Designing Ligands for Structure-Less Proteins

Abstract

Disordered proteins can play physiologically important roles, but can also lead to major diseases especially when they aggregate. Designing small molecule / peptide ligands for disordered proteins is a difficult challenge, since structure cannot be used as a guide. Here we demonstrate effective approaches for three major cases: 1) for partially disordered proteins, we identify key structured portions and design ligands for disrupting those, 2) for mostly disordered proteins, we measure spontaneous fluctuations and use that as an assay to quantify transient interactions with candidate ligands, and 3) for unfolded states of well-folded proteins, we use the folding nucleus to design an artificially stabilized exo-nucleus which can induce folding. We demonstrate the first strategy with amyloid beta (Aβ), whose aggregation is associated with Alzheimer's disease. Using a combination of fluorescence correlation spectroscopy (FCS), Forster resonance Energy transfer (FRET) and solid state NMR, we show that a small peptide designed to specifically disrupt a distal intra-molecular contact can drastically alter the membrane affinity of the Aβ oligomers. We demonstrate the second strategy with IAPP (amylin), whose aggregation is associated with Type II diabetes. We show that ns-timescale spontaneous fluctuations of IAPP, measured using FRET and FCS, can provide a sensitive assay for interactions of IAPP oligomers with small molecule drug candidates. The third case is demonstrated by designing small peptides which can specifically perturb the folding kinetics of GdnHCl-unfolded ubiquitin. We construct a chemically stabilized peptide fragment that mimics the folding nucleus of ubiquitin. We show that this peptide can accelerate the folding of ubiquitin. Overall, here we demonstrate “dynamics” based methods which can help in designing drugs for disordered proteins.