Expression, Function, and Regulation of E-Type Prostaglandin Receptors (EP3) in the Nonischemic and Ischemic Pig Heart
Abstract The action of prostacyclin, prostaglandin E1 (PGE1), and their mimetics on myocardial function includes changes in contractility, electrophysiological properties, and protection from injury caused by transient myocardial ischemia. This study was undertaken to investigate the basic properties of myocardial E-type prostaglandin (EP) receptors. Ligand binding studies using an enriched preparation of sarcolemmal membranes prepared from pig hearts revealed a single class of binding sites for [3H]PGE1, with a Kd of 3.7 nmol/L and a Bmax of 92 fmol/mg protein. Competition experiments indicated highest affinity for EPs, suggesting an EP receptor. In addition, the EP receptor subtype–selective agonists sulprostone (EP1 and EP3) and M&B 28.767 (EP3) were active, suggesting the presence of an EP3 receptor subtype. PGE1 stimulated sarcolemmal GTPase and inhibited sarcolemmal adenylyl cyclase activity, indicating EP3 receptor coupling to an inhibitory G protein (Gi). Additional in vivo experiments showed that intracoronary infusion of PGE1 (1 nmol/min) decreased isoprenaline-stimulated left ventricular contractile activity without altering systemic vascular resistance. This inhibition of β-adrenergic effects is compatible with the known myocardial anti-ischemic action of prostaglandins. Further experiments examined EP3 receptor density and G-protein coupling in sarcolemma from ischemic and reperfused ischemic myocardium. In anesthetized open-chest minipigs, occlusion of the left anterior descending coronary artery for 60 minutes increased EP3 receptor density by 50%, whereas receptor affinity was unchanged. This upregulation was prevented by pretreatment with colchicine (2 mg/kg IV), indicating microtubule-dependent receptor externalization. Northern hybridization showed comparable EP3 receptor mRNA expression in control and ischemic myocardium. The increase of receptor protein was reversed during 60 minutes of reperfusion. G-protein coupling proved to be intact in ischemic and reperfused ischemic myocardial tissue, as shown by preserved GTP-γ-S–induced decrease of [3H]PGE1 binding. These data demonstrate for the first time that myocardial receptors for PGE1 belong to the EP3 subtype. The properties of this receptor include inhibition of adenylyl cyclase and upregulation during regional myocardial ischemia, suggesting an involvement in the anti-ischemic activity of E- and I-type prostaglandins.
Prostaglandins exert various effects on the normal and ischemic heart. These include changes in contractility,1 energy metabolism,2 and, in the reperfused ischemic myocardium, a protective action2 3 in several experimental models, including the pig.4 5 6 In addition, a reduction by prostaglandins of arrhythmias induced by experimental ischemia has been repeatedly reported.7 8 9
The actions of prostaglandins are mediated by specific cell surface receptors that are coupled to defined postreceptor signal transduction pathways. Although the pharmacological characterization of prostaglandin receptors remains hampered by well-defined and high-affinity receptor antagonists, the knowledge about these receptors has dramatically increased after the sequencing and cloning of virtually all known prostaglandin receptors and clarification of their coupling to postreceptor signal transduction pathways.10
Surprisingly, little is known about the type, specificity, density, and postreceptor signaling of prostaglandin receptors in the heart, although ligand binding studies on bovine cardiac membranes11 have provided evidence that EP receptors are expressed in the heart and seem to be located on the sarcolemma. These receptors have been shown to inhibit myocardial adenylyl cyclase. This should be compatible with an EP3 receptor subtype but has not yet been confirmed by appropriate ligand binding studies. Subtypes of myocardial EP receptors may mediate at least some of the actions of E- and I-type prostaglandins on myocardial function in cardiac tissue under normoxic conditions and during myocardial ischemia and reperfusion.
The present work was designed to (1) characterize prostaglandin receptors located within the myocardium with respect to their agonist selectivity, affinity, and density and (2) examine the functional coupling of these receptors to postreceptor signal transduction pathways. Additional experiments were performed to (3) determine whether myocardial prostaglandin receptor activation alters myocardial contractile function. Since previous studies have shown that myocardial receptors with probable relevance for ischemic myocardial injury (eg, adrenergic receptors) may be upregulated during ischemia,12 13 14 this study was also designed to (4) evaluate myocardial prostaglandin receptor characteristics in ischemic and reperfused ischemic myocardial tissue.
Materials and Methods
All animal experiments were performed in accordance with the “Position of the American Heart Association on Research Animal Use” and were approved by the local governmental animal resource committee. Minipigs of either sex (Göttingen strain, 28 to 35 kg body weight) were anesthetized by intramuscular injection of 15 mg/kg ketamine and 3 mg/kg azaperon. Anesthesia was maintained by intravenous infusion of 5 mg/kg per hour metomidate, 0.8 mg/kg per hour azaperon, and 0.6 mg/kg per hour pancuronium into the left jugular vein. All animals were artificially ventilated with a mixture of room air and O2 and adjusted to an arterial Po2 between 100 and 150 mm Hg and Pco2 of 35 to 40 mm Hg. Thoracotomy was performed by midline incision of the sternum, followed by longitudinal pericardiotomy.
Fluid and electrolyte balance was maintained by intravenous infusion of Ringer’s solution supplemented with 5% glucose (100 mL/h). All animals were allowed to stabilize at least 30 minutes after the surgical preparation was completed.
Myocardial tissue was obtained by occlusion of the superior and inferior venae cavae and the ascending aorta, followed by intracoronary infusion of 1000 mL ice-cold (4°C) St Thomas’ Hospital cardioplegic solution. The hearts were then rapidly excised and frozen with clamps precooled in liquid nitrogen for further preparation of sarcolemmal membranes.
In the experiments designed to examine EP receptor properties in ischemic myocardial tissue, the LAD was carefully dissected free from the epicardium over a distance of 5 mm, halfway between its origin and apex. A silk suture (0.3-mm diameter) was placed around the vessel for later atraumatic occlusion. The animals were allowed to recover for at least 30 minutes after surgery. Then the LAD was occluded either for 15 minutes (n=6) or for 60 minutes (n=7). Thereafter, the hearts were arrested with cardioplegic solution as described above. To identify the ischemic left ventricular tissue, 20 mL of cardioplegic solution containing 20 mg of Evans blue was injected into the LAD at the site of the previous occlusion. Initial control experiments had excluded an interference of the dye with the preparation of sarcolemmal membranes and the biochemical parameters measured. Ischemic and control left ventricular myocardial tissue was separately frozen for preparation of sarcolemmal membranes.
In an additional group of pigs (n=4), regional myocardial ischemia (60 minutes) was induced as described, with the exception that the microtubule-disrupting agent colchicine (2 mg/kg IV) was administered 60 minutes before coronary occlusion. Another group of pigs (n=5) also underwent LAD occlusion for 60 minutes. Thereafter, the ligature was released, and reperfusion was allowed for 60 minutes.
Effect of Intracoronary PGE1 on Contractile Function
Further experiments (n=5) were designed to assess a potential effect of prostaglandin receptor activation on regional myocardial contractile activity. PGE1 (1 nmol/min) was infused via a 27-gauge needle into the LAD close to the origin of the first diagonal branch, a site where coronary blood flow is ≈20 mL/min. The infusion of PGE1 lasted for 10 minutes. Immediately before (control) and during the time of PGE1 administration (10 minutes), left ventricular contractility was measured by ultrasound crystals, which were placed in subepicardial and subendocardial positions (Schuessler & Associates sonomicrometer). One pair of crystals was implanted within the perfusion bed of the LAD; another was placed within the left lateral wall, apart from the perfusion bed of the LAD. Contractile function was quantified by the maximum systolic change of left ventricular wall thickness. Thirty minutes after the infusion of PGE1 was terminated, isoprenaline was infused intravenously at a dose of 3 μg/kg per minute. The inotropic action stabilized within 10 to 20 minutes. Thereafter, PGE1 infusion was repeated, and regional contractile function was recorded as before. Before and during intracoronary PGE1 infusion, cardiac output was measured by an electromagnetic blood flow probe placed around the ascending aorta. Left ventricular pressure was simultaneously measured by a catheter-tipped manometer introduced into the left ventricle via the left carotid artery. Systemic vascular resistance was calculated from peak left ventricular pressure and cardiac output using a Macintosh-based data analysis system (MacLab 16s, Wisstech).
Preparation of Sarcolemmal Membranes
Cardiac sarcolemma was prepared according to procedure II as described by Jones (1992).15 About 15 g of ischemic or nonischemic myocardial tissue was minced with scissors, suspended in 120 mL of buffer A (10 mmol/L histidine and 0.75 mol/L NaCl), and homogenized for 5 seconds at half-maximum speed with an Ultraturrax homogenizer (Janke & Kunkel). The homogenate was centrifuged at 14 000g for 20 minutes. Then the pellet was washed twice in buffer B (5 mmol/L histidine and 10 mmol/L NaHCO3), suspended in 5 mL of buffer C (10 mmol/L histidine and 300 mmol/L NaCl), and placed on top of a discontinuous sucrose gradient, consisting of two layers of buffer C (7 mL each) supplemented with of 0.25 and 0.6 mol/L sucrose, respectively. After 90 minutes of ultracentrifugation (300 000g), the membranes at the 0.25 to 0.6 mol/L sucrose interface were recovered, 3-fold diluted with sucrose-free buffer C, and sedimented (300 000g, 30 minutes). The sarcolemmal vesicles were then suspended in 200 μL aliquots (≈5 mg each) of 10 mmol/L Tris-HCl (pH 6.4) and stored frozen (−40°C) until use. All preparation steps were performed at 4°C in the presence of 0.1 mmol/L phenylmethylsulfonyl fluoride.
Ligand Binding Studies
Binding experiments were performed in 200 μL of binding buffer containing 100 mmol/L NaCl, 25 mmol/L HEPES (pH 6.5), and 25 to 50 μg membrane protein. Scatchard analysis was performed by the cold saturation technique in the presence of 3 nmol/L [3H]PGE1. Nonlabeled PGE1 was added in concentrations between 0.3 nmol/L and 1 μmol/L. Nonspecific binding was determined in the presence of 10 μmol/L PGE1. Equilibrium of binding was achieved after 60 minutes of incubation at 37°C. At this time, bound was separated from free activity by rapid filtration with a 24-well filtration device (Brandel) and by washing three times with 5 mL of ice-cold binding buffer. Radioactivity on the filters was counted with standard liquid scintillation techniques. The binding and displacement data were evaluated by computer-based nonlinear fitting analysis (Enzfitter, Biosoft). The best fit of the data was achieved when assuming a single class of binding sites. The binding data were referred to membrane protein, determined according to the method of Bradford.16
The kinetics of [3H]PGE1 binding to sarcolemmal membranes were studied according to the method described by Jaschonek et al.17 Kobs of the ligand-receptor complex was determined with [3H]PGE1 at a concentration of 5 nmol/L at 20°C. K−1 was measured as time course of [3H]PGE1 displacement after addition of 10 μmol/L PGE1 at 20°C. K+1 was calculated from the following equation: K+1=(Kobs−K−1)/ligand
Receptor-subtype specificity of binding was determined by competition of [3H]PGE1 binding with PGE0, PGE1, PGE2, sulprostone, butaprost, M&B 28.767, iloprost, cicaprost, PGD2, and SQ 29.548 (each 0.3 nmol/L to 10 μmol/L). The processing was identical to that for cold saturation with PGE1 (above).
Effect of GTP-γ-S on [3H]PGE1
Receptor–G-protein coupling was assessed by determination of the effect of GTP-γ-S on the specific binding of [3H]PGE1. Sarcolemmal membranes were incubated in binding buffer for 60 minutes at 37°C with or without 0.1 mmol/L GTP-γ-S, a concentration that can be expected to achieve complete saturation of guanine nucleotide binding sites of G-protein α subunits. Then saturation binding studies were carried out using [3H]PGE1 (0 to 20 nmol/L) as ligand. The effect of myocardial ischemia (15 or 60 minutes) and reperfusion (60 minutes) on G-protein coupling was assessed by determining the changes of specific [3H]PGE1 (3 nmol/L) binding to sarcolemmal membranes caused by a range of different GTP-γ-S concentrations (1 to 100 μmol/L).
Measurement of Na+,K+-ATPase Activity
Sarcolemmal Na+,K+-ATPase activity was determined by the K+-stimulated conversion of 4-nitrophenylphosphate to 4-nitrophenol at 400 nm using a double-beam spectrophotometer (UV-160, Shimadzu) by the procedure described by Danilenko et al.18
Measurement of GTPase Activity
Sarcolemmal GTP hydrolysis was measured as hydrolysis of [γ-32P]GTP, using the procedure of Fleming and Watanabe (1988).19
PGE1-Induced Inhibition of Sarcolemmal Adenylyl Cyclase
Sarcolemmal membranes (50 μg protein) were suspended in 100 μL of 50 mmol/L HEPES buffer (pH 7.4) containing (mmol/L) sodium EDTA 1, MgCl2 10, dithiothreitol 2, methylisobutylxanthine 2, ATP 2, creatine phosphate 4, and GTP 0.01. Creatine kinase was added at a concentration of 6.8 U/mL. Adenylyl cyclase was stimulated with forskolin (10 μmol/L). PGE1 was added at concentrations from 0.1 nmol/L to 1 μmol/L. Separate measurements showed that adenylyl cyclase inhibition was maximal at 1 μmol/L. The reaction was allowed to proceed for 10 minutes at 37°C and stopped by heat inactivation (90°C for 5 minutes). Membrane protein was sedimented at 10 000g for 10 minutes, and cAMP was determined by radioimmunoassay.
Northern Blot Analysis of Cardiac EP3 Receptor Message
Three additional pigs underwent 60 minutes of LAD occlusion. Total RNA was isolated from ischemic and control left ventricular tissue by the Trizol RNA isolation procedure (GIBCO-BRL). Poly(A+) RNA was isolated from total RNA by paramagnetic particle separation (PolyATract, Promega) and analyzed by electrophoresis on a 1% agarose/formaldehyde gel. Poly(A+) RNA (20 μg) was loaded on each lane. After transfer to Hybond N membrane (Amersham), RNA was fixed by baking at 80°C for 2 hours. The blot was hybridized overnight at high-stringency conditions (65°C, 5× SSPE, 5× Denhardt’s solution, 0.5% SDS, and 500 μg salmon sperm DNA) with a 32P random prime–labeled fragment of the porcine EP3 receptor (0.9-kb cDNA, EcoRI-BamHI; J. Meyer and K. Schrör, unpublished data, 1997) and a 1.3-kb human GAPDH20 EcoRI fragment. After hybridization, the membranes were washed twice at 37°C with 2× SSPE and 0.1% SDS and twice at 65°C with 0.2× SSPE and 0.1% SDS. The hybridization signals were detected on Kodak X-Omat film with exposition for 3 hours (GAPDH) or overnight (EP3 receptor). Densitometric analysis was performed on a PC-based gel evaluation system (Scan Pack, Biometra).
Drugs and Chemicals
PGE0, PGE1, and PGE2 were gifts from Dr P. Ney, Schwarz Pharma (Mannheim, Germany); sulprostone, iloprost and cicaprost, from Dr F. MacDonald, Schering (Berlin, Germany); butaprost, from Dr P. Gardiner, Bayer (Middlesex, UK); M&B 28.767, from Dr L. Caton, Rhone-Poulenc Rorer (Vitry sur Seine, France); and SQ 29.548, from Dr M. Ogletree, Squibb (Princeton, NJ). [3H]PGE1 and [γ-32P]GTP were purchased from DuPont de Nemours and Amersham, respectively. All other chemicals were from Sigma and Merck.
All data are mean±SEM. Different groups were compared by two-tailed t test for paired or unpaired samples, as required. More than two groups were compared by ANOVA (repeated measures) with Bonferroni’s correction. Differences between groups were considered significant at values of P<.05.
Characteristics of Sarcolemmal [3H]PGE1 Binding Sites
Saturation experiments with PGE1 indicated a Kd of 3.7±0.5 nmol/L and a density of binding sites (Bmax) of 92±15 fmol/mg protein (Fig 1⇓). At 3 nmol/L [3H]PGE1, nonspecific binding was ≈20% of total binding. Since the Scatchard transformation of the saturation binding data was best fit by linear regression, a single class of binding sites can be assumed. Hill transformation of the competition between labeled and unlabeled PGE1 revealed a slope of −0.95. This value is close to −1, indicating competition for independent binding sites.
Additional kinetic experiments evaluated the time course of association and dissociation of [3H]PGE1 on porcine sarcolemma. Kobs was 1.67×10−3 · s−1, and K−1 was 0.30×10−3 · s−1, resulting in K+1 of 2.74×105 mol−1 · s−1. The calculated Kd value was 1.1 nmol/L.
Specificity of PGE1 Binding to Porcine Sarcolemma
PGE2 displaced [3H]PGE1 (3 nmol/L) from the sarcolemmal binding sites with a potency similar to PGE1. (Fig 2⇓, Table 1⇓). The calculated pD2 values for PGE1 and PGE2 were almost identical. The PGI2 mimetics iloprost and cicaprost were at least one order of magnitude less active in displacing [3H]PGE1 but reached the same maximum as PGE1 and PGE2. In addition, PGE0 was equieffective to PGE1 and PGE2. PGD2 displaced [3H]PGE1 at micromolar concentrations. The thromboxane receptor antagonist SQ 29.548 did not displace [3H]PGE1 (not shown). These data collectively indicate specificity of the sarcolemmal binding site for EPs.
Since distinct subtypes of the EP receptor are known, it was of interest to identify the particular subtype located on porcine sarcolemma. This was studied by competition binding experiments with the EP receptor subtype–selective compounds sulprostone (EP1 and EP3), butaprost (EP2), and M&B 28.767 (EP3). In fact, sulprostone and M&B 28.767 displaced [3H]PGE1 at a potency similar to PGE1, whereas butaprost was inactive (Fig 3⇓, Table 1⇑). This indicates that the EP3 receptor subtype is present on porcine sarcolemma.
G-Protein Coupling of Sarcolemmal EP3 Receptors
[3H]PGE1 binding to sarcolemmal membranes was evaluated in the absence and presence of the stable GTP analogue GTP-γ-S (0.1 mmol/L) in order to assess coupling of sarcolemmal EP3 receptors to a G protein. As shown in Fig 4⇓, GTP-γ-S caused a distinct shift of the saturation binding curve to the right and a decreased negative slope in the Scatchard plot. The calculated Kd values were 2.3±0.1 and 5.7±0.5 nmol/L without and with GTP-γ-S, respectively, indicating that the GTP analogue decreased ligand affinity. This suggests that sarcolemmal EP3 receptors are coupled to a G protein. Obviously, GTP-γ-S did not alter the number of binding sites (Bmax), since the x-axis intercept of the Scatchard-transformed data was the same in saturation experiments with or without the addition of GTP-γ-S.
Further evidence in support of a G protein–coupled EP receptor was obtained from the effect of PGE1 on sarcolemmal hydrolysis of [γ-32P]GTP. PGE1 (1 nmol/L to 1 μmol/L) significantly increased GTPase activity. The EC50 for this effect was 4.7±0.3 nmol/L (n=4), and GTPase activity was stimulated to a maximum of 1 pmol/mg per minute at 100 nmol/L PGE1 (Fig 5⇓).
Since adenylyl cyclase inhibitory G proteins (Gi) represent an important, though not the only, class of EP receptor–coupled G proteins, it was of interest to determine the effect of PGE1 on sarcolemmal adenylyl cyclase activity. It was found that PGE1 inhibited forskolin (10 μmol/L)–stimulated activity of sarcolemmal adenylyl cyclase in a concentration-dependent manner (Table 2⇓). The effect was statistically significant at 10 nmol/L and higher. PGE1 did not alter basal adenylyl cyclase activity, determined in the absence of forskolin.
Functional Response to Myocardial EP3 Receptor Activation
Intracoronary infusion of PGE1 (1 nmol/min) did not alter regional contractile function, as determined by sonomicrometric measurement of left ventricular wall thickness in the perfusion bed of the LAD, distal to the site of PGE1 infusion (Fig 6⇓). However, in the presence of β-adrenergic stimulation by isoprenaline (3 μg/kg per minute IV), PGE1 caused a marked and significant delay of left ventricular systolic wall thickening. This effect was constant for the duration of intracoronary PGE1 infusion (10 minutes) and completely reversible within 30 minutes after the PGE1 infusion was stopped (not shown).
The negative inotropic action of PGE1 in the presence of β-adrenergic stimulation was not due to a change in peripheral vascular resistance (ie, systemic vasodilation), which was comparable before and during PGE1 infusion (3180±250 dyne · s · cm−5 before versus 3110±270 dyne · s · cm−5 after 2 minutes of intracoronary PGE1 infusion, P>.05). In contrast to isoprenaline, the cAMP-independent positive inotropic action of ouabain was not altered by PGE1 (not shown).
The present data collectively indicate that normoxic myocardium contains a homogeneous class of EP3 receptors coupled to an adenylyl cyclase inhibitory G protein, allowing for a functionally relevant attenuation of the inotropic response to β-adrenergic stimulation. This may contribute or even explain the pharmacologically important reduction of ischemic myocardial injury by PGE1 and other EP (or combined EP and IP) receptor agonists. Further experiments were therefore conducted to evaluate the regulation and G-protein coupling of EP3 receptors at short-term (15-minute) and long-term (60-minute) myocardial ischemia and reperfusion (60 minutes).
Effects of Acute Myocardial Ischemia on Sarcolemmal EP3 Receptor Density
In the experiments with 15 minutes of LAD occlusion, sarcolemmal EP receptor density (Bmax) was not significantly different between control myocardial tissue (74±13 fmol/mg protein) and the ischemic area (62±3 fmol/mg). Furthermore, receptor affinity (Kd) was comparable (3.0±1.6 and 2.6±1.4 nmol/L, respectively).
In contrast, LAD occlusion for 60 minutes caused a significant increase of Bmax from 83±8 (control myocardium distant from the ischemic area) to 126±21 fmol/mg protein, equivalent to an increase by 52±13% (P<.005, Fig 7⇓). This elevation of receptor density appeared to be specific for EP3 receptors, because the activity of K+-stimulated Na+,K+-ATPase, a marker enzyme of sarcolemmal membranes, was identical in membranes prepared from nonischemic and ischemic myocardium (Fig 7⇓). The Kd values remained unchanged in the control and ischemic areas (2.7±0.9 versus 2.9±1.8 nmol/L, respectively; P>.05). Thus, 60 minutes of regional ischemia increased the density without changing EP3 receptor affinity.
Another group of pigs also underwent a period of ischemia (60 minutes) but was pretreated with the microtubule-disrupting agent colchicine (2 mg/kg IV) 60 minutes before the LAD was occluded. Colchicine did not alter blood pressure, dP/dt, heart rate, or left ventricular end-diastolic pressure (not shown) but completely prevented the increase in receptor density (Fig 7⇑). In these experiments, receptor affinity was 4.9±1.0 nmol/L in control tissue and 4.1±1.9 nmol/L in ischemic tissue, which was not significantly different (P>.05).
When coronary occlusion (60 minutes) was followed by 60 minutes of reperfusion, sarcolemmal EP3 receptor density was 68±6 fmol/mg in nonischemic and 46±9 fmol/mg in reperfused ischemic myocardium. This decrease did not reach statistical significance (P>.05). This was accompanied by a similar decrease of myocardial K+-stimulated Na+,K+-ATPase activity. Reperfusion did not change sarcolemmal EP3 receptor affinity, as shown by identical Kd values in reperfused ischemic and nonischemic control myocardium (2.7±0.9 versus 2.7±0.5 nmol/L, respectively).
Expression of EP3 Receptor mRNA
In three experiments, poly(A+) RNA was prepared from left ventricular myocardial control tissue (left lateral wall) and from ischemic myocardium (left anterior wall) after LAD occlusion, separated on an agarose gel, and hybridized with a 32P-labeled probe of the porcine myocardial EP3 receptor. In all of these samples, distinct transcripts appeared at a size of 2.2 kb (Fig 8⇓). GAPDH was equal in control and ischemic myocardium. The densitometric analysis showed a comparable expression of EP3 receptor mRNA in the control and ischemic areas of the left ventricle. This suggests that the increased sarcolemmal EP3 receptor density during regional myocardial ischemia is not due to a transcriptional upregulation of EP3 receptor message.
G-Protein Coupling of Sarcolemmal EP3 Receptors in Ischemic Myocardium
Regional myocardial ischemia did not impair the GTP-γ-S–induced decrease of [3H]PGE1 (5 nmol/L) binding to sarcolemmal membranes, indicating intact receptor G-protein coupling during ischemia. This is shown by a similar GTP-γ-S (100 μmol/L)–induced decrease of [3H]PGE1 binding in membranes from tissue subjected to 60 minutes of LAD occlusion compared with membranes from nonischemic control tissue (Fig 9⇓). Similar results were obtained after 15 minutes of ischemia (not shown). Interestingly, GTP-γ-S at low concentration (1 μmol/L) caused a significant (P<.05) increase of bound activity in membranes from ischemic, but not from control, tissue.
Reperfusion (60 minutes) subsequent to regional ischemia (60 minutes) was also associated with a decrease in [3H]PGE1 binding after the addition of 100 μmol/L GTP-γ-S (Fig 9⇑). This indicates that G-protein coupling of EP3 receptors also remained intact in the reperfused ischemic myocardium. Similar to ischemia alone, a lower concentration of GTP-γ-S (1 μmol/L) increased [3H]PGE1 binding. In contrast to ischemia alone, this was also seen in the nonischemic control myocardium.
The present study demonstrates specific myocardial binding sites for EPs, extending earlier data11 by the identification of a G protein–coupled EP3 receptor subtype, located on cardiac sarcolemmal membranes. Scatchard analysis of [3H]PGE1 binding revealed a Kd of 3.7 nmol/L, which agrees with the Kd values previously reported for recombinant EP receptors.21 This Kd is also consistent with the observed EC50 value for PGE1-mediated sarcolemmal GTPase activation. Although the present work shows a single homogeneous class of binding sites, bovine hearts have previously been reported to display a high-affinity and a low-affinity site for EPs.11 Based on similar affinity, the binding site found in pig sarcolemma more likely corresponds to the low-affinity site of bovine sarcolemma, since saturation requires nanomolar concentrations of PGE1. The rank order of potency for displacement of [3H]PGE1 showed highest activity for M&B 28.767, which identifies an EP3 receptor subtype. Other than M&B 28.767 and sulprostone, PGE1 has some cross affinity toward IP receptors. However, the maximum of displacement was similar for the three ligands. A contribution of IP receptors to [3H]PGE1 binding is, therefore, unlikely.
The sarcolemmal membrane preparation used in the present study may be contaminated with membranes derived from other cellular compartments and different types of tissue. Since antibodies directed against EP3 receptors are not available, this receptor cannot be localized by immunohistochemistry. Several arguments, however, suggest that the EP3 receptors described in the present study are essentially located on cardiomyocytes. First, we have used a well-characterized technique for preparation of cardiac sarcolemmal membranes.15 Second, the functional measurements showed a marked inhibitory effect of intracoronary PGE1 on the β-adrenergic inotropic response to isoprenaline. This is supported by a recent study performed with neonatal cardiomyocytes, which showed that PGE1 in nanomolar concentration inhibits β-adrenergic arrhythmogenic effects.22 Finally, a number of studies have demonstrated that prostaglandins such as PGE1, PGI2, and their mimetics1 2 3 4 5 6 7 8 9 reduce ischemic myocardial injury.
In general, EP3 receptors may activate Gs, Gi, and Gp proteins.10 The present study demonstrates G-protein coupling of sarcolemmal EP3 receptors by a PGE1-induced increase in sarcolemmal GTPase activity and a GTP-γ-S–induced shift from one single class of high-affinity binding sites toward a single class of binding sites with decreased affinity. This, however, does not identify the type of G protein coupled to EP3 receptors.
It has been shown that G-protein coupling of EP3 receptors depends on posttranscriptional splicing. The known splice variants of this receptor mainly differ within their C-terminal amino acid sequences,21 23 24 which appear to determine the class of G proteins that responds to receptor stimulation.21 24 A GTP-γ-S–induced shift of ligand binding, similar to the one observed in the present study, has been reported for transfected EP3 receptors, which are coupled to inhibition of adenylyl cyclase.24 Evidence that the G protein, which couples to myocardial EP3 receptors, belongs to the class of inhibitory G proteins (Gi) is provided by the inhibition of sarcolemmal adenylyl cyclase by PGE1, which is in agreement with a previous study by Lopaschuk et al.11 This is also supported by the finding that PGE1 decreases the inotropic response to β-adrenergic stimulation by isoprenaline, without altering basal contractile function. Further, pertussis toxin prevents the GTP-γ-S–induced shift of [3H]PGE1 binding (T. Hohlfeld, unpublished data, 1996).
Whereas PGE1 completely inhibited the isoproterenol-induced inotropic effect, adenylyl cyclase was inhibited only partially (60%). One possible explanation is that the in vitro determination of adenylyl cyclase activity does not ideally reflect the cAMP formation in vivo. Since EP3 receptors and adenylyl cyclase require access from different sides of the sarcolemma, an incomplete penetration of PGE1 and/or ATP across sarcolemmal vesicle membrane may impair the in vitro inhibition by PGE1 as well as overall cAMP formation.
A number of studies have shown that PGE1, PGI2, and their mimetics improve postischemic myocardial function and reduce myocardial infarct size (for review, see References 2 and 32 3 ). In addition, Hide et al25 also demonstrated that PGE1 and PGE0 cause ischemic preconditioning. Therefore, it was of particular interest to evaluate whether myocardial EP3 receptors remain intact and coupled to G proteins in ischemic myocardial tissue.
Ligand binding clearly showed that receptor affinity was unchanged at short (15-minute) and prolonged (60-minute) durations of ischemia. Receptor density remained unchanged during short-term ischemia but was significantly increased during long-term ischemia in the presence of intact G-protein coupling, as indicated by a similar decrease of [3H]PGE1 binding by the GTP analogue GTP-γ-S. The increase in receptor density cannot be attributed to an enhanced yield of sarcolemmal membranes in ischemic tissue, since the activity of Na+,K+-ATPase, a marker of sarcolemmal membranes, was not altered.
The observed increase in receptor density during coronary occlusion might be caused by de novo receptor synthesis. However, EP3 receptor mRNA expression was similar in control and ischemic myocardium, arguing against a transcriptional increase of receptor synthesis. Alternatively, an intracellular receptor pool may have been externalized during ischemia, as has been shown for the ischemia-induced upregulation of myocardial β-adrenergic receptors.12 13 14 26 27 Notably, the time course and extent of upregulation appear to be similar for β-adrenergic12 and EP3 receptors.
A pool of intracellular (“light”) membranes is a source of myocardial β-adrenergic receptor upregulation.12 Several studies have shown that microtubule-disrupting agents, such as colchicine, interfere with the translocation of adrenergic receptors between cell surface and an intracellular compartment.28 29 It may be hypothesized, therefore, that EP3 receptors are upregulated by a similar mechanism. In fact, the present study shows that colchicine prevents EP3 receptor upregulation, suggesting that microtubule-dependent externalization of EP3 receptor protein is the explanation for EP3 receptor upregulation during ischemia. It should be noted in this context that intracellular eicosanoid receptors (ie, thromboxane A2/PGH2 receptors) have been described in platelets30 and that platelet thromboxane A2/PGH2 receptors are upregulated in patients with acute myocardial infarction.31
The upregulation of myocardial EP3 receptors suggests a role in prostaglandin-mediated protection against tissue injury in reperfused ischemic myocardium. This is supported by several arguments. First, EP3 receptor–mediated inhibition of adenylyl cyclase antagonizes the detrimental effects of catecholamines released during acute myocardial ischemia,32 eg, by reducing the deleterious effects of Ca2+ overflow. Second, inhibitory G proteins may open inwardly rectifying33 34 and ATP-sensitive35 K+ channels. There seems to be a consensus that glibenclamide-sensitive K+ channels are involved in myocardial protection by ischemic preconditioning.36 Third, G proteins that belong to the Gi species are known to activate sarcolemmal Na+-Ca2+ exchange37 and thus may contribute to the control of intracellular calcium concentration.
All types of prostaglandins are formed within the heart. These include PGE238 and PGI2,6 which are the most likely candidates for endogenous EP3 receptor agonists. However, under which circumstances and by which extent PGE2 and PGI2 achieve concentrations sufficiently high to interact with EP3 receptors is not entirely clear at present. Since we6 and others39 have found that cyclooxygenase inhibition adversely affects ischemic myocardial injury, it is likely that in reperfused ischemic myocardium prostaglandin concentrations are high enough to activate cardiac EP3 receptors.
In summary, we have shown that myocardial tissue expresses EP receptors that belong to the EP3 subtype. EP3 receptor activation causes in vitro inhibition of adenylyl cyclase and decreases the inotropic response to β-adrenergic activation. In addition, this receptor subtype is subject to upregulation during acute myocardial ischemia, whereas G-protein coupling remains intact. These data offer a rational basis for understanding the cardioprotective actions of EPs. In addition, they may stimulate the further development of therapeutic strategies to reduce or prevent ischemic myocardial injury by prostaglandin mimetics, since new and highly selective agonists of EP3 receptors become increasingly available.
Selected Abbreviations and Acronyms
|EP, IP||=||E-type and I-type prostaglandin|
|K −1||=||rate constant of dissociation|
|K +1||=||forward rate constant|
|K obs||=||observed time constant of association|
|LAD||=||left anterior descending coronary artery|
|M&B 28.767||=||15α-hydroxy-9-oxo-16-phenoxy-17,18,19,20-tetranorprost-13-transenoic acid|
|PG (as prefix)||=||prostaglandin|
|SQ 29.548||=||[1S-[1α, 2α (2), 3α,4α)]]-7-[3-[[2-[(phenylamino)carbonyl]hydrazine]methyl]-7-oxabicyclo[2,2.1]hept-2yl]heptanoic acid|
This study was supported by the Deutsche Forschungsgemeinschaft (SFB 242). The authors thank Kirsten Bartkowski and Elke Vasilescu for skillful technical assistance and Erika Lohmann for competent secretarial support. We thank Bayer UK (butaprost), Rhone-Poulenc Rorer (M&B 28.767), Schering AG (iloprost, sulprostone), Schwarz Pharma (PGE1), and Squibb (SQ 29.548) for their generous supply of compounds.
- Received November 25, 1996.
- Accepted August 7, 1997.
- © 1997 American Heart Association, Inc.
Schrör K, Hohlfeld T. Inotropic actions of eicosanoids. Basic Res Cardiol. 1993;87:2-11.
Schrör K. Eicosanoids and myocardial ischemia. Basic Res Cardiol. 1986;82(suppl 1):235-243.
Schrör K, Hohlfeld T. Eicosanoids and the ischemic myocardium. In: Piper HM, ed. Pathophysiology of Severe Ischemic Myocardial Injury. Dordrecht, Netherlands: Kluwer Academic Publishers; 1990:195-217.
Stürzebecher S, McDonald FM, Grundmann G, Hartmann S, Lammert C. Myocardial ischaemia and reperfusion in the anaesthetized pig: reduction of infarct size and myocardial enzyme release by the stable prostacyclin analogue iloprost. In: Schrör K, Sinzinger H, eds. Prostaglandins in Clinical Research. 301st ed. New York, NY: Alan R Liss Inc; 1988:155-160.
Hohlfeld T, Strobach H, Schrör K. Stimulation of endogenous prostacyclin protects the reperfused ischemic pig myocardium from ischemic injury. J Pharmacol Exp Ther. 1993;264:397-405.
Mest HJ, Schrör K, Förster W. Antiarrhythmic properties of PGE2: preliminary results. In: Bergström S, Bernhard S, eds. Advances in the Biosciences: International Conference on Prostaglandins, Vienna, 1972. Oxford, UK: Pergamon; 1973:385-393.
Lopaschuk GD, Michalak M, Wandler EL, Lerner RW, Piscione TD, Coceani F, Olley PM. Prostaglandin E receptors in cardiac sarcolemma: identification and coupling to adenylate cyclase. Circ Res. 1989;65:538-545.
Maisel AS, Motulsky HJ, Insel PA. Externalization of β-adrenergic receptors promoted by myocardial ischemia. Science. 1985;230:183-186.
Strasser RH, Krimmer J, Marquetant R. Regulation of beta-adrenergic receptors: impaired desensitization in myocardial ischemia. J Cardiovasc Pharmacol. 1988;12(suppl 1):S15-S24.
Danilenko MP, Turmukhambetova VC, Yesirev OV, Tkachuk VA, Panchenko MP. Na+/K+-ATPase-G protein coupling in myocardial sarcolemma: separation and reconstitution. Am J Physiol. 1991;261:H87-H91.
Fleming JW, Watanabe AM. Muscarinic cholinergic-receptor stimulation of specific GTP hydrolysis related to adenylate cyclase activity in canine sarcolemma. Circ Res. 1988;64:340-250.
Tso JY, Sun XH, Kao TH, Reece KS, Wu R. Isolation and characterization of rat and human glyceraldehyde-3-phosphate dehydrogenase cDNAs: genomic complexity and molecular evolution of the gene. Nucleic Acids Res. 1985;13:2485-2502.
Kang JX, Yunyuan LI, Leaf A. Differential effects of various eicosanoids on the contraction of cultured neonatal rat cardiac myocytes. Prostaglandins Leukot Essent Fatty Acids. 1996;55:128. Abstract.
Sugimoto Y, Negishi M, Hayashi Y, Namba T, Honda A, Watabe A, Hirata M, Narumiya S, Ichikawa A. Two isoforms of the EP3 receptor with different carboxyl-terminal domains. J Biol Chem. 1993;268:2712-2718.
Mukherjee AL, Bush LR, McCoy KE, Duke RJ, Hagler H, Buja LM, Willerson JT. Relationship between β-adrenergic receptor numbers and physiological responses during experimental canine myocardial ischemia. Circ Res. 1982;50:735-741.
Muntz KH, Olson EG, Lariviere GR, D’Souza S, Mukherjee A, Willerson JT, Buja LM. Autoradiographic characterization of beta adrenergic receptors in coronary blood vessels and myocytes in normal and ischemic heart. J Clin Invest. 1984;73:349-357.
Yonemochi H, Saikawa T, Takakura T, Ito S, Takaki R. Effects of calcium antagonists on β-receptors of cultured cardiac myocytes isolated from neonatal rat ventricle. Circulation. 1990;81:1401-1408.
Saussy DL, Mais DE, Baron DA, Pepkowitz SH, Halushka PV. Subcellular location of a thromboxane A2/prostaglandin H2 receptor antagonist binding site in human platelets. Biochem Pharmacol. 1987;37:647-654.
Dorn GW II, Liel N, Trask JL, Mais DE, Assey ME, Halushka PV. Increased platelet thromboxane A2/prostaglandin H2 receptors in patients with acute myocardial infarction. Circulation. 1990;65:538-545.
Schrör K, Funke K. Prostaglandins and myocardial noradrenaline overflow after sympathetic nerve stimulation during ischemia and reperfusion. J Cardiovasc Pharmacol. 1985;7(suppl 5):S50-S54.
Kurachi Y, Nakajima T, Ito H. Intracellular fluoride activation of muscarinic K channel in atrial cell membrane. Circulation. 1987;76(suppl IV):IV-105. Abstract.
Matesic DF, Manning DR, Luthin GR. Tissue-dependent association of muscarinic acetylcholine receptors with guanine nucleotide-binding regulatory proteins. Mol Pharmacol. 1991;40:347-353.
Kirsch GE, Codina J, Birnbaumer L, Brown AM. Coupling of ATP-sensitive K+ channels to A1 receptors by G proteins in rat ventricular myocytes. Am J Physiol. 1990;259:H820-H826.
Gross GJ, Auchampach JA. Blockade of ATP-sensitive potassium channels prevents myocardial preconditioning in dogs. Circ Res. 1992;70:223-233.
Brechler V, Pavoine C, Lotersztajn S, Garbarz E, Pecker F. Activation of Na+/Ca2+ exchange by adenosine in ewe heart sarcolemma is mediated by a pertussis toxin-sensitive G protein. J Biol Chem. 1990;265:16851-16855.
Jugdutt BI, Grover CB, Hutchins GM, Bulkley BH, Pitt B, Becker LC. Effect of indomethacin on collateral flow and infarct size in the conscious dog. Circulation. 1997;59:734-743.