Abstract PO-048: Visual nucleotyping identifies chromatin phenotypes triggered by genome editing

Le résumé fourni par la source

Abstract Optical microscopy has the potential to provide rapid phenotypic readout of cellular states. Here we show that automated optical phenotyping of nuclei is capable of rapidly and reliably discriminating the effects of targeted mutations to chromatin remodelers, which are major players in oncogenesis and targets for therapeutic intervention. Smarca4 is a chromatin remodeler frequently mutated in cancer. We investigated whether nuclear morphometry and concomitant supervised machine learning approaches on standard optical microscopy images of nuclear-DNA stained cells could reliably discern Smarca4-/- lung cancer cells from isogenic Smarca4+/+ clones that were derived by precision genome editing. We used TALEN-based editing on A549 lung cancer cells to generate isogenic Smarca4+/+ clones, performed conventional 2D 40x optical imaging of DAPI-stained cell nuclei from parental (Smarca4-/-) and edited (Smarca4+/+) cells, quantified a set of numerical parameters (features) that reflected various intuitive attributes of nuclear morphology and chromatin texture of the cell nuclei, and finally tested the efficacy of supervised machine learning algorithms containing these features to distinguish between the two cell categories. We found that the parental and edited A549 cells exhibited similar nuclear size and shape attributes but showed measurable differences in subtle chromatin texture. Additionally, classifiers based on these chromatin texture features were able to accurately distinguish the smarca4-rescued cells from the smarca4-mutant cells. Our results thus demonstrate the promise of quantitative chromatin phenotyping for genome editing applications in oncology. Citation Format: Vivek Nandakumar, Sandra Stehling-Sun, William Kerwin, Alister Funnell, John Stamatoyannopoulos. Visual nucleotyping identifies chromatin phenotypes triggered by genome editing [abstract]. In: Proceedings of the AACR Virtual Special Conference on Artificial Intelligence, Diagnosis, and Imaging; 2021 Jan 13-14. Philadelphia (PA): AACR; Clin Cancer Res 2021;27(5_Suppl):Abstract nr PO-048.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.