Increased Angiogenesis in Portal Hypertensive Rats: Role of Nitric Oxide

Transcription

1 Increased Angiogenesis in Portal Hypertensive Rats: Role of Nitric Oxide LAZAR T. SUMANOVSKI, 3 EDOUARD BATTEGAY, 2,3 MICHAEL STUMM, 3 MAAIKE VAN DER KOOIJ, 3 AND CORNEL C. SIEBER 1,3 Systemic and especially splanchnic arterial vasodilation accompany chronic portal hypertension. Different soluble mediators causing this vasodilation have been proposed, the strongest evidence being for nitric oxide (NO). No data exist if structural vascular changes may partly account for this vasodilatory state. Here, we developed a new in vivo quantitative angiogenesis assay in the abdominal cavity and determined if: 1) portal hypertensive rats show increased angiogenesis; and 2) angiogenesis is altered by inhibiting NO formation. Portal hypertension was induced by partial portal vein ligation (PVL). Sham-operated rats served as controls (CON). During the index operation (day 0), a teflon ring filled with collagen I (Vitrogen 100) was sutured in the mesenteric cavity. After 16 days, rings were explanted, embedded in paraffin, and ingrown vessels counted using a morphometry system. The role of NO was tested by adding an antagonist of NO formation (N -nitro-l-arginine [NNA], 3.3 mg/kg/d) into the drinking water. The mean number of ingrown vessels per implant was significantly higher in PVL rats compared with CON rats, i.e., 1, versus , respectively (P.05; N 5 per group). NNA significantly (P F.01) inhibited angiogenesis in PVL ( ; N 5) and in CON (174 25; N 6) rats, respectively. In contrast, the -adrenergic blocker, propranolol, did not prevent angiogenesis either in PVL or CON rats in a separate set of experiments (data not shown). The conclusions drawn from this study are that: 1) rats with portal hypertension show increased angiogenesis; and 2) inhibition of NO formation significantly prevents angiogenesis in both PVL and CON rats. Therefore, splanchnic vasodilation in chronic portal hypertension may also be a result of structural changes. (HEPATOLOGY 1999;29: ) Systemic and especially splanchnic vasodilation is the first step leading to the hyperdynamic circulation observed in chronic portal hypertension. 1 Different pathophysiological Abbreviations: PVL, partial portal vein ligation; CON, control rat; NNA, N -nitro-larginine; NDT, newly developed tissue; DV, detail view. From the 1 Division of Gastroenterology, 2 Medical Outpatent Clinic, and 3 Department of Research, University Hospital Basel, Basel, Switzerland. Received August 12, 1998; accepted December 1, Presented in part at the American Association for the Study of Liver Diseases Meeting, Chicago, IL, November 1997, and published in abstract form (HEPATOLOGY 1997;26: 943A). Supported by a Swiss National Science Foundation grant ( ). Cornel C. Sieber ( ) and Edouard Battegay ( ) are both recipients of a SCORE career development grant by the Swiss National Science Foundation. Address reprint requests to: Cornel C. Sieber, M.D., Division of Gastroenterology, University Hospital, CH-4031Basel, Switzerland. sieber@tmr.ch; fax: Copyright 1999 by the American Association for the Study of Liver Diseases /99/ $3.00/0 mechanisms have been postulated to be involved in this vasodilatory state, such as decreased responses to endogenous circulating vasopressors and/or increased levels of circulating vasodilators. 2 The endothelial- and neuralderived, locally acting vasodilator, nitric oxide (NO), 3 has been shown both in vivo 4,5 and in vitro 6-9 to play a major role in this vasodilation. Nevertheless, these functional alterations may not completely explain the sustained vasodilation observed. It is conceivable to postulate that besides functional changes, structural vascular changes also accompany chronic portal hypertensive states. Several lines of evidence seem to support such a hypothesis. We and others could previously show that maximally dilated in vitro perfused superior mesenteric arterial vascular beds of both prehepatic 6,7 and hepatic 8,9 portal hypertensive animals show significantly lower baseline perfusion pressures. Chronic treatment of normotensive rats with the vasodilator, minoxidil, inducing a long-term systemic vasodilation, has also been demonstrated to lead to structural changes in mesenteric arterioles. 10 In addition, many hemodynamic features of spontaneous hypertensive rats are reciprocal to those observed in portal hypertensive ones. It is tempting to speculate that this may also be true for structural vascular abnormalities, which are well described in mesenteric vessels of spontaneously hypertensive animals. 11 As stated above, NO plays a crucial role in the hemodynamic changes observed under portal hypertensive conditions. Besides its potent activity as a local vasodilator, 3 NO is 1044 an important angiogenic molecule and is involved in vascular remodeling. 15,16 To date, the quantification of angiogenesis in animal models has mostly been performed either in the chick embryo chorioallantoic membrane 17 or in the cornea 18 of different species. The drawbacks of these systems lie in the fact that the mesenteric hemodynamic alterations, thus the vascular area pertinent for the splanchnic vasodilation in portal hypertension, cannot quantitatively be studied. This was the background for us to develop and validate a new in vivo quantitative angiogenesis assay to study new vessel formation within the splanchnic area in normal and especially portal hypertensive rats. The aims of this set of experiments were therefore to: 1) establish a new in vivo quantitative angiogenesis assay in rats; 2) study if chronic portal hypertension is accompanied by structural vascular changes; and 3) explore if such changes may be modified by inhibiting NO biosynthesis. MATERIALS AND METHODS Rat Model of Portal Hypertension. The experiments were performed in male Sprague-Dawley rats (BRL Laboratories, Bubendorf, Switzer-

2 HEPATOLOGY Vol. 29, No. 4, 1999 SUMANOVSKI ET AL FIG. 1. A teflon ring (actually filled with collagen I) within the polyester mesh, which then is sutured into the mesentery. land) weighing 250 to 300 g at the time of the operation. Induction of portal hypertension was performed by partial portal vein ligation (PVL), as described elsewhere. 1,6 Briefly, overnight-fasted rats were anesthetized with nembutal (pentobarbital sodium, 40 mg/kg body weight). After a midline abdominal incision, the portal vein was freed from the surrounding tissue. A ligature (silk gut 3-0) was placed around a 20-gauge blunt-tipped needle lying alongside the portal vein. Subsequent removal of the needle yielded a calibrated stenosis of the portal vein. In controls (CON), the same operation was performed, except that after isolating the portal vein, no ligature was placed. All studies were performed 16 days after induction of portal hypertension, a period when the hyperdynamic circulation is known to be fully established. 1 Experiments were conducted according to the Guide for the Care and Use of Laboratory Animals. In Vivo Quantitative Angiogenesis Assay. Teflon rings (external diameter: 7 mm; internal diameter: 5 mm; height: 3 mm) were inserted into custom-tailored polyester mesh bags (PE 245 HS; Lanz-Anliker AG, Rohrbach, Switzerland) that were closed with a single-knot suture (Fig. 1). Subsequently, the implants were washed thoroughly in 70% ethanol, rinsed in deionized water, air-dried, and depyrogenized for 4 hours at 185 C. The following steps were all performed under strictly sterile conditions. Implants were laid on top of small teflon plates placed in Petri dishes. Afterward, implants were filled with bovine collagen I (Vitrogen 100, Collagen GmbH, Ismaning, Germany) to provide a matrix for the ingrowing tissue. Bovine collagen I was prepared with some modifications as described elsewhere. 17 Briefly, bovine collagen I was neutralized to a ph of 7.4 by the following steps: 80% collagen 10% 10 phosphatebuffered saline 10% 0.1 mol/l NaOH. Neutralized bovine collagen I was kept on ice throughout sample preparation to prevent premature gelation. Two thirds of the neutralized bovine collagen I solution was mixed with one third of 0.25% bovine serum albumin in phosphate-buffered saline. By using a 1-mL syringe with a 25-gauge needle, 80 µl of this gel solution was injected into each implant, carefully avoiding the generation of air bubbles. To maintain constant humidity, 8 ml of sterile water was added to each Petri dish. To allow gelation, implants in covered Petri dishes were transferred to an incubator at 37 C and 5% CO 2 for at least 1 hour. NO Biosynthesis Inhibition and Angiogenesis. To study the role of NO formation on angiogenesis, experimental groups of PVL and CON rats received N -nitro-l-arginine (NNA) (Sigma Ag, Buchs, Switzerland), an antagonist of NO formation, 19 into the drinking water at a concentration of 3.3 mg/kg/d. This dose has previously been shown to prevent the hyperdynamic circulation in portal hypertensive rats without having any significant hemodynamic effects in normal control animals. 20 -Adrenergic Blockade and Angiogenesis. In an attempt to explore if hemodynamic modulation interferes with angiogenesis, we studied separate groups of PVL and CON rats treated with and without the -adrenergic blocker, propranolol, at a dose of 10 mg/kg/d, known to effectively modulate hemodynamics in portal hypertensive rats. 21,22 Experimental Design. At day 0, rats were anesthetized and underwent surgery (PVL or sham operation) as described above. At the same time point, all PVL and CON rats received an implant, which was put between two mesenterial membranes and loosely sewed with single-knot sutures. After the operation, animals were earmarked for later identification. After 16 days, rats were killed. This time point was chosen for different reasons: Firstly, this is around the time when most hemodynamic studies are performed by different groups when using this model of portal hypertension. 1,4-8 Secondly, Nguyen et al. used the same time frame when studying angiogenesis in the chick embryo chorioallantoic membrane. 17 Thirdly, time-sequence studies with this new angiogenesis assay described here showed that until day 10, nearly no vessel formation could be observed within the teflon rings both in normal and portal hypertensive rats (data not shown). The implants were carefully removed, washed in 0.9% saline, and fixed in 4% formalin for 4 hours at room temperature. After fixation, implants were briefly rinsed in phosphate-buffered saline and paraffinized afterward. From each implant, 3-µm thin tissue sections were cut on a microtome (Medite AG, Nunningen, Switzerland), routinely stained with hematoxylin-eosin, and subsequently checked for quality under the microscope. The very first tissue sections from each implant were excluded from analysis because of altered morphology as a result of the polyester net. Only tissue sections cut at the same position in each implant were selected for further analysis. To further confirm the existence of vessels within the implants, we immunostained tissue sections with either rabbit polyclonal factor VIII (DAKO Diagnostics AG, Zug, Switzerland) (Fig. 2A), or with rabbit polyclonal flt-1 (Santa Cruz Biotechnology, Dr. Glaser AG, Basel, Switzerland) (Fig. 2B), known to be specific markers for endothelial cells. Videomorphometry. All tissue sections were analyzed by the same investigator (L.S.), who was blinded with regard to the pretreatment of the animal from which the sections originated. Representative hematoxylin-eosin stained tissue sections from each implant were examined using a microscope (Nikon) equipped with a videocamera (Panasonic F15 HS) and a videorecorder (Panasonic AG 7330). Only the newly developed tissue (NDT) inside the teflon ring was analyzed. First, overviews of NDT were filmed at 20 magnification. Subsequently, every adjacent lying detail view (DV) within the longest horizontal and vertical diameter of the NDT was recorded at 400 magnification. For further quantitative analysis, image data were transferred to a computer equipped with commercial image analysis software (MicroScale TC/TM, Version 2.2, Digithurst Ltd., Herts, UK). The following procedure was adopted: In a first step, areas of DVs were calculated. In a second step, all vessels were counted within every second DV, representing a minimum of 7% of the total NDT area, the mean number of DVs analyzed in each implant being Vessels were recognized histologically and counted only when they were cross-sectioned and showed an uninterrupted endothelial cell layer. In a third step, estimation of the total number of vessels ingrown into the NDT of one individual implant was performed by the following calculation: the ratio of NDT area to DV area was determined for each implant and subsequently multiplied by the arithmetic mean of the number of vessels counted in selected DVs. Statistical Analysis. Data are given as mean SEM. An unpaired Student t test was used to compare the different groups. A Bonferroni correction was applied for multiple comparisons. P.05 was considered statistically significant.

3 1046 SUMANOVSKI ET AL. HEPATOLOGY April 1999 FIG. 2. Representative stainings of tissue sections using polyclonal antibodies to the endothelial markers, factor VIII (A) and flt-1 (B), in a portal hypertensive animal. An impressive number of new vessels develops within 16 days in the teflon ring filled with collagen I as a matrix. RESULTS Baseline characteristics of the different studied groups are given in Table 1. Body weights were comparable between the eight studied groups both at the time of the index operation (day 0) as well as 16 days later, when the implants were analyzed (Table 1). After implantation of the teflon rings, the rats behaved normally and gained weight in the order of other rats without operation. Furthermore, there was no sign of an abscess or peritonitis in any of the animals at the time of explantation of the rings. Others 1,4,5 as well as our group 6,7,23,24 have previously used this model of prehepatic portal hypertension. The advantage of this model is that it provides a very reproducible portal hypertensive state. In a subset of animals (N 6 CON, N 7 PVL), we determined portal pressure by the splenic pulp pressure method 25,26 using a pressure transducer and a registration on a polygraph. Mean splenic pulp pressure corresponding to portal pressure was and mm Hg in CON and PVL animals, respectively (P.001). The impressive development of new vessels within the teflon ring is exemplified by staining endothelial cells with factor VIII (Fig. 2A), as well as flt-1 (Fig. 2B). Furthermore, a matrix of fibroblast and mononuclear cells besides the new vessel formation developed in the implants within the 16 days (Figs. 2 and 3). Number of Ingrown Vessels. Compared with CON (N 5), the number of ingrown vessels was significantly higher in PVL (N 5) rats, the mean number per implant being 1, for PVL rats and for CON rats,

4 HEPATOLOGY Vol. 29, No. 4, 1999 SUMANOVSKI ET AL TABLE 1. Baseline Characteristics of the Different Studied Groups Groups Number of Animals Portal Hypertension (yes/no) Treatment With NNA (yes/no) Mean Weight at Operation ( SD) Mean Weight at Analysis of Implants ( SD) Effect of NO-biosynthesis inhibition with NNA on angiogenesis* 1 5 no no yes no no yes yes yes Effect of -adrenergic blockade with propranolol on angiogenesis 1 6 no no yes no no yes yes yes *Animals treated with NNA received a daily dose of 3.3 mg/kg/d in the drinking water. Animals treated with propranolol received a daily dose of 10 mg/kg/d in the drinking water. respectively (P.05) (Fig. 4). Representative examples of a DV for both groups are given in Fig. 3A and 3B. Role of Chronic NO-Formation Inhibition on Angiogenesis. Inhibition of NO biosynthesis with the stereospecific antagonist, N -nitro-l-arginine (NNA), 19 significantly (P.01) reduced angiogenesis both in PVL (N 5) and CON (N 6) rats. The respective mean numbers of vessels per implant were (NNA) compared with 1, (no NNA) for PVL rats. For CON animals, the numbers were (NNA) compared with (no NNA), respectively (Fig. 4). Altogether, NNA prevented new vessel formation by more than two thirds both in normal and portal hypertensive animals. Again, representative examples for DVs are given in Fig. 3C and 3D. FIG. 3. Representative cuts (stained with hematoxylin-eosin) of mesenteric implants in chronic PVL (A) and CON (B) rats. PVL rats show significantly (P.05) increased angiogenesis. Furthermore, mesenteric implants after chronic inhibition of NO biosynthesis for PVL (C) and CON (D) rats are presented. A significantly (P.01) reduced angiogenesis is seen both in PVL (C) and CON (D) rats when compared with animals without NO-formation inhibition (A, B), respectively.

5 1048 SUMANOVSKI ET AL. HEPATOLOGY April 1999 FIG. 4. The total number of vessels (mean SEM) is given without ( ) and with ( ) chronic NO-formation inhibition with NNA. Implants from portal hypertensive animals (O) show a significantly (*) increased vessel formation compared with controls ( ). Furthermore, chronic NObiosynthesis inhibition significantly (#) inhibits new vessel formation in both groups. Role of -Adrenergic Blockade on Angiogenesis. Propranolol partly prevented angiogenesis in PVL rats. The mean number of vessels per implant was (N 8) without and (N 9) with propranolol. This difference did not reach statistical significance (P.07). Furthermore, propranolol did not modify angiogenesis in CON rats, the numbers being (N 6) without and (N 5) with propranolol. Finally, as already shown above for the other set of experiments, PVL rats ( ) showed a significantly increased number of vessels when compared with CON (405 65) rats, respectively (P.05). DISCUSSION Chronic portal hypertension goes along with a hyperdynamic circulatory state, initiated by a systemic and especially splanchnic vasodilation. 1 Research to date has been focused on putative functional mediators of these changes. Besides the classically described potential role of a diminished response to both endogenous vasopressors or increased levels of circulating vasodilators, 2 we 6-8 and others 4,5,9 have shown an important function of the locally acting vasodilator, NO, 3 as a mediator of the sustained vasodilation and vascular hyporesponsiveness. Besides endogenously applied vasoactive substances, neurally mediated vasomotion is also altered in the splanchnic bed in chronic portal hypertensive rats in a NO-dependent fashion. 23,24 Significantly increased baseline perfusion pressures have been reported in vessel preparations of spontaneous hypertensive rats. 27 Reciprocal to this finding, significantly lower baseline perfusion pressures are found in vessel preparations of portal hypertensive animals. 6-9,23,24 As in this system, vessels are maximally dilated as tested beforehand 6 ; we speculated that structural vascular changes including angiogenesis may account for this observation in portal hypertensive rats. To date, most studies exploring angiogenesis have used either the chick embryo chorioallantoic membrane 17 or the avascular cornea, e.g., of rabbits. 18 Even these techniques are effective in selectively studying the effect of angiogenic molecules; they do not really correspond to an in vivo situation with an intact circulation. For our purpose, to study the effect of chronic experimental portal hypertension on vascular structure, we needed to study angiogenesis in a reproducible quantitative way under in vivo conditions in the abdomen, where the vasodilation in chronic portal hypertension is especially prominent. 1 To accomplish this task, we developed a new technique with inert teflon rings filled with collagen as a matrix for the ingrowth of new vessels. This system lacks any biological tissue when being implanted into the abdominal cavity. Every biological tissue including vessels observed is therefore newly formed. The mean number of vessels within the implants was significantly increased in portal hypertensive compared with normal rats. This finding confirms our assumption that chronic portal hypertension induces structural, as well as the well-described functional, vascular changes. The finding of increased angiogenesis also parallels data in the gastric microvasculature of portal hypertensive rats. 28 Finally, it also fits with data that chronic systemic arterial hypotension induces structural vascular changes in splanchnic vessels. 10 Nevertheless, in vascular casts of gastric microvessels, significantly narrowed microvessels and a smaller amount of vessel buds as a sign of angiogenesis after toxic ethanol injury have been documented in portal hypertensive rats, when compared with control animals. 29 The authors of this study concluded that chronic portal hypertensive rats have an impairment of angiogenic responses. Even these findings somewhat contradict the results shown here; differences in the vascular bed studied, as well as the noxious stimulus in the other study, may partly explain this discrepancy. Besides its functional vasodilatory role, NO itself has recently been shown to be a potent angiogenic molecule Chronic inhibition of NO biosynthesis by NNA in a dose previously demonstrated to prevent the development of the hyperdynamic circulation in portal hypertensive rats without inducing significant hemodynamic changes in control rats 20 highly significantly prevented angiogenesis in PVL and CON rats in the present study. In fact, the mean number of new vessel formation after NO biosynthesis inhibition was practically identical in PVL rats compared with CON rats. To further distinguish if the antiangiogenic effect of NOformation inhibition is a result of the direct angiogenic potential of NO or is more indirectly linked to the hemodynamic alterations, we tested the -adrenergic blocker, propranolol, in separate experimental series. Propranolol partly prevented angiogenesis in PVL rats, but had no effect at all in CON rats, in which NO-formation inhibition significantly inhibited new vessel formation. It therefore seems that NO s exertion of its angiogenic stimulation is at least partly distinct from its vasodilatory action. The systemic and especially splanchnic vasodilation is the initial event leading to the hyperdynamic circulation in chronic portal hypertension, a major determinant of increased portal pressure. 30 Nevertheless, the development of a

6 HEPATOLOGY Vol. 29, No. 4, 1999 SUMANOVSKI ET AL collateral circulation is also an important feature, especially in this model of portal hypertension. 1,31,32 It is unknown inasmuch these collaterals are an opening of pre-existing vessels or even in this vascular bed are a result of new vessel formation, i.e., angiogenesis. Future studies using appropriate quantitative methods should also address the question of stimulated angiogenesis in these arterialized vessels. Nevertheless, as an impressive angiogenic potential was seen here in normal rats without a developed collateral vascular bed, as observed in chronic portal hypertension, this angiogenic process does not seem directly related to the collateralization of vessels in disease states such as chronic portal hypertension. In summary, we have developed and validated a new quantitative in vivo angiogenesis assay allowing for the exploration of angiogenic mechanisms in the splanchnic cavity. This new angiogenesis assay was then used to test the angiogenic potential in a disease state: chronic portal hypertension. By doing so, we could show that chronic portal hypertension is paralleled by significantly increased splanchnic angiogenesis, which may partly explain the splanchnic vasodilation in chronic portal hypertensive animals. Furthermore, this increase in angiogenesis could be prevented by chronically inhibiting NO biosynthesis. REFERENCES 1. Sikuler E, Kravetz D, Groszmann RJ. Evolution of portal hypertension and mechanisms involved in its maintenance in a rat model. Am J Physiol 1985;248:G618-G Groszmann RJ. Vasodilatation and hyperdynamic circulatory state in chronic liver diseases. In: Portal Hypertension. Pathophysiology and Treatment. Bosch J, Groszmann RJ, eds. Oxford: Blackwell Scientific Publications, 1994: Moncada S, Palmer MJ, Higgs EA. Nitric oxide: physiology, pathology and pharmacology. Pharmacol Rev 1991;43: Lee FY, Albillos A, Colombato L, Groszmann RJ. The role of nitric oxide in the vascular hyporesponsiveness to methoxamine in portal hypertensive rats. HEPATOLOGY 1992;16: Pizcueta MP, Piqué JM, Bosch J, Whittle BJR, Moncada S. Effects of inhibiting nitric oxide biosynthesis on the systemic and splanchnic circulation of rats with portal hypertension. Br J Pharmacol 1992;105: Sieber CC, Groszmann RJ. In vitro hyporeactivity to methoxamine in portal hypertensive rats: reversal by nitric oxide blockade. Am J Physiol 1992;262:G996-G Sieber CC, Groszmann RJ. Nitric oxide mediates the in vitro hyporeactivity to vasopressors in mesenteric vessels of portal hypertensive rats. Gastroenterology 1992;103: Sieber CC, Lopez-Talavera JC, Groszmann RJ. Role of nitric oxide in the in vitro splanchnic vascular hyporeactivity in ascitic cirrhotic rats. Gastroenterology 1993;104: Claria J, Jimenez W, Ros J, Rigol M, Angeli P, Arroyo V, Rivera F, et al. Increased nitric oxide dependent vasorelaxation in aortic rings of cirrhotic rats with ascites. HEPATOLOGY 1994;20: Tsoporis J, Fields N, Lee RMKW, Leenen FHH. Arterial vasodilation and cardiovascular structural changes in normotensive rats. Am J Physiol 1991;260:H1944-H Lee RMKW, Garfield RE, Forrest JB, Daniel EE. Morphometric study of structural changes in the mesenteric blood vessels of spontaneously hypertensive rats. Blood Vessels 1986;20: Ziche M, Morbidelli L, Masini E, Amerini S, Granger HJ, Maggi CA, Geppetti P, et al. Nitric oxide mediates angiogenesis in vivo and endothelial cell growth and migration in vitro promoted by substance P. J Clin Invest 1994;94: Ziche M, Morbidelli L, Choudhuri R, Zhang HT, Donnini S, Granger HJ. Nitric oxide synthase lies downstream from vascular endothelial growth factor induced but not basic fibroblast growth factor induced angiogenesis. J Clin Invest 1997;99: Montrucchio G, Lupia E, de Martino A, Battaglia E, Arese M, Tizzani A, Bussolino F, et al. Nitric oxide mediates angiogenesis induced in vivo by platelet-activating factor and tumor necrosis factor-. Am J Pathol 1997;151: Tronc F, Wassef M, Esposito B, Henrion D, Glagov S, Tedgui A. Role of NO in flow-induced remodeling of the rabbit common carotid artery. Arterioscler Thromb Vasc Biol 1996;16: Radu RD, Shesely EG, Maeda N, Smithies O, Segal SS, Sessa WC. Direct evidence for the importance of endothelium-derived nitric oxide in vascular remodeling. J Clin Invest 1998;101: Nguyen M, Shing Y, Folkman J. Quantitation of angiogenesis and antiangiogenesis in the chick embryo chorioallantoic membrane. Microvasc Res 1994;47: Folkman J, Shing Y. Angiogenesis. J Biol Chem 1992;267: Ishii K, Chang B, Kerwin JF, Huang Z, Murad F. N -nitro-l-arginine: a potent inhibitor of endothelium-derived relaxing factor formation. Eur J Pharmacol 1990;176: Lopez-Talavera JC, Sieber CC, Lee FY, Groszmann RJ. Chronic oral administration of N -nitro-l-arginine (NNA) ameliorates the development of the hyperdynamic circulation and portal-systemic shunting in portal hypertensive rats. Gastroenterology 1993;104:A Kroeger R, Groszmann RJ. The effect of the combination of nitroglycerin and propranolol on splanchnic and systemic hemodynamics in a portal hypertensive rat model. HEPATOLOGY 1985;5: Polio J, Sieber CC, Lerner E, Groszmann RJ. Cardiovascular hyporesponsiveness to norepinephrine, propranolol and nitroglycerin in portalhypertensive and aged rats. HEPATOLOGY 1993;18: Sieber CC, Sumanovski LT, Moll-Kaufmann C, Stalder GA. Hyposensitivity to nerve stimulation in portal hypertensive rats: role of nitric oxide. Eur J Clin Invest 1997;27: Moll-Kaufmann C, Sumanovski LT, Sieber CC. Neurally-mediated vasodilatation in normal and portal hypertensive rats: role of nitric oxide and calcitonin gene-related peptide. J Hepatol 1998;28: Chojkier M, Groszmann RJ. Measurement of portal-systemic shunting in the rat by using -labeled microspheres. Am J Physiol 1981;240:G371- G Cordoba J, Dupuis J, Gottstein J, Blei AT. Stenosis of a portacaval anastomosis affects circadian locomotor activity in the rat: a multivariable analysis. Am J Physiol 1997;273:G1218-G Criscione L, Nellis P, Riniker B, Thomann H, Burdet R. Reactivity and sensitivity of mesenteric vascular beds and aortic rings of spontaneously hypertensive rats to endothelin: effects of calcium entry blockers. Br J Pharmacol 1990;100: Albillos A, Colombato LA, Enriquez R, Ng OC, Sikuler E, Groszmann RJ. Sequence of morphological and hemodynamic changes of gastric microvessels in portal hypertension. Gastroenterology 1992;102: Ichikawa Y, Tarnawski A, Safreh IJ, Ishikawa T, Shimada H. Distorted microangioarchitecture and impaired angiogenesis in gastric mucosa of portal hypertensive rats. Gastroenterology 1994;106: Albillos A, Colombato LA, Groszmann RJ. Vasodilatation and sodium retention in prehepatic portal hypertension. Gastroenterology 1992;102: Lee FY, Colombato LA, Albillos A, Groszmann RJ. Administration of N -nitro-l-arginine ameliorates portal-systemic shunting in portalhypertensive rats. Gastroenterology 1993;105: Mosca P, Lee FY, Kaumann AJ, Groszmann RJ. Pharmacology of portal-systemic collaterals in portal hypertensive rats: role of the endothelium. Am J Physiol 1992;263:G544-G550.