existing genotoxicity assays (2)

8/12/2019 Existing Genotoxicity Assays (2)

1/16

21

CHAPTER 2: EXISTING GENOTOXICITY ASSAYS1

2.1 INTRODUCTION2

The utility of transgenic rodent mutation assays in genotoxic risk assessment may be considered3in the context of currently available assays. Typically, genotoxic effects can be manifested4through two mechanisms: induction of gene mutations and/or chromosomal aberrations. For the5purposes of this chapter, gene mutations are considered to be permanent single mutations6involving only one gene, while chromosomal aberrations are considered to be changes in the7structure of the chromosome (structural aberration), or a change in chromosome number8(numerical aberration). Generally, chromosomal aberrations affect more than one gene. A9mutagenic effect can be manifested through one or both of these mechanisms.10

Since experience with genetic toxicology testing over the past several decades has11demonstrated that no single assay is capable of detecting all genotoxic effects, the potential for a12

chemical to cause genotoxicity is typically determined through a battery of tests, both short and13long term, and both in vitroand in vivo. In vitroassays offer advantages in that they are relatively14inexpensive and easy to conduct, and do not directly involve the use of animals, but they usually15require supplementation with exogenous metabolic activation enzymes in order to simulate16mammalian metabolism, and generally do not allow for the consideration of factors including17pharmacokinetics and DNA-repair that are relevant in the in vivosystem. Nevertheless, in vitro18assays are typically used to provide an initial indication of the genotoxicity of a chemical, and the19results often serve as a guide to subsequent in vivostudies.20

In this chapter, existing genetic toxicology tests are reviewed and their benefits and21limitations are described (see also Table 2-1). For the purposes of this discussion, assays are22grouped into four categories: in vitromutation assays (briefly reviewed), in vivosomatic cell gene23

mutation assays, in vivosomatic cell chromosomal aberration assays and in vivoindicator assays.24

2.2 IN VITRO GENOTOXICITY ASSAYS25

2.2.1. IN VITRO CHROMOSOMAL ABERRATION ASSAYS. Chemicals causing chromosomal26aberrations may be identified with an in vitro cytogenetics assay [88]. Cell cultures are treated27with the test chemical and then mitosis is arrested in metaphase with an inhibitor, such as28colchicine. The metaphase spreads are examined by light microscopy to detect chromosome or29chromatid aberrations, or polyploid cells. A biologically significant increase in the frequency of30cells with structural or numerical aberrations compared with that of the concurrent control group31indicates the chemical is clastogenic or aneugenic, respectively. A major drawback of this assay,32in comparison with some of the other in vitroassays, is the subjectivity and cost of having the33

metaphase spreads scored by a highly trained observer.34

More recently, a harmonized protocol for the in vitro micronucleus test has been35developed by an IWGT working group [89,90]. A proposal for an OECD test guideline is under36development and the European Centre for the Validation of Alternative Methods is undertaking a37retrospective validation exercise. However, the transition of this assay from the development38stage to routine use for regulatory applications has not yet occurred.39

2.2.2 IN VITRO GENE MUTATION ASSAYS. The in vitro assays commonly used in genetic40toxicity testing include the Ames bacterial reverse mutation assay, which detects chemicals that41cause point mutations or frameshift mutations in histidine auxotrophic strains of Salmonella42typhimurium(e.g. strains TA100, TA98, TA102), and a reverse mutation assay using a tryptophan43


2/16

22

auxotrophic strain of Escherichia coli (e.g., WP2uvrA) [91]. Revertant cells grow on minimal1agar containing trace amounts of histidine or tryptophan, while wildtype cells rapidly deplete the2limiting amino acid and stop growing. If there is a toxicologically significant increase in the3

number of revertant colonies compared with the results of the concurrent negative control4(typically considered a two-fold or greater increase), the chemical is concluded to be mutagenic.5

Mammalian forward mutation assays, such as the thymidine kinase (Tk) assay or the6hypoxanthine-guanine phosphoribosyl transferase (Hprt) assay, detect mutations at the7heterozygous Tk or hemizygous Hprtgene [92]. Cells such as L5178Y mouse lymphoma cells8(Tklocus), several Chinese hamster cell lines (Hprtlocus), and human lymphoblastoid cells (Tk9locus) are most commonly used. Mutations are selected by incubation of the cell cultures with10the selective agents trifluorothymidine (Tk assay) or 6-thioguanine (Hprt assay). Cells having11forward mutations at the TKor Hprtgenes survive in the presence of the selective agent, while12wild-type cells accumulate a toxic metabolite and do not proliferate. Comparison of the mutant13frequency of the treatment groups with the concurrent negative control group allows the14

identification of a mutagenic chemical.15

2.3 IN VIVO ASSAYS FOR SOMATIC CELL CHROMOSOMAL ABERRATIONS16

In vivo chromosomal aberration assays assess the potential of a test chemical to cause DNA17damage that may affect chromosome structure, or interfere with the mitotic apparatus causing18changes in chromosome number. The in vivo assaystake into account DNA repair processes and19the effects of pharmacokinetic factors that do not play a role in in vitro systems. There are20several short-term assays that detect somatic cell chromosomal aberrations; these include the21rodent erythrocyte micronucleus assay and the chromosomal aberration assay in bone marrow.22These methods are described in the following sections.23

2.3.1 RODENT ERYTHROCYTE MICRONUCLEUS ASSAY. Because of its relative simplicity and24 its sensitivity to clastogens, the rodent erythrocyte micronucleus assay has now become the most25commonly conducted in vivo assay. It has achieved widespread acceptance and recommended26test methods have been published as OECD Test Guideline 474 [93].27

2.3.1.1 Principles. The micronucleus assay detects chromosome damage and whole28chromosome loss in polychromatic erythrocytes, and eventually in normochromatic erythrocytes29in peripheral blood as the red cells mature. A micronucleus is a small structure (1/5 to 1/20 the30size of the nucleus) containing nuclear DNA that has arisen from chromosome fragments or31whole chromosomes that were not incorporated into daughter nuclei at anaphase of mitosis.32Micronuclei can be found in cells of any tissue, but only form in dividing cells.33

There are four generally accepted mechanisms through which micronuclei can form: a)34 the mitotic loss of acentric chromosome fragments (forming structural aberrations), b) mechanical35consequences of chromosomal breakage and exchange, such as from lagging chromosomes, an36inactive centromere or tangled chromosomes (forming structural aberrations), c) mitotic loss of37whole chromosomes (forming numerical aberrations) and, d) apoptosis [94]. However, nuclear38fragments resulting from apoptosis are usually easy to identify because they are much more39numerous or pyknotic than those induced by clastogenic or aneugenic mechanisms. Structural40aberrations are believed to result from direct or indirect interaction of the test chemical with41DNA, while numerical aberrations are often a result of interference with the mitotic apparatus42preventing normal nuclear division.43


3/16

23

Bone marrow is the major haematopoietic tissue in the adult rodent. Administration of a1chemical during proliferation of haematopoietic cells may cause chromosome damage or2inhibition of the mitotic apparatus. These chromosome fragments or whole chromosomes may3

lag behind during cell division and form micronuclei. The erythrocyte is particularly well suited4to analysis for micronuclei because during maturation of the erythroblast to the polychromatic5erythrocyte (a period of about 6 hours following the final mitosis), the nucleus is extruded,6making detection of micronuclei easier [95]. In addition, the polychromatic erythrocyte (PCE)7still contains RNA, and so it stains blue-grey with Giemsa or reddish with acridine orange. This8allows differentiation from mature, haemoglobin-containing erythrocytes (NCE), which stain9orange with Giemsa or are unstained by acridine orange, and facilitates identification of the cells10where micronuclei induced by the test substance may be present [96]. Sampling of PCEs from11the bone marrow or peripheral blood prior to their differentiation to mature erythrocytes is12critical; once a PCE has matured, associating the presence of micronuclei in these cells with acute13chemical exposure is not possible. Mature erythrocytes persist in peripheral circulation for about141 month [95].15

The micronucleus assay is conducted using the bone marrow or peripheral blood of16rodents, typically mice, as the target tissue; the peripheral blood of species other than the mouse17can be used if there is evidence that demonstrates micronucleated erythrocytes are not rapidly18removed by the spleen [93]. The test chemical is administered to a minimum of five animals of19each sex per group orally by gavage or by intraperitoneal injection. Three doses are prepared so20that the dose range spans a range from the maximum tolerable dose to a dose level that does not21induce appreciable toxicity. One of two treatment schedules is recommended: the test chemical is22administered once and one group is sacrificed at 24 hours and another at 48 hours after treatment23(for bone marrow), or at 36 and 72 hours after treatment (for peripheral blood); if multiple24treatments are used, a single sample can be taken between 18-24 hours after the last25administration for bone marrow or between 36-48 hours for peripheral blood [93]. A delay26

between chemical administration and sampling of PCEs is necessary to allow sufficient time for27the number of micronucleated PCEs to rise to a peak. This corresponds to the time necessary for28absorption and metabolism of the chemical, the completion of the erythroblast cell cycle,29including any test chemical-induced cell-cycle delay, and for extrusion of the erythroblast nucleus30[95].31

Because the incidence of micronucleated PCEs is rare in untreated animals, to allow for32appropriate statistical power, at least 2000 PCEs per animal should be scored for the incidence of33micronuclei. The proportion of PCE among total erythrocytes should also be determined from at34least 200 total erythrocytes for bone marrow samples or from at least 1000 for peripheral blood35samples [93]. Data indicating the test chemical depressed the PCE : total erythrocyte (TE) ratio36are suggestive of cytotoxicity and provide evidence that the test chemical reached the bone37

marrow. However, significant clinical symptoms of toxicity or plasma test chemical38 concentrations may also be used as a measure of exposure, since the bone marrow is a relatively39well-perfused tissue and the plasma test substance concentration should approximate the level of40bone marrow exposure. However, if the test chemical is severely cytotoxic, finding and scoring412000 PCEs at higher concentrations may be difficult; in these circumstances it is particularly42important to have multiple dose groups.43

Because micronuclei are relatively rare, manual enumeration by light microscopy is time44consuming. For that reason, newer flow cytometric or image analysis methods have been adapted45for the rapid processing of slides. These methods have yet to undergo a thorough evaluation, but46they offer tremendous potential for improving the sensitivity and the efficiency of the assay.47


4/16

24

Any test chemical that induces an increase in the frequency of micronucleated PCE is1concluded to have induced chromosomal aberrations in vivo, but further mechanistic information2useful to distinguish micronuclei induced by clastogenic or aneugenic chemicals can be obtained.3

Micronuclei of aneugenic origin will contain centromeres, the presence of which can be verified4using one of two molecular cytogenetic methods: immunofluorescent CREST-staining or5fluorescence in situhybridization (FISH) with pancentromeric DNA probes (see [94,96]). This6information is useful in risk assessment because aneugens may exhibit threshold dose-responses7that are not typical of those chemicals that interact directly with DNA.8

2.3.1.2 Benefits & limitations. Based on the mechanism for micronucleus formation, the9micronucleus assay, in principle, is able to detect both clastogens and some aneugens. There is a10low spontaneous micronucleus frequency in erythrocytes (typically 50% reduction of the mitotic index).46


5/16

25

Following administration, sampling should usually be conducted twice. In order to1accumulate metaphase cells, cell division is arrested by administration of a mitotic inhibitor, such2as colchicine, 3-5 hours prior to sacrifice. Cells in their first metaphase after administration3

should be examined, so that any induced increase in the frequency of chromosomal aberrations4can be related to the treatment. Therefore, half the animals should be sacrificed for the first5sampling after a time period equivalent to 1.5x the normal cell cycle length (usually 12-18 hours6for most species used). Because of the potential for some chemicals to induce mitotic delay, the7remainder of the animals should be sacrificed 24 hours after the first sampling time. Bone8marrow cells are obtained immediately, exposed to hypotonic solution, fixed and stained. Using9light microscopy, an observer should score a minimum of 1000 cells/animal for mitotic index and10a minimum of 100 metaphase cells/animal for chromosomal aberrations. Cells scored for11chromosomal aberrations should have a number of centromeres within the range 2n 2, because12chromosomes can be lost during slide preparation[100].13

A test chemical that induces an increase in the frequency of structural aberrations,14

including chromosome-type and chromatid-type aberrations is considered to be clastogenic under15the test conditions. A test chemical that induces an alteration in the frequency of aneuploid cells16is considered to be aneugenic.17

2.3.2.2 Benefits and limitations. The chromosomal aberration assay detects clastogenic effects18of a test chemical by direct examination of metaphase cells; this often is more informative19because the types of aberrations can be described and classified, which allows the assay to20provide mechanistic information more readily than the micronucleus assay. The assay has been21in use for many years, it is accepted by regulatory agencies globally and a large database exists22for comparisons. Although bone marrow cells are the usual target, the assay can be adapted to23allow examination of other cell types, including spermatogonia. This allows the assay to be used24to evaluate the potential for germ cell chromosome aberrations that could lead to heritable genetic25

effects.26

However, conducting a chromosomal aberration assay is much more time consuming27than the micronucleus assay. It requires skilled personnel to correctly identify aberrations and is28not adaptable for automated scoring. As a result, it is necessarily subjective. Because of29cytotoxicity or chromosomal damage during slide preparation, the number of scoreable30metaphases may be low; this could make it difficult to find a sufficient number of intact31metaphases/animal, which would reduce the sensitivity of the test to detect small increases and32may also hinder the evaluation of polyploidy. Furthermore, like the micronucleus assay, the33chromosomal aberration assay does not, by design, provide information regarding whether the34test chemical induced gene mutations in the target cells or other tissues within the animal.35

2.4 IN VIVO ASSAYS FOR SOMATIC CELL GENE MUTATION IN36 ENDOGENOUS GENES37

In vivogene mutation assays in endogenous genes are rarely used for testing purposes38because of the lack of effective methods. The tests that currently exist are cumbersome and39generally not suitable for routine use. The mouse spot test is the only test with an OECD test40guideline, but some promising mutation tests using endogenous genes have also been developed.41These endogenous gene mutation assays include RetinoBlast,Hprt,Aprt, Tk

+/-, andDlb-1.42

2.4.1 MOUSE SPOT TEST. The mouse spot test was developed as a rapid screening test to43detect gene mutations and recombinations in somatic cells of mice. The recommended44methodology is described in OECD Test Guideline 484 [101].45


6/16

26

2.4.1.1 Principles. Although colour spots were induced experimentally by X-irradiation in 19571[102], the ability of chemicals to induce these genetic changes was not recognized until 1975,2when Fahrig induced colour spots by treatment of mice with ENU [103]. The mouse spot test is3

based on the observation that chemical mutagens can induce colour spots on the fur of mice4exposed in utero. The colour spots arise when mouse melanoblasts heterozygous for several5recessive coat colour mutations lose a dominant allele through a gene mutation, chromosomal6aberration or reciprocal recombination, allowing the recessive gene to be expressed [104].7Melanoblasts migrate from the neural crest to the midline of the abdomen during days 8-12 of8embryonic development, while continuing to divide to produce melanocytes [105]. Those9melanocytes that carry a coat colour mutation will result in differing pigmentation of the fur in a10band stretching from the back to the abdomen.11

Mice of the T-strain are mated with those of the HT or C57/B1 strain to produce embryos12with the desired genetic characteristics. In general, treatment of about 50 dams with the test13chemical would be sufficient to produce the desired ~300 F1animals per group. Dams are treated14

on days 8, 9 or 10 of gestation by gavage or intraperitoneal injection. Three or four weeks after15birth, the mice should be examined for coat spots under code. There are three classes of spots:16white ventral spots that are presumed to be the result of chromosomal aberrations leading to cell17death; yellow, agouti-like spots that are likely to be a result of misdifferentiations; and pigmented18black, grey, brown or near white spots randomly distributed over the whole coat, which are the19result of somatic mutations. However, only the last class of spot has genetic relevance. A20chemical that induces a biologically significant increase in the number of genetically relevant21(somatic mutation) spots is considered to be mutagenic in this test system [101].22

By examining fur from the spot using fluorescence microscopy, it is possible to23distinguish the classes of spots from each other and to distinguish, from somatic mutation spots,24the gene loci affected [106]. It is also possible to distinguish different types of genetic events.25

Identifiable gene mutations are caused by a mutation at the clocus that produces cells with the c26(albino) and c

ch(chinchilla) alleles, which causes light brown spots. Reciprocal recombinations27

result from a crossing-over involving the linked p (pink eyed dilution) and c (albino) loci; the28resulting recombinants are homozygous for either the wild type (visible as black spots) or mutant29alleles (visible as white spots) [105].30

2.4.1.2 Benefits and limitations. The mouse spot test is capable of detecting both gene31mutations and some types of chromosomal aberrations. It is relatively easy to conduct, does not32require specialized expertise, and the endpoint of interest (coat spots) can be directly identified by33visual examination. Further information regarding the causes of different types of mutagenic34events and the gene loci involved can be inferred by examining fur from the spots35microscopically. Basic information regarding the reproductive toxicity or teratogenicity of the36

test chemical can also be obtained by looking for obvious malformations or reduced numbers of37 pups in each litter. However, mutagenic activity is only detected within the small melanocyte38population very early in development. Furthermore, the test requires a large number of animals,39making it very costly to conduct. The high cost and the trend towards reduction of animals used40in toxicological testing has greatly limited the use of this assay.41

2.4.2 RETINOBLAST (EYE-SPOT) ASSAY. The retinoblast (eye-spot) assay is a variant of the42mouse spot test that identifies deletion mutations by scoring colour spots in the retinal pigment43epithelium instead of the coat. There is no methodology guideline for this assay and it is not yet44commonly used, other than for research applications.45


7/16

27

2.4.2.1 Principles. The eye-spot assay is similar in principle to the mouse spot test. It uses the1C57BL/6Jp

un/p

unstrain of mouse, which carries the pink eyed unstable mutation (p

un), a 70-kb2

tandem duplication at the pink-eyed dilution locus [107]. The pun

mutation carried by the test3

strain is an autosomal recessive mutation that produces a light grey coat colour and pink eyes.4Loss of one copy of the p

un tandem duplication causes reversion of the pun mutation to the5

wildtypep in a retinal pigment epithelial (RPE) precursor cell and leads to the production of a6RPE cell with black pigmentation, which is visible against the remaining non-pigmented RPE7cells [108]. With this assay, the frequency of mutations (deletions) affecting one copy of the8tandem duplication at thep

unlocus can be measured.9

The assay is most commonly conducted by treating dams from the C57BL/6J pun/ p

un10

strain with the test chemical by a relevant route of exposure (gavage, intraperitoneal injection or11inhalation) at approximately day 10 of gestation. Offspring are sacrificed at the age of 20 days,12the eyes are removed, the retina is placed on a slide and is examined by light microscopy. The13number of spots, which are observed as a single pigmented cell or groups of pigmented cells14

separated from each other by no more than one unpigmented cell, are counted. Each eye-spot15corresponds to onep

unmutation. A test chemical that induces a statistically significant increase in16

the frequency of eye spots compared with the negative control is concluded to have induced17mutations at thep

unlocus [109].18

Reversion of thepun

mutation arises from intrachromosomal recombination that results in19the deletion of one of the tandem fragments at the p

un loci. The deletion can occur by several20

mechanisms, such as intrachromosomal crossing-over, single strand annealing, unequal sister21chromatid exchange, and sister chromatid conversion [110]].22

2.4.2.2 Benefits and limitations. A number of carcinogens have been found to induce23intrachromosomal recombinations [110-113], so the endpoint scored by this assay is directly24

relevant to the assessment of carcinogenicity. The assay assesses a type of genetic effect25 (deletion mutation) that is not identified by many other assays. The eye-spot assay requires fewer26animals than the mouse spot test to achieve a high sensitivity [113] and it allows for direct27examination at the single cell level. However the target cells in this assay are RPE precursor28cells, which proliferate only during the period starting at about embryonic day 9 until shortly after29birth [114]. Because the assay is used to determine the frequency of deletions occurring in30embryos, it is not necessarily relevant to assessing deletions occurring in adult animals.31

2.4.3 GENE MUTATION ASSAYS USING ENDOGENOUS GENES WITH SELECTABLE32PHENOTYPES33

2.4.3.1 Hprt. TheHprtassay uses one of the few genes that are suitable for mutation analysis in34

wild type animalsin vivo. It has been widely used in research applications, but has yet to undergo35 a detailed performance assessment necessary for it to begin to be used for routine testing. No36guidelines exist for the conduct of this assay.37

2.4.3.1.1 Principles. TheHprtgene is located on the X-chromosome and spans 32 kb and 46 kb38in human and rodent cells, respectively. Both male and female cells carry only one active copy of39the Hprt gene; in female cells one copy of the X-chromosome is inactivated. The Hprt gene40codes for hypoxanthine-guanine phosphoribosyltransferase (HPRT), which plays a key role in the41purine salvage pathway. HPRT catalyses the transformation of purines (hypoxanthine, guanine,42or 6-mercaptopurine) to the corresponding monophosphate, which is cytotoxic to normal cells in43culture. The assay is based on the observation that cells with mutations in theHprtgene have lost44the HPRT enzyme and survive treatment with purine analogues.45


8/16

28

Mice or rats are treated with the test chemical by an appropriate route of exposure. After1a fixation period of several weeks, the spleens are removed from sacrificed animals and cultures2of splenic T-lymphocytes are established. T-lymphocytes are particularly useful to examine3

because they circulate throughout many tissues, which allows them a greater chance to come into4contact with an administered mutagen than cells that are permanently resident in a single tissue.5T-lymphocytes are also long-lived in circulation and they continue to undergo cell division,6which makes the identification of mutant cells possible. In addition to T-lymphocytes, mutant7frequency has also been determined in other cells, including those from the kidney, thymus and8lymph nodes.9

Mutant selection has been described by Tates et al. [115]. Cells are incubated in10microwell culture plates with the selective agent 6-thioguanine, a purine analogue that is a11substrate for HPRT and is toxic to non-mutant cells. After 8-9 days, cloning efficiency plates are12scored, while mutant frequency plates are scored after a 10-12 day expression period. The mutant13frequency is calculated as the ratio between the cloning efficiencies in selective media versus14

cloning media. A significant increase in mutant frequency in treatment cultures compared with15controls indicates the test chemical has induced mutation at the Hprt locus. The average16spontaneous mutant frequency at the Hprt locus is in the range of 10

-6 [116]. Using standard17techniques, further molecular analysis ofHprtmutations can be performed, if desired18

2.4.3.1.2 Benefits and limitations. The Hprtassay detects point mutations, frameshifts, small19insertions and small deletions. As an endogenous gene it is transcriptionally active and thus is20subject to transcription-coupled DNA repair [117]. However, becauseHprtis an X-linked gene21and is therefore functionally hemizygous, it is not particularly efficient at detecting large22deletions, chromosomal recombination and nondisjuncton events that may disrupt essential23flanking genes, and which are more effectively identified with assays using endogenous24autosomal genes [118]. Deletions extending into adjacent essential genes in hemizygous regions25

are usually lethal to the cell because there is no homologous region to compensate for the loss of26essential gene function. Because Hprt is not a neutral gene, there is selection pressure against27Hprtdeficient lymphocytes, particularly in young animals [119]. In addition, dilution of mutant28T-lymphocytes in circulation occurs as peripheral lymphocyte populations are renewed; this is29also affected by the age of the animal. The time from exposure to maximum average mutant30frequency was 2 weeks in the spleen of ENU exposed preweanling mice and 8 weeks in adult31mice [120]. As a result, sampling in the spleen must be carefully timed to detect the maximum32mutant frequency based on these factors. Although Hprtmutant frequency can be determined33from any tissue that can be subcultured, it is typically only assessed in T-lymphocytes, which34prevents identification of mutagenic effects that may arise preferentially in other target tissues.35

2.4.3.2 Aprt. TheAprtassay uses a constructedAprtheterozygous mouse model. Like theHprt36

model,Aprtis widely used in research, but is not yet used in routine genetic toxicology testing.37 No internationally accepted test guideline exists.38

2.4.3.2.1 Principles. Aprtis the gene coding for an enzyme (adenine phosphoribosyltransferase)39that catalyzes the conversion of adenine to AMP in the purine salvage pathway; it is expressed in40all tissues. The mouse Aprtgene is located on chromosome 8 [121], while the human gene is41located on chromosome 16 [122]. In the mouse, because of its location near the telomere, it is a42large target for chromosomal events such as translocation and mitotic recombination [123].43

Several Aprt knockout models have been created. A heterozygous Aprt+/-

mouse has44been developed by disrupting the Aprtgene in embryonic stem cells using a conventional gene45targeting approach [116]. This model can be used to investigate induced forward mutations46


9/16

29

leading to the loss of the autosomal dominant locus in T-lymphocytes and skin fibroblasts, as well1as mesenchymal cells from the ear [124] and epithelial cells from the kidney [125]. Using2methods similar to those used for theHprt model,Aprtheterozygous mice are treated with the test3

chemical. After a fixation period of several weeks, the animals are sacrificed and typically4splenic T-lymphocytes or skin fibroblasts are isolated and cultured. Aprtdeficient mutants are5selected using purine analogues such as 9-azaadenine or 2,6-diaminopurine, which are toxic to6Aprt proficient cells. After a 6-8 day expression period, the cloning frequency and mutant7frequency are determined using methods similar to those used for the Hprt model [115]. The8spontaneous mutant frequency at theAprtlocus in heterozygous mice is approximately 8.7 x 10

-69in T-lymphocytes [116] and 1.7 x 10-4 in skin fibroblasts [123]. Using standard techniques,10further molecular analysis ofAprtmutations can be performed, if desired11

An Aprt-/-

homozygous knockout mouse has also been developed, which is capable of12detecting chemicals that cause point mutations [126]. These mice haveAprtalleles inactivated by13reversible point mutations. Mice are administered the test chemical and held for a fixation period14

of several weeks, after which they are injected with14

C-adenine. Cells that have reverted toAprt

15 + have a functional adenine phosphoribosyltransferase enzyme and can sequester radiolabelled16

adenine by conversion to AMP, which is subsequently incorporated into nucleic acids. Using17autoradiography or scintillation counting, the frequency of revertant cells can be determined. The18in situmethod also enables identification of the cell types that are most susceptible to mutation19[126].20

2.4.3.2.2 Benefits and limitations. Because it is an autosomal heterozygous locus, Aprt can21detect, in addition to the point mutations, frameshifts and small deletions detectable by Hprt,22events that may lead to loss of heterozygosity, such as large deletions, mitotic non-disjunctions,23mitotic recombinations and gene conversions, if the function of the deleted essential gene is24provided by the homologous chromosome. The ability to detect the full spectrum of autosomal25

mutations allows Aprt to be a much more versatile biomarker than Hprt. However, the use of26Aprtas a mutational target is limited to only a few tissues due to the detection method, which27relies on culturing techniques. LikeHprt, Aprtis not a neutral gene and there may be negative28selection pressures on Aprtmutant cells, as well as influences on mutant frequency arising from29dilution as the T-lymphocyte pool is renewed.30

2.4.3.3 Tk+/-. A Tkheterozygous mouse model has been constructed. This model is currently31limited to research applications, as it remains in an early stage of development. No methodology32guidelines currently exist.33

2.4.3.3.1 Principles. The heterozygous Tk +/-

mouse was created by using a mouse embryonic34stem cell line with one allele of the Tk gene inactivated through targeted homologous35

recombination [127,128]. Tkis generally expressed only in dividing cells and encodes thymidine36 kinase, which is involved in pyrimidine salvage, catalyzing phosphorylation of thymine37deoxyriboside to form thymidylate. The Tk gene is located on the distal portion of mouse38chromosome 11 [129]. Cells that have lost the second Tkallele through mutation to become Tk

-/-39

are easily selected because they survive when cultured in the presence of a pyrimidine analogue40such as 5-bromo-2'-deoxyuridine (BrdU), while thymidine kinase-competent cells (Tk

+/+or Tk

+/-)41

do not.42

C57BL/6 Tk+/-

mice receive the test chemical by a relevant route of exposure either once43or in multiple administrations. Approximately 4-5 weeks after administration, mice are sacrificed44and splenic lymphocytes are isolated and cultured. Mutant selection is performed as described by45Dobrovolsky et al.[128]. Cells are incubated in microwell culture plates with the selective agent46


10/16

30

BrdU, a pyrimidine analogue that is toxic to non-mutant cells. After incubation for 10-12 days,1cloning efficiency in treatment and control plates is scored. The mutant frequency is calculated2by dividing the cloning efficiency of cells cultured in the presence of the selecting agent by the3

cloning efficiency of cells cultured in the absence of selection. A significant increase in mutant4frequency in treatment cultures compared with controls indicates the test chemical induced5mutation at the Tklocus. Using standard techniques, further molecular analysis of Tkmutations6can be performed, if desired. The spontaneous mutant frequency is approximately 2 x 10

-5[125].7

2.4.3.3.2 Benefits and limitations. The Tk +/-

mouse is the analogous in vivo model to the8commonly used mouse lymphoma assay. As such, it is a useful model to investigate the in vivo9responses of chemicals found to be mutagenic in the in vitro assay. The Tk model detects10intragenic mutations (point mutations, frameshifts and small deletions) as well as larger effects,11such as chromosome recombination, nondisjunction and large deletions that often lead to loss of12heterozygosity, for which it is particularly sensitive [128,130,131]. However, it too is limited to13examining tissues where cells can be easily cultured. In addition, the use of BrdU, which is itself14

a mutagen, as a selective agent can introduce the possibility that some mutants would be15produced by exposure to the selective agent. Dobrovolsky et al.[128] suggest this potential can16be minimized by keeping cultures under tight selection pressure during the culture phase, since17studies with Tk

+/-mouse lymphoma cells have indicated several cell divisions in the absence of18

the selective agent are required in order to fix mutations [132].19

2.4.3.4 Dlb-1. TheDlb-1specific locus test measures mutations occurring in the small intestine20of treatedDlb-1heterozygotes. The assay was created in 1990, but it remains in the development21stage and has not been widely used. There are no widely accepted methodology guidelines in22existence.23

2.4.3.4.1 Principles. The Dlb-1specific locus test identifies mutations occurring in the small24

intestine of mice heterozygous at the Dlb-1 locus. Dlb-1 is a polymorphic genetic locus that25 exists on chromosome 11 in the mouse [133] and has 2 alleles. Dlb-1bis an autosomal dominant26gene that determines the expression of binding sites for the lectin Dolichos biflorusagglutinin27(DBA) in intestinal epithelium, while Dlb-1

a determines DBA receptor expression in vascular28

endothelium [2]. Mice heterozygous at the Dlb-1locus (Dlb-1a/Dlb-1

b) that develop a mutation29of the Dlb-1

ballele in an intestinal stem cell can be detected by staining an intestinal epithelial30cell preparation with a peroxidase conjugate of DBA and scoring non-staining cell ribbons on the31villus; these cell ribbons are cells derived from a stem cell carrying aDlb-1mutation [134]. The32spontaneous mutation frequency of the Dlb-1gene has been reported to be approximately 1.6 x3310

-5mutants/villus per animal per week [61].34

2.4.3.4.2 Benefits and limitations. TheDlb-1assay can be used for studies of animals in utero,35

as well as adult animals. The number of animals used is consistent with most assays using an36 endogenous reporter gene, which is substantially fewer than are required for the mouse spot test.37However, the assay is restricted to analysis of mutations in the intestinal epithelium following38oral administration of the test chemical. Mutagens acting preferentially at another target tissue39may not be detected. In addition, theDlb-1gene has not yet been cloned, so the molecular nature40of any observed mutations cannot be determined. The assay is still in the development stage and41has not been used, other than for research purposes in a small number of laboratories.42

2.5 INDICATOR TESTS43

Indicator tests are those that do not directly measure consequences of DNA interaction44(i.e. mutation), but rely on other markers that suggest some type of interaction occurred. The two45


11/16

31

most commonly conducted tests are the sister-chromatid exchange assay and the unscheduled1DNA synthesis assay.2

2.5.1 SISTER-CHROMATID EXCHANGE ASSAY. The sister chromatid exchange (SCE) assay is3a widely used method for assessing chromosome breakage and repair, though it is much more4commonly conducted as an in vitro test. Methodology for the in vitro assay is described in5OECD Test Guideline 479, but there is no recommended methodology for the in vivoassay.6

2.5.1.1 Principles. Sister chromatid exchanges (SCE) are reciprocal exchanges of DNA7segments between sister chromatids of a chromosome that are produced during S-phase.8Although the molecular mechanism of these exchanges remains unknown, it is presumed to9require chromosome breakage, exchange of DNA at homologous loci and repair. Work with the10model genotoxin, ENU, has provided direct evidence that suggests the replication fork is the site11of SCE production [135]. However, SCE may not necessarily be caused by direct DNA12interaction in all cases. A chemical that does not damage DNA but instead creates intracellular13

conditions that favour inhibition of DNA replication could, in itself, create SCE. It is also14noteworthy that several strong clastogens, such as ionizing radiation and bleomycin, have failed15to produce an increase in SCE [136]. Because SCE induction does not, itself, indicate a chemical16is mutagenic, the interpretation of the toxicological relevance of SCE is often difficult.17

Rodents are most commonly used for SCE assays. Bone marrow cells are usually18sampled because there is always a pool of dividing cells and they are easy to prepare for scoring.19Generally, groups of five animals per sex are administered the test chemical once by gavage or by20intraperitoneal injection. At 2-3 hours following administration, the animals are administered a21single dose of BrdU, followed at twenty-one hours by an injection of the mitotic inhibitor22colchicine. Three hours later, all animals are sacrificed and slides of bone marrow cells are23prepared and scored by light microscopy. Other cell types, including spermatogonial cells, may24

also be used; the treatment and sampling times for other cells will depend on the cell cycle time25 of the target cells and what is known of the toxicokinetic factors specific to the test chemical.26

2.5.1.2 Benefits and limitations. The SCE assay offers a rapid and relatively inexpensive27assessment of potential test chemical-induced DNA damage; any tissue from which a cell28suspension can be made can be analyzed. A significant number of SCE assays for a wide variety29of chemicals have previously been conducted, which facilitates comparison of relative potencies30of test chemicals. However, the major drawback remains the unknown molecular basis of SCE31induction. Because factors other than direct DNA interaction can cause SCE, an increase in SCE32induction does not necessarily indicate mutagenicity, making interpretation in the context of33genetic toxicity testing difficult. As a result, this assay has fallen out of favour and is now rarely34conducted.35

2.5.2 UNSCHEDULED DNASYNTHESIS ASSAY. The unscheduled DNA synthesis (UDS) assay36is a commonly used method of assessing test chemical-induced DNA excision repair. The37induction of repair mechanisms is presumed to have been preceded by DNA damage. Measuring38the extent to which DNA synthesis occurred offers indirect evidence of the DNA damaging39ability of a test chemical. The recommended methodology is described in OECD Test Guideline40486 [137].41

2.5.2.1 Principles. The UDS assay measures DNA synthesis induced for the purposes of42repairing an excised segment of DNA containing a region damaged by a test chemical. DNA43synthesis is measured by detecting tritium-labelled thymidine (

3H-TdR) incorporation into DNA,44

preferably using autoradiography. The liver is generally used for analysis because, under normal45


12/16

32

circumstances, there are a low proportion of primary hepatocytes in S-phase of the cell cycle;1therefore an increase in DNA synthesis can be more easily attributed to repair of induced DNA2damage, rather than DNA synthesis supporting normal cell division. The liver is also the site of3

first-pass metabolism for chemicals administered orally or by intraperitoneal injection.4

The larger the number of nucleotides excised and repaired, the greater amount of5detectible 3H-TdR is incorporated into DNA. For that reason, the UDS assay is more sensitive in6detecting DNA damage that is repaired through nucleotide excision repair (removal of up to 1007nucleotides) as compared to base excision repair (removal of 1-3 nucleotides) [137]. Test8chemicals more prone to inducing nucleotide excision repair, such as those that form bulky DNA9adducts, have a greater potential to cause detectible UDS. However, the UDS assay does not, in10itself, indicate if a test chemical is mutagenic because it provides no information regarding the11fidelity of DNA repair, and it does not identify DNA lesions repaired by mechanisms other than12excision repair.13

The UDS assay is usually conducted using rats, though other species may be used if14justified. Two dose levels are selected on the basis of preliminary toxicity testing, with the15highest dose defined as that causing toxicity such that higher levels would be expected to produce16lethality. At least 3 animals/group (typically males only) are administered the test chemical once17by gavage. Intraperitoneal injection is not recommended because it could potentially expose the18liver directly to the chemical. A group of animals is sacrificed at 2-4 hours and another at 12-1619hours after treatment. Cultures of hepatocytes are prepared and incubated for 3-8 hours in

3H-20

TdR. Slides are prepared and processed for autoradiography using standard techniques. At least21100 cells per animal are examined and both nuclear and cytoplasmic grains should be counted to22determine the net nuclear grain count (cytoplasmic grains subtracted from nuclear grains).23Chemicals inducing a significant increase in net nuclear grain count for at least one treatment24group are considered to have induced UDS [137,138].25

2.5.2.2 Benefits and limitations. Theoretically, any tissue with a low proportion of cells in S-26phase can be used for analysis. Though only liver is routinely used, a UDS assay using27spermatocytes has been developed [139-141], allowing the measurement of potentially germ cell28specific DNA interactions. Because UDS is measured in the whole genome, it is potentially29much more sensitive than assays examining only specific loci. However, the extent of UDS gives30no indication of the fidelity of the repair process. For that reason, UDS does not provide specific31information on the mutagenic potential of a test chemical, but only information suggesting it does32or does not induce excision repair.33

2.6 RODENT CARCINOGENICITY BIOASSAY.34

The gold standard method for the determination of the carcinogenic activity of a test35 chemical is the chronic rodent carcinogenicity bioassay. Though expensive, the bioassay is36widely conducted and is prescribed by regulatory authorities for many products, such as drugs,37food additives and pesticides. The methodology is described in OECD Test Guideline 451 [142].38

2.6.1 PRINCIPLES. The rodent carcinogenicity bioassay determines whether a test chemical39causes neoplastic lesions in the tissues of experimental animals treated over the majority of their40lifetime. The bioassay is the most comprehensive evaluation that has been developed for41determining the potential for a test chemical to cause cancer. Using time-consuming but sensitive42histopathological examination of all major tissues, the bioassay determines whether the test43chemical caused an increase in the incidence of tumours in treatment groups relative to controls.44Those chemicals producing an increase in neoplastic lesions, in the absence of information45


13/16

33

suggesting the effect is a species-specific response, are considered carcinogenic for the purposes1of hazard identification within the human health risk assessment process.2

Usually rodents are used, but other species (i.e. dogs, monkeys) can be used if justified.3Often both mice and rats are selected for two studies run concurrently. Based on the results of4subchronic toxicity studies, a minimum of 3 dose levels are selected: the maximum tolerated5dose, defined as the dose that produces mild clinical evidence of toxicity without substantially6altering the normal life span due to effects other than tumours, as well as 2 additional dose levels,7the lowest of which is not expected to produce any clinical evidence of toxicity. Beginning no8later than 8 weeks of age, 50 animals/sex/group are administered the test chemical daily by a9relevant route of exposure (oral, dermal or inhalation). An additional 50 animals/sex are assigned10to the negative control group. The aim is to expose animals for a period of time equivalent to no11less than 80% of their total expected lifetimes (e.g. 24 months for rats, 18 months for mice). At12two intervals during the study and at sacrifice, blood samples are taken for haematological and13clinical chemistry analysis. At end of the treatment period all animals are sacrificed and tissue14

samples are taken from most tissues for gross and histopathological examination.15

The test chemical is considered to be carcinogenic if it produces a significant increase in16the incidence of tumours in at least one dose group of one sex of one species. For a negative17result to be considered valid, no more than 10% of any group may be lost due to autolysis,18cannibalism, or management problems, and survival in each group is no less than 50% at 1519months for mice and 18 months for rats. By the completion of the study, survival should not fall20below 25%.21

2.6.2 BENEFITS AND LIMITATIONS. The carcinogenicity bioassay directly determines whether22a test chemical causes the ultimate endpoint of interest in genetic toxicology testing cancer. In23contrast to the majority of short-term genotoxicity tests, the bioassay identifies neoplastic lesions24

produced by both genotoxic and non-genotoxic mechanisms. The bioassay has been conducted25 for many years and a large number of chemicals have been studied, particularly by the U.S.26National Toxicology Program, which has conducted a significant proportion of the studies27conducted world-wide.28

Despite these benefits, the bioassay is time-consuming, prohibitively expensive and29labour-intensive, which restricts its practical use to those agents that are of a high level of concern30with respect to human exposures. Moreover, the use of rodent carcinogenicity to predict31carcinogenicity in humans, or even in other species of rodent, is questionable for some chemicals.32Nevertheless, it remains the best method currently available for determining the carcinogenic33activity of chemicals.34

2.7 CONCLUSION35

In vivo assays are required components of any thorough genetic toxicity testing programme.36Unlike in vitroassays, in vivoassays are capable of accounting for toxicokinetic factors and DNA37repair processes that may, in some cases, modulate genotoxicity within the whole animal.38However, existing assays are seriously limited by a range of different factors, including cost of39the assay, the number of tissues in which genotoxicity may be measured, the state of40understanding of the endpoint, and the nature of the chemicals that will be detected.41


14/16


15/16

35

Study

TK+/-*

Dlb-1*

Sisterchromatid

exchange*

Unschedule

dDNA

synthe

sis

Carcinogenicity

bioassay

Species

Tkheterozygous

mouse

Dlb-1heterozygo

us

mouse

Ratormouse

Rat;alsomouse

Typicallyrodent

Number&sex

5/sex/doseshouldbe

sufficient

5/sex/doseshouldbe

sufficient

5/sex/groupshouldbe

sufficient

Atleast3/gro

up;

usuallymalesare

sufficient

Minimumof

50/dose/sex

Doselevels

Atleast2,plus

controls;maximum

doseshouldbethe

MTDorinducetarget

tissuecytotoxicity

Atleast2,plus

controls;maximu

m

doseshouldbeth

e

MTDorinducetarget

tissuecytotoxicit

y

Atleast2,plus

controls;maximum

doseshouldbethe

MTDorinducetarget

tissuecytotoxicity

Atleast2,plus

controls;max

imum

doseshouldb

ethe

MTDorindu

cetarget

tissuecytotoxicity

Aminimumof3,plus

controls;maximum

doseshouldbethe

MTDselectedfrom

preliminarytests

Routeof

administration

Therouteconsidered

mostrelevantto

humanexposure

Oraladministrati

on

Gavageor

intraperitoneal

injectionaremost

common

Oral(i.p.inje

ctionis

notrecommended)

Relevanttohuman

exposure:oral,

dermalorinhalation

aretypical

Treatmentschedule

Singleormultiple

administration

Singleormultiple

administration

Singleadministration

Singleadmin

istration

Dailyadministration

foraminimumof18

(mice)or24(rats)

months

Tissueanalyzed

Splenict-

lymphocytes

(primarily)

Intestinalepithelial

cells

Bonemarrow(other

celltypesmaybe

used)

Liver

Allmajortissuesand

organsexamined

histopathologically

Samplingtimes

Mutationfrequency

determinedaftera

selectionperiodof

10-12daysinBrdU

Sacrifice24hours

afteradministration;

animalsreceiveBrdU

at2-3hoursand

colchicineat21hours

2-4and12-16hours

afteradministration;

livercellsare

preparedand

incubatedfor

3-8

hoursin3H-T

dR

Atthecompletionof

theexposureperiod

Analysis

Calculationof

inducedmutant

frequency

Calculationof

inducedmutant

frequency

ScoringSCEby

examinationusing

lightmicroscopy

Determinatio

nofnet

nucleargrain

count

byautoradiographyin

atleast100c

ellsper

animal

Grosspathological

andhistopathological

examinationoftissues

In-lifephase

About2months

About2months

About1week

About1week

Twoyears

*nonguideline

study

1


16/16

36

existing genotoxicity assays (2)

Documents