This report details a strategic Structure-Based Drug Design (SBDD) project focused on developing novel inhibitors of UBE2N to address the significant clinical challenge of PARP inhibitor (PARPi) resistance in ovarian cancer. The project utilizes an integrated computational workflow, including BioSolveIT's SeeSAR, InfiniSee, FastGrow and HYDE assessment for rapid virtual screening and interactive lead design, followed by Molecular Dynamics (MD) simulations for rigorous validation. The goal is to identify new non-covalent potentially and stable UBE2N inhibitors for subsequent cell-based testing, ultimately contributing to more effective combination therapies for ovarian cancer patients who have developed resistance to PARP inhibitors. The next planned step is to advance the selected 10 candidates to purchased and do wet lab for binding affinity and cell-based assays. This work is expected to contribute to the development of more effective combination therapies for ovarian cancer patients who have become resistant to PARP inhibitors.
After 1 year, Shafi Ullah has achieved the following goals:
- Identify Initial Hits through High-Throughput Virtual Screening: This initial phase aimed to identify a pool of potential hit compounds through a combination of two complementary virtual screening methodologies: • A. Ligand-Based Virtual Screening (LBVS) o A library of potential inhibitors was compiled, guided by known UBE2Ni inhibitors such as ML307, Variabine B, and NSC697923. o Five large commercial databases (REAL Space, eXplore, Freedom Space, CHEMriya, and GalaXi) were screened. o The top 500 hits from each database were retained, and after removing duplicates, a final unique library of 2500 compounds were prepared from around 5 trillions compounds. • B. Detailed Binding Site Approach o Structural analysis of the UBE2Ni protein was initiated using SeeSAR. o From 11 identified potential PDB files in RCSB, seven PDBs were selected for further analysis after duplicate removal. o A total of 11 binding sites were identified across the original 11 files, most of which were found to contain covalent ligands in the form of small moiety instead of full ligand. o A novel tool Unoccupied pocket in SeeSAR was employed to identify unexplored binding pockets, leading to the detection of three distinct pockets which were previously unexplored for UBE2Ni designing. o A multi-stage analysis was conducted to select the most realistic and promising pocket for subsequent docking studies, involving individual analysis of each pocket, combinations of two pockets, and a comprehensive analysis of all three pockets, focused was on the crucial amino acid reported in literature for binding different partner especially in the vicinity of Cys87. • C. Structure-Based Virtual Screening (SBVS) o The 2500-compound library obtained from LBVS was screened against the UBE2N protein using an SBDD approach, targeting the most promising binding pocket identified from the prior detailed UBE2N structural analysis. o Initial top 500 hits were identified based on their HYDE assessment scores. o This list was further refined to a final set of 100 high-quality hit compounds, prioritizing key interactions with crucial amino acids in the active pocket, particularly in the vicinity of Cys87.
- Optimize and Design Lead Candidates: The 100 compounds identified from the initial screening served as the first set of hit candidates. This phase focused on optimizing and rationally designing these hits to improve their binding affinity properties. The process involved a detailed analysis of the top-ranking hits' scaffolds and the utilization of the FastGrow Inspirator tool to design new derivatives along with ADME assessment using Optibrium. While some newly designed compounds showed enhanced affinity, the majority did not, which may be due to predefined screening criteria from earlier steps.
- Rigorously Validate Leads: The final phase involves rigorous validation of the optimized lead candidates to minimize false positives and maximize translatability. This validation was conducted by Molecular Dynamics (MD) simulations of 25 selected compounds as leads for 100ns.