Constitutive and Lysophosphatidic Acid (LPA)-Induced LPA Production: Role of Phospholipase D and Phospholipase a 2

2482 Vol. 6, 2482–2491, June 2000 Clinical Cancer Research

Constitutive and Lysophosphatidic Acid (LPA)-induced LPA 1 Production: Role of Phospholipase D and Phospholipase A2

Astrid M. Eder, Takayo Sasagawa,2 Muling Mao, in ovarian cancer patients could identify novel targets for Junken Aoki, and Gordon B. Mills3 therapy. Department of Molecular Oncology, Division of Medicine, The University of Texas M. D. Anderson Cancer Center, Houston, Texas INTRODUCTION 77030 [A. M. E., T. S., M. M., G. B. M.], and Graduate School of In the United States, ovarian cancer is the fifth most com- Pharmaceutical Sciences, The University of Tokyo, Tokyo 113-0033, mon female malignancy and is the leading cause of death from Japan [J. A.] gynecological malignancies (1). In 1999, ϳ26,800 women will be newly diagnosed, and 14,500 will die from ovarian cancer ABSTRACT (1). The majority of patients are diagnosed with advanced epithelial ovarian cancer with widespread metastatic disease. The Ascitic fluid and plasma from ovarian cancer patients, dismal outcome for ovarian cancer results from an inability to but not from patients with nongynecological tumors, contain detect the tumor at an early curable stage. As 90% of stage IA elevated levels of the bioactive phospholipid lysophospha- and 70% of stage II tumors can be cured by current manage- tidic acid (LPA). We show that ovarian cancer cells consti- ment, ovarian cancer diagnosed at an early stage has a prognosis tutively produce increased amounts of LPA as compared similar to breast cancer. The most likely way to identify ovarian with normal ovarian epithelium, the precursor of ovarian cancer at an early, curable stage and to develop new, effective epithelial cancer, or breast cancer cells. In addition, LPA, therapies for advanced ovarian cancers is to improve our under- but not other growth factors, increases LPA production by standing of the processes leading to the initiation and progres- the OVCAR-3 ovarian cancer cell line but not by normal sion of this disease. ovarian epithelium or breast cancer cell lines. We show that Ascitic fluid from ovarian cancer patients, but not from phospholipase D activity contributes to both constitutive and patients with other cancers or with benign diseases such as LPA-induced LPA production by ovarian cancer cells. Con- hepatic disease, contains elevated levels of the phospholipid stitutive and LPA-induced LPA synthesis by ovarian cancer LPA4 (2–5). LPA levels are also significantly elevated in plasma cells is differentially regulated with respect to the require- from Ͼ90% of patients with ovarian cancer regardless of stage ment of specific phospholipase A2 (PLA ) subgroups. Group 2 (6). In contrast, LPA levels are not elevated in plasma of IB (pancreatic) secretory PLA plays a critical role in both 2 patients with breast cancer or leukemia or in healthy controls constitutive and LPA-induced LPA formation, whereas (6). LPA levels are also increased in patients with endometrial group IIA (synovial) secretory PLA contributes to LPA- 2 cancer and cervix cancer (6), multiple myeloma (7), and renal induced LPA production only. Calcium-dependent and/or dialysis (8), all of which can be clinically distinguished from -independent cytosolic PLA s are required for constitutive 2 ovarian cancer. This suggests that LPA in plasma might provide LPA synthesis but do not play a role in LPA-induced LPA a marker for diagnosis of ovarian cancer, establishing prognosis, formation. LPA increases the proliferation of ovarian cancer or monitoring response to therapy. Because LPA levels are also cells, decreases sensitivity to cisplatin, the most commonly elevated in the early stages of the disease (6), the plasma LPA used drug in ovarian cancer, decreases apoptosis and assay offers the possibility of earlier diagnosis of ovarian can- anoikis, increases protease production, and increases pro- cer, resulting in improved prognosis. duction of neovascularization mediators. Thus, an under- LPA displays a broad spectrum of biological activities standing of the source and regulation of LPA production (9–12). Its principle effects are growth related, such as induction of cellular proliferation and suppression of apoptosis, or involve the cytoskeleton or adhesive proteins contributing to aggregation, adhesion, contraction, secretion, and chemotaxis. Received 12/15/99; revised 2/22/00; accepted 2/23/00. LPA stimulates the growth (4, 13), prevents apoptosis (14) and The costs of publication of this article were defrayed in part by the anoikis (not presented), decreases sensitivity to chemotherapeu- payment of page charges. This article must therefore be hereby marked tic drugs (15), and increases invasiveness of ovarian cancer cells advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Supported by Grant PO1 CA64602 (to G. B. M.) and by a sponsored research grant from Atairgin Technologies, Irvine, California. 2 Current address: Department of Nutritional Science, Faculty of Health 4 The abbreviations used are: LPA, lysophosphatidic acid, 1-acyl-sn- and Welfare Science, Okayama Prefectural University, 111 Kuboki glycerol-3-phosphate; MAPK, mitogen-activated protein kinase; Edg, Soja, Okayama 719-1197, Japan. endothelial differentiation gene; PMA, phorbol 12-myristate 13-acetate; 3 To whom requests for reprints should be addressed, at Department of PLD, phospholipase D; PA, phosphatidic acid; PLA, phospholipase A;

Molecular Oncology, Division of Medicine, Box 92, University of sPLA2, secretory PLA2; cPLA2, cytosolic PLA2; iPLA2, calcium-inde- Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, pendent PLA2; EGF, epidermal growth factor; PDGF, platelet-derived Houston, TX 77030. Phone: (713) 792-7770; Fax: (713) 794-1807; growth factor; OOEPC, oleyloxyethylphosphocholine; AACOCF3, E-mail: [email protected]. arachidonyltrifluoromethyl ketone; LPC, lysophosphatidylcholine.

(12). These effects are associated with increased phosphoryla- bonds that confer structural rigidity. sPLA2s require millimolar tion of focal adhesion kinase, increased tyrosine phosphoryla- calcium concentrations for catalytic activity. cPLA2s are high tion of cellular proteins, increased intracellular calcium concen- molecular mass proteins (85 kDa) that contain calcium- and tration, and increased MAPK activity after treatment with LPA lipid-binding and pleckstrin homology domains that might con-

(13). In contrast, normal ovarian epithelial cells are resistant to fer regulation by phosphatidylinositol 4,5-bisphosphate. cPLA2s the effects of LPA (16),5, 6 suggesting that acquisition of LPA do not require calcium for their catalytic mechanism, but in responsiveness is associated with transformation. LPA acts on G response to elevated calcium levels found in stimulated cells, protein-coupled receptors encoded by the endothelial differen- they translocate to membranes or membrane vesicles where they tiation gene (Edg) subfamily (17). The LPA and sphingosine- encounter their phospholipid substrates. In addition, they are 1-phosphate receptor Edg1 (18–20) is expressed at high levels regulated by phosphorylation on serine residues by p38MAPK in normal and immortalized ovarian epithelial cells but at low family members. iPLA2s are also found in the cytosol and use a levels in most ovarian cancer cell lines (12). The LPA receptor similar catalytic mechanism as cPLA2s, but in contrast to Edg2 (21) is expressed by both normal ovarian epithelial cells cPLA2s, iPLA2s are not regulated by calcium. iPLA2s contain and ovarian cancer cell lines at varying levels (12, 22, 23). In ankyrin repeats that are involved in protein-protein interaction. contrast, the LPA receptors Edg4 (24) and Edg7 (25) are ex- Recently, a PLA2-independent pathway for LPA synthesis has pressed at relatively high levels in ovarian cancer cell lines but been described. In addition to generating PA, PLD directly only at very low levels in normal and immortalized ovarian generates LPA by hydrolysis of preexisting lysophosphatidyl- epithelial cells (12, 22). Binding of LPA to its receptor(s) choline (38). activates pertussis toxin-sensitive (Gi) and -insensitive (Gq and LPA has been demonstrated to activate PLD in a number of G12/13) pathways (10, 11), leading to the expression of growth systems (39–41) and is a potent activator of increases in cyto- factor-regulated genes that contain serum response elements. solic calcium and of MAPKs in ovarian cancer cells (2, 13). LPA is a normal constituent of serum (present at concen- Both increases in cytosolic calcium and MAPK activity activate trations ranging from 1 to 5 ␮M), where it is produced and cPLA2 (42). In view of the higher levels of LPA in the ascites released by activated platelets (26). LPA is also produced by and plasma of ovarian cancer patients and the ability of LPA to growth factor-stimulated fibroblasts (27), cytokine-stimulated activate the pathways mediating LPA production, we assessed leukocytes (11), PMA-activated ovarian cancer cells (28), and basal and LPA-induced LPA production by ovarian cancer cells. possibly by other cell types. Little is known, however, about We found that ovarian cancer cells, in contrast to normal ovarian LPA production in vivo and why LPA levels are elevated in epithelial cells or breast cancer cells, produce LPA either ovarian cancer patients. constitutively or in response to LPA. Both constitutive and LPA may be synthesized by cells either de novo from LPA-induced LPA production exhibited PLD-dependent and glucose through pathways of lipid metabolism in the endoplas- -independent components. Constitutive LPA production was mic reticulum or through liberation of precursor phospholipids primarily dependent on group IB (pancreatic) sPLA2 and on and subsequent enzymatic conversions in membrane mi- cPLA2 and/or iPLA2, whereas LPA-induced LPA production crovesicles (11, 29, 30). The latter pathway is considered the was dependent on both group IB (pancreatic) and group IIA principal source of production of free and secreted LPA. PLD (synovial) sPLA2, but not cPLA2 or iPLA2. first converts phosphatidylcholine to PA. Two distinct isoforms of PLD have been identified (31, 32). PLD1, but not PLD2, is activated by GTP-binding proteins and protein kinase C. Both MATERIALS AND METHODS isoforms use phosphatidylinositol 4,5-bisphosphate as cofactor. Reagents. LPA (oleoyl, 18:1), EGF, and PDGF were

During the subsequent step in LPA synthesis, PLA2 (or poten- purchased from Sigma Chemical Co. (St. Louis, MO). Fatty tially phospholipase A1) hydrolyzes the sn-2 (sn-1) ester bond of acid-free BSA was obtained from Boehringer Mannheim (Indi- PA to generate LPA. Various PLA2 enzymes displaying an anapolis, IN). Manoalide, OOEPC and AACOCF3 were ob- 32 exclusive or relative selectivity for PA have been characterized tained from Calbiochem (San Diego, CA). [ P]Pi (8810 Ci/ (30). The relative contribution of each PLA2 to LPA synthesis is mmol) was purchased from DuPont NEN (Boston, MA). not known. On the basis of nucleotide sequence comparisons, Cell Lines and Media. Cells were propagated in RPMI

PLA2s have been divided into 10 groups. On the basis of 1640 (Central Core Media Facility, University of Texas M. D. biological properties, PLA2s have been divided into three sub- Anderson Cancer Center) supplemented with 10% heat-inacti- groups: sPLA2, cPLA2, and iPLA2 (33–37). sPLA2s are low vated FCS (Sigma) and 1000 units/ml penicillin/streptomycin molecular mass proteins (ϳ14 kDa) with five to seven disulfide (Life Technologies, Inc., Grand Island, NY). The ovarian cancer cell lines OVCAR-3 and SK-OV-3 were obtained from the American Type Culture Collection (Rockville, MD). The ovarian cancer cell line HEY was kindly provided by Dr. Ron Buick 5 V. Estrella, T. Pustilnik, F. X. Claret, G. E. Gallick, G. B. Mills, and (University of Toronto, Toronto, Ontario, Canada). A2780.6.3 is J. R. Wiener. Lysophosphatidic acid induction of urokinase plasmino- a subclone of the ovarian cancer cell line A2780 (kindly pro- gen activator secretion requires activation of the p38MAPK pathway, vided by Dr. Thomas Hamilton, Fox Chase Cancer Center, submitted for publication. Philadelphia, PA) stably expressing Edg-2 (23). The breast 6 Y-L. Hu, E. Goetzl, G. B. Mills, N. Ferrara, and R. B. Jaffe. Induction of vascular endothelial growth factor expression by lysophosphatidic cancer cell lines MCF7, MDA-MB-231, and MDA-MB-468 acid in normal and neoplastic ovarian epithelial cells, submitted for were kindly provided by Dr. Janet Price (University of Texas publication. M. D. Anderson Cancer Center). Normal ovarian epithelial cells

(NOE35) were obtained in-house, and immortalized ovarian Table 1 Analysis of LPA production in unstimulated and LPA- epithelial cells (IOSE29, IOSE80) were kindly provided by Dr. stimulated ovarian and breast cancer cell linesa Nellie Auersperg (University of British Columbia, Vancouver, LPA produced British Columbia, Canada). (relative PhosphorImager units) In Vivo Labeling and Stimulation of Cells. Cells (0.2– Cell line unstimulated LPA-stimulated 0.8 ϫ 106) were plated in 60-mm dishes in complete medium. Ovarian cancer After 2 days, at 80% confluency, the cells were starved by OVCAR-3 199 885 removal of complete medium and addition of serum-free me- SK-OV-3 7550 4523 dium. Twenty-four h later, the cells were washed with phos- HEY 4519 1866 phate-free medium and incubated in phosphate-free medium for A2780.6.3 2057 1658 1 h. The cells were again washed with phosphate-free medium Breast cancer 32 MCF7 695 459 and incubated with 0.1 mCi [ P]Pi/ml in phosphate-free me- MDA-MB-231 1175 1147 dium. After 1 h, the labeling medium was removed, and the cells MDA-MB-468 646 859 were washed with serum-free medium and either treated with a 32P labeling and stimulation of cells, extraction of lipids, and inhibitor for 20 min or immediately stimulated with 25 ␮M LPA analysis of phospholipids by two-dimensional TLC were performed as for 2 h. Preliminary time course experiments had shown that described in “Materials and Methods.” The data shown are from one of maximum LPA production and release occurred after2hof at least two or three (OVCAR-3, SK-OV-3) independent experiments. LPA stimulation; therefore, this time point was used throughout. LPA was added to the cells in a solution of 1% fatty acid-free

BSA in PBS. We therefore routinely tested fatty acid-free BSA ϩ for the presence of trace amounts of LPA. gels and transferred to Hybond N membranes (Amersham, Lipid Extraction and Analysis of Phospholipids by Arlington Heights, IL). Edg-7 and 18S RNA probes were ra- TLC. In thrombin-activated platelets, 90% of newly generated diolabeled by random-prime labeling using the Redi-Prime la- LPA is released into the medium (26). Furthermore, preliminary beling kit (Amersham). Membranes were incubated with radio- ϫ experiments showed that LPA produced by ovarian cancer cells labeled probes in 50% formamide, 10 Denhardt’s solution, ϫ ␮ was not retained within the cells but released into the extracel- 0.1% SDS, 4 SSC, 10 mM EDTA, and 100 g/ml salmon- lular space. Therefore, cell supernatants were used as source for sperm single-strand DNA (Sigma) at 42°C for 18 h. The blots ϫ extraction of phospholipids. After stimulation, the cell superna- were washed at room temperature in 1 SSC, 0.1% SDS for 20 ϫ ϫ tant was removed and cleared by centrifugation at 14,000 ϫ g min three times and then at 50°C in 0.1 SSC, 0.1 SDS for 20 Ϫ for 5 min. Acetic acid was added to the samples to a final min three times prior to autoradiography at 80°C for 1–2 days concentration of 20 mM. The samples were then extracted with or analysis with PhosphorImager. Quality and comparable load- 1-butanol and centrifuged. The 1-butanol phase was removed, ing of RNA were confirmed by rehybridization of the mem- and the aqueous phase was again extracted. The 1-butanol branes with radiolabeled 18S RNA. phases were combined and washed twice with 1-butanol-satu- rated water. The extracted lipids contained in the 1-butanol RESULTS phases were dried, dissolved in chloroform:methanol (1:1), and Analysis of LPA Synthesis by Ovarian and Breast Can- loaded onto TLC plates (precoated silica gel 60 plates; EM cer Cell Lines. The bioactive phospholipid LPA is present at Separations Technology, Gibbstown, NJ). Phospholipids were elevated levels in the ascites and plasma of patients with ovarian separated by two-dimensional TLC with the first buffer system cancer (2–6) and has been shown to exhibit pleiomorphic ac- containing chloroform:methanol:ammonium hydroxide (13:7: tivities on ovarian cancer cells (12). Levels of LPA are higher in 1.1) and the subsequent buffer system containing chloroform: ascites (up to 80 ␮M) than in plasma (up to 10 ␮M) from ovarian methanol:88% formic acid:water (11:5.6:1:0.2). Phospholipids cancer patients, suggesting that LPA is produced in the perito- were detected by autoradiography and identified by comigration neal cavity and then migrates to the peripheral circulation. with nonradioactive marker lipids. Quantitation of LPA-contain- Indeed, LPA levels averaged 4-fold higher in matched ascites as ing spots was performed by PhosphorImager. PhosphorImager compared with plasma samples from ovarian cancer patients. In units were normalized with respect to the total amount of each case (n ϭ 10), ascites LPA levels were higher than plasma 32P-labeled phospholipids, which minimizes variability in cell LPA levels (not presented). Furthermore, ovarian cancer cells, numbers or in 32P labeling. Each experiment was performed at but not breast cancer cells, produce LPA in response to the least twice, and the repeat experiment(s) yielded similar results. tumor-promoting agent PMA, suggesting that ovarian cancer In lipid extracts from SK-OV-3 cells, we routinely saw a second cells may be the source of LPA in ascites and plasma of ovarian minor spot running slightly further than the major LPA spot in cancer patients (23). We thus asked whether LPA, at concen- the second dimension, which may represent alkenyl-LPA and trations found in the ascites of ovarian cancer patients, could was included in the LPA analysis. induce ovarian cancer cells to produce LPA. We incubated Total RNA Preparation and Northern Blot Analysis. ovarian cancer cell lines with or without 25 ␮M LPA for 2 h Total cellular RNA was isolated from normal and immortalized (optimal time and concentration for LPA production as assessed ovarian epithelial cells and various ovarian cancer cell lines in preliminary experiments) and determined the levels of LPA using a RNeasy Mini kit (Qiagen, Valencia, CA) according to present in the medium. One of four ovarian cancer cell lines the manufacturer’s instructions. Equal amounts of total RNA tested produced LPA in response to treatment with LPA were separated by electrophoresis on denaturing 1% agarose (OVCAR-3; see Table 1). The other three ovarian cancer cell

Fig. 1 Two-dimensional TLC analysis of constitutive LPA production by SK-OV-3 (upper panel) and MDA-MB-231 cells (lower panel). Cells were 32 labeled with [ P]Pi and incubated for2hinserum-free medium before the supernatant was removed. Lipids were extracted and analyzed by two-dimensional TLC. Newly synthesized LPA was identified by comigration with nonradioactive LPA. ori, origin of the sample on the TLC plate. Relative PhosphorImager units for LPA from this experiment are shown in Table 1.

lines (SK-OV-3, HEY, and A2780.6.3) constitutively produced of PA by PLD (43, 44). Incubation of SK-OV-3 cells with LPA in the absence of any exogenous stimulus, and LPA treat- 1-butanol at 0.5%, a concentration that completely inhibits PA ment did not further increase the amount of LPA produced by formation by PLD (45), caused a consistent 50% reduction in these cell lines (Table 1). One cell line, SK-OV-3, constitutively the amount of LPA that is constitutively produced and released produced particularly high levels of LPA (Table 1 and Fig. 1A). by SK-OV-3 cells (Fig. 2A). We conclude that there are PLD- In these cells, LPA was the predominant phospholipid released dependent and -independent components of newly synthesized into the medium. In contrast, the breast cancer cell lines MCF7, LPA release by SK-OV-3 cells. PLD-independent synthesis MDA-MB-231, and MDA-MB-468 produced only low levels of might involve the sequential action of PLC and diacylglycerol LPA, and treatment with LPA did not increase LPA formation kinase (11).

(Table 1 and Fig. 1B). Indeed, the SK-OV-3, HEY, and Involvement of PLA2 in Constitutive LPA Production. A2780.6.3 cell lines produced 5.6 times more LPA than the Conversion of PA to LPA by PLA2 has been implicated in LPA three breast cancer cell lines (4708 units Ϯ 1590 versus 838 synthesis by platelets. On the basis of their biological properties, Ϯ units 169). In contrast to the ovarian cancer cell lines tested, PLA2s have been classified into three subgroups: sPLA2, normal ovarian epithelial cells and immortalized ovarian epithe- cPLA2, and iPLA2. The relative contribution of members of lial cells produced very low amounts of LPA and could not be each of these subgroups to LPA production is not known. We

induced by LPA to produce more LPA (data not shown). The explored the role of secretory PLA2s in the constitutive produc-

effects of LPA were specific, because lysophosphatidylcholine, tion of LPA by SK-OV-3 cells by using inhibitors of sPLA2. EGF, and PDGF did not increase LPA production (presented SK-OV-3 cells were incubated in the presence of either mano-

herein). alide, an inhibitor of group IIA (synovial) sPLA2 (IC50, 0.02– Role of PLD in Constitutive LPA Formation. PLD, 0.2 ␮M; Ref. 46), or OOEPC, an inhibitor of group IB (pancre- ␮ which converts membrane phospholipids to PA, has been im- atic) sPLA2 (IC50, 6.2 M; Ref. 47), prior to extraction and plicated in LPA production by platelets. We explored the role of analysis of newly formed phospholipids in the medium. Mano-

PLD in constitutive LPA production by SK-OV-3 cells using the alide, at a concentration that inhibits group IIA sPLA2 (48), did ability of primary short-chain alcohols to inhibit the formation not alter the amount of LPA synthesized and released by SK-

markedly decreased the level of LPA being produced and released into the medium (90% inhibition). We conclude that

cytosolic, calcium-dependent, and/or -independent PLA2s play a critical role in the constitutive production of LPA by SK-OV-3 cells. LPA-induced LPA Production. LPA markedly up- regulated LPA production in one of four ovarian cancer cell lines tested (OVCAR-3; see Table 1 and Fig. 3). This response was dose and time dependent. Treatment with 3 ␮M LPA resulted in the formation and release of a small amount of LPA. Treatment with 10 and 30 ␮M LPA, concentrations present in ascites of ovarian cancer patients, however, caused the production and release of substantial amounts of LPA (Fig. 4). LPA formation in response to LPA was rapid; high levels of LPA were detected in the supernatant of OVCAR-3 cells within 30 min of incubation with LPA. Very little labeled LPA could be detected after 24 h of treatment with LPA (data not shown). Ovarian cancer patients are potentially exposed to many different growth factors in ascitic fluid, among them LPC, EGF, and PDGF (12, 53). In contrast to LPA, EGF (Fig. 5), PDGF (Fig. 5), and LPC (not presented) did not increase LPA production in OVCAR-3 cells. Role of PLD in LPA-induced LPA Formation. Pre- treatment of OVCAR-3 cells with 1-butanol, which inhibits PLD-mediated PA formation, resulted in a decrease of inducible LPA production by ϳ60% (Fig. 6A). Because basal LPA pro- Fig. 2 Inhibitors of PLD, pancreatic (group IB) sPLA2 and cPLA2/ duction was also sensitive to the presence of 1-butanol (44% iPLA2, but not a nonpancreatic (group IIA) sPLA2 inhibitor block 32 inhibition), PLD appears to be involved in both the induced and constitutive LPA production. SK-OV-3 cells were labeled with [ P]Pi, pretreated with 0.5% 1-butanol which inhibits PLD (A), 5 ␮M manoalide basal LPA production in OVCAR-3 cells. However, in each ␮ which inhibits group IIA (synovial) sPLA2 (B), 20 M OOEPC which case there was also a PLD-independent component of LPA ␮ inhibits group IB (pancreatic) sPLA2 (C), or 100 M AACOCF3 which production. inhibits cPLA and iPLA (D) for 20 min and stimulated with 25 ␮M 2 2 Involvement of PLA in LPA-induced LPA Synthesis. LPA for 2 h. Lipids were extracted and analyzed by two-dimensional 2 TLC. LPA levels were quantitated by PhosphorImager and normalized Both the group IIA sPLA2 inhibitor manoalide and the group IB with respect to total phospholipids. The efficacy of each of the inhibitors sPLA2 inhibitor OOEPC reduced LPA-induced production of was indicated by a marked shift in the pattern of secreted phospholipids LPA by ϳ40% (Fig. 6, B and C). This suggests that both types (not presented). of secretory PLA2 play a role in LPA-induced LPA production. However, there remains considerable sPLA2-independent LPA- induced LPA production. We therefore assessed the participa-

OV-3 cells (Fig. 2B). In contrast, OOEPC, at a concentration tion of cPLA2 and iPLA2 in LPA-induced LPA formation by that blocks group IB sPLA2 (49), reduced the amount of LPA using the cPLA2 and iPLA2 inhibitor AACOCF3. In contrast to being produced and released by SK-OV-3 cells by ϳ80% (Fig. its effect on constitutive LPA production, AACOCF3 did not

2C). This result indicates that group IB sPLA2(s) play a major alter LPA-induced LPA production by OVCAR-3 cells (Fig.

role in the constitutive formation of LPA by SK-OV-3 cells, 6D). Therefore, cytosolic PLA2s (both calcium-dependent and whereas group IIA sPLA2 activity is not required for LPA -independent) do not seem to participate in LPA-induced pro- formation in SKOV-3 cells. Both manoalide and OOEPC altered duction of LPA. the distribution of phospholipids in cell supernatants, demon- LPA Receptors Implicated in LPA-induced LPA Pro- strating efficacy of the inhibitors (not presented). duction. We have demonstrated previously, by Northern blot

Recently, iPLA2s have been identified that display either analysis, that normal and immortalized ovarian epithelial cells an absolute specificity or a high selectivity for PA, therefore express the LPA receptor Edg1, whereas ovarian cancer cell possibly playing a role in LPA synthesis (30). Moreover, most lines express only very low levels of Edg1 (12). mRNA levels

cell types contain cPLA2 that are specific for arachidonic acid at for Edg2 markedly vary among normal and immortalized ovar- the sn-2 position and that potentially are also involved in LPA ian epithelial cells as well as among ovarian cancer cell lines formation (36). The arachidonic acid analogue AACOCF3 in- (12, 22, 23). Edg4 mRNA is expressed in normal ovarian ␮ hibits both cPLA2 (IC50,50 M; Ref. 50) and iPLA2 (IC50,15 epithelial cells, and its mRNA levels are elevated in ovarian ␮M; Ref. 51) and thus can be used to explore the function of cancer cells (12, 22). The OVCAR-3 cell line expresses mod-

both types of cytosolic PLA2s. SK-OV-3 cells were treated with erately increased levels of Edg4 (12, 22). Recently, a novel LPA AACOCF3 before measuring the amount of newly synthesized receptor, Edg7, was identified and cloned (25). We determined LPA released into the medium. AACOCF3 at 100 ␮M, a con- its expression levels in normal and immortalized ovarian epi-

centration that completely blocks both cPLA2 and iPLA2 (52), thelial cells as well as in ovarian cancer cell lines by Northern

Fig. 3 Two-dimensional TLC analysis of LPA-inducible LPA production by OVCAR-3 cells. 32 Cells were labeled with [ P]Pi and incubated for2hinthe absence (upper panel) or presence (lower panel)of25␮M LPA before the supernatant was removed. Lipids were extracted and analyzed by two-dimensional TLC. Newly synthesized LPA was identified by comigration with nonradioactive LPA. ori, origin of the sample on the TLC plate. Relative PhosphorImager units for LPA from this experiment are shown in Table 1.

blot analysis. Normal and immortalized ovarian epithelial cells is produced in the peritoneal cavity and then migrates to the express barely detectable levels of Edg7 mRNA, whereas ovar- peripheral circulation. Ovarian cancer cells have been impli- ian cancer cells express Edg7 at varying levels (Fig. 7). Intrigu- cated to be the source of LPA production in ascites, because ingly, the highest level of Edg7 expression is found in the they have been shown to synthesize LPA in response to the OVCAR-3 cell line, which constitutively produces very low tumor-promoting agent PMA (28). However, whether PMA levels of LPA and which produces markedly increased levels of mimics a physiological process is not known. Ovarian cancer LPA in response to LPA. In OVCAR-3 cells, the change in cells in the patient might produce LPA either constitutively or Edg7 expression as compared with normal and immortalized after activation by the cellular milieu. Knowledge of regulation ovarian epithelial cells is much greater than the change in Edg4 of LPA production by ovarian cancer cells and the enzymes expression (Fig. 7; 19, 22). This suggests that Edg7 and poten- involved in LPA synthesis could lead to the development of tially Edg4 may play a role in LPA-induced LPA production by therapeutic measures that would interfere with LPA synthesis OVCAR-3 cells. and its deleterious effects on ovarian cancer cells. Here, we report that three of four ovarian cancer cell lines tested consti- DISCUSSION tutively produce LPA at much higher levels than breast cancer LPA stimulates growth, prevents apoptosis and anoikis, cells or normal ovarian epithelial cells, and that one ovarian decreases sensitivity to chemotherapeutic drugs, increases pro- cancer cell line can be induced by LPA to produce LPA. duction of neovascularization mediators, and increases invasive- Strikingly, the one ovarian cancer cell line induced to release ness of ovarian cancer cells (12). LPA levels are elevated in LPA by LPA constitutively produced the lowest level of LPA, ascites and plasma from ovarian cancer patients, implicating it raising the possibility that constitutive LPA production by the in ovarian tumorigenesis (2–6). Because LPA levels are higher other ovarian cancer cell lines played a role in amplifying LPA in ascitic fluid than in plasma, it has been hypothesized that LPA production. Normal ovarian epithelial cells, which constitutively

Fig. 4 LPA induces LPA production by OVCAR-3 cells in a dose- 32 dependent fashion. OVCAR-3 cells were labeled with [ P]Pi and left unstimulated (control) or stimulated with increasing concentrations of LPA for 2 h. Lipids were extracted from cell supernatants and analyzed by two-dimensional TLC. LPA levels were quantitated by PhosphorIm- ager and normalized with respect to total phospholipids.

Fig. 6 Inhibitors of PLD, pancreatic (group IB) sPLA2 and cPLA2/ iPLA2, but not a nonpancreatic (group IIA) sPLA2, inhibitor block LPA-induced LPA production. OVCAR-3 cells were labeled with 32 ␮ [ P]Pi, pretreated with 0.5% 1-butanol which inhibits PLD (A), 5 M manoalide which inhibits group IIA (synovial) sPLA2 (B), 20 ␮ ␮ Fig. 5 EGF and PDGF do not induce LPA production by OVCAR-3 M OOEPC which inhibits group IB (pancreatic) sPLA2 (C), or 100 M 32 AACOCF3 which inhibits cPLA2 and iPLA2 (D) for 20 min and cells. OVCAR-3 cells were labeled with [ P]Pi and left unstimulated or stimulated with 25 ␮M LPA for 2 h. Lipids were extracted and analyzed stimulated with 25 ␮M LPA, 10 ng/ml EGF, or 50 ng/ml PDGF for 2 h. Lipids were extracted from cell supernatants and analyzed by two- by two-dimensional TLC. LPA levels were quantitated by PhosphorIm- dimensional TLC. LPA levels were quantitated by PhosphorImager and ager and normalized with respect to total phospholipids. normalized with respect to total phospholipids.

contains at least 10 different isoforms. PLD is involved in the produce low levels of LPA, did not produce LPA in response to formation of the LPA precursor PA, and as shown herein, PLD exogenous LPA. Normal ovarian epithelial cells express low indeed plays a role in the production of LPA in ovarian cancer

levels of mRNA for the Edg4 and 7 LPA receptors (12, 22). cells. Little is known about the role that the various PLA2 mRNA levels for Edg4 and particularly Edg7 are markedly enzymes play in LPA formation. It has been shown that sPLA2 elevated in ovarian cancer cells (12, 22), suggesting that the is inactive on intact membrane bilayers but requires membrane novel expression of these receptors may mediate LPA-induced rearrangement and subsequent loss of membrane asymmetry to LPA production by ovarian cancer cells. Levels of Edg1 mRNA, mediate LPA production (35, 54–56). Such loss of membrane a putative LPA receptor, are high in normal ovarian epithelial asymmetry occurs during apoptosis or malignant transformation cells but low in most ovarian cancer cells (12), suggesting that (57). We have shown that LPA production by ovarian cancer

this receptor is not relevant to LPA-induced LPA production. cells requires group IB (pancreatic) sPLA2 activity, whereas Edg2 levels are not consistently altered between normal ovarian group IIA (synovial) sPLA2 does not seem to play a role in epithelial cells and cancer cells (12, 22, 23), and furthermore, constitutive LPA production by ovarian cancer cells and seems Edg2 appears to function as a negative receptor for LPA in to play only a minor role in the induction of LPA by LPA. There

ovarian cancer cells (23). seems to be a differential requirement for cPLA2 and/or iPLA2 The two main types of enzymes involved in LPA synthesis phospholipase A2 by cells that constitutively produce LPA

are PLD, which contains at least two isoforms, and PLA2, which versus cells that are induced by LPA to produce LPA. Consti-

seems to be a mutual interdependency between sPLA2 and

cPLA2, because group IIA sPLA2 also increases the expression

of cPLA2 (60). Recently, it was reported that sPLA2 may MAPK indirectly regulate cPLA2 by activating p38 (61), which in

turn phosphorylates cPLA2, contributing to its activation (42, 62). Taken together, these findings point to a complex interplay

between sPLA2 and cPLA2, possibly also iPLA2. Specific PLA2 isoform inhibitors should allow further elucidation of the effects

that the various PLA2 enzymes display on each other and on specific functions, such as LPA synthesis. LPA has been previously demonstrated to activate PLD

(39–41). The mechanism(s) by which LPA regulates PLA2 have not, however, been explored. We have demonstrated previously that LPA induces rapid increases in cytosolic free calcium and activates MAPK in ovarian cancer cells (13). It was

thus somewhat surprising that cPLA2, which is activated by increases in cytosolic calcium (42), does not seem to be involved in LPA-induced LPA production. We have shown recently that LPA activates p38MAPK in OVCAR-3 cells.5 This might be a mechanism by which constitutively produced LPA in

ovarian cancer cells contributes to cPLA2 activation.

PLA1, which cleaves at the sn-1 position of a glycerophos- pholipid, may be involved in the production of one particular Fig. 7 Ovarian cancer cell lines, but not normal or immortalized ovarian epithelial cells, express Edg7. Total RNA (10 ␮g) from the indicated species of LPA found in the ascites of ovarian cancer patients. cell lines was separated on a 1% denaturing agarose gel, and Northern LPA found in ascites consists of a mixture of sn-1 and sn-2 blot analysis was performed. The membrane was hybridized with 32P- species, with the sn-2 species exhibiting greater bioactivity than labeled Edg-7 probes, stripped, and rehybridized with 18S probes (upper the sn-1 counterpart (4). PA, the precursor of LPA, is the panel). Edg7 mRNA levels were quantitated by PhosphorImager and normalized with respect to 18S RNA (lower panel). preferred substrate of both a membrane-bound (63) and a cytosolic PLA1 (64), thus further implicating PLA1 in LPA synthe-

sis. It will be interesting to explore the contribution of PLA1 to

LPA production by ovarian cancer cells once inhibitors of PLA1 tutive LPA production has an absolute requirement for cPLA s become available. Interestingly, a recently described isoform of 2 ␤ and/or iPLA s, whereas they do not seem to play a role in LPA cPLA2, cPLA2- , prefers sn-1 cleavage to sn-2 cleavage (65), 2 ␥ induction by LPA. whereas another isoform, cPLA2- , efficiently cleaves at both The requirement for both secretory and cytosolic (calcium- positions (66). These cPLA2 enzymes could thus also contribute to sn-2 LPA formation in the ascites of ovarian cancer patients. dependent and/or -independent) PLA2 activity for constitutive LPA production might reflect a previously described cross-talk LPA levels in cell membranes are low (11), reflecting rapid conversion or degradation of LPA. Reduced rates of conversion between sPLA2s and cPLA2. Functionally active cPLA2 may be required to activate sPLA2 and to mediate LPA production as and/or degradation might contribute to the elevated levels of cPLA2 activation precedes that of sPLA2 (58). Furthermore, newly synthesized LPA in the supernatant of ovarian cancer blocking cPLA2 with specific inhibitors leads to a pronounced cells as compared with breast cancer cells. This may also con- reduction of arachidonic release from P388D1 macrophages, tribute to LPA-induced increases in LPA levels. LPA is con- which is greater than the expected change, considering that verted back to PA by LPA acyltransferase, whereas PA phos- sPLA2 is responsible for the majority of arachidonic acid re- phohydrolases and lysophospholipases rapidly degrade LPA leased (58). It has been postulated that an increase of free (11). Decreased expression or activity of these enzymes may arachidonic acid brought about by cPLA2 catalysis activates contribute to the increased LPA levels in ovarian cancer pa- sPLA2 (59), possibly by resulting in the membrane rearrange- tients. ment that appears to be required for sPLA2 activity (35, 54–56). In summary, we have shown that ovarian cancer cells, but In addition, AACOCF3, a cPLA2/iPLA2-specific inhibitor, not breast cancer cells or normal ovarian epithelial cells, release markedly reduced interleukin 1/tumor necrosis factor-induced high levels of LPA into the extracellular medium. We have group IIA sPLA2 expression at the mRNA and protein level further shown that PLD plays a role in LPA synthesis by ovarian

(59). This suggests that arachidonic acid released by cPLA2 at cancer cells, and that different PLA2 isoforms are required for the early stage of cytokine stimulation is required for the sub- constitutive and LPA-induced LPA production. These findings sequent induction of group IIA sPLA2 expression. The addition are clinically relevant because ascites and plasma of ovarian of exogenous arachidonic acid only partially reversed the (indi- cancer patients, but not of patients with nongynecological tu- rect) inhibition of group IIA sPLA2 by AACOCF3, which might mors, contain elevated levels of LPA. LPA in ovarian cancer reflect the requirement of additional cPLA2 or iPLA2 metabo- patients might be used as a marker for early diagnosis and as a lites for group IIA sPLA2 induction (59). Interestingly, there molecular target for therapeutic intervention.

ACKNOWLEDGMENTS similarities to G-protein-coupled receptors. J. Biol. Chem., 265: 9308– We thank Drs. Nellie Auersperg, Ron Buick, Thomas Hamilton, 9313, 1990. and Janet Price for providing cell lines. 19. Lee, M. J., Van Brocklyn, J. R., Thangada, S., Liu, C. H., Hand, A. R., Menzeleev, R., Spiegel, S., and Hla, T. Sphingosine-1-phosphate as a ligand for the G protein-coupled receptor EDG-1. Science (Wash- REFERENCES ington DC), 279: 1552–1555, 1998. 20. Lee, M. J., Thangada, S., Liu, C. H., Thompson, B. D., and Hla, T. 1. American Cancer Society. Cancer Facts and Figures-1998. Atlanta: Lysophosphatidic acid stimulates the G-protein-coupled receptor American Cancer Society, Inc., 1999. EDG-1 as a low affinity agonist. J. Biol. Chem., 273: 22105–22112, 2. Mills, G. B., May, C., McGill, M., Roifman, C. M., and Mellors, A. 1998. A putative new growth factor in ascitic fluid from ovarian cancer patients: identification, characterization, and mechanism of action. Can- 21. An, S., Dickens, M. A., Bleu, T., Hallmark, O. G., and Goetzl, E. J. cer Res., 48: 1066–1071, 1988. Molecular cloning of the human Edg2 protein and its identification as a functional cellular receptor for lysophosphatidic acid. Biochem. Bio- 3. Mills, G. B., May, C., Hill, M., Campbell, S., Shaw, P., and Marks, phys. Res. Commun., 231: 619–622, 1997. A. Ascitic fluid from human ovarian cancer patients contains growth factors necessary for intraperitoneal growth of human ovarian adeno- 22. Goetzl, E. J., Dolezalova, H., Kong, Y., Hu, Y-L., Jaffe, R. B., carcinoma cells. J. Clin. Investig., 86: 851–855, 1990. Kalli, K. R., and Conover, C. A. Distinctive expression and functions of the type 4 endothelial differentiation gene-encoded G protein-coupled 4. Xu, Y., Gaudette, D. C., Boynton, J. D., Frankel, A., Fang, X. J., Sharma, A., Hurteau, J., Casey, G., Goodbody, A., Mellors, A., et al. receptor for lysophosphatidic acid in ovarian cancer. Cancer Res., 59: Characterization of an ovarian cancer activating factor in ascites from 5370–5375, 1999. ovarian cancer patients. Clin. Cancer Res., 1: 1223–1232, 1995. 23. Furui, T., LaPushin, R., Mao, M., Kahn, H., Watt, S. R., Watt, 5. Westermann, A. M., Havik, E., Postma, F. R., Beijnen, J. H., Dalesio, M-A. V., Lu, Y., Fang, X., Tsutsui, S., Siddik, Z., Bast, R. C., Jr., and O., Moolenaar, W. H., and Rodenhuis, S. Malignant effusions contain Mills, G. B. Overexpression of Edg-2/vzg-1 induces apoptosis and lysophosphatidic acid (LPA)-like activity. Ann. Oncol., 9: 437–442, anoikis in ovarian cancer cells in a lysophosphatidic acid independent 1998. manner. Clin. Cancer Res., 5: 4308–4318, 2000. 6. Xu, Y., Shen, Z., Wiper, D. W., Wu, M., Morton, R. E., Elson, P., 24. An, S., Bleu, T., Hallmark, O. G., and Goetzl, E. J. Characterization Kennedy, A. W., Belinson, J., Markman, M., and Casey, G. Lysophos- of a novel subtype of human G protein-coupled receptor for lysophosphatidic acid as a potential biomarker for ovarian and other gynecologic phatidic acid. J. Biol. Chem., 273: 7906–7910, 1998. cancers [see comments]. J. Am. Med. Assoc., 280: 719–723, 1998. 25. Bandoh, K., Aoki, J., Hosono, H., Kobayashi, S., Kobayashi, T., 7. Sasagawa, T., Okita, M., Murakami, J., Kato, T., and Watanabe, A. Murakami-Murofushi, K., Tsujimoto, M., Arai, H., and Inoue, K. Mo- Abnormal serum lysophospholipids in multiple myeloma patients. Lip- lecular cloning and characterization of a novel human G-protein- cou- ids, 34: 17–21, 1999. pled receptor, EDG7, for lysophosphatidic acid. J. Biol. Chem., 274: 27776–27785, 1999. 8. Sasagawa, T., Suzuki, K., Shiota, T., Kondo, T., and Okita, M. The significance of plasma lysophospholipids in patients with renal failure 26. Eichholtz, T., Jalink, K., Fahrenfort, I., and Moolenaar, W. H. The on hemodialysis. J. Nutr. Sci. Vitaminol., 44: 809–818, 1998. bioactive phospholipid lysophosphatidic acid is released from activated platelets. Biochem. J., 291: 677–680, 1993. 9. Moolenaar, W. Lysophosphatidic acid, a multifunctional phospholipid messenger. J. Biol. Chem., 270. 12949–12952, 1995. 27. Fukami, K., and Takenawa, T. Phosphatidic acid that accumulates in platelet-derived growth factor-stimulated Balb/c 3T3 cells is a poten- 10. Moolenaar, W. H., Kranenburg, O., Postma, F. R., and Zondag, tial mitogenic signal. J. Biol. Chem., 267: 10988–10993, 1992. G. C. Lysophosphatidic acid: G-protein signalling and cellular re- sponses. Curr. Opin. Cell Biol., 9: 168–173, 1997. 28. Shen, Z., Belinson, J., Morton, R. E., Xu, Y., and Xu, Y. Phorbol 12-myristate 13-acetate stimulates lysophosphatidic acid secretion from 11. Goetzl, E. J., and An, S. Diversity of cellular receptors and func- ovarian and cervical cancer cells but not from breast or leukemia cells. tions for the lysophospholipid growth factors lysophosphatidic acid and Gynecol. Oncol., 71: 364–368, 1998. sphingosine 1-phosphate. FASEB J., 12: 1589–1598, 1998. 29. Jalink, K., Hordijk, P. L., and Moolenaar, W. H. Growth factor-like 12. Fang, X., Gaudette, D., Furui, T., Mao, M., Estrella, V., Eder, A., effects of lysophosphatidic acid, a novel lipid mediator. Biochim. Bio- Pustilnik, T., Sasagawa, T., Lapushin, R., Yu, S., Jaffe, R. B., Wiener, phys. Acta, 1198: 185–196, 1994. J. R., Erickson, J. R., and Mills, G. B. Lysophospholipid growth factors in the initiation, progression, metastases and management of cancer. 30. Gaits, F., Fourcade, O., Le Balle, F., Gueguen, G., Gaige, B., Ann. NY Acad. Sci., in press, 2000. Gassama-Diagne, A., Fauvel, J., Salles, J. P., Mauco, G., Simon, M. F., and Chap, H. Lysophosphatidic acid as a phospholipid mediator: path- 13. Xu, Y., Fang, X. J., Casey, G., and Mills, G. B. Lysophospholipids ways of synthesis. FEBS Lett., 410: 54–58, 1997. activate ovarian and breast cancer cells. Biochem. J., 309: 933–940, 1995. 31. Exton, J. H. Regulation of phospholipase D. Biochim Biophys Acta, 1439: 121–133, 1999. 14. Fang, X., Yu, S., Eder, A., Mao, M., Bast, R. C., Jr., Boyd, D., and Mills, G. B. Regulation of BAD phosphorylation at serine 112 by the 32. Frohman, M. A., Sung, T. C., and Morris, A. J. Mammalian phos- Ras-mitogen-activated protein kinase pathway. Oncogene, 18: 6635– pholipase D structure and regulation. Biochim. Biophys. Acta, 1439: 6640, 1999. 175–186, 1999. 15. Frankel, A., and Mills, G. B. Peptide and lipid growth factors 33. Dennis, E. A. Diversity of group types, regulation, and function of decrease cis-diamminedichloroplatinum-induced cell death in human phospholipase A2. J. Biol. Chem., 269: 13057–13060, 1994.

ovarian cancer cells. Clin. Cancer Res., 2: 1307–1313, 1996. 34. Dennis, E. A. The growing phospholipase A2 superfamily of signal 16. Pustilnik, T. B., Estrelle, V., Wiener, J., Mao, M., Eder, A., Watt, transduction enzymes. Trends Biochem. Sci., 1: 1–2, 1997. M. A., Bast, R. C., Jr., and Mills, G. B. Lysophosphatidic acid induces 35. Kudo, I., Murakami, M., Hara, S., and Inoue, K. Mammalian

urokinase secretion in ovarian cancer cells. Clin. Cancer Res., 11: non-pancreatic phospholipases A2. Biochim. Biophys. Acta, 1170: 217– 3704–3710, 1999. 231, 1993.

17. An, S., Goetzl, E. J., and Lee, H. Signaling mechanisms and 36. Gijon, M. A., and Leslie, C. C. Phospholipases A2. Semin. Cell molecular characteristics of G protein-coupled receptors for lysophos- Dev. Biol., 3: 297–303, 1997. phatidic acid and sphingosine 1-phosphate. J. Cell. Biochem. Suppl., 31: 37. Chaminade, B., Le Balle, F., Fourcade, O., Nauze, M., Delagebeau- 147–157, 1998. deuf, C., Gassama-Diagne, A., Simon, M-F., Fauvel, J., and Chap, H.

18. Hla, T., and Maciag, T. An abundant transcript induced in differ- New developments in phospholipase A2. Lipids Suppl., 34: S49–S55, entiating human endothelial cells encodes a polypeptide with structural 1999.

38. van Dijk, M. C., Postma, F., Hilkmann, H., Jalink, K., van Blitter- in thrombin-stimulated human platelets. J. Biol. Chem., 269: 15625– swijk, W. J., and Moolenaar, W. H. Exogenous phospholipase D gen- 15630, 1994. 2ϩ erates lysophosphatidic acid and activates Ras, Rho and Ca signaling 53. Westermann, A. M., Beijnen, J. H., Moolenaar, W. H., and Roden- pathways. Curr. Biol., 8: 386–392, 1998. huis, S. Growth factors in human ovarian cancer. Cancer Treat. Rev., 39. Malcolm, K. C., Elliott, C. M., and Exton, J. H. Evidence for 23: 113–131, 1997. Rho-mediated agonist stimulation of phospholipase D in rat1 fibro- 54. Fourcade, O., Simon, M. F., Viode, C., Rugani, N., Leballe, F., blasts. Effects of Clostridium botulinum C3 exoenzyme. J. Biol. Chem., Ragab, A., Fournie, B., Sarda, L., and Chap, H. Secretory phospholipase 271: 13135–13139, 1996. A2 generates the novel lipid mediator lysophosphatidic acid in mem- 40. Cross, M. J., Roberts, S., Ridley, A. J., Hodgkin, M. N., Stewart, A., brane microvesicles shed from activated cells. Cell, 80: 919–927, 1995. Claesson-Welsh, L., and Wakelam, M. J. O. Stimulation of actin stress 55. Fourcade, O., Le Balle, F., Fauvel, J., Simon, M. F., and Chap, H. fiber formation mediated by activation of phospholipase D. Curr. Biol., Regulation of secretory type-II phospholipase A2 and of lysophospha- 6: 588–597, 1996. tidic acid synthesis. Adv. Enzyme Regul., 38: 99–107, 1998. 41. Qi, C., Park, J. H., Gibbs, T. C., Shirley, D. W., Bradshaw, C. D., 56. le Balle, F., Simon, M. F., Meijer, S., Fourcade, O., and Chap, H. Ella, K. M., and Meier, K. E. Lysophosphatidic acid stimulates phos- Membrane sidedness of biosynthetic pathways involved in the produc- pholipase D activity and cell proliferation in PC-3 human prostate tion of lysophosphatidic acid. Adv. Enzyme Regul., 39: 275–284, 1999. cancer cells. J. Cell Physiol., 174: 261–272, 1998. 57. Zwaal, R. F., and Schroit, A. J. Pathophysiologic implications of 42. Leslie, C. C. Properties and regulation of cytosolic phospholipase membrane phospholipid asymmetry in blood cells. Blood, 89: 1121– A2. J. Biol. Chem., 272: 16709–16712, 1997. 1132, 1997. 43. Pai, J. K., Siegel, M. I., Egan, R. W., and Billah, M. M. Activation 58. Balsinde, J., and Dennis, E. A. Distinct roles in signal transduction of phospholipase D by chemotactic peptide in HL-60 granulocytes. for each of the phospholipase A2 enzymes present in P388D1 macro- Biochem. Biophys. Res. Commun., 150: 355–364, 1988. phages. J. Biol. Chem., 271: 6758–6765, 1996. 44. Bonser, R. W., Thompson, N. T., Randall, R. W., and Garland, 59. Kuwata, H., Nakatani, Y., Murakami, M., and Kudo, I. Cytosolic L. G. Phospholipase D activation is functionally linked to superoxide phospholipase A2 is required for cytokine-induced expression of type generation in the human neutrophil. [published erratum appears in IIA secretory phospholipase A2 that mediates optimal cyclooxygenase- Biochem. J., 265: 932, 1990]. Biochem. J., 264: 617–620, 1989. 2-dependent delayed prostaglandin E2 generation in rat 3Y1 fibroblasts. 45. Ohguchi, K., Kasai, T., and Nozawa, Y. Tyrosine phosphorylation J. Biol. Chem., 273: 1733–1740, 1998. of 100–115 kDa proteins by phosphatidic acid generated via phospho- 60. Murakami, M., Kuwata, H., Amakasu, Y., Shimbara, S., Nakatani, lipase D activation in HL60 granulocytes. Biochim. Biophys. Acta, Y., Atsumi, G., and Kudo, I. Prostaglandin E2 amplifies cytosolic 1346: 301–304, 1997. phospholipase A2- and cyclooxygenase-2-dependent delayed prostag- 46. Jacobson, P. B., Marshall, L. A., Sung, A., and Jacobs, R. S. landin E2 generation in mouse osteoblastic cells. Enhancement by

Inactivation of human synovial fluid phospholipase A2 by the marine secretory phospholipase A2. J. Biol. Chem., 272: 19891–19897, 1997. natural product, manoalide. Biochem. Pharmacol., 39: 1557–1564, 61. Hernandez, M., Burillo, S. L., Crespo, M. S., and Nieto, M. L. 1990. Secretory phospholipase A2 activates the cascade of mitogen-activated 47. Magolda, R. L., Ripka, W. C., Galbraith, W., Johnson, P. R., and protein kinases and cytosolic phospholipase A2 in the human astrocy- Rudnick, M. S. Novel synthesis of potent site-specific phospholipase A2 toma cell line 1321N1. J. Biol. Chem., 273: 606–612, 1998. inhibitors. In: J. M. Bailey (ed.), Prostaglandins, Leukotrienes and 62. Borsch-Haubold, A. G., Bartoli, F., Asselin, J., Dudler, T., Kramer, Lipoxins, pp. 669–676. New York: Plenum Publishing Corp., 1985. R. M., Apitz-Castro, R., Watson, S. P., and Gelb, M. H. Identification of

48. Miao, J. Y., Kaji, K., Hayashi, H., and Araki, S. Inhibitors of the phosphorylation sites of cytosolic phospholipase A2 in agonist- phospholipase promote apoptosis of human endothelial cells. J. Bio- stimulated human platelets and HeLa cells. J. Biol. Chem., 273: 4449– chem., 121: 612–618, 1997. 4458, 1998. 49. Lu, Y., Morley, P., and Durkin, J. P. Signalling events mediating the 63. Kucera, G. L., Sisson, P. J., Thomas, M. J., and Waite, M. On the activation of protein kinase C by interleukin-2 in cytotoxic T cells. Cell. substrate specificity of rat liver phospholipase A1. J. Biol. Chem., 263: Signal., 11: 275–285, 1999. 1920–1928, 1988. 50. Conde-Frieboes, K., Reynolds, L. J., Lio, Y-C., Hale, M. R., Was- 64. Higgs, H. N., Han, M. H., Johnson, G. E., and Glomset, J. A. serman, H. H., and Dennis, E. A. Activated ketones as inhibitors of Cloning of a phosphatidic acid-preferring phospholipase A1 from bovine 2ϩ 2ϩ intracellular Ca -dependent and Ca -independent phospholipase A2. testis. J. Biol. Chem., 273: 5468–5477, 1998. J. Am. Chem. Soc., 118: 5519–5525, 1996. 65. Song, C., Chang, X. J., Bean, K. M., Proia, M. S., Knopf, J. L., and ␤ 51. Ackermann, E. J., Conde-Frieboes, K., and Dennis, E. A. Inhibition Kriz, R. W. Molecular characterization of cytosolic phospholipase A2- . 2ϩ of macrophage Ca -independent phospholipase A2 by bromoenol lac- J. Biol. Chem., 274: 17063–17067, 1999. tone and trifluoromethyl ketones. J. Biol. Chem., 270: 445–450, 1995. 66. Underwood, K. W., Song, C., Kriz, R. W., Chang, X. J., Knopf,

52. Bartoli, F., Lin, H. K., Ghomashchi, F., Gelb, M. H., Jain, M. K., J. L., and Lin, L. L. A novel calcium-independent phospholipase A2, ␥ and Apitz-Castro, R. Tight binding inhibitors of 85-kDa phospholipase cPLA2- , that is prenylated and contains homology to cPLA2. J. Biol. A2 but not 14-kDa phospholipase A2 inhibit release of free arachidonate Chem., 273: 21926–19232, 1998.

Downloaded from clincancerres.aacrjournals.org on September 25, 2021. © 2000 American Association for Cancer Research. Constitutive and Lysophosphatidic Acid (LPA)-induced LPA Production: Role of Phospholipase D and Phospholipase A 2

Astrid M. Eder, Takayo Sasagawa, Muling Mao, et al.

Clin Cancer Res 2000;6:2482-2491.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/6/6/2482

Cited articles This article cites 56 articles, 27 of which you can access for free at: http://clincancerres.aacrjournals.org/content/6/6/2482.full#ref-list-1

Citing articles This article has been cited by 24 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/6/6/2482.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/6/6/2482. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.