Excretion of prostacyclin and thromboxane metabolites before, during, and after pregnancy-induced hypertension.

The aim of the present study was to assess whether changes in prostacyclin (PGI2) and thromboxane (TXA2) generation precede the manifestation of pregnancy-induced hypertension (PH). The metabolites 6-oxo-PGF1 alpha and TXB2 were measured in the urine of 69 randomly selected pregnant women from 16-20 weeks of gestation (wg) until delivery and more than 6 weeks postpartum. Between 16-20 and 21-24 wg 6-oxo-PGF1 alpha excretion did not change in patients who later developed PIH (n = 6) but increased significantly in the control group (n = 63). In contrast, a marked rise in TXB2 excretion was found in the PIH group but not in controls. Thereafter significant differences between both groups persisted from 25 wg until delivery. The 6-oxo-PGF1 alpha/TXB2 ratio was below the 10th percentile from 21-24 wg until delivery in patients with developing PIH. The excretion of both metabolites was substantially lower in the non-pregnant state without any difference between patient groups. These results show an altered urinary excretion of both 6-oxo-PGF1 alpha and TXB2 preceding the onset of the disease. A pathophysiological role of PGI2 deficiency and increased TXA2 formation in PIH appears substantiated.

Protection of reperfused ischemic pig myocardium by nexopamil, a new combined Ca2+ and serotonin antagonist.

We investigated the ability of a newly developed calcium and serotonin (5-HT2) antagonist, nexopamil, to protect the heart from ischemia- and reperfusion-induced myocardial injury. Anesthetized open-chest minipigs were subjected to 1 h left anterior descending coronary artery (LAD) occlusion and 3-h reperfusion. Thirty minutes before occlusion, one group of pigs (n = 7) received nexopamil (0.1 mg/kg intravenously, i.v.) and another group (n = 9) received vehicle. Nexopamil reduced infarct size (IS: tetrazolium stain) from 47 +/- 4% (vehicle) to 21 +/- 7% of the ischemic area (p < 0.05). In nexopamil-treated pigs, this was paralleled by reduced release of creatine kinase (CK) into coronary venous blood. In addition, nexopamil prevented reperfusion-associated myocardial contracture. Nexopamil decreased left ventricular peak pressure (LVPP) and pressure rate index (PRI) immediately before coronary occlusion by 11 and 18%, respectively. Coadministration of methoxamine (2 mg/kg, n = 6) with nexopamil increased LVPP and PRI to values of vehicle-treated pigs but did not prevent reduction in infarct size or CK activity in plasma. During reperfusion, neutrophil granulocytes showed increased formation of reactive oxygen metabolites (chemiluminescence) after stimulation with zymosan. Neutrophil counts in coronary venous blood were significantly reduced at 3 h reperfusion. Both changes were attenuated in nexopamil-treated pigs. Coronary occlusion resulted in increased platelet reactivity in coronary venous blood (collagen-induced aggregation) that was prevented by nexopamil. Nexopamil significantly increased the transcardiac (coronary venous-arterial) concentration gradients of 6-oxo-prostaglandin F1 alpha (PGF1 alpha) without changing thromboxane (B2 (TBX2) concentrations, indicating a selective increase in cardiocoronary PGI2 formation. Nexopamil reduces myocardial injury in reperfused ischemic myocardium. Besides calcium channel blocking activity, inhibition of ischemia-induced neutrophil activation and enhanced endogenous PGI2 formation may be factors contributing to the beneficial effects of nexopamil.

Incomplete Inhibition of Platelet Secretion by Low-dose Aspirin.

This study determines the effects of low-dose oral acetylsalicylic acid (Aspirin) on platelet aggregation and dense granule secretion in relation to inhibition of thromboxane formation. In addition, possible modifications of inhibition of platelet function by iloprost and linsidomine (SIN-1) were also studied. Aspirin (40 mg/day) was administered to 15 healthy male non-smoking volunteers for 8 consecutive days. Platelet function was measured ex vivo before and 12 h after the last drug intake. At the same times urinary excretion of thromboxane B(2) (TXB(2)) and 6-oxo-PGF(1α) were determined. Asirin treatment resulted in a significant (P < 0.01), by 65%, reduction of urinary TX B(2) excretion, whereas urinary excretion of 6-oxo-PGF(1α) remained unchanged, demonstrating the expected selectivity of low-dose aspirin for the platelet cyclooxygenase. There was also a nearly complete (93-95%) inhibition of collagen (0.3-5 µg/ml)-induced thromboxane formation in platelets. However, collagen-induced ATP-and 5-HT secretion was much less inhibited by aspirin and was reduced by only 20% at 5 µg/ml collagen. Similar results were obtained with ADP while thrombin (> 1.2 IU/ml)-induced 5-HT secretion was not antagonized at all. Also aspirin did not affect the inhibition of platelet function by iloprost and linsidomine (SIN-1). It is concluded that stimulation of platelet secretion by stronger stimuli, such as collagen or thrombin, is largely cyclooxygenase-independent and may not be detected by measuring serum thromboxane levels. This may be relevant for clinical situations with shear-stress-induced platelet activation, for example at carotis stenoses or other types of atherosclerotic plaques.

Direct inhibition of platelet function by organic nitrates via nitric oxide formation.

This study investigates the mechanisms of platelet inhibition by the nitrate esters isosorbide dinitrate, isoidide dinitrate, isomannide dinitrate, isosorbide 2-mononitrate and isosorbide 5-mononitrate as compared to the spontaneous nitric oxide (NO)-donor linsidomine, the active metabolite of molsidomine. Nitrates and linsidomine dose-dependently inhibited aggregation, ATP secretion and thromboxane formation of washed human platelets at a rank order of potency, identical with that for stimulation of cyclic GMP in cultured rat lung fibroblasts. While linsidomine (0.1 mM) caused a 3-fold platelet cGMP elevation, there was a weak (< or = 30%) but significant cGMP stimulation by organic nitroesters, which was tightly correlated with inhibition of platelet aggregation (r = 0.926, P = 0.008). Zaprinast (2 microM) potentiated, while methylene blue (1 microM) and oxyhemoglobin (10 microM) reversed the antiaggregatory effects. Linsidomine (0.5 microM-0.1 mM) dose-dependently released NO in a cell-free system. No spontaneous NO release was detected with organic nitroesters (0.1 mM). These data suggest that, to some extent, bioactivation of organic nitroesters occurs in platelets, resulting in platelet inhibition via the NO/cGMP system.

Equipotent inhibition by R(-)-, S(+)- and racemic ibuprofen of human polymorphonuclear cell function in vitro.

1. The effects of racemic (rac) ibuprofen and its S(+)- and R(-)-enantiomers on functions of human polymorphonuclear cells (PMN) and platelets were studied in vitro. 2. Rac-ibuprofen inhibited PMN functions (O2- generation, beta-glucuronidase release, LTB4 formation). Similar IC50 values (40-100 microM) were obtained for the S(+)- and R(-)-enantiomers. 3. All forms of ibuprofen inhibited cyclooxygenase-related platelet functions (aggregation, thromboxane formation). The S(+)-enantiomer was about twice as active as the racemate while the R(-)-enantiomer was at least 10-fold less active. This demonstrates that the S(+) is the only cyclooxygenase inhibitory component of the racemate. 4. The concentrations of rac-ibuprofen in PMN and platelets were similar to those in the incubation medium and represented equal concentrations of the enantiomers. This indicates that neither interconversion nor tissue accumulation of the compounds occurred. 5. These data indicate that antineutrophil effects of ibuprofen on human PMN are independent of cyclooxygenase inhibition. Therefore, R(-)-ibuprofen may be superior to the S(+)-isomer for the treatment of PMN-dependent inflammatory diseases. However, effective free drug concentrations may not be obtained in vivo.

Cardioprotective actions of wild garlic (allium ursinum) in ischemia and reperfusion.

The susceptibility to ventricular arrhythmias under the conditions of cardiac ischemia and reperfusion was investigated in the Langendorff heart preparation of rats fed for eight weeks a standard chow enriched with 2% of pulverized wild garlic leaves. The isolated hearts were perfused with a modified Krebs-Henseleit solution. The incidence of ventricular fibrillation (VF) during 20 min occlusion of the descending branch of the left coronary artery (LAD) was significantly reduced in the wild garlic group as compared to untreated controls (20% vs 88%). The same holds for the size of the ischemic zone (33.6% vs 40.9% of heart weight). In the reperfusion experiments (5 min after 10 min ischemia), ventricular tachycardia (VT) occurred in 70% of the wild garlic group vs 100% in untreated controls and VF in 50% vs 90%. The time until occurrence of extrasystoles, VT or VR was prolonged. No significant alterations in cardiac fatty acid composition could be observed. Although the prostacyclin production was slightly increased in hearts of the wild garlic group, inhibition of cyclooxygenase by acetylsalicylic acid (ASA; aspirin) could not completely prevent the cardioprotective effects suggesting that the prostaglandin system does not play a decisive role in the cardioprotective action of wild garlic. Furthermore, a moderate angiotensin converting enzyme (ACE) inhibiting action of wild garlic was found in vitro as well as in vivo that could contribute to the cardioprotective and blood pressure lowering action of wild garlic. Whether a free radical scavenging activity of wild garlic is involved in its cardioprotective effects remains to be established.

Oral naftidrofuryl prevents platelet hyperreactivity ex vivo and inhibits functional desensitization to prostacyclin in hypercholesterolemic rabbits.

Among other mediators, platelet-derived serotonin (5-HT) may contribute to thromboembolic complications of atherosclerosis. We determined whether long-term oral treatment with the 5-HT2 antagonist naftidrofuryl (NAF, 50 mg/kg daily for 12 weeks) alters platelet function in cholesterol-fed (1%) rabbits. Hypercholesterolemia resulted in marked platelet hyperreactivity to collagen and ADP. This included increased aggregation, ATP secretion, and thromboxane formation; e.g., collagen-induced (1.2 micrograms/ml) platelet aggregation was stimulated to 210 +/- 10 mm/30 s in cholesterol-fed rabbits as compared with 108 +/- 9 mm/30 s in rabbits fed a standard diet (p < 0.05). Inhibition of ADP-stimulated platelet activation by the prostacyclin mimetic iloprost was significantly reduced. NAF did not reduce plasma cholesterol in hypercholesterolemia, but prevented enhanced platelet aggregation, thromboxane formation, and ATP secretion. NAF treatment significantly reduced collagen-induced (1.2 micrograms/ml) aggregation to 81 +/- 20 mm/30 s in these animals (p < 0.05). NAF also inhibited functional desensitization of platelets to iloprost, but did not alter the impaired binding of [3H]iloprost to platelet membranes in hypercholesterolemic animals. NAF also did not change any of these parameters in normocholesterolemic rabbits. These data suggest beneficial effects of NAF on platelet hyperreactivity in experimental hypercholesterolemia which may also be relevant for its clinical use.

Stimulation of endogenous prostacyclin protects the reperfused pig myocardium from ischemic injury.

Several attempts have been undertaken to reduce the severity of ischemic myocardial injury by exogenous administration of eicosanoids and by modification of endogenous eicosanoid production. The present study investigates whether defibrotide, a compound that stimulates endogenous prostacyclin (PGI2), has a beneficial effect in experimental ischemic myocardial injury. Anesthetized, open-chest minipigs were subjected to 1 h of coronary artery occlusion, followed by 3 h of reperfusion. Defibrotide (32 mg/kg x h) or its vehicle were infused i.v. throughout the experiment. Defibrotide increased cardiac PGI2 formation 3- to 4-fold greater than control (P < .05). Thromboxane levels remained unchanged. Irreversible ischemic injury, as identified by negative tetrazolium staining, amounted to 44 +/- 6% of the area at risk in pigs receiving vehicle but was reduced to 23 +/- 4% by defibrotide (P < .05). This reduced tissue injury in defibrotide-treated pigs was associated with improved functional recovery (left ventricular pressure, + dP/dtmax), during early reperfusion. Recovery did not occur in vehicle-treated pigs. Collagen (2 micrograms/ml)-induced platelet aggregation ex vivo was increased in vehicle-treated pigs during ischemia and reperfusion, but not in animals treated with defibrotide. Polymorphonuclear neutrophil leukocyte accumulation in the ischemic border zone was reduced from 59 +/- 17 cells/mm2 in vehicle-treated pigs to 17 +/- 9 cells/mm2 by defibrotide (P < .05). Pretreatment of the animals with indomethacin (3 mg/kg) prevented the reduction of infarct size and polymorphonuclear neutrophil leukocyte infiltration by defibrotide. Indomethacin increased infarct size in vehicle- and defibrotide-treated pigs by 71 and 59%, respectively.(ABSTRACT TRUNCATED AT 250 WORDS)

Prostacyclin rather than nitric oxide lowers human umbilical artery tone in vitro.

This study was designed to determine vasodilator activities of two endothelium-derived relaxing factors: prostacyclin (PGI2) and nitric oxide (NO) in human umbilical arteries. Isolated vessel segments were contracted by submaximal concentrations of serotonin and bradykinin. These contractions were enhanced after inhibition of prostaglandin formation by the cyclooxygenase inhibitor indomethacin and after removal of the endothelium, both resulting in a pronounced decrease in PGI2 formation. Contractions remained unchanged after treatment of the vessels with nitro-L-arginine, a selective inhibitor of endogenous NO biosynthesis. The efficacy of inhibition of NO biosynthesis was established by a more than 60% reduction in cyclic GMP accumulation. Even inhibition of stimulated NO formation by histamine did not change vascular tone. These data suggest that PGI2 rather than NO is an endothelium-derived relaxing factor in human umbilical arteries.

Favourable effect of defibrotide in lipid A-induced shock in pigs.

Defibrotide (DEF), a compound previously found to stimulate vascular prostacyclin (PGI2) formation, has been investigated in an experimental model of septic shock. Anesthetized pigs were subjected to i.v. infusion of lipid A (1.5 mg/kg per hr for 4 hr). DEF (50 mg/kg per hr) or vehicle were infused i.v. throughout the experiments, starting 1 hr prior to lipid A. Two out of 7 pigs receiving vehicle survived lipid A infusion for 4 hr, whereas 6 out of 7 DEF treated animals survived this period (P less than 0.05). DEF delayed the shock-induced depression of platelet count and preserved platelet secretory function (collagen-induced ATP-secretion). DEF increased plasma PGI2 by 45% (P less than 0.05) during lipid A infusion and tended to reduce thromboxane levels. DEF did not change eicosanoid formation in sham-shock pigs (n = 4 per group). In vivo treatment with DEF significantly increased the stimulatory effect of bradykinin (1 microM) and arachidonic acid (100 microM) on PGI2 formation ex vivo of mesenteric and iliac artery segments. The improvement of survival in lipid A-induced shock by DEF may be related to an enhancement of vascular PGI2 generation, potentially due to a reduction of shock-induced platelet activation and microcirculatory dysfunction.

Nitric oxide but not prostacyclin is an autocrine endothelial mediator.

Using porcine aortic endothelial cells, the present study investigates whether stimulation of prostacyclin (PGI2) and nitric oxide also causes elevation of the respective second messengers cAMP and cGMP in the endothelial generator cells. The calcium ionophore A23187 at 0.3-3 microM increased endothelial cGMP levels up to 27-fold in an L-arginine-dependent manner as assessed through complete inhibition by NG-monomethyl-L-arginine (100 microM). The 36-fold PGI2 stimulation by 3 microM A23187 was not accompanied by an intracellular increase in cAMP or an enhanced cAMP efflux. Correspondingly, the PGI2 mimetic iloprost (10 pM-100 microM) did not change endothelial cAMP levels. However, forskolin (1-100 microM) and prostaglandin E2 (PGE2) (0.1-10 microM) produced concentration-dependent increases in cAMP with a 9-fold and 8-fold stimulation at 100 microM forskolin and 10 microM PGE2, respectively. These results demonstrate that in contrast to NO, PGI2 acts as a strictly paracrine hormone without affecting the respective second messenger cAMP in the endothelial generator cells.

Any physiological role for prostacyclin in regulation of fetal vessel tone?

This study investigates the contribution of endogenous prostacyclin and nitric oxide (NO) for maintenance of umbilical artery tone in vitro. Treatment with indomethacin as well as removal of the endothelium was followed by contractions of umbilical arteries and a reduced formation of prostacyclin. In contrast, vessel tone was not affected by inhibition of NO synthesis, using nitro-L-arginine. Thus, prostacyclin rather than nitric oxide appears to be the endogenous endothelium-derived factor in the umbilical circulation that is involved in regulation of vessel tone.

The role of thromboxane and prostacyclin in ciclosporin-induced nephrotoxicity.

The aim of the study was to determine the influence of ciclosporin (Cs) on prostacyclin and thromboxane generation. Four groups of patients were studied. Group 1: Controls, n = 10. Group 2: Patients without kidney disease treated with Cs, n = 10. Group 3: Patients after transplantation treated with azathioprine, n = 10. Group 4: Renal transplant recipients receiving Cs, n = 10. Parameters investigated: CIn, CPAH, TxB2 in plasma, serum and urine; 6-oxo-PGF1 alpha in plasma and urine, urinary 2,3-dinor-TxB2 excretion. CPAH and CIn were significantly decreased during Cs treatment. Plasma TxB2 levels were enhanced in patients without kidney disease receiving Cs (group 2) amounting to 189 +/- 106 pg/ml as compared to 12 +/- 4 pg/ml in controls (group 1). In patients without kidney disease (group 2), plasma 6-oxo-PGF1 alpha was increased (20 +/- 9 pg/ml) as compared to controls in group 1. Plasma TxB2 and plasma 6-oxo-PGF1 alpha were increased in renal graft recipients without any difference due to different immunosuppressive drugs. Treatment with Cs was associated with impaired renal function and resulted, in patients without kidney disease, in elevated plasma TxB2 and plasma 6-oxo-PGF1 alpha. This effect could not be proven in renal graft recipients. We suggest that the deleterious effect of Cs on kidney function is presumably not paralleled by corresponding changes in prostaglandin and thromboxane formation.

Fat emulsions containing medium chain triglycerides in patients with sepsis syndrome: effects on pulmonary hemodynamics and gas exchange.

Fat emulsions containing medium chain triglycerides (MCT) have recently been introduced into clinical practice as a component of total parenteral nutrition. Since several authors reported increased pulmonary artery pressure and impaired gas exchange during intravenous (i.v.) fat use, in particular in septic patients, we studied the pulmonary hemodynamic and gas exchange effects of i.v. fat containing MCT and long chain triglycerides (LCT) in patients with sepsis syndrome. As the effects of fat emulsions have been attributed to increased formation of prostanoids, the production of thromboxane A2 and prostacyclin was investigated by the determination of urinary thromboxane B2 and 6-keto-prostaglandin F2 alpha, respectively. The i.v. fat use did not induce any alterations in pulmonary hemodynamics and gas exchange, the distribution of ventilation and perfusion nor urinary prostaglandin content. We conclude that fat emulsions containing MCT induce little alterations in pulmonary hemodynamics and gas exchange. This result is probably due to reduced prostaglandin formation because fat emulsions containing MCT provide less prostaglandin precursors than pure LCT emulsions.

Oral cicaprost protects from hypercholesterolaemia-induced impairment of coronary vasodilation.

Isolated Langendorff-hearts prepared from cholesterol fed rabbits (1% cholesterol for 3 months) showed a significant impairment in endothelium-dependent relaxation after short-term infusion of bradykinin (0.05 mumol/l) and carbamoylcholine (0.1 mumol/l). Generation of the endothelial mediators nitric oxide and prostacyclin by bradykinin was enhanced in hypercholesterolemia. Cicaprost treatment (5 micrograms/kg x d) largely prevented the hypercholesterolemia-related impairment of coronary vasodilation and nitric oxide release. It is concluded that (i) impairment of endothelium-dependent relaxation in the coronary microcirculation of hypercholesterolemic rabbits is not due to diminished endothelium-dependent mediator release but rather to accelerated inactivation or reduced activity of the released mediators and that (ii) oral cicaprost beneficially influence these alterations.

Stimulation of prostacyclin synthesis by defibrotide: improved contractile recovery from myocardial "stunning".

Prostacyclin (PGI2) improves regional contractility of postischemically dysfunctional ("stunned") myocardium. We determined whether defibrotide, a fraction of mammalian DNA known to stimulate endogenous formation of PGI2, also improves contractile recovery of stunned myocardium. Anesthetized, open-chest minipigs were subjected to coronary occlusion of 5 min (left anterior descending branch, LAD) followed by 120 min of reperfusion. The animals were treated with defibrotide (32 mg x kg-1 x h-1, intravenously, i.v.) or vehicle throughout the experimental period. Defibrotide improved regional contractility in the ischemic reperfused area from 30 (vehicle) to 78% of the preischemic control without altering the contractility of nonischemic myocardium. Transcardiac PGI2 formation, determined from the difference between coronary venous and arterial plasma concentrations, was elevated from 437 (preischemic control) to 869 pmol x l-1 in defibrotide-treated animals, but was unchanged in the vehicle-treated and a sham-operated group. Thromboxane A2 (TXA2) release was not modified. Defibrotide reduced ischemia-induced formation of platelet aggregates but did not affect the activity of polymorphonuclear neutrophil granulocytes. The data demonstrate an improvement of contractile recovery from stunning by defibrotide that may be related to an inhibition of ischemia-induced platelet activation and (or) membrane protection owing to enhanced transcardiac formation of PGI2.

Interactions between nitric oxide and prostacyclin in myocardial ischemia and endothelial cell cultures.

This study investigates biochemical and functional interactions between NO and PGI2 that generate pathways in two different in vitro assays: porcine aortic endothelial cells (PAEC) and reperfused ischemic Langendorff hearts of rabbits. Using cGMP as an index of NO generation and 6-oxo-PGF1 alpha as an index for PGI2 production in endothelial cells, it is demonstrated that the two metabolic pathways for NO and prostacyclin formation act independent of each other. Moreover, NO appears to have an autocrine function in endothelial cells which does not exist with PGI2, probably because of a lack of PGI2 receptors. Endothelial damage in the course of myocardial ischemia is associated with a marked increase in mediator release whose inhibition has consequences for both myocardial and coronary function: inhibition of NO formation also inhibits PGI2 release and the recovery of coronary vessel tone with only minor if any effect on myocardial contractility. In contrast, inhibition of PGI2-generation results in marked deterioration of myocardial recovery with only minor changes in coronary perfusion. It is concluded from these data that PGI2 in endothelial injury is important for preservation of myocardial function while NO might mainly be involved in control of local vessel tone.

Helenalin and 11 alpha,13-dihydrohelenalin, two constituents from Arnica montana L., inhibit human platelet function via thiol-dependent pathways.

This study investigates the effect on human platelet function of two sesquiterpene lactones from Arnica montana L., helenalin (H) and 11 alpha,13-dihydrohelenalin (DH). Both compounds inhibited collagen-induced platelet aggregation, thromboxane formation and 5-hydroxytryptamine secretion in a concentration-dependent manner at 3-300 microM. When arachidonic acid was used as stimulus, thromboxane formation remained unaffected despite of inhibition of platelet aggregation. Both H and DH reduced the number of acid-soluble sulfhydryl groups in platelets, by up to 78% at anti-aggregatory concentrations. Moreover, H- and DH-induced platelet inhibition could be prevented by the thiol containing amino acid cysteine. It is concluded that H and DH inhibit platelet function via interaction with platelet sulfhydryl groups, probably associated with reduced phospholipase A2 activity.

Winner of the ESVS Prize 1988. Effects of prostaglandin E1 (PGE1) on experimental renal ischaemia.

A persisting incidence of acute renal failure has been observed after operative treatment of thoracoabdominal aortic aneurysm, ruptured abdominal aortic aneurysm and renal artery occlusive disease in patients with preoperative impairment of renal function. Because preservation of kidney function can play an important role in the outcome of these patients, the effects of prostaglandin E1 (PGE1) to prevent ischaemic renal failure were studied in an experimental model. Twenty dogs were exposed to 3 h warm ischaemia by clamping of the supra- and infrarenal aorta and both renal arteries. In 10 dogs PGE1 was given intravenously (100 ng/kg/min) for 15 min before clamping. Ten dogs treated with normal saline served as controls. Glomerular filtration rate, renal plasma flow, plasma creatinine, blood urea nitrogen, urine volume, free water clearance and renovascular resistance were calculated before and after renal ischaemia for both groups. The dogs were followed-up for 2 weeks and radionuclide studies with Tc-99m-MAG3, I-131-OIH and In-113m-DTPA were performed on the third postoperative day to calculate global and split renal clearance, tracer extraction fraction and mean transport time. After renal ischemia 9 dogs of the control group and 3 dogs of the PGE1-group developed acute renal failure (P less than 0.05 due to Fisher's exact text). PGE1 infusion significantly attenuated the postischaemic fall in glomerular filtration rate and renal concentrating ability as well as the postischaemic increase of plasma creatinine and blood urea nitrogen induced by 3 h warm renal ischaemia (P less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)