Enhanced and Classical MD Simulations Map Aggregation-Prone Regions in the Olfactomedin Domain of Myocilin

This study provides a comprehensive computational map of the aggregation-prone regions (APRs) in the olfactomedin (OLF) domain of human myocilin through an integrative approach combining classical MD at multiple temperatures, Gaussian-accelerated MD with three boost modes, and chemical denaturant simulations totaling 21 μs. Two structure-based aggregation propensity predictors identified five APRs enriched in hydrophobic and aromatic residues (residues 246-260, 276-283, 301-317, 364-379, and 430-451). Among these, the APR spanning residues 430-451 coincided with an experimentally amyloidogenic peptide of OLF. Furthermore, we found that the blade A interfaces (AB and AE) served as critical sites for destabilization, showing significantly earlier loss of contacts compared to other interfaces. Their destabilization was associated with two APRs, regions 276-283 and 301-317. All of the identified APRs are enriched in aromatic/hydrophobic amino acids that consistently became more solvent-exposed, with altered surface electrostatics in representative MD conformations compared to the crystal structure, suggesting that their exposure may act upstream of the OLF aggregation. Our extensive simulations linked the dynamics of the OLF blades to its aggregation propensity, highlighting the potential of the identified APRs as therapeutic targets to prevent myocilin aggregation.