Computational investigation of myoferlin-C2D-targeting potent peptides

Abstract

Myoferlin (MYOF) is a multidomain protein involved in cellular membrane dynamics and vesicular processes. Its C2D domain has emerged as a region of interest for molecular targeting of MYOF. However, the systematic design and molecular-level characterization of peptide candidates targeting C2D remain limited. Here, we aimed to computationally design short peptide candidates targeting the MYOF-C2D domain and evaluate their interactions with C2D. De novo peptide candidates were generated using RFdiffusion, followed by docking analysis, molecular dynamics simulations, and binding free energy calculations. Among 10,000 designs, 100 candidates were selected for docking, and the 10 candidates identified from the docking results were further evaluated using 100-ns MD simulations. Eight candidates maintained their interactions with C2D throughout the simulation, whereas two failed to form stable complexes. Among the peptide candidates, SIENIAAY exhibited the most favorable calculated MM/PBSA binding free energy and maintained its interaction with C2D throughout the simulation, supported by multiple polar and hydrophobic contacts. However, while SIENIAAY showed a predicted negative The blood–brain barrier (BBB) permeability, the second-ranked peptide, DLVAELYRKLLE, showed a positive BBB permeability prediction. Therefore, C2D-binding potential and predicted BBB accessibility were considered together, and DLVAELYRKLLE was selected as a potential peptide candidate targeting the MYOF-C2D domain.