Targets and mechanisms of chemically induced aneuploidy. Part 1 of the report of the 2017 IWGT workgroup on assessing the risk of aneugens for carcinogenesis and hereditary diseases

Targets and mechanisms of chemically induced aneuploidy. Part 1 of the report of the 2017 IWGT workgroup on assessing the risk of aneugens for carcinogenesis and hereditary diseases

Mutat Res Gen Tox En 847 (2019) 403025 Contents lists available at ScienceDirect Mutat Res Gen Tox En journal homepage: www.elsevier.com/locate/gent...

10MB Sizes 0 Downloads 31 Views

Mutat Res Gen Tox En 847 (2019) 403025

Contents lists available at ScienceDirect

Mutat Res Gen Tox En journal homepage: www.elsevier.com/locate/gentox

Minireview

Targets and mechanisms of chemically induced aneuploidy. Part 1 of the report of the 2017 IWGT workgroup on assessing the risk of aneugens for carcinogenesis and hereditary diseases

T

Anthony M. Lyncha,⁎, David Eastmondb, Azeddine Elhajoujic, Roland Froetschld, Micheline Kirsch-Volderse, Francesco Marchettif, Kenichi Masumurag, Francesca Pacchierottih, Maik Schuleri, David Tweatsj a

GSK, Ware, United Kingdom University of California, Riverside, CA, USA Novartis Institutes for Biomedical Research, Preclinical Safety, Basel, Switzerland d BfArM, Bonn, Germany e Vrije Universiteit Brussels, Brussels, Belgium f Environmental Health Science and Research Bureau, Health Canada, Ottawa, ON, Canada g Division of Genetics and Mutagenesis, National Institute of Health Sciences, Kanagawa, Japan h Health Protection Technology Division, Laboratory of Biosafety and Risk Assessment, ENEA, CR Casaccia, Rome, Italy i Pfizer, Groton, CT, USA j University of Swansea, Wales, United Kingdom b c

ABSTRACT

An aneuploidy workgroup was established as part of the 7th International Workshops on Genotoxicity Testing. The workgroup conducted a review of the scientific literature on the biological mechanisms of aneuploidy in mammalian cells and methods used to detect chemical aneugens. In addition, the current regulatory framework was discussed, with the objective to arrive at consensus statements on the ramifications of exposure to chemical aneugens for human health risk assessment. As part of these efforts, the workgroup explored the use of adverse outcome pathways (AOPs) to document mechanisms of chemically induced aneuploidy in mammalian somatic cells. The group worked on two molecular initiating events (MIEs), tubulin binding and binding to the catalytic domain of aurora kinase B, which result in several adverse outcomes, including aneuploidy. The workgroup agreed that the AOP framework provides a useful approach to link evidence for MIEs with aneuploidy on a cellular level. The evidence linking chemically induced aneuploidy with carcinogenicity and hereditary disease was also reviewed and is presented in two companion papers. In addition, the group came to the consensus that the current regulatory test batteries, while not ideal, are sufficient for the identification of aneugens and human risk assessment. While it is obvious that there are many different MIEs that could lead to the induction of aneuploidy, the most commonly observed mechanisms involving chemical aneugens are related to tubulin binding and, to a lesser extent, inhibition of mitotic kinases. The comprehensive review presented here should help with the identification and risk management of aneugenic agents.

1. Introduction As part of the 7th International Workshops on Genotoxicity Testing (IWGT) that were held in Tokyo, Japan in November 2017, a workgroup composed of 11 international experts from government, industry and academia was assembled to review and assess the risk of aneugens for human health. The proceedings of the workgroup are summarized in three companion manuscripts that respectively report on: (1) mechanisms and methods of detection of aneuploidy with consideration of the application of the adverse outcome pathway (AOP) approach for risk assessment of aneugens; (2) induction of aneuploidy in germ cells and its association with hereditary diseases; and, (3) role of aneuploidy in



the carcinogenic process. The present manuscript provides a review of the biological mechanisms of aneuploidy in mammalian cells and methods used to detect chemical aneugens. In addition, the current regulatory framework is discussed, along with the ramifications of chemical exposure to aneugens for human risk assessment. The workgroup explored the use of adverse outcome pathways (AOPs) to document mechanisms of chemically induced aneuploidy, with a focus on two molecular initiating events (MIEs), chemical binding to (a) tubulin and (b) the catalytic domain of aurora kinase B, which can lead to a number of adverse outcomes, including aneuploidy and polyploidy. The evidence linking chemically induced aneuploidy with carcinogenicity and hereditary disease was also reviewed. These discussions provided

Corresponding author at: David Jack Centre for Research & Development, GSK, Ware, Hertfordshire SG12 0DP, United Kingdom. E-mail address: [email protected] (A.M. Lynch).

https://doi.org/10.1016/j.mrgentox.2019.02.006 Received 17 December 2018; Received in revised form 22 January 2019; Accepted 20 February 2019 Available online 02 March 2019 1383-5718/ Crown Copyright © 2019 Published by Elsevier B.V. All rights reserved.

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

the basis for the IWGT consensus statements, which were presented at the 2017 ICEM and have been revised in the context of this and our two companion manuscripts.

some cases, tumour susceptibility (these models are described in more detail in the third manuscript of this series [10]). At the cellular level, the major consequences of aneuploidy (reviewed in [11,12]) are (i) reduction of cellular fitness and development of the organism by repression of cell proliferation; (ii) modification of the proteome leading to increased energy burden, alteration of metabolic capacities, increased drug sensitivity and increased likelihood of senescence; (iii) trading a reduction in proliferation rate for an increased ability of aneuploid cells to adapt and evolve, called the aneuploidy paradox, (iv) promotion of genome instability by creating imbalance in the levels of proteins required for DNA replication, repair and mitosis or by induction of chromosome breaks in the lagging chromosome trapped in the cleavage furrow during cytokinesis (for review see [13]), (v) cell death and (vi) altered differentiation. Lagging of chromosome(s) during the metaphase/anaphase transition may produce a micronucleus which may undergo rearrangements by two possible mechanisms: (i) locally defective DNA replication initiating serial, microhomology-mediated template switching (called chromoanagenesis) that produces local rearrangements with altered gene copy numbers and (ii), chromosome shattering with several breakage/replication cycles (chromothripsis) resulting from mitotic entry before completion or failure of DNA replication within the micronucleus (for review see [14]). At the cellular and tissue level, more than a century ago Theodor Boveri linked incorrect chromosome numbers in urchin embryos with abnormal development; and hypothesized that having the wrong number of chromosomes might cause cells to grow in an uncontrollable way and become the seeds of cancerous tumours [15]. In embryos, aneuploidy results essentially from events occurring in germ cells or in the early post fertilization cleavage cycles. In humans, it is estimated that at least 10–30% of all conceptuses and over 50% of preimplantation embryos carry a numerical chromosome abnormality (for review see our companion paper [6]). In human cancer, aneuploidy is a remarkably common feature present in ∼70–90% of solid tumours and > 50% of hematopoietic cancers [16]. The degree and spectrum of aneuploidies vary considerably among tumour types but many show recurrent whole-chromosome abnormalities. It is obvious that the tumorigenicity of an aneuploid cell is dependent on the gene content of the specific chromosomes that are lost or gained, and the balance of cancer suppressor/oncogenes. These chromosome imbalances can provide positive or negative selective effects on cell growth and survival, which are dependent on chromosome size (i.e. the amount of genetic material involved) and cell type/context. Whether aneuploidy is a cause, or a consequence of carcinogenesis has been debated for many decades. Although aneuploidy correlates with transformation, empirical tests of the hypothesis that aneuploidy drives tumorigenesis have been hampered by the difficulty of generating aneuploidy without causing other cellular defects, particularly cytotoxicity (cell death) and indirectly, genome instability. Nevertheless, aneuploidy is regarded as a hallmark of cancer, however, its role is complex with both pro- and anti-carcinogenic effects evident (for a more in-depth review of the role of aneuploidy in cancer please refer to the third manuscript of this series [10]).

2. Aneuploidy and genetic toxicology 2.1. What is aneuploidy? The spectrum of genetic changes assessed in genetic toxicology can be divided into four main categories: DNA sequence changes from one or few adjacent nucleotides up to several hundred nucleotides as a result of insertion or deletion of bases in the genome; copy number variations (CNV) where sections of the genome are present at a variable copy number in comparison with a reference genome and may arise, for example, by gene duplication; variations at the chromosomal level caused by chromosome breaks resulting in large deletions or rearrangements e.g. chromosome translocations; and numerical alterations involving whole chromosomes leading to aneuploidy or, in the instance of complete karyotypes, polyploidy. Therefore, numerical chromosome abnormalities refer to the gain or loss of a single or a few chromosomes at mitosis or meiosis, or to the formation of polyploid cells. Chromosome mis-segregation events are believed to originate from defects in one of the numerous pathways controlling cell division [1]. In humans, the most common chromosomal abnormality is aneuploidy [2]. Aneuploidy is defined as the alteration of chromosome number that is not a multiple of the haploid complement. This condition is different from polyploidy in which cells harbour a multiple of their haploid karyotype beyond diploidy, and in certain circumstances may represent an intermediate step leading to aneuploidy. Constitutional aneuploidy is a hereditary condition where specific numerical chromosome abnormalities are present in all cells of an individual and is found in ∼0.3% of new-born humans. This is because all autosomal aneuploidies, except for trisomies 13, 18, and 21, are embryo-lethal. Mosaic aneuploidy (i.e. where specific numerical chromosome abnormalities occur in only a fraction of an individual's cells) is also relatively uncommon (< 2%) in most adult mammalian tissues but is elevated in brain (∼4% in neurons) as part of the constitutional make-up of this organ [3]. Aneuploidy results mainly from chromosome non-disjunction, a chromosome distribution error occurring during mitosis or meiosis. The frequency of spontaneous numerical chromosome abnormalities in healthy subjects is relatively low i.e. a conservative estimate of 1.78% in peripheral blood lymphocytes was reported recently [4]. From a mechanistic point of view, aneuploidy can result from errors in the many processes controlling the fidelity of cell division: for example, cohesion defects of sister chromatids, merotelic attachment of chromosomes, hyper stabilization of kinetochore–microtubule interactions, tubulin depolymerisation, dysfunctional telomeres, defects in mitotic checkpoints and induction of unstable tetraploid intermediates by mitotic slippage, cytokinesis failure or viral-induced cell fusion. This means that there are targets, other than DNA, responsible for aneuploidy. In most normal cells, a surveillance system with effective checkpoints responds to the presence of abnormal chromosome content to halt cell cycling, causing cell death or inducing cell senescence. In cancer cells, aneuploidy is also associated with genome instability, a phenotype where cells display a high frequency of mutations and/or chromosomal instability (CIN). Indeed, aneuploidy, depending on the gene content of the chromosomes involved, may induce genomic instability directly [5]. During meiosis, the major molecular mechanisms that are involved in the induction of aneuploidy in somatic cells also play a role in the induction of aneuploidy in germ cells. The mechanisms that are unique to germ cells are discussed further in the second manuscript of this series [6]. At the molecular level, studies in mouse models genetically defective in genes encoding proteins involved in chromosome segregation like Mad2 [7], Bub1 [8] and CenpE [9] show increased aneuploidy and in

2.2. Cellular targets for aneuploidy During cell division, accurate cell cycle control and chromosome segregation require coordinated interaction of many cellular components including spindle microtubules, centrioles, protein kinases, topoisomerases, centrosomes and kinetochores, etc. (see Fig. 1). These individual components represent potential targets for the disruption of fidelity of chromosome segregation during mitosis and in turn, the induction of aneuploidy in daughter cells. Moreover, some environmental chemicals and drugs can interact with these components and impair their function. This may affect cell cycle progression and can lead to cell death. Cell cycle regulation has been the target of anti-cancer drugs and some are known aneugens, e.g. aurora kinase inhibitors. There is, 2

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

Fig. 1. Summary of key events in mitosis, protein targets and exemplary compounds in aneuploidy induction. Top row shows examples of TK6 cells at various stages of mitosis with chromatin stained blue, the mitosis specific antibody MPM2 in red and microtubules in green.

a single positive concentration is observed, additional studies may be required with narrower dose spacing to fully understand the dose response relationship. When considering the consequences of exposure to aneugens for human health, the same considerations hold true i.e. below a specific threshold concentration no effect is expected. In contrast at higher concentrations, there is a mounting evidence from human, rodent and in vitro studies which indicates that cells can only tolerate a certain level of chromosome instability [21]. Therefore, at high doses, cells with large genomic imbalances are most likely eliminated by cell death mechanisms. As such, cells can only express their genomic imbalance at doses that induce moderate levels of aneuploidy since these are compatible with cell survival. This is an important consideration for dose selection with cancer chemotherapies targeting the fidelity of cell division. It may also explain why cancer risk may be higher at relatively lower levels of aneuploidy than in highly perturbed genomes.

however, limited published mechanistic data in the public domain vis a vis the genotoxicity and carcinogenicity of newer chemical classes that have inhibitory effects on cell division e.g. CENP inhibitors, cell cycle inhibitors that target cyclin dependent kinases, polo-like-kinases, IκB kinases, etc. or agents that target the centrosome. The workgroup therefore focused on tubulin binding agents and aurora kinase proteins, pathways where much more information is available, adopting an adverse outcome pathway (AOP) framework to support mechanistic analyses [17,18]. 2.3. Genotoxins that induce aneuploidy (whole chromosome changes) are called aneugens Aneuploidy is considered to result from indirect mechanisms of genotoxicity. The relationships between exposure, target(s) engagement, and genotoxic endpoints have been modelled [19] and are reviewed in [20]. In contrast to DNA-binding mutagens, genotoxins that induce their effects via non-DNA binding mechanisms are expected to show sublinear concentration-effect response curves and act via a threshold. This means that for low dose aneugens, in general, no effects will be induced and/or seen because of the multiplicity of the targets involved in the division process. As the doses of aneugen increase, the cell division apparatus becomes more compromised resulting in cells with chromosome imbalances, and the potential for the effects of aneuploidy to be detected. At higher doses, the disturbances may be so severe as to prevent cell survival or propagation. Because of these effects, particular attention must be given to the dose-effect relationships when designing studies to assess aneuploidy as there is often only a narrow window of exposure where aneuploidy is induced, and higher doses will not further increase the effect. Therefore, exposures that cause moderate cell cycle alterations should be preferred. This has been the experience with both somatic and germ cell assays, suggesting that when

3. Aneuploidy detection in mammalian cells Aneuploidy can be detected in mammalian cells in a number of ways. The following paragraphs will give an overview on the detection of aneuploidy in standard genetic toxicology assays, the use of mode of action (MoA) screening assays to detect the ability of test compounds to act as aneugens and lastly methodologies that can identify mechanisms of action (MOA) that lead to aneuploidy induction. Many other approaches exist but will not be reviewed in this manuscript. 3.1. The chromosome aberration assay The chromosome aberration assay is designed to evaluate the clastogenic potential of a test compound in cultured cells (primary cells or established cell lines) in the presence and absence of an exogenous 3

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

metabolizing system (typically induced rat liver post-mitochondrial supernatant, S9). Cultured human lymphocytes are frequently used by default [22,23]. The methodology is well described as outlined in the Organisation for Economic Co-operation and Development (OECD) Test guideline TG 473 for the chromosome aberration assay [24]. In general, most chemical mutagens produce recognizable chromosomal damage, which is expressed as structural chromosome aberrations (deletions and rearrangements). The aberrations are best observed at the metaphase stage of cell division when chromosomes are condensed. Cell division can be arrested at metaphase by use of the mitotic inhibitor e.g. colcemid, which prevents formation of the mitotic spindle. The standard OECD test is not recommended for the detection of aneuploidy [24] since many cells are evaluated during their first mitosis before aneuploidy has occurred. By modifying the assay to score cells in their 2nd mitotic division after treatment, aneuploidy can be more reliably evaluated. Aneuploidy is generally assessed by scoring hyperdiploid cells (i.e. those cells with more than 46 chromosomes in the case of primary human lymphocytes) and polyploidy. An increase in hyperdiploidy provides direct evidence of aneuploidy whereas an increase in polyploidy can be indicative of a potential for aneugenicity. Hypodiploidy assessment is less reliable and methods must be used to control for technical artefacts [25]. One approach is the enumeration of specific chromosomes following exposure to potential chemical aneugens using whole chromosome paints and centromere specific probes (reviewed by [26]). FISH (fluorescence in situ hybridization) is one of the recommended methods for the assessment of aneuploidy as an apical endpoint in regulatory genotoxicity studies using cultured mammalian cells [27,28].

exposure. One indicator of aneugenic potential is the ability of a compound to induce micronucleated erythrocytes in the rodent bone marrow or blood micronucleus assays in vivo. The bone marrow, which is a highly proliferative cell compartment, is considered a sensitive target tissue when exposed. In the absence of an in vitro genotoxic signal, micronucleus assessments are routinely integrated into repeat dose toxicity studies in rodents. Therefore, if the in vivo somatic cell micronucleus assay is negative, with adequate evidence of exposure of the bone marrow to the test compound (for example, by determining exposure in plasma), it is considered reasonable to conclude that no further testing of aneuploidy-inducing potential would be warranted [37]. If the test is positive, CREST staining or FISH using pan-centromeric probes (performed in an in vitro MN test where practicable), can be used to determine the origin of the induced micronuclei. 3.3. Mode of Action (MoA) assays Recently, laboratories that conduct screening for the detection of genotoxic potential have focused on the development of assays that provide more mechanistic information and permit genotoxins to be distinguished based on their MoA. The 2 assays described below are currently undergoing intensive validation by the genetic toxicology community. 3.3.1. MultiFlow® The MultiFlow® DNA Damage Kit is an in vitro assay that provides genotoxic MoA information [38]. The assay has several features that make it attractive for early screening programmes: (i) compatibility with microtiter plates and add-and-read format keeps test article requirements and other resources low; (ii) its multiplexed nature provides several efficient readings related to DNA damage and overt cytotoxicity; and, (iii) the biomarker responses distinguish clastogenic from aneugenic MoA. The assay biomarkers include γH2AX, phospho-histone H3, nuclear p53 content, polyploidy, and cytotoxicity biomarkers, and each are quantified via one flow cytometric analysis. In general, γH2AX and early p53 responses are indicative of clastogenicity, whereas aneugenic activity is characterized by several combinations of increased phosphohistone H3, late p53 responses, and polyploidization. To date, the bulk of published work has been performed in TK6 cells. The assay allows for the identification of clastogens and aneugens using expert judgement or machine learning approaches. For example, a report by seven collaborating laboratories described an expert rule-based approach for interpreting MultiFlow® data [39]. This strategy utilizes a series of cut-off values, one for each biomarker. Depending on the combination of endpoints that exceed the threshold values, a clastogenic or aneugenic MoA is indicated. High accuracy, > 90%, was observed. In addition to this expert rule-based approach, multiplexed biomarker responses are amenable to machine learning strategies. For instance, two recent reports have demonstrated that reliable clastogen and aneugen MoA predictions could be made using logistic regression, random forest, and/or neural network algorithms derived from categorized training set data (n = 85 compounds) [39,40]. MultiFlow® may therefore represent a good starting point for early screening programmes, for instance to prioritize chemicals for further study, and/or to help guide the design of follow-up experiments that include other aneugen-detection biomarkers.

3.2. The micronucleus (MN) test Micronuclei in the cytoplasm of interphase cells may originate from acentric (i.e. lacking a centromere) chromosome fragments, or whole chromosomes that are unable to migrate to the poles during the anaphase stage of cell division. Micronuclei represent damage that has been transmitted to daughter cells, although the changes may not be compatible with cell survival. The in vitro mammalian cell MN test (OECD guideline 487) [29] is a default assay for regulatory testing and a powerful investigative and screening assay in exploratory research. The in vitro MN test provides a comprehensive basis for investigating chromosome damaging potential. This is because both aneugens and clastogens can be detected in cells that have undergone cell division during or after exposure to the test chemical in cell culture. The in vitro MN test has several advantages compared with the standard chromosome aberration assay, being more simple and rapid to score and with greater statistical power due to increased sample sizes [30]. As such, it is considered the standard for assessing aneuploidy-inducing agents, which are difficult to detect with the in vitro chromosomal aberration assays. The positive control aneugens recommended in the test are the tubulin inhibitors colchicine (CAS: 64-86-8) and vinblastine (CAS: 143-67-9). The discrimination between clastogens and aneugens is possible by combining the classic MN test with kinetochore (CREST) staining or FISH using pan-centromeric probes to identify whether the MN is resulting from chromosome fragments or whole chromosome loss [13,31–33]. According to the International Conference on Harmonisation guideline ICH S2(R1), colchicine and vinblastine typically produce 70–80% or more centromere-containing (kinetochore positive) micronuclei. Chromosome non-disjunction can also be analyzed by using chromosome-specific centromeric probes and interrogation of FISH signals in the main nuclei of daughter cells. In addition, the recent developments of automated analysis of micronuclei either by using microscopy based image analysis systems or using flow cytometry offers high detection capability and more rapid processing for chemical treatment procedures [34,35]. The in vivo MN test (OECD guideline 474) [36] is a key assay in genetic toxicology for compounds where there is significant human

3.3.2. ToxTracker The ToxTracker assay is a panel of six validated GFP-based mouse embryonic stem (mES) reporter cell lines that can be used to identify the activation of specific cellular signalling pathways for detection of biological reactivity and potential carcinogenic properties of newly developed compounds in a single test [41]. The ToxTracker GFP reporter cell lines represent three distinct and major biological responses that are associated with carcinogenesis, i.e. DNA damage, oxidative stress and the unfolded protein response. ToxTracker is based on a 4

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

panel of biomarker genes that are preferentially activated upon exposure to different classes of carcinogens and toxicants. These biomarkers have been identified from extensive toxicogenomics studies [42]. Green fluorescent protein (GFP) mES reporter cell lines were generated which enable the activation status of these biomarker genes to be determined. The reporters were created using bacterial artificial chromosomes (BAC) that contain the complete biomarker gene including promoter and regulatory elements ensuring physiological regulation of the GFP reporters [43]. Genotoxicity is detected by upregulation of the Bscl2-GFP reporter that is activated by pro-mutagenic DNA lesions and DNA replication stress and by the Rtkn-GFP reporter that is associated with DNA double strand breaks. The Srxn1-GFP and BlvrbGFP reporters indicate activation of the Nrf2 and Hmox1 antioxidant responses. The Ddit3-GFP reporter is directly associated with the unfolded protein response and Btg2-GFP is activated as part of a p53mediated stress response [41]. Aneugens that act by a tubulin binding mechanism can be detected in ToxTracker by assessing the fold change in the induction of the Bscl2GFP and Rtkn-GFP genotoxicity reporters. In [41], the authors indicate that in contrast to direct DNA damaging agents that induce both DNA damage reporters after approximately 8 h, mitotic spindle disruptors predominantly activate only the Rtkn-GFP reporter after a significantly longer time (approximately 12 h). Therefore, by comparing the differential induction and kinetics of the Bscl2-GFP and Rtkn-GFP DNA damage reporters, ToxTracker has the potential to discriminate between direct DNA damaging agents and non-DNA reactive aneugenic tubulin binding agents. However, the assay in its standard form lacks the ability to detect aneugenic aurora kinase inhibitors. The ToxTracker assay can only identify the aneugenic potential of cell cycle kinases by the inclusion of a DNA stain or antibody against phosphorylated histone H3 (to assess cell cycle inhibition and polyploidy).

vitro biochemical selectivity data to the cellular context [48], which has led to the need to develop cellular kinase selectivity assays. Assays that measure the cellular activity of a particular kinase usually rely on flow cytometry or imaging to measure a substrate phosphorylation site with a phosphorylation site specific antibody. For AURKA and AURKB, pLATS2(Ser 83) and the pH3(Ser 28; Ser10), respectively, are often used in mitotic cells using quantitative imaging [49] or flow cytometry [50]. The calculated IC50 values are predictive of a compound's ability to disturb mitotic division due to aurora kinase inhibition and can be used to determine the concentration range where aneuploidy induction would be expected. 4. Current regulatory framework for testing on aneuploidy The current regulatory guidance for mutagenicity testing requires an assessment of a chemical's potential to induce aneuploidy [28,51]. However, the standard battery of tests for chemicals or pharmaceuticals does not include a test that is specifically designed to assess induced aneuploidy alone. Instead, an in vitro cytogenetic study or in vitro micronucleus test are considered sufficient to detect chemicals that have the potential to induce aneuploidy. With the last revision of ICH S2(R1), two alternative standard test battery options can be used for pharmaceutical testing. The Ames test and the in vivo erythrocyte micronucleus test are mandatory in both options and the latter assay can detect potential in vivo aneugens. In addition, an in vitro cytogenetic assay (option 1) or a second in vivo endpoint e.g. Comet assay (option 2) is mandated. Therefore, the potential to detect chemically induced aneugenicity is present in option one in two assays (in vitro and in vivo) and in option 2 only in one assay in vivo. The reason for the second option was that the in vitro mammalian cell genotoxicity tests, including the micronucleus test, were considered to have poor specificity i.e. there are many compounds that are positive in vitro but negative in in vivo genotoxicity tests and/or in rodent carcinogenicity tests. Neither option, however, is expected to affect sensitivity to detect chemical aneugens and both testing strategies are considered acceptable by regulators. Thus, a negative test battery is considered sufficient to provide reassurance that a test compound does not possess relevant mutagenic activity (including aneuploidy). With respect to Option 1, the in vitro micronucleus test [29] is considered suitable for the detection of aneuploidy. The in vitro chromosome aberration test [24] can be used as an alternative when detection of hyperdiploidy and polyploidy is sought. However, the mandatory in vivo micronucleus assays in ICH S2(R1) ensures adequate detection of aneugens even without an in vitro micronucleus test. Both ICH and REACH guidelines require the rate of hyperdiploidy and polyploidy to be measured in the in vitro chromosome aberration assay. The REACH guidance requires for in vitro tests for gene mutations in bacteria and in mammalian cells for Annex VIII, IX and X, and a test for effects on chromosome structure and number. For the latter, REACH suggests two alternative assays, an in vitro chromosome aberration test or an in vitro micronucleus test, although it states that the in vitro chromosome aberration test is not optimal to measure numerical aberrations. REACH then recommends “to conduct a CAbvitro or preferably a MNTvitro”. Usually, REACH considers a negative result in the 3 tests required – in vitro tests for gene mutations in bacteria and in mammalian cells, and a test for effects on chromosome structure and number – as sufficient evidence to exclude genotoxic potential. This may be interpreted that under REACH a negative in vitro chromosome aberration assay is sufficient to conclude that no relevant aneugenic potential is present. However, a chromosome aberration assay positive only for polyploidy may not necessarily indicate aneugenic potential and can simply indicate cell cycle perturbation that is commonly associated with cytotoxicity. Indeed, there is experimental evidence to suggest that polyploidy does not always reflect a potential for aneuploidy [52], especially when observed in mammalian cell culture in vitro at high levels of cytotoxicity [53]. As such, it is recommended that

3.4. Measuring mechanism of action (MOA) events associated with aneuploidy 3.4.1. Microtubule polymerization assay The tubulin polymerization assay is an in vitro method used to determine the ability of a compound to interact with tubulin and in many cases to alter one or more of the characteristic phases of polymerization [44]. This assay is used as a follow up mechanistic study to identify one of the classic potential molecular initiating events of aneuploidy and provides insights on the MOA for aneuploidy inducing compounds. Microtubules are critical components of the cytoskeleton and play several major roles in cells, including mitosis. Tubulin polymerization can be monitored in vitro by an increase in absorbance at 340 nm over time. The resulting polymerization curve is representative of the three phases of microtubule polymerization: nucleation, growth and steady state equilibrium. Spindle inhibitors have the potential to alter the different phases of the polymerization curve. The addition of paclitaxel (microtubule stabilizer) is seen to eliminate the nucleation phase, enhance the Vmax and increase overall polymer mass of the reaction. The microtubule destabilizing drug, nocodazole, is expected to reduce the Vmax and decrease polymer mass. 3.4.2. Aurora kinase B (AURKB) kinase selectivity assays Kinase inhibitors are an important class of targets in drug development and kinase selectivity remains one of the big challenges. For this reason, there are currently a number of kinase selectivity panels (for review see [45]) and many include assays for either aurora kinase A (AURKA) and/or aurora kinase B (AURKB). One of the most commonly used approaches is the measurement of the inhibition of substrate phosphorylation using isolated enzymes involving radiolabelled ATP [46]. It is also possible to measure the binding affinity of kinase inhibitors for a particular kinase using a fluorescently labelled competitive inhibitor and measuring the displacement by the compound of interest [47]. However, it is sometimes difficult to translate these in 5

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

positive results for in vitro polyploidy alone should be interpreted with caution [54]. A study to investigate the practical consequence of polyploidy in human peripheral blood lymphocytes in vitro showed that of 20 tests showing increased polyploidy, none produced evidence of aneugenicity [55]. The authors concluded that monitoring polyploidy alone was not a reliable marker of aneuploidy. Therefore, in case of a chromosome aberration assay positive only for polyploidy and negative for structural chromosome aberrations, a negative in vivo micronucleus assay with evidence of bone marrow exposure is considered as sufficient evidence for a lack of aneugenic potential [28]. The mammalian cytogenetic assays do have some limitations. As indicated above, chromosome loss is not reliably detected in the chromosome aberration assay. Both the in vitro and in vivo micronucleus assay are not specific for aneugens. In the micronucleus assay, single lost chromosomes are expected to appear as micronuclei and detected in this assay. Standard protocols only require evaluation of the number of micronucleated cells as the endpoint. Further evaluation of any aneugenic/clastogenic potential of compounds with positive micronucleus results is only performed when deemed necessary for risk assessment. In such cases, centromere-specific staining in interphase cells can be used to detect micronuclei containing whole chromosomes and indicate chromosome loss. In general, the induction of aneuploidy is considered to represent a mechanism with a practical threshold and the possibility to define a dose level below which no aneugenic effect is expected. This rationale is based on the target molecules of known aneugens. As outlined in detail elsewhere in this manuscript, major target molecules of aneugens are proteins such as tubulin or aurora kinases, and do not involve direct chemical reaction of the aneugen with DNA. For risk assessment purposes, the major aspects to consider are therefore the dose response curve of the micronucleus assays and a sufficient predominance of centromere/kinetochore positive micronuclei (see ICH S2(R1)) to conclude that an aneugenic mechanism is responsible for the increase in micronuclei. Further evidence on the mechanism may be needed if the compound causes chromosome loss or chromosome nondisjunction. Currently, it is not clear from scientific evidence whether the points of departure for both mechanisms are substantially different [56–59]. In regulatory risk assessment, an aneugen detected in somatic cells is considered to also have the potential to cause effects in germ cells. Specific convincing testing is only required when, for instance in pharmaceutical labelling, effects on germ cells are to be excluded. There is precedent for regulatory threshold-based human risk assessments for aneugens [60,61]. For example, the United Kingdom Committee on Mutagenicity (COM) provided advice to the Pesticides Safety Directorate (PSD) and the Veterinary Medicines Directorate (VMD) on methyl benzimidazole carbamates (MBCs), including benomyl, carbendazim and thiophanate-methyl. These chemicals act by interfering with microtubule formation during mitosis and the COM concluded that aneuploidy inducing chemicals (particularly those that function by interfering with the spindle apparatus of cell division) have a threshold mode of action [27]. This view was supported by in vitro studies measuring micronucleus formation, chromosome loss and gain and non-disjunction in bi-nucleate human lymphocytes [56,62,63]. Updates to COM's position on benomyl, carbendazim and thiophanatemethyl and the spindle inhibitors mebendazole and nocodazole have been published [64,65]. Case examples also exist for pharmaceuticals. For example, a referral procedure was made to the European Medicines Agency (EMA) to assess the safety of thiocolchicoside (TCC) containing medicinal products. The original assessment of the genotoxicity of TCC and its major circulating metabolite, 3-O-glucuronidated aglycone (M1), identified M1 as an aneugen with a Margin of Exposure (MoE) of 20× fold in patients. Despite aneuploidy being recognized as a potential risk factor for cancer, embryotoxicity and teratogenicity when impacting somatic cells, and teratogenicity, embryotoxicity and spontaneous abortions, as well as fertility, when impacting germ cells, the precedence for a threshold-based mechanism with aneugens and safety exposure margin

were considered sufficient to support continued marketing authorization. In 2014, the EMA received new preclinical data on the genotoxicity for a second metabolite of TCC, 3-demethylthiocolchicine (M2). The EMA assessed (i) the genotoxicity of M2 with positive in vitro and in vivo micronucleus assays and centromere positive micronuclei and (ii) whether the data were sufficient to define a no effect level (NOEL) for the aneugenic potential of M2 with a sufficient safety margin over clinical exposure. In this assessment, preclinical as well as clinical data were reviewed. M2 induced aneuploidy was clearly confirmed by centromere staining, supporting the proposed mechanism, but a NOEL could not be established in female rats and the lowest effect level (LOEL) corresponded to a MoE of only 1.6-fold for Cmax and 4.1-fold for AUC compared to human plasma concentrations (based on 8 mg TCC bid). A review of the concentrations of M2 resulting in aneuploidy in vitro, where a NOEL could be established, resulted in MoE based on the 3 h in vitro exposures of 3.8-fold but was below parity (i.e. < 1-fold) for the 24 h in vitro exposure. Given these data, the EMA concluded that additional studies to establish a threshold dose in vivo would not change the risk assessment. The assessment of chromosome non-disjunction (CND) or chromosomal loss (CL) to determine a threshold for aneuploidy was not addressed by the sponsors although the EMA commented that “while acknowledging that CND is the most appropriate end-point to investigate low-dose effects of spindle poisons, no conclusion for threshold doses for aneuploidy induction were possible”. The outcome of the referral was that risk minimization measures should be undertaken, and the EMA recommended that treatment should be restricted to patients over 16 years old and treatment duration limited to 7 days for oral and 5 days for intra muscular injection. The full report is available on the EMA website [66]. In more recent years, new approaches have been developed to define thresholds and points of departure for aneugens based on modelling dose response curves [67]. An example of such modelling was used to evaluate the benzimidazole drug, flubendazole and its metabolites [68]. The benzimidazoles are also discussed in our companion papers [6,10]. 5. Examples of adverse outcome pathways associated with aneuploidy in somatic cells 5.1. The adverse outcome pathway (AOPs) framework The adverse outcome pathways (AOPs) framework, first proposed by Ankley, Bennett, Erickson, Hoff, Hornung, Johnson, Mount, Nichols, Russom, Schmieder, Serrrano, Tietge and Villeneuve [69] was further developed by Vinken [70] and Sturla, Boobis, FitzGerald, Hoeng, Kavlock, Schirmer, Whelan, Wilks and Peitsch [71] with the view of using AOPs to transform safety risk assessment. This had parallels with previous developments, such as MoA, but it was also recognized that the AOP framework could reduce reliance on animal models [72] and improve human translation [73], as well as inform target pathway assessment in the pharmaceutical industry [70]. Moreover, the AOP framework provides a compelling communication tool for scientific engagement. Essentially a modular construct [74], AOPs link existing knowledge along a pathway of causally connected key events (KEs) between two points, a molecular initiating event (MIE) and an adverse outcome (AO). The overall objective is to support regulatory decision making, (reviewed by [75]). The MIE represents the initial interaction between a molecule and a biomolecule (e.g. DNA-adduct formation, enzyme inhibition, receptor/ligand engagement) and is the primary anchor in the AOP. The AO is the concluding anchor and represents an event of regulatory significance typical of an established protection goal or apical endpoint in a regulatory test [73]. In between are the KEs – intermediate steps along the pathway, typically at different levels of biological organization (molecular, cellular, tissue/organ, organism and/or population), which are also relevant for risk assessment. By 6

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

design, these are essential steps in the pathway but are not sufficient to cause the AO and they must be measurable. The links between KEs are termed key event relationships (KERs) and represent scientifically established causal relationships. Tailored Bradford-Hill considerations can be used to define these and include biological plausibility, empirical support for the KERs (i.e. dose–response, temporality, and incidence), and essentiality of the KEs in the pathway for progression towards the AO. KER may be governed by several factors, e.g., dose, target affinity, genetics, environment, etc. and as such, can be dynamic and quantitative. The ideal KE is a biomarker that can support translational biology, although this is not always possible. AOPs support the development of biological relationships that are systematic, structured, quality controlled and causal [75]. As such, they provide a high level of confidence in the knowledge contained in individual AOPs for regulatory decision making. AOPs are chemically agnostic, however, knowledge of specific compounds, their targets and biological effects can be used during their construction. Ultimately each AOP should reflect underlying biology and not be chemical-specific. In other words, AOPs describe toxicodynamic relationships [75]. For example, knowledge that colchicine is an inhibitor of microtubule polymerization can be used to collate data from empirical studies with other tubulin inhibitors, such as vinca alkaloids and benzimidazoles. At this stage, the resulting pathway would remain a chemical-specific case study. However, once enough evidence is collected that supports the occurrence of downstream key events following tubulin inhibition, the pathway becomes agnostic for any specific chemical. In other words, the pathway would state that once tubulin polymerization is perturbed to a sufficient extent and for a sufficient duration (by any agent), this would elicit the sequence of KEs and adverse outcome events as defined by the specific adverse outcome pathways. The MIE within an AOP has potential utility for QSAR analysis [76,77]. A specific MIE can lead to multiple adverse outcomes and any given adverse outcome can arise from multiple MIEs. Since biological pathways do not operate in isolation from one another, Villeneuve et al. [73,74] suggest that networks of AOPs, sharing common KEs and KERs, could be amalgamated as functional units of toxicity prediction. The inclusion of ADME (toxicokinetics) linking the biological effects (toxicodynamics) described by a given AOP would support individual risk assessments (reviewed by [78]) and define an integrated approach to testing and assessment for potential toxicants [79]. Indeed, there is considerable potential for extrapolation between in vitro exposures, in vivo physiologic exposure, whole organism responses, and long-term health outcomes [79]. Ultimately, AOPs could be used to develop quantitative and dynamic computational (systems) models to predict toxicity and inform risk assessment [80]; but this will require considerable acceptance and investment by the scientific community [81]. The Organisation for Economic Co-operation and Development (OECD) has published a Users’ Handbook Supplement to the Guidance Document for Developing and Assessing AOPs [82]. In addition, the OECD has developed a web-based tool and repository known as the AOP-wiki that is used to develop, evaluate, and share AOPs (https:// aopkb.org/aopwiki). Recommendations for best-practices in AOP development (e.g. [73] and [75]) and application of Bradford Hill considerations (e.g. [83]) have been published. There have also been critical discussions on the limitations of the current AOP framework [75,84]. Recent updates to the OECD Handbook and AOP-wiki, address some of these challenges and support the future trajectory of the AOP framework in risk assessment. There are multiple molecular/cellular targets that can be linked with aneuploidy (Fig. 1). There is also precedence for the development of AOPs in genetic toxicology [85] and the IWGT workgroup was aware of OECD AOP 116, which describes the events leading to aneuploidy in offspring [18]. As such, the IWGT recognized the opportunity to develop complementary AOPs for aneuploidy in somatic cells. Indeed, for some targets e.g. inhibition of tubulin polymerization and aurora

kinases, there is a considerable body of empirical evidence in the public domain. For example, the ECETOC study [86] and others e.g. [87] provide a rich source of cheminformatics on chemical aneugens and outcomes in genetic toxicology studies. Furthermore, additional information on the fundamental mechanisms of mammalian cell division has emerged in recent years (e.g. [21,88]). In contrast, for other mechanisms, e.g., Polo kinase inhibition and CDK 1 inhibition, the empirical evidence is more limited, or access to data is restricted to the pharmaceutical industry (these targets are the focus for the development of novel chemotherapies). This led the IWGT workgroup to an early decision on where AOPs could be developed in this area and therefore, two MIEs i.e. chemical binding to (a) tubulin and (b) the catalytic domain of aurora kinase B, were selected. The second decision was to limit the presentation of the AOPs to defined AOs within the cell and a description of the KE, but that a full assessment of the evidence supporting the KERs, which is critical to AOP development, was considered beyond the scope of this paper. The intention is to fully develop these and submit to the AOP-wiki in due course. Lastly, there was a decision not to extend the AOPs to AOs in humans, such as cancer or hereditary disease, at this time. 5.2. Adverse outcome pathways related to tubulin binding Microtubules are polymers composed of α- and β-tubulin protein heterodimers arranged in a head-to-tail concatemer to form on average 13 (range 9–16, depending on cell type) liner and parallel protofilaments which combine, in a helical manner at a higher order structure, to form hollow cylindrical filaments [89]. Microtubules play an important role in several cellular processes including the maintenance of cell morphology and polarity and in supporting cell migration and intracellular trafficking. In cell division, microtubules play a crucial role in the formation of the mitotic spindle, where, together with other associated proteins and centrioles, they contribute to the stability of the spindle and dynamically to the separation of chromosomes during the process of mitosis [90]. The role of microtubules during meiosis is described in detail in the second paper from the aneugen workgroup [6]. The polymerization of microtubules is an extremely fluid process, with tubulin dimers added at one end (plus end), while dimers are removed from the other end (minus end); this process is critical for the proper functioning of microtubules [18]. Tubulin and the formation of mitotic-spindle microtubules represent a long established and successful target for cancer therapy. Various microtubule-targeting agents, such as vinca alkaloids and taxanes, have been used clinically for many decades, whereas other microtubuletargeting agents, such as colchicine, have been used in the treatment of acute gout and other inflammatory conditions. Microtubule-targeting agents can be differentiated according to their binding sites in tubulin. There are thought to be ∼5 binding sites exemplified by the vinca alkaloids, colchicine, the taxanes, the epothilones and laulimalide [91]. These agents can be further classified based on their effects on microtubule polymerization as microtubule-destabilizing agents or microtubule-stabilizing agents (reviewed by Perez [92]). Examples of microtubule-destabilizing compounds are shown in Table 1 and include colchicine and the vinca alkaloids. Colchicine was the first microtubule poison to be discovered and was originally extracted from the Meadow saffron, Colchicum autumnale. Colchicine binds to tubulin at the interface of the α and βtubulin, near the GTP-binding site of α-tubulin, and perturbs the assembly dynamics of microtubules [93,94]. The medicinal uses of colchicine and ADME in humans were recently reviewed by Marchetti et al. [18]. The vinca alkaloids, which bind to β-tubulin at a site which is distinct from colchicine, but still close to the GTP binding sites, were originally extracted from the Periwinkle, Catharanthus roseus, and include the first-generation compounds vinblastine and vincristine and the second-generation compounds, vindesine, vinorelbine, and vinflunine. Vinblastine and vincristine were shown to be potent 7

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

Table 1 Aneugens that bind tubulin and are known spindle poisons in mammalian cells. Compound

CAS number

In vitro/in vivo

Cell system

References

Colchicine

64-86-8

Carbendazim

10605-21-7

Nocodazole

31430-18-9

Mebendazole Benomyl

31431-39-7 17804-35-2 98-95-3 and 100-47-0 33069-62-4

Bisphenol-A

80-05-7

Rotenone

83-79-4

In vitro

Vincristine

57-22-7

Vinblastine

865-21-4

In vivo In vivo In vivo

Human lymphocytes MN in mouse peripheral blood Human lymphocytes Human lymphocytes MN in mouse bone marrow Human lymphocytes Human lymphocytes and mouse splenocyes Human lymphocytes Human lymphocytes MN in mouse bone marrow MN induction in V79 cells Human lymphocytes Mouse splenocytes MN assay in AHH-1, MCL-5 and V79 cell lines tripolar mitotic spindle induction in V79 cells MN assay in V79 cell line Spindle induction in V79 cells MN in mouse bone marrow MN in mouse peripheral blood MN in mouse peripheral blood

[62,63] [197] [62,63] [56] [198,199] [62,63] [200] [62,63] [56]

Nitrobenzene and benzonitrile Paclitaxel

In In In In In In In In In In In In In In

vitro vivo vitro vitro vivo vitro vitro vitro vitro vivo vitro vitro vitro vitro

[201] [200] [200] [107] [107] [202] [197] [197]

modes: dynamic instability, displayed by the plus end, and treadmilling, displayed at both ends. When guanosine triphosphate (GTP) is bound to β-tubulin subunits at the plus end of a microtubule, other GTP-bound heterodimers readily attach and the microtubule grows. Addition of a new dimer triggers the hydrolysis of GTP on the subunit that is no longer exposed and guanosine diphosphate (GDP) is formed. However, if a GDP-bound subunit becomes exposed at the plus end it will cause a conformational change resulting in the rapid depolymerization of the microtubule. This is referred to as dynamic instability. Treadmilling, however, signifies loss of subunits from the minus end and gain of subunits at the plus end, with no net change in the overall length of the microtubule [91]. Colchicine binds to the interface of the α and β-tubulin to give a soluble intermediate complex at low concentrations. The tubulin in this intermediate complex undergoes polymerization, but the colchicine slows down the polymerization rate and interferes with microtubule dynamics. At higher concentrations, colchicine induces a net depolymerization in microtubules which affects the subsequent elongation of the microtubule [101]. Vinca alkaloids are structurally distinct from colchicine and bind to a different site on tubulin. At low concentrations, binding occurs at the plus ends of microtubules producing a conformational change of dimers [102]. At higher concentrations, vinca alkaloids also have affinity for free tubulin heterodimers [103] which promotes microtubule depolymerization. Thus, both colchicine and vincas demonstrate a dual action, depending on concentration, which is common with other spindle poisons. The effects of spindle poisons on microtubule dynamics within the cell have been reviewed by [101]. The ability of chemicals to affect tubulin polymerization and impact microtubule dynamics can be evaluated in vitro using spectroscopy, typically using tubulin protein extracts derived from porcine brain tissues. Commercial assay kits are readily available (e.g. Tubulin Polymerization Assay Kit, Cytoskeleton Inc., Denver, USA or similar) or alternatively, studies can be commissioned from Contract Research Organizations. Both colchicine and vinblastine prevent porcine brain tubulin assembly [104–106]. KE2: Impaired chromosome attachment to the spindle. This occurs at the cellular level because of disruption to microtubule dynamics causing mitotic spindle abnormalities and impaired chromosome attachment to the spindle (KER2). Abnormalities in mitotic spindle structure become apparent in cells exposed to chemicals that bind to tubulin and disrupt microtubule dynamics. Several types of abnormality have been described including a reduction in microtubule density and the loss of the typical barrel shape

chemotherapeutic drugs over 50 years ago and vincristine was approved by the US Food and Drug Administration (FDA) in 1963. Antimicrotubule agents remain as front-line therapies against several malignancies, either by themselves or as combination therapies. Several AOPs have been developed that focus on mammalian somatic cells and share a common MIE – chemical binding to tubulin. The AOPs focus specifically on chemicals that bind to tubulin to cause the depolymerization (destabilization) of microtubules, rather than stabilization. Destabilization of microtubules leads to spindle abnormalities and malfunction, including abnormal mitosis [95] which are common events in all of the AOs. Fig. 2 presents these relationships as a network and Table 2 describes the KE(s) and KER(s) and AOs in that network. A description of the individual AOPs is presented below and is structured so as to reflect the sequence of events leading to the various AO. Each AOP was developed primarily using data derived from colchicine or the vinca alkaloids, although the mechanism of action is consistent with other microtubule-destabilizing agents listed in Table 1 (e.g. benzimidazoles). The genetic outcomes have been observed across several phyla, including genetic toxicology studies in human cells in vitro or clinical studies in human patients in vivo. The high homology in somatic cell tubulins suggests that the MIE and KEs are likely to be conserved and that they are therefore relevant for human risk assessment. 5.2.1. AOP 1: chemical binding to tubulin leading to somatic cell death MIE: Chemical binding to tubulin. This MIE occurs at the molecular level in the cytoplasm of cells. Tubulins represent a large superfamily, and several isotypes are described for both α and β tubulin in mammalian cells [96]. The currently known microtubule-disrupting agents bind to all isotypes, having only a slight preference for one over another [97]. The molecular binding properties of colchicine to tubulin, including a description of colchicine binding dynamics and characterization of the colchicine binding domain on tubulin have been reviewed in detail by Marchetti et al. [18]. The reader is referred to Lu et al. [98] for an overview of other tubulin inhibitors that interact with the colchicine binding site. The binding properties of various vinca alkaloids with tubulin have also been described in the literature [99,100]. KE1: Disruption of microtubule dynamics. This molecular event also occurs in the cytoplasm of cells as a result of chemical binding to tubulin causing microtubule depolymerization leading to disruption of microtubule dynamics (KER1). As mentioned above, the polymerization of microtubules is an extremely fluid process. Microtubule ends can undergo rapid polymerization and depolymerization. They behave in one of two different 8

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

Fig. 2. The AOP network that shares a common MIE – chemical binding to tubulin. Descriptions are provided in the body text for AOP-1 chemical binding to tubulin leading to somatic cell death, AOP-2 chemical binding to tubulin leading to somatic cell polyploidy and cell senescence, AOP-3 chemical binding to tubulin leading to somatic cell aneuploidy, AOP-4 chemical binding to tubulin leading to somatic cell clastogenicity and AOP-5 chemical binding to tubulin leading to chromothripsis. The KEs, KERs and AOs in that network are described in Table 2. Dashed lines represent some potential secondary consequences (described in the body text) whereas dotted lines represent a theoretical KER.

associated with the mitotic spindle, in part because of the reduction in the distance between the spindle poles. In addition, an increase in the formation of monopolar or multipolar spindles can often be observed [107]. These defects in mitotic spindle formation and microtubule dynamics impair normal chromosome attachment during pro-metaphase and the subsequent organization of chromosomes at the metaphase plate. A differential cell staining technique that allows for the simultaneous visualization of chromosome and spindle fibres [108] was used to investigate the action of aneugens on spindle formation in cultured cells [109–111]. The assay described both spindle abnormalities and impacts on chromosome segregation (e.g. lagging chromosomes) and its utility was further evaluated using a battery of aneuploidy and polyploidyinducing agents as part of the EEC 4th Environmental Research and Development Program [25]. However, the assay has largely been superseded by fluorescence microscopy methods using antibodies to proteins with a known role in spindle formation and function, e.g., fluorescence tagged anti-tubulin antibodies together with specific, fluorescent, nuclear and/or cytoplasmic dyes (reviewed by [112]). More recently, fluorescent speckle microscopy (FSM) has been intensively used to investigate macromolecular dynamics in vitro, such as microtubule flux during mitosis (reviewed by [113]) whereas live cell imaging has been used to investigate mitotic protein dynamics and function (reviewed by [114,115]). KE3: Mitotic arrest. This occurs at the cellular level when impaired chromosome(s) attachment to the spindle activates the Spindle Assembly Checkpoint (SAC) and leads to mitotic arrest (KER3). The SAC, also known as the mitotic checkpoint, serves as a surveillance mechanism at the metaphase–anaphase transition. During normal metaphase, the centromeres of chromosomes align on the metaphase plate at the equator of the cell before their chromatids will be separated into each of the two daughter cells. Balanced chromosome alignment is required for mitotic fidelity. This is achieved by correct connections and counterbalance of tensions generated by the action of

opposing kinetochore microtubules, and only when the conditions for faithful chromosome segregation have been met does the cell enter anaphase. Cells with defects in the mitotic spindle trigger the SAC. A single unattached or incorrectly attached chromosome is sufficient to activate the SAC in normal cells and block progression to anaphase until all of the chromosomes are correctly positioned at the metaphase plate [116]. The main components of the SAC represent a complex signalling network which operates by inhibiting the E3 ubiquitin ligase activity of the anaphase promoting complex (APC) which normally targets mitotic regulators such as cyclin B1 and securin for degradation to allow progression (reviewed by [116,117]). SAC activation leads to mitotic arrest until the conditions for successful progression are resolved. Mitotic arrest can be measured using the mitotic index in cytogenetic assays. Tubulin inhibitors, like colchicine and vincristine, are known to inhibit spindle formation and induce mitotic arrest in cultured mammalian cells; this results in increases in the mitotic index compared with vehicle or untreated control cultures [25]. KE4: Mitotic catastrophe. This occurs when continued mitotic arrest leads to mitotic catastrophe i.e. activation of pro-apoptotic signalling (KER4). If the conditions for successful metaphase-anaphase transition cannot be met, for example due to defects in the mitotic spindle, the cell undergoes mitotic catastrophe, a specific form of apoptosis associated with failed mitosis [118]. In some instances, arrested cells die without exiting mitosis in a process known as mitotic death. One key determinant that dictates the outcome of mitotic catastrophe involves cyclin B1 binding to cyclin-dependent kinase-1 (CDK1), required for mitotic arrest in metaphase, and its interaction with pro-survival and pro-apoptotic signalling pathways, especially those from the bcl-2 protein family. This cross-talk between the mitotic checkpoint and cell cycle signalling pathways controls cell fate. The ‘competing networksthreshold’ model provides a plausible explanation of how a cell might either die in mitosis or undergo mitotic slippage following SAC 9

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

Table 2 List of key events, key event relationships and adverse outcomes linked to the molecular initiating event: chemical binding to tubulin in mammalian somatic cells.

Colour schemes (MIE in green and AO in orange) are used to cross reference with Fig. 2. The KE terms mapped to ontology terms via Event Component Model are based on the conventions described by Ives et al. [203].

activation and mitotic arrest (reviewed by Topham and Taylor [119]). Downstream cell death can manifest in caspase-dependent or -independent apoptosis, mediated by p53 status with both involving mitochondria, depending on the balance between lethal and cytoprotective signals during mitotic arrest (see Miller et al., [97] and Nakayama and Inoue [120] for more detailed reviews). Clearly, mitotic catastrophe plays an important role in tumour suppression through the control of failed mitosis and avoidance of potential genomic instability [117]. Targeting tubulin and mitotic catastrophe remains an active area

of research in cancer therapy as tumour cells are often more sensitive to the lethal effects of anti-tubulin drugs compared with normal cells [121]. The AO of AOP-1 is cell death (KE5 and AO-1). This occurs at the level of the cell when activation of pro-apoptotic signalling leads to apoptosis, i.e. programmed cell death (KER5). Mitotic catastrophe involves hallmarks of apoptosis and the activation of caspase-2 prior to mitochondrial membrane permeabilization and the downstream events leading to cell death [118,121,122]. 10

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

Apoptosis is characterized by two major biochemical events, namely the activation of caspases and the permeabilization of the outer mitochondrial membrane, with the consequent release of multiple death effectors into the cytoplasm invoking cytoplasmic shrinkage, chromatin condensation and nuclear fragmentation which morphologically define programmed cell death. Measures of apoptosis include various cellular assays, e.g. Annexin V and Caspase signal detection or the Terminal deoxynucleotidyl transferase (TdT) dUTP Nick-End Labelling (TUNEL) assay, which detects apoptotic cells that have undergone extensive DNA degradation during the later stages of apoptosis (see https://www.sigmaaldrich. com/technical-documents/articles/biology/apoptosis-assays.html). MoA assays, such as Multiflow, include biomarkers that provide an index of cell death via apoptosis. Colchicine and vincristine [39] or nocodazole and carbendazim [123,124] are examples of tubulin inhibitors shown to induce concentration-dependent increases in apoptotic cell frequencies.

naturally in some cell types (e.g. hepatocytes, megakaryocytes, muscle cells, etc.) [61,132]. 5.2.3. AOP 3: Chemical binding to tubulin leading to somatic cell aneuploidy AOP 3 shares the same MIE, KE1, KE2 and KE3 as AOP 1. KE9: Asymmetric chromosome segregation. This occurs at the level of the cell. Following prolonged mitotic arrest, cells generally die in mitosis via apoptosis, i.e. the ‘mitotic death’ programme associated with mitotic catastrophe (KE4). However, in some instances, mitosis-incompetent cells can escape the mitotic arrest, either by ‘slipping’ into the next interphase (KE6) or they can progress to anaphase and undergo a round of aberrant mitosis [117]. This can occur because of silencing of the SAC due to cyclin B1 degradation, impaired checkpoint signalling or by inhibition of apoptosis. For example, APC-induced degradation of cyclin B1 can trigger cells to exit mitosis [133,134]. The suppression of mitotic catastrophe, e.g. by SAC silencing, directly promotes asymmetric chromosome segregation (KE9) if the onset of anaphase occurs in the absence of a functional spindle or incomplete kinetochore microtubule attachment(s). Alternatively, merotelic attachments in anaphase, i.e. when a single kinetochore becomes attached to microtubules from opposing poles of the spindle apparatus, can lead to the formation of lagging chromosomes and/or chromosome bridges if left unresolved [135]. Lagging chromosomes can also result from mono or multipolar spindles due to errors in centrosome replication which can be induced by pericentriolar matrix fragmentation in the case of tubulin inhibitors [95]. Lagging chromosomes, chromosome bridges and multipolar spindles can be visualized microscopically in mitotic cells exposed to tubulin inhibitors, such a colchicine and vinblastine, in vitro [25]. Thus, SAC silencing in conjunction with abnormal spindle function leads to asymmetric chromosome segregation (KER9). KE10: Micronucleus formation or chromosome non-disjunction. This occurs at the level of the cell. The failure of proper kinetochore attachments can result in asymmetric chromosome segregation and the formation of unbalanced complements of chromosomes at each pole. If the level of spindle damage is very severe, the cell can undergo multinucleation, an outcome that often prevents cytokinesis and leads to cell death (KE5 and AO-1). However, under less severe conditions asymmetric chromosome segregation can lead to (a) chromosome non-disjunction once the nuclear membrane is re-established or (b) the formation of micronuclei because the lagging chromosomes are not incorporated into the main nucleus [136]. Thus, following (a) or (b) above, the resulting asymmetric chromosome segregation, on completion of telophase and cytokinesis, leads to micronucleus formation or chromosome non-disjunction (KER10). In turn, chromosome non-disjunction and/or micronucleus formation results in aneuploidy (KE11 and AO-4) in one or both daughter cells, i.e. the generation of a daughter cell that contains an abnormal number of chromosomes. For example, in cells exposed to low concentrations of nocodazole, microtubule polymerization is disturbed resulting in a spindle with damaged microtubules. This can give rise to chromosome loss and non-disjunction [128]. Indeed, the induction of aneuploid cells has been determined following treatment with various tubulin inhibitors (see Table 1), including colchicine and vinca alkaloids. Evidence for the KEs and KERs in AOP-3 comes from a wide range of studies e.g. quantification of metaphase chromosomes in primary mammalian cell cultures [25], induction of micronuclei using kinetochore labelling in primary and cultured rodent cells, evidence for non-disjunction in cytokinesis-blocked human peripheral blood lymphocytes using fluorescent in situ hybridization labelling [62,63]. In vivo studies with rodent micronucleated erythrocytes using kinetochore labelling [137,138] or FISH analysis [137,139,140] have also been reported with colchicine and vinblastine, but in the experience of the

5.2.2. AOP 2: Chemical binding to tubulin leading to somatic cell polyploidy and cell senescence AOP 2 shares the same MIE, KE1, KE2 and KE3 as AOP 1. KE6: Mitotic slippage. This occurs at the level of the cell. Following pro-longed mitotic arrest, cells either die in mitosis via apoptosis (see AOP-1) or they can exit mitosis in a process called mitotic slippage [117]. Here, cells escape mitotic arrest and ‘slip’ to the next interphase without undergoing proper chromosome segregation or cytokinesis i.e. mitotic slippage (KE6). Mitotic slippage has been observed with aneugens [125] and various anti-tubulin drug therapies [126]. Mitotic slippage leads to the formation of tetraploid cells and the first adverse outcome of AOP-2 is polyploidy (KE7 and AO-2) since the resulting cells contain twice the haploid number of chromosomes. It is known that both colchicine and vinblastine are potent polyploidy inducing agents in mammalian cells in vitro [127]. A key molecular factor for mitotic slippage is TP53 deficiency, which is permissive for the bypass of mitotic catastrophe and the formation and survival of tetraploid cells. For example, exposure to high doses of the spindle poison nocodazole results in tetraploidy due to mitotic slippage in the absence of a functional spindle [128] whereas colcemid treatment often results in an increase of C-mitosis, an artificially induced abortive nuclear division in which the chromosome number is doubled [129]. In some instances, tetraploid cells proliferate but remain vulnerable to mitotic catastrophe, mitotic slippage or further aberrant cell divisions in subsequent mitosis, potentially leading to the generation of aneuploid “grand-daughter” cells at the next cell division [117]. Thus, tetraploidy can be viewed as a metastable intermediate between normal diploidy and aneuploidy [1]. Tetraploidy has been linked with cancer progression [128,130] and this is discussed further in the accompanying paper [10]. However, the extent to which aneuploidy occurs following exposure to chemicals that induce tetraploidy is unknown as the fate of polyploid cells is not normally tracked in standard cytogenetic assays used in genetic toxicology. As a consequence of sustained stress within tetraploid (polyploid) cells a number of cellular responses governed by the p53 and pRB tumour suppressor proteins become activated [131]. This triggers a signalling pathway that induces a stable cell cycle arrest preventing further cell proliferation i.e. cell senescence (KE8 and AO-3), and the second AO of AOP-2. Thus, polyploidy activates p53 and pRB cellular responses leading to cell cycle arrest i.e. cell senescence (KER 8). The prevalence of cell senescence in standard regulatory assays is unknown as it is not routinely measured in genetic toxicology. Nevertheless, AO-3 is an important outcome in cancer risk assessment because cell senescence is thought to limit tumourigenesis (reviewed by Cheng and Crasta [126]). Lastly, it should be recognized that polyploidy may be induced by other types of cellular insult like general cytotoxicity [55]. Somatic cell polyploidy is generally uncommon in mammalian cells but can occur 11

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

IWGT Workgroup these types of assay are seldom performed for regulatory purposes. In addition, there is evidence for micronucleus induction in vivo in rodents with other tubulin inhibitors and also some evidence in human studies [141]. Indeed, colchicine and the vinca alkaloids were used as positive controls for aneuploidy induction as part of the ECETOC research programme on aneugens [37] and in an interlaboratory study organized by the IWGT [87]. Indirect evidence also comes from rare human genetic disorders involving mutations in key checkpoint proteins in mitosis e.g. Mosaic Variegated Aneuploidy (MVA), which are characterized by high levels of somatic cell aneuploidy [142,143]. It is important to note that aneuploidy negatively impacts cellular fitness and micronucleated cells constitute a strong apoptotic signal [124]. Most aneuploid cells are therefore non-viable and trigger apoptosis (KE5 and AO-1) or growth inhibition via cell senescence (KE8 and AO-3). In rare instances, aneuploid cells can proliferate but they are often highly susceptible to further mitotic catastrophe (KE4), mitotic slippage (KE6) or another round of aberrant mitosis at the next cell division [117]. These serve as additional obstacles to cell survival and likely represent evolutionary back-ups to failed mitotic catastrophe, contributing to tumour suppression and avoidance of potential genomic instability associated with aberrant mitosis. As such they are important considerations for understanding the risks of somatic cell mutation associated with exposure to chemical aneugens and cancer susceptibility. This topic is explored further in our companion paper [10].

5.2.5. AOP 5: Chemical binding to tubulin leading to somatic cell chromothripsis AOP 5 shares the same MIE, KE1, KE2 and KE3 as AOP 1 and KE9 and KE10 of AOP 3. As mentioned above, lagging chromosomes can result in the formation of micronuclei in one or more daughter cells. Micronuclei can show abnormalities in DNA replication, transcription, and nuclear envelope (NE) structure, and often display DNA damage (reviewed in [14,151]). Importantly, micronuclei show frequent nuclear envelope collapse [152]. The membrane rupture of micronuclei containing lagging chromosomes can lead to chromothripsis (KER14). Thus chromothripsis (KE14 and AO-6) concludes AOP-5 and represents a process where one or more localized chromosomal regions undergo catastrophic shattering followed by haphazard and indiscriminate repair via DNA fragment ligation [153]. This results in chromosome(s) with random sequence order and orientation [60,154]. Genetic evidence for chromothripsis has been observed in various tumours and it is thought to occur during a single cell cycle. As such, chromothripsis can represent a super mutational event linked with tumour progression [155]. The loss of TP53 appears to be a pre-requisite, and most non-transformed cells do not survive such a catastrophic event (reviewed in Ly and Cleveland [60]). Another potential outcome of chromothrispsis is the production of “double minutes” [154], which have been observed in a wide range of tumours. The incidence of chromothripsis associated with exposure to tubulin inhibitors has not been characterized and therefore the AO remains theoretical in terms of the MIE–chemical binding to tubulin. In the absence of empirical evidence, further research will be required to substantiate AOP-5.

5.2.4. AOP 4: Chemical binding to tubulin leading to somatic cell clastogenicity AOP 4 shares the same MIE, KE1, KE2 and KE3 as AOP 1 and KE9 of AOP 3. KE12: DNA damage in lagging chromosomes and chromosome bridges. Cells that escape mitotic arrest at metaphase via SAC silencing can experience asymmetric chromosome separation if the structure/ function of the mitotic spindle remains compromised or if chromosomes possess improperly attached kinetochores (see AOP-3), leading to the formation of lagging chromosomes or chromosome bridges. The lagging chromosomes can become trapped in the cleavage furrow during cytokinesis if they fail to clear the spindle midzone. These chromosomes can be subject to DNA strand breaks leading to the formation of chromosome aberrations [144]. DNA damage can also occur as a result of mechanical damage to chromosome bridges resulting from impaired segregation of dicentric chromosomes or inappropriate resolution of DNA ultrafine bridges [145]. DNA damage in lagging chromosomes(s) or chromosome bridges results in clastogenicity, i.e. structural chromosome damage in one or both daughter cells (KER13). Thus clastogenicity (KE13 and AO-5) concludes AOP-4. Lagging chromosomes sometimes lead to the formation of chromosome bridges. In these cases, often, cell division cannot be completed, and cytokinesis fails. This can lead to the formation of tetraploid cells and the sequelae (KE6 and KE7/AO-2) described in AOP-2. The evidence linking aneuploidy to clastogenicity have often been described in a cancer setting. For example, following mitotic slippage, multinucleated cells have been shown to harbour DNA damage [146]. In terms of tubulin inhibitors, colchicine induces lagging chromosomes in human lymphocytes in vitro [127] and in the mouse [147] and was reported to induce structural chromosome aberrations in CHO cells in vitro [148] but not in mouse bone marrow cells after single intraperitoneal injection [149]. Vinorelbine, a synthetic vinca alkaloid had no effect on structural chromosome aberrations in human lymphocytes [150]. As such, the empirical data for AOP-4 and chemical tubulin inhibitors is somewhat conflicting with respect to somatic cell effects, and further research will be required to confirm the association with clastogenicity and understand any potential dose response relationships.

5.2.6. Summary It is clear from the five AOPs described above that a single MIE i.e. chemical binding to tubulin can have multiple adverse outcomes in somatic cells, i.e. cell death, polyploidy and cell senescence, aneuploidy, clastogenicity and (theoretically) chromothripsis. Of course, it is unlikely that the pathways perturbed by tubulin binding operate in isolation from one another; however, as a network (see Fig. 2) they represent a functional unit of prediction for many real-world scenarios of genotoxicity, especially given their shared and inter-related KEs and KERs. The extent to which each adverse outcome is dependent on exposure and other cell factors e.g. cell type and origin (wild type or tumour) is not fully understood, but it is clear that p53 status is an important determinant for several of the AOs described. 5.3. Adverse outcome pathway related to binding to the catalytic site of aurora kinase B (AURKB) 5.3.1. Introduction Mitotic kinases like CDKs, Polo-like kinases (PLKs), aurora and NEK kinases are essential for the proper timing and fidelity of mitosis. Kinases are involved in the regulation of the centrosome cycle, the formation of the mitotic spindle and the proper bipolar attachment of the sister chromatids mediated by the SAC [156]. As indicated above, the SAC makes sure that all chromosomes are properly aligned and is controlled by the BUB kinases (BUB1 and BUBR1), AURKB and the kinetochore kinase MPS1 [157]. In addition to their essential roles in early apoptosis, kinases like PLK1 and AURKB appear to co-ordinate mitotic progression and cytokinesis by activating the centralspindlin/ ECT2/RhoA pathway [158]. Because of their importance in key events of mitotic division, disruption of mitotic kinases could be a major factor leading the chemical induction of aneuploidy. Kinase targets have been pursued in great numbers for different indications and efforts have usually targeted the conservative ATP binding pocket with the goal to specifically inhibit a single kinase [159]. However, in addition to affecting the target kinase, most kinase inhibitors inhibit other kinases in the kinome [159]. A study with 12

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

Fig. 3. AOP-6: Chemical binding to catalytic domain of AURKA/B leading to somatic cell aneuploidy.

Table 3 List of key events, key event relationships and adverse outcome linked to the molecular initiating event: chemical binding to AURKA/B in mammalian somatic cells.

Colour schemes (MIE in green and AO in orange) are used to cross reference with Fig. 2. The KE terms mapped to ontology terms via Event Component Model are based on the conventions described by Ives et al. [203]. * Misalignment chromosomes end up in a binucleated polyploid cell in form of micronuclei or are incorporated in one of the 2 daughter nuclei.

commercially available kinase inhibitors tested against a panel of 300 recombinant protein kinases revealed that not all off-target kinases are hit with the same frequency, which allows for the discrimination of promiscuous and non-promiscuous kinases [160]. Some of the most frequently affected kinases in this context are the aurora kinases, which together with their fairly well understood biology make them a prime target for AOP development.

causes monopolar spindles and G2M arrest in mitosis [164,165]. AURKB is part of the chromosomal passenger complex (CPC) and is critical for chromosome condensation, chromosome bi-orientation on the mitotic spindle by correcting kinetochore-microtubule attachment errors, the spindle assembly checkpoint, as well as the final stages of cytokinesis [166–169]. AURKB inhibition leads to the inhibition of H3 phosphorylation at histone 3 Ser10 and Ser28, misaligned chromosomes and binucleated cells through the inhibition of cytokinesis [170]. AURKC has low expression levels in somatic cells but is highly expressed in meiotically dividing gametes [171]. Overexpression of AURKC in somatic cells but not ablation induces abnormal cell division resulting in centrosome amplification and multinucleation in cells [172]. So, which Aurora kinase is responsible for the effects observed after treatment with pan-aurora kinase inhibitors? Studies indicate that the cellular events that occur following exposure to aurora kinase inhibitors like ZM447439 and Hesperadin are phenotypically similar to the inhibition of AURKB, rather than AURKA [173,174]. For example, while

5.3.2. General overview of aurora kinases The aurora kinases are a family of highly conserved serine/threonine kinases that are important for the accurate distribution of chromosomal material during mitosis and meiosis [161]. Three aurora kinases, AURKA, AURKB, AURKC, with distinct biological functions and sub-cellular localization have been identified in mammalian cells to date [162]. AURKA plays a key role in centrosome function, spindle assembly in mitosis, the progesterone-induced oocyte maturation and metaphase I spindle orientation in meiosis [163]. Selective inhibition of AURKA 13

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

inhibition of AURKA by genetic approaches in cells results in centrosome separation defects, cells exposed to Aurora-kinase inhibitors can assemble bipolar spindles and enter mitosis with normal kinetics [165,175–177].

CPC through its roles in central spindle formation, regulation of furrow ingression, and abscission [184–186]. Depletion of AURKB by siRNA results in failure to establish a midbody matrix and accumulation of microtubule bundles with the plus ends of microtubules extending throughout the area where the midbody would normally form without clear boundaries [187]. The failure to form a functional midbody (i.e. abnormal mitosis) prevents cytokinesis and leads to binucleation (KER3) i.e. the formation of a binucleated daughter cell that is tetraploid [188,189]. KE4: Multipolar spindles in 2nd mitosis. Tetraploid cells generated through cytokinesis failure are relatively unstable compared to their diploid counterparts, in part as a result of doubling of their centrosomes. As such, cells frequently become aneuploid upon continued cell division [190] as they encounter a problem in the next mitosis with multipolar spindles and a high frequency of mitotic failure [191,192]. Consequently, cells that experienced cytokinesis failure subsequently suffer from mitotic abnormalities in the following cell cycle. This is because binucleation causes centrosome duplication and multipolar spindles in the 2nd mitotic division (KER4). Many of these cells are likely to be removed from the cycling population [193] by some of the same mechanisms described in AOPs 1 and 2. KE5: Micronucleus formation or chromosome non-disjunction. Interestingly, most cells with multipolar spindles will undergo bi-polar anaphases but show a high frequency of merotelic kinetochore attachments, which will produce chromosome mis-segregation due to anaphase lagging [192], resulting in the formation of micronuclei or chromosome non-disjunction (KER5) following telophase and cytokinesis. This results in the generation of one or more daughter cells that contains an abnormal number of chromosomes i.e. aneuploidy (KE6 and AO-1), which concludes AOP-6. Aurora kinase inhibitors are highly effective inducers of aneuploidy and polyploidy in a number of different cell systems. For example, Gollapudi, Hasegawa and Eastmond [194] showed that the two known aurora kinase inhibitors, VX-680 and ZM-447439, induced CREST positive micronuclei, hyperdiploidy and polyploidy in the human TK6 lymphoblastoid cell line. Similarly, treatment of HCT-116 cells with the AURKA specific inhibitor MLN8054 led to the induction of micronuclei and aneuploidy [195]. Overall the responses with aurora kinase inhibitors can be characterized as quite different from the tubulin binding aneugens. While tubulin binders block cells in mitosis, aurora kinase inhibitors have very little impact on the duration of mitosis. While tubulin binders often produce lagging chromosomes and micronuclei below the concentrations that produce polyploidy, aurora kinase inhibitors induce micronuclei at similar concentration as polyploidy induction. In addition, while tubulin binders often show significant induction of clastogenicity at concentrations that produce a prolonged mitotic arrest, aurora inhibitors lack the ability to arrest cells in mitosis and therefore do not produce significant clastogenicity.

5.3.3. AOP: Chemical binding to aurora kinases leads to somatic cell aneuploidy MIE: Chemical binding to the catalytic domain of aurora kinase. The eukaryotic protein kinases make up a large superfamily of homologous proteins, which are related by the structure of their kinase catalytic domains (Fig. 3; Table 3). The kinase domains that define this group of enzymes contain 12 conserved subdomains that fold into a common catalytic core structure that is involved in ATP binding. In the central part of the catalytic domain there is a conserved aspartic acid residue, which is important for the catalytic activity of the enzyme. Since most kinase inhibitors target the conserved ATP binding pocket, it is likely that most kinase inhibitors hit multiple off-target kinases that could be responsible for the observed biological effects [178]. KE1: Inhibition of the catalytic activity of aurora kinase A/B. Chemical binding to the catalytic domain leads to the inhibition of the catalytic activity of aurora kinase A/B (KER1). The three aurora paralogues A, B, and C are structurally very similar, in particular within the carboxyterminal catalytic domain, in which human AURKA and AURKB share 71% identity [163]. Taken this similarity into account, it is likely that most compounds will bind to all three aurora kinases with similar binding affinities. The ability of a compound to bind and selectively inhibit aurora kinases can easily be determined using different kinase selectivity panels [178]. The results from these selectivity panels usually give a very good indication whether or not a compound will inhibit cellular aurora kinases with a high degree of confidence. KE2: Abnormal mitosis (1st mitotic division). It has been shown that selective inhibition of AURKA causes defective centrosome separation and maturation, resulting in monopolar spindles and G2-M arrest [164,165,175]. Inhibition of AURKB prevents chromosome segregation and leads to misaligned chromosomes in mitosis [173]. This is not surprising since aurora-B in conjunction with the proteins INCENP, Borealin (also known as Dasra B) and Survivin, form the chromosome passenger complex (CPC). This complex targets different locations at differing times during mitosis, where it regulates early mitotic events such as the correction of chromosome microtubule attachment errors and the activation of the SAC until microtubule-to-kinetochore attachments are corrected [166]. Its inhibition leads to decreased CDK1 activity and premature mitotic exit without the resolution of aberrant microtubule-to kinetochore attachment. Consequently, this will lead to the appearance of misaligned chromosomes during the metaphase and lagging chromosomes during the anaphase stage of mitosis [179]. Thus, the inhibition of the catalytic activity of aurora kinase A/B causes abnormal mitosis in 1st mitotic division (KER2) following exposure. In contrast to tubulin binders, AURKB inhibitors do not appear to have a major effect on the microtubules of the mitotic spindle. Unlike tubulin binders, aurora kinase inhibitors do not induce a mitotic arrest but increase the proportion of prophases when compared with metaphase and anaphase cells [173]. KE3: Binucleation. It has been well described that inhibition of AURKB kinase leads to the prevention of cytokinesis and the formation of binucleated cells. Normally during the metaphase–anaphase transition, AURKB together with the other proteins of the CPC will leave the centromeres and locate to the central spindle [166]. Subsequently, the CPC will localize to the equatorial cortex, which is the region where the cytokinetic machinery is assembled [180]. This process is mediated by a decrease in cdk1 activity and requires both phosphatase and AURKB kinase activity [181–183]. In order for a cell to undergo cytokinesis, it has to assemble and form an equatorial contractile ring composed of actin, myosin and other cytoskeletal filaments. Both the location and the timing for the formation of the contractile ring are controlled by the

6. Summary and discussion Recent advances in our understanding of mitosis have provided insights into how mitotic errors contribute to cytotoxicity, tetraploidy (polyploidy), aneuploidy, and various types of structural chromosomal damage in somatic cells (see Fig. 1 for some of the potential targets involved). The disruption of mitotic fidelity leading to genomic instability, and the observations from cancer biology linking aneuploidy with intra-tumour heterogeneity, tumour progression, and metastasis, underpin current notions defining aneuploidy as a hallmark of cancer. There has also been considerable progress on the characterization of the risks of aneuploidy induction from chemical exposure in the past 20 years; see Kirsch-Volders et al. [196] for a recent review of collaborative research in Europe on this topic. This is particularly true for compounds that inhibit tubulin or specific cell cycle associated kinases e.g. aurora kinases. Both targets continue to be of significant interest as 14

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

drug developers seek to exploit differences in the ways that normal and tumour cells divide and/or cope with mitotic stress and chromosome segregation errors. The IWGT aneuploidy workgroup conducted a review of the scientific literature on the biological mechanisms of aneuploidy in mammalian cells and on the methods used to detect chemical aneugens in a regulatory context, including the recent development of assays, such as MultiFlow® and ToxTracker. That can provide MoA insight in support of human risk assessment. In addition, the group reviewed the current regulatory framework for aneugens and presented several case studies on aneugen risk assessments conducted by the UK COM and the EMA, respectively. The workgroup arrived at a consensus statement on the current regulatory test batteries, that, while not ideal, they are sufficient for the identification of aneugens and human risk assessment. The workgroup was also excited by the new approaches that have been developed to define thresholds and points of departure for aneugens based on modelling dose response curves [67] and they agreed that these endeavours are likely to significantly contribute to human risk assessment of aneugens in the future. Based on current biology, it is obvious that there are many different MIEs that could lead to the induction of aneuploidy. Therefore, the IWGT aneuploidy workgroup sought to use current knowledge of the molecular and cellular mechanisms associated with aneuploidy to develop AOPs. The AOP framework provides a pragmatic means to organize knowledge along a pathway of causally connected key events linking a MIE with an AO. As such, AOPs can support mechanistic understanding and identify definitive assays that can be used to measure KE/KER. The results of these assays can be used to help predict the likelihood of the various adverse outcomes described. Indeed, the use of a MIE/KE based toxicity testing battery that substitutes for classical toxicology endpoints like cancer has been widely deployed in genetic toxicology. The tests used by genetic toxicologists focus on endpoints such as mutations, chromosome loss and chromosome structural damage since multiple MIEs converge on these key events. Furthermore, the endpoints are highly relevant for hereditary disease and cancer, adverse outcomes of significant importance in human safety risk assessment. As such, genetic toxicology mechanisms describe known biological relationships responsible for DNA mutation and chromosomal damage that are highly suitable for assessment using the AOP framework. As mentioned above, there is precedence for the AOP framework in genetic toxicology. OECD AOP 15“the induction of mutations in premeiotic male germ cells following exposure to alkylating agents” was among the first AOPs published by the OECD [85]. More recently, OECD AOP 116 was published and describes the events in female germ cells leading to aneuploidy in offspring [18]. This AOP outlines the biological relationships whereby chemicals that bind to tubulin cause microtubule depolymerization resulting in spindle disorganization followed by altered chromosome alignment, and segregation, and the generation of aneuploidy. Colchicine, a prototypical chemical that binds to tubulin and causes microtubule depolymerization, was used to illustrate the AOP. This provided a template to develop AOPs in somatic cells. Published data on the role of aneuploidy in human health [86] and from other collaborations [87] provided a rich source of cheminformatics on chemical aneugens and their effects in various genetic toxicology assays used to assess aneuploidy, and other adverse effects, such as polyploidy, chromosome damage and cytotoxicity. This, together with published information on the fundamental mechanisms of mammalian cell division, was used to develop AOPs with a specific focus on two molecular initiating events (MIEs), chemical binding to (a) tubulin and (b) the catalytic domain of aurora kinase B. These MIEs can lead to several adverse outcomes, including aneuploidy and polyploidy in somatic cells, and it was interesting to see how KEs and KERs, in individual AOPs for the MIE - chemical binding to tubulin, are linked and how these connections become more apparent when the AOPs are presented as a network. Indeed, the results (see Fig. 2) show how

networks of AOPs, sharing common KEs and KERs, can be amalgamated and how they could potentially serve as functional units of toxicity prediction. This supports the conclusion of Villeneuve et al. [73,74]. Moreover, some of these connections (discussed above) shed light on the underlying biology pertinent to other outcomes, such as cancer or cytotoxicity. The network also suggests that common KEs and KERs exist across AOPs with different MIEs, for example, with the formation of tetraploid cells (polyploidy) in AOP-2 and AOP-6. A potential shortcoming of the current work is that the level of biological organization used in the development of the current AOPs has been limited to the cell and does not extend to higher levels of biological organization i.e. tissue/organ for cancer or the whole organism for hereditary disease. However, it should be noted that the cell is generally the unit of measure of concern in genetic toxicology and all six AOPs describe pathways that are potentially of concern for human risk assessment. For example, AOP-1 “chemical binding to tubulin leading to somatic cell death” may provide insight for the clinical adverse events associated with chemotherapy using tubulin inhibitors such as cytopenias, allopecia, neurotoxicity, gastrointestinal disorders and azoospermia (see https://toxnet.nlm.nih.gov). Extending the AOPs to cover endpoints at higher levels of biological organization i.e. tissue, organ, individual, which may be more consistent with outcomes linked with cytotoxicity, was beyond the scope of the current paper. The role of aneuploidy in hereditary mutagenesis and in cancer is reviewed in the accompanying papers. In terms of cancer biology, the evidence did not support extending the AOPs on tubulin inhibitors to cancer induction and there are no carcinogenesis data on aurora kinase inhibitors in the public domain. OECD AOP 116 provides evidence for linking tubulin inhibitors with germ cell aneuploidy and the potential for hereditary disease. Of course, there may be interest in extending the AOPs presented in the current manuscript to more complex areas in the future, as knowledge emerges. Nevertheless, based on the use of the AOP framework as described above, the IWGT workgroup arrived at the following additional consensus statements. The AOP framework provides a useful means to link evidence for molecular initiating events with somatic cell aneuploidy and polyploidy. The AOP framework provides a basis to explore the evidence linking aneuploidy with adverse events such as carcinogenesis and hereditary disease. Mechanistic understanding together with knowledge of thresholdbased dose response relationships are fundamental to current human risk assessment approaches used for aneugens. They inform regulatory and MoA assays used in testing strategies for assessing the potential of new chemical entities to induce genotoxicity (including aneuploidy) and cytotoxicity. This ensures human health and safety can be safeguarded. The comprehensive review presented in this manuscript should help with the identification and risk management of aneugenic agents, both in general, and specifically for chemicals that bind tubulin and aurora kinases. However, it is also fair to conclude that the longterm consequences of chemical-aneugen exposure remains poorly defined, especially in vivo, as does our understanding of the underlying biology regarding the fate of aneuploid cells. This is especially true about the effects of aneuploidy in early and late-stage tumour development and much remains to be elucidated about the consequences of mitotic errors in somatic cells and their role in cancer to better inform human risk assessment. Acknowledgements We thank Kenji Sugimoto for his participation in the workgroup up to the meeting in Tokyo. We also want to thank Elizabeth Rubitski for providing the immunofluorescence images in Fig. 1. 15

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

References

[28] ICH, Guidance on genotoxicity testing and data interpretation for pharmaceuticals intended for human use – S2(R1), (2011). [29] OECD, Test No 487: In Vitro Mammalian Cell Micronucleus Test, (2016). [30] S. Avlasevich, S. Bryce, M. De Boeck, A. Elhajouji, F. Van Goethem, A. Lynch, J. Nicolette, J. Shi, S. Dertinger, Flow cytometric analysis of micronuclei in mammalian cell cultures: past, present and future, Mutagenesis 26 (2011) 147–152. [31] D.A. Eastmond, J.D. Tucker, Kinetochore localization in micronucleated cytokinesis-blocked Chinese hamster ovary cells: a new and rapid assay for identifying aneuploidy-inducing agents, Mutat. Res. 224 (1989) 517–525. [32] D.A. Eastmond, J.D. Tucker, Identification of aneuploidy-inducing agents using cytokinesis-blocked human lymphocytes and an antikinetochore antibody, Environ. Mol. Mutagen. 13 (1989) 34–43. [33] M. Kirsch-Volders, T. Sofuni, M. Aardema, S. Albertini, D. Eastmond, M. Fenech, M. Ishidate Jr., S. Kirchner, E. Lorge, T. Morita, H. Norppa, J. Surralles, A. Vanhauwaert, A. Wakata, Report from the in vitro micronucleus assay, Mutat. Res. 540 (2003) 153–163. [34] S.M. Bryce, S.L. Avlasevich, J.C. Bemis, S.D. Dertinger, Miniaturized flow cytometry-based CHO-K1 micronucleus assay discriminates aneugenic and clastogenic modes of action, Environ. Mol. Mutagen. 52 (2011) 280–286. [35] J. Nicolette, M. Diehl, P. Sonders, S. Bryce, E. Blomme, In vitro micronucleus screening of pharmaceutical candidates by flow cytometry in Chinese hamster V79 cells, Environ. Mol. Mutagen. 52 (2011) 355–362. [36] OECD, Test No 474: Mammalian Erythrocyte Micronucleus Test, (2014). [37] M.J. Aardema, S. Albertini, P. Arni, L.M. Henderson, M. Kirsch-Volders, J.M. Mackay, A.M. Sarrif, D.A. Stringer, R.D. Taalman, Aneuploidy: a report of an ECETOC task force, Mutat. Res. 410 (1998) 3–79. [38] S.M. Bryce, D.T. Bernacki, J.C. Bemis, S.D. Dertinger, Genotoxic mode of action predictions from a multiplexed flow cytometric assay and a machine learning approach, Environ. Mol. Mutagen. 57 (2016) 171–189. [39] S.M. Bryce, D.T. Bernacki, J.C. Bemis, R.A. Spellman, M.E. Engel, M. Schuler, E. Lorge, P.T. Heikkinen, U. Hemmann, V. Thybaud, S. Wilde, N. Queisser, A. Sutter, A. Zeller, M. Guerard, D. Kirkland, S.D. Dertinger, Interlaboratory evaluation of a multiplexed high information content in vitro genotoxicity assay, Environ. Mol. Mutagen. 58 (2017) 146–161. [40] S.M. Bryce, D.T. Bernacki, S.L. Smith-Roe, K.L. Witt, J.C. Bemis, S.D. Dertinger, Investigating the generalizability of the MultiFlow® DNA damage assay and several companion machine learning models with a set of 103 diverse test chemicals, Toxicol. Sci. 162 (2018) 146–166. [41] G. Hendriks, R.S. Derr, B. Misovic, B. Morolli, F.M. Calleja, H. Vrieling, The extended Toxtracker assay discriminates between induction of DNA damage, oxidative stress, and protein misfolding, Toxicol. Sci. 150 (2016) 190–203. [42] G. Hendriks, M. Atallah, M. Raamsman, B. Morolli, H. van der Putten, H. Jaadar, I. Tijdens, R. Esveldt-van Lange, L. Mullenders, B. van de Water, H. Vrieling, Sensitive DsRed fluorescence-based reporter cell systems for genotoxicity and oxidative stress assessment, Mutat. Res. 709–710 (2011) 49–59. [43] G. Hendriks, M. Atallah, B. Morolli, F. Calleja, N. Ras-Verloop, I. Huijskens, M. Raamsman, B. van de Water, H. Vrieling, The ToxTracker assay: novel GFP reporter systems that provide mechanistic insight into the genotoxic properties of chemicals, Toxicol. Sci. 125 (2012) 285–298. [44] M. Mirigian, K. Mukherjee, S.L. Bane, D.L. Sackett, Measurement of in vitro microtubule polymerization by turbidity and fluorescence, Methods Cell Biol. 115 (2013) 215–229. [45] L.A. Smyth, I. Collins, Measuring and interpreting the selectivity of protein kinase inhibitors, J. Chem. Biol. 2 (2009) 131–151. [46] C.J. Hastie, H.J. McLauchlan, P. Cohen, Assay of protein kinases using radiolabeled ATP: a protocol, Nat. Protoc. 1 (2006) 968–971. [47] C.S. Lebakken, K. Hee Chol, K.W. Vogel, A fluorescence lifetime based binding assay to characterize kinase inhibitors, J. Biomol. Screen. 12 (2007) 828–841. [48] Z.A. Knight, K.M. Shokat, Features of selective kinase inhibitors, Chem. Biol. 12 (2005) 621–637. [49] C.O. de Groot, J.E. Hsia, J.V. Anzola, A. Motamedi, M. Yoon, Y.L. Wong, D. Jenkins, H.J. Lee, M.B. Martinez, R.L. Davis, T.C. Gahman, A. Desai, A.K. Shiau, A cell biologist's field guide to aurora kinase inhibitors, Front Oncol. 5 (2015) 285. [50] J. Yang, T. Ikezoe, C. Nishioka, T. Tasaka, A. Taniguchi, Y. Kuwayama, N. Komatsu, K. Bandobashi, K. Togitani, H.P. Koeffler, H. Taguchi, A. Yokoyama, AZD1152, a novel and selective aurora B kinase inhibitor, induces growth arrest, apoptosis, and sensitization for tubulin depolymerizing agent or topoisomerase II inhibitor in human acute leukemia cells in vitro and in vivo, Blood 110 (2007) 2034–2040. [51] ECHA, Guidance on Information Requirements and Chemical Safety Assessment – Chapter R.7a: Endpoint specific guidance, (2017). [52] S.M. Galloway, Cytotoxicity and chromosome aberrations in vitro: experience in industry and the case for an upper limit on toxicity in the aberration assay, Environ. Mol. Mutagen. 35 (2000) 191–201. [53] I.D. Mitchell, T.R. Lambert, M. Burden, J. Sunderland, R.L. Porter, J.B. Carlton, Is polyploidy an important genotoxic lesion? Mutagenesis 10 (1995) 79–83. [54] L. Muller, Y. Kikuchi, G. Probst, L. Schechtman, H. Shimada, T. Sofuni, D. Tweats, ICH-harmonised guidances on genotoxicity testing of pharmaceuticals: evolution, reasoning and impact, Mutat. Res. 436 (1999) 195–225. [55] P. Muehlbauer, R. Spellman, Elucidating the significance of polyploidy induction in the human lymphocyte chromosomal aberration assay by flow cytometry, Mutat. Res. 577 (Suppl. 1) (2005) e179. [56] K.S. Bentley, D. Kirkland, M. Murphy, R. Marshall, Evaluation of thresholds for benomyl- and carbendazim-induced aneuploidy in cultured human lymphocytes using fluorescence in situ hybridization, Mutat. Res. 464 (2000) 41–51.

[1] S.L. Thompson, D.A. Compton, Proliferation of aneuploid human cells is limited by a p53-dependent mechanism, J. Cell Biol. 188 (2010) 369–381. [2] C. Templado, L. Uroz, A. Estop, New insights on the origin and relevance of aneuploidy in human spermatozoa, Mol. Hum. Reprod. 19 (2013) 634–643. [3] K.A. Knouse, J. Wu, C.A. Whittaker, A. Amon, Single cell sequencing reveals low levels of aneuploidy across mammalian tissues, Proc. Natl. Acad. Sci. USA 111 (2014) 13409–13414. [4] G. Farkas, Z. Juranyi, G. Szekely, Z.S. Kocsis, S. Gundy, Relationship between spontaneous frequency of aneuploidy and cancer risk in 2145 healthy Hungarian subjects, Mutagenesis 31 (2016) 583–588. [5] S.A. Langie, G. Koppen, D. Desaulniers, F. Al-Mulla, R. Al-Temaimi, A. Amedei, A. Azqueta, W.H. Bisson, D.G. Brown, G. Brunborg, A.K. Charles, T. Chen, A. Colacci, F. Darroudi, S. Forte, L. Gonzalez, R.A. Hamid, L.E. Knudsen, L. Leyns, A. Lopez de Cerain Salsamendi, L. Memeo, C. Mondello, C. Mothersill, A.K. Olsen, S. Pavanello, J. Raju, E. Rojas, R. Roy, E.P. Ryan, P. Ostrosky-Wegman, H.K. Salem, A.I. Scovassi, N. Singh, M. Vaccari, F.J. Van Schooten, M. Valverde, J. Woodrick, L. Zhang, N. van Larebeke, M. Kirsch-Volders, A.R. Collins, Causes of genome instability: the effect of low dose chemical exposures in modern society, Carcinogenesis 36 (Suppl. 1) (2015) S61–S88. [6] F. Pacchierotti, K. Masumura, D. Eastmond, A. Elhajouji, R. Froetschl, M. KirschVolders, A. Lynch, M. Schuler, D. Tweats, F. Marchetti, Chemically induced aneuploidy in germ cells. Part II of the report of the 2017 IWGT workgroup on assessing the risk of aneugens for carcinogenesis and hereditary disease, Mutat. Res. (2019) (this issue). [7] R. Sotillo, J.M. Schvartzman, N.D. Socci, R. Benezra, Mad2-induced chromosome instability leads to lung tumour relapse after oncogene withdrawal, Nature 464 (2010) 436–440. [8] D.J. Baker, F. Jin, K.B. Jeganathan, J.M. van Deursen, Whole chromosome instability caused by Bub1 insufficiency drives tumorigenesis through tumor suppressor gene loss of heterozygosity, Cancer Cell 16 (2009) 475–486. [9] B.A. Weaver, A.D. Silk, C. Montagna, P. Verdier-Pinard, D.W. Cleveland, Aneuploidy acts both oncogenically and as a tumor suppressor, Cancer Cell 11 (2007) 25–36. [10] D. Tweats, K. Masumura, F. Pacchierotti, A. Elhajouji, R. Froetschl, M. KirschVolders, A. Lynch, M. Schuler, F. Marchetti, D. Eastmond, Role of aneuploidy in the carcinogenic process. Part III of the report of the 2017 IWGT workgroup on assessing the risk of aneugens for carcinogenesis and hereditary disease, Mutat. Res. (2019) (this issue). [11] A.J. Holland, D.W. Cleveland, Losing balance: the origin and impact of aneuploidy in cancer, EMBO Rep. 13 (2012) 501–514. [12] A.J. Holland, D.W. Cleveland, Chromoanagenesis and cancer: mechanisms and consequences of localized, complex chromosomal rearrangements, Nat. Med. 18 (2012) 1630–1638. [13] M. Kirsch-Volders, G. Plas, A. Elhajouji, M. Lukamowicz, L. Gonzalez, K. Vande Loock, I. Decordier, The in vitro MN assay in 2011: origin and fate, biological significance, protocols, high throughput methodologies and toxicological relevance, Arch. Toxicol. 85 (2011) 873–899. [14] M. Terradas, M. Martin, A. Genesca, Impaired nuclear functions in micronuclei results in genome instability and chromothripsis, Arch. Toxicol. 90 (2016) 2657–2667. [15] T. Bovery, Concerning the Origin of Malignant Tumours, (1914). [16] B.R. Williams, A. Amon, Aneuploidy: cancer's fatal flaw? Cancer Res. 69 (2009) 5289–5291. [17] C. Dominguez-Brauer, K.L. Thu, J.M. Mason, H. Blaser, M.R. Bray, T.W. Mak, Targeting mitosis in cancer: emerging strategies, Mol. Cell 60 (2015) 524–536. [18] F. Marchetti, A. Massarotti, C.L. Yauk, F. Pacchierotti, A. Russo, The adverse outcome pathway (AOP) for chemical binding to tubulin in oocytes leading to aneuploid offspring, Environ. Mol. Mutagen. 57 (2016) 87–113. [19] M. Kirsch-Volders, M. Aardema, A. Elhajouji, Concepts of threshold in mutagenesis and carcinogenesis, Mutat. Res. 464 (2000) 3–11. [20] A. Elhajouji, M. Lukamowicz, Z. Cammerer, M. Kirsch-Volders, Potential thresholds for genotoxic effects by micronucleus scoring, Mutagenesis 26 (2011) 199–204. [21] L.C. Funk, L.M. Zasadil, B.A. Weaver, Living in CIN: mitotic infidelity and its consequences for tumor promotion and suppression, Dev. Cell 39 (2016) 638–652. [22] H.J. Evans, M.L. O’Riordan, Human peripheral blood lymphocytes for the analysis of chromosome aberrations in mutagen tests, Mutat. Res. 31 (1975) 135–148. [23] D. Scott, Metaphase chromosome aberration assays in vitro, in: D.J. Kirkland (Ed.), Basic Mutagenicity Tests: UKEMS Recommended Procedures, Cambridge University Press, Cambridge, UK, 1990, pp. 62–86. [24] OECD, Test No 473: In vitro Mammalian Chromosome Aberration Test, (2016). [25] T.J. Warr, E.M. Parry, J.M. Parry, A comparison of two in vitro mammalian cell cytogenetic assays for the detection of mitotic aneuploidy using 10 known or suspected aneugens, Mutat. Res. 287 (1993) 29–46. [26] J.M. Parry, E.M. Parry, R. Bourner, A. Doherty, S. Ellard, J. O’Donovan, B. Hoebee, J.M. de Stoppelaar, G.R. Mohn, A. Onfelt, A. Renglin, N. Schultz, C. SoderpalmBerndes, K.G. Jensen, M. Kirsch-Volders, A. Elhajouji, P. Van Hummelen, F. Degrassi, A. Antoccia, D. Cimini, M. Izzo, C. Tanzarella, I.D. Adler, U. Kliesch, P. Hess, et al., The detection and evaluation of aneugenic chemicals, Mutat. Res. 353 (1996) 11–46. [27] COM, Annual Report of the Committees on Toxicity, Mutagenicity, Carcinogenicity of Chemicals in Food, Consumer Products and the Environment, (1993).

16

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al. [57] T. Ramirez, D.A. Eastmond, L.A. Herrera, Non-disjunction events induced by albendazole in human cells, Mutat. Res. 626 (2007) 191–195. [58] M. Schuler, P. Muehlbauer, P. Guzzie, D.A. Eastmond, Noscapine hydrochlorideinduced numerical aberrations in cultured human lymphocytes: a comparison of FISH detection methods and multiple end-points, Mutagenesis 18 (2003) 235–242. [59] A. Zijno, F. Marcon, P. Leopardi, R. Crebelli, Analysis of chromosome segregation in cytokinesis-blocked human lymphocytes: non-disjunction is the prevalent damage resulting from low dose exposure to spindle poisons, Mutagenesis 11 (1996) 335–340. [60] P. Ly, D.W. Cleveland, Rebuilding chromosomes after catastrophe: emerging mechanisms of chromothripsis, Trends Cell Biol. 27 (2017) 917–930. [61] W. Nagl, Electron microscopic observations of a possible tubulin synthesis-polymerization complex in oocytes of a pond-skater, Cytobios 21 (1978) 165–170. [62] A. Elhajouji, F. Tibaldi, M. Kirsch-Volders, Indication for thresholds of chromosome non-disjunction versus chromosome lagging induced by spindle inhibitors in vitro in human lymphocytes, Mutagenesis 12 (1997) 133–140. [63] A. Elhajouji, P. Van Hummelen, M. Kirsch-Volders, Indications for a threshold of chemically-induced aneuploidy in vitro in human lymphocytes, Environ. Mol. Mutagen. 26 (1995) 292–304. [64] COM, Annual Report of the Committees on Toxicity, Mutagenicity, Carcinogenicity of Chemicals in Food, Consumer Products and the Environment, (1995). [65] COM, Annual Report of the Committees on Toxicity, Mutagenicity, Carcinogenicity of Chemicals in Food, Consumer Products and the Environment, (1997). [66] EMA, European Medicines Agency recommends restricting use of thiocolchicoside by mouth or injection, (2013). [67] P.A. White, G.E. Johnson, Genetic toxicology at the crossroads-from qualitative hazard evaluation to quantitative risk assessment, Mutagenesis 31 (2016) 233–237. [68] D.J. Tweats, G.E. Johnson, I. Scandale, J. Whitwell, D.B. Evans, Genotoxicity of flubendazole and its metabolites in vitro and the impact of a new formulation on in vivo aneugenicity, Mutagenesis 31 (2016) 309–321. [69] G.T. Ankley, R.S. Bennett, R.J. Erickson, D.J. Hoff, M.W. Hornung, R.D. Johnson, D.R. Mount, J.W. Nichols, C.L. Russom, P.K. Schmieder, J.A. Serrrano, J.E. Tietge, D.L. Villeneuve, Adverse outcome pathways: a conceptual framework to support ecotoxicology research and risk assessment, Environ. Toxicol. Chem. 29 (2010) 730–741. [70] M. Vinken, The adverse outcome pathway concept: a pragmatic tool in toxicology, Toxicology 312 (2013) 158–165. [71] S.J. Sturla, A.R. Boobis, R.E. FitzGerald, J. Hoeng, R.J. Kavlock, K. Schirmer, M. Whelan, M.F. Wilks, M.C. Peitsch, Systems toxicology: from basic research to risk assessment, Chem. Res. Toxicol. 27 (2014) 314–329. [72] N. Burden, F. Sewell, M.E. Andersen, A. Boobis, J.K. Chipman, M.T. Cronin, T.H. Hutchinson, I. Kimber, M. Whelan, Adverse Outcome Pathways can drive non-animal approaches for safety assessment, J. Appl. Toxicol. 35 (2015) 971–975. [73] D.L. Villeneuve, D. Crump, N. Garcia-Reyero, M. Hecker, T.H. Hutchinson, C.A. LaLone, B. Landesmann, T. Lettieri, S. Munn, M. Nepelska, M.A. Ottinger, L. Vergauwen, M. Whelan, Adverse outcome pathway development II: best practices, Toxicol. Sci. 142 (2014) 321–330. [74] D.L. Villeneuve, D. Crump, N. Garcia-Reyero, M. Hecker, T.H. Hutchinson, C.A. LaLone, B. Landesmann, T. Lettieri, S. Munn, M. Nepelska, M.A. Ottinger, L. Vergauwen, M. Whelan, Adverse outcome pathway (AOP) development I: strategies and principles, Toxicol. Sci. 142 (2014) 312–320. [75] M. Leist, A. Ghallab, R. Graepel, R. Marchan, R. Hassan, S.H. Bennekou, A. Limonciel, M. Vinken, S. Schildknecht, T. Waldmann, E. Danen, B. van Ravenzwaay, H. Kamp, I. Gardner, P. Godoy, F.Y. Bois, A. Braeuning, R. Reif, F. Oesch, D. Drasdo, S. Hohme, M. Schwarz, T. Hartung, T. Braunbeck, J. Beltman, H. Vrieling, F. Sanz, A. Forsby, D. Gadaleta, C. Fisher, J. Kelm, D. Fluri, G. Ecker, B. Zdrazil, A. Terron, P. Jennings, B. van der Burg, S. Dooley, A.H. Meijer, E. Willighagen, M. Martens, C. Evelo, E. Mombelli, O. Taboureau, A. Mantovani, B. Hardy, B. Koch, S. Escher, C. van Thriel, C. Cadenas, D. Kroese, B. van de Water, J.G. Hengstler, Adverse outcome pathways: opportunities, limitations and open questions, Arch. Toxicol. 91 (2017) 3477–3505. [76] M. Al Sharif, I. Tsakovska, I. Pajeva, P. Alov, E. Fioravanzo, A. Bassan, S. Kovarich, C. Yang, A. Mostrag-Szlichtyng, V. Vitcheva, A.P. Worth, A.N. Richarz, M.T.D. Cronin, The application of molecular modelling in the safety assessment of chemicals: a case study on ligand-dependent PPARgamma dysregulation, Toxicology 392 (2017) 140–154. [77] T.E. Allen, J.M. Goodman, S. Gutsell, P.J. Russell, Defining molecular initiating events in the adverse outcome pathway framework for risk assessment, Chem. Res. Toxicol. 27 (2014) 2100–2112. [78] J.C. Caldwell, M.V. Evans, K. Krishnan, Cutting edge PBPK models and analyses: providing the basis for future modeling efforts and bridges to emerging toxicology paradigms, J. Toxicol. 2012 (2012) 852384. [79] K.E. Tollefsen, S. Scholz, M.T. Cronin, S.W. Edwards, J. de Knecht, K. Crofton, N. Garcia-Reyero, T. Hartung, A. Worth, G. Patlewicz, Applying Adverse Outcome Pathways (AOPs) to support Integrated Approaches to Testing and Assessment (IATA), Regul. Toxicol. Pharmacol. 70 (2014) 629–640. [80] C. Wittwehr, H. Aladjov, G. Ankley, H.J. Byrne, J. de Knecht, E. Heinzle, G. Klambauer, B. Landesmann, M. Luijten, C. MacKay, G. Maxwell, M.E. Meek, A. Paini, E. Perkins, T. Sobanski, D. Villeneuve, K.M. Waters, M. Whelan, How adverse outcome pathways can aid the development and use of computational prediction models for regulatory toxicology, Toxicol. Sci. 155 (2017) 326–336. [81] F. Sewell, N. Gellatly, M. Beaumont, N. Burden, R. Currie, L. de Haan,

[82] [83]

[84] [85]

[86] [87]

[88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104]

[105] [106] [107] [108] [109] [110] [111] [112] [113] [114]

17

T.H. Hutchinson, M. Jacobs, C. Mahony, I. Malcomber, J. Mehta, G. Whale, I. Kimber, The future trajectory of adverse outcome pathways: a commentary, Arch. Toxicol. 92 (2018) 1657–1661. OECD, Users’ Handbook Supplement to the Guidance Document for Developing and Assessing AOPs, (2018). R.A. Becker, G.T. Ankley, S.W. Edwards, S.W. Kennedy, I. Linkov, B. Meek, M. Sachana, H. Segner, B. Van Der Burg, D.L. Villeneuve, H. Watanabe, T.S. Barton-Maclaren, Increasing scientific confidence in adverse outcome pathways: application of tailored Bradford-hill considerations for evaluating weight of evidence, Regul. Toxicol. Pharmacol. 72 (2015) 514–537. H.M. Bolt, Adverse outcome pathways, Arch. Toxicol. 91 (2017) 4023–4024. C.L. Yauk, I.B. Lambert, M.E. Meek, G.R. Douglas, F. Marchetti, Development of the adverse outcome pathway “alkylation of DNA in male premeiotic germ cells leading to heritable mutations” using the OECD's users’ handbook supplement, Environ. Mol. Mutagen. 56 (2015) 724–750. ECETOC, Aneuploidy, Monograph No. 27, (1997). M. Kirsch-Volders, T. Sofuni, M. Aardema, S. Albertini, D. Eastmond, M. Fenech, M. Ishidate, S. Kirchner, E. Lorge, T. Morita, H. Norppa, J. Surrallés, A. Vanhauwaert, A. Wakata, Report from the in vitro micronucleus assay, Mutat. Res./Genet. Toxicol. Environ. Mutagen. 540 (2003) 153–163. J.M. Schvartzman, R. Sotillo, R. Benezra, Mitotic chromosomal instability and cancer: mouse modelling of the human disease, Nat. Rev. Cancer 10 (2010) 102–115. E. Nogales, M. Whittaker, R.A. Milligan, K.H. Downing, High-resolution model of the microtubule, Cell 96 (1999) 79–88. S. Meunier, I. Vernos, Microtubule assembly during mitosis – from distinct origins to distinct functions? J. Cell Sci. 125 (2012) 2805–2814. Y.M. Liu, H.L. Chen, H.Y. Lee, J.P. Liou, Tubulin inhibitors: a patent review, Expert Opin. Ther. Pat. 24 (2014) 69–88. E.A. Perez, Microtubule inhibitors: differentiating tubulin-inhibiting agents based on mechanisms of action, clinical activity, and resistance, Mol. Cancer Ther. 8 (2009) 2086–2095. B. Bhattacharyya, D. Panda, S. Gupta, M. Banerjee, Anti-mitotic activity of colchicine and the structural basis for its interaction with tubulin, Med. Res. Rev. 28 (2008) 155–183. J. Chen, T. Liu, X. Dong, Y. Hu, Recent development and SAR analysis of colchicine binding site inhibitors, Mini Rev. Med. Chem. 9 (2009) 1174–1190. J.G. Chen, S.B. Horwitz, Differential mitotic responses to microtubule-stabilizing and -destabilizing drugs, Cancer Res. 62 (2002) 1935–1938. R.F. Luduena, A hypothesis on the origin and evolution of tubulin, Int. Rev. Cell Mol. Biol. 302 (2013) 41–185. L.M. Miller, H. Xiao, B. Burd, S.B. Horwitz, R.H. Angeletti, P. Verdier-Pinard, Methods in tubulin proteomics, Methods Cell Biol. 95 (2010) 105–126. Y. Lu, J. Chen, M. Xiao, W. Li, D.D. Miller, An overview of tubulin inhibitors that interact with the colchicine binding site, Pharm. Res. 29 (2012) 2943–2971. S. Lobert, B. Vulevic, J.J. Correia, Interaction of vinca alkaloids with tubulin: a comparison of vinblastine, vincristine, and vinorelbine, Biochemistry 35 (1996) 6806–6814. W.D. Singer, R.H. Himes, Cellular uptake and tubulin binding properties of four Vinca alkaloids, Biochem. Pharmacol. 43 (1992) 545–551. R.A. Stanton, K.M. Gernert, J.H. Nettles, R. Aneja, Drugs that target dynamic microtubules: a new molecular perspective, Med. Res. Rev. 31 (2011) 443–481. R.J. Toso, M.A. Jordan, K.W. Farrell, B. Matsumoto, L. Wilson, Kinetic stabilization of microtubule dynamic instability in vitro by vinblastine, Biochemistry 32 (1993) 1285–1293. R.K. Warfield, G.B. Bouck, Microtubule-macrotubule transitions: intermediates after exposure to the mitotic inhibitor vinblastine, Science 186 (1974) 1219–1221. S. Albertini, U. Friederich, C. Holderegger, F.E. Wurgler, The in vitro porcine brain tubulin assembly assay: effects of a genotoxic carcinogen (aflatoxin B1), eight tumor promoters and nine miscellaneous substances, Mutat. Res. 201 (1988) 283–292. M. Brunner, S. Albertini, F.E. Wurgler, Effects of 10 known or suspected spindle poisons in the in vitro porcine brain tubulin assembly assay, Mutagenesis 6 (1991) 65–70. S.B. Hastie, Interactions of colchicine with tubulin, Pharmacol. Ther. 51 (1991) 377–401. G.E. Johnson, E.M. Parry, Mechanistic investigations of low dose exposures to the genotoxic compounds bisphenol – a and rotenone, Mutat. Res. 651 (2008) 56–63. E.M. Parry, N. Danford, J.M. Parry, Differential staining of chromosomes and spindle and its use as an assay for determining the effect of diethylstilboestrol on cultured mammalian cells, Mutat. Res. 105 (1982) 243–252. M. Kirsch-Volders, Differential staining of chromosomes and spindle cannot be used as an assay to determine the effect of cancer promoters on primary cultures of human fibroblasts, Mutat. Res. 171 (1986) 177–183. M. Nijs, M. Kirsch-Volders, Induction of spindle inhibition and abnormal mitotic figures by Cr(II), Cr(III) and Cr(VI) ions, Mutagenesis 1 (1986) 247–252. E.M. Parry, D.C. Sharp, J.M. Parry, The observation of mitotic division aberrations in mammalian cells exposed to chemical and radiation treatments, Mutat. Res. 150 (1985) 369–381. E.M. Parry, J.M. Parry, C. Corso, A. Doherty, F. Haddad, T.F. Hermine, G. Johnson, M. Kayani, E. Quick, T. Warr, J. Williamson, Detection and characterization of mechanisms of action of aneugenic chemicals, Mutagenesis 17 (2002) 509–521. M. Barisic, A.J. Pereira, H. Maiato, Fluorescent speckle microscopy in cultured cells, Methods Enzymol. 504 (2012) 147–161. I. Brust-Mascher, G. Civelekoglu-Scholey, J.M. Scholey, Analysis of mitotic protein dynamics and function in Drosophila embryos by live cell imaging and

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al. quantitative modeling, Methods Mol. Biol. 1136 (2014) 3–30. [115] A. Russo, F. Pacchierotti, D. Cimini, N.J. Ganem, A. Genesca, A.T. Natarajan, S. Pavanello, G. Valle, F. Degrassi, Genomic instability: crossing pathways at the origin of structural and numerical chromosome changes, Environ. Mol. Mutagen. 56 (2015) 563–580. [116] M.M. McGee, Targeting the mitotic catastrophe signaling pathway in cancer, Mediat. Inflamm. 2015 (2015) 146282. [117] I. Vitale, L. Galluzzi, M. Castedo, G. Kroemer, Mitotic catastrophe: a mechanism for avoiding genomic instability, Nat. Rev. Mol. Cell Biol. 12 (2011) 385–392. [118] M. Castedo, J.L. Perfettini, T. Roumier, K. Andreau, R. Medema, G. Kroemer, Cell death by mitotic catastrophe: a molecular definition, Oncogene 23 (2004) 2825–2837. [119] C.H. Topham, S.S. Taylor, Mitosis and apoptosis: how is the balance set? Curr. Opin. Cell Biol. 25 (2013) 780–785. [120] Y. Nakayama, T. Inoue, Antiproliferative fate of the tetraploid formed after mitotic slippage and its promotion: a novel target for cancer therapy based on microtubule poisons, Molecules 21 (2016). [121] A. Janssen, G.J. Kops, R.H. Medema, Elevating the frequency of chromosome missegregation as a strategy to kill tumor cells, Proc. Natl. Acad. Sci. USA 106 (2009) 19108–19113. [122] M. Castedo, J.L. Perfettini, T. Roumier, A. Valent, H. Raslova, K. Yakushijin, D. Horne, J. Feunteun, G. Lenoir, R. Medema, W. Vainchenker, G. Kroemer, Mitotic catastrophe constitutes a special case of apoptosis whose suppression entails aneuploidy, Oncogene 23 (2004) 4362–4370. [123] I. Decordier, E. Cundari, M. Kirsch-Volders, Influence of caspase activity on micronuclei detection: a possible role for caspase-3 in micronucleation, Mutagenesis 20 (2005) 173–179. [124] I. Decordier, L. Dillen, E. Cundari, M. Kirsch-Volders, Elimination of micronucleated cells by apoptosis after treatment with inhibitors of microtubules, Mutagenesis 17 (2002) 337–344. [125] A. Elhajouji, M. Cunha, M. Kirsch-Volders, Spindle poisons can induce polyploidy by mitotic slippage and micronucleate mononucleates in the cytokinesis-block assay, Mutagenesis 13 (1998) 193–198. [126] B. Cheng, K. Crasta, Consequences of mitotic slippage for antimicrotubule drug therapy, Endocr. Relat. Cancer 24 (2017) T97–T106. [127] S. Minissi, B. Gustavino, F. Degrassi, C. Tanzarella, M. Rizzoni, Effect of cytochalasin B on the induction of chromosome missegregation by colchicine at low concentrations in human lymphocytes, Mutagenesis 14 (1999) 43–49. [128] I. Decordier, E. Cundari, M. Kirsch-Volders, Survival of aneuploid, micronucleated and/or polyploid cells: crosstalk between ploidy control and apoptosis, Mutat. Res. 651 (2008) 30–39. [129] C.L. Rieder, R.E. Palazzo, Colcemid and the mitotic cycle, J. Cell Sci. 102 (Pt 3) (1992) 387–392. [130] A.J. Olaharski, R. Sotelo, G. Solorza-Luna, M.E. Gonsebatt, P. Guzman, A. Mohar, D.A. Eastmond, Tetraploidy and chromosomal instability are early events during cervical carcinogenesis, Carcinogenesis 27 (2006) 337–343. [131] J. Campisi, F. d’Adda di Fagagna, Cellular senescence: when bad things happen to good cells, Nat. Rev. Mol. Cell Biol. 8 (2007) 729–740. [132] T. Davoli, T. de Lange, Telomere-driven tetraploidization occurs in human cells undergoing crisis and promotes transformation of mouse cells, Cancer Cell 21 (2012) 765–776. [133] P. Lara-Gonzalez, F.G. Westhorpe, S.S. Taylor, The spindle assembly checkpoint, Curr. Biol. 22 (2012) R966–R980. [134] A. Musacchio, A. Ciliberto, The spindle-assembly checkpoint and the beauty of self-destruction, Nat. Struct. Mol. Biol. 19 (2012) 1059–1061. [135] S.L. Thompson, D.A. Compton, Examining the link between chromosomal instability and aneuploidy in human cells, J. Cell Biol. 180 (2008) 665–672. [136] S.L. Thompson, D.A. Compton, Chromosome missegregation in human cells arises through specific types of kinetochore-microtubule attachment errors, Proc. Natl. Acad. Sci. USA 108 (2011) 17974–17978. [137] H. Chen, D.S. Rupa, R. Tomar, D.A. Eastmond, Chromosomal loss and breakage in mouse bone marrow and spleen cells exposed to benzene in vivo, Cancer Res. 54 (1994) 3533–3539. [138] B.M. Miller, H.F. Zitzelsberger, H.U. Weier, I.D. Adler, Classification of micronuclei in murine erythrocytes: immunofluorescent staining using CREST antibodies compared to in situ hybridization with biotinylated gamma satellite DNA, Mutagenesis 6 (1991) 297–302. [139] J. Maki-Paakkanen, M. Hayashi, T. Suzuki, H. Tanabe, M. Honma, T. Sofuni, Analysis by fluorescence in situ hybridization with a mouse gamma satellite DNA probe of isolated micronuclei induced in mice by two clastogens and two spindle poisons, Mutagenesis 10 (1995) 513–516. [140] A. Takeiri, S. Motoyama, K. Matsuzaki, A. Harada, J. Taketo, C. Katoh, K. Tanaka, M. Mishima, New DNA probes to detect aneugenicity in rat bone marrow micronucleated cells by a pan-centromeric FISH analysis, Mutat. Res. 755 (2013) 73–80. [141] D. Hongping, L. Jianlin, Z. Meibian, W. Wei, J. Lifen, C. Shijie, Z. Wei, W. Baohong, H. Jiliang, Detecting the cytogenetic effects in workers occupationally exposed to vincristine with four genetic tests, Mutat. Res. 599 (2006) 152–159. [142] S. Matsuura, Y. Matsumoto, K. Morishima, H. Izumi, H. Matsumoto, E. Ito, K. Tsutsui, J. Kobayashi, H. Tauchi, Y. Kajiwara, S. Hama, K. Kurisu, H. Tahara, M. Oshimura, K. Komatsu, T. Ikeuchi, T. Kajii, Monoallelic BUB1B mutations and defective mitotic-spindle checkpoint in seven families with premature chromatid separation (PCS) syndrome, Am. J. Med. Genet. A 140 (2006) 358–367. [143] K. Snape, S. Hanks, E. Ruark, P. Barros-Nunez, A. Elliott, A. Murray, A.H. Lane, N. Shannon, P. Callier, D. Chitayat, J. Clayton-Smith, D.R. Fitzpatrick, D. Gisselsson, S. Jacquemont, K. Asakura-Hay, M.A. Micale, J. Tolmie,

[144] [145] [146] [147]

[148] [149] [150] [151] [152] [153] [154]

[155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168]

[169] [170] [171] [172] [173]

[174]

18

P.D. Turnpenny, M. Wright, J. Douglas, N. Rahman, Mutations in CEP57 cause mosaic variegated aneuploidy syndrome, Nat. Genet. 43 (2011) 527–529. A. Janssen, M. van der Burg, K. Szuhai, G.J. Kops, R.H. Medema, Chromosome segregation errors as a cause of DNA damage and structural chromosome aberrations, Science 333 (2011) 1895–1898. M.S. Levine, A.J. Holland, The impact of mitotic errors on cell proliferation and tumorigenesis, Genes Dev. 32 (2018) 620–638. Y. Zhu, Y. Zhou, J. Shi, Post-slippage multinucleation renders cytotoxic variation in anti-mitotic drugs that target the microtubules or mitotic spindle, Cell Cycle 13 (2014) 1756–1764. G. Schriever-Schwemmer, U. Kliesch, I.D. Adler, Extruded micronuclei induced by colchicine or acrylamide contain mostly lagging chromosomes identified in paintbrush smears by minor and major mouse DNA probes, Mutagenesis 12 (1997) 201–207. P. Arni, T. Hertner, Chromosomal aberrations in vitro induced by aneugens, Mutat. Res. 379 (1997) 83–93. W. Xu, I.D. Adler, Clastogenic effects of known and suspect spindle poisons studied by chromosome analysis in mouse bone marrow cells, Mutagenesis 5 (1990) 371–374. S. Celikler, R. Bilaloglu, N. Aydemir, Genotoxic effects induced by fotemustine and vinorelbine in human lymphocytes, Z. Naturforsch. C. 61 (2006) 903–910. E.M. Hatch, M.W. Hetzer, Linking micronuclei to chromosome fragmentation, Cell 161 (2015) 1502–1504. E.M. Hatch, A.H. Fischer, T.J. Deerinck, M.W. Hetzer, Catastrophic nuclear envelope collapse in cancer cell micronuclei, Cell 154 (2013) 47–60. C.Z. Zhang, A. Spektor, H. Cornils, J.M. Francis, E.K. Jackson, S. Liu, M. Meyerson, D. Pellman, Chromothripsis from DNA damage in micronuclei, Nature 522 (2015) 179–184. P.J. Stephens, C.D. Greenman, B. Fu, F. Yang, G.R. Bignell, L.J. Mudie, E.D. Pleasance, K.W. Lau, D. Beare, L.A. Stebbings, S. McLaren, M.L. Lin, D.J. McBride, I. Varela, S. Nik-Zainal, C. Leroy, M. Jia, A. Menzies, A.P. Butler, J.W. Teague, M.A. Quail, J. Burton, H. Swerdlow, N.P. Carter, L.A. Morsberger, C. Iacobuzio-Donahue, G.A. Follows, A.R. Green, A.M. Flanagan, M.R. Stratton, P.A. Futreal, P.J. Campbell, Massive genomic rearrangement acquired in a single catastrophic event during cancer development, Cell 144 (2011) 27–40. J.V. Forment, A. Kaidi, S.P. Jackson, Chromothripsis and cancer: causes and consequences of chromosome shattering, Nat. Rev. Cancer 12 (2012) 663–670. G.J. Kops, B.A. Weaver, D.W. Cleveland, On the road to cancer: aneuploidy and the mitotic checkpoint, Nat. Rev. Cancer 5 (2005) 773–785. M. Malumbres, M. Barbacid, Cell cycle kinases in cancer, Curr. Opin. Genet. Dev. 17 (2007) 60–65. A. Basant, S. Lekomtsev, Y.C. Tse, D. Zhang, K.M. Longhini, M. Petronczki, M. Glotzer, Aurora B kinase promotes cytokinesis by inducing centralspindlin oligomers that associate with the plasma membrane, Dev. Cell 33 (2015) 204–215. M. Vieth, J.J. Sutherland, D.H. Robertson, R.M. Campbell, Kinomics: characterizing the therapeutically validated kinase space, Drug Discov. Today 10 (2005) 839–846. T. Anastassiadis, S.W. Deacon, K. Devarajan, H. Ma, J.R. Peterson, Comprehensive assay of kinase catalytic activity reveals features of kinase inhibitor selectivity, Nat. Biotechnol. 29 (2011) 1039–1045. B. Goldenson, J.D. Crispino, The aurora kinases in cell cycle and leukemia, Oncogene 34 (2015) 537–545. V. Bavetsias, S. Linardopoulos, Aurora kinase inhibitors: current status and outlook, Front. Oncol. 5 (2015) 278. M. Carmena, W.C. Earnshaw, The cellular geography of aurora kinases, Nat. Rev. Mol. Cell Biol. 4 (2003) 842–854. P. Carpinelli, J. Moll, Aurora kinase inhibitors: identification and preclinical validation of their biomarkers, Expert Opin. Ther. Targets 12 (2008) 69–80. F. Girdler, K.E. Gascoigne, P.A. Eyers, S. Hartmuth, C. Crafter, K.M. Foote, N.J. Keen, S.S. Taylor, Validating Aurora B as an anti-cancer drug target, J. Cell Sci. 119 (2006) 3664–3675. M. Carmena, M. Wheelock, H. Funabiki, W.C. Earnshaw, The chromosomal passenger complex (CPC): from easy rider to the godfather of mitosis, Nat. Rev. Mol. Cell Biol. 13 (2012) 789–803. S.C. Sampath, R. Ohi, O. Leismann, A. Salic, A. Pozniakovski, H. Funabiki, The chromosomal passenger complex is required for chromatin-induced microtubule stabilization and spindle assembly, Cell 118 (2004) 187–202. M.E. Tanenbaum, L. Macurek, B. van der Vaart, M. Galli, A. Akhmanova, R.H. Medema, A complex of Kif18b and MCAK promotes microtubule depolymerization and is negatively regulated by Aurora kinases, Curr. Biol. 21 (2011) 1356–1365. Y. Terada, M. Tatsuka, F. Suzuki, Y. Yasuda, S. Fujita, M. Otsu, AIM-1: a mammalian midbody-associated protein required for cytokinesis, EMBO J. 17 (1998) 667–676. G. Vader, R.H. Medema, S.M. Lens, The chromosomal passenger complex: guiding Aurora-B through mitosis, J. Cell Biol. 173 (2006) 833–837. K.T. Yang, C.J. Tang, T.K. Tang, Possible role of Aurora-C in meiosis, Front. Oncol. 5 (2015) 178. J. Khan, F. Ezan, J.Y. Cremet, A. Fautrel, D. Gilot, M. Lambert, C. Benaud, M.B. Troadec, C. Prigent, Overexpression of active Aurora-C kinase results in cell transformation and tumour formation, PLoS One 6 (2011) e26512. C. Ditchfield, V.L. Johnson, A. Tighe, R. Ellston, C. Haworth, T. Johnson, A. Mortlock, N. Keen, S.S. Taylor, Aurora B couples chromosome alignment with anaphase by targeting BubR1, Mad2, and Cenp-E to kinetochores, J. Cell Biol. 161 (2003) 267–280. S. Hauf, R.W. Cole, S. LaTerra, C. Zimmer, G. Schnapp, R. Walter, A. Heckel, J. van

Mutat Res Gen Tox En 847 (2019) 403025

A.M. Lynch, et al.

[175] [176]

[177] [178] [179]

[180]

[181] [182] [183]

[184] [185] [186] [187] [188]

Meel, C.L. Rieder, J.M. Peters, The small molecule Hesperadin reveals a role for Aurora B in correcting kinetochore-microtubule attachment and in maintaining the spindle assembly checkpoint, J. Cell Biol. 161 (2003) 281–294. A.R. Barr, F. Gergely, Aurora-A: the maker and breaker of spindle poles, J. Cell Sci. 120 (2007) 2987–2996. M.G. Manfredi, J.A. Ecsedy, K.A. Meetze, S.K. Balani, O. Burenkova, W. Chen, K.M. Galvin, K.M. Hoar, J.J. Huck, P.J. LeRoy, E.T. Ray, T.B. Sells, B. Stringer, S.G. Stroud, T.J. Vos, G.S. Weatherhead, D.R. Wysong, M. Zhang, J.B. Bolen, C.F. Claiborne, Antitumor activity of MLN8054, an orally active small-molecule inhibitor of Aurora A kinase, Proc. Natl. Acad. Sci. USA 104 (2007) 4106–4111. T. Marumoto, S. Honda, T. Hara, M. Nitta, T. Hirota, E. Kohmura, H. Saya, AuroraA kinase maintains the fidelity of early and late mitotic events in HeLa cells, J. Biol. Chem. 278 (2003) 51786–51795. J.C. Uitdehaag, F. Verkaar, H. Alwan, J. de Man, R.C. Buijsman, G.J. Zaman, A guide to picking the most selective kinase inhibitor tool compounds for pharmacological validation of drug targets, Br. J. Pharmacol. 166 (2012) 858–876. W. Lan, X. Zhang, S.L. Kline-Smith, S.E. Rosasco, G.A. Barrett-Wilt, J. Shabanowitz, D.F. Hunt, C.E. Walczak, P.T. Stukenberg, Aurora B phosphorylates centromeric MCAK and regulates its localization and microtubule depolymerization activity, Curr. Biol. 14 (2004) 273–286. W.C. Earnshaw, C.A. Cooke, Analysis of the distribution of the INCENPs throughout mitosis reveals the existence of a pathway of structural changes in the chromosomes during metaphase and early events in cleavage furrow formation, J. Cell Sci. 98 (Pt 4) (1991) 443–461. S. Hummer, T.U. Mayer, Cdk1 negatively regulates midzone localization of the mitotic kinesin Mklp2 and the chromosomal passenger complex, Curr. Biol. 19 (2009) 607–612. G. Pereira, E. Schiebel, Separase regulates INCENP-Aurora B anaphase spindle function through Cdc14, Science 302 (2003) 2120–2124. Z. Xu, H. Ogawa, P. Vagnarelli, J.H. Bergmann, D.F. Hudson, S. Ruchaud, T. Fukagawa, W.C. Earnshaw, K. Samejima, INCENP-aurora B interactions modulate kinase activity and chromosome passenger complex localization, J. Cell Biol. 187 (2009) 637–653. M. Carmena, S. Ruchaud, W.C. Earnshaw, Making the Auroras glow: regulation of Aurora A and B kinase function by interacting proteins, Curr. Opin. Cell Biol. 21 (2009) 796–805. S. Ruchaud, M. Carmena, W.C. Earnshaw, Chromosomal passengers: conducting cell division, Nat. Rev. Mol. Cell Biol. 8 (2007) 798–812. M.S. van der Waal, R.C. Hengeveld, A. van der Horst, S.M. Lens, Cell division control by the Chromosomal Passenger Complex, Exp. Cell Res. 318 (2012) 1407–1420. Q. Wu, W. Zhang, T. Mu, T. Song, D. Li, Aurora B kinase is required for cytokinesis through effecting spindle structure, Cell Biol. Int. 37 (2013) 436–442. Y. Uetake, G. Sluder, Cell cycle progression after cleavage failure: mammalian somatic cells do not possess a “tetraploidy checkpoint”, J. Cell Biol. 165 (2004)

609–615. [189] C. Wong, T. Stearns, Mammalian cells lack checkpoints for tetraploidy, aberrant centrosome number, and cytokinesis failure, BMC Cell Biol. 6 (2005) 6. [190] N.J. Ganem, Z. Storchova, D. Pellman, Tetraploidy, aneuploidy and cancer, Curr. Opin. Genet. Dev. 17 (2007) 157–162. [191] N.J. Ganem, S.A. Godinho, D. Pellman, A mechanism linking extra centrosomes to chromosomal instability, Nature 460 (2009) 278–282. [192] W.T. Silkworth, I.K. Nardi, L.M. Scholl, D. Cimini, Multipolar spindle pole coalescence is a major source of kinetochore mis-attachment and chromosome missegregation in cancer cells, PLoS One 4 (2009) e6564. [193] M.T. Hayashi, J. Karlseder, DNA damage associated with mitosis and cytokinesis failure, Oncogene 32 (2013) 4593–4601. [194] P. Gollapudi, L.S. Hasegawa, D.A. Eastmond, A comparative study of the aneugenic and polyploidy-inducing effects of fisetin and two model Aurora kinase inhibitors, Mutat. Res. Genet. Toxicol. Environ. Mutagen. 767 (2014) 37–43. [195] K. Hoar, A. Chakravarty, C. Rabino, D. Wysong, D. Bowman, N. Roy, J.A. Ecsedy, MLN8054, a small-molecule inhibitor of Aurora A, causes spindle pole and chromosome congression defects leading to aneuploidy, Mol. Cell Biol. 27 (2007) 4513–4525. [196] M. Kirsch-Volders, F. Pacchierotti, E.M. Parry, A. Russo, U. Eichenlaub-Ritter, I.D. Adler, Risks of aneuploidy induction from chemical exposure: twenty years of collaborative research in Europe from basic science to regulatory implications, Mutat. Res. Rev. Mutat. Res. (2019) (in press). [197] Z. Cammerer, M.M. Schumacher, M. Kirsch-Volders, W. Suter, A. Elhajouji, Flow cytometry peripheral blood micronucleus test in vivo: determination of potential thresholds for aneuploidy induced by spindle poisons, Environ. Mol. Mutagen. 51 (2010) 278–284. [198] A.M. Sarrif, K.S. Bentley, L.J. Fu, R.M. O’Neil, V.L. Reynolds, R.G. Stahl, Evaluation of benomyl and carbendazim in the in vivo aneuploidy/micronucleus assay in BDF1 mouse bone marrow, Mutat. Res. 310 (1994) 143–149. [199] J.P. Seiler, The mutagenicity of benzimidazole and benzimidazole derivatives. VI. Cytogenetic effects of benzimidazole derivatives in the bone marrow of the mouse and the Chinese hamster, Mutat. Res. 40 (1976) 339–347. [200] G. Steiblen, T. Orsiere, C. Pallen, A. Botta, D. Marzin, Comparison of the relative sensitivity of human lymphocytes and mouse splenocytes to two spindle poisons, Mutat. Res. 588 (2005) 143–151. [201] D. Bonacker, T. Stoiber, K.J. Bohm, E. Unger, G.H. Degen, R. Thier, H.M. Bolt, Chromosomal genotoxicity of nitrobenzene and benzonitrile, Arch. Toxicol. 78 (2004) 49–57. [202] H. Tinwell, J. Ashby, Micronucleus morphology as a means to distinguish aneugens and clastogens in the mouse bone marrow micronucleus assay, Mutagenesis 6 (1991) 193–198. [203] C. Ives, I. Campia, R.L. Wang, C. Wittwehr, S. Edwards, Creating a structured AOP knowledgebase via ontology-based annotations, Appl. In Vitro Toxicol. 3 (2017) 298–311.

19