A structural and mechanistic model for BSEP dysfunction in PFIC2 cholestatic disease

BSEP (ABCB11) transports bile salts across the canalicular membrane of hepatocytes, where they are incorporated into bile. Biallelic mutations in BSEP can cause Progressive Familial Intrahepatic Cholestasis Type 2 (PFIC2), a rare pediatric disease characterized by hepatic bile acid accumulation leading to hepatotoxicity and, ultimately, liver failure. The most frequently occurring PFIC2 disease-causing mutations are missense mutations, which often display a phenotype with decreased protein expression and impaired maturation and trafficking to the canalicular membrane. To characterize the mutational effects on protein thermodynamic stability, we carried out biophysical characterization of 13 distinct PFIC2-associated variants using in-cell thermal shift (CETSA) measurements. These experiments reveal a cluster of residues localized to the NBD2-ICL2 interface, which exhibit severe destabilization relative to wild-type BSEP. A high-resolution (2.8 Å) cryo-EM structure provides a framework for rationalizing the CETSA results, revealing a novel, NBD2-localized mechanism through which the most severe missense patient mutations drive cholestatic disease. These findings suggest potential strategies for identifying mechanism-based small molecule correctors to address BSEP trafficking defects and advance novel therapies for PFIC2 and other cholestatic diseases.

Methods
CETSA

Adherent HEK293T cells were grown in a 6-well plate and transfected with BSEP-HiBiT constructs when they reached 70% confluency using PEI MAX® - Transfection Grade Linear Polyethyleneimine Hydrochloride (MW 40,000) (Polysciences) at a 1:3 DNA: PEI mass ratio. Cells were collected 24h after transfection by trypsinization, resuspended in fresh DMEM at a concentration of 1.2x106 cells/ml, and dispensed in PCR tubes (20 μl per tube, 2 replicates per condition). Tubes were placed in a multi-block thermal cycler and submitted to a thermal cycle of 10 minutes pre-incubation at 37°C, 3.5 minutes heating step, and fast cooling to 4°C. Cells were then lysed, and a luminescent signal was generated by adding Nano-Glo® HiBiT Lytic Detection mix prepared per manufacturer’s instructions (Promega) to the heated samples at a 1:1 vol/vol ratio. Cell lysates were subsequently dispensed in a white 384 well plate (Corning) and luminescence was detected on a plate reader (TECAN Spark®).

Melting curve analysis

For each independent CETSA experiment, the luminescence of two wells per temperature were averaged. The resulting melting curve was fitted with a non-linear regression model with GraphPad Prism (Version 9, Boston, MA) from which a melting/aggregation temperature was extracted. Mean Tagg values were obtained from at least three independent melting curves fitted individually (N=4 biological replicates for WT, R948C, R1128C, R1153C and R1268Q; N=3 for other variants).

Protein purification

A pellet of cells expressing WT BSEP or E297G mutant was resuspended in solubilization buffer (Tris 50 mM pH8, 150 mM NaCl, 10% glycerol, 2 mM DTT, 2 mM MgCl2, 1%DDM, 0.2% CHS, protease inhibitor tablet and DNase/RNase) for 2 hours at 4°C. The cell lysate was centrifuged at 200,000 g for 30 minutes at 4°C, and the supernatant was added to anti-GFP nanobody beads (Protein and Crystallography Facility, University of Iowa) for 1 hour at 4°C. Beads were washed with buffer A (150 mM NaCl, 50 mM Tris pH8, 2 mM DTT, 2 mM MgCl2, 0.02% GDN) and incubated with 3C protease (40 μM) for 3 hours at 4°C to remove the C-terminal GFP tag. The eluted protein was further purified by gel filtration chromatography using a Superose® 6 Increase 10/300 GL (Cytiva) equilibrated with Buffer A.

Differential scanning fluorimetry 

Freshly purified proteins were concentrated to ~0.4 mg/ml on a Pierce 100K MWCO concentrator (Thermo Fisher Scientific) catalog number 88503 for thermal stability analysis with Prometheus NT48 (NanoTemper Technologies). Tryptophan fluorescence 330/350 nm ratio was measured in high sensitivity capillaries submitted to a temperature ramp of 1°C per minute, from 20 to 80 °C. Resulting melting curves were analyzed with PR.Stability Analysis software (NanoTemper). Melting temperatures were obtained from at least three independent experiments (N=3 for WT and N=6 for E297G).

Activity assay and kinetic parameters for BSEP

Purified BSEP activity was measured using the ADP-Glo assay format (Promega, Madison, WI) according to the vendor protocol. The assay buffer contained 25 mM Hepes, pH 8.5, 150 mM NaCl, 2 mM MgCl2, 2 mM DTT, 0.02% GDN, 0.002% Triton X-100, 0.001% faf BSA. Experimental conditions: Purified BSEP 100 nM, ATP 1 mM, Sodium Taurocholate (TCA): 1 – 0 mM (eight concentrations, 1:2 dilution).