*4.7. Computational Methods*

We studied four ABCG5/G8 protein systems including the WT and the E146Q, A540F, and R543S mutants. Each MD system consisted of one copy of ABCG5/G8 heterodimer, 320 1,2-dimyristoyl-*sn*-glycero-3-phosphocholine (DMPC) lipids, 16 cholesterols, 43,621 TIP3P [46] water molecules, and 103 Cl<sup>−</sup> and 83 Na<sup>+</sup> to neutralize the MD systems. AMBER ff14SB [47], Lipid14 [48], and GAFF [49] force fields were used to model proteins, DMPC lipids, and cholesterols, respectively. The residue topology of cholesterol was prepared using the Antechamber module [48]. MD simulation was performed to produce isothermal–isobaric ensembles using the pmemd.cuda program in AMBER 18 [50]. The particle mesh Ewald (PME) method [51] was used to accurately calculate the electrostatic energies with the long-ranged correction taken into account. All bonds were constrained using the SHAKE algorithm [52] in both the minimization and MD simulation stages following a computational protocol described in our previous publication [21]. Briefly, there were three stages in a series of constant-pressure and -temperature MD simulations, including the relaxation phase, the equilibrium phase, and the sampling phase. In the relaxation phase, the simulation system was heated progressively from 50 K to 250 K at steps of 50 K, and a 1 ns MD simulation was run at each temperature. In the next equilibrium phase, the system was equilibrated at 298 K, 1 bar for 10 ns. Finally, a 100 ns MD simulation was performed at 298 K, 1 bar to produce isothermal–isobaric ensemble ensembles. In total, 1000 snapshots were recorded from the last phase simulation for post-analysis using the "cpptray" module implemented in the AMBER software package. Binding free energy decomposition and correlation analysis were performed using an internal program and the detailed elsewhere [53,54].

**Supplementary Materials:** Supplementary Materials can be found at http://www.mdpi.com/1422-0067/21/22/ 8747/s1.

**Author Contributions:** B.M.X. optimized the CHS-stimulated ATPase assay for ABCG5/G8; A.A.Z. generated and validated the mutant constructs; B.M.X., A.A.Z. and A.V. purified the proteins and carried out the ATPase assays and data analysis; J.W. performed the molecular dynamics simulation; J.-Y.L. oversaw the project; J.W. and J.-Y.L. wrote the manuscript. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by a startup grant from the University of Ottawa, a Discovery Grant from the Natural Sciences and Engineering Research Council (RGPIN 2018-04070), and a National New Investigator Award from the Heart and Stroke Foundation of Canada to J.-Y.L., as well as the National Science Foundation (NSF 1955260) and National Institutes of Health (R01-GM079383) grants to J.W. B.M.X. is a recipient of the Travel Awards from the Canadian Society of Molecular Biosciences (2018) and Biophysical Society of Canada (2019).

**Acknowledgments:** We thank William Jennings, Gloria Ihirwe, Midhet Hajira, and Chloé van de Panne for technical assistance. We also thank Donna Clary and Hui Li for their technical help in utilizing the common core facilities. We are indebted to critical feedback on reviewing and editing the manuscript from Vicky Brandt, and John Baenziger, Jean-François Couture, Gregory Graf, and Xiaohui Zha. This work is partially based on the theses that were submitted to fulfill in part the requirement for the degrees of Master of Science (A.A.Z.) and Honors Bachelor of Science (A.V.).

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
