Bax Apoptosis in Association with Suppression of 1-Phosphate

Home , Ceramide

Lysophosphatidic Acid and Sphingosine 1-Phosphate Protection of T Cells from Apoptosis in Association with Suppression of Bax This information is current as of September 26, 2021. Edward J. Goetzl, Yvonne Kong and Baisong Mei J Immunol 1999; 162:2049-2056; ; http://www.jimmunol.org/content/162/4/2049 Downloaded from

References This article cites 25 articles, 13 of which you can access for free at: http://www.jimmunol.org/content/162/4/2049.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on September 26, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1999 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Lysophosphatidic Acid and Sphingosine 1-Phosphate Protection of T Cells from Apoptosis in Association with Suppression of Bax1

Edward J. Goetzl,2 Yvonne Kong, and Baisong Mei

Members of a subfamily of G protein-coupled receptors (GPCRs), encoded by ﬁve different endothelial differentiation genes (edgs), speciﬁcally mediate effects of lysophosphatidic acid (LPA) and sphingosine 1-phosphate (S1P) on cellular proliferation and differentiation. Mechanisms of suppression of apoptosis by LPA and S1P were studied in the Tsup-1 cultured line of human T lymphoblastoma cells, which express Edg-2 and Edg-4 GPCRs for LPA and Edg-3 and Edg-5 GPCRs for S1P. At 10؊10 Mto10؊7 M, both LPA and S1P protected Tsup-1 cells from apoptosis induced by Abs to Fas, CD2, and CD3 plus CD28 in combination.

Apoptosis elicited by C6 ceramide was inhibited by S1P, but not by LPA, in part because ceramide suppressed expression of Edg-2 Downloaded from and Edg-4 surface receptors for LPA without affecting Edg-3 surface receptors for S1P. At 10؊9 Mto10؊7 M, LPA and S1P signiﬁcantly suppressed cellular levels of the apoptosis-promoting protein Bax, without altering the levels of Bcl-xL or Bcl-2 assessed by Western blots and immunoassays. Transfections of pairs of antisense plasmids for Edg-2 plus Edg-4 and Edg-3 plus Edg-5, and hygromycin selection of transfectants with reduced expression of the respective Edg R proteins in Western blots, inhibited both protection from apoptosis and reduction in cellular levels of Bax by LPA and S1P. Thus, LPA and S1P protection

from apoptosis is mediated by distinct Edg GPCRs and may involve novel effects on Bax regulatory protein. The Journal of http://www.jimmunol.org/ Immunology, 1999, 162: 2049–2056.

he lysophospholipid mediators lysophosphatidic acid ferentiation genes (edgs) 1–5. Edg protein GPCRs show amino (LPA)3 and sphingosine 1- phosphate (S1P) are generated acid sequence identity of 31–34% as a subfamily, but contain two T by complex enzymatic pathways from membranes of homology clusters with greater internal similarity of structures and many different types of stimulated cells (1–3). LPA and S1P are ligand specificity. Human Edg-2 and Edg-4 proteins constitute one both characterized by wide-spread cellular production, micromolar cluster, which are 46% identical and 72% similar in amino acid maximal concentrations in serum and some tissue fluids, high lev- sequence and are both GPCRs for LPA, but not S1P or other ly- els of binding to serum albumin, and biodegradation by multiple sosphingolipids (11, 12). Mouse Edg-2 GPCR also is a highly by guest on September 26, 2021 enzymatic mechanisms (4, 5). In extracellular fluids, these lipids specific LPA receptor (13). Edg-1, Edg-3, and Edg-5 constitute a are potent stimuli of cellular proliferation, differentiation, survival, second cluster of GPCRs, which are 45–60% identical in their adhesion, aggregation, and other specific functions (6–8). The ex- amino acid sequences and specifically bind and transduce signals istence of G protein-coupled receptors (GPCRs) for LPA and S1P from S1P and possibly other lysosphingolipids, but not LPA (14, was suggested initially by specific ligand structural dependence of 15). Xenopus oocytes and murine cells express a second type of their effects, ligand-induced desensitization of some cellular re- GPCR, termed psp24, which is not structurally homologous to Edg sponses, and pertussis toxin inhibition of their cellular Ca2ϩ mo- protein GPCRs, but specifically transduces LPA-evoked oscilla- bilizing and proliferative activities (9, 10). tory ClϪ currents by activation of the inositol trisphosphate-Ca2ϩ Two distinct types of GPCRs for the lysophospholipid media- system (16). tors recently have been defined structurally (11–16). One subfam- The capacities of LPA and S1P to enhance cellular survival ily of GPCRs for LPA and S1P consists of at least five homologous recently have been attributed in part to suppression of apoptosis seven-transmembrane domain proteins encoded by endothelial dif- (17, 18), but the complex mechanisms by which these lipids suppress apoptosis have not been elucidated fully. The present study was designed to delineate mechanisms central to LPA and S1P Departments of Medicine and Microbiology-Immunology, University of California protection of T cells from apoptosis and to identify associated Medical Center, San Francisco, CA 94143 alterations in cellular levels of proteins of the Bcl superfamily that Received for publication July 24, 1998. Accepted for publication November 9, 1998. may mediate suppression of apoptosis by these lipids. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Materials and Methods 1 The research described was supported by Grant HL31809 from the National Insti- Chemical reagents and Abs tutes of Health (E.J.G.). The sources of chemicals were: S1P, sphingosylphosphorylcholine, sphin- 2 Address correspondence and reprint requests to Dr. Edward J. Goetzl, Immunology gosine (S), and C6 ceramide (Biomol, Plymouth Meeting, PA); LPA, phos- and Allergy, UB8B, Box 0711, University of California Medical Center, 533 Parnas- ␤ sus, San Francisco, CA 94143-0711. E-mail address: [email protected] phatidic acid (PA), 1- -D-galactosyl-sphingosine (psychosine, PSSP), 3 hygromycin (Calbiochem, San Diego, CA), and fatty acid-free BSA (faf- Abbreviations used in this paper: LPA, 1-oleoyl-lysophosphatidic acid; S1P, sphin- BSA) (Sigma, St. Louis, MO). Hybridomas producing mouse mAbs spe- gosine 1-phosphate; S, sphingosine; PSSP, psychosine; PA, phosphatidic acid; GPCR, G protein-coupled receptor; edg, endothelial differentiation gene; Edg, GPCR en- cific for substituent peptides of human Edg-3 (amino acids 1–21), Edg-4 coded by an edg; RT, reverse transcription; G3PDH, glyceraldehyde 3-phosphate (9–27), and Edg-5 (303–322) were generated from splenocytes of female dehydrogenase; VIPR1, type I vasoactive intestinal peptide receptor; VIPR2, type II BALB/c mice, which had been immunized first in multiple s.c. and i.m. vasoactive intestinal peptide receptor; Rs, surface receptors; bp, base pair. sites with 100 ␮g of keyhole limpet hemocyanin conjugate (Pierce,

Rockford, IL) of the respective peptides in CFA, 3 wk later and weekly for 0.2 ␮g of anti-CD2 Ab, or a combination of 0.5 ␮g each of anti-CD3 and five additional weeks with 50 ␮g of the same conjugate in IFA, and then anti-CD28 Abs, and in others 5 ␮M C6 ceramide was the stimulus for with 100 ␮g of unconjugated peptide alone i.v. 2–3 days before removal of apoptosis. LPA, S1P, and other lipids were dispersed in 0.05 g/100 ml of the spleen (Ab Solutions, Palo Alto, CA). Each monoclonal IgG was pu- faf-BSA in medium. The principal assay for quantification of apoptosis was rified by protein A affinity-chromatography (Pierce) and used to develop reliable and sensitive endlabeling of free 3Ј-OH groups of newly-generated Western blots at 0.1–0.3 ␮g/ml. Other Abs were: monoclonal mouse anti- nucleosomal DNA (21). Cells from each well were pelleted at 200 ϫ g for

human Bax (clone 2D2, IgG1) and monoclonal mouse anti-Bcl-xL (clone 5 min at 4°C, resuspended in 0.5 ml of phosphate-buffered 4% formalde- 2H12, IgG2a) (Zymed, South San Francisco, CA), anti-human Bad (clone hyde, kept at room temperature for 10 min, repelleted, resuspended in 150 P3F6), anti-human Bak (G317–2), anti-Fas, anti-CD2, and anti-Bcl-2 ␮l of 80% ethanol, and immobilized and dried on poly-L-lysine precoated (PharMingen, Inc., San Diego, CA), anti-CD3 (Caltag Laboratories, South glass slides. Each slide was rehydrated in 20 mM Tris-130 mM NaCl, pH San Francisco, CA), and anti-CD28 (Bristol-Myers Squibb Pharmaceutical 7.6, and endogenous peroxidases were inactivated by treatment with 3%

Research Institute, Seattle, WA). Rabbit anti-Edg-2 Ab was a gift from H2O2 in 90% methanol for 5 min at room temperature before endlabeling Dr. Jerold Chun (University of California, San Diego). according to the procedures described in instructions for the Klenow- FragEL kit (Oncogene Research Products–Calbiochem, La Jolla, CA). Per- Cell culture and transfection centage apoptosis was calculated from the number of Tsup-1 cells with Human CD4ϩ8ϩ3low T lymphoblasts of the Tsup-1 line (19) were cultured stained nuclei of a total of 200 counted. Omission of Klenow fragment in RPMI 1640 medium (University of California at San Francisco Cell permitted assessment of those with residual endogenous peroxidase activ- Culture Facility) containing 10% (v:v) FBS, 100 U/ml of penicillin G, 100 ity, which never exceeded 1%. ϫ 7 ␮g/ml of streptomycin, and 1 mM ␤-mercaptoethanol (hereafter referred to For the radioactive assessment of DNA fragmentation, 0.5 10 as complete RPMI medium) at 37°C with 5% CO in air. Complete RPMI Tsup-1 cells/ml of complete RPMI medium with 0.5% FBS were incubated 2 ␮ 3 medium was added to cultures every 2–3 days to maintain a density of with 20 Ci/ml of [ H]thymidine (DuPont–New England Nuclear, Boston, 0.5–1 ϫ 106 Tsup-1 cells/ml. For all studies of effects of LPA and S1P, MA)for2hat37°C and washed twice in complete RPMI and twice in Downloaded from batches of 3–5 ϫ 107 Tsup-1 cells were conditioned in 30–50 ml of FBS-free RPMI before addition to 24-well plates as for the Klenow-Fra- RPMI/1% FBS for 24 h and RPMI/0.1% FBS for a minimum of 12 h. gEL assay. The processing of radioactively labeled Tsup-1 cell suspensions Transfections of replicate suspensions of 4 ϫ 106 Tsup-1 cells in 2 ml of was conducted as described (22), and percentage fragmentation of DNA ϩ ϩ ϩ ϫ RPMI/2% FBS were conducted by dropwise addition of a 250-␮l prein- was calculated by the formula: (S E)/(S E P) 100%, where ϭ ϭ cubated mixture of 5 ␮g of antisense plasmid DNA, 0.2 ␮g of DNA of the radioactivity was determined for S supernatant in the original well, E ϭ REP 4 plasmid (Invitrogen, San Diego, CA) encoding hygromycin-resis- Tris-EDTA-Triton extract of cell pellet, and P pellet dissolved in SDS. tance, and 15 ␮l of FuGENE 6 nonliposomal lipofection reagent (Boehr- After 16 h of incubation, the release of radioactivity without stimulation Ϯ Ϯ ϭ inger-Mannheim, Indianapolis, IN), incubation for 16–24 h, and washing was a mean ( SD) of 5.2 2.9% (n 12). The concomitants of apoptosis http://www.jimmunol.org/ once and incubation in 4 ml of RPMI/10% FBS with 800 ␮g/ml of hy- also were verified in some studies by fluorescence microscopic detection gromycin for seven additional days. Then the surviving transfectants were of characteristic morphological features of Tsup-1 cells stained with bis- ␮ washed and cultured in 4 ml of RPMI/0.1% FBS for 16 h. Antisense plas- benzimide trihydrochloride (Hoechst 33258; 10 g/ml in glycerol:PBS, mids containing full-length cDNA encoding Edg-2, Edg-3, Edg-4, and 30v:70v). Edg-5 surface receptors (Rs) in the reverse orientation relative to promoters Western blots were constructed in the expression vectors pRc/CMV2 for Edg-2, pcDNA 3.1 for Edg-3 and Edg-4 (Invitrogen, Carlbad, CA), and pSV.SPORT1 for Replicate suspensions of 1 ϫ 107 Tsup-1 cells, which had been incubated Edg-5 (Life Technologies, Gaithersburg, MD). with an inducer of apoptosis and/or LPA or S1P for 16 h, were washed three times with 10 ml of cold PBS, resuspended in 200 ␮l of cold 100 mM RT-PCR analysis of Edg protein receptors NaCl-50 mM Tris-HCl, pH 7.4, containing 5 mM EDTA, 1 mM PMSF, 20 ␮ by guest on September 26, 2021 Total cellular RNA was extracted from suspensions of Tsup-1 cells by the g/ml of leupeptin, 0.1 mM DL-thiorphan, 5% glycerol (v:v), and 1% TRIzol method (Life Technologies, Grand Island, NY), and a Superscript kit Nonidet P-40. After homogenization with a Teflon pestle on ice for 2 min ϫ (Life Technologies) was used for reverse transcription (RT) synthesis of at 250 rpm, each sample was centrifuged at 5000 g for 10 min at 4°C, ␮ Ϫ cDNAs. PCR began with a “hot start” at 94°C for 3 min, Taq DNA polymerase and the supernatant was divided into 20 l aliquots and stored at 70°C. ϫ ␮ was added, and amplification was conducted with 35 cycles of 30 s at 94°C, Aliquots of 5000 g supernatant containing 1–50 g of protein were ϫ 2 min at 55°C, and 1 min at 72°C. A total of 2 ␮Ci of [␣-32P]dCTP was added mixed with 4 Laemmli’s solution, heated to 100°C for 3 min, and elec- to some sets of reaction mixtures to allow quantification of mRNA encoding trophoresed in an SDS-12% polyacrylamide gel for 20 min at 100 v and 1 h each Edg receptor relative to that of the standard glyceraldehyde 3- phosphate at 140 v, along with a rainbow prestained set of molecular mass markers dehydrogenase (G3PDH) (20). Oligonucleotide primer pairs were: 5Ј-dCCT (DuPont–New England Nuclear or Amersham, Arlington Heights, IL). Pro- GGCCAAGGTCATCCATGACAAC and 5Ј-dTGTCATACCAGGAAAT teins in each gel were transferred electrophoretically to a nitrocellulose GAGCTTGAC for G3PDH; 5Ј-CTACACAAAAAGCTTGGATCACTCA membrane (Hybond; Amersham) for sequential incubation with 5 g/100 ml and 5Ј-CGACCAAGTCTAGAGCGCTTCCGGT for Edg-1 (1100 base pairs reconstituted nonfat milk powder for blocking unspecific sites, dilutions of (bp)); 5Ј-dGCTCCACACACGGATGAGCAACC and 5Ј-GTGGTCATT monoclonal mouse anti-Bax, anti-Bad, anti-Bak, anti-Bcl-xL or anti-Bcl-2, GCTGTGAACTCCAGC for Edg-2 (621 bp); 5Ј-dCAAAATGAGGCCT and then horseradish peroxidase-labeled goat anti-mouse IgG, before de- TACGACGCCA and 5Ј-dTCCCATTCTGAAGTGCTGCGTTC for Edg-3 velopment with a standard ECL kit (Amersham). (701 bp); 5Ј-dAGCTGCACAGCCGCCTGCCCCGT and 5Ј-dTGCTGTGC ELISA quantification of Bcl-2 CATGCCAGACCTTGTC for Edg-4 (775 bp); 5Ј-CTCTCTACGCCAAG CATTATGTGCT and 5Ј-ATCTAGACCCTCAGACCACCGTGTTGCCC Replicate suspensions of 1 ϫ 106 Tsup-1 cells, which had been incubated TC for Edg-5 (500 bp); 5Ј-dAGTCCTCAAATCATCCCACATCTGC and 5Ј- with an inducer of apoptosis and/or LPA or S1P for 16 h, were transferred dAAGTGGCACTTCCTGTCTCGTAATC for the type I vasoactive intestinal to 1.5 ml polypropylene Eppendorf tubes, pelleted at 400 ϫ g for 5 min at peptide receptor (VIPR1); and 5Ј-dTCCCAGCAGGTGTTTCCTGGCCTAC 4°C, and resuspended in 200 ␮l of 50 mM Tris-HCl, pH 7.4, containing 5 and 5Ј-dCGAGCCTCTTGTACTGTGACTGGTC for VIPR2. PCR products mM EDTA, 0.2 mM PMSF, 10 ␮g/ml of leupeptin, and 1 ␮g/ml of pep- were resolved by electrophoresis in a 2 g/100 ml agarose gel with ethidium statin. Then, 40 ␮l of Ag extraction detergent solution (Calbiochem) was bromide staining. G3PDH, VIPR, and Edg receptor bands were cut from gels added to each sample followed by 30 min of incubation on ice with vig- and solubilized for ␤-scintillation counting in 0.5 ml of sodium perchlorate orous mixing every 5 min and centrifugation at 3000 ϫ g for 10 min at 4°C. solution at 55°C for 1 h (EluQuick; Schleicher and Schuell, Keene, NH). Bcl-2 levels in duplicate 50-␮l portions of 1:3 dilutions of the 3000 ϫ g Initially, the G3PDH cDNA templates in several different-sized portions of supernatant for each sample were measured by an ELISA kit (Oncogene each sample were amplified to determine volumes that would result in G3PDH Research Products–Calbiochem) according to the procedures described. bands of equal intensity for each sample. Relative quantities of cDNA encod- ELISA plates were read at 450/540 nm by a Molecular Devices system ing each Edg receptor also were calculated by the ratio of radioactivity to that (Menlo Park, CA). in the corresponding G3PDH band (20). Results Induction and assessment of apoptosis Tsup-1 cell expression of Edg receptors After conditioning at low serum concentrations, replicate suspensions of ϩ ϩ low 5 ϫ 105 Tsup-1 cells in 0.5 ml of RPMI/0.1% FBS were incubated in The Tsup-1 line of human CD4 8 3 T lymphoblastoma cells is a useful model for studies of the regulation of human T cell apo- 24-well plastic plates (Falcon, Oxnard, CA) for 16 h at 37°C in 5% CO2 in air. Some wells were precoated overnight at 4°C with 30 ng of anti-Fas Ab, ptosis induced by different immunologically relevant stimuli (22). The Journal of Immunology 2051

receptor mRNA determined by RT-PCR and radioactive PCR was increased by each of the apoptosis-inducing Abs, but was un- changed by C6 ceramide (Fig. 1C and Table II).

Protection of Tsup-1 cells from apoptosis by LPA and S1P Both LPA and S1P prevented apoptosis induced by anti-Fas Ab and a combination of anti-CD3 and anti-CD28 Abs with differences only in lipid concentration dependence (Fig. 2). For anti-Fas Ab-stimulated apoptosis, LPA and S1P suppression were significant at 10Ϫ10 M and reached plateau levels of Ͼ80% inhibition by 10Ϫ9 Mto10Ϫ7 M. At 10Ϫ6 M, the maximal level of protection by S1P was maintained, whereas that observed with lower concentrations of LPA was lost completely. The net mean level of apoptosis evoked by anti-CD2 Ab was 20% (n ϭ 2), and this also was inhibited 94% by 10Ϫ8 M LPA, 71% by 10Ϫ8 M S1P, and 82% by 10Ϫ6 M S1P. As for apoptosis induced by anti-Fas Ab, that evoked by anti-CD2 Ab was not decreased by 10Ϫ6 M LPA. Similarly, 10Ϫ10 Mto10Ϫ7 M LPA and S1P significantly prevented apopto- Downloaded from sis induced by the combination of anti-CD3 and anti-CD28 Abs (Fig. 2). Again the protective effect of S1P was maintained at 10Ϫ6 M, whereas 10Ϫ6 M LPA did not prevent apoptosis. FIGURE 1. Detection of Edg receptor mRNA in Tsup-1 cells by RT- The radioactive assay of DNA fragmentation gave similar re- PCR. A, Equal aliquots of the RT cDNA product were amplified by sets of sults, with a mean background level of 4.6% that was increased to primers for Edg-1, Edg-2, Edg-3, Edg-4, and Edg-5, and concurrently for means of 34, 27, and 29%, respectively, by anti-Fas, anti-CD2, and

Ϫ http://www.jimmunol.org/ G3PDH. Each PCR product was electrophoresed in an agarose gel: lane 1, anti-CD3 plus anti-CD28 Abs without LPA or S1P. At 10 8 M, Edg-1; lane 2, Edg-2; lane 3, Edg-3; lane 4, Edg-4; and lane 5, Edg-5. LPA and S1P inhibited radioactively quantiﬁed apoptosis induced Edg-2 (B), Edg-3 (C), and Edg-4 (D) RT-PCR cDNA products were pre- by these three sets of Abs by means of 84, 94, and 73% and 87, 91, pared from RNA of equal numbers of Tsup-1 cells, which had been incu- and 79%, respectively. In contrast, in the same two studies 10Ϫ6 M bated with different apoptotic stimuli. B, C, and D, lane 1, medium alone; LPA only altered control levels of apoptosis by 5, Ϫ7, and Ϫ11%, lane 2, anti-Fas Ab; lane 3, anti-CD2 Ab; lane 4, combination of anti-CD3 Ϫ which all were insigniﬁcant, whereas 10 6 M S1P attained signif- and anti-CD28 Abs; and lane 5,5␮M C6 ceramide. The number below each lane is the ratio of 32P radioactivity in that Edg receptor RT-PCR icant levels of inhibition of 90, 88, and 82%, respectively. product to the 32P in the concurrent G3PDH product. Loss of protection from apoptosis at a high LPA concentration Ϫ6

was in part attributable to induction of apoptosis by 10 M LPA by guest on September 26, 2021 alone, which attained a mean level (ϮSD) of 4.7 Ϯ 0.9-fold higher Tsup-1 cells also bear Rs for many endogenous mediators that than the mean level of 3.7% for control Tsup-1 cells in medium, as influence thymocyte and T cell apoptosis, including prostaglandins contrasted with an increase of only 2.3 Ϯ 0.5-fold for 10Ϫ6 M S1P. and neuropeptides (22–24). The application of RT-semiquantita- Although this higher background of apoptosis for 10Ϫ6 M LPA tive PCR revealed expression of Edg-2, Edg-3, and Edg-4 Rs (Fig. was subtracted from the levels of apoptosis evoked by defined 1A). In this and two additional assays, a radioactive modification stimuli in the presence of 10Ϫ6 M LPA, LPA injury and the effects of PCR permitted assessment of the relative quantity of mRNA of anti-Fas and other apoptosis-inducing Abs appear to have been encoding each Edg receptor compared with that for G3PDH in more than additive. The levels of apoptosis attained means of 2.3-, unstimulated Tsup-1 cells (Table I). Radioactive PCR confirmed 2.7-, and 3.0-fold higher than medium-alone control at 10Ϫ9 M, the known predominant expression of VIPR2 and only marginally 10Ϫ8 M, and 10Ϫ7 M LPA, respectively, and 1.3-, 1.5-, and 1.6- detectable mRNA for VIPR1. The levels of mRNA encoding fold higher than medium at 10Ϫ9 M, 10Ϫ8 M, and 10Ϫ7 M S1P, Edg-2, Edg-3, and Edg-4 Rs were as high as or higher than that for which were significantly higher only for 10Ϫ7 M LPA and did not VIPR2 (n ϭ 3). In contrast, the amounts of mRNA encoding Edg-1 influence the profound protective effects of these concentrations of and Edg-5 Rs were just at the level of detection and less than half the lipids. In contrast to Ab-induced apoptosis, that evoked by C6 that of Edg-3 receptor, respectively (Fig. 1 and Table I). Of the ceramide was suppressed significantly by S1P, but not by either stimuli used to induce apoptosis in Tsup-1 cells, neither anti-Fas concentration of LPA (Fig. 2). Ab nor Abs to other surface protein Ags altered the levels of The capacities of LPA and S1P to prevent activation-induced mRNA encoding Edg-2 or Edg-4 Rs (Fig. 1, B and D, and Table apoptosis of Tsup-1 cells were specific for these lipids and did not II). In contrast, a concentration of C6 ceramide that evoked max- extend to structurally related compounds (Fig. 3). LPA at 10Ϫ7 M imal apoptosis reduced the apparent levels of Edg-2 and Edg-4 again evoked a modest, but significant, increase in background receptor mRNA (Fig. 1, B and D, and Table II). The level of Edg-3 apoptosis, and 10Ϫ7 M S1P in this study also enhanced apoptosis

Table I. Relative frequency of expression of principal Edg receptors by Tsup-1 cellsa

Receptor

VIPR1 VIPR2 Edg-1 Edg-2 Edg-3 Edg-4 Edg-5

VIPR/EdgR:G3PDH 32P 0.03 0.28 0.06 0.46 0.28 0.76 0.12

Table II. Effects of apoptosis-inducing stimuli on Tsup-1 cell expression of Edg receptorsa

Stimulus

Anti-CD3 ϩ Medium Anti-Fas Anti-CD2 anti-CD28 C6 ceramide

Edg-2:G3PDH 32P 0.47 0.39 0.31 0.35 0.05 Edg-3:G3PDH 32P 0.16 0.52 0.48 0.52 0.17 Edg-4:G3PDH 32P 0.63 0.57 ND 0.71 0.12

a All values are the mean of three RT-PCR determinations of mRNA-encoding Tsup-1 cell Edg receptors, of which only one is illustrated in Fig. 1. Each number shown is the ratio of 32P in the Edg receptor cDNA band to that in the G3PDH band. ND, not done. slightly and with borderline signiﬁcance. LPA and S1P, but not signiﬁcant protection from apoptosis, implies other functionally PA, S, or PSSP, prevented anti-Fas Ab induced apoptosis. S1P relevant effects of high concentrations of S1P. suppressed apoptosis elicited by C6 ceramide, but this protection was not mimicked by S or PSSP, and LPA again lacked protective Lack of effect of LPA or S1P on Bcl-xL, Bcl-2, Bad, or Bak activity against this stimulus (Fig. 3). Western blot analyses of equal amounts of protein from extracts of Tsup-1 cells that had been incubated with 10Ϫ7 M LPA or 10Ϫ7 M Downloaded from S1P alone or with anti-Fas Ab with and without 10Ϫ8 M and Suppression of Bax by LPA and S1P 10Ϫ6 M LPA and S1P failed to show any changes in cellular con-

Possible mechanisms by which LPA and S1P protect Tsup-1 tent of Bcl-xL for any of the samples (Fig. 4B). ELISA assays of cells from activation-induced apoptosis were examined first by extracts of Tsup-1 cells from similar experimental protocols also quantifying any associated changes in cellular content of pro- did not show significant changes in cellular levels of Bcl-2 (Table teins that enhance or inhibit apoptosis. Western blot analyses of IV). Western blot analyses of Tsup-1 cellular levels of the apop- http://www.jimmunol.org/ Bax extracted from anti-Fas Ab-treated Tsup-1 cells showed tosis-promoting proteins Bad and Bak did not reveal any alter- that incubation with 10Ϫ8 M LPA or 10Ϫ8 M S1P decreased the ations in association with exposures to LPA or S1P alone nor to level of Bax by Ͼ80% and Ͼ50%, respectively, as assessed by LPA or S1P in the presence of anti-Fas. dilutions of the positive control sample (Fig. 4A). In contrast, neither 10Ϫ6 M LPA nor 10Ϫ6 M S1P consistently affected Bax Reduction in effects of LPA and S1P on apoptosis by antisense levels in anti-Fas Ab-treated Tsup-1 cells despite the protective suppression of expression of Edg receptors effect of 10Ϫ6 M S1P. Densitometric evaluation of the results of The application of newly developed mouse mAbs to Edg-3, Edg-4, Western blots designed to define the concentration dependence and Edg-5 and a polyclonal Ab to Edg-2 demonstrated expression of LPA and S1P effects on Bax levels in Tsup-1 cells revealed of the respective 48- to 55-kDa receptor proteins by Tsup-1 cells by guest on September 26, 2021 significant suppression at 10Ϫ9 Mto10Ϫ7 M and maximal sup- (Fig. 5). Transfection of Tsup-1 cells with ligand-directed pairs of pression of Bax at 10Ϫ8 M for both lysophospholipids (Table plasmids encoding antisense messages for either the LPA Rs III). The lack of suppression of Bax by 10Ϫ6 M S1P, despite Edg-2 and Edg-4 or the S1P Rs Edg-3 and Edg-5, as well as the

FIGURE 2. Stimulus-dependence of suppression of apoptosis by LPA and S1P. Apoptosis was quantiﬁed by the Klenow-FragEL endlabeling kit method. Results with LPA or S1P were calculated as a percentage of the mean level of apoptosis in medium without either lipid (100%). Each bar and bracket is the mean Ϯ SD of the results of three different studies performed in duplicate. A paired Student’s t test was used to assess the level of signiﬁcance of the p Ͻ 0.01. Control levels of apoptosis (mean Ϯ SD) were 3.7 Ϯ ,ء ;difference between each mean value and that of the medium alone control: ϩ, p Ͻ 0.05 2.2% for medium alone, 29 Ϯ 4.1% for anti-Fas Ab, 34 Ϯ 9.6% for anti-CD3 and anti-CD28 Abs combined, and 17 Ϯ 3.5% for 5 ␮M C6 ceramide. The Journal of Immunology 2053

FIGURE 3. Lipid-specificity of the apoptotic effects of LPA and S1P. Quantification of apoptosis and calculation of results were the same as in Fig. 2. Downloaded from Each bar is the mean of the results of two different studies performed in triplicate. A paired Student’s t test was used to assess the level of significance of p Ͻ 0.01. Control levels of apoptosis (mean Ϯ SD) were ,ء ;the difference between each mean value and that of the medium alone control: ϩ, p Ͻ 0.05 7.9 Ϯ 3.7% for medium alone, 61 Ϯ 13% for anti-Fas Ab, and 48 Ϯ 10% for 5 ␮M C6 ceramide.

REP 4 hygromycin-resistance plasmid, followed by 7 days of se- and Edg-4 antisense plasmids or Edg-3 and Edg-5 antisense plas- http://www.jimmunol.org/ lection in hygromycin, selectively reduced respective levels of the mids, followed by hygromycin selection, compared with effects on targeted Edg receptor proteins detectable in Western blots, with apoptosis of sham transfectants, also were evaluated functionally. little or no suppression of those of the complementary set speciﬁc At 10Ϫ9 M and 10Ϫ8 M, LPA and S1P characteristically protected for the other ligand (Fig. 5). Sham-transfected Tsup-1 cells, which Tsup-1 cell sham transfectants from anti-Fas Ab-induced apoptosis received only empty vectors, responded to 10Ϫ8 M LPA or S1P (Table V). Protection from anti-Fas Ab-induced apoptosis by LPA with respective reductions of ϳ85 and ϳ90% in the level of im- was signiﬁcantly less in Tsup-1 cells transfected with Edg-2 and munoreactive Bax protein observed after incubation with anti-Fas Edg-4 antisense plasmids, without a change in the effectiveness of Ab alone, as assessed with dilutions of the latter sample (Fig. 6). S1P. Protection from anti-Fas Ab-induced apoptosis by S1P was

In contrast, transfection with a mixture of expression plasmids significantly less in Tsup-1 cells transfected with Edg-3 and Edg-5 by guest on September 26, 2021 encoding Edg-2 and Edg-4 antisense messages prevented the de- antisense plasmids, without a change for LPA (Table V). crease in Bax elicited by LPA, but not by S1P (Fig. 6). In the reciprocal study, transfection with a mixture of expression plas- Discussion mids encoding Edg-3 and Edg-5 antisense messages prevented the The ability of LPA and S1P to suppress activation-induced apo- decreases in Bax elicited by S1P, but not by LPA. The effects of ptosis of Tsup-1 lymphoblastoma cells is dependent on cellular LPA and S1P on apoptosis of Tsup-1 cells transfected with Edg-2 expression of the receptors for both lipids and on receptor transduction of alterations in specific mediators or inhibitors of apoptosis. Tsup-1 cells expressed mRNA encoding predominantly Edg-2, Edg-3, and Edg-4 Rs (Fig. 1), which are the most promi- nent members of the subfamily found on lymphoid cells and other leukocytes of the immune system (11, 12, 14). Radioactive RT- PCR semiquantification of mRNA specific for each of these Edg

Table III. Suppression of Tsup-1 cell Bax by LPA and S1Pa

FIGURE 4. Western blot analyses of the effects of LPA and S1P on Mean % Control apoptosis-regulating proteins. Proteins extracted from Tsup-1 cells after Concentration Ϫlog M Density incubation with anti-Fas Ab and/or LPA or S1P were resolved by electrophoresis in a 12% polyacrylamide-SDS gel and developed with 1 ␮g/ml of LPA 10 84 the respective mAbs. A, Bax protein. Each lane received 5 ␮g of total 953 proteins from cells treated with: anti-Fas Ab alone (lane 1), anti-Fas Ab 826 with 10Ϫ8 M S1P (lane 2), anti-Fas Ab with 10Ϫ6 M S1P (lane 3), anti-Fas 735 Ϫ Ϫ 691 Ab with 10 8 M LPA (lane 4), anti-Fas Ab with 10 6 M LPA (lane 5), 1/2 dilution of anti-Fas Ab alone (sample 1; lane 6). B, Bcl-xL protein. Each S1P 10 77 ␮ lane received 10 g of total proteins from cells treated with: medium alone 942 Ϫ7 Ϫ7 (control; lane 1), 10 M LPA (lane 2), 10 M S1P (lane 3), anti-Fas Ab 813 Ϫ Ϫ Ϫ alone (lane 4), anti-Fas Ab with 10 8 M LPA, 10 6 M LPA, 10 8 M S1P, 725 and 10Ϫ6 M S1P (lanes 5–8, respectively), and a duplicate of medium 688 control (sample 1; lane 9). Prestained protein molecular mass standards a Each value is the mean of results of three Western blots for which Bax immu- were carbonic anhydrase (30 kDa) and trypsin inhibitor (21.5 kDa; noreactive protein was quantiﬁed by densitometry with an analytical system that has Amersham). been described (23, 24). 2054 LYSOPHOSPHOLIPID SUPPRESSION OF T CELL APOPTOSIS

Table IV. Lack of effect of LPA and S1P on Tsup-1 cell levels of immunoreactive Bcl-2a

Medium Anti-Fas Anti-CD2 Anti-CD3 ϩ Concentration Ϫlog M Alone Ab Ab Anti-CD28 C6 Ceramide

LPA 9 86 98 ND 85 ND 8 84 101 90 85 105 7 80 115 ND 109 ND 6 119 94 84 70* 88

S1P 9 94 97 ND 74 ND 8 97 93 88 94 120 7 103 104 ND 81 ND 68196858893

a Each level of Bcl-2 presented is the mean percentage of the concentration in Tsup-1 cells exposed to the same apoptosis-inducing stimulus or medium alone without LPA or S1P (100%). Mean values were calculated from the results of ELISA determinations of Bcl-2 extracted from duplicate samples of Tsup-1 cells in two separate studies. *, One level was signiﬁcantly different (p Ͻ 0.05) from that of the control exposed to both anti-CD3 Ab and anti-CD28 Ab without LPA or S1P.

Rs (Table I) showed relative levels similar to or higher than those apoptosis. The first was a loss of protection by LPA but not S1P at Downloaded from found for the VIPR2, which has a mean density of 89,500 recep- the highest concentration examined of 10Ϫ6 M, which was attrib- tors/Tsup-1 cell (23). The level of mRNA-encoding Edg-5 receptor uted in part to direct injury of Tsup-1 cells by this level of LPA but was much lower and that for Edg-1 receptor was undetectable not S1P. The second was lack of protection by LPA from ceram- (Table I). The conditions and stimuli used to induce apoptosis of ide-induced apoptosis (Figs. 2 and 3), which appeared to correlate Tsup-1 cells either had no effect or increased the levels of Edg-2, with the suppression of expression of the Edg-2 and Edg-4 Rs for Edg-3, and Edg-4 R mRNA, with the exception of C6 ceramide LPA by C6 ceramide (Fig. 1 and Table II). In contrast, S1P pro- http://www.jimmunol.org/ (Fig. 1 and Table II). Ceramide strikingly reduced the levels of tected Tsup-1 cells from apoptosis induced by ceramide to the mRNA encoding the Edg-2 and Edg-4 Rs for LPA, without af- same extent as it afforded protection from apoptosis evoked by the fecting mRNA encoding the Edg-3 R for S1P. The recent produc- other stimuli (Figs. 2 and 3), and ceramide did not suppress ex- tion of mAbs specific for Edg-3, Edg-4, and Edg-5 Rs and a poly- pression of the Edg-3 receptor for S1P below the level detected on clonal anti-Edg-2 antiserum permitted confirmation but not unstimulated Tsup-1 cells (Fig. 1 and Table II). Although exoge- accurate quantification of expression of these Edg Rs by Tsup-1 nous S is converted to S1P in many types of cells and PA is sim- cells (Fig. 5). ilarly metabolized to LPA (1–3), neither S nor PA exhibited the

The abilities of LPA and S1P to potently suppress activation- protective effects of LPA and S1P from apoptosis induced by any by guest on September 26, 2021 induced apoptosis of Tsup-1 cells were demonstrated by a highly of the agonists employed (Fig. 3). reproducible primary assay, which quantifies newly-generated Initial investigations of possible mechanisms by which LPA and 3Ј-OH groups of nucleosomal fragments of DNA (Figs. 2 and 3), S1P protect Tsup-1 cells from apoptosis employed anti-Fas Abs as and were confirmed by quantification of the release of radioactive the inducing agent, because its effect was suppressed by both lipids prelabeled fragments of DNA. Both assays revealed elements of with similar potency and activity. The level of the proapoptotic stimulus specificity of LPA and S1P protection from apoptosis. At 10Ϫ10 Mto10Ϫ7 M, LPA and S1P suppressed with similar con- protein Bax (25) extracted from Tsup-1 cells undergoing apoptosis centration-dependence apoptosis induced by anti-Fas Abs, anti- after anti-Fas Ab stimulation was suppressed by maximally pro- Ϫ9 Ϫ7 Ϫ6 CD3 plus anti-CD28 Abs (Fig. 2), and anti-CD2 Abs. LPA and tective concentrations of 10 Mto10 M, but not by 10 M S1P showed two differences in their patterns of protection from LPA or S1P (Fig. 4 and Table III). Thus, effects of S1P on elements of apoptosis other than Bax must explain its protective effect at 10Ϫ6 M. In contrast, there were no changes in the levels of other

FIGURE 6. Reduction in LPA- and S1P-induced suppression of Tsup-1 cell Bax by antisense-diminished expression of Edg receptors. Proteins extracted from Tsup-1 cell transfectants after incubation with anti-Fas Ab and/or LPA or S1P were resolved by electrophoresis in a 12% polyacrylamide-SDS gel and developed with 1 ␮g/ml of anti-Bax mAb. Each lane FIGURE 5. Western blot analysis of antisense suppression of expres- received 5 ␮g of total proteins from: sham (vector only)-transfected Tsup-1 sion of Edg Rs by Tsup-1 cells. Lanes 1–3 were loaded with 20 ␮g(A), 30 cells incubated with anti-Fas Ab and medium alone (lane 1), 10Ϫ8 M LPA ␮g(B), 10 ␮g(C), and 50 ␮g(D) of total proteins extracted from Tsup-1 (lane 2), and 10Ϫ8 M S1P (lane 3); Edg-2 plus Edg-4 antisense-transfected cells. Proteins were from Tsup-1 cells that had been sham transfected (lane Tsup-1 cells incubated with anti-Fas Ab and medium alone (lane 4), 1), transfected with antisense plasmids for Edg-2 and Edg-4 (lane 2), or 10Ϫ8 M LPA (lane 5), or 10Ϫ8 M S1P (lane 6); and Edg-3 plus Edg-5 transfected with antisense for Edg-3 and Edg-5 (lane 3). Blots A-D were antisense-transfected Tsup-1 cells incubated with anti-Fas Ab and medium developed with rabbit Ab to Edg-2 and mAbs to Edg-3, Edg-4, and Edg-5, alone (lane 7), 10Ϫ8 M LPA (lane 8), or 10Ϫ8 M S1P (lane 9). The respectively. The marginal line shows the position of a prestained 45 kDa prestained protein molecular mass standard was trypsin inhibitor (21.5 marker (chicken OVA; New England Nuclear). kDa; Amersham). The Journal of Immunology 2055

Table V. Prevention of LPA and S1P protection from apoptosis by suggest that LPA- and S1P-induced changes in Bax contribute antisense suppression of Edg Rsa to protection from apoptosis predominantly at lower levels of the lipids. Similar antiapoptotic decreases in cellular levels of Antisense LPA/S1P Mean % Control Bax have been observed only rarely as a consequence of the Pretreatment Concentration Anti-Fas Ab Apoptosis protective effects of some drugs on neural cells (26). A greater Sham LPA (10Ϫ9 M) ϩ 51 understanding of the separate and concerted roles of stimulation Ϫ Sham LPA (10 8 M) ϩ 31 of growth, suppression of Bax, and inhibition of other proapop- Ϫ9 ϩ Sham S1P (10 M) 30 totic activities such as some caspases (17) will require more Sham S1P (10Ϫ8 M) ϩ 18 deﬁnitive dissection of these mechanisms and their interactions. Edg 2 ϩ 4 LPA (10Ϫ9 M) ϩ 69* The recently reported capacity of S1P to inhibit caspases 3, 6, Edg 2 ϩ 4 LPA (10Ϫ8 M) ϩ 61† and 7, but not 8, in the Jurkat line of human T cells, as assessed Ϫ8 Edg 2 ϩ 4 S1P (10 M) ϩ 22 by cleavage of speciﬁc protein substrates (17), suggests the po- Ϫ tential importance of S1P and LPA in regulating activities of Edg 3 ϩ 5 S1P (10 9 M) ϩ 53* Edg 3 ϩ 5 S1P (10Ϫ8 M) ϩ 39* this family of proteases in T cell apoptosis. Thus, caspase in- Ϫ Edg 3 ϩ 5 LPA (10Ϫ8 M) ϩ 33 hibition by 10 6 M S1P may explain the Bax-independent pro-

a Each value is the mean of the results of two analyses corrected for the level in tection from apoptosis by this highest concentration. These ear- medium without anti-Fas Ab and expressed as a percentage of net apoptosis induced lier studies were restricted to S1P and thus should ﬁrst be by anti-Fas Ab in medium alone without LPA or S1P (100%). These control levels of extended to an analysis of effects of LPA on caspases. The data

apoptosis evoked by anti-Fas Ab alone in the two analyses were 28 and 34%. The Downloaded from significance of the effects of transfection with antisense encoding vectors, followed by available suggest differences between LPA and S1P in their hygromycin-selection, as contrasted with empty vector (sham)/selection was calcu- specificity, potency, and mechanisms of antiapoptotic activities. pϽ 0.05; †, p Ͻ 0.01). The prevention of ceramide-induced apoptosis by S1P, but not,ء) lated by a paired Student’s t test LPA, may reflect principally the demonstrated reduction in expression of Edg-2 and Edg-4 Rs for LPA, without parallel changes in S1P Rs. However, other contributing factors may proapoptotic proteins, including Bad and Bak, or in the antiapop- include the known antagonism of S1P and ceramide at sites, http://www.jimmunol.org/ totic proteins Bcl-xL and Bcl-2 (25) in Tsup- 1 cells undergoing such as protein kinase C, that are critical mediators or regula- apoptosis after exposure to anti-Fas Abs (Fig. 4 and Table IV). tors of apoptosis. The dependence of inhibition of activation-induced apoptosis by LPA and S1P on expression of Edg Rs by Tsup-1 cells was ex- Acknowledgments amined next by transfection of ligand-related combinations of antisense plasmids directed to the LPA Rs Edg-2 and Edg-4 and the We thank Bethann Easterly for preparing the illustrations and for editorial S1P Rs Edg-3 and Edg-5. The extent and selectivity of suppression assistance. of Tsup-1 cell content of Edg-2 and Edg-4 and Edg-3 and Edg-5 receptor proteins supported the effectiveness of this antisense ap- References by guest on September 26, 2021 proach (Fig. 5). Two of the primary indicators of the effects of 1. Moolenaar, W. H. 1995. Lysophosphatidic acid, a multifunctional phospholipid LPA and S1P on Tsup-1 cell apoptosis were assessed in parallel messenger. J. Biol. Chem. 270:12949. 2. Spiegel, S., and S. Milstein. 1995. Sphingolipid metabolites: members of a new for cells transfected with antisense plasmids, as contrasted with class of lipid second messengers. J. Membrane Biol. 146:225. those sham transfected with plasmids lacking the antisense inserts. 3. Brindley, D. N., A. Abousalham, Y. Kikuchi, C.-N. Wang, and D. W. Waggoner. The pair of antisense plasmids directed to Edg-2 Rs and Edg-4 Rs 1996. “Cross talk” between the bioactive glycerolipids and sphingolipids in sig- nal transduction. Biochem. Cell Biol. 74:469. for LPA reduced significantly the protective effect of LPA on anti- 4. Fourcade, O., M. F. Simon, C. Viode, N. Rugani, F. Leballe, A. Ragab, Fas Ab-induced apoptosis observed in sham-transfected controls, B. Fournie, L. Sarda, and H. Chap. 1995. Secretory phospholipase A2 generates but had no effect on protection by S1P (Table V). A pair of anti- the novel lipid mediator lysophosphatidic acid in membrane microvesicles shed from activated cells. Cell 80:919. sense plasmids directed to Edg-3 Rs and Edg-5 Rs for S1P reduced 5. Wang, A., R. Loo, Z. Chen, and E. A. Dennis. 1997. Regiospecificity and cata- significantly the protective effect of S1P on anti-Fas Ab-induced lytic triad of lysophospholipase I. J. Biol. Chem. 272:22030. apoptosis observed in sham transfectants, but did not affect the 6. Moolenaar, W. H., O. Kranenburg, F. R. Postma, and G. Zondag. 1997. Lyso- phosphatidic acid: G protein signaling and cellular responses. Curr. Opin. Cell protection afforded by LPA (Table V). Suppression of the level of Biol. 9:168. Ϫ Ϫ Bax induced by 10 8 M LPA or 10 8 M S1P in sham transfectants 7. Gomez-Munoz, A., D. W. Waggoner, L. O’Brien, and D. N. Brindley. 1995. similarly was reduced by combined transfection with antisense Interaction of ceramides, sphingosine, and sphingosine 1-phosphate in regulating DNA synthesis and phospholipase D activity. J. Biol. Chem. 270:26318. plasmids directed to Edg-2 plus Edg-4 and Edg-3 plus Edg-5, re- 8. Wu, J., S. Spiegel, and T. W. Sturgill. 1995. Sphingosine 1-phosphate rapidly spectively (Fig. 6). Thus, LPA and S1P effects on activation-in- activates the mitogen-activated protein kinase pathway by a G protein-dependent duced apoptosis of Tsup-1 cells depend on expression of a relevant mechanism. J. Biol. Chem. 270:11484. 9. Durieux, M. E., and K. R. Lynch. 1993. Signaling properties of lysophosphatidic complement of the Edg Rs specific for each lysophospholipid li- acid. Trends Pharmacol. Sci. 14:249. gand. That antisense reduction in Edg receptor expression reduced 10. Goodemote, K. A., M. E. Mattie, A. Berger, and S. Spiegel. 1995. Involvement homologous ligand protection from apoptosis, and the suppression of a pertussis toxin-sensitive G protein in the mitogenic signaling pathways of sphingosine 1-phosphate. J. Biol. Chem. 270:10272. of Bax in parallel supported a causal relationship. The discrepancy 11. An, S., M. A. Dickens, T. Bleu, O. G. Hallmark, and E. J. Goetzl. 1997. Mo- between protection and Bax reduction by 10Ϫ6 M S1P suggests a lecular cloning of the human Edg-2 protein and its identification as a functional cellular receptor for lysophosphatidic acid. Biochem. Biophys. Res. Commun. dominant role for non-Bax mechanisms. 231: 619. The capacity of LPA and S1P to suppress apoptosis often is 12. An, S., T. Bleu, O. G. Hallmark, and E. J. Goetzl. 1998. Characterization of a linked intimately with their abilities to stimulate cellular pro- novel subtype of human G protein-coupled receptor for lysophosphatidic acid. J. Biol. Chem. 273:7906. liferation, as both effects may share biochemical prerequisites 13. Hecht, J. H., J. A. Weiner, S. R. Post, and J. Chun. 1996. Ventricular zone gene-1 and signaling mechanisms. The present findings implicate al- (vzg-1) encodes a lysophosphatidic acid receptor expressed in neurogenic regions teration in Bax concentration as one of the mechanisms through of the developing cerebral cortex. J. Cell Biol. 135:1071. 14. An, S., T. Bleu, W. Huang, O. G. Hallmark, S. R. Coughlin, and E. J. Goetzl. which LPA and S1P protect some types of cells from apoptosis. 1997. Identification of cDNAs encoding two G protein-coupled receptors for The concentration-dependent relationships (Table III) further lysosphingolipids. FEBS Lett. 417:279. 2056 LYSOPHOSPHOLIPID SUPPRESSION OF T CELL APOPTOSIS

15. Lee, M.-J., J. R. Van Brocklyn, S. Thangada, C. H. Liu, A. R. Hand, tion of neuropeptides and neuropeptide receptors in a murine model of immune R. Menzeleev, S. Spiegel, and T. Hla. 1998. Sphingosine-1-phosphate as a ligand inflammation in lung parenchyma. Am. J. Resp. Cell. Mol. Biol. 16:133. for the G protein-coupled receptor EDG-1. Science 279:1552. 21. Gavriela, Y., Y. Sherman, and S. A. Ben-Sasson. 1992. Identification of pro- 16. Guo, Z., K. Liliom, D. J. Fischer, I. C. Bathurst, L. D. Tomei, M. C. Kiefer, and grammed cell death in situ via specific labeling of nuclear DNA fragmentation. G. Tigyi. 1996. Molecular cloning of a high-affinity receptor for the growth J. Cell Biol. 119:493. factor-like lipid mediator lysophosphatidic acid from Xenopus oocytes. Proc. 22. Goetzl, E. J., S. An, and L. Zeng. 1995. Specific suppression by prostaglandin E2 Natl. Acad. Sci. USA 93:14367. of activation-induced apoptosis of human CD4ϩ8ϩ T lymphoblasts. J. Immunol. 17. Cuvillier, O., D. S. Rosenthal, M. E. Smulson, and S. Spiegel. 1998. Sphingosine 154:1041. 1-phosphate inhibits activation of caspases that cleave poly(ADP-ribose) poly- 23. Leppert, D., S. L. Hauser, J. L. Kishiyama, S. An, L. Zeng, and E. J. Goetzl. 1995. merase and lamins during Fas- and ceramide-mediated apoptosis in Jurkat T Stimulation of matrix metalloproteinase-dependent migration of T cells by eico- lymphocytes. J. Biol. Chem. 273:2910. sanoids. FASEB J. 9:1473. 18. Levine, J. S., J. S. Koh, V. Triaca, and W. Lieberthal. 1997. Lysophosphatidic 24. Xia, M., D. Leppert, S. L. Hauser, S. P. Sreedharan, P. J. Nelson, A. M. Krensky, acid: a novel growth and survival factor for renal proximal tubular cells. and E. J. Goetzl. 1996. Stimulus-specificity of matrix metalloproteinase-depen- Am. J. Physiol. 273:F575. dence of human T cell migration through a model basement membrane. J. Im- 19. Xia, M., S. P. Sreedharan, and E. J. Goetzl. 1996. Predominant expression of type munol. 156:160. II vasoactive intestinal peptide receptors by human T lymphoblastoma cells: 25. Penninger, J. M., and G. Kroemer. 1998. Molecular and cellular mechanisms of transduction of both Ca2ϩ and cAMP signals. J. Clin. Immunol. 16:21. T lymphocyte apoptosis. Adv. Immunol. 68:51. 20. Kaltreider, H. B., S. Ichikawa, P. K. Byrd, D. A. Ingram, J. L. Kishiyama, 26. Paterson, I. A., and W. G. Tatton. 1998. Antiapoptotic actions of monoamine S.P. Sreedharan, M. L. Warnock, J. M. Beck, and E. J. Goetzl. 1996. Upregula- oxidase B inhibitors. Adv. Pharmacol. 42:312. Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021