(587g) Breaking Boundaries: Free Energy Landscapes for Membrane Pore Formation Induced By Antimicrobial Peptides

Antimicrobial peptides (AMPs) are attractive drug candidates to combat the antimicrobial resistance crisis because of their ability to directly disrupt microbial cell membranes.¹ Although thousands of AMPs have been discovered over the past few decades, their molecular mechanisms of action are still poorly understood because membrane disruption depends upon several properties of both lipid membranes (lipid charge, tail saturation, tail length) and AMPs (length, hydrophobicity, amphiphilicity) which affect whether AMPs form membrane-spanning pores.² Better understanding the pore formation process is of great interest to researchers in developing new AMPs that not only have high activity against target microbial membranes, but also have low affinity for human cells (e.g., human red blood cells (RBCs)) for therapeutic applications in vivo.

To corroborate experimental data on AMP pore formation, molecular dynamics (MD) simulations have been used to resolve corresponding mechanistic pathways. The formation of metastable pores in lipid bilayers is a long timescale process that requires the system to overcome a large energetic barrier for pore creation; consequently, these systems are typically biased to enhance sampling.³ However, picking a collective variable (CV) that accurately represents the pore formation process is nontrivial to avoid unphysical hysteresis during sampling. Recently, Awasthi and Hub developed a CV (referred to as ξ) that more realistically captures the formation of hydrophilic pores in membranes. ξ has been shown to be hysteresis-free and quickly converge in umbrella sampling simulations of pore formation in pure membranes⁴, however, it is yet to be tested extensively in systems in the presence of pore forming agents such as AMPs.

In this study, we investigate affinities for pore formation for 3 AMPs studied in the literature – aurein 1.2, melittin, and magainin 2 – using MD simulations to provide insight into variations in pore formation free energies as a function of peptide and lipid properties. We developed a robust, generalizable methodology for pre-biasing and sampling AMP pore formation events that involves long equilibration of peptides around a pore using the MARTINI coarse-grained force field followed by a backmapping procedure to achieve final pore-lined system configurations for the all-atom CHARMM36 force field. Using this method, we compute free energy barriers for pore formation without directly biasing peptides or whole lipids, allowing us to investigate mechanisms of pore formation for these 3 AMPs that are a consequence of natural lateral diffusion and clustering. Trajectory analysis of atomistic bilayer structural perturbations and peptide-peptide interactions provides further insight into variations in pore formation free energies for these three peptides. Building on these results, we then investigate the consequences of pore formation in representative asymmetric and cholesterol-containing RBC membranes as a model of exploring hemolysis mechanisms of AMPs.

(1) Benfield, A. H.; Henriques, S. T. Mode-of-Action of Antimicrobial Peptides: Membrane Disruption vs. Intracellular Mechanisms. Frontiers in Medical Technology 2020, 2, Mini Review.

(2) Li, J.; Koh, J.-J.; Liu, S.; Lakshminarayanan, R.; Verma, C. S.; Beuerman, R. W. Membrane Active Antimicrobial Peptides: Translating Mechanistic Insights to Design. Frontiers in Neuroscience 2017, 11, Review.

(3) Leontiadou, H.; Mark, A. E.; Marrink, S. J. Molecular dynamics simulations of hydrophilic pores in lipid bilayers. Biophys J 2004, 86 (4), 2156-2164. DOI: 10.1016/s0006-3495(04)74275-7 From NLM.

(4) Hub, J. S.; Awasthi, N. Probing a Continuous Polar Defect: A Reaction Coordinate for Pore Formation in Lipid Membranes. Journal of Chemical Theory and Computation 2017, 13 (5), 2352-2366. DOI: 10.1021/acs.jctc.7b00106.