Primary biliary cirrhosis (PBC) is a liver-specific auto-immune disorder which affects about 1 in 1000 women above 40 years (female:male ratio is 9:1) [1, 2]. PBC is characterized by progressive immune-mediated destruction of small interlobular bile ductules, portal and periportal inflammation, ductular proliferation, hepatic fibrosis and cirrhosis, and finally liver failure. In early stages, patients are often asymptomatic and only diagnosed by elevated serum markers of cholestasis (alkaline phosphatase, gamma-glutamyl transferase) and the presence of serum antimitochondrial antibodies directed against the E2 subunit of the pyruvate dehydrogenase complex (anti-PDC-E2) and other dehydrogenases [3, 4]. During the course of the slowly progressive disease, the majority of patients will develop fatigue and pruritus. Today, ursodeoxycholic acid (UDC)  is approved as standard treatment by actual European and American guidelines [1, 2] and may halt disease progression and allow a normal life expectancy for up to two of three patients with PBC.
The aetiology of PBC still remains enigmatic, although key pathogenetical mechanisms have been unravelled during the last decade [3, 4]. A genetic risk profile is obvious as confirmed by sibling studies and recent genome-wide association studies3. Environmental factors that might trigger PBC have been identified [3, 4]. Dr. Gershwin's and Dr. Leung's group recently proposed the following initiation of PBC, based on extensive immunological work over the past decade . Initial exposure to chemicals reactive to the lipoyl domain of PDC-E2 may lead to a primary IgM-specific immune response against the anti-PDC-E2 antigen, e.g. xenobiotic-modified PDC-E2, in genetically predisposed individuals. Subsequently, the similarity between the lipoyl domain of PDC-E2 and the xenobiotic-modified lipoyl domain of PDC-E2 may generate cross-reactive immune responses against the self-antigen, leading to self-tolerance breakdown to the lipoyl domain of mitochondrial PDC-E2. Later, through the process of affinity maturation and isotype switching, the secondary immune response produces IgGs that are highly specific for mitochondrial PDC-E2 .
But why, once self-tolerance to PDC-E2 is broken, is the subsequent immune destruction restricted to cholangiocytes of small interlobular bile ductules in PBC? Here, endogenous pathogenetical factors might come into play. Intriguingly, patients with PBC show a micro-RNA-506-suppressed expression of the Cl-/HCO3− exchanger, AE2 (SLC4A2), a key exporter of HCO3− in human cholangiocytes, and, thereby, impaired biliary HCO3− formation [7-9]. This might be crucial for the development of bile ductular injury in PBC according to the concept of a ‘biliary HCO3− umbrella’  which protects human cholangiocytes against the toxic effects of human hydrophobic bile acids , particularly the major hydrophobic bile acid in cholestasis, glycochenodeoxycholic acid (GCDC) . Thus, a defective biliary HCO3- umbrella in PBC as a result of impaired AE2 expression might render small bile ductules (which do not express mucins, another protective factor) particularly vulnerable to GCDC-induced cholangiocyte damage, e.g. apoptosis. In vitro, intracellular GCDC is a strong apoptotic stimulus in cholangiocytes with impaired AE2 expression and a defective biliary HCO3- umbrella . Dr. Gershwin's group convincingly demonstrated that immune destruction of small bile ductules is exacerbated by retention/exposure of PDC-E2 in ‘apoptotic blebs’ from apoptotic cholangiocytes . An exacerbation of the orchestrated immunological response could then be the consequence rather than the cause of GCDC-induced cholangiocyte damage in PBC.
In an extensive set of publications, now culminating in another contribution to Liver International, Sasaki and coworkers focus on the importance of autophagy and cell senescence in PBC [14, 15]. Although the authors do not touch upon the initial trigger of the disease, their work provides additional clues to the pathogenesis of PBC.
Autophagy (from the Greek ‘autos’ and ‘phagein’, self-eating) is a cellular process that occurs at a low rate under basal conditions. It is stimulated when cells are under stress and require the regeneration of metabolical precursors (for energy or as building blocks), or need to clear subcellular debris. Enhanced autophagy is observed during starvation, depletion of growth factors or when high amounts of energy are suddenly required. The process starts with the formation of an autophagosome, a vesicle that engulfs intracellular large objects, such as (damaged) mitochondria, parts of other organelles such as endoplasmic reticulum and peroxisomes, or larger oligomeric or aggregated proteins. Autophagosomes, thus, allow the disposal of polyubiquitinated protein aggregates that do not fit into the barrel-like structure of the proteasome. Autophagosomes eventually fuse with lysosomes (forming autolysosomes) to have their content degraded and allow the cell to re-use valuable metabolites. Autophagy primarily is a protective mechanism to prevent cell death under stress. However, upon chronical stress, cells eventually can go into senescence, a state in which the cell no longer can proliferate, or even enters apoptosis.
In their actual work , Sasaki et al. have shown that the autophagy markers, LC3 (microtubule-associated protein light chain 3) and p62, are detected significantly more frequently in cholangiocytes of PBC livers than of livers from healthy controls or NASH patients. They could also demonstrate that the staining for PDC-E2 is more abundant in PBC livers and shows overlap of the autophagy markers and this intense PDC-E2 signal in small bile ductules. Using in vitro cultured cholangiocytes, they showed that cellular stress, induced by serum starvation or, notably, incubation with high concentrations of GCDC, induced autophagy and also resulted in colocalization of the mitochondrial enzymes PDC-E2 and cytochrome c oxidase, subunit 1 (CCO) with LC3. This data implies a role for autophagy of mitochondria (mitophagy) in the pathogenesis of PBC. Sasaki et al. also performed similar studies to assess cell senescence using a large number of PBC and control livers and observed that cholangiocytes in damaged small bile ductules in PBC show multiple features of senescence, including increased expression of senescence-associated beta-galactosidase, p16(INK4a) and p21(WAF1/Cip1). Furthermore, telomere length was significantly decreased in cholangiocytes in the damaged small bile ductules in PBC compared with normal-looking bile ducts in PBC, chronical viral hepatitis and normal livers, consistent with increased senescence in small bile ductules in PBC.
At present, it remains unclear whether increased senescence ultimately contributes to ductopenia, the disappearance of the damaged small bile ductules, seen in PBC, as senescence and (programmed) cell death (e.g. apoptosis) are distinct processes. Sasaki and Nakanuma suggest that perhaps also hepatic stem/progenitor cells show increased senescence, eventually reducing the formation of cholangiocytes to replenish dead cells .
What is the appropriate order of events? Sasaki et al. proposed that autophagy precedes senescence in PBC, as they evaluated both processes by immunohistological staining in a large number of PBC livers in early and later stages of the disease. The onset of senescence might be related to oxidative stress, as senescence in early stage PBC is associated with expression of the oxidative stress marker 8-OHdG. The expression of the polycomb protein Bmi1 is decreased in damaged bile ducts in PBC, and expression of this protein is also reduced by oxidative stress, providing another line of evidence for oxidative stress and the onset of PBC. Autophagy itself does not induce oxidative stress, but insufficient autophagy capacity could, as inefficient removal of damaged mitochondria can possibly lead to accumulation of reactive oxygen species. Notably, the p62 signal is clearly increased in PBC. The p62 protein is an adaptor protein that binds to poly-ubiquitinylated proteins and targets them to the autophagosome. When autophagy is inhibited, p62 signal accumulates, as a result of the impaired degradative capacity of the cell. Therefore, it is possible that the increased p62 staining seen by Sasaki et al. is consistent with inefficient autophagy, resulting in increased presence of damaged mitochondria and oxidative stress. The increased signal of PDC-E2 and CCO in PBC livers also points to the direction of a mitochondrial component of the pathogenesis. Obviously, the striking presence of anti-PDC-E2 antibodies in PBC patients makes a contribution of mitochondria to the PBC aetiology very presumable. The authors speculate on the contribution of autophagy on the appearance of anti-PDC-E2 antibodies in PBC. Autophagy has been implicated in innate and adaptive immunity, as autophagosomes can engulf intracellular pathogens, or pathogen containing vacuoles, can deliver viral nucleic acids to endosomal compartments, resulting in TLR7 signalling and antigen loading onto MHC-class II for antigen presentation occurs via autophagosomes. PDC-E2 immunoreactivity on fixed cholangiocytes that have not been permeabilized was increased by GCDC and other stimuli that induced autophagy, providing a novel hypothesis that autophagy could play a role in the development of anti-PDC-E2 autoimmunity in PBC by mediating antigen exposure (Fig. 1). Whether the PDC-E2 at the cell surface actually leads to recognition by (cytotoxic) autoreactive T cells that specifically target this PDC-E2-self-antigen remains to be shown.