Am J Physiol Gastrointest Liver Physiol 309: G301–G309, 2015. First published July 2, 2015; doi:10.1152/ajpgi.00010.2015.

Metformin reduces hepatic resistance and portal pressure in cirrhotic rats Dinesh M. Tripathi,1,2* Eva Erice,1* Erica Lafoz,1 Héctor García-Calderó,1 Shiv K. Sarin,2 Jaime Bosch,1 Jordi Gracia-Sancho,1* and Juan Carlos García-Pagán1* 1

Barcelona Hepatic Hemodynamic Laboratory, Liver Unit, Hospital Clínic, Institut d’Investigacions Biomèdiques August Pi i Sunyer, and Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas, Barcelona, Spain; and 2 Institute of Liver and Biliary Sciences, New Delhi, India Submitted 12 January 2015; accepted in final form 27 June 2015

Tripathi DM, Erice E, Lafoz E, García-Calderó H, Sarin SK, Bosch J, Gracia-Sancho J, García-Pagán JC. Metformin reduces hepatic resistance and portal pressure in cirrhotic rats. Am J Physiol Gastrointest Liver Physiol 309: G301–G309, 2015. First published July 2, 2015; doi:10.1152/ajpgi.00010.2015.—Increased hepatic vascular resistance is the primary factor in the development of portal hypertension. Metformin ameliorates vascular cells function in several vascular beds. Our study was aimed at evaluating the effects, and the underlying mechanisms, of metformin on hepatic and systemic hemodynamics in cirrhotic rats and its possible interaction with the effects of propranolol (Prop), the current standard treatment for portal hypertension. CCl4-cirrhotic rats received by gavage metformin 300 mg/kg or its vehicle once a day for 1 wk, before mean arterial pressure (MAP), portal pressure (PP), portal blood flow (PBF), hepatic vascular resistance, and putative molecular/cellular mechanisms were measured. In a subgroup of cirrhotic rats, the hemodynamic response to acute Prop (5 mg/kg iv) was assessed. Effects of metformin ⫾ Prop on PP and MAP were validated in common bile duct ligated-cirrhotic rats. Metformin-treated CCl4-cirrhotic rats had lower PP and hepatic vascular resistance than vehicle-treated rats, without significant changes in MAP or PBF. Metformin caused a significant reduction in liver fibrosis (Sirius red), hepatic stellate cell activation (␣-smooth muscle actin, platelet-derived growth factor receptor ␤ polypeptide, transforming growth factor-␤R1, and Rho kinase), hepatic inflammation (CD68 and CD163), superoxide (dihydroethidium staining), and nitric oxide scavenging (protein nitrotyrosination). Prop, by decreasing PBF, further reduced PP. Similar findings were observed in common bile duct ligated-cirrhotic rats. Metformin administration reduces PP by decreasing the structural and functional components of the elevated hepatic resistance of cirrhosis. This effect is additive to that of Prop. The potential impact of this pharmacological combination, otherwise commonly used in patients with cirrhosis and diabetes, needs clinical evaluation. fibrosis; liver; portal hypertension; LSEC; cirrhosis

consequence of liver cirrhosis that results in life-threatening complications with elevated morbidity and mortality. The initial factor for its development is the presence of an increased vascular resistance in the cirrhotic liver, due to hepatic architectural distortion and increased vascular tone (5). Histologically, cirrhosis is mainly characterized by increased deposition and altered composition of the extracellular matrix, and by the development of regenerative nodules. In the fibrotic liver, hepatic stellate cells (HSC) become activated, and their phenotype change to be

PORTAL HYPERTENSION IS A SERIOUS

* D. M. Tripathi and E. Erice are co-first authors; J. Gracia-Sancho and J. C. García-Pagán are co-last authors. Address for reprint requests and other correspondence: J. C. García-Pagán, Barcelona Hepatic Hemodynamic Lab, Liver Unit, Hospital Clínic de Barcelona, Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Villarroel 170, 08036 Barcelona, Spain (e-mail: [email protected]). http://www.ajpgi.org

myofibroblast like, with contractile properties and elevated production of collagen (13, 38). The increased hepatic vascular tone in cirrhosis is considered to be mainly due to endothelial dysfunction, as a consequence of an insufficient nitric oxide (NO) bioavailability in sinusoidal endothelial cells (15). Low NO bioavailability has been related both to a decreased endothelial NO synthase (eNOS) activity, which occurs despite normal eNOS protein expression, and to increased NO scavenging by superoxide radicals (19, 20). The biguanide drug metformin, besides having metabolic effects as an insulin sensitizer in type 2 diabetes mellitus, also confers vascular protection, improving the endothelial dysfunction in several cardiovascular diseases (7, 8, 10, 25, 31, 45). Indeed, previous studies in experimental models of arterial hypertension, diabetes, or fatty liver disease proposed that metforminderived vasoprotection may be due to an increase in NO bioavailability, resulting from enhanced eNOS activity (42) and reduced oxidative stress (33, 34), as well as by an improvement in the phenotype of vascular smooth muscle cells (2, 24). Then it is reasonable to propose that metformin treatment may have a potential role reducing intrahepatic resistance and, therefore, portal hypertension in cirrhosis. Interestingly, metformin has also been suggested to have an effect inhibiting the ␣-adrenergic tone, which was shown to be synergistic with the effect of propranolol controlling arterial hypertension (35, 37). Therefore, the primary purpose of the present study was to evaluate the hemodynamic and molecular effects of metformin administration on rats with liver cirrhosis. In addition, we also tested whether metformin influences the hemodynamic response to nonselective ␤-blockade. MATERIALS AND METHODS

The study was carried out in male Wistar and Sprague-Dawley rats (Charles River Laboratories, Barcelona, Spain). All procedures were approved by the Laboratory Animal Care and Use Committee of the University of Barcelona and were conducted in accordance with the European Community guidelines for the protection of animals used for experimental and other scientific purposes (EEC Directive 86/ 609). The personnel who prepared and administered treatments and those that performed the experimental studies were different. Treatments’ codes were not open for interpretation of the results until the inclusion of all animals. Experimental Models of Liver Cirrhosis and Metformin Administration Induction of cirrhosis by carbon tetrachloride. Cirrhosis with intrahepatic portal hypertension was experimentally induced in Wistar rats by inhalation of carbon tetrachloride (CCl4) three times a week, and phenobarbital (0.3 g/l) was added to the drinking water (12). When cirrhotic rats developed ascites, after ⬃12–15 wk of CCl4 inhalation, administration of CCl4 and phenobarbital was discontin-

0193-1857/15 Copyright © 2015 the American Physiological Society

G301

G302

METFORMIN IN PORTAL HYPERTENSION

ued, and treatments were started 1 wk later. A group of control rats, which only received phenobarbital, was included in the present study. Induction of cirrhosis by common bile duct ligation. Secondary biliary cirrhosis with intrahepatic portal hypertension was experimentally induced in rats by common bile duct ligation (CBDL), as described (41). After 3 wk of surgery, treatments started. Metformin administration. Control and cirrhotic rats received by gavage, once a day, either metformin hydrochloride (300 mg·kg⫺1·day⫺1; Sigma-Aldrich; n ⫽ 6 control, n ⫽ 12 in CCl4, and n ⫽ 9 in CBDL) or vehicle (0.9% NaCl; n ⫽ 6 control, n ⫽ 12 in CCl4, and n ⫽ 9 in CBDL) during 1 wk. Hemodynamic experiments were initiated 24 h after the administration of the last dose of metformin or vehicle (26, 43, 47). The dose of metformin has previously been shown to improve vascular function in experimental models of arterial hypertension, diabetes, and liver steatosis, among others (3, 9, 36, 43, 45). Experimental Studies In vivo hemodynamic study. All rats had free access to food and water until 12 h before the study. Methods for the hemodynamic evaluation in portal hypertensive rat models have been extensively described in previous studies (1, 17, 41). Briefly, under anesthesia and body temperature maintained at 37 ⫾ 0.5°C, portal pressure (PP; mmHg; ileocolic vein), mean arterial pressure (MAP, mmHg; femoral artery), portal blood flow (PBF; ml/min; portal vein as close as possible to the liver), and superior mesenteric artery blood flow (SMABF; ml/min; superior mesenteric artery) were measured using perivascular ultrasonic transit-time flow probes connected to a flow meter (Transonic Systems, Ithaca, NY), recorded using a PowerLab apparatus (4SP), and analyzed using the Chart v5.01 software (ADInstruments, Mountain View, CA). Hepatic vascular resistance (HVR, mmHg·ml⫺1·min·g) was calculated as PP/PBF. In a subgroup of cirrhotic animals (n ⫽ 9 in CCl4 and n ⫽ 6 in CBDL), after baseline hemodynamic data were obtained, rats received propranolol (5 mg/kg iv) for 10 min through the femoral vein catheter (39), and MAP and PP were recorded again. At the end the hemodynamic study, serum samples from cirrhotic rats were collected by cardiac puncture to subsequently evaluate alanine aminotransferase, aspartate aminotransferase, and bilirubin, all by standard protocols. Evaluation of the hepatic endothelial phenotype. NO BIOAVAILABILITY. Cyclic guanosine monophosphate, a marker of NO bioavailability, was determined in liver homogenates from cirrhotic rats treated with metformin or vehicle using an enzyme immunoassay (Cayman Chemical, Tallin, Estonia), as previously described (1). In addition, realtime levels of NO were determined using the 4-amino-5-methylamino-2=,7=-difluorofluorescein diacetate staining in liver sinusoidal endothelial cell (LSEC) freshly isolated from cirrhotic rats treated with metformin or vehicle, and in isolated LSEC from cirrhotic livers and treated in vitro with metformin (1 mM; 24 h) or its vehicle (PBS) (n ⫽ 3 per condition) (18, 19). NO MOLECULAR PATHWAY. Protein expression of eNOS, phosphorylated eNOS (p-eNOS), AMP-activated protein kinase (AMPK), and phosphorylated AMPK (p-AMPK) were assessed by Western blot in livers from cirrhotic rats treated with metformin or vehicle (17). Primary antibodies comprehended a mouse antibody recognizing eNOS (BD Transduction Laboratories, Lexington, KY), a rabbit anti-p-eNOS at Ser1176 (Cell Signaling Technology, Beverly, MA), a rabbit anti-AMPK (Cell Signaling Technology), or a rabbit p-AMPK at Thr172 (Cell Signaling Technology). Protein expression was determined by densitometric analysis using the Science Lab 2001, Image Gauge (Fuji Photo Film, Düsseldorf, Germany). After being stripped, blots were assayed for GAPDH content as standardization of sample loading. Quantitative densitometry values of eNOS and AMPK were normalized to GAPDH. The degree of eNOS phosphorylation and

AMPK phosphorylation was calculated as the ratio between the densitometry readings of p-eNOS/eNOS and p-AMPK/AMPK. ⫺ NO SCAVENGING BY SUPEROXIDE. Superoxide (O2 ) levels were quantified with the oxidative fluorescent dye dihydroethidium (10 ␮M; Molecular Probes, Eugene, OR) in liver slices and LSEC isolated from metformin- or vehicle-treated cirrhotic animals, as described (19, 21). Specificity of the assay was ensured using SOD (200 U/ml) as negative control. Nitrotyrosine content, as secondary marker of NO scavenging by O⫺ 2 to form peroxynitrite, was analyzed in liver sections (10 ␮M) through fluorohistochemistry (22) and in liver tissue through Western blot (1:1,000, Sigma) (19). Fluorescence images were obtained with a fluorescence microscope (Olympus BX51, Tokyo, Japan), and quantitative analysis of at least 10 images per slide was performed with Image J 1.44m software (National Institutes of Health, Bethesda, MD). SOD ACTIVITY. Total SOD activity was measured in liver homogenates using a commercially available immunoassay (Sigma, Tres Cantos, Madrid, Spain), according to the manufacturer’s instructions (22). Evaluation of hepatic fibrosis. QUANTIFICATION OF HEPATIC FIBROSIS. Semiquantitative analysis of hepatic fibrosis was performed in liver tissue blocks fixed in 10% formalin, embedded in paraffin, sectioned, and stained with 0.1% Sirius red. Eight fields from each slide were randomly selected and photographed, and the red-stained area per total area was measured using AxioVision software (Carl Zeiss MicroImaging) (41). Additionally, fibrosis was further quantified determining hepatic hydroxyproline content using a commercial kit (BioVision, Milpitas, CA), as previously described (4). Briefly, 40 mg of snap-frozen livers were hydrolyzed in HCl (12 N) at 120°C for 3 h and centrifuged to remove precipitates. Aliquots (10 ␮l) from each sample were evaporated to dryness and incubated with chloramine T (2.5 mM) for 5 min and Ehrlich’s reagent (410 mM) for 90 min at 60°C. Absorption was measured at 560 nm and referred to a standard curve. Results are expressed as micrograms per milligram liver tissue. HSC AND PORTAL MYOFIBROBLASTS PHENOTYPES. Hepatic protein expression of ␣-smooth muscle actin (␣-SMA; surrogate marker of HSC activation) and desmin (structural marker of HSC) were analyzed by immunohistochemistry. Immunostaining of paraffin-embedded liver sections was performed with a mouse anti-␣-SMA antibody (1:1,000, Sigma) and a mouse anti-desmin antibody (1:50, DAKO) or, as a negative control, with phosphate-buffered saline. Bound antibodies were visualized using Dako Real Envision Detection System Peroxidase/DAB⫹, and slides were then counterstained with hematoxylin. ␣-SMA and desmin relative volume was determined by point-counting morphometry using a point grid to obtain the number of intercepts over ␣-SMA- and desmin-positive cells over the tissue. Twelve fields were counted in each liver. All measurements were performed by two blinded observers (12). Fibrosis was further characterized in livers determining the expression of ␣-SMA, and platelet-derived growth factor receptor ␤ polypeptide (PDGFRB), and the Rho kinase activity by Western blot in hepatic samples using a mouse antibody against ␣-SMA (1:1,000, Sigma), a goat antibody against PDGFRB (1:500, Santa Cruz Biotechnology), a mouse antibody recognizing moesin (1:200, Santa Cruz Biotechnology), and a mouse anti-phosphorylated moesin at Thr558 antibody (1:200 Santa Cruz Biotechnology). Rho kinase activity was calculated as the ratio of p-moesin/moesin (41). Taqman expression assays were used to determine hepatic mRNA expression of procollagen I, marker of fibrosis, collagen 15A1, marker of portal myofibroblasts (30), and metalloproteinases (MMP)-2, MMP-9, and MMP-13 and its inhibitors TIMP-1 (tissue inhibitor of metalloproteinase-1) and TIMP-2, as markers of fibrinolysis. Specific effects of metformin on HSC phenotype was investigated using the human activated HSC cell line LX-2, kindly provided by Dr. Bataller. Cells were treated with 1 mM metformin for 24 h, and the expression of ␣-SMA, PDGFRB, transforming growth factor

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00010.2015 • www.ajpgi.org

G303

METFORMIN IN PORTAL HYPERTENSION

Table 1. Effects of metformin on hepatic and systemic hemodynamics in CCl4-cirrhotic rats Parameter

Vehicle

Metformin

P Value

n MAP, mmHg PP, mmHg PBF, ml/min SMABF, ml/min HVR, mmHg·ml⫺1·min·g HR, beats/min Body weight, g

12 99 ⫾ 22 13.9 ⫾ 2.6 13.8 ⫾ 6.1 8.9 ⫾ 3.6 12.0 ⫾ 4.0 338 ⫾ 40 369 ⫾ 9

12 102 ⫾ 17 10.2 ⫾ 2.8 14.5 ⫾ 5.4 8.1 ⫾ 3.2 7.9 ⫾ 2.7 365 ⫾ 46 373 ⫾ 8

0.92 0.003 0.774 0.662 0.009 0.139 0.75

Values are means ⫾ SD; n, no. of rats. CCl4, carbon tetrachloride; MAP, mean arterial pressure; PP, portal pressure; PBF, portal blood flow; SMABF, superior mesenteric artery blood flow; HVR, hepatic vascular resistance; HR, heart rate.

(TGF)-␤, TGF-␤ receptor type 1 (TGF-␤R1), and pro-collagen I was analyzed using Western blot or quantitative PCR, as described above. ANALYSIS OF HEPATIC INFLAMMATION. Hepatic macrophages M1 subtype (CD68⫹) and M2 subtype (CD163⫹) were evaluated by Western blot in hepatic tissue samples using a mouse monoclonal antibody against CD68 (1:1,000 Bio-Rad) and CD163 (1:1,000 Bio-Rad). Statistics Statistical analysis was performed using the SPSS 19.0 statistical package (IBM, Armonk, NY). Comparisons between groups were performed with U Mann-Whitney for continuous variables and Fisher test for categorical variables. Comparisons of intragroup results before and after propranolol administration were performed using paired t-test. All data are reported as means ⫾ SD, unless otherwise specified. Differences were considered significant at a P value ⬍ 0.05. RESULTS

Effects of Metformin in CCl4 Cirrhotic Rats Metformin lowers PP in CCl4 cirrhotic animals. CCl4 cirrhotic animals exhibited portal hypertension compared with control rats (Table 1). CCl4 cirrhotic rats receiving metformin exhibited statistically significant lower PP than rats receiving vehicle (10.2 ⫾ 2.8 vs. 13.9 ⫾ 2.6 mmHg; ⫺27%; P ⫽ 0.003). This reduction was not associated with a significant change in PBF reflecting a fall in HVR (7.9 ⫾ 2.7 vs. 12.0 ⫾ 4.0 mmHg·ml⫺1·min·g; ⫺34%; P ⫽ 0.009). MAP, SMABF, and heart rate (HR) were not modified by metformin (Table 1). Propranolol was given intravenously after obtaining baseline values. This produced a significant reduction in HR and PP in both groups. However, while the reduction in HR was similar (⫺30% in metformin-pretreated rats vs. ⫺27% in the vehicle group), the reduction in PP was significantly greater in the metformin group (⫺32 vs. ⫺17% in vehicle rats; P ⫽ 0.03). As a consequence, the final PP in the metformin ⫹ propranolol group was markedly lower than that in the vehicle ⫹ propranolol group (6.6 ⫾ 2.2 vs. 10.9 ⫾ 1.8 mmHg; 40% difference; P ⫽ 0.001). No differences in biochemical parameters were found comparing both groups of CCl4-cirrhotic rats (Table 2). Metformin increases liver endothelial NO bioavailability in CCl4 cirrhotic rats. No differences in cyclic guanosine monophosphate content were found in liver homogenates from cirrhotic rats treated with metformin or vehicle (18.3 ⫾ 2.9 pmol/ml in metformin vs. 19.2 ⫾ 3.4 pmol/ml in vehicle; P ⫽ nonsignificant). However, LSEC isolated from metformintreated cirrhotic animals exhibited significantly higher NO levels than those from animals receiving vehicle (Fig. 1A), as

demonstrated by measuring specific endothelial NO bioavailability using 4-amino-5-methylamino-2=,7=-difluorofluorescein staining. Such increased NO bioavailability was also observed in LSEC isolated from cirrhotic livers treated in vitro with metformin (Fig. 1B). Metformin does not modify eNOS and AMPK pathways but exerts an antioxidant effect within the cirrhotic liver. No effects of metformin administration were found analyzing eNOS, peNOS, AMPK, or p-AMPK protein expression (data not shown). However, livers from cirrhotic rats treated with metformin exhibited markedly lower O⫺ 2 levels compared with the vehicle group (⫺76% in tissue, ⫺71% in isolated LSEC; Fig. 1, C and D), which was associated with marked diminished levels of nitrotyrosinated proteins (⫺43% by fluorohistochemistry and ⫺28% by Western blot; Fig. 1, E and F), surrogate marker of NO scavenging by O⫺ 2 . Liver SOD activity was significantly higher in metformin-treated cirrhotic rats, indicating greater liver antioxidant capacity than those treated with vehicle (1.27 ⫾ 0.12 vs. 2.06 ⫾ 0.13 U/ml; ⫹61%; P ⬍ 0.05). Metformin stimulates reduction of fibrosis in CCl4 cirrhotic rats. As expected, CCl4 cirrhotic rats exhibited marked distortion of hepatic parenchyma with abundant fibrosis, as evaluated by Sirius red staining. As shown in Fig. 2, metformin produced a significant reduction in hepatic fibrosis compared with vehicle administration (⫺41% in Sirius red; ⫺17% in hydroxyproline; ⫺34% in pro-collagen I); however, it did not reach normal values. Importantly, fibrosis improvement was associated with significant diminutions in the expression of ␣-SMA (⫺74% by Western blot; ⫺61% by immunohistochemistry), desmin (⫺46%), and PDGFRB (⫺39%), and in Rho kinase activity (⫺55% in p-moesin-to-total moesin ratio), altogether suggesting decreased activation and abundance of HSC. Fibrinolysis characterization revealed no changes in the mRNA expression of MMP-2, MMP-9, and TIMP-2 (data not shown). However, we observed a trend to reduction in MMP-13 (⫺50%; P ⫽ 0.3) and TIMP-1 (⫺30%; P ⫽ 0.2). In addition, livers from metformin-treated animals showed reduced levels of the recently proposed marker of portal myofibroblasts, collagen 15A1, although it did not reach statistical significance (Fig. 2F). As shown in Fig. 3, cellular studies confirmed the effects of metformin improving HSC phenotype. Indeed, LX-2 cells treated with metformin showed decreased expression of procollagen I (⫺50%), ␣-SMA (⫺48%), PDGFRB (⫺41%), and TGF-␤R1 (⫺16%), without significant differences in TGF-␤ expression (data not shown).

Table 2. Effects of metformin on biochemical parameters in CCl4-cirrhotic and CBDL-cirrhotic rats CCl4 AST, U/l ALT, U/l Bilirubin, mg/dl CBDL AST, U/l ALT, U/l Bilirubin, mg/dl

Vehicle

Metformin

P Value

269 ⫾ 120 148 ⫾ 41 0.20 ⫾ 0.07

264 ⫾ 65 103 ⫾ 37 0.18 ⫾ 0.06

0.97 0.46 0.87

486 ⫾ 222 106 ⫾ 19 0.9 ⫾ 0.2

460 ⫾ 201 67 ⫾ 36 0.7 ⫾ 0.2

0.91 0.075 0.11

Values are means ⫾ SD; n ⫽ 12 CCl4 and 9 common bile duct ligation (CBDL) rats. AST, aspartate aminotransferase; ALT, alanine aminotransferase.

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00010.2015 • www.ajpgi.org

G304

METFORMIN IN PORTAL HYPERTENSION

A

B

Metformin

Vehicle

3.0

NO levels (RLU)

*

3.5 2.5 2.0 1.5 1.0 0.5 0.0

Vehicle

Metformin

C

Metformin

2.0

*

1.5 1.0 0.5 0.0

Vehicle

Metformin

D

Vehicle

Metformin

1.2 0.8

*

0.4 0.0

Vehicle

Metformin

E

Metformin

Vehicle

O2- levels (RLU)

O2- levels (RLU)

Fig. 1. Metformin improves endothelial nitric oxide (NO) bioavailability and oxidative stress in carbon tetrachloride (CCl4)-cirrhotic livers. In vivo real-time production of NO was determined in liver sinusoidal endothelial cell (LSEC) freshly isolated from cirrhotic rats treated with metformin or vehicle (A), or in cirrhotic LSEC treated for 24 h in vitro with metformin or vehicle (B). Intrahepatic superoxide (O⫺ 2 ) was determined in cirrhotic rats treated with metformin or vehicle (C), or in LSEC isolated from cirrhotic rats receiving metformin or vehicle (D). Hepatic levels of nitrotyrosinated (N-Tyr) proteins [fluorohistochemistry (E) and Western blot (F)] were determined in tissue samples from control rats (c) and cirrhotic rats treated with metformin (m) or vehicle (v). Values are means ⫾ SE. Depicted results were derived from n ⫽ 3 independent experiments for cellular studies, and from n ⫽ 6 (control) and n ⫽ 12 (cirrhotic) animals per group for tissular determinations. *P ⬍ 0.05 vs. cirrhotic-vehicle. #P ⬍ 0.05 vs. control. All images are ⫻20 magnification. RLU, relative light units.

NO levels (RLU)

Vehicle

1.2 0.8

*

0.4 0.0

Vehicle

Metformin

F N-Tyr GAPDH c

m

#

5

1.2 1.0 0.8

*

0.6 0.4 0.2 0.0

v

Metformin

Vehicle

Metformin reduces hepatic inflammation in CCl4 cirrhotic rats. Analysis of metformin effects on hepatic inflammation showed that M1 subtype macrophages (CD68⫹) were decreased (⫺31%; P ⫽ 0.04) compared with vehicle-treated cirrhotic rats. No differences in the expression of M2 subtype (CD163⫹) macrophages were observed.

Metformin

N-Tyr relative to control

N-Tyr protein (RLU)

Vehicle

4

*

3 2 1 0

Control

Vehicle

Metformin

Effects of Metformin in CBDL Cirrhotic Rats Metformin ameliorates portal hypertension in CBDL cirrhotic rats. Animals with cirrhosis of the liver caused by CBDL were used as a validation group. In this experimental model, animals treated with metformin also exhibited signifi-

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00010.2015 • www.ajpgi.org

G305

#

25 20

*

15 10 5 0

Control

Vehicle

Metformin

C

B p=0.14 α-SMA

0.9 0.6

GAPDH

0.3 0.0

Vehicle Control

Vehicle

Metformin

10.0

#

8.0

p=0.1

6.0 4.0 2.0 0.0

Control

Vehicle

1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

* Vehicle

Metformin

E

Vehicle

GAPDH

* Vehicle

Metformin

Metformin

Vehicle

p-moesin/moesin relative to vehicle

p=0.1

PDGFRB mRNA relative to vehicle

Desmin relative to vehicle

1.2 1.0 0.8 0.6 0.4 0.2 0.0

Metformin

1.5 1.0

*

0.5 0.0

Vehicle

Metformin

F

p-Moesin Moesin

Vehicle

Metformin

Metformin

D 1.2 1.0 0.8 0.6 0.4 0.2 0.0

Vehicle

Metformin

α-SMA relative to vehicle

Sirius Red (% total area)

Metformin

1.2

1.2 1.0 0.8 0.6 0.4 0.2 0.0

Metformin

* Vehicle

Collagen 15A1 mRNA relative to vehicle

Vehicle

Control

#

1.5

α-SMA relative to vehicle

A

pro-collagen I mRNA Hydroxyproline relative to control (μg/mg)

METFORMIN IN PORTAL HYPERTENSION

1.2 1.0 0.8 0.6 0.4 0.2 0.0

p=0.12

Vehicle

Metformin

Metformin

Fig. 2. Metformin promotes liver fibrosis regression in CCl4-cirrhotic animals. A, left: representative histological images of livers stained with Sirius red from control rats and cirrhotic rats treated for 1 wk with vehicle or metformin, and quantification of fibrosis [Sirius red staining area per total area; n ⫽ 6 (control) and n ⫽ 12 (cirrhotic) per group]. Right: hydroxyproline levels (top) and pro-collagen I mRNA expression (bottom) determined in livers described before. B: representative ␣-smooth muscle actin (␣-SMA) immunohistochemistry and Western blot with quantifications from livers described in A. C: desmin detection and quantification in livers described above. D: platelet-derived growth factor receptor ␤ polypeptide (PDGFRB) mRNA expression in livers described in A. E: Rho kinase activity expressed as the ratio between phosphorylated (p)-moesin and moesin in livers described above. F: mRNA expression of collagen 15A1 determined in livers described above. Slide quantifications were derived from 8 (Sirius Red) or 12 (immunohistochemistry) pictures per preparation. Values are means ⫾ SE. *P ⬍ 0.05 vs. vehicle. #P ⬍ 0.05 vs. control. All images are ⫻10 magnification.

cantly lower PP than the vehicle-treated group (17.2 ⫾ 2.3 vs. 19.1 ⫾ 2.7 mmHg; ⫺10%; P ⫽ 0.009), with no differences in MAP or HR (Table 3). Similar to what was observed in CCl4-cirrhotic rats, intravenous propranolol produced a significant reduction in PP in CBDL rats treated with either metformin or vehicle. As a consequence, the final PP in the metformin ⫹ propranolol group was markedly lower than that in the vehicle ⫹ propranolol group (14.8 ⫾ 1.7 vs. 17.5 ⫾ 1.4 mmHg; ⫺15%; P ⫽ 0.01). No differences in biochemical parameters were found comparing CBDL-cirrhotic rats receiving metformin or vehicle (Table 2). Metformin improves eNOS activity but does not affect liver fibrosis in CBDL-cirrhotic animals. Metformin treatment enhanced the phosphorylation of eNOS, suggesting improvement in its enzymatic activity, without modifying eNOS, AMPK, or p-AMPK (Fig. 4, A and B). Fibrosis evaluation using Sirius red staining revealed no differences between CBDL cirrhotic rats receiving metformin or vehicle (Fig. 4C). For quantification of portal myofibroblasts using collagen 15A1, although 18% lower in the metformin-treated rats, the difference did not reach statistical significance between both groups (data not shown). DISCUSSION

In liver cirrhosis, increased HVR, due to an increased hepatic vascular tone and to architectural abnormalities of the liver parenchyma, is the main player in the development of portal hypertension. Different studies have evaluated the pos-

sibility of reducing HVR by enhancing hepatic NO bioavailability and reducing hepatic vascular tone using several experimental strategies (14, 21, 29). Although these studies showed beneficial effects, novel therapeutic strategies based on European Medicines Agency/US Food and Drug Administration approved drugs with no systemic adverse effects are required to improve treatments for patients with portal hypertension. The present study shows that 1-wk metformin administration decreases PP in two different models of liver cirrhosis. In fact, in CCl4-cirrhotic rats treated with metformin, PP was 27% lower than in those receiving vehicle. The decrease in PP was not associated with modifications in PBF, pointing to a decreased HVR. Importantly, the beneficial effects of metformin reducing portal hypertension were confirmed, although of less magnitude, in CBDL-cirrhotic rats. Interestingly, metformin appears to ameliorate HVR differently in the two cirrhotic models used. Indeed, in CCl4cirrhotic animals treated with metformin, a marked amelioration in fibrosis was observed, which was associated with an improvement in HSC phenotype and reductions in the hepatic content of HSC and probably, although to a less extent, in portal myofibroblasts. This finding is in agreement with previous studies suggesting that long-term treatment with metformin ameliorates mild fibrosis in liver and heart (6, 40, 50); nevertheless, our report describes for the first time the effects of metformin reducing fibrosis in a pathology where exaggerated collagen deposition exists. Characterization of the underlying molecular mechanisms leading to fibrosis diminution in terms of fibrosis regression revealed a decrease in the expression of

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00010.2015 • www.ajpgi.org

G306

METFORMIN IN PORTAL HYPERTENSION

B

GAPDH v

1.0 0.8

*

0.6 0.4 0.2 0.0

Vehicle

Metformin

1.4 1.2 1.0 0.8

*

0.6 0.4 0.2

Vehicle

C

D

Metformin

TGFβR1 GAPDH

1.2

v

1.0 0.8

*

0.4 0.2 0.0

Metformin

*

1.0 0.8 0.6 0.4 0.2 0.0

Vehicle

the MMP-13 and MMP inhibitor TIMP-1, although it did not achieve significance, with no differences in MMP-2, MMP-9, and TIMP-2 compared with vehicle-treated cirrhotic rats. Overall, these data suggest that matrix degradation, at the dose and duration of treatment, may not play a major role in the reduction of fibrosis. Metformin-mediated HSC deactivation may result from its capability to reduce oxidative stress, a well-known profibrogenic stimulus, but also from the inhibition of the proliferative and profibrogenic pathways PDGF, TGF-␤, and Rho kinase (38, 47). Indeed, our study agrees with previous data demonstrating that inhibition of Rho kinase results in HSC phenotype amelioration and senescence (27). Metformin-derived inhibition of Rho kinase may occur through, at least, two different mechanisms: increment in NO bioavailability, and enhancement of the Rac1-cdc42 signaling pathway. Both of them have been described as Rho kinase inhibitors (28, 32). Table 3. Effects of metformin on hepatic and systemic hemodynamics in CBDL-cirrhotic rats Parameter

Vehicle

Metformin

P Value

n MAP, mmHg PP, mmHg HR, beats/min Body weight, g

9 98 ⫾ 11 19.1 ⫾ 2.7 340 ⫾ 35 336 ⫾ 4

9 100 ⫾ 19 17.2 ⫾ 2.3 335 ⫾ 49 331 ⫾ 5

1.00 0.009 0.83 0.48

m

1.2

TGFβR1 relative to vehicle

0.6

Vehicle

Values are means ⫾ SD; n, no. of rats.

m

0.0

PDGFRB mRNA relative to vehicle

Fig. 3. Metformin promotes hepatic stellate cell (HSC) deactivation. Human HSC LX-2 were treated with metformin (m), or its vehicle (v), for 24 h, and the expression of pro-collagen I (A), ␣-SMA (B), PDGFRB (C), and transforming growth factor (TGF)-␤R1 (D) were analyzed. Values are means ⫾ SE derived from n ⫽ 3 independent experiments. *P ⬍ 0.05 vs. vehicle.

α-SMA

1.2

α-SMA relative to vehicle

pro-collagen I mRNA relative to vehicle

A

Metformin

By contrast, we were unable to demonstrate an effect of metformin on liver fibrosis in the CBDL model. This discrepancy may be due to the different characteristics of fibrogenesis in the two models; the CBDL model is characterized by a very rapid progression of fibrosis with no possibility of spontaneous regression due to the persistence of the bile obstruction, while the CCl4 is much slower and susceptible to regression upon stopping administration of the toxic. Importantly, discrepancies in the mechanisms explaining the beneficial effects of a certain drug when comparing different experimental models of cirrhosis are not new. In fact, previous works from our team and others using the thromboxane A2 receptor antagonist terutroban or the farnesoid X receptor agonist obeticholic acid already demonstrated such phenomena (41, 44). In addition, our data also show that metformin enhances liver NO bioavailability, which contributes to reduce HVR and PP. In the CCl4-cirrhotic model, we did not detect any increase in eNOS phosphorylation, a marker of increased eNOS activity, but we observed an increased endothelial NO bioavailability that was mainly derived from an upregulated SOD activity, thus reducing NO scavenging and formation of peroxynitrite. The use of antioxidants to increase liver NO was primarily described by our group and validated using different antioxidants (11, 12, 14, 19, 48, 49). So, our results agree with previous reports demonstrating the antioxidant effects of metformin in other vascular beds (33, 34). On the other hand, in the CBDL-cirrhosis model, we found that metformin induces the activation of the NO-generating enzyme eNOS. Previous in

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00010.2015 • www.ajpgi.org

G307

METFORMIN IN PORTAL HYPERTENSION

A

B

C

p-eNOS

p-AMPK

eNOS

AMPK

GAPDH

GAPDH Metformin

Vehicle

1.5 1.0 0.5 0.0

Vehicle

Metformin

Metformin

Vehicle

1.5 1.0 0.5 0.0

Metformin

30

2.0

Sirius Red (% total area)

*

2.0

p-AMPK/AMPK relative to vehicle

p-eNOS/eNOS relative to vehicle

Vehicle

Vehicle

Metformin

25 20 15 10 5 0

Vehicle

Metformin

Fig. 4. Effects of metformin in common bile duct ligation (CBDL)-cirrhotic animals. A: endothelial NO synthase (eNOS) and p-eNOS expression determined in CBDL-cirrhotic animals treated for 1 wk with vehicle (left) or metformin (right, n ⫽ 9 per group). B: expression of AMP-activated protein kinase (AMPK) and its phosphorylated form in livers described above. C: representative images of fibrosis detection using Sirius red staining (⫻10 magnification) with their quantification from livers described in A. Slide quantifications are derived from 8 pictures per preparation. Values are means ⫾ SE. *P ⬍ 0.05 vs. vehicle.

vitro studies showed that metformin is able to increase eNOSdependent NO production in endothelial cells (11–13). The reason for such discrepancy between experimental models of cirrhosis is not clear. However, this has been previously reported by our group and others (16, 41, 44), emphasizing the need of extending studies to different experimental models to increase the chances of translating to human beings the findings observed at the bench-side. It is important to note that the reduction in PP due to metformin was observed in both models of cirrhosis, although the magnitude of the reduction, probably because of the observed reduction in fibrosis, was greater in the CCl4 than in the CBDL model. Although NO certainly plays a central role modulating vascular tone within the cirrhotic liver (46), we cannot discard that other mechanisms that have been implicated in metforminderived vasoprotection, such as inhibition of proinflammatory responses (31, 40), suppression of vasoconstrictor prostanoids (34), or modulation of calcium flux within endothelial cells (23, 34), may also contribute to improve the intrahepatic vascular tone in cirrhosis. It is worthy to note that no detrimental systemic effects, analyzed as changes in MAP, SMABF, or in HR, were observed in cirrhotic animals receiving metformin. This observation is in agreement with a recently published study demonstrating no contraindications of metformin when administered to patients with liver cirrhosis (51). Such differential effects of metformin, according to the vascular bed, could be attributed to the known affinity of metformin for the damaged/dysfunctional endothelium, evidenced by previous studies demonstrating no

ACKNOWLEDGMENTS Authors are indebted to Montse Monclús and Sergi Vila for excellent technical assistance.

Table 4. Effects of metformin on hepatic and systemic hemodynamics in control rats

GRANTS

Parameter

Vehicle

Metformin

P Value

n MAP, mmHg PP, mmHg HR, beats/min

6 124 ⫾ 18 7.8 ⫾ 1.1 346 ⫾ 42

6 135 ⫾ 17 7.9 ⫾ 1.5 337 ⫾ 43

0.23 0.72 0.89

Values are means ⫾ SD; n, no. of rats.

effect of metformin on vascular function in normal rats (25, 33, 35, 45). In fact, our results support this hypothesis, since no significant changes in PP or in systemic hemodynamics were observed when metformin was administered to control rats (Table 4). Considering that ␤-blockers are accepted therapeutic agents for primary and secondary prophylaxis in portal hypertensive patients, we further evaluated the possible synergistic effects of an acute administration of propranolol in animals under metformin treatment. Metformin-receiving animals exhibited a further decrease in PP after propranolol administration, reaching much lower levels than in animals treated with placebo, with no differences in systemic hemodynamics, thus demonstrating that metformin may add a significant advantage to the established treatment with propranolol. The additive effects of both treatments reducing PP may result from the capability of metformin to reduce the HVR, together with the reduction in portal blood inflow due to nonselective ␤-blockade. These results are in accordance with previous reports describing the combined effect of metformin and propranolol ameliorating arterial hypertension (35, 37). In conclusion, the present study provides novel information showing that metformin administration to cirrhotic animals decreases PP, thus ameliorating portal hypertension. Moreover, the metformin-derived improvement has a synergistic effect with ␤-blockers reducing PP, suggesting a new therapeutic approach to treat portal hypertensive patients.

This work was funded by the Ministerio de Economía y Competitividad [SAF (2013-44723-R) and FIS (PI13/00341 and PI14/00029)], the European Union (Fondos Europeos de Desarrollo Regional, “una manera de hacer Europa”), and by an Indo-Spanish collaborative grant from the Ministerio de Economía y Competitividad-Spain and the Department of Science and Technology-India (ACI2009-0938). D.M.T. has a Sheila Sherlock Fellowship from the European Association for the Study of the Liver (EASL). J. G.-S. has a

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00010.2015 • www.ajpgi.org

G308

METFORMIN IN PORTAL HYPERTENSION

Ramón y Cajal contract from the Ministerio de Economía y Competitividad. Centro de Investigación Biomédica en Red de Enfermedades Hepáticas y Digestivas is funded by the Instituto de Salud Carlos III. DISCLOSURES No conflicts of interest, financial or otherwise, are declared by the author(s). AUTHOR CONTRIBUTIONS Author contributions: D.M.T., E.E., J.G.-S., and J.C.G.-P. conception and design of research; D.M.T., E.E., E.L., and H.G.-C. performed experiments; D.M.T., E.E., E.L., and H.G.-C. analyzed data; D.M.T., J.G.-S., and J.C.G.-P. prepared figures; D.M.T., E.E., J.G.-S., and J.C.G.-P. drafted manuscript; D.M.T., E.E., E.L., H.G.-C., S.K.S., J.B., J.G.-S., and J.C.G.-P. approved final version of manuscript; S.K.S., J.B., J.G.-S., and J.C.G.-P. interpreted results of experiments; J.B., J.G.-S., and J.C.G.-P. edited and revised manuscript; J.G.-S. and J.C.G.-P. codirected the study. REFERENCES 1. Abraldes JG, Rodriguez-Vilarrupla A, Graupera M, Zafra C, GarciaCaldero H, Garcia-Pagan JC, Bosch J. Simvastatin treatment improves liver sinusoidal endothelial dysfunction in CCl(4) cirrhotic rats. J Hepatol 46: 1040 –1046, 2007. 2. Agard C, Rolli-Derkinderen M, Dumas-de-La-Roque E, Rio M, Sagan C, Savineau JP, Loirand G, and Pacaud P. Protective role of the antidiabetic drug metformin against chronic experimental pulmonary hypertension. Br J Pharmacol 158: 1285–1294, 2009. 3. Bergheim I, Guo L, Davis MA, Lambert JC, Beier JI, Duveau I, Luyendyk JP, Roth RA, Arteel GE. Metformin prevents alcohol-induced liver injury in the mouse: Critical role of plasminogen activator inhibitor-1. Gastroenterology 130: 2099 –2112, 2006. 4. Bergman I, Loxley R. Two improved and simplified methods for the spectrophotometric determination of hydroxyproline. Anal Chem 35: 1961–1965, 1963. 5. Bosch J, Pizcueta P, Feu F, Fernandez M, Garcia-Pagan JC. Pathophysiology of portal hypertension. Gastroenterol Clin North Am 21: 1–14, 1992. 6. Bugianesi E, Gentilcore E, Manini R, Natale S, Vanni E, Villanova N, David E, Rizzetto M, Marchesini G. A randomized controlled trial of metformin versus vitamin E or prescriptive diet in nonalcoholic fatty liver disease. Am J Gastroenterol 100: 1082–1090, 2005. 7. Caballero AE, Delgado A, Guilar-Salinas CA, Herrera AN, Castillo JL, Cabrera T, Gomez-Perez FJ, Rull JA. The differential effects of metformin on markers of endothelial activation and inflammation in subjects with impaired glucose tolerance: a placebo-controlled, randomized clinical trial. J Clin Endocrinol Metab 89: 3943–3948, 2004. 8. Calvert JW, Gundewar S, Jha S, Greer JJ, Bestermann WH, Tian R, Lefer DJ. Acute metformin therapy confers cardioprotection against myocardial infarction via AMPK-eNOS-mediated signaling. Diabetes 57: 696 –705, 2008. 9. Choi YH, Kim SG, Lee MG. Dose-independent pharmacokinetics of metformin in rats: hepatic and gastrointestinal first-pass effects. J Pharm Sci 95: 2543–2552, 2006. 10. Davis BJ, Xie Z, Viollet B, Zou MH. Activation of the AMP-activated kinase by antidiabetes drug metformin stimulates nitric oxide synthesis in vivo by promoting the association of heat shock protein 90 and endothelial nitric oxide synthase. Diabetes 55: 496 –505, 2006. 11. Delgado MG, Gracia-Sancho J, Marrone G, Rodriguez-Vilarrupla A, Deulofeu R, Abraldes JG, Bosch J, Garcia-Pagan JC. Leptin receptor blockade reduces intrahepatic vascular resistance and portal pressure in an experimental model of rat liver cirrhosis. Am J Physiol Gastrointest Liver Physiol 305: G496 –G502, 2013. 12. Di Pascoli M, Divi M, Rodriguez-Vilarrupla A, Rosado E, GraciaSancho J, Vilaseca M, Bosch J, Garcia-Pagan JC. Resveratrol improves intrahepatic endothelial dysfunction and reduces hepatic fibrosis and portal pressure in cirrhotic rats. J Hepatol 58: 904 –910, 2013. 13. Friedman SL. Liver fibrosis–from bench to bedside. J Hepatol 38, Suppl 1: S38 –S53, 2003. 14. Garcia-Caldero H, Rodriguez-Vilarrupla A, Gracia-Sancho J, Divi M, Lavina B, Bosch J, Garcia-Pagan JC. Tempol administration, a superoxide dismutase mimetic, reduces hepatic vascular resistance and portal pressure in cirrhotic rats. J Hepatol 54: 660 –665, 2010.

15. Garcia-Pagan JC, Gracia-Sancho J, Bosch J. Functional aspects on the pathophysiology of portal hypertension in cirrhosis. J Hepatol 57: 458 – 461, 2012. 16. Grace JA, Klein S, Herath CB, Granzow M, Schierwagen R, Masing N, Walther T, Sauerbruch T, Burrell LM, Angus PW, Trebicka J. Activation of the MAS receptor by angiotensin-(1–7) in the renin-angiotensin system mediates mesenteric vasodilatation in cirrhosis. Gastroenterology 145: 874 –884, 2013. 17. Gracia-Sancho J, Laviña B, Rodriguez-Vilarupla A, Brandes RP, Fernandez M, Bosch J, Garcia-Pagan JC. Evidence against NADPH oxidase modulating hepatic vascular tone in cirrhosis. Gastroenterology 133: 959 –966, 2007. 18. Gracia-Sancho J, Lavina B, Rodriguez-Vilarrupla A, Garcia-Caldero H, Bosch J, Garcia-Pagan JC. Enhanced vasoconstrictor prostanoid production by sinusoidal endothelial cells increases portal perfusion pressure in cirrhotic rat livers. J Hepatol 47: 220 –227, 2007. 19. Gracia-Sancho J, Lavina B, Rodriguez-Vilarupla A, Garcia-Caldero H, Fernandez M, Bosch J, Garcia-Pagan JC. Increased oxidative stress in cirrhotic rat livers: a potential mechanism contributing to reduced nitric oxide bioavailability. Hepatology 47: 1248 –1256, 2008. 20. Gracia-Sancho J, Russo L, Garcia-Caldero H, Garcia-Pagan JC, Garcia-Cardena G, Bosch J. Endothelial expression of transcription factor Kruppel-like factor 2 and its vasoprotective target genes in the normal and cirrhotic rat liver. Gut 60: 517–524, 2011. 21. Guillaume M, Rodriguez-Vilarrupla A, Gracia-Sancho J, Rosado E, Mancini A, Bosch J, Garcia-Pagan JC. Recombinant human manganese superoxide dismutase reduces liver fibrosis and portal pressure in CCl4cirrhotic rats. J Hepatol 58: 240 –246, 2013. 22. Hide D, Ortega-Ribera M, Fernandez-Iglesias A, Fondevila C, Salvado MJ, Arola L, Garcia-Pagan JC, Mancini A, Bosch J, GraciaSancho J. A novel form of the human manganese superoxide dismutase protects rat and human livers undergoing ischemia and reperfusion injuries. Clin Sci (Lond) 127: 527–537, 2014. 23. Isoda K, Young JL, Zirlik A, MacFarlane LA, Tsuboi N, Gerdes N, Schonbeck U, Libby P. Metformin inhibits proinflammatory responses and nuclear factor-kappaB in human vascular wall cells. Arterioscler Thromb Vasc Biol 26: 611–617, 2006. 24. Kalender A, Selvaraj A, Kim SY, Gulati P, Brule S, Viollet B, Kemp BE, Bardeesy N, Dennis P, Schlager JJ, Marette A, Kozma SC, Thomas G. Metformin, independent of AMPK, inhibits mTORC1 in a rag GTPase-dependent manner. Cell Metab 11: 390 –401, 2010. 25. Katakam PV, Ujhelyi MR, Hoenig M, Miller AW. Metformin improves vascular function in insulin-resistant rats. Hypertension 35: 108 –112, 2000. 26. Kita Y, Takamura T, Misu H, Ota T, Kurita S, Takeshita Y, Uno M, Matsuzawa-Nagata N, Kato K, Ando H, Fujimura A, Hayashi K, Kimura T, Ni Y, Otoda T, Miyamoto K, Zen Y, Nakanuma Y, Kaneko S. Metformin prevents and reverses inflammation in a non-diabetic mouse model of nonalcoholic steatohepatitis. PLoS One 7: e43056, 2012. 27. Klein S, Klosel J, Schierwagen R, Korner C, Granzow M, Huss S, Mazar IG, Weber S, van d V, Pieper-Furst U, Furst DO, Nattermann J, Lammert F, Sauerbruch T, Trebicka J. Atorvastatin inhibits proliferation and apoptosis, but induces senescence in hepatic myofibroblasts and thereby attenuates hepatic fibrosis in rats. Lab Invest 92: 1440 –1450, 2012. 28. Kumar A, Al-Sammarraie N, DiPette DJ, Singh US. Metformin impairs Rho GTPase signaling to induce apoptosis in neuroblastoma cells and inhibits growth of tumors in the xenograft mouse model of neuroblastoma. Oncotarget 5: 11709 –11722, 2014. 29. Lavina B, Gracia-Sancho J, Rodriguez-Vilarrupla A, Chu Y, Heistad DD, Bosch J, Garcia-Pagan JC. Superoxide dismutase gene transfer reduces portal pressure in CCl4 cirrhotic rats with portal hypertension. Gut 58: 118 –125, 2009. 30. Lemoinne S, Cadoret A, Rautou PE, El MH, Ratziu V, Corpechot C, Rey C, Bosselut N, Barbu V, Wendum D, Feldmann G, Boulanger C, Henegar C, Housset C, Thabut D. Portal myofibroblasts promote vascular remodeling underlying cirrhosis formation through the release of microparticles. Hepatology 61: 1041–1055, 2015. 31. Lin HZ, Yang SQ, Chuckaree C, Kuhajda F, Ronnet G, Diehl AM. Metformin reverses fatty liver disease in obese, leptin-deficient mice. Nat Med 6: 998 –1003, 2000. 32. Liu Y, Huang C, Ceng C, Zhan H, Zheng D, Han W. Metformin enhances nitric oxide production and diminishes Rho kinase activity in rats with hyperlipidemia. Lipids Health Dis 13: 115, 2014.

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00010.2015 • www.ajpgi.org

METFORMIN IN PORTAL HYPERTENSION 33. Majithiya JB, Balaraman R. Metformin reduces blood pressure and restores endothelial function in aorta of streptozotocin-induced diabetic rats. Life Sci 78: 2615–2624, 2006. 34. Matsumoto T, Noguchi E, Ishida K, Kobayashi T, Yamada N, Kamata K. Metformin normalizes endothelial function by suppressing vasoconstrictor prostanoids in mesenteric arteries from OLETF rats, a model of type 2 diabetes. Am J Physiol Heart Circ Physiol 295: H1165–H1176, 2008. 35. Muntzel MS, Abe A, Petersen JS. Effects of adrenergic, cholinergic and ganglionic blockade on acute depressor responses to metformin in spontaneously hypertensive rats. J Pharmacol Exp Ther 281: 618 –623, 1997. 36. Pasarin M, Abraldes JG, Rodriguez-Vilarrupla A, La Mura V, Garcia-Pagan JC, Bosch J. Insulin resistance and liver microcirculation in a rat model of early NAFLD. J Hepatol 55: 1095–1102, 2011. 37. Peuler JD. Opposing adrenergic actions of intravenous metformin on arterial pressure in female spontaneously hypertensive rats. Cardiovasc Res 43: 237–247, 1999. 38. Pinzani M, Rosselli M, Zuckermann M. Liver cirrhosis. Best Pract Res Clin Gastroenterol 25: 281–290, 2011. 39. Pizcueta MP, de Lacy AM, Kravetz D, Bosch J, Rodes J. Propranolol decreases portal pressure without changing portocollateral resistance in cirrhotic rats. Hepatology 10: 953–957, 1989. 40. Raso GM, Esposito E, Iacono A, Pacilio M, Cuzzocrea S, Canani RB, Calignano A, Meli R. Comparative therapeutic effects of metformin and vitamin E in a model of non-alcoholic steatohepatitis in the young rat. Eur J Pharmacol 604: 125–131, 2009. 41. Rosado E, Rodriguez-Vilarrupla A, Gracia-Sancho J, Tripathi D, Garcia-Caldero H, Bosch J, Garcia-Pagan JC. Terutroban, a TPreceptor antagonist, reduces portal pressure in cirrhotic rats. Hepatology 58: 1424 –1435, 2013. 42. Sartoretto JL, Melo GA, Carvalho MH, Nigro D, Passaglia RT, Scavone C, Cuman RK, Fortes ZB. Metformin treatment restores the altered microvascular reactivity in neonatal streptozotocin-induced dia-

43. 44.

45. 46. 47. 48. 49.

50.

51.

G309

betic rats increasing NOS activity, but not NOS expression. Life Sci 77: 2676 –2689, 2005. Sena CM, Matafome P, Louro T, Nunes E, Fernandes R, Seica RM. Metformin restores endothelial function in aorta of diabetic rats. Br J Pharmacol 163: 424 –437, 2011. Verbeke L, Farre R, Trebicka J, Komuta M, Roskams T, Klein S, Elst IV, Windmolders P, Vanuytsel T, Nevens F, Laleman W. Obeticholic acid, a farnesoid X receptor agonist, improves portal hypertension by two distinct pathways in cirrhotic rats. Hepatology 59: 2286 –2298, 2014. Verma S, Yao L, Dumont AS, McNeill JH. Metformin treatment corrects vascular insulin resistance in hypertension. J Hypertens 18: 1445–1450, 2000. Wiest R, Groszmann RJ. The paradox of nitric oxide in cirrhosis and portal hypertension: too much, not enough. Hepatology 35: 478–491, 2002. Xiao H, Ma X, Feng W, Fu Y, Lu Z, Xu M, Shen Q, Zhu Y, Zhang Y. Metformin attenuates cardiac fibrosis by inhibiting the TGFbeta1-Smad3 signalling pathway. Cardiovasc Res 87: 504 –513, 2010. Yang YY, Lee KC, Huang YT, Wang YW, Hou MC, Lee FY, Lin HC, Lee SD. Effects of N-acetylcysteine administration in hepatic microcirculation of rats with biliary cirrhosis. J Hepatol 49: 25–33, 2008. Yang YY, Lee TY, Huang YT, Chan CC, Yeh YC, Lee FY, Lee SD, Lin HC. Asymmetric dimethylarginine (ADMA) determines the improvement of hepatic endothelial dysfunction by vitamin E in cirrhotic rats. Liver Int 32: 48 –57, 2012. Zhang CX, Pan SN, Meng RS, Peng CQ, Xiong ZJ, Chen BL, Chen GQ, Yao FJ, Chen YL, Ma YD, Dong YG. Metformin attenuates ventricular hypertrophy by activating the AMP-activated protein kinaseendothelial nitric oxide synthase pathway in rats. Clin Exp Pharmacol Physiol 38: 55–62, 2011. Zhang X, Harmsen WS, Mettler TA, Kim WR, Roberts RO, Therneau TM, Roberts LR, Chaiteerakij R. Continuation of metformin use after a diagnosis of cirrhosis significantly improves survival of patients with diabetes. Hepatology 60: 2008 –2016, 2014.

AJP-Gastrointest Liver Physiol • doi:10.1152/ajpgi.00010.2015 • www.ajpgi.org