The role of microcirculatory dysfunction, the innate immune system and intestinal permeability in the early onset of acute-on-chronic liver failure.

Amongst digestive diseases, cirrhosis is the most common non-neoplastic cause of mortality in the Western world. Furthermore, its prevalence is likely to increase in the coming decades because of the increasing incidence of nonalcoholic fatty liver disease. In this project, we focused on the pathophysiology of acute-on-chronic liver failure (ACLF, chapter IA) to gain a further understanding of several of the lethal complications in cirrhosis: bacterial translocation from the gut (BTL), intrahepatic microcirculatory dysfunction and portal hypertension, immune dysfunction and the process of hepatic fibrogenesis. In a first set of experiments, we evaluated different translational rat models of hepatic disease to evaluate the presence of gut barrier dysfunction and ultimately intestinal bacterial translocation (Chapter III). Surprisingly, in a well validated model of toxic cirrhosis and portal hypertension (thioacetamide-induced cirrhosis, TAA), BTL did not occur, despite the presence of increased gut permeability. However, in both models of short-term and long-term cholestasis, BTL was clearly present, consistent with the clinical presentation of patients suffering from ACLF (jaundice) and data by other groups showing that also in humans, bacterial translocation occurs only in the context of decompensated cirrhosis (again with jaundice) and even acute cholestatic liver injury. These findings led to the hypothesis that the farnesoid-X receptor (FXR), a chief regulator of bile acid metabolism, is critically involved in hepatic decompensation, because it not only reverses cholestasis as such, but appears also expressed by intestine, immune cells and systemic endothelium, with increasing reports of diverse functionality potentially important in the context of ACLF (Chapter IB). For this reason, we preferred the use of a highly potent and selective FXR agonist (obeticholic acid, INT-747) above loss-of-function experiments to better resemble the human condition and to avoid the interference of compensatory pathways as seen after genetic knock-down. In a first set of experiments, we gained proof of concept that FXR is indeed deficient in the ileum of cholestatic rats (by means of bile-duct ligation, BDL), and by restoration of its functionality with INT-747, gut permeability is restored selectively in the ileum, resulting in reduced BTL (Chapter IV). Interestingly, this is not associated with major alterations in neither bile acid pool nor composition, suggesting another functional interaction than purely via FXR-mediated alterations in bile acid metabolism. From this perspective, we were able to document an important decrease in both the systemic and local intestinal immune response after INT-747 treatment in BDL rats (especially targeting natural killer cells and interferon-γ expression) with secondary beneficial effects on gut dysfunction, thus preventing BTL observed in this model. Given the fact that we also found a decrease in hepatic fibrosis upon FXR agonist treatment in these BDL rats, we then aimed to investigate whether this is only secondary to the reported decrease in BTL (and subsequently decreased Kupffer cell activation) or through a direct effect within the hepatic micro-environment. For these studies, to avoid a possible interference of BTL, we evaluated TAA rats subjected to 4 weeks of INT-747 treatment, both in a prophylactic and a therapeutic setting (Chapter VI). As such, we were able to document a decrease in hepatic fibrosis by FXR agonists, again related to an overall decrease in hepatic inflammation and subsequent decrease in hepatic stellate cell (HSC) activation as well as cytokines involved in HSC-related fibrogenesis, fibrinolysis, apoptosis and proliferation. A direct effect on HSC, as reported by other groups, was excluded in vitro. This reduction in hepatic fibrosis was in turn associated with a decrease in the structural component of intrahepatic vascular resistance (IHVR) and thus decreased portal hypertension. In a final set of experiments, we then aimed to investigate the pure hemodynamic effects in the hepatic microcirculation by INT-747, by applying only short-term (24 hour) treatment in both BDL and TAA cirrhotic rats (Chapter V). In both models, we were able to document hepatic FXR pathway down-regulation/deficiency. Conversely, following FXR re-activation by INT-747, we were able to document decreased portal hypertension by reduction also of the functional component of IHVR. This was related to an increased hepatic expression of phosphorylated endothelial nitric oxide synthase (P-eNOS) by liver sinusoidal endothelial cells. Again in vitro, a direct effect on HSC contractility by INT-747 could be excluded. Interestingly, the upstream mechanisms of increased P-eNOS differed upon the etiology of the underlying liver disease: in TAA rats this was related to alterations in the Rho kinase pathway, while in BDL rats this depended upon decreased asymmetric dimethylarginine (ADMA), secondary to increased dimethylarginine dimethylaminohydrolase-2 (DDAH-2). 
In conclusion, we have shown that FXR deficiency in the gut-liver axis is critically associated with many of the lethal complications in cirrhosis: bacterial translocation from the gut, hepatic and intestinal inflammation, portal hypertension and the development of hepatic fibrosis. We have also demonstrated that functional re-activation of the FXR pathway by means of INT-747 is able to attenuate or reverse many of these detrimental processes in different experimental settings of liver injury. These data explain in part the beneficial findings in clinical trials at this time conducted in patients suffering from primary biliary cirrhosis, non-alcoholic fatty liver disease and portal hypertension and supports extending these trials with FXR agonists to other forms of chronic liver injury.