Saphenous Vein. um.3-9 In experimental animals, nitric oxide accounts. at least in part for the biological activity of endothelium-derived

Transcription

1 52 Different Ativation of the Endothelial L-Arginine and Cylooxygenase Pathway in the Human Internal Mammary Artery and Saphenous Vein Zhihong Yang, Ludwig von Segesser, Erwin Bauer, Peter Stulz, Marko Turina, and Thomas F. Luisher Downloaded from by guest on April 7, 218 The endothelium releases substanes ontrolling vasular tone and platelet funtion. We investigated mediators of endothelium-dependent responses in human internal mammary arteries and saphenous veins. The inhibitor of nitri oxide formation, NG-monomethyl L-arginine, enhaned the sensitivity to norepinephrine (fivefold) and evoked more pronouned endothelium-dependent ontrations in internal mammary arteries (19+6% of 1 mm KCI) than in saphenous veins (2±1%; p<.5). In internal mammary arteries, NG-monomethyl L-arginine, but not indomethain, markedly redued endothelium-dependent relaxations to aetylholine (from 95±2% to 39+7%; p<.5) and prevented those to histamine (78±6% to 4+3%; p <.5). In saphenous veins, endothelium-dependent relaxations to aetylholine were weak (24+11%), while nitri oxide aused omparable relaxations (85+±3%) as in internal mammary arteries (8±5%; NS). NG-Monomethyl L-arginine prevented the relaxations to aetylholine and unmasked endothelium-dependent ontrations (3±1%o). Indomethain and the thromboxane synthetase inhibitor CGS-138 augmented relaxations of saphenous veins to aetylholine from 24±11% to 46±91% (p<.5). Histamine-evoked ontrations were onverted to endothelium-dependent relaxations by indomethain and the thromboxane A2/endoperoxide reeptor antagonist SQ-3741 (38+±3% and 4±+6%; p<.5) but not CGS Thus, 1) nitri oxide mediates endothelium-dependent relaxations in human arteries and veins; 2) internal mammary arteries release more nitri oxide than do saphenous veins, and 3) in saphenous veins, the effets of nitri oxide are redued by endothelium-derived ontrating fators originating from the ylooxygenase pathway. (Cirulation Researh 1991; 68:52-6) T he endothelium is a soure of several vasoative substanes that an affet vasular tone and platelet funtion.12 Aetylholine, bradykinin, histamine, and other agents an release endothelium-derived relaxing fators from endothelial ells in ulture and from intat arteries with endotheli- From the Department of Researh (Z.Y., T.F.L.), Laboratory of Vasular Researh; Department of Mediine (T.F.L.), Division of Cardiology; and Department of Thorai and Cardia Surgery (P.S.), University Hospital, Basel; and Department of Cardiovasular Surgery (L.v.S., E.B., M.T.), University Hospital, Zurih, Switzerland. Supported by grants from the Swiss National Researh Foundation (No ), the Swiss Cardiology Foundation, and the Shweizerishe Rentenanstalt. T.F.L. is a reipient of a areer development award of the Swiss National Researh Foundation (SCORE grant No ). Address for orrespondene: Thomas F. Lusher, MD, Division of Cardiology, University Hospital, CH-431 Basel, Switzerland. Reeived May 1, 199; aepted August 9, 199. um.3-9 In experimental animals, nitri oxide aounts at least in part for the biologial ativity of endothelium-derived relaxing fators released by aetylholine and bradykinin.1,11 In endothelial ells, nitri oxide is leaved from its preursor, the amino aid L-arginine, by speifi enzymes.'2-14 By means of the methylated amino aid NG-monomethyl L-arginine (LNMMA), this pathway for the formation of nitri oxide an be inhibited.15-'7 In addition to nitri oxide, the endothelium an produe other relaxing fators suh as prostaylin1819 and yet unidentified substanes suh as endothelium-derived hyperpolarizing fator.2-24 Furthermore, at least in blood vessels of dogs and rats, the endothelium an mediate ontrations While the endothelium-derived ontrating fator released during hypoxia is resistant to most pharmaologial interventions,26 the ontrations to arahidoni aid and aetylholine an be prevented by inhibitors of ylooxygenase The ylooxygenase

2 Yang et al Endothelium-Derived Nitri Oxide in Human Bypass Vessels 53 Downloaded from by guest on April 7, 218 pathway an produe a variety of substanes suh as thromboxane A232 and other prostaglandinsl2'8 as well as superoxide anions33 that evoke ontrations in isolated blood vessels. Endothelium-dependent relaxations differ in arteries and veins.4,9,25 This differene ould be related to the release of different amounts of endotheliumderived nitri oxide or other endothelium-derived relaxing fators,4'34'35 a different vasular responsiveness to the relaxing fators, or a different release of endothelium-derived ontrating fators in the two blood vessels. This ould have important onsequenes for the regulation of blood flow and that of platelet funtion1,2,36-38 in arteries and veins. In addition, this may ontribute to funtional differenes of human arteries and veins when they are used as bypass grafts to treat patients with oronary artery disease.',4,39,4 Thus, the present experiments were designed to haraterize endothelium-dependent relaxations and ontrations in the human internal mammary artery and saphenous vein and to delineate the mediators involved in these responses. Materials and Methods Preparation of Blood Vessels Internal mammary arteries and saphenous veins were obtained from patients undergoing oronary bypass surgery for oronary artery disease. Care was taken during harvesting of the blood vessels not to touh the inner surfae of the blood vessels. During surgial preparation of the saphenous vein, the dilatation proedure was avoided. Veins stained with methylene blue were disarded. Immediately after removal, the vessels were plaed into modified Krebs-Ringer biarbonate solution of the following omposition (millimolar): NaCI 118, KCl 4.7, CaCl2 2.5, MgSO4 1.2, KH2PO4 1.2, NaHCO3 25., edetate alium disodium.26, and gluose 11.1 (ontrol solution). The blood vessels were disseted free, ut into rings of about 5 mm in length, and suspended in organ hambers filled with ontrol solution (37 C, aerated with a gas mixture ontaining 95% 2 and 5% C2). Experimental Setup The blood vessel rings were suspended in organ hambers filled with ontrol solution, and hanges in isometri fore were reorded with fore transduers (Statham Universal UC2, Gould, In., Cleveland, Ohio, or Showa Sokki, Rikadenki GmbH, Freiburg, FRG). Before the experiment, the vasular rings were strethed and exposed to norepinephrine (1-` or 3 x 1- M in veins and arteries, respetively) at eah level of streth, until the optimal point of the length-tension relation was reahed. Then the vessels were allowed to equilibrate for 45 minutes. Endothelium Removal The presene or absene of the endothelium was onfirmed in eah ring (ontrated with norepinephrine 1` to 3 x 1`7 M) at the beginning of the experiment by a relaxation to aetylholine or bradykinin (1-6 M), respetively. For ertain experiments (n=8), the endothelium was deliberately removed from the vasular rings by gentle rubbing of the intraluminal surfae of the blood vessels with a otton swab. Drugs The following drugs were used (unless otherwise indiated the drugs were obtained from Sigma Chemial Co., St. Louis): aetylholine hydrohloride, bradykinin, CGS-1384` (Ciba-Geigy Ltd., Basel, Switzerland), imetidine (Smith, Kline and Frenh Ltd., Welwyn Garden, Herts, UK), histamine, indomethain, LNMMA14-17 (Calbiohem, Luerne, Switzerland), l-norepinephrine bitartrate, SQ (Squibb Institute for Medial Researh, Prineton, N.J.), and superoxide dismutase. The onentrations of the drugs are expressed as final molar onentrations in the bath solution. Indomethain was dissolved in 1- M Na2CO3. CGS-138 (1-4 M) was dissolved in dimethyl sulfoxide and then further diluted in distilled water. LNMMA (1-3 M) was dissolved in methanol and diluted in distilled water. SQ-3741 (1-` M) was dissolved and diluted in ethanol. The final onentration of the solvents in the organ bath was less than.6%. Nitri Oxide Solution Nitri oxide gas (99.9% pure) was obtained from L'Air Liquide Belge, Antewerp, Belgium. A glass bulb (125 ml in volume) was flushed with helium to remove oxygen. The flask was then filled with nitri oxide gas and losed. Separate glass bulbs were filled with 1 ml double-distilled water deoxygenated with helium over 4 hours.1 Appropriate volumes of nitri oxide gas were then injeted into the water-filled gas bulbs, and the onentration of nitri oxide in the fluid was alulated taking into aount its solubility in water and air. All gas bulbs were interonneted in a losed gastight system to prevent any entry of oxygen (Arbor Ventil and Fitting AG, Niederrohrdorf, Switzerland). Nitri oxide was injeted with gastight syringes (Hamilton, Kontron AG, Zurih) into the organ hamber near the suspended vessels. Protools When onentration-response urves were onstruted, the drugs were added in a umulative fashion exept for nitri oxide (beause of rapid breakdown and evaporation of the gas in the aerated organ hambers). Contrations are expressed as perent of the inrease in tension indued by 1 mm potassium hloride. The onentration of an agonist exhibiting 5% of the ontration to potassium hloride (EC5 value) was alulated for eah ring separately and expressed as negative log molar (pd2). To study endothelium-dependent relaxations, the rings were ontrated with norepinephrine (3x 1-7 M in arteries and 1-7 M in veins). The relaxations are expressed as perent derease in tension of that

3 Downloaded from by guest on April 7, Cirulation Researh Vol 68, No 1, January 1991 C 2. QCL 15., 2 o E ) 1. _ 5-.o_' 5. O. C C.-.- IMA sv L-NMMA (1 4M) FIGURE 1. Inrease in tension aused by the inhibitor of nitr oxide formation, NG-monomethyl L-arginine (L-NMMA, 1-4 M in the presene of indomethain 1`5 M) in the human intemal mammary artery (IMA) and saphenous vein (SV) with endothelium. The ontrations are expressed asperent of the inrease in tension obtained with 1 mm KCl. *Statistially signifiant differene between the internal mammary artery and saphenous vein (p<o.5). ontration indued by norepinephrine. For analysis of the relaxations, the maximal relaxation, the negative log molar onentration of a given vasodilator exhibiting 5% relaxation (pd2 value), was used. In ertain experiments, the vessels were preinubated with drugs that interfere with the release of endothelium-derived vasoative substanes (i.e., LNMMA, indomethain, and CGS-138 for 3 minutes; SQ-3741 and superoxide dismutase for 15 minutes). Data are given as mean±sem. In all experiments, n equals the number of patients from whom the blood vessels were obtained. Whenever possible, onentration-response urves of the ^ _ U E ~ Q C _ -, od CL- C) 12 1_ * Control n=4 o L-NMMA n=4 n= tested agonists were performed in the presene and absene of the inhibitors in the same ring or in the rings obtained from the same patient. Beause LNMMA by itself aused an inrease in tension, the onentrations of agonists used to preontrat the vessels were redued to math the ontrations obtained under ontrol onditions. The t test for paired or unpaired observations and analysis of variane (ANOVA) followed by Sheffe's test were used for statistial analysis. A two-tailed value of p<.5 was onsidered to indiate a statistial differene. Results Basally Released Endothelium-Derived Nitri Oxide Endothelium-dependent ontrations to LNMMA. In quiesent internal mammary arteries with, but not in those without, endothelium, LNMMA (1-4 M in the presene of indomethain 1` M) evoked a ontration averaging 19±6% of the response to 1 mm KCl (Figure 1; p<.1 as ompared with rings without endothelium, n =4). In ontrast, in the saphenous vein with endothelium, LNMMA (1-4 M in the presene of indomethain 1- M) aused only a small ontration (2+1%; Figure 1; p<.5 as ompared with the artery, n =4). Contrations to norepinephrine. Norepinephrine (1-9 to 1-5 M) aused onentration-dependent ontrations in internal mammary artery. In vasular rings with endothelium, LNMMA (1-4 M) enhaned the sensitivity to norepinephrine (Figure 2, left panel; pd2 ontrol, ; pd2 LNMMA, 7.±.2; onentration shift, fivefold; p<.5). In artery rings without endothelium, the methylated L-arginine analogue was without any effet (Figure 2, right panel; n=4). a = a E _ r - ~ Ci * Control n=4 o L-NMMA n3= Norepinephrine (-iogm) Norepinephrlne (-logm) FIGURE 2. Contrations indued by inreasing onentrations of norepinephrine in internal mammary arteries with endothelium (left panel) or without endothelium (right panel). The inhibitor of nitri oxide formation, NG-monomethyl L-arginine (L-NMMA, 1` M), augmented the ontrations in preparations with endothelium (left panel; onentration shift, fivefold, p<o.5), while it had no effet in arteries without endothelium (right panel).

4 Yang et al Endothelium-Derived Nitri Oxide in Human Bypass Vessels 55 Downloaded from by guest on April 7, 218 o Q. =a OE C C) * Control n=4 n L-NMMA n= Norepinephnne (-logm) FIGURE 3. Contrations indued by norepinephrine in human saphenous veins with endothelium under ontrol onditions and in the presene of the inhibitor of nitri oxide formation, NG-monomethyl L-arginine (L-NMMA, 1-4 M). Note that L-NMMA does not affet the ontrations to norepinephrine in the human saphenous vein. Norepinephrine (1-` to 1` M) aused ontrations in saphenous veins with endothelium (Figure 3; pd2 6.3±.1, n=4). LNMMA (1-4 M in the presene of indomethain 1` M) did not augment the ontrations to norepinephrine (Figure 3; n=4). Stimulated Release ofendothelium-denved Substanes Aetylholine. In arteries with endothelium, aetylholine (1-` to 1` M) aused potent relaxations (Figure 4; n=8). The maximal response averaged 95+2%, and the pd2 value was 7.7±.1. Indomethain (1-5 M), used to inhibit the prodution of prostaylin, did not affet the sensitivity or the maximal response to the musarini agonist. LNMMA (1-4 M) markedly redued the sensitivity and maximal response (39±7%) to aetylholine (Figure 4; p<.5 as ompared with ontrol). If indomethain and LNMMA were added together, a further redution of the response to 23±8% was ahieved (Figure 4; p<.5 as ompared with LNMMA alone, n=5). In saphenous veins, aetylholine (1-9 to 1-4 M) aused only weak endothelium-dependent relaxations (Figure 5, left panel; n=5). The maximal response averaged 24±11% (p<.5 as ompared with rings without endothelium, n=4). The relaxation obtained was signifiantly less than in the mammary artery (p<.5). LNMMA (1-' M) prevented the relaxations to aetylholine in the saphenous vein and unmasked endothelium-dependent ontrations (Figure 6). The maximal ontration in LNMMA-treated rings with endothelium averaged 3±1% (Figure 5, left panel; p<.5 as ompared with rings without endothelium, n=4). In ontrast to the mammary artery, indomethain (1-5 M) augmented endothelium-dependent relaxations to aetylholine from 24±11% to 46±9% in the saphenous vein (Figure 5, left panel; p<.5 as ompared with ontrol, n =5). The thromboxane synthetase inhibitor CGS-138 (1-5 M) augmented the relaxations to aetylholine to a omparable degree as indomethain from 24+8% to 46±5% (Figure 5, right panel; p<.5, n =8). Histamine. Histamine (1-9 to 1-6 M in the presene of imetidine 1-5 M and indomethain 1-5 M) aused endothelium-dependent relaxations with a maximal response of 78+±6% in internal mammary artery (Figure 7; pd2 7.±.2, n=4). LNMMA (1-4 M) ompletely prevented the relaxations indued by the monoamine (maximal response, 4+3%; p<.5 as ompared with ontrol). In rings of saphenous vein, histamine (1-9 to 1-6 M) did not ause signifiant endothelium-dependent relaxations under ontrol onditions (i.e., in the absene of any inhibitor; maximal response, 9+ 5%) but evoked marked ontrations at higher onentrations of the monoamine (3 x 1-7 to 1x 1-6 M; maximal response, 72±16%; Figure 8, left panel; n=4). Endothelium-dependent relaxations were unmasked in the presene of both indomethain (1-5 M) and the thromboxane A2/endoperoxide reeptor antagonist SQ-3741 (3 x 1-6 M; Figure 8, right panel). The maximal relaxations obtained in the presene of indomethain and SQ-3741 averaged -.2. C z g ~ a * a l l Aetylholine (-logm) L-NMMA plus Indomethain n=5 L-NMMA n=4 Indomethain n=6 Control n=8 FIGURE 4. Endothelium-dependent relaxations to aetylholine in the human intemal mammary artery. Under ontrol onditions, the musarini agonist evokes potent relaxations that are unaffeted by the inhibitor of ylooxygenase, indomethain (1-` M). The inhibitor of nitri oxide formation, NG-monomethyl L-arginine (L-NMMA, i` M), signifiantly redues the relaxations indued by the musarini agonist (p<o.oos). In the presene of L-NMMA, indomethain further redues the relaxations indued by aetylholine (p<o.o5).

5 56 Cirulation Researh Vol 68, No 1, January 1991,_ CD - eb CD C ( CL._ z * Control n=5 o Indomethain n=5 * L-NMMA n=4 o Without endothellum n= C C) -. 4) q-. CL.a z " * Control n=8 O CGS138 n= g g * * g w w Downloaded from by guest on April 7, 218 Aetylholine (-iogm) Aetyiholine (-logm) FIGURE 5. Endothelium-dependent effets of aetlholine in the human saphenous vein. Left panel: Under ontrol onditions, aetylholine evokes only weak endothelium-dependent relaxations. The inhibitor of nitri oxide fonration, NG-monomethyl L-arginine (L-NMMA, 1C4M), unmasks endothelium-dependent ontrations to the musarini agonist. In ontrast, the inhibitor of ylooxygenase, indomethain (1 5 M), signifiantly augments endothelium-dependent relaxations to aetylholine (p<.5). Right panel: The inhibitor ofthromboxane A2 synthetase, CGS-138 (1-5 M), augments the relaxations indued by aetylholine to a similar degree as indomethain (p<.5 as ompared with ontrol). 38+3% and 4±t6%, respetively (p<.5 as ompared with ontrol, n=4 to 5). In ontrast, the thromboxane synthetase inhibitor CGS-138 (1-5 M) did not signifiantly affet the response. The endothelium-dependent relaxations to histamine that ourred in the presene of indomethain were abolished by LNMMA (1-4 M; Figure 8, left panel; maximal response, 1+2%; p<.5 as ompared with indomethain alone, n=5). In addition, LNMMA (1-4 M) markedly enhaned the ontrations to histamine (1-6 M; 2±64%; p<.5 as ompared with ontrol, n=4). Superoxide dismutase With endothelium and L-NMMA (15 units/ml in the presene of LNMMA i` M) did not inhibit the response (29±1%; p=ns as ompared with LNMMA alone, n =4). Relaxations to Nitri Oxide Exogenous nitri oxide (1-9 to 3 x 1-6 M) aused onentration-dependent relaxations of internal +2 C). E -2 - C. -4-9g] NE 1-7M Without endothelium = NE 17M 2min Aetylholine (-HogM) FIGURE 6. Original reording ofendothelium-dependent ontrations to aetylholine in the human saphenous vein. In the presene of the inhibitor of nitri oxide formation, NG-monomethyl L-arginine (L-NMMA, 1-4 M), the musarini agonist evokes onentration-dependent ontrations in the ring with endothelium (top panel) but no signifiant hanges in tension in that without endothelium (bottom panel). NE, norepinephrine.. Z-6 a Control n=4-8 ~ - * L-NMMA n=4-1 CL. 4) -1 * Histamine (-1ogM) FIGURE 7. Endothelium-dependent relaxations to histamine in the human intemal mammary artery. In the presene of indomethain (1`5 M to inhibit prostaglandin prodution) and imetidine (1` M to prevent inhibitory effets of histamine on vasular smooth musle), the monoamine auses potent relaxations (ontrol), whih are prevented by the inhibitor of nitri oxide formation, NG-monomethyl L-arginine (L-NMMA, 1-4 M; p<.5).

6 Yang et al Endothelium-Derived Nitri Oxide in Human Bypass Vessels 57 a Q, 'U a) C 'C ~. - C a 4)Q C - - ~ I= CL (ft a1) * Control n=4 o Indomethain n=4 * Indomethain plus L-NMMA n=5 I._ C) - s CJr. CL C a.,a-- '- O 4DZ ~ O- CL a. (f S * Control n=11 * o CGS n=6 SQ n=5-5- a l , Downloaded from by guest on April 7, 218 Histamine (-logm) FIGURE 8. Endothelium-dependent effets of histamine in the human saphenous vein. Left panel Under ontrol onditions, histamine evokes only very weak relaxations, but marked ontrations at higher onentrations. Indomethain (1` M) to inhibit the formation ofprostaglandins unmasked marked endothelium-dependent relaxations (p <. 5 as ompared with ontrol). These relaxations ould be prevented by the inhibitor ofnitri oxide formation, N'-monomethyl L-arginine (L-NMMA, 1 4M; p <. 5). Right panel In ontrast to aetylholine (see Figure 6), the thromboxane synthetase inhibitor CGS-138 (1J M) did not affet the response to histamine, but the thromboxanea2jendoperoxide reeptor antagonist, SQ-3741, unmasked endothelium-dependent relaxations (p<.5). mammary arteries and saphenous veins. The maximal relaxation did not differ in the two blood vessels and averaged 85+3% in the artery and 8 +5% in the vein (n = 11 and 16, respetively; NS). The saphenous vein exhibited a slightly enhaned response to lower onentrations of nitri oxide (1-8 to 1i M) as ompared with the mammary artery (p<.5), while the pd2 value did not differ statistially ( and 7.+.2, NS; n=11 and 16, respetively). Disussion The present study demonstrates a differential ativation of the endothelial L-arginine and ylooxygenase pathway in the human internal mammary artery and saphenous vein. In the artery, potent endothelium-dependent relaxations were evoked by aetylholine and histamine; the relaxations are mediated by endothelium-derived nitri oxide, while the ylooxygenase produts do not ontribute to the response. In ontrast, veins release less endotheliumderived nitri oxide in response to both autaoids than do arteries, and the nitri oxide effets are markedly inhibited by onomitantly released endothelium-derived ontrating fators originating from the ylooxygenase pathway. The human internal mammary artery must ontinuously release nitri oxide sine the inhibitor of nitri oxide formation, LNMMA,15-17evoked endotheliumdependent ontrations in this preparation. Similarly, LNMMA augmented the ontrations indued by norepinephrine in arterial rings with, but not in those without, endothelium. This indiates that the basal release of endothelium-derived nitri oxide importantly modulates the effets of vasoonstritor hormones in human arteries. In ontrast to the arteries, the saphenous vein appears to form very little nitri oxide under basal onditions, sine LNMMA evoked only weak endothelium-dependent ontrations in quiesent preparations and had no effet on the onentration-response urve to norepinephrine. The fat that aetylholine and histamine evoked potent endothelium-dependent relaxations in the human internal mammary artery onfirms previous observations.4,9 This study demonstrates that both agonists ativate the endothelial L-arginine pathway and that nitri oxide fully aounts for the relaxations indued by histamine and in large part for that aused by aetylholine. Indeed, the inhibitor of the formation of nitri oxide, LNMMA, prevented the endothelium-dependent relaxation to histamine. Sine prostaylin has been exluded,9 nitri oxide formed from L-arginine must be the only mediator of the response to histamine. Similarly, LNMMA markedly redued endothelium-dependent relaxations to aetylholine. In ontrast to histamine, however, the musarini agonist still evoked a small relaxation in the presene of the inhibitor of the L-arginine pathway, indiating that other relaxing fators are released as well. Indeed, although the ylooxygenase inhibitor indomethain18 had no effet on endothelium-dependent relaxations to aetylholine under ontrol onditions, in the presene of LNMMA, the drug further inhibited the response. Beause prosta-

7 Downloaded from by guest on April 7, Cirulation Researh Vol 68, No 1, January 1991 ylin is the major ylooxygenase produt of the internal mammary artery43 and a potent vasodilator in this preparation,44 this indiates that the prostanoid ontributes to the endothelium-dependent relaxations only during inhibition of the L-arginine pathway. Indeed, in endothelial ells in ulture, nitri oxide inhibits the formation of prostaylin.45 Thus, it is likely that nitri oxide formed during stimulation with aetylholine redues the formation of the prostanoid under ontrol onditions, while prostaylin ontributes to the endothelium-dependent relaxations during blokade of the L-arginine pathway. The fat that a small relaxation to aetylholine persisted, even in the presene of LNMMA and indomethain, may be related to an inomplete inhibition of the L-arginine pathway or due to the release of a relaxing substane distint from nitri oxide and prostaylin suh as endothelium-derived hyperpolarizing fator.2-24 It is of interest that, although all patients from whom the internal mammary arteries were obtained suffered from oronary artery disease, the vessels exhibited potent endothelium-dependent relaxations to aetylholine and histamine. Indeed, atheroslerosis is assoiated with a dereased release of endothelium-derived relaxing fator both in experimental animals and humans Although it annot be exluded that even better responses would be observed in normal subjets, this may be related to the known resistane of the human internal mammary artery against atheroslerosis. Even in patients with oronary artery disease, the inidene of atheroslerosis in the human internal mammary artery is less than 5%.49,5 In line with previous observations from this laboratory,4,9 endothelium-dependent relaxations to aetylholine and histamine were markedly less pronouned in the saphenous vein than in the mammary artery. In ontrast to the artery, in the veins indomethain aused a pronouned augmentation of the relaxations indued by aetylholine and unmasked those to histamine. The relaxations to aetylholine and histamine that ourred after inhibition of ylooxygenase were prevented by LNMMA, indiating that, in ontrast to dogs,34 nitri oxide is the only relaxing fator released in human veins after stimulation with the autaoids. A redued apaity of venous vasular smooth musle to relax in response to nitri oxide an be exluded as a ause of the muh weaker endothelium-dependent relaxations in veins as ompared with arteries. Indeed, the relaxations indued by exogenous nitri oxide, as those to nitrovasodilators,4,4451 were slightly enhaned in the human saphenous vein as ompared with the internal mammary artery. Thus, veins must form less nitri oxide in their endothelial ells as ompared with arteries. In addition to nitri oxide, the saphenous vein onomitantly releases an endothelium-derived ontrating fator that offsets, at least in part, the relaxing effets of the endogenous nitrate. Indeed, in the presene of LNMMA, aetylholine evoked endothelium-dependent ontrations. Sine both the inhibitor of ylooxygenase, indomethain, and that of thromboxane synthetase, CGS-138, enhaned the endothelium-dependent relaxations to aetylholine to a similar degree, thromboxane A2 must mediate the response. Similar onlusions have been reahed for the basilar artery of dogs.29'32 In venous oronary bypass grafts studied in situ in the atheterization laboratory, seletive infusion of aetylholine does ause a ontration, while arterial grafts exhibit a dilatation in response to the musarini agonist.52 Thus, this response may also be operative in vivo. In ontrast, indomethain and the thromboxane A2/endoperoxide reeptor antagonist SQ-3741, but not the thromboxane synthetase inhibitor, unmasked the endothelium-dependent relaxations to histamine in the saphenous vein. Thus, it is likely that endoperoxide intermediates (prostaglandin H253) mediate the endothelium-dependent ontrations to histamine. Superoxide radials that mediate endothelium-dependent ontrations in the anine basal artery33 an be exluded beause the free radial savenger superoxide dismutase had no effet. The prodution and release of prostaglandins from the endothelium of the human saphenous vein that ativate platelets and ause ontration may be related to its native anatomial site. Indeed, in a superfiial blood vessel prone to injury, the presene of a prooagulatory and ontrating system may be important to limit blood loss during hemorrhage. The present findings may have important onsequenes for oronary bypass surgery and graft funtion. In arterial grafts, endothelial damage during surgial preparation would augment the effets of vasoonstritor hormones suh as norepinephrine and serotonin and may in turn preipitate postoperative vasospasm.54 If the endothelial layer is preserved, however, the internal mammary artery has a potent endogenous nitrate system that an limit or prevent vasospasm, thrombus formation, and eventually graft olusion. In ontrast, in saphenous veins used as oronary bypass grafts, the endothelial L-arginine pathway is muh less ative and its effets are inhibited by onomitantly released endotheliumderived ontrating fators. Partiularly, thromboxane A2 may failitate platelet adhesion and thrombus formation and may endanger loal blood flow.4.55 In venous grafts, ylooxygenase inhibitors favorably orret the imbalane between endothelium-derived nitri oxide and the ontrating fators, and this may ontribute to the benefiial effets of aspirin and aspirinlike drugs on venous graft funtion.56 A greater release of endothelium-derived nitri oxide in arterial grafts may not only protet against vasospasm and thrombus formation, but also against the development of atherosleroti hanges that ommonly develop in venous grafts.49 Indeed, in vasular smooth musle ells in ulture, stimulation of yli GMP by nitri oxide and other nitrovasodilators inhibits mitogenesis and proliferation.57 Thus, the

8 Yang et al Endothelium-Derived Nitri Oxide in Human Bypass Vessels 59 Downloaded from by guest on April 7, 218 differential biologial properties of venous and arterial grafts may importantly ontribute to graft funtion and at least in part explain their differential pateny rate and linial outome. Aknowledgments The authors wish to thank Bernadette Libsig and Sabine Bohnert, Department of Researh, University Hospital Basel, for their assistane in the preparation of the manusript and Draguljub Popovi and Silvia Distel, Operating Room Laboratory, University Hospital Zurih, for tehnial assistane. We are grateful to Dr. L. Crisione, (Ciba-Geigy, Basel, Switzerland) for kindly supplying CGS-138 and to Dr. J. Wunderlin and M. Ogletree (Squibb, Zurih, and Prineton, N.J.) for kindly providing SQ Referenes 1. Ltisher TF: Endothelial Vasoative Substanes and Cardiovasular Disease. Basel, Switzerland, S Karger, AG, 1988 pp Furhgott RF, Vanhoutte PM: Endothelium-derived relaxing and ontrating fators. FASEB J 1989;3: Furhgott RF, Zawadzki JV: The obligatory role of endothelial ells in the relaxation of arterial smooth musle by aetylholine. Nature 198;299: Ltisher TF, Diederih D, Siebenmann R, Lehmann K, Stulz P, von Segesser L, Yang Z, Turina M, Griidel E, Weber E, Btihler FR: Differene between endothelium-dependent relaxations in arterial and in venous oronary bypass grafts. N Engl J Med 1988;319: Coks TM, Angus JA, Campbell JH, Campbell GR: Release and properties of endothelium-derived relaxing fator (EDRF) from endothelial ells in ulture. J Cell Physiol 1985;123: Cherry PD, Furhgott RF, Zawadzki JV, Jothianandan D: Role of endothelial ells in relaxation of isolated arteries to bradykinin. Pro Natl Aad Si USA 1982;79: Van de Voorde J, Leusen I: Role of the endothelium in the vasodilator response of rat thorai aorta to histamine. Eur J Phannaol 1983;87: Toda N: Mehanism of histamine ations in human oronary arteries. Cir Res 1987;61: Yang Z, Diederih D, Shneider K, Siebenmann R, Stulz P, von Segesser L, Turina M, Buihler FR, Ltisher TF: Endothelium-derived relaxing fator and protetion against ontrations indued by histamine and serotonin in the internal mammary artery and saphenous vein. Cirulation 1989;8; Palmer RMJ, Ferrige AG, Monada S: Nitri oxide release aounts for the biologial ativity of endothelium-derived relaxing fator. Nature 1987;327: Chen W, Palmer RMJ, Monada S: Release of nitri oxide from rabbit aorta. J Vas Med Biol 1989;1: Palmer RMJ, Ashton DS, Monada S: Vasular endothelial ells synthesize nitri oxide from L-arginine. Nature 1988;333: Palmer RMJ, Monada S: A novel itrulline-forming enzyme impliated in the formation of nitri oxide by vasular endothelial ells. Biohem Biophys Res Commun 1989;158: Sakuma I, Stuehr D, Gross SS, Nathan C, Levi R: Identifiation of arginine as a preursor of endothelium-derived relaxing fator. Pro Natl Aad Si USA 1988;85: Rees DD, Palmer RMJ, Hodson HF, Monada S: A speifi inhibitor of nitri oxide formation from L-arginine attentuates endothelium-dependent relaxation. Br J Pharmaol 1989;96: Hibbs JB Jr, Taintor RR, Vavrin Z: Marophage ytotoxiity: Role for L-arginine deiminase and imino nitrogen oxidation to nitrite. Siene 1987;235: Marletta MA: Nitri oxide: Biosynthesis and biologial signifiane. Trends Biohem Si 1989;14: Monada S, Vane JR: Pharmaology and endogenous roles of prostaglandin endoperoxides, thromboxane A2 and prostaylin. Pharmaol Rev 1979;3: Baenziger NL, Fogerty FJ, Mertz LF, Chernuta LF: Regulation of histamine-mediated prostaylin synthesis in ultured human vasular endothelial ells. Cell 1981;24: Feletou M, Vanhoutte PM: Endothelium-dependent hyperpolarization of anine oronary smooth musle. Br J Pharmaol 1988;93: Vanhoutte PM: The end of the quest? Nature 1987;327: Komori K, Lorenz RR, Vanhoutte PM: Nitri oxide, aetylholine, and eletrial and mehanial properties of anine arterial smooth musle. Am J Physiol 1988;255:H27-H Bdny J-L, Brunet PC: Neither nitri oxide nor nitroglyerin aounts for all the harateristis of endothelially mediated vasodilatation of pig oronary arteries. Blood Vessels 1988;25: Taylor SG, Weston AH: Endothelium-derived hyperpolarizing fator: A new endogenous inhibitor from the vasular endothelium. Trends Pharm Si 1988;9: De Mey JG, Vanhoutte PM: Heterogenous behavior of the anine arterial and venous wall: Importane of the endothelium. Cir Res 1982;51: Rubanyi GM, Vanhoutte PM: Hypoxia releases a vasoonstritor substane from the anine vasular endothelium. J Physiol (Lond) 1985;364: Katusi ZS, Shepherd JT, Vanhoutte PM: Anoxi ontrations in isolated erebral arteries: Contribution of endotheliumderived fators, metabolites of arahidoni aid and alium entry. J Cardiovas Pharmaol 1986;8:(suppl 8): Miller VM, Vanhoutte PM: Endothelium-dependent ontrations to arahidoni aid are mediated by produts of ylooxygenase in anine veins. Am J Physiol 1985;248:H432-H Katusi ZS, Shepherd JT, Vanhoutte PM: Endotheliumdependent ontrations to alium ionophore A23187, arahidoni aid and aetylholine in anine basilar arteries. Stroke 1988;19: Ltisher TF, Vanhoutte PM: Endothelium-dependent ontrations to aetylholine in the aorta of the spontaneously hypertensive rat. Hypertension 1986;8: Ltisher TF, Diederih D, Vanhoutte PM, Weber E, Bflhler FR: Endothelium-dependent responses in the ommon arotid and renal artery of normotensive and spontaneously hypertensive rats. Hypertension 1988;11: Usui H, Kurahashi K, Ashida K, Fujiwara M: Aetylholineindued ontratile response in anine basilar artery with ativation of thromboxane A2 synthesis sequene. IRCS Med Si Physiol 1983;1I: Katusi ZS, Vanhoutte PM: Superoxide anion is an endothelium-derived ontrating fator. Am J Physiol 1989;257: H33-H Miller VM, Vanhoutte PM: Is nitri oxide the only endothelium-derived relaxing fator in anine femoral veins? Am J Physiol 1989;257:H191-H Thulesius, Ugaily-Thulesius L, Neglen P, Shuhaiber H: The role of the endothelium in the ontrol of venous tone: Studies on isolated human veins. Clin Physiol 1988;8: Radomski MW, Palmer RMJ, Monada S: Endogenous nitri oxide inhibits human platelet adhesion to vasular endothelium. Lanet 1987;2: Radomski MW, Palmer RMJ, Monada S: The antiaggregating properties of vasular endothelium: Interations between prostaylin and nitri oxide. BrJPharmaol 1987;92: Monada S, Herman SG, Higgs EA, Vale C: Differential formation of prostaylin (PGX or PGI2) by layers of the arterial wall: An explanation for the anti-thromboti properties of vasular endothelium. Thromb Res 1977;11: Loop FD, Lytle BW, Cosgrove DM, Stewart RW, Goormasti M, Williams GW, Golding LAR, Gill CG, Taylor PC, Sheldon WC, Proudfit WL: Influene of the internal mammary-artery

9 Downloaded from by guest on April 7, Cirulation Researh Vol 68, No 1, January 1991 graft on 1-year survival and other ardia events. N Engl J Med 1986;314: Yang Z, Lusher TF: Endothelium-dependent regulatory mehanisms in human oronary bypass grafts: Possible linial impliations. Z Kardiot 1989;78: Mihelson JK, Simpson PJ, Gallas MT, Luhesi BR: Thromboxane synthetase inhibition with CGS 138 improves oronary blood flow after streptokinase-indued thrombolysis. Am Heart J 1987;113: Shumaher WA, Heran CL, Allen GT, Ogletree ML: Leukotrienes ause mesenteri vasoonstrition and hemoonentration in rats without ativating thromboxane reeptors. Prostaglandins 1989;38: Subramian VA, Hernandez Y, Tak-Goldman K, Grabowski EF, Weksler BB: Prostaylin prodution by internal mammary artery as a fator in oronary bypass grafts. Surgery 1986;1: Yang Z, Buhler FR, Diederih D, Lusher TF: Different effets of endothelin on yli AMP- and yli GMPmediated vasular relaxation in human arteries and veins: Comparison with norepinephrine. J Cardiovas Pharmaol 1989;13(suppl 5): Doni MG, Whittle BJR, Palmer RMJ, Monada S: Ations of nitri oxide on the release of prostaylin from bovine endothelial ells in ulture. Eur J Pharmaol 1988;151: Brunkwall J, Bergqvist D, Stjernquist U: Prostaylin and thromboxane release from the vessel wall: Comparison between an inubation and a perfusion model. Prostaglandins 1987;34: Forstermann U, Mugge A, Alheid U, Haverih A, Frolih JC: Seletive attenuation of endothelium-mediated vasodilation in atherosleroti human oronary arteries. Cir Res 1988;62: Shimokawa H, Vanhoutte PM: Impaired endotheliumdependent relaxation to aggregating platelets and related vasoative substanes in porine oronary arteries in hyperholesterolemia and atheroslerosis. Cir Res 1989;64: Grondin CM, Campeau L, Lesperene J, Enjalbert M, Bourassa MG: Comparison of late hanges in internal mammary artery and saphenous vein grafts in two onseutive series of patients 1 years after operation. Cirulation 1984; 7(suppl I):I-28-I Singh RN: Atheroslerosis and the internal mammary arteries. Cardiovas Intervent Radiol 1983;6: Lusher TF, Rihard V, Yang Z: Interation between endothelium-derived nitri oxide and SIN-1 in human and porine blood vessels. J Cardiovas Phartnaol 1989;14(suppl 11): S76-S8 52. Werner GS, Wiegand V, Kreuzer H: Effet of aetylholine on arterial and venous grafts and oronary arteries in patients with oronary artery disease. Eur Heart J 199;11: Kato T, Iwana Y, Okumura K, Hashimoto H, Ito T, Satake T: Prostaglandin H2 may be the endothelium-derived ontrating fator released by aetylholine in the aorta of the rat. Hypertension 199;15: Sarabu MR, MClung JA, Fass A, Reed GE: Early postoperative spasm in left internal mammary artery bypass grafts. Ann Thora Surg 1987;44: Fuster V, Dewanjee MK, Kaye MP, Josa M, Metke MP, Chesboro JH: Noninvasive radioisotopi tehnique for detetion of platelet deposition in oronary artery bypass grafts in dogs and its redution with platelet inhibitors. Cirulation 1979;6: Chesebro JH, Fuster V, Elvebak LR, Clements IP, Smith HC, Holmes DR Jr, Bardsley WT, Plutz JR, Wallae RB, Puga FJ, Orszulak TA, Piehler JM, Danielson GK, Shaff HV, Frye RL: Effet of dipyridamole and aspirin on late vein-graft pateny after oronary bypass operation. N Engl J Med 1984;31: Garg UC, Hassid A: Nitri oxide-generating vasodilators and 8-bromo-yli guanosine monophosphate inhibit mitogenesis and proliferation of ultured rat vasular smooth musle ells. J Clin Invest 1989;83: KEY WORDS * aetylholine * endoperoxides * histamine N-monomethyl L-arginine * thromboxane A2

10 Different ativation of the endothelial L-arginine and ylooxygenase pathway in the human internal mammary artery and saphenous vein. Z H Yang, L von Segesser, E Bauer, P Stulz, M Turina and T F Lüsher Downloaded from by guest on April 7, 218 Cir Res. 1991;68:52-6 doi: /1.RES Cirulation Researh is published by the Amerian Heart Assoiation, 7272 Greenville Avenue, Dallas, TX Copyright 1991 Amerian Heart Assoiation, In. All rights reserved. Print ISSN: Online ISSN: The online version of this artile, along with updated information and servies, is loated on the World Wide Web at: Permissions: Requests for permissions to reprodue figures, tables, or portions of artiles originally published in Cirulation Researh an be obtained via RightsLink, a servie of the Copyright Clearane Center, not the Editorial Offie. One the online version of the published artile for whih permission is being requested is loated, lik Request Permissions in the middle olumn of the Web page under Servies. Further information about this proess is available in the Permissions and Rights Question and Answer doument. Reprints: Information about reprints an be found online at: Subsriptions: Information about subsribing to Cirulation Researh is online at: