New insights from recurrent primary biliary cirrhosis in liver transplantation: The paradox of BEComing a fibroblast?


  • See Article on Page 977

  • Potential conflict of interest: Nothing to report.

In this issue of HEPATOLOGY, Robertson and colleagues take observations from an emerging area of cell biology, referred to as the epithelial-to-mesenchymal transition (EMT),1 and draw together several threads that illuminate the early pathophysiology of primary biliary cirrhosis (PBC). In this case report, a liver transplant recipient with recurrent PBC developed characteristic features of EMT in biliary epithelial cells (BECs). The plot thickens, however, as cell markers indicative of EMT were observed at 24 days following transplantation, prior to any histological or symptomatic evidence of recurrent PBC, which was confirmed later by liver biopsy at 9 months. Also, of clinical significance, the patient was switched from tacrolimus to cyclosporine at 9 months with subsequent improvement of symptoms and hepatic biochemistry. This brings up the provocative thought that we may be able to predict early recurrent disease and possibly treat patients in a timely fashion with either ursodeoxycholic acid or even an experimental challenge with change of immunosuppression to cyclosporine.2–6 Lest we get carried away, however, EMT has only been reported in one PBC patient. Also, the positive effects of cyclosporine in transplant recipients with PBC have only been shown in retrospective studies; we will tackle the implications of this at the end of the editorial.


BEC, biliary epithelial cell; EMT, epithelial-to-mesenchymal transition; PI3K, phosphatidylinositol 3-kinase; TGF-β, transforming growth factor-beta.

First of all, what is EMT? The term EMT has been coined to define a cellular reprogramming process whereby epithelial cells adopt the phenotypic and molecular characteristics of mesenchymal cells.1 We know that epithelial cells usually form a tightly bound, single layer of polarized and structured cells; whereas mesenchymal cells lack tight junctions, polarity, and a consistent structure. Accordingly, phenotypic markers that signify the loss of epithelial morphology and adhesion properties or the gain of mesenchymal cell markers and migratory capacity are currently used to monitor EMT (Fig. 1). This process is essential for organ morphogenesis7 and plays an important role in repair and wound healing.8 It is now accepted that EMT is critically involved in a wide spectrum of pathophysiological disorders that include cataract formation, fibrosis in pulmonary and renal disease, as well as carcinogenesis and development of metastasis.1, 8–10 This means the onset of EMT not only provides early prognostic data on the potential for disease progression but also an insight into specific disease processes.

Figure 1.

Epithelial-to-mesenchymal transition is defined by disruption of architectural characteristics of epithelium, such as tight junctions, with the development of features characteristic of mesenchymal cells, such as migration. Thus the loss of signal for cytokeratin, occludin, or E-cadherin on bile ducts with a gain of vimentin or S100A4 fibroblast lineage markers and activation of signal transduction pathways with nuclear accumulation of β-catenin and Smad 2/3, for example, would signify EMT.

So how does EMT potentially affect the induction and progression of PBC? Two mechanisms immediately spring to mind; the transition of BEC into portal fibroblasts could potentially cause both ductopenia and fibrosis. The authors observed the fibroblast lineage marker S100A4 in BECs at 24 days and 9 months, as well as smooth muscle actin–positive myofibroblasts adjacent to bile ducts in the 9-month biopsy with recurrent PBC. This implies that BEC-to-mesenchymal transition may be contributing to the resident portal fibroblast cell population. In support of this, BECs have been shown to undergo phenotypic conversion to myofibroblasts and migration in the mouse bile duct ligation model.11 Also, human BECs can be triggered to undergo EMT in vitro, with the production of the myofibroblast markers on cells, such as smooth muscle actin and fibronectin.11, 12 These data suggest that BECs may have the inherent capacity to fully differentiate into myofibroblasts and contribute to the production of extracellular matrix as seen with the classical onion skin fibrosis observed surrounding bile ducts of patients with primary sclerosing cholangitis, for example.

EMT may also directly result in ductopenia by converting BECs to portal fibroblasts. The mechanism, however, provides us with a conundrum, because the EMT process is essentially antiapoptotic.1 At present, it is thought that the balance between cholangiocyte proliferation and cell death controls disease progression, which favors apoptosis as ductopenia ensues.13, 14 However, activation of prosurvival pathways associated with EMT has been found in PBC BECs. For example, activation of AKT increases dramatically in interlobular biliary epithelium throughout the progression of PBC with highest expression in stage 4 disease when apoptosis is far less prominent.14 This suggests that the apoptosis observed in PBC may, in part, be triggered by immune attack on bile ducts, and this may well be counterbalanced by AKT and other prosurvival pathways observed in PBC that can directly promote the mesenchymal transition in BECs.14, 15 The paradoxical notion that epithelial cells demonstrate both increased proliferation and increased apoptosis is also observed in carcinogenesis, a process involving EMT.1, 16

One pathway implicated in EMT is the highly conserved canonical Wnt/β-catenin pathway1, 16 that plays an important role in embryogenesis and specification of the liver.17 Indeed, most Wnt factors and their receptors are expressed in hepatocytes, stellate cells, sinusoidal endothelial cells, and biliary epithelial cells.18 Liver-specific loss of β-catenin results in diminished regeneration following partial hepatectomy with increased apoptosis providing a testimony to the reparative and prosurvival activity of the pathway.19 One microarray study identified a number of Wnt pathway mediators differentially up-regulated in PBC liver, but the implications of Wnt signaling in PBC have yet to be determined.20 Activation of the Wnt/β-catenin pathway is also involved in the development of intestinal, liver, and other cancers.16 In fact, the wnt-1 gene, previously referred to as the int-1 gene, was originally discovered as a proto-oncogene transactivated by the integration of the mouse mammary tumor virus in murine breast tumors.21 This virus is identical to the human betaretrovirus, recently characterized in patients with PBC.22 However, it remains to be resolved whether the human betaretrovirus also transactivates the Wnt/β-catenin pathway.

Another pathway possibly involved in BEC EMT is the phosphatidylinositol 3-kinase (PI3K)-Akt pathway that is activated by a variety of growth factors.1, 10 Known for its role in survival and EMT, the activation of the PI3K-Akt pathway also regulates metabolism and mitochondrial function.10, 23 In cancer cells, PI3K-Akt activation results in a metabolic switch whereby energy is preferentially generated by aerobic glycolysis instead of oxidative phosphorylation, a process first described by Otto Warburg in 1930 in cancer cells.24 There is now renewed interest in the metabolic hypothesis of cancer whereby reliance on aerobic glycolysis provides a proliferative advantage to cells by bypassing mitochondria as the main source of cellular energy. One of the mitochondrial gatekeepers regulating the switch in glucose metabolism from glycolysis to oxidative phosphorylation is none other than the PBC autoantigen, pyruvate dehydrogenase complex (PDC). We know that PDC-E2 is both overexpressed and aberrantly located on PBC biliary epithelium,22 but the actual mitochondrial defect in PBC is poorly characterized. Further investigations of factors that promote EMT and mitochondrial dysfunction in PBC are clearly warranted to understand the actual cellular defects in the BEC.

This brings up the question of what triggers BECs to undergo mesenchymal transition? The initiation of EMT is in part mediated by the ability of epithelial cells to detect changes in the extracellular milieu that set in motion the transition. A number of receptor and integrin signaling cascades have been identified that participate in establishment and maintenance of EMT,7, 10 such as the profibrinogenic cytokine transforming growth factor-beta (TGF-β). In this case report, intense nuclear localization of Smad2/3 was found within BECs, suggesting a role for the TGF-β signaling pathway.25, 26 Although this observation requires further assessment, a number of experimental observations support a role for TGF-β in BEC mesenchymal transition. TGF-β is a key mediator and activator of hepatic fibrosis, and strategies that disrupt TGF-β, also prevent liver fibrosis in vivo.8, 26–30 The expression TGF-β and its receptor are up-regulated in BECs from patients with biliary fibrosis and biliary atresia, as well as animal models of bile duct ligation.31, 32In vitro, human BECs undergo mesenchymal transition in the presence of TGF-β.11, 12 Thus, the EMT model fits with our current understanding of progressive fibrosis in liver disease.1, 8–10

Finally, the clinical improvement observed in the transplant recipient following the change of calcineurin inhibitor merits comment. Whereas cyclosporine has shown benefit in the nontransplant setting with PBC,33 therapy has not been adopted due to side effects. Several liver transplant centers reported that the recurrence of PBC appears earlier and is more severe in those patients maintained on tacrolimus versus cyclosporine.2–6 The reason remains obscure but there are parallels here with calcineurin inhibitor use and recurrent HCV infection in liver transplant recipients. In this setting, other factors such as older donors, cold ischemic time, the “era effect”, and immunosuppressant use for controlling acute rejection have all emerged as important factors with regard to onset and severity of recurrent HCV.34 In fact many of these factors appear to be associated with recurrent PBC as well.2 For recurrent HCV infection, there is an emerging interest in the antiviral role of cyclosporine. Indeed, it has been recently shown that cyclosporine imparts antiviral activity by binding cyclophillin B, a necessary cofactor for function of the HCV RNA polymerase NS5B.35 Cyclosporine also has antiviral activity against several viruses including HIV, where cyclophillin activity is necessary for viral assembly. As a result, the cyclosporine analog, NIM811, which lacks immunosuppressive activity, is currently being developed as an antiviral agent.35 So, even though it is by no means clear why cyclosporine should provide preferential outcomes to tacrolimus in PBC, mechanisms have been described for HCV infection. Nevertheless, the potential therapeutic benefits of cyclosporine should be pursued further in liver transplant recipients with PBC. Perhaps we can also monitor recurrent disease with markers for EMT? Watch this space.