Insights into the Tunnel Mechanism of Cholesteryl Ester Transfer Protein through All-atom Molecular Dynamics Simulations

Abstract

Cholesteryl ester transfer protein (CETP) mediates cholesteryl ester (CE) transfer from the atheroprotective high density lipoprotein (HDL) cholesterol to the atherogenic low density lipoprotein cholesterol. In the past decade, this property has driven the development of CETP inhibitors, which have been evaluated in large scale clinical trials for treating cardiovascular diseases. Despite the pharmacological interest, little is known about the fundamental mechanism of CETP in CE transfer. Recent electron microscopy (EM) experiments have suggested a tunnel mechanism, and molecular dynamics simulations have shown that the flexible N-terminal distal end of CETP penetrates into the HDL surface and takes up a CE molecule through an open pore. However, it is not known whether a CE molecule can completely transfer through an entire CETP molecule. Here, we used all-atom molecular dynamics simulations to evaluate this possibility. The results showed that a hydrophobic tunnel inside CETP is sufficient to allow a CE molecule to completely transfer through the entire CETP within a predicted transfer time and at a rate comparable with those obtained through physiological measurements. Analyses of the detailed interactions revealed several residues that might be critical for CETP function, which may provide important clues for the effective development of CETP inhibitors and treatment of cardiovascular diseases.

Keywords: cholesterol; cholesterol metabolism; cholesterol regulation; cholesterol-binding protein; lipid metabolism; lipid transport; lipid-protein interaction; lipoprotein metabolism; molecular dynamics.

FIGURE 1.

Simulation system for studying CE transfer. A, a representative negative staining EM image shows that CETP bridged HDL and LDL, forming a ternary complex (shown in schematic in B). C, a simplified simulation system (shown in schematic) was used to simulate the ternary complex to elucidate CE transfer from HDL to LDL at an atomic level. D, the N-terminal β-barrel domain of the CETP was inserted ∼35 Å deep into a POPC monolayer adhered to a CE pool, whereas the C-terminal β-barrel domain penetrated ∼32 Å inside an opposing POPC monolayer attached to a TG pool. The region between the two opposing lipid monolayers was filled with water molecules. The POPC headgroups and fatty tails are colored yellow and green, respectively, and the CE, TG, CETP, and water molecules are colored pink, gray, blue, and orange, respectively. The CE molecule is highlighted using van der Waals spheres.

FIGURE 2.

Equilibration of the simulation system through all-atom MD simulations. A, three repeated equilibrations of CETP were monitored on the basis of the RMSD (top panel), the radius of gyration (Rgyr) against the mass center of the molecule (middle panel), and the volume of the entire molecule (bottom panel). B, the equilibration of the CE and TG pools was achieved after a stable measured pool thickness was reached (top panel). The equilibration of the POPC monolayers was achieved after the distance (Dist.) between POPC molecules was stable (indicated by the first peak of the phosphorus radial distribution function (bottom panel)). C, the fluctuation of CETP in lipid monolayers (blue line), as indicated by the RMSF of the Cα atom, was calculated from the last 20 ns of simulations and compared with the RMSF of this molecule in solution (orange line). D, the CETP structure in the lipid monolayer is colored according to the RMSF values (the color and value relationship are shown as a color bar).

FIGURE 3.

Conformational changes in the CETP structure after equilibration. A, the conformational changes in the CETP structure were determined after computing the shifts of the Cα atoms in the average structures from the last 20 ns of simulations against the crystal structure. The CETP structure is colored according to the value of the shift (the color and value relationship are shown as a color bar). Zoomed-in views of two predominant conformational changes located in the distal ends of the N- and C-terminal β-barrel domains are shown. These regions (blue) were magnified and compared with the crystal structure (yellow). Two perpendicular views show the CETP internal cavities and pores in the crystal structure (B) and the structure after equilibration in lipid monolayers (C).

FIGURE 4.

CE transfer through the CETP molecule through all-atom steered molecular dynamics simulations. A, snapshots of a CE molecule (shown in van der Waals spheres) during transfer through a CETP molecule (shown in ribbon) under a representative force of 11 kcal/mol/Å. The POPC headgroups and fatty tails are colored yellow and green, respectively, and the CE, TG, CETP, and water molecules are colored pink, gray, blue, and orange, respectively. Shown is a representative snapshot image of CETP when the CE molecule was in the middle of the N-terminal β-barrel domain (B), leading to the rotation of the N-terminal β-barrel domain (C, blue arrows). D, each residue shifts from its original position during CE transfer. For each Cα atom, the maximum shift from four simulations was averaged and colored on the CETP structure (the color and value relationship are shown as a color bar). E, CE transfer through CETP was simulated four times under the representative force of 11 kcal/mol/Å. The RMSD from each simulation versus the CE position along the CE transfer pathway is shown.

FIGURE 5.

CE transfer pathways and the relationship between transfer time and driving force. A, the transfer processes described above were repeated four times. The CE transfer pathways are shown by the trajectories of the C10–C19 bond of the CE steroid ring during its transfer within the CETP tunnel. Each pathway of CE transfer is shown in different colors. The rotational angle of the C10–C19 bond against the paper surface in the side view was used as an indicator of the CE steroid ring orientation (B). Error bars represent S.D. The CE steroid ring would rotate ∼90° to pass through the central cavity and exit the C-terminal distal pore. After repeating the CE transfer 72 times under a series of 18 driving forces (from 6 to 23 kcal/mol/Å) (C), the relationship between CE transfer time and driving force is plotted according to the negative log and subsequently fitted using a linear equation (D). The best fit line has a slope of 2.75 and an R-factor of 0.96, corresponding to the following time and force relationship: Time (ns) = 475 × Force (kcal/mol/Å)^−2.75.

FIGURE 6.

Energy barriers, energy wells, and residues required for CE transfer through CETP. A, energy distributions along the CE transfer pathway under a series of driving forces (ranging from 6 to 23 kcal/mol/Å). Each curve was averaged from four experiments repeated with the same driving force. The combination energy (black line) includes the interaction energies, i.e. the energy between the transferring CE molecule and lipid pools (orange line) and the energy between the transferring CE molecule and the CETP molecule (blue line). The energy distribution was calculated within a 6-Å step along the transfer pathway. The combined energy under a representative driving force of 11 kcal/mol/Å is highlighted in purple. The energy under the driving forces of 6 and 23 kcal/mol/Å is indicated by arrows. B, the representative energy distributions under a driving force of 11 kcal/mol/Å are shown with error bars indicating S.D. calculated from four repeated simulations within a 1-Å step along the transfer pathway. The central energy barrier and low energy wells within the β-barrel domains are indicated by purple and green arrows, respectively. C, the hydrophobicity of the tunnel wall surrounding the transferred CE molecule (distance <2.4 Å) is displayed as the average hydrophobicity within a 1-Å step against the CE transfer pathway. Error bars represent S.D. D, the residues that predominantly contributed to the energy wells in the N-terminal and C-terminal β-barrel domains (left and right panels) and to the energy barrier in the central region (central panel) are highlighted in green. E, a schematic showing the residues that might regulate CE transfer via strong interactions.