Secretoneurin facilitates endothelium-dependent relaxations in porcine coronary arteries

Secretoneurin enhances the adhesion and transendothelial migration properties of monocytes and is a part of the peptide family encoded by the secretogranin II gene. The expression of the secretogranin II gene is upregulated in senescent endothelium. The present study was designed to examine the effects of secretoneurin on endothelium-dependent responsiveness. Isometric tension was measured in rings (with or without endothelium) of porcine coronary arteries. Secretoneurin did not induce contraction of quiescent or contracted rings. In preparations contracted by U-46619, relaxation was observed with high concentrations of the peptide. This relaxation was endothelium dependent and reduced by the nitric oxide synthase inhibitor N(ω)-nitro-l-arginine methyl ester (l-NAME). It was abolished when the preparations were incubated with l-NAME in combination with the cyclooxygenase inhibitor indomethacin. The relaxation was not affected by the combination of 1-[(2-chlorophenyl)diphenylmethyl]-1H-pyrazole (TRAM-34) and 6,12,19,20,25,26-hexahydro-5,27:13,18:21,24-trietheno-11,7-etheno-7H-dibenzo[b,m][1,5,12,16]tetraazacyclotricosine-5,13-diiumditrifluoroacetate hydrate (UCL 1684), which abrogates endothelium-dependent hyperpolarizations. These results indicate that secretoneurin acutely induces relaxation through the activation of endothelial nitric oxide synthase (eNOS) and cyclooxygenase, with nitric oxide playing the dominant role. Prolonged (24 h) incubation with physiological concentrations of secretoneurin enhanced the relaxations to bradykinin and to the calcium ionophore A-23187, but this difference was not observed in preparations incubated with l-NAME or the calmodulin antagonist calmidazolium. Under these conditions, the relaxation to sodium nitroprusside remained unchanged. Incubation with secretoneurin significantly augmented the expression of eNOS and calmodulin as well as the dimerization of eNOS in cultures of porcine coronary arterial endothelial cells. These observations suggest that secretoneurin not only acutely causes but also, upon prolonged exposure, enhances endothelium-dependent relaxations.     

Retention of endothelium-dependent vasodilatory responses in canine coronary arteries following cryopreservation

In the present study, the cryoprotective effect of dimethyl sulfoxide (Me2SO) and fetal calf serum (FCS) on coronary endothelium and endothelium-dependent relaxation (EDR) responses was studied in isolated canine coronary arteries following cryostorage at -75 degrees C. Compared to the freshly isolated coronary arteries, the EDR responses to acetylcholine, thrombin, and calcium ionophore were not significantly altered following 1 day storage at -75 degrees C in the presence of 1.8 M Me2SO and 20% FCS. Prolonged cold storage to 7 days, however, resulted in a slight, but significant, rightward shift of the concentration-response curves of acetylcholine and thrombin, but not calcium ionophore. The maximum relaxant response after 7-day cryostorage was 80 to 85% of fresh controls. Omission of FCS from the cryostorage incubation medium further accentuated the loss of EDR responses to all three endothelium-dependent vasodilators tested. Scanning electron microscopic examinations of the intimal surface of the Me2SO and FCS cryostored canine coronary arteries confirmed the preservation of intimal endothelial cells following 1 or 7 days of storage at -75 degrees C, while significant patches of loss of endothelial cells were observed in the arteries cryostored only in the presence of Me2SO. No significant inhibitory effect of cryostorage was observed for the direct, endothelium-independent relaxation induced by isoproterenol, regardless of the presence or absence of FCS. These results demonstrate that slow freezing of canine coronary arteries to -75 degrees C in Krebs-Henseleit solution containing Me2SO and FCS provides good preservation of the vascular smooth muscle function and endothelium-dependent vasodilatory responses.     

Guanosine 5'-O-(3-thiotriphosphate) causes endothelium-dependent, pertussis toxin-sensitive relaxations in porcine coronary arteries.

To determine whether direct stimulation of endothelial G-proteins causes relaxations of the underlying vascular smooth muscle, the effects of guanosine 5'-O-(3-thiotriphosphate) (GTP gamma S) and sodium fluoride were studied in porcine coronary arteries and endothelial cells. Isometric tension was measured in coronary rings contracted with prostaglandin F2 alpha. GTP gamma S (in the presence of saponin) and sodium fluoride (in the presence of AlCl3) relaxed rings with, but not those without endothelium. The responses were inhibited by nitro-L-arginine and pertussis toxin. In membrane fractions of coronary endothelial cells, GTP gamma S and sodium fluoride inhibited the ADP-ribosylation of G-proteins catalyzed with [32P]-NAD and pertussis toxin. These data suggest that direct stimulation of G-proteins in endothelial cells by GTP gamma S and sodium fluoride causes a pertussis toxin-sensitive relaxation which may be attributed to the release of nitric oxide.     

Inhibitory effect of valves on endothelium-dependent relaxations to calcium ionophore in canine saphenous vein

The present study was designed to evaluate endothelium-dependent relaxation to the calcium ionophore A-23187 in isolated canine saphenous veins. Isometric force recordings and cGMP measurements using isolated veins with and without valves were performed. During contractions to U-46619 (3 x 10(-7) M), endothelium-dependent relaxations to A-23187 (10(-9)-10(-6) M) were significantly reduced in rings with valves compared with rings without valves. Endothelial removal abolished A-23187-induced relaxation. Relaxations to forskolin (FK; 10(-8)-10(-5) M) and diethylaminodiazen-1-ium-1,2-dionate; DEA-NONOate, 10(-9)-10(-5) M) were identical in rings with and without valves. In rings without valves, a nitric oxide synthase inhibitor, N(G)-nitro-L-arginine methyl ester (L-NAME; 3 x 10(-4) M), and a cyclooxygenase inhibitor, indomethacin (10(-5) M), partially reduced A-23187-induced relaxation. However, in rings with valves, L-NAME had no effect, whereas indomethacin abolished the relaxation to A-23187. A selective soluble guanylate cyclase inhibitor, 1H-[1,2,4]-oxadiazolo [4,3-a]quinoxalin-1-one (ODQ; 3x10(-6) M), had no effect on the relaxation to A-23187 in either group. In contrast, ODQ abolished the A-23187-induced increase in cGMP levels, suggesting that relaxation to nitric oxide released by A-23187 is independent of increases in cGMP. These results demonstrate that endothelium-dependent relaxation to A-23187 is reduced in regions of veins with valves compared with relaxation in the nonvalvular venous wall. Lower production of nitric oxide in endothelial cells of valvular segments appears to be a mechanism responsible for reduced reactivity to A-23187.     

Thrombin induces endothelium-dependent relaxation of pig coronary arteries

At nanomolar concentrations, the proteolytic enzyme thrombin caused a reversible concentration-dependent relaxation of PGF2 alpha-precontracted pig coronary artery ring segments with intact endothelium. After mechanical removal of the endothelium both thrombin- and bradykinin-induced relaxation disappeared. The thrombin-induced relaxation was inhibited by the tightbinding inhibitor hirudin in a concentration-dependent manner.     

Exercise preserves endothelium-dependent relaxation in coronary arteries of hypercholesterolemic male pigs.

We tested the hypothesis that exercise training (Ex) attenuates hypercholesterolemia-induced impairment of endothelium-dependent relaxation (EDR) in male porcine coronary arteries [left anterior descending coronary arteries (LAD)] by increasing nitric oxide (NO) release [due to increased endothelial NO synthase (NOS) expression] and/or increased bioactivity of NO. Adult male pigs were fed a normal-fat (NF) or high-fat (HF) diet for 20-24 wk. Pigs were Ex or remained sedentary (Sed) for 16-20 wk, beginning after 4 wk on diet. Four groups of pigs were used: NF-Sed, NF-Ex, HF-Sed, and HF-Ex. HF enhanced LAD contractions induced by KCl, aggregating platelets (AP), and serotonin (5-HT). AP and 5-HT produced EDR after blockade of cyclooxygenase with indomethacin (Indo) and smooth-muscle 5-HT(2) receptors with ketanserin. HF impaired EDR induced by AP, 5-HT, and bradykinin. Results indicate a decreased contribution of NO to EDR in HF-Sed LADs, because the percentage of bradykinin-induced EDR inhibited by N(G)-nitro-L-arginine methyl ester was 27% in NF-Sed and 34% in NF-Ex but only 17% in HF-Sed. Also, N(G)-nitro-L-arginine methyl ester + Indo results indicate that release of an Indo-sensitive vasoconstrictor contributes to blunted EDR in HF-Sed LAD. Immunoblot and immunohistochemistry results indicate the following: 1) LAD endothelial NOS protein content was similar among groups; 2) HF decreased LAD superoxide dismutase (SOD) but increased caveolin-1 content; and 3) Ex increased SOD content of HF LADs. We conclude that HF impairs EDR by impairing the contribution of NO released from NOS (due to decreased SOD and increased caveolin-1 protein content) and by production of an Indo-sensitive vasoconstrictor. Ex preserves EDR in HF LADs by decreasing the production of the constrictor and increasing NO-release by NOS and/or NO bioactivity and bioavailability.     

Passive smoking impairs endothelium-dependent relaxation of isolated rabbit arteries

The aim of the present study was to examine the effect of prolonged passive smoking (lasting 3 weeks) on plasma catecholamine levels and reactivity of isolated rabbit arteries. Plasma noradrenaline, adrenaline and dopamine levels were determined radioenzymatically. Isolated rings of the thoracic aorta and carotid artery were suspended in organ chambers and connected to a force transducer for the recording of isometric tension. Plasma noradrenaline levels were found to be significantly elevated in rabbits subjected to passive smoking for 3 weeks. Plasma adrenaline and dopamine levels were not changed. Transmural nerve stimulation of arterial rings evoked frequency-dependent contractions. Prolonged passive smoking did not affect neurogenic contractions of the arteries tested. On the other hand, endothelium-dependent relaxations of phenylephrine-precontracted arteries were significantly impaired. Furthermore, hypertrophy of the left ventricle was observed. In conclusion, passive smoking impairs endothelium-dependent relaxations but not neurogenic contractions of systemic arteries. The impaired relaxations of arteries may be, at least in part, mediated through the degradation of released nitric oxide by superoxide anions derived from cigarette smoke.     

Short-term training enhances endothelium-dependent dilation of coronary arteries, not arterioles.

Our objective was to test the hypothesis that short-term exercise training (STR) of pigs increases endothelium-dependent dilation (EDD) of coronary arteries but not coronary arterioles. Female Yucatan miniature swine ran on a treadmill for 1 h, at 3.5 mph, twice daily for 7 days (STR; n = 28). Skeletal muscle citrate synthase activity was increased in STR compared with sedentary controls (Sed; n = 26). Vasoreactivity was evaluated in isolated segments of conduit arteries (1-2 mm ID, 3-4 mm length) mounted on myographs and in arterioles (50-100 microm ID) isolated and cannulated with micropipettes with intraluminal pressure set at 60 cmH(2)O. EDD was assessed by examining responses to increasing concentrations of bradykinin (BK) (conduit arteries 10(-12)-10(-6) M and arterioles 10(-13)-10(-6) M). There were no differences in maximal EDD or BK sensitivity of coronary arterioles from Sed and STR hearts. In contrast, sensitivity of conduit arteries (precontracted with PGF(2alpha)) to BK was increased significantly (P < 0.05) in STR (EC(50), 2.33 +/- 0.62 nM, n = 12) compared with Sed animals (EC(50), 3.88 +/- 0.62 nM, n = 13). Immunoblot analysis revealed that coronary arteries from STR and Sed animals had similar levels of endothelial nitric oxide synthase (eNOS). In contrast, eNOS protein was increased in STR aortic endothelial cells. Neither protein nor mRNA levels of eNOS were different in coronary arterioles from STR compared with Sed animals. STR did not alter expression of superoxide dismutase (SOD-1) protein in any artery examined. We conclude that pigs exhibit increases in EDD of conduit arteries, but not in coronary arterioles, at the onset of exercise training. These adaptations in pigs do not appear to be mediated by alterations in eNOS or SOD-1 expression.     

Kinins mediate the antiproliferative effect of ramipril in rat carotid artery.

Angiotensin-converting enzyme (ACE) inhibitors have been shown to inhibit neointimal proliferation in response to endothelial injury in the rat carotid artery. Since ACE inhibitors block degradation of kinins, our objective in this study was to determine whether kinins mediate the antiproliferative effect of the ACE inhibitor ramipril. Endothelial denudation was achieved in the left carotid artery of male Sprague-Dawley rats using a balloon catheter. The rats were divided into four groups: a) vehicle (saline); b) DuP 753 10 mg/kg/day; c) ramipril 5 mg/kg/day; and d) ramipril 5 mg/kg/day plus Hoe 140 70 micrograms/kg/day. Ramipril markedly reduced neointimal proliferation compared to control (vehicle) (p less than 0.05) and DuP 753-treated groups (p less than 0.05). When ramipril was given together with Hoe 140 its effect was significantly blunted (p less than 0.05). These results show that kinins are important mediators in the antiproliferative effect of ACE inhibitors.     

Activated neutrophils inhibit cerebrovascular endothelium-dependent relaxations in vitro

The effect of formyl-Met-Leu-Phe- (fMLP-) activated feline neutrophil granulocytes on endothelium-dependent and independent relaxations was studied in the middle cerebral artery of the cat in vitro. Endothelium-dependent relaxations caused by acetylcholine and ATP were markedly inhibited after 30 minutes of incubation of the vessels with neutrophils (5000 cells/microliter) in the presence of 5 microM fMLP, followed by a replacement of the bath solution in order to remove the neutrophils from the medium. Direct vasorelaxations in response to the nitric oxide donor compound SIN-1, however, remained unchanged. Both neutrophils and fMLP caused transient contractions during the incubation period. The present study provides direct evidence for the ability of activated neutrophils to cause an inhibition of vascular endothelium-dependent responses in vitro.     

Role of ceramide in TNF-alpha-induced impairment of endothelium-dependent vasorelaxation in coronary arteries

The present study tested the hypothesis that ceramide, a sphingomylinase metabolite, serves as an second messenger for tumor necrosis factor-alpha (TNF-alpha) to stimulate superoxide production, thereby decreasing endothelium-dependent vasorelaxation in coronary arteries. In isolated bovine small coronary arteries, TNF-alpha (1 ng/ml) markedly attenuated vasodilator responses to bradykinin and A-23187. In the presence of N(G)-nitro-L-arginine methyl ester, TNF-alpha produced no further inhibition on the vasorelaxation induced by these vasodilators. With the use of 4,5-diaminofluorescein diacetate fluorescence imaging analysis, bradykinin was found to increase nitric oxide (NO) concentrations in the endothelium of isolated bovine small coronary arteries, which was inhibited by TNF-alpha. Pretreatment of the arteries with desipramine (10 microM), an inhibitor of acidic sphingomyelinase, tiron (1 mM), a superoxide scavenger, and polyethylene glycol-superoxide dismutase (100 U/ml) largely restored the inhibitory effect of TNF-alpha on bradykinin- and A-23187-induced vasorelaxation. In addition, TNF-alpha activated acidic sphingomyelinase and increased ceramide levels in coronary endothelial cells. We conclude that TNF-alpha inhibits NO-mediated endothelium-dependent vasorelaxation in small coronary arteries via sphingomyelinase activation and consequent superoxide production in endothelial cells.     

Exercise training and responsiveness of isolated coronary arteries.

Exercise is associated with release of catecholamines and vasoactive intestinal polypeptides. Recurrent exposure to catecholamines modifies the sensitivity of adrenoceptors. To test the hypothesis that exercise training may affect the sensitivity of the epicardial coronary arteries, we performed studies on isolated coronary arteries from male dogs capable of running on a treadmill. The animals were separated randomly into two groups: sedentary and exercise training. After 11 wk, rings of left circumflex and left anterior descending coronary arteries were studied in vitro. Contractions to alpha 1-adrenergic agonists (norepinephrine and phenylephrine) were not affected by exercise training. During contractions with prostaglandin F2 alpha, endothelium-dependent relaxations to alpha 2-adrenergic agonists (norepinephrine and UK 14304) were not reduced significantly by exercise training. The concentration-relaxation curves to beta-adrenergic agonists (norepinephrine, isoproterenol, and epinephrine) were shifted to the right after training. The concentration-response curves to vasoactive intestinal polypeptide, but not that to substance P, were shifted to the right in rings with endothelium from exercise-trained animals. These findings demonstrate a decrease in responsiveness of canine vascular smooth muscle to beta-adrenergic agonists and to vasoactive intestinal polypeptide after exercise training.     

Time-dependent potentiation of contractile response to norepinephrine in canine isolated cerebral arteries.

The time course of contractile responses to alpha-adrenoceptor agonists was investigated using various arteries isolated from dogs and monkeys. The contractile response to norepinephrine was increased during the time course of the experiment in canine basilar and internal carotid arteries, whereas the response of isolated canine external carotid arteries and monkey internal carotid arteries did not change significantly. Treatment with 10(-7) M propranolol, 5 x 10(-6) M cocaine plus 10(-5) M hydrocortisone, or 5 x 10(-5) M acetylsalicylic acid did not significantly affect the time-dependent potentiation of the norepinephrine-induced contraction in canine internal carotid arteries. The time-dependent enhancement in the response to norepinephrine was also observed in the arterial preparations from which the endothelial cells were removed. The contractile response of canine internal carotid arteries to phenylephrine did not alter significantly throughout the experiments. On the other hand, the responses to clonidine and xylazine were markedly enhanced with time. Significant potentiation of the norepinephrine-induced contraction was observed in canine internal carotid arteries treated with 10(-8) M prazosin, whereas 10(-8) M yohimbine attenuated the time-dependent potentiation. These results suggest that the contractile responses of isolated canine basilar and internal carotid arteries to norepinephrine are potentiated during the course of the experiment, which is likely to be related, in part, to an enhancement in alpha 2-adrenoceptor mediated contraction.     

Alterations in canine coronary arteries produced by chronic cigarette smoking

Segments of left circumflex coronary artery were obtained from beagles that had been exposed to cigarette smoke for 2 years and from nonsmoking controls. A ring sample from each was used to determine passive (0-Ca2+ and 2 mM EGTA) and active (75 mM K+) force-length relations. Other contiguous samples were used to determine collagen and elastin content, water content and distribution, and total electrolyte content. No significant differences were found in passive mechanical properties or the collagen and elastin content of arteries from the two groups of animals. Maximum values of active force development were found to be increased in arteries from smoking dogs as were values on the ascending part of the active force-length curve. Total water content and the distribution space of an extracellular marker (60Co-EDTA) were not significantly different among arteries from the two groups. This along with the similarity of total electrolyte contents suggests that values of cell volume were not different between the two groups and that the above differences in active force development represent alterations in smooth muscle cells at the subcellular level. These results suggest that the contribution of chronic smoking as one of a group of cardiac risk factors may be due to an effect of augmenting coronary artery smooth muscle contractility.     

Exercise attenuates the effects of hypercholesterolemia on endothelium-dependent relaxation in coronary arteries from adult female pigs.

We tested the hypothesis that exercise training (Ex) attenuates the effects of hypercholesterolemia on endothelium-dependent relaxation in left anterior descending coronary arteries. Adult female pigs were fed a normal-fat (NF) or high-fat (HF) diet for 20 wk. Four weeks after the diet was initiated, pigs were trained or remained sedentary (Sed) for 16 wk, yielding four groups of pigs: 1) NF-Sed, 2) NF-Ex, 3) HF-Sed, and 4) HF-Ex. Sensitivity (EC(50)) to bradykinin (BK) was impaired in HF-Sed arteries. Ex improved BK-induced relaxation such that the EC(50) and maximal response to BK in HF-Ex arteries was not different from that in NF-Sed and NF-Ex. ACh-induced constriction was less in HF-Ex arteries than in HF-Sed, NF-Sed, and NF-Ex. To determine the mechanism(s) by which HF and Ex affected responses to BK and ACh, vasoactive responses were assessed in the presence of N(G)-nitro-L-arginine methyl ester [L-NAME; to inhibit nitric oxide (NO) synthase], indomethacin (Indo; to inhibit cyclooxygenase), and L-NAME + Indo. L-NAME inhibited BK-induced relaxation in NF (not HF) arteries. Indo did not significantly alter relaxation to BK in NF arteries; however, relaxation was enhanced in HF-Sed arteries. Double blockade with L-NAME + Indo attenuated BK-induced relaxation in NF arteries and eliminated relaxation in HF arteries. Neither L-NAME nor Indo altered constrictor responses to ACh in NF or HF arteries; however, double blockade with L-NAME + Indo attenuated constriction to ACh in NF-Ex arteries. Endothelium-independent relaxation to sodium nitroprusside was enhanced in HF-Sed and HF-Ex arteries. Collectively, these results indicate that HF impaired endothelial function in coronary arteries by impairing production of NO and by enhancing production of a constrictor that was inhibited by Indo. Ex attenuated the effects of hypercholesterolemia by improving NO-mediated, endothelium-dependent relaxation and by reducing the influence of the Indo-sensitive constrictor.     

Kinins mediate the inhibition of atrial natriuretic peptide diuretic effect induced by pepsanurin.

Pepsanurin is a peptidic fraction resulting from pepsin digestion of plasma globulins, that inhibits ANP renal excretory actions. We studied whether kinin-like peptides mediate the anti-ANP effect by testing if pepsanurin: 1) was blocked by the kinin B2 receptor antagonist HOE-140, 2) was produced from kininogen, and 3) was mimicked by bradykinin. Anti-ANP activity was assessed in anesthetized female rats by comparing the excretory response to two ANP boluses (0.5 microgram i.v.) given before and after i.p. injection of test samples. Pepsanurin from human or rat plasma (1-5 mL/kg), and bradykinin (5-20 micrograms/kg), dose-relatedly inhibited ANP-induced water, sodium, potassium and cyclic GMP urinary excretion, without affecting arterial blood pressure. The same effect was exerted by pepsin hydrolysates of purified kininogen, whereas hydrolysates of kininogen-free plasma had no effect. HOE-140 (5 micrograms, i.v.) did not alter baseline, or ANP-induced excretion, but blocked the anti-ANP effects of pepsanurin. Histamine (15 micrograms/kg) plus seroalbumin hydrolysates did not affect ANP response, despite inducing larger peritoneal fluid accumulation as compared with pepsanurin or bradykinin. We concluded that kinins cleaved from kininogen mediate the anti-ANP effects of pepsanurin by activation of kinin B2 receptors, independently of changes in systemic arterial pressure or peritoneal fluid sequestration.     

Substance P distribution and effects in the canine epicardial coronary arteries

Substance P (SP), a vasoactive neuropeptide detected in animal and human hearts has been reported to increase coronary blood flow in animals. However, no data are available on SP effects on epicardial coronary arteries, the site of coronary disease. To determine the amount and distribution of SP and its action in the large coronary vessels, we studied two groups of dogs. One group was anesthetized for collecting three 1 cm segments of the circumflex coronary artery (CX) and left anterior descending artery (LAD) through a left thoracotomy. These segments represented proximal (I), middle (II), and distal (III) portions of the two arteries. Concentrations (ng/g) of SP-like immunoreactivity (SP-LI) were determined by radioimmunoassay. SP-LI was present in LAD (I: 1.17 +/- 0.20, II: 1.08 +/- 0.36, III: 1.14 +/- 0.25) and CX (I: 1.44 +/- 0.38, II: 1.51 +/- 0.47, III: 0.70 +/- 0.20). SP differences among segments of LAD and segments I and II of CX were not significant, but there was a significant difference between segment III of CX and the others. In the second group of closed chest anesthetized dogs, we examined the effects of intracoronary SP infusion before and during administration of serotonin (5HT). LAD and CX artery responses (% area change) to SP and to SP plus 5HT were examined using quantitative coronary angiography. Intracoronary 133Xe in saline provided coronary flow data. SP infusion produced significant vasodilation in segment II (15% area increase) and III (17%) during the highest dose (1 microgram/min). The three SP doses infused with 5HT (0.05 mg/min) did not produce vasodilation, although LAD segment III constriction from 5HT was abolished during the highest dose of SP infusion. The presence of SP, and its dilatory effect on the coronary arteries, suggests a role in maintaining vasodilator tone in the coronary arteries.     

Arachidonic acid evokes epithelium-dependent relaxations in canine airways

Responses to arachidonate were examined in rings with and without epithelium of lobar, segmental, and subsegmental canine bronchi. Arachidonate evoked epithelium-dependent relaxations, which were less pronounced in subsegmental bronchi and abolished by indomethacin and meclofenamate. Nordihydroguairetic acid (NDGA) and nafazatrom reduced epithelium-dependent relaxations only in lobar but unmasked epithelium-independent relaxations to arachidonate in all bronchi. Prostaglandin E2 and prostacyclin relaxed all tissues similarly. In lobar bronchi without epithelium, basal release of prostaglandin E2 was reduced by indomethacin but unaffected by NDGA. Arachidonate augmented prostaglandin E2 release more in subsegmental than in lobar bronchi with epithelium; in bronchi without epithelium the rise was absent (lobar) or attenuated (subsegmental). Arachidonate augmented the release of 6-ketoprostaglandin F1 alpha more in lobar bronchi with than without epithelium; this was inhibited by indomethacin, but not NDGA. Thus arachidonate releases prostaglandin E2 (possibly produced by cyclooxygenase inaccessible to inhibitors and activated by lipoxygenase products) but not prostacyclin from the epithelium. Heterogeneity in response to arachidonate is not due to different sensitivity to, or production of, prostaglandins.