Pharmacological and expression profile of the prostaglandin I2 receptor in the rat craniovascular system

Vascular Pharmacology 55 (2011) 50–58

Contents lists available at ScienceDirect

Vascular Pharmacology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / v p h

Pharmacological and expression profile of the prostaglandin I2 receptor in the rat craniovascular system Maja Myren, Jes Olesen, Saurabh Gupta ⁎ Danish Headache Center, Department of Neurology, Glostrup Research Institute, Glostrup Hospital, Faculty of Health Sciences, University of Copenhagen, DK-2600 Glostrup, Denmark

a r t i c l e

i n f o

Article history: Received 22 December 2010 Received in revised form 8 June 2011 Accepted 27 June 2011 Keywords: PGI2 IP receptor Migraine Craniovascular system PGI2 receptor antagonist

a b s t r a c t Activation of the trigeminal nerve terminals around cerebral and meningeal arteries is thought to be an important patho-mechanism in migraine. Vasodilatation of the cranial arteries may also play a role in increasing nociception. Prostaglandin I2 (PGI2) is capable of inducing a headache in healthy volunteers, a response that is likely to be mediated by the prostaglandin I2 receptor (IP). This study investigates the functional and molecular characteristics of the IP receptor in the rat craniovascular system. In the closed cranial window model, iloprost, an IP receptor agonist, dilated the rat middle meningeal artery (MMA) (Emax = 170% ± 16%; pED50 = 6.5 ± 0.2) but not the rat cerebral artery (CA) in vivo. The specific antagonist of the IP receptor, CAY10441, significantly blocked the iloprost-induced response dose-dependently, with the highest dose attenuating iloprost (1 μg kg− 1) induced dilatations by 70% (p b 0.05). CAY10441 did not have any effect on the prostaglandin E2-induced vasodilatory response, thus suggesting no interaction with EP2 and EP4 receptors. IP receptor mRNA transcripts and protein were present in meningeal as well as in cerebral rat vasculature, and localized the IP receptor protein to the smooth vasculature of the cranial arteries (MMA, MCA and basilar artery). Together, these results demonstrate that the IP receptor mediates the dilatory effect of PGI2 in the cranial vasculature in rats. Antagonism of this receptor might be of therapeutic relevance in acute migraine treatment. © 2011 Elsevier Inc. All rights reserved.

1. Introduction Prostaglandin I2 (PGI2) or prostacyclin is an important mediator of pain and inflammation (Zeilhofer, 2007). Recently, it been shown that PGI2 is capable of inducing a mild to moderate headache in healthy volunteers (Wienecke et al., 2008) and migraine-like attacks in migraineurs (Wienecke et al., 2010). Furthermore, Sarchielli et al. (2000) reported increased ictal levels of 6-keto-PGF1α, the stable product of PGI2, in the internal jugular venous blood in migraineurs during an attack. PGI2 acts on the seven trans-membrane G-proteincoupled receptor called the IP receptor (Mitchell et al., 2008; Stitham et al., 2007), also known as the IP1 receptor (Abramovitz et al., 2010), which on activation increases 3′-5′-cyclic adenosine monophosphate (cAMP) formation (Stitham et al., 2007), thus dilating the blood vessels. Migraine is a very complex disease and to date numerous mechanisms are thought to be involved in this cephalic pain. According

Abbreviations: PGI2, prostaglandin I2; IP, prostaglandin I2 receptor; PGE2, prostaglandin E2; EP2, prostaglandin E2 receptor 2; EP4, prostaglandin E2 receptor 4; MMA, middle meningeal artery; MCA, middle cerebral artery; CA, cerebral arteries; MAP, mean arterial pressure; CGRP, calcitonin gene-related peptide. ⁎ Corresponding author at: Glostrup Research Park, Department of Neurology, University of Copenhagen, Glostrup University Hospital, Nordre Ringvej 69, DK-2600 Glostrup, Denmark. Tel.: + 45 3863 3056; fax: + 45 3863 3983. E-mail address: [email protected] (S. Gupta). 1537-1891/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.vph.2011.06.004

to the trigemino-vascular theory, activation of the neurons in the trigeminal circuit may facilitate nociception. This activation may assist the release of neuropeptides stored in the perivascular nerves surrounding the blood vessels, thus in turn dilating the large cerebral and meningeal arteries. In the process mast cells are also activated, and by their degranulation PGI2 is released along with other signaling molecules. Together these cascades generate a positive feedback loop by depolarization of the neuropeptide-releasing nerves and dilatation of the vessels. As a result the pain signals are relayed via the trigeminal system to the cortex as the final destination where the pain is perceived (Brennan and Charles, 2010). Non-steroidal anti-inflammatory drugs are effective in the treatment of headache and migraine. However, regular and long-term use of these drugs is associated with adverse effects such as ulceration of the gastrointestinal tract, renal failure and cardiovascular risks (Wolfe et al., 1999; Bombardier et al., 2000) due to the non-selective blockage of the cyclooxygenase enzymes and thus prostaglandin synthesis. Therefore, a target that is more specific than cyclooxygenase inhibitors is needed to avoid these adverse effects. The objective of the present study was to pharmacologically characterize the vascular effects of the PGI2 agonist, iloprost, in rat cranial vasculature in the genuine closed cranial window model. Furthermore, we studied the in vivo potential of the specific IP receptor antagonist, CAY10441 (RO1138452), to block iloprost- and PGE2-induced responses in the rat cranial arteries. Finally, we investigated the mRNA

M. Myren et al. / Vascular Pharmacology 55 (2011) 50–58

and protein expression profiles and localization of the IP receptor in the rat middle meningeal artery (MMA) as a part of the dura mater and in the cerebral arteries (CA). 2. Material and methods All experiments were performed according to the guidelines and regulations of the Danish Animal Experimentation Inspectorate (file: 2009/561-1664) on the care and use of experimental animals. In all experiments adult male Sprague–Dawley rats were used (Taconic Europe, Ejby, Denmark). The rats were maintained in cages with a 12-h light/dark cycle and had free access to food and water. 2.1. In vivo pharmacological improved ‘closed cranial window’ model studies Twenty-four rats (276–410 g) were anesthetized with pentobarbital (Mebumal® 65 mg kg − 1) i.p. and depth of anesthesia was tested by suppression of the hind paw reflex. Anesthesia was continuously supplemented with pentobarbital (Mebumal® 20 mg kg − 1 h − 1) i.v. during the experiment. The body temperature was maintained at 37.0 ± 0.5 °C throughout the experiments using an automatic regulated heating plate (Letica HB101, Panlab, Barcelona, Spain). Following intubation the animal was mechanically ventilated by a respirator (SAR-830/AP, CWE, Ardmore, PA, USA) with 30/70% air mixture of O2/N2O, a stroke volume of 3.5–4.0 ml and a stroke rate of 55–65 per minute. The improved closed cranial window introduces an indwelling catheter (Portex, Fine Bore Polythene Tubing, inner diameter 0.4 mm, Astratech, Taastrup, Denmark) glued to a tip (outer diameter 0.3 mm) in the right carotid artery for infusion of test substance instead of the femoral artery. This improvement introduces the test molecule directly to the local cranial circulation (Gupta et al., 2010). Other catheters were placed in the left and right femoral arteries and veins for infusion of anesthetic, measurement of mean arterial blood pressure (MAP) and sampling of blood for gas tension analysis. Arterial blood samples were collected prior to, during, and at the end of each experiment, for analysis of partial pressure of oxygen (PaO2), carbon dioxide (PaCO2) and pH (ABL520, Radiometer, Brønshøj, Denmark). All values were kept within normal limits (pH 7.35–7.45, MAP 81.7–127.5 mm Hg and PaCO2 35.2– 42.7 mm Hg). The closed cranial window model experiments were performed as described previously (Gupta et al., 2010). In brief, the animal was placed in a stereotactic frame. Skin covering the dorsal surface of the skull was retracted and the connective tissue and muscle removed, leaving the right parietal bone exposed. The bone was thinned, making a window (10 × 7 mm2) by carefully drilling with a dental drill cooled by application of ice-cold synthetic interstitial fluid (SIF composition in mM: 108 NaCl, 3.48 KCl, 3.5 MgSO4, 26 NaHCO3, 11.7 NaH2PO4, 1.5 CaCl2, 9.6 sodium gluconate, 5.55 glucose and 7.6 sucrose; pH 7.4) until the middle meningeal and/or the branch of the middle cerebral artery were clearly visible while still leaving the skull intact. For visualization of the arteries, an intravital microscope (model MZ 16; Leica, Heerbrugg, Switzerland) was positioned above the cranial window. A video dimension analyser (V94, Living Systems Instrumentation©, USA) continuously monitored and measured the target artery diameter. MAP was measured constantly as well. All data were continuously displayed on a computer monitor by the data acquisition and analysis software Perisoft (Version 2.0; PerimedAB, Järfälla, Sweden). At the end of the experiments the animals were sacrificed by an overdose of pentobarbital (150 mg kg− 1 i.v.) or 1 M potassium chloride i.v. 2.1.1. Experimental design The thinning of the skull induces mechanical stress therefore after surgery the animal was left to recover for at least 60 min. The recovery period resulted in a natural constriction of the target artery and its baseline diameter was determined. In the beginning of all experi-

51

ments 150 μl of vehicle was infused via the internal carotid artery (i.c.) route at the rate of 50 μl min − 1. Each test-dose used was constituted in a volume corresponding to one-eighth of the weight of the animal in order to minimize the injected volume (Gupta et al., 2010). The dilatory capacity of the target artery was then tested with one single dose of calcitonin gene-related peptide (CGRP) (100–500 ng kg − 1) (PolyPeptide Laboratories, Strasbourg, France) (Myren et al., 2010; Gupta et al., 2010). To study the dilatory effect of iloprost or PGE2 (both from Cayman Chemicals Europe, Tallinn, Estonia) doses (at halflog increments) were administered via either i.c. infusion or topical application onto the MMA or CA. We used the IP receptor agonist, iloprost, as a substitute for PGI2, since PGI2 is chemically very unstable (Boie et al., 1994). The artery was allowed to return to baseline between each measurement. The reproducibility of the chosen iloprost dose (1 μg kg − 1) was tested by repeating the doses three times prior to use in the actual experiments. The reproducibility of the PGE2 dose (300 ng kg − 1) has been already established in a previous study (Myren et al., 2010). PGE2 was studied to explore if there was interaction between the vasodilatory receptors activated by PGE2, EP2 and EP4, with IP receptor antagonist. Using the antagonist, the %inhibition was calculated by comparison between the first (before antagonist) and the second dose of iloprost or PGE 2 (after antagonist infusion). The antagonist, CAY10441 (pKi 8.33 [IP] (Bley et al., 2006)) (Cayman Chemicals Europe, Tallinn, Estonia) also known as RO1138452, was administered via i.c. infusion as follows: infusion of the antagonist for 1 min followed by 150 μl saline over 2 min. Infusion of CAY10441 dilated the MMA and further infusion of the test substances was not initiated until the diameter of the MMA had returned to its baseline (ten minutes). Infusion of iloprost or PGE2 for 1 min was followed by 150 μl saline over 2 min. After completion of an experimental series the capacity of the artery was tested with a single dose of CGRP (100–500 ng kg− 1). The CGRP-induced dilatation was compared before and after the experiment series. 2.1.2. Analysis Analysis of the in vivo studies was performed as previously described (Myren et al., 2010). In brief, they were based on measurements of two parameters: the changes in the diameter of the artery and changes in MAP. Dilatation of the vessels (measured in arbitrary units) and changes in MAP (mm Hg) were calculated as percentage change from the baseline, which was defined as the average of the 60 s prior to the administration of the test substance. Vessel diameter was measured at the peak response occurring 1 to 2 min after drug administration. All data are expressed as the mean± S.E.M. of the percentage change from baseline values with n indicating the number of experiments. The dose response curves were analyzed to establish the maximum response (Emax) and the negative logarithm of the dose (pED50) that led to 50% of the maximal response. In case of topical application pEC50 was calculated as the concentration required to increase the artery diameter by 50% of Emax. The antagonizing effect ‘%-inhibition’ was calculated as the percentage difference between the mean maximal inhibition compared to mean iloprost response (1 μg kg− 1). To compare the consecutive responses to either saline or iloprost a one-way Anova was used followed by Dunnett's post hoc test. To compare agonist response before and after antagonist treatment a Student's paired t-test was used. 2.2. mRNA expression studies Nine rats (350–400 g) were anesthetized with pentobarbital and perfused transcardially with 300 ml ice-cold Na +-Krebs buffer (in mM: NaCl 119, NaHCO3 15, KCl 4.6, CaCl2 1.5, NaH2PO4 1.2, MgCl2 1.2 and glucose 5.5) to wash out blood from the craniovasculature. The middle cerebral and basilar arteries (pooled collectively and termed cerebral arteries) and the MMA as a part of the dura mater, were

52

M. Myren et al. / Vascular Pharmacology 55 (2011) 50–58

carefully dissected out using a microscope and immediately pooled in sterile eppendorf tubes containing a RNA stabilization solution (RNAlater; Ambion, Woodward, Austin, TX, USA). This way each tube contained tissue from three rats. The tissues were disrupted and homogenized thoroughly with ceramic beads (1.4 mm, MoBio Laboratories, Copenhagen Biotech Supply, Brønshøj, Denmark) in RTL buffer (Qiagen, Hilden, Germany) during centrifugation. Total RNA was purified from the tissues using an RNeasy Mini Kit (Qiagen, Hilden, Germany). After purification, yield and purity (our cut-off criteria was ~ 2.0) of the RNA were assessed spectrophotometrically by measuring the absorbance at 260 nm and by determining the ratio 260/280 nm, respectively. A total of three samples, each sample containing total RNA from three rats, were prepared for analysis. Quantitative real-time PCR (qPCR) was performed with LightCycler technology (Roche, Hvidovre, Denmark) and Quanti Fast SYBR Green I dye (Qiagen, Hilden, Germany). mRNA was quantified with a calibrator-normalized relative quantification approach. The detection ratios between the target gene (IP receptor) and a non-regulated reference gene (β-actin) were used to calculate the relative abundance of mRNA in each sample. A calibrator (rat heart tissue) was included in each run to provide a constant calibration point for all samples: both within the run and between related runs. Consequently, the final expression data are presented as target/reference ratio of experimental samples normalized by the target/reference ratio of the calibrator. Pre-validated gene-specific primers from Qiagen's QuantiTect Primer Assay were used to detect each mRNA transcript (IP-R: Cat. No. QT01291563, size 68 bp., NCBI GenBank Acc. No. NM_001077644) (β-actin: Cat. No. QT00193473, size 145 bp., NCBI GenBank Acc. No. NM_031144). Master mix was prepared from a mix of SYBR Green, appropriate primers, and 2 μl cDNA template according to the Qiagen protocol. Light cycler amplification program was run as previously described (Myren et al., 2010). Primer-generated PCR products from calibrator cDNA template were evaluated on an agarose (2%) gel to verify end product size and purity. All fluorescence data were processed with the LightCycler Relative Quantification software from Roche (Version 1.0). An un-paired t-test was used to determine the statistical difference between the two groups. 2.3. Western blotting Nine rats (350–400 g) were anesthetized with pentobarbital and perfused transcardially with 300 ml ice-cold Na +-Krebs buffer (in mM: NaCl 119, NaHCO3 15, KCl 4.6, CaCl2 1.5, NaH2PO4 1.2, MgCl2 1.2 and glucose 5.5) to wash out blood from the craniovasculature. The middle cerebral and basilar arteries (pooled collectively and termed cerebral arteries) and the MMA as a part of the dura mater, were carefully dissected out using a microscope and immediately pooled in sterile eppendorf tubes and quickly frozen on dry ice after dissection. This way each tube contained tissue from three rats. The frozen tissues were crushed into powder on dry ice and added to 50 μl of ice-cold lysis buffer ((10 mM Tris (pH = 7.4), 0.5% Triton X-100) including one Complete Mini EDTA-free protease inhibitor cocktail tablet and one PhosSTOP phosphatase inhibitor cocktail tablet (Roche Diagnostics, Manheim, Germany)). The samples were centrifuged at 10,000 g for 10 min at 4 °C and the supernatants stored at −80 °C. Protein concentrations were determined using a Bio-Rad protein assay (BioRad, CA, USA). Whole rat brain lysate (20 μg) were loaded as a positive control to ensure the specificity of the IP receptor antibody. In each lane 15 μg MMA/dura or 15 μg cerebral arteries lysate was loaded. The proteins were separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to a polyvinylidenedifluoride (PVDF) membrane (Biorad, CA, USA). The PVDF membranes were blocked in 5% non-fat dry milk in Tris-Buffered Saline (TBS) with 0.1% Tween-20 (TBS-T) for 1 h at room temperature.

The blot was incubated at 4 °C overnight with primary antibodies diluted in TBS-T containing 3% non-fat dry milk. The primary antibodies were directed against the murine IP receptor (1:1000, #160070, Cayman Chemicals Europe, Tallinn, Estonia) and β-actin (1:10.000, #A5441, Sigma-Aldrich, Brøndby, Denmark). The specificity of the IP receptor antibody was studied with an IP receptor blocking peptide (antibody and blocking peptide were mixed in 1:9 ratio) (Cayman Chemicals Europe, Tallinn, Estonia). The PVDF blots were washed three times in TBS-T and incubated 1 h at room temperature with secondary horseradish peroxidase (HRP) antibodies (polyclonal goat anti-rabbit immunoglobins/HRP 1:20000 (DakoCytomation, Denmark) or goat anti-mouse/HRP 1:20000, (Pierce, Rockford, IL, USA)) diluted in TBS-T. The PVDF blots were washed in TBS-T before being processed for analysis using the ECL Advanced ChemiLuminiscence detection kit (GE healthcare, UK). The blots were scanned by a LAS-4000 Electronically Cooled CCD Camera System (GE Healthcare, Brøndby, Denmark) and the chemifluorescent images were captured and stored digitally. To compare the expression profile, densitometry of the antibody signals was measured in ImageGauge 4.0 software (Fujifilm) and held relative to the β-actin signals. 2.4. Immunohistochemistry Three rats were deeply anesthetized with pentobarbital and perfused via the aorta with firstly 300 ml of 1× PBS (pH 7.4) and secondly fixated with 300 ml 4% paraformaldehyde in PBS. The MMA, MCA and basilar arteries (BA) were isolated and kept in 4% paraformaldehyde overnight. For dehydration the arteries were transferred to 30% sucrose for 24 h. The arteries were embedded in Tissue-Tek O.C.T. Compound (Sakura Finetek, Alphen aan den Rijn, Netherlands) and snap frozen on dry ice. The frozen arteries were cut at 10 μm sections in a cryostat and mounted on glass slides and kept at − 80 °C. The sections were rinsed three times with 1×PBS + 0.1% Tween20, and incubated with 1% human serum albumin in 1× PBS for one hour followed by rinsing. The sections were incubated with mouse anti-β-actin antibody (1:500; AB11003, Abcam, Cambridge, U.K.), mouse anti-endothelial nitric oxide synthase (eNOS) antibody (1:50; #610297, BD Biosciences, Brøndby, Denmark) and/or rabbit anti-IP receptor antibody (1:200; #160070, Cayman Chemicals Europe, Tallinn, Estonia) overnight at 4 °C. After removal of the primary antibody, the sections were incubated with Alexa Flour 488-conjugated donkey anti-mouse IgG (1:250) and Alexa Flour 594-conjugated donkey anti-rabbit IgG (1:250) (both from Invitrogen, Taastrup, Denmark) for 1 h at room temperature. Control staining was conducted without primary antibody, otherwise treated as described above. No staining was observed on control sections. The antibodies were detected at the appropriate wavelength on a confocal microscope (Nikon, C1 plus, Nikon Instruments Inc., USA). 2.5. Data analysis GraphPad Prism® Version 5.0 (GraphPad Software Inc., CA, USA) was used for all statistical analysis of the performed experiments and for the construction of the graphs. Statistical significance was assumed when p b 0.05. 3. Results 3.1. In vivo effects of iloprost and PGE2 on rat middle meningeal artery In the closed cranial window model, the IP receptor agonist, iloprost, (0.01–10 μg kg − 1) caused a dose-dependent dilatation of the rat MMA at half log increments (Fig. 1A) (Emax = 170 ± 16%) (mean pED50 = 6.5 ± 0.2, n = 5). The highest doses of iloprost caused a

M. Myren et al. / Vascular Pharmacology 55 (2011) 50–58

53

Fig. 1. Effects of iloprost, PGE2, ethanol, CGRP and CAY10441 on the middle meningeal artery (MMA) in anesthetized rats. (A) Internal carotid artery (i.c.) administration of iloprost (n = 5) and (B) topical application of iloprost (n = 4). (C) Effects of iloprost- (left) and calcitonin gene-related peptide (CGRP)- (right) induced dilatation on MMA after doses of ethanol (EtOH) (10% and 50%) in anesthetized rats (n = 4). (D) Effects of CAY10441 on iloprost-induced changes (1 µg/kg) in MMA in anesthetized rats (n = 5). (E) Effects of CAY10441 on PGE2-induced changes (300 ng/kg) in MMA in anesthetized rats (n = 3). Data are mean ± S.E.M. The * and # indicate statistically significant differences versus the respective controls with p b 0.05 (one-way Anova and Dunnett's post hoc test). n.s.: not significant. CGRP was used as an internal control in all the studies to monitor the viability of the experimental preparation.

transient but a significant drop in MAP (Table 1). Topical application of iloprost (0.3–100 μg ml − 1) caused concentration-dependent dilatation of the rat MMA (Fig. 1B) (Emax = 117 ± 7%) (mean pEC50 = 5.5 ± 0.1, n = 4). Topical administration did not alter MAP at any concentration. PGE2 also caused a dose-dependent dilatation of the rat MMA at half log increments (Fig. 1A) (Emax = 137 ± 9%) (mean pED50 = 7.4 ± 0.3, n = 3).

3.2. In vivo effects of iloprost on rat cerebral artery Administration of iloprost i.c. (0.1–10 μg kg − 1) did not induce any dilatation of the rat cerebral arteries, except at the highest doses (3– 10 μg kg − 1), however, the dilatory response was accompanied by a significant drop in MAP (Table 1). Topical application of iloprost (0.3– 100 μg ml − 1) did not induce any changes in the cerebral artery diameter.

Table 1 Effects of iloprost on mean arterial pressure (MAP, %) administered via the intra carotid artery. Data are mean ± S.E.M. Dose (μg kg− 1)

Middle meningeal artery

Cerebral artery

0 0.01 0.03 0.1 0.3 1 3 10

− 0.25 ± 0.8 (5) − 1.42 ± 1.9 (5) − 1.42 ± 1.5 (5) − 0.69 ± 0.5 (5) − 3.70 ± 1.6 (5) − 12.71 ± 2.3⁎ (5) − 33.60 ± 2.6⁎ (5) − 47.83 ± 3.2⁎ (5)

7.8 ± 2.8 (4) Nd Nd − 0.19 ± 0.2 (4) − 3.03 ± 3.1 (4) − 13.44 ± 4.7⁎ (4) − 32.37 ± 4.6⁎ (4) − 49.86 ± 3.5⁎ (4)

Nd: not determined. (n) = number of experiments. ⁎ p b 0.05, significantly different from MAP-vehicle (one-way ANOVA and Dunnett's post hoc test).

54

M. Myren et al. / Vascular Pharmacology 55 (2011) 50–58

Table 2 Effect of iloprost (1 μg kg− 1) and PGE2 (300 ng kg− 1) alone, and iloprost and PGE2 after a dose of CAY10441 on mean arterial pressure (MAP, %) via intra carotid administration. Data are mean ± S.E.M. CAY10441 (μg kg− 1)

Iloprost (μg kg− 1)

MAP response after Iloprost

PGE2 (μg kg− 1)

MAP response after PGE2

– 100 300

1 1 1

− 15.3 ± 5.6 (5) − 21.1 ± 5.9 (5) − 13.0 ± 4.6 (5)

0.3 0.3 0.3

− 8.8 ± 7.6 (3) – − 24.9 ± 2.6 (3)

studied. A representative recording of the experiment showing the changes in MAP and MMA diameter after CGRP (0.5 μg kg − 1), iloprost (1 μg kg − 1), and CAY10441 (300 μg kg − 1) is shown in Fig. 2. After the experimental series was complete CGRP (0.5 μg kg − 1) was capable of dilating the MMA (Fig. 2). There was no difference between the CGRPinduced dilatation of the MMA before and after the completion of experiments in the presence of CAY10441 (Fig. 1D) thus demonstrating that the effect of the IP receptor antagonist did not apply to the general vascular reactivity of the artery.

(n) = number of experiments.

3.3. In vivo effects of the IP antagonist, CAY10441, on iloprost-induced dilatation in rat middle meningeal artery

3.4. In vivo effects of the IP antagonist, CAY10441, on PGE2-induced dilatation in rat middle meningeal artery

The iloprost-induced dilatations were characterized using the specific IP receptor antagonist, CAY10441. In order to study CAY10441 a sub-maximal dose of iloprost (1 μg kg − 1) was used and the reproducibility of this dose was tested. The maximum dilatory responses were 145 ± 28%, 145 ± 35%, 153 ± 39% for the consecutive iloprost challenges (n = 3). As CAY10441 was only soluble in ethanol, a control study was made to investigate the effect of ethanol on iloprost- and CGRP-induced dilatation on the rat MMA. The results do not show any significant difference in 1 μg kg − 1 iloprost- or 0.5 μg kg − 1 CGRP-induced dilatation before and after ethanol (0– 50%; 50 μl min − 1) (Fig. 1C) (n = 4). Ethanol caused a dose-dependent but transient increase in MMA diameter and a concurrent decrease in MAP (17% when 50% ethanol was administered) from basal levels. Both these parameters reverted to basal levels within minutes after which only iloprost or CGRP was administered (50 μl min − 1). CAY10441 (100 μg kg − 1) significantly blocked the iloprost-induced dilatation and 300 μg kg − 1 CAY10441 blocked the iloprostinduced dilatation by ~ 70% on rat MMA (Fig. 1D) and attenuated the hypotensive response induced by iloprost at this dose (1 μg kg − 1) but not significantly (Table 2). Administration of 100–300 μg kg − 1 CAY10441 alone resulted in a transient increase of the MMA diameter and MAP dropped to 44 ± 5% (100 μg kg − 1 ) and 56 ± 4% (300 μg kg − 1) transiently before the iloprost-induced response was

To investigate if there was an interaction between CAY10441 and the EP2 and EP4 receptors, we tested if PGE2-induced dilatations were attenuated by pre-treatment with CAY10441. CAY10441 (300 μg kg − 1) did not affect the PGE2-induced dilatation (300 ng kg− 1) (n = 3) (Fig. 1E). Similarly, there was no difference between the CGRP-induced dilatation of the MMA before and after the infusion of CAY10441 (Fig. 1E).

3.5. mRNA expression of the IP receptor in the craniovascular system The IP receptor mRNA expression was analyzed by RT-PCR. IP receptor mRNA transcript was detected yielding a 68 bp PCR product in tissues from the rat heart (Fig. 3A), which serves as positive control. β-Actin was included to serve as a positive control and yielded a 145 bp product in the rat heart (Fig. 3A). Quantitative real-time PCR was performed to explore the IP receptor mRNA transcripts levels in the rat cranial arteries. Under identical conditions, mRNA expression was identified in the rat meningeal and cerebral arteries. The Ct (threshold cycle) for IP receptor ranged between 28 and 29 in the tested tissues. The IP receptor was significantly more expressed in the meningeal arteries than in the cerebral arteries (p b 0.05) (n = 3) (Fig. 3B).

Fig. 2. A representative trace of an iloprost vs. CAY10441 experiment. Upper panel: mean arterial pressure (MAP) in mm Hg. Lower panel: diameter of the middle meningeal artery (MMA) in arbitrary units (AU). (1) The control: calcitonin gene-related peptide (CGRP) response (0.5 μg kg− 1), (2) the iloprost response (1 μg kg− 1), (3) the effect of CAY10441 (300 μg kg− 1) given 2 min before iloprost, (4) the blocked effect of iloprost response (1 μg kg− 1), and (5) the control: CGRP response (0.5 μg kg− 1) in the end of the experiment series. Each drug was infused over a min followed by 150 μl of saline before addition of another drug.

M. Myren et al. / Vascular Pharmacology 55 (2011) 50–58

55

Fig. 3. mRNA expression of the IP receptor in the cranial arteries. (A) Expression of the IP receptor (68 bp) and β-actin (145 bp) mRNA in rat heart tissue (control). 100 bp marker (M). (B) Quantification of the IP receptor mRNA expression in rat cerebral and meningeal arteries (n = 3). The asterisk (*) indicates statistically significant difference between the two groups, p b 0.05 (Unpaired t-test). Data are mean ± S.E.M.

3.6. Protein expression of the IP receptor in the craniovascular system Western blotting was employed to verify if the IP receptor mRNA transcripts were translated into protein in the rat meningeal and cerebral arteries. The IP receptor protein was detected at the expected size, at approximately 67 kDa in the positive control tissue (whole rat brain lysate) and in the rat meningeal and cerebral arteries (n = 3–6) (Fig. 4A). β-actin was used as a loading control at 42 kDa confirming equal loading volume among the lanes (15 μg per lane) (Fig. 4A). An IP receptor blocking peptide was used to check the specificity of the antibody for the IP receptor (Fig. 4B). The measurement of the IP receptor protein relative to the β-actin protein indicates that there was a tendency towards more IP receptor protein in the meningeal arteries than in the cerebral arteries (Fig. 4C) however it was not significant. 3.7. Localization of the IP receptor in the meningeal and cerebral arteries Immunohistochemical examination using anti-IP antibody showed that the IP receptor localized to the rat MMA, MCA and BA (n = 3) (Fig. 5). Double immunostaining revealed co-localization of the IP receptor and smooth muscle β-actin (Fig. 5C, I, O), indicating that IP receptor is expressed in smooth muscle cells in the three cranial arteries. IP receptor immunoreactivities were not co-localized with the endothelial marker, eNOS, in any of the tested arteries (Fig. 5F, L, R), indicating that the IP receptor is restricted to the smooth muscle cells. 4. Discussion The trigeminal afferents innervating the cranial vasculature are important in migraine pathophysiology. Recently, it was shown that PGI2 was able to induce headache in both healthy subjects and migraineurs resulting in a concomitant dilatation of the superficial temporal artery (Wienecke et al., 2008, 2010). Therefore, we sought to explore the vasodilatory properties of the PGI2 receptor, IP, in the rat

Fig. 4. Protein expression of the IP receptor in the cranial arteries. (A) A representative blot of IP receptor protein (67 kDa) expression in rat cerebral and meningeal arteries. Whole brain lysate from rat was used as a positive control. β-Actin (42 kDa) was used as a loading control. (B) Use of a blocking peptide showed that the band at 67 kDa was specific for the IP receptor. Lane 1 (from left): blocking peptide (BP), lane 2: marker, and lane 3: IP antibody (AB) control on whole rat brain lysates. β-Actin was used as a loading control (below). (C) Densiometric measurement of the IP receptor protein relative to the β-actin protein expression in rat cerebral and meningeal arteries (n = 3–6). Data are mean ± S.E.M.

cranial arteries. Using in vivo pharmacology and molecular biological methods we for the first time show that the IP receptor agonist, iloprost, was able to cause dilatation of the rat meningeal blood vessels. The vascular responses of iloprost are very likely mediated by the IP receptor since these could be blocked by the IP receptor antagonist CAY10441. Furthermore, this antagonist had no effect on EP2 and EP4 receptors thus indicating selectivity for the IP receptor. We also demonstrated the presence and localization of IP receptor in the cranial arteries. Both topical application as well as intra-carotid administration of iloprost dose-dependently dilated the rat MMA in vivo. To best of our knowledge there are no previous studies demonstrating PGI2-induced dilatation of the MMA. However, the presented results are comparable with in vitro studies showing both iloprost- and cicaprost-induced concentration-dependent relaxations shown in isolated human cerebral arteries (pEC50 8.3 ± 0.1) (Davis et al., 2004; Boullin et al., 1979). Dilatation of MMA by topical application required higher concentrations of iloprost as compared to intra-carotid doses. The same was observed with PGE2 in the same model (Myren et al., 2010). The results show that PGI2 can penetrate the thinned skull and act on the underlying MMA hence mimicking a physiological condition where perivascular nerves could release PGI2 to act on the smooth muscle cells of the

56

M. Myren et al. / Vascular Pharmacology 55 (2011) 50–58

Fig. 5. Localization of the IP receptor in the rat middle meningeal artery (MMA), middle cerebral artery (MCA) and basilar artery (BA). IP receptors are expressed in rat MMA (A, D), MCA (G, J) and BA (M,P). Double immunostaining showed that IP receptors co-localize with smooth muscle cell marker, β-actin (B, H, N), in the merged pictures (C, I, O) in all three arteries. The endothelial cell marker endothelial nitric oxide synthase, eNOS (E, K, Q), did not co-localize with the IP receptor in the merged pictures (F, L, R) (n = 3).

craniovasculature. We also re-produced the PGE2-induced dosedependent dilatation of rat MMA, showing similar dilatory potency as reported previously (pED50 7.0 ± 0.3) (Myren et al., 2010). Iloprost was not able to induce any significant dilatations in the cerebral arteries using the same doses that dilated the MMA significantly. Although the higher doses of iloprost did induce dilatations with a concurrent fall in MAP, suggesting that the

dilatations are a result of activation of auto-regulatory mechanisms (Petersen et al., 2005). Previously, we also observed that the blood– brain barrier may prevent prostaglandin E2 from inducing a dilatory effect on rat cerebral arteries in the closed cranial window model (Myren et al., 2010). Therefore, the absence of iloprost-induced responses indicates that functional IP receptors are not found in the cerebral endothelium but are instead likely to reside in the smooth

M. Myren et al. / Vascular Pharmacology 55 (2011) 50–58

muscles of the artery. Our immunohistochemical localization of the IP receptor to the smooth muscle cells of the MCA, and not the endothelium, confirms our findings. The cerebral arteries did not change vascular tone in response to topical iloprost application and likely the meninges posed as a barrier for iloprost penetration to the cerebral vasculature. Indeed, our protein and mRNA data demonstrate that IP receptors are present in the meningeal as well as cerebral rat vasculature. Furthermore, as shown in a previous study of an open cranial window in newborn pigs, the pial artery dilated in response to PGI2 even at very low doses (1–100 ng ml − 1) (Armstead, 1995). Several other studies also demonstrate the vasodilatory capacity of cerebral arteries in various species in vitro (Davis et al., 2004; Whalley et al., 1989; Hogestatt and Uski, 1987) again indicating functional IP receptors on cerebral vasculature. We have shown that the specific IP receptor antagonist, CAY10441, dose-dependently attenuates the iloprost-induced response in the MMA. However, other studies have shown that iloprost has agonist affinity for mouse and human EP1–4 receptors in vitro (Kiriyama et al., 1997; Abramovitz et al., 2000). EP2 and EP4 are the only EP receptors capable of inducing dilatations, hence they may affect the response in the present study. Regarding rat receptors, only the EP2 receptor has been shown to have affinity for iloprost in vitro (Boie et al., 1997). It is noteworthy, that in each of these studies the affinity of iloprost for the EP2 and EP4 receptors are several log units higher than for the IP receptor. Indeed CAY10441 did not have any effect on the PGE2-induced response on the meningeal artery in the closed cranial window model, indicating lack of interaction between EP2 and EP4 receptors and the IP receptor antagonist. Thus, CAY10441 seem to be a specific IP receptor antagonist on rat receptors (Bley et al., 2006) in our hands also. CAY10441 weakly attenuated the mild hypotensive response induced by iloprost proving the point further. We employed intra-carotid infusion in order to infuse iloprost and CAY10441 directly to the local cranial circulation. Interestingly, Clark et al. (2004) have reported a pain reduction in the writhing assay after giving CAY10441 (0.3– 10 mg kg− 1 i.v.). Another group studying the role of PGI2/IP receptor in a rat model of endotoxic shock used the similar concentrations (1– 100 mg kg− 1 i.p.) (Höcherl et al., 2008). Importantly, in both these studies CAY10441 had no effect on MAP. The differences are difficult to explain as we have shown that the effect of vehicle per se did not alone cause such a substantial MAP effect. Other reasons accounting for the differences could be due to different routes of administration of the antagonist or first pass metabolism by the liver (in the case of the i. p. route). Notwithstanding the adverse effect of CAY10441 on MAP, CAY10441-induced antagonism of the IP receptor provides a potential target for therapeutic intervention in migraine. It is worth mentioning that IP receptors are present in the coronary vasculature (Oida et al., 1995; Merritt et al., 1991), therefore the possible coronary side effect profile should be monitored carefully while developing IP antagonist as an anti-migraine agent. There is similar side effect profile of prevalent anti-migraine agents, i.e. triptans (5-HT1B/1D agonists) and triptans are contraindicated in the migraine patients with cardiovascular conditions (Chan et al., 2011). Interestingly, other prostaglandin receptors also have proved to have a similar dilatory function as the IP receptor, like EP2 and EP4 (Myren et al., 2010; Maubach et al., 2009). Therefore, a further development of the IP receptor antagonist, perhaps with a modified or different chemical structure, or an antagonist with affinity for more than one target of prostaglandin receptors may prove to be rewarding. Currently, non-steroidal anti-inflammatory drugs (cyclooxygenase inhibitors) are effective in relieving migraine but several adverse effects of these drugs have been reported with long-term use (Bombardier et al., 2000; Wolfe et al., 1999). Why the severity of the side-effects in these drugs is so wide spread, is easily recognizable since one cyclooxygenase inhibitor potentially can inhibit the synthesis of all the prostanoids (prostaglandin D2, E2, F2, I2, and thromboxane) thereby proving to be a very unspecific target.

57

However, it is plausible that targeting specific and more down stream pathways in the prostanoid signaling cascade will minimize the adverse effects associated with the prostanoid synthesis blockade. To our knowledge quantitative real-time PCR studies of the IP receptor mRNA expression have not previously been described in rat meningeal and cerebral arteries. By conventional RT-PCR, molecular expression of the IP receptor was shown in rat heart tissue (Oida et al., 1995). In the present study the IP receptor protein was detected in rat meningeal and cerebral arteries by western blotting technique. However, in this study there was no significant difference between the protein expressions in the two tested tissues. This divergence between the qPCR data and the western blot data may merely point to a resting-state level of IP receptor protein in the rat meningeal arteries. It is plausible that an increase in PGI2-signaling or other factors could boost IP receptor protein expression since the receptor mRNA already exists in significant levels. We co-localized the IP receptor with smooth muscle β-actin and endothelial NOS. Our results clearly demonstrate that the IP receptor immunoreactivities are localized in the smooth muscles of the rat MMA, MCA and BA. Interestingly, topical administration of iloprost to the MMA was demonstrated to promote sensitization of both Aδ- and C-fibers on meningeal nociceptors (Zhang et al., 2007). Therefore, given our results showing that iloprost topically can dilate the MMA, it is plausible that the positive feedback cycle of the trigemino-vascular theory may consist of an initial trigger that may induce the dilatation of the meningeal vasculature and thereby the sensitizing of the nociceptors. In this scenario, nearby mast cells could then in turn be activated and degranulated releasing PGI2, among other substances, hence activating the IP receptors in the vicinity on blood vessels, astrocytes and nerve terminals. Downstream signaling from the IP receptor works mainly via the cAMP-PKA pathway which has also been implicated in the generation of headache and mechanical sensitization of dural nociceptors (Levy and Strassman, 2002; Lassen et al., 2002). Consequently, stimulation of cultured rat trigeminal ganglia neurons in vitro with iloprost or prostacyclin caused a significant release of CGRP (Jenkins et al., 2001). Likewise, it was shown in rat sensory neurons that the PGI2 release of substance P and CGRP was cAMP dependent (Hingtgen et al., 1995). Thus, the release of neuropeptides, like CGRP and substance P, may be one of the downstream effects of the cAMP-PKA pathway proving that prostaglandins may be directly involved in the well accepted CGRPmediated migraine pathway. In conclusion, the present study shows that the IP receptor is one of the major dilatory receptors in the rat meningeal artery. The IP receptor was detected on mRNA and protein levels in the rat cerebral and meningeal vasculature, and localized in the smooth muscle cells of these arteries. The dilatory response of the meningeal artery was attenuated by a selective IP receptor antagonist. The vasodilator effects of PGI2 could be one of the causes of the clinically documented headache-inducing effect of this compound.

Acknowledgments We thank Michael Baun for the skilful technical aid regarding isolation of the rat cranial arteries, and Marianne N. P. Rasmussen and Lars Kruse at the Department of Experimental Medicine for the professional support with the immunofluorescence staining and imaging. The study was supported by grants from the Lundbeck Foundation (LUCENS), the Augustinus Foundation, the A.P. Møller Foundation for the Advancement of Medical Science, the Danish Medical Council, the Danish agency for Science Technology and Innovation (No-271-07-0773), and the Novo Nordisk Foundation. The study was also funded by a PhD scholarship from the Faculty of Health Sciences, University of Copenhagen.

58

M. Myren et al. / Vascular Pharmacology 55 (2011) 50–58

References Abramovitz, M., Adam, M., Boie, Y., Carriere, M., Denis, D., Godbout, C., Lamontagne, S., Rochette, C., Sawyer, N., Tremblay, N.M., Belley, M., Gallant, M., Dufresne, C., Gareau, Y., Ruel, R., Juteau, H., Labelle, M., Ouimet, N., Metters, K.M., 2000. The utilization of recombinant prostanoid receptors to determine the affinities and selectivities of prostaglandins and related analogs. Biochim. Biophys. Acta 1483, 285–293. Abramovitz, M., Breyer, R.M., Coleman, R.A., Hills, R., Jones, R.L., Narumiya, S., Woodward, D.F., 2010. Prostanoid receptors: IP1.IUPHAR database. http://www.iuphar-db.org/ DATABASE/ObjectDisplayForward?familyId=58&objectId=345 (Last accessed June 8, 2011). Armstead, W.M., 1995. Role of nitric oxide and cAMP in prostaglandin-induced pial arterial vasodilation. Am. J. Physiol. 268, H1436–H1440. Bley, K.R., Bhattacharya, A., Daniels, D.V., Gever, J., Jahangir, A., O'Yang, C., Smith, S., Srinivasan, D., Ford, A.P., Jett, M.F., 2006. RO1138452 and RO3244794: characterization of structurally distinct, potent and selective IP (prostacyclin) receptor antagonists. Br. J. Pharmacol. 147, 335–345. Boie, Y., Rushmore, T.H., Darmon-Goodwin, A., Grygorczyk, R., Slipetz, D.M., Metters, K. M., Abramovitz, M., 1994. Cloning and expression of a cDNA for the human prostanoid IP receptor. J. Biol. Chem. 269, 12173–12178. Boie, Y., Stocco, R., Sawyer, N., Slipetz, D.M., Ungrin, M.D., Neuschafer-Rube, F., Puschel, G.P., Metters, K.M., Abramovitz, M., 1997. Molecular cloning and characterization of the four rat prostaglandin E2 prostanoid receptor subtypes. Eur. J. Pharmacol. 340, 227–241. Bombardier, C., Laine, L., Reicin, A., Shapiro, D., Burgos-Vargas, R., Davis, B., Day, R., Ferraz, M.B., Hawkey, C.J., Hochberg, M.C., Kvien, T.K., Schnitzer, T.J., VIGOR Study Group, 2000. Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. N. Engl. J. Med. 343 (1520–8), 2. Boullin, D.J., Bunting, S., Blaso, W.P., Hunt, T.M., Moncada, S., 1979. Responses of human and baboon arteries to prostaglandin endoperoxides and biologically generated and synthetic prostacyclin: their relevance to cerebral arterial spasm in man. Br. J. Clin. Pharmacol. 7, 139–147. Brennan, K.C., Charles, A., 2010. An update on the blood vessel in migraine. Curr. Opin. Neurol. 23, 266–274. Chan, K.Y., Vermeersch, S., de Hoon, J., Villalon, C.M., Maassenvandenbrink, A., 2011. Potential mechanisms of prospective antimigraine drugs: a focus on vascular (side) effects. Pharmacol. Ther. 129, 332–351. Clark, R.D., Jahangir, A., Severance, D., Salazar, R., Chang, T., Chang, D., Jett, M.F., Smith, S., Bley, K., 2004. Discovery and SAR development of 2-(phenylamino) imidazolines as prostacyclin receptor antagonists [corrected] Bioorg. Med. Chem. Lett. 14, 1053–1056. Davis, R.J., Murdoch, C.E., Ali, M., Purbrick, S., Ravid, R., Baxter, G.S., Tilford, N., Sheldrick, R.L., Clark, K.L., Coleman, R.A., 2004. EP4 prostanoid receptor-mediated vasodilatation of human middle cerebral arteries. Br. J. Pharmacol. 141, 580–585. Gupta, S., Bhatt, D., Boni, L., Olesen, J., 2010. Improvement of the closed cranial window model in rats by intracarotid infusion of signalling molecules implicated in migraine. Cephalalgia 30, 27–36. Hingtgen, C.M., Waite, K.J., Vasko, M.R., 1995. Prostaglandins facilitate peptide release from rat sensory neurons by activating the adenosine 3′,5′-cyclic monophosphate transduction cascade. J. Neurosci. 15, 5411–5419. Höcherl, K., Schmidt, C., Kurt, B., Bucher, M., 2008. Activation of the PGI2/IP system contributes to the development of circulatory failure in a rat model of endotoxic shock. Hypertension 52, 330–335.

Hogestatt, E.D., Uski, T.K., 1987. Actions of some prostaglandins and leukotrienes on rat cerebral and mesenteric arteries. Gen. Pharmacol. 18, 111–117. Jenkins, D.W., Feniuk, W., Humphrey, P.P., 2001. Characterization of the prostanoid receptor types involved in mediating calcitonin gene-related peptide release from cultured rat trigeminal neurones. Br. J. Pharmacol. 134, 1296–1302. Kiriyama, M., Ushikubi, F., Kobayashi, T., Hirata, M., Sugimoto, Y., Narumiya, S., 1997. Ligand binding specificities of the eight types and subtypes of the mouse prostanoid receptors expressed in Chinese hamster ovary cells. Br. J. Pharmacol. 122, 217–224. Lassen, L.H., Haderslev, P.A., Jacobsen, V.B., Iversen, H.K., Sperling, B., Olesen, J., 2002. CGRP may play a causative role in migraine. Cephalalgia 22, 54–61. Levy, D., Strassman, A.M., 2002. Distinct sensitizing effects of the cAMP-PKA second messenger cascade on rat dural mechanonociceptors. J. Physiol. 538, 483–493. Maubach, K.A., Davis, R.J., Clark, D.E., Fenton, G., Lockey, P.M., Clark, K.L., Oxford, A.W., Hagan, R.M., Routledge, C., Coleman, R.A., 2009. BGC20-1531, a novel, potent and selective prostanoid EP receptor antagonist: a putative new treatment for migraine headache. Br. J. Pharmacol. 156, 316–327. Merritt, J.E., Brown, A.M., Bund, S., Cooper, D.G., Egan, J.W., Hallam, T.J., Heagerty, A.M., Hickey, D.M., Kaumann, A.J., Keen, M., 1991. Primate vascular responses to octimibate, a non-prostanoid agonist at the prostacyclin receptor. Br. J. Pharmacol. 102, 260–266. Mitchell, J.A., Ali, F., Bailey, L., Moreno, L., Harrington, L.S., 2008. Role of nitric oxide and prostacyclin as vasoactive hormones released by the endothelium. Exp. Physiol. 93, 141–147. Myren, M., Baun, M., Ploug, K.B., Jansen-Olesen, I., Olesen, J., Gupta, S., 2010. Functional and molecular characterization of prostaglandin E2 dilatory receptors in the rat craniovascular system in relevance to migraine. Cephalalgia 30, 1110–1122. Oida, H., Namba, T., Sugimoto, Y., Ushikubi, F., Ohishi, H., Ichikawa, A., Narumiya, S., 1995. In situ hybridization studies of prostacyclin receptor mRNA expression in various mouse organs. Br. J. Pharmacol. 116, 2828–2837. Petersen, K.A., Dyrby, L., Williamson, D., Edvinsson, L., Olesen, J., 2005. Effect of hypotension and carbon dioxide changes in an improved genuine closed cranial window rat model. Cephalalgia 25, 23–29. Sarchielli, P., Alberti, A., Codini, M., Floridi, A., Gallai, V., 2000. Nitric oxide metabolites, prostaglandins and trigeminal vasoactive peptides in internal jugular vein blood during spontaneous migraine attacks. Cephalalgia 20, 907–918. Stitham, J., Arehart, E.J., Gleim, S.R., Douville, K.L., Hwa, J., 2007. Human prostacyclin receptor structure and function from naturally-occurring and synthetic mutations. Prostaglandins Other Lipid Mediators 82, 95–108. Whalley, E.T., Schilling, L., Wahl, M., 1989. Cerebrovascular effects of prostanoids: invitro studies in feline middle cerebral and basilar artery. Prostaglandins 38, 625–634. Wienecke, T., Olesen, J., Oturai, P.S., Ashina, M., 2008. Prostacyclin (epoprostenol) induces headache in healthy subjects. Pain 139, 106–116. Wienecke, T., Olesen, J., Ashina, M., 2010. Prostaglandin I(2) (epoprostenol) triggers migraine-like attacks in migraineurs. Cephalalgia 30, 179–190. Wolfe, M.M., Lichtenstein, D.R., Singh, G., 1999. Gastrointestinal toxicity of nonsteroidal antiinflammatory drugs. N. Engl. J. Med. 340, 1888–1899. Zeilhofer, H.U., 2007. Prostanoids in nociception and pain. Biochem. Pharmacol. 73, 165–174. Zhang, X.C., Strassman, A.M., Burstein, R., Levy, D., 2007. Sensitization and activation of intracranial meningeal nociceptors by mast cell mediators. J. Pharmacol. Exp. Ther. 322, 806–812.