Patients with chronic cholestatic diseases of the biliary tree, such as primary sclerosing cholangitis (PSC), have long suffered from lack of efficacious medical treatments.1 PSC is a rare disease that is characterized by chronic inflammation and fibrotic obliteration of the intrahepatic and/or extrahepatic bile ducts.2 Initially, the destruction of the bile ducts leads to cholestasis and hepatic fibrosis, which ultimately progresses to cirrhosis and end-stage liver disease.3 Liver transplantation is currently the only life-extending therapeutic option for patients with end-stage disease.4 PSC is also complicated by an elevated potential for the development of malignancies of the biliary system (i.e., cholangiocarcinoma).5 Although the etiology of PSC is unknown, the majority of cases occurs in association with inflammatory bowel disease,6 and there is an increasing body of evidence that suggests an autoimmune component of the pathogenesis of PSC.3 However, PSC lacks the features common to typical autoimmune disease and responds poorly to immunosuppressive therapies.4
To date, ursodeoxycholic acid (UDCA) is the most well-studied therapy for PSC.3 UDCA has been shown to improve serum liver tests; however, controlled trials of UDCA therapy (12-15 mg/kg body weight/day) did not have consistent effects on the improvement of liver histology and did not show a survival benefit.7 Several small trials have suggested that higher doses of UDCA (25-30 mg/kg body weight/day) might be more effective than the standard dose.8–10 However, a recent randomized trial in 198 patients demonstrated no effect of higher UDCA dosages on symptoms and other clinical parameters.11 Due to the lack of effective therapies, there is a dire need to develop novel therapeutic modalities for PSC and other cholangiopathies.
In this issue, Halilbasic and colleagues present findings that 24-norursodeoxycholic acid (norUDCA) has unique physiologic and therapeutic properties in the multidrug resistance protein 2 null (Mdr2−/−) mouse, which is an in vivo model of cholangiopathies that closely resembles sclerosing cholangitis with obliterative fibrosis of the bile ducts.12 NorUDCA is a C23 homolog of UDCA with one less methyl group on its side chain.13 In humans, norUDCA undergoes little N-acyl amidation and rather undergoes biotransformation (i.e., glucuronidated and highly hydroxylated) as a metabolized drug.13 The unconjugated form of norUDCA is a weak acid that can be reabsorbed from the bile by cholangiocytes and resecretion by hepatocytes in a process termed cholehepatic shunting, which is associated with a HCO3−-rich hypercholeresis from cholangiocytes.13 The parent compound UDCA can also trigger HCO3−-rich hypercholeresis from cholangiocytes, but this will only occur in vitro once taurine stores are depleted and unconjugated UDCA is the predominant form of the bile acid.14 Taurine-conjugated bile acids can only undergo cholehepatic shunting via the apical sodium-dependent transporter Asbt expressed by cholangiocytes,15 whereas the unconjugated norURSO is permeable to the membrane and readily available for the cholehepatic shunt. This would suggest that norUDCA would be more efficacious in the stimulation of the cholehepatic shunt. In fact, the same group previously demonstrated that norUDCA was superior to UDCA in the treatment of sclerosing cholangitis pathology observed in Mdr2−/− mice.16
Mdr2−/− mice lack the canalicular phosphatidylcholine flippase MDR2 (MDR3/ABCB4 in humans), which is the only known transporter that contributes to the phospholipid content of the bile that is critical in the formation of mixed micelles.17 Knockout of MDR2 in mice results in an elegant animal model of sclerosing cholangitis that closely resembles the human cholangiopathy.18 Interestingly, mutations of MDR3 have been linked to a number of hepatobiliary disorders including progressive familial intrahepatic cholestasis, intrahepatic cholestasis of pregnancy, intrahepatic cholelithiasis, biliary cirrhosis, and sclerosing cholangitis.19 In the Mdr2−/− mouse cholangiopathy model, it is postulated that regurgitation of bile from leaky ducts into the portal tracts, leading to induction of periductular inflammation due to the absence of phospholipids and increased buildup and damage by “toxic bile”, followed by activation of periductular fibrogenesis, finally causes obliterative cholangitis due to atrophy and death of cholangiocytes.18 The current thought is that the accumulation of “toxic bile” by alterations in bile secretion at the hepatocellular and cholangiocellular levels may play a key role in the pathogenesis of cholestatic liver diseases.19 In Mdr2−/− mice, norUDCA ameliorated, as evidenced by markedly improved liver tests and liver histology, and significantly reduced hydroxyproline content and the number of infiltrating neutrophils and proliferating hepatocytes and cholangiocytes.16 This therapeutic effect of norUDCA in Mdr2−/− mice was thought to involve the stimulation of bile flow and the accompanied effect of flushing of injured bile ducts of toxic bile acids and by the induction of detoxification and elimination routes for bile acids.16 In summary, norUDCA had antifibrotic, anti-inflammatory, and antiproliferative effects that resulted in healing of the cholangitis and fibrosis observed in Mdr2−/− mice.16
In the current study, Halilbasic and colleagues clearly demonstrate that the side chain structure of norUDCA determines the unique physiological and therapeutic properties of the bile acid in mice.12 These unique properties were not possessed by its derivatives tauro-norUDCA (taurine-conjugated norUDCA) or dinorUDCA (a norUDCA homolog with a shorter side chain).12 NorUDCA stimulated an increase in hepatic and serum bile acid levels and the stimulation of biliary HCO3− output in vivo and stimulated fluid secretion in isolated bile duct units, which supports the concept of cholehepatic shunting of conjugation-resistant bile acids.12 Only norUDCA improved cholangitis and fibrosis present in the liver tissues of the Mdr2−/− mice, whereas dinorUDCA actually worsened cholangitis and fibrosis.12 The findings with dinorUDCA indicate the importance of in vivo evaluation of potential therapeutic compounds and their metabolites in animal models of liver disease. In addition, UDCA also induced choleresis in Mdr2−/− mice resulting in disruption of cholangioles and aggravation of liver injury, which was in direct contrast with the beneficial effects of norUDCA. The mechanisms for the differences observed between norUDCA and UDCA and dinorUDCA are not yet known, but one could postulate that it is due to the activation of cholehepatic shunting by norUDCA resulting in the activation of increased canalicular bile flow and ductal biliary HCO3− output that washes toxic bile acids out of the liver. Their work points to the potential importance of designing therapeutics for cholestatic liver diseases that will have increased capacity to undergo cholehepatic shunting, thereby stimulating both canalicular bile flow and biliary HCO3− output as a method to rid the liver of toxic bile acids.
The findings by Halilbasic and colleagues support the expectation that norUDCA will prove to be effective in the treatment of cholestatic disorders with a background of MDR3 mutation or deficiency. Several caveats remain in the translation of these findings from the animal model to the human patient, because it is important to note that there are differences between the Mdr2−/− mouse cholangiopathy model compared to the pathology of PSC in humans. Among these differences are that the Mdr2−/− mouse model progresses to hepatocellular carcinoma rather than cholangiocarcinoma and the lack of the alteration of bile acid lipid profiles in patients with PSC.20 However, these findings do not take away from the overall importance that norUDCA, and similar compounds that undergo or can target cholehepatic shunting, are promising drug candidates for the management of PSC and other cholestatic liver diseases by stimulating bile flow and bicarbonate excretion through activation of the cholehepatic shunt, stimulating the protective effect of toxic bile elimination (Fig. 1). As these drug candidates are evaluated it will also be important to determine what effects they have on the expression of bile salt and drug transporters such as, apical sodium/bile acid cotransporter (SLC10A2), ileal binding protein (FABP6), and multi drug resistance associated protein 3 (ABCC3) on cholangiocytes. The protective effect may not only be due to an increase in bicarbonate excretion by cholehepatic shunting, but also might be due to an alteration of cholangiocyte-specific transporters. The jury is still out on norUDCA, and we expectantly await clinical trials to determine if norUDCA will be effective and safe in patients with PSC or other cholestatic liver diseases.