Integrating Ion-Mobility and MD Simulations with Enhanced Sampling Techniques to Characterize Gas-Phase Compaction of Protein Structures

VIRAJ GANDHI, Carlos Larriba-Andaluz, Purdue University

     Abstract Number: 565
     Working Group: Aerosol Physics

Abstract
This work addresses significant challenges associated with characterizing protein structures in the gas phase. The compaction of proteins in the gas phase is a phenomenon that is widely acknowledged and observed experimentally. However, there are no theoretical and simulation approaches that truly explain the process. Among the possible reasons, two stand out, the first is because most forcefields are optimized in aqueous environments and the second is because the timescale to cover such broad ergodic spaces is too large to sample every scenario. Our contribution lies in extending molecular dynamics(MD) simulations beyond conventional approaches, using enhanced sampling techniques such as Adaptive Biasing Forces to navigate the protein’s configuration space more effectively so reaching the equilibrium gas-phase structure observed experimentally. Collision cross-sections (CCS) of six different proteins (Immunoglobulin, Concanavalin, BSA, PK, SAP, and AD) were experimentally measured and found to align with literature values. Contrarily, structures generated through unbiased MD simulations in the gas-phase produced CCS values 10-30% larger. To reconcile this discrepancy, a harmonic biasing potential in the direction of the radius of gyration was employed to efficiently map all possible configurations which might be hidden behind energy barriers not easily bypassed. Periods alternating between applied-force and force-free intervals, akin to simulated annealing, are employed. Upon force application, a significant decrease in CCS was observed, indicative of instantaneous compression. During subsequent force-free intervals, relaxation was noted without reverting to unbiased values, suggesting the attainment of more compact configurations. With successive cycles, the CCS during the force-free intervals gradually stabilizes, reaching an equilibrium value. Convergence was achieved regardless of the strength of the force applied, and the convergence CCS was within ~4% of the experimental CCS for different charge states. The protein structures tend to become globular, even for the Y-shaped immunoglobulin, which seems a likely outcome under capillarity effects.