Small molecule inhibitors of protein-protein interactions
The development of orally bioavailable small molecule drugs targeting protein-protein interactions (PPIs) has been challenging1. Unlike conventional targets, PPIs’ extended, open surface makes it difficult for small molecules to bind. In order to achieve strong binding, it is frequently necessary to use larger molecules, which traditionally is considered to disfavor druglikeness2. However, PPIs possess great therapeutic potential due to their abundance and regulatory roles in cells3. More extensive studies are needed to identify larger chemotypes that retain good druglike properties and therefore might have utility against PPI targets. NF-κB Essential Modulator (NEMO), interacting with IκB Kinase subunit β (IKKβ), is an important PPI target because of its regulatory role in NF-κB signaling4. Literature suggests that the N-terminal domain of NEMO is intrinsically disordered in the absence of bound ligand5. To test this hypothesis, I developed variants of the NEMO N-terminal domain, and studied their secondary structure, stability, and affinity for IKKβ, showing that the N-terminal domain of NEMO is intrinsically structured (Chapter Two). I also characterized partially peptidic NEMO inhibitors from our collaborator, Carmot Therapeutics. We tested the binding of these compounds and their peptidic fragments to full-length NEMO using fluorescence anisotropy (FA)6 and surface plasmon resonance (SPR). The results provided information about hit validity, binding affinity and kinetics (Chapter Three). Macrocycles are of interest for inhibiting PPIs partly because of their proposed good membrane permeability7. To evaluate this hypothesis, I implemented a membrane permeability assay, tested the permeability of a set of macrocyclic compounds, and used the results to develop a multiple linear regression model to predict permeability from macrocycles’ physicochemical properties. The model suggests that hydrophobicity correlates positively with good permeability, while high polarity or high aromatic ring count renders macrocycles less permeable (Chapter Four). Finally, in a separate project, to elucidate the origins of protein-ligand binding energy between interleukin-2 (IL-2) and its known small molecule inhibitors8, I developed a SPR based binding assay, and validated it by showing that the KD value of known inhibitor Ro26-45508 agrees with the literature value (Chapter Five). The assay will be useful in future studies of IL-2 inhibitors and their fragments.