4-phenylbutyrate enhances the cell surface expression and the transport capacity of wild-type and mutated bile salt export pumps

Authors

  • Hisamitsu Hayashi,

    1. Department of Molecular Pharmacokinetics, Graduate School of Pharmaceutical Sciences, The University of Tokyo, Tokyo, Japan
    Search for more papers by this author
  • Yuichi Sugiyama

    Corresponding author
    1. Department of Molecular Pharmacokinetics, Graduate School of Pharmaceutical Sciences, The University of Tokyo, Tokyo, Japan
    • Professor and Chair, Department of Molecular Pharmacokinetics, School of Pharmaceutical Sciences, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan
    Search for more papers by this author
    • fax: (81) 3-5841-4766


  • Potential conflict of interest: Nothing to report.

  • See Editorial on Page 1340

Abstract

Progressive familial intrahepatic cholestasis type 2 (PFIC2) is caused by a mutation in the bile salt export pump (BSEP/ABCB11) gene. We previously reported that E297G and D482G BSEP, which are frequently found mutations in European patients, result in impaired membrane trafficking, whereas both mutants retain their transport function. The dysfunctional localization is probably attributable to the retention of BSEP in endoplasmic reticulum (ER) followed by proteasomal degradation. Because sodium 4-phenylbutyrate (4PBA) has been shown to restore the reduced cell surface expression of mutated plasma membrane proteins, in the current study, we investigated the effect of 4PBA treatment on E297G and D482G BSEP. Transcellular transport and cell surface biotinylation studies using Madin-Darby canine kidney (MDCK) II cells demonstrated that 4PBA treatment increased functional cell surface expression of wild-type (WT), E297G, and D482G BSEP. The prolonged half-life of cell surface–resident BSEP with 4PBA treatment was responsible for this result. Moreover, treatment of Sprague-Dawley rats with 4PBA resulted in an increase in BSEP expression at the canalicular membrane, which was accompanied by an increase in the biliary excretion of [3H]taurocholic acid (TC). Conclusion: 4PBA treatment with a clinically achievable concentration enhances the cell surface expression and the transport capacity of WT, E297G, and D482G BSEP in MDCK II cells, and also induces functional BSEP expression at the canalicular membrane and bile acid transport via canalicular membrane in vivo. 4PBA is a potential pharmacological agent for treating not only PFIC2 patients with E297G and D482G mutations but also other cholestatic patients, in whom the BSEP expression at the canalicular membrane is reduced. (HEPATOLOGY 2007;45:1506–1516.)

Progressive familial intrahepatic cholestasis type 2 (PFIC2), an inherited autosomal recessive liver disease, is characterized by cholestasis and jaundice in the first year of life. Without liver transplantation, this disease leads to cirrhosis and death before adulthood.1 Many studies have been performed in PFIC2 patients, and the hereditary defect in the expression of the bile salt export pump (BSEP/ABCB11) results in the acquisition of PFIC2.1, 2 BSEP is an adenosine triphosphate (ATP)-binding cassette transmembrane transporter located on the bile canalicular membrane, responsible for the biliary excretion of monovalent bile acids (such as taurocholic acid).3–8 It is likely that impaired biliary bile acid secretion causes accumulation of bile acids in hepatocytes and progressive severe hepatocellular damage because of the toxicity of a high concentration of bile acids. Genomic analysis of PFIC2 patients has revealed many kinds of missense, premature termination, frame shift, and splicing junction mutations in the BSEP gene.1 Among them, E297G and D482G, two missense mutations in the second intracellular loop and in the first ATP-binding domain, respectively, are frequently observed in PFIC2 patients. Indeed, each of these two mutations is present in 30% of European PFIC2 families.9 We have shown that the introduction of these two mutations causes a reduction in BSEP expression at the cell surface because of impaired membrane trafficking of BSEP followed by proteasomal degradation. However, both mutated BSEPs per se exhibit normal transport functions.10 Accordingly, restoration of the reduced cell surface expression of these mutated BSEPs is an important task for achieving the therapeutic goal for PFIC2 patients with E297G and D482G mutations.

Sodium 4-phenylbutyrate (4PBA) has been shown to be capable of restoring the reduced cell surface expression of cystic fibrosis transmembrane conductance regulator (CFTR/ABCC7) with the deletion of phenylalanine at 508 (CFTR ΔF508).11 CFTR ΔF508 is the most common mutation in cystic fibrosis patients12 and has similar features to E297G and D482G BSEP in that this mutated molecule accumulates in the endoplasmic reticulum (ER), followed by degradation in the proteasomes, but maintains its normal function as a chloride channel.13–15 4PBA is a nontoxic butyrate analog that was originally approved for clinical use as an ammonia scavenger in subjects with urea cycle disorders.16 Clinical trials of this agent in cystic fibrosis patients with ΔF508 demonstrated that CFTR function in the nasal epithelia is induced by 4PBA therapy.17 Considering that a surgical procedure such as liver transplantation is the only therapy to cure PFIC2, this compound may offer a new medical therapy for PFIC2.

In the current study, we examined whether 4PBA treatment with a clinically achievable concentration can restore the reduced cell surface expression of BSEP caused by the two common mutations and also investigated the mechanism by which 4PBA increases BSEP expression at the cell surface. We evaluated the effectiveness and mechanism of action of 4PBA by biological and transport functional experiments with Madin-Darby canine kidney (MDCK) II cells expressing wild-type (WT), E297G, and D482G BSEP. Surprisingly, 4PBA treatment enhanced the cell surface expression and the transport capacity of WT BSEP as well as that of both mutated BSEPs. Therefore, the effectiveness of 4PBA treatment was further determined in vivo by a [3H]TC infusion study into Sprague-Dawley (SD) rats and biological and transport functional experiments using canalicular membrane vesicles (CMVs) from SD rats.

Abbreviations

4PBA, sodium 4-phenylbutyrate; ATP, adenosine triphosphate; BFA, brefeldin A; BSEP, bile salt export pump; CF/CFTR ΔF508, cystic fibrosis /cystic fibrosis transmembrane conductance regulator with the deletion of phenylalanine at 508; CMVs, canalicular membrane vesicles, DPPIV, dipeptidyl peptidase IV; ER, endoplasmic reticulum; GFP, green fluorescent protein; I, infusion rate; MDCK, Madin-Darby canine kidney; MOI, multiplicity of infection; PFIC, progressive familial intrahepatic cholestasis; SD, Sprague-Dawley; SDS-PAGE, sodium dodecyl sulfate polyacrylamide gel electrophoresis; WT, wild-type.

Materials and Methods

Materials.

Pharmaceutical grade 4PBA was purchased from Scandinavian Formulas Inc. (Sellersville, PA). We obtained [3H]taurocholic acid ([3H]TC) (2 Ci/mmol) from NEN Life Science Products (Boston, MA). Antibodies against the human BSEP, P-glycoprotein (P-gp) (C219), and dipeptidyl peptidase IV (DPPIV) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA), Signet (Dedham, MA), and BD Biosciences (Mountain View, CA), respectively. Antiserum for rBsep was raised in rabbits against an oligopeptide (the carboxyl terminal of rBsep; AYYKLVITGAPIS).18 All other chemicals were of analytical grade. MDCK II cells were cultured in Dulbecco's modified Eagle medium (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum, penicillin (100 U/ml), and streptomycin (100 U/ml) at 37°C with 5% CO2 and 95% humidity.

Generation of Recombinant Adenovirus.

The BD Adeno-X Adenoviral Expression System (BD Biosciences, Palo Alto, CA) was used to establish the human WT, E297G, and D482G BSEP recombinant adenoviruses as previously described.10 The virus titer was checked with an Adeno-X Rapid Titer Kit (Clontech). As a control, recombinant adenoviruses containing green fluorescence protein (GFP) were used.

4PBA Treatment.

MDCK II cells were seeded in six-well plates at a density of 2.5 × 105 cells per well. After a 24-hour culture, confluent cells were infected by recombinant adenovirus containing cDNAs for WT, E297G, D482G BSEP, and GFP at 200 multiplicity of infection (MOI). 4PBA treatment was carried out for various periods and at various concentrations. After the end of the 4PBA treatment period, crude membrane fractions were prepared as described.19 The specimens were separated by 6% sodium dodecyl polyacrylamide gel electrophoresis (SDS-PAGE) and subjected to western blot analysis.

Western Blot Analysis.

Specimens were loaded per well onto a 6% SDS-PAGE plate with a 3.75% stacking gel, and subjected to western blot analysis as described.10 Immunoreactivity was detected with an ECL Advance Western Blotting Detection Kit (Amersham Biosciences, Piscataway, NJ). The intensity of the band indicating mature BSEP was quantified by Multi Gauge software Ver 2.0 (Fujifilm, Tokyo, Japan).

Transcellular Transport Assay.

MDCK II cells were seeded on transwell membrane inserts (pore size of 3 μM; Falcon, Bedford, MA) in 24-well plates at a density of 1.5 × 105 cells per insert. After 2 days' culture, confluent cells were infected by recombinant adenovirus containing cDNAs for WT, E297G, D482G BSEP, and GFP at 50 MOI. Cells were cultured for 24 hours after infection and subsequently treated with 1 mM 4PBA. Then, 24 hours after treatment, the transcellular transport assays were performed as described.19

The apparent efflux clearance across the apical membrane (PSapical) was calculated by dividing the steady-state rate for the transcellular transport of [3H]TC determined over 2 hours by the cellular concentration of [3H]TC determined at the end of the experiments (2 hours).

Cell Surface Biotinylation and Determination of Degradation Rate of Cell Surface Expressing Protein.

MDCK II cells were seeded in 6-well plates at a density of 2.5 × 105 cells per well. After a 24-hour culture, confluent cells were infected by recombinant adenovirus containing cDNAs for WT, E297G, D482G BSEP, and GFP at 200 MOI. Cells were cultured for 24 hours after infection and subsequently treated with 1 mM 4PBA. Then, 24 hours after treatment, cell surface biotinylation was performed as described.10

When the degradation rate of the cell surface–resident protein was examined, biotinylated MDCK II cells were incubated for various periods at 37°C, with or without 1 mM 4PBA, before solubilization. The remaining biotinylated protein was isolated as described and separated by 6% SDS-PAGE and subjected to western blot analysis.

Determination of BSEP Messenger RNA Levels.

MDCK II cells were seeded in six-well plates at a density of 2.5 × 105 cells per well. After a 24-hour culture, confluent cells were infected by recombinant adenovirus containing cDNAs for WT, E297G, D482G BSEP, and GFP at 50 MOI. Cells were cultured for 24 hours after infection and subsequently treated with 1 mM 4PBA. Then, 24 hours after treatment, RNA was isolated using ISOGEN (Wako Pure Chemical Industries, Osaka, Japan) according to the manufacturer's instructions. Reverse transcription was performed as previously reported10 after DNase I (Takara Shuzo, Tokyo, Japan) treatment at 37°C for 1 hour. BSEP mRNA levels were determined by real-time quantitative PCR using a LightCycler and the appropriate software (version 3.53; Roche Diagnostics, Mannheim, Germany) as described previously.10 Quantitative PCR was performed with 5′-dAGTGGGGGAGCTGAATACAA-3′and 5′-dCCAATGGTGGCTGCTCCAAT-3′ (BSEP) and 5′-AACGACCCCTTCATTGAC-3′ and 5′-TCCACGACATACTCAGCAC-3′ (GAPDH) as primers. BSEP gene expression in each reaction was normalized by the expression of GAPDH.

Actinomycin D (Act D) and Cycloheximide (CHX) Treatment.

MDCK II cells were seeded in 6-well plates at a density of 2.5 × 105 cells per well. After a 24-hour culture, confluent cells were infected by recombinant adenovirus containing cDNAs for WT, E297G, and D482G BSEP at 200 MOI. Cells were cultured for 36 hours after infection and subsequently treated, with or without 5 μg/ml Act D (Sigma, St. Louis, MO), to inhibit mRNA synthesis and, with or without 20 μg/ml CHX (Sigma, St. Louis, MO), to inhibit protein synthesis. Then, 2 hours after the treatment, 4PBA was added to medium at 1 mM, in the presence or absence of Act D and CHX. Subsequently, 6 hours (E297G, D482G BSEP) or 8 hours (WT) after 4PBA treatment, the crude membrane fraction was prepared as described previously.19 The specimens were separated by 6% SDS-PAGE and subjected to western blot analysis.

Brefeldin A Washout Study.

MDCK II cells were seeded in 6-well plates at a density of 2.5 × 105 cells per well. After a 24-hour culture, confluent cells were infected by recombinant adenovirus containing cDNAs for WT, E297G, and D482G BSEP at 200 MOI. Cells were cultured for 12 hours after infection, and subsequently treated, with or without 5 μg/ml brefeldin A (BFA) (Sigma, St. Louis, MO), to inhibit the translocation of BSEP from ER to the Golgi complex. Then, 2 hours after the treatment, 4PBA was added into media at 1 mM, in the presence or absence of BFA. Subsequently, 12 hours after 4PBA treatment, BFA was washed out for various periods, with or without 1 mM 4PBA, before the preparation of the crude membrane fraction as described previously.19 The specimens were separated by 6 % SDS-PAGE and subjected to western blot analysis.

Animals.

Male SD rats (6-7 weeks old) were purchased from Nippon SLC (Shizuoka, Japan). All animals were maintained under standard conditions with a reverse dark-light cycle and treated humanely. Food and water were available ad libitum. The studies reported in this manuscript were carried out in accordance with the guidelines provided by the Institutional Animal Care Committee (Graduate School of Pharmaceutical Sciences, The University of Tokyo, Tokyo, Japan).

In Vivo Infusion Study in Sprague-Dawley Rats.

Male SD rats (6-7 weeks old) were fed 4PBA or vehicle by gavage for various periods (5, 10, 15 days) and at various doses (0.2, 0.6, 2.4 g/kg/day). 4PBA or vehicle was given as 3 divided doses. After 4PBA treatment, the in vivo [3H]TC infusion study was performed as described.20 The [3H]TC in saline was infused through the femoral vein cannula at a rate for 70 ng/min/kg for 90 minutes. Blood and bile specimens were obtained at specified times. The radioactivity associated with the plasma, bile and liver was determined in a liquid scintillation counter as described.20

Pharmacokinetic Analysis.

The total plasma clearance (CLtotal), biliary clearance normalized by circulating plasma (CLbile, plasma), and biliary clearance normalized by the liver concentration (CLbile, liver) were calculated from the equations CLtotal = I/Css, plasma, CLbile, plasma = Vss, bile/Css, plasma, and CLbile, liver = Vss, bile/Css, liver, where I, Css, plasma, Vss, bile, and Css, liverrepresent the infusion rate (in ng/min/kg), plasma concentrations at steady-state (in ng/ml), biliary excretion rate at steady-state (in ng/min/kg), and average hepatic concentration at steady-state (in ng/ml), respectively. Plasma concentrations at steady-state (Css, plasma) was determined as the mean value of the plasma [3H]TC concentrations at 15, 30, 60, and 90 minutes. Biliary excretion rate at steady-state (Vss, bile) was determined as the mean value of the biliary excretion rate of [3H]TC from 10 to 15 minutes, 25 to 30 minutes, from 55 to 60 minutes, and from 85 to 90 minutes. Css, liver was determined as the average hepatic [3H]TC concentration at the end of the in vivo experiment.

Transport Assays with Canalicular Membrane Vesicles.

Male SD rats (6-7 weeks old) were given by gavage 0.6 g/kg/d 4PBA or vehicle in 3 divided doses for 10 days. CMVs were prepared from the liver of the treated rats as described.18 Prepared CMVs were subjected to western blot analysis and used for transport assays. Transport assays were performed using the rapid filtration method previously reported.4

Results

4PBA-Mediated Up-regulation of BSEP Expression at the Cell Surface.

MDCK II cells expressing WT, E297G, and D482G were treated with 4PBA for various periods and at various concentrations (Fig. 1A,B). In the previous study, we suggested that the mature cell surface–resident form and immature ER-resident form of BSEP were detected as 170 and 150 kDa, respectively.10 The effectiveness of 4PBA treatment was examined by investigating the expression level of the mature form of BSEP (170 kDa) via western blot analysis of crude membrane fractions. 4PBA treatment altered the expression level of the mature form of not only E297G and D482G BSEP but also WT BSEP in a concentration-dependent and time-dependent manner, the optimal condition being 1 mM for 24 hours. The expression level of the mature form of WT, E297G, and D482G BSEP was increased 2.5-fold to 3-fold in response to 1 mM 4PBA treatment for 24 hours, which is a clinically achievable concentration.21–23

Figure 1.

The time-, and concentration-dependent effect of 4PBA treatment on BSEP expression in MDCK II cells. (A) Time-dependent effect of 4PBA treatment on BSEP expression. MDCK II cells expressing WT, E297G, and D482G BSEP were treated for the indicated periods with 1 mM 4PBA before the preparation of crude membrane fractions. Prepared specimens (40 μg) were separated by 6% SDS-PAGE and subjected to western blot analysis. (B) Concentration-dependent effect of 4PBA treatment on BSEP expression. MDCK II cells expressing WT, E297G, and D482G BSEP were treated with the indicated 4PBA concentration for 24 hours before the preparation of crude membrane fractions. Prepared specimens (40 μg) were separated by 6% SDS-PAGE and subjected to western blot analysis.

We then examined the increase in cell surface expression of BSEP by 4PBA treatment using transcellular transport assay and cell surface biotinylation methods. In the transcellular transport assay, the function of BSEP, which mediates the efflux of [3H]TC into the apical side, was studied by measuring the transcellular transport of [3H]TC across MDCK II monolayers (Fig. 2A). Vectorial transport of [3H]TC in the apical direction was observed in MDCK II monolayers expressing WT, E297G, and D482G BSEP, and hardly detected in MDCK II monolayers expressing GFP. Similar to hepatocyte, coexpression of Na+-taurocholate cotransporting polypeptide (NTCP), basolateral uptake transporter, and BSEP, apical efflux transporter, is needed to detect the vectorial transport of [3H]TC in the apical direction in LLC-PK1 and MDCK monolayers.19, 24 The finding that the expression of BSEP alone is sufficient to induce the vectorial transport of [3H]TC in the apical direction suggests that MDCK II cells endogenously express uptake transporter for bile acids in the basolateral membrane as suggested.25 The basal-to-apical flux of [3H]TC across MDCK II monolayers expressing WT, E297G, and D482G BSEP was increased 1.5-fold, 2.5-fold, and 3-fold, respectively, by 4PBA treatment under optimal conditions (1 mM, 24 hours), whereas the increase in the basal-to-apical flux of [3H]TC was not detected in MDCK II cells expressing GFP. 4PBA treatment did not affect the apical-to-basal flux of [3H]TC across any type of MDCK II monolayers. To directly confirm the effect of 4PBA treatment on the function of BSEP, PSapical, a kinetic parameter essential for BSEP function,10, 26 was calculated (Fig. 2B). PSapical was increased in MDCK II cells expressing WT, E297G, and D482G BSEP, but not affected in MDCK II cells expressing GFP by 4PBA treatment. BSEP-dependent PSapical (PSapical, BSEP) in MDCK II cells expressing WT, E297G, and D482G BSEP, which was calculated by subtracting the PSapical value in MDCK II cells expressing GFP from that in MDCK II cells expressing WT, E297G, and D482G BSEP, was enhanced 1.7-, 3.0-, and 2.8-fold, respectively, by 4PBA treatment under optimal conditions (1 mM, 24 hours).

Figure 2.

The effect of 4PBA treatment on cell surface expression of BSEP in MDCK II cells. MDCK II cells expressing WT, E297G, D482G BSEP, and GFP were treated for 24 hours, with or without 1 mM 4PBA, before the experiments. The cell surface expression of BSEP was determined by transcellular transport assay and cell surface biotinylation analysis. (A) Time profiles of the transcellular transport of [3H]TC across MDCK II monolayers. Transcellular transport of 1 μM [3H]TC across MDCK II monolayers expressing WT, E297G, D482G BSEP, and GFP was examined as a function of time. Closed (black circle, black square) and Open (white circle, white square) symbols represent MDCK II monolayers treated, with and without 4PBA, respectively. Circles and squares represent the transcellular transport in the basal-to-apical and apical-to-basal directions, respectively. Each point and vertical bar represents the mean ± SE of triplicate determinations. Where vertical bars are not shown, the SE was contained within the limits of the symbols. (B) The transport of [3H]TC across the apical membrane of MDCK II monolayers. The clearance of the transport of [3H]TC across the apical membrane of MDCK II monolayers expressing WT, E297G, D482G BSEP, and GFP (PSapical) was determined as described in Materials and Methods. Each bar represents the mean ± SE of triplicate determinations. Significantly different from control by Student t test (*P < 0.05; **P < 0.01). (C, D) Determination of cell surface expression by the biotinylation analysis. The cell surface fractions were isolated by the biotinylation method as described in Materials and Methods. Prepared specimens were separated by 6% SDS-PAGE and subjected to western blot analysis.

Furthermore, we examined the cell surface expression of WT and mutated BSEPs by cell surface biotinylation in MDCK II cells (Fig. 2C). Cell surface expression of WT, E297G, and D482G BSEP was increased 1.8-fold, 3.1-fold, and 2.6-fold, respectively, by 4PBA treatment under optimal conditions (1 mM, 24 hours), whereas cell surface expression of exogenously expressing P-gp and endogenously expressing DPPIV was not affected (Fig. 2D). The increase in cell surface expression of BSEP by 4PBA treatment was to an equivalent degree to that in PSapical, BSEP, suggesting that 4PBA treatment with a clinically achievable concentration can enhance the transport capacity of BSEP in MDCK II cells expressing WT, E297G, and D482G BSEP by increasing the cell surface expression of WT, E297G, and D482G BSEP.

The Transcriptional and Translational Effect of 4PBA Treatment on the Increased Cell Surface Expression of BSEP.

We then examined the mechanism by which 4PBA treatment may increase the cell surface expression of BSEP. A possible mechanism is that 4PBA treatment promotes transcription or translation of WT, E297G, and D482G BSEP and consequently increases the cell surface expression of WT, E297G, and D482G BSEP by mass action. This hypothesis is also supported by the fact that 4PBA is an analog of butyrate, a known transcriptional regulator, which upregulates β- and γ-globin.27

The expression of BSEP mRNA in MDCK II cells, with or without 4PBA treatment, was quantified by real-time PCR. The expression of BSEP mRNA was normalized by that of GAPDH, because the GAPDH mRNA expression level was not affected by 4PBA treatment (data not shown). WT, E297G, and D482G BSEP mRNA expression levels were slightly increased by 4PBA treatment, but the difference was not statistically significant [P = 0.10 (WT), 0.20 (E297G, D482G)] (Fig. 3A).

Figure 3.

The transcriptional and translational effect of 4PBA treatment on BSEP expression in MDCK II cells. (A) Determination of expression levels of BSEP mRNA. MDCK II cells expressing WT, E297G, and D482G BSEP were treated for 24 hours, with or without 1 mM 4PBA, before the experiments. Real-time quantitative PCR was performed as described in Materials and Methods. BSEP expression in each reaction was normalized by the expression of GAPDH. Each bar represents the mean ± SE of triplicate determinations. (B, C) 4PBA-induced BSEP up-regulation during transcriptional inhibition with Act D (B) and translational inhibition with CHX (C). MDCK II cells expressing WT, E297G, and D482G BSEP were treated for 6 hours (E297G, D482G BSEP) or 8 hours (WT BSEP), with or without 1 mM 4PBA, in the presence or absence of 5 μg/ml Act D (B), and 20 μg/ml CHX (C) before the preparation of crude membrane fractions. Prepared specimens (40 μg) were separated by 6% SDS-PAGE and subjected to western blot analysis.

We further examined the transcriptional and translational effect of 4PBA treatment on the increased expression of mature BSEP by pre-treating MDCK II cells with Act D and CHX (Figs. 3B,C). The inhibition of transcription with Act D and translation with CHX did not affect the 4PBA-mediated increase in the mature form of BSEP in MDCK II cells expressing WT, E297G, and D482G BSEP. These results indicate that a post-translational mechanism mainly contributes to the 4PBA-mediated increase in the cell surface expression of WT, E297G, and D482G BSEP.

The Effect of 4PBA Treatment on BSEP Maturation.

One possible post-translational mechanism is the promotion of BSEP maturation from the ER-resident immature form to the mature cell surface-resident form. This hypothesis is also supported by previous reports that 4PBA is considered to promote the trafficking of the CFTR ΔF508 from the ER to the plasma membrane and to restore activity in the plasma membrane.11, 28, 29 To examine the effect of 4PBA treatment on BSEP maturation, we blocked the export of BSEP from the ER to the Golgi complex by using BFA, which results in the accumulation of the immature ER-resident form of BSEP, and determined the maturation rate in MDCK II cells after BFA washout (Fig. 4A,B). The maturation rate was evaluated by measuring the increased band density of mature BSEP by western blot analysis at the indicated time points after BFA washout. The band density of the mature form was calculated by subtracting the density detected at time 0 from that at the designated time points (Fig. 4B). The expression level of the immature ER-resident form of WT, E297G, and D482G BSEP was almost identical in non– and 4PBA-treated MDCK II cells just after BFA washout (Fig. 4A; 0 hours), and that of the mature form was similar in non– and 4PBA-treated MDCK II cells until 3 hours after BFA washout, at which time the mature form of BSEP linearly increased. The expression level of the mature form was higher in 4PBA-treated MDCK II cells compared with nontreated cells 8 hours after BFA washout. This result suggests that 4PBA treatment does not promote WT, E297G, and D482G BSEP maturation, but stabilizes the mature form of these proteins.

Figure 4.

The effect of 4PBA treatment on BSEP maturation in MDCK II cells. (A) Determination of BSEP maturation. MDCK II cells expressing WT, E297G, and D482G BSEP were treated for 12 hours, with or without 1 mM 4PBA, in the presence or absence of 5 μg/ml BFA. To show the conversion of immature low-molecular-weight BSEP accumulated by BFA to mature high-molecular-weight BSEP, BFA was washed out for various periods, with or without 1 mM 4PBA, before the preparation of crude membrane fractions. Prepared specimens (40 μg) were separated by 6% SDS-PAGE and subjected to western blot analysis. (B) Quantification of band density indicating mature BSEP in (A). The intensity of the band indicating mature BSEP was quantified by Image Gauge software. Closed and open circles represent the immunosignal of mature BSEP in MDCK II cells treated, with and without 4PBA, respectively.

4PBA-Mediated Prolongation of the Half-life of Cell Surface-Resident BSEP.

Cell surface–resident BSEP constitutively cycles between the canalicular membrane and the intracellular compartment, and is finally removed from this cycle to the degradation pathway.30, 31 Therefore, blocking the entry to the degradation pathway is also a possible mechanism for the increased cell surface expression of BSEP. To examine whether 4PBA treatment can inhibit the cell surface–resident BSEP from entering the degradation pathway, we measured the degradation rate of the cell surface–resident BSEP using biotin-labeling methods in MDCK II cells expressing WT, E297G, and D482G BSEP. The half-life of cell surface–resident WT and E297G BSEP was prolonged 1.8- and 2.5-fold, respectively, by 4PBA treatment, whereas those of exogenously expressing P-gp and endogenously expressing DPPIV was not affected (Fig. 5A,B). The degradation rate of cell surface–resident D482G BSEP could not be determined under the same conditions as WT and E297G because of its low expression level. Because low-temperature treatment has been documented to be capable of correcting the trafficking of the mutated misfolded protein,32, 33 we examined its effect on D482G (Fig. 5C). The cell surface expression of D482G increased 4.0-fold by 24-hour culture at a lower temperature (27°C). We further examined the degradation rate of cell surface–resident D482G at 37°C after 24 hours' culture at a lower temperature (Fig. 5D,E). The half-life of the cell surface–resident D482G BSEP was prolonged 3.3-fold by 4PBA treatment such as WT and E297G BSEP. The prolonged half-life of the cell surface–resident WT, E297G, and D482G BSEP is consistent with the result of the BFA washout study, which suggests that 4PBA treatment stabilizes the mature form of BSEP (Fig. 4A,B).

Figure 5.

The effect of 4PBA treatment on the degradation rate of cell surface–resident BSEP in MDCK II cells. (A) The degradation rate of cell surface–resident WT and E297G BSEP. MDCK II cells expressing WT and E297G BSEP were treated for 24 hours, with or without 1 mM 4PBA, before the experiments. After cell surface biotinylation, cells were incubated for the indicated time at 37°C, with or without 1 mM 4PBA, as described in Materials and Methods. Remaining biotinylated proteins isolated with streptavidin beads were separated by 6% SDS-PAGE and subjected to western blot analysis. (B) Quantification of band density indicating WT and E297G BSEP in (A). The intensity of the band indicating WT and E297G BSEP was quantified by Image Gauge software and expressed as a percentage of the BSEP present at 0 hours, respectively. Closed and open circles represent remaining cell surface WT BSEP or E297G BSEP in MDCK II cells treated, with and without 4PBA, respectively. Each bar represents the mean ± SE of triplicate determinations. By the same methods as WT and E297G BSEP, the densitometric analysis was also performed for exogenously expressing P-gp and endogenously expressing DPPIV. (C) Cell surface expression of D482G BSEP at a low temperature. MDCK II cells expressing D482G BSEP were cultured for 24 hours at 27°C before the cell surface biotinylation. Biotinylated proteins were isolated with streptavidin beads as described in Materials and Methods. Prepared specimens were separated via 6% SDS-PAGE and subjected to the western blot analysis. (D) The degradation rate of cell surface–resident D482G BSEP. MDCK II cells expressing D482G BSEP were cultured at 27°C for 24 hours, with or without 1 mM 4PBA, before the experiments. The specimens prepared by the same methods as (A) were separated by 6% SDS-PAGE and subjected to western blot analysis. (E) Quantification of band density indicating D482G BSEP in (D). The intensity of the band indicating D482G BSEP was quantified by Image Gauge software and expressed as a percentage of the BSEP present at 0 hours, respectively. Closed and open circles represent remaining cell surface D482G BSEP in MDCK II cells treated, with and without 4PBA, respectively. Each bar represents the mean ± SE of triplicate determinations.

4PBA-Mediated Up-Regulation of BSEP Expression at the Canalicular Membrane Under In Vivo Conditions.

4PBA treatment can increase the functional cell surface expression of WT, E297G, and D482G BSEP in MDCK II cells (Fig. 2). Therefore, we examined whether 4PBA treatment has the same effect under in vivo conditions using SD rats. The dosage of 4PBA in SD rats was determined according to the approved dosage in children (0.45-0.6 g/kg/d). In a [3H]TC infusion study after repeated treatment with 0.6 g/kg/day 4PBA for 10 days, no significant differences were observed in Css, plasma and Vss, bile of [3H]TC between vehicle-treated and 4PBA-treated SD rats (Tables 1 and 2), whereas Css, liver of [3H]TC in 4PBA-treated rats was 2.0-fold lower than that in vehicle-treated rats (Tables 1 and 2). Consequently, CLtotal and CLbile, plasma for [3H]TC were not significantly different between vehicle-treated and 4PBA-treated SD rats. Conversely, CLbile, liver for [3H]TC in 4PBA-treated rats was 2.0-fold higher than that in vehicle-treated rats (Tables 1 and 2). These results suggest that the biliary excretion of [3H]TC via the canalicular membrane is enhanced.

Table 1. Pharmacokinetic Parameters of [3H]TC During Constant Infusion into 4PBA- or Vehicle-Treated SD Ratsat 0.6 g/kg/day for Indicated Time Periods
 Vehicle (n = 9)4PBA (5 days) (n = 6)4PBA (10 days) (n = 9)4PBA (15 days) (n = 3)
  • NOTE. Data represent the mean ± SE.

  • Significantly different from vehicle-treated SD rats by Student's t test (

  • *

    P < 0.05;

  • P < 0.01;

  • P < 0.001).

Css, plasma (ng/ml)2.0 ± 0.22.3 ± 0.32.0 ± 0.22.1 ± 0.2
Css, liver (ng/ml)5.1 ± 0.63.1 ± 0.52.6 ± 0.52.3 ± 0.2
Vss, bile (ng/min/kg)64.4 ± 2.164.9 ± 7.867.9 ± 6.868.9 ± 2.4
CLtotal (ml/min/kg)29.4 ± 3.826.6 ± 3.830.7 ± 3.727.5 ± 2.7
CLbile, plasma (ml/min/kg)31.4 ± 4.232.3 ± 8.532.3 ± 2.933.2 ± 2.0
CLbile, liver (ml/min/kg)14.2 ± 3.021.7 ± 3.0*29.3 ± 6.129.9 ± 1.5
Table 2. Pharmacokinetic Parameters of [3H]TC During Constant Infusion into 4PBA- or Vehicle-Treated SD Rats at Indicated Dose for 10 Days
 Vehicle (n = 9)4PBA (0.2g/kg/day) (n = 3)4PBA (0.6g/kg/day) (n = 9)4PBA (2.4g/kg/day) (n = 3)
  • NOTE. Data represent the mean ± S.E.

  • Significantly different from vehicle-treated SD rats by Student's t-test (

  • *

    P < 0.001).

Css, plasma (ng/ml)2.0 ± 0.22.6 ± 0.22.0 ± 0.23.3 ± 0.7
Css, liver (ng/ml)5.1 ± 0.63.8 ± 0.72.6 ± 0.5*5.9 ± 1.8
Vss, bile (ng/min/kg)64.4 ± 2.166.5 ± 1.667.9 ± 6.873.3 ± 6.4
CLtotal (ml/min/kg)29.4 ± 3.826.6 ± 3.830.7 ± 3.729.6 ± 3.5
CLbile, plasma(ml/min/kg)31.4 ± 4.226.3 ± 2.032.3 ± 2.923.9 ± 4.7
CLbile, liver (ml/min/kg)14.2 ± 3.019.0 ± 3.929.3 ± 6.1*19.2 ± 3.2

To investigate the optimal 4PBA treatment conditions, we also performed a [3H]TC infusion study for various periods and at various doses (Tables 1 and 2). The optimal conditions were 0.6 g/kg/day 4PBA for 10 or 15 days. No apparent side effects were observed under all tested conditions.

To examine whether the increased BSEP expression at the canalicular membrane is associated with the enhanced biliary excretion of [3H]TC via the canalicular membrane, we performed vesicle uptake studies and western blot analysis using CMVs. CMVs were prepared from SD rats after repeated treatment with 0.6 g/kg/d 4PBA or vehicle for 10 days. The ATP-dependent uptake of [3H]TC by CMVs from 4PBA-treated rats was linear up to 0.5 minutes and was 3.3-fold higher at 0.5 minutes than that from vehicle-treated rats in the linear range of uptake (Fig. 6A). A kinetic analysis revealed that the initial ATP-dependent uptake of [3H]TC into CMVs from 4PBA-treated and vehicle-treated rats could be described by a single saturable component with apparent Km values of 13.5 ± 2.4 and 9.5 ± 1.0 μM, respectively (Fig. 6B). These Km values of [3H]TC were consistent with previously reported values.34 The Vmax value of [3H]TC was increased from 270 ± 20 to 1,000 ± 110 pmol/min/mg protein by 4PBA treatment (Fig. 6B). Furthermore, western blot analysis with CMVs demonstrated that the Bsep expression level was increased 3.2-fold by 4PBA treatment, whereas P-gp and DPPIV expression levels were not affected (Fig. 6C). The change in the Bsep expression level seems close to the change in its Vmax values for [3H]TC. These results suggest that 4PBA treatment with clinically relevant dosage in humans can increase functional Bsep expression at the canalicular membrane, and consequently, enhance bile acid transport via the canalicular membrane in SD rats.

Figure 6.

The effect of 4PBA treatment on BSEP expression at the canalicular membrane in SD rats. Male SD rats (6-7 weeks old) were given 4PBA or vehicle by gavage for 10 days and at 0.6 g/kg/d before the experiments. Either 4PBA or vehicle was given as 3 divided doses. (A) Time-dependent uptake of [3H]TC by CMVs. CMVs prepared from 4PBA- (black circles) and vehicle- (white circles) treated SD rats were incubated at 37°C with 5 mmol/L ATP or AMP. The concentration of [3H]TC was 1 μM. The uptake of [3H]TC was obtained by subtracting the value in the presence of AMP from that in the presence of ATP. Each point and vertical bar represents the mean ± SE of triplicate determinations. (B) Eadie-Hofstee plot of the uptake of [3H]TC by CMVs. CMVs prepared from 4PBA- (black circle) and vehicle- (white circle) treated SD rats were incubated at 37°C for 0.5 minute and 1 minute, respectively, with 5 mmol/L ATP or AMP. The concentration of [3H]TC was 1 μM. The uptake of [3H]TC was obtained by subtracting the value in the presence of AMP from that in the presence of ATP. Each point and vertical/horizontal bar represents the mean ± SE of triplicate determinations. (C) Results of western blot analysis with CMVs. CMVs prepared from 4PBA- and vehicle-treated SD rats (10 μg) were separated by 6% SDS-PAGE and subjected to western blot analysis.

Discussion

We previously reported that E297G and D482G BSEP, which are frequently found mutants in European patients, result in impaired membrane trafficking, whereas the transport function of both mutated BSEPs remains almost normal.10 Restoration of the reduced cell surface expression of these mutated BSEPs is a therapeutic goal for PFIC2 patients with E297G and D482G mutations, because both mutated BSEPs per se exhibit normal transport functions. In the current study, we analyzed the potential therapeutic effect of 4PBA against PFIC2 by examining the effect of 4PBA treatment on the cell surface expression of E297G and D482G BSEP. In addition, we investigated the in vivo effect of 4PBA in SD rats.

Initially, we examined the effect of 4PBA treatment on WT, E297G, and D482G BSEP using MDCK II cells. The transcellular transport assays showed that 4PBA treatment significantly increases the PSapical, BSEP in MDCK II cells expressing WT BSEP as well as both mutated BSEPs (Fig. 2B). Together with the finding that the increase in PSapical, BSEP by 4PBA treatment correlates with the cell surface expression of WT, E297G, and D482G BSEP (Fig. 2C), 4PBA treatment with a clinically achievable concentration can enhance the cell surface expression and the transport capacity of WT, E297G, and D482G BSEP in MDCK II cells.

Next, we examined the underlying mechanism of 4PBA action by which the cell surface expression of WT and both mutated BSEPs is increased. Quantitative PCR demonstrated that BSEP mRNA levels were hardly changed by 4PBA treatment (Fig. 3A). The effect of 4PBA treatment was not affected by the inhibition of transcription with Act D and translation with CHX (Fig. 3B,C). Furthermore, the expression levels of the immature ER-resident form of WT, E297G, and D482G BSEP were almost identical in non– and 4PBA-treated MDCK II cells just after BFA washout (Fig. 4A; 0 hours). These results indicate that a post-translational mechanism mainly contributes to the 4PBA-mediated increase in the cell surface expression of WT, E297G, and D482G BSEP.

Although we investigated whether 4PBA treatment promotes the maturation of WT, E297G, and D482G BSEP as the possible post-translational mechanism, such an effect of 4PBA treatment was not observed in the BFA washout study (Fig. 4A,B). However, the results of this study showed that 4PBA treatment could stabilize the mature form of WT, E297G, and D482G BSEP (Fig. 4A,B). This finding is consistent with the results of the biotin labeling method in which 4PBA treatment prolonged the half-life of the cell surface–resident WT, E297G, and D482G BSEP (Fig. 5A-E). Because cell surface–resident BSEP constitutively cycles between the canalicular membrane and the intracellular compartment, and is finally removed from this cycle to the degradation pathway,30, 31 4PBA treatment may interrupt the internalization process or promote the recycling process and, consequently, increase the cell surface expression of WT, E297G, and D482G BSEP. The stabilization of cell surface–resident mature protein is a novel effect of 4PBA. Although the molecular mechanism of this effect remains unknown, taking into the consideration that the prolonged half-life with 4PBA treatment was only detected in cell surface–resident WT, E297G, and D482G BSEP, but not for other membrane proteins such as P-gp and DPPIV (Fig. 5B,E), 4PBA treatment possibly induces a specific interaction of WT, E297G, and D482G BSEP with adaptor proteins.

We also examined the effect of 4PBA treatment on increased Bsep expression at the canalicular membrane in vivo by using SD rats. The [3H]TC infusion study and vesicle study using CMVs demonstrated that 4PBA treatment results in an increase in the biliary excretion of [3H]TC via the canalicular membrane (Fig. 6A,B; Tables 1 and 2). Furthermore, western blot analysis with CMVs indicated that this is attributable to an increase in Bsep expression by 4PBA treatment (Fig. 6C). Therefore, as observed in vitro, 4PBA treatment also can increase functional Bsep expression at canalicular membrane, and consequently, increase bile acid transport via the canalicular membrane in SD rats. If the prolonged half-life of cell surface–resident Bsep with 4PBA treatment is responsible for this result in vivo as well as in vitro, the protective effect of 4PBA treatment against hepatocellular damage by increasing biliary excretion of bile acids and lowering intrahepatic bile acids may be remarkably observed in PFIC2 patients with E297G and D482G mutations. This can be implied from the result of the in vitro study showing that 4PBA treatment has a greater effect on the increase in the functional cell surface expression of mutated BSEPs compared with WT BSEP (Fig. 2B,C). Whether the action of 4PBA in vivo is caused by the same mechanism as in vitro remains unknown. 4PBA treatment also may have modified the post-translational process of Bsep in vivo, because the Bsep mRNA expression level in the liver was similar in vehicle- and 4PBA-fed SD rats (data not shown). The fact that chronic treatment is required for exerting the action of 4PBA in the liver also supports this hypothesis. The half-life of the mature form of Bsep in vivo and in MDCK II cells is 4 days and 6 hours, respectively.31, 35 If the prolonged half-life of cell surface–resident Bsep is involved in the effect of 4PBA, chronic treatment will be needed to increase the functional cell surface expression of Bsep in vivo. A long-term treatment was actually needed in SD rats as opposed to a short-term treatment being effective in vitro (Tables 1 and 2; Fig. 1A).

The BSEP expression level at the canalicular membrane is also reduced in not only PFIC2, but also in several cholestatic models, examples of which are endotoxin- or sepsis-induced cholestasis,36–38 and cholestasis in pregnancy.37, 39 In these cholestatic conditions, it seems that BSEP is internalized from the canalicular membrane to the pericanalicular intracellular compartment followed by degradation. Although whether the reduction in BSEP expression at the canalicular membrane is the direct cause or the consequent result of these diseases remains unknown, the accumulation of bile acids in hepatocyte caused by impaired BSEP function leads to a deterioration of these diseases. A possible mechanism that 4PBA treatment increases the cell surface–resident WT BSEP is the interruption of the internalization process from cell surface to intracellular compartment or promotion of the recycling from intracellular compartment back to cell surface. Because this mechanism is a combating mechanism for these diseases, this agent may be a potential candidate to treat or to inhibit the progression of these diseases.

In conclusion, we demonstrated that 4PBA treatment with a clinically achievable concentration induces the cell surface expression and the transport capacity of WT, E297G, and D482G BSEP in MDCK II cells and also enhances functional Bsep expression at the canalicuar membrane and bile acid transport via canalicular membrane in vivo. The prolonged half-life of the cell surface–resident BSEP by 4PBA treatment was suggested to be responsible for the results at least in vitro. These results suggest that 4PBA is a potential pharmacological agent not only for PFIC2 patients with E297G and D482G mutations but also for other cholestatic patients, in whom the BSEP expression at the canalicular membrane is reduced. Clinical trials will be required to determine the utility and safety of 4PBA as a therapy for these cholestatic diseases.

Acknowledgements

We thank Dr. Hiroyuki Kusuhara and Ms. Misaki Kuroda for advice on the manuscript.

Ancillary