Prostaglandin E2 Induces Vascular Relaxation by E-Prostanoid 4 Receptor-Mediated Activation of Endothelial Nitric Oxide Synthase

Home , Thromboxane receptor

Ana-Marija Hristovska, Lasse E. Rasmussen, Pernille B.L. Hansen, Susan S. Nielsen, Rolf M. Nu¨sing, Shuh Narumiya, Paul Vanhoutte, Ole Skøtt, Boye L. Jensen

Abstract—The present experiments were designed to test the hypothesis that prostaglandin (PG) E2 causes vasodilatation through activation of endothelial NO synthase (eNOS). Aortic rings from mice with targeted deletion of eNOS and

E-prostanoid (EP) receptors were used for contraction studies. Blood pressure changes in response to PGE2 were measured in conscious mice. Single doses of PGE2 caused concentration-dependent relaxations during contractions to ϭ Ϫ8 phenylephrine (EC50 5*10 mol/L). Relaxation after PGE2 was absent in rings without endothelium and in rings from Ϫ Ϫ eNOS / mice and was abolished by NG-nitro-L-arginine methyl ester and the soluble guanylate cyclase inhibitor 1,2,4 1H -oxadiazolo-[4,3-a]quinoxalin-1-one. In PGE2-relaxed aortic rings, the cGMP content increased significantly. Ϫ8 PGE2-induced relaxations were abolished by the EP4 receptor antagonist AE3–208 (10 mol/L) and mimicked by an EP4 agonist (AE1–329, 10Ϫ7 mol/L) in the presence of endothelium and eNOS only. Relaxations were attenuated significantly in rings from EP4Ϫ/Ϫ mice but normal in EP2Ϫ/Ϫ. Inhibitors of the cAMP-protein kinase A pathway Ϫ8 attenuated, whereas the inhibitor of protein phosphatase 1C, calyculin (10 mol/L), abolished the PGE2-mediated 495 Ϫ/Ϫ relaxation. In aortic rings, PGE2 dephosphorylated eNOS at Thr . Chronically catheterized eNOS mice were hypertensive (137Ϯ3.6 mm Hg, nϭ13, versus 101Ϯ3.9 mm Hg, nϭ9) and exhibited a lower sensitivity of blood

pressure reduction in response to PGE2 compared with wild-type mice. There was no difference in the blood pressure response to nifedipine. These findings show that PGE2 elicits EP4 receptor-mediated, endothelium-dependent stimulation of eNOS activity by dephosphorylation at Thr495 resulting in guanylyl cyclase–dependent vasorelaxation and accumulation of cGMP in aortic rings. (Hypertension. 2007;50:525-530.) Key Words: cyclooxygenase Ⅲ cGMP Ⅲ hypertension Ⅲ phosphorylation Ⅲ thromboxane

rostaglandin (PG) E2 is a physiologically important va- precipitates hypertension, which would be compatible with a 1 10,11 Psodilator in several vascular beds. PGE2 can bind to and serial coupling between the pathways also at this site. stimulate a family of specific E-prostanoid (EP) receptors However, it is not known whether there exists a direct with Ն4 distinct subtypes, EP1 to EP4. Of those, the EP2 and interplay between PGE2 and NO formation in the vasculature, 12 EP4 receptors are functionally associated with vasodilata- as shown previously to be the case for PGD2. PGE2 EP2 and tion.1 EP2 receptors are less abundant than EP4 in most EP4 receptors are coupled to adenylyl cyclase.1 The EP4 tissues but are involved in the systemic vasodepressor re- receptor also activates a phosphatidylinositol 3-kinase and 13 sponse to PGE2 infusion, and targeted deletion of EP2 results protein kinase B/Akt. Both the cAMP/protein kinase A 2,3 in salt-sensitive hypertension. PGE2 signaling through the (PKA) and phosphatidylinositol 3-kinase-protein kinase EP4 receptor contributes to the systemic vasodepressor re- B/Akt pathways may augment endothelial NO synthase 4,5 sponse to PGE2 in mice, attenuates ischemia-reperfusion (eNOS) activity at the posttranslational level. Thus, recipro- injury in the heart,6 and maintains the patency of the ductus cal phosphorylation at eNOS-Ser1177 and dephosphorylation at arteriosus in fetal life.7 Experiments with inhibitors of NO eNOS-Thr495 (human sequences) control enzyme activity.14–16 synthase have provided data to suggest that NO is an Whether the phosphorylation state of eNOS in the vasculature extracellullar mediator for PGE2-mediated vasodilatation in is altered by PGE2 is unknown. The present study was various vascular beds.8,9 Local infusion of cyclooxygenase designed to test the hypothesis that vasodilatation mediated inhibitors or NO synthase inhibitors in the renal medulla by PGE2 is caused by EP4 receptor-mediated phosphoryla-

Received February 14, 2007; first decision March 12, 2007; revision accepted June 21, 2007. From the Department of Physiology and Pharmacology (A-M.H., L.E.R., P.B.L.H., S.S.N., O.S., B.L.J.), University of Southern Denmark, Odense, Denmark; Institute of Clinical Pharmacology (R.M.N.), Johann Wolfgang Goethe-University, Frankfurt, Germany; Department of Pharmacology (P.V.), University of Hong Kong, Hong Kong; and the Department of Pharmacology (S.N.), Kyoto University Faculty of Medicine, Kyoto, Japan. Correspondence to Boye L. Jensen, Department of Physiology and Pharmacology, University of Southern Denmark, Winslowparken 21, 3, DK-5000 Odense C, Denmark. E-mail: [email protected] © 2007 American Heart Association, Inc. Hypertension is available at http://hyper.ahajournals.org DOI: 10.1161/HYPERTENSIONAHA.107.088948 525 526 Hypertension September 2007 tion/dephosphorylation of eNOS with subsequent increased production of NO. The experiments were performed on rings of isolated aortae from wild-type mice and mice with targeted disruption of eNOS and EP receptors. In addition, PGE2 was infused to conscious unstressed wild-type mice and eNOSϪ/Ϫ mice with chronic indwelling catheters to study, at the systemic level, the dependency of PGE2-mediated blood pressure changes on eNOS.

Materials and Methods Isometric Force Measurements in Aorta Rings Please see the online data supplement available at http://hyper. ahajournals.org for details on mouse strains. Two rings were prepared per mouse. Four rings (Ϸ3-mm long) per experiment were suspended in a Halpern-Mulvany myograph (model 610 mol/L, Danish Myo Technology A/S) using 200-␮m pins for recording of isometric force (PowerLab, ADI Instruments). Rings incubated at

37°C in physiological salt solution equilibrated with 5% CO2 in air at a pH of 7.4. The rings were normalized at resting tension of Ϸ23 mN and allowed to equilibrate for 15 minutes. Please see the data supplement for details, drugs, and reagents. Measurements of eNOS Phosphorylation and cGMP Content Aortic rings were mounted as described and contracted with phen- ϫ Ϫ7 Ϫ7 ylephrine (3 10 mol/L). PGE2 (10 mol/L) or vehicle was added, Ͼ and preparations that relaxed 50% to PGE2 and their controls were frozen rapidly in liquid nitrogen. Five rings were pooled per condition. The tissue was homogenized and divided in 2 fractions, 1 for cyclic nucleotide measurement (enzyme immunoassay kits 5810213 and 581001, Cayman Chemical) and 1 for Western blotting (anti-phospho-eNOS-Thr495 1:2000, Upstate; anti-phospho-Ser1177 1:200, Cell Signaling; and anti-␣ actin 1:1000, Abcam). Please see the data supplement for details on Western blotting and PCR. Blood Pressure Measurement Catheters were placed in the femoral artery and vein for measurement of arterial blood pressure and intravenous infusions, respectively. The catheters were produced as described by Mattson.17 Mice Figure 1. A, Concentration-response curve to PGE2 on phenyl- recovered for 4 days before the experiments were performed. On the Ϫ ephrine-constricted mouse aortic rings. At 10 9 mol/L there was day of the experiment, the artery line was connected to a pressure signiﬁcant relaxation (Pϭ0.01; 1 sample t test; nϭ6). B, Bar transducer (Fo¨hr Medical Instruments GMBH) connected to an Ϫ7 graphs illustrate average PGE2-induced (10 mol/L) relaxations amplifier. The measured arterial blood pressure was converted into a of aortic rings from wild-type control mice (WT; nϭ7); in digital signal by an analog to digital converter (National Instruments, denuded rings (without endothelium; nϭ6); in the presence of Ϫ Ϫ USB 6008) and read by a computer using Laboratory View (National the eNOS inhibitor L-NAME (nϭ11); in rings from eNOS / mice Instruments) software. Data were collected at 200 Hz. Systolic, (nϭ6); and in the presence of soluble guanylase cyclase inhibi- diastolic, and mean blood pressure and heart rate (HR) were tor 1H1,2,4-oxadiazolo-[4,3-a]quinoxalin-1-one (nϭ6). *Statistically determined. For detailed protocols, please see the data supplement. signiﬁcant difference between the respective series and WT (PϽ0.001; ANOVA with posthoc unpaired t test with Bonferroni Data Analysis and Statistics correction). C, cGMP concentration in pooled mouse aortic Ͻ Constriction was expressed either relative to the maximal constric- rings exposed to PGE2 and vehicle. *P 0.001, unpaired t test; nϭ5 separate preparations at each condition. tion of the respective substance set to be 100% or relative to phenylephrine-induced constriction. Relaxation was expressed as percentage of phenylephrine-induced tone. Data are shown as 7.45Ϯ0.19 (nϭ5; data not shown). Acetylcholine (10Ϫ6 meanϮSEM. For comparison of single and group mean values, mol/L) caused relaxation of preconstricted rings (78.5Ϯ3.1%; 1-way and 2-way ANOVA followed by Bonferroni’s multiple nϭ5; data not shown). Phenylephrine at EC (3ϫ10Ϫ7 mol/L) comparison test were performed, respectively, using GraphPad Prism 80 software (GraphPad Prism). One sample t test was used for compar- was used in subsequent experiments to constrict the rings. ing a single mean with a hypothetical value. Values of PϽ0.05 were Addition of PGE2 as a single bolus to constricted rings caused considered to indicate statistically significant differences. a rapid and transient relaxation that was present over the full concentration range tested (Յ10Ϫ5 mol/L; Figure 1A). At 10Ϫ7 Results mol/L, PGE2 caused a relaxation of 45.5Ϯ6.3% (nϭ6), and Concentration-Response to PGE2 and Roles of this concentration was used in subsequent experiments to Endothelium and NO Synthase identify the receptor(s) and cellular transduction pathway Phenylephrine caused concentration-dependent, sustained involved. Maximal relaxation (69Ϯ7.5%) was observed at Ϫ6 contractions of mouse aortic rings with a mean –log EC50 of 10 mol/L of PGE2. Hristovska et al PGE2 Stimulates eNOS Through EP4 527

Figure 2. A, left, Effect of PGE2 in rings from control mice (wild-type [WT]; nϭ7), in the presence of EP4 receptor antagonist (AE3-208; nϭ8) and in rings from EP4Ϫ/Ϫ mice (nϭ4). Right, Effect of an EP4-receptor agonist (AE1-329; 10Ϫ7 mol/L) in rings from WT mice Ϫ Ϫ (nϭ11), from eNOS / mice (nϭ4), and from WT mice in the presence of the NOS inhibitor L-NAME (nϭ8). *PϽ0.001 (ANOVA and posthoc unpaired t test with Bonferroni correction). #PϽ0.05, 1 sample t test. B, Effect of PGE2 in aortic rings from control mice (WT) and in rings prepared from EP2Ϫ/Ϫ mice (nϭ4). Effect of the EP2-agonist butaprost (10Ϫ6 mol/L) in constricted rings from WT mice (nϭ7). *PϽ0.001 (ANOVA and posthoc unpaired t test with Bonferroni correction). #PϽ0.05, 1 sample t test. C, Lanes 1 to 3, PCR ampliﬁcation of cDNA from whole aorta for EP receptors and ␤-actin (positive control). Lane 4: negative control, omission of reverse transcriptase. Lane 5: positive control, mouse kidney cDNA. Lane 6: size marker, ␾X174DNA/HaeIII fragments.

Ϫ/Ϫ When PGE2 was added cumulatively with no washing steps rings from eNOS mice (nϭ4) and in rings from wild-type Ϫ11 Ϫ5 Ϫ between each concentration (10 to 10 mol/L, 10 minutes mice treated with L-NAME (10 4 mol/L; 30 minutes; nϭ8; at each concentration), PGE2 had no dilatory effect but Figure 2A, right). In rings prepared from EP2 receptor- Ϫ/Ϫ Ϫ7 elicited contraction at high concentrations (in quiescent rings: deficient mice (EP2 ), PGE2 (10 mol/L) caused relax- Ϯ ϭ pEC50 4.89 0.06, n 5; and in preconstricted rings: pEC50 ations that were indistinguishable from those observed in 5.89Ϯ0.06, nϭ4; data not shown). As observed previously,18 rings from wild-type mice (Figure 2B; nϭ4). The EP2 Ϫ6 PGE2-induced contraction was inhibited by thromboxane receptor-specific agonist butaprost (10 mol/L) marginally ϭ Ϯ ϭ receptor antagonists (S18886; –log IC50 7.91 0.10, n 5; affected tone in preconstricted rings (Figure 2B; 4.8Ϯ1.4% ϭ Ϯ ϭ and SQ29548: –log IC50 7.47 0.02, n 4; data not shown). relaxation; nϭ7). PCR analysis of RNA from whole aorta PGE2-induced contractions were absent in rings from mice showed significant expression of predominantly EP4 recep- with targeted disruption of the thromboxane receptor, al- tors (Figure 2C). though they contracted to phenylephrine (data not shown; nϭ5). In subsequent experiments, a TP antagonist (S18886) Signaling Pathways for PGE -Induced Activation Ϫ 2 was present (10 7 mol/L), and only bolus applications of of eNOS prostanoids and analogs were given. The cAMP-dependent protein kinase A inhibitors Rp-8-Br- Removal of the endothelium abolished relaxations to ace- cAMPs (10Ϫ4 mol/L; 45 minutes) and KT5720 (10Ϫ6 mol/L; Ϫ6 tylcholine (10 mol/L, data not shown) and to PGE2 (Figure 45 minutes; nϭ7) significantly attenuated the PGE2-induced 1B). Subsequent exposure to the NO donor sodium nitroprus- relaxation (Figure 3A; nϭ6). The adenylate cyclase inhibitor Ϫ6 side (SNP) (10 mol/L) relaxed the same rings to the level of SQ22536 (10Ϫ4 mol/L; 45 minutes; nϭ6) yielded similar ϭ basal tone (data not shown; n 6). Incubation of endotheli- results (Figure 3A). Relaxation to acetylcholine was not um-intact rings with NG-nitro-L-arginine methyl ester (L- Ϫ4 affected significantly by blockers of the cAMP-PKA pathway NAME; 10 mol/L, 30 minutes) eliminated PGE2-induced (data not shown). In PGE -relaxed rings (50.3Ϯ3%; nϭ25), relaxations (Figure 1B; nϭ11). In rings prepared from 2 Ϫ/Ϫ cAMP was not significantly elevated compared with control eNOS mice, acetylcholine (data not shown) and PGE2 did Ϯ Ϯ ϭ Ͼ ϭ (38.5 7 pmol/mg versus 30.1 9 pmol/mg; n 5; P 0.1). not alter vascular tone (Figure 1B; n 6). The same rings Ϫ5 Ϫ6 Inhibitors of phosphatidylinositol 3-kinase (LY294002: 10 displayed complete relaxation after application of SNP (10 Ϫ Ϫ mol/L, nϭ4; and wortmannin: 10 7 and 10 6 mol/L, nϭ4 mol/L; data not shown). Inhibition of soluble guanylate Ϫ each; data not shown) and mitogen-activated protein kinase cyclase with 1H1,2,4-oxadiazolo-[4,3-a]quinoxalin-1-one (10 6 (PD98059; 2ϫ10Ϫ5 mol/L; 30 minutes; nϭ3) did not affect mol/L) prevented PGE2-mediated relaxation in a reversible the PGE2-induced relaxation significantly (Figure 3B). Incu- fashion (Figure 1B; nϭ6). In PGE2-dilated rings (48Ϯ3%; nϭ40), the cGMP content increased significantly compared bation of rings with an inhibitor of protein phosphatase 1C, 19 Ϫ8 with control rings (Figure 1C; nϭ5). calyculin A, (10 mol/L; 30 minutes) had no effect on relaxations to acetylcholine (data not shown) but abolished those to PGE (Figure 3B; nϭ11). In phenylephrine- EP Receptors Involved in the Relaxation to PGE2 2 Ϫ8 1177 An EP4 receptor-selective antagonist (AE3–208 10 mol/L; contracted mouse aortic rings, Ser phospho-eNOS was

15 minutes) abolished the relaxation to PGE2 (Figure 2A; hardly detectable by Western blotting (data not shown), Ϫ/Ϫ 495 nϭ8). In rings from EP4 mice, the relaxation to PGE2 was whereas Thr -phospho-eNOS was readily detectable (Figure attenuated significantly but not abolished (Figure 2A; nϭ4). 3C). Exposure of preconstricted aortae rings to PGE2 caused Ϫ7 495 ϭ The EP4 receptor-specific agonist AE1–329 (10 mol/L) significant dephosphorylation at Thr (Figure 3C; n 5), produced significant relaxation in rings from wild-type mice whereas Ser1177 phospho-eNOS did not change significantly (Figure 2A, right; nϭ11) but had no effect on vascular tone in (data not shown). 528 Hypertension September 2007

Figure 3. A, Effect of PGE2 in phenylephrine-constricted aortae in the presence of vehicle (nϭ7) and inhibitors of the cAMP-PKA pathway: Rp-8-Br-cAMPs (nϭ6), SQ22536 (nϭ6), and KT5720 (nϭ7). *PϽ0.001 control vs vehicle; 1-way ANOVA and posthoc t test with Bonferroni correction. B, Effect of PGE2 on force in phenylephrine-constricted vessels in the presence of the phosphatidylinositol 3-kinase inhibitor LY249002 (nϭ4); mitogen-activated protein kinase inhibitor PD98058 (nϭ3); and PP1C inhibitor calyculin A (nϭ11). *PϽ0.001 vehicle vs control, 1-way ANOVA and posthoc t test with Bonferroni correction. C, Bar graph shows the effect of PGE2 on phospho-eNOS Thr495 level in pooled aortic rings. Mean optical densityϮSEM; nϭ5 preparations per condition each consisting of 5 separate rings. *PϽ0.001, 1 sample t test.

Ϯ ϭ Arterial Blood Pressure Response to PGE2 in 83 6.7 mm Hg; n 5). There was no difference in HR in ؊/؊ ␮ Wild-Type and eNOS Mice response to 25 g/kg of PGE2 in wild-type compared with In wild-type mice, the resting mean systolic blood pressure eNOSϪ/Ϫ mice (11.5Ϯ3.4% versus 6.9Ϯ2.2% increase; ϭ ␮ was 101Ϯ3.9 mm Hg, and HR was 640Ϯ23 bpm (nϭ9). In P 0.29). A bolus of 50 g/kg of PGE2 decreased blood Ϫ Ϫ eNOSϪ/Ϫ mice, arterial pressure was elevated significantly pressure significantly in both wild-type and eNOS / mice (137Ϯ3.6 mm Hg; nϭ13; Figure 4A and 1B) and HR was not (Figure 4C). The relative blood pressure decrease to 25 ␮g/kg ␮ different from control (625Ϯ23 bpm). Injection of a saline and 50 g/kg of PGE2 was significantly larger in wild-type Ϫ Ϫ bolus had no effect on arterial blood pressure or HR in either mice compared with eNOS / mice (Figure 4C). There was wild-type or eNOSϪ/Ϫ (Figure 4B). Two consecutive applica- no significant difference in the blood pressure response to nifedipine, which lowered arterial blood pressure by 37Ϯ4% tions of PGE2 (bolus of 50 ␮g/kg; 1-hour interval) led to a Ϫ/Ϫ similar decrease in arterial blood pressure during first and and 46Ϯ6% in eNOS and wild-type mice (Figure 4C). second application in wild-type mice (pressure after PGE There was no significant difference in the time to recovery of 2 Ϯ in mm Hg; 56Ϯ2.5 and 57Ϯ0.9; nϭ4; first and second blood pressure: in wild-type mice, 3.2 1.0 minutes and Ϫ Ϫ Ϫ Ϫ / Ϯ application, respectively) and in eNOS / mice (100Ϯ9.5 and eNOS mice, 3.4 1.3 minutes. 99Ϯ4.6; nϭ6; Figure 4B). The maximal response occurred within the first minutes after the bolus injection. Dose- Discussion response experiments showed that PGE2 had no significant The present study demonstrates an endothelium- and eNOS- effect in either strain at 10 ␮g/kg (data not shown). At 25 dependent, EP4 receptor-mediated relaxation of an isolated ␮ g/kg, PGE2 caused a significant decrease in arterial blood vessel preparation by PGE2 that is associated with cGMP pressure in wild-type mice only (Figure 4C; from 100Ϯ7.0 to accumulation and dephosphorylation of eNOS at Thr495 (Fig-

Figure 4. Arterial blood pressure responses as measured by indwelling catheters in the femoral artery of conscious wild-type and Ϫ/Ϫ eNOS mice after administration of PGE2. A, Original traces depict typical responses to the bolus injection of PGE2 in wild-type mice and eNOSϪ/Ϫ mice. Mice with targeted disruption of eNOS exhibited significantly elevated baseline blood pressure. B, Bar graphs depict mean arterial blood pressure in wild-type mice (WT) and eNOSϪ/Ϫ mice. The pressure was determined in both strains at baseline, after administration of saline bolus, and in response to 2 consecutive administrations of PGE2 (50 ␮g/kg). *Statistically significant difference between wild-type and eNOSϪ/Ϫ mice at PϽ0.05. C, Bar graphs show average, relative decrease in arterial pressure in Ϫ/Ϫ response to PGE2 administration (in micrograms per kilogram) in wild-type and eNOS mice. Nif indicates nifedipine. *Statistically significant difference between wild-type and eNOSϪ/Ϫ mice at PϽ0.05 (unpaired Student t test). Hristovska et al PGE2 Stimulates eNOS Through EP4 529

It may also reflect that NO-mediated responses display segmental heterogeneity; they predominate in conduit vessels like the aorta, and eNOS is also more strongly expressed in this segment compared with smaller vessels.23,24 The in vivo experiments demonstrated a fall in arterial Ϫ/Ϫ blood pressure in response to PGE2, both in eNOS and

wild-type mice. The response to PGE2 was less pronounced in mice with targeted disruption of eNOS than in wild-type animals. This differential sensitivity is compatible with a significant contribution of eNOS to the systemic vasodilator

response to PGE2. However, the effect of PGE2 on blood pressure also involves mechanisms clearly not dependent on eNOS. Both peripheral vascular and central nervous mechanisms may be involved.25 In the mouse, EP2 and EP4

receptors contribute to the acute depressor effect of PGE2 on arterial blood pressure.4,5 Thus, EP2 receptors could be respon-

sible for the residual systemic vasodepressor action of PGE2 in eNOSϪ/Ϫ mice, and EP4 receptors associated with vascular smooth muscle in the arteriolar segments could also participate. Vasoconstriction because of cross-activation of thromboxane

Figure 5. Schematic outline of pathways activated by PGE2 in receptors by prostanoids and isoprostanes has been observed in mouse aortic rings. various preparations.18,26 No significant expression of EP receptors associated with constriction (EP1 and EP3) was observed in ure 5). The novel finding may provide an explanation for the mouse aorta in the present and a previous study.27 Moreover, the rapid effects of cyclooxygenase inhibitors or prostanoids on dissociation constant values for PGE2 at EP1 and EP3 receptors

NO-dependent physiological responses in certain vascular are significantly lower than the concentrations of PGE2 required beds.8–11 The use of an integrated vascular preparation to mediate constriction in the present study.28 These observa- showed that, after prolonged exposure to PGE2, the EP4 tions permit the conclusion that PGE2 is a low-affinity agonist at receptor-mediated relaxation waned, and at high concentra- the thromboxane receptor in the mouse aorta, as observed 18 tions of PGE2, cross-reactivity with thromboxane receptors previously in the rat aorta. was apparent, as reported also in rat aorta.18 It is well established that EP4 receptors exhibit ligand-mediated desen- Perspectives sitization primarily because of internalization.20,21 When The present data provide evidence for a novel pathway vessels had not been exposed to PGE2 before, the vessels between PGE2 and eNOS activity initiated by endothelial EP4 responded by relaxation at the full range of concentrations. receptor occupation and mediated by PP1C-dependent de- Removal of the endothelium, L-NAME treatment, or dele- phosphorylation of eNOS. This pathway is relevant not only tion of eNOS abolished the ability of PGE2 to cause relax- for physiological regulation of vascular tone and blood flow ation, so we infer that smooth-muscle EP4 or EP2 receptors but may also contribute to other vascular processes influ- are of minor significance for relaxation after PGE2 in the enced by eNOS activity, such as platelet aggregation and mouse aorta. The initial sequence of events must depend on smooth muscle cell proliferation. EP4 receptor-mediated activation of eNOS, followed by NO-mediated stimulation of soluble guanylate cyclase with Acknowledgments accumulation of cGMP in the smooth muscle (Figure 5). The We thank Lis Teusch and Mette Fredenslund for excellent technical assistance and Prof Ulf Simonsen, Aarhus University, for assistance data indicate that PGE2 activates eNOS through formation of with Western blotting protocols. We thank Prof Peter Bie and Ole cAMP by the endothelium; vasorelaxation was inhibited by 3 Madsen for assistance with blood pressure measurements. separate blockers of PKA and adenylate cyclase; calcium- coupled PGE2 EP1 receptors were not detected in the aorta, Sources of Funding and the response depended on phosphoprotein phosphatase The present work was supported by grants from the Danish Research 1C (PP1C), which is activated predominantly by the cAMP Council for Health and Disease, Danish Heart Foundation, NOVO 19 Nordisk Foundation, AJ Andersen and Hustrus Foundation, Else pathway (Figure 5). cAMP did not increase significantly in Poulsens Legacy, Danish Kidney Association Research Fund, and PGE2-relaxed vascular rings. This finding does not exclude Helen and Ejnar Bjørnows Foundation. that PGE2 induces formation of cAMP restricted to the limited pool of endothelial cells in the preparation. The notion Disclosures of cAMP-PKA–dependent activation of eNOS by PP1C- None. mediated dephosphorylation at Thr495 is consistent with data 14,19 obtained in cultured endothelial cells. PGE2 phosphory- References lates eNOS at Ser1177 in cultured human endothelial cells at 1. Narumiya S, Sugimoto Y, Ushikubi F. Prostanoid receptors: structures, properties, and functions. Physiol Rev. 1999;79:1193–1226. 22 passage 3 to 6. This discrepancy with the present study is 2. Kennedy CR, Zhang Y, Brandon S, Guan Y, Coffee K, Funk CD, likely because of the very different experimental conditions. Magnuson MA, Oates JA, Breyer MD, Breyer RM. Salt-sensitive hyper- 530 Hypertension September 2007

tension and reduced fertility in mice lacking the prostaglandin EP2 16. Lin MI, Fulton D, Babbitt R, Fleming I, Busse R, Pritchard KA Jr, Sessa WC. receptor. Nat Med. 1999;5:217–220. Phosphorylation of threonine 495 in endothelial nitric-oxide synthase coor- 3. Tilley SL, Audoly LP, Hicks EH, Kim HS, Flannery PJ, Coffman TM, dinates the coupling of L-arginine metabolism to efficient nitric oxide pro- Koller BH. Reproductive failure and reduced blood pressure in mice duction. J Biol Chem. 2003;278:44719–44726. lacking the EP2 prostaglandin E2 receptor. J Clin Invest. 1999;103: 17. Mattson DL. Long-term measurement of arterial blood pressure in con- 1539–1545. scious mice. Am J Physiol. 1998;274:R564–R570. 4. Audoly LP, Tilley SL, Goulet J, Key M, Nguyen M, Stock JL, McNeish 18. Dorn GW 2nd, Becker MW, Davis MG. Dissociation of the contractile JD, Koller BH, Coffman TM. Identification of specific EP receptors and hypertrophic effects of vasoconstrictor prostanoids in vascular responsible for the hemodynamic effects of PGE2. Am J Physiol. 1999; smooth muscle. J Biol Chem. 1992;267:24897–24905. 277:H924–H930. 19. Michell BJ, Chen Zp, Tiganis T, Stapleton D, Katsis F, Power DA, Sim 5. Zhang Y, Guan Y, Schneider A, Brandon S, Breyer RM, Breyer MD. AT, Kemp BE. Coordinated control of endothelial nitric-oxide synthase Characterization of murine vasopressor and vasodepressor prostaglandin phosphorylation by protein kinase C and the cAMP-dependent protein E(2) receptors. Hypertension. 2000;35:1129–1134. kinase. J Biol Chem. 2001;276:17625–17628. 6. Xiao CY, Yuhki K, Hara A, Fujino T, Kuriyama S, Yamada T, Takayama 20. Nishigaki N, Negishi M, Ichikawa A. Two Gs-coupled prostaglandin E K, Takahata O, Karibe H, Taniguchi T, Narumiya S, Ushikubi F. Pros- receptor subtypes, EP2 and EP4, differ in desensitization and sensitivity to taglandin E2 protects the heart from ischemia-reperfusion injury via its the mtetabolic inactivation of the agonist. Mol Pharmacol. 1996;50: receptor subtype EP4. Circulation. 2004;109:2462–2468. 1031–1037. 7. Nguyen M, Camenisch T, Snouwaert JN, Hicks E, Coffman TM, 21. Bastepe M, Ashby B. The long cytoplasmic carboxyl terminus of the Anderson PA, Malouf NN, Koller BH. The prostaglandin receptor EP4 prostaglandin E receptor EP4 subtype is essential for agonist-induced Nature 2 triggers remodelling of the cardiovascular system at birth. . desensitization. Mol Pharmacol. 1997;51:343–349. 1997;390:78–81. 22. Namkoong S, Lee SJ, Kim CK, Kim YM, Chung HT, Lee H, Han JA, Ha 8. Dumont I, Hou X, Hardy P, Peri KG, Beauchamp M, Najarian T, KS, Kwon YG, Kim YM. Prostaglandin E2 stimulates angiogenesis by Molotchnikoff S, Varma DR, Chemtob S. Developmental regulation of activating the nitric oxide/cGMP pathway in human umbilical vein en- endothelial nitric oxide synthase in cerebral vessels of newborn pig by dothelial cells. Exp Mol Med. 2005;37:588–600. prostaglandin E . J Pharmacol Exp Ther. 1999;291:627–633. 2 23. Davis RJ, Murdoch CE, Ali M, Purbrick S, Ravid R, Baxter GS, Tilford N, 9. Najarian T, Marracetylcholinee AM, Dumont I, Hardy P, Beauchamp Sheldrick RL, Clark KL, Coleman RA. EP4 prostanoid receptor-mediated MH, Hou X, Peri K, Gobeil F Jr, Varma DR, Chemtob S. Prolonged ϩ vasodilatation of human middle cerebral arteries. Br J Pharmacol. 2004;141: hypercapnia-evoked cerebral hyperemia via K channel- and prostaglan- 580–585. din E2-dependent endothelial nitric oxide synthase induction. Circ Res. 2000;87:1149–1156. 24. Shimokawa H, Yasutake H, Fujii K, Owada MK, Nakaike R, Fukumoto 10. Zewde T, Mattson DL. Inhibition of cyclooxygenase-2 in the rat renal Y, Takayanagi T, Nagao T, Egashira K, Fujishima M, Takeshita A. The medulla leads to sodium-sensitive hypertension. Hypertension. 2004;44: importance of the hyperpolarizing mechanism increases as the vessel size 424–428. decreases in endothelium-dependent relaxations in rat mesenteric circu- 11. Mattson DL, Lu S, Nakanishi K, Papanek PE, Cowley AW Jr. Effect of lation. J Cardiovasc Pharmacol. 1996;28:703–711. chronic renal medullary nitric oxide inhibition on blood pressure. Am J 25. Feuerstein G, Adelberg SA, Kopin IJ, Jacobowitz DM. Hypothalamic Physiol. 1994;266:H1918–H1926. sites for cardiovascular and sympathetic modulation by prostaglandin E2. Brain Res. 1982;231:335–342. 12. Braun M, Schror K. Prostaglandin D2 relaxes bovine coronary arteries by endothelium-dependent nitric oxide-mediated cGMP formation. Circ Res. 26. Audoly LP, Rocca B, Fabre JE, Koller BH, Thomas D, Loeb AL, 1992;71:1305–1313. Coffman TM, FitzGerald GA. Cardiovascular responses to the iso- 13. Fujino H, Regan JW. EP4 prostanoid receptor coupling to a pertussis prostanes iPF2␣-III and iPE2-III are mediated via the thromboxane A2 toxin-sensitive inhibitory G protein. Mol Pharmacol. 2006;69:5–10. receptor in vivo. Circulation. 2000;101:2833–2840. 14. Gallis B, Corthals GL, Goodlett DR, Ueba H, Kim F, Presnell SR, Figeys D, 27. Fujino T, Yuhki K, Yamada T, Hara A, Takahata O, Okada Y, Xiao CY, Ma Harrison DG, Berk BC, Aebersold R, Corson MA. Identification of flow- H, Karibe H, Iwashima Y, Fukuzawa J, Hasebe N, Kikuchi K, Narumiya S, dependent endothelial nitric-oxide synthase phosphorylation sites by mass Ushikubi F. Effects of the prostanoids on the proliferation or hypertrophy of spectrometry and regulation of phosphorylation and nitric oxide production cultured murine aortic smooth muscle cells. Br J Pharmacol. 2002;136: by the phosphatidylinositol 3-kinase inhibitor LY294002. J Biol Chem. 530–539. 1999;274:30101–30108. 28. Kiriyama M, Ushikubi F, Kobayashi T, Hirata M, Sugimoto Y, Narumiya 15. Fleming I, Fisslthaler B, Dimmeler S, Kemp BE, Busse R. Phosphory- S. Ligand binding specificities of the eight types and subtypes of the lation of Thr495 regulates Ca2ϩ/calmodulin-dependent endothelial nitric mouse prostanoid receptors expressed in Chinese hamster ovary cells. oxide synthase activity. Circ Res. 2001;88:E68–E75. Br J Pharmacol. 1997;122:217–224.