Arachidonic acid induced coronary reactions and their inhibition by docosahexaenoic acid

The objective of the present study was to further investigate the influence exerted by docosahexaenoic acid (DHA) on the coronary reactions induced in isolated perfused hearts of rats and guinea pigs by bolus doses of arachidonic acid (AA). As in previous studies, we found that AA produced a coronary constriction followed by a longer lasting dilatation. The present data demonstrate that a 5-min infusion of DHA at 0.17-0.68 microM caused a concentration-dependent inhibition of the AA-induced constriction. The vasodilatation determined by AA was also depressed, but only after about 30 min of a sustained DHA infusion. The precursor of AA, linoleic acid (LA), was also infused for about 30 min, and like DHA it inhibited the coronary reactions induced by AA. LA is not converted into AA by the isolated heart, but like DHA, was probably incorporated into the cells of the coronary vascular compartment. It is known that LA, administered "in vivo" to mammals, is converted into AA and increases the production of eicosanoids, whereas DHA does not follow this metabolic pathway. The incorporation of these essential polyunsaturated fatty acids by the isolated perfused heart would inhibit the cyclooxygenase in the coronary vessel walls, interfering with the generation of vasomotor metabolites from AA. We postulate that the systemic administration of DHA, by inhibiting the synthesis of a constrictor metabolite, could be beneficial in reducing the damage due to microvascular constriction in myocardial ischaemia.     

Functional and cellular interactions between nitric oxide and prostacyclin

Nitric oxide (NO) and prostacyclin (PGI(2)) can be released by vascular agents to synergize their effects on vascular relaxation. In the present study we assess whether NO could affect PGI(2) production. We evaluated the effect of NO on PGI(2)-mediated arachidonic acid (AA)-induced relaxation in the perfused heart. We used cultured endothelial cells to characterize the mechanism involved in the NO effect on PGI(2) synthesis. AA-induced PGI(2) synthesis was enhanced when NO synthesis was inhibited. NO inhibited AA-induced relaxation and PGI(2) release in the coronary circulation. S-Nitroso-acetyl-DL-penicillamine (SNAP) decreased PGI(2) production in cultured endothelial cells. The SNAP effect was blunted by the inhibitor of soluble guanylate cyclase (LY-83,583) and the blocker of cGMP-dependent protein kinases (H-9). Specific cyclooxygenase-1 (COX-1) immunoprecipitation was associated to co-precipitation of four proteins. COX-1 showed neither serine nor threonine phosphorylation. One of the proteins that co-precipitated with COX-1 presented increased serine phosphorylation in the presence of SNAP. This effect was inhibited by the H-9. We suggest that NO, through cGMP-dependent protein kinases, produces the phosphorylation of a 104-kDa protein that is associated with inhibition in the activity of the COX-1, decreasing PGI(2) synthesis and thereby decreasing coronary PGI(2)-mediated vasodilatation.     

H1 and H2 histamine actions on lung vessels; their relevance to hypoxic vasoconstriction.

Pulmonary vasomotor actions of histamine and the possible relationship of histamine to hypoxic pulmonary vasconstriction were studied in anaesthetized cats with one lobe of lung perfused at constant flow and in isolated perfused rat and ferret lungs. In the cat histamine caused dilatation, biphasic responses and constriction with increasing doses. Histamine induced dilatation was better demonstrated during hypoxic vasoconstriction and was reduced by an H2 histamine antagonist; constriction with histamine was abolished by an H1 antagonist. Histamine also caused both vasodilatation and vasoconstriction in ferret lungs. A mast cell stabilizing agent had no effect on hypoxic pulmonary vasoconstriction in cats or rats. This response was unaffected in cats but greatly reduced in rats and ferrets by cyproheptadine, a combined histamine and 5-hydroxy-tryptamine inhibitor. It was unaffected in cats but abolished in ferrets an H1 histamine inhibitor. It was again unaffected in cats but greatly reduced in rats and ferrets by an H2 histamine inhibitor. These species differences may reflect differences in mechanism but more probably reflect non-specific effects of the inhibitors in certain circumstances. However, when drugs nearly abolished hypoxic vasoconstriction, ATP still caused vasoconstriction.     

Effect of noradrenaline on reactive myocardial hyperemia in dogs

The effect of norepinephrine (NE) on reactive myocardial hyperemia (RH) after a brief (20 sec) flow arrest was studied in 9 closed-chest anesthetized dogs. NE infusion into the left ventricle resulted in the increase of RH peak during coronary flow and the elevation of the left ventricular pressure and its first derivative. The increased coronary dilatation may be explained by the activation of beta 1-adrenergic myocardial receptors. RH peak decreased later, with the coronary flow, left ventricular pressure and its first derivative returning to the control level. This RH decrease can be explained by the activation of alpha-adrenergic receptors of the coronary vessels and the competition between alpha-receptors for vasoconstriction and metabolic vasodilatation.     

Effects of cyclooxygenase inhibition on canine coronary artery blood flow and thrombosis

This study was designed to determine the effect of inhibitors of cyclooxygenase (COX)-1, COX-2, and the nonselective COX inhibitor naproxen on coronary vasoactivity and thrombogenicity under baseline and lipopolysaccharide (LPS)-induced inflammatory conditions. We hypothesize that endothelial COX-1 is the primary COX isoform in the canine normal coronary artery, which mediates arachidonic acid (AA)-induced vasodilatation. However, COX-2 can be induced and overexpressed by inflammatory mediators and becomes the major local COX isoform responsible for the production of antithrombotic prostaglandins during systemic inflammation. The interventions included the selective COX-1 inhibitor SC-560 (0.3 mg/kg iv), the selective COX-2 inhibitor nimesulide (5 mg/kg iv), or the nonselective COX inhibitor naproxen (3 mg/kg iv). The selective prostacyclin (IP) receptor antagonist RO-3244794 (RO) was used as an investigational tool to delineate the role of prostacyclin (PGI(2)) in modulating vascular reactivity. AA-induced vasodilatation of the left circumflex coronary artery was suppressed to a similar extent by each of the COX inhibitors and RO. The data suggest that AA-induced vasodilatation in the normal coronary artery is mediated by a single COX isoform, the constitutive endothelial COX-1, which is reported to be susceptible to COX-2 inhibitors. The effect of the COX inhibitors on thrombus formation was evaluated in a model of carotid artery thrombosis secondary to electrolytic-induced vessel wall injury. Pretreatment with LPS (0.5 mg/kg iv) induced a systemic inflammatory response and prolonged the time-to-occlusive thrombus formation, which was reduced in the LPS-treated animals by the administration of nimesulide. In contrast, neither SC-560 nor naproxen influenced the time to thrombosis in the animals pretreated with LPS. The data are of significance in view of reported adverse cardiovascular events observed in clinical trials involving the use of selective COX-2 inhibitors, thereby suggesting that the endothelial constitutive COX-1 and the inducible vascular COX-2 serve important functions in maintaining vascular homeostasis.     

Changes in coronary vascular resistance associated with prolonged hypoxia in isolated rat hearts: a possible role of prostaglandins

Hypoxia caused an initial dilatation in the coronary circulation of perfused male rat hearts but within 15 minutes the coronary vessels became strongly constricted. In hearts from very young (30 days) animals only dilatation was seen. Physiological levels of progesterone in the perfusate prevented the constriction whereas estradiol and testosterone had little effect. Blockade of adrenergic alpha receptors or angiotensin receptors did not prevent the constriction. Two structurally different inhibitors of prostaglandin synthesis, indomethacin and aspirin, and three drugs which can interfere with prostaglandin action, chloroquine, procaine, and propranolol blocked the constriction. Thromboxane A2, a product of PG synthesis, had been reported to be a coronary vasoconstrictor but four drugs which inhibit thromboxane A2 synthesis, dipyridamole, benzydamine, N-0164 and imidazole were not able to prevent the hypoxia-induced constriction. This form of hypoxic coronary constriction seems not to be related to α-adrenergic, angiotensin or thromboxane A2 effects. It may depend on some other product of the prostaglandin pathway.     

Disruption of the endoplasmic reticulum and increases in cytoplasmic calcium are early events in cell death induced by the natural triterpenoid Asiatic acid

Triterpenoids are a novel class of compounds being investigated as potential therapeutic agents for the treatment of prostate cancer and other malignancies. Asiatic acid (AA) is a member of the ursane family of triterpenoids and has anticancer activity, but its mechanism of action is not completely understood. To investigate its mechanism of action, PPC-1 prostate cancer cells were treated with AA at increasing concentrations and times. AA induced rapid caspase-dependent and independent cell death that peaked within 8 h of treatment. AA-induced death was associated with early activation of caspases 2, 3, and 8, but not caspase 9. Within 2.5 h of treatment, release of calcium from intracellular stores and dilatation of the endoplasmic reticulum was observed. Thus, disruption of the endoplasmic reticulum and alterations in calcium homeostasis are early events in AA-induced death.     

Reduced glutathione modulates the arachidonic acid induced coronary reactions

Coronary flow was recorded from spontaneously beating isolated perfused hearts of rats and guinea pigs. Arachidonic acid (AA), in single bolus doses, produced a fast short lasting coronary constriction followed by a slow developing but persisting vasodilation. These reactions (biphasic type) were characteristic of the guinea pig heart. In about 50% of the rat hearts the vasoconstrictor action predominated while the biphasic response was obtained in the rest of the experiments. Pretreatment of rats with aspirin prevented the responses to AA in the isolated heart. The administration of reduced glutathione (GSH) (about 1 mM to the rat or 0.5-0.75 mM to the guinea pig hearts) produced a marked development and (or) enhancement of the vasodilator action of AA. Repeated or single large doses of AA produced a change of pattern of responses from biphasic to constrictor type; the addition of GSH restored the vasodilator phase. Since GSH directs the endoperoxide metabolism towards the synthesis of prostaglandin E2 (PGE2), we postulate that the coronary dilatation of resistance vessels produced by AA would be due to a great extent to PGE2.     

Pharmacological profile of a novel carbacyclin derivative with high metabolic stability and oral activity in the rat

A novel carbacyclin derivative (16S)-13,14-dehydro-16,20-dimethyl-3-oxa-18,18,19,19-tetradehydro- 6a- carbaprostaglandin-I2 (3-oxa-analogue) has been synthesized in order to find chemically and metabolically stable prostacyclin-mimetics with a potency equal or even superior to PGI2. The 3-oxa-analogue was found to be stabilized against beta-oxidation, a main metabolic degradation step also for chemically stable PGI2-analogues. The compound is orally available and displays a long duration of 4.5-48 h of antiaggregatory and hypotensive action. The 3-oxa-analogue inhibits ADP-induced platelet aggregation with an IC50 of 3.0 nM. Following intravenous application the 3-oxa-analogue lowers diastolic blood pressure in a dose dependent manner, the ED20 being 0.1-0.2 micrograms/kg after injection and less than or equal to 0.05 micrograms/kg/min after infusion respectively. In vivo platelet aggregation is inhibited after i.v. infusion of the 3-oxa-analogue with an IC50 of 0.037 micrograms/kg/min. As compared to Iloprost, the 3-oxa-analogue is 5-12 fold more potent with respect to in vivo hypotensive and anti-aggregatory effects. The results of the present studies indicate that the 3-oxa-analogue has a pharmacological profile comparable to prostacyclin (PGI2) and Iloprost. Due to the fact that the 3-oxa-analogue is chemically and metabolically stable, long term oral treatment can be achieved in clinical conditions in which PGI2 and Iloprost have already been shown to be therapeutically useful principles.     

Effect of PGI2 on the response of the ovine placenta to norepinephrine

We have examined the effects of PGI2, 50 microgram/kg, on norepinephrine induced placental vasoconstriction in 6 chronically catheterized near-term sheep. Regional blood flows were measured with radioactive microspheres. Control flows were measured. Norepinephrine was than infused at 50 microgram/min throughout the experiment. After 15 min the blood flows were again measured and PGI2 was then added to the infusate at 50 microgram/min. In 15 min regional blood flows were again measured and the PGI2 infusion was stopped. Regional blood flows were measured for the last time 15 min later. The renal and nonplacental uterine vasculatures behaved in a predictable manner. There was constriction with norepinephrine but PGI2 opposed the effects of norepinephrine and decreased the resistance towards the normal levels. The placenta did not behave as did the other organs. Norepinephrine increased placental resistance but PGI2 did not decrease the resistance and severely depress the placental blood flows. PGI2 does not appear to oppose norepinephrine induced placental vasoconstriction.     

Endothelial NO Synthase (eNOS) phosphorylation regulates coronary diameter during ischemia-reperfusion in association with oxidative stress

The link between endothelial nitric oxide synthase (eNOS) activation and vascular diameter during ischemia-reperfusion was investigated in the rat heart. After short (<30 min) and long (>45 min) time of ischemia conferred by coronary artery occlusion of the rats, reperfusion caused dilatation and constriction of arterioles, respectively. Partial oxygen pressure (pO2) measurement of the heart by the electrode confirmed the hyper-perfusion and no-reflow phenomena during reperfusion, as well as myocardial ischemia. The vascular diameter was correlated with phosphorylation of Akt and serine 1177 residue of eNOS, and formation of NO-bound guanylate cyclase (GC) by immuoflorescence study. Western blotting confirmed the phosphorylation of eNOS-Ser1177 depending on ischemia time. The constriction during reperfusion after 45 min of ischemia is supposedly caused by the inhibition of Akt-mediated eNOS-Ser1177 phosphorylation, which was suppressed by a PKC inhibitor chelerythrine, or ROS scavengers N-2-mercaptopropionyl glycine (MPG) and 4,5-Dihydroxy-1, 3-benzenedisulfonic acid disodium salt (Tiron). However, an endothelin receptor antagonist BQ123 alleviated the vasoconstriction by increasing NO availability but not eNOS-Ser1177 phosphorylation. Thus, vascular patency is correlated with eNOS-Ser1177 phosphorylation in association with ROS, and PKC during reperfusion. Endothelin inhibits vasodilatation by reducing NO availability during reperfusion.     

Endothelium-derived hyperpolarizing factor in coronary microcirculation: responses to arachidonic acid

In coronary resistance vessels, endothelium-derived hyperpolarizing factor (EDHF) plays an important role in endothelium-dependent vasodilation. EDHF has been proposed to be formed through cytochrome P-450 monooxygenase metabolism of arachidonic acid (AA). Our hypothesis was that AA-induced coronary microvascular dilation is mediated in part through a cytochrome P-450 pathway. The canine coronary microcirculation was studied in vivo (beating heart preparation) and in vitro (isolated microvessels). Nitric oxide synthase (NOS) (N(omega)-nitro-L-arginine, 100 microM) and cyclooxygenase (indomethacin, 10 microM) or cytochrome P-450 (clotrimazole, 2 microM) inhibition did not alter AA-induced dilation. However, when a Ca(2+)-activated K(+) channel channel or cytochrome P-450 antagonist was used in combination with NOS and cyclooxygenase inhibitors, AA-induced dilation was attenuated. We also show a negative feedback by NO on NOS-cyclooxygenase-resistant AA-induced dilation. We conclude that AA-induced dilation is attenuated by cytochrome P-450 inhibitors, but only when combined with inhibitors of cyclooxygenase and NOS. Therefore, redundant pathways appear to mediate the AA response in the canine coronary microcirculation.     

The platelet rebound phenomenon during PGI2-infusion occurs at the receptor level

In 6 patients treated with continuous prostaglandin I2 (PGI2)-infusion using a portable pump at a rate of 5 ng/kg/min for 7 days drug receptor interaction of [3H]Iloprost, a stable PGI2 analogue, with a particulate platelet membrane fraction was investigated. Saturation binding experiments of the high affinity platelet prostacyclin receptor performed before and at the end of PGI2 infusion revealed a significant increase of dissociation constant (Kd) and increase in maximal number of binding sites (Bmax). These findings suggest that continuous long-term PGI2 infusion results in a functional desensitization of the membrane-bound PGI2 platelet receptor with a decrease in receptor affinity and an increase in number of binding sites as suggested earlier based upon platelet sensitivity behaviour.     

The adrenergic regulation of the cardiovascular system in the South American rattlesnake, Crotalus durissus

The present study investigates adrenergic regulation of the systemic and pulmonary circulations of the anaesthetised South American rattlesnake, Crotalus durissus. Haemodynamic measurements were made following bolus injections of adrenaline and adrenergic antagonists administered through a systemic arterial catheter. Adrenaline caused a marked systemic vasoconstriction that was abolished by phentolamine, indicating this response was mediated through alpha-adrenergic receptors. Injection of phentolamine gave rise to a pronounced vasodilatation (systemic conductance (G(sys)) more than doubled), while injection of propranolol caused a systemic vasoconstriction, pointing to a potent alpha-adrenergic, and a weaker beta-adrenergic tone in the systemic vasculature of Crotalus. Overall, the pulmonary vasculature was far less responsive to adrenergic stimulation than the systemic circulation. Adrenaline caused a small but non-significant pulmonary vasodilatation and there was tendency of reducing this dilatation after either phentolamine or propranolol. Injection of phentolamine increased pulmonary conductance (G(pul)), while injection of propranolol produced a small pulmonary constriction, indicating that alpha-adrenergic and beta-adrenergic receptors contribute to a basal regulation of the pulmonary vasculature. Our results suggest adrenergic regulation of the systemic vasculature, rather than the pulmonary, may be an important factor in the development of intracardiac shunts.     

The role of adrenergic receptors in Ni2+-induced coronary vasoconstriction

The possible involvement of adrenergic receptors in nickel ion (Ni2+)-induced coronary vasoconstriction was studied on isolated perfused rat hearts and on isolated canine coronary artery strips. The experiments on both models showed that (i) alfa-adrenergic blockade by phenoxybenzamine or phentholamine caused only partial depression of Ni2+-induced coronary vasoconstriction: (ii) beta-adrenergic receptor blockade by propranolol totally prevented Ni2+-action, and (iii) Ni2+ (1 microM) caused significant inhibition of coronary vasodilatation induced by isoproterenol. The experimental results indicate that alfa-adrenoceptors play minor role (if any) in the coronary action mechanism of Ni2+ but it may be mediated by beta-adrenergic mechanisms. Nickel was found to alter the reactivity of coronary beta-adrenoceptors suggesting a possible modulatory role of this trace metal in coronary adrenergic mechanisms.     

Identification and blockade of vascular H2 receptors.

Experiments were conducted in anesthetized dogs to determine the nature of receptors mediating vascular actions of histamine. In the perfused gracilis muscle histamine caused vasodilatation that was attenuated in part by mepyramine, an H1-receptor blocker. Metiamide, an H2 blocker, given alone had no effect on dilatation. However, the combination of mepyramine and metiamide resulted in a large attenuation of dilatation. Histamine caused constriction of the perfused saphenous vein that was totally blocked by mepyramine suggesting that venoconstriction by histamine involves only H1 receptors. Histamine infusion caused a fall in arterial pressure and a large reduction in peripheral resistance. Mepyramine attenuated the fall in pressure but not the reduction in resistance. Combined H1- and H2-receptor blockade largely eliminated the effects of histamine infusion further documenting the existence of H1 and H2 receptors. The effects of H1 and H2 antihistamines on a variety of physiological vasodilator responses were examined. Evidence was obtained to indicate that H1- and H2-histamine receptors are involved in the active component of baroreceptor-mediated reflex vasodilatation, poststimulation vasodilatation, sympathetic vasodilatation in the guanethidine-treated dog, and vasodilator responses following compound 48/80. No evidence for the participation of either H1- or H2-histamine receptors in reactive hyperemia or the dilatation accompanying exercise was found. It is concluded that in the dog both endogenously-released and exogenous histamine exert vascular effects by activation of both H1 and H2 receptors.     

Recycling of platelet phosphorylation and cytoskeletal assembly

The shape change and aggregation of washed platelets induced by 10 microM arachidonic acid (AA) can be reversed by 20 ng/ml prostacyclin (PGI2), but these platelets can be reactivated by treatment with 30 microM epinephrine and subsequent addition of 10 microM AA mixture. These events may be modulated by cAMP since 2 mM dibutyryl cAMP also reversed activation without reactivation by epinephrine and AA. We examined protein phosphorylation and formation of cytoskeletal cores resistant to 1% Triton X-100 extraction of these platelets and correlated these processes with aggregation, fibrinogen binding, and changes in ultrastructure. Unactivated platelet cores contained less than 15% of the total actin and no detectable myosin or actin-binding protein. AA-induced cytoskeletal cores, which contained 60-80% of the total actin, myosin, and actin-binding protein as the major components, were disassembled back to unactivated levels by PGI2 and then fully reassembled by epinephrine and AA. Phosphorylation of myosin light chain and a 40,000-dalton protein triggered by AA (two- to fivefold) was reversed to basal levels by PGI2 but was completely restored to peak levels upon addition of the epinephrine and AA mixture. The reversibility of actin-binding protein phosphorylation could not be established clearly because both PGI2 and dibutyryl cAMP caused its phosphorylation independent of activation. With this possible exception, cytoskeletal assembly with associated protein phosphorylation, aggregation, fibrinogen binding, and changes in ultrastructure triggered by activation are readily and concertedly recyclable.     

低氧,山莨菪碱对肺内小动脉平滑肌细胞产生血栓素A2和前列环素的影响

本工作观察了急性重度低氧及山莨菪碱（anisodamine,Am）对分离培养家兔肺内小动脉平滑肌细胞（PASMCS）培养液中TXB2和6-酮-PGF1a含量及它们的比值的影响。结果是：重度低氧明显增加PASMCS培养液中TXB2和6-酮-PGF1a含量及它们的比值。常氧和低氧条件下,终浓度为2.5×10－5×mol／L的山莨菪碱显著降低TXB2的含量,但6-酮-PGF1a的含量无变化。这提示：急性重度低氧可能通过增加PASMCS产生和释放TXA2和PGI2及它们的比值而致肺血管收缩,山莨菪碱可能通过抑制PASMCs产生ThA2和降低ThB2与PGI2的比值逆转低氧所致的肺血管收缩效应。     

Influence of chemotherapeutic agents on prostacyclin synthesis. II. Effects of tetracycline and penicillin G on prostacyclin synthesis by the rat thoracic aorta and myometrial tissues

The influence of the two antibiotics tetracycline hydrochloride (T) and penicillin G sodium (P) on PGI2 synthesis by the male rat thoracic aorta and day-20 pregnant rat myometrium was investigated in vitro using a rat platelet antiaggregatory bioassay method. Pretreatment of the tissues for 30 min at 37 degrees C with T (21-168 microM) or P (28-224 microM) significantly inhibited PGI2 synthesis in absence or presence of exogenous arachidonic acid (AA) (16.6 microM), (P less than 0.01, n = 5-6). Furthermore, pretreatment of rats with the two drugs (T 11 and P 175 mu mole kg-1 for 30 min) significantly antagonised AA (4 n mole kg-1)-induced hypotension in urethane-anaesthetised rats. They also (T 0.5-4 and P 1-6 microM) antagonised AA-induced aggregation in rabbit citrated platelet-rich plasma. T failed to affect ADP-induced aggregation to any significant level whereas P (3-6 microM) reduced ADP-induced aggregation. The drugs seemed to interfere with the action of the PG endoperoxide synthase (or PG cyclooxygenase) enzyme resulting in decreased formation of PGG2 and PGH2. Such an effect may have resulted from the induced formation of toxic [OH-] radicals and/or inhibition of O2 uptake by the tissues under the influence of the drugs. The demonstrated inherent property of these two antibiotics to inhibit the synthesis of the potent vasodilator, platelet antiaggregatory, anticonvulsant and inhibitor of gastric acid secretion--PGI2, may partly contribute towards better understanding of the biochemical mechanisms that underlie some of the previously known but poorly understood actions of these antibiotics. Furthermore, since good evidence exists for the involvement of excessive uterine prostaglandin synthesis in dysmenorrhoea and premature deliveries, it is suggested that the potential benefits of T or P in these two disorders be investigated.     

Prostaglandin production in cultured gastric mucosal cells: role of cAMP on its modulation

The effect of cAMP on prostaglandin production may depend on cell types. To clarify the relationship between PG and cAMP, we examined arachidonate's effects on PG synthesis and intracellular cAMP accumulation in monolayers of rat gastric mucosal cells. These cells produced PGE2, PGI2 and thromboxaneA2 (TXA2) in amounts of 316 +/- 18, 100 +/- 7 and 30 +/- 5 pg per 10(5) cells in 10 min, respectively, in response to 10 microM arachidonic acid (AA). The production of these PG, however, leveled off subsequently. Cells initially exposed to AA responded poorly to a subsequent stimulation by AA. AA simultaneously stimulated intracellular cAMP accumulation; this stimulatory effect on cAMP production was abolished by the pretreatment with indomethacin. Nevertheless, the pretreatments with dibutyryl cAMP (0.1-5 mM) did not alter the amount of subsequent AA-induced PGE2 production. Furthermore, the preincubation with 1mM isobutyl methyl xanthine also failed to affect PGE2 synthesis, while it increased intracellular cAMP accumulation. Our studies suggest AA stimulates intracellular cAMP formation in cultured gastric mucosal cells, linked with conversion of AA to cyclooxygenase metabolites, AA-induced PG production is limited in these cells, and it seems, however, unlikely that intracellular cAMP modulates AA metabolism to PG.