Synopsis of the in vivo results obtained with the 10 known or suspected aneugens tested in the CEC collaborative study

Mutation Research, 287 (1993) 131-137 131 © 1993 Elsevier Science Publishers B.V. All rights reserved 0027-5107/93/$06.00 MUT 00065 Synopsis of the in vivo results obtained with the 10 known or suspected aneugens tested in the CEC collaborative study I l se -Dore Ad ler GSF-lnstitut fiir Siiugetiergenetik, D-8042 Neuherberg, Germany (Accepted 3 November 1992) Keywords: CEC Aneupl~Jidy Project; Test results, synopsis Summary The synopsis of the in vivo test results in the first collaborative CEC Aneuploidy Project with 10 selected chemicals, colchicine (COL), econazole (EZ), chloral hydrate (CH), hydroquinone (HQ), diazepam (DZ), thiabendazole (TB), cadmium chloride (CD), thimerosal (TM), pyrimethamine (PY) and vinblastine (VBL), allowed several conclusions. (1) The spindle poisons, COL and VBL, were positive in all bone marrow and germ cell tests; (2) the clastogen HQ also induced aneuploidy in somatic and germinal cells; (3) the other seven compounds gave contradictory results either between laboratories or between test systems which require further experimental clarification; (4) CREST labeling or in situ hybridization for centromere identification showed about 70% fluorescent signals in micronuclei induced by COL or VBL but only about 15% in HQ induced micronuclei; (5) the tests for induction of a delay in cell division progression can be recommended as a prescreen for possible aneugens; (6) all test methods applied in these experiments require standardization with respect to sample size, sampling times and statistical treatment of the data. A second CEC Aneuploidy Programme has started recently to answer some of the questions raised by the first study regarding tissue and sex specificities. In 1987, a coordinated program was initiated under the auspices of the Commission of the European Communities (CEC) with the aim of developing a data base which would allow the recommendation of screening tests for the assess- ment of aneuploidy induction (Parry, 1987). Dur- ing the first 4 years the in vivo studies of three laboratories were supported in part by CEC fund- Correspondence: Dr. Ilse-Dore Adler, GSF Forschungszen- trum ftir Umwelt und Gesundheit, Institut fiir Siiugetiergene- tik, D-8042 Neuherberg, Germany. ing (I.-D. Adler, Neuherberg; F. Pacchierotti, Rome; R. Barale, Pisa) and one laboratory partic- ipated voluntarily (M. Kirsch-Volders, Brussels). The assays ranged from a simple test of changes in the mitotic index in bone marrow preparations via chromosome counts and micronucleus counts in bone marrow cells of mice to germ cell studies in the mouse. The micronucleus test was supple- mented by CREST and in situ hybridization anal- yses of the origin of micronuclei induced by se- lected chemicals. Chromosomal aberrations were also analyzed. In male germ cells meiotic delay was assessed and chromosomes in second meiotic 132 metaphases were counted. The 10 chemicals tested in the coordinated exercise were colchicine (COL), econazole (EZ), chloral hydrate (CH), hydroquinone (HQ), diazepam (DZ), thiabenda- zole (TB), cadmium chloride (CD), thimerosal (TM), pyrimethamine (PY) and vinblastine (VBL). All experiments were performed with mice, each laboratory employed their own stocks or strains and the protocols used, even though often de- signed to assess the same endpoint, i.e., micronu- cleus induction, differed from laboratory to labo- ratory particularly with respect to sample sizes and sampling times. Furthermore, statistical treatments of the data were not identical. The conflicting results indicate that these parameters of testing require harmonization. Results Bone marrow (1) Mitotic index changes, c-mitotic effects and mitotic delay In mouse bone marrow cells the induction of mitotic arrest was measured as increased mitotic index within short intervals (1-3 h) after treat- ment. In the same experiments mitotic metaphase cells were classified into five groups according to their increasing degrees of chromatid separation and contraction to determine the c-mitotic (col- chicine-like) effect. Furthermore, the frequencies of mitotic metaphases and anaphases were deter- mined to assess any delay in the progression of the mitotic division. Significant effects for two or three parameters were observed with five of the chemicals, COL, EZ, CH, HQ and vincristine (as replacement for VBL in the early experiments), while the other five compounds showed no effects (Miller and Adler, 1989). The mitotic delay was determined as average cell generation time (AGT) based on counting the proportions of first, second and third mitoses in BrdUrd-labeled bone marrow mitoses. The AGT was prolonged by five compounds, COL, CH, HQ, TB and VBL (Leopardi et al., 1993; Manca et al., 1990; Pacchierotti et al., 1991), while the other five chemicals, EZ, DZ, CD, TM and PY, had no effect on the AGT when com- pared with the concurrent solvent controls. (2) Hyperploidy and polyploidy Chromosome counts in BrdUrd-labeled second mitoses of mouse bone marrow cells showed in- duction of hyperploidy by COL, CH, HQ and VBL (Gustavino et al., 1991; Leopardi et al., 1993; Manca et al., 1990; Marrazzini et al., 1993; Pacchierotti et al., 1991). One participating laboratory also found signifi- cant increases in hyperploid cells with EZ and DZ (Marrazzini et al., 1993). Comparing all treated groups with the combined controls of all their experiments these authors also found a sig- nificant increase in hyperploid bone marrow cells with CD, TB and TM (Marrazzini et al., 1993). The differences in results from the two labora- tories for these five chemicals are due to three experimental parameters: doses, sampling inter- vals and sample sizes. In the case of EZ one laboratory found no hyperploidy induction when they treated four animals with 80 mg/kg and sampled bone marrow cells only once at 18 h after treatment (Leopardi et al., 1993). The other laboratory found an increased frequency of hy- perploidy at 24 and 48 h after treatment with 320 mg/kg, using three animals for each time interval (Marrazzini et al., 1993). The same correlation between protocol differences and results was true for TB. With 62.5 mg/kg sampled at 18 h one laboratory found no effect (Leopardi et al., 1993). With 250 mg/kg TB the other laboratory ob- tained a significant induction of hyperploidy, though only when the results from the 24 and 48 h sampling times were pooled (Marrazzini et al., 1993). For DZ and TM the sampling time (18 h vs. 24 h) seemed to be the determining differ- ence. For CD the different results in the two laboratories could be attributed to sample size. One group only tested a dose of 5 mg/kg at the 18 h sampling interval with four treated animals (Leopardi et al., 1993). The other laboratory tested 5, 10 and 20 mg/kg of CD at 24 and 48 h. The results were only significant when all treated groups were pooled and compared with all con- trol animals in the whole study so that the total sample size was actually 18 treated and 26 control animals (Marrazzini et al., 1993). These compar- isons demonstrate that a single dose and a single sampling interval are not sufficient to obtain a reliable result. TABLE 1 RESULTS OF TESTING FOR STRUCTURAL CHROMOSOME ABERRATIONS BY U. KLIESCH (NEUHERBERG) 133 Control TB CD TM PY Dose (mg/kg) - 500 10 20 20 Mitotic index (%) 4.7 6.1 5.0 5.9 4.9 Cells with aberrations (% ± SE) 0.25 ± 0.2 0.4 + 0.2 0.6 + 0.4 0.4 ± 0.2 1.0 ± 0.5 Each treated group consisted of five males; bone marrow was sampled 24 h after treatment; 100 cells per animal were scored for gaps, breaks and exchanges; mitotic indices were determined as frequencies of mitotic stages among 500 cells per animal. For abbreviations of chemicals see text. Polyploidies were found with COL and VBL in both laborator ies (Gustavino et al., 1991; Manca et al., 1990; Marrazz in i et al., 1993; Pacchierott i et al., 1991). None of the other chemicals which increased the frequencies of hyperplo id cells also significantly increased the frequency of polyploid cells; however, some polyploid cells were ob- served after HQ and PY t reatment in one of the laborator ies (Marrazzini et al., 1993). (3) Structural chromosome aberrations Of all 10 chemicals tested only HQ showed a strong clastogenic effect (Wang_Xu and Adler , 1990; Marrazz in i et al., 1993). A slight clastogenic response was also seen with the two spindle poi- sons COL and VBL as well as with CD in one laboratory (Marrazzini et al., 1993) but not in the other with COL and CD (Wang Xu and Adler , 1990; Table 2). Apar t from varying factors in test protocols the main reason for the dif ferences in the assessment of dastogenic i ty is most likely to be the dif ferent criteria used for discr iminating chromat id gaps and breaks. None of the other chemicals induced structural chromosome aberra- tions in mouse bone marrow cells under the ex- per imenta l condit ions used. (4) Micronucleus test Micronuclei (MN) were found in mouse poly- chromatic erythrocytes (PCE) after t reatment with COL, HQ and VBL by all part ic ipat ing laborato- ries (Adler et al., 1991; Marrazzin i et al., 1993; TABLE 2 SYNOPSIS OF IN VIVO RESULTS COL EZ CH HQ DZ TB CD TM PY VBL Bone marrow C-mitoses + 1 + i + 1 + 1 _ 1 _ 1 _ 1 _ 1 _ 1 + 1 AGT +2 2 +2 +2 2 +2 2 2 2 +2 Hyperploidy +2,+3 _2,+3 +2,+3 +2,+3 _2,+3 _2,+3 _2,+3 _2,+3 2, 3 +2,+3 Polyploidy +2,+3 _2,_3 2, 3 _2,+3, _2,_3 2, 3 2, 3 _2, 3 2, 3 +2,+3 SA 1 ,+3 1 3 I 3 +1,+3 1 3 1 3 1 +3 1 3 1 3 +3 1' 1' MN +1,+2, 1, 2 , I , 2 , + 1,+2, , 2 , 1 , 2 , 1 , 2 , 1 , 2 , , 2 , +1,+2, +3,+4 +34 +3,_4 +3,+4 +3, 4 +3,+4 +3, 4 +3,_4 +3,+4 +3,+4 CREST + MN (%) 66 15 70 Sgermatocytes Meiotic delay + 1 + 1 + i + 1 + 1 _ 1 + 1 _ 1 _ 1 + 1 Hyperploidy +1,+2 +1,_2 +1,_2 +1,+2 +1,_2 _1,+2 +1,_2 1, 2 1, 2 +1 AGT, average generation time; SA, structural chromosome aberrations; MN, micronuclei. For abbreviations of chemicals see text. Each plus or minus represents the results from one laboratory according to the author's conclusion. Laboratories: 1 I.-D. Adler, Neuherberg, Germany; 2 F. Pacchierotti, Rome, Italy; 3 R. Barale, Pisa, Italy; 4 M. Kirsch-Volders, Brussels, Belgium. 134 Pacchierotti et al., 1991). The data of two of the laboratories showed no micronucleus induction with any of the other seven chemicals despite the fact that they were tested quite extensively (Adler et al., 1991). The sample sizes were large (six and 10 animals per dose-time group, 1000 or 2000 PCE per animal, respectively), at least three doses per chemical were tested in one of the two labo- ratories (Neuherberg) with two sampling times in each of the two laboratories (Neuherberg and Brussels). One other laboratory obtained a positive mi- cronucleus response with TB and PY (Leopardi et al., 1993). However, the micronucleus induc- tion was determined in the experiments that were performed for aneuploidy scoring, i.e., the ani- mals were implanted with BrdUrd tablets under anesthesia and the bone marrow cells were treated with a hypotonic solution (Leopardi et al., 1993). The authors concluded that their experi- ments were not designed to accurately determine micronucleus induction and the high micronu- cleus yields obtained could have been influenced by the variant experimental protocol. Another laboratory obtained a significant mi- cronucleus induction with CH, DZ, TB, CD, TM and PY (Marrazzini et al., 1993). The effect was only observed with one of several doses for each of the compounds, usually in the lower dose range at the 24 h interval, and represented about a doubling of the concurrent control values (Mar- razzini et al., 1993). The positive effects were confirmed in follow-up experiments for three of the chemicals, CD, CH and DZ. The positive results for DZ, in particular, were due to the low concurrent control values, 0.08% micronucleated PCE in both the first and the second experiment, 0.13% micronucleated PCE in the third experi- ment (Marrazzini et al., 1993). Elaborate statistical evaluations by Marrazzini et al. (1993) fitted time and dose responses to weighted polynomial regressions stepwise with in- creasing order up to the third degree. This proce- dure eventually lead to a positive call for EZ even though none of the individual dose-time groups showed a significant difference from the controls. The differences in calls between the three laboratories (Brussels, Neuherberg and Pisa) can- not be solely explained by differing protocols since the dose ranges and sampling times were similar. The small individual sample size of three animals in one laboratory was compensated by the follow-up experiments (Marrazzini et al., 1993). The major difference was seen in the sta- tistical treatment of the data. Therefore, the bone marrow micronucleus results with DZ, EZ, TB, CD, TM and PY require clarification. The micronucleus results with CH were called negative in two (Adler et al., 1991; I~opardi et al., 1993) and postive in one laboratory (Marraz- zini et al., 1993). After reanalysis of the data obtained with CH treatment of males only apply- ing the linear trend test (Margolin and Risko, 1988) over dose regardless of sampling time the results of three of the laboratories (Adler et al., laboratory Neuherberg; Leopardi et al., 1993; Marrazzini et al., 1993) were significantly differ- ent from the control. Therefore, it was concluded that CH induced micronuclei in bone marrow cells, although none of the present results show as clear a positive response as did the experi- ments of Russo et al. (1992). (5) Identification of centromeres in micronuclei Immunological identification of centromeric proteins by CREST antibodies was applied to determine the rate of micronuclei produced by lagging chromosomes. As expected, the two spin- dle poisons COL and VBL elevated the frequen- cies of CREST-positive micronuclei to 66-70% as compared with the control level of 42%. Russo et al. (1992) found 62% CREST-positive mi- cronuclei after CH treatment. Thus, micronuclei induced by COL, CH and VBL arose predomi- antly by lagging chromosomes. Only 15% of the HQ induced micronuclei were CREST-positive (Miller and Adler, 1990; Miller et al., 1991) indicating that at the doses tested the clastogenic effects of HQ dominated the spindle effects in the micronucleus assay. In situ hybridization with the major gamma satellite DNA probe resulted in almost identical frequencies of micronuclei with fluorescent sig- nals produced by COL (67%), VBL (66%) and HQ (15%) as the CREST labeling (Miller et al., 1991). Thus, both ways of identifying micronuclei formed by complete chromosomes have similar powers of detection. These labeling techniques 135 applied to micronuclei are able to identify chemi- cals which may cause aneuploidy due to lagging chromosomes; however, they are not suited to detect aneugens that also have a pronounced clastogenic activity, such as HQ. Spermatocytes (1) Meiotic delay Meiotic delay was determined by counting the frequencies of first and second meiotic divisions in slide areas where 1000 mid-pachytene nuclei per animal were found. Seven of the chemicals induced meiotic delay in mouse spermatocytes, COL, EZ, CH, HQ, DZ, CD and VBL (Miller and Adler, 1992). (2) Hyperploidy The seven chemicals which caused meiotic de- lay also increased the frequencies of hyperploid second meiotic divisions in the same experiments. The remaining three chemicals showed neither effect (Miller and Adler, 1992). The second labo- ratory obtained increases in hyperploid second meiotic divisions with COL, HQ and TB but not with CH, DZ and CD (Leopardi et al., 1993). Again, the differences in results from the two laboratories are most likely to be due to different test protocols. One of the major variations be- tween the experimental protocols concerned the arrest of meiotic metaphases by colchicine treat- ment 2.5 h prior to sacrifice (Leopardi et al., 1993). This is unnessesary and not required when chemicals are being tested for aneuploidy induc- tion. Another difference in test protocols con- cerned sampling times which were 6 and 8 h for all chemicals (except for HQ which was also sampled at 18 h) in one laboratory (Leopardi et al., 1993), and 6, 14 and 22 h in the other (Miller and Adler, 1992). The time response data for COL, CH and VBL showed that the effect in- creased significantly between 6 and 14 h (Miller and Adler, 1992). Both groups chose the sampling times arbitrarily. A suitable but flexible test pro- tocol can be based on optimizing the sampling times for chromosome counting according to the time course of meiotic delay, i.e., during the onset or the recovery period. Discussion One of the main difficulties in testing for ane- uploidy induction arises through the many targets that can be involved in the malsegregation of chromosomes during cell division. Effects on tubulin, tubulin associated proteins and micro- tubule condensation as well as effects on the centrioles imply that there may be a very narrow dose range over which aneuploidy can occur. Low doses may not impair spindle function at all while high doses may block the spindle formation com- pletely so that ceils are permanently arrested. Therefore, it is not surprising that dose-response curves show a plateau or even a decrease of the effect at higher doses. This may not be the case for chemicals that act primarily on the cen- tromeres or centromere associated proteins since these represent chromosomal structures and may affect individual chromosomes rather than the overall process of chromosome distribution as such. These chemicals may primarily result in lagging chromosomes which can be detected in the micronucleus assay. With these mechanistic considerations it is mandatory to test several doses and sampling times. The two positive control chemicals which are known spindle inhibitors and have a general ap- plication in cytogenetic techniques due to their metaphase-arresting capacity were readily identi- fied in the micronucleus assay. With additional CREST staining or in situ hybridization the pres- ence of complete chromosomes in the majority of micronuclei induced by COL and VCR could be demonstrated. Whether or not the remaining fraction of micronuclei is due to chromosome breakage remains to be determined. The clasto- genic effect of COL and VBL is weak if at all detectable by routine cytogenetic assays. There- fore, it could be assumed that the labeling meth- ods applied so far are not detecting all chromo- some bearing micronuclei. The cell shape and the cell membrane of erythrocytes pose considerable problems for the adhesion of the cells on slide surfaces and for the penetration of antibodies and DNA probes into the erythrocytes. Experi- ments are under way in various laboratories to flow-sort micronucleated erythrocytes and to use 136 other centromere specific DNA probes such as the mouse minor satellite DNA. The common phenomenon that chemicals with aneugenic properties also alter the progession of cell division has been confirmed in both mitotic and meiotic cells. One laboratory used the deter- mination of the average generation time (AGT) for bone marrow cells. However, the AGT is determined at 17 h after treatment. A change in AGT can also be caused by clastogens without aneugenic properties and may be due to length- ening of G 2 rather than to an effect on the progression of mitosis. It seems more feasible to determine the colchicine-like effect (increases in mitotic index and c-mitotic metaphases) in bone marrow cells at relatively short intervals after treatment, i.e., within a few hours. In spermato- cytes there was a complete correlation between meiotic delay and hyperploidy induction. There- fore, it is suggested that the assessment of mei- otic delay may be used to prescreen chemicals for possible aneugenic effects in male germ cells. More questions arose from the results of the first set of experiments than could be answered in the time period of the project. Table 2 shows a synopsis of the results obtained in the different laboratories for each of the 10 chemicals as dis- cussed in the previous section. Uniformly positive results were obtained with the two spindle poi- sons and the results with HQ are also in good agreement between laboratories and test systems. Problems have to be solved in the bone marrow tests with the other seven chemicals since there were interlaboratory disagreements. EZ seems to elevate the mitotic index, cause hyperploidy in bone marrow cells (at least in one laboratory) and also induces meiotic delay and hyperploidy in spermatocytes. The same is true for CH, although the micronucleus data show only weak responses using certain statistical considerations. Differ- ences in response between somatic and germinal cells were most obvious for DZ, TB, CD, TM and PY which can only be solved by further testing with improved experimental protocols. A new CEC Aneuploidy Programme com- menced in late 1991. It focuses primarily on the most prominent issues that arose from the previ- ous program. The role of cellular and organ spe- cific metabolism in the interaction of environ- mental chemicals with cellular targets of rele- vance will be studied. Labeled chemicals will be employed to determine their distribution in ro- dents. The mechanisms of action and relative activities of aneugenic chemicals in male and female germ ceils and somatic cells will be ana- lyzed in rodents. Research will be undertaken to determine the potential value of transgenic ani- mals in the detection of induced aneuploidy in specific rodent tissues. Application of in situ hy- bridization will be extended to a variety of DNA probes that allow chromosome painting and thereby visualization of chromosomes in inter- phase nuclei of various tissues. The results discussed demonstrate that experi- ments for standardization of aneuploidy tests and for an adequate statistical design are urgently needed to derive reliable protocols particularly for the micronucleus assay and chromosome counting. As a preliminary recommendation a simplified in vivo screening procedure can be envisaged consisting of two steps: (1) analysis of mitotic arrest and c-mitoses in rodent bone mar- row cells, and (2) determination of meiotic delay in rodent spermatocytes. Chemicals that show positive effects in either of these two tests should be subjected to aneuploidy testing by counting chromosomes in somatic and germinal cells, re- spectively. Acknowledgement The research was supported by the CEC Envi- ronmental Programme. References Adler, I.-D., U. Kliesch, P. van Hummelen and M. Kirsch- Voiders (1991) Mouse micronucleus tests with known and suspect spindle poisons: results from two laboratories, Mutagenesis, 6, 47-53. Gustavino, B., B. Bassani and F. Pacchierotti (1991) Vinblas- tine-induced numerical chromosome changes and selec- tion processes in mouse bone marrow cells, Mutation Res., 248, 45-50. Leopardi, P., A. Zijno, B. Bassani and F. Pacchierotti (1993) In vivo studies on chemically induced aneuploidy in mouse somatic and germ cells, Mutation Res., 287, 119-130. Manca, A., B. Bassani, A. Russo and F. Pacchierotti (1990) Origin of aneuploidy in relation to disturbances of cell cycle progression. I. Effects of vinblastine on mouse bone marrow cells, Mutation Res., 229, 29-36. Margolin, B.H., and K.J. Risko (1988) The statistical analysis of in vivo genotoxicity data, in: J. Ashby, F.J. de Serres, M.D. Shelby, B.H. Margolin, M. Ishidate Jr. and G.C. Becking (Eds.), Evaluation of Short Term Tests for Car- cinogens, Vol. 1, Cambridge University Press, Cambridge, pp. 1.29-1.42. Marrazzini, A., C. Betti, F. Bernacchi, I. Barrai and R. Barale (1993) Mouse micronucleus and metaphase analysis with known and suspect spindle poisons, Mutation Res., in press. Miller, B.M., and I.-D. Adler (1989) Suspect spindle poisons: analysis of c-mitotic effects in mouse bone marrow cells, Mutagenesis, 4, 208-215. Miller, B.M., and I.-D. Adler (1990) Application of antikine- tochore antibody staining (CREST staining) to micronuclei in erythrocytes induced in vivo, Mutagenesis, 5, 411-415. Miller, B.M., and I.-D. Adler (1992) Aneuploidy induction in mouse spermatocytes, Mutagenesis, 7, 69-76. 137 Miller, B.M., H.F. Zitzelsberger, H.-U.G. Weier and I.-D. Adler (1991) Classification of micronuclei in murine ery- throcytes: immunofluorescent staining using CREST-anti- bodies compared to in situ hybridization with biotinylated gamma satellite DNA, Mutagenesis, 6, 297-302. Pacchierotti, F., B. Bassani, P. Leopardi and A. Zijno (1991) Origin of aneuploidy in relation to disturbances on cell cycle progression. II: Cytogenetic analysis of various pa- rameters in mouse bone marrow cells after colchicine and hydroquinone treatment, Mutagenesis, 6, 307-312. Parry, J.M. (1987) The validation of test systems for the detection of chemicals capable of inducing numerical chromosome aberrations, Mutagenesis, 2, 309. Russo, A., A. Stocco and F. Majone (1992) Identification of kinetochore containing (CREST + ) micronuclei in mouse bone marrow erythrocytes, Mutagenesis, 7, 195-197. Wang Xu, and I.-D. Adler (1990) Clastogenic effects of known and suspect spindle poisons studied by chromosome analy- sis in mouse bone marrow cells, Mutagenesis, 5, 371-374.