Normalization of Nitric Oxide Production Corrects Arterial Vasodilation and Hyperdynamic Circulation in Cirrhotic Rats

Transcription

1 GASROENEROLOGY 1995;109: Normalization of Nitric Oxide Production Corrects Arterial Vasodilation and Hyperdynamic Circulation in Cirrhotic Rats MICHEL NIEDERBERGER, PIERRE-YVES MARIN, PERE GINI~S, KENNEH MORRIS, PHOEBE SAI, DING-LI XU, IVAN MCMURRY, and ROBER W. SCHRIER Department of Medicine, University of Colorado School of Medicine, Denver, Colorado Background & Aims: Recent studies suggest that production of nitric oxide is increased in cirrhosis. his study determines to what extent this increased production contributes to arterial vasodilation and hyperdynamic circulation in cirrhosis. Methods: Mean arterial pressure (MAP), cardiac index, and Systemic vascular resistance (SVR) were determined in cirrhotic rats with ascites undergoing long-term treatment with different doses of the NO synthesis inhibitor /V~-nitro-L-arginine methyl ester (L-NAME) (3 mg or 0.5 mg. kg -1. day=z). Untreated cirrhotic rats with ascites and controls were also studied. he vascular production of NO was estimated by the aortic concentration of guanosine 3',5'- cyclic monophosphate (cgmp). Results: Untreated cirrhotic rats had significantly lower MAP and SVR and higher cardiac index and aortic cgmp concentration than controls. When administrated to cirrhotic rats, an L-NAME dose of 3 mg- kg -1. day -1 induced a reduction of cgmp concentration less than normal levels. In these rats, MAP and SVR increased to greater than and cardiac index decreased to less than values in controls. By contrast, cirrhotic rats treated with 0.5 rag. kg -1. day -1 L-NAME had similar aortic cgmp concentrations as controls, suggesting a normalization of NO production. his was associated with a normalization of MAP, cardiac index, and SVR and a reduction in the elevated plasma renin activity and vasopressin concentration. Conclusions: Normalization of vascular NO production corrects systemic hemodynamic abnormalities in cirrhotic rats with ascites. A terial vasodilation and hyperdynamic circulation are almost invariably present in human and experimental cirrhosis. 1-5 his arterial vasodilation, mainly splanchnic, occurs early in cirrhosis, resulting in a relative underfilling of the arterial circulation. his underfilling activates a baroreceptor-mediated neurohumoral response, i.e., renin-angiotensin-aldosterone axis, sympathetic nervous system, and nonosmotic release of arginine vasopressin, to counterregulate this vasodilation. hese hormones eventually lead to renal sodium and water retention in an attempt to compensate for the arterial underfilling by plasma volume expansion. 6-9 hese hemo- dynamic abnormalities are proposed to play an important role in the development of sodium and water retention and ascites formation in cirrhosis. 2'1 he pathogenic mechanism responsible for these abnormalities is still unknown. A variety of endogenous substances with vasodilator activity have been proposed to account for the arterial vasodilation, but the exact mediators are unknown, l,ti Recently, attention has been focused on nitric oxide, a very potent vasodilator factor synthesized in normal conditions in the arterial wall by the vascular endothelium. NO released from endothelial cells diffuses into the extracellular space and activates the soluble guanylate cyclase of the underlying vascular smooth muscle ceils. his results in the generation of the second messenger of guanylate cyclase, guanosine 3',5'-cyclic monophosphate (cgmp), and vascular relaxation. 12'13 NO plays an important role in the regulation of blood flow and arterial pressure in normal circumstances because the inhibition of endogenous NO synthesis in experimental animals and healthy subjects is associated with vasoconstriction and arterial hypertension In 1991, Vallance and Moncada proposed that an increased vascular production of NO could account for the arterial vasodilation in cirrhosis. 19 Since then, several studies have been published investigating this hypothesis in animals with portal hypertension he approach used in these investigations was to assess the short-term effects of the pharmacological inhibition of NO synthesis with L-arginine analogues on systemic hemodynamics. As in normal animals, NO-synthesis inhibition in animals with portal hypertension caused a marked increase in arterial pressure and total systemic vascular resistance and reduction in cardiac output. 23 In these studies, however, doses of L-arginine analogues were used that could have suppressed NO production to less than normal lev- Abbreviations used in this paper: AVP, plasma arginine vasopressin; L-NAME, N~-nitro-L-arginine methyl ester; PRA, plasma renin activity by the American Gastroenterological Association /95/$3.00

2 November 1995 NORMALIZAION OF NO PRODUCION IN CIRRHOSIS 1625 els. In addition, the effect of L-arginine analogue treat- ment on vascular NO production was not estimated. herefore, the precise role of NO in the pathogenesis of the hyperdynamic circulation in cirrhosis remains to be defined conclusively. he present study was aimed at investigating the ef- fects of oral long-term administration of different doses of the NO synthesis inhibitor NC;-nitro-L-arginine methyl ester (L-NAME) on systemic hemodynamics and vascular NO production in rats with experimental cirrho- sis and ascites. Vascular NO production was estimated by measuring aortic cgmp concentration, which is a sensitive index of NO synthase activity in the arterial wall. 24'25 Previous studies from our laboratory have shown that rats with cirrhosis and ascites have increased aortic cgmp concentration, suggesting an enhanced vascular NO production. 26 he results of the current study indi- cate that the administration of L-NAME at a dose that normalizes aortic cgmp concentration reverses arterial vasodilation and hyperdynamic circulation in rats with cirrhosis and ascites. hese findings indicate that NO plays a major role in the pathogenesis of the arterial vasodilation and hyperdynamic circulation in experimen- tal cirrhosis. Materials and Methods Animals Male Sprague-Dawley rats (Sasco, Omaha, NE) housed in a controlled environment were used in this investigation. Animals were allowed free access to food and water until the time of study. Cirrhosis was induced by weekly intragastric administration of carbon tetrachloride and phenobarbital in the drinking water (0.35 g/dl) as described elsewhere. 27 Cirrhotic rats were studied after ascites developed. Cirrhosis was confirmed by histological examination and ascites by visual examination at laparotomy. In these cirrhotic rats, the study was performed 9-11 days after the last dose of carbon tetrachloride. he investigation included the following groups of rats: phenobarbital-treated control rats (n = 8), cirrhotic rats with ascites (n = 8), and cirrhotic rats with ascites treated with either 3 mg" kg 1. day-1 (n = 8) or 0.5 mg" kg -~ day 1 (n = 8) L-NAME (Bachem, orrance, CA) administered by gavage twice a day for 7 days. hese two doses of L-NAME were selected to cause a reduction in aortic cgmp concentration to subnormal (3 mg'kg 1-day-l) or normal (0.5 mg'kg-*" day -1) levels, respectively. Doses were chosen from a pilot study in which the effect of several doses of L-NAME on aortic cgmp concentration in cirrhotic rats with ascites was assessed. o investigate the effects of long-term NO synthesis inhibition in control rats, two groups of phenobarbital-treated rats administered L-NAME (3 or 0.5 mg" kg -~'day-1) were also studied (n = 8 in both groups). o assess the effects of longterm NO-synthesis inhibition on hormonal systems, plasma renin activity (PRA) and plasma arginine vasopressin (AVP) levels were measured in an additional group of cirrhotic rats with ascites treated with L-NAME 0.5 mg" kg -1 day -1 for 1 week (n = 6) as well as in untreated cirrhotic rats with ascites (n = 6). Experimental Procedures he hemodynamic evaluation, including mean arterial pressure, cardiac output, and total systemic vascular resistance, was performed in all rats. Rats were anesthetized with ether, and a carotid artery and the two jugular veins were cannulated with a polyethylene catheter (PE 30). he catheters were tunneled subcutaneously through the back of the neck and exteriorized. After 2 days of recovery from anesthesia, the conscious rats were placed in a small, well-ventilated, Plexiglas chamber (Narco Bio-systems, Houston, X) fluxed with room air. Cardiac output was then measured using the standard indocyanine green dye technique as described in detail elsewhere. 2s-3 Briefly, 5 mg of indocyanine green dye was injected into a jugular vein catheter while blood was simultaneously pumped through the shunt from carotid artery to jugular vein at 3 ml/ rnin through a densitometer cuvette. his signal was imputed to a computer, physiograph, and oscilloscope to generate a dye curve. he arterial pressure signal was imputed to a physiograph and microcomputer. Cardiac index was determined by dividing cardiac output by the animal's body weight. Systemic vascular resistance was calculated by mean arterial pressure/ cardiac index. After the hemodynamic evaluation, rats were allowed to recover for 24 hours and then were killed by decapitation to obtain the thoracic aorta for cgmp concentration determination. he aorta was cleaned from adjacent tissue, rinsed in cold phosphate-buffered saline, and snap-frozen in liquid nitrogen in a very short period of time (<3 minutes). issue samples were stored at -70 C until cgmp concentration determination. For hormonal determination, a 2-mE blood sample was obtained from a right femoral artery catheter in conscious animals placed in individual cages. Blood was immediately centrifuged at 4 C, and the plasma was frozen at -70 C for further analysis. Determination of Aortic cgmp Concentration and Hormonal Measurements Concentration of cgmp in aortic tissue was measured as previously described. 26 Briefly, the thoracic aorta was homogenized in 1 ml of0.1n HC1 with an all-glass homogenizer at 4 C. Homogenates of aortic tissue were then centrifuged at 3000g for 60 minutes, and the supernatant was stored at -20 C until assayed. Protein concentration from each sample was determined with the Bradford method (Bio-Rad Laboratories, Richmond, CA). cgmp concentration was measured by radioimmunoassay (Amersham Corp., Arlington Heights, IL) using an acetylation protocol and was expressed as femtomoles per milligram of protein. PRA was estimated by radioimmunoassay of Angiotensin I (Incstar, Stillwater, MN) generated and was expressed as nanograms per milliliter per hour. Plasma

3 1626 NIEDERBERGER E AL. GASROENEROLOGY Vol. 109, No. 5 able 1. Systemic Hemodynamic Parameters and Aortic cgmp Concentration in Untreated Control Rats and Rats reated With L-NAME for 7 Days by Gavage II. L-NAME ill. L-NAME I. Untreated 0.5 mg. kg -1. day -~ 3 mg. kg -i. day 1 (n = 8) (n = 8) (n = 8) I vs. II a I vs. III a Mean arterial pressure b (ram Hg) Cardiac index (ml. min -~. 100 g=~) Systemic vascular resistance ~ (ram Hg. min. ml -~. 100 g-l) Aortic cgmp concentration e (fmol/mg) 128 _ <0.05 < _ <0.05 < _ <0.01 < <0.01 <0.001 NOE. Values are expressed as mean + SEM. anewman- Keuls test. ~Analysis of variance, F = 13.2; P = CAnalysis of variance, F = 9.2; P = ~Analysis of variance, F = 18.1; P < eanalysis of variance, F = 15.8; P = AVP levels were measured by radioimmunoassay as described elsewhere 27 and were expressed as picograms per milliliter. Statistical Methods Statistical analysis of the results was performed using analysis of variance and Newman-Keuls tests, unpaired Student's t test, and linear regression analysis. Results are expressed as mean SEM. Values were considered significant when P was <0.05. he investigation was approved by the University of Colorado Health Sciences Center Animal Care and Use Committee. Results Effects of L-NAME on Systemic Hemodynamics and Aortic cgmp Concentration in Control Rats Long-term NO-synthesis inhibition in control rats caused marked changes in systemic hemodynamics characterized by an increase in arterial pressure and total systemic vascular resistance and a reduction in cardiac index (able 1). hese effects were more marked in the group of rats treated with the highest dose of L-NAME (3 mg" kg -l" day-l), confirming that progressive suppression of endogenous NO synthesis is associated with gradual arterial vasoconstriction. In these animals, aortic cgmp concentration correlated with the L-NAME dose. Rats treated with the lower dose of L-NAME (0.5 mg" kg -1" day -1) had a mean aortic cgmp concentration less than untreated control rats. In rats treated with the highest dose, the decrease in aortic cgmp concentration was most marked (able 1). In these three groups of rats, aortic cgmp concentration correlated inversely with arterial pressure (r = -0.51; P = 0.02) and total systemic vascular resistance (r = -0.65; P = 0.002) and directly with cardiac index (r = 0.65; P = 0.002). Effects of L-NAME on Systemic Hemodynamics and Aortic cgmp Concentration in Cirrhotic Rats As previously reported, 26 cirrhotic rats with ascites had higher concentrations of aortic cgmp than control rats ( vs fmol/mg; P < 0.001) (Figure 1). he long-term administration of L-NAME (3 mg" kg -1" day -1) to cirrhotic rats with ascites was associated with a marked reduction of aortic cgmp concentration to less than the levels observed in control animals ( fmol/mg vs. controls; P < 0.05). By 0 a..m t- O.< N controls a (72s+~) cirrhosis with ascites untreated L-Name L.Name O.$m~kg/d 3mg/kg/d Figure 1. Aortic concentration of cgmp (mean _+ SEM) in control rats (n = 8) and cirrhotic rats with ascites either untreated (n = 8) or treated with 0.5 mg.kg-l-day -1 (n = 8) or 3 mg.kg-l.day -1 (n = 8) L-NAME for 7 days by gavage (mean + SEM). Analysis of variance; F = 37.3; P < ap < vs. all groups; bp < 0.01 vs. 3 mg.kg-l.day-1; L-NAME cp < 0.05 vs. controls (Newman-Keuls test).

4 November 1995 NORMALIZAION OF NO PRODUCION IN CIRRHOSIS 1627 =,, A ~ 120- i 110" B "5--' C 6 b (131.-0) (128.=+4) 4 (4.0;L-0.3) d (2.2~-0.2) g (49+3) e (3.7±0.2) C (147~-4) / / /,/x,//,.// :/t,.// :.//,./1 /// f (5.3~-o.4) r//, /za 1 1, #.lie with ascites, and cirrhotic rats with ascites treated with the two different doses of L-NAME are shown. As expected, cirrhotic rats with ascites had a characteristic hemodynamic pattern of arterial vasodilation and hyperdynamic circulation as indicated by lower arterial pressure and systemic vascular resistance and higher cardiac index than control rats ( vs mm Hg, P < 0.01; 2.2 _+ 0.2 vs mm Hg" min'ml -1"100 g 1, p < 0.001; vs ml" min-* 100 g-l, p < 0.001; respectively). Cirrhotic rats with ascites treated with the highest dose of L- NAME had significantly increased arterial pressure and systemic vascular resistance compared with control rats ( mm Hg and mm Hg'min" ml -1" 100 g-l; p < 0.05 and P < 0.01, respectively). Cardiac index was lower, although the difference did not reach statistical significance ( ml'min -1" 100 g-i; NS). Cirrhotic rats with ascites treated with the lowest dose of L-NAME (0.5 mg" kg -l" day -1) had significantly higher arterial pressure and systemic vascular resistance and lower cardiac index than untreated cirrhotic rats. he hemodynamic parameters in rats treated with this lower dose of L-NAME were similar to those in control rats ( mm Hg, mm Hg'min'mL -l" 100 g-l, and ml'min -l" 100 g-l, respectively; NS). Effects of L-NAME on Hormonal Parameters in Cirrhotic Rats Cirrhosis with ascites Controls Untreated L-Name L-Name O~mg/kg/d 3mg/kg/d Figure 2. Systemic hemodynamic parameters (mean _+ SEM) in control rats (n = 8) and cirrhotic rats with ascites either untreated (n = 8) or treated with 0.5 mg. kg -~. day =1 or 3 mg kg 1. day-~ L-NAME for 7 days by gavage. Mean arterial pressure; F = 11.0; P < ap < 0.01 vs. all groups; bp < 0.05 VS. L-NAME 3; P < 0.05 vs. controls. Systemic vascular resistance: F = 18.7; P < ep < 0.01 vs. all groups; ep < 0.01 VS. L-NAME 3; fp < 0.01 vs. controls, Cardiac index: F = 17.5; P < gp < 0.01 vs. all groups; hp < 0.05 VS. 3 mg kg -1"day -~ L-NAME (analysis of variance and Newman-Keuls test). contrast, cirrhotic rats with ascites treated with L-NAME (0.5 mg" kg -1" day -1) had aortic cgmp concentrations iess than untreated cirrhotic rats with ascites and similar to control rats ( fmol/mg; P < and not significant, respectively). In Figure 2, the systemic hemodynamic parameters in control rats, cirrhotic rats he effects of long-term L-NAME administration (0.5 mg'kg-l'day -*) on hormonal parameters are shown in Figure 3. L-NAME treatment in cirrhotic rats with ascites was associated with a marked decrease in PRA and AVP (PRA: cirrhosis, ng'ml-*'h-1; cirrhosis plus L-NAME, 2.1 +_ 0.8 ng" ml -1 h-l; P < 0.05; AVP: cirrhosis, pg/ ml; cirrhosis plus L-NAME, pg/ml; P < 0.05). Discussion Recently, several lines of evidence have been presented suggesting the existence of an increased synthesis of NO in experimental cirrhosis. First, the increment in blood pressure with short-term NO-synthesis inhibition with the L-arginine analogue NC;-nitro-L-arginine is greater in rats with cirrhosis and ascites than in control rats) 1 Next, studies using mesenteric arterial preparations or aortic rings of rats with cirrhosis and ascites have shown that the impaired pressor responsiveness to vasoconstrictors characteristic of cirrhosis can be corrected by NO-synthesis inhibition) 2-34 his impaired

5 1628 NIEDERBERGER E AL. GASROENEROLOGY Vol. 109, No. 5 A 12-! l 0... B 10] (9.3+_3.4) Cirrhosis with ascites ( ) s (2,.1:L-0,8). Cirrhosis with MeRes + L-NAME O.$mg/kg/d Cirrhosis Cirrhosis with ascites with asdtes + L.NAME 0.$mg/kg/d Figure 3. (A) PRA and (B) AVP levels in untreated cirrhotic rats with ascites (n = 6) and cirrhotic rats with ascites treated with 0.5 mg-kg 1.day-1 L-NAME for 7 days (n = 6). *P < 0.05 (unpaired Student's t test). Normal values in control rats for PRA and AVP in our laboratory are ng oml -1. h -1 and pg/ml, respectively. responsiveness in aortic rings is also abolished by endothelial denudatio'n. 33'34 Furthermore, aortic rings from cirrhotic rats with ascites show an enhanced relaxing effect of NO-dependent vasodilators. 35 Lastly, NO pro- duction from vascular tissue, as assessed either by the release of NO from in vitro preparation of aorta or mesenteric artery or by aortic concentration of cgmp, is increased in cirrhotic rats compared with controls. 26'36 he finding of high serum nitrite and nitrate levels in patients with cirrhosis and ascites also suggests the existence of an increased NO production in human cirrhosis. 37 On this background, the research on the role of NO in the pathogenesis of hemodynamic abnormalities in experimental cirrhosis is now mainly focused on three different aspects: the identification of the specific isoform of NO synthase (i.e., constitutive or inducible) that accounts for the increased NO synthesis, the mechanism(s) responsible for the activation of NO synthase, and the role that NO plays in the pathogenesis of arterial vasodilation and hyperdynamic circulation. he present study was designed to investigate the role of NO in the hyperdynamic circulation in cirrhosis. Specifically, the hypothesis was examined that if the arterial vasodilation and hyperdynamic circulation in cirrhosis is caused by overproduction of NO, the normalization of NO synthesis would result in the correction of abnormalities in the arterial circulation. In the present investigation, vascular NO production was assessed by measuring aortic cgmp concentration. Previous investigations have shown that cgmp content in the arterial wall estimates the degree of NO-synthase activation In these latter studies, normal rats that received long-term treatment with the NO synthesis inhibitor L-NAME showed a dose-dependent decrease in arterial wall cgmp concentration and an increase in arterial pressure. Changes in aortic cgmp concentration and arterial pressure were inversely correlated and could be reversed after the short-term administration of L-arginine. he results of the current study confirmed these findings. In control rats treated with different doses of t-name, a dose-dependent decrease in aortic cgmp concentration was observed (able 1). Furthermore, changes in aortic cgmp concentration correlated with modifications in systemic hemodynamics, thus suggesting comparable effects in conductance and resistance vessels. herefore, these findings further support that cgmp content in the arterial wall is a sensitive index of in vivo NO-synthase activation and may be used to monitor changes in the activity of this enzyme. With this approach of estimating the activity of NO synthase in the arterial wall, the long-term administration of a dose of L-NAME (0.5 rag" kg -1 day -1) normalized the aortic cgmp production in cirrhotic rats. Moreover, this dose restored the mean arterial pressure, the systemic vascular resistance, and the cardiac output to the level of the control rats. he cirrhotic rats treated

6 November 1995 NORMALIZAION OF NO PRODUCION IN CIRRHOSIS 1629 with this low dose of L-NAME, 0.5 mg" kg -1" day -1, also showed a marked suppression of PRA and AVP levels (Figure 3). his decrease in PRA and AVP is also compatible with the reversal of arterial underfilling, secondary to vasodilation. 4'6-1 Cirrhotic rats with ascites were studied because the hemodynamic alterations are more pronounced than cirrhotic rats without ascites. As we previously showed, 26 however, aortic cgmp concentration is also significantly increased in cirrhotic rats without ascites compared with the controls but to a lesser extent than cirrhotic rats with ascites. his suggests a smaller amount of NO production. hus, in cirrhotic rats without ascites, the dose of L-NAME needed to normalize the hemodynamics should be smaller. herefore, the present results emphasize the major role of increased NO production in experimental cirrhosis and are consistent with the hypothesis that arterial vasodilation is a crucial event in cirrhosis, as proposed by the arterial vasodilation hypothesis. 4-1 It has been postulated that the increased cardiac output in human cirrhosis was not a consequence of arterial vasodilation and afterload reduction but is determined primarily by an increase in vascular volume according to echocardiographic estimation of left ventricular size during the cardiac cycle. 38 hese findings with indirect measurements have to be interpreted cautiously because central circulatory underfilling, when using central catheterization to determine the central blood volume, has been shown to be an integral part of the hemodynamic derangement in human cirrhosis. 39 Moreover, during volume expansion, blood pressure should increase, and this does not explain the progressive decrease in arterial pressure that characterizes cirrhosis. he present results show that by normalizing a locally produced vasodilator, namely NO, the elevated cardiac output and the decrease in blood pressure are corrected, thus reinforcing the role of arterial vasodilation in experimental cirrhosis. hese results also raise the possibility of using longterm vascular NO inhibition to correct hemodynamic abnormalities in cirrhosis and, therefore, to prevent their complications. However, this encouraging possibility needs further studies to assess the effect of NO inhibition on renal sodium and water retention and eventually on mortality. hus, it was not possible to evaluate in the present study because the rats were killed at the end of the experiment to measure cgmp concentration. he long-term effect of this selective modulation of NO production on splanchnic hemodynamics should also be assessed because short-term administration of NG-mono - methyl-l-arginine, another NO synthase inhibitor, decreased portal vein but not portal pressure because of a concomitant increase in portal vascular resistance. 21 In conclusion, the results of the present study show that the normalization of aortic cgmp concentration by long-term NO-synthesis inhibition is associated with a normalization of the arterial vasodilation and hyperdynamic circulation in experimental cirrhosis. hus, an overproduction of NO seems to play a major role in the pathogenesis of these hemodynamic perturbations in cirrhosis. References 1. Schrier RW, Caramelo C. Hemodynamics and hormonal alterations in hepatic cirrhosis. In: Epstein M, ed. he kidney in liver disease. 3rd ed. Baltimore: Williams & Wilkins, 1988: Arroyo V, Gin(~s P, JimOnez W, RodOs J. Ascites, renal failure, and electrolyte disorders in cirrhosis. Pathogenesis, diagnosis, and treatment. In: Mclntyre N, Benhamou JP, Bircher J, Rizet~o M, Rod6s J, eds. Oxford textbook of clinical hepatology. Oxford, England: Oxford Medical, 1991: Genecin P, Groszmann R J. Portal hypertension. In: Schiff E, Schiff L, eds. Diseases of the liver. 7th ed. Philadelphia: Lippincott, 1993: Shapiro MD, Nichols KM, Groves BM, Kluge R, Chung HM, Bichet DG, Schrier RW. Interrelationship between cardiac output and vascular resistance as determinants of effective arterial blood volume in cirrhotic patients. Kidney Int 1985;28: Albillos A, Columbato LA, Groszmann RJ. Vasodilation and sodium retention in prehepatic portal hypertension. Gastroenterology 1992; 102: Schrier RW. Pathogenesis of sodium and water retention in highoutput and low-output cardiac failure, nephrotic syndrome, cirrhosis and pregnancy. N Engl J Med 1988;319: , Schrier RW. Body fluid volume regulation in health and disease: a unifying hypothesis. Ann Intern Med 1990; 113: Schrier RW. An odyssey into the milieu int6rieur: pondering the enigmas. J Am Soc Nephrol 1992;2: Martin PY, Schrier RW. Renal sodium excretion and edematous disorders. Endocrinol Metab Clin North Am 1995;24: Schrier RW, Arroyo V, Bernardi M, Epstein M, Henriksen JH, RodOs J. Peripheral arterial vasodilation hypothesis: a proposal for the initiation of renal sodium and water retention in cirrhosis. Hepatology 1988;8: Groszmann RJ. Vasodilatation and hyperdynamic circulatory state in chronic liver diseases, in: Bosch J, Groszmann R J, eds. Portal hypertension. Pathophysiology and treatment. Oxford, England: Blackwell Scientific, 1994; Moncada S, Palmer RM, Higgs EA. Nitric oxide: physiology, pathophysiology, and pharmacology. Pharmacol Rev 1991;43: Lowenstein C J, Dinerman JL, Snyder SH. Nitric oxide: a physiologic messenger. Ann Intern Med 1994; 120: Rees DD, Palmer RM, Moncada S. Role of endothelium-derived nitric oxide in the regulation of blood pressure. Proc Natl Acad Sci USA 1989;86: Ribeiro MO, Antunes E, de Nucci G, Lovisolo SM, Zatz R. Chronic inhibition of nitric oxide synthesis. A new model of arterial hypertension. Hypertension 1992;20: Dananberg J, Sider RS, Grekin RJ. Sustained hypertension induced by orally administered nitro-l-arginine. Hypertension 1993;21: Haynes WG, Noon JP, Walker BR, Webb DJ. Inhibition of nitric oxide synthesis increases blood pressure in healthy subjects. J Hypertens 1993; 11: Stamler JS, Loh E, Roddy MA, Currie KE, Creager MA. Nitric oxide

7 1630 NIEDERBERGER E AL. GASROENEROLOGY Vol. 109, No. 5 regulates basal systemic and pulmonary vascular resistance in healthy humans. Circulation 1994;89: Vallance P, Moncada So Hyperdynamic circulation in cirrhosis: a role for nitric oxide? Lancet 1991;337: Pizcueta MP, Piqu6 JM, Bosch J, Whittle B J, Moncada S. Effects of inhibiting nitric oxide biosynthesis on the systemic and splanchnic circulation of rats with portal hypertension. Br J Pharmacol 1992; 105: Pizcueta P, Piqu6 JM, Fernandez M, Bosch J, Rod6s J, Whittle B J, Moncada S. Modulation of the hyperdynamic circulation of cirrhotic rats by nitric oxide inhibition. Gastroenterology 1992; 103: Lee FY, Albillos A, Colombato LA, Groszmann RJ. he role of nitric oxide in the vascular hyporesponsiveness to methoxamine in portal hypertensive rats. Hepatology 1992;16: Lee FY, Colombato LA, Albillos A, Groszmann RJ. N~-nitro-L-argi - nine administration corrects peripheral vasodilation and systemic capillary hypotension and ameliorates plasma volume expansion and sodium retention in portal hypertensive rats. HepatoiogS/1993; 17: Arnal JF, Warin L, Michel JB. Determinants of aortic cyclic guanosine monophosphate in hypertension induced by chronic inhibition of nitric oxide synthase. J Clin Invest 1992;90: Arnal JF, El Amrani AI, Michel JB. Atrial natriuretic factor influences in vivo plasma, lung and aortic wall cgmp concentrations differently. Eur J Pharmacol 1993;237: Niederberger M, Gin6s P, sai P, Martin PY, Morris K, Weigert A, McMurtry I, Schrier RW. Increased aortic cyclic guanosine monophosphate concentration in experimental cirrhosis in rats: evidence for a role of nitric oxide in the pathogenesis of arterial vasoditation in cirrhosis. Hepatology 1995;21: Kim JK, Summer SN, Howard RL, Schrier RW. Vasopressin gene expression in rats with experimental cirrhosis. Hepatology 1993;17: Coleman G. Cardiac output by dye dilution in the conscious rat. J Appl Physiol 1974;37: Haynes J, Chang SW, Morris KG, Voelkel NF. Platelet-activating factor antagonists increase vascular reactivity in perfused rat lungs. J Appl Physiol 1988;65: Stevens, Morris K, McMurtry I, Zamora M, ucker A. Pulmonary and systemic vascular responsiveness to NF-(z in conscious rats. J Appl Physiol 1993; 74: Cl~ria J, Jim6nez W, Ros J, Asbert M, Castro A, Arroyo V, Rivera F, Rodes J. Pathogenesis of arterial hypotension in cirrhotic rats with ascites: role of endogenous nitric oxide. Hepatology 1992; 15: Sieber CC, L6pez-aiavera JC, Grozmann RJ. Role of nitric oxide in the in vitro splanchnic vascular hyporeactivity in ascitic cirrhotic rats. Gastroenterology 1993; 104: Castro A, Jim6nez W, Cl~ria J, Ros J, Martlnez JM, Bosch M, Arroyo V, Piulats J, Rivera F, Rodes J. Impaired responsiveness to angiotensin II in experimental cirrhosis: role of nitric oxide. Hepatology 1993; 18: Weigert A, Niederberger M, Gines P, Martin PY, Higa EMS, McMurtry Y, Schrier RW. Endothelium dependent vascular hyporesponsiveness without nitric oxide synthase (NOS) induction in aorta of cirrhotic rats (abstr). Hepatology 1994;20:21A. 35. Cl~ria J, Jim6nez W, Ros J, Rigol M, Angeli P, Arroyo V, Rivera F, Rodes J. Increased nitric oxide-dependent vasorelaxation in aortic rings of cirrhotic rats with ascites. Hepatology 1994;20: Ros J, Jim~nez W, Lamas S, Cl~ria J, Arroyo V, Rivera F, Rod6s J. Nitric oxide production in arterial vessels of cirrhotic rats. Hepatology 1995;21: Guarner C, Soriano G, omas A, Bulbena O, Novella M, Balanzo J, Vilardell F, Mourelle M, Moncada S. Increased serum nitrite and nitrate levels in patients with cirrhosis: relationship to endotoxemia. Hepatology 1993;18: Lewis FW, Adair O, Rector WG Jr. Arterial vasodilation is not the cause of increased cardiac output in cirrhosis. Gastroenterology 1992; 102: Henriksen JH, 8endtsen F, Sorensen A, Stadeager C, Ring- Larsen H. Reduced central blood volume in cirrhosis. Gastroenterology 1989; 97: Received March 30, Accepted July 3, Address requests for reprints to: Robert W. Schrier, M.D., Department of Medicine, University of Colorado Health Sciences Center, 4200 East 9th Avenue, Denver, Colorado Fax: (303) Supported by grant DK from the National Institutes of Health. Dr. Niederberger received a grant from the Swiss National Science Foundation. Dr. Gin,s received grant CIRI BE from the Consell Interdepartmental de Recerca i Innovaci6 ecnol6gica, Generalitat de Catalunya, Catalunya, Spain, and a grant from the Asociacibn Espafiola para el Estudio del H gado. Dr. Martin received a grant from the University Hospital of Geneva, Geneva, Switzerland. Presented in part at the 27th annual meeting of the American Society of Nephrology on October 26-27, 1994, and published in abstract form (J Am Soc Nephrol 1994; 5:585A).