Purification and Characterization of Rat Kidney Sphingosine Kinase*

Sphingosine kinase catalyzes the formation of the bioactive sphingolipid metabolite sphingosine 1-phosphate, which plays important roles in numerous physiological processes, including growth, survival, and motility. We have purified rat kidney sphingosine kinase 6 × 105-fold to apparent homogeneity. The purification procedure involved ammonium sulfate precipitation followed by chromatography on an anion exchange column. Partially purified sphingosine kinase was found to be stabilized by the presence of high salt, and thus, a scheme was developed to purify sphingosine kinase using sequential dye-ligand chromatography steps (since the enzyme bound to these matrices even in the presence of salt) followed by EAH-Sepharose chromatography. This 385-fold purified sphingosine kinase bound tightly to calmodulin-Sepharose and could be eluted in high yield with EGTA in the presence of 1 m NaCl. After concentration, the calmodulin eluate was further purified by successive high pressure liquid chromatography separations on hydroxylapatite, Mono Q, and Superdex 75 gel filtration columns. Purified sphingosine kinase has an apparent molecular mass of ∼49 kDa under denaturing conditions on SDS-polyacrylamide gel, which is similar to the molecular mass determined by gel filtration, suggesting that the active form is a monomer. Sphingosine kinase shows substrate specificity ford-erythro-sphingosine and does not catalyze the phosphorylation of phosphatidylinositol, diacylglycerol, ceramide,dl-threo-dihydrosphingosine, orN,N-dimethylsphingosine. However, the latter two sphingolipids were potent competitive inhibitors. With sphingosine as substrate, the enzyme had a broad pH optimum of 6.6–7.5 and showed Michaelis-Menten kinetics, with K m values of 5 and 93 μm for sphingosine and ATP, respectively. This study provides the basis for molecular characterization of a key enzyme in sphingolipid signaling.

class of lipid second messengers (1)(2)(3)(4). Ceramide is an important regulatory component of stress responses and programmed cell death, known as apoptosis (2,5,6). In contrast, we have implicated a further metabolite of ceramide, SPP, as a second messenger in cellular proliferation and survival induced by platelet-derived growth factor, nerve growth factor, and serum (7)(8)(9). Previously, we showed that SPP protects cells from apoptosis resulting from elevations of ceramide (7,9) and proposed that the dynamic balance between levels of the sphingolipid metabolites (ceramide and SPP) and consequent regulation of opposing signaling pathways is an important factor that determines whether a cell survives or dies (7). Recently, we demonstrated that this ceramide/SPP rheostat is an evolutionarily conserved stress regulatory mechanism influencing growth and survival of yeast (10). A variety of stress stimuli, including Fas ligand, tumor necrosis factor-␣, interleukin-1, growth factor withdrawal, anticancer drugs, oxidative stress, heat shock, and ionizing radiation, stimulate sphingomyelinase, leading to increased ceramide levels (2,5,6), whereas platelet-derived growth factor and other growth factors stimulate ceramidase and sphingosine kinase and elevate SPP levels (3,4,8,11). Progress in determining the importance of these sphingolipid metabolites has been hampered because most of the relevant metabolic enzymes have not yet been purified or cloned.
The level of SPP in cells is low and determined by the relative contributions of its formation, mediated by sphingosine kinase (12), and its degradation, catalyzed by an endoplasmic reticulum pyridoxal phosphate-dependent lyase and specific phosphatases (10,13,14). SPP was initially described as an intermediate in the degradation of long-chain sphingoid bases (15). However, the roles of SPP in cellular proliferation, survival, and other cellular responses (reviewed in Ref. 16), as well as the observations that SPP triggers novel signal transduction pathways of calcium mobilization (17,18), activation of phospholipase D (19), and the Raf/MKK/ERK signaling cascade (20,21), suggest that the importance of sphingosine kinase is not restricted to the catabolism of sphingolipids as was originally proposed nearly 20 years ago (22).
Sphingosine kinase is a ubiquitous enzyme found in yeast (22); Tetrahymena pyriformis (23); rat liver, kidney and brain (24,25); bovine brain (26); and human and porcine platelets (25,27). Although it is known that sphingosine kinase activity increases in response to certain growth-promoting agents, such as platelet-derived growth factor (8,28,29), phorbol esters (3,7,30), the B subunit of cholera toxin (31), and nerve growth factor (9), little is yet known about the properties or mechanism of regulation of sphingosine kinase. Thus, purification and characterization of sphingosine kinase are important goals to gain understanding of its physiological roles. We have now succeeded in purifying sphingosine kinase from rat kidneys by Ͼ6 ϫ 10 5 -fold to apparent homogeneity.
Assay of Sphingosine Kinase Activity-Sphingosine kinase activity was determined as described previously with minor modifications (12). Samples (up to 40 g) and 10 l of 1 mM sphingosine (dissolved in 5% Triton X-100) were mixed with buffer A (20 mM Tris (pH 7.4), 20% glycerol, 1 mM mercaptoethanol, 1 mM EDTA, 1 mM sodium orthovanadate, 40 mM ␤-glycerophosphate, 15 mM NaF, 10 g/ml leupeptin, 10 g/ml aprotinin, 10 g/ml soybean trypsin inhibitor, 1 mM phenylmethylsulfonyl fluoride, and 0.5 mM 4-deoxypyridoxine) in a total volume of 190 l, and reactions were started by addition of 10 l of [ 32 P]ATP (10 Ci, 20 mM) containing MgCl 2 (200 mM) and incubated for 5 or 15 min at 37°C. Reactions were terminated by addition of 20 l of 1 N HCl followed by 0.8 ml of chloroform/methanol/HCl (100:200:1, v/v). After vigorous vortexing, 240 l of chloroform and 240 l of 2 M KCl were added, and phases were separated by centrifugation. The labeled lipids in the organic phase were resolved by TLC on Silica Gel G60 with 1-butanol/ethanol/acetic acid/water (80:20:10:20, v/v) and visualized by autoradiography. The radioactive spots corresponding to authentic SPP were identified as described (32), scraped from the plates, and counted in a scintillation counter or, alternatively, quantified with a Molecular Dynamics Storm PhosphorImager. Sphingosine kinase specific activity is expressed as pmol of SPP formed per min (unit)/mg of protein. At each purification step, linearity of the enzymatic reaction with time of incubation and protein concentration was observed.
Extraction and Ammonium Sulfate Fractionation-Frozen rat kidneys (0.5 kg) were thawed in 0.5 liter of cold 20 mM Tris (pH 7.4) containing 20% glycerol, 1 mM dithiothreitol, 1 mM EDTA, 0.5 g/ml leupeptin, 0.5 g/ml aprotinin, 0.5 g/ml soybean trypsin inhibitor, and 0.2 mM phenylmethylsulfonyl fluoride (buffer B), decapsulated, transferred to fresh buffer (2 ml/g), minced, and then homogenized in a blender. After centrifugation at 5000 ϫ g for 15 min, the supernatant fraction was filtered through glass wool, centrifuged at 20,000 ϫ g for 1 h, filtered through glass wool again, and centrifuged at 100,000 ϫ g for 90 min. This supernatant was then centrifuged at 100,000 ϫ g for 1 h to obtain the cytosolic fraction, which was fractionated by precipitation with ammonium sulfate. The 25-45% ammonium sulfate precipitate was resuspended in 150 ml of buffer B and dialyzed overnight against the same buffer.
DEAE-cellulose Chromatography-The dialysate was clarified by centrifugation for 15 min at 100,000 ϫ g and applied overnight at a flow rate of 70 ml/h to a DEAE-cellulose column (5-cm diameter, 500-ml bed volume) equilibrated with buffer B. After washing with buffer B, proteins were eluted with a linear gradient to 0.5 M NaCl in the same buffer at a flow rate of 250 ml/h. Sphingosine kinase activity was determined on aliquots of each fraction, and peak activity fractions were pooled.
Dye-Ligand Affinity Chromatography on Green A-and Blue A-Sepharose Columns-The pooled DEAE fractions were applied to a green A-Sepharose column (5-cm diameter, 200 -250-ml bed volume) equilibrated with buffer B containing 0.2-0.22 M NaCl at a flow rate of 120 ml/h. After the sample was applied, the flow rate was increased to 300 ml/h, and stepwise elution was performed with 3 bed volumes of buffer B containing 0.2 and 0.4 M NaCl, respectively, and then 6 bed volumes of buffer B containing 1 M NaCl. The 1 M NaCl fraction contained most of the sphingosine kinase activity and was diluted 1:1 with buffer B and applied at 120 ml/h to a blue A-Sepharose column (5-cm diameter, 125-ml bed volume) equilibrated with buffer B containing 0.5 M NaCl. The blue A-Sepharose was washed at 250 ml/h with 3 bed volumes of buffer B containing 0.5 M NaCl and with 3 bed volumes of buffer B containing 0.7 M NaCl, and the sphingosine kinase activity was eluted with 6 bed volumes of buffer B containing 2 M NaCl.
EAH-Sepharose Chromatography-Half of the 2 M NaCl fraction was concentrated 100-fold in Centricon Plus-20 concentrators (M r 10,000 cutoff) and then dialyzed against buffer B. The dialysate was centrifuged to remove precipitated proteins and, after addition of Triton X-100 to a final concentration of 0.05%, was loaded onto a 20-ml EAH column pre-equilibrated with buffer B containing 0.05% Triton X-100. The EAH column was washed stepwise with 60-ml fractions of buffer B and 0.05% Triton X-100 containing 0, 30, 150, 200, and 600 mM NaCl. Most of the sphingosine kinase activity was eluted with 150 mM NaCl.
Affinity Chromatography on Calmodulin-Sepharose 4B-CaCl 2 was added to the EAH fraction to a final concentration of 4 mM and immediately applied to a calmodulin-Sepharose 4B column (12 ml) preequilibrated with buffer B without EDTA and containing 100 mM NaCl, 0.05% Triton X-100, 2 mM CaCl 2 , and 10% sucrose. The column was washed successively with 30 ml of equilibration buffer and 60 ml of the same buffer containing 2 mM EGTA. Sphingosine kinase activity was TABLE I Comparison of sphingosine kinase activity from various sources Cytosolic fractions of various tissues and S. cerevisiae were obtained after ultracentrifugation at 100,000 ϫ g essentially as described for rat kidneys under "Experimental Procedures." The specific activity of sphingosine kinase from each source is expressed as pmol of SPP formed per min/mg relative to rat kidneys. Percent cytosolic ϭ (activity in 100,000 ϫ g supernatant/total activity in homogenate) ϫ 100. TABLE II Purification of sphingosine kinase from rat kidney Sphingosine kinase was purified from 2.5 kg of rat kidneys (ϳ2500 kidneys) as described under "Experimental Procedures." One unit of sphingosine kinase activity is 1 pmol of SPP formed from sphingosine per min. Up to the ammonium sulfate fractionation step, sphingosine kinase activity was determined by incubation for 5 min at 37°C to ensure linearity of reactions. With more purified fractions, assays were carried out for 15 min. After the blue A-Sepharose step, the activity was frozen and stored at Ϫ70°C. This fraction was usually thawed within several weeks and used for the next purification steps. It should be noted that concentration and dialysis of the blue A-Sepharose eluate resulted in loss of 30 -50% of the sphingosine kinase activity. then eluted with 30 ml of equilibration buffer containing 2 mM EGTA and 1 M NaCl, and five fractions of 6 ml each were collected. The sphingosine kinase-containing fractions (fractions 2 and 3) were concentrated in a Centriprep-10 concentrator, diluted with buffer B containing 0.05% Triton X-100 and 10% sucrose, and reconcentrated until the final salt concentration was 50 mM or less.
Hydroxylapatite Chromatography-The concentrated calmodulin-Sepharose fraction was injected onto a hydroxylapatite column equilibrated with buffer B containing 0.05% Triton X-100, 10% sucrose, and 25 mM potassium phosphate and was eluted at a flow rate of 0.4 ml/min with a linear gradient of 0.025-0.5 M potassium phosphate in the same buffer (Waters HPLC system). Fractions (0.4 ml) were collected in tubes containing 100 l of 5 M NaCl, and sphingosine kinase activity was determined.
Mono Q Anion Exchange Chromatography-The pooled hydroxylapatite fraction was concentrated and desalted using Microcon-10 concentrators and injected onto a Mono Q 5/5 column (Waters HPLC system) equilibrated with buffer B containing 0.05% Triton X-100, 10% sucrose, and 15 mM potassium phosphate. The column was washed for 10 min with equilibration buffer at 1 ml/min, and then a linear gradient of 0.015-0.5 M potassium phosphate was applied. Fractions of 1 ml were collected in tubes containing 250 l of 5 M NaCl. Sphingosine kinase activity was eluted as two broad peaks. Each peak was pooled and reapplied to a small calmodulin-Sepharose column (1 ml) to concentrate the sample since concentration by ultrafiltration usually resulted in a marked loss of activity. Furthermore, this second calmodulin-Sepharose column decreased the Triton X-100 concentration, which would have resulted from ultrafiltration of the Mono Q fractions and which would interfere with further purification.
Gel Filtration Chromatography on Superdex 75-The eluate from the calmodulin affinity column was concentrated to 200 l using Microcon-10 concentrators and then injected onto a Superdex 75 gel filtration column (Waters HPLC system) pre-equilibrated with buffer B containing 0.05% Triton X-100, 10% sucrose, and 1 M NaCl. Proteins were eluted at a flow rate of 0.4 ml/min, and 0.4-ml fractions were collected. The column was calibrated using bovine serum albumin, ovalbumin, chymotrypsinogen, and lysozyme as standard proteins to determine the apparent molecular mass of native sphingosine kinase.
SDS-Polyacrylamide Gel Electrophoresis-SDS-polyacrylamide gel electrophoresis was performed essentially as described by Laemmli (33). Protein samples were boiled for 5 min in sample buffer with or without a reducing agent and loaded onto 12% gels. Molecular masses of the various protein bands were estimated with the low molecular mass range prestained standard proteins from Bio-Rad. Proteins were visualized by silver staining. Sample buffer containing ␤-mercaptoethanol often produced artificial bands using the sensitive silver stain procedure, and thus, most gels were run under nonreducing conditions.
Protein Determination-Proteins were determined with either the Coomassie dye binding method (Pierce) or the Lowry procedure after precipitation with 7% trichloroacetic acid in the presence of 0.015% deoxycholate (Peterson variation (34)). After Superdex 75 gel filtration, the protein concentration was too low for determination by these methods and was estimated from optical densities obtained by scanning silver-stained gels using bovine serum albumin as a standard.
Characterization-Purified sphingosine kinase obtained after the gel filtration step was used for characterization studies. To determine pH dependence, the following buffers were used: pH 4 -5, 200 mM sodium acetate; pH 6.0 -6.  The dialyzed 25-45% ammonium sulfate precipitate was applied to a DEAE-cellulose column preequilibrated with buffer B, and the proteins were eluted with a linear gradient (0 -0.5 M) of NaCl, collecting 20-ml fractions. Sphingosine (Sph) kinase activity (units/fraction; q) and protein (mg/fraction; OO) were measured as described under "Experimental Procedures." Similar results were obtained in at least six experiments.

FIG. 2. Purification of sphingosine kinase on green A-and blue A-Sepharose columns.
A, sphingosine (Sph) kinase activity eluted from the DEAE-cellulose column was applied to a green A-Sepharose column equilibrated with buffer B containing 0.2 M NaCl and eluted stepwise with the indicated concentrations of NaCl as described under "Experimental Procedures." B, the sphingosine kinase-containing fraction from the green A column was diluted 1:1 with buffer B, applied to a blue A-Sepharose column equilibrated with buffer B containing 0.5 M NaCl, and eluted stepwise with increasing NaCl concentrations as described under "Experimental Procedures." Protein concentration and sphingosine kinase activity were measured in each fraction. Results are expressed as percentage of total protein or activity applied to each of the columns. Similar results were obtained in at least six experiments. RT, run-through.

RESULTS AND DISCUSSION
Sphingosine Kinase Activity in Various Tissues-It has been reported that sphingosine kinase is a ubiquitous enzyme (12, 22-27, 35, 36). Using our recently developed quantitative sphingosine kinase assay (12, 35), we measured sphingosine kinase activity in various rat tissues to determine which source would be most appropriate for purification. In agreement with a previous qualitative study (24), we found that spleen and kidney have higher specific activities than liver, which has about twice the activity of brain (Table I). More than 50% of the

FIG. 3. Separation of sphingosine kinase by chromatography on EAH-Sepharose and calmodulin-Sepharose columns.
A, a portion of the blue A-Sepharose 2 M NaCl eluate was concentrated and dialyzed. Triton X-100 was added to a final concentration of 0.05% before application to EAH-Sepharose. The column was eluted stepwise with increasing NaCl concentrations. B, the sphingosine (Sph) kinase-containing fractions from the EAH column were applied to a calmodulin-Sepharose column after addition of CaCl 2 , and the proteins were then eluted stepwise with buffer B containing 10% sucrose, 0.05% Triton X-100, and 1 mM EGTA without and then with 1 M NaCl added. Results are expressed as percentage of the total protein or sphingosine kinase activity applied to each of the columns. Similar results were obtained in at least six experiments. RT, run-through. sphingosine kinase activity in kidney, liver, and brain was in the cytosolic fraction, independently of ionic strength of the extraction buffer. SPP levels have also been measured in these tissues (37). However, there does not appear to be a good correlation between sphingosine kinase activity and SPP levels, as the highest levels of SPP were found in brain and spleen. Furthermore, kidney has more SPP than liver, which contains only very low levels (37,38). Although we found that spleen has high sphingosine kinase activity and SPP levels, we selected kidneys as the source for the purification of sphingosine kinase since initial experiments demonstrated that the majority of the sphingosine kinase activity in spleen was associated with membranes (Table I) and unstable. Bovine kidneys were also examined as a potential source for purification of sphingosine kinase since they are much larger. However, the specific activity in the cytosolic fraction from bovine kidneys was lower than that in rat kidney cytosol (Table I), and very poor recoveries were found after preliminary ammonium sulfate fractionations (data not shown). Interestingly, we have found that most of the sphingosine kinase in Saccharomyces cerevisiae is in the microsomal fraction and not in the cytosol (Table I). These results suggest that various forms of sphingosine kinase may exist depending on the tissue and species.
Purification of Rat Kidney Sphingosine Kinase- Table II summarizes the purification of sphingosine kinase from 2500 rat kidneys. Sphingosine kinase was purified ϳ6 ϫ 10 5 -fold to near homogeneity, with a total activity recovery of 0.6%. After an initial 20,000 ϫ g centrifugation of the homogenate, it was important to carry out two subsequent centrifugations at 100,000 ϫ g since after ammonium sulfate precipitation of the 20,000 ϫ g supernatant and consequent dialysis, 60% of the sphingosine kinase activity became insoluble and could not be solubilized by detergents or in different buffers with pH values ranging from 6.0 to 8.0. However, when the second 100,000 ϫ g supernatant was fractionated with ammonium sulfate, most of the sphingosine kinase activity could be resolubilized, and very little was irreversibly associated with insoluble material.
In contrast to bovine brain sphingosine kinase (26), rat kidney sphingosine kinase binds to DEAE-cellulose at pH 7.5 in Tris buffer, but not in phosphate buffer. The activity was eluted with 0.2 M NaCl from DEAE-cellulose as a single broad peak (Fig. 1). As expected for a nucleotide-binding protein, sphingosine kinase binds tightly to both green A and blue A dye-matrix columns (Fig. 2, A and B), even in the presence of relatively high salt concentrations (0.2 M for the green A column and 0.5 M for the blue A column). This makes it possible to directly apply the pooled DEAE fractions containing sphingosine kinase activity to these dye-matrix columns in a sequential manner. This was advantageous since either concentration or dialysis of sphingosine kinase activity resulted in considerable loss of activity. Sphingosine kinase was purified 270 -400-fold after the dye-ligand chromatography steps, with a yield of 30 -50%. Sphingosine kinase activity eluted from the blue A column in 2 M NaCl could be stored for several weeks at Ϫ70°C after quick freezing in liquid nitrogen. It should also be noted that concentration and dialysis of the blue A column eluate resulted in a loss of ϳ30 -50% of the activity. Subsequent purification steps were then repeated two times with half of this fraction since poor recoveries were found when larger columns were used for the chromatography separations described below. Addition of 0.05% Triton X-100, but not Nonidet P-40 or ␤-octyl glucopyranoside, markedly improved both the recovery and the resolution of proteins and was thus included in subsequent steps. Triton X-100 has previously been successfully used to stabilize a number of other lipid enzymes, such as phospholipase A 2 , ceramidase, acid sphingomyelinase, and phosphoinositide 4-kinase, since it prevents aggregation and nonspecific adsorption to surfaces (39 -42).
Several attempts were made to purify sphingosine kinase by affinity chromatography on immobilized sphingosine (43) or ATP and by hydrophobic chromatography, but they were un- successful probably due to the necessity of maintaining the enzyme in solutions containing high salt concentrations and detergent. We found that sphingosine kinase binds tightly to 1,6-diaminohexane covalently linked to Sepharose 4B (EAH-Sepharose), but could not be eluted from this matrix by substrates containing primary amino groups, such as sphingosine or choline, in a similar manner, as was previously found for purification of choline kinase (44). In contrast, at least 70% of the applied sphingosine kinase activity was eluted from EAH-Sepharose with 0.15 M NaCl, whereas only 14 -18% of the applied proteins were eluted in this fraction, resulting in 3-5fold purification (Fig. 3A).
This EAH fraction was immediately applied to a calmodulin-Sepharose column after addition of CaCl 2 . Most of the protein applied (96 -98%) did not bind and was eluted with the wash buffer. In contrast, most of the sphingosine kinase activity was tightly retained, and only a small fraction of the activity could be eluted with EGTA in the absence of CaCl 2 (Fig. 3B). However, Ͼ95% of the activity was eluted when the calmodulin column was eluted with 2 mM EGTA solution containing 1 M NaCl (Fig. 3B), resulting in at least 20-fold purification. Similarly, another calmodulin-binding lipid kinase, inositol-1,4,5-trisphosphate 3-kinase, from either human platelets (45) or rat brain (46) could not be eluted from a calmodulin-Sepharose column unless the elution buffer contained EGTA as well as 0.5% Triton X-100 or 0.2% SDS, respectively. Despite purification of sphingosine kinase by Ͼ9000-fold at this stage, several protein bands were still evident on silver-stained SDS-polyacrylamide gels. Detectable amounts of calmodulin also coeluted with the sphingosine kinase activity (Fig. 4B). One of the difficulties we encountered in later steps of the purification of sphingosine kinase was that the kinase rapidly lost activity. However, we found that addition of 10% sucrose, which has been shown to stabilize other lipid enzymes, including squalene synthetase (47), to buffers containing 0.05% Triton X-100 and high salt concentrations further increased stability at these stages of purification.
Sphingosine kinase was then further purified by hydroxylapatite chromatography followed by HPLC on a Mono Q column. As shown in Fig. 4, sphingosine kinase activity was eluted from a hydroxylapatite column as a single peak with a gradient of increasing concentrations of potassium phosphate at ϳ0.06 M, resulting in 6-fold additional purification. Although sphingosine kinase is tightly bound to the strong anion exchanger Mono Q and can be eluted at an ionic strength of 0.15 M with a salt gradient (data not shown), it is not as tightly bound when applied in the presence of 15 mM potassium phosphate. We found that most of the activity could be eluted with the wash buffer alone in this case, greatly improving separation from other more tightly bound proteins (Fig. 5). Sphingosine kinase was resolved into two activity peaks by chromatography on Mono Q. For further purification, only the first activity peak was utilized since the major protein band in this fraction was a 49-kDa polypeptide (Fig. 5B), and preliminary experiments utilizing several different sequences of purification suggested that this 49-kDa polypeptide correlated with the sphingosine kinase activity.
Because SDS-polyacrylamide gel electrophoresis analysis suggested that sphingosine kinase was highly purified at this stage, we decided to do a final purification by gel filtration chromatography. However, it was necessary to concentrate the Mono Q eluate to a small volume before injection onto a Superdex 75 gel filtration HPLC column. Ultrafiltration at this stage resulted in large losses of activity and high concentrations of Triton X-100, which interfere with the gel filtration. Thus, we used a 1-ml calmodulin-Sepharose column to concentrate the sample 9-fold, eluted exactly as above described for the largescale calmodulin-Sepharose column. Further concentration of the calmodulin eluate by ultrafiltration did not result in major losses of activity, and the resulting concentration of Triton X-100 did not interfere with the resolution of the gel filtration column.
As shown in Fig. 6, the sphingosine activity was eluted from a Superdex 75 gel filtration column at a volume corresponding to an apparent native molecular mass of ϳ59 kDa when compared with standard proteins (Fig. 6B). Silver-stained SDSpolyacrylamide gels revealed that the fractions with the highest sphingosine kinase activity (fractions 23-25) contained a single 49-kDa polypeptide under both reducing and nonreducing conditions. Thus, sphingosine kinase isolated from rat kidney is likely active as a monomer since the apparent native molecular mass of sphingosine kinase was similar to its molecular mass on SDS-polyacrylamide gel. However, the possibility that sphingosine kinase purified to homogeneity from rat kidney cytosol may be an active fragment of membrane-bound sphingosine kinase cannot be excluded. It should be noted that the native molecular mass of Swiss 3T3 fibroblast sphingosine kinase determined by gel filtration was identical to that of highly purified rat kidney sphingosine kinase (data not shown). The specific activity of highly purified rat kidney sphingosine kinase (100 mol/min/mg) is ϳ10 -100-fold higher than that of several other highly purified lipid kinases, including phosphoinositide 3-kinase (48,49), phosphoinositide 4-kinase (40), phospholipase D (50), and phospholipase A 2 (41), but it is the same order of magnitude as that reported for acid sphingomyelinase from human urine (51).
Characterization of Sphingosine Kinase-To characterize purified sphingosine kinase, a number of experiments were carried out to examine the pH dependence, substrate specificity, and enzyme kinetics. Purified sphingosine kinase was active from pH 6 to Ͼ8, with a broad pH optimum between pH 6.6 and 7.5 (Fig. 7A). At pH 7.4, activity increased linearly with the incubation time for the first 60 min of the reaction and then gradually decreased. Sphingosine kinase activity was maximal at a concentration of 5-10 mM MgCl 2 (Fig. 7B), whereas physiological concentrations of calcium (1-100 M) had no effect on its activity. Activity with D-erythro-sphingosine showed typical Michaelis-Menten kinetics, with K m ϭ 5.1 Ϯ 1.7 M and V max ϭ 101 Ϯ 22 mol/mg/min (Fig. 7C). The K m for ATP was 93 M (Fig. 7D), similar to K m values for other lipid kinases (40,48). Next, we examined the substrate specificity for sphingosine kinase. The naturally occurring D(ϩ)-erythro-trans-isomer was the best substrate for purified rat kidney sphingosine kinase. DL-erythro-Dihydrosphingosine was also phosphorylated, but to a lesser extent (30% compared with D(ϩ)-erythro-sphingosine), whereas DL-threo-dihydrosphingosine, L-threo-dihydrosphingosine, ceramide, diacylglycerol, and phosphatidylinositol were not phosphorylated (Fig. 8A). N,N-Dimethylsphingosine and DL-threo-dihydrosphingosine have previously been used to decrease SPP levels stimulated by various physiological stimuli (7,8,18). We have now found that both N,N-dimethylsphingosine and DL-threo-dihydrosphingosine are potent competitive inhibitors of purified sphingosine kinase, with K i values of 9.9 Ϯ 1.0 and 5.2 Ϯ 0.5 M, respectively (Fig. 8, B and C). These results further substantiate the usefulness of these compounds as tools to inhibit sphingosine kinase activity in vivo and to examine the role of SPP in diverse cellular responses.
In summary, rat kidney sphingosine kinase has been purified to homogeneity by Ͼ6 ϫ 10 5 -fold, indicating that sphingosine kinase is a low abundance protein in rat kidney and probably in other tissues as well. Similarly, another kinase, choline kinase, has been partially purified from rat kidney by Ͼ200,000-fold (44). This study provides the basis for molecular characterization of sphingosine kinase and will aid in elucidation of its role in various physiological processes.