MA(R/S)TINI 3: An Enhanced Coarse-Grained Force Field for Accurate Modeling of Cyclic Peptide Self-Assembly and Membrane Interactions

Abstract

Self-assembled nanotubes (SCPNs) formed by alternating chirality α-Cyclic Peptides (d,l-α-CPs) have presented interesting biological applications, such as antimicrobial activity or ion transmembrane transport. Due to difficulties to follow these processes with experimental techniques, Molecular Dynamics (MD) simulations have been commonly used to understand the mechanism that led to their biological activity. However, the high computational cost of atomic resolution simulations makes them unsuitable for simulating dynamic processes involving multiple units like their self-assembly in different environments. In this regard, coarse-grain (CG) models might represent a more feasible option. However, general coarse-grained force fields such as MARTINI do not explicitly account for noncovalent interactions, such as hydrogen bonding, which are essential for secondary structure formation and the self-assembly of proteins and peptides. This problem becomes particularly important when simulating CPs due to the specific directionality of their interactions. In a previous work, it has been proven how MARTINI classical parametrization overestimated the self-assembly of CPs not distinguishing parallel and antiparallel interactions as well as allowing forbidden rotational angles. The so-called MA(R/S)TINI force field fixed the problem by including two extra particles into the backbone bead while preserving the behavior of several CP sequences in the presence of different membrane models. However, this new parametrization presented a much higher CP–CP interaction energy, being another critical issue for self-assembly overestimation. The release of MARTINI 3 expanded the scope of the force field by introducing new particles and labels specifically tailored to improve the representation of noncovalent interactions. Nevertheless, since it uses the same mapping strategy for protein backbones, this new version also failed at capturing the specific directionality of CPs. Taking advantage of the new possibilities offered by MARTINI 3, MA(R/S)TINI has been updated in the present work. This new version uses a new mapping of CPs based on original beads of the force field and releases the restraints previously imposed on the lateral side chains of the CPs. This new parameterization fixes the formerly overestimated interaction energy between CPs in both parallel and antiparallel orientations, while preserving the advantages of the previous version of MA(R/S)TINI. The new parametrization provided in the present work is aimed to facilitate the understanding, design, and optimization of new bioactive CPs based on CG-MD simulations.