chain bases. Based on the fact that the ysr2 mutant accumulates DHS-1-P and glycerolipids when exogenous dihydrosphingosine is added to the medium, it does not appear that YSR2 acts on uptake of sphingoid bases. Rather, dihydrosphingosine may need to be phosphorylated to facilitate its uptake and delivery to the cellular compartment where it will be dephosphorylated by YSR2 and then incorporated into sphingolipids.

Although YSR2-1 has DHS-1-P phosphatase activity as indicated in the in vitro assay (Table III) and in the study where the ysr2-1 mutant accumulated DHS-1-P at early time points (Fig. 2D), defects of YSR2-1 only slightly reduced synthesis of sphingolipids from exogenous [³H]dihydrosphingosine (Fig. 4C). Therefore, while sharing similarity in sequence and biochemical function, YSR2 and YSR2-1 appear to play different physiological functions. This observation requires further study of the YSR2-1 gene.

In conclusion, this study clearly demonstrates that the sphingolipid and the glycerolipid pathways are closely connected through phosphorylation of dihydrosphingosine. Dephosphorylation of DHS-1-P then provides a substrate for sphingolipid biosynthesis whereas cleavage of DHS-1-P by the lyase provides a substrate for glycerolipid biosynthesis. These studies, therefore, demonstrate that YSR2 is a DHS-1-P phosphatase that plays a critical role in these metabolic interconnections.