Identification of Human Plasma Lysophospholipase D, a Lysophosphatidic Acid-producing Enzyme, as Autotaxin, a Multifunctional Phosphodiesterase
2002; Elsevier BV; Volume: 277; Issue: 42 Linguagem: Inglês
10.1074/jbc.m205623200
ISSN1083-351X
AutoresAkira Tokumura, Eiji Majima, Yuko Kariya, Kyoko Tominaga, Kentaro Kogure, Katsuhiko Yasuda, Kenji Fukuzawa,
Tópico(s)Protein Kinase Regulation and GTPase Signaling
ResumoWe purified human plasma lysophospholipase D that produces physiologically active lysophosphatidic acid and showed that it is a soluble form of autotaxin, an ecto-nucleotide pyrophosphatase/phosphodiesterase, originally found as a tumor cell motility-stimulating factor. Its lower K mvalue for a lysophosphatidylcholine than that for a synthetic substrate of nucleotide suggests that lysophosphatidylcholine is a more likely physiological substrate for autotaxin and that its predicted physiological and pathophysiological functions could be mediated by its activity to produce lysophosphate acid, an intercellular mediator. Recombinant autotaxin was found to have lysophospholipase D activity; its substrate specificity and metal ion requirement were the same as those of the purified plasma enzyme. The activity of lysophospholipase D for exogenous lysophosphatidylcholine in human serum was found to increase in normal pregnant women at the third trimester of pregnancy and to a higher extent in patients in threatened preterm delivery, suggesting its roles in induction of parturition. We purified human plasma lysophospholipase D that produces physiologically active lysophosphatidic acid and showed that it is a soluble form of autotaxin, an ecto-nucleotide pyrophosphatase/phosphodiesterase, originally found as a tumor cell motility-stimulating factor. Its lower K mvalue for a lysophosphatidylcholine than that for a synthetic substrate of nucleotide suggests that lysophosphatidylcholine is a more likely physiological substrate for autotaxin and that its predicted physiological and pathophysiological functions could be mediated by its activity to produce lysophosphate acid, an intercellular mediator. Recombinant autotaxin was found to have lysophospholipase D activity; its substrate specificity and metal ion requirement were the same as those of the purified plasma enzyme. The activity of lysophospholipase D for exogenous lysophosphatidylcholine in human serum was found to increase in normal pregnant women at the third trimester of pregnancy and to a higher extent in patients in threatened preterm delivery, suggesting its roles in induction of parturition. lysophosphatidic acid endothelial differentiation gene lysophospholipase D lysophosphatidylcholine autotaxin phosphodiesterase hexanoyl decanoyl lauroyl myristoyl palmitoyl stearoyl oleoyl linoleoyl α-linoleoyl phosphatidylcholine butyryl octanoyl arachidonoyl acetyl p-nitrophenyl thymidine-5′-monophosphate autotaxin from human teratocarcinoma cell line ATX from an A2058 human melanoma cell line Recent rapid advance in research on phospholipidic mediators has revealed new members that have diverse sets of signaling through their specific G-protein-coupled receptors (1Fukushima N. Ishii I. Contos J.J.A. Weiner J.A. Chun J. Annu. Rev. Pharmacol. Toxicol. 2001; 41: 507-534Crossref PubMed Scopus (325) Google Scholar, 2Goetzl E.J. An S. FASEB J. 1998; 12: 1589-1598Crossref PubMed Scopus (490) Google Scholar, 3Moolenaar W.H. Exp. Cell Res. 1999; 253: 230-238Crossref PubMed Scopus (371) Google Scholar). Particularly, lysophosphatidic acid (LPA,1radyl-sn-glycerol-3-phosphate) has received interest because of its identification as an active component in platelet-aggregating activity and vasoactivity in incubated animal sera and plasma (4Tokumura A. Prog. Lipid Res. 1995; 34: 151-184Crossref PubMed Scopus (166) Google Scholar). Later, its principle effects were found to be growth-related; that is, cellular proliferation, alterations in differentiation, and suppression of apoptosis (2Goetzl E.J. An S. FASEB J. 1998; 12: 1589-1598Crossref PubMed Scopus (490) Google Scholar, 3Moolenaar W.H. Exp. Cell Res. 1999; 253: 230-238Crossref PubMed Scopus (371) Google Scholar, 4Tokumura A. Prog. Lipid Res. 1995; 34: 151-184Crossref PubMed Scopus (166) Google Scholar, 5Tigyi G. Prostaglandins Other Lipid Mediators. 2001; 64: 47-62Crossref PubMed Scopus (92) Google Scholar). Other important effects are its cytoskeletal response, cell aggregation, contraction, adhesion, or chemotaxis (1Fukushima N. Ishii I. Contos J.J.A. Weiner J.A. Chun J. Annu. Rev. Pharmacol. Toxicol. 2001; 41: 507-534Crossref PubMed Scopus (325) Google Scholar, 2Goetzl E.J. An S. FASEB J. 1998; 12: 1589-1598Crossref PubMed Scopus (490) Google Scholar, 3Moolenaar W.H. Exp. Cell Res. 1999; 253: 230-238Crossref PubMed Scopus (371) Google Scholar, 4Tokumura A. Prog. Lipid Res. 1995; 34: 151-184Crossref PubMed Scopus (166) Google Scholar, 5Tigyi G. Prostaglandins Other Lipid Mediators. 2001; 64: 47-62Crossref PubMed Scopus (92) Google Scholar). Its diverse activities appear to be regulated by the relative distribution of its specific receptors in a family of endothelial differentiation genes (EDG), desensitization of the receptors, and the local concentration of LPA determined by balance between LPA-producing and -degrading enzyme activities (1Fukushima N. Ishii I. Contos J.J.A. Weiner J.A. Chun J. Annu. Rev. Pharmacol. Toxicol. 2001; 41: 507-534Crossref PubMed Scopus (325) Google Scholar, 2Goetzl E.J. An S. FASEB J. 1998; 12: 1589-1598Crossref PubMed Scopus (490) Google Scholar, 3Moolenaar W.H. Exp. Cell Res. 1999; 253: 230-238Crossref PubMed Scopus (371) Google Scholar, 4Tokumura A. Prog. Lipid Res. 1995; 34: 151-184Crossref PubMed Scopus (166) Google Scholar, 5Tigyi G. Prostaglandins Other Lipid Mediators. 2001; 64: 47-62Crossref PubMed Scopus (92) Google Scholar, 6Gaits F. Fourcade O., Le Balle F. Gueguen G. Gaigé B. Gassama-Diagne A. Fauvel J. Salles J. Mauco G. Simon M. Chap H. FEBS Lett. 1997; 410: 54-58Crossref PubMed Scopus (148) Google Scholar). Although LPA is produced intracellularly in phospholipid biosynthesis and turnover, this should be differentiated from its actual pool as a first messenger unless there are efficient mechanisms for its release into extracellular fluid or leaflets of plasma membranes (3Moolenaar W.H. Exp. Cell Res. 1999; 253: 230-238Crossref PubMed Scopus (371) Google Scholar, 4Tokumura A. Prog. Lipid Res. 1995; 34: 151-184Crossref PubMed Scopus (166) Google Scholar, 5Tigyi G. Prostaglandins Other Lipid Mediators. 2001; 64: 47-62Crossref PubMed Scopus (92) Google Scholar, 6Gaits F. Fourcade O., Le Balle F. Gueguen G. Gaigé B. Gassama-Diagne A. Fauvel J. Salles J. Mauco G. Simon M. Chap H. FEBS Lett. 1997; 410: 54-58Crossref PubMed Scopus (148) Google Scholar). In this sense, secretory or ecto-type LPA-producing enzymes can supply LPA directly to its receptors. The first study on extracellular production of LPA was our own, detecting lysophospholipase D (lysoPLD) activity that produces bioactive LPA from lysophosphatidylcholine (LPC) in rat plasma (7Tokumura A. Harada K. Fukuzawa K. Tsukatani H. Biochim. Biophys. Acta. 1986; 875: 31-38Crossref PubMed Scopus (154) Google Scholar). Its distinct substrate specificity and cation requirement from known intracellular lysoPLD activity suggested that it is a hitherto uncharacterized metalloenzyme (7Tokumura A. Harada K. Fukuzawa K. Tsukatani H. Biochim. Biophys. Acta. 1986; 875: 31-38Crossref PubMed Scopus (154) Google Scholar, 8Tokumura A. Miyake M. Yoshimoto O. Shimizu M. Fukuzawa K. Lipids. 1998; 33: 1009-1015Crossref PubMed Scopus (52) Google Scholar, 9Tokumura A. Nishioka Y. Yoshimoto O. Shinomiya J. Fukuzawa K. Biochim. Biophys. Acta. 1999; 1437: 235-245Crossref PubMed Scopus (32) Google Scholar). To understand its physiological and pathological significance, the identity of lysoPLD activity in blood circulation should be clarified. Therefore we purified and characterized lysoPLD in human plasma and obtained strong evidence that it is a soluble form of autotaxin (ATX), a member of the ecto-nucleotide pyrophosphatase/phosphodiesterase (PDE) family (10Stracke M.L. Clair T. Liotta L.A. Adv. Enzyme Regul. 1997; 37: 135-144Crossref PubMed Scopus (80) Google Scholar). LPC (1-hexanoyl (6:0), decanoyl (10:0), lauroyl (12:0), myristoyl (14:0), palmitoyl (16:0), stearoyl (18:0), oleoyl (18:1), linoleoyl (18:2), and α-linolenoyl (18:3)), LPC plasmalogen from beef heart (alkenyl-LPC), 1-O-hexadecyl-LPC (alkyl-LPC), phosphatidylcholine (PC) (1,2-dibutyryl (4:0/4:0), dihexanoyl (6:0/6:0), dioctanoyl (8:0/8:0), didecanoyl (10:0/10:0), dilauroyl (12:0/12:0), dimyristoyl (14:0/14:0) and diarachidonoyl (20:4/20:4)), and sn-glycero-3-phosphocholine were commercial products of Sigma and Funakoshi (Tokyo, Japan). 1–20:4-LPC was prepared by the hydrolysis of 20:4/20:4-PC with phospholipase A2 from bee venom, as described (11Satouchi K. Sakaguchi M. Shirakawa M. Hirano K. Tanaka T. Biochim. Biophys. Acta. 1994; 1214: 303-308Crossref PubMed Scopus (19) Google Scholar). For preparation of 2–20:4-LPC, 1-alkenyl-LPC from beef heart was reacted with arachidonic anhydride, and the resultant PC was subjected to mild acid hydrolysis, as described (11Satouchi K. Sakaguchi M. Shirakawa M. Hirano K. Tanaka T. Biochim. Biophys. Acta. 1994; 1214: 303-308Crossref PubMed Scopus (19) Google Scholar). 1–20:4–2-acetyl (2:0)-PC and 1–2:0–2-20:4-PC were prepared by reacting of 1–20:4-LPC and 2–20:4-LPC with acetic anhydride, respectively, in the presence of pyridine at 35 °C for 12 h. Human blood (120–170 ml) was withdrawn from the antecubital vein of healthy volunteers of both genders aging 21–26 and mixed with 0.15 volume of citrate-phosphate-dextrose solution (102 mm sodium citrate, 17 mm citrate, 129 mm dextrose, and 16 mm NaH2PO4). Plasma was prepared by centrifugation of the blood at 1,500 × g for 30 min, and the proteins in pooled human plasma (1150 ml) were fractionated between 30 and 60% saturation of ammonium sulfate. Fractions were subjected to six chromatographies as follows: 1) anion exchange chromatography on a HiPrep 16/10 Q XL column (16 × 100 mm) with an NaCl gradient (0–1 m) for 40 min in 20 mmTris-HCl (pH 8.0) at a flow rate of 2.5 ml/min, 2) hydroxyapatite chromatography on a Bio-Scale ceramic hydroxyapatite column (5 ml) with a phosphate gradient (10–250 mm) for 50 min in sodium phosphate (pH 6.8) at a flow rate of 1.5 ml/min, 3) heparin affinity chromatography on a HiTrap heparin column (5 ml) in 50 mmTris-HCl (pH 8.0) in a stepwise fashion with 0.15, 0.25, and 1m NaCl each for 5 min at a flow rate of 2 ml/min, 4) concanavalin A affinity chromatography on a Con A-Sepharose HR column (16 × 83 mm) with 0.5 mmethyl-α-d-mannopyranoside for 67 min in 20 mm Tris-HCl (pH 7.4) containing 0.5 m NaCl at a flow rate of 1.5 ml/min, 5) hydrophobic interaction chromatography on a TSKgel Phenyl-5PW (7.5 × 75 mm) with an ammonium sulfate gradient (1–0 m) for 50 min in 20 mm Tris-HCl (pH 8.0) at a flow rate of 1 ml/min, and 6) anion exchange chromatography on a TSKgel Bio-Assist Q (4.6 × 50 mm) in 20 mm Tris-HCl (pH 8.0) containing 20% glycerol in a linear gradient of NaCl concentration from 0 to 0.5 m for 50 min and then 0.5 to 1m for 5 min at a flow rate of 0.5 ml/min. Proteins in chromatographic fractions were separated by 7.5% SDS-PAGE. Fractions with lysoPLD activity in the final purification step by the second anion exchange chromatography were combined and subjected to SDS-PAGE on 7.5% gel. After staining the gel with SyproRuby, the 110-kDa band was cut and treated with trypsin. The tryptic digest was directly analyzed by nanoscale high performance liquid chromatography on a C18 column (0.1 × 50 mm) coupled to tandem mass spectrometer (Q-Tof2) equipped with a nanoelectrospray ionization source. Positive ion tandem mass spectra were measured. The amino acid sequence from the NH2 terminus of the purified enzyme was determined in a protein sequencer (Procise cLC). LysoPLD activity was estimated by determining choline liberated from exogenously added LPC, PC, or sn-glycero-3-phosphocholine. In the standard assay, diluted sample solutions (0.1 ml) were first incubated with 0.05 ml of the substrate dispersion in saline containing 0.05% fatty acid-free bovine serum albumin at 37 °C for various time intervals up to 24 h. Choline in 0.1-ml aliquots of the incubation mixture was oxidized to betaine with 1 units/ml choline oxidase (12Takayama M. Itoh S. Nagasaki T. Tanimizu I Clin. Chim Acta. 1977; 79: 93-98Crossref PubMed Scopus (1000) Google Scholar), and hydrogen peroxide concomitantly generated was determined by fluorimetry (13Zaitu K. Ohkura Y. Anal. Biochem. 1980; 109: 109-113Crossref PubMed Scopus (187) Google Scholar). In brief, the second assay mixture consisted of 0.1 ml of the first assay mixture, 2.6 ml of 0.1m Tris-HCl buffer (pH 8.5), 0.2 ml of 7.5 mm3-(4-hydroxyphenyl)propionic acid (Dojindo, Kumamoto, Japan), 0.1 ml of 2 units/ml horseradish peroxidase (Funakoshi), and 0.01 ml of 300 units/ml choline oxidase (Toyobo, Osaka). After incubation at 37 °C for 15 min, fluorescent intensity from the mixtures was measured at 404 nm with the excitation of 320 nm. Nucleotide PDE activity was measured with a synthetic substrate of nucleotide, p-nitrophenyl thymidine-5′-monophosphate (pNP-TMP, Sigma) essentially as described previously (14Ruzzel W.E. Method Enzymol. 1963; 6: 236-258Crossref Scopus (159) Google Scholar). In the standard assay, 0.1-ml aliquots of diluted sample solutions were incubated with 0.05 ml of the pNP-TMP solution (0.45 mm) at 37 °C for different time intervals. The assay solution was mixed with 1.0 ml of 0.1 n NaOH, and the increase in optical density was measured at 400 nm. Both lysoPLD and nucleotide PDE activities of conditioned medium of Chinese hamster ovary K1 cells transfected with a vector containing full-length cDNA of rat ATX corresponding to human ATX-t but not phosphodiesterase I/nucleotide pyrophosphatase brain-specific (15Narita M. Goji J. Nakamura H. Sano K. J. Biol. Chem. 1994; 269: 28235-28242Abstract Full Text PDF PubMed Google Scholar) or empty vector (pcDNA3) with the aid of LipofectAMINETM were measured by the methods described above. The activities of the recombinant ATX were calculated by subtracting the value for the medium of control cells from that for the medium of ATX-transfected cells. The conditioned media were provided by Dr. J. Aoki, the University of Tokyo, Japan, who constructed rat cDNA for ATX by reverse transcription-polymerase chain reaction using rat liver cDNA library as template DNA. Human cDNA for ATX-t amplified from human liver DNA library was kindly provided by Dr. T. Katada, the University of Tokyo, and was transfected in Chinese hamster ovary K1 cells as described for rat ATX. Both its lysoPLD and nucleotide PDE activities were measured as described above. Re-solubilized proteins from human plasma precipitated with between 30–60% ammonium sulfate were fractionated by sequential HPLC. A summary of results on purification of lysoPLD is presented in Table I. We obtained 14,478-fold purified lysoPLD with 6.1% recovery at the seventh purification step (Phenyl-5PW). By the second anion exchange chromatography, the lysoPLD was further highly purified with 1.2% yield; a major peak with lysoPLD activity was separated by a minor peak with the enzyme activity (Fig.1). The first major peak (fractions 47–51) showed a single band at 110 kDa stained with silver on SDS-PAGE under nonreducing conditions (Fig. 2). This is in agreement with the estimated 110 kDa for a protein peak with lysoPLD activity by its gel filtration chromatography on a Superdex 200 HR 10/30 column (20 mm Tris-HCl (pH 8.0) with 0.2m NaCl). The major peak fraction had the same relative activities for various LPCs with those of the minor peak fraction (data not shown). The peptides digested from the 110-kDa protein band with trypsin were separated by nanoscale reverse phase chromatography on a C18 column and directly analyzed by nanoelectrospray ionization tandem mass spectrometry. A data base search of tandem mass spectra with a Mascot Search Program revealed that purified lysoPLD shared four fragments of amino acid sequences with ATX from the human teratocarcinoma cell line, Ntera2D1 (ATX-t) (16Lee H.Y. Murata J. Clair T. Polymeropoulos M.H. Torres R. Manrow R.E. Liotta L.A. Stracke M.L. Biochem. Biophys. Res. Commun. 1996; 218: 714-719Crossref PubMed Scopus (95) Google Scholar), and human PD-Iα (17Kawagoe H. Soma O. Goji J. Nishimura N. Narita M. Inazawa J. Nakamura H. Sano K. Genomics. 1995; 30: 380-384Crossref PubMed Scopus (65) Google Scholar). The fitted peptides correspond to318YGPFGPEMTNPLR330,335IVGQLMDGLK344,440RIEDIHLLVER450, and854TYLHTYESEI863 of ATX-t (Fig.3). The amino acid sequence from Tyr318 to Arg330 was not included in ATX from an A2058 human melanoma cell line (ATX-m) (18Murata J. Lee H.Y. Clair T. Krutzsch H.C. Årestad A.A. Sobel M.E. Liotta L.A. Stracke M.L. J. Biol. Chem. 1994; 269: 30479-30484Abstract Full Text PDF PubMed Google Scholar) originally identified for a factor with tumor cell motility-stimulating activity at nanomolar concentrations in a pertussis toxin-sensitive mechanism (19Stracke M.L. Krutzsch H.C. Unsworh E.J. Årestad A.A. Cioce V. Schiffmann E. Liotta L.A. J. Biol. Chem. 1992; 267: 2524-2528Abstract Full Text PDF PubMed Google Scholar). cDNA for ATX-t (863 amino acids, Fig. 3) and PD-Iα (863 amino acids) were thought to be a spliced form of cDNA for ATX-m (915 amino acids) (16Lee H.Y. Murata J. Clair T. Polymeropoulos M.H. Torres R. Manrow R.E. Liotta L.A. Stracke M.L. Biochem. Biophys. Res. Commun. 1996; 218: 714-719Crossref PubMed Scopus (95) Google Scholar, 17Kawagoe H. Soma O. Goji J. Nishimura N. Narita M. Inazawa J. Nakamura H. Sano K. Genomics. 1995; 30: 380-384Crossref PubMed Scopus (65) Google Scholar); in the amino acid sequence of ATX-m, 52 amino acids were inserted between Glu324 and Met325 in that of ATX-t or PD-Iα (18Murata J. Lee H.Y. Clair T. Krutzsch H.C. Årestad A.A. Sobel M.E. Liotta L.A. Stracke M.L. J. Biol. Chem. 1994; 269: 30479-30484Abstract Full Text PDF PubMed Google Scholar). Thus, our present result showed that human plasma lysoPLD is a soluble form of human ATX-t or PD-Iα but not ATX-m. Consistent with this suggestion, the elution pattern in the second anion exchange chromatography of human plasma lysoPLD was found to be similar with that of nucleotide PDE activity (Fig. 1). The COOH terminus of purified lysoPLD was identified as the isoleucine residue, since the peptide from Thr854 to Ile863corresponds to the COOH-terminal region of ATX-t (Fig. 3). In addition, the NH2-terminal sequences of the 110-kDa band were determined as AEGWEEGPPTVLSD and RLHTKGSTEERHLLY, which were identical to the sequences of ATX-t from Ala36 to Asp49and from Arg585 to Tyr599, respectively (Fig.3). Because two additional protein bands of about 75 and 30 kDa were detected by SDS-PAGE of purified lysoPLD under reducing conditions (data not shown), the 110-kDa protein should be constructed by cross-linking via a disulfide bridge(s) of these two peptides, generated by multiple proteolytic cleavages of its precursor protein with a putative transmembrane domain. One of the proteolytic cleavage sites should be between Arg35 and Ala36, releasing the NH2-terminal sequence containing the transmembrane domain (Fig. 3). The site was closer to the NH2 terminus than that of the deduced cleavage site between Ser48 and Asp49 for ATX-m (18Murata J. Lee H.Y. Clair T. Krutzsch H.C. Årestad A.A. Sobel M.E. Liotta L.A. Stracke M.L. J. Biol. Chem. 1994; 269: 30479-30484Abstract Full Text PDF PubMed Google Scholar).Table IPurification of human plasma lysoPLDStepTotal proteinTotal activitySpecific activityYieldPurificationmgunitsaUnits are defined in nanomolars of choline formed/h at 37 °C.units/mg%-FoldPlasma64,90041,6000.641001.0Ammonium sulfate (30–60%)14,60022,3001.53532.4Anion exchange (1Fukushima N. Ishii I. Contos J.J.A. Weiner J.A. Chun J. Annu. Rev. Pharmacol. Toxicol. 2001; 41: 507-534Crossref PubMed Scopus (325) Google Scholar)1,70018,30010.84416.8Hydroxyopatite86.38,30096.220150Heparin10.16,15060814949Concanavalin A0.5033,4006,7588.210,559Phenyl-5PW0.2732,5309,2676.114,478Anion exchange (2Goetzl E.J. An S. FASEB J. 1998; 12: 1589-1598Crossref PubMed Scopus (490) Google Scholar)5171.2a Units are defined in nanomolars of choline formed/h at 37 °C. Open table in a new tab Figure 2Electrophoretic analysis of purified lysoPLD from human plasma. The proteins in fractions by anion exchange chromatography of human plasma lysoPLD on TSKgel Bio-Assist Q were stained with silver on 7.5% SDS-PAGE in the absence of dithiothreitol. The arrow shows 110-kDa protein bands with lysoPLD activity.View Large Image Figure ViewerDownload (PPT)Figure 3Fitted amino acid sequences of human plasma lysoPLD with those of ATX-t. The amino acid sequence of cloned human ATX-t is given in the one-letter code. Four peptides derived from purified human plasma lysoPLD, which was found to fit with those of ATX-t by liquid chromatography-mass spectrometry/mass spectrometry, aredouble-underlined. Two NH2-terminal sequences of the purified lysoPLD are underlined. An open triangle shows the putative amino acid at the NH2terminus of the soluble form of ATX-t (18Murata J. Lee H.Y. Clair T. Krutzsch H.C. Årestad A.A. Sobel M.E. Liotta L.A. Stracke M.L. J. Biol. Chem. 1994; 269: 30479-30484Abstract Full Text PDF PubMed Google Scholar).View Large Image Figure ViewerDownload (PPT) We examine the enzymic properties of purified lysoPLD from human plasma. The time courses of hydrolysis of 16:0-LPC and pNP-TMP by the purified lysoPLD are shown in Fig.4 A. The rate of hydrolysis of pNP-TMP by the purified lysoPLD was higher than that of 16:0-LPC. Both the K m and V max values of its lysoPLD activity to 16:0-LPC calculated from the Lineweaver-Burk plot were 0.26 ± 0.5 mm and 103 ± 22 nmol/h/ml (n = 3), lower than that of its PDE activity to pNP-TMP (5.5 ± 0.5 mm and 752 ± 236 nmol/h/ml,n = 3), respectively. Co2+ increased not only lysoPLD activity of the purified enzyme, as in rat plasma lysoPLD (8Tokumura A. Miyake M. Yoshimoto O. Shimizu M. Fukuzawa K. Lipids. 1998; 33: 1009-1015Crossref PubMed Scopus (52) Google Scholar), but also its nucleotide PDE activity. As shown in Fig.4 B, both ATP and pNP-TMP inhibited the lysoPLD activity in a concentration-dependent manner. This result indicates that purified plasma lysoPLD has nucleotide pyrophosphatase/PDE activity. To obtain direct evidence that human plasma lysoPLD is a soluble form of ATX-t, we measured lysoPLD activity in conditioned media from Chinese hamster ovary K1 cells transfected with a vector containing full-length cDNA of human ATX-t or rat ATX. The time courses of lysoPLD and nucleotide PDE activities of the soluble forms of recombinant ATX (Fig. 4, C andE) were similar to those of purified human plasma lysoPLD (Fig. 4 A), respectively. As in the case of the purified lysoPLD from human plasma, the K m andV max values of the lysoPLD activity of the recombinant human (0.090 ± 0.001 mm and 223 ± 2.6 nmol/h/ml, n = 3) and rat (0.11 ± 0.01 mm and 647 ± 133 nmol/h/ml, n = 3) ATX to 16:0-LPC was less than that of its PDE activity to pNP-TMP (human ATX 1.78 ± 0.07 mm and 2865 ± 62 nmol/h/ml, n = 3; rat ATX 3.63 ± 0.37 mm and 9696 ± 303 nmol/h/ml, n = 3). The addition of Co2+ to the assay solution resulted in an increase in both its lysoPLD and nucleotide PDE activities, like the case of the purified plasma lysoPLD. Both ATP and pNP-TMP inhibited the lysoPLD activity of the recombinant human and rat ATX in a concentration-dependent manner (Fig. 4, D andF). We next examined the substrate specificity of purified plasma lysoPLD for various LPCs. The lysoPLD hydrolyzed three subclasses of LPC (acyl > alkyl > alkenyl, Fig.5 A) but failed to hydrolyzesn-glycero-3-phosphocholine, indicating that hydrophobic interaction of amino acid residues in its active site with the fatty acyl moiety is necessary for activity. 1–14:0-LPC was the best of the saturated fatty acyl LPCs, and unsaturated species of C18-fatty acyl LPCs were better than the saturated ones. The enzyme hydrolyzed 2–20:4-LPC acetylated to minimize acyl migration during the assay for lysoPLD activity at a higher rate than that for the acetylated 1–20:4-LPC (Fig. 5 A). Thus, lysoPLD supplies LPAs that are preferable for the LPA3(EDG-7) LPA receptor more efficiently, since LPA3 responds to LPA with a 2-unsaturated acyl group more than LPAs, with a 1-saturated or unsaturated acyl group (20Bandoh K. Aoki J. Hosono H. Kobayashi S. Kobayashi T. Murakami-Murofushi K. Tsujimoto M. Arai H. Inoue K. J. Biol. Chem. 1999; 274: 27776-27785Abstract Full Text Full Text PDF PubMed Scopus (469) Google Scholar). The lysoPLD hydrolyzed PCs with two saturated medium chain acyl groups but not with two saturated short or long chain acyl groups (Fig. 5 A). Co2+ increased the activities for acyl LPCs and all related analogs except 18:2-LPC and shifted the optimal chain lengths of saturated acyl LPC and PC from 14 to 12 and from 10 and 8, respectively, and the optimal numbers ofcis-double bonds of unsaturated acyl LPA from 2 to 1. The Co2+-induced alteration in the substrate specificity of purified lysoPLD suggested that this metal ion was replaced with natively bound metal ions, playing an important role in two-step catalytic reactions of ATX (21Zimmermann H. Braun N. Prog. Brain Res. 1999; 120: 371-385Crossref PubMed Google Scholar, 22Bollen M. Gijisbers R. Ceulemans H. Stalmans W. Stefan C. Crit. Rev. Biochem. Mol. Biol. 2000; 35: 393-432Crossref PubMed Scopus (255) Google Scholar). The substrate specificity of purified human plasma lysoPLD for phospholipids coincided with that of the recombinant human ATX (data not shown), very similar to that of the recombinant ATX (Fig. 5 B). Taken altogether, the current results indicate that human plasma lysoPLD is a soluble form of human ATX-t. An early investigation showed that several fractions with nucleotide PDE activity were separated by electrophoresis of human serum, one of which had a much higher K m value (30 mm) for pNP-TMP than those suggested for others (less than 0.3 mm) and increased during pregnancy (23Lüthje L. Pickert S. Ogilvie A. Hornemann E. Siegfried W. Waldherr A. Domschke W. Clin. Chim. Acta. 1988; 177: 131-140Crossref PubMed Scopus (4) Google Scholar). In our preliminary work, serum lysoPLD activity measured with a radioactive 16:0-LPC was found to increase in humans in pregnancy (24Tokumura A. Yamano S. Aono T. Fukuzawa K. Ann. N. Y. Acad. Sci. 2000; 905: 347-350Crossref PubMed Scopus (30) Google Scholar). In this study with a more convenient lysoPLD assay, we confirmed our previous finding and found that the elevated activity decreased to near the level of nonpregnant women after delivery (Fig.6). Interestingly, its activity in serum from patients with threatened preterm delivery at the third trimester of pregnancy was significantly higher than that of normal pregnant women (Fig. 6). These results suggest that lysoPLD-mediated production of LPA plays physiological roles in parturition, since LPA injected intravenously into rats was reported to increase intrauterine pressure because of contraction of uterine smooth muscle in vivo(25Tokumura A. Fukuzawa K. Yamada S. Tsukatani H. Arch. Int. Pharmacodyn. 1980; 245: 74-83PubMed Google Scholar). In the current study, by liquid chromatography-mass spectrometry/mass spectrometry, we found that the amino acid sequences of trypsin-digested peptides from purified human plasma lysoPLD were fitted to those of human ATX-t and PD-Iα. However, we could not exclude the possibility that human plasma lysoPLD is an uncharacterized isoform of ATX because of limited numbers of its partial amino acid sequences, revealed by liquid chromatography/mass spectrometry/mass spectrometry. Judging by its apparent molecular mass (110 kDa), the plasma lysoPLD would be a soluble form of ATX-t or PD-Iα, although there is no information on its cell sources, mode of its processing within the cells, and mechanism of its secretion. The soluble form of ATX-t appears to be somewhat structurally different from the soluble form of ATX-m; the molecular mass of the soluble ATX-m was estimated to be 125 kDa and appears to be cleaved at a distinct site from that for ATX-t when processed, probably before its secretion (18Murata J. Lee H.Y. Clair T. Krutzsch H.C. Årestad A.A. Sobel M.E. Liotta L.A. Stracke M.L. J. Biol. Chem. 1994; 269: 30479-30484Abstract Full Text PDF PubMed Google Scholar, 19Stracke M.L. Krutzsch H.C. Unsworh E.J. Årestad A.A. Cioce V. Schiffmann E. Liotta L.A. J. Biol. Chem. 1992; 267: 2524-2528Abstract Full Text PDF PubMed Google Scholar). The enzymic properties of the plasma lysoPLD (time courses, kinetics, potentiation by Co2+, and substrate specificity) were found to be similar to those of recombinant human ATX-t and rat ATX examined in this study, whereas the amino acid sequences of human ATX-t and rat ATX are not identical. Thus, this minor structural inconsistency between human and rat ATX seems not to affect their enzymic property appreciably. Our results with the recombinant ATX strongly indicate that lysoPLD purified from human plasma is a soluble enzyme of ATX-t isoform. It has been postulated that the product by ATX of an unknown phosphorylated substrate besides nucleotides may stimulate tumor cell motility (10Stracke M.L. Clair T. Liotta L.A. Adv. Enzyme Regul. 1997; 37: 135-144Crossref PubMed Scopus (80) Google Scholar, 22Bollen M. Gijisbers R. Ceulemans H. Stalmans W. Stefan C. Crit. Rev. Biochem. Mol. Biol. 2000; 35: 393-432Crossref PubMed Scopus (255) Google Scholar). The present study has thrown insight to this issue; the tumor cell motility-stimulating effect of ATX may not be mediated by its postulated agonistic action or nucleotide-metabolizing effect (10Stracke M.L. Clair T. Liotta L.A. Adv. Enzyme Regul. 1997; 37: 135-144Crossref PubMed Scopus (80) Google Scholar, 19Stracke M.L. Krutzsch H.C. Unsworh E.J. Årestad A.A. Cioce V. Schiffmann E. Liotta L.A. J. Biol. Chem. 1992; 267: 2524-2528Abstract Full Text PDF PubMed Google Scholar, 22Bollen M. Gijisbers R. Ceulemans H. Stalmans W. Stefan C. Crit. Rev. Biochem. Mol. Biol. 2000; 35: 393-432Crossref PubMed Scopus (255) Google Scholar, 26Clair T. Lee H.Y. Liotta L.A. Stracke M.L. J. Biol. Chem. 1997; 272: 996-1001Abstract Full Text Full Text PDF PubMed Scopus (141) Google Scholar), but the pathophysiological substrate for ATX may be LPC, since LPA is known to induce tumor cell metastasis (27Imamura F. Horai T. Mukai M. Shinkai K. Sawada M. Akedo H. Biochem. Biophys. Res. Commun. 1993; 193: 497-503Crossref PubMed Scopus (176) Google Scholar, 28Stam T. Michiels F. van der Kammen R.A. Moolenaar W.H. Collard J.G. EMBO J. 1998; 17: 4066-4074Crossref PubMed Scopus (203) Google Scholar) through its specific receptors coupled to G proteins including pertussis toxin-sensitive Gi/o (1Fukushima N. Ishii I. Contos J.J.A. Weiner J.A. Chun J. Annu. Rev. Pharmacol. Toxicol. 2001; 41: 507-534Crossref PubMed Scopus (325) Google Scholar, 2Goetzl E.J. An S. FASEB J. 1998; 12: 1589-1598Crossref PubMed Scopus (490) Google Scholar, 3Moolenaar W.H. Exp. Cell Res. 1999; 253: 230-238Crossref PubMed Scopus (371) Google Scholar). By combining the conclusion of the identity of plasma lysoPLD with both the wide distribution of abundant levels of LPC (4Tokumura A. Prog. Lipid Res. 1995; 34: 151-184Crossref PubMed Scopus (166) Google Scholar) and relatively low levels of ATP in animal body fluids (22Bollen M. Gijisbers R. Ceulemans H. Stalmans W. Stefan C. Crit. Rev. Biochem. Mol. Biol. 2000; 35: 393-432Crossref PubMed Scopus (255) Google Scholar) (for example, human serum contains about 200 μm LPC (29Tokumura A. Miyake M. Nishioka Y. Yamano S. Aono T. Fukuzawa K. Biol. Reprod. 1999; 61: 195-199Crossref PubMed Scopus (119) Google Scholar) and about 1 μm ATP (30Park W. Masuda I. Cardenal-Escarcena A. Palmer D.W. McCarty D.J. J. Rheumatol. 1996; 23: 1233-1236PubMed Google Scholar)) we speculate that lysoPLD/ATX exerts its diverse physiological effects by producing bioactive LPA. Indeed, there is a close fit between the postulated physiological functions of ATX and LPA that have been separately investigated. ATX is implicated in cartilage matrix calcification (22Bollen M. Gijisbers R. Ceulemans H. Stalmans W. Stefan C. Crit. Rev. Biochem. Mol. Biol. 2000; 35: 393-432Crossref PubMed Scopus (255) Google Scholar, 31Goding J.W. J. Leukocyte Biol. 2000; 67: 285-311Crossref PubMed Scopus (91) Google Scholar, 32Bachner D. Ahrens M. Betat N. Schroder D. Gross G. Mech. Dev. 1999; 84: 121-125Crossref PubMed Scopus (71) Google Scholar), prenatal and postnatal development of the central nervous system (22Bollen M. Gijisbers R. Ceulemans H. Stalmans W. Stefan C. Crit. Rev. Biochem. Mol. Biol. 2000; 35: 393-432Crossref PubMed Scopus (255) Google Scholar, 31Goding J.W. J. Leukocyte Biol. 2000; 67: 285-311Crossref PubMed Scopus (91) Google Scholar, 32Bachner D. Ahrens M. Betat N. Schroder D. Gross G. Mech. Dev. 1999; 84: 121-125Crossref PubMed Scopus (71) Google Scholar, 33Fuss B. Baba H. Phan T. Tuohy V.K. Macklin W.B. J. Neurosci. 1997; 17: 9095-9103Crossref PubMed Google Scholar), and new blood formation (34Nam S.W. Clair T. Kim Y. McMarlin A. Schiffmann E. Liotta L.A. Stracke M.L. Cancer Res. 2001; 61: 6938-6944PubMed Google Scholar), although definite evidence is lacking on the fundamental molecular mechanism. LPA was shown to accelerate mouse embryonic neurogenesis and postnatal myelin maturation (1Fukushima N. Ishii I. Contos J.J.A. Weiner J.A. Chun J. Annu. Rev. Pharmacol. Toxicol. 2001; 41: 507-534Crossref PubMed Scopus (325) Google Scholar), proliferate primary osteoblasts (35Grey A. Banovic T. Naot D. Hill B. Callon K. Reid I. Cornish J. Endocrinology. 2001; 142: 1098-1106Crossref PubMed Scopus (50) Google Scholar), and induce proliferation of vascular endothelial cells (36Panetti T.S. Chen H. Misenheimer T.M. Getzler S.B. Mosher D.F. J. Lab. Clin. Med. 1997; 129: 208-216Abstract Full Text PDF PubMed Scopus (93) Google Scholar). In particular, their involvement in the postnatal development of the central nervous system should be addressed. The intermediate peak of mRNA expression of LPA1(EDG-2) LPA receptor around the time of active myelination in neonatal rat brain suggested its involvement in oligodendrocyte maturation and myelination (37Allard J. Barron S. Diaz J. Lubetski C. Zalc B. Schwartz J.C. Sokoloff P. Eur. J. Neurosci. 1998; 10: 1045-1054Crossref PubMed Scopus (81) Google Scholar). Similar spatiotemporal expression of mRNA for ATX was observed in developing rat brain after birth; the mRNA was expressed in oligodendrocytes and in cells of the choroid plexus, indicating its relevance of maintenance of myelin sheath and blood-brain barrier (33Fuss B. Baba H. Phan T. Tuohy V.K. Macklin W.B. J. Neurosci. 1997; 17: 9095-9103Crossref PubMed Google Scholar). Although lysoPLD exists in animal serum and plasma (7Tokumura A. Harada K. Fukuzawa K. Tsukatani H. Biochim. Biophys. Acta. 1986; 875: 31-38Crossref PubMed Scopus (154) Google Scholar, 8Tokumura A. Miyake M. Yoshimoto O. Shimizu M. Fukuzawa K. Lipids. 1998; 33: 1009-1015Crossref PubMed Scopus (52) Google Scholar, 9Tokumura A. Nishioka Y. Yoshimoto O. Shinomiya J. Fukuzawa K. Biochim. Biophys. Acta. 1999; 1437: 235-245Crossref PubMed Scopus (32) Google Scholar), human follicular fluid (29Tokumura A. Miyake M. Nishioka Y. Yamano S. Aono T. Fukuzawa K. Biol. Reprod. 1999; 61: 195-199Crossref PubMed Scopus (119) Google Scholar), hen egg white (38Nakane S. Tokumura A. Waku K. Sugiura T. Lipids. 2001; 36: 413-419Crossref PubMed Scopus (53) Google Scholar), human saliva, human amniotic fluid, and guinea pig peritoneal washings, 2A. Tokumura, K. Tominaga, S. Yamano, T. Dohi, and K. Fukuzawa, unpublished data. there is no evidence for its source. The physiological role for LPA generated by lysoPLD in the body fluids including blood plasma is still unclear except that LPA in human follicular fluid was shown to participate in the stimulation of oocyte maturation (39Hinokio K. Yamano S. Nakagawa K. Irahara M. Kamada M. Tokumura A. Aono T. Life Sci. 2002; 70: 759-767Crossref PubMed Scopus (38) Google Scholar). The physiological meaning of the elevation in the serum lysoPLD activity in human pregnancy (24Tokumura A. Yamano S. Aono T. Fukuzawa K. Ann. N. Y. Acad. Sci. 2000; 905: 347-350Crossref PubMed Scopus (30) Google Scholar) is particularly intriguing and could be clarified in near future, since its higher activity of patients in threatened preterm delivery compared with that of women in normal pregnancy found in this study suggested the involvement of LPA produced by circulating lysoPLD in human parturition. Previously, we found that serum lysoPLD activity increased in rabbits fed on a high cholesterol diet, which suggested its pathological role in the induction of atherosclerosis (40Tokumura A. Kanaya Y. Kitahara M. Miyake M. Yoshioka Y. Fukuzawa K. J. Lipid Res. 2002; 43: 307-316Abstract Full Text Full Text PDF PubMed Google Scholar), since LPA induces attachment of monocytes to vascular endothelial cells (40Tokumura A. Kanaya Y. Kitahara M. Miyake M. Yoshioka Y. Fukuzawa K. J. Lipid Res. 2002; 43: 307-316Abstract Full Text Full Text PDF PubMed Google Scholar) and stimulates proliferation (41Tokumura A. Iimori M. Nishioka Y. Kitahara M. Sakashita M. Tanaka S. Am. J. Physiol. 1994; 267: C204-C210Crossref PubMed Google Scholar) and dedifferentiation (42Hayashi K. Takahashi M. Nishida W. Yoshida K. Ohkawa Y. Kitabatake A. Aoki J. Arai H. Sobue K. Circ. Res. 2001; 89: 251-258Crossref PubMed Scopus (169) Google Scholar) of vascular smooth muscle cells. The pathophysiological significance of LPA generated by lysoPLD besides its promotion of tumor metastasis is a matter of debate. In this study, we provided evidence that human plasma lysoPLD is identical to ATX-t. Information accumulated by previous studies on the molecular biology of ATX is expected to trigger accelerated advances in the evaluation of the physiological and pathological significance of LPA production by lysoPLD in animal body fluids. We thank Dr. J. Aoki for kindly providing conditioned media of ATX-transfected and control cells, Dr. T. Katada for the generous gift of cDNA for human ATX-t, and Y. Maruyama and W. Hatsuyama for technical assistance.
Referência(s)