Kinetics of γH2AX and phospho-histone H3 following pulse treatment of TK6 cells provides insights into clastogenic activity

Abstract

The desire for in vitro genotoxicity assays to provide higher information content, especially regarding chemicals’ predominant genotoxic mode of action, has led to the development of a novel multiplexed assay available under the trade name MultiFlow^®. We report here on an experimental design variation that provides further insight into clastogens’ genotoxic activity. First, the standard MultiFlow DNA Damage Assay—p53, γ H2AX, phospho-histone H3 was used with human TK6 lymphoblastoid cells that were exposed for 24 continuous hours to each of 50 reference clastogens. This initial analysis correctly identified 48/50 compounds as clastogenic. These 48 compounds were then evaluated using a short-term, ‘pulse’ treatment protocol whereby cells were exposed to test chemical for 4 h, a centrifugation/washout step was performed, and cells were allowed to recover for 20 h. MultiFlow analyses were accomplished at 4 and 24 h. The γ H2AX and phospho-histone H3 biomarkers were found to exhibit distinct differences in terms of their persistence across chemical classes. Unsupervised hierarchical clustering analysis identified three groups. Examination of the compounds within these groups showed one cluster primarily consisting of alkylators that directly target DNA. The other two groups were dominated by non-DNA alkylators and included anti-metabolites, oxidative stress inducers and chemicals that inhibit DNA-processing enzymes. These results are encouraging, as they suggest that a simple follow-up test for in vitro clastogens provides mechanistic insights into their genotoxic activity. This type of information will contribute to improve decision-making and help guide further testing.

Introduction

Clastogenicity is a genotoxic mode of action (MoA) characterised by alterations in DNA structure. Double-strand DNA breaks and subsequent changes to chromosome structure that include deletions, inversions and translocations are classic lesions associated with clastogen exposure (1). This type of damage is fundamentally different from that caused by exposure to an aneugen which would result in numerical changes in the overall chromosomal content of cells. In the context of product development, clastogenicity can be a challenging situation to manage. Thus, the identification of not only the presence of clastogenic activity but also insights into the ways by which test substances elicit this DNA damage can provide important information that would contribute to improved decision-making and guide further testing.

There are numerous methodologies either on the market or in development that address genotoxic MoA in vitro. The ToxTracker assay (2) uses multiple cell lines expressing specific reporter systems that convey information on genotoxicity and aneugen/clastogen characterisation. Genomic signatures of genotoxicity can be evaluated using the TGx-DDI multiple gene biomarker assay (3). Kopp et al. (4) review the use of γH2AX as a marker of clastogenicity with high utility in cell-based studies. Finally, our laboratory has developed an in vitro, multiplexed, flow cytometric methodology (5), named MultiFlow^®, which examines several nuclear biomarkers to identify test articles’ genotoxic MoA. The MultiFlow platform is highly adaptable and a variation has recently been described that can discriminate among several aneugenic mechanisms—tubulin destabilisation vs. tubulin stabilisation vs. kinase inhibition (6). We report here on another adaptation to the method created for the characterisation of clastogens based on their interaction with DNA.

For this purpose, we examined the persistence or loss of the γ H2AX and phospho-histone H3 (p-H3) biomarker responses in the context of a short-term exposure/washout protocol. This approach was based on observations that γ H2AX and p-H3 responses of certain clastogenic compounds, measured after 4 h of exposure, were maintained or even increased at 24 h, whereas other chemicals showed early responses that were lost upon the recovery period. The latter profile appeared to primarily consist of non-alkylator types of DNA damage, for instance, enzyme inhibition, disruption of nucleotide pools or generation of reactive oxygen species (ROS).

The value of gaining this level of insight about clastogenesis, even in broad terms as described above, is related to the fact that some classes, e.g., enzyme inhibition, are expected to exhibit sub-linear dose-response relationships. As explained by Kirkland and colleagues (7), ‘…chemicals may exert their in vitro genotoxic effects via primary damage to a non-DNA target, and as such there would be an NOEC (no observed effect concentration) below which there would be no damage either to cellular target or DNA’. Also, ‘…chemicals (or their metabolites) may induce direct damage to DNA, but only at certain concentrations above a threshold defined, for example, by detoxification or other protective processes such as those that occur at extreme or non-physiological conditions’. Thus, in some product categories and use cases, it may be appropriate to manage clastogenic agents that match these profiles differently than those that do not share these characteristics.

The experiments described here were performed with 50 diverse clastogens and were designed to test whether the kinetics of appearance and disappearance of γ H2AX and p-H3 following short-term, pulse treatment of TK6 are capable of providing mechanistic insights into clastogens’ genotoxic activity.

Materials and Methods

Chemicals, cells, culture conditions

The identity and source of 50 reference clastogens, as well as presumed molecular mechanisms responsible for their genotoxicity, are provided in Table 1.

Table 1.

List of clastogenic chemicals studied and their proposed activities

Chemical	CAS No.	Source	Presumed clastogenic activities	References
1,3-Propane sultone	1120-71-4	Millipore Sigma	Alkylator	(8)
2,4-Diaminotoluene	95-80-7	NTP	Alkylator	(9)
3-Phenylprop-2-enal (trans-cinnamaldehyde)	14371-10-9	NTP	ROS generator	(10)
4-Nitroquinoline 1-oxide	56-57-5	Millipore Sigma	Alkylator	(11)
5-Fluorouracil	51-21-8	Millipore Sigma	Anti-metabolite, DNA/RNA synthesis	(11)
6-Thioguanine	154-42-7	NTP	Anti-metabolite, DNA/RNA synthesis	(12)
8-Hydroxyquinoline	148-24-3	NTP	ROS generator	(13)
Adriamycin HCl (a.k.a. doxorubicin HCl)	23214-92-8	Millipore Sigma	Intercalator, topoisomerase II inhibitor	(14)
Aphidicolin	38966-21-1	Millipore Sigma	DNA/RNA synthesis inibitor	(15)
Azathioprine	446-86-6	Millipore Sigma	DNA/RNA synthesis inibitor	(16)
Azidothymidine	30516-87-1	Millipore Sigma	Nucleoside analog, DNA/RNA synthesis inhibitor	(11)
β-Lapachone	4707-32-8	Selleckchem	Topoisomerase I inhibitor	(17)
Bleomycin sulfate	9041-93-4	Millipore Sigma	ROS generator, intercalator	(18)
Cadmium chloride	10108-64-2	Millipore Sigma	ROS generator	(19)
Camptothecin	7689-03-4	Millipore Sigma	Topoisomerase I inhibitor	(20)
Chlorambucil	305-03-3	Millipore Sigma	DNA cross-linker	(21)
Ciprofloxacin	85721-33-1	Millipore Sigma	Topoisomerase II inhibitor	(22)
Cisplatin	15663-27-1	Millipore Sigma	Alkylator	(11)
Cytosine arabinoside	147-94-4	Millipore Sigma	Nucleoside analog	(11)
Emodin	518-82-1	Millipore Sigma	Topoisomerase II inhibitor	(23)
Ethyl methanesulfonate	62-50-0	Millipore Sigma	Alkylator	(24)
Etoposide	33419-42-0	Millipore Sigma	Topoisomerase II inhibitor	(11)
Flumequine	42835-25-6	Millipore Sigma	Topoisomerase II inhibitor	(25)
Genistein	446-72-0	Millipore Sigma	Topoisomerase II inhibitor	(26)
Glutaraldehyde	111-30-8	Millipore Sigma	DNA cross-linker	(27)
Glycidamide	5694-00-8	Millipore Sigma	Alkylator, metabolite of acrylamide	(28)
Hydralazine HCl	304-20-1	Millipore Sigma	Nucleoside analog; DNA methylation inhibitor	(29)
Hydrogen peroxide	7722-84-1	Millipore Sigma	ROS generator	(30)
Hydroquinone	123-31-9	Millipore Sigma	ROS generator	(31)
Hydroxyurea	127-07-1	Millipore Sigma	Anti-metabolite, DNA/RNA synthesis inhibitor	(21)
Irinotecan	100286-90-6	Selleckchem	Topoisomerase I inhibitor	(32)
Melphalan	142-82-3	Millipore Sigma	DNA cross-linker	(21)
Menadione	58-27-5	Millipore Sigma	ROS generator	(33)
Mercuric chloride	7487-94-7	NTP	Possible ROS-mediated action	(34)
Methotrexate	59-05-2	Millipore Sigma	Anti-metabolite, DNA/RNA synthesis inhibitor	(35)
Methyl methanesulfonate	66-27-3	Millipore Sigma	Alkylator	(11)
Mitomycin C	50-07-7	Millipore Sigma	DNA cross-linker	(11)
Mitoxantrone	70476-82-3	Selleckchem	Topoisomerase II inhibitor	(36)
Methylnitronitrosoguanidine	70-25-7	Millipore Sigma	Alkylator	(37)
N-Ethyl-N-nitrosourea	759-73-9	Millipore Sigma	Alkylator	(11)
Olaparib	763113-22-0	Millipore Sigma	PARP inhibitor	FDA Approved Label
Propyl gallate	121-79-9	Millipore Sigma	ROS generator	(38)
Resorcinol diglycidyl ether	108-46-3	Millipore Sigma	Alkylator	(39)
SN-38	86639-52-3	Selleckchem	Topoisomerase I inhibitor	(40)
Stavudine	3056-17-5	Millipore Sigma	Nucleoside analog	(41)
Temozolomide	85622-93-1	Millipore Sigma	Alkylator	(42)
Teniposide	29767-20-2	Millipore Sigma	Topoisomerase II Inhibitor	(43)
Thiotepa	52-24-4	Millipore Sigma	DNA cross-linker	(21)
Topotecan	123948-87-8	Millipore Sigma	Topoisomerase I inhibitor	(44)
Ziram	137-30-4	Millipore Sigma	PARP activator	(45)

Abbreviations: NTP, National Toxicology Program; ROS, reactive oxygen species; PARP, poly(ADP-ribose) polymerase.

TK6 cells were purchased from ATCC^® (cat. no. CRL-8015). Cells were grown in a humidified atmosphere at 37°C with 5% CO₂ and maintained at or below 1 × 10⁶ cells/ml. The culture medium consisted of RPMI 1640 with 200 µg/ml sodium pyruvate (both from Sigma-Aldrich, St. Louis, MO, USA), 200 µM l-glutamine, 50 units/ml penicillin and 50 µg/ml streptomycin (from Mediatech Inc., Manassas, VA, USA) and 10% v/v heat-inactivated horse serum (Gibco^®, a Thermo Fisher Scientific Company, Waltham, MA, USA).

DNA damage assay: cell treatments

Chemicals were tested in U-bottom 96 well plates containing 198 µl TK6 cell suspension (2 × 10⁵/ml) plus 2 µl of DMSO-solubilised test chemical per well. In the absence of excessive cytotoxicity (defined below), the top concentration was 1 mM or the lowest precipitating concentration, whichever was lower. Nineteen additional concentrations were tested using a square root dilution scheme—i.e., each concentration differed from the one above by a factor of 70.71%. In this manner, a wide range of concentrations was evaluated (i.e. nearly 3 orders of magnitude, 0.0014 to 1 mM). For the single replicate study performed here, each of the 20 concentrations was tested in a single well, whereas solvent was evaluated in 4 replicate wells. Upon addition of test chemicals, the plates were immediately incubated in a humidified atmosphere at 37°C with 5% CO₂ for 24 h.

DNA damage assay: sample processing

After exposure, TK6 cells were prepared for flow cytometric analysis using reagents and instructions included in the MultiFlow^® DNA Damage Kit—p53, γ H2AX, phospho-histone H3 (Litron Laboratories, Rochester, NY, USA). Components and preparation of the MultiFlow working solution were described in detail previously (46,47). At the 4- and 24-h sampling times, cells were resuspended with pipetting, then 25 µl were removed from each well and added to a new 96-well plate containing 50 µl/well of pre-aliquoted working MultiFlow reagent solution. Mixing was accomplished by pipetting the contents of each well several times. After incubation at room temperature for 30 min, samples were analysed via flow cytometry.

DNA damage assay: flow cytometric analysis

Flow cytometric analysis was carried out using a Miltenyi Biotec MACSQuant^® Analyzer 10 flow cytometer with integrated 96-well MiniSampler device. Stock photomultiplier tube (PMT) detectors and associated optical filter sets were used to detect fluorescence emissions associated with the fluorochromes: FITC (detected in the B1 channel, to use Miltenyi instrument parlance), PE (B2 channel), propidium iodide (B3 channel) and Alexa Fluor^® 647 (R1 channel). Flow cytometry data were analysed using FlowJo software v10.5.0 (Ashland, OR, USA).

Representative bivariate graphs, gating logic and position of regions were described in detail in earlier reports (1,46,47). Whereas γ H2AX and p53 responses were based on the shift in median channel fluorescence intensity relative to same-plate solvent controls, p-H3 and polyploidy data were based on their frequency among other nuclei. Nuclei to counting bead ratios were calculated for each sample, and these ratios were used to determine absolute nuclei counts (those with 2n and greater DNA-associated propidium iodide fluorescence). Nuclei counts were used to derive relative nuclei count (RNC), and %cytotoxicity was calculated as 100% minus %RNC at 24 h.

DNA damage assay: data analysis

Data analyses described herein were restricted to those concentrations that did not exceed the MultiFlow assay’s cytotoxicity limit. That is, as described by Dertinger et al. (48), the top concentration of each chemical had to exhibit ≤80% reduction to RNC at the 24-h time point, and a maximum of two concentrations within the cytotoxicity range of 70–80% were permitted.

The initial genotoxic potential and genotoxic MoA (i.e. clastogen vs. aneugen) predictions were made as described in a recent publication (48): via a machine learning (ML) ensemble and by comparing fold-increase values against a global evaluation factor (GEF) rubric. These GEFs are essentially threshold values for the various biomarkers that, when exceeded, signify a positive response (46). Similar to the ML analysis, a pre-determined set of rules for the application of the GEFs was applied to establish the clastogenicity calls. As described previously, an overall genotoxic call was made when either the GEF rubric or ML ensemble provided evidence of genotoxicity (48). Fold-increase data used to generate the ML ensemble are available as a Supplementary File.

Clastogen kinetics assay: cell treatments

The chemicals identified as exhibiting a clastogenic (n = 48) MoA were tested in a follow-up assay designed to investigate the kinetics of appearance and disappearance of γ H2AX and p-H3 biomarkers. The cells were seeded into wells of a U-bottom 96 well plate (5 × 10⁵/ml; 198 µl/well). Test chemicals were added over a range of concentrations in a volume of 2 µl. The top concentration was the same as the highest acceptable concentration from the initial MultiFlow DNA Damage Assay. As with the initial assay, a single replicate experiment was performed using a square root 2 dilution scheme to test 20 concentrations, and 4 solvent control wells were included on each 96 well plate. After the 4-h pulse treatment, cells were washed out of the compound and left to incubate for an additional 20 h. At both 4- and 24-h time points, cell aliquots were collected and processed through the standard MultiFlow protocol.

Clastogen kinetics assay: flow cytometric analysis

Flow cytometric analysis was performed using a Miltenyi Biotec MACSQuant^® Analyzer 10 flow cytometer with integrated 96-well MiniSampler device. The user-defined mixing and fluidics parameters for the MACSQuant were as follows: 40 µl of the sample were mixed, and then 20 µl were analysed at a flow rate of 50 µl/min until the 20 µl volume was exhausted. The ‘screen mode’ setting ensured that the probe was rinsed with 208.4 µl of sheath fluid between samples. Kit-supplied Sphero™ Multi-Fluorophore Particles were used to set and monitor PMT voltages, and electronic compensation was used to eliminate fluorescence spillover. Flow cytometry data were analysed using FlowJo software v10.5.0.

Clastogen kinetics assay: data analysis

Unsupervised clustering was performed with JMP software (v15) based on area under curve (AUC) data. To convert the biomarker responses to AUC, feature scaling was applied to every test article concentration to bring the values into the range 0–1 (49). AUC values were also calculated for γ H2aX and p-H3 at both 4- and 24-h time points. In all cases, AUC was calculated using Microsoft Excel via the trapezoidal rule as described at https://calculushowto.com/find-the-area-under-the-curve-in-excel. Note that one (1) was subtracted from every biomarker’s fold-change value before AUC calculations were made in order to set the no effect (baseline) value to zero. With this offset in place, AUC values were zero or nearly so in the case of no response, positive in the case of an increase and negative in the case of a reduction. The cluster analysis options were set as follows: clustering method = hierarchical; method for calculating distances between clusters = ‘Ward’; data as usual = ‘Standardise Robustly’; data visualisation = ‘Dendrogram’, with ‘Two-Way Clustering’.

Results and Discussion

ML algorithm outputs generated from the initial MultiFlow assay results are shown in Figure 1. Specifically, clastogen probability scores are shown for all three models, i.e., logistic regression, artificial neural network and Random Forest. Positive call criteria for clastogenicity are one or more concentrations achieving greater than 90% probability or two or more concentrations greater than 80%. A majority vote rule was then applied whereby two or more models must agree on a positive call in order to be considered a clastogen. In addition to the ML-based classifications, we also employed a set of GEFs that were developed as part of an interlaboratory validation trial of the MultiFlow DNA Damage Assay (46). Based on the outputs from both the GEF and ML analyses a positive call in either system was considered a definitive clastogenicity call and 48 out of the 50 clastogens were correctly identified. The two compounds classified as non-genotoxicants were ciprofloxacin and glutaraldehyde, both of which did not exhibit robust enough responses to trigger either GEF or ML clastogenicity predictions. Aneugenicity predictions were also calculated for these test chemicals, but little to no activity was evident (data not shown).

Fig. 1.

Manhattan plots depicting ML-based probability scores for prediction of clastogenicity of the 50 test compounds. Abbreviations: ANN, artificial neural network (blue); RF, Random Forest (red); LR, logistic regression (green).

The γ H2AX responses following exposure to two representative clastogens in the context of a 4-h washout protocol are shown in Figure 2. When comparing the induction of γ H2AX at 4 h with that observed at 24 h of exposure, there are clear differences between methyl methanesulfonate (MMS), an established DNA alkylating agent, and etoposide (ETO), a well-known inhibitor of topoisomerase II. The histogram for MMS shows persistence, and even slight increases, of the γ H2AX response after washout. Conversely, the histogram for ETO reveals a clear reduction in the γ H2AX values obtained at 24 h compared to those seen at the time of washout at 4 h. As described below, this served as one of the two main signatures used to group clastogens according to the mechanism of action.

Fig. 2.

Histograms depicting kinetics of γ H2AX responses following exposure to the prototypical clastogens MMS or ETO. 4-h values in blue are elevated in a dose-related fashion for each compound, however, 24-h responses in red are not maintained after washout at 4 h for ETO.

Figure 3 shows the unsupervised clustering of 48 clastogens based on γ H2AX and p-H3 responses at 4 h and following 20 h of recovery. This analysis identified three main clusters, designated as Groups 1–3. The heatmap conveys the range of responses of the respective endpoints, while the tree-like dendrogram depicts relationships between compounds that behave similarly at varying levels of detail. The cluster means for the respective groups shown in Figure 3 provide some detail into the general responses of the two biomarkers that make up the three clusters of clastogens. Another approach for visualising these data is by the parallel plots depicted in Figure 4. These plots readily show the different patterns in response across the three groups: reductions in γ H2AX from 4 to 24 h in Groups 1 and 3; elevations/maintenance of γ H2AX from 4 to 24 h in Group 2; increases in p-H3 from 4 to 24 h for Groups 1 and 3, etc.

Fig. 3.

Unsupervised clustering of 48 compounds identified as clastogens via preliminary MoA assessment. Clusters are identified by group names in the dendrogram to the right and via colour coding of the chemical names: Group 1 in red, Group 2 in green and Group 3 in blue. Mean biomarker responses for clusters/groups are displayed in the cluster means section.

Fig. 4.

Parallel plots for Group 1 (red), Group 2 (green) and Group 3 (blue) clastogens as identified by unsupervised clustering. The relationships between biomarkers responses for each chemical (line) reveals profiles that are consistent within each group.

An examination of the specific compounds within each group reveals broad similarities in regards to clastogenic activity(ies). Thus, Group 2 was predominantly made up of clastogens known to be (i) alkylators with electrophilic moieties that attack DNA at nitrogen and oxygen within bases and/or the phosphodiester backbone, e.g., melphalan, thiotepa, temozolomide and (ii) DNA-reactive chemicals that cause inter- and/or intra-stand crosslinks, e.g., cisplatin and mitomycin C (50). Collectively we will refer to those chemicals clustered into Group 2 as ‘alkylators’.

Groups 1 and 3 shared similarities in terms of response profiles with the differences mainly arising in the magnitude(s) of the biomarker responses. This is perhaps best observed in the parallel plots shown in Figure 4 where the changes in biomarker responses are modest in Group 3 compared to Group 1, but the overall direction of change is the same. An examination of the members of these groups as seen in Figure 3 shows several compounds that inhibit DNA-processing enzymes such as topoisomerases (e.g. ETO, topotecan) or enzymes involved in DNA synthesis (e.g. cytosine arabinoside). The compounds in Groups 1 and 3 also include examples of anti-metabolites and nucleoside analogs, the clastogenicity of which is often attributed to their ability to disrupt nucleotide pool balance through inhibition of key enzymes in de novo synthesis of nucleic acid precursors, for instance, ribonucleotide synthase, thymidylate synthase, dihydrofolate reductase, etc. This set of diverse clastogenic activities is perhaps best generalised by referring to the chemicals in Groups 1 and 3 as ‘non-alkylators’, in contrast to the alkylators in Group 2.

Based on these activities, most of the compounds in Groups 1 and 3 would be expected to have sub-linear dose-response curves (7) and as such would be managed differently with regard to prioritisation and decision-making compared to Group 2 compounds. The inclusion of compounds in Groups 3 that are known to elicit their DNA damaging activity via the production of ROS, i.e., hydrogen peroxide and menadione, is also supported by Kirkland et al. (7). The authors state that while such agents would be expected to interact directly with DNA, these compounds ‘…would be expected to have a threshold…’ and are thus fundamentally different from the alkylators. The expectation of linear vs. non-linear dose-response relationships is complicated by the fact that some of these agents, especially non-canonical nucleobases and nucleoside analogs, may exhibit multiple modes of action. For example, beyond disrupting nucleotide pools, 5-fluorouracil can be incorporated into mammalian nuclear DNA to a minor extent (51). This can often make it challenging to perfectly classify clastogens into discrete categories based on DNA reactivity, but we attempted to use information based on the predominant mechanism of action on target molecules and then consider additional factors such as repair kinetics, etc.

As mentioned above, several of the 48 clastogens did not appear to segregate according to the broad classification scheme of Group 2 being alkylators and Groups 1 and 3 as non-alkylators. For instance, 1,3-propane sultone and 4-nitroquinoline 1-oxide appeared in Group 1, despite the fact that both are reported to be alkylating agents (11,52). In our experience, these can be considered weak clastogens and thus the loss of γ H2AX at the 24-h time point may have been mediated by this and concurrent repair of damage. Bleomycin was placed in Group 3 (non-alkylators), and while it is often characterised as a radiomimetic which closely interacts with DNA (18), the production of ROS is highly implicated in its chromosome breaking activity. This may or may not functionally categorise it with other ROS producers as seen in Group 3. Finally, 6-thioguanine (6-TG) was placed in Group 2 (alkylators) and while it may be mediating some of its DNA damaging effects via disruption of DNA synthesis, it has also been reported to incorporate into DNA and lead to single-strand breaks, cross-linking and other related damage (53). Therefore, as the case with bleomycin, 6-TG is rather challenging to characterise. Finally, 2,4-diaminotoluene is typically considered to be DNA-reactive but was categorised in Group 3. This may have been because this compound’s genotoxic activity is generally reported in the context of metabolic activation, which was not employed in this study. The overall weak responses observed may have been sufficient to trigger an overall clastogen MoA call but do not reflect the full extent of its activity. This observation highlights the need to investigate the use of S9 or other metabolically competent test systems in the context of this proposed methodology.

Overall, using the groups as defined above the following performance can be assessed. Based on the presumed clastogenic activities as shown in Table 1, 11 out of the 13 members in Group 2 share the characteristics of being DNA alkylators. If we combine the chemicals in Groups 1 and 3, we find that 30 out of 35 have clastogenic activities consistent with non-alkylators, e.g., enzyme inhibitors, ROS inducers. Thus, in total, the Clastogen Kinetics Assay correctly categorised 41/48 clastogens based on our alkylator vs. non-alkylator categories. No single method is going to be perfect and the influence of timing and alternate repair pathways likely has an impact on the overall performance of this approach. Future work could potentially investigate the use of additional time point(s) to more accurately define the kinetics of damage and repair and perhaps better characterise certain types of genotoxic agents. Even so, the method as it stands is simple, efficient and performs well.

The current study highlights the enhanced utility of the MultiFlow platform with regard to providing information not only on genotoxic MoA but also insight into clastogens’ genotoxic mechanisms. In a previous publication, Dertinger et al. (48) investigated the ability of the standard MultiFlow DNA Damage Assay data to further categorise aneugens and clastogens once they were identified. While there was clear functionality in this approach for grouping subclasses of aneugens, there was not much utility in the case of clastogens (see Figure 5 in that publication). Based on the data shown here, once a clastogenic MoA has been identified by the standard assay, a follow-up assessment incorporating a washout at 4- and 20-h recovery period provides additional important information about compounds’ DNA targeting activity.

The value of genotoxic MoA information in contributing to risk assessments is well established in the literature (54). Indeed Snyder (55) states that

It would be of potentially great importance to distinguish clastogenicity arising from direct covalent drug/DNA interaction from that arising secondarily through topoisomerase inhibition. In the latter case, clastogenicity would be expected to be a threshold phenomenon occurring only after a high enough cellular drug concentration was attained to shift the equilibrium of non-covalent DNA- or ternary complex-binding to increase DNA residence time.

While Snyder was speaking directly to topoisomerase inhibition, this concept is generalisable to other non-DNA targets. Thus, this level of information from an in vitro screening situation would clearly benefit decision-making and focus resources on viable candidate compounds.

We acknowledge that this approach to characterising clastogens is by no means complete with regard to exact identification of mechanism(s) of action or target molecules. Thus, we are continuing to develop approaches that take advantage of the flexibility and high information content provided by the MultiFlow platform. Our latest methodology in development explores specific DNA repair pathways, seeks to provide the deepest level(s) of mechanistic information available via this system. Such an approach, when combined with the information provided by the Clastogen Kinetics Assay, should have great utility in rapid, early hazard screening as well as in the generation of adverse outcome pathways which are becoming an important part of modern risk assessment strategies.

Conclusions

The high information content of current in vitro multiplexed assays provides opportunities to obtain deeper levels of insight into mechanism and molecular targeting of compounds of interest. We demonstrate here the utility of γ H2AX and p-H3 kinetics in categorising clastogens into groups that appear to inform on their clastogenic mechanisms, i.e., alkylators vs. non-alkylators. Such information will be useful in high-throughput screening activities for prioritisation and decision-making as well as prediction and design of required downstream testing paradigms.

Funding

This work was funded in part by a grant from the National Institute of Health/National Institute of Environmental Health Sciences (NIEHS; grant no. R44ES029014).

Conflict of interest statement: S.M.B., S.D.D. and J.C.B. are employed by Litron Laboratories. Litron holds a patent for flow cytometry-based analyses described herein. Litron currently sells and provides testing services based on the MultiFlow DNA Damage Kit—p53, γ H2AX, phospho-histone H3 described herein.