Title: Autotaxin Is Overexpressed in Glioblastoma Multiforme and Contributes to Cell Motility of Glioblastoma by Converting Lysophosphatidylcholine TO Lysophosphatidic Acid
Abstract: Autotaxin (ATX) is a multifunctional phosphodiesterase originally isolated from melanoma cells as a potent cell motility-stimulating factor. ATX is identical to lysophospholipase D, which produces a bioactive phospholipid, lysophosphatidic acid (LPA), from lysophosphatidylcholine (LPC). Although enhanced expression of ATX in various tumor tissues has been repeatedly demonstrated, and thus, ATX is implicated in progression of tumor, the precise role of ATX expressed by tumor cells was unclear. In this study, we found that ATX is highly expressed in glioblastoma multiforme (GBM), the most malignant glioma due to its high infiltration into the normal brain parenchyma, but not in tissues from other brain tumors. In addition, LPA1, an LPA receptor responsible for LPA-driven cell motility, is predominantly expressed in GBM. One of the glioblastomas that showed the highest ATX expression (SNB-78), as well as ATX-stable transfectants, showed LPA1-dependent cell migration in response to LPA in both Boyden chamber and wound healing assays. Interestingly these ATX-expressing cells also showed chemotactic response to LPC. In addition, knockdown of the ATX level using small interfering RNA technique in SNB-78 cells suppressed their migratory response to LPC. These results suggest that the autocrine production of LPA by cancer cell-derived ATX and exogenously supplied LPC contribute to the invasiveness of cancer cells and that LPA1, ATX, and LPC-producing enzymes are potential targets for cancer therapy, including GBM. Autotaxin (ATX) is a multifunctional phosphodiesterase originally isolated from melanoma cells as a potent cell motility-stimulating factor. ATX is identical to lysophospholipase D, which produces a bioactive phospholipid, lysophosphatidic acid (LPA), from lysophosphatidylcholine (LPC). Although enhanced expression of ATX in various tumor tissues has been repeatedly demonstrated, and thus, ATX is implicated in progression of tumor, the precise role of ATX expressed by tumor cells was unclear. In this study, we found that ATX is highly expressed in glioblastoma multiforme (GBM), the most malignant glioma due to its high infiltration into the normal brain parenchyma, but not in tissues from other brain tumors. In addition, LPA1, an LPA receptor responsible for LPA-driven cell motility, is predominantly expressed in GBM. One of the glioblastomas that showed the highest ATX expression (SNB-78), as well as ATX-stable transfectants, showed LPA1-dependent cell migration in response to LPA in both Boyden chamber and wound healing assays. Interestingly these ATX-expressing cells also showed chemotactic response to LPC. In addition, knockdown of the ATX level using small interfering RNA technique in SNB-78 cells suppressed their migratory response to LPC. These results suggest that the autocrine production of LPA by cancer cell-derived ATX and exogenously supplied LPC contribute to the invasiveness of cancer cells and that LPA1, ATX, and LPC-producing enzymes are potential targets for cancer therapy, including GBM. Autotaxin (ATX) 2The abbreviations used are: ATX, autotaxin; GBM, glioblastoma multiforme; LPA, lysophosphatidic acid; LPC, lysophosphatidylcholine; lysoPLD, lysophospholipase D; RT, reverse transcription; BBB, blood-brain barrier; NPP, nucleotide pyrophosphatases/phosphodiesterases; CNS, central nervous system; siRNA, small interfering RNA; GAPDH, glyceraldehyde-3-phosphate dehydrogenase. is a 125-kDa glycoprotein and a potent tumor cell motogen that was originally isolated from the conditioned medium of A2058 human melanoma cells as a cell motility-stimulating factor for melanoma cells (1Stracke M.L. Krutzsch H.C. Unsworth E.J. Arestad A. Cioce V. Schiffmann E. Liotta L.A. J. Biol. Chem. 1992; 267: 2524-2529Abstract Full Text PDF PubMed Google Scholar). ATX was subsequently identified as a member of a family of ecto/exoenzymes referred to as nucleotide pyrophosphatases/phosphodiesterases (NPPs) (2Goding J.W. Terkeltaub R. Maurice M. Deterre P. Sali A. Belli S.I. Immunol. Rev. 1998; 161: 11-26Crossref PubMed Scopus (143) Google Scholar, 3Stefan C. Gijsbers R. Stalmans W. Bollen M. Biochim. Biophys. Acta. 1999; 1450: 45-52Crossref PubMed Scopus (46) Google Scholar). The three cloned members of this family (PC-1/NPP1, ATX/NPP2, and B-10/NPP3) share a 47-55% amino acid sequence identity. PC-1/NPP1 and B-10/NPP3 hydrolyze 5′-phosphodiester bonds in nucleotides in vitro, whereas ATX/NPP2 shows only weak activity at hydrolyzing such bonds. ATX is synthesized as a type II membrane protein and is released from cells in a soluble form by an unknown mechanism (3Stefan C. Gijsbers R. Stalmans W. Bollen M. Biochim. Biophys. Acta. 1999; 1450: 45-52Crossref PubMed Scopus (46) Google Scholar, 4Stracke M.L. Clair T. Liotta L.A. Adv. Enzyme Regul. 1997; 37: 135-144Crossref PubMed Scopus (80) Google Scholar). Enhanced expression of ATX in Ras-transformed NIH3T3 cells greatly enhances their invasive, tumorigenic, and metastatic potentials (5Nam S.W. Clair T. Campo C.K. Lee H.Y. Liotta L.A. Stracke M.L. Oncogene. 2000; 19: 241-247Crossref PubMed Scopus (162) Google Scholar). In addition, enhanced expression of ATX has been repeatedly demonstrated in various malignant tumor tissues including non-small cell lung cancer (6Yang Y. Mou L.j. Liu N. Tsao M.S. Am. J. Respir. Cell Mol. Biol. 1999; 21: 216-222Crossref PubMed Scopus (110) Google Scholar), breast cancer (7Euer N. Schwirzke M. Evtimova V. Burtscher H. Jarsch M. Tarin D. Weidle U.H. Anticancer Res. 2002; 22: 733-740PubMed Google Scholar, 8Yang S.Y. Lee J. Park C.G. Kim S. Hong S. Chung H.C. Min S.K. Han J.W. Lee H.W. Lee H.Y. Clin. Exp. Metastasis. 2002; 19: 603-608Crossref PubMed Scopus (148) Google Scholar), renal cell cancer (9Stassar M.J. Devitt G. Brosius M. Rinnab L. Prang J. Schradin T. Simon J. Petersen S. Kopp S.A. Zoller M. Br. J. Cancer. 2001; 85: 1372-1382Crossref PubMed Scopus (141) Google Scholar), hepatocellular carcinoma (10Zhao Z. Xu S. Zhang G. Zhonghua Gan Zang Bing Za Zhi. 1999; 7: 140-141PubMed Google Scholar, 11Zhang G. Zhao Z. Xu S. Ni L. Wang X. Chin. Med. J. (Engl. Ed.). 1999; 112: 330-332PubMed Google Scholar), and thyroid cancer (12Kehlen A. Englert N. Seifert A. Klonisch T. Dralle H. Langner J. Hoang V.C. Int. J. Cancer. 2004; 109: 833-838Crossref PubMed Scopus (99) Google Scholar), suggesting that ATX confers the tumorigenic and metastatic potentials of cancer cells. However, there is no direct evidence to show such a hypothesis so far. The mechanism by which ATX exhibits its biological activity toward various cancer cells was unknown. An ATX point mutant that is deficient in 5′-nucleotide phosphodiesterase activity was found to abolish the cell motility-stimulating activity of ATX (13Lee H.Y. Clair T. Mulvaney P.T. Woodhouse E.C. Aznavoorian S. Liotta L.A. Stracke M.L. J. Biol. Chem. 1996; 271: 24408-24412Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar), indicating that the migratory response to ATX requires an intact catalytic site. Recently, we and others showed that ATX has lysophospholipase D (lysoPLD) activity, which catalyzes a reaction to produce a bioactive lysophospholipid, lysophosphatidic acid (LPA), from lysophosphatidylcholine (LPC) (14Umezu-Goto M. Kishi Y. Taira A. Hama K. Dohmae N. Takio K. Yamori T. Mills G.B. Inoue K. Aoki J. Arai H. J. Cell Biol. 2002; 158: 227-233Crossref PubMed Scopus (803) Google Scholar, 15Tokumura A. Majima E. Kariya Y. Tominaga K. Kogure K. Yasuda K. Fukuzawa K. J. Biol. Chem. 2002; 277: 39436-39442Abstract Full Text Full Text PDF PubMed Scopus (618) Google Scholar). ATX has a significantly lower Km for LPC than the Km for the classical nucleotide substrate. Because LPA has long been defined as a cell motility-stimulating factor for various cell types including glioblastomas (16Manning T.J. Parker J.C. Sontheimer H. Cell Motil. Cytoskeleton. 2000; 45: 185-199Crossref PubMed Scopus (100) Google Scholar, 17Sturm A. Sudermann T. Schulte K.M. Goebell H. Dignass A.U. Gastroenterology. 1999; 117: 368-377Abstract Full Text Full Text PDF PubMed Scopus (104) Google Scholar, 18Itoh K. Yoshioka K. Akedo H. Uehata M. Ishizaki T. Narumiya S. Nat. Med. 1999; 5: 221-225Crossref PubMed Scopus (560) Google Scholar), ATX has been suggested to regulate motility by producing LPA through the G-protein-coupled receptor. Indeed, recent studies have shown that ATX stimulates the cell motility of various cancer cells in vitro through one of the LPA receptors, LPA1 (19Van Leeuwen F. Olivo C. Grivell S. Giepmans B.N. Collard J.G. Moolenaar W.H. J. Biol. Chem. 2003; 278: 400-406Abstract Full Text Full Text PDF PubMed Scopus (155) Google Scholar, 20Shida D. Kitayama J. Yamaguchi H. Okaji Y. Tsuno N.H. Watanabe T. Takuwa Y. Nagawa H. Cancer Res. 2003; 63: 1706-1711PubMed Google Scholar, 21Hama K. Aoki J. Fukaya M. Kishi Y. Sakai T. Suzuki R. Ohta H. Yamori T. Watanabe M. Chun J. Arai H. J. Biol. Chem. 2004; 279: 17634-17639Abstract Full Text Full Text PDF PubMed Scopus (248) Google Scholar, 22Yamada T. Sato K. Komachi M. Malchinkhuu E. Tobo M. Kimura T. Kuwabara A. Yanagita Y. Ikeya T. Tanahashi Y. Ogawa T. Ohwada S. Morishita Y. Ohta H. Im D.S. Tamoto K. Tomura H. Okajima F. J. Biol. Chem. 2004; 279: 6595-6605Abstract Full Text Full Text PDF PubMed Scopus (162) Google Scholar). Taking account of the fact that elevated ATX expression has been detected in various tumors (6Yang Y. Mou L.j. Liu N. Tsao M.S. Am. J. Respir. Cell Mol. Biol. 1999; 21: 216-222Crossref PubMed Scopus (110) Google Scholar, 7Euer N. Schwirzke M. Evtimova V. Burtscher H. Jarsch M. Tarin D. Weidle U.H. Anticancer Res. 2002; 22: 733-740PubMed Google Scholar, 8Yang S.Y. Lee J. Park C.G. Kim S. Hong S. Chung H.C. Min S.K. Han J.W. Lee H.W. Lee H.Y. Clin. Exp. Metastasis. 2002; 19: 603-608Crossref PubMed Scopus (148) Google Scholar, 9Stassar M.J. Devitt G. Brosius M. Rinnab L. Prang J. Schradin T. Simon J. Petersen S. Kopp S.A. Zoller M. Br. J. Cancer. 2001; 85: 1372-1382Crossref PubMed Scopus (141) Google Scholar, 10Zhao Z. Xu S. Zhang G. Zhonghua Gan Zang Bing Za Zhi. 1999; 7: 140-141PubMed Google Scholar, 11Zhang G. Zhao Z. Xu S. Ni L. Wang X. Chin. Med. J. (Engl. Ed.). 1999; 112: 330-332PubMed Google Scholar, 12Kehlen A. Englert N. Seifert A. Klonisch T. Dralle H. Langner J. Hoang V.C. Int. J. Cancer. 2004; 109: 833-838Crossref PubMed Scopus (99) Google Scholar), it is possible that certain cancer cells utilize the ATX-LPC-LPA-LPA1 system for their motility. In these cells, a possible regulatory factor that remains to be characterized is LPC. LPC is always present in plasma. In human plasma, its concentration ranges from 100 to 300 μm. LPC is also detected in other biological fluids such as seminal fluids and cerebrospinal fluids and in tissues and various types of cells but at much lower concentrations than in plasma (23Tanaka M. Kishi Y. Takanezawa Y. Kakehi Y. Aoki J. Arai H. FEBS Lett. 2004; 571: 197-204Crossref PubMed Scopus (98) Google Scholar, 24Fang X. Gaudette D. Furui T. Mao M. Estrella V. Eder A. Pustilnik T. Sasagawa T. Lapushin R. Yu S. Jaffe R.B. Wiener J.R. Erickson J.R. Mills G.B. Ann. N. Y. Acad. Sci. 2000; 905: 188-208Crossref PubMed Scopus (218) Google Scholar). Glioblastoma multiforme (GBM) is a highly malignant brain tumor. Removal of the tumor mass transiently improves the condition of the patient, but the ability of GBM cells to infiltrate normal brain tissue invariably almost always leads to tumor recurrence. Thus, most patients experience recurrence within 1 year (25Fujimaki T. Matsutani M. Nakamura O. Asai A. Funada N. Koike M. Segawa H. Aritake K. Fukushima T. Houjo S. et al.Cancer. 1991; 67: 1629-1634Crossref PubMed Scopus (37) Google Scholar), and less than 20% of the patients survive more than 2 years (26Fujimaki T. Neurol. Med.-Chir. (Tokyo). 2000; 40: 1-106Crossref Scopus (1) Google Scholar). GBM cells (glioblastomas) are highly motile and invade the normal brain parenchyma diffusely (27Burger P.C. Dubois P.J. Schold S.J. Smith K.J. Odom G.L. Crafts D.C. Giangaspero F. J. Neurosurg. 1983; 58: 159-169Crossref PubMed Scopus (305) Google Scholar). Several factors responsible for their invasive phenotype have been reported, such as certain extracellular matrix proteins including laminin, fibronectin, and/or collagen can promote glioma cell migration (28Giese A. Rief M.D. Loo M.A. Berens M.E. Cancer Res. 1994; 54: 3897-3904PubMed Google Scholar, 29Hauglan H. Tysnes B. Tysnes O. Anticancer Res. 1997; 17: 1035-1043PubMed Google Scholar). Secreted matrix metalloproteinases remodel the extracellular matrix, creating pathways more conducive to migration through normal brain tissue (30Westermarck J. Khri V. FASEB J. 1999; 13: 781-792Crossref PubMed Scopus (1400) Google Scholar). Investigations of the factors that affect the motility of glioblastomas are of particular interest because an understanding of these factors is needed for valid GBM therapy. Because the blood-brain barrier (BBB) is disrupted in GBM tissue, some components in plasma might affect the cell motility of glioblastomas (31Wolff M. Boker D.-K. Clin. Neuropathol. 1999; 8: 72-78Google Scholar, 32Seitz R. Wechsler W. Acta Neuropathol. 1987; 73: 145-152Crossref PubMed Scopus (67) Google Scholar). In this study, we examined the expression of ATX and LPA1 in brain tumor tissues and various tumor cell lines and found that both ATX and LPA1 are predominantly expressed in glioblastomas and GBM tissue. Using glioblastomas as a model system, we evaluated the effects of ATX, LPA1, LPA, and LPC on the motility of the cells. Reagents—1-oleoyl-LPA (18:1) and 1-oleoyl-LPC (18:1) was purchased from Avanti Polar Lipids Inc. (Alabaster, AL). Ki16425 was kindly provided by Dr. Hideo Ohata (Kirin Brewery Co., Takasaki, Japan). Cell Lines—All human tumor cell lines were maintained in RPMI 1640 (Sigma) supplemented with 2 mm glutamine, 1× penicillin/streptomycin, and 5% (v/v) heat-inactivated fetal bovine serum as described previously (33Yamori T. Matsunaga A. Sato S. Yamazaki K. Komi A. Ishizu K. Mita I. Edatsugi H. Matsuba Y. Takezawa K. Nakanishi O. Kohno H. Nakajima Y. Komatsu H. Andoh T. Tsuruo T. Cancer Res. 1999; 59: 4042-4049PubMed Google Scholar). The cell lines used in this study were NCl-H23 (lung), NCl-H226 (lung), NCl-H522 (lung), NCl-H460 (lung), A549 (lung), DMS273 (lung), DMS114 (lung), HCC-2998 (colon), HT-29 (colon), WiDr (colon), HCT-15 (colon), DLD1 (colon), SW480 (colon), LOVO (colon), CaRI (rectum), WiDr (colon), CaCo2 (colon), Colo320 (colon), Colo201 (colon), HCT-116 (colon), KM12 (colon), HT1080 (colon), RXF-631L (renal), ACHN (renal), OVCAR3 (ovary), OVCAR4 (ovary), OVCAR5 (ovary), OVCAR8 (ovary), SKOV-3 (ovary), U251 (CNS), SF295 (CNS), SF539 (CNS), SF268 (CNS), SNB75 (CNS), SNB78 (CNS), MKN45 (stomach), MKN28 (stomach), St4 (stomach), MKN1 (stomach), MKN7 (Stomach), MKN74 (stomach), KatoIII (stomach), MKN28 (stomach), MKN45 (stomach), MKN74 (stomach), MDA-MB231 (breast), HBC4(breast), BSY1 (breast), MCF7 (breast), DU145 (prostate), PC3 (prostate), HPC5 (others), A2058 (melanoma), and HeLa (uterine cervix). Mouse 203G glioma cells were maintained in RPMI 1640 supplemented with 2 mm glutamine, 1× penicillin/streptomycin, and 10% (v/v) heat-inactivated fetal bovine serum. This cell line was kindly provided by Dr. Koji Adachi (Nippon Medical School Tokyo, Japan). Quantitative Real-time RT-PCR—From various cancer tissues and cancer cell lines, total RNA from cells was isolated using ISOGEN (Nippongene, Toyama, Japan) and reverse-transcribed using the SuperScript first-strand synthesis system for RT-PCR (Invitrogen). Oligonucleotide primers for PCR were designed using Primer Express Software (Applied Biosystems, Foster City, CA). The sequences of the oligonucleotides used in PCR were as follows: ATX (human), forward, 5′-GGGTGAAAGCTGGAACATTCTT-3′; ATX (human), reverse, 5′-GCCACCGCAATATGGAATTATAAG-3′; LPA1 (human), forward, 5′-AATCGGGATACCATGATGAGTCTT-3′; LPA1 (human), reverse, 5′-CCAGGAGTCCAGCAGATGATAAA-3′; LPA2 (human), forward, 5′-CGCTCAGCCTGGTCAAGACT-3′; LPA2 (human), reverse, 5′-TTGCAGGACTCACAGCCTAAAC-3′; LPA3 (human), forward, 5′-AGGACACCCATGAAGCTAATGAA-3′; LPA3 (human), reverse, 5′-GCCGTCGAGGAGCAGAAC-3′; LPA4 (human), forward, 5′-CCTAGTCCTCAGTGGCGGTATT-3′; LPA4 (human), reverse, 5′-CCTTCAAAGCAGGTGGTGGTT-3′; ATX (mouse), forward, 5′-GGAGAATCACACTGGGTAGATGATG-3′; ATX (mouse), reverse, 5′-ACGGAGGGCGGACAAAC-3′; LPA1 (mouse), forward, 5′-GAGGAATCGGGACACCATGAT-3′; LPA1 (mouse), reverse, 5′-ACATCCAGCAATAACAAGACCAATC-3′; LPA2 (mouse), forward, 5′-GACCACACTCAGCCTAGTCAAGAC-3′; LPA2 (mouse), reverse, 5′-CTTACAGTCCAGGCCATCCA-3′; LPA3 (mouse), forward, 5′-GCTCCCATGAAGCTAATGAAGACA-3′; LPA3 (mouse), reverse, 5′-AGGCCGTCCAGCAGCAGA-3′; LPA4 (mouse), forward, 5′-CAGTGCCTCCCTGTTTGTCTTC-3′; LPA4 (mouse), reverse, 5′-GAGAGGGCCAGGTTGGTGAT-3′; GAPDH (human/mouse), forward, 5′-GCCAAGGTCATCCATGACAACT-3′; GAPDH (human/mouse), reverse, 5′-GAGGGGCCATCCACAGTCTT. PCR reactions were performed using an ABI Prism 7000 sequence detection system (Applied Biosystems). The transcript number of human GAPDH was quantified, and each sample was normalized on the basis of GAPDH content. LysoPLD Assay—Samples were incubated with 2 mm LPC (14:0) in the presence of 100 mm Tris-HCl (pH 9.0), 500 mm NaCl, 5 mm MgCl2, and 0.05% Triton X-100 during indicated hours at 37 °C. The liberated choline was detected by an enzymatic photometric method as described (14Umezu-Goto M. Kishi Y. Taira A. Hama K. Dohmae N. Takio K. Yamori T. Mills G.B. Inoue K. Aoki J. Arai H. J. Cell Biol. 2002; 158: 227-233Crossref PubMed Scopus (803) Google Scholar). Briefly, the liberated choline was oxidized by choline oxidase, and the hydrogen peroxide generated was quantified using horseradish peroxidase and TOOS reagent (N-ethyl-N-(2-hydroxy-3-sulfoproryl)-3-methylanikine, Dojin, Tokyo, Japan). SDS-PAGE/Western Blotting—Both cells and cell media were used to test for ATX protein expression. Cells were homogenized in phosphate-buffered saline, and then the cell homogenates were separated into membrane and soluble fractions by centrifugation at 100,000 × g for 60 min. Protein samples were separated by SDS-PAGE and transferred to nitrocellulose membranes using the Bio-Rad protein transfer system. The membranes were blocked with 10 mm Tris-HCl (pH 7.5) containing 150 mm sodium chloride, 5% (w/v) skimmed milk, and 0.05% (v/v) Tween 20, incubated with anti-lysoPLD monoclonal antibody (3D1) (23Tanaka M. Kishi Y. Takanezawa Y. Kakehi Y. Aoki J. Arai H. FEBS Lett. 2004; 571: 197-204Crossref PubMed Scopus (98) Google Scholar), and then treated with anti-rat IgG-horseradish peroxidase. Proteins bound to the antibody were visualized with an enhanced chemiluminescence kit (ECL, Amersham Biosciences). Isolation of Stable ATX Transfectant—The plasmid vector pCAGGS (kindly provided by Dr. Junichi Miyazaki (34Niwa H. Yamamura K. Miyazaki J. Gene (Amst.). 1991; 108: 193-199Crossref PubMed Scopus (4618) Google Scholar)) was utilized to create rat ATX (ATX-t) tagged with the Myc epitope at the C terminus (pCAGGS-rATX-Myc). To establish stable transfectants, 203G glioma cells were transfected with pCAGGS-rATX-Myc using Lipofectamine Plus reagents (Invitrogen) as recommended by the manufacturer and were grown in RPMI 1640 containing 800 μg/ml G418 (Wako). Individual G418-resistant clones were isolated by limiting dilution and screened by immunocytochemistry using Myc antibody and by measuring the lysoPLD activities of the culture media. Recombinant ATX Preparations—Rat ATX was expressed and partially purified using baculovirus system as described previously (14Umezu-Goto M. Kishi Y. Taira A. Hama K. Dohmae N. Takio K. Yamori T. Mills G.B. Inoue K. Aoki J. Arai H. J. Cell Biol. 2002; 158: 227-233Crossref PubMed Scopus (803) Google Scholar). The purified ATX was dialyzed in phosphate-buffered saline and used for cell motility assays. Boyden Chamber Assay—Chemotaxis was assayed as described previously (21Hama K. Aoki J. Fukaya M. Kishi Y. Sakai T. Suzuki R. Ohta H. Yamori T. Watanabe M. Chun J. Arai H. J. Biol. Chem. 2004; 279: 17634-17639Abstract Full Text Full Text PDF PubMed Scopus (248) Google Scholar). In brief, polycarbonate filters with 8-μm pores (Neuro-Probe, Inc., Gaithersburg, MD) were coated with 0.001% fibronectin (Sigma). Cells (1 × 105 cells in 200 μl/well) were loaded into the upper chambers and incubated at 37 °C for 6 h to allow migration. Cell migration to the bottom of the filter was evaluated by measuring optical density at 590 nm. For Ki16425 treatment, cells were preincubated with 1 μm Ki16425 for 30 min. Wound Healing Assay—Cells were plated in cell culture plates (12-well) using cell growth media containing fetal bovine serum. After the cells had reached semiconfluence, fetal bovine serum was removed from the media and replaced with serum-free media. A plastic pipette tip was drawn across the center of the plate to produce a clean wound area 24 h after serum depletion. Medium was then replaced with serum-free medium containing different concentrations of 1-oleoyl-LPA, 1-oleoyl-LPC, Ki16425 (1 μm) and lysoPLD. After the cells were cultured for 12, 24, or 36 h, cell movement into the wound area was examined. The migration distances between the leading edge of the migrating cells and the edge of the wound were compared. Quantification of LPC—LPC concentration in cell culture supernatant was determined as described previously (35Kishimoto T. Soda Y. Matsuyama Y. Mizuno K. Clin. Biochem. 2002; 35: 411-416Crossref PubMed Scopus (83) Google Scholar). Briefly, cells were cultured in serum-free RPMI 1640 containing 0.1% fatty acid-free bovine serum albumin (Sigma) for 2 days. LPC was extracted from culture media using the Bligh and Dyer method (51Bligh E.G. Dyer W.J. Can. J. BioChem. Physiol. 1959; 37: 911-917Crossref PubMed Scopus (43133) Google Scholar) and resuspended in phosphate-buffered saline containing 0.1% bovine serum albumin. LPC concentration was determined by a recently developed enzymatic colorimetric method as described. Briefly, samples were treated with lysophospholipase, glycerophosphorylcholine, phosphodiesterase, and choline oxidase. The resulting hydrogen peroxide generated was quantified using horseradish peroxidase and TOOS reagent. RNA Interference—SNB-78 glioblastoma cells were transfected with siRNA oligonucleotide duplexes 1 day after confluence (day -1) with trans-it TKO (Takara, Kyoto, Japan) according to the manufacturer's instructions. Generally 20 nm siRNA was transfected with 0.5 μl of Lipofectamine per well of a 24-well plate with fresh media. Each experiment contained equivalent samples transfected with a nontargeting control siRNA pool and samples not treated with trans-it TKO. siRNA oligonucleotide duplexes for each gene of interest were purchased from WAKO (Osaka, Japan) as optimized single duplexes (ATX1, sense, 5′-gccguuggagucaauaucuGC-3′, antisense, 5′-agauauugacuccaacggcAA-3′ and ATX2, sense, 5′-gggagacugcuguaccaauTA-3′, antisense, 5′-auugguacagcagucucccCT-3′). Transfection efficiency was monitored using fluorescent (Cy3)-tagged oligonucleotides (Blockit, Invitrogen) transfected as described above and visualized with a mercury lamp fluorescent microscope. Expression of ATX and LPA Receptors in 50 Tumor Cell Lines—We examined the expression of ATX in 50 cultured human tumor cell lines derived from various tumors using the quantitative RT-PCR technique. We found that some cells expressed a significant amount of ATX at both the mRNA and the protein levels (Figs. 1 and 2). High ATX expression was detected in DMS273 (lung cancer), colo320 (colon cancer), SKOV3 (ovarian cancer), MKN1 (stomach cancer), and most of the brain cancer cells (SF295, SF539, SF268, SNB-75, and SNB-78). The expression was highest in SNB-78 cells. In good agreement with this observation, both ATX protein and lysophospholipase D activity were detected in the culture supernatants of these ATX-positive cells (Fig. 2). Most of the protein was detected in the culture cell supernatants, whereas a small amount was detected in cells (Fig. 2A). These results confirm that ATX is secreted from cells, although ATX is initially biosynthesized in cells as a type II membrane protein.FIGURE 2ATX protein and lysoPLD activity in glioblastoma cell lines. A, Western blot of cell membrane fractions (ppt), soluble fraction (sup), and cell culture medium using anti-human ATX monoclonal antibody 3D1. B, lysoPLD activity of culture media from glioblastoma and other cancer cell lines. LPC was used as the substrate.View Large Image Figure ViewerDownload Hi-res image Download (PPT) We also examined the expression of the LPA receptors (LPA1, LPA2, LPA3, and LPA4) in the 50 human tumor cell lines using quantitative RT-PCR. Although the expression pattern of the four LPA receptors does not necessarily reflect the tissue origin of the tumor cells, restricted LPA receptor expression patterns were obtained (Fig. 1). LPA2 was predominantly expressed in cells from colon, stomach, and breast cancers (36Kitayama J. Shida D. Sako A. Ishikawa M. Hama K. Aoki J. Arai H. Nagawa H. Breast Cancer Res. 2004; 6: R640-R646Crossref PubMed Scopus (92) Google Scholar, 37Shida D. Watanabe T. Aoki J. Hama K. Kitayama J. Sonoda H. Kishi Y. Yamaguchi H. Sasaki S. Sako A. Konishi T. Arai H. Nagawa H. Lab. Investig. 2004; 84: 1352-1362Crossref PubMed Scopus (121) Google Scholar). LPA3 expression was relatively low. However, LPA3 was expressed by certain ovarian and prostate cancer cell lines. Expression of LPA4 was fairly low. By contrast, LPA1 was dominant in brain tumor cells (Fig. 2). Expression of ATX and LPA Receptors in GBM—Because SF295, SF539, SF268, SNB-75, and SNB-78 are defined as glioblastomas (gliomas derived from GBM), we attempted to examine the expression of ATX and LPA receptors in tissues from various brain tumors. We found that expression of ATX was markedly high in GBM tissues (Fig. 3). Three of four GBM tissue samples showed extremely high ATX expression. ATX expression is apparently lower in tissues from other brain tumors. One exception is a patient of astrocytoma (case AS2 number 4) whose tissue showed high ATX expression. He experienced early recurrence after only 16 months, and the tumor progressed to GBM at recurrence (case GBM number 1). Among the four LPA receptors, LPA1 was dominantly expressed in most brain tumor tissues tested including GBM with low expression of LPA2, LPA3, and LPA4, which may reflect the expression pattern in normal brain tissues (38An S. Bleu T. Hallmark O.G. Goetzl E.J. J. Biol. Chem. 1998; 273: 7906-7910Abstract Full Text Full Text PDF PubMed Scopus (492) Google Scholar, 39Bandoh K. Aoki J. Hosono H. Kobayashi S. Kobayashi T. Murakami M.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 expression pattern of ATX and LPA receptors in GBM tissues indicates that ATX contributes to the invasive property of glioblastomas by producing LPA. LPC Stimulates Cell Motility of ATX-expressing Cells—To test the possibility that glioblastomas acquire their high invasiveness through autocrine production of LPA by ATX, we first examined the effect of enhanced ATX expression on cell motility. We used mouse glioma cell line 203G that expressed LPA1 (Fig. 4A) but not a detectable amount of ATX (Fig. 4B). 203G glioma cells that stably express ATX (203G-ATX) were established by transfecting ATX cDNA and by selecting neomycin-resistant clones. The established three lines expressed significant levels of ATX as judged by both lysoPLD activity (data not shown) and Western blotting (Fig. 4B). In addition, these cell lines showed similar expression pattern of LPA receptors to the parental 203G cells (data not shown). The effects of LPA on the motility of transfected cells and mock-transfected 203G cells in the Boyden chamber were similar (Fig. 4C). LPA had a similar effect on the motility of parental 203G cells (not shown). The effect of LPA on the motility of these cells was abolished by the LPA1 antagonist, Ki16425 (Fig. 4C). By contrast, these cells showed quite distinct responses to LPC. LPC significantly stimulated the migration of 203G-ATX cells in Boyden chamber assay (Fig. 4D). However, a similar response was not induced in mock-transfected 203G cells (Fig. 4D) or in parental 203G cells (not shown). In addition, the stimulatory effect of LPC in 203G-ATX cells was completely abolished by the addition of Ki16425 (Fig. 4D), showing that LPA mediates the LPC-stimulated cell migration of the cells through LPA1. We confirmed that platelet-derived growth factor induced similar migratory response in ATX-overexpressing, mock-transfected, and parental 203G cells (data not shown). We further examined the role of endogenously expressed ATX in the cell motility of ATX-expressing cells, which has not been previously demonstrated so far. For this experiment, we used SNB-78, which has the highest ATX expression among the 50 tumor cell lines (Fig. 1). As shown in Fig. 5A, SNB-78 cells, like other LPA1-positive cells, showed a migratory response to exogenous LPA in the Boyden chamber. LPC also stimulated the migration of SNB-78 (Fig. 5B). Ki16425 blocked not only the LPA-induced migratory response but also the LPC-induced migratory response (Fig. 5). This indicates that: 1) LPC is converted to LPA by the lysoPLD activity of endogenous ATX, 2) the LPA generated subsequently stimulates cell migration through LPA1, and 3) LPC behaves as a chemotactic factor toward ATX-expressing cells. We previously showed that LPC is released from cells and could be converted to LPA to induce cell migration by exogenously added ATX