Alpha-synuclein (aSyn), a presynaptic intrinsically disordered protein whose aggregation and accumulation of its fibrillar form in intracellular inclusion bodies is a hallmark of synucleinopathies, a group of neurodegenerative disorders (NDs) that includes Parkinson's disease and multiple system atrophy. Although the etiology of NDs remain elusive, growing evidence point toward the critical roles of aSyn fibrils in disease progression and their prion-like disease propagation. Besides, their early involvement in disease pathogenesis, aSyn fibrils represent promising targets for early-stage diagnostics and therapeutic interventions in synucleinopathies.

Advances in cryogenic electron microscopy (cryo-EM) have enabled the resolution of aSyn fibrils structures derived from both patient tissue and in vitro preparations, revealing a diverse landscape of fibrillar folds. Notably, distinct synucleinopathies have been associated with specific fibril conformations, offering opportunities for disease differentiation through the development of polymorph-selective binders. However, designing such molecules remains challenging due to the limited availability of patient-derived fibrils, structural divergence between pathological and in vitro fibrils, and the difficulty of reproducing disease-relevant folds in vitro.

This dissertation leverages computational strategies to address key challenges in amyloid biology and protein binder design. In the first aim, we developed FibrilSite, a surface-level comparative analysis framework for amyloid fibrils. This tool bridges the structural gap between the distinct in vitro and pathological aSyn fibrils and guides structure-based drug discovery toward conserved, targetable sites while avoiding regions prone to cross-reactivity with fibrils from other amyloidogenic proteins. In the second aim, we established a de novo binder design pipeline tailored for amyloid fibrils, successfully generating the first computationally designed binders against aSyn fibrils with demonstrated target and/or polymorph specificity. In the third aim, we explored the potential of implementing state-of-the-art ML-based protein design tools to enhance the experimental success of our in-house binder design pipeline, MaSIF-seed, providing practical insights into their current capabilities and limitations for mini-protein binder development. Together, these contributions lay the foundation for future efforts to develop structure-guided, polymorph-specific molecules for amyloid targeting, offering a promising path toward early-stage differential diagnosis and more precise disease-modifying strategies for neurodegenerative diseases.