A high content screening platform for live imaging and deep learning-assisted quantitative assessment of 3D cell culture models of hepatotoxicity – DEEP-HEPATOSCREEN

3D cell cultures have become the gold standard in cell biology. By recapitulating the functions and the cellular architecture of tissues in a miniature version, they represent a good alternative to animal models and have promising applications in regenerative medicine, cell therapy and toxicology. Yet, current quality control approaches based on genomics or imaging are ill-adapted because intrinsically low throughput, and generally expensive and destructive.
DEEP-HEPATOSCREEN aims at overcoming this bottleneck by developping a ground-breaking high content screening instrument dedicated to 3D cell cultures, integrating miniaturized 3D cell-culture holders and artificial intelligence-based quantitative phenotypic analysis.
We propose to focus on 3D cell cultures relevant for hepatotoxicity assays, and perform a proof-of-concept dose-response study to investigate the effect of paracetamol, which can cause severe and irreversible liver damage when absorbed at high concentration.
To achieve this goal, we will perform massively parallelized 3D live monitoring of hepatic 3D cell culture models stably expressing histone H2B-GFP, under different drug concentrations, from which quantitative nuclear profiling and molecular footprints will be extracted and used for the phenotypic classification by artificial intelligence.
DEEP-HEPATOSCREEN relies on cutting edge core technologies and know-how developed by our consortium, including quantitative microscopy, artificial intelligence, proteomics, bioinformatics and cell biology.
We expect our project to become a breakthrough to organotypic cell culture research and drug discoveries.