| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
Departments of
* Biochemistry and Molecular Biology and
Medicine, and the
Ralph H. Johnson Veterans Administration Medical Center, Medical University of South Carolina, Charleston, South Carolina, USA;
Department of Biology, University of Bologna, Bologna, Italy; and
¶ Department of Medicine and Pharmacology, Vanderbilt University School of Medicine, Nashville, Tennessee, USA
1Correspondence: Department of Biochemistry and Molecular Biology, Room 501, Basic Science Bldg., Medical University of South Carolina, 173 Ashley Ave., P.O. Box 250509, Charleston, SC 29425, USA. E-mail: hannun{at}musc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
(TNF) induced prostaglandin E2 (PGE2) production by 4 h and cyclooxygenase-2 (COX-2) induction as early as 2 h. This TNF-induced PGE2 production was inhibited by NS398, a COX-2 selective inhibitor. GC-MS analysis revealed that only COX-2-generated prostanoids were produced in response to TNF, thus providing further evidence of COX-2 selectivity. As sphingolipids have been implicated in mediating several actions of TNF, their role in COX-2 induction and PGE2 production was evaluated. Sphingosine-1-phosphate (S1P) induced both COX-2 and PGE2 in a dose-responsive manner with an apparent ED50 of 100–300 nM. The related sphingolipid sphingosine also induced PGE2, though with much less efficacy. TNF induced a 3.5-fold increase in sphingosine-1-phosphate levels at 10 min that rapidly returned to baseline by 40 min. Small interfering RNAs (siRNAs) directed against mouse SK1 decreased (typically by 80%) SK1 protein and inhibited TNF-induced SK activity. Treatment of cells with RNAi to SK1 but not SK2 almost completely abolished the ability of TNF to induce COX-2 or generate PGE2. By contrast, cells treated with RNAi to S1P lyase or S1P phosphatase enhanced COX-2 induction leading to enhanced generation of PGE2. Treatment with SK1 RNAi also abolished the effects of exogenous sphingosine and ceramide on PGE2, revealing that the action of sphingosine and ceramide are due to intracellular metabolism into S1P. Collectively, these results provide novel evidence that SK1 and S1P are necessary for TNF to induce COX-2 and PGE2 production. Based on these findings, this study indicates that SK1 and S1P could be implicated in pathological inflammatory disorders and cancer.—Pettus, B. J., Bielawski, J., Porcelli, A. M., Reames, D. L., Johnson, K. R., Morrow, J., Chalfant, C. E., Obeid, L. M., Hannun, Y A. The sphingosine kinase 1/sphingosine-1-phosphate pathway mediates COX-2 induction and PGE2 production in response to TNF-.
Key Words: prostaglandin E2 • inflammation • ceramide • sphingosine • RNA interference • tumor necrosis factor
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, and COX-2, an inducible isoform thought to mediate inflammatory events in response to inflammatory stimuli, such as the cytokines interleukin-1ß and tumor necrosis factor- (TNF) (2
, 3)
. Much evidence has established COX-2 as an important therapeutic target for the treatment of inflammatory disorders such as arthritis (4)
. More recently, it has been found that inhibitors of COX-2 also reduce the risk and spread of some cancers. Moreover, in mouse models of familial adenomatous polyposis it has been shown that crosses with knockout mice in either cytosolic phospholipase A2 (cPLA2) or COX-2 will reduce tumor size and metastasis (5
, 6)
.
TNF is a pleiotropic cytokine that elicits a wide spectrum of physiologic and pathogenic events such as proliferation, differentiation, cell death, modulation of gene transcription, and inflammation (7)
. The different cellular responses to TNF occur through binding of TNF to either 55 or 75 kDa membrane receptors and subsequent engagement of adaptor proteins (7
, 8)
. The proapoptotic role of TNF is mediated by recruitment of TRADD and FADD/MORT1 adaptor proteins and subsequent recruitment and activation of various caspases (9)
. The inflammatory reaction and anti-apoptotic function of the TNF receptor are primarily mediated by another class of adaptor proteins, TNF receptor-associated factors (TRAF) (10
, 11)
, which differentially activate distinct downstream signals such as NF-
B and JNK.
Although the intracellular signaling pathways by which TNF induces COX-2 expression are largely unresolved, it has been reported that TNF might induce COX-2 expression through the activation of tyrosine kinases (12)
or protein kinase C (13)
. Chen et al. demonstrated that TNF regulates the expression of COX-2, at least in part, through NF-B (14)
.
Sphingolipids constitute one important class of TNF mediators. These molecules, once regarded as inert structural components of cell membranes, are now known to serve as a particularly rich source of bioactive compounds, including ceramide, sphingosine, and sphingosine-1-phosphate (S1P). They are used as signaling molecules within or among cells (15)
, and thus play essential roles in the regulation of many different cellular events such as cell growth, differentiation, stress responses, and apoptosis (16)
.
Indeed, in many cell types, TNF increases intracellular ceramide levels through hydrolysis of sphingomyelin and/or activation of the de novo synthetic pathway (17
18
19
20)
. Whereas ceramide has been implicated as a mediator of TNF induced apoptosis (21
, 22)
, S1P has emerged as an anti-apoptotic and mitogenic factor with importance in pathological disease states such as cancer and angiogenesis (23
24
25
26
27
28
29)
. In certain cells, such as human umbilical vein endothelial cells (30)
, hepatocytes (31)
, and neutrophils (32)
, TNF stimulates the production of S1P upon activation of sphingosine kinase (SK). Xia et al. recently demonstrated a physical and functional interaction between TRAF2 and SK1 that transduces the TNF signal resulting in the activation of NF-B and anti-apoptosis (33)
. Thus, sphingolipids appear to regulate various responses to TNF.
Previous studies have shown that some sphingolipid mediators, such as ceramide, sphingosine, or S1P, are able to induce production of the inflammatory mediator prostaglandin E2 (PGE2) (34
35
36
37)
. Moreover, the expression level of COX-2, a key enzyme in the biosynthesis of PGE2, is increased in response to exogenous sphingosine or ceramide (37
38
39
40)
. However, little insight is available as to which of these sphingolipids, if any, may function as an endogenous regulator of COX-2 and/or PGE2. Discerning the roles of individual sphingolipids is complicated by their metabolic interconversion. Thus, whereas each sphingolipid signals distinct biological actions, the metabolic enzymes that regulate the formation of these lipids, serve as switches between these signaling modules (41)
.
The goal of this study therefore was to define which specific pathway of sphingolipid metabolism may regulate the inflammatory response to TNF. In this report, we demonstrate that in murine fibroblast L929 cells and human lung adenocarcinoma A549 cells, TNF-induced COX-2 expression and PGE2 production are dependent on activation of sphingosine kinase 1. These results illustrate an important combined approach at delineating bioactive lipids involved in specific cell responses and identify a novel and specific role for sphingosine kinase 1.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Measurement of PGE2 levels
Media were taken at the indicated time points and assayed according to manufacturer’s instructions (note: for most treatment conditions, a 1:2 dilution ratio was optimal for analysis) using the prostaglandin E2 monoclonal EIA Kit from Cayman Chemical (Ann Arbor, MI, USA; Catalog No. 514010). Media containing PGE2 competes with PGE2-acetylcholinestaerase conjugate for a limited amount of PGE2 monoclonal antibody. The antibody-PGE2 conjugate binds to a goat anti-mouse antibody previously attached to the wells. The plate is washed to remove any unbound reagents, then the substrate to acetylcholinesterase is provided. The concentration of PGE2 in a sample is inversely proportional to the yellow color produced.
Reverse transcription-PCR
mRNA from L929 cells was isolated using the Qiagen mini kit for mRNA extraction. cDNA was then generated using AMV-RT (Promega). RT-PCR reaction conditions were optimized on a Stratagene Gradient 96 Robocycler to determine a linear cDNA range using Taq polymerase (Promega) and 30 cycles of 94°C 1 h, primer dependent (58–68°C) 1 h, 72°C 2 h, and normalized to rRNA controls. Specific forward and reverse primers were designed for each mouse or human gene to yield 
500 bp fragments. Mouse SK1 (58°C, 10 µL Q solution, 502 bp): (CTTCTGGGCTGCGGCTCTATTCTG) and (GGAAAGCAACCACGCGCACA); mouse SK2 (68°C, 10 µL Q solution, 504 bp): (TTGCCCTCACCCTCACAACACAAG) and (CCTCGTAAAGCAGCCCGTCTCCA); mouse S1P lyase (66°C, 5 µL Q solution, 602 bp): (TACCGGGACTTGGCGTTAG) and (TGCTTGGAGATGCGTAGACAC); mouse S-1-P phosphatase (66°C, 524 bp): (CAACTTGCCGCTCTACTACCT) and (GAAGCCCGATGATGATGA); mouse COX-2 (58°C, 504 bp): (AGTTGTCAAACTGCGAGCTAAG) and (GCTTCCCAGCTTTTGTAACC); and human SK1 (58°C, 550 bp): (CCGACGAGGACTTTGTGCTAAT) and (GCCTGTCCCCCCAAAGCATTAAC).
Immunoblotting
Lysates were normalized to protein concentration using a Bio-Rad protein assay reagent (Hercules, CA, USA). Each sample (40 µg) was resolved on 7.5% polyacrylamide gels (SDS-PAGE) under denaturing conditions, then transferred to 0.45µm nitrocellulose membranes. After blocking overnight at 4°C with 5% nonfat milk in Tris-buffered saline/0.01% Tween 20 and washing, membranes were incubated with the relevant antibodies for 3 h at room temperature. The membranes were washed extensively in Tris-buffered saline/0.1% Tween 20 (washing buffer). The blots were probed using specific polyclonal primary antibodies (anti-mouse COX-2; (M-19) anti-human COX-2 (C-20)) Santa Cruz Biotechnology, anti-SK made in-house as described (42)
. Bands were visualized using the appropriate horseradish peroxidase-conjugated secondary antibody and the ECL Immunoblotting detection system (Amersham Pharmacia Biotech). In all cases, equal loading was verified by poststaining with 0.1% Amido black in methanol/acetic acid/water.
Sphingosine kinase activity assay
Activation of sphingosine kinase was measured as described previously (43
, 44)
with slight modifications. Cells were resuspended in ice-cold 0.1M phosphate buffer (pH 7.4) containing 20% glycerol, 1 mM mercaptoethanol, 1 mM EDTA, phosphate inhibitors 1 mM sodium orthovanadate, 15 mM sodium fluoride), protease inhibitors (10 µg/mL each leupeptin, aprotinin, trypsin, chymotrypsin, and 1 mM phenylmethylsulfonyl fluoride), and 0.5 mM 4-deoxypyridoxine, then disrupted by three 5 s pulses with a Fisher 550 sonic dismembranator at setting 2. Each sample was assayed for sphingosine kinase activity by incubation with sphingosine (Sigma) and [
-32P]ATP (5 µCi, 50 µM final) for 30 min at 37°C and the products were separated on TLC on Silica Gel G60 (Whatman) using 1-butanol/methanol/acetic acid/water (80:20:10:20) and visualized by autoradiography. The radioactive spots corresponding to sphingosine phosphate were scraped and counted in a scintillation counter.
Prostaglandin measurements by GC-MS
Various prostaglandins were identified and quantified in medium obtained from incubation of L929 cells in the absence and presence of TNF using stable isotope dilution techniques employing gas chromatography/negative ion chemical ionization mass spectrometry as described (45)
.
Sphingolipid measurements by positive mode MRM ESI/MS/MS
Electrospray Ionization (ESI)/MS/MS analysis of sphingosine bases, sphingosine-1-phosphates, and ceramides was performed on a Thermo Finnigan TSQ 7000 triple quadruple mass spectrometer, operating in a multiple reaction monitoring (MRM) positive ionization mode. Total cells, fortified with internal standards, were extracted with ethyl acetate/iso-propanol/water (60/30/10 v/v), an aliquot taken for inorganic phosphate determination, and the remainder evaporated to dryness and reconstituted in 100 µL of methanol. The reconstituted samples were injected on the Surveyor/TSQ 7000 LC/MS system and gradient was eluted from the BDS Hypersil C8, 150 x 3.2 mm, 3 µm particle size column, with 1.0 mM methanolic ammonium formate/2 mM aqueous ammonium formate mobile phase system. Peaks for the target analytes and internal standards were collected and processed using the Xcalibur software system.
For quantitative analysis, eight point calibration curves were generated by spiking an artificial matrix, mimicking the protein content of cells, with known amounts of target analyte synthetic standards and the same amount of internal standards as for cell samples. The calibration standard samples then underwent an identical extraction procedure to compensate for incomplete or varying recoveries of the target lipids. C17 base D-erythro-sphingosine (17C Sph) was used for the quantitation of sphingosine (Sph; C18 base) and dihydrosphingosine (DhSph) whereas analogous C17 sphingosine-1-phosphate (17CS1P) served the same purpose for the corresponding sphingosine-1-phosphate (S1P; C18 base) and dihydro-sphingosine-1-phosphate (DhS1P). Ceramides were quantitated using N-palmitoyl-D-erythro-sphingosine, C13 base (13C Cer) and N-heptadecanoyl-D-erythro-sphingosine, C18 base (C17 Cer) as internal calibration standards. Calibration curves were constructed by plotting peak area ratios of synthetic standards corresponding to each target analyte with respect to the appropriate internal standard. The target analyte peak areas from the samples were similarly normalized to their respective internal standard, then compared with the calibration curves using a linear regression model.
RNA interference
Gene silencing of mouse SK1, SK2, sphingosine-1-phosphate lyase, sphingosine-1-phosphate phosphatase, and human SK1 was performed essentially as described using sequence specific siRNA reagents (46)
(http:www.mpibpc.gwdg.de/abteilungen/100/105/sirna.html) (Xeragon, a Qiagen Company): mouse SK-1 starting 76 nucleotides from start codon (GGGCAAGGCUCUGCAGCUCTdTd and GAGCUGCAGAG-CCUUGCCCTdTd); mouse SK-2 starting 110 nucleotides from start codon (GCCCUACACAUACAGCGACTdTd and GUCGCUGUAUGUGUAGGGCTdTd); mouse sphingosine-1-phosphate lyase starting 107 nucleotides from the start codon (UAUGAGCCCUGGCAGCUCATdTd and UGAGC-UGCCAGGGCUCAUATdTd); mouse sphingosine-1-phosphate phosphatase starting 63 nucleotides from the start codon (GCUUCCAGCGUCUGUGUGGTdTd and CCACACAGACGCUGGAAGCTdTd); human sphingosine kinase 1 (47)
starting 70 nucleotides from the start codon (GGGC-AAGGCCUUGCAGCUCTdTd and GAGCUGCAAGGCCUUGCCTdTd); scrambled sequence (SCR) (CUGGUUACGA-AGCGAAUCCTdTd and GGAUUCGCUUCGUAACCAGTdTd). All sequences were evaluated against the database using the NIH blast program to test for specificity. L929 cells were transfected with the 21-nucleotide duplexes using OligofectAMINE (Invitrogen, San Diego, CA, USA) as recommended by the manufacturer. After 48 h cells were treated with TNF and assayed as described.
Statistical analysis
Statistical analysis was performed using the nonparametric Mann-Whitney-Wilcoxon (MWW) test on combined data (48
, 49)
. The null hypothesis was rejected for MWW 2 values equal to or smaller than 0.10 (P
0.10), which is equal to a confidence interval of 
0.90.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, TNF induced cell death in L929 cells in a dose- and time-dependent manner (data not shown). Dose-dependent induction of PGE2 always occurred before the onset of death. Figure 1
A shows that treatment of L929 cells with 5 nM TNF induced a significant increase in PGE2 as early as 4 h, whereas cell death, as measured by trypan blue and MTT (data not shown), did not commence until 16 h (36)
. PGE2 levels continued to accumulate up to 24 h after stimulation. Owing to the remarkable chemical stability of PGE2 (50)
, levels accumulate in culture media over the time course of production (Fig. 1A
). Taking this cumulative nature of PGE2 levels into account, the maximal rate of formation occurs from 4 to 8 h (75 pg/mL/hr) (Fig. 1A
, inset). Thus, the key regulatory events in PGE2 production occur before and up to 4 h after TNF stimulation.
|
To test whether the PGE2 produced was through COX-2, a COX-2-selective inhibitor NS398 was used (51)
. At 300 nM, there was complete inhibition of TNF-induced PGE2 production, suggesting a key role for COX-2 in the PGE2 response to TNF in this system (Fig. 1B
). To determine whether COX-2 was induced, cells were lysed for Western immunoblotting. Figure 1C
shows COX-2 induction in response to TNF as early as 2 h and increasing at 4 and 8 h. Semiquantitative RT-PCR showed similar changes in COX-2 mRNA levels at 4 and 8 h after TNF (data not shown). To determine which prostanoids were generated in response to TNF, media were taken for GC-MS analysis. Only the COX-2-associated prostaglandins PGE2 (3.5-fold increase), PGD2 (3.5-fold increase), and PGF2 (2.0-fold increase), but not TXB2 and 6-keto PGF1, were generated in response to TNF (Fig. 1D
). Thus, TNF induction of prostanoids is COX-2 selective.
Exogenous sphingosine-1-phosphate induces COX-2 in a lipid-specific manner
To investigate whether S1P has a role in the induction of COX-2 in response to TNF, COX-2 levels were measured by immunoblotting, followed by densitometry analysis in response to exogenous lipids. S1P delivered in 0.1% ethanol/dodecane (BSA complexed S1P gave identical results; data not shown) increased COX-2 dose dependently at concentrations ranging from 10 nM to 1 µM, with little further induction from 1 to 10 µM (Fig. 2
A). The effective dose for 50% maximal induction was determined to be 100–300 nM based on these data. To determine lipid specificity, the effects of ceramide, sphingosine, and S1P each at 5 µM in 0.1% ethanol/0.002% dodecane were compared for PGE2 production and COX-2 induction. As shown in Fig. 2B, C
, all sphingolipids tested were able, with different degrees of effectiveness, to induce COX-2 and PGE2 production. However, of those tested, S1P appeared to be the most effective. Thus, S1P was sufficient to induce COX-2 and lead to PGE2 production in L929 cells.
|
Sphingosine-1-phosphate accumulates in response to TNF
To determine whether endogenous S1P formation plays a role in induction of COX-2 and PGE2, we analyzed sphingolipid levels by ESI tandem mass spectrometry. Basal ceramide levels were determined to be 3.8 pmol/(nmol phosphate (Pi)), basal sphingosine 30.0 fmol/(nmol Pi), and basal S1P 1.0 fmol/(nmol Pi). Figure 3
shows a 3.5-fold increase in S1P levels peaking at 10 min (3.5 fmol/(nmol Pi)), with a rapid return to baseline by 40 min and no further alteration for up to 4 h (2–4 h data not shown). Neither total ceramide levels nor the ratios of individual species (data not shown) changed significantly during this time course although sphingosine levels decreased 50%, with a slow return to baseline by 4 h (2–4 h data not shown). Similarly, dihydrosphingosine levels dropped 40% from a baseline of 17.5 fmol/(nmol Pi), and dihydrosphingosine-1-phosphate levels increased from below the detection threshold to 1.5 fmol/(nmol Pi) at 10 min and 1.0 fmol/(nmol Pi) at 20 min. Using immunoblot analysis, we found induction of sphingosine kinase 1 (SK1), the enzyme responsible for S1P formation, which was significant at 2, 4, and 8 h (a representative 4 h induction can be seen in Fig. 4
). As confirmation, mRNA levels were assessed by semiquantitative RT-PCR; again, there was significant induction at 2, 4, and 8 h (data not shown).
|
|
Sphingosine kinase 1 is required for TNF-induced COX-2 induction
As a precursor to testing the necessity of SK for COX-2 induction and PGE2 production, SK1 was first down-regulated using RNA interference (RNAi) technology. Initially a time course of 24, 48, and 72 h post-siRNA treatment was used. mRNA levels were followed by RT-PCR and protein levels by immunoblotting, with the finding that 48 h was optimal for inhibition of SK1 with an siRNA ED50 of 50 nM (data not shown). Thus, for all experiments shown, cells were treated with 200 nM small interfering RNA oligonucleotides (siRNAs) for 48 h before treatment with 5 nM TNF for 4 h. This approach resulted in significant attenuation of the levels of SK1 as evaluated by immunoblot and total cell SK activity assays (Fig. 4)
. L929 cells possessed a basal SK activity of 19 pmol·mg-1·min-1 before TNF-mediated SK1 induction. SK1 RNAi partially inhibited this basal activity and completely inhibited TNF-induced increases (to 45 pmol·mg-1·min-1) in total SK activity. These changes were highly specific to SK1 RNAi as scrambled (SCR) siRNAs and siRNAs directed against SK2, S1P lyase and S1P phosphatase did not decrease SK1 levels (data not shown). This basal SK activity suggested the additional presence of SK2, which is an integral membrane SK isoform often suggested to play a "constitutive" role in cells. To evaluate SK2, therefore, the 100,000 x g membrane pellet was taken for SK activity assay and found to contain 1/3 of the total activity. Furthermore, SK activity in this membrane fraction was inhibited by 50% in response to 48 h 200 nM RNAi directed against SK2. Taken together, these finding suggests that both SK1 and SK2 are present in L929 cells and that RNAi causes significant isoform specific inhibition.
To determine whether either SK1 or SK2 was necessary for TNF induction of S1P accumulation observed at 10 min, cells were treated with 200 nM SCR siRNAs or siRNAs against SK1 or SK2. After 48 h, cells were treated with 5 nM TNF for 10 min and S1P levels were measured by ESI tandem mass spectrometry. A significant 70% inhibition of S1P induction by 10 min was observed with SK1 but not SK2 RNAi (data not shown). To determine whether S1P induction mediated by SK1 was necessary for TNF induction of COX-2, cells were treated with 200 nM SCR siRNAs or siRNAs against SK1 or SK2. After 48 h, cells were treated with 5 nM TNF for 4 h. As shown in Fig. 5
, cells treated with RNAi against SK1 did not induce COX-2 in response to TNF and had minimal PGE2 production in response to TNF. By contrast, cells treated with RNAi against SK2 gave results nearly identical to control scrambled siRNAs with respect to fold induction of COX-2 and levels of PGE2 production in response to TNF (Fig. 5)
. Therefore, SK1 but not SK2 is necessary for TNF induction of COX-2 and resultant PGE2 production.
|
Taken together, these results confirmed a role for SK1-generated S1P but did not rule out a downstream metabolite such as palmitaldehyde or sphingosine. To distinguish these possibilities, cells were treated with RNAi to S1P lyase and S1P phosphatase. RNAi of S1P lyase and S1P phosphatase enhanced the ability of TNF to induce COX-2 leading to elevated PGE2 (Fig. 5)
. These results confirm that SK1-dependent S1P and not a downstream metabolite is necessary for COX-2 induction. Moreover, they suggest that both S1P lyase and S1P phosphatase serve to attenuate the effects of S1P on COX-2 and PGE2 production by clearing intracellular S1P.
The actions of exogenous sphingosine are due to intracellular metabolism to S1P
The effectiveness of RNAi in attenuating the levels of SK1 provided a unique opportunity to reexamine the lipid specificity of exogenous sphingolipids in regulating PGE2 production. Thus, to determine whether the actions previously attributed to sphingosine and ceramide were due to metabolic conversion to S1P, cells were treated with siRNAs to SK1 and S1P lyase, then cells were treated with 5 µM of each lipid for 4 h and assayed for PGE2. Cells treated with sphingosine responded with elevated PGE2 levels in control and in SCR RNAi transfected cells (Fig. 6
). This effect of sphingosine was blocked in cells treated with siRNAs to SK1, demonstrating a requirement for conversion of sphingosine into S1P in order to induce PGE2. Reciprocally, PGE2 responses to sphingosine were enhanced in cells treated with siRNAs to S1P lyase. By contrast, all cells responded without inhibition to exogenous S1P, consistent with the hypothesis that responsiveness to sphingosine (and perhaps ceramide also) can be explained by metabolism of these lipids to S1P.
|
As seen in Fig. 6
, ceramide did not have statistically significant effects on its own at 4 h. However, in cells treated with siRNA to S1P lyase, increased sensitivity to ceramide-induced PGE2 production was observed, perhaps due to metabolic trapping of S1P generated by metabolism of ceramide. To further determine whether ceramide could indeed induce significant PGE2 production, we treated L929 cells with 5 µM C6 and C16 ceramide and found that by 6 h ceramide yielded significant 2.4- and 2.7-fold increases in PGE2 levels, respectively. When pretreated with SK1 but not scrambled control RNAi, this effect was abolished, consistent with the hypothesis that ceramide (similar to sphingosine) exerts its effects on PGE2 production through intracellular metabolism to S1P (data not shown).
The SK1/S1P pathway mediates TNF-induced COX-2 induction in human A549 lung adenocarcinoma cells
Because of the importance of the observation that SK1 was necessary for TNF-induced COX-2 induction, we sought to reproduce our observations in a human cell line. In the human lung adenocarcinoma A549 cells, even though these cells possessed significant basal COX-2 levels (52)
, TNF was found to further induce COX-2 within 2 h, leading to significant PGE2 within a time frame similar to that found in L929 cells (Fig. 7
A). TNF is not cytotoxic to A549 cells (data not shown); thus, the regulation of PGE2 and COX-2 is independent of the pathways regulating cell death. To test the sufficiency of S1P, exogenous S1P was added and found to dose-responsively lead to induction of COX-2 with a slightly higher ED50 (300 nM–1 µM) than found in L929 cells (Fig. 7B
). Similarly, S1P led to a dose-responsive elevation of PGE2 at 4 h (Fig. 7C
). Furthermore, TNF led to a potent threefold increase in S1P levels within 10 min, decreasing to baseline by 40 min in a manner that mirrored the time course observed in L929 cells (data not shown).
|
To determine the necessity of SK1 as a mediator of TNF-induced COX-2 elevation, RNAi was generated against human SK1 (47)
. Figure 7D
shows that treatment with SK1 RNAi significantly silenced SK1 mRNA production leading to a decrease in SK1 protein levels in both control and TNF-treated cells. Figure 7D
also shows that RNAi against SK1 significantly inhibited TNF-induced COX-2 elevation (80% inhibition by densitometry) and led to an 81% inhibition of PGE2 production (data not shown). Thus, we conclude that the necessity of SK1 for COX-2 induction observed in response to TNF is not restricted to L929 cells and therefore may represent a novel aspect of the COX-2 regulatory mechanism.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Several lines of evidences support the conclusion that S1P is an important mediator of COX-2 induction. 1) Exogenous S1P is sufficient to increase COX-2 with a potent ED50 in two cell lines. 2) Although related sphingolipids also produce similar effects, they do so with less efficacy and only when SK1 is available. 3) S1P rather than a downstream metabolite is the necessary product of SK1 activity as confirmed by the enhanced COX-2 elevation and PGE2 production observed in response to inhibition of S1P lyase or S1P phosphatase clearance pathways using RNAi. 4) TNF induces potent generation of S1P in a relevant time course. 5) The presence of SK1 protein is necessary for TNF-mediated induction of COX-2 leading to the generation of PGE2, as revealed using RNAi gene silencing studies. 6) TNF generation of S1P, induction of COX-2 and PGE2, as well as S1P sufficiency for PGE2 production and the necessity of SK1 for TNF-induced COX-2 induction were all observed in a distinct human cell line (A549) using distinct SK1 RNAi oligonucleotides.
This concerted approach using molecular and pharmacological tools is proving to be a useful strategy for identifying the key lipid mediator in a given system and for a specific response. A similar strategy was recently used to study the regulation of sterol regulatory element binding protein by phosphatidylethanolamine in Drosophila (53)
. Both studies highlight the utility of RNAi technology as a means of sequence-specific silencing of target genes; as shown in Scheme 1
, the enzymes targeted by RNAi in our study are the upstream and downstream enzymes regulating the generation and removal of the putative lipid mediator in this COX-2 response. Whereas SK1 RNAi inhibited S1P formation and blocked COX-2 induction by TNF and sphingosine, RNAi to S1P phosphatase and S1P lyase enhanced COX-2 accumulation, leading to enhanced responsiveness of these agonists. The sequence-specific nature of RNAi silencing allows evaluation of isoform specificity as well. Whereas the basal constitutive SK activity present in L929 cells was probably comprised of both SK1 and SK2 (as evaluated by membrane vs. soluble activities and effects of RNAi), only activation of SK1 in response to TNF was necessary for COX-2 induction. This finding is not surprising, however, as SK1 is consistently implicated in agonist-stimulated S1P generation; although much less is known about SK2, it is thought to be constitutively active. This finding demonstrates the need for further investigation into the topological and regulatory differences between these isoenzymes of SK.
|
The observation that S1P is a critical mediator of COX-2 induction has potentially important therapeutic implications. The importance of COX-2 in inflammation and cancer implies equal significance for the role attributed to S1P in regulating this pathway. Earlier independent lines of evidence have implicated S1P in pathophysiological disease states such as cancer and angiogenesis. A connection for S1P in inflammation was suspected by observations that S1P can protect leukocytes and lymphocytes from apoptosis induced by Fas, TNF-, and ceramide (38)
. In mast cells the relative levels of S1P and sphingosine may determine, in part, allergic responsiveness (39)
; in some cells, activation of SK was shown to be responsible for mediating Fc
RI-triggered mast cell degranulation (40)
. Together with the observations described here, SK increasingly offers therapeutic potential. The effectiveness of RNAi itself as a vehicle for targeted inhibition offers therapeutic possibilities that also warrant investigation.
However, the upstream and downstream mechanism(s) by which S1P is involved in this process are still unknown and need to be determined. Xia et al. recently demonstrated a physical and functional interaction between TRAF2 and SK1. TRAF2 is a component of the TNF signaling pathway that associates with TNF and transduces the TNF signal, resulting in the activation of NF-B and anti-apoptosis (33)
. Such an interaction is the most likely mechanism for the rapid increase in S1P observed in both cell lines in response to TNF. The additional delayed induction of SK1 protein is also interesting and may play a mechanistic role as well even though it does not appear to contribute significantly to intracellular S1P levels. The S1P levels in the media are currently below the detection threshold of our MS method, and so we cannot rule out a late-phase extracellular production of S1P or comment on whether the actions of S1P involve intracellular or extracellular means.
Several studies have suggested that COX-2 is not regulated by NF-B alone, but that some combination of pathways such as NF-B, ERK, p38, AP-1, and PKC (54
, 55)
may be required. Recent studies have implicated S1P in activation of NF-B (33
, 56)
, AP-1 (57)
, and ERK (58)
. Continued investigation of S1P signaling pathways related to COX-2 are vital and may help develop mechanism-based therapeutic strategies for treatment of inflammatory disorders and cancer.
In conclusion, this study clarifies the confusion in the literature surrounding studies implicating related sphingolipids in COX-2 induction by proving that their actions are dependent on metabolic conversion to S1P in order to be sufficient for this process. This study demonstrates the applicability of RNAi for the important but extremely difficult task of identifying and establishing a specific bioactive lipid, and not related metabolites, as the endogenous mediator of a process. Before this technology, the possibility of intracellular conversion after exogenous addition of lipids and the lack of a complete set of specific pharmacological inhibitors were potential obstacles to such identification.
|
ACKNOWLEDGMENTS |
|---|
Received for publication November 20, 2002. Accepted for publication March 25, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Sattler and B. Levkau Sphingosine-1-phosphate as a mediator of high-density lipoprotein effects in cardiovascular protection Cardiovasc Res, May 1, 2009; 82(2): 201 - 211. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Maceyka, S. Milstien, and S. Spiegel Sphingosine-1-phosphate: the Swiss army knife of sphingolipid signaling J. Lipid Res., April 1, 2009; 50(Supplement): S272 - S276. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. E. Schnitzer, A. Weigert, J. Zhou, and B. Brune Hypoxia Enhances Sphingosine Kinase 2 Activity and Provokes Sphingosine-1-Phosphate-Mediated Chemoresistance in A549 Lung Cancer Cells Mol. Cancer Res., March 1, 2009; 7(3): 393 - 401. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kawamori, T. Kaneshiro, M. Okumura, S. Maalouf, A. Uflacker, J. Bielawski, Y. A. Hannun, and L. M. Obeid Role for sphingosine kinase 1 in colon carcinogenesis FASEB J, February 1, 2009; 23(2): 405 - 414. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Zhang, S. Liu, J. Liu, T. Zhang, Q. Shen, Y. Yu, and X. Cao Immune Complex/Ig Negatively Regulate TLR4-Triggered Inflammatory Response in Macrophages through Fc{gamma}RIIb-Dependent PGE2 Production J. Immunol., January 1, 2009; 182(1): 554 - 562. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. J. Snider, T. Kawamori, S. G. Bradshaw, K. A. Orr, G. S. Gilkeson, Y. A. Hannun, and L. M. Obeid A role for sphingosine kinase 1 in dextran sulfate sodium-induced colitis FASEB J, January 1, 2009; 23(1): 143 - 152. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. M. Bivol, M. Hultstrom, O. A. Gudbrandsen, R. K. Berge, and B. M. Iversen Tetradecylthioacetic acid downregulates cyclooxygenase 2 in the renal cortex of two-kidney, one-clip hypertensive rats Am J Physiol Regulatory Integrative Comp Physiol, December 1, 2008; 295(6): R1866 - R1873. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Ohama, M. Okada, T. Murata, D. L. Brautigan, M. Hori, and H. Ozaki Sphingosine-1-phosphate enhances IL-1{beta}-induced COX-2 expression in mouse intestinal subepithelial myofibroblasts Am J Physiol Gastrointest Liver Physiol, October 1, 2008; 295(4): G766 - G775. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Serrano-Sanchez, Z. Tanfin, and D. Leiber Signaling Pathways Involved in Sphingosine Kinase Activation and Sphingosine-1-Phosphate Release in Rat Myometrium in Late Pregnancy: Role in the Induction of Cyclooxygenase 2 Endocrinology, September 1, 2008; 149(9): 4669 - 4679. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L.-Y. Chen, G. Woszczek, S. Nagineni, C. Logun, and J. H. Shelhamer Cytosolic phospholipase A2{alpha} activation induced by S1P is mediated by the S1P3 receptor in lung epithelial cells Am J Physiol Lung Cell Mol Physiol, August 1, 2008; 295(2): L326 - L335. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Lebman and S. Spiegel Thematic Review Series: Sphingolipids. Cross-talk at the crossroads of sphingosine-1-phosphate, growth factors, and cytokine signaling J. Lipid Res., July 1, 2008; 49(7): 1388 - 1394. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Takabe, S. W. Paugh, S. Milstien, and S. Spiegel "Inside-Out" Signaling of Sphingosine-1-Phosphate: Therapeutic Targets Pharmacol. Rev., June 1, 2008; 60(2): 181 - 195. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Anelli, C. R. Gault, A. B. Cheng, and L. M. Obeid Sphingosine Kinase 1 Is Up-regulated during Hypoxia in U87MG Glioma Cells: ROLE OF HYPOXIA-INDUCIBLE FACTORS 1 AND 2 J. Biol. Chem., February 8, 2008; 283(6): 3365 - 3375. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Culver and S. M. Laster Adenovirus Type 5 Exerts Multiple Effects on the Expression and Activity of Cytosolic Phospholipase A2, Cyclooxygenase-2, and Prostaglandin Synthesis J. Immunol., September 15, 2007; 179(6): 4170 - 4179. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Kusner, C. R. Thompson, N. A. Melrose, S. M. Pitson, L. M. Obeid, and S. S. Iyer The Localization and Activity of Sphingosine Kinase 1 Are Coordinately Regulated with Actin Cytoskeletal Dynamics in Macrophages J. Biol. Chem., August 10, 2007; 282(32): 23147 - 23162. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. G. Payne, C. A. Oskeritzian, R. Griffiths, P. Subramanian, S. E. Barbour, C. E. Chalfant, S. Milstien, and S. Spiegel The immunosuppressant drug FTY720 inhibits cytosolic phospholipase A2 independently of sphingosine-1-phosphate receptors Blood, February 1, 2007; 109(3): 1077 - 1085. [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]
|
|
|
|
|
|
M.-N. Raymond, C. Bole-Feysot, Y. Banno, Z. Tanfin, and P. Robin Endothelin-1 Inhibits Apoptosis through a Sphingosine Kinase 1-Dependent Mechanism in Uterine Leiomyoma ELT3 Cells Endocrinology, December 1, 2006; 147(12): 5873 - 5882. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Oskouian, P. Sooriyakumaran, A. D. Borowsky, A. Crans, L. Dillard-Telm, Y. Y. Tam, P. Bandhuvula, and J. D. Saba Sphingosine-1-phosphate lyase potentiates apoptosis via p53- and p38-dependent pathways and is down-regulated in colon cancer PNAS, November 14, 2006; 103(46): 17384 - 17389. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. H. Zhang, J. C. Fehrenbacher, M. R. Vasko, and G. D. Nicol Sphingosine-1-Phosphate Via Activation of a G-Protein-Coupled Receptor(s) Enhances the Excitability of Rat Sensory Neurons J Neurophysiol, September 1, 2006; 96(3): 1042 - 1052. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Dragusin, S. Wehner, S. Kelly, E. Wang, A. H. Merrill Jr., J. C. Kalff, and G. van Echten-Deckert Effects of sphingosine-1-phosphate and ceramide-1-phosphate on rat intestinal smooth muscle cells: implications for postoperative ileus FASEB J, September 1, 2006; 20(11): 1930 - 1932. [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]
|
|
|
|
 |
|
 Y. H. Zhang, M. R. Vasko, and G. D. Nicol Intracellular sphingosine 1-phosphate mediates the increased excitability produced by nerve growth factor in rat sensory neurons J. Physiol., August 15, 2006; 575(1): 101 - 113. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
X. Jiang and X. Han Characterization and direct quantitation of sphingoid base-1-phosphates from lipid extracts: a shotgun lipidomics approach J. Lipid Res., August 1, 2006; 47(8): 1865 - 1873. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. E. Skaznik-Wikiel, T. Kaneko-Tarui, A. Kashiwagi, and J. K. Pru Sphingosine-1-Phosphate Receptor Expression and Signaling Correlate with Uterine Prostaglandin-Endoperoxide Synthase 2 Expression and Angiogenesis During Early Pregnancy Biol Reprod, March 1, 2006; 74(3): 569 - 576. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Yang, B. E. Castle, A. Hanidu, L. Stevens, Y. Yu, X. Li, C. Stearns, V. Papov, D. Rajotte, and J. Li Sphingosine Kinase 1 Is a Negative Regulator of CD4+ Th1 Cells J. Immunol., November 15, 2005; 175(10): 6580 - 6588. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Maceyka, H. Sankala, N. C. Hait, H. Le Stunff, H. Liu, R. Toman, C. Collier, M. Zhang, L. S. Satin, A. H. Merrill Jr., et al. SphK1 and SphK2, Sphingosine Kinase Isoenzymes with Opposing Functions in Sphingolipid Metabolism J. Biol. Chem., November 4, 2005; 280(44): 37118 - 37129. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Wu, R. A. Silbajoris, Y. E. Whang, L. M. Graves, P. A. Bromberg, and J. M. Samet p38 and EGF receptor kinase-mediated activation of the phosphatidylinositol 3-kinase/Akt pathway is required for Zn2+-induced cyclooxygenase-2 expression Am J Physiol Lung Cell Mol Physiol, November 1, 2005; 289(5): L883 - L889. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. E. Chalfant and S. Spiegel Sphingosine 1-phosphate and ceramide 1-phosphate: expanding roles in cell signaling J. Cell Sci., October 15, 2005; 118(20): 4605 - 4612. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. C. Hait, S. Sarkar, H. Le Stunff, A. Mikami, M. Maceyka, S. Milstien, and S. Spiegel Role of Sphingosine Kinase 2 in Cell Migration toward Epidermal Growth Factor J. Biol. Chem., August 19, 2005; 280(33): 29462 - 29469. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. J. Pettus, K. Kitatani, C. E. Chalfant, T. A. Taha, T. Kawamori, J. Bielawski, L. M. Obeid, and Y. A. Hannun The Coordination of Prostaglandin E2 Production by Sphingosine-1-phosphate and Ceramide-1-phosphate Mol. Pharmacol., August 1, 2005; 68(2): 330 - 335. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Osawa, H. Uchinami, J. Bielawski, R. F. Schwabe, Y. A. Hannun, and D. A. Brenner Roles for C16-ceramide and Sphingosine 1-Phosphate in Regulating Hepatocyte Apoptosis in Response to Tumor Necrosis Factor-{alpha} J. Biol. Chem., July 29, 2005; 280(30): 27879 - 27887. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. D. Church, G. Hessler, J. E. Goodall, D. A. Rider, C. J. Workman, D. A. A. Vignali, P. A. Bacon, E. Gulbins, and S. P. Young TNFR1-induced sphingomyelinase activation modulates TCR signaling by impairing store-operated Ca2+ influx J. Leukoc. Biol., July 1, 2005; 78(1): 266 - 278. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. A. Taha, K. Kitatani, J. Bielawski, W. Cho, Y. A. Hannun, and L. M. Obeid Tumor Necrosis Factor Induces the Loss of Sphingosine Kinase-1 by a Cathepsin B-dependent Mechanism J. Biol. Chem., April 29, 2005; 280(17): 17196 - 17202. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Pannu, A. K. Singh, and I. Singh A Novel Role of Lactosylceramide in the Regulation of Tumor Necrosis Factor {alpha}-mediated Proliferation of Rat Primary Astrocytes: IMPLICATIONS FOR ASTROGLIOSIS FOLLOWING NEUROTRAUMA J. Biol. Chem., April 8, 2005; 280(14): 13742 - 13751. [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]
|
|
|
|
 |
|
 A. Bai, G. P. Meier, Y. Wang, C. Luberto, Y. A. Hannun, and D. Zhou Prodrug Modification Increases Potassium Tricyclo[5.2.1.02,6]-decan-8-yl Dithiocarbonate (D609) Chemical Stability and Cytotoxicity against U937 Leukemia Cells J. Pharmacol. Exp. Ther., June 1, 2004; 309(3): 1051 - 1059. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. A. Taha, W. Osta, L. Kozhaya, J. Bielawski, K. R. Johnson, W. E. Gillanders, G. S. Dbaibo, Y. A. Hannun, and L. M. Obeid Down-regulation of Sphingosine Kinase-1 by DNA Damage: DEPENDENCE ON PROTEASES AND p53 J. Biol. Chem., May 7, 2004; 279(19): 20546 - 20554. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), 0.01 (*; 6 independent experiments at this concentration). B, C) L929 cells grown in 24-well plates (
