HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: M500321-JLR200v1 47/2/268    most recent Alert me when this article is cited Alert me if a correction is posted Citation Map Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Van Overloop, H. Articles by Van Veldhoven, P. P. Search for Related Content PubMed PubMed Citation Articles by Van Overloop, H. Articles by Van Veldhoven, P. P. Social Bookmarking                      What's this?

Journal of Lipid Research, Vol. 47, 268-283, February 2006
Copyright © 2006 by American Society for Biochemistry and Molecular Biology

Further characterization of mammalian ceramide kinase: substrate delivery and (stereo)specificity, tissue distribution, and subcellular localization studies

Helena Van Overloop, Sofie Gijsbers,¹ and Paul P. Van Veldhoven²

Katholieke Universiteit Leuven, Faculteit Geneeskunde, Departement Moleculaire Celbiologie, Afdeling Farmacologie, Leuven, Belgium

Published, JLR Papers in Press, November 3, 2005.

¹ Present address of S. Gijsbers: Katholieke Universiteit Leuven, Rega Instituut, Afdeling Virologie en Chemotherapie, Minderbroederstraat 10, B-3000 Leuven, Belgium.

³ Because of this fact, and by using another reading frame, this clone was predicted to code for KIAA1646 protein (AB051433).

⁴ Several lipophilic phosphorylatable compounds, including all four sphinganine and sphingenine stereoisomers, 4-hydroxysphinganine, (truncated) (dihydro)ceramides, (truncated) 1,2-sn-diacylglycerol, 1- or 2-monoacylglycerols, 1-monoalkylglycerols, and phosphatidylinositol, as well as less well-studied kinase substrates (tocopherol, cholesterol, farnesol) or putative substrates (galactosylceramide, anandamide), have been tested as substrates under different delivery modes (EtOH/BSA, octyl-ß-D-glucopyranoside/cardiolipin, CHAPS, DMSO) in the presence of Ca or Mg ions, with bacterially expressed kinase, lysates of CHO cells stably transfected with HsLK2-Myc-His, mitochondria isolated from such cells, or in vitro-transcribed/translated kinase. Compared with the appropriate controls, no remarkable differences were noticed. In addition, in lipid extracts from CHO cells cultured in the presence of labeled glucose (precursor of glycerolipids) or serine (precursor of sphingolipids) or uploaded with ³²P, no additional or enhanced radiolabeled products were revealed when analyzed by two-dimensional TLC/autoradiography and compared with cells stably transfected with the empty vector. On the other hand, these experiments showed that E. coli diacylglycerol kinase can phosphorylate anandamide and that coupled transcription/translation systems phoshorylate sphingoid bases, diacylglycerol, and ceramide (P. P. Van Veldhoven and K. De Greef, unpublished data).

² To whom correspondence should be addressed. e-mail: paul.vanveldhoven{at}med.kuleuven.be

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Recombinant human ceramide kinase (HsCERK) was analyzed with regard to dependence on divalent cations and to substrate delivery, spectrum, specificity, and stereoselectivity. Depending on the chain length of the ceramide, either albumin for short-chain ceramide or a mixed micellar form (octylglucoside/cardiolipin) for long-chain ceramide was preferred for the substrate delivery, the former resulting in higher activities. Bacterially expressed HsCERK was highly dependent on Mg²⁺ ions, much less on Ca²⁺ ions. A clear preference for the D-erythro isomer was seen. Various N-acylated amino alcohols were no substrate, but N-hexanoyl-1-O-hexadecyl-2-desoxy-2-amino-sn-glycerol and N-tetradecanoyl-2S-amino-1-butanol were phosphorylated, suggesting that the secondary hydroxy group is not required for recognition. The properties of HsCERK, expressed in CHO cells, were similar to those of the bacterially expressed protein, including the Mg²⁺ dependence. In mouse, the highest activities were found in testis and cerebellum, and upon subcellular fractionation the activity was recovered mainly in the microsomal fraction. This fits with the plasma membrane localization in CHO cells, which was mediated by the N-terminal putative pleckstrin domain. No evidence for phosphorylation of ceramide by the recently described multiple lipid kinase was found. The latter kinase is localized in the mitochondria, but no firm conclusions with regard to its substrate could be drawn.

Supplementary key words N-acyl serinol • sphingosine • lipid kinase • ceramide phosphate • pleckstrin • testis • anandamide

Abbreviations: CERK, ceramide kinase; Cer1P, ceramide-1-phosphate; CL, cardiolipin; CMC, critical micellar concentration; DMS, N,N-dimethylsphingenine; EST, expressed sequence tag; EtOH, ethanol; GFP, green fluorescent protein; HsCERK, human ceramide kinase; LDAO, lauryldimethylammonium N-oxide; MuLK, multisubstrate lipid kinase; C₂-ceramide, N-acetyl-sphingenine; C₆-ceramide; N-hexanoyl-sphingenine; OG, octyl-ß-D-glucopyranoside; SeP, sphingenine-1-phosphate

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Sphingolipids, ubiquitous constituents of eukaryotic cells, are important signaling molecules with essential roles in cell proliferation, differentiation, and apoptosis (1). Ceramide, the initial product of the sphingomyelin cycle, functions as a second messenger in a variety of biological processes, such as cell differentiation and apoptosis (2). Ceramide can be hydrolyzed by ceramidase(s) to produce sphingenine. Ceramide and sphingenine can be phosphorylated to ceramide-1-phosphate (Cer1P) and sphingenine-1-phosphate (SeP), respectively. The phosphorylated lipids influence various biological processes and are recognized by specific phosphatases (3, 4). The responsible kinases, ceramide kinase (CERK) (5) and sphingosine kinase [type 1 (6–9) and type 2 (10)], belong to a family of kinases related to diacylglycerol kinases.

Cer1P was initially described as the product of a Ca²⁺-stimulated kinase that copurified with brain synaptic vesicles (11). A regulatory function in the secretion of neurotransmitters, by promoting membrane fusion, was suggested. Phagocytosis of IgG-coated erythrocytes by human neutrophils was found to be associated with an increase of Cer1P (12). A mitogenic, as well as an antiapoptotic, effect was also attributed to Cer1P (13, 14). By inhibition of acid sphingomyelinase, leading to an increase in ceramide level, Cer1P blocks both DNA fragmentation and caspase activation (15). Cer1P was also shown to mediate arachidonic acid release and prostanoid synthesis and to be a direct activator of cytosolic phospholipase A₂ (16). Recently, the stimulation of Ca²⁺ entry through voltage-operated calcium channels by Cer1P was demonstrated in rat pituitary cells (17).

Compared with the effects of Cer1P, less is known about CERK. Reports about the subcellular distribution of the kinase activity are contradictory (5, 11), and detailed information about CERK substrate specificity and stereoselectivity has not been reported. Given the recent report about another mammalian kinase, multisubstrate lipid kinase (MuLK), acting on ceramide and diacylglycerol as the main substrate (18), additional information on CERK, and MuLK, will be beneficial for understanding ceramide signaling. Here, we report on some properties of these enzymes, whose cDNAs were assembled during the cloning of human sphingosine kinases (19).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
Synthesis of lipids
Sphinganine, sphingenine stereoisomers, and N-octyl-sphingenine were obtained from Avanti%20Polar%20Lipids">Avanti Polar Lipids, Sigma, or Toronto Research Chemicals. Other lipids, amino alcohols, and amino acids (analytical grade) were from Sigma, Aldrich, or Acros Organics, except for N,N-dimethylsphingenine (DMS; from Alexis), 1-O-hexadecyl-2-desoxy-2-amino-sn-glycerol, 1-O- and 3-O-octadecyl-sn-glycerol (from Bachem), anandamide (Larodan) and CAY1044 (from Cayman). FTY720 was a gift from Novartis Pharma (Basel, Switzerland).

N-[1-¹⁴C]C₆-sphingenine was prepared as follows. Na-[1-¹⁴C]hexanoate (4.5 µmol; specific radioactivity, 55.4 µCi/µmol; ARC), diluted with cold Na-hexanoate (6.5 µmol), was dried overnight with P₂O₅ under vacuum. After adding 240 µl of tetrahydrofuran containing 11 µmol of 2,6-dichlorobenzoic acid, the solution was stirred for 2 h at room temperature followed by the addition of 660 µl of 146 mM triethylamine in dichloromethane containing sphingenine (16.5 µmol) and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (55 µmol). After a N₂ flush, the reaction mixture was stirred overnight at room temperature. Phase separation was induced by adding 4.84 ml of chloroform, 5.17 ml of methanol, and 4.95 ml of water, and the lower phase was washed with 5.5 ml of methanol and 10 mM NaOH (1:1, v/v). The lower phase was dried, and the ceramide produced was further purified by preparative TLC (silica G60; Merck) with chloroform-methanol-acetic acid (95:5:1, v/v) as solvent. The same method was used for the synthesis of N-[1-¹⁴C]C₂-sphingenine using [1-¹⁴C]acetate (sodium salt; specific radioactivity, 56.5 µCi/µmol; ARC).

Other amides were made by acylation of the aminoalkanols/diols with the appropriate acylchlorides, obtained by thionylchloride treatment of the fatty acid, adopted from Weis and Raizman (20). Briefly, the acylchloride (2 mmol) was dissolved in 1 ml of dimethylformamide, to which the aminoalkanol/diol (2 mmol), dissolved in 4 ml of dimethylformamide-pyridine (95:5, v/v), was added. After 2 h at room temperature, the mixture was quenched with 5 ml of 1 N HCl, and the amides were extracted twice into 5 ml of diethylether. To remove any formed O-acylation, the dried ether extract was dissolved in 2 ml of 40% methylamine-ethanol (EtOH; 1:1) and heated to 60°C for 1 h. Water (4 ml) and EtOH (4 ml) were added, and the amides were extracted into 2 x 5 ml of diethylether. After drying, the crude amides were dissolved in chloroform-diisopropylether (6:4, v/v), applied to silica-SepPAK (5 g; Waters), and eluted with chloroform-diisopropylether (6:4 or 9:1, v/v). Purity was analyzed by TLC in chloroform-methanol-acetic acid (90:10:1 and 60:40:1, v/v) and mass spectrometry analysis (Finnigan MAT TSQ 70; electrospray ionization; positive and negative mode). Structures were confirmed by daughter scan analysis.

N-Acylated amino acids were obtained by treating the amino acid (1 mmol, dissolved in 5 ml of 0.2 M NaHCO₃) with N-tetradecyl-hydroxysuccinimide (1 mmol, dissolved in 5 ml of tetrahydrofuran), adopted from Lapidot, Rappoport, and Wolman (21). After overnight reaction, 5 ml of 1 M HCl was added and amides were extracted into diethylether. The dried organic layer was dissolved in chloroform-methanol (2:1, v/v) and applied on a NH₂-BondElut cartridge (500 mg; Varian). After a wash with chloroform-methanol (2:1 and 1:9, v/v), amides were eluted with methanol-acetic acid (95:5). TLC analysis was done in chloroform-methanol-acetic acid (80:18:2, v/v).

Cloning and expression of lipid kinases
By BLAST homology searches (22), human and murine expressed sequence tags (ESTs) that code for proteins related to yeast sphingosine kinases (23) were identified and clustered into five different groups (19) (unpublished data). Related to the human ESTs, the inserts of IMAGE (24) clones 2392073 and 6185601 (UK-HGMP Resource Center, Hinxton, England), followed later by clone hk01650 (KIAA1646; Kazusa DNA Research Institute, Chiba, Japan), which all belonged to cluster 4 [lipid kinase 4 (LK4)], were sequenced (ALF DNA sequencer) with fluorescent plasmid primers and designed internal primers. Clone hk01650, despite its large insert (4,171 bp) (25), did not contain the putative start codon³; clone 2392073 missed both 5' and 3' information and is likely produced by erratic splicing, as it contains some intron sequence; clone 6185601 contained both the start and stop codons but also two deletions, 11 and 64 bp, the first resulting in a premature stop (at bases 449–451 of the open reading frame). The insert of IMAGE clone 649534, which belonged to cluster 2 [lipid kinase 2 (LK2)], was sequenced in a similar way and contained the full open reading frame (1,269 bases). The mouse counterpart was derived from IMAGE clone 2631716.

Expression vectors were made as described in Table 1, according to standard protocols (26). For PCR, Pfx polymerase (GibcoBRL) was used, whereas Top10F' Escherichia coli cells were used as host for recombinant work and for protein expression. For the expression of pETM-30-encoded proteins, E. coli BL21(DE3) was used.

For eukaryotic expression of human ceramide kinase (HsCERK), CHO cells were cultured on 100 mm plates in -Minimal Essential Medium (GibcoBRL) supplemented with 10% (v/v) fetal bovine serum (GibcoBRL), L-glutamine (Invitrogen), and antibiotics/antimycotics (Invitrogen). The CHO cells were transfected using the Lipofectamine Plus Reagent (Invitrogen) or the polyethylenimine procedure (27). The cells were harvested 24 h after transfection, dissolved in 0.5 ml of PBS containing a mix of protease inhibitors, sonicated on ice (3 x 10 s), frozen in liquid nitrogen, and stored at –80°C. The kinase activity, corrected for activity in cells transfected with the appropriate empty vector, with N-[¹⁴C]C₆-D-erythro-sphingenine bound to BSA, varied between 5.4 (Lipofectamine transfection) and 0.92 (polyethylenimine transfection) nmol/min/mg protein, being severalfold higher than the endogenous activity (60 ± 20 pmol/min/mg protein).

The other lipid kinase, HsLK2/MuLK, was expressed in bacteria transformed with pSG004 and in CHO cells transfected with pSG005 (or empty vector). For the latter, stable transfectants were selected by growing the cells in the presence of 400 µg/ml G418.

Antisera
His₆-GST-HsCERK_269–537 fusion protein was expressed in E. coli BL21 cells, transformed with pHVO008, and induced with 3 mM isopropylthio-ß-galactoside for 4 h at 37°C (1 liter of culture medium). The harvested bacteria (Kontron; 6,200 g for 10 min) were resuspended in lysis buffer (6 M guanidinium chloride, 100 mM NaH₂PO₄, 10 mM Tris, and10 mM imidazole, pH 8.0) and gently stirred for 1 h at 4°C. After removal of debris (Kontron; 7,750 g for 15 min), nickel-nitrilotriacetic acid agarose resin (Qiagen) was added to the supernatant, and the mixture was gently stirred for 30 min at 4°C. Beads were collected and washed with 8 M urea, 100 mM NaH₂PO₄, 10 mM Tris, and 10 mM imidazole, pH 7.0. Bound fusion proteins were eluted with elution buffer (8 M urea, 100 mM NaH₂PO₄, 10 mM Tris, and 250 mM imidazole, pH 7.0). Eluted proteins were mixed with Freund's complete adjuvant (or incomplete for boosters) and injected subcutaneously into rabbits. Ten days after the fifth booster, the rabbit was bled. For affinity purification, the serum was incubated with blot strips containing a His₆-tagged truncated HsCERK, obtained from Top10F' E. coli cells transformed with pHVO006 and induced with arabinose 0.2% (w/v), and purified via nickel-nitrilotriacetic acid agarose Sepharose and SDS-PAGE/blotting. Specific antibodies were eluted from the strips with Gentle Elute Buffer (Pierce).

Anti-HsLK2/MuLK was obtained by immunizing rabbits with the bacterially expressed and purified His₆-tagged protein. The generation of anti-human Pex14p (28), anti-green fluorescent protein (GFP) (29), and anti-myc (30) has been described.

Tissue distribution and subcellular studies
Tissues obtained from mouse (male C57/Bl6 N) or rat (male Wistar) were homogenized in 0.25 M sucrose, 5 mM Mops/NaOH, pH 7.2, and 0.1% (v/v) EtOH (HM) using Dounce (Potter-Elvehjem) homogenizers or a Polytron device (Kinematica). Subcellular fractions enriched in nuclei, mitochondria, lysosomes and peroxisomes, microsomal vesicles, and cytosol were prepared from testis or brain tissue homogenates, made with Potter-Elvehjem homogenizers, as described previously for rat liver (31). For the fractionation of brain tissue (cerebellum and medulla oblongata), myelin was removed (32) by resuspending the lysosome and peroxisome fraction (derived from 5 g starting wet weight) in 2 ml of 0.85 M sucrose, 5 mM Mops/NaOH, pH 7.2, and 0.1% (v/v) EtOH, followed by overlaying with 2 ml of homogenization medium. After centrifugation at 108,000 g (Beckman SW55) during 45 min, the supernatant, containing interfering myelin, was discarded. Marker enzyme and protein measurements were done as described previously for glutamate dehydrogenase (33), catalase (31), acid phosphatase (33), glucose-6-phosphatase (33), carboxylesterase (31), lactate dehydrogenase (33), proteins (31), and CERK.

For subcellular localization studies by immunofluorescence microscopy, CHO cells were grown on cover slips on 12-well plates and transfected with selected vectors. Monolayers were fixed at 24 h after transfection with 4% (w/v) paraformaldehyde, permeabilized with 1% Triton X-100 in PBS, blocked with 5% (w/v) BSA in PBS, and processed for immunostaining as described previously (34). Antibodies used were diluted in 1% (w/v) BSA in PBS. Primary antibodies were anti-FLAG M2 (Stratagene), anti-KDEL (Affinity BioReagents), anti-biotin (Rockland), anti-GFP, anti-HsCERK, anti-myc, and anti-Pex14p; secondary antibodies were anti-rabbit IgG Cy3, anti-mouse IgG FITC, and anti-rabbit AP (Sigma). To visualize mitochondria, cells were exposed to 2 µM Red Mitotracker (Red.CM-H₂Xros; Molecular Probes) for 45 min before fixation. Fluorescence was observed with a Leica DMR microscope equipped with FITC/RSGFP/Bodipy/Fluo3/DIO and Texas Red filters.

Kinase assays
To measure CERK, different assays, based on the recovery of labeled Cer1P in the organic phase when the assay mixture was subjected to a lipid extraction under acid conditions, were used. The substrate was presented bound to albumin (EtOH/BSA assay), mixed with CHAPS (detergent assay), or inserted in a mixed detergent/phospholipid micelle [octyl-ß-D-glucopyranoside/cardiolipin (OG/CL) assay].

For the standard EtOH/BSA assay, a 50 µl aliquot of bacterial or CHO lysates, or homogenate or subcellular fraction (appropriately diluted in HM), was mixed with 150 µl of reaction mixture. To prepare the reaction mixture, 20 nmol of substrate was dried at the bottom of a screw-capped glass tube, dissolved in EtOH (1% of final assay volume), and mixed with 4 volumes of BSA (resulting in a molar ceramide/BSA ratio of 2.5). Final concentrations were 100 µM substrate, 40 µM BSA, 5 mM ATP, 50 mM Mops/NaOH, pH 7.2, 3 mM MgCl₂, 40 mM NaF, 1 mM DTT, and 100 µM orthovanadate. Either labeled substrate {N-[1-¹⁴C]C₆-sphingenine (50,000 dpm/nmol), N-[1-¹⁴C]C₂-sphingenine (50,000 dpm/nmol), N-[1-¹⁴C]C₁₆-sphingenine (40,000 dpm/nmol)} or [-³²P]ATP (7,000 dpm/nmol) was used. Purity and specific activity of the ATP were determined by ultraviolet light absorbance and TLC on polyethyleneimine-cellulose plates (35). For the other assays, the dried ceramide was solubilized in 40 µl of 2% CHAPS (detergent assay) or 40 µl of 3.75% OG/2.5 mM CL (OG/CL assay), followed by sonication in a water bath and addition of the other solutions. For inhibition studies, the lipid substrate concentration was decreased to 25 µM and the inhibitor was tested at 100 µM, and both were dissolved together.

After an incubation of 15 min at 37°C, reactions were stopped by the addition of 1.5 ml of chloroform-methanol (1:2, v/v) and phase-separated by adding 1 ml of chloroform and 0.7 ml of a 0.5 N HCl/2 M NaCl solution. When using labeled substrate, a 1 ml aliquot of the lower phase was transferred to a glass tube and dried with nitrogen. The residue was dissolved in 60 µl of chloroform-methanol (4:1, v/v), immediately followed by spotting 20 µl on a Silica 60G glass plate with a concentration area (Merck). When using [-³²P]ATP, 250 µl of the lower phase was semiautomatically spotted. Plates were developed in chloroform-acetone-methanol-acetic acid-water (10:4:3:2:1, v/v). Spots of the phosphorylated products, visualized by autoradiography (Amersham Biosciences MP film), were scraped from the plates into 0.5 ml of 1% SDS and counted by liquid scintillation chromatography.

All measurements were corrected for the presence of endogenous kinase substrates, possessing after phosphorylation similar relative mobility values as (truncated) Cer1P (assays done plus/minus substrate) and/or endogenous kinase (transfections/transformations with empty vector), unless they proved to be negligible under the particular assay condition.

The phosphorylation of other lipids was determined in a similar manner, using one of the delivery modes described above. When sphingoid bases/amino alcohols were tested, alkaline phase separations were used as described previously for sphingosine kinase measurements (19, 36).

The activity of the putative lipid kinase was analyzed by similar procedures in the presence of MgCl₂ or CaCl₂. In addition, substrate was also delivered to the enzyme dissolved in DMSO (final concentration, 2%).

Analysis of cellular lipids
CHO cells were grown on six-well plates to 50% confluence in the medium described above and transfected with plasmid pHVO001 or the appropriate empty vector pCMV-Tag 2B using the Lipofectamine Plus Reagent (Invitrogen). The medium was replaced by phosphate-free Dulbecco's modified Eagle's medium (ICN), supplemented with 0.2% (v/v) Ultroser (GibcoBRL), Glutamax (GibcoBRL), and antibiotics/antimycotics (Invitrogen) 20 h after transfection. After 6 h, cells were given 2 ml of fresh phosphate-free medium containing 20 µCi of ³²P and C₆- or C₂-ceramides, added as EtOH/4% BSA (1:4, v/v) complexes (final concentration, 5 µM), or N-C₁₆-ceramide, dissolved in dodecane-methanol (37). After 15 h of incubation, medium was removed and monolayers were washed twice with PBS, scraped in 2 x 1 ml of methanol (38), and transferred to extraction tubes. Phase separation was introduced by the addition of 2 ml of chloroform and 1.8 ml of a 0.5 N HCl/2 M NaCl solution to the combined methanol solutions. An aliquot (250 µl) of the chloroform layer was spotted onto a silica plate and developed in chloroform-methanol-25% NH₄OH (60:35:8, v/v) (first dimension), allowed to dry, turned 90° and run in acetone, and subsequently in the same direction developed in chloroform-acetone-methanol-acetic acid-water (10:4:3:2:1, v/v). Radioactive spots were revealed by scanning (P-Imager; Molecular Dynamics) or by autoradiography.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

 
By homology with the yeast sphingosine kinases, up to five different sequences coding for related proteins were found in the human and mouse EST databases by different groups. Two cDNAs have been shown to encode sphingosine kinases: our cluster 1 (19) corresponded to type 1 sphingosine kinase (7–9), and our cluster 3 corresponded to type 2 (10). The proteins encoded by two other clustered ESTs were tentatively called lipid kinases (cluster 4, accession number AJ457828; cluster 2, accession numbers AJ278150 and AJ401619). The first one, as shown by Sugiura et al. (5), appeared to be CERK, a protein of 60 kDa (537 amino acids in human; 531 in mouse); the other one was recently claimed to be a MuLK (18), a protein of 47 kDa (422 amino acids in human; 421 in mouse). Both contain the conserved domains initially described for sphingosine kinases (6), and their relationship to sphingosine kinases has been discussed in the cited references. In the meantime, Tuson, Marfany, and Gonzalez-Duarte (39) found evidence for still another kinase when characterizing the RP26 locus on chromosome 2q31.2-q32.3, one of the loci mutated in autosomal recessive retinitis pigmentosa. Based on the similarity of the predicted gene product, the gene was named ceramide kinase-like gene (CERKL).

To understand some discrepancies between our preliminary data, using bacterially expressed HsCERK, and the data reported by Sugiura et al. (5), based on CERK expressed in HEK293 cells, we further investigated in depth the enzyme characteristics. Unless indicated otherwise, the data reported below were all obtained with fusion proteins containing the full-length protein. In addition, substantial efforts were made to define the substrate of the other kinase (LK2), which we will refer to as MuLK to minimize confusion.

Effect of chain length and substrate delivery on CERK activity
In a first approach, the length of the N-acyl chain, using albumin as a carrier for the ceramides, was varied. With His-tagged bacterially expressed CERK and in the presence of Mg²⁺ ions (see below), the highest activity was observed for N-hexanoyl-sphingenine (N-C₆-Se), somewhat less for N-acetyl-sphingenine (N-C₂-Se) (Fig. 1A). The natural substrate, C₁₆-ceramide, was poorly phosphorylated (Fig. 1A). With N-C₆-Se as substrate, almost normal Michaelis-Menten kinetics were observed for recombinant HsCERK (Fig. 2). A K_m value of 30 µM and a V_max of 35 nmol/min/mg were calculated. For N-C₂-Se, an apparent K_m of 22 µM was obtained, but substrate inhibition was seen at >100 µM. The same chain length dependence was seen when using CHO-expressed kinase, in contrast to Sugiura et al. (5), who reported low activity with N-C₂-Se. Because the data reported by these authors were obtained in a micellar assay, the effect of substrate dissolution and delivery on CERK activity was examined more closely. EtOH/BSA, detergents such as Triton X-100 and CHAPS, and, to a lesser extent, OG and mixed micelles (OG/CL) turned out to be useful for the dissolution/delivery of C₆-ceramide to the bacterially expressed kinase (Fig. 1A). High concentrations of OG and CHAPS [five times the critical micellar concentration (CMC)], however, almost completely inhibited CERK activity. Low concentrations of lauryldimethylammonium N-oxide (LDAO) and N-laurylsarcosine were also inhibitory. This inhibitory effect was not attributable to an ineffective dissolution of the ceramide (data not shown) (40); moreover, inhibition was also obvious when LDAO (1x CMC), OG (5x CMC), or CHAPS (5x CMC) was added to assays containing BSA-bound ceramide as substrate. With both EtOH/BSA and OG/CL as carriers, C₂-ceramide was less well phosphorylated than C₆-ceramide. Interestingly, C₁₆-ceramide was a better substrate than C₆-ceramide when OG/CL instead of EtOH/BSA was used for its delivery, and the activities toward C₁₆-ceramide or ceramide containing a 2-hydroxy fatty acid (ceramide IV) were comparable (data not shown).

View larger version (16K): [in this window] [in a new window]   Fig. 1. Effect of substrate delivery on ceramide kinase (CERK) activity. A: Phosphorylation of ceramide (100 µM) was determined in lysates from E. coli Top10F' cells expressing human ceramide kinase (HsCERK; pPVV072) using different solubilization methods and [-32P]ATP. The dried C6-ceramide (gray bars) was dissolved either in ethanol (EtOH) followed by BSA (EtOH/BSA), or with mixed octyl-ß-D-glucopyranoside/cardiolipin micelles (OG/CL), or with OG, CHAPS, or Triton X-100, all tested at one or five times their critical micellar concentration (CMC). With lauryldimethylammonium N-oxide and N-laurylsarcosine, both at 1x and 5x CMC, or with EtOH/BSA combined with detergent, phosphorylation of C6-ceramide was <1% of the highest activity (data not shown). Activities are expressed as nanomoles of phosphorylated product per milligram of lysate protein per minute. For C16-ceramide (white bar) and C2-ceramide (black bars), only OG/CL and EtOH/BSA were tested. B: Phosphorylation of ceramide (100 µM) was measured in lysates from CHO cells transfected with HsCERK (pHVO002) in the presence of [-32P]ATP, using the EtOH/BSA (black bars) or OG/CL (gray bars) assay as described in Materials and Methods. Activities are expressed as picomoles of phosphorylated product per milligram of lysate protein per minute. ANA, anandamide; Cx, N-Cx-sphingenine; DAG, 1,2-sn-dioleoylglycerol; 4-OH C6, N-hexanoyl-4-OH-sphinganine.

View larger version (15K): [in this window] [in a new window]   Fig. 2. Enzyme kinetics of recombinant HsCERK. The activity of recombinant (His)6-tagged HsCERK, expressed in Top10F' E. coli, was measured at increasing concentrations of C2-ceramide (open diamonds), C6-ceramide (closed squares), and C16-ceramide (open triangles) in the presence of [-32P]ATP and solubilized in EtOH/BSA (A) or OG/CL (B). Activities are expressed as nanomoles of phosphorylated product per milligram of protein per minute. Because of borderline activities, data for C16- and C2-ceramide are not shown in A and B, respectively.

Because the activities are substantially higher with truncated ceramides complexed to BSA than in the mixed micellar assay, the former assay was mainly used to study the CERK properties.

Role of divalent ions for CERK activity
In older reports concerning CERK measurements in tissue/cells, a stimulation by Ca²⁺ was reported (11, 41). Also, Sugiura et al. (5) reported a Ca²⁺ dependence for the recombinant CERK, reaching a maximum at 0.5 mM free Ca²⁺. In our preliminary trials, performed in the absence of Ca²⁺ but in the presence of Mg²⁺ (see above), high kinase activity was found. Therefore, the influence of Ca²⁺ and Mg²⁺ on HsCERK activity was determined with C₆-ceramide as substrate. In the absence of Ca²⁺, the activity of CERK, bacterially as well as eukaryotically expressed, was increased by Mg²⁺ in a concentration-dependent manner (Fig. 3A). In the absence of Mg²⁺, CERK was stimulated by Ca²⁺, although to a lesser extent than by Mg²⁺, and inhibited at >1 mM Ca²⁺ (0.04 mM free Ca²⁺) (Fig. 3B). The highest HsCERK activity was measured in the presence of Mg²⁺ and without Ca²⁺ for both forms of expressed kinase (Fig. 3C). Sugiura et al. (5) also described a stimulation by Mg²⁺, which was less than by Ca²⁺. The reason for this discrepancy with our data is not clear. Because the Mg²⁺ dependence is seen for both the EtOH/BSA and the OG/CL assay, it is not attributable to the presence of acidic phospholipids, known to bind Mg²⁺ as well as Ca²⁺ ions. It is also not caused by differences in the ATP concentrations used. At all ATP concentrations tested (0.5–20 mM), Mg²⁺ resulted in 10- to 20-fold higher activities than Ca²⁺ (data not shown). In the presence of 3 mM Mg²⁺, a K_m of 400 µM was obtained for ATP, substantially higher than the reported value of 32 µM using natural ceramides (5). Also, the pH profiles were analyzed with 3 mM Mg²⁺ with or without Ca²⁺ (1 or 3 mM). A broad pH optimum between 6.2 and 7.8 was seen, and the activity of the (bacterially expressed) kinase was always higher in the presence of Mg²⁺, using either the EtOH/BSA or the OG/CL assay (data not shown). As a consequence, the regulation of CERK activity by calcium levels might occur to a lesser extent than proposed (42).

View larger version (14K): [in this window] [in a new window]   Fig. 3. Effect of Mg2+ and Ca2+ on CERK activity. CERK activity was measured with 100 µM [14C]C6-ceramide as substrate and solubilized with OG/CL under standard conditions except for the presence of 0.2 mM EDTA. Activities are expressed as picomoles of phosphorylated C6-ceramide formed per milligram of lysate protein per minute. A, B: CERK activity in lysate from CHO cells transfected with HsCERK (pHVO002) was measured in the presence of increasing Mg2+ concentrations in the absence of Ca2+ (A) or in the presence of increasing Ca2+ concentrations in the absence of Mg2+ (B). Profiles similar to these, but at higher specific activities, were observed when using lysate from E. coli Top10F' cells expressing HsCERK (pPVV072) (data not shown). C: CERK activity in the absence or presence of Mg2+ (3 mM) and/or Ca2+ (1 mM) in a membrane fraction prepared from E. coli Top10F' cells expressing HsCERK (pPVV072) (open bars) or lysate from CHO cells transfected with HsCERK (pHVO002) (closed bars).

View larger version (37K): [in this window] [in a new window]   Fig. 4. Substrate specificity of CERK. Phosphorylation of the different ceramide stereoisomers (A) and ceramide analogs (B), each at 100 µM, was determined in lysate from E. coli Top10F' cells expressing HsCERK (pPVV072) in the presence of [-32P]ATP using the EtOH/BSA procedure. Activities are expressed as nanomoles of phosphorylated product per milligram of protein per minute. Unless specified otherwise, amino alcohols and amides are racemic, whereas sphingoid bases and derivatives possess the D-erythro configuration. N-tetradecanoyl-3-amino-1-propanol, N-tetradecanoyl-4-amino-1-butanol, N-tetradecanoyl-3-hydroxypyrrolidin, N-tetradecanoyl-R-serine, N-tetradecanoyl-S-serine, N-tetradecanoyl-2-benzylserine, N-tetradecanoyl-3-phenylserine, N-tetradecanoyl-3-amino-tyrosine, N-tetradecanoyl-1-amino-2-naphthol-4-sulfonic acid, N-tetradecanoyl-3-hydroxy-anthranilic acid, N-tetradecanoyl-4-aminophenol, N-tetradecanoyl-diethanolamine, and N-hexadecanoyl-ethanolamine were not phosphorylated (data not shown).

Only a few of the tested analogs were substrates for CERK (Fig. 4B). If the sphingoid base was shortened to the first three C-atoms (N-C₁₄-2-amino-1,3-propanediol), the analog was still a substrate. According to Bieberich, Kawagushi, and Yu (46), the N-C₁₆ analog of this compound (called N-acylated serinol) acts as a ceramide mimic. When added to cultured NG108-15 or F-11 cells, it causes ceramide increase and apoptosis. N-Acyl serinol is a moderate inhibitor of acidic ceramidase and can compete with ceramide for glucosyl transfer (46), but competition with the kinase, as shown here, could also contribute to the changes in ceramide. When reversing the amino and secondary hydroxy group (N-C₁₄-3-amino-1,2-propanediol), the amide is poorly phosphorylated. Interestingly, the hydroxy group at carbon 3 of the base is not necessary for recognition by CERK, because N-C₄-2-amino-1-hexadecanol, N-C₁₄-2-amino-1-butanol, and N-C₆-1-O-hexadecyl-2-desoxy-2-amino-sn-glycerol are good substrates. Indirectly, these data support the notion that the primary hydroxy group of ceramide is phosphorylated by the kinase. To document the importance of the amino group configuration, both R- and S-isomers of N-C₁₄-2-amino-1-butanol were synthesized and tested: the 2S(D) isomer is clearly a better substrate than the 2R(L) isomer (Fig. 4B). Mixing experiments suggest that the 2R compound, although a poor substrate, can have a stimulatory effect (data not shown). These data agree with the selectivity seen toward the D-erythro isomer of C₆-ceramide and explain the recognition of the L-threo isomers, both corresponding to the configuration of the S isomer at carbon 2. The recognition of N-C₆-1-O-hexadecyl-2-desoxy-2-amino-sn-glycerol is interesting (Fig. 4B). Previously, it was shown that the etherlipid analog, 1-O-hexadecyl-2-desoxy-2-amino-sn-glycerol, was efficiently used by human sphingosine kinase type 1 (19). Hence, it is striking that both this base and sphingosine fit into the catalytic site of sphingosine kinase, and after N-acylation, into that of CERK, despite some structural differences, underlining the evolutionary relationship between both kinases. On the other hand, the N-acylated form of the immunosuppressive drug FTY720, another sphingosine analog that can be phosphorylated by sphingosine kinase types 1 and 2 (47–49), was not recognized by CERK.

Finally, we were interested to learn whether anandamide (N-arachidonoyl-ethanolamine) was phosphorylated or not. Anandamide, an endocannabinoid, can be considered as structurally related to ceramide, but unlike the other analogs used here, it is naturally occurring (50). It was a substrate, albeit a very poor one, of CERK (Fig. 1B).

The selectivity for the natural D-erythro substrates and its substrate spectrum, as analyzed here in detail, accentuates the physiological role of this kinase, classifying it as a true CERK. It should be emphasized that short-chain ceramides such as N-C₂-Se, so often used as a water-soluble ceramide analog in cellular studies, and ceramide mimics such as N-acyl serinol can be efficiently phosphorylated by CERK. According to Sugiura et al. (5), N-C₂-sphingenine is a poor substrate; however, as shown here, the affinity is very much dependent on the way the substrate is delivered to the enzyme (see below).

Tissue distribution of HsCERK
Having established optimum assay conditions for CERK using recombinant proteins, its presence was measured in different mouse tissues. In all tissues, Cer1P was formed, suggesting a ubiquitous expression of CERK. The level of kinase activity varied substantially, however, being markedly high in testis and cerebellum, followed by pancreas and cerebrum, and negligible in skeletal muscle, using either EtOH/BSA (Fig. 5) or mixed micelles (data not shown) to deliver the substrate. This distribution, however, does not fully reflect the expression of CERK, which, based on Northern analysis, is high in mouse testis, heart, and spleen (5). The lack of correlation between mRNA, protein, and activity levels can be attributable to many factors, including the presence of an inhibitor or activities affecting the substrate or its product. In heart and lung homogenates, unless diluted appropriately, residual ATPase activity interfered with the kinase measurements, despite the use of NaF and orthovanadate (H. Van Overloop, L. Van den Bossche, and P. P. Van Veldhoven, unpublished data). When analyzing the kinase activity in testis, we noticed the formation of an additional ³²P-labeled spot that, based on relative mobility value, represents SeP, but this was only seen in BSA-based assays (data not shown). Presumably, ceramide bound to BSA is more accessible to hydrolysis (by ceramidases) than when inserted in a mixed micelle. If formed, sphingenine is efficiently phosphorylated in testis homogenates, a tissue in which we previously reported a high sphingosine kinase activity when using BSA-bound substrates (36). Unfortunately, the amounts of CERK in tissues are too low to be convincingly visualized by immunoblotting (data not shown).

View larger version (14K): [in this window] [in a new window]   Fig. 5. Tissue distribution of CERK activity in mouse. CERK activity was determined in mouse tissue homogenates using 100 µM C6-ceramide, solubilized with EtOH/BSA, and using 200 µg of protein per assay. Activities are expressed as picomoles of phosphorylated C6-ceramide per milligram of protein per minute. The values shown are means ± SEM of separate measurements performed on three male mice, 60–90 days of age. The same activity distribution pattern was seen with C2-ceramide, complexed to BSA, or C16-ceramide, solubilized with OG/CL, as substrate (data not shown).

The kinase described as MuLK was cloned some years ago in our laboratory and named putative lipid kinase 2 (accession numbers AJ278150 and AJ401619). Despite many attempts (52), both with bacterially and eukaryotically expressed proteins, no substrate, when corrected for background or endogenous kinase, could be demonstrated.⁴ The reason for the lack of activity is not clear; no PCR-introduced errors appear to be present in the plasmids used, and the cDNAs cloned by Waggoner et al. (18) were based on our database entries. It is possible that the tag, although at different ends in the bacterial and eukaryotic constructs, interferes with the kinase activity. Because of the lack of experimental detail on the way MuLK was constructed (18), we cannot explore this possibility. Another consideration is that the majority of our data were obtained with the human kinase, whereas those reported by Waggoner et al. (18) relate to the murine kinase, characterized by one additional amino acid (Q₂₇₁).

The expression of MuLK in human tissues, estimated by RT-PCR techniques by Waggoner et al. (18), is highest in pancreas, followed by brain, liver, and kidney. Although MuLK, expressed in CHO cells, was recognized by our antiserum, we failed to see signals of the correct size (50 kDa) upon immunoblotting of various rat or mouse tissues, indicating low expression. Given the tissue distribution of MuLK, if active, it is rather unlikely that it contributes to ceramide phosphorylation under our assay conditions. In addition, the fact that the same tissue distribution was seen when using EtOH/BSA or mixed micelles to deliver the substrate, either short- or long-chain ceramides (Fig. 5; data not shown), also suggests that the measured activity resides in a single enzyme. Moreover, the subcellular localization of MuLK is different from that of CERK (see below).

Subcellular localization of HsCERK and HsMuLK
The enzyme(s) catalyzing the phosphorylation of ceramide was (were) previously reported to be membrane-associated (11, 41). Upon overexpression of CERK in HEK293 cells, again a membrane-bound activity was found (5). Mitsutake et al. (42) recently showed that HsCERK was predominantly localized in the cytosol of RBL-2H3 cells but was present at the plasma membrane in a few cells, and Carre et al. (53) reported on a Golgi localization in Cos-1 and HUVEC cells and translocation of HsCERK to the plasma membrane upon osmotic swelling. In our hands, tagged forms of HsCERK (FLAG-HsCERK and EGFP-HsCERK fusions), upon expression in CHO cells, were mainly localized to the plasma membrane, based on colocalization with appropriate markers (Fig. 6A), whereas no evidence for an ER association was found (Fig. 6B).

View larger version (41K): [in this window] [in a new window]   Fig. 6. Subcellular localization of HsCERK. CHO cells were analyzed 24 h after Lipofectamine-mediated transfection with plasmids encoding a full-length FLAG-HsCERK fusion (pHVO001) (A, B, D1), green fluorescent protein (GFP)-HsCERK (pHVO002) (C), N-truncated FLAG-HsCERK116-537 (pHVO010) (D2), or appropriate control vectors (data not shown) by direct (C1) or indirect (A1, B1, B2, C2, D1, D2) immunofluorescence microscopy. FLAG fusions were stained with monoclonal mouse anti-FLAG M2 (Stratagene)/anti-mouse FITC (A1, B1, D1, D2); plasma membrane was visualized with the Image-iTTM LIVE plasma membrane labeling kit (Molecular Probes) (A2); endoplasmic reticulum was visualized with rabbit anti-KDEL (Affinity BioReagents)/anti-rabbit Cy3 (B2), and CERK was visualized with affinity-purified rabbit anti-CERK/anti-rabbit Cy3 (C2). Overlays are shown in A3, B3, and C3. In E, CHO cells transfected with pEGFP-C1 (lane 1) or pHVO002, coding for a GFP-HsCERK fusion (lane 2), were analyzed 24 h after transfection by immunoblotting using anti-CERK antibody or anti-GFP. Both antibodies immunodecorated a band of the expected size for the fusion protein (87.7 kDa). Presumably, anti-CERK also recognizes the endogenous CERK in CHO cells (left panel, arrowhead); the asterisk in the right panel indicates a protein cross-reacting with anti-GFP. Migration of protein standards, expressed in kDa, is indicated.

View larger version (49K): [in this window] [in a new window]   Fig. 7. C6-ceramide is phosphorylated by CHO cells overexpressing HsCERK. C6-ceramide was added to CHO cells without (A) or with (B) overexpression of HsCERK, uploaded with 32P, followed by lipid extraction after overnight incubation. Lipids were analyzed by two-dimensional TLC [first dimension, chloroform-methanol-NH4OH, 60:35:8 (v/v); second dimension, acetone wash followed by chloroform-acetone-methanol-acetic acid-water, 10:4:3:2:1 (v/v)], followed by detection by autoradiography. Phosphorylated C6-ceramide is indicated by a solid arrow. Phosphorylation of endogenous ceramides (asterisks) was also seen in CHO cells overexpressing HsCERK without the addition of ceramide (data not shown). The arrowheads indicate the position of sphingenine-1-phosphate, resulting from C6-ceramide breakdown by ceramidase(s) and subsequent action of endogenous sphingosine kinase. Similar findings were obtained when fortifying the cells with C2-ceramide (data not shown).

View larger version (63K): [in this window] [in a new window]   Fig. 8. Subcellular localization of human multisubstrate lipid kinase. CHO cells, stably transfected with pSG005 coding for HsLK2-Myc-His, were analyzed by indirect immunofluorescence microscopy, after permeabilization with 1% (A, B), 0.03% (w/v) (C–E), or 0.015% (w/v) (F–H) Triton X-100, using monoclonal mouse anti-myc as primary antibody and anti-mouse FITC as secondary antibody (A, C, F). Colocalization studies were performed using Mitotracker Red CM-H2xRos (Molecular Probes) (B) or polyclonal rabbit anti-biotin and anti-rabbit-Cy3 (D, G). In E and H, cells were stained with polyclonal rabbit anti-Pex14p and anti-rabbit Cy3. Control incubations with cells stably transfected with pcDNA/Myc-His showed that the staining with anti-myc was attributable to the expressed fusion protein.

View larger version (18K): [in this window] [in a new window]   Fig. 9. Subcellular distribution of ceramide phosphorylating activities in rat brain and testis. Fresh rat brain tissue (cerebellum + medulla oblongata) (left panels) and testis (right panels) homogenates were fractionated into a nuclear (N), a heavy mitochondrial (M), a light mitochondrial (L), a microsomal (P), and a cytosolic (S) fraction. In each fraction, protein, marker enzymes, and ceramide phosphorylation (100 µM C6-ceramide, solubilized with BSA/EtOH) were measured. Results are expressed as relative specific activities (RSA) versus cumulative percentage of total protein according to de Duve et al. (64). RSA is defined as the percentage of total activity present in a particular fraction divided by the corresponding percentage of total protein. Recoveries varied between 82% and 102%. CERK activity in rat brain and testis amounted to 27 and 26 pmol/min/mg protein, respectively. Other markers (not shown) included catalase, acid phosphatase, glucose-6-phosphatase, 5'-nucleotidase, and lactate dehydrogenase, with the highest relative specific activity in the L, L, P, P, and S fractions, respectively. Although glutamate dehydrogenase is generally considered a marker for mitochondria, a soluble activity has been reported in nonhepatic tissues (65), likely explaining the rather high amount in the S fraction of testis.

Note. Very recently, Bektas et al. (62) published their work on MuLK and claimed that this kinase is not at all active on ceramide, in agreement with our negative data, but is able to phosphorylate monoacylglycerols and diacylglycerols. During the reviewing process, a paper by Wijesinghe et al. (63) on the substrate specificity of CERK appeared in this journal. Using Triton X-100 micelles to deliver the substrate, similar conclusions about D-erythro specificity were reached. However, the reported requirements for a sphingosine backbone, at least 12 carbon atoms long, and the secondary hydroxy group for substrate recognition are not supported by our study.

	ACKNOWLEDGMENTS

 
This work was supported by grants from the Flemish Fonds voor Wetenschappelijk Onderzoek (G.0405.02) and from the Belgian Ministry of Federaal Wetenschapsbeleid Interuniversitaire Attractiepolen (IAP-P5/05). H.V.O. is an aspirant from the Flemish Fonds voor Wetenschappelijk Onderzoek. Technical support by S. Asselberghs, V. Brys, K. De Greef, and G. Van der Hoeven to various parts of this work is highly appreciated. The authors thank Dr. T. Nagase (Kazusa DNA Research Institute, Chiba, Japan) for access to clone hk01650, Dr. G. Stier (EMBL, Heidelberg, Germany) for the pETM-30 expression vector, and Dr. V. Brinkmann (Novartis Pharma, Basel, Switzerland) for the sample of FTY720.

Manuscript received July 25, 2005 and in revised form November 2, 2005.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION REFERENCES

1. Futerman, A. H., and Y. A. Hannun. 2004. The complex life of simple sphingolipids. EMBO Rep. 5: 777–782.[CrossRef][Medline]

2. Pettus, B. J., C. E. Chalfant, and Y. A. Hannun. 2002. Ceramide in apoptosis: an overview and current perspectives. Biochim. Biophys. Acta. 1585: 114–125.[Medline]

3. Spiegel, S., and S. Milstien. 2002. Sphingosine 1-phosphate, a key cell signaling molecule. J. Biol. Chem. 277: 25851–25854.[Free Full Text]

4. Gomez-Munoz, A. 2004. Ceramide-1-phosphate: a novel regulator of cell activation. FEBS Lett. 562: 5–10.[CrossRef][Medline]

5. Sugiura, M., K. Kono, H. Liu, T. Shimizugawa, H. Minekura, S. Spiegel, and T. Kohama. 2002. Ceramide kinase, a novel lipid kinase. J. Biol. Chem. 277: 23294–23300.[Abstract/Free Full Text]

6. Kohama, T., A. Olivera, L. Edsall, M. M. Nagiec, R. Dickson, and S. Spiegel. 1998. Molecular cloning and functional characterization of murine sphingosine kinase. J. Biol. Chem. 273: 23722–23728.[Abstract/Free Full Text]

7. Nava, V. E., E. Lacana, S. Poulton, H. Liu, M. Sugiura, K. Kono, S. Milstien, T. Kohama, and S. Spiegel. 2000. Functional characterization of human sphingosine kinase-1. FEBS Lett. 473: 81–84.[CrossRef][Medline]

8. Pitson, S. M., P. A. B. Moretti, J. R. Zebol, P. Xia, J. R. Gamble, M. A. Vadas, R. J. D'Andrea, and B. W. Wattenberg. 2000. Expression of a catalytically inactive sphingosine kinase mutant blocks agonist-induced sphingosine kinase activation. A dominant-negative sphingosine kinase. J. Biol. Chem. 275: 33945–33950.[Abstract/Free Full Text]

9. Melendez, A. J., E. Carlos-Dias, M. Gosink, J. M. Allen, and L. Takacs. 2000. Human sphingosine kinase: molecular cloning, functional characterization and tissue distribution. Gene. 251: 19–26.[CrossRef][Medline]

10. Liu, H., M. Sugiura, V. E. Nava, L. C. Edsall, K. Kono, S. Poulton, S. Milstien, T. Kohama, and S. Spiegel. 2000. Molecular cloning and functional characterization of a novel mammalian sphingosine kinase type 2 isoform. J. Biol. Chem. 275: 19513–19520.[Abstract/Free Full Text]

11. Bajjalieh, S. M., T. F. Martin, and E. Floor. 1989. Synaptic vesicle kinase. A calcium-stimulated lipid kinase that co-purifies with brain synaptic vesicles. J. Biol. Chem. 264: 14354–14360.[Abstract/Free Full Text]

12. Hinkovska-Galcheva, V. T., L. A. Boxer, P. J. Mansfield, D. Harsh, A. Blackwood, and J. A. Shayman. 1998. The formation of ceramide-1-phosphate during neutrophil phagocytosis and its role in liposome fusion. J. Biol. Chem. 273: 33203–33209.[Abstract/Free Full Text]

13. Gomez-Munoz, A., P. A. Duffy, A. Martin, L. O'Brien, H. S. Byun, R. Bittman, and D. N. Brindley. 1995. Short-chain ceramide-1-phosphates are novel stimulators of DNA synthesis and cell division: antagonism by cell-permeable ceramides. Mol. Pharmacol. 47: 833–839.[Abstract]

14. Gomez-Munoz, A., L. M. Frago, L. Alvarez, and I. Varela-Nieto. 1997. Stimulation of DNA synthesis by natural ceramide-1-phosphate. Biochem. J. 325: 435–440.[Medline]

15. Gomez-Munoz, A., J. Y. Kong, B. Salh, and U. P. Steinbrecher. 2004. Ceramide-1-phosphate blocks apoptosis through inhibition of acid sphingomyelinase in macrophages. J. Lipid Res. 45: 99–105.[Abstract/Free Full Text]

16. Pettus, B. J., A. Bielawska, S. Spiegel, P. Roddy, Y. A. Hannun, and C. E. Chalfant. 2003. Ceramide kinase mediates cytokine- and calcium ionophore-induced arachidonic acid release. J. Biol. Chem. 278: 38206–38213.[Abstract/Free Full Text]

17. Tornquist, K., T. Blom, R. Shariatmadari, and M. Pasternack. 2004. Ceramide-1-phosphate enhances calcium entry through voltage-operated calcium channels by a protein kinase C-dependent mechanism in GH4C1 rat pituitary cells. Biochem. J. 380: 661–668.[CrossRef][Medline]

18. Waggoner, D. W., L. B. Johnson, P. C. Mann, V. Morris, J. Guastella, and S. M. Bajjalieh. 2004. MuLK, a eukaryotic multi-substrate lipid kinase. J. Biol. Chem. 279: 38228–38235.[Abstract/Free Full Text]

19. Gijsbers, S., S. Asselberghs, P. Herdewijn, and P. P. Van Veldhoven. 2002. 1-O-Hexadecyl-2-desoxy-2-amino-sn-glycerol, a substrate for human sphingosine kinase. Biochim. Biophys. Acta. 158: 1–8.

20. Weis, B., and P. Raizman. 1958. II. Synthesis of long chain fatty acid amines of sphingosine and dihydrosphingosine. J. Am. Chem. Soc. 80: 4657–4658.

21. Lapidot, Y., S. Rappoport, and Y. Wolman. 1967. Use of esters of N-hydroxysuccinimide in the synthesis of N-acylamino acids. J. Lipid Res. 8: 142–145.[Abstract]

22. Altschul, S. F., T. L. Madden, A. A. Schaffer, J. Zhang, Z. Zhang, W. Miller, and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25: 3389–3402.[Abstract/Free Full Text]

23. Nagiec, M. M., M. Skrzypek, E. E. Nagiec, R. L. Lester, and R. C. Dickson. 1998. The LCB4 (YOR171c) and LCB5 (YLR260w) genes of Saccharomyces encode sphingoid long chain base kinases. J. Biol. Chem. 273: 19437–19442.[Abstract/Free Full Text]

24. Lennon, G. G., C. Auffray, M. Polymeropoulos, and M. B. Soares. 1996. The I.M.A.G.E. Consortium: an integrated molecular analysis of genomes and their expression. Genomics. 33: 151–152.[CrossRef][Medline]

25. Hirosawa, M., T. Nagase, Y. Murahashi, R. Kikuno, and O. Ohara. 2001. Identification of novel transcribed sequences on human chromosome 22 by expressed sequence tag mapping. DNA Res. 28: 1–9.

26. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

27. Boussif, O., F. Lezoualc'h, M. A. Zanta, M. D. Mergny, D. Scherman, B. Demeneix, and J. P. Behr. 1995. A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: polyethylenimine. Proc. Natl. Acad. Sci. USA. 92: 7297–7301.[Abstract/Free Full Text]

28. Amery, L., M. Fransen, K. De Nys, G. P. Mannaerts, and P. P. Van Veldhoven. 2000. Mitochondrial and peroxisomal targeting of 2-methylacyl-CoA racemase in humans. J. Lipid Res. 41: 1752–1759.[Abstract/Free Full Text]

29. Fransen, M., T. Wylin, C. Brees, G. P. Mannaerts, and P. P. Van Veldhoven. 2001. Human Pex19p binds peroxisomal integral membrane proteins at regions distinct from their sorting sequences. Mol. Cell. Biol. 21: 4413–4424.[Abstract/Free Full Text]

30. Ghys, K., M. Fransen, G. P. Mannaerts, and P. P. Van Veldhoven. 2002. Functional studies on human Pex7p: subcellular localization and interaction with proteins containing a peroxisome-targeting signal type 2 and other peroxins. Biochem. J. 365: 41–50.[CrossRef][Medline]

31. Van Veldhoven, P. P., and G. P. Mannaerts. 1991. Subcellular localization and membrane topology of sphingosine-1-phosphate lyase in rat liver. J. Biol. Chem. 266: 12502–12507.[Abstract/Free Full Text]

32. Kovacs, W. J., P. L. Faust, G. A. Keller