Cell Morphological Profiling Enables High-Throughput Screening for PROteolysis TArgeting Chimera (PROTAC) Phenotypic Signature

Abstract

PROteolysis TArgeting Chimeras (PROTACs) use the ubiquitin–proteasome system to degrade a protein of interest for therapeutic benefit. Advances made in targeted protein degradation technology have been remarkable, with several molecules having moved into clinical studies. However, robust routes to assess and better understand the safety risks of PROTACs need to be identified, which is an essential step toward delivering efficacious and safe compounds to patients. In this work, we used Cell Painting, an unbiased high-content imaging method, to identify phenotypic signatures of PROTACs. Chemical clustering and model prediction allowed the identification of a mitotoxicity signature that could not be expected by screening the individual PROTAC components. The data highlighted the benefit of unbiased phenotypic methods for identifying toxic signatures and the potential to impact drug design.

Advancing Targeted Protein Degradation via Multiomics Profiling and Artificial Intelligence

“…In this context, identification of degraders could be enabled by searching for compounds that trigger a cellular response that is functionally linked to the ubiquitin-proteasome system or any other protein degradation pathway. Indeed, PROTAC phenotypic signatures have already been identified through cell morphological profiling …”

Using Transcriptomics and Cell Morphology Data in Drug Discovery: The Long Road to Practice

“…Also combining information from multiple domains, such as chemical structure and imaging information, has been performed using Bayesian matrix factorization (BMF Macau) methods and compared to Random Forest . It was found that both methods performed similarly when ECFP fingerprints were used as compound descriptors.…”

“…For new modalities, such as PROteolysis TArgeting Chimeras (PROTACs), it is not clear how to profile compounds for safety, and the Cell Painting assay has been evaluated for this purpose recently as well. 66 It was found that the signal contained in Cell Painting readouts with respect to mitochondrial toxicity of PROTACs was concentrationdependent, with 10 μM and 1 μM offering better signals for model generation than a concentration of 0.1 μM, and with the model based on 1 μM concentration data offering virtually perfect classification on a prospective test set.…”

Self-Supervised Learning of Phenotypic Representations from Cell Images with Weak Labels

“…The features extracted from these kinds of images are used to build profiles which can assess bioactivity and MOA of the cells in response to treatments with chemical compounds and/or genetic perturbations. In drug discovery these phenotypic profiles are used to compare new treatments with known ones, and downstream applications include routine lead compound identification, drug target screening through CRISPR technology and toxicity assessment [13].…”