(654e) High-Throughput Activity Reprogramming of Proteases (HARP)

Proteases play a crucial role in various physiological and pathological processes, making the discovery of potent and selective inhibitors a major objective in biomedicine, medicinal chemistry, and chemical biology. Despite decades of research, this endeavor remains arduous, iterative, costly, and often unsuccessful. Traditional small-molecule inhibitors frequently lack selectivity due to their inability to distinguish among the conserved active sites of related proteases. Recently, there has been a growing shift towards protein-based inhibitors, including macrocyclic peptides, ankyrin repeats, antibodies and their fragments (nanobodies, ScFvs, Fabs). Inhibitory macromolecules are not only extremely potent and exquisitely selective but also can uncover inhibition mechanisms not found in natural inhibitors, including allosteric regulation. Unfortunately, current selection and screening methods for these macromolecules are primarily based on binding, requiring extensive downstream biochemical characterization to isolate inhibitors for large pools of binders.

To tackle this challenge, we developed HARP, a yeast-based functional screening platform designed to isolate protease-inhibitory macromolecules from large synthetic libraries. HARP links macromolecule-mediated inhibition of an endoplasmic reticulum (ER)-resident protease target to a robust, quantifiable cell-surface phenotype, enabling selection via fluorescence-activated cell sorting. Using HARP, we successfully identified nanomolar-potency and highly selective inhibitory nanobodies against Tobacco Etch Virus Protease (TEVp) and human Kallikrein 6, including a rare TEVp uncompetitive inhibitor with an inhibitory constant of 7.6 nM. HARP's design principles suggest that conventional binding-first platforms would likely overlook these inhibitors, particularly uncompetitive inhibitors and those with strong inhibition despite moderate binding affinities. To validate our findings, we developed a rigorous characterization pipeline incorporating biochemical assays, binding affinity measurements, deep sequencing, and structural analysis. Notably, coupling HARP with deep sequencing revealed a linear correlation between yeast-derived inhibition phenotypes and in vitro performance, reinforcing HARP’s reliability as a quantitative inhibitor discovery platform. This study introduces HARP as the first yeast-based inhibitor discovery platform of its kind, boasting high dynamic range, precision, and versatility in enzyme targets. For example, we are concurrently working on expanding the HARP concept to other scaffolds (endogenous inhibitors, ScFvs, DARPins, Fabs, cyclic peptides), other protease targets, and other post-translational modification enzymes (kinases, acetyltransferases), thereby positioning it as a premier platform for discovering modulatory macromolecules.