If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Corresponding author. , Department of Gastroenterology, Hepatology and Infectious Diseases, Medical Faculty of Otto-von-Guericke University Magdeburg, University Hospital Magdeburg AöR, Leipziger Str. 44, 39120 Magdeburg, Germany, Fax: +49 391 6713105. Phone: +49 391 6713100. .
Footnotes
# contributed equally
Affiliations
Department of Gastroenterology, Hepatology and Infectious Diseases, Medical Faculty, University Hospital Düsseldorf, Heinrich Heine University Düsseldorf, Germany
Department of Gastroenterology, Hepatology and Infectious Diseases, Medical Faculty, University Hospital Düsseldorf, Heinrich Heine University Düsseldorf, GermanyDepartment of Gastroenterology, Hepatology and Infectious Diseases, University Hospital Magdeburg, Medical Faculty, Otto-von-Guericke University, Magdeburg, Germany
Pediatric Gastroenterology and Hepatology, Department for Pediatric Kidney, Liver and Metabolic Diseases, Hannover Medical School, Hannover, GermanyDepartment of Human Genetics, Hannover Medical School, Hannover, Germany
Department of Gastroenterology, Hepatology and Infectious Diseases, Medical Faculty, University Hospital Düsseldorf, Heinrich Heine University Düsseldorf, Germany
Department of Gastroenterology, Hepatology and Infectious Diseases, Medical Faculty, University Hospital Düsseldorf, Heinrich Heine University Düsseldorf, Germany
Department of Gastroenterology, Hepatology and Infectious Diseases, University Hospital Magdeburg, Medical Faculty, Otto-von-Guericke University, Magdeburg, Germany
Department of Gastroenterology, Hepatology and Infectious Diseases, Medical Faculty, University Hospital Düsseldorf, Heinrich Heine University Düsseldorf, Germany
Department of General Pediatrics, Neonatology and Pediatric Cardiology, Medical Faculty, University Hospital Düsseldorf, Heinrich Heine University, Düsseldorf, Germany
Department of General Pediatrics, Neonatology and Pediatric Cardiology, Medical Faculty, University Hospital Düsseldorf, Heinrich Heine University, Düsseldorf, Germany
Department of Otorhinolaryngology, Heinrich Heine University School of Medicine, Düsseldorf, GermanyDepartment of Pediatrics III, University Children’s Hospital, University of Duisburg-Essen, Essen, Germany
Department of Gastroenterology, Hepatology and Infectious Diseases, Medical Faculty, University Hospital Düsseldorf, Heinrich Heine University Düsseldorf, Germany
Corresponding author. , Department of Gastroenterology, Hepatology and Infectious Diseases, Medical Faculty, University Hospital Düsseldorf, Heinrich Heine University, Moorenstr. 5, 40225 Düsseldorf, Germany. Phone: +49 211 8113509.
Department of Gastroenterology, Hepatology and Infectious Diseases, Medical Faculty, University Hospital Düsseldorf, Heinrich Heine University Düsseldorf, GermanyDepartment of Gastroenterology, Hepatology and Infectious Diseases, University Hospital Magdeburg, Medical Faculty, Otto-von-Guericke University, Magdeburg, Germany
Development of a functional diagnostic assay for antibody-induced BSEP deficiency (AIBD)
•
Confirmation of BSEP inhibition by serum antibodies corroborates AIBD diagnosis
•
Updated diagnostic workflow for AIBD confirming presence of inhibitory anti-BSEP antibodies
•
Monitoring of onset and treatment response of AIBD using this functional test
Abstract
Background&Aims
Antibody-induced bile salt export pump deficiency (AIBD) is an acquired form of intrahepatic cholestasis, which may develop following orthotopic liver transplantation (OLT) for progressive familial intrahepatic cholestasis type 2 (PFIC-2). Approximately 8-33% of transplanted PFIC-2 patients develop bile salt export pump (BSEP) antibodies which trans-inhibit this bile salt transporter from the extracellular, biliary side. AIBD is diagnosed by demonstration of BSEP-reactive and -inhibitory antibodies in patient serum. We developed a cell-based test directly measuring BSEP trans-inhibition by antibodies in serum samples to confirm AIBD diagnosis.
Methods
Sera from healthy controls and cholestatic Non-AIBD or AIBD cases were tested I) for anti-canalicular reactivity by immunofluorescence (IF) staining of human liver cryosections, II) for anti-BSEP reactivity by IF staining of HEK293 cells expressing BSEP-EYFP and immunodetection of BSEP-EYFP on Western blot, and III) for BSEP trans-inhibition using HEK293 cells stably expressing NTCP-mCherry and BSEP-EYFP. The trans-inhibition test uses [3H]-taurocholate as substrate and is divided into an uptake phase dominated by NTCP followed by BSEP-mediated export. For functional analysis, sera were bile salt depleted.
Results
We found BSEP trans-inhibition by seven sera containing anti-BSEP antibodies, but not by five cholestatic or nine control sera, all lacking BSEP reactivity. Prospective screening of a post-OLT PFIC-2 patient showed seroconversion to AIBD and the novel test method allowed monitoring of treatment response. Notably, we identified a post-OLT PFIC-2 patient with anti-BSEP antibodies yet without BSEP trans-inhibition activity, in line with asymptomatic presentation at serum sampling.
Conclusions
Our cell-based assay is the first direct functional test for AIBD and allows confirmation of diagnosis as well as monitoring under therapy. We propose an updated workflow for AIBD diagnosis including this functional assay.
Lay Summary
Antibody-induced BSEP deficiency (AIBD) is a potentially serious complication that may affect PFIC-2 patients after liver transplantation. To improve its early diagnosis and thus immediate treatment, we developed a novel functional assay to confirm AIBD diagnosis using patient’s serum and propose an updated diagnostic algorithm for AIBD.
E.L. has received honoraria for lectures from Albireo and Mirum and has served on advisory boards for both companies and on the GPGE board of directors. U.B. has received grants from Mirum Pharma, Albireo Pharma, Alexion Pharma, and has received consulting fees and honoraria from Mirum, Albireo, Alexion, Nestle, Vivet. C.K. has received travel grant from Falk. T.B. has received grants, consulting fees, and/or honoraria from Abbvie, Alexion, Bayer, BMS, Gilead, GSK, Eisai, Enyo Pharma, Falk Foundation, HepaRegeniX GmbH, Humedics, Intercept, Ipsen, Janssen, Medupdate GmbH, MSD/Merck, Merz, Norgine, Novartis, Orphalan, Roche, Sequana Medical, SIRTEX, SOBI, and Shionogi. V.K. has received honoraria for speaker’s bureau from Falk, Albireo, CSL, Sanofi, Gilead, Abbvie, Intercept and has served on one advisory board for Astra Zeneca. All other authors declare that there is no potential conflict of interest.
Financial support statement
This study was supported by the BMBF through HiChol (01GM1904A to V.K. and C.D.; 01GM1904B to U.B., A.S., E.P.).
The NTCP-, BSEP- and NTCP-/BSEP-expressing HEK293 cell lines and all datasets generated during the current study are available from the corresponding authors upon request.
Introduction
Severe BSEP deficiency, also termed Progressive familial intrahepatic cholestasis type 2 (PFIC-2) is caused by inherited pathogenic variants in the ABCB11 gene encoding the canalicular bile salt export pump (BSEP).
BSEP is an ATP binding cassette (ABC) transporter exclusively expressed in hepatocytes, which utilizes the energy obtained from ATP hydrolysis to export bile salts (BS) against their concentration gradient from hepatocytes into the canalicular lumen.
ABCB11 variants resulting in reduced or absent functional BSEP expression at the canalicular membrane lead to retention of bile salts (BS) in the hepatocyte (Fig. 1A, left panel).
Serum BS levels are drastically increased, and the overall clinical presentation is that of a low-gamma glutamyl transferase (gGT) intrahepatic cholestasis.
Left untreated, PFIC-2 may be associated by severe pruritus and progress to liver cirrhosis and hepatocellular carcinoma, necessitating orthotopic liver transplantation (OLT) of a BSEP-competent donor organ, which restores hepatic BS excretion and enterohepatic BS circulation.
ATP8B1 and ABCB11 analysis in 62 children with normal gamma-glutamyl transferase progressive familial intrahepatic cholestasis (PFIC): phenotypic differences between PFIC1 and PFIC2 and natural history.
Fig. 1Development of a cell-based BSEP trans-inhibition assay for AIBD diagnosis. A, left panel Maintaining the enterohepatic circulation of bile salts (BS) is an essential function of the liver. Severe BSEP deficiency (PFIC-2) is caused by reduced or absent BSEP expression and results in disruption of the enterohepatic BS circulation and accumulation of BS within hepatocytes and the circulation. Right panel Functional BSEP expression in the transplant restores enterohepatic BS circulation. Some patients, however, show a recurrence of symptoms due to development of BSEP-inhibitory antibodies, termed antibody-induced BSEP deficiency (AIBD). After entering the canalicular space either by a paracellular route or by transcytosis, they may bind and can trans-inhibit BSEP. B Since vectorial BS transport is not necessary in order to recapitulate BSEP trans-inhibition, unpolarized cells may be used as long as BSEP activity can be separated from NTCP activity. C HEK293 cells were transduced with recombinant lentiviral vectors for constitutive expression of either human NTCP-mCherry, BSEP-EYFP, or both transporters (also see Fig. S1A). Confocal microscopy images of the established cell lines. Nuclei were stained with DAPI. Bar=10 μm. D Assay principle. Cells are incubated with [3H]-taurocholate (TC) during an import phase dominated by NTCP (also see Fig. S2A) which is followed by an export phase mediated by BSEP. During export, re-uptake of exported [3H]-TC by sodium-dependent NTCP is prevented by replacing extracellular sodium with choline. Radioactivity contained in supernatant S1, supernatant S2, and cell lysate (L) is measured by liquid scintillation counting. E Proof of assay principle using empty control (HEK293; n=7), BSEP (n=6), NTCP (n=7), and NTCP/BSEP (n=7) cells. Data are shown as mean ± SD. Significances were calculated using the two-sided, unpaired Student´s t-test.
In 2009, we and others described a novel, acquired form of intrahepatic cholestasis, which results from development of an immunoglobulin gamma (IgG) anti-BSEP response after successful OLT for PFIC-2 and was named antibody-induced BSEP deficiency (AIBD) (Fig. 1A, right panel).
PFIC-2 patients without any residual BSEP expression in their native liver are at increased risk of developing AIBD, since they congenitally lack immunological BSEP tolerance.
The clinical presentation of AIBD mimics PFIC-2 and is caused by antibody-mediated BSEP inhibition from the extracellular side in the canalicular lumen (trans-inhibition; Fig. 1A).
In the majority of reported cases, this phenotypic disease recurrence was at least temporarily manageable by either intensified immunosuppressive therapy, antibody removal via immunoadsorption or plasmapheresis and/or depletion of antibody-producing B cells by rituximab treatment.
Notably, an AIBD patient refractory to these therapeutic approaches was cured by allogeneic hematopoietic stem cell transplantation as an individualized therapeutic salvage approach.
Allogeneic haematopoietic stem cell transplantation eliminates alloreactive inhibitory antibodies after liver transplantation for bile salt export pump deficiency.
based on I) post-OLT recurrence of a low-gGT intrahepatic cholestasis phenotype in PFIC-2 patients accompanied by II) the presence of anti-canalicular antibodies in the patient´s serum, III) which are directed against BSEP.
their detection requires a liver biopsy for immunostaining and is thus not regularly used for diagnosis. We could previously demonstrate that preincubation of BSEP-containing inside-out vesicles (IOV) with purified IgG from AIBD patient sera resulted in BSEP transport inhibition.
Because the intravesicular side (corresponding to the biliary side) of the IOV membrane is not directly accessible to the antibodies, IgG need to be included into the IOVs in a time-consuming freeze-thaw procedure
To facilitate a more direct functional readout, we developed a simplified, cell-based assay which permits direct measurement of BSEP trans-inhibition by serum antibodies, the pathophysiological cause of AIBD.
Additional method descriptions can be found in the online supplementary information.
Patient material
This study was performed according to the guidelines of the declaration of Helsinki. All serum donors gave written consent for the investigation (approved by the local ethics committee study number 5350). AIBD sera were obtained from patients for diagnostic workup of which seven were included in this study (one female, six male). Sera of nine healthy individuals (five female, four male) and of five patients presenting with cholestasis and increased serum BS but lacking anti-BSEP antibodies (Non-AIBD, three female, two male) were used as reference samples throughout this study. All sera were stored at -80 °C. The non-tumorous liver metastasis resection margin of a patient with colorectal cancer was used as control liver tissue for detecting anti-canalicular reactivity in serum samples.
Generation of stable HEK293 cell lines expressing human NTCP and BSEP
Preparation of lentiviral particles and subsequent transduction was carried out as described.
Briefly, lentiviral particles were generated in HEK293T cells using either puc2CL6EGIP-NTCP-mCherry or -BSEP-EYFP vector plasmid and separately used to transduce HEK293 cells (Fig. S1A). After expansion, cells were sorted by EYFP-(BSEP) or mCherry-(NTCP) fluorescence on a FACS-Aria-III cell sorter (Becton Dickinson, Heidelberg, Germany), followed by clonal separation by dilution into hybridoma dishes. Suitable clones for both NTCP-mCherry and BSEP-EYFP were picked, and activity of both transporters was confirmed by cellular [3H]-taurocholate (TC) uptake for NTCP-mCherry (see results) and a vesicular transport assay for BSEP-EYFP (Figs. S1B and C). Finally, the monoclonal HEK293-BSEP-EYFP cell line was transduced with NTCP-mCherry before sorting and subsequent clonal isolation in hybridoma dishes.
Bile salt depletion of serum samples and bile salt analysis
All steps were carried out at room temperature (RT). 200 μL serum was diluted tenfold with PBS before concentration back to 200 μL in a spin concentrator with a molecular weight cutoff of 10 kDa (Amicon Ultra 2 mL, Merck, Darmstadt, Germany). This dilution/concentration was repeated four times. After the last concentration, the volume was adjusted to exactly 200 μL with PBS. Extraction and analysis of BS levels in original and BS-depleted samples was carried out by liquid chromatography-tandem mass spectrometry (LC-MS/MS) as described.
Peak fractions were pooled, buffer-exchanged to PBS to the volume of the original serum sample using 10 kDa cutoff spin concentrators. NTCP/BSEP cells were preincubated with comparable amounts of total IgG as described below (∼325 μg per well).
Cell-based BSEP trans-inhibition test
NTCP/BSEP cells were kept under Puromycin selection and were passaged every 4-5 days in high glucose DMEM (Biochrom, Berlin, Germany). At near-confluence, cells were trypsinized and resuspended in 10 mL medium (per 75 cm
culture surface), diluted twofold in medium, and seeded onto poly-L-lysine- (PLL, Sigma-Aldrich, Heidelberg, Germany) coated 12-well plates at 1 mL per well. After overnight cultivation, medium on the confluent monolayer was changed to 470 μL of 25 mM HEPES/TRIS (pH7.4) buffered low glucose DMEM. 30min later, 25 μL of either BS-depleted sera, purified IgG, or PBS were added per well and plates incubated for 1 h. For complement inactivation, BS-depleted sera were incubated for 30min at 56 °C before being assayed. All samples were assayed in triplicates. Transport was started by addition of 5 μL [3H]-TC (freshly diluted into ddH2O, final concentration 20 nM per well). Plates were kept at 37 °C for 3min, after which supernatants (supernatant 1, S1) were transferred to reaction tubes. Cells were washed three times with 500 μL ice-cold-PBS. After addition of 500 μL prewarmed export buffer (20 mM HEPES/TRIS pH7.4, 1.8 mM CaCl2, 1.2 mM KH2PO4, 1.2 mM MgSO4, 5 mM KCl, 144 mM choline chloride), plates were kept at 37 °C for 3min, put on ice, and supernatants (S2) were transferred to fresh reaction tubes. Cells were washed as above and lysed in 500 μL PBS with 0.5% (w/v) SDS and 1μL/mL Benzonase (Merck, Darmstadt, Germany) to reduce sample viscosity. Each well yields three samples: supernatant S1 (= import phase), supernatant S2 (export phase), and cell lysate L (intracellular BS retention). 400 μL of each sample was mixed with 4 mL of liquid scintillation cocktail (Ultima Gold™, BD Biosciences, Heidelberg, Germany) by vigorous vortexing for 20sec and then measured on a liquid scintillation counter (Packard Instruments, Frankfurt, Germany). The sum of averaged counts per minute (cpm) for S1, S2, and L (triplicates) was set to 100% for the HEK293 cell line in Fig. 1E to visualize the distribution of total radioactivity between S1, S2, and L during the test. All other cpm data for transduced cell lines was transformed into percent of HEK293 cells. All serum-based BSEP trans-inhibition test values are expressed in percent of cpm (S1 + S2 + L) of the respective no serum measurements. All graphs were prepared using Graphpad Prism 5.0a (Graphpad Software, San Diego, USA). S2 and L data from sera AIBD7-7.3 and from the no serum controls were used to calculate the thresholds (L/S2 as mean±SD; max. Inhibition, 3.28±0.34; no inhibition, 0.36±0.12) of maximum and no inhibition, respectively, which are shown in Fig. 3 C and E.
Fig. 3Disease onset and response to treatment in AIBD patient 7. A IHC staining of a normal (left) and the patient’s liver (right) for BSEP. Bar = 50 μm. B BSEP reactivity of serum samples on Western blots using plasma membrane preparations from HEK293 cells (∅) and BSEP-EYFP-expressing cells (BSEP). C BSEP trans-inhibition by serum antibodies was plotted as L/S2 (both averages from triplicate measurements). Thresholds for maximum (AIBD7-7.3) and no inhibition (no serum) are indicated in red and green, respectively. Ritux, rituximab; PP, plasmapheresis. D Detection of canalicular IgG deposits (left, red) alongside normal BSEP expression (right, red) in transplant biopsy by IF staining. MRP2 (green) served as canalicular marker. Bar = 10 μm. E Changes in BSEP trans-inhibition in response to multiple sessions of immunoadsorption (n=4) and plasmapheresis (n=8). Significances were calculated using Wilcoxon´s matched-pairs test (one-sided).
Briefly, methanol-fixed cryosections of non-cholestatic human liver were immunostained with human serum samples at 1:100 for anti-canalicular reactivity and a murine monoclonal antibody against multidrug resistance-associated protein 2 (MRP2; M2I-4; Alexis, Grünberg, Germany) as a canalicular marker (1:25). Goat anti-human-IgG-Cy3 (Jackson ImmunoResearch Laboratories, West Grove, PA) and goat anti-mouse-IgG-Alexa Fluor 488 (Invitrogen, Karlsruhe, Germany) served as secondary antibodies (both at 1:500). All images were acquired via the ZEN software (black edition, Ver.2.3 SP1) on a LSM 880 confocal laser scanning microscope (Zeiss, Jena, Germany) using Immersol 518 F (Zeiss) oil on a Zeiss Plan-APOCHROMAT 63x Oil/DIC objective at RT.
Immunofluorescence staining of HEK293 cells
For staining with sera, empty HEK293 and BSEP cells were trypsinized and mixed at a ratio of 1:5 before seeding onto coverslips in 12-well plates and overnight cultivation. Cells on coverslips were washed with PBS, fixed with ice-cold methanol for 30sec, and blocked for 30min (UltraVision Block, Thermo, Germany). Cells were stained with patient sera (1:50) for 1 h followed by goat anti-human-IgG-Cy3 (1:500) and DAPI (1:20,000) for 1 h. Coverslips were mounted on microscopic slides using Dako fluorescence mounting medium (Dako, Waldbronn, Germany).
Live cell immunofluorescence staining
NTCP/BSEP cells were seeded into PLL-coated IBIDI dishes (ibiTreat μ-dish 35 mm, high, ibidi, Gräfelfing, Germany) and incubated overnight to reach confluency. Medium was changed to 25 mM HEPES/TRIS buffered low-glucose DMEM without puromycin. After 30min the medium was aspirated and the edge around the observation area in the dish was carefully dried with a cleaning tissue. The cell layer was then covered with 25 μL of BS-depleted serum or PBS in 500 μL buffered medium to conserve serum (leaving out the dried edge), and dishes were incubated for 1 h. Dishes were put on ice, medium aspirated and the cell layer was gently washed three times with ice-cold PBS containing Ca2+ and Mg2+ before fixation for 30sec with ice-cold methanol. Cells were washed twice with PBS, blocked for 30min (UltraVision Block) and stained for 1 h either for human IgG (1:500; Goat anti-human-IgG (H+L) Alexa Fluor 647 Fab fragment, Jackson ImmunoResearch Laboratories) or Na+/K+-ATPase (1:200; clone M7-PB-E9, Sigma-Aldrich) followed by goat anti mouse-IgG1 Alexa Fluor 647 IgG for 1 h (1:500; Jackson ImmunoResearch Laboratories). All secondary antibody dilutions contained DAPI at 1:20,000. Cells were gently washed three times with PBS and once with H2O before mounted with a 18 mm round coverslip using DAKO mounting medium.
Western blot analysis of patient sera
Membrane vesicle preparations (10 μg protein per lane) from HEK293 and BSEP-EYFP cells were separated by SDS-PAGE and transferred onto nitrocellulose membranes. Blot membranes were cut into segments and blocked for 1 h in TBS-T containing 10% (w/v) non-fat dried milk. Segments were probed with human serum samples (1:1,000 in block) for 1 h followed by goat anti-human IgG-HRP conjugate (1:10,000 in TBS-T) for 1 h before detection using Western Lightning Plus chemiluminescent substrate (Perkin Elmer, Rodgau, Germany) in a ChemiDoc MP Imaging System (Biorad, Feldkirchen, Germany).
Results
Development of a cell-based BSEP trans-inhibition assay for AIBD diagnosis
To confirm AIBD diagnosis, a suitable transport inhibition assay needs to recapitulate BSEP trans-inhibition. This could either be achieved by use of a polarized cell line allowing spatial separation of NTCP-mediated uptake and BSEP-mediated export (Fig. 1B), or by use of an assay strategy which allows temporal separation of NTCP from BSEP activity. The latter was achieved by transduction of HEK293 cells with recombinant lentiviral vectors allowing constitutive expression of both human NTCP-mCherry and BSEP-EYFP (Fig. 1C, Fig. S1A). Cells transduced with either transporter alone served as references [3H]-taurocholate (TC) served as a model BS substrate for both transporters. First, cells were incubated with 20 nM [3H]-TC for 3min (Fig. 1D). After incubation, remaining radioactivity in supernatant S1 was measured by liquid scintillation counting. Lacking any endogenous BS uptake system, both HEK293 (grey in Fig.1E) and BSEP cells (green) failed to show any [3H]-TC import from the supernatant into the cytosol during the first incubation step. Thus, all [3H]-TC remained in S1 (Fig. 1E). In contrast, NTCP cells (red) showed significant [3H]-TC import as indicated by reduced radioactivity in S1. In the double-transduced NTCP/BSEP cell line (orange) [3H]-TC was detectable in S1 to similar levels as observed for the NTCP cells (Fig. 1E), demonstrating a higher capacity for NTCP import than BSEP export. This both indicates successful [3H]-TC loading of cells via NTCP and necessitates subsequent NTCP inactivation to measure BSEP export. Accordingly, after the first incubation phase cells were washed and incubated in export buffer containing choline instead of sodium. Since TC uptake by NTCP depends on sodium symport, removal of extracellular sodium renders NTCP inactive. After 3min export, the second supernatant (S2) was obtained, cells washed and lysed (L). Without [3H]-TC import into HEK293 and BSEP cells there was no detectable radioactivity in the supernatant after the BSEP export phase (S2) or in their cell lysate (L) (Fig. 1E). In contrast, NTCP cells showed only minimal amounts in S2, indicating lack of endogenous [3H]-TC export and thus substantial intracellular [3H]-TC accumulation (L). In NTCP/BSEP cells, the amount of [3H]-TC was significantly higher in S2 and significantly lower in L as compared to NTCP cells. This demonstrates a significant, BSEP-mediated [3H]-TC export in the NTCP/BSEP cells. Taken together, redistribution of [3H]-TC between S1, S2 und L during both incubation (= import and export) phases clearly demonstrated NTCP-mediated import followed by BSEP-mediated export in the NTCP/BSEP cells.
Validation of the BSEP trans-inhibition assay using serum samples from different cohorts
To validate the assay, we used serum samples from non-cholestatic healthy controls (n=9), from cholestatic patients without presence of anti-BSEP antibodies (Non-AIBD, n=5), and from patients diagnosed with AIBD (n=7) (Fig. 2). AIBD cases were identified by: I) detection of anti-canalicular antibodies in patient’s serum by immunofluorescence (IF) staining on cryosections of normal human liver tissue (Fig. S2A); II) demonstration of anti-BSEP reactivity by IF staining of BSEP-EYFP-expressing HEK293 cells (Fig. S2B) and III) by western blot detection of BSEP in plasma membrane preparations from BSEP-EYFP cells using the patient’s serum (Fig. 2A, Fig. S2C). Complete diagnostic datasets of each cohort member including individual BSEP trans-inhibition data can be found in Fig. S2 (AIBD), Fig. S3 (Control) and Fig. S4 (Non-AIBD). Of all patients included in this study only AIBD6 and Non-AIBD3 have previously been published.
All identified BSEP variants underlying PFIC-2 in this cohort were either predicted or confirmed to cause absent BSEP expression in native liver (corresponding to type 3 in the NAPPED classification) in all but one case (AIBD1: c.2944G>A = p(Gly982Arg) homozygous).
Fig. 2Validation of the BSEP trans-inhibition assay using serum samples from different cohorts. A Western blot data of representative sera used in the test. Crude plasma membranes were prepared from HEK293 cells (∅) and BSEP-EYFP-expressing cells (BSEP), separated via SDS-PAGE and transferred onto nitrocellulose. Membranes were then probed with indicated sera as primary antibody. B Anti-BSEP antibodies (huIgG) are present in BSEP-reactive (AIBD) but absent from Non-AIBD and Control sera and bind to BSEP-EYFP on the surface of live HEK293 cells. NTCP/BSEP cells seeded onto PLL-treated dishes were incubated with sera, fixed and immunostained for human IgG with an AlexaFluor (AF) 647-coupled secondary antibody (shown in red). Complete sets of original image channels can be found in Fig. S6A. Bar = 20 μm. C Binding of anti-BSEP antibodies to their target on the surface of live cells does not trigger BSEP internalization. NTCP/BSEP cells seeded onto PLL-treated IBIDI dishes were incubated with indicated sera, fixed and immunostained for Na+/K+-ATPase as a plasma membrane marker (using AF 647, channel shown in red (see Fig. S6B)). Bar = 20 μm. D Validation of cell-based BSEP trans-inhibition test using BS-depleted Control (n=9), Non-AIBD (n=5) and AIBD (n=7) sera and untreated cells (no serum, n=15). Data are shown as mean ± SD. Significances were calculated using the two-sided, unpaired Student´s t-test. Radioactivity measurements in S1, S2, L after AIBD preincubation resemble cells expressing only NTCP (inset). E Cell-based BSEP trans-inhibition test using serum IgG. Total IgG was purified from serum Control1 and AIBD6. NTCP/BSEP cells were preincubated with the IgG amount present in the serum volume normally assayed. Data are shown as mean ± SD of triplicate measurements.
Serum samples were first depleted of excess BS which may compete with the radioactive reporter substrate for transport by both NTCP and BSEP during the assay, thus effectively mimicking BSEP inhibition (false positive assay result). To characterize the extent of this substrate competition in NTCP/BSEP cells, increasing amounts of unlabeled TC were added at assay start (Fig. S5A). NTCP-driven [3H]-TC import remained unaffected up to a concentration of 500 μM unlabeled TC (S1). In contrast, reduction of [3H]-TC export by BSEP was already observed between 50 and 100 μM TC (S2), leading to gradually increasing intracellular retention of the radioactive reporter substrate (L). This indicated a higher import than export capacity of this cell line for BS, which corresponds to the physiological situation where BSEP mediates the rate-limiting step of serum-to-bile transport of BS.
Consequently, high serum BS concentrations characteristic of both PFIC-2 and AIBD may lead to misinterpretation of BSEP transport assays (false positive result). Total serum BS (TBS) levels in our AIBD cohort ranged from 153.7 to 512.5 μM with an average concentration of 268.9±119.7 μM (n=5, 2 samples were too small for TBS measurement). In the Non-AIBD cohort, TBS levels ranged from 52.4 to 467.8 μM (203.6±158.7 μM; n=5). Accordingly, BS were depleted from all serum samples by repeated dilution and concentration steps using spin concentrators retaining anti-BSEP antibodies. BS, which are largely bound to serum (lipo)proteins,
, were sufficiently decreased after five rounds of dilution/concentration to 28.9±20.4 μM in AIBD sera (n=5) and to 7.5±5.1 μM in Non-AIBD cholestatic sera (n=5) as determined by mass spectrometric analysis (Fig. S5B). BS in control sera (n=9) were decreased from 8.0±5.6 μM (range 2.4 to 19.6 μM) to 0.9±0.5 μM after depletion. Following BS depletion, BS concentrations in all three cohorts were below the inhibitory concentration of BSEP activity.
Preincubation of live NTCP/BSEP cells with BS-depleted AIBD sera, but not with Non-AIBD or Control sera led to staining of the cell surface with human IgG (Fig. 2B, Fig. S6A). This staining was caused by BSEP-reactive antibodies, since all AIBD sera only stained BSEP-EYFP cells but not empty HEK293 cells (Fig. S2B). To establish whether this antibody decoration triggered BSEP internalization, which would reduce BS export and thus could be misinterpreted as BSEP inhibition (false positive result), Na+/K+-ATPase was stained as a plasma membrane marker (Fig. 2C). Subsequent colocalization analysis demonstrated that extracellular antibody decoration of BSEP-EYFP did not change its colocalization with Na+/K+-ATPase (Figs. S6B and C), indicating that induction of BSEP internalization upon antibody binding was negligible.
Next, BSEP trans-inhibition was tested using BS-depleted serum samples from all cohorts. Neither Control nor Non-AIBD sera showed any reduction in BSEP-mediated transport in comparison to the no serum reference (Fig. 2D), preserving the substrate distribution S2>L. In contrast, preincubation with all seven AIBD sera resulted in decreased export (S2) and increased cellular retention (L) of [3H]-TC (red arrows in Fig. 2D and Fig. S2D), shifting the substrate distribution to L>S2. Notably, this distribution resembled that of cells expressing only NTCP (Fig. 2D, inset). Preincubation of NTCP/BSEP cells with total purified IgG from AIBD but not Control serum resulted in BSEP trans-inhibition, demonstrating that BSEP trans-inhibition was antibody-mediated (Fig. 2E). Since antibody decoration of cells is the first step in the classical complement pathway, it may lead to cell lysis triggered by incubation with fresh AIBD sera and compromise the test. Fresh samples of a BS-depleted AIBD and Control serum, however, did not show any difference in BSEP trans-inhibition between native and heat-inactivated states (Fig. S7). Thus, we could rule out that measured BSEP inhibition in this assay was an artifact of BSEP internalization from the plasma membrane, of excess serum BS competing with the reporter substrate, or of cell lysis due to complement activation. Therefore, the BSEP trans-inhibition measured in our test with AIBD sera was antibody-mediated and caused by direct inhibition of BSEP-driven BS transport.
Functional testing substantiates diagnosis of AIBD and provides a readout of treatment response: example of an affected patient.
The functional readout of BSEP trans-inhibition not only corroborates AIBD diagnosis but can be used to monitor antibody-depleting therapeutic approaches and thus help guide subsequent interventions. A boy of 11 years (AIBD7) initially presented with pruritus and strongly elevated serum BS (204.5 μM). A biopsy of his native liver showed fibrosis stage III and evidence of chronic intrahepatic cholestasis. IHC staining for BSEP was negative, suggesting severe BSEP deficiency (Fig. 3A), which was confirmed by detection of a pathogenic homozygous splice-site variant in the BSEP (ABCB11) gene. Partial external biliary diversion was performed six months after diagnosis and temporally alleviated symptoms. 12 months later (at 12.5 years of age) the patient received the full organ of an unrelated deceased donor after which his clinical condition improved and serum BS normalized (from 133.5 to 8.1 μM). Since this patient was at high risk of developing AIBD due to congenital absence of BSEP,
, he was screened for emergence of serum anti-BSEP antibodies. Two serum samples taken within the first year after OLT showed no BSEP-reactivity on Western blots (Fig. 3B, sera 1 and 2). A transplant biopsy was performed for suspected acute rejection one month after OLT and found portal inflammation, endothelitis and ductulitis. 11 months after OLT, an elevated ALT (163U/L) level prompted a second biopsy to rule out rejection and showed liver fibrosis stage I and uncharacteristic portal inflammation. A third biopsy prompted by transaminase elevation 6 months later (17 months after OLT) showed progression of fibrosis to stage II-III, and thus immunosuppression was increased (tacrolimus, mycophenolate mofetile, prednisolone). One month later (18 months after OLT), donor-specific antibodies were detected. 23 months after OLT, anti-BSEP antibodies were detected for the first time in the patient´s serum (Fig. 3B, serum 3). Pruritus recurred and ALT (160U/L) was persistently elevated despite normal gGT (11U/L). Serum TBS increased from normal levels (4.3 μM) to 230.7 μM over the course of 1.5 months. BSEP trans-inhibition was first detected from serum drawn two years after OLT (Fig. 3C, serum 4). The fourth liver biopsy was performed to rule out other differential diagnoses besides AIBD and showed normal canalicular BSEP expression alongside canalicular IgG deposits (Fig. 3D), which are highly characteristic of AIBD.
Based on absence of other causes of liver disease, presence of canalicular IgG deposits and BSEP trans-inhibition, he was diagnosed with AIBD and treatment was initiated (Fig. 3C). The patient received a first round of three steroid pulses (10-20 mg/kg/d prednisolone) followed by three plasmapheresis treatments and two doses of rituximab (375mg/m2 each). This was followed by another treatment with immunoadsorption and plasmapheresis, both of which reduced serum anti-BSEP antibody levels (Fig. 3B) and BSEP trans-inhibition (Fig. 3C). To further reduce systemic anti-BSEP antibody load, a series of immunoadsorption and plasmapheresis treatments (12 instances over 3.5 months) was initiated (Fig. 3E). Each treatment instance resulted in a reduction of BSEP trans-inhibition in vitro, coinciding with reduced serum anti-BSEP reactivity on Western blot (Fig. S8). While stable, the patient continues to require intensified immunosuppression and regular plasmapheresis treatment alongside intravenous IgG supplementation to manage his AIBD.
An updated diagnostic workflow for AIBD
To illustrate the additional diagnostic value of a functional assay, we compiled data regarding anti-BSEP antibody detection
from three post-OLT PFIC-2 patients (Fig. 4A-C). In contrast to control serum, sera from all three patients showed anti-canalicular reactivity on human liver tissue (Fig. 4A). In AIBD7 and patient 1, but not patient 2, this reactivity was directed against BSEP as shown by IF and Western blot based on BSEP cells (Fig. 4B, C). Accordingly, AIBD7 and patient 1 would have been diagnosed with AIBD by demonstration of anti-BSEP antibodies in their sera. Applying the BSEP trans-inhibition assay, we demonstrated that BS-depleted serum from AIBD7, but not patient 1 and 2 trans-inhibited BSEP (Fig. 4D). Patient 1, the sibling of another AIBD patient from a previous study,
, lacked congenital BSEP expression and also had detectable levels of serum anti-BSEP antibodies (Fig. 4B, C) which could bind to extracellular BSEP epitopes (Fig. S9A). However, at time of analysis, the anti-BSEP antibodies detected in patient 1’s serum did not significantly inhibit BSEP transport activity (Fig. 4D), even after doubling the serum amount during preincubation (Fig. S9B). In line with this, total BS levels in his serum were normal (4.8 μM) and patient 1 was symptom-free at the time of serum sampling. In contrast, serum of patient 2 contained anti-canalicular antibodies, which were not reactive towards BSEP (Fig. 4B, C), accordingly showing no BSEP trans-inhibition (Fig. 4D). Thus, we could exclude AIBD in patient 1 and 2 at the time of serum sampling. However, as shown for AIBD7 (see Fig. 3) inhibitory antibodies may develop in patient 1 over time. Incorporation of the BSEP trans-inhibition assay into the AIBD diagnostic algorithm allows functional confirmation of AIBD in post-OLT PFIC-2 patients (Fig. 5).
Fig. 4A-C Current diagnostic workup of AIBD is based on the detection of BSEP-reactive antibodies in patient serum samples. A Cryosections of normal human liver tissue were immunostained with patient sera alongside normal control serum for detection of anti-canalicular antibodies (human IgG, red). MRP2 (green) stained as canalicular marker. Bar = 10 μm. B Immunostaining of HEK293 cells expressing BSEP-EYFP (green) with control or patient sera (red). Nuclei were stained with DAPI. Bar = 10 μm. C BSEP reactivity of patient sera was tested by Western blot using plasma membrane preparations from HEK293 cells (∅) and BSEP-EYFP-expressing cells (BSEP). D Summary of diagnostic workup for (1) IF testing for anti-canalicular reactivity as depicted in A, (2) IF and WB data for BSEP specificity as shown in B and C, and (3) BSEP trans-inhibition test data. Data are shown as mean ± SD of triplicate measurements.
Fig. 5Updated proposed diagnostic workup for AIBD incorporating the cell-based BSEP trans-inhibition test (also see Fig. 4D). According to this procedure, patients AIBD1-7 were conclusively diagnosed with AIBD, while AIBD was excluded for patients 1 and 2 at the time of serum sampling.
In the present study, we developed and validated a cell-based assay that directly detects antibody-mediated BSEP inhibition from the extracellular side (trans-inhibition) using a BS-depleted serum sample, recapitulating the central causative factor of AIBD pathogenesis.
Allogeneic haematopoietic stem cell transplantation eliminates alloreactive inhibitory antibodies after liver transplantation for bile salt export pump deficiency.
Relapsing features of bile salt export pump deficiency after liver transplantation in two patients with progressive familial intrahepatic cholestasis type 2.
Cholestasis After Pediatric Liver Transplantation-Recurrence of a Progressive Familial Intrahepatic Cholestasis Phenotype as a Rare Differential Diagnosis: A Case Report.
Recurrent low gamma-glutamyl transpeptidase cholestasis following liver transplantation for bile salt export pump (BSEP) disease (posttransplant recurrent BSEP disease).
It is less likely that newly synthesized BSEP is decorated by IgG at the sinusoidal membrane, since it seems to be directly targeted from the Golgi apparatus to the canalicular membrane.
This together with the demonstration that AIBD serum stained BSEP in unpermeabilized freshly isolated hepatocytes, in which the canalicular membrane is accessible,
Allogeneic haematopoietic stem cell transplantation eliminates alloreactive inhibitory antibodies after liver transplantation for bile salt export pump deficiency.
Relapsing features of bile salt export pump deficiency after liver transplantation in two patients with progressive familial intrahepatic cholestasis type 2.
Cholestasis After Pediatric Liver Transplantation-Recurrence of a Progressive Familial Intrahepatic Cholestasis Phenotype as a Rare Differential Diagnosis: A Case Report.
Five of the six previously unreported AIBD cases in this study carry variants either predicted or confirmed to cause complete loss of BSEP expression which further corroborates that congenital absence of BSEP expression poses a risk factor for AIBD development.
Many pharmacological studies use polarized cell lines in which BS undergo vectorial transcellular transport from the basal to the apical side of the cell monolayer.
Vectorial transport of unconjugated and conjugated bile salts by monolayers of LLC-PK1 cells doubly transfected with human NTCP and BSEP or with rat Ntcp and Bsep.
Am J Physiol Gastrointest Liver Physiol.2006; 290: G550-G556
These tests require a confluent cell layer and handling of cells on cell culture inserts. However, for measuring BSEP trans-inhibition vectorial BS transport is not required. Therefore, our novel approach makes use of unpolarized cells (Fig. 1B, D). In this assay, the only function of NTCP is to preload the cells with BS during the import phase. This is possible since NTCP import by far supersedes BSEP export in the assay cell line (Fig. S5A), resulting in a substantial net uptake of [3H]-TC in the first assay phase as depicted by comparable S1 levels in NTCP and NTCP/BSEP cells (Fig. 1E). After preloading the cells, NTCP activity is diminished by removal of extracellular sodium to exclusively measure BSEP transport (and its inhibition) in the export phase. Here, the bulk of accumulated [3H]-TC is exported from the cells into S2, resulting in the substrate distribution S2>L. Intracellular BS retention caused by BSEP inhibition, a characteristic of AIBD, was exclusively observed in all tested AIBD sera, causing a shift in substrate distribution to L>S2. Accordingly, a shift from S2 to L indicates BSEP trans-inhibition and AIBD. Moreover, substrate distribution and thus trans-inhibition can readily be expressed as L/S2 (Figure 3C, E).
Since AIBD leads to cholestasis accompanied by high serum BS levels, BS depletion is a prerequisite for robust test performance and prevention of false positive results. A simple approach using spin concentrators sufficed to reduce BS levels below inhibitory values (Fig. S5B). Furthermore, complement inactivation by heat treatment was not necessary for proper test function (Fig. S7). Cell surface decoration with anti-BSEP antibodies could be expected to trigger the classical complement cascade pathway leading to membrane attack complex formation and cell lysis during preincubation. Previous studies have not found any impact of human complement on human cell lines, which has been attributed to the general presence of cell surface complement regulatory factors such as decay accelerating factor (DAF, CD55) CD59 and MCP (CD46), all of which are expressed on the surface of HEK293 cells,
, to protect cells and tissues from homologous complement activation.
Transplanted patients are treated with immunosuppressants and a variety of other medications. While small molecular weight compounds should be removed along with the excess BS during spin column depletion, a substantial modulation of NTCP transport activity during import should raise a red flag of concern since some compounds that modulate NTCP may also modulate BSEP activity.
Inhibition of bile acid transport across Na+/taurocholate cotransporting polypeptide (SLC10A1) and bile salt export pump (ABCB 11)-coexpressing LLC-PK1 cells by cholestasis-inducing drugs.
This problem could be overcome by purification of total IgG from the serum sample (Fig. 2E). We do not routinely purify serum antibodies since this requires more hands-on time and introduces more variation in terms of recovery due to the small serum volumes routinely obtained from pediatric patients. Antibody-mediated BSEP internalization during preincubation is another potential confounder of the assay which could be excluded by colocalization analysis (Figs. S6B and C). This lack of BSEP internalization in vitro is in line with observations of normal canalicular BSEP localization in biopsy material from AIBD livers (also see Fig. 3D), supporting direct functional BSEP inhibition at the canaliculus in vivo.
By prospective screening of a patient transplanted for PFIC-2, who was at high risk of developing AIBD due to congenital absence of BSEP expression, we were able to functionally monitor seroconversion and thus prompt immediate treatment while evaluating its outcome by assaying BSEP trans-inhibition activity (Fig. 3). Ultimately, the ongoing treatment regime led to stabilization of the patient. Notably, another PFIC-2 patient (Fig. 4, patient 1) developed anti-BSEP antibodies after OLT, which recognized extracellular BSEP epitopes (Fig. S9A) yet failed to show any BSEP trans-inhibition (Fig. 4D), even after doubling the amount of BS-depleted serum during preincubation (Fig. S9B). While we could previously demonstrate that both intra- and extracellular BSEP epitopes are recognized by AIBD-derived anti-BSEP antibodies,
, information on the complexity of the anti-BSEP allo-immune response with respect to polyclonality is elusive. Based on the recent BSEP cryo-EM structure,
, only eight percent of its amino acids (∼110 of 1321) are accessible from the extracellular side and therefore pathophysiological targets of binding and trans-inhibition. Thus, only a fraction of the various anti-BSEP antibody specificities contained in a given humoral response can bind to their target from the canalicular lumen. Antibodies that do not confer trans-inhibition upon BSEP binding could leave its carrier essentially asymptomatic, since canalicular antibody decoration does not trigger any changes in canalicular BSEP localization in affected livers (see above).
In summary, transplanted PFIC-2 patients may develop AIBD. We strongly encourage immediate diagnostic testing whenever phenotypic symptom recurrence post OLT is suspected since this enables a timely therapeutic intervention. The "classical" AIBD diagnosis is strongly corroborated by functional demonstration of BSEP-reactive antibodies with trans-inhibitory capacity and may help to rule out other differential diagnoses of post-OLT cholestasis. Using NTCP/BSEP co-expressing cells and a temporal separation of their respective activity into import and export phases allows testing for BSEP trans-inhibition. The AIBD diagnostic workflow proposed here unites detection of anti-BSEP antibodies with functional testing for BSEP trans-inhibitory antibodies.
Acknowledgments
Expert technical assistance by Nicole Eichhorst, Annette Tries, and Nathalie Walter is gratefully acknowledged as well as outstanding management of patient data by Nicole Weyandt. We thank “Hamburg macht Kinder gesund e.V.” and the “YAEL-Stiftung” for supporting the database on liver-transplanted children at the University Medical Center Hamburg-Eppendorf and "Billy Rubin Förderverein Kindergastroenterologie MHH e.V.".
Appendix A. Supplementary data
The following is/are the supplementary data to this article:
ATP8B1 and ABCB11 analysis in 62 children with normal gamma-glutamyl transferase progressive familial intrahepatic cholestasis (PFIC): phenotypic differences between PFIC1 and PFIC2 and natural history.
Allogeneic haematopoietic stem cell transplantation eliminates alloreactive inhibitory antibodies after liver transplantation for bile salt export pump deficiency.
Relapsing features of bile salt export pump deficiency after liver transplantation in two patients with progressive familial intrahepatic cholestasis type 2.
Cholestasis After Pediatric Liver Transplantation-Recurrence of a Progressive Familial Intrahepatic Cholestasis Phenotype as a Rare Differential Diagnosis: A Case Report.
Recurrent low gamma-glutamyl transpeptidase cholestasis following liver transplantation for bile salt export pump (BSEP) disease (posttransplant recurrent BSEP disease).
Vectorial transport of unconjugated and conjugated bile salts by monolayers of LLC-PK1 cells doubly transfected with human NTCP and BSEP or with rat Ntcp and Bsep.
Am J Physiol Gastrointest Liver Physiol.2006; 290: G550-G556
Inhibition of bile acid transport across Na+/taurocholate cotransporting polypeptide (SLC10A1) and bile salt export pump (ABCB 11)-coexpressing LLC-PK1 cells by cholestasis-inducing drugs.
00021-6&id=ga1.jpg){kind=link}
00021-6&id=gr1.jpg){kind=link}
00021-6&id=gr3.jpg){kind=link}
00021-6&id=gr2.jpg){kind=link}
00021-6&id=gr4.jpg){kind=link}
00021-6&id=gr5.jpg){kind=link}