Virtual Screening in Search of Inhibitors of UBR1 E3 Ligase

Multiple myeloma (MM) is a haematological malignancy that accounts for approximately 10% of all neoplasms of this type. The disease is characterized by a clonal proliferation of plasma cells, leading to an abnormal increase in monoclonal immunoglobulins in the serum. Initial therapy for active MM involves a combination of drugs, including proteasome inhibitors, immunomodulatory agents and corticosteroids. The management of relapsed/refractory MM is more complex as MM does not follow a single pattern of oncogenic pathway alteration. Although advances have significantly improved prognosis, patients have a five-year survival rate of around 60%. For this reason, MM remains a malignancy in continuous therapeutic evolution, and the current challenge is to develop new strategies that reduce minimal residual disease. In this sense, PROTAC (PROteolysis TArgeting Chimera) is part of the paradigm shift adopted in recent decades, which is based on targeted protein degradation and exploits the mechanism of the ubiquitin-proteasome system. It allows proteins to be degraded through the action of an enzymatic system followed by destruction within the 26S proteasome. This technology consists of a synthetic molecule composed of a warhead and an anchor, connected by a linker. This project aimed to select potential compounds that could successfully bind to the UBR1 E3 Ligase protein, in order to constitute the anchor component of a PROTAC. To achieve this goal, a combination of computational methods was employed. Since the crystallographic structure of UBR1 has not yet been fully resolved, the UBR-box domain was modeled by homology, using the 3NY3 structure as template. In order to obtain more meaningful results, the model of the entire UBR1 protein predicted by AlphaFold2 was also taken into consideration. Molecular dynamics simulations were performed on both models to extract the most representative structures of the trajectory for use in subsequent stages. Potential binding sites that could accommodate the ligands were then identified. After selecting the most promising pocket on the homology model of the UBR-box domain of UBR1 from among those found, also supported by visualization of the electrostatic potential map, a structure-based pharmacophore model was constructed. The virtual screening phase was carried out using the NCI Diversity Set and the pharmacophore hypothesis was used as a filter. The hits found were preliminarily analyzed in terms of pharmacokinetic properties and safety profile, which allowed the identification of 10 potential ligands. Molecular docking represented another key step, enabling the evaluation of the binding affinity between the UBR1 E3 Ligase target and the potential ligands. To support a more robust selection of final candidates, the most promising hits were used as a starting point for a 95% similarity search on ZINC-22, which led to the construction of a docking set of more than 130 compounds. The properties of all of them were then analyzed in detail. Through the evaluation of the features and the results obtained in the docking phase, 28 compounds were selected, divided by priority, for the subsequent experimental testing phase.