Computational Discovery of Aggregation-Mediated Dipeptide Inhibitors Targeting PHLDA1 for Cardiovascular Therapy.

Cardiovascular disease remains a leading cause of global mortality, underscoring the urgent need for novel therapeutic strategies. This study focuses on the pleckstrin homology-like domain family A member 1-encoded protein (PEP), a regulator of cardiomyocyte apoptosis and a promising yet underexplored therapeutic target. Leveraging an integrated computational approach, we combined AlphaFold3-based structure prediction with extensive molecular dynamics simulations to characterize PEP's dynamic architecture and identify high-affinity dipeptide inhibitors. Our results reveal that PEP possesses significant intrinsic disorder, with molecular dynamics simulations refining its conformation and highlighting force field-dependent behaviors, particularly the superior performance of CHARMM-class force fields in preserving functional secondary structures. High-throughput screening of phenylalanine-based dipeptides identified FF as the strongest binder, exhibiting a unique aggregation-mediated binding mechanism that engages both primary and secondary sites on PEP through multivalent interactions. Concentration-dependent simulations further confirmed the robustness of FF binding and revealed residue-specific interaction hotspots. Notably, we demonstrate that low-confidence regions in AF3 predictions (pLDDT <50) frequently participate in functional binding, challenging the conventional lock-and-key paradigm. These findings not only establish PEP as a tractable drug target but also provide a novel framework for designing aggregation-prone peptide therapeutics against cardiovascular diseases.