Linetal.CellCommunicationandSignaling2013,11:8 http://www.biosignaling.com/content/11/1/8 RESEARCH Open Access Up-regulation of COX-2/PGE by endothelin-1 via 2 κ MAPK-dependent NF- B pathway in mouse brain microvascular endothelial cells Chih-Chung Lin1†, Hsi-Lung Hsieh2†, Ruey-Horng Shih3, Pei-Ling Chi3, Shin-Ei Cheng3 and Chuen-Mao Yang3* Abstract Background: Endothelin-1 (ET-1) is a proinflammatory mediator and elevated intheregions of several brain injury and inflammatorydiseases.The deleterious effects ofET-1 on endothelial cells may aggravate brain inflammation mediated through theregulationof cyclooxygenase-2 (COX-2)/prostaglandinE (PGE ) system invarious cell types. 2 2 However, thesignaling mechanismsunderlying ET-1-induced COX-2expression inbrainmicrovascular endothelial cellsremain unclear. Herein we investigated theeffects of ET-1 inCOX-2 regulation in mouse brain microvascular endothelial (bEnd.3) cells. Results: The data obtained withWestern blotting, RT-PCR,and immunofluorescent staining analysesshowed that ET-1-induced COX-2 expression was mediatedthrough an ET -dependent transcriptional activation. Engagementof B G-and G -protein-coupled ET receptors by ET-1 led to phosphorylation of ERK1/2, p38 MAPK, and JNK1/2 and i q B then activated transcription factor NF-κB. Moreover, the data of chromatin immunoprecipitation (ChIP) and promoter reporter assay demonstrated that the activated NF-κB was translocated into nucleus and bound to its corresponding binding sites inCOX-2promoter,thereby turning on COX-2 gene transcription.Finally,up-regulation ofCOX-2 by ET-1 promoted PGE release inthese cells. 2 Conclusions: These results suggestedthat inmouse bEnd.3 cells, activationof NF-κB by ET -dependent MAPK B cascades is essential for ET-1-induced up-regulationof COX-2/PGE system. Understandingthemechanismsof 2 COX-2 expression and PGE release regulated by ET-1/ET system onbrain microvascular endothelial cellsmay 2 B provide rationally therapeutic interventions for brain injury or inflammatory diseases. Keywords: Endothelin-1, COX-2,MAPK, NF-κB, Brain microvascular endothelial cells Background endothelial cells are regulated at transcription and trans- Cerebral capillary and microvascular endothelial cells lation levels by avariety of chemical and physical stimuli play an active role in maintaining cerebral blood flow, and the levels of ET, ET-1 especially, are elevated in microvascular tone and blood-brain barrier (BBB) func- shock, myocardial infarction, and kidney failure indica- tions [1]. In the development of various vascular dis- tive of enhanced formation in these diseases [4]. More- eases, an early finding is dysfunction of the vascular over, the bioactivity of ET-1 triggers vasoconstriction endothelium that is closely related to clinical events in and pro-inflammatory action which have been impli- patients with atherosclerosis and hypertension [2,3]. The cated in the pathogenesis of hypertension and vascular vasoactive mediators such as endothelin (ET) could be diseases [3-6]. The effects of ET-1 are mediated through produced by endothelial cells to maintain hemodynamic aGprotein-dependentregulation, includingtwo types of responses. Production and release of ETs from cultured ET receptors: ET type A (ET ) and type B (ET ) [7]. A B ET is involved in constriction and proliferation of A *Correspondence:[email protected] vascular smooth muscle cells, whereas ETB on endothe- †Equalcontributors lial cells mediates the generation of nitric oxide, which 3DepartmentofPharmacology,CollegeofMedicine,ChangGungUniversity, acts as vasodilator and inhibits platelet aggregation [8]. 259Wen-Hwa1stRoadKwei-San,Tao-Yuan,Taiwan Fulllistofauthorinformationisavailableattheendofthearticle Moreover, ET-1 also plays a substantial role in the ©2013Linetal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreative CommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,and reproductioninanymedium,providedtheoriginalworkisproperlycited. Linetal.CellCommunicationandSignaling2013,11:8 Page2of14 http://www.biosignaling.com/content/11/1/8 normal development or in the central nervous system inflammatory mediators in neurodegenerative disor- (CNS) diseases. In brain, endothelial cells [9] and astro- ders[19]. cytes [10] are potential sources of ET-1 release in re- ET-1 is known to activate ET receptors (ET or ET ), A B sponse to hypoxic/ischemic injury of the brain. A report a heterotrimeric G protein-coupled receptor (GPCR), has shown that the ET receptors are located on brain which stimulate multiple signaling pathways and regu- B endothelial and vascular smooth muscle cells, and late diverse cellular functions [7,20-22]. The principal modulate post-injury responses of these cells in the CNS mechanism underlying activation by ET-1 is mediated [6].Thus,thereisanincreasinginterestintheregulatory through ET receptors coupling G proteins, resulting in B q roleofendothelialcellsinneurovascularcoupling,which activation of phospholipase C (PLC)-β, phosphoinositide matches adequate supply of cerebral blood flow with the (PI) hydrolysis, and formation of inositol trisphosphate local metabolic demands that are imposed by neural ac- (IP ) and diacylglycerol, leading to Ca2+ increase and 3 tivity [11]. As a fundamental component of the neuro- protein kinase C (PKC) activation [22]. Activation of a vascular unit, endothelial dysfunction has been shown to G protein-coupled ET receptor has been also shown to i B be implicated in neurodegenerative diseases [11,12]. Cir- inhibit adenylylcyclase activity [23].Additionally, several cumstantial evidence has further demonstrated that studies have demonstrated that activation of G and G q i overexpression of ET-1 on endothelial cells has deleteri- protein-coupled receptors via different signal pathways ous effects on ischemic brain [1]. It has been demon- could activate diverse mitogen-activated protein kinases strated that endothelial ET-1 induces cytokines or (MAPKs) [24]. It has been shown that ET-1-stimulated chemokines (e.g., interleukine-1 or interleukin-8) pro- MAPKs activation to regulate various cellular responses duction and secretion by non-neuronal cells, including including cell survival, growth, proliferation, and cellular astrocytes and human brain-derived endothelial cells, hypertrophy in several cell types [21,25]. Several studies which directly contributes to BBB breakdown during have suggested that up-regulation of COX-2 requires ac- CNS inflammation [11,13]. These findings suggest that tivation of MAPKs and related transcription factors in ET-1 might be involved in neuroinflammation. However, various cell types [22,26,27]. Our previous reports also the detailed mechanisms responsible for ET-1 action are demonstrate that several GPCR agonists (e.g. sphingo- still limited. sine 1-phosphate, thrombin, and bradykinin) stimulate Cyclooxygenase (COX), known as prostaglandin- MAPKs and NF-κB activation associated with COX-2 endoperoxide synthase, is a rate-limiting key enzyme in expression in rat VSMCs and astrocytes [28,29]. Al- the synthesis of prostaglandins (PGs). In this process, though several pro-inflammatory mediators have been phospholipase A catalyzes the release of arachidonic extensively confirmed to rapidly up-regulate NF-κB- 2 acid (AA) from membrane phospholipids, while COX dependent genes such as COX-2 and play a critical role catalyzes the conversion of AA into PGs [14,15]. COX in inflammation [29,30], the signaling mechanisms by exists two isoforms: COX-1, which is constitutively which ET-1-induced MAPKs activation linked to COX-2 expressed under normal conditions in most tissues, expression and PGE production are not completely 2 mediates regulating normal physiological responses and defined inbrainmicrovascularendothelial cells. controls vascular homeostasis; COX-2, is not detectable In this study, we investigated the molecular mechan- in most normal tissues or cells, but its expression can be isms underlying ET-1-induced COX-2 expression in induced by a variety of stimuli such as cytokines, endo- mouse brain microvascular endothelial (bEnd.3) cells. toxin, and growth factors to produce PGs during inflam- These findings suggested that ET-1 induces COX-2 ex- matory responses in various cell types like vascular pression at the transcriptional and translational levels, endothelial and smooth muscle cells [16,17]. Previous which is mediated through the ET receptor (coupling B reports have shown that COX-2 immunoreactivity is a to G and G )-dependent activation of ERK1/2, p38 i q characteristic finding in the synovial macrophage and MAPK, JNK1/2, and NF-κB pathway, leading to PGE 2 vascular cells of patients with arthritis and atheroscler- biosynthesis in mouse bEnd.3 cells. These results pro- osis, respectively. Moreover, several studies have indi- vide new insights into the mechanisms of ET-1 action cated COX-2 as a major therapeutic target for the which may be therapeutic value in brain inflammatory treatment of inflammatory disorders like arthritis [14]. diseases. The mice with homozygous deletion of the cox-2 gene lead to a striking reduction of endotoxin-induced in- Results flammation [18]. Accordingly, COX-2 may play a cru- ET-1inducesCOX-2expressionandPGE releasein 2 cial role in the development of various inflammatory bEnd.3cells responses including vascular inflammation. In the To investigate the effect of ET-1 on COX-2/PGE sys- 2 CNS, several studies have indicated that up-regulation tem, bEnd.3 cells were incubated with various concen- of COX-2 leads to production of PGs which are potent trations of ET-1 for the indicated time intervals. The Linetal.CellCommunicationandSignaling2013,11:8 Page3of14 http://www.biosignaling.com/content/11/1/8 data showed that ET-1 induced COX-2 expression in a biosynthesis, we collected the conditioned media and time- and concentration-dependent manner (Figure 1A). determined PGE levels by using an EIA kit. The results 2 There was a significant increase within 2-4 h, reached a showed that ET-1 time-dependently stimulated PGE re- 2 maximal response within 6 h, and declined within 24 h. lease (Figure 1D) and a significant PGE production was 2 ET-1 also time-dependently induced COX-2 mRNA ex- observed within 4 h, reached a maximal response within pression in bEnd.3 cells, determined by RT-PCR 6 h and slightly declined within 24 h. These results sug- (Figure 1B). There was a significant increase in COX-2 gested that ET-1 induces COX-2/PGE system via up- 2 mRNA within 30 min, and reached a maximal response regulatingCOX-2gene expression inbEnd.3 cells. within 2 h. Moreover, to confirm whether ET-1 induces COX-2 expression via the transcription activity of COX- ET-1upregulatesCOX-2expressionviaanET receptor B 2 promoter, cells were transiently transfected with COX- ET-1exertsits biological effects via ETreceptors,including 2 promoter-luciferase reporter construct and then sti- ET andET ,whicharemembersofGPCRsuperfamily[7]. A B mulated with ET-1 for the indicated time intervals. As First, we determined which subtypes of ET receptors are shown in Figure 1C, ET-1 time-dependently induced expressedonbEnd.3cellsbyRT-PCR.Thedatashowedthat COX-2 promoter-luciferase activity in bEnd.3 cells. A ET but not ET receptors are expressed on bEnd.3 cells B A maximalresponsewasobtainedwithin4h.Ourprevious (Figure 2A). Next, to identify the subtypes of ET receptors studies have shown that COX-2 expression induced by involved in ET-1-induced COX-2 expression, pretreatment BK or sphingosine 1-phosphate is mainly responsible for with BQ-788 (an ET antagonist), but not BQ-123 (an ET B A prostanoid (PGE ) release in various cell types [28,29]. antagonist), attenuated the ET-1-induced COX-2 protein 2 Thus, to determine whether ET-1 could induce PGE (Figure 2B) and mRNA (Figure 2C) expression, suggesting 2 Figure1ET-1inducedCOX-2expressionandPGE release.(A)TimeandconcentrationdependenceofET-1-inducedCOX-2expression,cells 2 weretreatedwithvariousconcentrationET-1fortheindicatedtimeintervals.(B)TimedependenceofET-1-inducedCOX-2mRNAexpression, cellsweretreatedwith10nMET-1fortheindicatedtimeintervals.COX-2mRNAwasanalyzedbyRT-PCR.(C)TimedependenceofET-1-induced COX-2promotertranscriptionactivity,cellsweretransfectedaCOX-2promoter–luciferasereportergeneandthenincubatedwithET-1forthe indicatedtimes.Thepromoterreporterassaywasperformedasdescribedin“MaterialsandMethods”.(D)PGE releaseinducedbyET-1,the 2 conditionedmediawerecollectedtoassayPGE levelbyEIAasdescribedin“MaterialsandMethods”(n=3ineachgroup;*P<0.05,#P<0.01 2 comparedwithvehicle). Linetal.CellCommunicationandSignaling2013,11:8 Page4of14 http://www.biosignaling.com/content/11/1/8 Figure2InvolvementofET receptorsinET-1-inducedCOX-2expression.(A)RT-PCRanalysisofETreceptorexpressioninbEnd.3cells.Lane B 1:maker.Lane2:ET primersandbEnd.3RNA.Lane3:ET primersandbEnd.3RNA.(B,C)CellswerepretreatedwithBQ-123orBQ-788for1h A B andthenincubatedwithET-1for(B)6hand(C)1h.(D)CellsweretransfectedwithsiRNAforET receptorfor24handthenexposedtoET-1 B for6h.The(B,D)COX-2proteinand(C)mRNAwereanalyzedbyWesternblotandRT-PCR,respectivelyasdescribedinFigure1.Dataare expressedasmean±SEMofthreeindividualexperiments(n=3ineachgroup,#P<0.01ascomparedwithcellsstimulatedbyET-1alone). that ET receptor is predominantly involved in these [7,22]. To further determine which of G proteins was B responses.Tofurtherconfirmthisnote,transfectionofcells involved in ET-1-induced COX-2 expression, pretreatment with ET siRNA significantly down-regulated ET protein with either G protein antagonist GP antagonist-2 (GPA2) B B i expression and inhibited ET-1-induecd COX-2 expression or G protein antagonist GP antagonist-2A (GPA2A) con- q (Figure2D).ThesedatasuggestedthatET-1-inducedCOX- centration-dependently attenuated ET-1-induced COX-2 2 expression is mediated through an ET receptor- protein(Figure3A)andmRNA(Figure3B)expression.Fur- B dependentmannerinthesecells. thermore, to confirm these results, as shown in Figure 3C and D,transfectionwith eitherGαorG αdown-regulated i q InvolvementofaG andG protein-coupledET receptor Gα or G α protein, respectively, and attenuated ET-1- i q B i q inET-1-induecdCOX-2expression induced COX-2 expression. These data demonstrated that ET receptor has been shown to be a pleiotropic GPCR for ET-1-inducedCOX-2expressionismediatedthrougheither ET-1 which is coupled to G proteins including G and G G orG protein-coupledET receptorsinbEnd.3cells. i q i q B Linetal.CellCommunicationandSignaling2013,11:8 Page5of14 http://www.biosignaling.com/content/11/1/8 Figure3ET-1inducesCOX-2expressionviaaG andG proteins-coupledET receptor.(A,B)CellswerepretreatedwithG antagonist i q B i (GPA2)orG antagonist(GPA2A)for1handthenexposedtoET-1for(A)6hand(B)1h.(C,D)CellsweretransfectedwithsiRNAfor(C)Gαor q i (D)G αproteinfor24handthenexposedtoET-1for6h.The(A,C,D)COX-2proteinand(B)mRNAwereanalyzedbyWesternblotandRT- q PCR,respectivelyasdescribedinFigure1.Dataareexpressedasmean±SEMofthreeindividualexperiments(n=3ineachgroup,#P<0.01as comparedwithET-1alone). ET-1-inducedCOX-2expressionismediatedthrough 2phosphorylationwhichwasattenuatedbypretreatment MAPKs with U0126, SB202190, or SP600125 during the period Activation of MAPKs by ET-1 could modulate cellular of observation. Moreover, to ensure the roles of MAPKs functions of endothelial cells [31]. To investigate the in ET-1-induced COX-2 expression, transfection with roles of ERK1/2, p38 MAPK, and JNK1/2 in ET-1- siRNA of ERK2, p38 MAPK, or JNK1 down-regulated induced COX-2 expression, pretreatment with the in- the expression of total ERK2, p38 MAPK, or JNK1 pro- hibitor of MEK1/2 (U0126), p38 MAPK (SB202190), or tein and attenuated ET-1-induced COX-2 expression JNK1/2 (SP600125) attenuated ET-1-induced COX-2 (Figure 4D). These data indicated that phosphorylation protein (Figure 4A) and mRNA (Figure 4B) expression of ERK1/2, p38 MAPK, and JNK1/2 is involved in ET-1- in bEnd.3 cells, suggesting the involvement of ERK1/2, induced COX-2 expression in bEnd.3 cells. To demon- p38 MAPK, and JNK1/2 in ET-1-induced responses. To strate whether ET-1 stimulates ERK1/2, p38 MAPK, and further determine whether ET-1-stimulated ERK1/2, p38 JNK1/2phosphorylation viaaGprotein-coupledET re- B MAPK, and JNK1/2 phosphorylation is involved in ceptor cascade, pretreatment with BQ-788, GPA2, or COX-2 expression, as shown in Figure 4C, ET-1 time- GPA2A attenuated ET-1-stimulated ERK1/2, p38 dependently stimulated ERK1/2, p38 MAPK, and JNK1/ MAPK, and JNK1/2 phosphorylation during the period Linetal.CellCommunicationandSignaling2013,11:8 Page6of14 http://www.biosignaling.com/content/11/1/8 Figure4(Seelegendonnextpage.) Linetal.CellCommunicationandSignaling2013,11:8 Page7of14 http://www.biosignaling.com/content/11/1/8 (Seefigureonpreviouspage.) Figure4ET-1-inducedCOX-2expressionismediatedthroughMAPKsphosphorylation.(A,B)Cellsweretreatedwith10nMET-1for(A)6 hand(B)1hintheabsenceorpresenceofU0126,SB202190,orSP600125.TheCOX-2proteinandmRNAexpressionweredeterminedby WesternblotandRT-PCR.(C)TimedependenceofET-1-stimulatedERK1/2,p38MAPK,andJNK1/2phosphorylation,cellswereincubatedwith10 nMET-1fortheindicatedtimesintheabsenceorpresenceofU0126(1μM),SB202190(300nM),orSP600125(300nM).(D)Cellswere transfectedwithsiRNAofERK2,p38MAPK,orJNK1andthenexposedtoET-1for6h.(E)CellswerepretreatedwithBQ-788(1μM),GPA2(1μM), orGPA2A(1μM)for1handthenincubatedwithET-1(10nM)fortheindicatedtimes.ThecelllysateswerecollectedandanalyzedbyWestern blottingusingananti-COX-2,anti-phospho-ERK1/2,anti-phospho-p38MAPK,anti-phospho-JNK1/2,anti-ERK2,anti-p38MAPK,anti-JNK1,oranti- GAPDH(asaninternalcontrol)antibody.Dataareexpressedasmean±SEMofatleastthreeindividualexperiments(n=3ineachgroup;*P<0.05, #P<0.01ascomparedwithET-1alone). of observation (Figure 4E). These results demonstrated essential for ET-1-induced COX-2 gene up-regulation. that G protein-coupled ET -dependent activation of The transcriptional activity of NF-κB was evaluated by a (i/q) B ERK1/2, p38 MAPK, and JNK1/2 by ET-1 is, at least in promoter (containing NF-κB binding sites)-luciferase ac- part,requiredforCOX-2expression inbEnd.3cells. tivity assay. As shown in Figure 6A, ET-1 enhanced NF- κB transcriptional activity in a time-dependent manner NF-κBisrequiredforET-1-inducedCOX-2expression with a maximal response within 60 min, which was sig- ET-1 has been shown to modulate cellular functions nificantly inhibited by pretreatment with an inhibitor of through activation of NF-κB signaling in various cell NF-κB (Bay11-7082, 10 nM). Moreover, pretreatment types [32]. To examine whether activation of NF-κB is with BQ-788 (1 μM), GPA2 (1 μM), GPA2A (1 μM), required for ET-1-induced COX-2 expression, as shown U0126 (U0, 1 μM), SB202190 (SB, 300 nM), or in Figure 5A and B, pretreatment with a selective NF-κB SP600125(SP,300nM)attenuatedNF-κBtranscriptional inhibitor Bay11-7082, which blocks activation of NF-κB activity stimulated by ET-1 (Figure 6B), demonstrating signaling, attenuated ET-1-induced COX-2 protein and that ET-1 enhances the NF-κB transcriptional activity mRNA expression in bEnd.3 cells. To determine through an ET -dependent activation of MAPKs. Subse- B whether the involvement of NF-κB in ET-1-induced quently, we determined that ET-1-stimulates NF-κB p65 responses mediated through NF-κB (p65 subunit) trans- binding activity in a time-dependent manner by ChIP- location, as shown in Figure 5C, ET-1 time-dependently PCR analysis (Figure 6C). ET-1-stimulated NF-κB p65 stimulated translocation of NF-κB p65 from cytosol into binding activity was inhibited by pretreatment with nucleus determined by Western blot. A maximal re- U0126, SB202190, SP600125, Bay11-7082, or BQ-788 sponse was obtained within 90 min and sustained over (Figure 6C, lower part). In addition, we have demon- 120 min (upper part). Moreover, we also confirmed the strated that ET-1 time-dependently induces COX-2 pro- NF-κB p65 translocation by an immunofluorescence moter activity (Figure 1C). We further demonstrated staining. The imaging data confirmed that ET-1 stimu- that ET-1-increased the COX-2 promoter activity was lated the p65 translocation at 90 min, which was inhib- significantly inhibited by pretreatment with BQ-788, ited by pretreatment with Bay11-7082 (Figure 5C, lower GPA2, GPA2A, U0126, SB202190, SP600125, or Bay11- part). We further demonstrated that ET-1 stimulated 7082 (Figure 6D), suggesting that ET-1 stimulates COX- translocation of NF-κB p65 was attenuated by pretreat- 2 promoter activity via the ET -dependent activation of B ment with the inhibitor of ET receptor (BQ-788), MAPKs and NF-κB in bEnd.3 cells. To further ensure B MEK1/2 (U0126), p38 MAPK (SB202190), JNK1/2 that NF-κB indeed mediates ET-1-induced COX-2 pro- (SP600125), or NF-κB (Bay11-7082) (Figure 5D). To fur- moter activity through binding to its regulatory element ther verify that NF-κB p65 is essential for ET-1-induced within the COX-2 promoter region, the wild-type (WT) COX-2 expression, as shown in Figure 5E, transfection and mutated by a single-point mutation of the NF-κB with p65 siRNA significantly reduced the p65 protein binding site (mt-κB) COX-2 promoters were constructed expression (upper panel) and the ET-1-induced COX-2 (as illustrated in Figure 6E, upper panel). ET-1- expression (lower panel). The results suggested that ET- stimulated COX-2 promoter activity was significantly 1-stimulated NF-κB translocation mediated through ET attenuated in bEnd.3 cells transfected with mt-κB-COX- B receptor, ERK1/2, p38 MAPK, and JNK1/2 is required 2 (Figure 6D, lower panel), indicating that NF-κB elem- forCOX-2inductioninbEnd.3cells. ent was essential for ET-1-induced COX-2 promoter ac- tivity. These results further confirmed that ET-1 induces InvolvementofNF-κBinCOX-2genepromoteractivity COX-2 promoter activity via enhancing NF-κB binding stimulatedbyET-1 to the κB binging site within COX-2 promoter region in We have found that ET-1 stimulates translocation of bEnd.3 cells. NF-κB p65 leading to COX-2 expression (Figure 5). We have found that ET-1 time-dependently induces Next, we examined whether activation of NF-κB is PGE release (Figure 1D). Here, we further determined 2 Linetal.CellCommunicationandSignaling2013,11:8 Page8of14 http://www.biosignaling.com/content/11/1/8 Figure5NF-κB(p65)isessentialforET1-1-inducedCOX-2expression.(A,B)Cellsweretreatedwith10nMET-1for6hintheabsenceor presenceofBay11-7082.TheCOX-2proteinandmRNAexpressionweredeterminedbyWesternblotandRT-PCRasdescribedinFigure1. (C)TimedependenceofET-1-stimulatedp65NF-κBtranslocationbysubcellularisolation,Westernblot,andimmunofluorescentstain.(D)Cells werepretreatedwithU0126(1μM),SB202190(300nM),SP600125(300nM),BQ-788(1μM),orBay11-7082(10nM)for1handthenincubated withET-1(10nM)for90min.ThenuclearfractionwasanalyzedbyWesternblot.(E)Cellsweretransfectedwithp65siRNAandthenexposedto ET-1for6h.ThecelllysateswerecollectedandanalyzedbyWesternblottingusingananti-COX-2,anti-p65,anti-LaminA,oranti-GAPDH (asaninternalcontrol)antibody.Dataareexpressedasmean±SEMofatleastthreeindividualexperiments(n=3ineachgroup;#P<0.01as comparedwithET-1alone). the involvement of these signaling components in ET-1- involved in inflammatory responses. The up-regulation induced PGE release, as shown in Figure 6F, ET-1- of COX-2 has been shown to display a wide range of 2 induced PGE release was markedly attenuated by pre- biological activities in different tissues, including devel- 2 treatment with BQ-788, GPA2, GPA2A, U0126, opment, proliferation, cancers, and inflammation SB202190, SP600125, Bay11-7082, or transfection with [14,15]. Moreover, ET-1 is elevated in the regions of vas- p65 siRNA. These results demonstrated that ET - cular injuries andinflammation[7,8]. Circumstantialevi- B mediated activation of MAPKs (i.e., ERK, p38 MAPK, dence has further demonstrated that overexpression of and JNK) and NF-κB by ET-1 is essential for COX-2 up- ET-1 on endothelial cells has deleterious effects on is- regulationandPGE releaseinbEnd.3cells. chemic brain [1,5,6]. Reid et al. (1995) suggest that the 2 ET-1 model provides new insights into the mechanisms Discussion of cerebral ischemia and reperfusion injury, and evalu- Several lines of evidence have demonstrated that high ates the usefulness of novel strategies of neuroprotection levels of PGs, synthesized by inducible COX-2, are [33]. ET-1 has been shown to up-regulate the expression Linetal.CellCommunicationandSignaling2013,11:8 Page9of14 http://www.biosignaling.com/content/11/1/8 Figure6ET-1-stimulatedCOX-2promoteractivityismediatedthroughNF-κB-dependentpathway.(A)TimedependenceofET-1- enhancedNF-κBtranscriptionactivity,cellsweretransfectedwithaNF-κB-luciferasereportergeneandthenexposedtoET-1fortheindicated times.(B)Aftertransfection,thecellswerepretreatedwithBQ-788(1μM),GPA2(1μM),GPA2A(1μM),U0126(1μM),SB202190(300nM), SP600125(300nM),or(A)Bay11-7082(10nM)for1handthenincubatedwithET-1(10nM)for60min.(C)CellswerepretreatedwithBQ-788, U0126,SB202190,orSP600125for1handthenincubatedwithET-1.Thep65NF-κBbindingactivitywasanalyzedbyChIP-PCRasdescribedin Methods.(D)ForCOX-2promoteractivity,cellsweretransfectedwithaCOX-2-promoter-luciferasereportergeneandthenexposedtoET-1.After transfection,thecellswerepretreatedwithBQ-788,GPA2,GPA2A,U0126,SB202190,SP600125,orBay11-7082for1handthenincubatedwith ET-1for6h.(E)Schematicrepresentationofa50-promoterregionsofthemousedifferentCOX-2promoterconstructs,eitherwild-type(WT)or mutationoftheκB-bindingsite(mt-κB)fusedtothepGL-luciferasereportergene,thetranslationalstartsite(+1)oftheluciferasereportergene wasindicatedbyanarrow.CellsweretransfectedwithWTCOX-2promoterreportergene(WT-COX-2)orNF-κBmutatedCOX-2promoter reportergene(mt-κB-COX-2)andthenincubatedwithorwithoutET-1for6h.Thepromoterreporterassaywasperformedasdescribedin Methods.(F)CellswerepretreatedwithBQ-123,BQ-788,GPA2,GPA2A,U0126,SB202190,SP600125,Bay11-7082,ortransfectedwithp65siRNA andthenincubatedwithET-1for6h.ThePGE levelswereanalyzedbyEIA.Dataareexpressedasmean±SEMofatleastthreeindividual 2 experiments(n=3ineachgroup;#P<0.01ascomparedwithET-1alone). Linetal.CellCommunicationandSignaling2013,11:8 Page10of14 http://www.biosignaling.com/content/11/1/8 of COX-2 through MAPKs in various cell types signaling pathways [24]. MAPKs activation by ET-1 has [26,27,34]. However, little is known about the effect of been shown to modulate various cellular responses in ET-1 on COX-2 expression in brain vascular endothelial several cell types [22,25]. Activation of ERK1/2 (p44/p42 cells. Here, we applied cultured models of mouse bEnd.3 MAPK) mightbe implicated in the expression of inflam- cells coupled with Western blot analysis, selective matory genes in several models of vascular injury and pharmacological inhibitors, transfection with siRNAs, inflammation [17,28]. In this study, we demonstrated immunofluorescenct staining, and promoter assay to in- thatET-1stimulatedanET receptor-dependentcascade B vestigate the molecular mechanisms underlying ET-1- ofsequential ERK1/2phosphorylation (Figure 4E), which induced COX-2 expression and PGE release. Our contributes to induction of COX-2 protein and mRNA 2 results demonstrate that in bEnd.3 cells, activation of levels (Figure 4A and 4B), promoter activity (Figure 6D), ET receptor-dependent MAPKs (ERK1/2, p38, and and PGE release (Figure 6F).Theinvolvement ofERK1/ B 2 JNK1/2) and NF-κB signaling cascade is essential for 2 in COX-2 expression and PGE release was furthe 2 ET-1-inducedCOX-2geneexpressionandPGE release. confirmed by transfection of cells with p42 siRNA 2 ET-1activatesETreceptorsubtypes(ET andET )which (Figure 4D). These results are consistent with those of A B are coupled to various G proteins such as G and G and obtained with COX-2 expression induced by BK, throm- q i then lead to multiple signaling pathways and regulate di- bin,orET-1invariouscelltypes[17,26,28].Additionally, verse cellular functions [7,20-22]. Thus, we first demon- we found that expression of COX-2 and release of PGE 2 strated a significant expression of ET receptor in mouse induced by ET-1 were also attenuated by the inhibitor of B bEnd.3cells(Figure2A).TheinvolvementofET receptors p38 MAPK or JNK1/2. Pretreatment with SB202190 or B in these responses is confirmed by that pretreatment with SP600125 both markedly reduced ET-1-induced ex- BQ-788 (an ET receptor antagonist) reduced the ET-1- pression of COX-2 protein and mRNA (Figure 4A and B induced COX-2 protein and mRNA expression (Figure 2B 4B), promoter activity (Figure 6D), and PGE release 2 and 2C), promoter activity (Figure 6D), and PGE release (Figure 6F). Moreover, we also demonstrated that ET-1 2 (Figure6F),butnotbyanET receptorantagonistBQ-123. stimulates phosphorylation of p38 MAPK and JNK via A Subsequently, we confirmed these results by transfection anET -dependent manner(Figure4Cand4E).Similarly, B with ET siRNA (Figure 2D), suggesting that ET receptor we further confirmed these results by transfection with B B predominantly mediates ET-1-induced COX-2 expression siRNA for p38 MAPK or JNK1 that attenuated ET-1- and PGE release in bEnd.3 cells. Next, several subtypes of induced COX-2 expression (Figure 4D). These data 2 GproteinsarepotentiallyimplicatedinET-1-inducedCOX- clearly indicated that in bEnd.3 cells, three MAPK cas- 2 expression. We use GPA2 (a G protein antagonist) and cades (i.e. ERK1/2, p38 MAPK, and JNK1/2) are i GPA2A(aG proteinantagonist)tointerruptGproteinsig- required for ET-1-induced COX-2 expression and q naling and consequent COX-2 expression (Figure 3A). PGE release. These results are consistent with those 2 Moreover, the inhibitory effects of GPA2 and GPA2A on of obtained with up-regulation of COX-2 by ET-1 via COX-2inductionbyET-1werealso observedinitsmRNA p38 MAPK in glomerular mesangial cells or esophageal (Figure3B),promoteractivity(Figure6D),andPGE release smooth muscle cells [27,34]. For the role of JNK1/2, 2 (Figure6F),indicatingthatET-1-inducedCOX-2expression we are the first presented that JNK1/2 plays a critical and PGE release is mediated through a GPCR (i.e. ET ) role in induction of COX-2 by ET-1 in endothelial 2 B coupling to either G or G protein in bEnd.3 cells,consist- (bEnd.3) cells. i q ent with previous studies from esophageal smooth muscle It has been well established that inflammatory responses cells[34]andratbrainastrocytes[22].Incontrast,previous following exposure to extracellular stimuli are highly reports have shown that ET-1 induces COX-2 expression dependent on activation of NF-κB transcription factor, via ET receptors in peripheral lung microvascular smooth which plays an important role in regulation of several gene A muscle cells [25] and ET-1 (ET ) receptors linked to expression [40]. The 5’-flanking region of the COX-2 pro- A phospholipaseCandphospholipaseA activationandpros- moterhasbeenshowntocontainseveralbindingsequences 2 tanoidsecretion(e.g.PGE )inculturedhumanbrainmicro- for various transcription factors including NF-κB [41]. 2 vascular endothelial cells [35,36]. However, in respiratory Therefore, the regulation of COX-2 transcription may be andcardiovascularsystems,bothETreceptorsubtypes,ET mediatedbyaberrantactivationofseveraldistincttranscrip- A in particular, are involved in progression of several diseases tion factors dependent on agonists [29,42]. These reports [37,38].Theredifferencesmaybeduetocelltypespecificor suggest that NF-κB plays a critical role in the regulation of differentexperimentalconditions. COX-2 expression in the development of the inflammatory Abnormal MAPK regulation might be implicated in responses.OurdatashowedthatET-1-inducedCOX-2gene several models of CNS injury and inflammation [39]. expressionandPGE releasewassignificantlyabolishedbya 2 Several linesof evidence demonstrate that MAPKs could selective NF-κB inhibitor Bay11-7082 (Figures 5 and 6) or be activated by GPCR agonists through different NF-κBp65siRNA(Figures5Eand6F),suggestingthatNF-