Differentiation of Aneugens and Clastogens in the In Vitro Micronucleus Test by Kinetochore Scoring Using Automated Image Analysis

Abstract

The in vitro micronucleus test according to OECD Test Guideline 487 (TG 487) is widely used to investigate the genotoxic potential of drugs. Besides the identification of in vitro genotoxicants, the assay can be complemented with kinetochore staining for the differentiation between clastogens and aneugens. This differentiation constitutes a major contribution to risk assessment as especially aneugens show a threshold response. Thus, a novel method for automated MN plus kinetochore (k+) scoring by image analysis was developed based on the OECD TG 487. Compound‐induced increases in MN frequency can be detected using the cytokinesis‐block (cytochalasin B) method in V79 cells after 24 h in a 96‐well format. Nuclei, MN, and kinetochores were labeled with nuclear counterstain and anti‐kinetochore antibodies, respectively, to score MN in binuclear or multinuclear cells and to differentiate compound‐induced MN by the presence of kinetochores. First, a reference data set was created by manual scoring using two clastogens and aneugens. After developing the automated scoring process, a set of 14 reference genotoxicants were studied. The automated image analysis yielded the expected results: 5/5 clastogens and 6/6 aneugens (sensitivity: 100%) as well as 3/3 non‐genotoxicants (specificity: 100%) were correctly identified. Further, a threshold was determined for identifying aneugens. Based on the data for our internally characterized reference compounds, unknown compounds that induce ≥53.8% k+ MN are classified as aneugens. The current data demonstrate excellent specificity and sensitivity and the methodology is superior to manual microscopic analysis in terms of speed and throughput as well as the absence of human bias. Environ. Mol. Mutagen. 60:227–242, 2019. © 2018 Wiley Periodicals, Inc.

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

“…With the goal to develop a predictive micronucleus assay that is also providing mechanistic information, we decided to base the design F I G U R E 8 (Continued) Bryce et al, 2016;Bryce et al, 2017;Wilde et al, 2019;Wilson et al, 2021).…”

“…On the other hand, aneugens induce increases of micronuclei containing a kinetochore signal (i.e., CENPA + MN) with aneugenic tubulin binders induce a mitotic arrest (i.e., increase in pH 3 and MPM2‐positive cells) at the 4‐h timepoint and compound that act through aurora inhibition are characterized by large increases in polyploid cells. Based on our extensive experience with slides, flow, and imaging‐based micronucleus screens, we believe that this approach might have several advantages over the combination of MicroFlow® and MultiFlow™ assays that have tried to provide higher throughput micronucleus capabilities combined with mechanistic endpoints (Bryce et al, 2013; Bryce et al, 2016; Bryce et al, 2017; Wilde et al, 2019; Wilson et al, 2021).…”

“…In addition, it does not provide the necessary mechanistic information to decide whether a compound or chemical series could be progressed further in development. However, the endpoint is amenable to automation by flow cytometry (Bryce et al, 2008; Bryce et al, 2010; Bryce et al, 2011; Bryce et al, 2013) and image analysis (Wilde et al, 2019; Wilson et al, 2021), which allows moving toward more high‐throughput and high‐content approaches to gain statistical power and enable quantitative dose–response analysis. In addition, newer assay platforms aim to provide more mechanistic information to guide decision‐making and risk‐management strategies.…”

“…Using a customized Cellomics ArrayScan™ VTi image analysis algorithm for automated MN scoring, the assay provided good concordance to microscopic scoring in a test set of 46 compounds. Subsequently, Wilde et al (2019) advanced a high throughput version of cytokinesis‐block micronucleus test that integrates with kinetochore staining to aid in compound classification. The customized MetaXpress™ image analysis algorithm identifies bi‐ and multinucleated cells by calculating the distance from neighboring nuclei.…”

“…Wilde et al (2019) andWilson et al (2021) that use a single timepoint.Although the image analysis approach byWilde et al (2019) provides information on MN frequency and kinetochore staining is helpful in classifying unknown compounds, these established protocols lack biomarkers detection that is crucial for understanding the MOA for many compounds. Especially for weak inducers of micronuclei or in situations where the micronucleus induction can be confounded by large increases in apoptosis.…”