Journal of Hepatology 43 (2005) 791–798 www.elsevier.com/locate/jhep
Neutral endopeptidase (EC 3.4.24.11) in cirrhotic liver: A new target to treat portal hypertension?*,** Giovanni Sansoe`1,*, Manuela Aragno2, Raffaella Mastrocola2, Francesca Restivo2, Giulio Mengozzi3, Antonina Smedile4, Floriano Rosina1, Oliviero Danni2, Maurizio Parola2, Mario Rizzetto4 1 Gastroenterology Unit, Gradenigo Hospital, Turin, Italy Department of Experimental Medicine and Oncology, University of Turin, Turin, Italy 3 Institute of General Pathology, Molinette Hospital, Turin, Italy 4 Department of Gastroenterology, Molinette Hospital, University of Turin, Turin, Italy 2
Background/Aims: In liver cirrhosis atrial natriuretic peptide (ANP) decreases portal vascular resistance and tributary flow. The enzyme neutral endopeptidase (NEP) degrades ANP and bradykinin and generates endothelin-1 from big-endothelin. We determined the effects of NEP inhibition by candoxatrilat on hormonal status, liver function and arterial and portal pressures in rats with CCl4-induced cirrhosis. Methods: Two groups of seven control rats received 1 ml 5% glucose solution alone or containing 10 mg/kg candoxatrilat; three groups of 10 ascitic cirrhotic rats received placebo, 5 or 10 mg/kg candoxatrilat. NEP protein concentration and immunostaining were analyzed in normal and cirrhotic livers. Results: In cirrhotic rats 10 mg/kg candoxatrilat significantly increased steady-state indocyanine green clearance (a parameter reflecting liver plasma flow) (P!0.01), decreased portal pressure (P!0.01), had no effect on arterial pressure and plasma renin activity but increased ANP plasma levels (P!0.05) and urinary excretions (P!0.01) of ANP and cGMP. In the cytosol fraction of rat cirrhotic livers a 280% increase in NEP content was found (P!0.01), chiefly localized in desmin-positive myofibroblast-like cells of fibrous septa. Conclusions: Candoxatrilat has few effects on systemic hemodynamics and hormonal status; its portal hypotensive action depends on effects exerted on intrahepatic vascular resistance. q 2005 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. Keywords: Liver cirrhosis; Portal pressure; Animal model; Atrial natriuretic peptide; Neutral endopeptidase; Candoxatrilat; Liver myofibroblast-like cells
Received 23 November 2004; received in revised form 20 February 2005; accepted 5 April 2005; available online 20 June 2005 * This study was presented orally at the Research Forum on Portal Hypertension which took place at the 2004 Digestive Disease Week (DDW), New Orleans, LA, USA, May 2004. It was also presented in poster form at the 2004 annual meeting of the American Association for the Study of Liver Diseases (AASLD), Boston, MA, USA, November 2004. ** The authors who have taken part in this study declared that they have not a relationship with the manufacturers of the drugs involved either in the past or present and did not receive funding from the manufacturers to carry out their research. * Corresponding author. Address: Gastroenterology and Hepatology, Room 9N/983, Toronto General Hospital, 200 Elizabeth Street, Toronto, ON, M5G 2C4 Canada. Tel.: C1 416 816 2545; fax: C1 416 340 5019. E-mail address:
[email protected] (G. Sansoe`). Abbreviations: ANP, atrial natriuretic peptide; AVP, arginine vasopressin; BNP, brain natriuretic peptide; c-GMP, guanosine 3 0 ,5 0 -cyclic monophosphate; CICG, indocyanine green plasma clearance; CNP, C-type natriuretic peptide; ET-1, endothelin-1; HSC, hepatic stellate cells; ICG, indocyanine green; IL-1, interleukin-1; IL-6, interleukin-6; MFs, liver myofibroblast-like cells; NEP, neutral endopeptidase; NPCR, natriuretic peptide clearance receptor; PRA, plasma renin activity; RAP, receptor-associated protein; TGF-b, transforming growth factor-b; TNF-a, tumor necrosis factor-a. 0168-8278/$30.00 q 2005 European Association for the Study of the Liver. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.jhep.2005.04.017
792
G. Sansoe` et al. / Journal of Hepatology 43 (2005) 791–798
1. Introduction In liver cirrhosis, augmented portal tributary blood flow secondary to splanchnic vasodilatation, and increased intrahepatic vascular resistance determine portal hypertension [1]. Endogenous natriuretic peptides inhibit sympathetic nervous system, renin-angiotensin-aldosterone axis, and arginine vasopressin (AVP) secretion, leading to vasodilatation and natriuresis [2]. When infused in cirrhotic patients, natriuretic peptides decrease hepatic venous pressure gradient, i.e. portal pressure [3]. In cirrhotic rats the infusion of brain natriuretic peptide (BNP) and ring-deleted ANP analogues reduced portal tributary blood flow [4,5], but either physiological or pharmacological doses of ANP did not affect intrahepatic vascular resistance [6]. Intraportal injection of ANP did not alter portal flow resistance in anaesthetized dogs [7]. However, in cirrhotic patients, ANP infusion reduced portal pressure and increased hepatic blood flow, a finding indicative of ANP-induced decrease in intrahepatic resistance to portal flow [3]. Finally, recent data suggest that C-type natriuretic peptide (CNP) may counteract both liver fibrogenesis and portal hypertension in chronic liver diseases [8]. Since ANP administration causes systemic vasodilatation, decrease in cardiac index, and hypotension in normal subjects [9] and, to a larger extent, in patients with cirrhosis [10,11], the use of natriuretic peptides as portal hypotensive agents is not recommended. Clearance of natriuretic peptides occurs mainly in the kidney, lung, brain, and heart through binding to natriuretic peptide clearance receptors (NPCRs) or proteolysis by neutral endopeptidase (NEP) [12], a membrane-bound Zn-metalloendopeptidase also called atriopeptidase or EC 3.4.24.11. NEP has been found in the brush border membrane of kidney tubule cells, in bronchial fibroblasts and epithelial cells, in B lymphoid-progenitors, and in glial cells [13]. NEP degrades opioid-peptides [14], bradykinin [15], bombesin-like peptide [16], substance P [17], endogenous natriuretic peptides [18,19], and adrenomedullin [20], as well as endothelin-1 (ET-1) [21] and angiotensin II [17]. NEP can produce the vasoconstrictor polypeptide ET-1 from circulating precursors (i.e. big ET-1 and ET-11– 31) [22,23]: the balance between degradation and generation of ET-1 by NEP depends on whether the predominant substrates available for proteolysis are vasodilators, vasoconstrictors, or precursors of vasoconstrictors [24]. NEP is inhibited by candoxatrilat and its prodrug candoxatril, by thiorphan, sinorphan, and phosphoramidon [17,25,26]. Although NEP inhibitors increase plasma concentrations of natriuretic peptides and cause natriuresis [25,26] they do not lower arterial pressure in normotensive subjects [26]. Indeed, both candoxatril and candoxatrilat may raise arterial pressure in healthy humans [27]. NEP inhibitors have been reported to lower blood pressure in patients with essential hypertension [28], a finding later disputed [29].
This study aimed: (a) to analyze NEP protein content and immuno-localization in normal and cirrhotic rat livers and (b) to determine the effects of i.v. candoxatrilat on hemodynamics (i.e. portal pressure and mean arterial pressure), liver function (measured through indocyanine green clearance [CICG]), and hormonal status in cirrhotic rats with ascites induced by chronic administration of CCl4.
2. Methods Studies were performed on anaesthetized male adult Wistar rats with cirrhosis and ascites, and on anaesthetized male adult Wistar control rats. Both groups were fed ad libitum with standard chow and water as drinking fluid. Cirrhosis was induced by CCl4 (Riedel de Haen, Sigma-Aldrich, Seelze, Germany), which was administered by gavage twice weekly [30]. Cirrhotic rats were studied 12 weeks after starting the cirrhosis induction program, when ascites was developed. Control rats were studied following a similar period of standardized diet. Experiments in rats were performed in compliance with the procedures outlined in the Italian Ministry of Health guidelines (no. 86/609/EEC) and according to the Principles of Laboratory Animal Care (NIH no. 85-23, revised in 1985). The protocol assessed the urinary, hormonal and portal hemodynamic responses to the i.v. administration of different doses of the NEP inhibitor candoxatrilat (Pfizer Central Research, Sandwich, Sussex, England) [25], in 14 control rats and 30 rats with CCl4-induced ascitic cirrhosis.
2.1. Animal groups Candoxatrilat was dissolved in 5% glucose solution, obtaining two different doses, F5 (5 mg/kg) or F10 (10 mg/kg), to be administered intravenously to the rats in the same volume of fluid (1 ml). The rats were divided into five groups: seven control rats receiving diluent alone (G1), seven control rats receiving F10 (G2), 10 cirrhotic rats receiving diluent alone (G3), 10 cirrhotic rats receiving F5 (G4), and 10 cirrhotic rats receiving F10 (G5).
2.2. Study protocol The rats were anaesthetized with a mixture of Ketavet 100 (Farmaceutici Gellini, Sabaudia, Italy) and Rompum (Xylazine, Bayer A.G., Leverkusen, Germany) (4:1 v:v) by intraperitoneal injection (0.5 ml mixture/200 g b.wt). A constant infusion of indocyanine green (ICG) (Cardiogreen; Hyson Wescott and Dunning, Baltimore, USA) was carried out through the caudal vein for 180 min, in order to assess its steady-state plasma clearance at different times [31,32]. The ICG infusion rate was 5 mg/ min (1 ml of ICG-containing solution infused per hour). After 90 min of infusion, laparotomy was performed and the urinary bladder was emptied; a clamp was positioned on the urethral orifice. Cardiac blood was sampled by cardiac puncture (0.5 cc) (time 1) to assess basal values of CICG. In order to directly measure portal pressure, a polypropylene catheter (0.7 mm diameter) was inserted into a small ileal vein and gently advanced to the bifurcation of the superior mesenteric and the splenic veins. After 10-min stabilization, basal portal pressure was measured. Then, either diluent alone or candoxatrilat (F5 or F10) was given as a single bolus in the right femoral vein and portal pressure measurements were repeated after 5 and 15 min. Cardiac blood was sampled (0.5 cc) 15 and 30 min (times 2 and 3) after candoxatrilat or vehicle injection. Each blood withdrawal was replaced with an equal volume of i.v. saline. Blood samples withdrawn at time 2 were used to measure the new ICG steady-state plasma levels. Blood samples at time 3 were used for hormonal status determination. Blood samples withdrawn at time 3 from five controls and five cirrhotic rats infused only with vehicle were used to determine tumor necrosis factor-a (TNF-a) and interleukin-1 (IL-1) serum levels. One hour after candoxatrilat or vehicle injection, all rats were killed by abdominal aorta exsanguination, and all urine produced during the 60 min period after candoxatrilat or vehicle administration was drained through bladder puncture. This urine volume was used to determine hourly excretions of sodium, potassium,
G. Sansoe` et al. / Journal of Hepatology 43 (2005) 791–798 ANP and cGMP. Livers were excised and immediately frozen at K80 8C for bio-molecular determinations. In a further group of five anaesthetized cirrhotic rats, mean arterial pressure was evaluated by means of tail sphygmomanometry (Blood Pressure Recorder 8005, WCW Electronic, Milan, Italy) before and 40 min after 10 mg/kg candoxatrilat i.v. administration in the caudal vein, without performing laparotomy. The blood pressure recorder was equipped with a high sensitivity pulse transducer coupled with a microprocessor program which allowed to distinguish between systolic and diastolic blood pressure, as described in literature [33,34].
793
measured using RIA (ANP Shionoria, Cis Bio International, Gif-sur-Yvette Cedex, France). Plasma renin activity (PRA) was determined using RIA for angiotensin I after an incubation period of 2 h (Renin Maia Kit, Biodata, Rome, Italy). Endothelin-1 plasma concentrations were assayed with a commercially available kit (Endothelin-1 Radioimmunoassay, Peninsula laboratories, King of Prussia, PA, USA). Urinary concentrations of cGMP were determined using RIA (Immunotech, Marseille, France).
2.7. Calculations
2.3. NEP protein concentration in rat liver
ICG clearance was calculated by the steady-state plasma clearance formula:
For Western blot analysis, tissue specimens were prepared from the livers of five rats from each experimental group (G1–G5): 500 mg slices were homogenized at 10% (w/v) in Tris-buffer and centrifuged at 1000 g for 10 min at 4 8C. The supernatants were centrifuged at 100,000 g for 45 min; the second supernatants were the cytosolic fraction. NEP was detected on 10% homogenates and cytosolic extracts, as described by Laemmli et al. [35]. Equal amounts of proteins (60 mg) were separated on 10% SDS-polyacrylamide gels, and blotted on nitrocellulose membranes. Non-specific binding was blocked with 5% (w/v) non-fat dried milk in 5 mM Tris–HCl, pH 7.4, containing 200 mM NaCl and 0.05% (v/v) Tween 20 (TBS-Tween) overnight at 4 8C. Blots were incubated with a rabbit polyclonal antibody raised against NEP (CD10, Santa Cruz Biotechnology) and then with a goat polyclonal antibody against the receptor associated protein (RAP, Santa Cruz Biotechnology) diluted at 1:1000 in TBS-Tween containing 2% (w/v) non-fat dried milk for 1 h at room temperature. After washing three times for 10 min in TBS-Tween, blots were reacted with peroxidase-labeled secondary antibody in TBS-Tween containing 2% (w/v) non-fat dried milk. Following three washing steps in TBS-Tween, immunoreactive proteins were detected with the chemiluminescence assay (ECL, Amersham) and subsequent exposure to film for 2–10 min. Specific bands were quantified by densitometry using analytic software (Bio-Rad, Multi-Analyst, Munich, Germany); before any comparison the net intensity of bands in each experiment was normalized to the intensity of the corresponding RAP band, used as an internal standard to evaluate the degree of non-specific protein expression in this homogenate [36].
CICG Z Infusion rate ðICGÞ=ssP ICG
2.4. Liver NEP indirect immunofluorescence Indirect immunofluorescence was performed in a humidified chamber at room temperature, essentially as described elsewhere [37], on liver cryostat sections (6 mm thick). Mouse monoclonal anti-desmin or anti-aSMA (DakoCytomation, Denmark) and rabbit polyclonal anti-NEP (CD10, Santa Cruz Biotechnology, Santa Cruz, California, USA) primary antibodies were diluted (1:250 and 1:100, respectively) in phosphate buffered saline, pH 7.4, containing 0.1% bovine serum albumin (PBS-BSA). As secondary antibodies either anti-mouse or anti-rabbit C3y-conjugated antibodies (Amersham Biosciences, Braunschweig, Germany), diluted 1:1000 in PBS-BSA, as well as FITC-conjugated anti-mouse antibodies (SigmaAldrich, Milan, Italy) diluted 1:200 in PBS-BSA, were used. At the end of the conventional immunofluorescent staining procedure, liver sections were washed twice in PBS and stained with the DNA fluorescent dye DAPI to visualize liver cell nuclei [38].
where ssP-ICG is the steady-state plasma concentration of ICG [31,32]. Mean arterial pressure (MAP) was calculated by the usual formula: 1=3ðsystolic blood pressure K diastolic blood pressureÞ C diastolic blood pressure:
2.8. Statistical analysis Results are expressed as meansGSD. All comparisons between groups of rats were made by Wilcoxon rank sum test. Comparisons between measurements carried out in the same animal but at different times were made by Wilcoxon signed rank test. Correlation coefficients were derived using Spearman’s rank correlation. Significance was accepted at the 5% probability level.
3. Results 3.1. Liver NEP protein expression We found over-expression of neutral endopeptidase in the whole homogenate from cirrhotic livers (Fig. 1); NEP protein content in cytosolic fraction of cirrhotic livers was roughly three-fold that of healthy animals (Fig. 2).
2.5. Serum cytokine determinations Serum TNF-a levels were determined using a sensitive ELISA kit (Biosource International, Camarillo, California, USA) specific for rat TNFa, according to the manufacturer’s instructions. Serum IL-1 levels were evaluated using an ELISA kit (Maoyuan Technique Co, Shanghai, China).
2.6. Plasma and urinary determinations Urinary concentrations of sodium and potassium were measured with a flame photometer. ICG plasma concentrations were determined colorimetrically. Arginine vasopressin concentrations were measured on EDTA plasma using RIA (Vasopressin Direct RIA, Buhlmann Laboratories AG, Postfach, Switzerland). ANP urinary and plasma concentrations were
Fig. 1. Western blot of a representative experiment showing NEP content in control and cirrhotic rat livers (upper panel: whole liver homogenate 10% diluted in buffer; lower panel: cytosolic fraction of liver homogenate). NEP: neutral endopeptidase; RAP: receptor associated protein (see text). C: control; Cir: cirrhotic rat.
G. Sansoe` et al. / Journal of Hepatology 43 (2005) 791–798
794 Relative densities of band 140
P
