Extension of CAVS coarse‐grained model to phospholipid membranes: The importance of electrostatics

Abstract

It is evident from experiment that electrostatic potential (or dipole potential) is positive inside PC or PE lipid bilayers in the absence of ions. MARTINI coarse-grained (CG) model, which has been widely used in simulating physical properties of lipid bilayers, fails to reproduce the positive value for the dipole potential in the membrane interior. Although the total dipole potential can be correctly described by the BMW/MARTINI model, the contribution from the ester dipoles, playing a nontrivial role in the electrostatic potential across lipid membranes, is neglected by this hybrid approach. In the ELBA CG model, the role of the ester dipoles is considered, but it is overweighed because various atomistic models have consistently shown that water is actually the leading contributor of dipole potential. Here, we present a CG approach by combining the BMW-like water model (namely CAVS model) with the ELBA-like lipid model proposed in this work. Our CG model was designed not only to correctly reproduce the positive values for the dipole potential inside PC and PE lipid bilayers but also to properly balance the individual contributions from the ester dipoles and water, surmounting the limitations of current CG models in the calculations of dipole potential. © 2017 Wiley Periodicals, Inc.

Effect of Cholesterol and 6-Ketocholestanol on Membrane Dipole Potential and Sterol Flip-Flop Motion in Bilayer Membranes

“…In our previous works, − we have demonstrated the importance of electrostatic interactions in the prediction of the dipole potential. In fact, the dipole potential can generate about 10 8 to 10 9 V/m of electric field, which would substantially promote or prevent the transportation of molecules into lipid bilayer membranes. − However, the relation between the dipole potential and the sterol transportation in membranes is less well studied and understood.…”

“…Finally, some popular CG models fail to capture the sterol effect on the dipole potential owing to an inaccurate treatment of electrostatic interactions. 37,41,42 In our previous work, 48 we presented the charge attached to virtual site (CAVS) model for water (Figure S1 of the Supporting Information). In this approach, a CG bead represents four real water molecules.…”

“…In the CAVS model for phospholipids (Figure S2 of the Supporting Information), electrostatic interaction sites were introduced into the ester groups of phospholipids in order to take into account the contribution of ester dipoles to the dipole potential. 41 Moreover, considering the hydrophilic nature of cholesterol, electrostatic interaction sites were introduced into the CAVS model for cholesterol. 42 In this work, a comparative study is made between the effects of cholesterol (CHOL) and 6-ketocholestanol (6-KC) on the physical properties of dimyristoylphosphatidylcholine (DMPC) bilayer membranes as well as its effect on the sterol flip-flop motion in DMPC bilayer membranes.…”

“…First, atomistic MD simulations would overestimate the dipole potential while using nonpolarizable water models. ,, Second, it is very expensive to use polarizable models to detect the sterol flip-flop motion in lipid membranes. Finally, some popular CG models fail to capture the sterol effect on the dipole potential owing to an inaccurate treatment of electrostatic interactions. ,, …”

“…Each CG bead is composed of two positively charged sites (namely, CGPs), one van der Waals (vdW) interaction center (namely, CGM), and one virtual site (namely, CGN). In the CAVS model for phospholipids (Figure S2 of the Supporting Information), electrostatic interaction sites were introduced into the ester groups of phospholipids in order to take into account the contribution of ester dipoles to the dipole potential . Moreover, considering the hydrophilic nature of cholesterol, electrostatic interaction sites were introduced into the CAVS model for cholesterol …”