Abstract

Progressive familial cholestasis (PFIC) 2 and benign recurrent intrahepatic cholestasis (BRIC) 2 are caused by mutations in the bile salt export pump (BSEP, ABCB11) gene; however, their prognosis differs. PFIC2 progresses to cirrhosis and requires liver transplantation, whereas BRIC2 is clinically benign. To identify the molecular mechanism(s) responsible for the phenotypic differences, eight PFIC2 and two BRIC2 mutations were introduced in rat Bsep, which was transfected in MDCK II cells. Taurocholate transport activity, protein expression, and subcellular distribution of these mutant proteins were studied in a polarized MDCK II monolayer. The taurocholate transport activity was approximately half of the wild-type (WT) in BRIC2 mutants (A570T and R1050C), was substantially less in two PFIC2 mutants (D482G and E297G), and was almost abolished in six other PFIC2 mutants (K461E, G982R, R1153C, R1268Q, 3767–3768insC, and R1057X). Bsep protein expression levels correlated closely with transport activity, except for R1057X. The half-life of the D482G mutant was shorter than that of the WT (1.35 h vs. 3.49 h in the mature form). BRIC2 mutants and three PFIC mutants (D482G, E297G, and R1057X) were predominantly distributed in the apical membrane. The other PFIC2 mutants remained intracellular. The R1057X mutant protein was stably expressed and trafficked to the apical membrane, suggesting that the COOH-terminal tail is required for transport activity but not for correct targeting. In conclusion, taurocholate transport function was impaired in proportion to rapid degradation of Bsep protein in the mutants, which were aligned in the following order: A570T and R1050C > D482G > E297G > K461E, G982R, R1153C, R1268Q, 3767–3768insC, and R1057X. These results may explain the phenotypic difference between BRIC2 and PFIC2.

  • bile salt export pump
  • bile secretion
  • cholestasis
  • ABCB11
  • mutation

bile acids are amphipathic steroidal compounds produced from cholesterol in hepatocytes and secreted into bile across the canalicular membrane as glyco- or tauroconjugates. Bile acids are required for intestinal absorption of dietary fat and hydrophobic vitamins, and predominantly return to the liver through the enterohepatic circulation. Bile salt export pump (BSEP/Bsep, ABCB11) is the major canalicular ATP-dependent bile acid transporter (3). Its plasma membrane distribution is restricted to the canalicular domain (3, 12, 39, 40). Stimulation with taurocholate (TC) (13, 20, 22) or cAMP (10, 13, 22) increases bile acid secretion in part by recruitment of Bsep to the canalicular membrane in a phosphoinositide 3-kinase-dependent manner (21).

There are three genetically identified forms of progressive familial cholestasis (PFIC). PFIC1, PFIC2, and PFIC3 are caused by mutations in FIC1 (ATP8B1), BSEP, and MDR3 (ABCB4) genes, respectively (6). More than 10 different mutations in BSEP have been reported in patients having PFIC2 (9, 32). Of these, D482G and E297G mutations are most frequent (35). Studies of TC transport activity and intracellular trafficking by D482G and E297G mutants have reported inconstant results. Expression of these mutants in Sf9 cells resulted in reduced TC transport activity (25, 41); however, similar studies in HEK 293 cells revealed normal TC transport (8). D482G mutant Bsep was localized in the apical membrane of HepG2 (28) and Madin-Darby canine kidney (MDCK) cells (41), and in the cytoplasm of MDCK II cells (8). These results are inconsistent probably because different BSEP/Bsep species (human vs. rat vs. mouse) and assay systems (Sf9 vs. HEK 293, MDCK vs. HepG2) were used. Defective glycosylation (28) and decreased protein expression of Bsep (8, 28) were also reported. However, turnover and stability investigations of Bsep have not been described.

Benign recurrent intrahepatic cholestasis (BRIC) 1 and BRIC2 are caused by mutations in the FIC1 and BSEP genes, respectively. The phenotypes of PFIC2 and BRIC2 differ although both are caused by mutations in the same gene (BSEP). PFIC2 is characterized by progressive liver damage usually requiring transplantation. In contrast, BRIC2 is presumably manifested by intermittent and usually nonprogressive cholestasis. A similar situation exists regarding mutations in FIC1 (PFIC1 vs. BRIC1) (14, 38). How mutations in BSEP cause different phenotypes (i.e., BRIC2 vs. PFIC2) is unknown. In cystic fibrosis, cystic fibrosis transmembrane regulator (CFTR) gene mutations are classified into the following effects: defective protein production, processing, regulation, or conduction (30). Comparable identification of molecular mechanism(s) involved in PFIC2 and BRIC2 mutants has not been accomplished.

To elucidate such mechanisms, we introduced eight PFIC2 and two BRIC2 mutants in rat Bsep, transfected them in polarized MDCK II monolayer (23), and determined bile acid transport activity, subcellular localization, and protein stability. Results parallel clinical severity of PFIC2 and BRIC2.

MATERIALS AND METHODS

Materials.

[3H]TC was purchased from PerkinElmer Life Sciences (Boston, MA). TC and MG132 were from Calbiochem (San Diego, CA). Cycloheximide, tunicamycin, and bafilomycin A1 were from Sigma (St. Louis, MO). Sulfo-NHS-LC-biotin and streptavidin-agarose beads were purchased from Pierce (Rockford, IL).

cDNA.

pBK-Bsep vector expressing rat Bsep (GenBank accession number: U69487) (3) was provided by P. J. Meier (University Hospital, Zurich, Switzerland). pBsep-EYFP vector expressing Bsep fused with an enhanced yellow fluorescent protein (EYFP) was constructed by subcloning full-length Bsep into the KpnI/ApaI site of pEYFP-N1 (BD Biosciences Clontech, Palo Alto, CA) after removing the stop codon by polymerase chain reaction (PCR)-based mutagenesis. Vectors expressing Bsep mutants were constructed by site-directed mutagenesis (QuikChange XL site-directed mutagenesis kit, Stratagene, La Jolla, CA). In mutants anticipated to give rise to a stop codon, EFYP was fused immediately downstream of the truncated Bsep sequence. Mutagenesis was verified by DNA sequence analysis.

Cell culture.

MDCK II cells were maintained in a humidified atmosphere at 37°C with 5% CO2 in complete minimum essential medium (Life Technologies, Rockville, MD) supplemented with 10% fetal bovine serum, 100 units/ml penicillin, and 100 μg/ml streptomycin.

TC transport assay.

TC transport activity was studied as previously described (23). Briefly, MDCK II cells were replated on Transwell membrane inserts (no. 3470, diameter: 6.5 mm, pore size: 0.4 μm, Corning Costar, Cambridge, MA) at a concentration of 5 × 104 cells/membrane. After 24 h, cells were transfected with vectors expressing Na+-TC cotransporting polypeptide (Ntcp) (33), Bsep, or both by use of Lipofectamine 2000 (Invitrogen). To evaluate TC transport activity of Bsep mutants, mutants were used instead of the wild type (WT). Total amount of DNA used for transfection was adjusted to the same by adding β-gal-expressing vector. In 2 days, cells were incubated with 10 μM TC (containing 1 μM [3H]TC) in the basal compartment. After incubation for 1 h at 37°C, radioactivity in the apical medium (0.1 ml) was determined by scintillation counter (Beckman Coulter; model LS 1801). Transcellular TC flux was calculated from radioactivity in the apical medium.

Immunoblot analysis.

MDCK II cells grown on 100-mm plates were transfected as described above. After 48 h, cells were washed three times with PBS, incubated for 20 min at 4°C in 200 μl of lysis buffer (50 mM Tris, pH 7.5, 2 mM CaCl2, 1% Triton X-100) containing protease inhibitors (Complete Protease Inhibitor Cocktail Tablets, Roche Diagnostics, Indianapolis, IN), and harvested by scraping with a rubber policeman. Cell lysates were transferred to 1.5-ml tubes and centrifuged for 10 min (1,500 g, 4°C). Supernatants were transferred to new tubes and protein concentrations were determined by the method of Lowry et al. (17). Laemmli buffer was added to a final concentration of 1× and samples were incubated for 30 min at room temperature. Proteins were separated by SDS-PAGE on 7.5% gels, transferred onto polyvinylidene difluoride membranes (Millipore, Bedford, MA), and probed with the corresponding antibody against green fluorescent protein (GFP) (Invitrogen) or β-actin (Sigma). The antibody against GFP recognizes EYFP. The membranes were incubated with appropriate horseradish peroxidase-coupled secondary antibodies (Cell Signaling Technology, Beverly, MA) followed by ECL reaction (PerkinElmer Life Sciences). Blot images were captured by ATTO LightCapture System (ATTO, Tokyo, Japan) and band density was quantified by CS Analyzer (ATTO).

Biotinylation of cell surface proteins.

Biotinylation of cell surface Bsep proteins was performed according to the manufacturer's instruction. Briefly, MDCK II cells grown on 24-mm Transwell membrane inserts were transfected with WT or D482G mutant Bsep. After 2 days, cells were washed three times with ice-cold PBS (pH 8.0) and incubated with freshly made biotinylation solution (0.5 mg/ml sulfo-NHS-LC-biotin) in the upper or lower chamber at room temperature for 30 min. Cells were washed three times with PBS containing 100 mM glycine, scraped off, and suspended in 200 μl lysis buffer. Cell lysates were transferred to 1.5-ml tubes and centrifuged for 10 min (1,500 g, 4°C), and 50 μl of packed streptavidin-agarose beads were added to the supernatants. After incubation overnight with rotation at 4°C, beads were washed three times with lysis buffer. Proteins were eluted from the beads with Laemmli buffer, separated by SDS-PAGE, and immunoblotted.

Determination of protein degradation half-life.

MDCK II cells were transiently transfected with WT or mutant Bsep. After 24 h, cells were incubated with cycloheximide (20 μg/ml) to inhibit further protein synthesis. MG132 (10 μM) or bafilomycin A1 (1 μM) was added with cycloheximide, when necessary. Following incubation for indicated time periods, cells were harvested, lysed, and subjected to SDS-PAGE and immunoblot analysis. The band density was quantified as described above and normalized so that the density at time t = 0 was 100%. Log10 of the percentage of density was plotted vs. time, and half-life was calculated from the log10 of 50% for the protein.

Quantification of Bsep mRNA expression in transiently transfected MDCK II cells.

MDCK II cells were transiently transfected with an empty vector or WT or mutant Bsep, and total RNA was extracted 24 h after transfection. After digestion with DNase I, cDNA was obtained by reverse transcription with SuperScript III and random hexamers (Invitrogen). Quantitative real-time PCR was performed using TaqMan technology on ABI 7700 sequence detection system (Applied Biosystems, Foster City, CA). The primer and FAM dye-labeled probe for rat Bsep were purchased from Applied Biosystems. The ΔCT value was obtained by normalizing to an endogenous reference (rRNA), and the abundance of Bsep mRNA relative to the wild type was determined by calculating the value, 2Math.

Confocal laser scanning microscopy.

The subcellular distribution of Bsep protein was studied as previously described (23). Briefly, MDCK II cells grown on the Transwell membrane inserts (no. 3460, diameter: 12 mm, pore size: 0.4 μm, Corning) were transfected with Bsep expression vectors. After 48 h, membranes were cut, washed twice with PBS, fixed in 4% paraformaldehyde, and mounted with DABCO (1,4-diazabicyclo[2.2.2]octane, triethylenediamine) (Sigma). Nuclei were counterstained with DAPI (4′6-diamidino-2-phenylindole dihydrochloride, hydrate) (D8417, Sigma). After 1 h incubation at room temperature, Bsep localization was determined by confocal laser scanning microscopy (LSM-410; Carl Zeiss, Jena, Germany) with projection and three-dimensional orthogonal reconstruction modes.

Statistical analysis.

Mann-Whitney U-test was used to compare Bsep protein expression and protein degradation half-life. Correlation between TC transport activity and Bsep protein expression was tested for significance by Spearman rank correlation. P value <0.05 was considered significant.

RESULTS

TC transport activity, protein expression, and mRNA expression of the D482G mutant.

The D482G mutant is the most frequently observed mutation in PFIC2. The mutation lies in the first nucleotide-binding domain and the flanking amino acid sequences are identical among rat, mouse, and human orthologs (Fig. 1, A and B). In a polarized MDCK II monolayer, the D482G mutation revealed 32.3 ± 5.8% (mean ± SD) TC transport activity of that observed in WT (Fig. 2A). Bsep protein expression was quantified by immunoblot analysis of cell lysates from MDCK II cells transiently transfected with WT or mutated Bsep (Fig. 2B). WT Bsep yielded two bands of ∼170 and ∼145 kDa, which represent the mature (band C) and core-glycosylated (band B) forms of Bsep, respectively, as previously demonstrated (23). The D482G mutant produced the same two bands, suggesting normal maturation of Bsep protein. But the expression level of this mutant was significantly lower than that of WT. The density of the mature form of Bsep, which harbors TC transport activity, was 26.0 ± 8.2% of that of WT. These results reveal that the observed decrease in TC transport activity of the D482G mutation is attributable to decrease in Bsep protein expression. To exclude the possibility that transcription was altered, mRNA expression levels were quantified by real-time PCR. The mRNA expression of D482G was slightly higher than that of WT; however, the difference was not statistically significant (Fig. 2C), which suggests that the D482G mutant Bsep protein may be degraded faster than is the WT. Next, we tested the subcellular distribution of the D482G mutant (Fig. 3). MDCK II cells were transfected with WT or D482G Bsep, and the fluorescent signals were observed by confocal laser scanning microscopy. The D482G mutant was predominantly distributed along apical membranes similar to that of WT (Fig. 3A). To confirm this result, we studied localization by surface biotinylation. As expected, WT and D482G Bsep were not detected in the basolateral membrane (Fig. 3B). Only band C was present in the apical membrane, suggesting that Bsep protein required glycosylation to traffic onto the apical membrane (19). D482G mutant protein was present in the apical membrane at 30.6 ± 14.4% of the level of WT. This result is similar to that observed for TC transport activity (Fig. 2A).

Fig. 1.

Predicted BSEP/Bsep topology and the positions of PFIC2 and BRIC2 mutations. A: BSEP/Bsep are proposed to have 12 transmembrane domains and 2 sets of Walker A and B motifs with ABC signature sequence. Four N-linked glycosylation sites reside in the first extracellular loop. The positions of 8 PFIC2 mutations (E297G, K461E, D482G, G982R, R1057C, R1153C, 3767–3768insC, and R1268Q) and 2 BRIC2 mutations (A570T and R1050C) are indicated by ☆ and ★, respectively. B: amino acid sequences flanking mutation sites in human BSEP are aligned with mouse and rat Bseps. The PFIC2 and BRIC2 mutations in human BSEP and the corresponding residues in rat and mouse Bseps are shown in bold.

Fig. 2.

Taurocholate (TC) transport activity and expression of Bsep protein and mRNA of the D482G mutant expressed in MDCK II cells. A: MDCK II cells grown on Transwell membrane inserts were cotransfected with Ntcp and β-gal or Ntcp and either wild-type (WT) or the D482G mutant Bsep. In 2 days, cells were incubated for 1 h at 37°C with [3H]TC (1 μM) and cold TC (10 μM) in the basal compartment, and transcellular TC flux was determined on the basis of radioactivity in the apical medium. Bsep-dependent TC transport activity was calculated from the difference in transcellular TC flux between the value obtained in cells cotransfected with Ntcp and β-gal and that with Ntcp and Bsep and normalized to WT (62.3 ± 9.3 pmol/well) as 100%. Each value represents mean ± SD of 6 independent experiments performed in duplicate. B: MDCK II cells were transiently transfected with WT or D482G mutant Bsep and lysed. Five micrograms of cell lysates were separated by SDS-PAGE and subjected to immunoblot analysis using antibody to GFP. The expression of β-actin was studied as the control. A representative result is shown at top. Bottom: Bsep protein expression levels relative to those of β-actin were calculated and normalized to WT as 100%. Each value represents mean ± SD of 6 independent experiments performed in duplicate. C: MDCK II cells were transiently transfected with an empty vector or WT or D482G mutant Bsep, and total RNA was extracted 24 h after transfection. cDNA was obtained by reverse transcription with SuperScript III and random hexamers, and quantitative real-time PCR was performed using TaqMan technology on ABI 7700 sequence detection system. The ΔCT value was obtained by normalizing to an endogenous reference (rRNA) and the abundance of Bsep mRNA relative to WT was determined by calculating the value 2−ΔΔCT . Each value represents mean ± SD of 3 independent experiments.

Fig. 3.

Subcellular localization of the WT and D482G mutant Bsep. A: MDCK II cells grown on Transwell membrane inserts were transfected with the WT or D482G mutant Bsep. In 2 days cells were observed with confocal laser scanning microscopy. Nuclei were counterstained with DAPI. Bsep-EYFP-derived fluorescence is colored by yellow and nuclei by blue. Top, center, and right panels show x–z, x–y, and y–z plane images, respectively. The scale bars are 10 μm. B: MDCK II cells grown on 24-mm Transwell membrane inserts were transfected with the WT or the D482G mutant Bsep. After 2 days cells were washed with ice-cold PBS (pH 8.0) 3 times and incubated with freshly made biotinylation solution in either the upper (apical) or lower (basolateral) chamber at room temperature for 30 min. Cells were washed 3 times with PBS containing 100 mM glycine, scraped off, and suspended in lysis buffer. After rotation overnight with packed streptavidin-agarose beads at 4°C, proteins were eluted from the beads with Laemmli buffer, separated by SDS-PAGE, and immunoblotted. A representative blot is shown at top. Bottom: Bsep protein expression levels were determined and normalized to the apical WT as 100%. Each value represents mean ± SD of 3 independent experiments in duplicate.

Biochemical half-life of the D482G mutant.

To explore the mechanism(s) for attenuation of protein expression of the D482G mutant, the biochemical half-life was determined in MDCK II cells by analyzing Bsep protein expression after inhibiting further protein synthesis with cycloheximide treatment (Fig. 4). Figure 4A shows a representative blot. Band C of the WT was still present in 12 h, whereas, in D482G, bands disappeared in 4 h. The calculated half-life of band C in the D482G was 1.35 h, which was significantly shorter than that in WT (3.49 h, P < 0.05, Fig. 4B). Likewise the half-life of band B in the D482G was shorter than that in WT (0.41 h vs. 1.16 h, P < 0.05, Fig. 4B). These results suggest that the D482G protein is degraded faster than is the WT.

Fig. 4.

Biochemical half-life of the WT and D482G mutant Bsep in MDCK II cells. A: MDCK II cells were transiently transfected with the WT or D482G mutant Bsep. After 24 h cells were further cultured in the presence of cycloheximide (20 μg/ml) and lysed at 0, 3, 6, and 12 h (WT) or at 0, 1, 2 and 4 h (D482G). Cell lysates were separated by SDS-PAGE and subjected to immunoblot analysis using the antibody to GFP. Loaded protein amounts: 5 μg. A representative blot is shown. B: band density was measured and normalized so that the density at time t = 0 was 100%. The log10 of the percentage of density was plotted vs. time, and the half-life was calculated from the log10 of 50% for bands C and B. Each value represents mean ± SD of 6 independent experiments performed in duplicate.

To investigate where Bsep protein is degraded, we studied the effect of inhibitors on its biochemical half-life. Addition of MG132, a proteasome inhibitor (29), greatly stabilized Bsep protein. A representative blot is shown in Fig. 5A. Bands C of WT at 12 h and D482G at 4 h were denser in the presence of MG132. The half-life of each band was significantly extended by MG132 in WT and D482G; from 3.49 to 14.1 h in WT band C, from 1.35 to 13.2 h in D482G band C, from 1.16 to 2.95 h in WT band B, and from 0.41 to 1.73 h in D482G band B (Fig. 5C). In contrast, bafilomycin A1, a lysosomal proton pump inhibitor (43), did not affect stability of Bsep protein (Fig. 5, B and C). These results suggest that the WT and D482G mutant Bsep proteins are degraded in proteasomes rather than in lysosomes.

Fig. 5.

Effect of MG132 and bafilomycin A1 on the biochemical half-life of the Bsep protein. A and B: MDCK II cells were transiently transfected with the WT or D482G mutant Bsep. After 24 h, cells were cultured with cycloheximide (20 μg/ml) in the absence or presence of either MG132 (10 μM) (A) or bafilomycin A1 (1 μM) (B) and lysed at 0, 3, 6, and 12 h (WT) or at 0, 1, 2, and 4 h (D482G). Cell lysates were separated by SDS-PAGE and subjected to immunoblot analysis using the antibody to GFP. Loaded protein amounts: 5 μg. A representative blot is shown in A and B. C: band density was measured and normalized so that the density at t = 0 was 100%. The log10 of the percentage of density was plotted vs. time, and the half-life was calculated from the log10 of 50% for bands C and B. Each value represents mean ± SD of 3–6 independent experiments.

TC transport activity, protein expression, and mRNA expression of the PFIC2 and BRIC2 mutants.

We constructed seven additional PFIC2 (32) and two BRIC2 mutants (37) by site-directed mutagenesis (Fig. 1, A and B). The amino acid sequence flanking the indicated mutation is highly conserved among rat, mouse, and human orthologs. The 3767-3768insC mutation was expected to yield a truncated protein with additional 39 amino acids at the COOH terminus in human BSEP. The corresponding mutation in rat Bsep was expected to give rise to a truncated protein with 39 new amino acids at the COOH terminus similar to the human counterpart. The MDCK II monolayer was transfected with vectors expressing various mutations. Transfection efficacy was 5–7%. TC transport activity was determined as described. The E297G mutation revealed ∼10.2 ± 6.6% TC transport activity of WT (Fig. 6A). Other PFIC2 mutants (K461E, G982R, R1153C, R1268Q, 3767–3768insC, and R1057X) did not show significant TC transport activity. In contrast, two BRIC2 mutants exhibited considerable transport activity (A570T: 52.2 ± 13.0%, R1050C: 58.7 ± 9.9% of the WT). Protein expression levels (band C) were then studied in lysates of MDCK II cells transfected with the various mutants. The E297G mutant was expressed at 16.1 ± 6.8% of that of WT, whereas expression of the other mutants, except for R1057X, was at trace level (Fig. 6, B and C). R1057X was expressed as much as was the WT at ∼120 kDa, which is consistent with the molecular size anticipated after truncated mutant protein. A570T and R1050C mutants were expressed at 51.0 ± 17.2 and 50.0 ± 18.0% of WT, respectively. mRNA expression levels examined in MDCK II cells expressing these mutants except for A570T were slightly higher than those with WT, but the difference was not statistically significant (Fig. 6D). TC transport activity was significantly correlated with Bsep protein expression levels when R1057X was excluded from analysis (Fig. 6E, P < 0.001). Therefore, the decrease in the TC transport activity is probably attributable to rapid Bsep protein degradation.

Fig. 6.

Activity and expression of the PFIC2 and BRIC2 mutant Bsep. A: MDCK II cells grown on Transwell membrane inserts were cotransfected with Ntcp and β-gal or Ntcp and either WT or mutant Bsep. In 2 days, cells were incubated for 1 h at 37°C with [3H]TC (1 μM) and cold TC (10 μM) in the basal compartment and transcellular TC flux was determined based on radioactivity in the apical medium. Bsep-dependent TC transport activity was calculated from the difference in transcellular TC flux between the value obtained in cells cotransfected with Ntcp and β-gal and that with Ntcp and Bsep and normalized to WT (62.3 ± 9.3 pmol/well) as 100%. Each value represents mean ± SD of 3 independent experiments performed in duplicate. B: MDCK II cells were transiently transfected with an empty vector or WT or mutant Bsep and lysed. Five micrograms of cell lysates were separated by SDS-PAGE and subjected to immunoblot analysis using the antibody to GFP. The expression of β-actin was studied as the control. A representative result is shown. C: Bsep protein expression levels relative to those of β-actin were calculated and normalized to WT as 100%. Each value represents mean ± SD of 3–6 independent experiments. D: MDCK II cells were transiently transfected with an empty vector or WT or various mutant Bsep, and total RNA was extracted 24 h after transfection. cDNA was obtained by reverse transcription with SuperScript III and random hexamers, and quantitative real-time PCR was performed using TaqMan technology on ABI 7700 sequence detection system. The ΔCT value was obtained by normalizing to an endogenous reference (rRNA), and the abundance of Bsep mRNA relative to WT was determined by calculating the value, 2Formula. Each value represents mean ± SD of 3 independent experiments. E: TC transport activity and Bsep protein expression of WT and various mutant Bsep, which were normalized to WT as 100%, are plotted. Correlation between TC transport activity and Bsep protein expression of WT and mutants Bsep excluding R1057X was tested for significance by Spearman rank correlation.

Subcellular distribution study revealed that E297G, R1057X, A570T, and R1050C mutants were predominantly located along the apical membrane, whereas the other PFIC2 mutants (K461E, G982R, R1153C, R1268Q, and 3767-3768insC) remained intracellular (Fig. 7). In summary, BRIC2 mutants (A570T and R1050C) were expressed substantially, trafficked correctly, and maintained half of transport activity. Two PFIC2 mutants (D482G and E297G) were expressed at 10–30% of WT, trafficked correctly, and exhibited 10–30% activity. The R1057X mutant was well expressed and trafficked correctly, but it lacked activity. Other PFIC2 mutant proteins (K461E, G982R, R1153C, R1268Q, and 3767-3768insC) were unstable and lost all transport activity.

Fig. 7.

Subcellular localization of the PFIC2 and BRIC2 mutant Bsep. MDCK II cells grown on Transwell membrane inserts were transfected with the WT or mutant Bsep. In 2 days cells were observed with confocal laser scanning microscopy. Nuclei were counterstained with DAPI. Top, center, and right panels show x–z, x–y, and y–z plane images, respectively. The scale bars are 10 μm.

We determined whether proteasomal degradation was increased in PFIC2 and BRIC2 mutants. MDCK II cells were transiently transfected with WT or various mutant Bsep. After 24 h, cells were cultured with cycloheximide (20 μg/ml) in the absence or presence of either MG132 (10 μM) or bafilomycin A1 (1 μM). Bsep protein expression at 6 h was compared between the three groups (Fig. 8). In the WT and each mutant, Bsep protein degradation was significantly spared when cells were cultured with MG132 (P < 0.05) but not bafilomycin A1. Ammonium chloride, another lysosome inhibitor, did not affect Bsep protein expression (data not shown). These results suggest that the PFIC2 and BRIC2 mutants we studied are degraded in proteasomes rather than in lysosomes.

Fig. 8.

Effect of MG132 and bafilomycin A1 on protein expression of the PFIC2 and BRIC2 mutant Bsep. A: MDCK II cells were transiently transfected with the WT or various mutant Bsep. After 24 h, cells were cultured with cycloheximide (20 μg/ml) in the absence or presence of either MG132 (10 μM) or bafilomycin A1 (1 μM) and lysed at 6 h. Cell lysates were separated by SDS-PAGE and subjected to immunoblot analysis using the antibody to GFP. Loaded protein amounts: 5 μg. A representative blot is shown. C, control; MG, MG132; BA, bafilomycin A1. B: band density was measured and normalized to control as 1 in each mutant. Each value represents mean ± SD of 3–6 independent experiments. The expression of WT and mutant Bsep protein when cells were cultured in the presence of MG132 was significantly greater than that of the control (P < 0.05). In contrast, the expression in the presence of bafilomycin A1 was similar to that of the control.

DISCUSSION

In this study we demonstrated that the TC transport activity corresponded to Bsep protein level in most PFIC2 and BRIC2 mutants, indicating that the impaired function is derived from decreased protein expression. From the observation that a representative mutant, D482G, had a shorter biochemical half-life than the WT, rapid degradation of Bsep protein may be responsible for impaired function.

Vesicles extracted from Sf9 cells expressing a relevant mutant were previously used for TC transport activity assay (3). We chose a polarized MDCK II monolayer for this purpose because MDCK II, a canine cell line, is likely to be more physiological than Sf9, an insect cell. Posttranslational modification in Sf9 cells is different from that expressed in mammalian cells. Furthermore, the polarized MDCK II monolayer system permits us simultaneous observation of activity, expression, and subcellular distribution (23).

The biochemical half-life of the mature (band C) and core-glycosylated forms (band B) of the WT was 3.49 and 1.16 h, respectively (Fig. 4B), which are similar to values previously reported (3.76 and 1.20 h) (23). Rapid decay of the core-glycosylated form, presumably in the endoplasmic reticulum (ER), also occurs in CFTR expressed in HEK cells (42). The biochemical half-life of both forms of the D482G mutant was significantly shorter than that of the WT (mature form: 1.35 h, and core-glycosylated form: 0.41 h, Fig. 4B). These results suggest that after translation D482G Bsep protein is unstable and rapidly degraded. In contrast to a previous study (28), D482G was completely glycosylated (Fig. 2B). This difference may be explained by differences in species (mouse vs. rat Bsep) and cells used (MDCK II vs. HepG2 cells). The inhibition study revealed that degradation occurs predominantly in proteasomes rather than lysosomes, in agreement with the previous study (41). The core-glycosylated form is likely to be misfolded and degraded by the ER-associated degradation system (19). As shown by confocal laser scanning microscopy and cell surface biotinylation (Fig. 3), the mature form of Bsep traffics to the apical membrane, is internalized by endocytosis, and subsequently degraded mainly in proteasomes. The finding that lysosomes have little role in degradation of Bsep was unexpected, because CFTR is endocytosed from the apical membrane and degraded in the lysosome (31). Proteasomes are involved in endocytotic degradation of interleukin-2 receptor complex (44), growth hormone receptor (36), δ-opioid receptor (2), V2 vasopression receptor (18), and GABAA receptor subunits (1). The mechanism(s) by which proteasome inhibitors delay degradation of these receptors is unknown. Possibilities include depletion of the free ubiquitin pool thus reducing ubiquitination and endocytosis from the plasma membrane, impaired ubiquitination of essential components of the endocytic mechanism, and existence of a direct pathway between the plasma membrane and the proteasome. HAX-1 and cortactin participate in clathrin-mediated endocytosis of Bsep (26). Further research is required to elucidate the regulatory mechanism(s) whereby unstable Bsep is endocytosed and degraded.

In the eight PFIC2 mutants studied, D482G and E297G were predominantly distributed in the apical membrane and exhibited TC transport activity (D482G: 32.3% and E297G: 10.2% of WT). The mature form of Bsep protein (band C) of K461E, G982R, R1153C, R1268Q, and 3767–3768insC mutants was hardly detected (Fig. 6B) and, consequently, TC transport activity was abolished (Fig. 6A). These mutant proteins are highly unstable and rapidly degraded, which accounts for the defect of activity. Our observation that the D482G and E297G mutants exhibited less impaired transport activity than other mutants is in agreement with the clinical phenotype. The response to biliary diversion is better in PFIC2 patients carrying D482G and E297G mutations than in those with other mutations (27).

Mechanisms responsible for Bsep protein stability have not been investigated in detail. We demonstrated that rapid degradation of Bsep protein is responsible for the defect of TC transport activity with exception of the R1057X mutation. This mutant protein was expressed as well as the WT, and trafficked correctly but did not exert transport activity (Fig. 6). This mutant lacks 264 amino acids of the COOH terminus that contains the second ATP-binding domain (Fig. 1A), which is essential for transport activity. In CFTR (24, 34) and MRP2 (5), the PDZ-interacting domain located at the COOH terminus is critical for the apical targeting. Bsep does not have this domain; however, the COOH-terminal tail is unlikely to contain an apical membrane polarization signal.

The two BRIC2 mutants revealed more than 50% transport activity, which was significantly higher than that of PFIC2 mutants (Fig. 6). The reduced activity of BRIC2 may be caused by rapid degradation of Bsep protein similarly with PFIC2 mutants. In BRIC2 patients, the absence of immunodetectable BSEP protein in the canalicular membrane of hepatocytes was reported (15) as well as PFIC2 (9, 11). Our in vitro study reinforces this observation. PFIC2 patients progress to cholestatic liver cirrhosis and require liver transplantation, whereas, although BRIC2 patients experience cholestatic episodes, their prognosis is good. TC transport activity was 50–60% in BRIC2 mutants (A570T and R1050C) and 0–30% in PFIC2 mutants. Although TC transport activity in MDCK II cells correlated with phenotype of PFIC2 and BRIC2, other factors probably participate. In several patients who had clinical and histopathological characteristics of BRIC, progressed to PFIC (37, 38), which suggests possible phenotypic continuum between BRIC2 and PFIC2. Furthermore the E297G mutation, which is responsible for PFIC2, also occurs in BRIC2 patients (37). Therefore, although the BSEP genotype appears to play an important role in influencing clinical severity, other precipitating factors, including viral infection and pregnancy (37), may participate.

In conclusion, our study demonstrates that rapid protein degradation is responsible for decreased TC transport activity in all PFIC2 and BRIC2 mutations except R1057X. From the view of maintenance of TC transport activity, the mutants could be aligned in the following order: A570T and R1050C > D482G > E297G > K461E, G982R, R1153C, R1268Q, 3767-3768insC, and R1057X. R1057X is expressed stably and localized correctly but loses its activity (defective activity). Other mutations are categorized into defective processing. In cystic fibrosis, pharmacological chaperones that stabilize mutant CFTR protein have been extensively studied (4, 16). Drugs that stabilize mutant Bsep protein may be candidates for the treatment of PFIC2 and can be evaluated in the MDCK II monolayer system.

When we were preparing the manuscript, a paper appeared in which 4-phenylbutyrate (4PBA) was demonstrated to extend the half-life of cell surface-resident WT, E297G, and D482G BSEP by 1.8-, 2.5-, and 3.3-fold, respectively, in MDCK II cells (7). The authors also showed that the administration of 4PBA in Sprague-Dawley rats increased the biliary excretion of TC along with the enhanced BSEP expression at the canalicular membrane. Interestingly, the stabilizing effect of 4PBA was specific for BSEP protein because other apical proteins such as P-glycoprotein and dipeptidyl peptidase IV were not affected. The mechanism(s) by which 4PBA stabilizes BSEP protein should be elucidated in the future. Given that rapid protein degradation is responsible for decreased TC transport activity in most PFIC2 and BRIC2 mutations, this drug is expected to improve transport activity of these mutants.

GRANTS

This work was supported in part by Grants DK-54785 and 35652 (to I. M. Arias) from the National Institutes of Health and by Japanese Grant-in-Aid for Scientific Research (to T. Kagawa).

Acknowledgments

We thank H. Tsukamoto and H. Kamiguchi for technical support and Y. Wakabayashi for helpful advice.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

View Abstract