Combination of Molecular Dynamics Simulations and Machine Learning Reveals Structural Characteristics of Stereochemistry-Specific Interdigitation of Synthetic Monomycoloyl Glycerol Analogs.

Abstract

Synthetic monomycoloyl glycerol (MMG) analogs possess robust immunostimulatory activity and are investigated as adjuvants for subunit vaccines in preclinical and clinical studies. These synthetic lipids consist of a glycerol moiety attached to a corynomycolic acid. Previous experimental studies have shown that the stereochemistry of the lipid acid moiety affects whether the MMG analogs self-assemble into interdigitated or noninterdigitated structures below the main phase transition temperature (Tm). In this study, we elucidated possible thermodynamic mechanisms governing the phase behavior of MMG analogs by exploring their conformations, interactions, and dynamics using a combination of machine learning (ML) and molecular dynamics (MD) simulations. We compared two analogs, MMG-1 and MMG-6, which differ only by the stereochemistry of the lipid acid moiety; the former has a configuration different from the natural MMG, and the latter displays a native-like stereochemistry. Three different membrane states were simulated: (1) a noninterdigitated single bilayer, (2) a noninterdigitated double bilayer, and (3) a fully interdigitated double bilayer. Our results indicate that the propensity for interdigitation of the MMG analogs in a bilayer is linked to the degree to which their hydrocarbon chains are ordered and oriented. This study demonstrates how combining MD simulations with ML can enhance the molecular understanding of lipid-based pharmaceutical formulations.