*2.2. The Lipid Environments Fine-Tune ABCG5*/*G8 ATPase Activity*

− ABC transporters need to function in a phospholipid-embedded environment. However, it is unknown whether the ABCG5/G8 function is controlled by phospholipids of specific headgroups or in specific lipid compositions. Because a high concentration of bile acids is required to activate ABCG5/G8 ATPase activity, attempts to use reconstituted proteoliposomes failed due to the immediate solubilization of the reconstituted proteins. To facilitate the assessment of mutant functions, we evaluated the lipid environments to obtain the most optimal assay conditions. To study the effect of lipid conditions and phospholipid species on the ABCG5/G8 function, we analyzed the CHS-coupled ATPase activity in the presence of two polar lipid extracts under conditions of fixed concentrations of sodium cholate and CHS (see Section 4). Using *E. coli* polar lipids, we carried out an ATP concentration-dependent ATPase assay to determine the Michaelis–Menten kinetic parameters of CHS-stimulated ATP hydrolysis. We observed the maximal ATP hydrolysis by ABCG5/G8 at concentrations slightly over 2.5 mM of ATP with a Vmax of 677.1 <sup>±</sup> 25.6 nmol/min/mg, a KM(ATP) of 0.60 mM, and a kcat of 1.69 s−<sup>1</sup> . When using bovine liver polar lipids, we observed ~3.5-fold lower catalytic rate of ATP hydrolysis and ~50% higher KM(ATP) (Figure 4A and Table 1). In the current study, polar lipids, cholate (bile acid), and CHS were all present in the reaction, indicating that the presence of *E. coli* polar lipids results in higher ATP association and, consequently, better stimulates ABCG5/G8 ATPase activity. When comparing the calculated values of kcat and kcat/KM, we indeed observed an overall fivefold higher turnover rate in the presence of *E. coli* polar lipids than liver polar lipids (Table 1).

To determine the dependence of phospholipid headgroups, we tested the three most abundant phospholipids in either lipid extract on the ATP hydrolysis by ABCG5/G8, i.e., phosphatidylethanolamine (PE), phosphatidylcholine (PC), and phosphatidylglycerol (PG) (see Section 4). Preincubation with egg PE resulted in the highest specific activity, while the use of soy PC or egg PG only led to slightly higher ATP hydrolysis than the basal activity (Figure 4B). Interestingly, PE, the phospholipid found in both *E. coli* and liver lipids, is sufficient to stimulate ATP hydrolysis in ABCG5/G8 to almost the highest specific activity, as reported here. In the meantime, using PC or PG alone, the specific activity of ABCG5/G8 was also higher than that obtained with the liver polar lipid mixture. These results suggest phospholipid headgroups in regulating the ABCG5/G8 ATPase activity. Further investigations are necessary to pinpoint the effects of individual types of phospholipids on the sterol transporter function.

**Figure 4.** Lipid dependence of ABCG5/G8 ATPase activity. (**A**) Purified ABCG5/G8 was assayed in the presence of either *Escherichia coli* or bovine liver polar lipids, and the specific activities of ATP hydrolysis were obtained by the ATP concentration-dependent experiments (0–5 mM ATP). Both curves are fitted to the Michaelis–Menten equation (Prism 8), and, using two independently purified proteins, the means of at least three independent experiments along with standard deviations are plotted here. The kinetic parameters are listed in Table 1. (**B**) In conditions of 5 mM ATP and 4.1 mM CHS, ATP hydrolysis of purified ABCG5/G8 was assayed in the presence of egg phosphatidylethanolamine (PE), soy phosphatidylcholine (PC), or egg phosphatidylglycerol (PG), where *p*-values of 0.0006 and 0.0003 (marked as \*\*\*), respectively, were obtained using ordinary one-way ANOVA (Prism 8).



<sup>a</sup> Standard errors were calculated from the fits shown in Figures 3A and 5 using Prism 8 (GraphPad Software, San Diego, CA, USA). <sup>b</sup> Turnover rates, kcat, were calculated using the following formula: Vmax <sup>=</sup> <sup>k</sup>cat <sup>×</sup> [E], where [E] is the protein concentration of ABCG5/G8 (363.1 nM). <sup>c</sup> Differential Gibbs free energy was calculated according to the following formula: ∆∆GMUT <sup>=</sup> <sup>−</sup>RTln(kMUT/kWT), where kMUT is the kcat of mutants, kWT is the kcat of WT, <sup>R</sup> <sup>=</sup> 8.314 J·mol−<sup>1</sup> ·K−<sup>1</sup> (R: gas constant), and T = 310.15 K (37 ◦C). <sup>d</sup> Mutants were all assayed in the presence of *E. coli* polar lipids. <sup>e</sup> Number of independent experiments.

−

**Figure 5.** ATP dependence of ABCG5/G8 ATPase activity. Purified proteins were assayed in the presence of *E. coli* polar lipids, and the specific activities of ATP hydrolysis were obtained from the ATP concentration-dependent experiments (0–5 mM ATP). The curves are fitted to the Michaelis–Menten equation (Prism 8), and, using two-to-four independently purified proteins, the means of at least three independent experiments along with standard deviations are plotted here. The kinetic parameters are listed in Table 1.
