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

J. Biol. Chem., Vol. 275, Issue 44, 34628-34633, November 3, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/44/34628    most recent M006393200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Davaille, J. Articles by Lotersztajn, S. Search for Related Content PubMed PubMed Citation Articles by Davaille, J. Articles by Lotersztajn, S. Social Bookmarking                      What's this?

Antiproliferative Properties of Sphingosine 1-Phosphate in Human Hepatic Myofibroblasts

A CYCLOOXYGENASE-2 MEDIATED PATHWAY^*

Julien Davaille, Cyrille Gallois, Aïda Habib^§, Liying Li, Ariane Mallat, Jiangchuan Tao, Thierry Levade^¶, and Sophie Lotersztajn

From INSERM U99, Hopital Henri Mondor, Créteil 94010, France, ^§ INSERM U348, Hopital Lariboisière, Paris 75010, France and ^¶ INSERM U466, CHU Rangueil, Toulouse 31403, France

Received for publication, July 18, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Proliferation of hepatic myofibroblasts (hMF) is central for the development of fibrosis during liver injury, and factors that may limit their growth are potential antifibrotic agents. Sphingosine 1-phosphate (S1P) is a bioactive sphingolipid with growth-regulating properties, either via Edg receptors or through intracellular actions. In this study, we examined the effects of S1P on the proliferation of human hMF. Human hMF expressed mRNAs for the S1P receptors Edg1, Edg3, and Edg5. These receptors were functional at nanomolar concentrations and coupled to pertussis toxin-sensitive and -insensitive G proteins, as demonstrated in guanosine 5'-3-O-(thio)triphosphate binding assays. S1P potently inhibited hMF growth (IC₅₀ = 1 µM), in a pertussis toxin-insensitive manner. Analysis of the mechanisms involved in growth inhibition revealed that S1P rapidly increased prostaglandin E₂ production and in turn cAMP, two growth inhibitory messengers for hMF; C₂-ceramide and sphingosine, which inhibited hMF proliferation, did not affect cAMP levels. Production of cAMP by S1P was abolished by NS-398, a selective inhibitor of COX-2. Also, S1P potently induced COX-2 protein expression. Blocking COX-2 by NS-398 blunted the antiproliferative effect of S1P. We conclude that S1P inhibits proliferation of hMF, probably via an intracellular mechanism, through early COX-2-dependent release of prostaglandin E₂ and cAMP, and delayed COX-2 induction. Our results shed light on a novel role for S1P as a growth inhibitory mediator and point out its potential involvement in the negative regulation of liver fibrogenesis.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Liver fibrosis is characterized by increased deposition and altered composition of extracellular matrix during chronic injury. This fibrogenic response is characterized by intense proliferation and accumulation of myofibroblasts that actively synthesize extracellular matrix and proinflammatory cytokines, as demonstrated in experimental models and culture studies (1). Evidence for heterogeneity in the liver myofibroblast population has been provided recently, and it has been described in rat that two populations of myofibroblasts with fibrogenic potential, hepatic stellate cells and hepatic myofibroblasts, accumulate during chronic liver injury (2, 3). Given their greatly enhanced mitogenic properties, identification of agents that may regulate the growth of these cells has been the topic of several recent studies. Accordingly, we have described the growth inhibitory properties of ET-1, TNF-,¹ and C-type natriuretic peptides (4-7) in human hepatic myofibroblasts (hMF). Moreover, we have recently shown that proliferation of hMF is tightly controlled by cyclooxygenases (4-6), the rate-limiting enzymes in the conversion of arachidonic acid into prostaglandins and thromboxane. Indeed, ET-1 and TNF- markedly inhibit proliferation of hMF through a pathway that involves induction of cyclooxygenase-2 (COX-2) and production of prostaglandin E₂ and prostaglandin I₂ (4-6).

Recent lines of evidence suggest that sphingolipids, in addition to being structural constituents of cell membranes, play a key role as signaling molecules. In particular, metabolites of sphingolipids, including ceramide, sphingosine, and sphingosine 1-phosphate (S1P), have recently emerged as a new class of lipid messengers that regulate cell proliferation, differentiation, and survival (8-11). Ceramide can be generated following receptor-coupled activation of sphingomyelinase and has been linked to cell growth arrest and apoptosis (8). In contrast, S1P, a downstream metabolite of ceramide, produced by phosphorylation of sphingosine following activation of sphingosine kinase, is mitogenic and regulates apoptosis in diverse cell types (12). S1P modulates cell function by two distinct mechanisms, either as an intracellular second messenger or by activating the Edg (endothelial differentiation gene) receptors, a newly identified family of G protein-coupled receptors, (9, 10, 12). The S1P second messenger action has been suggested by the observations that sphingosine kinase activity is increased in response to various stimuli, such as oxidized low density lipoproteins (13), or to ligands that bind to various receptors, including tyrosine kinase receptors (12, 13), the TNF receptor (14), or G protein-coupled receptors (15, 16). Intracellular effects of S1P include cell proliferation, positive or negative regulation of apoptosis, expression of adhesion molecules, and cytoskeletal remodeling (12) and are mimicked by micromolar concentrations of the sphingolipid. On the other hand, extracellular actions of S1P are mediated by Edg receptors, which form a large family of at least eight members, five of them being high affinity receptors for S1P (Edg1, Edg3, Edg5, Edg6, and Edg8), and the others which are selective for lysophosphatidic acid (17). Activation of Edg receptors by S1P has been linked to stimulation of K⁺ channels in myocytes, neurite retraction, and cell rounding of neurons and stimulation of cell migration and of proliferation (10, 18, 19).

Although S1P is generally described as a mitogenic agent, a few studies recently showed that S1P also increases cAMP (20, 21), which has antiproliferative properties in many cell types, including hMF (4). This led us to investigate the role of S1P as a possible growth inhibitory factor for these cells. In this study, we report that human hMF express functional S1P receptors. S1P inhibits proliferation of hMF, probably via an intracellular mechanism involving early COX-2-dependent release of PGE₂ and cAMP and delayed COX-2 induction. Our results shed light on a novel role for S1P as a growth inhibitory mediator and point out its potential involvement in the negative regulation of liver fibrogenesis.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Sphingosine 1-phosphate, sphingosine, C₂-ceramide, dihydrosphingosine 1-phosphate, and NS-398 were from Biomol (Tebu, France). Bacterial (Streptomyces sp.) sphingomyelinase was from Sigma (France). Stock solutions of S1P and dihydro-S1P were dissolved in methanol and stored at 80 °C. Freshly made dilutions were performed in 0.4% fatty acid-free bovine serum albumin, following evaporation of methanol, resuspension in 0.4% fatty acid-free bovine serum albumin, and sonication. Stock solutions of C₂-ceramide and sphingosine were dissolved in Me₂SO, and further dilutions were made in Waymouth medium. Fetal calf serum was from JBio Laboratories (France), and pooled human AB-positive serum were supplied by the National Transfusion Center. [methyl-³H]Thymidine (25 Ci/mmol) and [³⁵S]GTPS were from ICN (France). cAMP radioimmunoassay was from Immunotech (France). The protein assay kit was from Bio-Rad. All other chemicals were from Sigma (France). CellTiter 96 AQ_ueous One Solution Cell Proliferation Assay was from Promega.

Cell Isolation and Culture-- Human hMF were obtained by outgrowth of explants prepared from surgical specimens of normal liver, as described previously (22). This procedure was performed in accordance with ethical regulations imposed by the French legislation. Cells were used between the 3rd and 7th passage, and all experiments were performed on hMF made quiescent by a 3-day incubation in serum-free Waymouth medium, unless otherwise indicated.

The myofibroblastic nature of the cells was routinely evaluated by electron microscopy and positivity for smooth muscle -actin by immunohistochemistry, as described previously (22). The cultures were also found to express two markers of rat hepatic myofibroblasts, fibulin-2 and interleukin-6, and not the protease P100, a marker for rat hepatic stellate cells (2).

Reverse Transcription and Amplification by Polymerase Chain Reaction for S1P Receptors-- Total RNA was extracted from confluent quiescent hMF, using RNeasy kit (Qiagen, Promega, France). RT-PCR was performed on 25 ng of RNA using the access RT-PCR System (Promega, WI) and two oligonucleotides, the 3' reverse transcription (RT) primer and the 5' PCR primer. Oligonucleotide primers are as follows: 5', 5'-GGC TGG AAC TGC ATC AGT GCG C-3' and 3', 5'-GAG CAG CGC CAC ATT CTC AGA GC-3' for Edg1; 5', 5'-GTG TTC ATC GCC TGC TGG TCC C-3' and 3', 5'-GCA GGC TGG ATG GGT GAG GC-3' for Edg3; 5', 5'-ATG GGC AGC TTG TAC TCG GAG TAC CTG-3' and 3', 5'-CAC TCG GCA ATG TAC CTG TTT C-3' for Edg5 and the predicted PCR products of 223, 242, and 496 bp, respectively. The RT-PCRs for Edg1, Edg3, and Edg5 were performed as follows: 45 min of reverse transcription at 48 °C, denaturation for 3 min at 94 °C, 5 cycles of "touchdown" (1 min at 94 °C, 1 min at 67 °C, 1 min at 72 °C), and varying cycles of PCR as shown in Fig. 1 (1 min at 94 °C, 1 min at 60 °C, 1 min at 72 °C). PCR products were size-fractionated in a 2% agarose gel and blotted onto Hybond N+ membrane. After a prehybridization in Express Hyb buffer (CLONTECH) for 3 h at 37 °C, the membrane was hybridized in the same buffer overnight at 37 °C with 50 ng of oligonucleotides complementary to sequences within the cDNAs flanked by the PCR primers, labeled with T4 polynucleotide kinase (Edg1 probe, 5'-CCA CGG TCT TCA CTC TGC TTC TG-3'; Edg3 probe, 5'-CAT TGA TGT GGC CTG CAG GG-3'; Edg5 probe, 5'-CAC TCG GCA ATG TAC CTG TTT C-3'). After washing in 2× SSC, 0.1% SDS for 1 h at 37 °C and then for 1 h at 48 °C, membranes were subjected to PhosphorImager analysis (Molecular Dynamics, France).

[³⁵S]GTPS Binding Assay-- Confluent hMF were made quiescent by incubation in Waymouth medium without serum for 48 h. Cells were then washed with ice-cold phosphate-buffered saline (PBS) medium, scraped in buffer A (20 mM Tris, pH 7.4, 500 µM phenylmethylsulfonyl fluoride, 1 µg/ml leupeptin, 1 µg/ml aprotinin, 0.5 µg/ml pepstatin), and homogenized with a Polytron homogenizer. The homogenate was spun 5 min at 5,500 × g, and the pellet was discarded. The supernatant was centrifuged for 40 min at 43,000 × g, and the resulting pellet was resuspended in buffer A and frozen at 80 °C until use. Membranes (10 µg of protein/assay) were incubated for 45 min at 30 °C in buffer B (20 mM Tris, pH 7.4, 5 mM MgCl₂, 1 mM EDTA, 1 mM EGTA, 100 µM phenylmethylsulfonyl fluoride, 1 µg/ml leupeptin, 1 µg/ml aprotinin, 0.5 µg/ml pepstatin) containing 3 µM GDP unless otherwise indicated, 50 pM [³⁵S]GTPS (0.5 µCi/assay), and varying concentrations of ligands. The samples were then rapidly filtered on GF/B glass microfiber filters (Whatman) presoaked in buffer C (20 mM Tris, pH 7.4, 10 mM MgCl₂, 100 mM NaCl, and 1 mM -mercaptoethanol). The filters were washed three times with 2 ml of buffer C and counted. Specific binding was calculated as the difference in bound radioactivity in the absence or presence of 10 µM unlabeled GTPS and did not exceed 20% of the total binding.

Pertussis Toxin Treatment of Human hMF-- Confluent hMF were made quiescent by incubation in Waymouth medium without serum for 48 h. Cells were then treated for 24 h with either 3 or 100 ng/ml PTX or vehicle, and DNA synthesis was measured as described below. Alternatively, membranes were prepared as described above and used in GTPS binding assays.

Cyclic AMP Assay-- Cells were seeded in 24-well plates, grown to confluence, and serum-deprived for 48 h in Waymouth medium. Following preincubation with 0.6 mM isobutylmethylxanthine for 15 min, hMF were stimulated with PBS containing effectors. When indicated, cells were preincubated for 60 min with either the cyclooxygenase-2 inhibitor NS-398 (5 µM) or vehicle (Me₂SO). Samples were processed as described previously (4), and cyclic AMP was assayed by a commercial radioimmunoassay (Immunotech, France).

Prostaglandin Release-- Confluent monolayers in 96-well plates were serum-deprived for 48 h, washed in PBS, and further incubated in Waymouth medium. Following pretreatment with 5 µM NS-398 or vehicle, cells were stimulated with the indicated effectors over various times. Release of PGE₂ was assayed by a specific enzyme immunoassay as described previously (23).

DNA Synthesis and Cell Proliferation Assays-- DNA synthesis was measured in triplicate wells by incorporation of [³H]thymidine, as described previously (7). Confluent hMF were made quiescent by incubation in Waymouth medium without serum for 48 h. Cells were then stimulated for 30 h with human serum following pretreatment for 30 min with NS-398, dexamethasone, or vehicle. [³H]Thymidine (0.5 µCi/well) was added during the last 8 h of incubation.

Cell proliferation was assessed as described previously (7). Briefly, cells were seeded in 96-well plates at low density (5000/well) in Dulbecco's modified Eagle's medium containing 5% human serum and 5% fetal calf serum (5:5), allowed to attach overnight, and made quiescent by a 48-h incubation in serum-free medium. Cells were then incubated for 3 days with 5% human serum in the absence or presence of 10 µM S1P that was renewed daily. The medium was changed for PBS, and CellTiter 96 AQ_ueous One Solution reagent was added to each well, and absorbance was recorded at 490 nm.

Western Blotting Analysis of COX Protein Expression-- Confluent hMF in 12-well plates were made quiescent in serum-free Waymouth medium over 3 days and were then incubated for various times with 10 µM S1P. Whole cell extracts were prepared, and Western blotting analysis was performed as described previously (6).

Statistics-- Results are expressed as mean ± S.E. of n experiments. Results were analyzed by repeated measures analysis of variance or two-tailed Student's t test, as appropriate, with p < 0.05 considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Characterization of Receptors for S1P in Human hMF-- Identification of the S1P receptor subtypes present in human hMF was performed by RT-PCR analysis, with varying PCR cycles. Bands of 223, 242, and 496 bp corresponding to the size of the Edg1, Edg3, and Edg5 products were identified and quantified. The results presented in Fig. 1 indicate that transcripts for the three Edg receptors are present in human hMF (Fig. 1).

View larger version (23K): [in this window] [in a new window]   Fig. 1.   Expression of Edg1, Edg3, and Edg5 mRNAs by human hMF. RT-PCR for Edg1, Edg3, and Edg5 were performed with varying PCR cycles. The PCR products were then size-fractionated and blotted, and the membranes were hybridized with a labeled oligonucleotide complementary to the respective S1P receptor sequences within the cDNA flanked by the PCR primers. Bands with 223, 242, and 496 bp corresponding to the size of the Edg1, Edg3, and Edg5 products were identified and quantified by PhosphorImager analysis.

Coupling of S1P receptors to G proteins was evaluated in [³⁵S]GTPS binding assays, which measure GDP-GTP exchange on the  subunit of the G protein and, therefore, the initial steps of G protein activation by a receptor ligand. Optimal conditions for S1P-induced stimulation of [³⁵S]GTPS binding were defined, in particular with respect to time dependence and to GDP concentration, which is added to keep the G protein in the non-dissociated form. Binding was linear up to 45 min. GDP dose-dependently decreased basal [³⁵S]GTPS binding (Fig. 2, inset), and the concentration required for optimal activation by S1P was in the range of 1-10 µM (not shown). As shown in Fig. 2, S1P dose-dependently increased binding of [³⁵S]GTPS to G proteins in membranes of human hMF, with an EC₅₀ of 5 nM, in agreement with the reported K_d of S1P for its receptors. Sphinganine 1-phosphate (dihydro-S1P), which binds and signals through all S1P receptors with a potency similar to that of S1P (24), was as efficient as S1P in enhancing GTPS binding (Fig. 2A). We also investigated the coupling of S1P receptors to G_i/G_o proteins in human hMF by using pertussis toxin (PTX), which ADP-ribosylates and inactivates the  subunit of G_i/G_o, thereby maintaining G_i/Go in a non-dissociated form. Optimal GDP concentrations required to keep the G proteins in a non-dissociated form were lower in PTX-treated than in control membranes (Fig. 2, inset). This is probably due to the fact that PTX excludes G_i proteins from the pool of G proteins accessible to GDP. Under optimal conditions of GDP, treatment with PTX 3 ng/ml partially abolished the stimulatory effect of S1P on GTPS binding (Fig. 2B).

View larger version (24K): [in this window] [in a new window]   Fig. 2.   S1P and dihydro-S1P enhance GTPS binding in human hMF, via a pertussis toxin-sensitive and -insensitive pathway. A, membranes from human hMF were assayed for [35S]GTPS binding assays as described under "Experimental Procedures," with varying concentrations of S1P and dihydro-S1P, in the presence of 3 µM GDP. Results represent the mean ± S.E. of six experiments and are expressed as percent of control. p < 0.05 compared with control. B, effect of PTX treatment on GTPS binding in response to S1P. Human hMF were treated with 3 ng/ml pertussis toxin or vehicle for 24 h, and membranes were prepared as described under "Experimental Procedures." GTPS binding was assayed in the presence of 3 µM GDP for vehicle membranes and 0.1 µM GDP for pertussis toxin-treated membranes (see inset). Results are the mean of two experiments. Inset, inhibitory effect of GDP on GTPS binding in pertussis toxin-treated and vehicle membranes. [35S]GTPS binding assay was assayed in pertussis toxin-treated and vehicle membranes, with varying concentrations of GDP. Results represent the mean ± S.E. of three experiments.

These results demonstrate that human hMF express receptors for S1P that are functionally coupled to pertussis toxin-sensitive and -insensitive G proteins.

Sphingosine 1-Phosphate Inhibits Proliferation of Human hMF in a Pertussis Toxin-insensitive Manner-- We investigated the effects of S1P and of other sphingolipid metabolites on the proliferation of human hMF. S1P had no mitogenic effects on human hMF (not shown) but dose-dependently reduced serum-stimulated DNA synthesis (Fig. 3A). A maximal 55% inhibition was observed at 10 µM S1P, with an IC₅₀ of 1 µM. The inhibitory effect of S1P was fully reproduced by sphinganine 1-phosphate (dihydro-S1P) (Fig. 3A). However, the IC₅₀ values of S1P and dihydro-S1P were 1 and 1.5 µM, respectively, i.e at least 200 times higher than their EC₅₀ values to stimulate GTPS binding (Fig. 2) and higher than the reported K_d value of S1P for its receptors (9, 10, 12). Direct assessment of cell growth confirmed the antiproliferative effect of S1P, with a 35% inhibition of cell growth being observed after 3 days (Fig. 3B). The growth inhibitory effects of S1P were insensitive to pertussis toxin (Fig. 3C), although pertussis toxin treatment was effective, since it reduced the mitogenic effect of serum. Moreover, pertussis toxin was maximally effective since DNA synthesis in response to serum was reduced to the same extent at 3 and 100 ng/ml of the toxin (Fig. 3C, inset). Finally, hMF DNA synthesis was also inhibited by exogenous addition of other sphingolipids, C₂-ceramide, a permeant analog of ceramide, sphingosine, the immediate precursor/metabolite of S1P, and by bacterial sphingomyelinase, the enzyme that hydrolyzes sphingomyelin to ceramide (Fig. 3D). Concentrations higher than 10 µM C₂-ceramide caused hMF death (not shown).

View larger version (38K): [in this window] [in a new window]   Fig. 3.   S1P and other sphingolipids inhibit the proliferation of human hMF. A, effects of s1P or dihydro-S1P on DNA synthesis of human hMF. Confluent cells were made quiescent by incubation in serum-free medium over 3 days. Cells were stimulated for 30 h with varying concentrations of S1P or dihydro-S1P, in the presence of 5% human serum. [3H]Thymidine incorporation into DNA was measured as described under "Experimental Procedures." Results represent the mean ± S.E. of six experiments and are expressed as percent of control. p < 0.05 compared with control for concentrations over 0.3 µM. B, effect of S1P on hMF growth. Cell growth was assayed at day 0 and day 3 as described under "Experimental Procedures," without or with 5% human serum in the absence or presence of 10 µM S1P or vehicle, which was added every day. Results are the mean ± S.E. of three experiments. p < 0.05 for S1P versus control. C, inhibition of DNA synthesis by S1P is pertussis toxin-insensitive. Confluent quiescent cells were pretreated for 24 h with 3 ng/ml pertussis toxin or vehicle. DNA synthesis was then measured as described in A. Results represent the mean ± S.E. of three experiments. The inhibitory effect of S1P was not significantly different in hMF treated with PTX or vehicle. The inset shows the effects of 3 and 100 ng/ml pertussis toxin on DNA synthesis measured with 5% human serum. D, effects of other sphingolipids on DNA synthesis of hMF. DNA synthesis was assayed as described in A with varying concentrations of either C2-ceramide, sphingomyelinase (in units/ml), or sphingosine. Results represent the mean ± S.E. of six experiments and are expressed as percent of control. p < 0.05 compared with control for concentrations over 0.3 µM lipid or 0.1 units/ml sphingomyelinase.

The next series of experiments were performed to analyze the signaling events that mediate inhibition of hMF proliferation by S1P.

COX-2 Mediates the Growth Inhibitory Effects of Sphingosine 1-Phosphate in Human hMF-- We first examined the effects of S1P on PGE₂ and cAMP levels, two growth inhibitory mediators for human hMF (4-6). Following addition of S1P, cAMP production rose within 5 min, maximally increased to 20-fold by 10 min, declined after 15 min, and returned to basal levels after 60 min (Fig. 4A). Stimulation of cAMP production by S1P was totally blocked by the selective COX-2 inhibitor NS-398 (Fig. 4A) and unaffected by PTX treatment (Fig. 4B). These data suggested that S1P raises cAMP following PGE₂ production by COX-2. Accordingly, S1P caused a rapid 3-5-fold activation of PGE₂ production, which correlated with the time course of cAMP production (Fig. 4C), and was blocked by NS-398 (Fig. 4C, inset). It should be noted that NS-398 decreased basal PGE₂ production, in agreement with our previous data showing that COX-2 is constitutively expressed in hMF (see Fig. 5A) and accounts for basal PGE₂ production in these cells (6). Therefore, these results demonstrate that S1P stimulates an early COX-2-dependent production of PGE₂, leading to cAMP synthesis.

View larger version (34K): [in this window] [in a new window]   Fig. 4.   S1P enhances cyclic AMP production through prostaglandin synthesis in human hMF. A, S1P stimulates cyclic AMP production via a COX-2-dependent pathway. Confluent quiescent cells were pretreated for 1 h with 5 µM NS-398 or vehicle (Me2SO) and were further stimulated with 10 µM S1P for the indicated times. Results are the mean ± S.E. of three experiments and are expressed as fold over basal levels. NS-398, by lowering levels of endogenous prostaglandins, reduced basal cAMP levels from 209 ± 29 to 68 ± 14 fmol/mg. p < 0.05 for NS-398 effect. B, effect of pertussis toxin treatment on the elevation of cAMP by S1P. Confluent quiescent cells were pretreated for 24 h with 3 ng/ml pertussis toxin or vehicle and further stimulated with 10 µM S1P. Results are the mean ± S.E. of three experiments and are expressed as fold over basal levels. p < 0.05 compared with vehicle for S1P and with PTX for S1P + PTX; values for S1P + PTX were not significantly different from values for S1P. C, S1P enhances release of PGE2. Confluent quiescent cells were stimulated with 10 µM S1P over various times. PGE2 release was assayed by enzyme immunoassay as described under "Experimental Procedures." Results show a typical experiment repeated twice. The inset shows the effects of 5 µM NS-398 on PGE2 production. Results are the mean ± S.E. of three experiments and are expressed as fold over basal levels. p < 0.05 for NS-398 effect. D, effects of other sphingolipids on cAMP production. Confluent quiescent cells were exposed for 10 min to 10 µM S1P, 10 µM C2-ceramide (C2-Cer), 1 unit/ml sphingomyelinase (SMase), 10 µM sphingosine (Sph), or the respective concentrations of vehicle (Me2SO for C2-Cer and Sph). Cyclic AMP was extracted and measured as described under "Experimental Procedures." Results are the mean ± S.E. of four experiments and are expressed as fold over basal levels. p < 0.05 compared with control for S1P; values for all other sphingolipids were not significantly different from control.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   S1P induces COX-2 protein expression and activity in human hMF. A, S1P induces COX-2 protein expression but has no significant effect on COX-1. Whole cell extracts obtained from quiescent hMF treated by 10 µM S1P for various times were analyzed by Western blot of COX-2 or COX-1 proteins. Results show a typical experiment repeated three times. B, S1P stimulates PGE2 release. Confluent quiescent cells were pretreated for 60 min with 5 µM NS-398 or vehicle (Me2SO) and were further stimulated with 10 µM S1P over various periods. PGE2 release was assayed by enzyme immunoassay as described under "Experimental Procedures." Results show a typical experiment repeated twice.

The effects of other sphingolipids were assessed on cAMP levels. Sphingosine, which had no significant effect after 10 min of incubation (Fig. 4D), caused only a 3-fold maximal increase in cAMP levels after 30 min (not shown); sphingomyelinase caused a maximal 1.7-fold cAMP elevation only after 60 min of incubation, whereas C₂-ceramide did not affect cAMP production. Therefore, these late effects of sphingosine and sphingomyelinase, observed at delayed incubation time, are probably related to some conversion into S1P.

Since COX-2 is also an inducible enzyme, and given its central role in the inhibition of hMF proliferation (4, 6), we investigated the effect of S1P on COX-2 protein expression. As shown in Fig. 5A, S1P caused a strong induction of COX-2, which was maximal after 3 h and remained elevated for at least 8 h. In contrast, S1P did not significantly affect COX-1 expression. Induction of COX-2 was associated with a 3-fold increase in PGE₂ production, an effect totally blocked by the COX-2 inhibitor NS-398 (Fig. 5B).

Finally, we used NS-398 and dexamethasone, which selectively inhibit COX-2 activity and transcription (6), respectively, in order to determine the role of COX-2 in the growth inhibitory effect of S1P. As shown in Fig. 6, both agents blunted the antiproliferative effect of S1P.

View larger version (31K): [in this window] [in a new window]   Fig. 6.   COX-2 mediates inhibition of hMF proliferation by S1P. Confluent cells were made quiescent by incubation in serum-free medium over 3 days. hMF were pretreated for 60 min with 5 µM NS-398 or 18 h with 0.1 µM of dexamethasone (DEX) or their vehicles, Me2SO and ethanol, respectively, and were further stimulated with 5% human serum, together with 3 or 5 µM S1P. [3H]Thymidine incorporation into DNA was measured as described under "Experimental Procedures." Results represent the mean ± S.E. of four experiments and are expressed as percent of respective control (control, 11,654 ± 4749 cpm; NS-398, 22,761 ± 7629 cpm; dexamethasone, 27,358 ± 2229 cpm). p < 0.05 for NS-398 and dexamethasone effects.

Taken together, these results suggest that S1P inhibits the growth of human hMF by a pathway involving both early PGE₂ and cAMP productions, as well as delayed COX-2 induction.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Proliferation of myofibroblasts is central for the development of liver fibrosis during chronic liver diseases. The present study shows that the growth of human hepatic myofibroblasts is inhibited by S1P and constitutes the first evidence for an antiproliferative effect of this sphingolipid. These data suggest that S1P may play a key role as a negative regulator of liver fibrogenesis.

Data concerning expression and function of Edg receptors in liver are scarce. Hepatic messenger RNAs for Edg1 and Edg3 have been detected (25, 26), but their cell distribution is unknown. Regarding hepatic actions of S1P, a receptor-mediated glycogenolytic effect was described in rat hepatocytes (27). We show here that human hMF express the mRNAs for the S1P receptors Edg1, Edg3, and Edg5. These receptors are functional and coupled to G proteins with an affinity of 5 nM, in keeping with the apparent affinity of S1P for its receptors. In addition, dihydro-S1P, which binds the three S1P receptors with equal affinity (9, 10, 12), also elicits G protein activation, with an affinity similar to that of S1P. Finally, the incomplete effect of pertussis toxin on [³⁵S]GTPS binding indicates that Edg receptors are coupled to both G_i/G_o proteins and to pertussis toxin-insensitive G proteins. These findings are in agreement with the reported exclusive coupling of Edg1 to G proteins of the G_i/G_o family and the association of Edg3 and Edg5 to G_i/G_o and also to G_q, G_s, and G₁₃ (9, 10, 12).

A major point of the present study is that S1P exhibits potent antiproliferative effects in human hMF, which contrasts with its mitogenic properties in many other cells. Whether this growth inhibitory effect of S1P is linked to an intracellular effect or to its plasma membrane receptors remains to determined, although both effects are not mutually exclusive. The intracellular role of S1P has been generally evaluated by using micromolar concentrations of the sphingolipid exogenously applied to cell cultures and confirmed by microinjection of the sphingolipid. Because of its highly lipophilic nature, S1P is taken up by the intracellular compartments and acts as a second messenger, as described for its precursor ceramide. In contrast, the effects linked to activation of S1P receptors occur at nanomolar concentrations of the lipid (9, 10, 12). In human hMF, several lines of evidence argue for an intracellular mechanism for S1P action. First, growth inhibitory effects of S1P and dihydro-S1P are observed at micromolar concentrations, whereas G protein activation is observed at nanomolar concentrations. Second, in a separate study, we obtained evidence suggesting that ET-1 inhibits hMF growth via an intracellular increase of S1P. Indeed, ET-1 stimulates sphingosine kinase in human hMF (28). Moreover, the increase in cAMP elicited by ET-1 is blocked by selective inhibitors of sphingosine kinase, dimethyl sphingosine and DL-threo-dihydrosphingosine, suggesting a second messenger role for S1P in ET-1 action in these cells (28).

Growth inhibitory properties of S1P on human hMF rely on a novel signaling pathway for the sphingolipid that involves both early and delayed production of PGE₂. Rapid production of PGE₂ is ensured by constitutive COX-2, as shown by the dramatic reduction of basal PGE₂ production by the selective COX-2 inhibitor NS-398 (Fig. 4C). These data suggest that, although COX-1 is constitutively expressed in human hMF (Fig. 5), the major part of PGE₂ levels found in basal conditions derives from constitutive COX-2. The rapid increase in PGE₂ production in turn leads to elevation of cAMP, a growth inhibitory mediator that blocks early events involved in hMF proliferation (4). Recent studies indicate that S1P generally decreases cAMP production (9, 10, 12), but S1P-induced cAMP elevation has also been reported in a few instances (20, 29, 30). Interestingly, in human vascular smooth muscle cells, S1P is devoid of mitogenic properties, and stimulates cAMP production (20). Moreover, in these cells, PGE₂ and cAMP show antiproliferative properties (31, 32), but whether S1P could provoke growth arrest was not investigated. Similarly, lysophosphatidic acid, which shares structural homology with S1P and binds to receptors of the Edg family, may also regulate cell growth positively or negatively. Although generally promitogenic (10), lysophosphatidic acid inhibits the growth of myelocytes via a cAMP-dependent pathway (33).

In human hMF, S1P also ensures delayed PGE₂ secretion, following COX-2 induction. These data constitute the first description of a regulation of COX-2 by S1P. Blocking COX-2 by either dexamethasone or NS-398 abrogates the antiproliferative effect of S1P, indicating that COX-2 mediates the growth inhibitory effect of S1P. Our recent studies indicated that activation of COX-2 is a key event in the inhibition of hMF proliferation. Indeed, we found that the antiproliferative effects of ET-1 and TNF- involve COX-2 (6). We also showed that the mitogenic effects of PDGF-BB and thrombin result from a balance between a promitogenic and a COX-2-dependent growth inhibitory pathway (5). Therefore, the present data further support a central role of COX-2 in hMF growth inhibition. According to our previous data, early COX-2-dependent cAMP will block early steps involved in hMF proliferation, such as extracellular signal-regulated kinase and c-Jun NH₂-terminal kinase activations (4). The consequences of COX-2 induction by S1P remain to be determined but may involve regulation of more distal events, such as cell cycle components (34).

In summary, the present study brings out two major findings, (i) the presence of Edg receptors in human hMF and (ii) a novel role for S1P as a growth inhibitory bioactive sphingolipid, via COX-2 activation. At present, our data rather favor the hypothesis of an intracellular antiproliferative action of S1P. In this context, biological functions associated with the presence of Edg receptors in human hMF remain to be explored. Accumulation of hepatic myofibroblasts is central in the development of liver fibrosis, as shown in experimental models of liver injury and in human chronic liver diseases (35). Therefore, our results point to a key role of S1P in the control of liver fibrogenesis. An important issue would be to characterize the distribution of sphingosine kinase in the various liver cell types. Evidence for secretion of S1P by other cells and tissues is limited, probably due to its lipophilic nature and/or a high S1P degradative activity. Nevertheless, S1P release has been detected in activated platelets and in allergically stimulated mast cells, and S1P concentration in serum is around 0.5 µM (36, 37). Whether hepatic production of S1P is altered during chronic liver diseases is under investigation.

	ACKNOWLEDGEMENTS

We thank J Hanoune for permanent support and O. Cuvillier, F. Pecker, C. Pavoine, Y. Laperche, and G. Guellaen for critical reading of the manuscript. We thank C. Feral for help in setting PCR experiments.

	FOOTNOTES

^* This work was supported in part by INSERM, the Université Paris-Val-de-Marne, and by grants from the Association pour la Recherche sur le Cancer (to S. L.) and the Ligue Départementale du Val d'Oise de la Recherche Contre le Cancer (to S. L.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Supported by fellowships from the Ministère de l'Enseignement et de la Recherche. Both authors contributed equally to this work.

To whom correspondence should be addressed: Unité INSERM 99, Hôpital Henri Mondor, 94010 Créteil, France. Tel.: 33-1-49 81 35 34; Fax: 33-1-48 98 09 08; E mail: loterszt@im3.inserm.fr.

Published, JBC Papers in Press, August 14, 2000, DOI 10.1074/jbc.M006393200

	ABBREVIATIONS

The abbreviations used are: TNF, tumor necrosis factor; RXTT-PCR, reverse transcription-polymerase chain reaction; PBS, phosphate-buffered saline; PGE₂, prostaglandin E₂; COX, cyclooxygenase; PTX, pertussis toxin; GTPS, guanosine 5'-3-O-(thio)triphosphate; S1P, sphingosine 1-phosphate; hMF, hepatic myofibroblasts; bp, base pairs.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				H. Hasegawa, Y. Yamada, K. Komiyama, M. Hayashi, M. Ishibashi, T. Sunazuka, T. Izuhara, K. Sugahara, K. Tsuruda, M. Masuda, et al. A novel natural compound, a cycloanthranilylproline derivative (Fuligocandin B), sensitizes leukemia cells to apoptosis induced by tumor necrosis factor related apoptosis-inducing ligand (TRAIL) through 15-deoxy-{Delta}12, 14 prostaglandin J2 production Blood, September 1, 2007; 110(5): 1664 - 1674. [Abstract] [Full Text] [PDF]

				V. Serriere-Lanneau, F. Teixeira-Clerc, L. Li, M. Schippers, W. de Wries, B. Julien, J. Tran-Van-Nhieu, S. Manin, K. Poelstra, J. Chun, et al. The sphingosine 1-phosphate receptor S1P2 triggers hepatic wound healing FASEB J, July 1, 2007; 21(9): 2005 - 2013. [Abstract] [Full Text] [PDF]

				J. Gillibert-Duplantier, V. Neaud, J.-F. Blanc, P. Bioulac-Sage, and J. Rosenbaum Thrombin inhibits migration of human hepatic myofibroblasts Am J Physiol Gastrointest Liver Physiol, July 1, 2007; 293(1): G128 - G136. [Abstract] [Full Text] [PDF]

				S. H. Ki, M. J. Choi, C. H. Lee, and S. G. Kim G{alpha}12 Specifically Regulates COX-2 Induction by Sphingosine 1-Phosphate: ROLE FOR JNK-DEPENDENT UBIQUITINATION AND DEGRADATION OF I{kappa}B{alpha} J. Biol. Chem., January 19, 2007; 282(3): 1938 - 1947. [Abstract] [Full Text] [PDF]

				Y. H. Zeidan, B. J. Pettus, S. Elojeimy, T. Taha, L. M. Obeid, T. Kawamori, J. S. Norris, and Y. A. Hannun Acid Ceramidase but Not Acid Sphingomyelinase Is Required for Tumor Necrosis Factor-{alpha}-induced PGE2 Production J. Biol. Chem., August 25, 2006; 281(34): 24695 - 24703. [Abstract] [Full Text] [PDF]

				S. K. Goparaju, P. S. Jolly, K. R. Watterson, M. Bektas, S. Alvarez, S. Sarkar, L. Mel, I. Ishii, J. Chun, S. Milstien, et al. The S1P2 Receptor Negatively Regulates Platelet-Derived Growth Factor-Induced Motility and Proliferation Mol. Cell. Biol., May 15, 2005; 25(10): 4237 - 4249. [Abstract] [Full Text] [PDF]

				M. Graupera, S. March, P. Engel, J. Rodes, J. Bosch, and J.-C. Garcia-Pagan Sinusoidal endothelial COX-1-derived prostanoids modulate the hepatic vascular tone of cirrhotic rat livers Am J Physiol Gastrointest Liver Physiol, April 1, 2005; 288(4): G763 - G770. [Abstract] [Full Text] [PDF]

				A. Damirin, H. Tomura, M. Komachi, M. Tobo, K. Sato, C. Mogi, H. Nochi, K. Tamoto, and F. Okajima Sphingosine 1-Phosphate Receptors Mediate the Lipid-Induced cAMP Accumulation through Cyclooxygenase-2/Prostaglandin I2 Pathway in Human Coronary Artery Smooth Muscle Cells Mol. Pharmacol., April 1, 2005; 67(4): 1177 - 1185. [Abstract] [Full Text] [PDF]

				J. I. Kim, E. J. Jo, H.-Y. Lee, M. S. Cha, J. K. Min, C. H. Choi, Y. M. Lee, Y.-A. Choi, S.-H. Baek, S. H. Ryu, et al. Sphingosine 1-Phosphate in Amniotic Fluid Modulates Cyclooxygenase-2 Expression in Human Amnion-derived WISH Cells J. Biol. Chem., August 22, 2003; 278(34): 31731 - 31736. [Abstract] [Full Text] [PDF]

				B. J. PETTUS, J. BIELAWSKI, A. M. PORCELLI, D. L. REAMES, K. R. JOHNSON, J. MORROW, C. E. CHALFANT, L. M. OBEID, and Y. A. HANNUN The sphingosine kinase 1/sphingosine-1-phosphate pathway mediates COX-2 induction and PGE2 production in response to TNF-{alpha} FASEB J, August 1, 2003; 17(11): 1411 - 1421. [Abstract] [Full Text] [PDF]

				Y. Jin, E. Knudsen, L. Wang, Y. Bryceson, B. Damaj, S. Gessani, and A. A. Maghazachi Sphingosine 1-phosphate is a novel inhibitor of T-cell proliferation Blood, June 15, 2003; 101(12): 4909 - 4915. [Abstract] [Full Text] [PDF]

				J. Davaille, L. Li, A. Mallat, and S. Lotersztajn Sphingosine 1-Phosphate Triggers Both Apoptotic and Survival Signals for Human Hepatic Myofibroblasts J. Biol. Chem., September 27, 2002; 277(40): 37323 - 37330. [Abstract] [Full Text] [PDF]

				B. Saile, N. Matthes, K. Neubauer, C. Eisenbach, H. El-Armouche, J. Dudas, and G. Ramadori Rat liver myofibroblasts and hepatic stellate cells differ in CD95-mediated apoptosis and response to TNF-alpha Am J Physiol Gastrointest Liver Physiol, August 1, 2002; 283(2): G435 - G444. [Abstract] [Full Text] [PDF]