Several mechanisms are involved in regulating NO production following eNOS activation. Studies in relation to decreased hepatic eNOS activity have, to date, focused on inhibitors of eNOS activity such as asymmetric dimethylarginine (ADMA)[26,27] and caveolin-1 (major coat protein of endothelial caveolae)[28], or on the process affecting the post translational modification of eNOS[29]. Furthermore, trafficking and proper subcellular localization of eNOS are also critical for regulation of its activity[29,30]. One such protein, eNOS trafficking inducer (NOSTRIN), identified in a yeast two-hybrid approach, has been demonstrated to interact physically with eNOS via its C-terminal SH3 domain and regulated its function[29-31].

An emerging body of evidence indicated that ADMA, a deleterious endogenous inhibitor of NO synthases and thus presumed to be a marker of hepatic dysfunction in cirrhosis with PHT. One mechanism thought to be partially responsible for the reduction in NO and resultant ED in liver disease is an increase in the levels of the endogenous inhibitor of NOS, ADMA[19,32]. In this regard, our previous studies have shown evidence that increased ADMA contributes to reduced hepatic NO biosynthesis as a consequence of altered hepatic vascular function in cirrhosis[26,27,33]. In critically ill patients to whom admitted in ICU, hepatic dysfunction was associated with elevated ADMA levels and was identified as an independent predictor of mortality[34]. Furthermore, increased plasma ADMA was reported recently in biopsy proven non-alcoholic fatty liver disease (NAFLD) patients[35], hepatic vein of patients with compensated cirrhosis[36] and decreased following liver transplantation; thus the significant improvement of liver function[37] propose an important role for ADMA in clinical medicine. Elevated plasma ADMA has also recognised as an important risk factor for cardiovascular disease[38], coronary heart disease[39] and chronic renal failure[40]. Consequently, increased ADMA may be related to elevated activity of protein methyltransferase (PRMT), which is responsible for the methylation of arginine residues in cellular proteins[41]. Moreover, increased ADMA level was correlated with the severity of inflammation and levels of increased proinflammatory cytokine such as tumor necrosis factor (TNF)[42,43]. Laleman et al[44] showed in cirrhotic animals infusion of ADMA and NG-nitro-L-arginine methyl ester (L-NAME), other known inhibitors of NOS synergistically aggravated and resulted in paradoxical vasoconstriction, which associated with a further decrease in NOx levels. The pathophysiological increase in hepatic ADMA concentration observed in cirrhotic rats manifest decrease of hepatic eNOS activity.

Furthermore, the intracellular levels of these methylated arginines are regulated through their metabolism to citrulline and dimethylamine by its hepatic specific enzyme called, dimethylarginine diaminohydrolase (DDAH)[41,45] (Figure 1). Two isoforms of DDAH have been identified and are widely expressed in human and rodent liver. DDAH 1 is an important isoform for the regulation of hepatic and systemic ADMA concentration and is present higher levels in tissue that expressing neuronal NOS (nNOS)[41,45]. The other DDAH isoform (DDAH 2) has an important effect in regulating NO activity, and is mainly found in tissue expressing eNOS and inducible NOS (iNOS)[41,45,46]. It is well known that increased intracellular DDAH plays a critical role in regulating tissue ADMA concentration[26,27,33,47], therefore alterations of DDAH activity and expression lead to change in intracellular ADMA concentrations and concomitant NO synthesis. In vitro, human umbilical vein endothelial cells (HUVECS) exposed to prolonged (48 h) TNF-α show eight fold increase of ADMA, compared to control medium and associated DDAH activity was decreased to almost 60% of baseline values[48]. DDAH is a redox sensitive enzyme and is thus subject to inhibition by oxidants derived from endothelial superoxide[47,49], and antioxidant treatment corrects DDAH inhibition in in vivo[33,50]. Cirrhotic rat livers showed an increased O₂^- content compared to control rat livers and was ameliorated by adenoviral gene delivery of superoxide dismutase, increases NO bioavailability, improves intrahepatic endothelial function and reduces portal pressure[51]. In contrast to ADMA, SDMA, its vasoactive stereoisomer, has no effect on inhibition of NO synthases but competes with arginine for cellular transport across the y+ transporter[38,52]. Recently, Siroen and co-workers have shown that the human liver takes up substantial amounts of SDMA from the portal and systemic circulation and suggested that high plasma levels of SDMA may have hemodynamic consequences similar to those reported for ADMA[53]. However, in patients with alcoholic cirrhosis, noted plasma SDMA level was within the normal limit[54].

Toll-like receptors (TLRs) belong to the family of transmembrane pattern-recognition receptors that recognize pathogen-associated molecular patterns, which including LPS[87]. TLR4 expressed on both parenchymal and non-parenchymal cell types in the liver and its activation trigger hepatic innate immune signaling, and may contribute to endothelial dysfunction and intrahepatic vascular tone in patients with cirrhosis[8,87,88]. In addition, the potential role of TLR4 in mediating renal injury in patients with cirrhosis was described recently[89]. The seminal observations made in this study were that the renal expression of TLR4 and the excretion of TLR4 protein was significantly higher in patients with cirrhosis who presented with acute deterioration and had renal dysfunction compared with those that did not[89]. Furthermore, urinary TLR4 was associated with significantly greater risk of death in patients with renal dysfunction and in those with superimposed inflammation[89]. TLR4 expressed on the surface of several cell types, including endothelial cells and its activation shown to reduce NO concentration, resulting ED[90]. Accordingly, anti-TLR4 treatment improved endothelium-dependent relaxation, and improved NO[91]. TLR4 also contributes to the increased ROS production and ED in hypertension, diabetes and obesity[92]. Recently, Benhamou et al[93] revealed that both TLR2 and TLR4 in mediating endothelial dysfunction and vascular remodeling in primary arterial antiphospholipid syndrome. TRL4 signalling leads to activation of NF-κB[94], a pathway associated with endothelial injury[95], and increased TLR4 expression has also shown in advanced liver disease[89,96]. LPS induced TLR4 mediated proinflammatory signalling has also showed in human hepatic stellate cells (HSC)[97], and functional expression of TLR9 has detected in sinusoidal hepatic endothelial cells and hepatocytes[98]. Furthermore, several animal studies support the importance of TLR4 in hepatic fibrosis, and TLR4 knockout mice showing less fibrosis induced by BDL or carbon tetrachloride (CCl₄) compared to wild type[99]. These results suggested an important function of TLRs on the development of inflammatory pathology in hepatic cirrhosis. Stadlbauer et al[96] observed that in AH patients, the increased TLR 2, 4 and 9 expressions correlated with neutrophil dysfunction and endotoxemia, albumin an endotoxin scavenger attenuated these complaints by decreasing TLRs expression. Several lines of evidence exist implicating gut derived endotoxemia in the pathogenesis of portal hypertension[100]. Administration of norfloxacin, a selective gut decontaminant prophylactic reduced endotoxin levels, TLR4 expression and decreased NO-mediated forearm vasodilatation and improved survival in cirrhosis[94,101].

Angiotensin II

The renin-angiotensin system (RAS) plays a key physiological role in regulating vascular function. In the pathophysiology, RAS has also shown to promote vascular injury by triggering ED, vascular remodelling and vascular inflammation[102,103]. Angiotensin (Ang) II the core composition of the RAS involved in many chronic diseases, which including hepatic cirrhosis[104]. Increased Ang II causes endothelial dysfunction, vasoconstriction, sodium water retention, elevated blood pressure, ROS generation, inflammatory mediators and pro-fibrotic cytokines[105]. The adverse effects of Ang II induced ED is mediated by interaction with the plasma membrane AT 1 receptors (Ang II type 1 receptors) and causes NO reduction by inducing eNOS enzyme dysfunction and promoting NOS uncoupling. Thus, pharmacological inhibition of the production or actions of Ang II receptor blockers now represents an effective strategy to delay the progression of endothelial dysfunction in experimental models and humans[105,106]. In this context, previous clinical studies have shown evidence that RAS play an important role in the elevation of the ADMA concentration in hypertensive patients, and blockade of Ang II by ACEI or Ang II receptor blocker (ARB) significantly attenuates the elevated level of ADMA, resulting in endothelial protection[107,108]. Pharmacological blockade of angiotensin II receptors using the drug Candesartan cilextil (CC) may attenuate the progression of liver cirrhosis and endothelial dysfunction. In HUVECS, CC increased the eNOS protein level, inhibited the expression of nicotinamide adenine dinucleotide phosphate oxidase (Nox) subunits and Ang II induced intracellular ROS and nitric oxide, and promoted the extracellular release of nitric oxide[109] suggesting that it augmented the bioavailability of nitric oxide. Thus, CC administration may attenuate ED and for the future therapeutic approach in portal hypertensive patients with cirrhosis. Ang II can also activate NADPH oxidase (Nox), leading to increased ROS generation and commencing ED in cirrhosis[104].

TXA₂, a vasoactive prostanoid and COX metabolite, increased intrahepatic vascular tone in cirrhosis, more specifically in the phenomenon of hyper responsiveness to vasoconstrictors and caused intrahepatic ED[110,111]. Administration of COX inhibitors such as flurbiprofen and nitroflurbiprofen (NO releasing COX inhibitor) to cirrhotic rats result in decreased hepatic TXA₂ production and an increased intrahepatic nitrate/nitrite (an index of NO synthesis) concentration thereby by attenuating intrahepatic vascular resistance, endothelial dysfunction, and hepatic hyper reactivity to vasoconstrictors[111]. TXA₂ and PGI_2, the other COX 2 derivatives may act simultaneously, producing a compensatory effect that reduces NO release and may limit the hyperdynamic circulation in cirrhosis[112]. The use of indomethacin, a COX inhibitor has shown to prevent liver fibrosis, and rise in portal hypertension in liver cirrhosis[113]. Indeed, study also shows that COX inhibitors could worsen the hyperdynamic circulation associated with liver cirrhosis[112]. TXA₂ induces vasoconstriction by activating the TXA₂/prostaglandin-endoperoxide (TP) receptor[114]. TP receptor ligands include TXA₂, PGH₂, and isoprostanes[115,116]. TXA₂ acts through its G-protein-coupled receptor leading to vasoconstriction by activating the RhoA/Rho-kinase pathway, and by increasing calcium levels in HSC[117]. Oral administration of terutroban, a specific antagonist of the TP-receptor[118] has shown to attenuates inflammation and oxidative stress and reduce RhoA/Rho-kinase-dependent signaling and restore NO bioavailability in endothelial cells[119] may represent a useful agent in the treatment of endothelial dysfunction in cirrhosis with portal hypertension.

A most potent vasoconstrictor endothelin (ET) -1 regarded as a key player in ED, primarily binding to G-protein coupled receptors such as ETA and ETB and acts in a paracrine fashion[120]. Previous study has shown that ETB receptors present in endothelial cell can elicit endothelium-dependent relaxation by improving NO release by contrast, ETA and ETB receptors present on the fibroblasts and smooth muscle cells trigger vasoconstriction and inflammation[121]. Cirrhotic rats with diabetes showed higher intrahepatic ET-1 vasoresponsiveness than normoglycemic cirrhotic rats[122]. ED also found in patients with insulin resistance (IR)[123,124]. In this context, IR can be triggered by ET-1 infusion in rats by activating phosphatidylinositol (PI) 3-kinase activity in smooth muscle cells in an ETA dependent manner and treatment with ETA receptor antagonists results in improvement of insulin sensitivity and associated endothelial function via an inhibition of PI 3-kinase activity[125]. In this context, a very recent study pointed out that LPS stimulation to portal hypertensive rats showed enhanced renal vascular response to ET-1 through ETA overexpression[126].

Lectin-like oxidised LDL receptor-1 (LOX1) is a key receptor for oxidised low-density lipoprotein (Ox-LDL) and identified in endothelial cells, and considered as a marker of ED in various pathological setting[127,128]. LOX-1 promotes ROS generation augments endothelial adhesiveness to monocytes and inhibits NO synthesis[129]. The recent review by Lubrano et al[130] described in detail about the relationship between LOX1 and ROS. Furthermore, increased expression of LOX1 was found in the placenta of women with intrahepatic cholestasis during pregnancy[131]. LOX-1 polymorphism also associated with liver disease severity in non-alcoholic steatohepatitis[132] and could be a biomarker for patients with endothelial dysfunction in liver cirrhosis.

Protease activated receptors (PARs) are G protein-coupled receptors which, mediating cellular effects of some proteases of the activated coagulation system such as thrombin, trypsin or metalloproteinase[133]. ECs express PAR1, PAR2 and PAR4[134]. Endothelial PARs play important roles in the crosstalk between coagulation and inflammation in sepsis[133]. Acute PAR1 activation causes an increase in vascular permeability, presumably due to direct endothelial contractile responses[133]. PAR1 deficiency and blockade has shown to reduce inflammation in a mouse model of colitis[135]. Moreover, activation of protease through its receptor following thrombus formation, hemorrhage and inflammation led to the conversion of ECs to a proinflammatory phenotype and may result in vascular lesion development[133]. Garcia et al[136] have shown in cultured ECs, PAR1 activation stimulates the production of prostacyclin and NO, consisting with other reports shown in in vivo that PAR1 activators cause hypotension when injected intravenously and cause NO mediated vasodilation[137]. In this context, Knight et al[138] and Sakata et al[139] reported that PAR2 promotes experimental liver fibrosis by increase of TGF β production in mice and to induce a profibrogenic phenotype in human HSCs. Thus, targeting PARs and using its antagonists in endothelial dysfunction with cirrhosis may represent a novel therapeutic approach in preventing portal hypertension in cirrhosis.

Adiponectin is a protein hormone synthesized by adipose tissue. Plasma physiological concentration of adiponectin represents 0.05% of all plasma proteins and is a key component in the relationship between adiposity, inflammation and insulin resistance[140]. Previous studies have shown evidence that the association between hypoadiponectinemia and ED[141,142]. Wang et al[142] showed in ECs that adiponectin stimulating NO synthesis by activating AMPK mediated pathway. Furthermore, adiponectin knockout mice exhibited impaired endothelium dependent vasodilation and NO production[143]. Earlier published studies have pointed out that adiponectin administration significantly increases NO production by regulating eNOS enzyme activity and its phosphorylation and maintain endothelial function[141]. However, despite the hepatoprotective effect of adiponectin have shown in NAFLD and other chronic liver ailments, the plasma concentration of which increased in patients with cirrhosis of different aetiologies. Indeed, the factors related to elevated levels of adiponectin in cirrhosis are not yet completely understood. In this context, Tietge et al[144] showed an evidence that increased adiponectin levels in cirrhotic patients correlate exclusively with reduced liver function and altered hepatic haemodynamics. Furthermore, Salman et al[145] and Tacke et al[146] reported that an elevated adiponectin correlated with inflammation and liver damage and high levels were found in human cholestasis as well as in an animal model of cirrhosis[145,146]. Thus, adiponectin may serve as a novel biomarker for cholestasis in liver cirrhosis and agents that modifying adiponectin concentration in liver cirrhosis may use as a potential diagnostic tool but also as therapeutic target for ED in cirrhosis.

Liver plays a significant role in lipid homeostasis including several stages of lipid synthesis and transportation. Thus, it is reasonable to anticipate an abnormal lipid profile associated with the progression of hepatic dysfunction. Furthermore, hyperlipidaemia is main risk factor for ED, which is a common indicator in patients with hepatic cirrhosis[147]. Accumulated evidence indicates that increased hepatic and plasma free fatty acid (FFA) concentration led to hyperlipidemia and may cause ED in cirrhosis. Previously it has been shown that FFA trigger HUVECs apoptosis and inhibit cell cycle progression[148]. In this regard, palmitic acid a key FFA in the bloodstream, exposure to ECs causes cell necrosis and the release of proinflammatory cytokines[149,150] consistent with other report showing in cultured bovine retinal pericytes, in which palmitate can induce apoptosis by promoting oxidative stress[151]. Recently Ristic-Medic et al[152] observed in cirrhotic patients, increased levels of palmitic acid and total saturated fatty acids when compared to healthy controls. Thus, FFA play an important role in cirrhosis with ED and agents that reduce palmitic acid concentration would be used as a possible future target for therapy. One such agent, eicosapentaenoic acid (EPA), ω-3 polyunsaturated fatty acid (PUFA) is abundant in fish oil has shown to enhance the production of NO, via activating eNOS and improve normal vascular endothelium[153]. Very recently Lee et al[154] have demonstrated that treatment with EPA protects against palmitic acid induced ED through activation of the AMPK-eNOS mediated pathway. Highly purified EPA also shown to prevent the development of inflammation and hepatic fibrosis in rats[155,156]. In addition, peroxisome proliferator-activated receptor (PPAR)-α, a member of the nuclear receptor superfamily and a key regulator of fatty acid homeostasis[157], has been shown to improve endothelial dysfunction and portal pressure in cirrhotic rats[158]. The above indices, therefore, provide a rationale for novel insights into the pathophysiology of ED and the potential for the development of novel biomarkers and therapeutic approaches in patients with endothelial dysfunction and advanced liver disease.

Ascorbic acid (vitamin C) has been shown to improve the NO-dependent vasodilatation in vascular beds of patients with conditions characterized by marked ED in cirrhosis[159]. The other antioxidant α-tocopherol (vitamin E) has also shown to improvement of hepatic ED by suppressed hepatic ADMA and oxidative stress and improved hepatic NO in cirrhotic rats[160]. In addition, folic acid, a superoxide scavenging vitamin B9 and its active metabolite 5-methyltetrahydrofolate (5-MTHF) has been shown to restore ED in patients with many cardiovascular diseases[161]. Folic acid mainly involved in downregulating eNOS derived superoxide and eNOS uncoupling thereby improving regeneration of BH₄ from BH₂, by preventing BH4 oxidation, which results in increased NO[161]. The beneficial effects of 5-MTHF have shown in decompensated cirrhotic patients recently[162]. Superoxide dismutase (SOD) gene transfer also has shown to reduce portal pressure in CCl₄ cirrhotic rats with portal hypertension through reducing oxidative stress and increased NO bioavailability[61]. Tempol, a SOD mimetic reduces superoxide, increases nitric oxide, and reduces portal pressure in sinusoidal endothelial cells of cirrhotic livers[163]. Furthermore, flavonoids, an integral part of the human diet have been shown to confer protective effects on vascular endothelial function in humans[164], and its protective effects are chiefly ascribed to their antioxidant and vasodilatory actions[165]. Very recently, Hsu et al[15] found that green tea polyphenol decreases the severity of portosystemic collaterals and mesenteric angiogenesis in rats with liver cirrhosis. Furthermore, De Gottardi et al[166] observed in cirrhotic patients with portal hypertension that dark chocolate blunted the postprandial increase in hepatic venous pressure gradient (HVPG) by improving flow-mediated hepatic vasorelaxation and ameliorated systemic hypotension. In addition, Lin et al[167] showed quercetin supplementation has associated with multifactorial potential as well as down-regulation of NF-κB and TGF-β/Smad signalling, probably via interference with TLR signalling. Resveratrol, a natural polyphenolic flavonoid present higher amount in grapes has shown to reduce portal pressure by attenuating ED. Moreover, resveratrol supplementation also results in reducing oxidative stress and upregulating eNOS expression without affecting systemic hemodynamics in cirrhotic rats[168].

Statins (HMG-CoA reductase inhibitors) lower serum cholesterol concentrations and exhibits beneficial therapeutic effects in cirrhotic patients, as evidenced by various clinical trials[169-171]. Abraldes et al[169] demonstrated that simvastatin improve hepatic NO generation and endothelial function and lowers portal pressure in patients with cirrhosis. Moreover, atorvastatin has been shown to prevent liver inflammation and HSC activation induced by Ang-II infusion[172]. Schwabl et al[172] found that pioglitazone, an insulin sensitiser decreases portosystemic shunting by modulating inflammation and angiogenesis in cirrhotic and non-cirrhotic portal hypertensive rats. Additionally, Sorafenib, a tyrosine kinase inhibitor approved in the treatment of hepatocellular carcinoma, has shown to have beneficial effects in reducing portal pressure in cirrhosis[173]. The other multitarget receptor tyrosine kinase inhibitors such as Sunitinib and Imatinib were also shown to use for the treatment of portal hypertension[174] and may have a potential role in regulating ED in cirrhosis through NO mediated mechanisms. However, further encouraging clinical studies of statins strategy to ED in cirrhosis needs to be explored.

Human serum albumin (HSA) is one of the most frequent treatments in patients with decompensated cirrhosis[175]. It also reduced the severity of other chronic liver diseases such as HE and HRS and improved survival of patients with spontaneous bacterial peritonitis (SBP)[175-177]. In addition, albumin has been demonstrated to have a clinically significant beneficial effect on ED and survival during experimental endotoxemia[178]. It also reduced sequential organ failure assessment (SOFA) score in critically ill patients with hypoalbuminemia[179]. Furthermore, pentoxifylline and N-acetylcysteine, the other known anti-inflammatory agents have shown to associated with reduced the risk of inflammation and ED in cirrhosis[180-182]. Antibiotics such as quinolones and rifaximin and high-density lipoprotein (HDL) treatment have shown to associate with the reduction of inflammation and portal pressure by neutralising portal bacterial endotoxin load[20,101,183]. Thus, anti-inflammatory agents would improve NO bioavailability and reduce ED, considered a potential therapeutic approach for the management of portal hypertension in cirrhosis.

Obeticholic acid, a synthetic farnesoid X receptor (FXR) ligand belongs to a nuclear receptor superfamily of transcription factor, which plays an important role in bile acid and lipid metabolism[184,185] has also been the subject of considerable attention over recent years[50]. FXR agonists have numerous target genes including DDAH1[186]. Our previous study has shown evidence that obeticholic acid significantly increases hepatic DDAH-1 and eNOS activity and improved NO bioavailability in cirrhotic rats, leading to improvement in endothelial function and associated drop in portal pressure[33]. Similarly, a multi-centre phase 2a trial of obeticholic acid in decompensated cirrhotic patients show a trend towards a drop in portal pressure[187]. Another promising approach of transfection of cirrhotic liver with DDAH-1 decreased ED and portal hypertension in BDL rats[188]. Furthermore, other member of the nuclear receptor superfamily, PPAR- α also has shown to improve endothelial dysfunction and portal pressure in cirrhosis[158].

AST-120, an oral adsorbent carbon microspheres and ammonia-lowering agent[189] has shown to reduce ED in adenine-induced uremic rats[190]. AST-120 treatment was also shown to prevent the progression of HE in cirrhotic rats[189] and chronic kidney disease (CKD) in a clinical setting[191], which may be used for the treatment of ED in cirrhosis. Indeed, further studies would be needed for describing the potential role of AST-120 on NO mediated ED in cirrhosis. Furthermore, we established a new promising therapy such as OCR- 002 (ornithine-phenylacetate) for the treatment of HE[27,192]. OCR-002 is currently being advanced in the clinic and also effective in reducing PHT by lowering ammonia mediated inflammation and improving NO bioavailability[193] and may consider for the future therapeutic approach for the management of ED and associated PHT in patient with cirrhosis.

Non-selectiveβblockers

Carvedilol is a non-selective β blocker with alpha-1 adrenergic blocker activity has been shown to amelioration of oxidative stress and restoration of antioxidant enzyme activities, and attenuation of NF-κB mediated inflammation in chronic liver disease[194]. Interestingly, Reiberger et al[195] observed in BDL cirrhotic rat that nebivolol, a third generation beta-blocker increased splanchnic blood flow and portal pressure via NO mediated signalling. In this context, Ma et al[196] demonstrated in myocardial infraction that targeting NO with a nebivolol treatment improves diastolic dysfunction through reducing myocardial oxidative stress by enhancing 5’-AMP-activated protein kinase and Akt activation of NO biosynthesis. Moreover, long-term nebivolol administration reduces renal fibrosis and prevents endothelial dysfunction in a rat hypertensive model[197]. In this context, Mookerjee et al[11] have pointed out that specific controlled studies are addressing the use of β-blockers in patients with severe decompensation of cirrhosis with high risk of sepsis and renal dysfunction are inadequate. Hence, further clinical studies on the effect of β-blockers through NO mediated pathway are challenging in liver cirrhosis.

Additionally, Candesartan cilextil (CC), a selective angiotensin II type I (AT1) receptor antagonist widely used as an antihypertensive in clinical practice has shown to improve ED[198,199]. In addition, recent clinical study has shown evidence that CC administration was safe and well tolerated to compensated cirrhotic patients, without an underlying cause of renal failure or hepatic decompensation[200]. Given the substantial experimental evidence that CC has the potential beneficial effect in distinct human diseases by blocking Ang-II mediated AT1 receptor and may improve ED and NO bioavailability and associated mechanism in cirrhosis.