Smart Peptide Design of a Novel Melittin-Based Inhibitor of Tyrosinase and MITF Using Machine-Learning Techniques

Melanogenesis is a tightly regulated biological process essential for photoprotection; however, its dysregulation contributes to pigmentary disorders and melanoma progression. Tyrosinase (TYR), the rate-limiting enzyme in melanin synthesis, and microphthalmia-associated transcription factor (MITF), the master regulator of melanocyte function, represent complementary therapeutic targets for effective modulation of melanin production. This study aimed to design melittin-derived peptide inhibitors capable of targeting both TYR and MITF using an integrated computational strategy.

A comprehensive in silico pipeline combining machine learning–guided peptide optimization, protein–peptide structural modeling, molecular docking, MM/GBSA binding free energy calculations, and molecular dynamics simulations was employed. A dual-branch neural network trained on antibody–antigen interaction datasets demonstrated robust predictive performance for IC₅₀ estimation (R² = 0.63; Pearson r = 0.88), enabling efficient pre-screening of 340 rationally designed melittin mutants. High-confidence complexes were generated using AlphaFold-Multimer prior to docking and dynamic evaluation.

Several mutants exhibited improved binding affinity, enhanced structural stability, and favorable interaction energetics compared to wild-type melittin. Mutant 42 (Val5Pro) demonstrated the strongest stabilization toward TYR, whereas Mutant 15 (Val5Asp) and Mutant 68 (Val5Gly) exhibited enhanced interaction stability with MITF. Molecular dynamics analyses revealed mutation-dependent reductions in conformational flexibility, improved peptide retention, enhanced interfacial packing, and favorable energetic stabilization across the selected systems.

The findings demonstrate the feasibility of rationally engineering melittin-derived peptides for dual targeting of major melanogenic regulators. Overall, the proposed computational framework provides a scalable and cost-effective strategy for peptide-based anti-melanogenic inhibitor discovery and establishes a foundation for future experimental validation and therapeutic development in dermatology, cosmetic science, and melanoma-associated research.