Human intrahepatic biliary epithelial cells engulf blebs from their apoptotic peers

Authors


Correspondence: M. E. Gershwin, Division of Rheumatology, Allergy and Clinical Immunology, University of California at Davis School of Medicine, 451 Health Sciences Drive, Suite 6510, Davis, CA 95616, USA.

E-mail: megershwin@ucdavis.edu

Summary

The phagocytic clearance of apoptotic cells is critical for tissue homeostasis; a number of non-professional phagocytic cells, including epithelial cells, can both take up and process apoptotic bodies, including the release of anti-inflammatory mediators. These observations are particularly important in the case of human intrahepatic biliary cells (HiBEC), because such cells are themselves a target of destruction in primary biliary cirrhosis, the human autoimmune disease. To address the apoptotic ability of HiBECs, we have focused on their ability to phagocytize apoptotic blebs from autologous HiBECs. In this study we report that HiBEC cells demonstrate phagocytic function from autologous HiBEC peers accompanied by up-regulation of the chemokines CCL2 [monocyte chemotactic protein-1 (MCP-1)] and CXCL8 [interleukin (IL)-8]. In particular, HiBEC cells express the phagocytosis-related receptor phosphatidylserine receptors (PSR), implying that HiBECs function through the ‘eat-me’ signal phosphatidylserine expressed by apoptotic cells. Indeed, although HiBEC cells acquire antigen-presenting cell (APC) function, they do not change the expression of classic APC function surface markers after engulfment of blebs, both with and without the presence of Toll-like receptor (TLR) stimulation. These results are important not only for understanding of the normal physiological function of HiBECs, but also explain the inflammatory potential and reduced clearance of HiBEC cells following the inflammatory cascade in primary biliary cirrhosis.

Introduction

Clearance of apoptotic cells is critical for tissue haemostasis and resolution of inflammatory lesions. A major physiological manifestation of apoptosis is the breakdown of the cytoskeleton resulting in the formation of blebs. These blebs are normally removed by macrophages without a detectable inflammatory reaction [1-3]. However, defects in the apoptotic pathways have been linked to oncogenesis [4], and while the mechanisms of apoptotic bleb formation are well studied, the implication of such apoptotic blebs in autoimmune disorders is a subject of intense interest [5-7]. This issue is particularly important for human intrahepatic biliary epithelial cells (HiBECs), the major target of the autoimmune liver disease primary biliary cirrhosis (PBC) [8, 9]. Recent data suggest that epithelial cells can take up and process apoptotic bodies from their neighbouring apoptotic cells as non-professional phagocytes and/or present pathological epitopes onto their surface as non-professional antigen-presenting cells (APC) [10-14]. In the current study we explored this issue by addressing whether HiBECs have the ability to engulf apoptotic blebs derived from their apoptotic peers.

Materials and methods

Cell culture

HiBECs (ScienCell, San Diego, CA, USA) were derived from two normal human donors and cryopreserved immediately after purification. HiBECs were cultured at 37°C in a humidified 5% CO2 environment in sterile epithelial cell medium (EpiCM) supplemented with 2% fetal bovine serum (FBS), epithelial cell growth supplement and 1% penicillin/streptomycin (ScienCell). HiBECs were characterized using a previously standardized immunofluorescence-assisted microscopic method using antibodies to cytokeratin 18, cytokeratin 19 and vimentin, which stained > 90% of the cultured cells. Experiments were performed using cells between passages 2 and 8 [15-18].

Generation of monocyte-derived macrophages

Monocytes were isolated from human peripheral blood mononuclear cells (PBMCs) using anti-CD14 microbead-assisted magnetic cell sorting (Miltenyi Biotec, Bergisch Gladbach, Germany). Aliquots of monocytes (0·5 × 106/ml) were then resuspended in RPMI-1640 culture medium containing 10% heat-inactivated FBS, 90 U/ml penicillin and 90 μg/ml streptomycin supplemented with 100 ng/ml granulocyte–macrophage colony-stimulating factor (GM-CSF). These cells were cultured for at least 5 days at 37°C in a humidified 5% CO2 incubator to generate macrophage [19].

Cell staining with fluorochromes

The lipophilic fluorochromes PKH-26 and PKH-67 were purchased from Sigma-Aldrich (St Louis, MO, USA). HiBECs were stained with the lipophilic dyes PKH-67 (green) or PKH-26 (red). Briefly, cells (5 × 106) were washed in serum-free EpiCM and the cell pellet resuspended in 0·5 ml of 2·3 × 10−6 M PKH-26 or 5 × 10−7 M PKH-67 staining solution and incubated for 5 min at room temperature. After stopping the labelling by adding 0·5 ml FBS and incubating for 1 min, the cells were washed three times with complete medium. Similarly, single-cell suspension of thymocytes, isolated from 8-week-old C57/BL6 (purchased from Jackson Laboratory, Bar Harbor, ME, USA), was also labelled in parallel with PKH-26 for induction of apoptosis.

Apoptosis induction and bleb isolation

Apoptosis was induced with bile salts as described previously [15, 16], with minor modifications. Briefly, PKH-26-labelled or -unlabelled HiBECs were incubated in culture medium containing 1 mM sodium glycochenodeoxycholate (GCDC; Sigma-Aldrich) at 37°C for 4 h. The apoptosis rate was assessed by flow cytometry after staining with 7-aminoactinomycin D (7-AAD) and fluorescein isothiocyanate (FITC)-annexin V (BD Biosciences, San Jose, CA, USA). More than 70% of the cells undergo early apoptosis (annexin V-positive and 7-AAD-negative) after treatment. To isolate blebs from apoptotic HiBECs, the culture supernatants were collected and centrifuged twice at 500 g for 5 min to remove the remaining cells. The supernatants were then centrifuged at 100 000 g for 45 min at 4°C. The pellets containing blebs were resuspended in EpiCM. The bleb numbers were counted as we have described previously [19]. For obtaining apoptotic thymocytes, PKH-26-stained thymocytes were irradiated at 30 Gy (caesium-137 irradiator, 627 R/min), washed and analysed using the fluorescence-based apoptosis assay.

Phagocytosis assay

Fluorescent microbeads (FluoSpheres® carboxylate-modified microspheres, 1·0 μm, yellow-green fluorescent; Life Technologies, Carlsbad, CA, USA) were utilized for phagocytosis. Briefly, an aliquot of 5 × 104 PKH-26-labelled HiBECs or macrophages were dispensed into individual wells of a 24-well plate and incubated overnight at 37°C, with the addition of the microbeads at a cell/bead ratio of 1:5. After culture for the indicated times, the microbead-engulfed cells were washed twice with culture medium, and then dissociated with 0·25% trypsin–ethylenediamine tetraacetic acid (EDTA) (Life Technologies) for 2 min. The cells were then washed with culture medium and analysed by flow cytometry. For analysis of the engulfment of apoptotic cells, PKH-26-labelled apoptotic thymocytes were added to the culture of PKH-67-labelled live HiBECs (5 × 104 cell/well) at a ratio of 2:1. After culture for 16 h, the apoptotic cell-engulfed HiBECs (red PKH-26 and green PKH-67 double-positive) were analysed by flow cytometry. For phagocytosis of apoptotic HiBECs, apoptosis was induced with bile salts in PKH-26-labelled apoptotic HiBECs. The apoptotic HiBECs were added to PKH-67-labelled live HiBECs (5 × 104 cell/well) at a ratio of 1:1. After culturing for 16 h, the cells were analysed by flow cytometry. In nested studies, the Toll-like receptor (TLR) ligand polyI:C (TLR-3, 10 μg/ml), lipopolysaccharide (LPS) (TLR-4, 10 μg/ml), cytosine–phosphate–guanosine (CpG)-B (TLR-9, 1 μM), Pam3CSK4 (TLR-1/2, 10 μg/ml) and peptidoglycan (PGN) (TLR-2, 10 μg/ml) were added to individual cultures of HiBECs incubated with either microbeads or apoptotic cells for 16 h. The ability of the cells to phagocytize was analysed by flow cytometry, as described above.

Flow cytometry

Aliquots of 5 × 105 HiBECs were stained with one of the following FITC-conjugated antibodies: anti-CD51, anti-CD61, anti-CD93, anti-CD36, anti-CD16, anti-CD32 or phycoerythrin (PE)-conjugated anti-CD91, anti-CD14, anti-CD64 (all from BioLegend, San Diego, CA, USA) or anti-human leucocyte antigen (HLA)-ABC (eBioscience, San Diego, CA, USA). The cells were stained for 30 min at 4°C, washed twice, then analysed on a FACScan flow cytometer (BD Immunocytometry System, San Jose, CA, USA).

Western blotting

To determine the expression of phosphatidylserine receptors (PSR) in HiBECs [3], HiBECs and human PBMCs (as positive control) were lysed by incubation in radio immunoprecipitation assay (RIPA) buffer containing a protease inhibitor cocktail (Cell Signaling Technology, Danvers, MA, USA). Protein concentration was determined by the bicinconic acid assay (Thermo Scientific, Rockford, IL, USA). The expression of PSR in HiBEC was detected by standard Western blotting techniques using anti-human PSR (H-300; Santa Cruz Biotechnology, Santa Cruz, CA, USA). Anti-human β-actin antibody was used as positive control (C4; Santa Cruz).

Confocal microscopy

To visualize the engulfment of microbeads by HiBECs, aliquots of 1–5 × 103 PKH-26 (red)-labelled cells were co-cultured with carboxylate-modified fluorescent microspheres (green) at 37°C in an eight-well Lab-TekTM II CC2 chamber slide (Fisher Scientific, Waltham, MA, USA) for 16 h. The cells were washed twice with culture medium and fixed with 4% paraformaldehyde for 30 min, followed by 4,6-diamidino-2-phenylindole (DAPI; Invitrogen, Carlsbad, CA, USA) staining to visualize nuclear degeneration. The slides were examined using a Zeiss LSM 700 confocal microscope to identify fluorescent microbead-containing cells.

Cytokine and chemokine assays

HiBECs were seeded in a 24-well culture plate at 5 × 104 cells per well. The cells were incubated overnight at 37°C to allow the cells to attach to the culture plates and the medium was decanted to remove dead cells and cell debris. Freshly isolated blebs were added to the HiBEC cultures at a bleb/cell ratio of 10:1 and the cells were cultured for 2 days. The culture supernatants were collected and the cytokine and chemokine concentration was determined by the human inflammatory cytokine cytometric bead array (CBA) kit and human chemokine CBA kit (BD Biosciences), as described previously [15].

Reverse transcription–polymerase chain reaction (RT–PCR)

Total RNA was extracted from HiBECs utilizing the RNeasy kit (Qiagen, Valencia, CA, USA). After DNase I treatment, first-strand cDNA was synthesized using SuperscriptTM II (Invitrogen) and random hexamer primer (Invitrogen) in the presence of an RNase inhibitor. Levels of mRNA transcripts and the internal control gene β-actin in the HiBECs were analysed by RT–PCR in a GeneAmp PCR System 9700 (PE Applied Biosystems, Foster City, CA, USA) with AmpliTaqGold enzyme and buffer (Perkin-Elmer). The primers are described in Table 1.

Table 1. Primer sequences used for real-time polymerase chain reaction (PCR) analyses.
GeneForward primerReverse primer
  1. HLA: human leucocyte antigen.
CD40ACTGTACGAGTGAGGCCTGTGAGACGAAAGCAGATGACACATTGGAG
CD80CAAAGCTGACTTCCCTACACCCTCCATTTTCCAACCAGGAGAG
CD86ACAGCAGAAGCAGCCAAAATTCCTGGTCCTGCCAAAATAC
HLA-ABCACGGACCCCCCCAAGACACCGCTGCCAGGTCAGTGTGAT
HLA-DRCGGGGTTGGTGAGAGCTTCAACCACCTGACTTCAATGCTG
β-ActinAGACCTGTACGCCAACACAGAGGAGGAGCAATGATCTTG

Statistical analysis

All results are expressed as means ± standard error (s.e.m.). Statistical analysis was performed using the Prism statistical package (GraphPad Software, Inc., La Jolla, CA, USA). Normally distributed data were analysed using one-way analysis of variance (anova) with post-test. A value of P < 0·05 was deemed statistically significant.

Results

Phagocytic receptor expression on HiBECs

To explore the phagocytic potential of HiBECs, we first examined their expression of phagocytosis-related receptors. As shown in Fig. 1, we observed the cell surface expression of CD51 (integrin αV), CD61 (integrin β3) and CD93 (C1q receptor) on HiBECs by flow cytometry (Fig. 1a), as well as the expression of PSR by immunoblot analysis (Fig. 1c). However, HiBECs did not show any detectable expression levels of the scavenger receptor A, CD31, CD34, CD344, CD14 and CD11b, all of which have been implicated in uptake of apoptotic cells by macrophages (data not shown); nor were the Fc-receptors (i.e. CD16, CD32 and CD64) expressed on the HiBEC surface (Fig. 1b). These results indicate that HiBECs do not express all the phagocytosis-related molecules of professional phagocytotic cells, such as macrophages.

Figure 1.

Expression of phagocytic receptors on intrahepatic biliary epithelial cells (HiBECs). (a,b) The phagocytosis-related cell surface markers on HiBECs were analysed by flow cytometry. Human leucocyte antigen (HLA)-ABC was used as a positive control. Grey peaks represent HiBECs stained with isotype control antibody. (c) The expression of phosphatidylserine receptors (PSR) in HiBEC was analysed by immunoblotting.

Phagocytic ability of HiBECs

As shown in Fig. 2, whereas approximately 10% of the macrophages show engulfed microbeads after co-culturing with the beads for 1 h, very few HiBECs demonstrate microbead uptake at this time-point. However, the percentage of HiBECs with engulfed microbeads increased substantially after 8 h and increased further after 16 h to 30%, although the percentage remained at approximately 70% of the values seen in cultures of macrophages. The fluorescent photomicrographs in Fig. 3 demonstrate that the fluorescent beads were indeed engulfed by HiBECs, rather than attached to the HiBECs surface. Of note, most HiBECs with engulfed beads had internalized only one bead after 16 h of incubation, whereas macrophages usually engulf multiple microbeads (Fig. 3). These results indicate that HiBECs clearly possess the ability to engulf microbeads, although at a lower level than the professional phagocytic macrophages. As HiBECs expressed PSR, as shown in Fig. 1, we next examined their ability to phagocytize apoptotic cells, which are known to express the ‘eat-me’ signal phosphatidylserine. HiBECs and macrophages were labelled with the green dye PKH-67 and co-incubated with the red dye PKH-26-labelled apoptotic mouse thymocytes 16 h, and then examined by flow cytometry. As seen in Fig. 4a, the apoptotic thymocytes were ingested at similar levels by HiBECs and macrophages.

Figure 2.

Uptake of microbeads by intrahepatic biliary epithelial cells (HiBECs). HiBECs and macrophages were cultured with carboxylate-modified fluorescent microbeads for 1–16 h, then analysed by flow cytometry. (a) Representative dot-plots of HiBECs and macrophages analysed after culturing with the microbeads for 1 and 16 h. (b) Percentage of HiBECs and macrophages with engulfed microbeads at different time-points after culturing with microbeads.

Figure 3.

Confocal microscopic analysis of intrahepatic biliary epithelial cells (HiBECs)and macrophages with engulfed microbeads. The cells were labelled with PKH-26 (red) and cultured with carboxylate-modified fluorescent microbeads (green) for 16 h in an eight-well chamber slide.

Figure 4.

Phagocytosis of apoptotic cells by intrahepatic biliary epithelial cells (HiBECs). Mouse thymocytes (a) and HiBECs (b) were labelled with PKH-26, followed by induction of apoptosis as described in Materials and methods. The apoptotic cells were added to PKH-67-labelled live HiBECs and macrophages and cultured for 16 h. The cells were analysed by flow cytometry to identify HiBECs or macrophages with engulfed apoptotic cells, which were PKH-67 and PKH-26 double-positive. (a) Engulfment of apoptotic thymocytes (Ap-thymocyte) by HiBECs and macrophages. (b) Engulfment of autologous apoptotic cells (Ap-HiBEC) by HiBECs. (c) Effects of Toll-like receptor (TLR) ligands on the uptake of fluorescent microbeads or autologous apoptotic HiBECs by live HiBECs. PKH-67-labelled HiBECs were co-cultured with fluorescent microbeads or PKH-26-labelled apoptotic HiBECs in the presence of polyI:C, lipopolysaccharide (LPS), cytosine–phosphate–guanosine (CpG)-B, Pam3CSK4 and peptidoglycan (PGN), respectively, followed by flow cytometric analysis. **P < 0·01 (one-way analysis of variance with post-test).

We then examined whether or not HiBECs could engulf autologous apoptotic human cells in addition to apoptotic murine thymocytes. As we have demonstrated previously, the apoptotic HiBECs induced by bile salt expressed phosphatidylserine [15, 16]. We therefore used bile salt to induce apoptosis of PKH-26-labelled HiBECs, which were then added to PKH-67-labelled live HiBECs and cultured for 16 h. The phagocytic cells were analysed by flow cytometry. As shown in Fig. 4b, the HiBECs indeed exhibited phagocytic uptake of apoptotic HiBECs. Taken together, these results indicated that similar to the phagocytic ability of macrophages, HiBECs have the ability of clearing apoptotic cells, which is probably mediated by the phosphatidylserine signalling.

TLR-ligands and phagocytosis of apoptotic cells by HiBECs

In previous studies, exposure to LPS was shown to enhance the phagocytosis of apoptotic cells by professional phagocytotic cells [20-22]. To address whether LPS and/or other microbial constituents can exert a similar effect on HiBECs, a panel of TLR ligands in appropriate concentrations, including polyI:C (TLR-3), LPS (TLR-4), CpG-B (TLR-9), Pam3CSK4 (TLR-1/2) and PGN (TLR-2), were added to cultures of HiBECs along with either microbeads or apoptotic cells for 16 h. Of interest was the finding that not only did these TLR ligands fail to show increased uptake of microbeads or apoptotic cells by HiBECs (Fig. 4c), select TLR ligands such as polyI:C and CpG-B even suppressed the engulfment of autologous apoptotic cells (Fig. 4c).

HiBECs respond to apoptotic cells by secreting the chemokines CXCL8 and CCL2, but not proinflammatory cytokines

To investigate the innate immune response of HiBECs to autologous apoptotic cells, HiBECs were co-cultured for 2 days with apoptotic cells. The supernatants were collected and the levels of proinflammatory cytokines and chemokines were measured. The proinflammatory cytokines interleukin (IL)-1β, IL-10, tumour necrosis factor (TNF)-α and IL-12p70 were undetectable by a CBA assay in the culture system used in this study. IL-6 were detected in media from HiBEC cultures, but the level was not increased when the HiBECs were co-cultured with apoptotic cells. However, two of the five chemokines tested, CXCL-8 and CCL-2, were found to be secreted at significantly increased levels in HiBECs co-cultured with apoptotic cells (Fig. 5a).

Figure 5.

Chemokine production by intrahepatic biliary epithelial cells (HiBECs) exposed to apoptotic cells. HiBECs were co-cultured with apoptotic HiBECs or mouse thymus cells (a); apoptotic blebs isolated from apoptotic HiBECs (b) for 2 days. The culture supernatants were collected and the levels of secreted chemokines were measured by cytometric bead array kit. The experiment was performed in triplicate and repeated twice with similar results. **P < 0·01 (one-way analysis of variance with post-test).

To exclude the possibility that the two chemokines were secreted by the bile salt-induced apoptotic HiBECs in the co-culture, we isolated apoptotic blebs from apoptotic cells as reported previously [15, 16]. The blebs were added to HiBEC cultures at a bleb/cell ratio of 10:1 and the cells were cultured for 2 days. As shown in Fig. 5b, addition of the blebs increased the levels of secreted CXCL-8 and CCL2, indicating that engulfment of the apoptotic blebs indeed enhanced secretion of these chemokines from HiBECs.

HiBECs and APC function-involved surface markers after engulfment of apoptotic blebs

To determine whether HiBEC acquire APC function after processing the autoantigens from bleb, we analysed the expression of the co-stimulatory molecules and MHCs on the surface of HiBECs co-cultured with blebs. As shown in Fig. 6a, the fluorescent intensity of cell surface CD40, CD80, CD86 and HLA-DR on the HiBECs was not changed markedly after co-culturing for 24 h with apoptotic blebs. Similar results were obtained when the HiBECs were co-cultured with blebs in the presence of PolyI:C and LPS, a condition that mimics local inflammation around biliary cells in vivo (data not shown). In agreement with the cell surface staining results, differences were not detected in the mRNA levels of these genes determined using standard RT–PCR (Fig. 6b).

Figure 6.

Expression of activation and co-stimulation markers by phagocytotic intrahepatic biliary epithelial cells (HiBECs). HiBECs were cultured with isolated bleb for 24 h, then analysed for the expression of cell surface markers CD40, CD80, CD86 and human leucocyte antigen (HLA)-DR by flow cytometry (a), or for the mRNA levels of these genes by reverse transcription–polymerase chain reaction (RT–PCR (b). (a) Antibody isotype control staining is shown in the grey histogram. (b) Bleb-engulfed HiBECs (bleb-HiBEC) are compared with HiBECs that had not been exposed to blebs. HLA-ABC and β-actin were used as control genes for RT–PCR assay.

Discussion

The data presented in this study demonstrate that the phagocytic function of HiBECs for apoptotic blebs derived from autologous HiBECs is accompanied by the up-regulated expression of the chemokines CCL2 and CXCL8. We also demonstrate that the HiBECs express the phagocytosis-related receptor PSR, suggesting that the phagocytic ability of HiBECs involves the ‘eat-me’ signal phosphatidylserine expressed by the blebs. Interestingly, HiBECs express only PSR and CD93, which is a regulator of the phagocytosis of apoptotic cells, but do not express the other cell surface molecules found in macrophages. As a result, there is a reduced ability to engulf microbeads but a similar ability to phagocytize blebs of apoptotic cells after 16 h of contact. This characteristic is similar to that of other non-professional phagocytic cells [23, 24]. Of note, we should emphasize that TLR ligands do not enhance the phagocytosis ability of HiBECs, but that after processing apoptotic cells HiBECs secrete the chemokines CXCL8 and CCL2 but not proinflammatory cytokines.

Monocyte chemotactic protein-1 (MCP-1/CCL2) is a β-chemokine responsible, in part, for the chemotaxis of mononuclear phagocytes. Macrophages from CCL2 knock-out (KO) and CCR2 KO mice demonstrate reduced phagocytic function [25]. CCL2-dependent recruitment of phagocytes in sites of inflammation improves protective immunity to bacterial infection and timely clearance of dead cells [26-28]. In addition, CXCL8 is a potent attractant of neutrophils to the site of tissue injury and induces phagocytosis once they have arrived [29, 30]. Therefore, the post-engulfment response of HiBECs to secrete CCL2 and CSCL8 may attract more monocytes and macrophages to the site. The recruited phagocytes could help local live HiBECs in the clearance of apoptotic cells [31].

Both professional and non-professional phagocytes participate in the the clearance of the massive number of cells that undergo apoptosis during development [32, 33]. Although there are differences between professional phagocytes and non-professional tissue cells [34], the mechanisms of uptake as well as the accompanied immune responses in these two categories of phagocytes are still not completely clear. In the present study, we demonstrate that after engulfing autologous apoptotic blebs, HiBECs secrete the chemokines CCL2 and CXCL8, but not any of the proinflammatory cytokines tested. In addition, we demonstrate that engulfment of blebs by HiBECs does not up-regulate expression of the co-stimulatory molecules and other markers of functional APCs, even in the presence of potent inflammatory inducers such as polyI:C and LPS, suggesting that the phagocytic HiBECs do not serve as APCs in physiological conditions.

It is accepted widely that inefficient clearance of apoptotic cells has proinflammatory consequences that can lead to the development of autoimmune and persistent inflammation [35-40]. Although intravenous administration of apoptotic cells has been shown to result in the development of autoantibodies, due presumably to inefficient clearance of excessive apoptotic cells [41], there are also data indicating that defects in engulfment-related genes leads to autoimmune disease [37]. The release of the intracellular contents from blebs is thought to provoke an inflammatory response, particularly towards intracellular antigens. This may provide the immunogenic impetus for the onset of autoimmune disorder. Indeed, we have shown recently that the apoptotic blebs derived from HiBECs contained the mitochondrial autoantigens recognized in PBC [16]. Furthermore, these antigen-contained blebs can induce pathogenic immune responses of macrophages derived from PBC patients [19]. This raises the possibility that the clearance of apoptotic cells by normal HiBECs actually helps to reduce inflammation by competing for the blebs with resident professional phagocytes in PBC, whereas producing more proinflammatory cytokines by the post-engulfed phagocytes around bile ducts may recruit more immune cells to the inflammatory site. However, in this study our data were collected from normal HiBECs. The role of HiBECs in the pathogenesis of PBC and its relation to aetiology is still unclear [42-44]. However, it has been demonstrated that BECs have the ability to present novel mitochondrial self-peptides derived from phagocytosed apoptotic BECs. Apoptotic cell phagocytosis by non-professional phagocytes may also influence the tissue specificity of autoimmune diseases [45]. The resolution of this issue is critical to the understanding of biliary specificity.

Acknowledgements

The authors thank Shanie McCarty for cell irradiation. We also thank Ms Nikki Phipps for support in preparing this paper. Financial support for this work was provided by National Institutes of Health grant DK39588 and Chinese National Natural Science Foundation (Grant No. 81202362).

Disclosure

The authors have no conflicts of interest to disclose.

Ancillary