Abstract
Immunoglobulin light chain (AL) amyloidosis is the most common form of systemic amyloidosis. The disease correlates with the formation of insoluble aggregates mostly composed of patient-specific antibody light chains, whose hypervariable regions make each case unique and highlight the need for personalized therapeutics. In this study, we focused on a pathogenic homodimer we previously obtained from a patient-derived light chain. By analyzing the dynamics and the interface of this dimer, we identified a peptide with potential inhibitory activity. The peptide was then refined using a computational mutagenesis protocol that iteratively improved its sequence to maximize complementarity with the protein interface, taking into account shape, electrostatics, and hydropathy. The resulting optimized peptide is found to bind the monomer with a binding affinity comparable to that of the full pathogenic interface. These results suggest that the designed peptide could act as an effective antagonist of the pathogenic dimer, and demonstrate that our computational strategy could provide a general framework for designing patient-specific inhibitory peptides against aggregation-prone proteins.