Flow cytometric micronucleus assay and TGx-DDI transcriptomic biomarker analysis of ten genotoxic and non-genotoxic chemicals in human HepaRG™ cells

Abstract

Background: Modern testing paradigms seek to apply human-relevant cell culture models and integrate data from multiple test systems to accurately inform potential hazards and modes of action for chemical toxicology. In genetic toxicology, the use of metabolically competent human hepatocyte cell culture models provides clear advantages over other more commonly used cell lines that require the use of external metabolic activation systems, such as rat liver S9. HepaRG™ cells are metabolically competent cells that express Phase I and II metabolic enzymes and differentiate into mature hepatocyte-like cells, making them ideal for toxicity testing. We assessed the performance of the flow cytometry in vitro micronucleus (MN) test and the TGx-DDI transcriptomic biomarker to detect DNA damage-inducing (DDI) chemicals in human HepaRG™ cells after a 3-day repeat exposure. The biomarker, developed for use in human TK6 cells, is a panel of 64 genes that accurately classifies chemicals as DDI or non-DDI. Herein, the TGx-DDI biomarker was analyzed by Ion AmpliSeq whole transcriptome sequencing to assess its classification accuracy using this more modern gene expression technology as a secondary objective.

Methods: HepaRG™ cells were exposed to increasing concentrations of 10 test chemicals (six genotoxic chemicals, including one aneugen, and four non-genotoxic chemicals). Cytotoxicity and genotoxicity were measured using the In Vitro MicroFlow® kit, which was run in parallel with the TGx-DDI biomarker.

Results: A concentration-related decrease in relative survival and a concomitant increase in MN frequency were observed for genotoxic chemicals in HepaRG™ cells. All five DDI and five non-DDI agents were correctly classified (as genotoxic/non-genotoxic and DDI/non-DDI) by pairing the test methods. The aneugenic agent (colchicine) yielded the expected positive result in the MN test and negative (non-DDI) result by TGx-DDI.

Conclusions: This next generation genotoxicity testing strategy is aligned with the paradigm shift occurring in the field of genetic toxicology. It provides mechanistic insight in a human-relevant cell-model, paired with measurement of a conventional endpoint, to inform the potential for adverse health effects. This work provides support for combining these assays in an integrated test strategy for accurate, higher throughput genetic toxicology testing in this metabolically competent human progenitor cell line.

Keywords: Genetic toxicology; Micronucleus; RNA-Seq; TGx-28.65 genomic biomarker; Toxicogenomics.

Fig. 1

Cytotoxicity assessment in human HepaRG™ cells following exposure to: (a) DDI chemicals in μM concentrations; and (b) non-DDI chemicals in mM concentrations (except COL, which was in μM) using the In Vitro MicroFlow® assay (Litron Laboratories). See Table 1 for specific concentrations (C1 = lowest concentration and C6 = highest concentration). Percent relative survival is depicted 96 h following the last exposure (n = 2). DDI chemical abbreviations: 2-nitrofluorene (2NF), cisplatin (CISP), etoposide (ETP), aflatoxin B1 (AFB1), and methyl methanesulfonate (MMS). Non-DDI chemical abbreviations: 2-deoxy-D-glucose (2DG), sodium chloride (NaCl), ampicillin trihydrate (AMP), sodium ascorbate (ASC), and colchicine (COL). Control represents the vehicle control (DMSO for 2NF, CISP, ETP, AFB1, and COL; water for MMS; media for 2DG, NaCl, AMP, and ASC). Error bars depict standard error, but are too small to see in all but one data point. * P < 0.05 compared to the vehicle control

Fig. 2

Measurement of MN frequency in human HepaRG™ cells following exposure to: (a) DDI chemicals in μM concentrations; and (b) non-DDI chemicals in mM concentrations (except COL, which was in μM) using the In Vitro MicroFlow® assay (Litron Laboratories). Percentage of MN induction is depicted 96 h following the last exposure (n = 2). See Table 1 for specific concentrations (C1 = lowest concentration and C6 = highest concentration). DDI chemical abbreviations: 2-nitrofluorene (2NF), cisplatin (CISP), etoposide (ETP), aflatoxin B1 (AFB1), and methyl methanesulfonate (MMS). Non-DDI chemical abbreviations: 2-deoxy-D-glucose (2DG), sodium chloride (NaCl), ampicillin trihydrate (AMP), sodium ascorbate (ASC), and colchicine (COL). Control represents the vehicle control (DMSO for 2NF, CISP, ETP, AFB1, and COL; water for MMS; media for 2DG, NaCl, AMP, and ASC). Error bars depict standard error, but are too small to see for many data points. * P < 0.01 compared to the vehicle control

Fig. 3

(a) The heatmap on the left depicts the responses of the TGx-DDI biomarker genes in the 28 reference chemicals used to generate it by DNA microarray analysis in TK6 cells, and the test chemicals assessed with AmpliSeq in HepaRG™ cells are shown in the subsequent columns. The labels on the far right hand side are Gene Symbols corresponding to the GenBank accession numbers for the biomarker genes. The color scale indicates fold changes relative to control: up-regulated genes are in red, down-regulated genes in green, and genes exhibiting no changes relative to controls are in black. Predictions of DDI/non-DDI and NSC classification probabilities for all treatment conditions are shown using red (DDI) and blue (non-DDI) bars above each heatmap. (b) Principal component analysis using the TGx-DDI biomarker for TK6 cells exposed to the training set of chemicals (red text = DDI training set; blue text = non-DDI training set) and for HepaRG™ cells exposed to 10 test chemicals at low, mid, and high concentrations 7 h following the last exposure (green text = replicates of test agent). The line drawn at 0 on the PCA plot divides the DDI and non-DDI agents and was used for classification. (c) Hierarchical clustering of the chemicals with TGx-DDI, with color coding as in panel B. The main branch on the dendrogram separates the DDI and non-DDI agents and was used for classification of the test agent