Although most studies have shown that leukotrienes

Transcription

1 Endothelial-Dependent Relaxation Induced by Leukotrienes C 4, D 4, and E 4 in Isolated Canine Arteries 93 Roberta J. Secrest and Barry M. Chapnick Leukotriene D 4 has been shown to possess the capacity to relax canine superior mesenteric and renal arterial rings in an endothelial-dependent manner. The present study was designed to determine if the remaining peptidoleukotrienes, leukotrienes C 4 and E 4, share this property. In addition, influences of atropine and of inhibitors of cyclooxygenase and lipoxygenase activities on relaxation produced by leukotriene D 4 and acetylcholine were determined to characterize further leukotriene D 4 -induced relaxation and to compare these properties with those of acetylcholine. Vasomotor tone was measured with isometric force transducers. Following induction of tone with norepinephrine, leukotriene C 4 and acetylcholine produced concentration-dependent relaxation of renal and superior mesenteric arterial rings in which the endothelium was intact. Only minimal decreases in tone were produced in response to leukotriene E 4. Neither acetylcholine nor leukotriene C 4 altered tone after the endothelium had been intentionally disrupted. Nitroglycerin relaxed rings both before and after rubbing the endothelium. These results demonstrate that, similar to leukotriene D 4, leukotriene C, possesses the capacity to produce endothelial-dependent relaxation in canine renal and superior mesenteric arteries. Relaxation of the superior mesenteric artery produced in response to acetylcholine, but not leukotriene D 4, was inhibited in presence of atropine. Incubation of the rings with meclofenamate had no effect on relaxation induced by either acetylcholine or leukotriene D 4. Thus, it appears that endothelialdependent relaxation induced by leukotriene D 4 is neither dependent on muscarinic receptor activation nor related to generation of cyclooxygenase metabolites of arachidonic acid. In contrast, 5,,11,14- eicosatetraynoic acid and nordihydroguaiaretic acid attenuated relaxation in response to leukotriene D 4 and acetylcholine, suggesting that lipoxygenase-derived products may participate in leukotriene D 4 -induced as well as acetylcholine-induced relaxation. (Circulation Research 19; 62:93-991) Although most studies have shown that leukotrienes contract vascular smooth muscle, divergent effects have been observed. Leukotrienes C 4 (LTC 4 ), D 4 (LTD 4 ), and E 4 (LTE 4 ) have been shown to produce marked vasoconstriction in the mesenteric vascular bed of both the anesthetized dog and cat but to have little vasoactivity in the canine kidney. 13 In addition, LTD 4 decreased coronary blood flow in anesthetized dogs, 4 while in anesthetized pigs, LTD 4 -induced coronary vasoconstriction was associated with concomitant release of a vasodilator substance. 5 In vitro, LTQ, and LTD 4 produced dosedependent contractions of helical strips of rabbit coronary arteries. 6 Other blood vessels obtained from rabbits were either unresponsive (renal artery, mesenteric artery, thoracic aorta) or only weakly responsive (pulmonary artery). 6 In contrast to the observations of Kito et al, 6 LTQ and LTD 4 relaxed spiral segments of dog coronary artery precontracted with 27 mm K. 7 From the Department of Pharmacology, St. Louis University School of Medicine, St. Louis, Missouri. Supported by National Institutes of Health grant HL Conducted while B.M.C. was a recipient of National Heart, Lung, and Blood Institute Research Career Development Award K Address for correspondence: Dr. Barry M. Chapnick, Department of Pharmacology, St. Louis University School of Medicine, 1402 South Grand Boulevard, St. Louis, Missouri R.J.S.'s present address is Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, IN Received December 17, 196; accepted December 3, 197. Thus, while leukotrienes are generally considered to be vasoconstrictor substances, it appears that these compounds also possess the capacity to produce vasodilation or vasomotor relaxation. Furthermore, LTD 4 has been shown to relax precontracted canine renal and superior mesenteric arterial rings in an endothelialdependent manner. Arterial relaxation induced by a variety of vasoactive substances has been shown to depend, at least partially, on the presence of a functionally intact endothelium. In addition, this vasomotor relaxation has been associated with the release of an unknown mediator postulated to be formed in the endothelium. Although the precise nature of this endothelium-derived relaxing factor (EDRF) remains undefined, cyclooxygenase products of arachidonic acid metabolism do not appear to be responsible for this activity. 9 However, two chemically different lipoxygenase inhibitors, 5,,11, 14-eicosatetraynoic acid (ETYA) and nordihydroguaiaretic acid (NDGA), attenuated or reversed endothelialdependent relaxation produced by acetylcholine (ACh), arachidonic acid, or bradykinin. Therefore, it was postulated that a product of arachidonic acid metabolism formed via the lipoxygenase pathway subserved endothelial-dependent relaxation. 910 Because peptidoleukotrienes failed to relax strips of rabbit aorta, it was concluded that endothelial-dependent relaxation was not mediated by these lipoxygenase metabolites of arachidonic acid."' 2 In addition, we have demonstrated that LTD 4 produced relaxation of

2 94 Circulation Research Vol 62, No 5, May 19 precontracted canine renal and superior mesenteric arterial rings in which the endothelial cell layer was intact. No change in vasomotor tone was induced in vascular rings following intentional disruption of the intima. These observations suggested that LTD 4, a lipoxygenase-derived product, probably was not a component of EDRF. Indeed, whether an EDRF was released by LTD 4 remains unknown. In view of the fact that the other members of the peptide-containing leukotriene family, LTC 4 and LTE 4, are well known to possess vasoactive properties and because LTD 4 is intermediate in the conversion of LTC 4 to LTE 4, the purpose of the present study was to determine whether LTC 4 and LTE 4 produce endothelial-dependent relaxation of canine arteries in vitro. An additional goal was to characterize further the influences of LTD 4 on vasomotor tone of isolated canine arteries in which the endothelial cell layer was intact. Because ACh is the prototypical endothelialdependent relaxing agent, vasomotor effects of LTD 4 were compared with those of ACh. Materials and Methods Adult male mongrel dogs (17-24 kg), fasted overnight but allowed free access to water, were anesthetized with sodium pentobarbital (30 mg/kg i.v.). The renal and superior mesenteric arteries were carefully excised, trimmed free of adhering fat and connective tissue, and cut into rings 2-3 mm long. Four to six ring segments were obtained from each artery. During preparation of the ring segments, special care was taken to avoid touching the luminal surface to ensure integrity of the endothelium. The rings were mounted as previously described in 20-ml jacketed glass organ chambers containing ml of the incubation medium, modified Krebs-Ringer bicarbonate solution. Isometric force was measured with Grass FT.03 force displacement transducers coupled to a Grass polygraph (model 7, Grass Instruments, Quincy, Massachusetts). All equilibrations and experimental procedures were conducted at 37 C. The modified Krebs-Ringer bicarbonate solution was continuously gassed with 95% O 2-5% CO 2 to maintain ph at 7.4 and had the following composition (mm): NaCl 11.3, KC1 4.7, CaCl 2 2.5, MgSO 4 1.2, KH 2 PO 4 1.2, NaHCO , Ca-EDTA 0.026, glucose Renal and superior mesenteric arterial rings were suspended under basal tensions of 4 g and g, respectively. These loads were near optimum for maximal isometric contraction in response to 20 mm KC1, as assessed from predetermined length-tension relations. An equilibration period of 60 minutes was allowed, during which the bath was rinsed every 15 minutes with the control buffer solution and the tension was adjusted to maintain the appropriate basal level. Vasomotor activity of leukotrienes was observed solely under conditions of active tone. Therefore, when basal tension was stable, tone was induced with concentrations of norepinephrine (NE) ranging from 10~ 7 to 3xlO" 6 M. These concentrations of NE produced a level of contractile activity that was 40-75% of the maximal response previously determined from dose-response curves relating accumulated concentration of NE to force of contraction. When the NE-induced contraction reached a steady state, ACh, an individual leukotriene, and nitroglycerin were each added to the incubation medium over a wide range of concentrations in a cumulative manner. The tissues were washed with 60 ml buffer and allowed to reequilibrate for 30 minutes after addition of each individual agonist. Effects of only one particular leukotriene were determined in each ring preparation. Interrelations between endothelial function and vasomotor activity of both LTC 4 and LTE 4 were evaluated in the following manner. In the presence of induced tone, control responses to ACh (10~ 7 M), nitroglycerin (10"' M), and the particular leukotriene were obtained before disrupting the endothelium. The tissues were then washed with 60 ml buffer and allowed to reequilibrate for 30 minutes. Following equilibration, the luminal surface of the ring was.rubbed gently with a cotton-tipped applicator for seconds to intentionally remove or disrupt the endothelial cell layer. Thirty minutes later, tone was induced with NE, and influences of ACh, nitroglycerin, and the leukotriene were again determined. If after this procedure, relaxation produced by ACh was decreased by at least 0% of the control response in absence of a decrease in relaxation produced by nitroglycerin, an endothelial - independent relaxing agent, the endothelium was considered to be functionally disrupted. Therefore, in this complete series of experiments, vasomotor responses to ACh, nitroglycerin, and an individual leukotriene were obtained and compared in a single ring segment both before and after rubbing the luminal surface. Because LTD 4 had previously been shown to produce endothelial-dependent relaxation of canine superior mesenteric and renal arterial rings, a separate series of studies was conducted to determine effects of atropine, meclofenamate, ETYA, and NDGA on LTD 4 -induced vasomotor relaxation. Analogous to the protocol described above, submaximal tone was induced with NE, and control responses to ACh, nitroglycerin, and LTD 4 were obtained prior to addition of an individual antagonist to the incubation medium. After obtaining the control responses, the ring preparations were washed with 60 ml buffer and allowed to reequilibrate for 30 minutes. Following the reequilibration period, the rings were studied in pairs. Thus, individual rings were exposed to either an inhibitor or its vehicle for minutes prior to induction of tone with NE. Effects of cumulative additions of ACh, nitroglycerin, and LTD 4 were then redetermined in the presence of either the antagonist or its vehicle. Influence of only one blocking agent on vasomotor activity was determined in an individual ring preparation. NE (/-norepinephrine hydrochloride; Sigma Chemical, St. Louis, Missouri; dose in terms of base) was dissolved and diluted in saline containing ascorbic acid (1 mg/mj). ACh (acetylcholine hydrochloride; Sigma; dose in terms of base) was dissolved and diluted in

3 Secrest and Chapnick Vasomotor Relaxation Induced by Leukotrienes 95 saline. LTD 4, LTC 4, and LTE 4 (Merck Frosst Canada, Pointe-Claire-Dorval, Quebec, Canada) were obtained as stock solutions in water. The stock solutions were divided into aliquots and stored under an atmosphere of nitrogen in a freezer (So-Low) at 0 C. For each experiment, an aliquot was thawed, diluted with saline to an appropriate concentration, and kept on ice. Meclofenamate (Parke-Davis, Morris Plains, New Jersey) was dissolved in 100 mm sodium carbonate. Nitroglycerin (Parke-Davis) and atropine (atropine sulfate; Sigma; dose in terms of base) were dissolved in saline. ETYA was prepared by dissolving the acid in absolute ethanol and then diluting with 1.00 mm sodium carbonate to a final concentration of 3 mg/ml of the acid in 10% ethanol and sodium carbonate. NDGA was diluted in 100 mm sodium carbonate to a final concentration of 0.3 mg/ml. All solutions, with the exception of ETYA, which required addition of 240 \i\, were prepared in concentrations such that only small volumes, no greater than 100 u.1, were added to the incubation chamber. All values were determined as peak change from control level. Relaxation responses were expressed as percent reduction of NE-induced tone. The results are reported as mean±sem. Values of n indicate the number of rings studied, each obtained from an individual animal. Statistical analyses were performed with Student's t test for paired or unpaired observations. 13 A p value of 0.05 or less was considered statistically significant. Results Influence of Leukotrienes on Vasomotor Tone Vasomotor activity of ACh and peptidoleukotrienes was observed solely after induction of active tone. Therefore, in all studies, the agonists were added to the incubation medium after the rings were precontracted to a stable plateau with NE (10" 7 to 3 x 10" 6 M) as described above. In unrubbed ring preparations, ACh (10~ 7 M) produced relaxations of 69.6 ±4.0% (n = 32) and 67.9 ±2.2% («= 33) in renal and superior mesenteric arteries, respectively. Relaxation induced by ACh was immediate in onset and reached a plateau within 2-4 minutes. Relaxation of the rings in response to LTC 4 (10~ to 3x 10~ 7 M) was somewhat slower in onset (15-20 seconds) than that produced by ACh and reached a plateau approximately 3 minutes later. Concentration-response curves relating relaxation produced by the leukotrienes in superior mesenteric and renal arterial rings as a function of NE-induced tone are shown in Figures 1 and 2, respectively. LTD 4 and LTC 4 (10~ to 3 x 10" 7 M) elicited concentration-dependent relaxations in both vessels. Relaxation of the rings in response to LTD 4 ranged from 13.2 ±2.% to 33.7±6.2% in the renal artery and from 6.6± 1.1% to 34.9 ±3.1% in the superior mesenteric artery. In comparison, LTC 4 produced decreases in tone ranging from 5.2 ± 1.7% to 32.9 ± 7.0% in the renal artery and 3.9 ±0.9% to 31. ±4.2% in the superior mesenteric artery. Although LTD 4 appeared to possess a greater 3 I e I 50-i LTD4 LTC, 'LTE. ^10) 10" 9 io~ 7 Molar Concentration FIGURE 1. Concentration-response curves comparing vasomotor effects of leukotrienes D, (LTDJ, C, (LTCJ, and E, (LTEJ on canine superior mesenteric arteries (endothelium intact) after induction of submaximal tone with norepinephrine (NE). Concentrations of NE used to induce tone were those required to obtain 40-75% of maximum contraction and ranged between 10~'and 3 x 10'" M. In five rings, relaxation was not produced in response to the lower concentrations (i.e., 10~' and 3x]0~'M LTEJ. All values expressed as mean ± SEM. Number adjacent to each symbol indicates number of rings studied, each obtained from an individual animal. 'Significantly different from UC/induced relaxation (p < 0.05). capacity to produce relaxation than LTC 4, statistically significant differences between responses were observed only at concentrations of 3 x 10 ~ M and of both 3 x 10" M and 10~ 7 M in the superior mesenteric and renal arteries, respectively. In contrast to both LTD 4 and LTC 4, LTE 4 had very little, if any, influence on the precontracted rings. The lower concentrations of LTE 4 (10~ and 3 x 10~ M) had no effect on either renal or superior mesenteric arterial rings, while concentrations of 10~ 7 and 3 x 10~ 7 M produced minimal decreases in tone. To determine whether the LTC 4 -induced relaxation described above was dependent on an intact functional endothelium, vasomotor influences of LTC 4, ACh, and nitroglycerin were determined in a separate group of renal and superior mesenteric arterial rings both before and after disruption of the intimal surface. In this group of tissues, tone induced in response to similar concentrations of NE before and after rubbing the intimal surface was not altered (Table 1). Prior to rubbing the intimal surface, addition of ACh (10~ 7 M) produced a fall in NE-induced tone of 6.9 ± 3.% in the superior mesenteric artery and 72. ±4.6% in the renal artery. LTC 4 (lo" 7 M) decreased tone by 2.7 ±4.9% in the superior mesenteric and 1.7 ± 2.9% in the renal artery. However, following perturbation of the intimal layer, vasomotor activity of both ACh and LTC 4 was completely abolished in both vessels. In contrast to these observations, vasomotor relaxation produced in response to nitroglycerin (10~ 6 M) was slightly, but not significantly, enhanced following endothelial disrup-

4 96 Circulation Research Vol 62, No 5, May ^ ^ ACh RA 75-i # 2 it") 1CT» 3x10^ 1CT 7 Motor Cootnitration FIGURE 2. Concentration-response curves comparing vasomotor effects of leukotrienesd, (LTDJ, C 4 (LTCJ, ande, (LTEJ on canine renal arteries (endothelium intact) after induction of submaximal tone with norepinephrine (NE). Concentrations of NE used to induce tone were those required to obtain 40-75% of maximum contraction and ranged between 10' 7 and 3x10'' M. In five rings, relaxation was not produced in response to the lower concentrations (L e., 10'' and 3 x 10~' M LTEJ. All values expressed as mean ± SEM. Number adjacent to each symbol indicates number of rings studied, each obtained from an individual animal. ^Significantly different from LTC 4 - induced relaxation (p < 0.05). tion. Thus, in the superior mesenteric artery, nitrogrycerin produced a fall in tone of 6.±4.1% and 90.6 ±3.6% before and after rubbing, respectively. In the renal artery, nitrogrycerin decreased tone 74.3 ± 10.2% before and 0.6 ± 6.0% after endothelial disruption. Effect of Atropine on LTD 4 -Induced Vasomotor Relaxation To determine whether the LTD 4 -induced relaxation was mediated via interaction with the muscarinic receptor, the influence of atropine, a muscarinic receptor blocking agent, on responses produced by ACh and LTD 4 was determined. Prior to incubation of the rings with atropine, ACh (10~ to 10~ 7 M) produced dose-dependent decreases in NE-induced tone (Figure 3). Atropine (10~ 7 M) was added to the bath 30 minutes TABLE 1. Effect of Inttmai Surface Rubbing on Contraction Produced by Norepinephrine Renal artery (n = 6) Concentration of norepinephrine (M) Contraction () + Endo (2.O±O.7)xlO ±1.0 -Endo (2.0±0.4)xl0" ±1.0 Superior mesenteric artery (n = ) + Endo (12.6±5.1)xl ±0.6 -Endo (12.±4.3)xlO Endo, endothelium not rubbed; - Endo, endothelium rubbed. Values represent mean ± SEM. n, number of animals. * * * f"'t r- VP FIGURE 3. Effect of atropine on relaxation of canine renal (RA) and superior mesenteric (SMA) arteries produced by acetylcholine (ACh). Following a 30-minute incubation period with either atropine (10' 1 M) or vehicle (0.9% NaCl), submaximal tone was induced with norepinephrine (NE), and responses to ACh were then obtained. Concentrations of NE ranged from 10~'to3x 10' 6 Mand were those required to obtain 40-75% of maximum contraction. All values expressed as mean ± SEM (n = 7). *Significantty different from vehicle after incubation with atropine (p<0.05). prior to reinduction of tone with NE. During incubation of the rings solely with atropine, no effect on baseline tone was observed. In addition, the antimuscarinic agent did not alter contractile responses to NE in either the superior mesenteric or renal arterial rings (Table 2). However, in the presence of atropine, ACh-induced relaxation was virtually abolished in both vessels (Figure 3). In contrast, relaxation induced by LTD 4 (10~ 7 M) (Figure 4) was unchanged following pretreatment with the muscarinic antagonist. TABLE 2. Effect of Atropine and Meclofenamate on Contraction Produced by Norepinephrine Force of norepinephrineinduced contraction Cg) Superior Renal artery mesenteric artery Atropine Meclofenamate ±1.5 I4.3m ±1.0* (M) ±1.0* Atropine, 10~ 7 M; meclofenamate, 10 5 M. Values represent mean ± SEM. Concentrations of norepinephrine used ranged between 10" 7 and 3x 10" 6 M and were those required to produce 40-75% of maximal contraction, n. number of animals. Significantly different from vehicle (p<0.05).

5 Secrest and Chapnick Vasomotor Relaxation Induced by Leukotrienes I i I I RA SMA 50-1 I I UTU l NWide FIGURE 4. Effect of atropine on relaxation of canine renal (RA) and superior mesenteric (SMA) arteries produced by leukotriene D, (LTDJ (]0~ 7 M). Following a 30-minute incubation period with either atropine (10~ 7 M) or vehicle (0.9% NaCl), submaximal tone was induced with norepinephrine (NE), and responses to LTD, were then obtained. Concentrations of NE ranged from 10''to 3 x 10~ 6 M and were those required to obtain 40-75% of maximum contraction. Values expressed as mean±sem (n = 7). Effect of Inhibitors of Cyclooxygenase Activity on LTD 4 -Induced Vasomotor Relaxation Superior mesenteric and renal arterial ring preparations with intact endothelium were incubated with meclofenamate (10~ 5 M) or its vehicle for 30 minutes. Submaximal contraction was then induced with NE, and effects of ACh and LTD 4 on vasomotor tone were determined. During incubation with meclofenamate, baseline tone tended to increase slightly, but the change was not significant. However, contractile responses to NE were significantly enhanced in the presence of the inhibitor of cyclooxygenase activity (Table 2). In contrast to its effect on these contractile responses, meclofenamate did not alter relaxation produced by LTD 4 in either renal or superior mesenteric arterial rings (Figure 5). In addition, relaxation induced by either ACh (10' to 10" 7 M) or nitroglycerin (10" 6 M) was unchanged following pretreatment with meclofenamate (Table 3). Effect of Inhibitors of Lipoxygenase Activity on LTD 4 -Induced Vasomotor Relaxation Influence of ETYA (3 x 10" 4 M) and NDGA (3 x 10~ 5 M), two chemically different lipoxygenase inhibitors, on LTD 4 -induced relaxation was evaluated to determine whether vasomotor responses produced by LTD 4 were related to the formation and/or release of lipoxygenase metabolites of arachidonic acid. In this series of experiments, unrubbed ring preparations of superior mesenteric artery were used. After a 60- minute incubation period with ETYA, NDGA, or their respective vehicles, tone was induced with NE as previously described, and vasomotor responses to ACh, LTD 4, or nitrogh/cerin were obtained. During incubation of the vessel rings in the presence of either inhibitor of lipoxygenase activity, basal tone was RA SMA FIGURE 5. Effect of meclofenamate (Meclo) on relaxation of canine superior mesenteric (SMA) and renal (RA) arterial rings produced by leukotriene D f (LTDJ (10~ 7 M). Following a 30-minute incubation period with either meclofenamate (10' s M) or vehicle (loomm NafiOJ, submaximal tone was induced with norepinephrine (NE), and responses to LTD, were then obtained. Concentrations ofne ranged from 10' 7 to 3 x 10~ 6 M and were those required to obtain 40-75% of maximum contraction. Values expressed as mean ± SEM (n = 7 for SMA; n = forra). unaltered. However, these substances differentially affected the vasomotor response of the rings to NE. In the presence of ETYA, the contractile response produced by NE was slightly enhanced (i.e., approximately 20%), while NDGA had no effect on NEinduced contraction (Table 4). In contrast to these divergent influences on contractile responses, addition of either ETYA or NDGA to the incubation medium resulted in attenuation of relaxation produced by both LTD 4 (Figures 6 and 7) and ACh (Table 5). Relaxation in response to nitroglycerin was enhanced in the presence of ETYA and, although not statistically significant, tended to be larger after addition of NDGA to the incubation fluid (Table 6). Discussion In the present study, it was observed that LTC 4 produced relaxation of precontracted renal and superior TABLE 3. Effect of Meclofenamate on Relaxation Produced by Acetylchollne and Nitroglycerin Percent decrease in norepinephrine-induced tone Renal artery Superior mesenteric (n = ) artery Meclo- Meclofenamate fenamate Acetylcholine (M) 10"«3X10" 10" ± ± ±7.7 5O.0± ± ±7.2 Nitroglycerin 10-6 (M) 69.4± ± ± ± Meclofenamate, 10" 3 M. Values represent mean ± SEM. n, number of animals. 25.7± ± ±.6 1.±6.0

6 9 Circulation Research Vol 62, No 5, May 19 TABLE 4. Effect of ETYA and NDGA on Norepinephrine- Induced Contraction in Isolated Canine Superior Mesenterlc Arteries Force of contraction (g) (n=16) 17.3±1.0 ETYA 20.6 ±0.* 17.6±0. NDGA ETYA, 3 x 10-4 M; NDGA, 3 x 10" 5 M. ETYA, 5,,11,14-eicosatetraynoic acid; NDGA, nordihydroguaiaretic acid. Values represent mean ± SEM. Concentrations of norepinephrine used ranged between 10 ~7 and 3x 10" 6 M and were thoserequiredto produce 40-75% of maximal contraction, n, number of animals. Significantly different from vehicle (p<0.05). mesenteric arterial rings in which the endothelial layer was intact, while LTE 4 produced little, if any, reduction in vasomotor tone. In contrast, after rubbing the luminal surface of the ring, neither relaxation nor contraction was produced in response to the leukotrienes. Effectiveness of endothelial removal in each ring was assessed pharmacologically. Thus, the influence of ACh and nitrogrycerin on vasomotor tone of both rubbed andunrubbed rings was determined. ACh produced relaxation of rings prior to, but not after, rubbing the intimal surface. In contrast, nitrogrycerin, a substance known to produce vasomotor relaxation in an endothelial-independent manner, 9 decreased tone in both rubbed and unrubbed ring preparations. These data strongly support the interpretation that rubbing the luminal surface with a cotton swab removed the endothelial cell layer, or at least markedly decreased its capacity to influence vasomotor tone. Results obtained xi< r» HT 7 LTD«<M) FIGURE 6. Effect of 5,,11,14-eicosatetraynoic acid (ETYA) on relaxation of canine superior mesenteric arterial rings produced by leukotriene D t (LTDJ. Following a 60-minute incubation period with either ETYA (3 x 10' 4 M) or vehicle (10% ethanol in 100 mm Na^OJ, submaximal tone was induced with norepinephrine (NE), and responses tq-ltd, were then obtained. Concentrations of NE ranged from W 7 to Jx 10' M and were those required to obtain 40-75% ofmaximum contraction. Values expressed as mean ± SEM (n = 7). *Significantfy different from vehicle after incubation with ETYA (p<0.05). I I # 50-, i 10" 7 3X10-7 LTD^M) FIGURE 7. Effect of nordihydroguaiaretic acid {NDGA) on relaxation of canine superior mesenteric arterial rings produced by leukotriene D, (LTDJ. Following a 60-minute incubation period with either NDGA (3 x 10'1 M) or vehicle (100 mm Na 2 COJ, submaximal tone was induced with norepinephrine (NE), and responses to LTD, were then obtained. Concentrations of NE ranged from 10' 7 to 3 x 10' M and were those required to obtain 40-75% of maximum contraction. Values expressed as mean + SEM (n = 9). *Significantly different from vehicle after incubation with NDGA (p < 0.05). in this study extend our previous observations in that these studies demonstrate LTC 4 also possesses the capacity to produce endothelial-dependent relaxation of canine superior mesenteric and renal arteries, while vasomotor activity of LTE< was only marginal. The endothelial-dependent relaxations produced by the leukotrienes in precontracted canine superior mesenteric and renal arterial rings were dependent on concentration, which ranged from 10~ to 3 x 10" 7 M. In addition, the data obtained in the present study suggest that the vasomotor activity of these products differed. Thus, LTD 4 appeared to be most active, while addition of LTE 4 to the incubation fluid induced TABLE 5. Effect of ETYA and NDGA on Relaxation Produced by Acetylcholine in Isolated Canine Superior Mesenteric Arteries Acetylcholine 10"' 3x10"' 10-3x10- io- 7 Percent decrease in norepinephrine-induced tone (n==10) 10)14) ETYA NDGA (M) 17.1 ± ± ± ± ±6.2.1 ±2.5* 20. ±2.6* 2.+2.* * 40.3±3.7* * * 30.± ±2.7* 39.± * 43.1 ± ±4.1* ETYA, 3X 10" 4 M; NDGA, 3x 10" 3 M. ETYA, 5,,11,14-eicosatetraynoic acid; NDGA, nordihydroguaiaretic acid. Concentrations of norepinephrine ranged from 10 ~7 to 3 x IO- 6 M and were those required to produce 40-75% of maximal contraction. Values represent mean ± SEM. n, number of animals. Significantly different from vehicle (p<0.05).

7 Secrest and Chapnick Vasomotor Relaxation Induced by Leukotrienes 99 TABLE 6. Effect of ETYA and NDGA on Relaxation Produced by Nitroglycerin in Isolated Canine Superior Mesenteric Arteries Percent decrease in norepinephrine-induced tone produced by nitroglycerin (M) io-«3x10- io- 7 3xlO" 7 lo" ±6.7 ETYA 71 NDGA n * 7 7.5± ± ± ±3.* ± ±5.3* 39.0± ± ± ±5.7* ±6.3 6.± ±5.5* 69.0± ±5. ETYA, 3 x lo" 4 M; NDGA, 3 x 10" 5 M. ETYA, 5,,11,14-eicosatetraynoic acid; NDGA, nordihydroguaiaretic acid. Concentrations of norepinephnne ranged between 10~ 7 and 3 x 10" 6 M and were those required to produce 40-75% of maximal contraction. Values represent mean±sem. n, number of animals. Significantly different from vehicle (p<0.05). minimal changes in tone that were observed only at the higher concentrations studied (10" 7 and 3 x 10" 7 M). The relaxant capacity of LTC 4 appeared to be intermediate between that of LTD 4 and LTE 4. These differences in activity were most apparent in the renal artery. Although it must be realized that the present study was not designed to reflect differences in potency between the individual leukotrienes, these observations are in agreement with the hypothesis suggested by Bernstrom and Hammarstrom 14 that LTC 4 may serve as an active precursor of LTD 4, while conversion of LTD 4 to LTE 4 results in substantial loss of biological activity. Results obtained in other studies under both in vivo and in vitro conditions also are compatible with this hypothesis. For example, in the anesthetized dbg and cat, LTD 4 was more active than LTC 4 in producing mesenteric vasoconstriction, and both LTD 4 and LTC 4 were considerably more active than LTE 4.'- 3 Additionairy, both LTD., and LTC 4 produced equivalent contractile responses in the distal pulmonary artery of the guinea pig, while LTE 4 was much less potent. 12 Furthermore, when the capacities of LTD 4 and LTC 4 to produce contraction of helical strips of rabbit aorta were compared, LTD 4 was more potent than LTC 4. 6 Thus, while the present observations were not correlated with these previously reported hemodynamic or vasomotor responses, the rank order of ability to produce relaxation was consistent with that found in earlier studies. Although most studies have shown that leukotrienes produce contraction of both vascular and nonvascular smooth muscle, 1516 recent observations suggest these substances also have the capacity to relax vascular smooth muscle. Both LTC 4 and LTD 4 have been reported to relax spirally cut segments of dog coronary artery incubated in the presence of 27 mm K. 7 Additionally, in the anesthetized pig, an unidentified coronary vasodilator substance was generated during continuous intracoronary infusion of LTD 4. 5 Because no attempt was made to determine the presence or functional integrity of the endothelium, it is unknown whether either of these responses was dependent on an intact intimal layer. However, the observations of Burke et al 7 and Ezra et al 5 are consistent with our findings. Results of our initial study suggested that indomethacin, an inhibitor of cyclooxygenase activity, had no effect on endothelial-dependent relaxation produced by LTD 4 and ACh. To investigate further the possibility that a vasodilator prostaglandin, such as prostacyclin, may subserve LTD 4 -induced relaxation, influence of another chemically different inhibitor of cyclooxygenase activity, meclofenamate, on LTD 4 -induced decreases in vasomotor tone was determined. In agreement with the previous observation, relaxation of either renal or superior mesenteric arterial rings in response to ACh or LTD 4 was not altered in the presence of meclofenamate. Thus, although LTC 4 and LTD 4 possess the capacity to release cyclooxygenase products, of which prostacyclin is the major component, from human endothelial cells, 17 it does not appear that the reduction in tone produced by LTD 4 is dependent on release of a cyclooxygenase-derived product of arachidonic acid metabolism. While relaxation of the rings in response to LTD 4 was not altered in the presence of meclofenamate, NE-induced contraction was enhanced. Similarly, indomethacin also enhanced the contractile response to NE. These observations are consistent with the hypothesis that released vasodilator prostaglandins serve as modulators of vasoconstrictor hormones. 1 Both ACh, the prototypical endothelial-dependent relaxing agent, and the peptidoleukotrienes produced relaxation solely in vessels in which the intima was unperturbed; however, significant differences between these responses were observed. Relaxation produced by LTD 4 was not altered in the presence of atropine, a muscarinic receptor antagonist. In contrast, AChinduced relaxation was markedly inhibited following incubation of the rings with atropine. This observation is in agreement with results of other studies in which endothelial-dependent relaxation produced by ACh was blocked in the presence of atropine.' The lack of effect of atropine on LTD 4 -induced relaxation suggests that LTD 4 does not interact with the muscarinic receptor to produce vasomotor relaxation. In our previous study, it was found that FPL55712, a putative leukotriene antagonist," attenuated LTD 4 -induced relaxation but had no effect on that produced by ACh. When taken

8 990 Circulation Research Vol 62, No 5, May 19 together, these results suggest that the two endothelialdependent relaxing substances act through different receptors to initiate vascular smooth muscle relaxation. Whether the secondary mechanism(s) and/ or mediators) involved in the relaxation process are similar for the two agents cannot be determined from these studies. Relaxation produced by LTD 4 and ACh, but not nitrogtycerin, was attenuated after incubation with either ETYA, an inhibitor of both cyclooxygenase and lipoxygenase activities, or NDGA, a lipoxygenase inhibitor. These observations are consistent with previous studies demonstrating that both ETYA and NDGA inhibit endothelial-dependent relaxation When considered together, these results and the lack of effect of blockers of cyclooxygenase activity on either leukotriene- or ACh-induced changes in vasomotor tone support the hypothesis that lipoxygenase metabolites participate in the process of endothelialdependent relaxation. While the identity of such a metabolite remains neither known nor isolated, such a substance should possess the capacity to produce vasomotor relaxation after removal of the endothelial layer. Although the present observations indicate that endothelial-dependent relaxation is not mediated by these peptidoleukotrienes, they do not preclude a role for some other lipoxygenase-derived product(s) in mediation of endothelial-dependent relaxation. Indeed, Forstermann and Neufang 11 and Berkowitz et al 12 arrived at the same conclusion when they found that peptidoleukotrienes failed to relax precontracted rabbit aorta. In contrast to the effects of ETYA and NDGA on decreases in vasomotor tone elicited by LTD 4 and ACh, relaxation produced in response to nitrogrycerin, an endothelial-independent relaxing agent, was enhanced slightly in the presence of ETYA and, although not statistically significant, tended to be larger during incubation of the arterial rings with NDGA. Consistent with these observations, it has recently been found that either removal of the endothelium 21 or inhibition of lipoxygenase activity 22 augments vasomotor relaxation produced by nitrovasodilators. Although the effect of NDGA on nitroglycerin-induced responses was not significant, these data appear to be compatible with the hypothesis proposed by Pohl and Busse 22 that EDRF acts as an inhibitor of relaxation produced in response to nitrovasodilators. Alternatively, because ETYA also possesses the capacity to inhibit cyclooxygenase activity, these data may be interpreted to suggest that a cyclooxygenase product(s) and not EDRF may be responsible for inhibition or attenuation of nitroglycerininduced relaxation. If this is the case, inhibitors of cyclooxygenase activity, such as indomethacin or meclofenamate, should possess the capacity to enhance relaxation produced in response to nitrogrycerin. Results obtained in both this and our previous investigations suggested that neither indomethacin nor meclofenamate significantly enhanced relaxation of superior mesenteric or renal arterial rings produced in response to a single concentration (10~ 6 M) of nitrogrycerin. It, therefore, did not appear that a cyclooxygenase product of arachidonic acid was a participant in the relaxation. However, these.studies were not specifically designed to answer this question, and, clearly, further studies to evaluate more completely the hypothesis of Pohl and Busse 22 are required. While ETYA and NDGA had similar inhibitory effects on relaxation produced by LTD 4 and ACh, their influence on NE-induced contraction differed. ETYA enhanced NE-induced contraction, while NDGA had no effect on the contractile response. These findings were most likely due to a reduction in cyclooxygenase activity produced by ETYA and were consistent with results obtained in the presence of the cyclooxygenase inhibitors, meclofenamate (described above) and indomethacin. Indeed, it may be possible that the slightly enhanced superior mesenteric and renal arterial relaxation produced by nitrogrycerin in the presence of ETYA was a consequence of enhanced NE-induced tone. In addition to serving as a substrate for cyclooxygenase and lipoxygenase enzymatic systems, arachidonic acid may be oxidized via cytochrome P-450 monooxygenases. 23 " 25 Indeed, the presence of a cytochrome P-450-dependent mixed-function oxidase has been described in vascular tissue, including the endothelium Although the identities and biological activities or roles of arachidonic acid metabolites arising from cytochrome P-450-dependent monooxygenases are virtually unknown, initial studies suggest that these products may have the ability to alter vasomotor tone. 2 Evidence supporting the hypothesis that products of the cytochrome P-450 system may participate in endothelial-dependent relaxation has been obtained in rabbit aorta 29 and pulmonary artery. 30 In these studies, endothelial-dependent relaxation induced by methacholine, A2317, or arachidonic acid was attenuated in the presence of either metyrapone or SKF 5A, two inhibitors of cytochrome P-450. In addition, in vivo induction or depletion of vascular cytochrome P-450-dependent enzymatic activity enhanced or attenuated, respectively, endothelial-dependent vasomotor relaxation in response to arachidonic acid. 30 At high concentrations, ETYA has been shown to inhibit cytochrome P-450-mediated arachidonic acid metabolism in the renal cortex. 24 It has also been suggested that high concentrations of both ETYA and NDGA may inhibit arachidonic acid metabolism by cytochrome P-450 in anterior pituitary eel Is. 31 Thus, the inhibitory effects of ETYA and NDGA could conceivably be mediated, at least in part, by actions on cytochrome P-450. In summary, our observations indicate that relaxation produced by the leukotrienes in isolated canine superior mesenteric and renal arteries is dependent on an intact endothelium. However, it must be realized that actual release of an endogenous relaxant factor from endothelial cells in response to the leukotrienes has not been demonstrated in these studies. Indeed, if the leukotrienes do release a vasodilator from endothelial cells, it is unknown whether this substance and the

9 Secrest and Chapnick Vasomotor Relaxation Induced by Leukotrienes 991 mechanism of relaxation are the same as that described for acetylcholine. Although the vasomotor effects of leukotrienes described in the present study do not appear to be consistent with results obtained in intact anesthetized animals, it must be emphasized that the present study was performed on relatively large arteries and may not reflect influences of these substances on small arteries and veins as well as on resistance vessels. Acknowledgments The secretarial help of Linda Russell is gratefully acknowledged. The gifts of leukotrienes C 4, D 4, and E 4 (provided by Dr. Joshua Rokach of Merck Frosst) and 5,,11,14-eicosatetraynoic acid (provided by Dr. James Hamilton of Hoffmann-LaRoche, Inc, Nutley, New Jersey) are greatly appreciated. References 1. Feigen LP: Differential effects of leukotriene C 4, D 4) and E 4 in the canine renal and mesenteric vascular beds. J Pharmacol Exp Ther 193;225: Chapnick BM: Divergent influences of leukotrienes C 4, D 4, and E 4 on mesenteric and renal blood flow. Am J Physiol 194; 246:H51-H Lippton HL, Armstead WM, Hyman AL, Kadowitz PJ: Vasoconstrictor effects of leukotrienes C, and D 4 in the feline mesenteric vascular bed. Prvstagtandins 194;27: Panzenbeck MJ, KaleyG: Leukotriene D 4 reduces coronary blood flow in the anesthetized dog. Prostaglandins 193;25: Ezra D, Feuerstein G, Czaja JF, Laurindo FRM, Finton CK, Yeh GC, Goldstein RE: Unique coronary vasodilator induction by leukotriene D 4. Am J Physiol 195;249:H69-H Kito G, Okuda H, Ohkawa S, Terao S, Kikuchi K: Contractile activities of leukotrienes C 4 and D 4 on vascular strips from rabbits. Life Sci 191;29: Burke JA, Levi R, Guo Z, Corey FJ: Leukotrienes C 4, D 4, and E 4 : Effects on human and guinea-pig cardiac preparations in vitro. / Pharmacol Exp Ther 192;221: Secrest RJ, Olsen EJ, Chapnick BM: Leukotriene D 4 relaxes canine renal and superior mesenteric arteries. Ore Res 195; 57: Furchgott RF, Zawadzki JV: The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature 190;2: Chand N, Altura BM: Inhibition of endothelial cell-dependent relaxations to acetylcholine and bradykinin by lipoxygenase inhibitors in canine isolated renal arteries. Microciradation 191:1: Forstermann U, Neufang B: C-6-sulfidopeptide leukotrienes are unlikely to be involved in the endothelium dependent relaxation of rabbit aorta by acetylcholine. Prostaglandins 194; 27: Berkowitz BA, Zabko-Potapovich B, Valocik R, Gleason JR: Effects of leukotrienes on the vasculature and blood pressure of different species. / Pharmacol Exp Ther 194;229: Daniel WW: Biostatistics: A Foundation for Analysis in the Health Sciences, ed 2. New York, John Wiley & Sons, Inc, Bernstrom K, Hammarstrom S: Metabolism of leukotriene D by porcine kidney. / Biol Chem 191;256: Piper PJ: Pharmacology of leukotrienes. Br Med Bull 193; 39: Feuerstein G: Leukotrienes and the cardiovascular system. Prostaglandins 194;27: Cramer ED, Pologe L, Pawlowski NA, Conn ZA, Scott WA: Leukotriene C promotes prostacyclin synthesis by human endothelial cells. Immunology 193;0: Brody MJ, Kadowitz PJ: Prostaglandins as modulators of the autonomic nervous system. Fed Proc 1974;33: Augstein J, Fanner JB, Lee TB, Sheard P, Tattersall ML: Selective inhibitor of slow-reacting substance of anaphylaxis. Nature 1973;245: Singer HA, Peach MJ: Endothelium-dependent relaxation of rabbit aorta II: Inhibition of relaxation stimulated by methacholine and A2317 with antagonists of arachidonic acid metabolism. J Pharmacol Exp Ther 193;226: Shirasakj Y, Su C: Endothelium removal augments vasodilation by sodium nitroprusside and sodium nitrite. Eur J Pharmacol 195;114: Pohl U, Busse R: Endothelium-derived relaxing factor inhibits effects of nitrocompounds in isolated arteries. Am J Physiol 197;252:H307-H Capdevila J, Chacos N, Werringloer J, Prough RA, Estabrook RV: Liver microsomal cytochrome P-450 and the oxidative metabolism of arachidonic acid. Proc NadAcad Sci USA 191; 7: Morrison AR, Pascoe N: Metabolism of arachidonate through NADPH-dependent oxygenase of renal cortex. Proc Natl Acad Sci USA 191;7: Schwartzman ML, Abraham NG, Carroll MA, Levere RD, McGiff JC: Regulation of arachidonic acid metabolism by cytochrome P-450 in rabbit kidney. Biochem J 196;23: Juchau MR, Bond JA, Benditt EP: Aryl 4-monooxygenase and cytochrome P-450 in the aorta: Possible role in atherosclerosis. Proc Natl Acad Sci USA 1976;73: Abraham NG, Pinto A, Mullane KM, Levere RD, Spokas E: Presence of cytochrome P^t50-dependent monooxygenase in intimal cells of the hog aorta. Hypertension 195;7: Schwartzman M, Ferreri NR, Carroll MA, Songu-Mize E, McGiff JC: Renal cytochrome P-450 related arachidonate metabolite inhibits (Na + +K + )ATPase. Nature 195;314: Singer HA, Saye JA, Peach MJ: Effects of cytochrome P-450 inhibitors on endothelium-dependent relaxations in rabbit aorta. Blood Vessels 194;21: Pinto A, Abraham NG, Mullane KM: Cytochrome P-450 dependent monooxygenase activity and endothelial-dependent relaxations induced by arachidonic acid. J Pharmacol Exp Ther 196;236:445^ SnyderGD, Capdevila J, Chacos N, Manna S,FalckJR: Action of luteinizing hormone-releasing hormone: Involvement of novel arachidonic acid metabolites. Proc Natl Acad Sci USA 193;0: KEY WORDS canine mesenteric artery leukotrienes canine renal artery endothelial-dependent relaxation