A new imaging platform (iScreen) allows for the concurrent assessment of micronucleus induction and genotoxic mode of action in human A375 cells

• Published in: Environmental and molecular mutagenesis 2022-06, Vol.63 (5), p.230-245
• Main Authors: Sun, Xiaowen, Rubitski, Elizabeth, Spellman, Richard A., Engel, Maria, Schuler, Maik
• Format: Article
• Language: English
• Subjects: Algorithms; aneuploidy; aurora kinase inhibitor; Cell lines; Centromere protein A; clastogen; Clastogenicity; Cytotoxicity; Genotoxicity; Genotoxicity testing; Histone H2A; Histones; Image analysis; Image processing; Kinases; Mode of action; Polyploidy; Protein A; Staining; Toxicity; Tubulin; tubulin binder
• ISSN: 0893-6692; 1098-2280
• DOI: 10.1002/em.22496

Genotoxicity testing guidelines require the assessment of the clastogenic and aneugenic potential of compounds. While in vitro micronucleus assays detect both types of endpoints, it requires labor‐intensive microscopic scoring and does not discriminate between the two modes of actions. Here, we present a novel high‐content imaging platform in A375 human cells that addresses the need for rapid scoring while providing additional mechanistic information. We evaluated the new platform with 12 compounds, three compounds from each mechanistic class (clastogen, aneugen tubulin binder, aneugen aurora inhibitor, and nongenotoxicant) following 4‐ and 24‐h compound treatments. The approach we developed is first discriminating between genotoxicant and nongenotoxicant using an image analysis algorithm to quantify micronucleus induction below a 60% cytotoxicity cutoff. Then it uses centromere protein A (CENPA) staining for the genotoxic compounds to discriminate between aneugens and clastogens. Lastly, we use phosphorylated histone H2AX Ser139 (γH2AX) staining to confirm clastogenicity and changes in phosphorylated histone 3 Ser10 (pH 3) and increases in polyploidy in mitotic cells to discriminate between aneugens that bind tubulin from those that affect aurora kinases. All compounds were correctly classified, and we showed by using benchmark dose–response analysis that the imaging platform in A375 cells is at least as sensitive as the MicroFlow® assay in TK6 cells for genotoxicant but appears to be more specific for the nongenotoxicants. A detailed comparison of the cell lines and a more comprehensive validation with a much larger compound set, predictive and dose–response modeling will be presented in the future.