Differential induction of cyclooxygenase-2 in human arterial and venous smooth muscle: role of endogenous prostanoids.

Two isoforms of cyclooxygenase (COX) have been identified: a constitutive isoform (COX-1), found in abundance in platelets and the vascular endothelium, and an "inflammatory" cytokine-inducible isoform (COX-2). Because COX metabolites regulate vascular smooth muscle cell (SMC) function and the interaction between the vessel and circulating components, we have investigated the possibility that COX-2 can be induced in human arterial or venous SMC. Untreated venous or arterial cells contained undetectable levels of COX-1 or COX-2 and released low levels of metabolites. After stimulation with interleukin-1beta, tumor necrosis factor-alpha, interferon-gamma, and bacterial lipopolysaccharide, both venous and arterial SMC expressed COX-2 protein and released increased amounts of prostaglandins. In addition, the induced release of PGE2 was inhibited by the COX-2-selective inhibitor, L-745,337. When cells were treated with the mixture of cytokines, venous SMC expressed greater amounts of COX-2 protein and released more prostaglandins than arterial SMC. Furthermore, when COX-2 activity was blocked by L-745,337, COX-2 expression in arterial SMC, but not in venous SMC, increased. Thus, this article describes, for the first time, that COX-2 is expressed in greater amounts in venous SMC than in arterial SMC. Moreover, we show that this "differential induction" is due to a negative-feedback pathway for COX-2 expression in arterial SMC but not in venous SMC. The ability of COX-2 activity to limit COX-2 expression in some cells but not others may contribute to the highly developed mechanisms involved in prostanoid release.

Induction of cyclooxygenase-2 in human saphenous vein and internal mammary artery.

Within vessels, cyclooxygenase (COX) is expressed constitutively (COX-1) in endothelial cells where its production of prostacyclin is thought to contribute to the maintenance of vascular integrity. Recently, a novel isoform of COX, COX-2, has been described that is induced in animal arterial vessels after physical damage or exposure to proinflammatory cytokines. However, induction of COX-2 in human vessels has not been characterized. Moreover, the relative ability of arteries and veins to express COX-2 has not been addressed. Thus, we have compared the ability of segments of human saphenous vein and internal mammary artery, obtained from the same patient, to express COX-2 activity and mRNA after organ culture in the presence and absence of interleukin-1 beta. COX-2 metabolites, measured by radioimmunoassay, were released by both the internal mammary artery and saphenous vein in the following rank order: prostaglandin E2 > or = prostacyclin thromboxane A2. Inclusion of interleukin-1 beta in the culture medium increased the release of prostanoids by the saphenous vein but not by the internal mammary artery. However, the selective COX-2 inhibitor NS-398 significantly attenuated prostacyclin release from both tissues. Northern blot analysis showed no detectable COX-2 mRNA in freshly prepared saphenous vein or internal mammary artery. In contrast, after 48 hours in organ culture, low levels of COX-2 mRNA were detected in both internal mammary artery and saphenous vein, an effect that was greatly increased by interleukin-1 beta. These observations show that COX-2 is induced in human saphenous vein and internal mammary artery and suggest that this may occur in humans after coronary artery bypass graft surgery. The induction of COX-2 and subsequent release of prostacyclin may represent an endogenous defense mechanism against endothelial damage incurred during surgical preparation of these vessels for bypass.

Role of nitric oxide in the dilator actions of capsaicin-sensitive nerves in the rabbit coronary circulation.

Perivascular sensory nerves release calcitonin gene-related peptide (CGRP) and substance P, the dilator actions of which can be regulated by nitric oxide (NO). This study investigated the role of NO in the vasodilation caused by sensory nerve stimulation, by capsaicin, or exogenous CGRP and substance P in the isolated perfused coronary circulation of the rabbit. Coronary perfusion pressure (CPP) was raised in order to observe vasodilator responses, using the thromboxane mimetic, U46619. Capsaicin (3 x 10(-6) moles), alpha CGRP (3 x 10(-11) moles) and substance P (3 x 10(-12) moles) caused comparable reductions in CCP. At these concentrations, responses to capsaicin and CGRP were inhibited by the antagonist CGRP(8-37) but unaffected by the neurokinin-1 receptor antagonist, CP 96,345. The nitric oxide synthase inhibitor, NG nitro L-arginine methyl ester inhibited the effects of substance P and capsaicin but not CGRP. These results suggest that CGRP release following capsaicin-induced sensory nerve activation is modulated by NO.

Characterization of the induction of nitric oxide synthase and cyclo-oxygenase in rat aorta in organ culture.

1. Within vessels, the formation of nitric oxide (NO) or prostaglandins is normally catalysed in the endothelium by constitutive isoforms of NO synthase (eNOS) and cyclo-oxygenase (COX-1), respectively. However, during inflammatory conditions, the underlying smooth muscle acquires the ability to release NO and prostaglandins after the expression of inducible isoforms of NOS (iNOS) and COX (COX-2). The co-induction of iNOS and COX-2 has been studied over 24 h in isolated vascular smooth muscle cells in vitro. However, due to the limitation of using cultured cells, the relationship between the activities of iNOS and COX over longer periods has not been addressed. Moreover, the relative contribution of the endothelium to the production of NO and prostaglandins under inflammatory conditions is not completely understood. 2. Here using an organ culture system, we have determined the profile of COX (6-keto prostaglandin F1 alpha (6-keto PGF1 alpha), PGE2, thromboxane B2 (TXB2) and NOS (nitrite and nitrate) metabolites released over a period of 10 days from segments of rat aorta. In each case, segments from the same animal were left untreated or treated with bacterial lipopolysaccharide (LPS; 10 micrograms ml-1) in order to induce iNOS and COX-2. Prostaglandins were measured by radioimmunoassay whilst nitrite and nitrate were measured, respectively, by Greiss reaction alone, or following a nitrate reductase step. The isoforms of NOS and COX responsible for metabolite release were characterized pharmacologically by use of inhibitors and at the molecular level by reverse transcription polymerase chain reaction with specific primers for iNOS, eNOS, COX-1 and COX-2. In separate experiments the role of the endothelium in the release of nitrite, nitrate and prostaglandins and in the expression of iNOS, eNOS, COX-1 and COX-2 was determined by comparing responses in endothelium denuded and endothelium-intact segments of rat aorta. 3. Under control culture conditions vessels released prostaglandins in the following rank order 6-keto PGF1 alpha = PGE2 > > TXB2. LPS increased the release of 6-keto PGF1 alpha and PGE2 but not of TXB2, an effect that was inhibited by the protein synthesis inhibitor cycloheximide (1 microM), the anti-inflammatory steroid dexamethason (1 microM), the nonsteroidal anti-inflammatory drug indomethacin (30 microM) and, where tested, the selective COX-2 inhibitor NS-398 (30 microM). Similarly, segments of rat aorta released detectable levels of nitrite and nitrate, which were reduced by NG-nitro-L-arginine methyl ester (L-NAME, 1 mM), which inhibits all isoforms of NOS, and by dexamethasone (1 microM), which inhibits the induction of iNOS. The proportion of nitrate to nitrite released over the 10 day period varied greatly from approximately 1:1 on days 5 to 8 to 5:1 on day 9. However, the sum of nitrite and nitrate (NOx) as well as PGE2 remained elevated over the whole 10 day period. The formation of 6-keto PGF1 alpha peaked on days 1 and 2. 4. In freshly prepared tissue, mRNAs for eNOS, COX-1, iNOS and COX-2 were detected. After 24 h in culture, there was an apparent increase in the level of mRNAs for iNOS and COX-2 but not for eNOS or COX-1, an effect that was further enhanced when LPS was included in the culture medium. The expressions of mRNA for eNOS, COX-1, iNOS or COX-2 were not greatly different in vessels with intact or disrupted endothelium. Similarly the release of NOx or PGE2 by vessels after the 1st or 9th day in culture were not significantly different from vessels prepared with or without endothelium. 5. Thus, COX-2 and iNOS are co-induced in intact vessels in culture, with the vascular smooth muscle being the main site of mediator generation. In contrast to data from isolated cells in culture (observed usually over 1 day), both COX and NOS activities in cultured blood vessels were elevated for at least 10 days. Also, unlike isolated cells in culture, the COX and NOS pathways were active independently; L-NAME had little effect on the activity of COX and indomethacin had little effect on the activity of NOS.

An anti-platelet-endothelial cell adhesion molecule-1 antibody inhibits leukocyte extravasation from mesenteric microvessels in vivo by blocking the passage through the basement membrane.

Platelet-endothelial cell adhesion molecule-1 (PECAM-1, CD31) plays an active role in the process of leukocyte migration through cultured endothelial cells in vitro and anti-PECAM-1 antibodies (Abs) inhibit accumulation of leukocytes into sites of inflammation in vivo. Despite the latter, it is still not clear at which stage of leukocyte emigration in vivo PECAM-1 is involved. To address this point directly, we studied the effect of an anti-PECAM-1 Ab, recognizing rat PECAM-1, on leukocyte responses within rat mesenteric microvessels using intravital microscopy. In mesenteric preparations activated by interleukin (IL)-1 beta, the anti-PECAM-1 Ab had no significant effect on the rolling or adhesion of leukocytes, but inhibited their migration into the surrounding extravascular tissue in a dose-dependent manner. Although in some vessel segments these leukocytes had come to a halt within the vascular lumen, often the leukocytes appeared to be trapped within the vessel wall. Analysis of these sections by electron microscopy revealed that the leukocytes had passed through endothelial cell junctions but not the basement membrane. In contrast to the effect of the Ab in mesenteric preparations treated with IL-1 beta, leukocyte extravasation induced by topical or intraperitoneal administration of the chemotactic peptide formyl-methionyl-leucyl-phenylalanine was not inhibited by the anti-PECAM-1 Ab. These results directly demonstrate a role for PECAM-1 in leukocyte extravasation in vivo and indicate that this involvement is selective for leukocyte extravasation elicited by certain inflammatory mediators. Further, our findings provide the first in vivo indication that PECAM-1 may have an important role in triggering the passage of leukocytes through the perivascular basement membrane.

Interleukin-1-induced leukocyte extravasation across rat mesenteric microvessels is mediated by platelet-activating factor.

Although our understanding of the molecular interactions that mediate the adhesion of leukocytes to venular endothelial cells has greatly expanded, very little is known about the mechanisms that mediate the passage of leukocytes across the vessel wall in vivo. The aim of the present study was to investigate the role of endogenously formed platelet-activating factor (PAF) in the process of leukocyte extravasation induced by interleukin-1 (IL-1). To determine at which stage of emigration PAF was involved, we studied the behavior of leukocytes within rat mesenteric microvessels by intravital microscopy. Rats were injected intraperitoneally with saline, recombinant rat IL-1 beta (IL-1 beta), or the peptide N-formyl-methionyl-leucyl-phenylalanine (FMLP) 4 hours before the exteriorization of the mesenteric tissue. In animals treated with IL-1 beta there was a significant increase in the number of rolling and adherent leukocytes within venules (20- to 40-micron diameter) and in the number of extravasated leukocytes in the tissue. Pretreatment of rats with the PAF receptor antagonist UK-74,505 had no effect on the leukocyte responses of rolling and adhesion, but significantly inhibited the migration of the leukocytes across the vessel wall induced by IL-1 beta (76% inhibition). A structurally unrelated PAF antagonist, WEB-2170, produced the same effect (64% inhibition). However, in contrast, UK-74,505 had no effect on the leukocyte extravasation induced by FMLP, indicating selectivity for the response elicited by certain mediators. These results provide the first line of direct evidence for the involvement of endogenously formed PAF in the process of leukocyte extravasation induced by IL-1 in vivo.

Prolonged microvascular vasodilation induced by leukotriene B4 in human skin is cyclooxygenase independent.

When the potent chemoattractant leukotriene B4 (LTB4) is applied topically to human skin it causes delayed onset, long-lasting leukocyte accumulation and erythema. We investigated the role of prostaglandins in the increase in local blood flow by applying LTB4 topically to the forearm skin of 22 healthy male volunteers and measuring the effect of the anti-inflammatory compounds tenidap, naproxen and indomethacin. Local microvascular blood flow responses were measured by laser Döppler flow probe and planimetry. LTB4 induced dose-dependent increases in blood flow which were maximal at 48 hr and lasted 4 days. Laser Döppler flow (% flux) at 48 hr was 2.7 +/- 0.1, 20.6 +/- 3.1, 28.7 +/- 2.4 and 30.2 +/- 2.3% in control and 3, 10, 30 ng/site LTB4, respectively (mean +/- S.E.M.). In eight subjects the intradermal injection of indomethacin, at a dose (3 x 10(-9) mol/site) that inhibited significantly the increased flow induced by intradermally injected arachidonic acid (1 x 10(-9) mol/site, n = 6), had no effect on the increased skin blood flow response induced by LTB4 (10 ng/site) at 48 hr. Blood flow in vehicle-injected LTB4 sites was 810 +/- 150% above basal and 819 +/- 149% in sites injected with indomethacin. In 20 subjects, the effect of the anti-inflammatory drugs naproxen and tenidap given orally on the skin blood flow responses to LTB4 were compared in a double-blind crossover design. The 1085 +/- 98% increase in local blood flow induced by 30 ng of LTB4 at 48 hr was unaltered at the end of the treatment periods with either naproxen or tenidap, where blood flow in the LTB4-treated sites was increased 1018 +/- 131% and 1034 +/- 130%, respectively. Because the vasodilator response to exogenous LTB4 was not altered by nonsteroidal anti-inflammatory drugs either injected locally or taken orally, we suggest that endogenous prostaglandins are not involved in this response.

Evidence for sensory nerve involvement in cutaneous reactive hyperemia in humans.

To study the involvement of local sensory nerves in reactive hyperemia, laser-Doppler measurements of skin blood flow were recorded in locally anesthetized and untreated forearm sites in eight volunteers after 90, 180, and 360 seconds of arrested forearm blood flow. The reactive hyperemia increased in magnitude and duration in response to increasing occlusion periods. However, maximum postocclusion flows in the untreated site of 31 +/- 5%, 38 +/- 6%, and 49 +/- 5% (mean +/- SEM) flux were significantly greater than the 14 +/- 3% (P < .005), 20 +/- 4% (P < .005), and 25 +/- 5% (P < .001) flux seen in the anesthetized sites. The duration of the hyperemia was also shortened from 139 +/- 26 seconds in the untreated site to 61 +/- 17 seconds (after the 360-second occlusion, P < .02) in the anesthetized sites. The anesthesia did not alter the increase in local blood flow induced by intradermally injected calcitonin gene-related peptide. Topically applied capsaicin induced a localized increase in blood flow that was unaffected by anesthesia and a surrounding flare that was abolished by the treatment. The results show that local anesthesia can significantly inhibit reactive hyperemia by a mechanism that does not alter the vasodilation induced by exogenous calcitonin gene-related peptide or the localized capsaicin-induced release of vasodilators from sensory nerves. Indomethacin was also found to be effective in suppressing reactive hyperemia. The evidence suggests that postocclusion reactive hyperemia in human forearm skin is mediated by a local reflex involving sensory nerves and a cyclooxygenase product, probably a vasodilator prostaglandin.

Pituitary adenylate cyclase activating polypeptide is a potent vasodilator in humans.

The vasodilator effect of the novel peptide pituitary adenylate cyclase activating polypeptide (PACAP) was investigated in humans. Forearm blood flow was measured in six healthy men by venous occlusion plethysmography. Infusion of PACAP into the brachial artery at 0.01, 0.1, 1, 3, and 10 pmol/min produced a dose-related increase in forearm blood flow in the cannulated arm from 2.8 +/- 0.6 to 8.6 +/- 2.4 ml/100 ml/min at the highest dose (mean +/- SEM, p less than 0.05). In a subsequent experiment, where the highest dose of PACAP was repeated after a 36 min interval, there was no tachyphylaxis of the forearm blood flow response, with the forearm blood flow increasing by 129 +/- 9% during the first infusion and 128 +/- 31% during the second infusion (N.S.). In further experiments, microvascular blood flow was measured by a laser-Doppler flow probe to compare the effects of intradermally injected PACAP, vasoactive intestinal polypeptide (VIP), and calcitonin gene-related peptide (CGRP). When injected into the skin of normal volunteers at 10(-12) to 10(-11) mol/site, each peptide caused a rapid flare lasting 2-3 min, which became erythematous after 5 min. At 10(-12) mol/site, intradermally injected PACAP and VIP caused a maximum increase in skin blood flow at 15 min of 379 +/- 96 and 307 +/- 121% (% increase above basal +/- SEM), respectively, and these responses were not significantly affected by oral aspirin (600 mg) taken 1.5 h beforehand. The vasodilation induced by PACAP at 10(-12) mol/site lasted approximately 6 h, whereas the effect of the same dose of CGRP and VIP lasted less than 2 h. These data suggest that PACAP is a potent and long-lasting vasodilator in humans.

Pituitary adenylate cyclase activating polypeptide is a potent vasodilator and oedema potentiator in rabbit skin in vivo.

1. The effects of pituitary adenylate cyclase activating polypeptide (PACAP) on microvascular blood flow and plasma protein leakage were investigated in rabbit skin in vivo. 2. Intradermal injection of PACAP38, the 38 amino acid form of the peptide, caused a dose-dependent increase in blood flow measured by a 133Xe clearance technique. An equivalent increase in blood flow was induced by 10(-12) mol per site of PACAP38, 10(-12) mol per site of human alpha-calcitonin gene-related peptide (CGRP) and 10(-10) mol per site of vasoactive intestinal polypeptide (VIP). 3. The vasodilator activity of PACAP38 was not significantly different from that of the 27 amino acid form of the peptide, PACAP27, when measured with a laser Doppler flow meter, causing a 104 +/- 14% compared with 110 +/- 18% increase above basal blood flow at 10(-12) mol per site respectively. 4. At 10(-12) mol per site the effect of PACAP38 was longer lasting than that of CGRP. Blood flow remained significantly increased above control at 2 h with PACAP38 (P less than 0.05) whereas blood flow after intradermal CGRP had returned to control values by this time. 5. PACAP38 injected alone had no significant effect on microvascular leakage of 125I-labelled albumin. However, PACAP38 significantly potentiated bradykinin-induced oedema where it was approximately 100 fold more potent than VIP. 6. Oedema potentiation induced by PACAP38 was not inhibited by indomethacin at a dose which did inhibit potentiation of bradykinin-induced oedema by arachidonic acid.7. PACAP38 is at least as potent as other peptides which have been postulated to be involved in the inflammatory response when tested in rabbit skin in vivo. PACAP may contribute to both the hyperaemia and oedema components of inflammation.

Responses of atherosclerotic human coronary arteries in vivo to the endothelium-dependent vasodilator substance P.

BACKGROUND: The effects of the endothelium-dependent vasodilator substance P (SP) on atherosclerotic human coronary arteries was studied. METHODS AND RESULTS: [125I]-SP binding to luminal cells was shown to be preserved in the atherosclerotic epicardial coronary arteries of four patients. No binding to medial smooth muscle cells was demonstrated. Intracoronary infusions of SP were undertaken in patients with coronary artery disease. SP was infused for 2-minute periods starting at a dose of 2.8 pmol/min rising by doubling increments to 22.4 pmol/min. Analysis of the epicardial coronary artery diameter, using a computerized analysis system (CAAS) of the angiograms, was performed at the end of each infusion. Analysis of seven smooth vessel segments from seven coronary vessels, which were stenosed at more proximal sites, was performed. Significant dose-dependent dilatation was seen (p = 0.04), which was maximal at 5.6 pmol/min SP. No additional dilatation was produced with 2 mg intracoronary isosorbide dinitrate (ISDN). Two of these seven patients showed no response to SP, and only one of these appeared to sustain dilatation with ISDN (2 mg intracoronary). In a second group of six patients with discrete coronary stenoses, analysis at the site of the stenosed segments appeared to reveal dilatation in response to SP in only one instance. One other stenotic segment dilated with isosorbide dinitrate but failed to dilate with SP; the remaining four were fixed. The segment immediately proximal to the stenosis preserved a dose-dependent vasodilator response. CONCLUSIONS: These findings demonstrate that the endothelium-dependent vasodilator substance P can still produce epicardial vasodilatation in vivo in the presence of coronary atherosclerosis.

Endothelin is a potent long-lasting vasoconstrictor in men.

Endothelin, a 21-amino acid peptide synthesized by cultured porcine aortic endothelial cells, has recently been identified and shown to produce a potent and prolonged constriction of mammalian blood vessels in vitro. We have studied the effect of local infusion of this peptide on resistance and capacitance vessels of normal volunteers. Infusion of endothelin (5 pmol/min) reduced forearm blood flow by 39 +/- 7% from control observations. The maximum response was seen after approximately 55 min of infusion. After stopping the infusion, return of flow to basal values took approximately 120 min. This contrasts with the short onset and duration of action observed when angiotensin II was infused. During coinfusion studies, reversal by nicardipine (0.3-10 micrograms/min) occurred at similar concentrations in both endothelin-induced and angiotensin-induced flow reduction. This observation suggests that nicardipine nonspecifically antagonizes the flow-reducing effects of endothelin. A pattern of slow onset of constriction was found on local infusion of endothelin (5 pmol/min) into dorsal hand veins. During coinfusion of nicardipine (1.5 microgram/min), no reversal of endothelin-induced (5 pmol/min) constriction of dorsal hand veins occurred. The pharmacological profile of this peptide in the peripheral circulation of humans suggests that it may be involved in long-term regulation of vascular tone.

Substance P dilates epicardial coronary arteries and increases coronary blood flow in humans.

The effects of intracoronary infusions of substance P in conscious humans on epicardial vessel diameter and coronary sinus oxygen saturation were investigated in 13 patients who had angiographically normal coronary arteries. The dose of substance P ranged from 2.8 to 89.6 pmol/min, starting at 2.8 pmol/min and rising by doubling increments. All infusions were performed for 2-minute periods. Epicardial vessel diameter was measured by a computerized analysis system (CAAS) of the angiogram performed at the end of each infusion period. Coronary sinus oxygen saturation (CSO2) was measured continuously by a fiber-optic oximeter catheter in the coronary sinus. A reduction of arterial pressure, indicative of a systemic effect of substance P, occurred only when infusions were 44.8 pmol/min or more. At considerably smaller dosages, a dose-dependent increase in the left anterior descending artery diameter was seen with substance P; maximal dilation occurred at a dosage of 11.2 pmol/min. The percent increase in vessel diameter for this dosage was 29.4 +/- 19.8% (mean +/- SD) at the distal site of analysis and 16.7 +/- 10% at the proximal site of analysis. CSO2 rose in a dose-dependent manner and was maximal at a dosage of 11.2 pmol/min, which caused a 14.6 +/- 7.2% O2 rise above the preinfusion saturation of 43.5 +/- 11.1% O2. As the heart rate-blood pressure double product showed no change, it is argued that these changes in CSO2 probably reflect rises in coronary flow that were seen with dosages of substance P without systemic effect. In three patients, preconstriction induced by ergonovine was reversed by substance P infusions, which produced a degree of dilation equivalent to that of isosorbide dinitrate. We conclude that substance P at very small dosages, which are without peripheral effects, causes in vivo epicardial coronary artery dilatation as well as resistive vessel dilatation in humans.

Influence of endothelin on human coronary arteries and localization of its binding sites.

Endothelin, a 21 amino acid peptide synthesized by cultured porcine aortic endothelial cells, has recently been identified and shown to produce a potent and prolonged constriction of mammalian blood vessels in vitro. Using tissue obtained from explanted hearts at the time of cardiac transplantation, the response of isolated human epicardial coronary arteries to endothelin was studied. The presence of endothelin binding sites was demonstrated in these vessels using an autoradiographic technique. Endothelin produced a dose-dependent increase of tension in the isolated coronary vessels with a maximal tension achieved equal to 135% of that induced by 90 mM of potassium. The maximal response was slow to develop and had a prolonged duration of 15 to 20 minutes. Nicardipine (4 microM) failed to affect the contraction induced by low doses of endothelin, but decreased the tension obtained at high doses. However, adenosine, substance P and glyceryl trinitrate were all effective in reversing the contraction induced by endothelin, while indomethacin and acetylcholine were ineffective. These features differ from those of other endogenous constrictor agents and make endothelin a potential candidate for long-term modulation of vascular tone.

Endothelin-1 is a potent long-lasting vasoconstrictor in dog peripheral vasculature in vivo.

Endothelin-1 (ET-1), a 21 amino acid peptide, has recently been identified and shown to produce a potent and prolonged constriction of mammalian blood vessels in vitro. We have studied the effect of local infusion of this peptide on resistance vessels in the hindlimb of the anesthetized greyhound dog. Incremental doses of ET-1 (3-200 pmol/min) were infused into the left femoral artery. Doses above 10 pmol/min produced a slowly progressive reduction in hindlimb blood flow in a dose dependent fashion, maximally reducing flow by 79.5% +/- 3.2 from 152.3 +/- 29.1 ml/min to 27.8 +/- 5.8 ml/min (+/- SEM, p less than 0.015), with a concomitant rise in vascular resistance. In the control vessel (right femoral artery) there were no statistically significant changes in blood flow observed. Onset time of the response to ET-1 was 3 min, whereas spontaneous recovery of the flow occurred at 30 min following cessation of the infusion. We demonstrated transient reversal of constriction in this arterial model during coinfusion with endothelin-1 (100 pmol/min) of dihydropyridine calcium channel blocking agent nicardipine (0.5-20 nmol/min), substance P (0.5-50 fmol/min), adenosine (10-10,000 pmol/min), and isosorbide dinitrate (0.001-0.1 mg/min).