Teratogenesis,Carcinogenesis, and Mutagenesis 12:97-112 (1992)

Further In Vitro and In Vivo Mutagenicity Assays With Thiram and Ziram Fungicides: Bacterial Reversion Assays and Mouse Micronucleus Test R. Crebelli, A. Zijno, L. Conti, B. Crochi, P. Leopardi, F. Marcon, L. Renzi, and A. Carere lstituto Superiore di Sanita (R.C., A.Z., L.C., B.C., I? L., F: M., A. C.), and ENEA C.R.E. Casaccia (L.R.), Rome, Italy The fungicides thiram and ziram have been assayed in a battery of nine bacterial strains of different genetic specificity. The results obtained suggest the induction of excisable DNA lesion(s), and indicate similar mutability of strains with AT or GC base pairs at target sites. This mutagenic profile is clearly distinct from that of oxidative mutagens, and it does not support the proposed role of oxidative stress in the mechanism of dithiocarbamates mutagenicity in bacteria. Furthermore, the bone marrow micronucleus test has been carried out in B6C3F1 mice with intraperitoneal administration of high grade thiram (12.5-50 mgkg) and ziram samples (2.5-10 mg/kg in males, and 5-20 mgkg in females). Thiram produced a significant increase of micronucleated PCEs in male mice sampled 48 h after treatment with 25, 37.5, and 50 mg/kg. No significant increase was detected in treated females. Ziram, tested in a lower range of doses because of its higher toxicity, resulted negative in both sexes. Both the acute toxicity and the ratio polychromatic/normochromatic erythrocytes indicated some sex specificity in the toxic effects induced by these dithiocarbamates in the B6C3F1 mouse. o 1992Wiley-Liss, Inc. Key words: dithiocarbamates, mutagenicity, short-term tests, bacteria, micronuclei

INTRODUCTION

Thiram (tetramethylthiuram disulfide, TMTD) and ziram (zinc dimethyldithiocarbamate) are dithiocarbamatederivativesextensively employed as contact fungicides for the protection of vegetable crops and seeds and for mold control in the manufacture of paper, plastics, and textiles. Both compounds are also widely used as vulcanization accelerators in the rubber industry [ 11. Address reprint requestsandcorrespondencetoDr.RiccardoCrebelli,IstitutoSuperiorediSanitB, Laboratory of ComparativeToxicology and Ecotoxicology, Wale Regina Elena, 299-0016 1 Rome, Italy.

01992Wiley-Liss,Inc.

98

Crebelli et al.

Dithiocarbamates are very reactive compounds, with strong metal-binding characteristics, and are capable of interacting with protein sulfhydryl groups. Toxic and teratogenic effects have been observed in experimental animals exposed to thiram and ziram, as well as some tumorigenic effects that were insufficient for an evaluation of their carcinogenicity [2]. Numerous mutagenicity short-term tests indicate that both compounds are genotoxic. Thiram was directly positive in bacterial reversion tests with both base substitution and, at a lower degree, frameshift sensitive strains of Salmonella typhimurium [3-71, in repair tests in bacteria [8-91, in forward mutation [9] and chromosome malsegregation assays in Aspergillusnidulans [lo], and in the sex-linked recessive lethals in Drosophila melanogaster [ 111. In mammalian cells in vitro, thiram produced mixed results: it was positive in assays for the induction of sister-chromatid exchanges (SCE) and unscheduled DNA synthesis on human lymphocytes in vitro [12-131, and negative for SCE induction in Chinese hamster ovary cells [ 111. It was positive (at a toxic dose) in one of two studies in Chinese hamster lung V79 cells (hprtlocus) [ 11,141. In vivo, administration of thiram (commercial formulations) at high, toxic doses induced micronuclei in mouse bone marrow [ 14, 151, increased the frequency of sex and autosomal univalents and polyploidy in mouse spermatocytes [ 161, and produced morphologically abnormal sperms in mice [ 16,171. Negative results were obtained in a limited micronucleus test in Chinese hamsters [ 111. Ziram was directly mutagenic in S. typhimurium (mainly in base substitution sensitive strains) [33,181 and in a repair test in Bacillus subtilis [8]. Samples of both technical ziram (as rubber chemical) and of a formulate containing ziram as the active ingredient were positive in the sex linked recessive lethals (SLRL) test and in somatic and germ-line mosaic assays in Drosophila [ 1 1,19,20]. In in vitro cultured mammalian cells, ziram (technical grade) induced high frequencies of chromosome aberrations (and no SCE) in Chinese hamster ovary cells both with and without S9 1211and-tested only without S9-forward mutations at the tk locus in mouse lymphoma cells [22]. It failed to induce unscheduled DNA synthesis (UDS) in rat hepatocytes in vitro [23] and it was inactive in a point mutation assay in Chinese hamster V79 cells [ 111. In vivo, administration of Cuman L (a formulate containing 27% ziram as the active ingredient) to Swiss mice at sublethal doses resulted in the induction of significant increases in sex and autosomal univalents and polyploidy in the germ cells and in the number of micronuclei in polychromatic erythrocytes 1241. Finally, significant increases in chromosomal aberrations were observed in peripheral blood lymphocytes of rubber industry workers occupationally exposed to TMTD and ziram [25,26]. In spite of the abundancy of studies which indicate a mutagenic potential for both thiram and ziram, the data available are regarded as inadequate for evaluating the genotoxic risks posed by these dithiocarbamate compounds [27]. In particular the lack of information on the mechanism of action and uncertainties concerning the effects observed in vivo prevent a definitive evaluation [28]. As far as the mechanism of action is concerned, difficulties in explaining the mutagenicity of dithiocarbamates in terms of electrophilicity [29] lead some authors to postulate an indirect mechanism, based on the inhibition of cellular defenses to oxidative damage [6]. In this study we have addressed this question by assaying both thiram and ziram in a wide battery of bacterial strains, comprising those which show specific sensitivity

Thiram and Ziram Mutagenicity Testing

99

to oxidative damages, in order to investigate the proposed role of oxidative stress in thiram and ziram genotoxicity. Furthermore, in view of the limitations inherent to the published investigations on the in vivo clastogenic activity of dithiocarbamates, mainly related to the use of low grade technical products and/or inadequate experimental design, we have carried out a micronucleus test in the mouse with high grade thiram and ziram samples in compliance with the EEC/OECD guidelines. The results obtained in this study confirm the direct mutagenicity of both compounds in bacteria, but do not support a role for oxidative damage in the in vitro mutagenicity of dithiocarbamates. A clastogenic effect of thiram to mouse bone marrow is revealed by the micronucleus test. MATERIALSAND METHODS Chemicals

Thiram (CAS 137-26-8) technical product (99.7%) and ziram (CAS 137-30-4) technical product (98.5%) were provided by UCB S.A., Brussels. Positive control substances were provided as follows: 9-aminoacridinehydrochloride (CAS n. 524 17-22-8)and 4-nitro-o-phenylendiamine (99-56-9) were from Aldrich Chimica S.r.l., Milan; 2-aminoanthracene (613-13-8) was from Fluka AG, Buchs; sodium azide (26628-22-8), mitomycin C (50-07-7), 4-nitroquinoline N-oxide (56-57-5), cumene hydroperoxide (80-15-9), and danthron (1,8-dihydroxyanthraquinone,117-10-2) were from Sigma Chemical Co., St. Louis, MO; ethyl methanesulfonate (62-50-0) was from Kodak, Rochester, NY; hydroquinone (123-31-9) was from C. Erba, Milan, Italy. Bacterial Mutagenicity Assays

A battery of nine bacterial strains of different genetic specificity [30-321 was used. Salmonella typhimurium strains TA15355, TA1537, TA98, TA100, TA102, TA1950, and TA 1975were kindly supplied by Dr. B .N. Ames, University of California, Berkeley, CA. Escherichia coli strains WP2 (pKM101) and WP2 uvrA @KMlOl) were a gift of Dr. P. Wilcox, GlaxoGroupResearch Ltd., Ware, Hertfordshire, UK. Plate incorporation assays were carried out following the method of Maron and Ames [31]. For the E . coli strains, the top agar was supplemented with 0.0176 mM L-tryptophan [32]. Exogenous metabolic activation was provided by liver postmitochondrialfractions (S9)obtainedfromAroclor 1254 (Monsanto, St. Louis, MO) induced Sprague-Dawley rats. The S9 level routinely used was 50 pl/plate, corresponding to 2 mg of proteidplate. Tested compounds were dissolved in dimethylsulfoxide (DMSO, C. Erba), and assayed in a range of doses, up to a toxic or inhibitory concentration. All determinations were made in triplicate plates. Mutagenic activities (expressed as the number of induced revertants per mg) were calculated by linearregression analysis using the experimental data set with best r value. When no reproducible, dose-related increase of revertant colonies was observed on treated plates, the result was evaluated as negative. Positive controls were included in all the experiments in order to confirm strain sensitivity and S9 activity, as well as routine controls of samples and S9 sterility. Mouse Bone Marrow MicronucleusTest

The in vivo mutagenicity of thiram and ziram was evaluated in the mouse bone marrow micronucleus test with B6C3Fl male and female mice provided by Charles

100

Crebellietal.

River Italia (Calco, Como). The animals, weighing 20-26 g, were maintained on a balanced chow with tap water ad libitum and were acclimatized for 1 wk under standard conditions before treatment. Five males and five females were used for each dose and for solvent and positive controls. Treatments were by single intraperitoneal(i.p.) injection. Fresh solutionsof thiram and ziram in spectrophotometricgrade DMSO were prepared for each experiment. Controls received DMSO alone. The final volume administered was 5 ml/kg body weight. Preliminary dose range finding experiments were carried out to establish the dose levels to be assayed in the main test. In the absence of signs of overt toxicity, the lethal effect of the administered chemicals was used to define the high dose. After thiram administration at 100 mg/kg all treated mice (four of four) died i n 24 h, whereas with 50 mg/kg three of five survived at 48 h. Ziram killed all treated mice (five of five) at 50 mglkg, and produced some lethality in males (three of five treated mice died at 48 h) but not in females (all five treated animals survived at 48 h) at 20 mg/kg. Consequently, according to the recommended use of >50% of the LD5o as the high dose level in the micronucleustest [33,34],thiramwasadministeredat 12.5,25.0,and50.0mg/kg to both sexes and ziram was given at 2.5,5.0, and 10.0mg/kg to males and 5.0, 10.0, and 20.0 mg/kg to females. As positive controls, ten males and ten females were injected with 80 mg/kg hydroquinone (HQ) dissolved in distilled water. Two sampling times, at 24 and 48 h, were used for solvent and for treated groups. Positive controls were sampled 24 h after treatment. Animals were sacrificed by cervical dislocation. Both femours of each mouse were flushed out in fetal calf serum (FCS) and centrifugated at 1,000rpm (140g) for 5 min. The pellet was resuspended in few drops of FCS and smeared on cleaned slides. The day after slides were fixed 5 min in absolute methanol and stained 20 min with 5% Giemsa in 0.01 M Sorensen buffer (pH 6.8). Two-thousand polychromatic erythrocytes (PCE) per mouse were scored for the presence of micronuclei on blind slides. For each animal, the number of normochromatic erythrocytes (NCE) in the microscope fields that showed 2,000 PCEs was determined in order to calculate the PCE percentage on total erythrocytes (PCEs + NCEs) as a measure of bone marrow toxicity. Statistical analysis was performed according to Love11 et al. [35]. Each dose group was tested by chi-square for evidence of heterogeneity between animals. Homogeneous groups were pooled over individual animals and compared (treated vs. control) by 2 x 2 chi-square test including the Yates’ correction factor. Heterogeneous groups were compared using a variance ratio value calculated from the between- and the within-group chi-square value. Furthermore analysis of variance, after the square root transformation d(x+l)of the individual number of micronuclei, was applied to examine the effects of sex, time of sample, and treatment. Student’s t test was used to compare the reduction of PCE percentage in treated groups against paired controls. RESULTS BacterialAssays

The results of bacterial reversion assays within thiram and ziram are summarized in Table I, where the specific mutagenic activities (expressed as the number of induced revertants per mg) in each tester strain are shown. Raw data used to calculate regression values are shown in Table 11.

+

190 410

120" 555

TA1535

-

0 0 0 0

+

TA1537 0 0

-

TA98

0 0

+

+

TAlOO

2,390 1,700 5,570 3,630

0 0

-

+

0 0

TA 102 0

ND ND

+

TA1975

ND

250 ND

ND ND

+

TA1950 -

"Induced revertants/mg. Mutagenic activities were calculated by linear regression analysis from raw data shown in Table 11. 0, not mutagenic (see Methods). ND, not determined.

Thiram Ziram

Strain: s9:

TABLE I. Mutagenic Activity of Thiram and Ziram in S. typhimurium and E . coli Strains

0 0

-

WP2 0 0

+

+ 3,240 3,030 4,000 4,140

-

WP2 uvrA

1st exp.

Positive control

Thiram 2nd exp.

Positive control

Ziram

Positive controlb

Thirarn 1st exp.

30 60 120

15

0

0 7.5 15 30 60

0 60 120 180 240

Dose (Fgiplate)

7 2 3 2222 2024 24f2 32f4 774 f 29

9 2 1 13 ? 2 21 2 1 3423 43 2 6 893 2 106

921" 2923 30 2 7 2724 tox 893 2 106

TA1.535

8?1 15t6 14t5 13t1 l o t 1 911 t 54

17 t 2 nt 15 t 3 2323 17 t 6 386 ? 46

4629 26f6 30 f 4 25f8 17 t 4 180 t I 1

TA1537

TAlOO

TA102

37 36 45 44 42 520 3 5 5 6 t6 2 59 2 2 2

?

33 t 5 nt 41 2 7 46 f 8 51 ? 6 430 f 65

20 ? 4 23 t 3 30 t 9 15t3 14 2 5 404 2 36

17 52 59 114

15 7

153 ? 15 247 2 27 266 t 22 349 t 27 323 t 27 1773 ? 114

153 2 221 2 223 2 337 2 408 2 1773 ?

337 370 343 335 281 713

69 30 25 35 t 30 f 73 2 2 2

?

249 f 87 303 f 39 249 t 17 187 i- 20 261 t 23 >2000

9 9 ? 14 235 2 12 231 ? 11 188 t 26 268 2 11 187 2 4 308264 18525 tox 135 2 5 1316 2 196 2 1 0 0 2 288

Without Metabolic Activation

TA98

TABLE 11. Results of Bacterial Reversion Assays With Thiram and Ziram Fungicides

15 2 4 12 2 2 14 2 4 13 t 3 19 t 3 417 f 59

nt

nt

TA1975

10 2 2 2125 26 ? 5 28 f 5 37 f 3 1120 f 168

nt

nt

TA1950

5 7 6 6 t7 t 33 2 It 2 2

34? 3 4323 34 t 1 1 37 t 6 30 ? 2 141 ? 29

67 71 51 44 50 313

nt

WP2

i- i9

t4 t 19 t 20 t 22 f 106

1 6 6 ? 21 210219 247 2 46 354 2 19 tox >2500

157 201 221 267 408 701

nt

WP2uvrA

Positive control

Thiram 2nd exp.

Positive control

Ziram I st exp.

Positive control

Thiram 1st exp.

Positive control

Ziram 2nd exp.

0 15 30 60 120

30 60 90

15

0 7.5

0 60 120 180 240

0 7.5 15 30 60

14 2 29 k 36 k 36 % 49 2 93 2 8 6 6 13 6

1

6 2 1 nt 25 +- 7 24 2 6 4 0 t 11 31 t 3 104 2 12

10 5 1 25 f 5 41 2 7 24 t 2 29 f 3 98 2 6

7 k 3 13 2 2 22 t 1 33 f 5 33 2 2 714 t 29

6f1 9 2 2 11 t 2 8t2 11 f 6 1 0 6 2 18

10 t 2 nt 18 f 4 12 f 0 16 2 2 13f3 185 f 16

372 9 34 2 1 36 f 7 27 2 3 38 f 17 161 f 13

15 k 4 10 t 2 11f3 911 k 54

8f1 11 t 4

16 f 3 2320 29t 3 26 t 10 25 f 7 857 t 67

39 f 3 40 f 5 54 t 4 4728 5 1 4 f 142

nt

*6

17 f 1 28 t 6 31 f 5 21 2 2 19 t 1 472 f 53

28

1404 k 220

nt

133 f 9 195 f 6 281 5 27 299 f 28

337 418 347 437 319 713

f 20

nt

332 & 64 124 k 22 363 2 64 nt 295 2 46 327 f 58 363 t 45 331 +- 13 351 & 4 2 373 f 87 1 3 4 2 k 162 1956 -t 84

*

289 t 10 2 3 4 2 16 200 +- 5 >I600

267

26 t 60 f 84 f 61 2 73

f 69 2

137 t 13 332 t 64 249 k 20 364 f 14 353 f 5 281 Tt 20 358 f 14 393 2 33 377 f 15 335 2 23 nt 288 f 44 656 -t- 130 1956 84

101 t 14 315 -t 75 345 f 35 277 -t 32 tox 866 Tt 60

With Metabolic Activation

37 f 3 33 f 2 36 f 4 43 f 3 4220 520 t 59

nt

nt

nt

nt

nt

nt

nt

nt

1 6 0 2 32 291 t 12 319 k 12 3 6 0 k 16 nt 1530 k 160

nt

nt

nt

(Table continued on next page.)

49 % 2 45 f 1 4 6 t 10 3724 52 f 3 700 f 20

nt

nt

nt

0 3.75 7.5 15 30 60

14 2 1 nt 22 2 3 22 f 3 35 f 2 42 f 3 83 f 6

62 1 nt 7r 1 7 2 3 621 6 2 2 108 2 18

16 f 3 nt 23 f 3 29+2 2 8 ~ 3 24+2 857 + 67

128 2 10 179 2 12 324 _t 53 303f28 nt nt 958 f 86 f

nt

11

339 2 34 414211 317 + 12 321 rfr 54 >2000

397

nt

nt

+ 10 nt 35f 8 37 f 1 48 2 9 24 f 9 347 2 30 50

185 2 29 nt 231 t 13 239 5 30 3 1 5 f 75 439 f 71 1250 ? 230

aRevertantcolonies per plate (mean and standard deviation from counts of three plates). bWithout S9: TA1535 and WP2uvrA. 4-nitroquinoline-N-oxide, 0.5 pg; TA1537, 9-aminoacridine, 50 kg: TA98, 4-nitro-o-phenylendiamine,5 pg ; TA100, sodium azide, 5 pg; TA102, mitomycin C , 0.5 pg (1st exp.) and cumene hydroperoxie 0.1 p1 (2nd exp.); TA1975, ethylmethane sulfonate, 2 pl;TA1950, sodium axide, 10 pg; WP2, mitomycin C, 0.5 pg. With S9, 2-aminoanthracene, 1 pg (TA1535, TA1537, TA98, TA100) and 50 pg (WP2 and WP2 uvrA); danthron, 100 pg (TA102). nt, not tested; tox, toxic.

Positive control

Ziram 2nd exp.

TABLE 11. Results of Bacterial Reversion Assays With Thiram and Ziram Fungicides Continued.

Thiram and Ziram Mutagenicity Testing

105

Thiram and ziram displayed a similar mutagenic profile in bacteria, both reverting, both in the presence and in the absence of exogenous activation, base substitution sensitive strains that are deficient in the DNA excision repair system, i.e. S . typhimurium TA1535 (hisG46 rfa uvrB) and TAlOO (hisG46 fla uvrB pKMIOI) and E . coli WP2 uvrA (pKMIOI). Experiments performed with thiram in S. typhimurium strains TA1950 (hisG46, uvrB) and TA1975 (hisG46, rfu) confirmed the requirement for an uvr genetic background to elicit a mutagenic response. Comparison of the results in strains TAlOO and WP2 uvrA, both of which are exicision repair defective and harbour the pKMlOl plasmid, did not indicate a preferential mutagenic activity of dithiocarbamates vs. GC (as at the target site in TA100) or AT base pairs (as in the ochre mutation of the WP2 derivatives). Rather, comparison of the results in WP2 uvrA (pkMIOI) with the excision proficient strains WP2 (pkMlOI)andTA102 (hisG428rfapKMlOI)--allofwhich carry an ochre mutation at the reversion site-suggests the induction of premutagenic lesion(s) which can be efficiently handled by the excision DNA repair system. No significant mutagenicity was detected in the frameshift sensitive strains TA1537 and TA98. Mouse MicronucleusTest

Thiram. In the main test (Table 111) four males treated with the highest dose (one to be sacrificed at 24 h and three at 48 h) died. Slight increases in the frequency of micronucleated PCEs (MnPCEs) were observed in males treated with 25 and 50 mg/kg, and in females treated at the highest dose. This effect attained statistical significance only in males sacrificed 48 h after treatment (P< 0.05 andP