Jerry A. Smith

Endothelium‐derived relaxing factor inhibits in vitro platelet aggregation

1 We studied the effects of endothelium‐derived relaxing factor (EDRF), bovine retractor penis muscle inhibitory factor and sodium nitroprusside, three stimulants of guanylate cyclase, on the in vitro aggregation of washed human platelets. 2 Platelet aggregation induced either by collagen or by the thromboxane A2 analogue U46619 was inhibited by all three agents. 3 The anti‐aggregatory effect of each agent was inhibited by haemoglobin. 4 The anti‐aggregatory effect of EDRF was potentiated by superoxide dismutase. 5 These findings are discussed in relation to a potential role for EDRF in haemostasis.

Factors released from endocardium of the ferret and pig modulate myocardial contraction.

1. In isolated heart muscle preparations, selective removal of the endocardium results in a characteristic and unusual negative inotropic effect. Possible mechanisms for this effect were investigated in this study. 2. In endocardium‐intact preparations of ferret papillary muscle, 8‐bromo‐cyclic GMP, sodium nitroprusside, atrial natriuretic peptide (ANP) and substance P each induced changes in contractile behaviour similar to selective endocardial removal, and each significantly elevated myocardial cyclic GMP levels. Substance P failed to elevate myocardial cyclic GMP levels following removal of endocardium or in the presence of haemoglobin, suggesting that it may act by releasing endothelium‐derived relaxing factor (EDRF) from endocardium. However, there was no change in myocardial cyclic GMP levels following endocardium removal alone. 3. In cascade bioassay experiments, it was confirmed that porcine cultured endocardial cells released an unstable humoral agent whose effects on an endothelium‐denuded pig coronary artery were indistinguishable from EDRF. 4. The negative inotropic effects of endocardium removal were reversed in bioassay experiments where an endocardium‐denuded papillary muscle was exposed to the effluent from a column of porcine cultured endocardial cells on microcarrier beads. This demonstrates for the first time the release of a ‘contraction prolonging factor’ from endocardium, the tonic release of which would explain the negative inotropic effect of endocardium removal. 5. It is concluded that elevation of ferret papillary muscle cyclic GMP (as for example with EDRF) produces changes in contractile performance similar to those induced by endocardium removal. We also demonstrate that superfused porcine cultured endocardial cells release a humoral agent (provisionally named ‘endocardin’) which causes reversal of the changes in mechanical properties seen after endocardial removal.

The mechanisms by which haemoglobin inhibits the relaxation of rabbit aorta induced by nitrovasodilators, nitric oxide, or bovine retractor penis inhibitory factor

1 The mechanisms by which haemoglobin and methaemoglobin inhibit the vasodilator actions of glyceryl trinitrate, sodium azide, nitric oxide, and the bovine retractor penis inhibitory factor (IF) were studied on rabbit endothelium‐denuded aortic rings. 2 Methaemoglobin was less effective than haemoglobin against each vasodilator, it was more effective at inhibiting the relaxation to azide than that to glyceryl trinitrate. 3 Glyceryl trinitrate was neither bound nor inactivated when passed through columns of haemoglobin‐agarose or methaemoglobin‐agarose. Azide was reversibly bound but less by haemoglobin‐agarose than by methaemoglobin‐agarose. Inhibition of the vasodilator actions of glyceryl trinitrate is not attributable therefore to a direct interaction with the haemoproteins, although a small part of the inhibition of azide‐induced relaxation by methaemoglobin is likely to be due to a direct interaction. 4 Columns of haemoglobin‐agarose were more effective than columns of methaemoglobin‐agarose in removing nitric oxide from solution. The greater ability of haemoglobin, compared to methaemoglobin, to inhibit vasodilatation induced by nitrovasodilators may therefore reflect the greater ability of haemoglobin to bind nitric oxide which is the active principle of the nitrovasodilators. 5 Neither the acid‐activated nor the inactive forms of IF were bound or inactivated when passed through columns of methaemoglobin‐agarose. Neither form of IF was retained on passage through colomns of haemoglobin‐agarose, but the resulting activity in the eluates was less than control, was unstable and, unlike the original activity, decayed rapidly on ice. The greater ability of haemoglobin, compared to methaemoglobin, to inhibit vasodilatation induced by IF might therefore reflect the greater ability of haemoglobin to interact with this vasodilator and inactivate it.

Nitric oxide synthase in cultured endocardial cells of the pig

1 Endocardial cells release factors which regulate myocardial contractility and guanosine 3′:5′‐cyclic monophosphate (cyclic GMP) levels. One of these factors is indistinguishable from endothelium‐derived relaxing factor (EDRF). 2 The effluent from pig heart endocardial cells cultured on microcarrier beads caused the relaxation of a pig coronary artery ring denuded of endothelium. This relaxation was enhanced by a combination of superoxide dismutase and catalase and was attenuated by haemoglobin, which binds nitric oxide (NO), and by inhibitors of NO synthase, NG‐monomethyl‐l‐arginine (l‐NMM A) or NG‐nitro‐l‐arginine. 3 A Ca2+‐, l‐arginine‐ and NADPH‐dependent enzyme activity which generated NO was detected by a specific spectrophotometric assay in cytosol prepared from endocardial cells. The formation of NO was inhibited in a concentration‐dependent manner by l‐NMMA (but not d‐NMMA) and this could be partially reversed upon addition of excess l‐arginine. 4 Like endothelial cells from the blood vessels, the endocardial cells possess the ability to synthesize NO, which may act to regulate myocardial contractility.

Porcine ventricular endocardial cells in culture express the inducible form of nitric oxide synthase

1 We have investigated whether porcine endocardial cells in culture express the inducible, Ca2+‐independent form of nitric oxide (NO) synthase. 2 NO synthase activity in cytosolic extracts of endocardial cells was measured by estimation of the rate of formation of l‐[14C]‐citrulline from l‐[14C]‐arginine. 3 Treatment of the cells in culture with lipopolysaccharide or cytokines induced a Ca2+‐independent NO synthase activity in the cell cytosol. The combination of tumour necrosis factor (TNFα, 10 ng ml−1) and interleukin‐1β (IL‐1β, 10 ng ml−1) induced the greatest enzyme activity. 4 The increased Ca2+‐independent NO synthase activity following exposure to cytokines was parallelled by an increase in guanosine 3′:5′‐cyclic monophosphate (cyclic GMP) levels in the endocardial cell cytosol. 5 Simultaneous addition of dexamethasone (0.01–1 μm) or cycloheximide (0.03–3 μm) inhibited in a concentration‐dependent manner TNFα‐ and IL‐1β‐induced expression of Ca2+‐independent NO synthase activity. Neither dexamethasone (1 μm) nor cycloheximide (3 μm) had any effect on the activity of the constitutive NO synthase. 6 The possible pathophysiological consequences of endocardial expression of the inducible NO synthase are discussed.

Evidence that inhibitory factor extracted from bovine retractor penis is nitrite, whose acid-activated derivative is stabilized nitric oxide.

1 Unactivated extracts of bovine retractor penis (BRP) contain 3–7 μm nitrite. Acid‐activation of these extracts at pH 2 for 10 min followed by neutralization generates the active form of inhibitory factor (IF; assayed by its vasodilator action on rabbit aorta), and is associated with partial loss of nitrite. 2 Increasing the time of acid‐activation at pH 2 from 10 to 60 min with intermittent vortex mixing generates greater vasodilator activity and increases nitrite loss. 3 When acid‐activated and neutralized extracts are incubated at 37°C or 30 min or boiled for 5 min, vasodilator activity is lost and nitrite content increased. Reactivation of these samples at pH 2 for 10 min followed by neutralization leads to partial recoveries of vasodilator activity with loss in nitrite content. 4 Addition of sodium nitrite to BRP extracts increases acid‐activatable vasodilator activity pro rata. 5 Acid‐activation of aqueous sodium nitrite solutions results in less loss of nitrite and generation of less vasodilator activity than BRP extracts. Vasodilatation is only transient and is rapidly abolished on neutralization, whereas responses to acid‐activated BRP extracts are more prolonged and activity is stable on ice. 6 Bovine aortic endothelial cells yield vasodilator activity that is indistinguishable from that isolated from BRP. It is activated by acid, stable on ice, abolished by boiling or by haemoglobin, and appears to be due to the generation of nitric oxide (NO) from nitrite. 7 The data provide confirmatory evidence that nitrite in BRP extracts is IF, that acid‐activation of BRP extracts yields NO which is responsible for its vasodilator action, and that inactivation occurs by decay of NO to nitrite and nitrate. They further suggest that BRP extracts contain a NO‐stabilizing agent which favours conversion of nitrite to NO. 8 The finding that bovine aortic endothelial cells yield an agent indistinguishable from IF suggests that nitrite in endothelial cells may likewise be the precursor of endothelium‐derived relaxing factor (EDRF), itself identified as NO.

Release of endothelium‐derived relaxing factor from pig cultured aortic endothelial cells, as assessed by changes in endothelial cell cyclic GMP content, is inhibited by a phorbol ester

1 Cultured aortic endothelial cells of the pig respond to the endothelium‐derived relaxing factor (EDRF) they release with an increase in cyclic GMP content. This response is inhibited by haemoglobin or by l‐NG‐monomethyl‐arginine (l‐NMMA), and has been used to investigate the effects of phorbol esters on EDRF release. 2 Pretreatment with phorbol‐12,13‐dibutyrate (PDB) but not the inactive 4α‐phorbol‐12,13,‐didecanoate (PDD), inhibited increases in cyclic GMP induced by substance P (10−8 m) in a time and concentration‐dependent manner. PDB did not affect basal cyclic GMP levels. 3 PDB (3 × 10−7 m), but not PDD (3 × 10−7 m), also inhibited ATP (10−5 m)‐induced increases in cyclic GMP, but did not affect those induced by bradykinin (10−7 m). 4 Increases in cyclic GMP induced by low (10−7 m) but not high (10−6 m) concentrations of the calcium ionophore A23187 were inhibited by PDB (3 × 10−7 m). This inhibitory effect was due to enhanced destruction of EDRF by superoxide anions rather than inhibition of EDRF release, as the inhibition was abolished in the presence of superoxide dismutase (SOD, 30 u ml−1) and catalase (CAT, 100 u ml−1). 5 SOD and CAT did not affect the inhibitory action of PDB on substance P or ATP‐induced increases in cyclic GMP. 6 Increases in endothelial cell cyclic GMP content induced by sodium nitroprusside (10−5 m) were unaffected by PDB pretreatment. 7 The inhibitory effects of PDB are probably a result of an action of protein kinase C on the steps between receptor occupation and phospholipase C activation.

Induction of a calcium-independent NO synthase by hypercholesterolaemia in the rabbit

Endothelium‐dependent and ‐independent relaxation of aortic ring preparations was assessed and nitric oxide (NO) synthase activity measured in the lung, and cerebellum of cholesterol‐fed and normal rabbits. Endothelium‐dependent relaxation of acetylcholine and ATP was depressed while that to the calcium ionophore, A23187, was unaltered in the cholesterol‐fed group. Relaxation to sodium nitroprusside was however greater in aortae from the cholesterol‐fed animals. Neither Ca2+‐dependent nor Ca2+‐independent NO synthase activity could be detected in aortae or hearts taken from either group of animals. Activity of both enzymes was unaltered in cerebellae from both groups of animals. Activity of the Ca2+‐independent enzyme was however significantly greater (ca. 2 fold) in lungs from the cholesterol‐fed rabbits though the activity of the Ca2+‐dependent NO synthase was not significantly altered. This finding may account for the increased production of nitrogen oxides previously observed in this model of hypercholesterolaemia.

Oxidation of low density lipoprotein by bovine and porcine aortic endothelial cells and porcine endocardial cells in culture

Oxidation of low density lipoprotein (LDL) in atherosclerotic lesions may be involved in converting macrophages into cholesterol-laden foam cells, a major characteristic of atherosclerotic lesions. It has been reported, and is widely believed, that endothelial cells derived from rabbit, pig and human aortas, but not those derived from bovine aortas, are capable of oxidising LDL in vitro. We have re-investigated this subject and found that during a 48-h incubation period bovine aortic endothelial cells (both in primary culture and in subcultures) were capable of consistently modifying LDL, increasing its uptake and degradation by macrophages by more than 4-fold. Incubation of LDL with bovine aortic endothelial cells for only 24 h, however, produced inconsistent modification of the LDL, whereas mouse peritoneal macrophages consistently modified LDL in 24 h. The modification of LDL by bovine aortic endothelial cells was an oxidative process, as the chain-breaking antioxidants, alpha-tocopherol and probucol, completely or greatly inhibited it. Thus, bovine aortic endothelial cells are capable of oxidising LDL but they are slower at doing so than are certain other types of cells. Nitric oxide generated by activated macrophages has very recently been shown to inhibit their oxidation of LDL. We have therefore investigated whether or not the inhibition of the constitutive nitric oxide synthase of bovine or porcine aortic endothelial cells would increase their rate of oxidation of LDL.(ABSTRACT TRUNCATED AT 250 WORDS)

Atriopeptin II-induced relaxation of rabbit aorta is potentiated by M&B 22,948 but not blocked by haemoglobin

1 We examined the effects of haemoglobin (which inhibits the vascular responses to stimulation of soluble guanylate cyclase) and of M&B 22,948 (which selectively inhibits cyclic GMP phosphodiesterase) on the relaxation induced in rabbit aorta by the atrial natriuretic peptide, atriopeptin II (which stimulates particulate guanylate cyclase). 2 Pretreatment with M&B 22,948 (100 μM) produced a 2.3 fold potentiation of atriopeptin II‐induced relaxation of endothelium‐denuded rings of rabbit aorta. 3 Pretreatment with haemoglobin (10 μM) had no effect on the relaxation or the 10.9 fold increase in cyclic GMP content induced by atriopeptin II in endothelium‐denuded rings of rabbit aorta. 4 The potentiation by M&B 22,948 suggests a causal role for cyclic GMP in mediating atriopeptin II‐induced vasodilatation of rabbit aorta. 5 The inability of haemoglobin to block the atriopeptin II‐induced rise in cyclic GMP suggests that it does not block stimulation of particulate guanylate cyclase. Thus, it is unlikely that a ferrous haem‐containing receptor site is involved in the activation of the particulate form of guanylate cyclase as it is with soluble guanylate cyclase.