Novel non-CRBN based molecular glue identification and corresponding E3 ligase target deconvolution through multi-omics analysis

Targeted Protein Degradation (TPD) is considered one of the new drug discovery strategies that could address previously “undruggable” disease targets through the ubiquitin-proteasome system (UPS). Although the current Target Protein Centric approach has proven successful in many cases, especially with well-validated E3 ligases such as CRBN and VHL, there is still a strong need to identify potential target protein degraders (PROTACs, molecular glues, etc.) through both target-centric and target-agnostic approaches. This would not only explore and identify potential tissue-specific novel E3 ligases but also greatly expand the range of potential degradable disease targets. In this study, we first developed a molecular glue-focused library consisting of approximately 5,000 compounds through various medicinal chemistry evaluations. A GSPT1-HiBit cell line and a GSPT1-CRBN-KO counter cell line were developed for primary hit screening purposes. More than 100 potential hits were identified through the initial screening, which included both CRBN-dependent and CRBN-independent hits. A GSPT1-based proximity labeling cell line was then constructed for further target deconvolution purposes via proteomics. Potential candidate E3 ligases were subsequently validated using an array-based expression manipulation strategy. Additionally, we tested novel synthetic lethal target discovery through both in vitro and in vivo screens, followed by omics-based target validation. Using a CRBN-focused glue library, we combined both transcriptomics and proteomics assays to evaluate potential “degradable” targets. To sum up, our study demonstrates that integrated target discovery and target deconvolution strategies could be useful for TPD drug discovery as well as other disease scenarios.