Potential conflict of interest: Nothing to report.
This study was supported by the Program for Promotion of Fundamental Studies in Health Sciences of the National Institute of Biomedical Innovation and Grant-in-Aid for Young Scientists (B) (23790175).
After our revision, Xu et al. (Yale Liver Center, Yale University School of Medicine, New Haven, CT) also demonstrated the importance of the tyrosine motif in the carboxyl terminus of BSEP for BSEP internalization in the annual meeting of the American Association for the Study of Liver Diseases (Xu et al., HEPATOLOGY 54;S1:453A).
The bile salt export pump (BSEP) mediates the biliary excretion of bile salts and its dysfunction induces intrahepatic cholestasis. Reduced canalicular expression of BSEP resulting from the promotion of its internalization is one of the causes of this disease state. However, the molecular mechanism underlying BSEP internalization from the canalicular membrane (CM) remains unknown. We have shown previously that 4-phenylbutyrate (4PBA), a drug used for ornithine transcarbamylase deficiency (OTCD), inhibited internalization and subsequent degradation of cell-surface-resident BSEP. The current study found that 4PBA treatment decreased significantly the expression of α- and μ2-adaptin, both of which are subunits of the AP2 adaptor complex (AP2) that mediates clathrin-dependent endocytosis, in liver specimens from rats and patients with OTCD, and that BSEP has potential AP2 recognition motifs in its cytosolic region. Based on this, the role of AP2 in BSEP internalization was explored further. In vitro analysis with 3×FLAG-human BSEP-expressing HeLa cells and human sandwich-culture hepatocytes indicates that the impairment of AP2 function by RNA interference targeting of α-adaptin inhibits BSEP internalization from the plasma membrane and increases its cell-surface expression and transport function. Studies using immunostaining, coimmunoprecipitation, glutathione S-transferase pulldown assay, and time-lapse imaging show that AP2 interacts with BSEP at the CM through a tyrosine motif at the carboxyl terminus of BSEP and mediates BSEP internalization from the CM of hepatocytes. Conclusion: AP2 mediates the internalization and subsequent degradation of CM-resident BSEP through direct interaction with BSEP and thereby modulates the canalicular expression and transport function of BSEP. This information should be useful for understanding the pathogenesis of severe liver diseases associated with intrahepatic cholestasis. (HEPATOLOGY 2012;55:1889–1900)
The bile salt export pump (BSEP) is an ATP-binding cassette transmembrane transporter located on the bile canalicular membrane (CM) and plays an indispensable role in the biliary secretion of monovalent bile salts.1, 2 Functional suppression of BSEP by mutations in the gene coding for BSEP,3-5 alcoholic hepatitis, autoimmune hepatitis, and drug-induced liver injury6, 7 causes intrahepatic cholestasis, a state of impaired biliary secretion, which results in hepatocellular damage because chronic exposure of cells to high concentrations of bile salts has a detergent effect and induces apoptosis or necrosis by activating signaling pathways.8 In severe cases, liver transplantation (LTx) is necessary because there are no established medical therapies for these diseases.
We and other groups have indicated previously that reduced expression of BSEP on the CM is at least partly responsible for BSEP dysfunction in patients with these kinds of intrahepatic cholestasis.7, 9-11 However, the detailed molecular mechanism responsible for these effects remains to be elucidated. Because disappearance of BSEP from the CM has been detected immediately after the stimulation of various cholestatic animal models with cholestatic factors,12-14 the molecular mechanism underlying BSEP internalization from the CM has generated considerable interest in trying to discover the cause of BSEP dysfunction in intrahepatic cholestasis and in developing pharmaceutical agents to treat this disease state.
Clathrin-mediated endocytosis is the most widely studied of the endocytic pathways. This process is initiated by nucleation through binding of FCH-domain-only proteins to the plasma membrane, then proceeds to cargo selection by adaptor proteins and coat assembly by the formation of clathrin triskelia and, finally, to dynamin-mediated scission.15 In these processes, the AP2 adaptor complex (AP2) plays an important role in cargo selection. AP2 is comprised of four nonidentical subunits: α-, β2-, μ2-, and σ2-adaptin. The 70-kDa trunk domains of α- and β2-adaptin, together with the 50-kDa μ2-adaptin and the 17-kDa σ2-adaptin, form the 200-kDa membrane-proximal core. The AP2 pore is the site of binding to the major endocytic motifs in cargo, tyrosine motif, YXXϕ (where X is any amino acid and ϕ is a hydrophobic residue), and acidic dileucine motifs, [ED]xxxL[LI], as well as to the plasma membrane, phosphatidyl inositol-4,5-bisphosphate, thereby mediating cargo selection. In addition, AP2 functions as a central hub in the maturation of the clathrin-coated pit through its binding motifs with clathrin and the accessory proteins. Given that clathrin-mediated endocytosis has been assumed to mediate BSEP internalization16 and that the cytoplasmic domains of BSEP have several potential AP2 interaction sites, it is possible that BSEP is internalized from the CM in a clathrin-mediated manner through the action of AP2.
We previously reported that 4-phenylbutyrate (4PBA), a drug used for ornithine transcarbamylase deficiency (OTCD), is promising for the treatment of intrahepatic cholestasis associated with reduced BSEP expression on the CM, because rats treated with this drug showed an increase in the hepatocanalicular expression and transport function of BSEP associated with the inhibition of its internalization and subsequent degradation from the cell surface.17 This current study explored the change in expression of α- and μ2-adaptin in liver specimens of Sprague-Dawley (SD) rats and patients with OTCD after 4PBA administration and showed those significant decreases, supporting the implication of AP2 in BSEP internalization.
Based on this background, this hypothesis was tested using HeLa cells, McA-RH7777 cells, human sandwich-culture hepatocytes (HSCH), and rat liver. The data show that AP2 mediates BSEP internalization from the CM and its subsequent degradation through direct interaction with a tyrosine motif in the carboxyl terminus of BSEP, and that it negatively regulates the hepatocanalicular expression and transport function of BSEP.
4PBA, 4-phenylbutyrate; AP2, AP2 adaptor complex; BEI, biliary excretion index; BFA, brefeldin A; BSEP, bile salt export pump; CHC, clathrin heavy chain; CM, canalicular membrane; CMVs, canalicular membrane vesicles; EGF, epidermal growth factor, EPS15, epidermal growth factor receptor pathway substrate 15; ER, endoplasmic reticulum; GST, glutathione S-transaminase; hBSEP, human bile salt export pump; HSCH, human sandwich-culture hepatocytes; ICC, immunocytochemistry; LTx, liver transplantation; mRNA, messenger RNA; OTCD, ornithine transcarbamylase deficiency; OTCD-4PBA(+), OTCD patients given 4PBA treatment; OTCD-4PBA(–), OTCD patients who were not given 4PBA treatment; PFIC2, progressive familial intrahepatic cholestasis type 2; P-gp, P-glycoprotein; qPCR, quantitative polymerase chain reaction; rBSEP, rat bile salt export pump; RNAi, RNA interference; SD, Sprague-Dawley; SE, standard error; shRNA, short hairpin RNA; siRNA, short interfering RNA; SM, sinusoidal membrane; SMVs, sinusoidal membrane vesicles; TC, taurocholate; Tf, transferrin; Ub, ubiquitin.
Materials and Methods
A detailed description of experimental procedures is presented in the Supporting Materials and Methods. All materials and methods used standard techniques and commercially available reagents of analytical grade.
Animals and 4PBA Administration.
Male SD rats (6-7 weeks old) were purchased from Nippon SLC (Shizuoka, Japan). All animals were maintained under standard conditions and treated humanely. Food and water were available ad libitum. To examine the effect of 4PBA on messenger RNA (mRNA) and protein expression, male SD rats were given 0.6 g/kg/day of 4PBA or vehicle by gavage, divided into three doses for 12 days, and were sacrificed to collect their livers for further analysis. The studies reported in this article were carried out in accord with the guidelines of the Institutional Animal Care Committee of the Graduate School of Pharmaceutical Sciences at The University of Tokyo (Tokyo, Japan).
Sampling of Liver Tissue and Measurement of Laboratory Parameters in OTCD Patients.
Small liver samples were collected from 7 OTCD patients who received 4PBA treatment (OTCD-4PBA[+]) when they underwent LTx. The first control group consisted of organ donors whose liver specimens were obtained before subsequent transplantation. Because the amount of donor liver tissue was matched with the body size of the recipient before transplantation, the liver segment that remained after size matching was available for this study. For the second control group, patients who did not receive 4PBA treatment (OTCD-4PBA[-]), samples were obtained by liver-needle biopsy from 3 OTCD patients (5-R, 6-R, and 7-R in Table 2 of a previous study18) from the test group mentioned above before 4PBA treatment. This second control group also included a liver specimen taken from an OTCD patient (8-R in Table 2 of a previous study18) who underwent LTx without 4PBA treatment. Liver specimens were immediately snap-frozen in liquid nitrogen and preserved at -70°C after sampling. Serum was collected on two occasions: first, on the day before beginning 4PBA treatment and, second, on the day before LTx after daily administration of 4PBA to OTCD patients or from organ donors who did not receive 4PBA treatment. The results of laboratory analyses of these samples are summarized in Table 2 of a previous study.18 The study was approved by the institutional ethics review boards and informed consent was obtained from all patients' parents before assessment.
Experiments were repeated at least three times, and the data in the figures are presented as the mean ± standard error (SE). The significance of differences between two variables and between multiple variables was calculated at the 95% confidence level by the Student t test and one-way analysis of variance with Dunnett's test, respectively, using Prism software (GraphPad Software, Inc., La Jolla, CA).
Decrease in Hepatic α-Adaptin Expression After 4PBA Administration to Rat.
After administration of 4PBA and following the approved dosage for children, which is optimal for increasing BSEP function,17 total RNA, sinusoidal membrane vesicles (SMVs) and canalicular membrane vesicles (CMVs) prepared from livers of SD rats were subjected to quantitative polymerase chain reaction (qPCR) and western blotting analysis, respectively. mRNA expression of α2-adaptin, a predominant isoform of α-adaptin in the liver,19 was significantly decreased by 71% by 4PBA treatment, whereas that of μ2-adaptin was not significantly affected (Fig. 1A). The decreased mRNA expression of α2-adaptin was reflected in the total amount of α-adaptin mRNA, which was measured using the common sequence of α1- and α2-adaptin, and showed a 67% decrease. No change in the mRNA expression of BSEP was observed after 4PBA administration, as described previously (data not shown).17 Western blotting analysis using SMVs and CMVs showed that 4PBA decreased α-adaptin expression to 47% of that observed with vehicle in SMVs and to 41% of that observed with vehicle in CMVs (Fig. 1B). Although mRNA expression of μ2-adaptin was not altered by 4PBA administration (Fig. 1A), its protein expression was reduced to 51% of that observed with vehicle in SMVs and to 55% of that observed with vehicle in CMVs (Fig. 1B), which is consistent with previous reports showing that the depletion of α-adaptin by short interfering RNA (siRNA) severely reduces μ2-adaptin expression.20, 21 As reported previously, BSEP expression in CMVs was increased by 4PBA (Fig. 1B).17
The purity of the SMV and CMV preparations was assessed according to levels of BSEP and P-glycoprotein (P-gp), which are CM markers, in SMVs and the Na+/K+-ATPase alpha 1 subunit, a sinusoidal membrane (SM) marker, in CMVs (Fig. 1B).
Alteration in the Hepatic Expression of α-Adaptin and BSEP in OTCD Patients After 4PBA Treatment.
4PBA was approved originally for the treatment of OTCD patients. We investigated the effect of 4PBA on the expression of α-adaptin and BSEP in human liver specimens from OTCD patients before and after beginning 4PBA treatment. These OTCD patients had serum concentrations of aspartate aminotransferase, alanine aminotransferase, and gamma-glutamyl transferase, which are liver-function indices, mostly within the normal range before 4PBA treatment.18 4PBA treatment did not change these values significantly and did not induce any adverse events.
mRNA and protein expression levels of α-adaptin and BSEP were compared using liver specimens obtained from the same OTCD patients before and after they were given 4PBA treatment and samples from organ donors. qPCR analysis showed that total mRNA expression of both isoforms of α-adaptin in OTCD-4PBA(+) was 65% and 60% lower than that in donors and OTCD-4PBA(-), respectively, whereas that of BSEP was similar in donors, OTCD-4PBA(-), and OTCD-4PBA(+) (Fig. 2A). Western blotting analysis of membrane fractions showed that the α-adaptin expression level in the OTCD-4PBA(+) group was lowered to 30% and 37% in donors and the OTCD-4PBA(-) group, respectively, whereas BSEP expression level in the OTCD-4PBA(+) group was elevated to 270% and 216% in donors and the OTCD-4PBA(-) group, respectively (Fig. 2B).
Interaction of α-Adaptin With BSEP.
3×FLAG-hBSEP and 3×FLAG-rBSEP, which are human and rat BSEPs harboring the 3×FLAG epitope tag at the N-terminus, were correctly trafficked to the plasma membrane, as previously reported for native BSEP (Fig. 3D-F).10 It also showed a degradation rate from the cell surface similar to that of native BSEP (data not shown), suggesting that the intracellular sorting and trafficking machinery of BSEP is not impaired by this epitope tag. Therefore, in this report, 3×FLAG-hBSEP or 3×FLAG-rBSEP was employed to enhance the immunosignal from BSEP in the experiments in which ectopically expressed BSEP was required.
The presence of α-adaptin in coimmunoprecipitates with anti-FLAG antibodies from the lysate of 3×FLAG-hBSEP-expressing (3×FLAG-hBSEP) HeLa cells and 3×FLAG-rBSEP-expressing (3×FLAG-rBSEP) McA-RH7777 cells and with anti-BSEP antibodies from the CMVs of SD rats was confirmed by western blotting analysis with anti-α-adaptin antibodies (Fig. 3A-C), which reacts with both isoforms of α-adaptin, α1-adaptin, and α2-adaptin.19 This result indicates that α-adaptin interacts with BSEP on the plasma membrane. A cellular localization study supported this proposed interaction. In 3×FLAG-hBSEP HeLa cells, both 3×FLAG-hBSEP and α-adaptin were detected at the plasma membrane with a colocalized dot-like structure (Fig. 3D, arrow). Immunohistochemical analysis showed that α-adaptin was present in the SM and CM of rat hepatocytes and 3×FLAG-rBSEP McA-RH7777 cells and colocalized with BSEP at the CM (Fig. 3E,F), which is consistent with the results of western blotting analysis of SMVs and CMVs (Fig. 1B). These results suggest that AP2 interacts with BSEP on the CM.
Increase in BSEP Expression Caused by the Depletion of α-Adaptin.
In this report, the RNA interference (RNAi) method was used to examine the influence of AP2 function on BSEP, because overexpression of proteins involved in clathrin-mediated endocytosis has been shown to induce indirect effects.20, 21
Depletion of α-adaptin severely impairs AP2 functioning.20, 21 In HeLa cells, α-adaptin suppression using siRNA designed for the knockdown of both isoforms of α-adaptin (α-adaptin siRNA) decreased the expression of another subunit of AP2 (μ2-adaptin), as reported previously,20, 21 and increased that of cell-surface-resident 3×FLAG-hBSEP by 3.5-fold (Fig. 4A). No effect was observed for the Na+/K+-ATPase alpha 1 subunit, a plasma-membrane marker. Endogenous BSEP expression in a crude membrane fraction of HSCH was also increased 2.0-fold by treatment with siRNA targeting α1- and α2-adaptin (α1,2-adaptin siRNA) (Fig. 4B). The increase in BSEP expression was accompanied by an enhanced biliary excretion index (BEI) for [3H]-taurocholate ([3H]-TC) (control siRNA, 43.9%; α1,2-adaptin siRNA, 69.2%) (Fig. 4C), which is an indicator of BSEP-transport function, as demonstrated by the near absence of biliary bile salt concentration in progressive familial intrahepatic cholestasis type 2 (PFIC2) patients.3
Depletion of the clathrin heavy chain (CHC), a basic subunit of the clathrin coat, induced 3×FLAG-hBSEP expression on the plasma membrane of HeLa cells (Fig. 4D). Given that AP2 functions as an adaptor protein that mediates the formation of clathrin-coated pits by interacting with both cargoes and clathrin, all of these results suggest that BSEP is internalized in a clathrin-dependent manner through AP2 and subsequently degrades.
Involvement of AP2 in BSEP Internalization.
HeLa cells stably expressing shRNA for the knockdown of both isoforms of α-adaptin (α-adaptin shRNA HeLa cells) were constructed (Fig. 5A) and used for endocytosis assay. In these constructed cells, the uptake of [125I]-transferrin (Tf), which is mediated in an AP2-dependent manner,21 was severely inhibited (Fig. 5B), whereas the uptake of [125I]-epidermal growth factor (EGF), which is mediated in an AP2-independent manner,21 was as efficient as in HeLa cells stably expressing control shRNA (control shRNA HeLa cells) (Fig. 5C), suggesting a disruption of AP2 functioning in α-adaptin shRNA HeLa cells. After this had been confirmed (Fig. 5A-C), the internalization rate of 3×FLAG-hBSEP was measured with cell-surface biotinylation (Fig. 5D,E). 3×FLAG-hBSEP internalized at a slow rate (1% per minute in control shRNA HeLa cells), consistent with the results of immunocytochemistry (ICC) showing a small number of the dot structures on the plasma membrane with the colocalization of 3×FLAG-hBSEP and α-adaptin (Fig. 3D). Suppression of AP2 function significantly inhibited the internalization of cell-surface-resident 3×FLAG-hBSEP. The amounts of internalized 3×FLAG-hBSEP after 5- and 15-minute incubations were 7.1- and 3.2-fold lower, respectively, in α-adaptin shRNA HeLa cells than in cells with control shRNA. Together, these results demonstrate that AP2 mediates the internalization of BSEP from the plasma membrane.
AP2 Interaction With a Tyrosine Motif in the Carboxyl Terminus of the BSEP.
Analysis of the primary structures of the cytoplasmic domains of the BSEP revealed three canonical AP2 interaction sites: an acidic dileucine motif in the amino terminus and an acidic dileucine motif and a tyrosine motif in the carboxyl terminus (Fig. 6A). Therefore, a glutathione S-transferase (GST) pull-down assay was performed to identify the putative binding sites in BSEP. We found that α- and μ2-adaptin were pulled down by GST-fusion proteins with the amino acid sequences of the carboxyl terminus (amino acid residues 1283-1321) of hBSEP (GST-hBSEP-Cter), but not by those with the amino acid sequences of the amino terminus (amino acid residues 1-61) of hBSEP (GST-hBSEP-Nter) or by GST itself (Fig. 6B). In addition, a point mutant of GST-hBSEP-Cter, in which a tyrosine residue within the potential AP2-binding sequence (1311YKL1314V) was substituted for alanine (GST-hBSEP-Cter Y1311A), did not interact with α- and μ2-adaptin (Fig. 6B). This suggested that the tyrosine motif in the carboxyl terminus of BSEP, a sequence preserved in BSEP found in species ranging from humans to skates, is a direct binding site for AP2.
In HeLa cells, 3×FLAG-hBSEP containing the Y1311A mutation (3×FLAG-hBSEP Y1311A) was correctly trafficked to the plasma membrane in a cell-surface biotinylation study (Fig. 6C) and showed similar cellular localization to that of 3×FLAG-hBSEP in an ICC study (data not shown). Therefore, 3×FLAG-hBSEP Y1311A was used for further analysis. The cell-surface biotinylation study demonstrated that the ratio of the amount of 3×FLAG-hBSEP Y1311A in the cell surface to that in the cell lysate was 2.5-fold higher than that for 3×FLAG-hBSEP (Fig. 6C,D), suggesting that the Y1311A mutation inhibits internalization of 3×FLAG-hBSEP and thereby induces its cell-surface expression. The inhibitory effect of the Y1311A mutation on 3×FLAG-hBSEP internalization was confirmed by the endocytosis assay with HeLa cells. The amounts of internalized 3×FLAG-hBSEP Y1311A for the 2- and 5-minute incubations were 8.1- and 4.1-fold lower, respectively, than those of 3×FLAG-hBSEP (Fig. 6E,F). Taken together, these results suggest that AP2 interacts with BSEP through the 1311YKL1314V motif, which mediates its internalization from the plasma membrane.
Involvement of AP2 in BSEP Internalization in a Polarized Hepatoma Cell Line.
We explored the implications of AP2 function in BSEP internalization in hepatocytes using McA-RH7777 cells, a polarized cell line from rat hepatoma, because McA-RH7777 cells develop CM through the formation of couplets like hepatocytes do and because, in this cell line, the interaction and colocalization of BSEP with α-adaptin were confirmed (Fig. 3B,E). It is practically difficult to label the CM of McA-RH7777 cells with sulfo-NHS-SS-biotin as is done for HeLa cells (Figs. 5D and 6E) because of the presence of tight junctions. Therefore, to measure the internalization of BSEP from CM, we constructed a fusion protein of rBSEP with Dronpa (Dronpa-rBSEP), a photochromic fluorescent protein,22 and conducted time-lapse imaging to measure the disappearance of its fluorescent intensity from the CM after switching the CM-resident Dronpa-rBSEP to a bright form (Fig. 7A). The fluorescence decay of Dronpa-rBSEP from the CM was significantly inhibited by the introduction of the Y1311A mutation (Fig. 7B), which is an essential amino acid residue for the AP2 interaction with BSEP (Fig. 6B). The fluorescent intensity on the CM 15 minutes after photoactivation was lowered to 65.1% and 79.9% for Dronpa-rBSEP and Dronpa-rBSEP-Y1311A, respectively. This result suggests that the internalization of Dronpa-rBSEP-Y1311A was slower than for Dronpa-rBSEP, and that AP2 mediates BSEP internalization through its interaction with BSEP in hepatocytes as well as in HeLa cells.
In this study, we showed that in SD rats and OTCD patients, the expression of α- and μ2-adaptin decreased significantly after treatment with 4PBA (Figs. 1 and 2), which can inhibit internalization and subsequent degradation of cell-surface-resident BSEP.17 Based on this, and the fact that potential AP2 recognition sites are present in the cytosolic region of BSEP, the current study explored whether AP2 mediates the internalization of BSEP from the CM through its direct interaction with BSEP. In HeLa cells, disruption of AP2 function by the suppression of α-adaptin significantly delayed the internalization rate of 3×FLAG-hBSEP (Fig. 5D,E), and, in rat hepatocytes, α-adaptin was colocalized with BSEP at the CM and coimmunoprecipitated with anti-BSEP antibody from CMVs (Fig. 3C,F). These findings are further supported by a GST pull-down assay and mutagenesis analysis of potential AP2-binding motifs in BSEP, because the introduction of the Y1311A mutation eliminated the interaction of GST-hBSEP-Cter with α- and μ2-adaptin (Fig. 6B) and significantly inhibited the internalization of 3×FLAG-hBSEP in HeLa cells (Fig. 6E,F) and Dronpa-rBSEP in McA-RH7777 cells, a polarized hepatoma cell line (Fig. 7). Considering that depletion of α-adaptin induced the expression and transport function of BSEP in HSCH (Fig. 4B,C), AP2 function is suggested to negatively regulate the hepatocanalicular expression and transport function of BSEP through an acceleration of the internalization of BSEP and its subsequent degradation. To our knowledge, this is the first study to suggest that AP2 is implicated in endocytosis from the CM of hepatocytes and to directly demonstrate part of the molecular mechanism responsible for BSEP internalization. Given that AP2 functions as an adaptor protein that mediates the formation of clathrin-coated pits through interactions with both cargoes and clathrin,15 and that the knockdown of CHC increases the cell-surface expression of 3×FLAG-hBSEP (Fig. 4D), this finding is consistent with a previous report showing that BSEP expression in the apical membranes of MDCK cells is increased by expression of the dominant negative form of cortactin,16 which is implicated in the invagination process of clathrin-mediated endocytosis,23 and by depletion of its binding partner, Hax-1.16
We reported previously that short-chain ubiquitination of cell-surface-resident BSEP is responsible for its degradation, because fusion of ubiquitin (Ub) to the carboxyl terminus of BSEP accelerates degradation of the cell-surface-resident BSEP.24 Ub modifications at one or several different lysine residues in cell-surface-resident membrane proteins act as a signal for lysosomal degradation. This can be achieved by either their accelerated internalization or their incorporation into multivesicular bodies, as opposed to being recycled to the plasma membrane in the early endosome.25 Together with our recent study showing that the ubiquitination of BSEP accelerates the internalization of BSEP,26 two molecular mechanisms-ubiquitination and an AP2-dependent mechanism-appear to modulate the internalization process of BSEP. Furthermore, considering a previous study by Ortiz et al., which showed that the dominant negative form of epidermal growth factor receptor pathway substrate 15 (EPS15) increased the cell-surface expression of BSEP,16 both mechanisms are likely to be implicated in BSEP internalization and to have mutual dependency. EPS15 contains binding sites for the α-adaptin and the two Ub-interacting motifs at the C-terminal domain, and this adaptor protein is involved in the initial steps of clathrin-coated pit formation through the recruitment of ubiquitinated receptors on the plasma membrane and through AP2 through each interacting site.27 Therefore, it is possible that ubiquitination, together with AP2 and EPS15, represents an internalization mechanism for BSEP, thereby modulating its degradation (Supporting Fig. 1). Further studies on AP2, ubiquitination, and EPS15 are needed to detail the molecular mechanism underlying BSEP internalization under physiological conditions and in severe liver diseases with intrahepatic cholestasis, in which BSEP internalization is accelerated.
Because expression of α-adaptin mRNA was significantly decreased in the liver of SD rats and OTCD patients after 4PBA treatment, and was accompanied by a reduction in the protein levels of α- and μ2-adaptin (Figs. 1 and 2), it is possible that 4PBA treatment reduces the function of AP2 in the liver by transcriptional modulation of α-adaptin, thereby inhibiting BSEP internalization and subsequent degradation and, consequently, elevating the canalicular expression and transport function of BSEP. Our previous finding that 4PBA inhibits internalization and subsequent degradation of cell-surface-resident BSEP supports this further.17 On the other hand, we have shown previously that susceptibility of the cell-surface-resident BSEP to ubiquitination is also associated with the underlying mechanism of 4PBA action.24 Therefore, future studies would be needed to understand the contribution of each aspect of molecular machinery incorporating EPS15 to 4PBA-mediated induction of BSEP. There are other possible mechanisms for the 4PBA-mediated induction of BSEP on the canalicular membrane: For example, an increase in the de novo synthesis of BSEP by the endoplasmic reticulum (ER), induction of its trafficking to the plasma membrane, or inhibition of Ub-mediated sorting leading to lysosomal degradation of the internalized BSEP. The former two pathways are unlikely, because we have shown previously that the level of the ER-resident immature form of BSEP is not increased by 4PBA treatment under brefeldin A (BFA)-treated conditions, when transport from the ER to the Golgi complex is prevented.17 Moreover, after BFA washout, the rate of maturation of the ER-resident immature form of BSEP to the cell-surface-resident mature form of BSEP was not accelerated by 4PBA treatment. We have also shown that the short-chain ubiquitination of BSEP, which is inhibited by 4PBA treatment, accelerates the internalization of BSEP, but not the degradation of internalized BSEP.26 Taken together, the most likely hypothesis is that 4PBA inhibits the function of AP2 through a reduced expression of α-adaptin and reduces the susceptibility of cell-surface-resident BSEP to the short-chain ubiquitination, thereby preventing BSEP internalization and its subsequent degradation.
As demonstrated by a loss-of-function study with mice in which μ2-adaptin-deficient embryos died before day 3.5 postcoitus,28 AP2 plays an essential role in cell viability. However, in our current and previous study,17 4PBA treatment caused no apparent adverse effects in OTCD patients or SD rats, despite the reduced expression of α- and μ2-adaptin. Given that μ2-adaptin heterozygotes exhibit decreased expression of α- and μ2-adaptin, but are viable and normal,28 and that cultured cell lines, in which suppression of α- and μ2-adaptin was achieved by RNAi treatment remained viable, even though the uptake of Tf was impaired (Fig. 5B),20, 21 it is likely that 4PBA treatment suppresses the function of AP2 to the extent allowed by the maintenance of cell viability. It is also conceivable that 4PBA affects the expression of α- and μ2-adaptin only in hepatocytes, and that the roles of α- and μ2-adaptin in hepatocytes are relatively minor, compared with their roles in other organs. The former hypothesis is supported by the preferred distribution of 4PBA to the liver after oral administration.29
The other principal finding of this study is that BSEP expression in OTCD patients is markedly increased after 4PBA administration without apparent adverse effects (Fig. 2B). This fact generates considerable interest in evaluating a protective effect of 4PBA against severe liver diseases with intrahepatic cholestasis. Clinical trials will determine the utility and safety of 4PBA as a therapy for these diseases.
In conclusion, we have shown that AP2 mediates the internalization and subsequent degradation of cell-surface-resident BSEP through direct interaction with BSEP, thereby regulating the canalicular expression and transport function of BSEP. A tyrosine motif in the carboxyl terminus of BSEP is essential for the interaction with AP2. This information should be helpful in understanding the pathogenic mechanism of severe liver diseases associated with intrahepatic cholestasis, in which BSEP internalization is accelerated, and for the development of potent drugs for these diseases.
The authors thank Dr. Larissa Kogleck for her advice on the manuscript and Mr. Naoki Kotani for his excellent technical advice on the studies with HSCH.