Model-free estimation of the effective correlation time for C–H bond reorientation in amphiphilic bilayers: ¹H–¹³C solid-state NMR and MD simulations

Molecular dynamics (MD) simulations give atomically detailed information on structure and dynamics in amphiphilic bilayer systems on timescales up to about 1 μs. The reorientational dynamics of the C–H bonds is conventionally verified by measurements of ¹³C or ²H nuclear magnetic resonance (NMR) longitudinal relaxation rates R₁, which are more sensitive to motional processes with correlation times close to the inverse Larmor frequency, typically around 1-10 ns on standard NMR instrumentation, and are thus less sensitive to the 10-1000 ns timescale motion that can be observed in the MD simulations. We propose an experimental procedure for atomically resolved model-free estimation of the C–H bond effective reorientational correlation time τ_e, which includes contributions from the entire range of all-atom MD timescales and that can be calculated directly from the MD trajectories. The approach is based on measurements of ¹³C R₁ and R_1ρ relaxation rates, as well as ¹H−¹³C dipolar couplings, and is applicable to anisotropic liquid crystalline lipid or surfactant systems using a conventional solid-state NMR spectrometer and samples with natural isotopic composition. The procedure is demonstrated on a fully hydrated lamellar phase of 1-palmitoyl-2-oleoyl-phosphatidylcholine, yielding values of τ_e from 0.1 ns for the methyl groups in the choline moiety and at the end of the acyl chains to 3 ns for the g₁ methylene group of the glycerol backbone. MD simulations performed with a widely used united-atom force-field reproduce the τ_e-profile of the major part of the acyl chains but underestimate the dynamics of the glycerol backbone and adjacent molecular segments. The measurement of experimental τ_e-profiles can be used to study subtle effects on C–H bond reorientational motions in anisotropic liquid crystals, as well as to validate the C–H bond reorientation dynamics predicted in MD simulations of amphiphilic bilayers such as lipid membranes