Dihydropyridine agonist Bay K 8644 inhibits platelet activation by competitive antagonism of thromboxane A2-prostaglandin H2 receptor

The critical role of myosin IIA in platelet internal contraction

BACKGROUND

Shape change and centralization of granules surrounded by a microtubular coil (internal contraction) are among the earliest morphologic changes observed following platelet activation. Myosin IIA contributes to initiation of platelet shape change, but its role in internal contraction has not been defined.

OBJECTIVE

To define the contribution of myosin IIA to platelet internal contraction.

METHODS

Aspirin-treated platelets suspended in calcium-free buffer were activated with a low concentration (25 nm) of the thromboxane A(2) analog U46619 which initiated shape change and internal contraction via a Rho kinase pathway. Shape change and internal contraction were assessed by aggregometry and transmission electron microscopy (TEM), and Rho activation and myosin regulatory light chain (MRLC) phosphorylation were studied concurrently.

RESULTS AND CONCLUSIONS

Low-concentration blebbistatin (10 microm) inhibited internal contraction in the majority of platelets with minimal inhibition of shape change without significant suppression of MRLC phosphorylation. Higher blebbistatin concentrations (25-100 microm) produced concentration-dependent inhibition of aggregation, shape change, Rho activation, and MRLC phosphorylation. These data demonstrate: (i) direct platelet myosin IIA participation in internal contraction; and (ii) inhibition of Rho activation and MRLC phosphorylation by >10 microm blebbistatin.

Platelet prostanoid receptors

Prostaglandins (PGs) and thromboxanes are important modulators of platelet activation, and there is strong evidence to support the existence of distinct thromboxane, prostacyclin, PGD2 and PGE2 receptors on the platelet plasma membrane. In this review, each of these platelet prostanoid receptors is discussed in detail, with respect to their receptor pharmacology, molecular biology and signal transduction, and as to any therapeutic implications of the development of specific agonists and/or antagonists. In addition, it considers the possibility that there are separate vascular receptors for 8-epi PGF2 alpha, which are not present on the platelet.

Characteristics of protein kinase C-independent exocytosis in human platelets

We evaluated the characteristics of the protein kinase C (PKC)-independent mechanism for ATP release in platelet-rich plasma. When ADP (10 microM) and U46619 (1 microM) were both added as agonists, a significant release was observed immediately after stimulation. The PKC inhibitor, Ro-31-7549 (10 microM), or a cyclooxygenase inhibitor, aspirin (400 microM) or indomethacin (20 microM), partially inhibited ATP release with little effect on platelet aggregation. The ATP release observed in the presence of Ro-31-7549 was abolished by a cyclooxygenase inhibitor or by preventing aggregation without stirring. In the nonstirred condition, thromboxane B2 formation was reduced by 93%. When sodium arachidonate (1 mM) rather than U46619 was used with ADP, ATP release in the presence of Ro-31-7549 was abolished by stopping the stirring with no effect on thromboxane B2 formation. In contrast, ADP/U46619-induced ATP release observed in the presence of aspirin was only partially inhibited when the stirring was stopped. This release was also inhibited dose-dependently by Ro-31-7549 at concentrations between 1 and 10 microM. These results suggest that PKC-independent ATP-release in this system requires aggregation and is inhibited by a cyclooxygenase inhibitor, while PKC-dependent exocytosis is insensitive to aggregation and a cyclooxygenase inhibitor.

Inhibitory effects of efonidipine hydrochloride on contraction induced by several vasoconstrictors in porcine coronary artery: comparison with effects of nifedipine and nisoldipine

We studied the effects of efonidipine hydrochloride (efonidipine), a 1,4-dihydropyridine derivative, on contractions induced by high-K+ solution (high K+), serotonin (5-HT), U46619, which is a stable analog of thromboxane A2, and endothelin-1 (ET-1) in comparison with those of nifedipine and nisoldipine in porcine coronary arteries. The effects of the drugs were compared after 1- and 3-h incubations. Efonidipine, nifedipine, and nisoldipine each inhibited the contractions induced by these vasoconstrictors. The inhibition of high-K(+)- and 5-HT-induced contractions by efonidipine, but not by nifedipine and nisoldipine, increased when the incubation time was prolonged, whereas the inhibition of U46619- and ET-1-induced contractions was not altered. The potency of efonidipine on U46619- and ET-1-induced contractions was greater than that of nifedipine and equivalent to that of nisoldipine. Thus the inhibitory effect of efonidipine on U46619- and ET-1-induced contractions seems to be stronger than its effects on high-K(+)- or 5-HT-induced contractions, in contrast to the effects of other dihydropyridines. In an additional series of experiments, efonidipine did not inhibit U46619-induced contractions in Ca2(+)-free solution or in the presence of nifedipine. Moreover, efonidipine did not inhibit the specific binding of [3H]SQ 29,548, a thromboxane A2 antagonist, to porcine coronary arterial membrane. Therefore we think that the inhibitory effect of efonidipine on contractions induced by vasoconstrictors was caused by blockade of Ca2+ influx through L-type Ca2+ channels. However, some unknown mechanism(s) in addition to this effect on Ca2+ channels may contribute to the effect of efonidipine on U46619- and ET-1-induced contractions.

Two thromboxane A2 receptor isoforms in human platelets. Opposite coupling to adenylyl cyclase with different sensitivity to Arg60 to Leu mutation

Thromboxane A2 (TXA2) receptor is a key molecule in hemostasis as its abnormality leads to bleeding disorders. Two isoforms of the human TXA2 receptor have been cloned; one from placenta and the other from endothelium, here referred to as TXR alpha and TXR beta, respectively. These isoforms differ only in their carboxyl-terminal tails. We report that both isoforms are present in human platelets. The two isoforms expressed in cultured cells show similar ligand binding characteristics and phospholipase C (PLC) activation but oppositely regulate adenylyl cyclase activity; TXR alpha activates adenylyl cyclase, while TXR beta inhibits it. The Arg60 to Leu mutant of TXR alpha, which has been shown to impair PLC activation (Hirata, T., A. Kakizuka, F. Ushikubi, I. Fuse, M. Okuma, and S. Narumiya. 1994. J. Clin. Invest. 94: 1662-1667), also impairs adenylyl cyclase stimulation, whereas that of TXR beta retains its activity to inhibit adenylyl cyclase. These findings suggest that the pathway linked to adenylyl cyclase inhibition might be involved in some of the TXA2-induced platelet responses such as shape change and phospholipase A2 activation which remain unaffected in the patients with this mutation.

Effects of bepridil on silent myocardial ischemia and eicosanoid metabolism in chronic stable angina pectoris after healing of myocardial infarction

To investigate the effects of bepridil on silent myocardial ischemia and on eicosanoid metabolism, 10 patients with chronic stable angina underwent exercise treadmill testing and 48-hour ambulatory electrocardiographic monitoring both before and after 4 weeks of bepridil administration (150 mg/day). Fasting venous levels of thromboxane B2, 6-keto-prostaglandin F1 alpha, and leukotriene C4 were measured by radioimmunoassay. Bepridil decreased heart rate responses to daily activities during ambulatory monitoring, and significantly (p < 0.05) reduced the median frequency and duration of silent myocardial ischemic episodes (from 5.5 to 0 events/48 hours and from 86 to 0 minutes/48 hours respectively). Bepridil significantly decreased the blood pressure heart rate product at peak exercise and significantly prolonged the mean exercise tolerance time (from 456.6 to 527.0 second). Bepridil also significantly decreased the plasma levels of thromboxane B2 and leukotriene C4 at rest. These results suggest that bepridil may reduce silent myocardial ischemic episodes either by the reduction of cardiac oxygen demand during daily activities and exercise stress, or by controlling coronary and systemic vasomotor tone. The drug also has a salutary effect on eicosanoid metabolism, to which its efficacy on silent myocardial ischemic episodes may be related.

Comparative effects of the dihydropyridine-type calcium-agonists (-)-S-Bay K 8644, (+/-)-Bay-W 5035 and (+/-)-Bay-T 5006 on human platelet aggregability

1. Human platelet aggregation induced by collagen is concentration-dependently inhibited by dihydropyridine (DHP)-type calcium(Ca)-agonists. 2. There was no significant difference between the maximal anti-aggregatory effects or the anti-aggregatory potencies of (-)-S-Bay-K 8644 (EC50: 5.3 +/- 1.5 x 10(-5) M), (+/-)-Bay-W 5035 (EC50: 14.9 +/- 8.8 x 10(-5) M) or (+/-)-Bay-T 5006 (EC50: 2.7 +/- 1.9 x 10(-5) M) (P > 0.05). 3. Antiaggregatory effects of DHP-type Ca-agonists seem to be independent of Ca-channel activation.

Thromboxane-insensitive dog platelets have impaired activation of phospholipase C due to receptor-linked G protein dysfunction

Human platelet thromboxane A2/prostaglandin H2 (TXA2/PGH2) receptors are linked to phosphoinositide-specific phospholipase C (PI-PLC) via a G protein tentatively identified as a member of the Gq class. In contrast, platelet thrombin receptors appear to activate PI-PLC via other unidentified G proteins. Platelets from most dogs are TXA2 insensitive (TXA2-); i.e., they do not aggregate irreversibly or secrete although they bind TXA2, but they respond normally to thrombin. In contrast, a minority of dogs have TXA2-sensitive (TXA2+) platelets that are responsive to TXA2. To determine the mechanism responsible for TXA2- platelets, we evaluated receptor activation of PI-PLC. Equilibrium binding of TXA2/PGH2 receptor agonists, [125I]BOP and [3H]U46619, and antagonist, [3H]SQ29,548, revealed comparable high-affinity binding to TXA2-, TXA2+, and human platelets. U46619-induced PI-PLC activation was impaired in TXA2- platelets as evidenced by reduced (a) phosphorylation of the 47-kD substrate of protein kinase C, (b) phosphatidic acid (PA) formation, (c) rise in cytosolic calcium concentration, and (d) inositol 1,4,5 trisphosphate (IP3) formation, while thrombin-induced PI-PLC activation was not impaired. GTPase activity stimulated by U46619, but not by thrombin, was markedly reduced in TXA2- platelets. Antisera to Gq class alpha subunits abolished U46619-induced GTPase activity in TXA2-, TXA2+, and human platelets. Direct G protein stimulation by GTP gamma S yielded significantly less PA and IP3 in TXA2- platelets. Immunotransfer blotting revealed comparable quantities of Gq class alpha-subunits in all three platelet types. Thus, TXA2- dog platelets have impaired PI-PLC activation in response to TXA2/PGH2 receptor agonists secondary to G protein dysfunction, presumably involving a member of the Gq class.

Response of a human megakaryocytic cell line to thrombin: increase in intracellular free calcium and mitogen release

The CHRF-288-11 cell line has been previously shown to exhibit properties consistent with a megakaryocytic origin. The response of these cells to thrombin has now been investigated. Thrombin treatment of CHRF-288-11 cells results in both an increase in intracellular free calcium levels and secretion of mitogenic activity and beta-thromboglobulin. Cell viability is not affected. The mitogenic activity released from the cells is due primarily to the presence of basic fibroblast growth factor. Immunohistochemical data indicate a packaging of basic fibroblast growth factor into granular structures. Trypsin and phorbol 12-myristate 13-acetate also initiate release of mitogenic activity from this cell line, whereas under non-stirred conditions collagen and ADP do not. Through measurements of intracellular calcium levels it was determined that thrombin pretreatment of cells ablates a further response to thrombin, but does not block an increase in intracellular calcium levels due to trypsin. This suggests that these two agonists may act through different mechanisms. The thrombin-induced release reaction is inhibited almost completely by the reagents hirudin and dipyridamole, and only partially by indomethacin. These data indicate that the CHRF-288-11 cell line should provide an excellent model system in which to study the packaging of factors into granules which undergo regulated release.

Purification of the human blood platelet thromboxane A2/prostaglandin H2 receptor protein

The human platelet thromboxane A2/prostaglandin H2 receptor has been purified 6100-fold to apparent homogeneity by a three-step chromatographic procedure with an overall yield of 6%. A 6-fold purification of the receptor was first achieved by chromatography of 3-[(3-cholamidopropyl)dimethyl-ammonio]-1-propanesulfonate (CHAPS)-solubilized membrane proteins from human platelets on a diethylaminoethyl (DEAE)-Sepharose column. The DEAE eluate fractions containing receptor activity were then applied to a newly developed affinity column using the cyclohexyl derivative of SQ30,741 (SQ31,491) as the immobilized ligand. Elution of the receptor from the affinity column with BM13.177 yielded a further purification of 1700-fold. An additional 4-fold receptor purification from the affinity column eluate was achieved by HPLC using GPC 500 and GPC 100 columns connected in tandem. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis and silver staining of the HPLC eluate containing purified receptor revealed a single, distinct band with a molecular weight of 55,000. The receptor binding activity was detected with [3H]SQ29,548 using a newly developed binding assay which involved immobilization of the receptor on polyethyleneimine-treated glass fiber filters. The binding of [3H]SQ29,548 to the purified receptor was time dependent, saturable, reversible and highly specific. Unlabeled SQ29,548, BM13.505, and U46619 (but not thromboxane B2 or 6-keto prostaglandin F1 alpha) competed for [3H]SQ29,548 binding to the purified receptor in a concentration-dependent manner. Scatchard analysis of [3H]SQ29,548 binding to the purified receptor revealed the presence of a single class of high-affinity binding sites, with a Kd of 4 nM and a Bmax of 17 nmol/mg protein.

Effects of dihydropyridines and inorganic calcium blockers on aggregation and on intracellular free calcium in platelets

[Ca2+]i increase is necessary in physiological platelet activity, particularly aggregation and release. The increase of [Ca2+]i observed during platelet activation depends in part on Ca2+ influx from the extracellular medium. The participation of voltage-operated Ca2+ channels as a pathway for Ca2+ entry is controversial. In the present study we have attempted to reinvestigate this problem by measuring aggregation and [Ca2+]i changes in platelets activated by ADP or thrombin and incubated with organic or inorganic blockers of calcium channels. The main findings of the present paper can be summarized as follows: (i) Ni2+, Co2+ and Mn2+, well known inorganic blockers of Ca2+ channels, inhibited platelet aggregation induced by ADP or thrombin in a dose-dependent manner, Ni2+ being the most effective agent. (ii) Thrombin induced a rise in free [Ca2+]i in platelets incubated both in 1 mmol/l Ca(2+)-containing medium and in nominally Ca(2+)-free medium; the rise of free [Ca2+]i was in the first case up to 370 +/- 31 nmol/l and in the second case up to 242 +/- 26 nmol/l, indicating that this observed difference was due to Ca2+ entry from the extracellular medium. Co2+ and Ni2+ abolished that difference by inhibiting Ca2+ influx. (iii) Nisoldipine, nitrendipine and nimodipine (10-50 nmol/l) inhibited in a dose-dependent manner platelet aggregation induced by either ADP or thrombin in platelets incubated in normal-Ca2+ normal-K+ medium, also, aggregation was inhibited to a similar extent in platelets incubated in normal-Ca2+ high-K+ medium. (iv) Nisoldipine--the most effective dihydropyridine to inhibit platelet aggregation--also inhibited Ca2+ influx in platelets incubated in normal-Ca2+ medium, either in normal-K+ or high-K+ media. Our data support the existence of voltage-operated, dihydropyridine-sensitive calcium channels (L-type) and a physiological role for them in platelet function.

Interactions of dihydropyridine Ca2+ channel agonists with the human platelet thromboxane A2/prostaglandin H2 receptor

The specific interactions at the human platelet thromboxane A2/prostaglandin H2 (TXA2/PGH2) receptor by four 1-4 dihydropyridine (DHP) agonists were studied. Using competition equilibrium binding assays with the TXA2/PGH2 receptor agonist [125I]BOP and the antagonist [125I]PTA-OH, the affinities of racemic BAY K 8644 (BAY), racemic CGP 28392 (CGP) and (+) and (-) SDZ 202-791 (SDZ) for the TXA2/PGH2 receptor were determined. The rank order potencies for competition were BAY greater than (-)SDZ greater than CGP greater than or equal to (+)SDZ. Bay, CGP and SDZ (stereoselectively) inhibited specific incorporation of the TXA2/PGH2 receptor photoaffinity probe [125I]PTA Azido into three proteins associated with the TXA2/PGH2 receptor with Mr of 43, 39 and 27 kDa as assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis autoradiography. Using the fluorescent Ca2+ probe Fura-2, it was observed that I-BOP failed to stimulate classical divalent cation channels. However, SDZ stereoselectively inhibited the rise in [Ca2+]i induced by I-BOP while not affecting that induced by thrombin. Although DHPs specifically and stereoselectively interact with the TXA2/PGH2 receptor on human platelets, the TXA2/PGH2 receptor-mediated rise in [Ca2+]i is not through stimulation of a classical divalent cation channel.

Receptor-mediated mitogenic effect of thromboxane A2 in vascular smooth muscle cells

The effects of thromboxane A2 (TXA2) on the proliferation of vascular smooth muscles cells (VSMC) were examined using primary cultures of VSMC from rat aorta. U46619, a stable TXA2 mimetic, stimulated DNA synthesis of VSMC only in the presence of insulin. The effect was concentration-dependent with a half-maximal effect obtained at approximately 1 x 10(-8) M. The mitogenic effect of U46619 was larger than that of endothelin, another mitogen derived from endothelium. Among several TXA2/PGH2 analogs, the proliferative activity was detected only in the agonists, and not in the antagonists or in the metabolite of TXA2. A series of TXA2/PHG2 receptor antagonists completely suppressed the U46619-stimulated DNA synthesis as well as the [3H]SQ29,548 binding to the TXA2/PGH2 receptors in VSMC. The rank order of binding affinities to the receptors among the respective antagonists correlated well with the potencies for suppression of the proliferative effects of U46619. The mitogenic effects of U46619 were also attenuated by the presence of calcium antagonists. U46619 caused activation of phospholipase C with the production of inositol trisphosphate, leading to increases in the intracellular free Ca2+ concentration as measured with the fluorescent indicator fura-2. These results suggest that TXA2 induces mitogenic effects on VSMC through binding to its specific receptors. This effect of TXA2 on the proliferation of VSMC may be related to the development of atherosclerosis.

Adenosine potentiates the inhibitory effects of calcium channel antagonists on human platelet aggregation induced by thromboxane A2 or U46619

Calcium channel antagonists inhibit platelet function in vitro and ex vivo, but the mechanism responsible has not been clearly defined. The concentrations of these agents required to inhibit platelet aggregation in vitro are several fold higher than those attained in vivo. Adenosine, a known inhibitor of platelet function, is produced in large quantities in ischemic myocardium. In order to test the hypothesis that adenosine may potentiate the platelet-inhibitory effects of calcium channel antagonists, we studied the effect of adenosine plus nifedipine, verapamil or diltiazem on human platelet aggregation induced by thromboxane A2 or the stable endoperoxide/thromboxane A2 mimic, U46619 +/- epinephrine. Adenosine, in concentrations achieved in the plasma during myocardial ischemia (0.01-0.1 microM), enhanced the inhibitory effects of nifedipine, verapamil and diltiazem on platelet aggregation 5-100 fold. The same concentrations of adenosine alone did not inhibit platelet aggregation. In the presence of non-inhibitory concentrations of adenosine, nifedipine, in concentrations approaching those attained in vivo following standard therapeutic doses (as low as 0.29 microM), significantly inhibited thromboxane A2-induced platelet aggregation. Therefore, adenosine potentiates the in vitro inhibitory effects of calcium channel antagonists on platelet aggregation induced by thromboxane A2 or thromboxane A2 plus epinephrine. These results suggest that adenosine production by ischemic myocardium may augment the inhibitory effect of calcium channel antagonists on platelets.

Distinct platelet thromboxane A2/prostaglandin H2 receptor subtypes. A radioligand binding study of human platelets

Thromboxane A2 (TXA2) and prostaglandin H2 (PGH2) may aggregate platelets via a common membrane receptor(s). To further characterize this receptor, binding of the radiolabeled TXA2/PGH2 mimetic [125I]BOP to washed human platelets (WP) was investigated. [125I]BOP was competitively displaced from its platelet binding site by stable TXA2/PGH2 analogues. Competition curves were shallow with Hill coefficients of -0.73 +/- 0.05 (P less than 0.001 different from unity) (90 +/- 1% specific binding). Scatchard plots were curvilinear and most consistent with two binding sites; a high-affinity site with Kd of 234 +/- 103 pM, Bmax of 0.7 +/- 0.3 pM/mg protein (180 +/- 87 sites/WP), and a lower affinity site with Kd of 2.31 +/- 0.86 nM, Bmax of 2.2 +/- 0.3 pM/mg protein (666 +/- 65 sites/WP). [125I]BOP association and dissociation kinetics gave a Kd of 157 pM without evidence of negative cooperativity. The EC50 for I-BOP-induced initial Ca2+ increase was 209 +/- 24 pM, shape change was 263 +/- 65 pM, and aggregation was 4.4 +/- 0.5 nM. Parallel binding studies using the TXA2/PGH2 receptor antagonist [125I]PTA-OH showed a single binding site. The rank order for TXA2/PGH2 analogues to displace [125I]PTA-OH was identical to that for [125I]BOP. These studies indicate that [125I]BOP binds to two distinct sites on human platelets that may represent platelet TXA2/PGH2 receptor subtypes. The close correlation of IC50 values for I-BOP-induced platelet shape change and aggregation with the two Kds for [125I]BOP binding suggests that these platelet responses may be independently mediated by the two putative receptors.

Mechanisms of vascular graft thrombosis: role of altered canine platelet sensitivity to thromboxane

Using the standard turbidimetric method of platelet aggregation and quantitation of platelet secretion with 14C-Serotonin, we have examined the responsiveness of the platelets of mongrel dogs to arachidonic acid (AA), and the thromboxane agonist U46619 in the presence and absence of a subthreshold concentration of epinephrine. In response to stimulation with 750 microM AA, the platelets of 18 dogs produced irreversible aggregation (Group I), the platelets of 22 dogs showed, at most, reversible aggregation (Group II), while the platelets of 8 dogs demonstrated no aggregatory response (Group III). In the presence of AA and a subthreshold concentration of epinephrine (0.5 microM), the platelets of all three groups demonstrated enhanced aggregatory and secretory responses although the extent of 14C-Serotonin secretion differed significantly between all three groups. These in vitro differences in platelet aggregation correlate with the in vivo deposition of platelets onto synthetic vascular grafts and the maintenance of graft patency. When stimulated with 0.5 microM U46619 and a subthreshold concentration of epinephrine, the platelets of 97% Group I dogs and 75% of Group II dogs exhibited irreversible aggregation, while the platelets of all Group III dogs showed only reversible aggregation. In addition, significant differences in the extent of 14C-Serotonin secretion to this combination of agonists were observed between groups. Further examination of the specific effects of U46619 on canine platelets revealed that although the aggregatory and secretory responses to U46619 vary between the different canine platelet populations, the threshold concentration of U46619 required to produce platelet shape change is identical among all groups. Quantitation of the stable metabolite of AA produced via the cyclooxygenase pathway, thromboxane B2 (TxB2), revealed no significant differences in the production of TxB2 by the platelets of these different populations upon stimulation with AA. Our results suggest that the mechanisms underlying the differences in responsiveness of canine platelets to AA, are likely due to differences in sensitivity of canine platelets to TxA2, and may be localized to the mechanism responsible for mediating platelet aggregation and secretion in response to TxA2.

A common binding site for primary prostanoids in vascular smooth muscles: a definitive discrimination of the binding for thromboxane A2/prostaglandin H2 receptor agonist from its antagonist

Differences in binding characteristics between agonists and antagonists for the thromboxane A2/prostaglandin H2 (TXA2/PGH2) receptor were examined in rat cultured vascular smooth muscle cells (VSMC). Scatchard analysis indicated the existence of two binding sites for the TXA2/PGH2 agonist, whereas a single class of recognition sites for the receptor antagonists were observed with approximately the same maximum binding capacity (Bmax) as a high-affinity binding site of the agonist. Weak binding inhibition by approx. 100 nM of primary prostanoids (PGE1, PGF2 alpha and PGD2) was detected only with the TXA2/PGH2 agonist, and not with the antagonist. Primary prostanoids as well as TXA2/PGH2 agonists (U46619 and STA2) suppressed the [3H]PGF2 alpha and [3H]PGE1 binding with almost the same potency, whereas TXA2/PGH2 antagonists (S-145, SQ29,548 and ONO3708) did not. The Bmax value of the binding sites was roughly identical in PGF2 alpha, PGE1 and a low-affinity binding site of U46619. These results suggest the existence of two binding sites for TXA2/PGH2 in VSMC, i.e., a high-affinity binding site corresponding to that of the TXA2/PGH2 antagonists and a low-affinity binding site in common with primary prostanoids.