Discovery and Development of Novel Deubiquitinase Inhibitors via Parallel High-Throughput Screening

Abstract

Ubiquitination is a critical post-translational modification in cell biology which results in a myriad of functional outcomes including proteasomal degradation of substrate proteins. The regulation of protein homeostasis based on ubiquitination state is fundamental for cell function and survival. Ubiquitin is appended onto proteins substrates via the concerted action of an E1-E2-E3 enzymatic cascade and the removal of these ubiquitin marks is catalyzed by the action of deubiquitinating enzymes (DUBs). DUBs are a class of isopeptidases subdivided into seven families based on sequence homology and common folds within their catalytic domains. They are responsible for the regulation of ubiquitin dynamics through catalytic cleavage of ubiquitin from protein substrates as well as maintaining the cellular ubiquitin pool through processing of ubiquitin precursors and free ubiquitin chains. As ubiquitination is involved in numerous cellular processes, dysregulation of the ubiquitin system leads to the formation and progression of multiple diseases. The ubiquitin-proteasome system (UPS) has been the source of extensive research since its discovery to better understand the functional role of all the molecular players involved in this complex system. While multiple diseases can be tied to dysfunction and aberrant regulation of different components of this system, there are relatively few validated chemical probes or clinically approved therapies to further understand their role in disease. While recent advancements have led to the development of novel therapies targeting components of the UPS such as proteasome inhibitors and degraders which utilize E3 ligases for the depletion of neo-substrates, therapeutics targeting DUBs has been limited. Despite being linked to many diseases and possessing defined catalytic domains with putative sites for small molecule inhibitor binding, few selective small molecule inhibitors and no approved DUB drugs currently exist. A common pitfall of current DUB inhibitor development has been the lack of selectivity of hit compounds identified in high throughput screening (HTS) campaigns for individual DUBs within and among the various DUB families. This lack of selective DUB inhibitors that can be utilized as probes to further explore DUB biology has hindered the understanding of DUB function and relevance in disease. This prompted us to redesign a small molecule screening cascade which prioritized selectivity along with compound potency in order to establish a broader framework for future DUB HTS development across this gene-family. Using a fluorogenic ubiquitin-rhodamine assay previously utilized for assessing DUB activity, we optimized and carried out a high-throughput screen of a ~50,000 member, chemically diverse small molecule library against eight different DUBs, spanning three well-characterized DUB families. Hit compounds were confirmed in dose-response as well as through a series of counter-screens and orthogonal assays including expanded selectivity assessments against a larger panel of DUBs. Through these efforts we have generated a large DUB-ligand interaction dataset that has been made publicly available for future DUB inhibitor development. This work also led to the identification of multiple highly selective DUB inhibitors and provide a roadmap for rapidly identifying and validating selective inhibitors of related enzymes. From the panel of DUBs screened, USP28, had multiple potent and selective analogs which were prioritized for medicinal chemistry efforts to improve potency and better understand compound activity. Through these efforts we generated a co-crystal structure of one analog bound USP28, representing the first USP28 co-crystal structure with a small molecule that has been obtained to date. Mutating amino acid residues in the binding pocket has led to the identification of key residues that mediate compound activity. These results provide tool compounds and resistance mutants to enable better exploration of USP25 and USP28 function.