Role of lipoxygenase products in the effects of angiotensin II in the isolated aorta and perfused heart of the rat

The objective of this study was to determine whether arachidonate metabolites are involved in the vasoconstrictive effects of angiotensin II in rats. In the isolated perfused heart, dexamethasone (4 mg/kg) significantly suppressed the maximal decreases in coronary flow induced by angiotensin II and vasopressin (reference drug). In the heart, the nonselective lipoxygenase inhibitor nordihydroguaiaretic acid (NDGA, 1 μM) markedly suppressed the angiotensin II-induced decreases in coronary flow. NDGA (10 μM) inhibited both angiotensin II- and methoxamine- (reference drug) induced contractions in aortic rings with (in the presence of L-NAME) and without endothelium. In the heart, the leukotriene synthesis inhibitor MK-886 (0.3 μM) significantly reduced the maximal effects to angiotensin II, but the leukotriene antagonist FPL 55712 (0.1 and 0.3 μM) had no effect. We conclude that in the isolated perfused rat heart angiotensin II-induced decreases in coronary flow are in part mediated by Hpoxygenase products, which might be derived from the 5-Hpoxygenase pathway, but are probably not leukotrienes. Furthermore, endothelium independent Hpoxygenase products mediate part of the contractile responses to angiotensin II in the isolated rat aorta.


Introduction
Vascular angiotensin II is involved in the longterm regulation of blood vessel function and structure and is a major pathophysiological factor in hypertension, atherosclerosis and restenosis. There is accumulating evidence for the presence of a complete renin-angiotensin system in the heart, and a functional role has been suggested. [2][3][4][5] Locally generated angiotensin II could be of importance in the regulation of coronary flow, as has for instance been demonstrated in patients with microvascular angina (Syndrome X). 6 Furthermore, angiotensin II has a direct effect on protein synthesis, which contributes to cardiac hypertrophy. 7 It is well known that angiotensin II exerts its effects through stimulation of inositol phospholipid metabolism. Angiotensin II activates phospholipase C, which hydrolyses phosphatidylinositol biphosphate into inositol triphosphate and 1,2-diacylglycerol. 8 In addition, it has been demonstrated recently that 12-1ipoxygenase activation plays a key role in, e.g. angiotensin IIinduced vascular smooth muscle cell hypertrophy. 9 Other studies have shown that arachidonate metabolites contribute to various responses to angiotensin II, including inflammatory, contractile and secretory actions, m-6 This paper deals with the contribution of arachidonate metabolites to the vasoconstrictive effect of angiotensin II in the coronary circulation and aorta of the rat. If arachidonate metabolites play a role in coronary constrictions evoked by angiotensin II, it could be hypothesized that drugs inhibiting arachidonate metabolism have a beneficial effect in the prevention of angiotensin II-induced vasospasms, that can lead to angina pectoris or hypertension. To assess the contribution of arachidonate metabolites to angiotensin II-induced responses, we tested several drugs that interfere with arachidonate metabolites in isolated Langendorff perfused rat hearts and isolated aortas. To study whether arachidonate dependent effects were selective for angiotensin II we also tested methoxamine and vasopressin as agonists, which like angiotensin II exert their effects through stimulation of inositol phospholipid metabolism. We examined the cyclo-oxygenase pathway in the effects of angiotensin II by using the cyclooxygenase inhibitor indomethacin, the lipoxygenase pathway was studied by using the nonselective lipoxygenase (and cyclo-oxygenase) inhibitor  In the heart, left ventricular pressure was measured by a water filled balloon (HSE, Freiburg, Germany) and recorded on a Gould recorder.
The diastolic left ventricular pressure was adjusted between 5 and 10mm Hg. The heart was -|,-electrically driven at a frequency of 5Hz (32C) AA861 and equilibrated for  Experiments with endothelium were con-equilibration time of 30 min and during the conducted with and without 100 l.tM L-NAME added struction of the dose-response curves. One to the incubation medium, dose-response curve was constructed per heart Finally, concentration-response curves for in all treatment and control groups. angiotensin II (0.1-1000 nM) were constructed after 30 min of incubation with losartan (0.01, Data analysis: 0.1 and 1 laM) in aortic rings with endothelium in the presence of l O01.tM L-NAME or in tings without endothelium.
Drugs and drug solutions.. Angiotensin II (acetate salt), indomethacin, NDGA, methoxamine HCl, phenylephrine HCl, acetylcholine HCl, L-NAME and sodium nitroprusside (SNP) were obtained from Sigma Chemical (St. Louis, MO, USA). AA 861 (2,3,4,5-trimethyl-6-(12-hydroxy-5,10-dodecadiynyl)-l,4-benzoquinone) was obtained from Takeda Chem. Ind. Ltd. (Osaka, Japan). Losartan (DuP 753) was a gift from du Pont de Nemours & Company (Wilmington, Del., USA). Heparin was from Leo (Weesp, the Netherlands) and sodium pentobarbitone was from Ceva. (Paris, Heart. When baseline coronary flow was significantly enhanced by drugs (NDGA, AA 861, MK-886 or FPL 55712), we used control groups with SNP in sufficient concentration to reach a baseline coronary flow that was equal to the baseline coronary flow in the drag groups. These control groups with SNP were added to exclude the possibility that the enhancing effects of the drugs on baseline coronary flow might obscure the possible effects of these drugs on angiotensin II-induced constrictions. Baseline coronary flow values were analysed by Student's t-test or ANOVA followed by Duncan's Range Test for multiple comparison.
Coronary vasoconstriction induced by angiotensin II and vasopressin is expressed as the per-France). Vasopressin (Vasopressin-Sandoz: lypressin) was a gift from Sandoz, Wander centage reduction of baseline coronary flow. Pharma (Uden, the Netherlands). FPL 55712 Dose-response curves of angiotensin II and vaso-(sodium 7-[3-(4-acetyl-3-hydroxy-2-propylphe-pressin were analysed by MANOVA with a noxy)-2-hydroxypropoxy]-4-oxo-8-propyl-4H-l-b-repeated measures design. enzopyran-2-carboxylate) was a gift from Fisons Aorta. Contractions are expressed as percentage Pharmaceuticals (Loughborough, England). MKof the maximal contraction of 100 mM KCl. Con-886 was a gift from Merck Frosst (Quebec, centration-response curves for angiotensin II, Canada) and dexamethasone phosphate (decamethoxamine and vasopressin were analysed by dron) was obtained from O.P.G. (Utrecht, the MANOVA using a repeated measures design. Netherlands). All other chemicals were of analy-Heart and aorta. A p-value < 0.05 was contical grade (Merck, Darmstadt, Germany). sidered statistically significant. Data are expressed Unless otherwise specified, drugs were disas means + S.E.M. In the figures, a significant solved in saline. Dexamethasone phosphate difference-between the entire dose/concen-(4mg/kg) and saline (2ml/kg) were administration-response curves is indicated with an tered into the tail vein 5.5h prior to testing, asterisk. NDGA, AA 861, and MK-886 were dissolved in ethanol. For the aorta experiments NDGA and AA 861 were diluted with saline (final concentra-Rult tion ethanol, < 0.1%). For the heart experiments NDGA, AA 861 and MK-886 were diluted to the Isolated Langendorffperfused heart: appropriate concentration in the perfusion fluid (final concentration ethanol, <0.005%). FPL 55712, SNP and losartan were dissolved in bidistilled water and diluted to the appropriate concentration in the perfusion fluid. Indomethacin NaHCO. heart, 27mM; aorta, 24.9mM). Vaso-Angiotensin II and vasopressin effects. Angiopressin was diluted with saline. Solutions were tensin II induced a dose-dependent decrease in freshly prepared every day and kept on ice and coronary flow with a pD2 of 11.5 4-0.1 and an protected from light. Emax of 55.9 4-2.4% (n 8) (Fig. 2). Losartan In experiments using isolated hearts, NDGA, (0.01-101.tM) reduced the angiotensin II-induced AA 861, MK-886, FPL 55712, SNP, losartan (and decreases in coronary flow in a concentration indomethacin) were already present in the perfu-dependent manner. The effect of losartan 10-5M sion fluid before mounting the hearts, during the on the response to angiotensin II is shown in  Modulation by dexamethasone. After pretreatment with 4mg/kg dexamethasone for 5.5 h, the maximal angiotensin IIand vasopressin-induced reductions in coronary flow were significantly reduced from 53.7 _ _ _ 3.8% (n 5, saline pretreated rats) to 40.5 3.6% (n 5; Fig. 3, left panel) and from 74.9 +_ 7.2% (n 6, saline pretreated rats) to 57.5 3.6% (n 6; Fig. 3 All,-log dose (moles) 16  All,-log dose (moles) 16  there was no change in maximal effect (Fig. 7c).
The lipoxygenase inhibitor had no effect on the baseline tension of these rings (data not shown).
Effects of indomethacin and AA 86I on angiotensin II-, and methoxamine-induced contractions in aortic rings without endothelium. AA 861 (10 btM) and indomethacin (10 l.tM) had no effect on the contractile responses to both angiotensin II and methoxamine ( Table 2).
Effects of NDGA on angiotensin II-, and methoxamine-induced contractions in aortic rings with endothelium in the presence of L-NAME. In this set of experiments, angiotensin II and methoxamine induced concentration dependent contractions in aortic rings with endothelium with a maximum of 78.0 _ _ _ 9.9% (eleven rings, six rats) and 199.0 __+ 10.4% (eleven rings, six rats) respectively (Fig. 8, left" panels).
In the presence of -NAME a second concentration-response curve was constructed for angiotensin II and methoxamine after incubation with either 0.1% ethanol (control) or 10M NDGA. In contrast to methoxamine, the maximal effect to angiotensin II of the second concentration-response curve was significantly lower compared to the first concentration-response curve. The lipoxygenase inhibitor significantly attenuated the maximal contraction to angiotensin II and shifted the concentration-response curve of methoxamine to the right (Fig. 8, right panels).

Effects of losartan on angiotensin II-induced
contractions in aortic rings without endothelium and with endothelium in the presence of L-NAME. Incubation of denuded aortic rings with losartan (10 nM, 100 nM and 1 M) shifted the concentration-response curve to angiotensin II to the right in a concentration dependent manner. With the highest concentration of losartan (1 btM) the angiotensin II-induced contractions (up to 300 nM angiotensin II) were virtually abolished. Similar effects were seen in intact aortic rings in the presence of 100 l.tM t-NAME (data not shown). The pAi-value calculated for losartan in rat aorta (with and without endothelium) was 8.2.

Discussion
The objective of this study was to determine whether arachidonate metabolites play a role in the vascular effects of angiotensin II in rats.  angiotensin II-induced decreases in coronary flow and inhibited the angiotensin II-induced From our results we can conclude that: (1) in contractions in the aorta in a concentration the coronary system of the rat, part of the anglo-dependent manner, while the cyclo-oxygenase tensin II-induced decreases in coronary flow are inhibitor indomethacin had no effect in both mediated by lipoxygenase products and not by tissues. NDGA, which is known to be a dual cyclo-oxygenase products; (2) these lipoxygenase cyclo-oxygenase and lipoxygenase inhibitor 22 as products might be derived from the 5-1ipoxy-well as an anti-oxidant, has been reported to genase pathway of arachidonate metabolism but have various nonspecific effects. Inhibition of are probably not leukotrienes; and (3) endotheendothelium-dependent relaxations 2 and of lium independent lipoxygenase products (not kinin and noradrenaline-induced contractions, 24 cyclo-oxygenase products) mediate part of the possibly related to the inhibition of the transcontractile responses to angiotensin II and to a membrane calcium influx, 25 have been described. lesser extent to methoxamine and vasopressin in However, based on our results we have reason the isolated rat aorta, to believe that lipoxygenase products (not cyclo-In the isolated perfused heart and aorta, angio-oxygenase products) mediate at least part of the tensin II acted via the AT1 receptor as was shown angiotensin II-induced constrictions, since in the by the inhibition by losartan. 8 We found that heart several inhibitors acting at different levels losartan reduced the angiotensin II-induced of the arachidonate metabolism (NDGA, dexadecreases in coronary flow and angiotensin IImethasone, MK-886) significantly reduced angioinduced contractions in the aorta, in a concentratensin II-induced decreases in coronary flow. tion dependent manner. In the presence of losar-Furthermore, we have shown that NDGA inhibtan 10 -5 M, the angiotensin II effect is almost ited the contractions to angiotensin II in a conabolished in the heart. However, we did not centration dependent manner in the aorta. perform experiments with a specific AT2-receptor The exact lipoxygenase pathway (5-, 12-, or antagonist and therefore cannot exclude a 15-1ipoxygenase) cannot be determined from our results. The leukotriene synthesis inhibitor MKaorta were also attenuated by NDGA; however, to 886 significantly reduced the maximal decrease a much lesser extent compared to angiotensin II. in coronary flow induced by angiotensin II in the Therefore, for vasopressin and methoxamine, heart. In contrast, in both aorta and heart, the 5lipoxygenase products may also be involved in lipoxygenase inhibitor AA 861 had no effect on (coronary) constrictions. The effects of angiotenthe responses to angiotensin II. MK-886 inhibits sin II in aortic rings (denuded) were characterthe membrane translocation of 5-1ipoxygenase by ized by a marked reduction of the maximal an interaction with the 5-1ipoxygenase activating effect, while for inhibition of the methoxamine protein and subsequently prevents the activation and vasopressin contractions a rightward shift of of 5-1ipoxygenase. 228 MK-886 is selective for the the concentration-response curve was observed 5-1ipoxygenase pathway of arachidonate acid and no effect or a minor effect on the maximal metabolism since it has no effect on 12-1ipoxycontractions was seen. In accordance with our genase or cyclo-oxygenase, 26 whereas AA 861 can results are observations in aorta strips of the also inhibit 12-hydroxyeicosatetraenoic acid (12rabbit where 10btM NDGA caused a 10% depres-HETE) formation. 29 On the basis of the selectivsion of the noradrenaline-induced tension and ity of MK-886, we suggest that the 5-1ipoxygenase AA 861 (up to 30 tM) had no effect. 24 Others pathway is the candidate that most likely med-have demonstrated that the constrictor response iates decreases in coronary flow induced by to a maximal concentration of norepinephrine angiotensin II.
was not affected by the lipoxygenase inhibitor The leukotriene antagonist FPL 55712 did not baicaleine, 6 and this is compatible with our change the angiotensin II-induced decreases in results. Additional studies are required to detercoronary flow, indicating that the end products mine whether the different profiles of NDGA from the 5-1ipoxygenase pathway, the leukoinhibition for angiotensin II and methoxamine/ trienes, are not involved in angiotensin II-induced vasopressin are related to differences in mechandecreases in coronary flow, leaving 5-hydroperism of action of the agonists. It has been shown oxyeicosatetraenoic acid (5-HPETE) and 5-hy-that 10M NDGA inhibits the stimulated influx droxyeicosatetraenoic acid (5-HETE)as potential of extracellular calcium in rabbit neutrophils, mediators of the angiotensin II-induced effects in without affecting cellular calcium redistribution. 25 the isolated perfused rat heart. This may contribute to the inhibitory action by Saito et al. also suggested that in cultured rat NDGA at 10/.tM on agonist-induced arterial convascular smooth muscle cells angiotensin II traction. In rat aortic rings, however, we already induced constrictions are mediated by H(P)ETEs. observed an inhibitory effect of NDGA at 3 tM.
However, in contrast with our results and in NDGA (10 tM) inhibited both angiotensin IIaccordance with those of other investigators with and methoxamine-induced contractions in aortic respect to smooth muscle cell growth' and ster-rings with (in the presence of >NAME) and oidogenesis, '4 Saito et aL emphasized the without endothelium, indicating that lipoxygenase role of 12H(P)ETE. Further studies are necessary products that mediate angiotensin II-and to determine which of these two pathways, the methoxamine-induced contractions do not origi-5-, or the 12-1ipoxygenase pathway is predominate from the endothelium. nant under specified conditions. NDGA, AA 861, MK-886 and FPL 55712 Indomethacin alone had no effect on angioten-induced a significant increase in baseline corosin II-induced constrictions in both aorta and nary flow suggesting that rat coronary flow is heart. The lack of effect for indomethacin on controlled in part by endogenously produced angiotensin II-induced contractions in rat aorta leukotrienes and perhaps other lipoxygenase was also shown in other studies. 3 However, in products. In accordance, several investigators the presence of AA 861 we observed a marked have found that AA 861, FPL 55712 and NDGA inhibitory effect of indomethacin on angiotensin increased baseline coronary flow in perfused rat II-induced effects in the isolated perfused heart, and guinea pig hearts. 33'34 A role for cyclo-oxyge-In guinea-pig hearts indomethacin induced shuntnase products in the regulation of rat coronary ing of the arachidonate metabolism through the flow is not likely, since indomethacin did not lipoxygenase pathway. 32 Such a mechanism affect baseline coronary flow in rats. We used might enhance the synthesis of lipoxygenase pro-control groups with SNP in sufficient concentraducts leading to e.g. 5-HPETEs or 5-HETEs when tion to reach a baseline coronary flow that was indomethacin is present, equal to the baseline coronary flow in the drug The inhibitory effect of NDGA was not selecgroups. To our knowledge none of the reported tive for angiotensin II. Vasopressin-induced arachidonic acid metabolites operate through the decreases in coronary flow and methoxaminesame mechanism as SNP, i.e. elevation of cGMP and vasopressin-induced contractions in the production. In addition, we performed some pilot experiments with NDGA in the presence of 10 nM SNP. In those experiments the angiotensin II-induced decreases in CF were not significantly different from the decreases in CF in the presence of NDGA alone, indicating that SNP did not interfere with the effect of NDGA.
In summary, lipoxygenase products and not cyclo-oxygenase products mediate at least part of the angiotensin II-induced decreases in coronary flow in the isolated perfused heart of the rat. These products also play a small role in the constrictions to vasopressin in the isolated rat heart. The lipoxygenase products involved in the effects of angiotensin II might be derived from the 5-1ipoxygenase pathway of arachidonate metabolism but are probably not leukotrienes. Furthermore, endothelium independent lipoxygenase products (not cyclo-oxygenase products) mediate part of the contractile responses to angiotensin II and to a lesser extent to methoxamine and vasopressin in the isolated rat aorta. Therefore, we can conclude that intermediary (endothelium-independent) lipoxygenase products other than leukotrienes play a role in the vascular effects of angiotensin II in the rat.
Experiments with 5-HPETE and/or 5-HETE to study whether the effects of these mediators are comparable with those of angiotensin II are an interesting subject for further investigation.