researcher Johns Hopkins University, United States
Introduction/Rationale: Aggregation of α-synuclein into amyloid fibrils is a key pathological feature of Parkinson’s disease and other synucleinopathies. Existing inhibitors of α-synuclein aggregation, including small molecules and short peptides, exhibit limited binding affinities and specificity, restricting their therapeutic potential. To address this gap, we developed a de novo protein design pipeline to generate high-affinity inhibitors targeting the steric zipper interface of α-synuclein fibrils.
Methods: The active site of α-synuclein was identified through literature review and structural inspection and used as input for RFdiffusion to generate both unconstrained and scaffolded backbones. ProteinMPNN then inferred amino acid sequences for each backbone, producing over 3,000 unique candidates. AlphaFold-Multimer was employed to predict inhibitor-fibril complexes, and candidates were filtered based on structural stability and interfacial complementarity. Top sequences were further refined and ranked using HADDOCK and PRODIGY docking analyses to estimate binding affinities.
Results: Our best designs exhibited dissociation constants (Kd) as low as 2.8 × 10⁻¹² M, demonstrating significantly stronger predicted binding than previously reported peptide inhibitors such as KISFRV (Kd = 7.5 × 10⁻⁵ M). Notably, the top-performing binders displayed diverse sequence motifs and binding modes, indicating that our design framework effectively explores a wide conformational and chemical space. These results highlight the potential of generative protein design to create tailored inhibitors against amyloids with prion-like propagation behavior.
Conclusion: While the therapeutic translation of these designs remains constrained by challenges such as blood-brain barrier permeability, our study establishes a scalable computational strategy for designing peptide-based inhibitors with high affinity and specificity. Future work will focus on experimentally validating binding and exploring methods to disrupt pre-formed fibrils.