Modeling Ethers with Molecular Dynamics: Updated CHARMM Force Field Parameters for Ethers in Model Compounds and Lipid Membranes

Abstract

Listen

Translate article

Small ether molecules such as polyethylene glycol (PEG) have extensive industrial and medical applications. Additionally, phospholipids containing ether linkages make up 30% of the glycerophospholipids in the human brain. Preliminary research suggests that the CHARMM All-Atom Additive Force Field does not accurately recreate experimental results for bilayers composed of ether lipids such as 1,2-di-O-hexadecyl-sn-glycero-3-phosphocholine (DHPC). 100-ns simulations of pure DHPC bilayers do not reproduce essential target data such as surface area per lipid and electron density profile, which describe the structural properties of the membrane. Additionally, the initial parameterization of ethers tested a limited population of linear ethers and did not accurately reproduce potential energy scans about the O-C-C-O dihedral (J. Chem. Theory Comput., 3(3):1120 - 1133). We have used the MP2 density and Dunning diffuse aug-cc-pVQZ basis sets to compute the charge distributions of model ethers, including various lengths of PEG. We have also used Hybrid Methods for Interaction Energies (HM-IE) to compute potential energy scans about the O-C-C-O angle effectively at the CCSD(T)/aug-cc-pVQZ level (J. Phys. Chem. A, 108: 107 - 112). We found that the charge on ether oxygens is substantially more negative than the current CHARMM force field suggests. Adjusting this charge allows more water to penetrate a DHPC bilayer, increasing the surface area per lipid and improving the bilayer's structural representation. Additionally, our new dihedral parameters for the central O-C-C-O angle more accurately reproduce the quantum potential energy scan and bring the free energies of solvation of model compounds closer to experimental values. Further research will involve evaluating the radius of gyration of PEG in salt solutions to improve the interaction of ether oxygens with ions in simulation (Colloid Jour., 72(2): 279 - 281).