Reappraisal of eight representative carcinogenic and non-carcinogenic compounds in a new medium-term rat liver bioassay using d -galactosamine

Cancer Letters 104 ( 1996)X5-90

CAMCER LETTERS

Reappraisal of eight representative carcinogenic and non-carcinogenic compounds in a new medium-term rat liver bioassay using D-galactosamine Hyoung-Chin Kim’ -+,*.I,Shin-Woo Chaa, Chang-Su Haa, Jung-Koo Roha, Yong-Soon Leeb, Fumio Furukawac, Akiyoshi Nishikawac, Michihito Takahashic “Toxicology Reseurch Center; Korea Reseurch Institute of Chemical Technology, 100 Jungdong, Yusongku, Duejon 305-343, South Korea ‘Cbllege of Veterinary Medicine, Seoul Nationul University, 108 Sodundong, Kwonsunku, Suwon 441-744, South Korea ‘Divi.Gon of Path&g?; National Institute of Health Sciences, I -18-l Kamiyogu, Setuguyu-ku, Tokyo 1%. Japan

Received 14 February 1996;revision received 14 March 1996; accepted 14 March 1996

Abstract

The carcinogenic potential of eight different compounds was assayed in a new medium-term carcinogenicity bioassay using D-gaiactosamine (DGA) as a non-surgical method to induce regeneration in place of partial hepatectomy (PH). Male rats were initially given a single i.p. injection (200 mg/kg) of diethylnitrosamine (DEN) and after 2 weeks on basal diet, received two i.p. injections of DGA (300 mg/kg) at the end of weeks 2 and 5. They were treated with one of the test compounds aflatoxin HI. hcnzo[a]pyrene, diethylstilbestrol, urethane, sodium saccharin, bucetin, D-mannitol and sodium chloride in the diet or basal diet alone for weeks 3-8. Carcinogenic potential was assessed by comparing the numbers and areas per cm2 of glutathione S-transferase placental form-positive (GST-P+) foci in the livers of treated animals with those of the control animals given DEN/DGA alone. Positive estimations of carcinogenicity were obtained for the hepatocarcinogens aflatoxin B,, diethylstilbestrol or urethane, and for the non-liver carcinogen benzo[a]pyrene. Negative values were shown in rats treated with the non-carcinogens, D-mannitol and sodium chloride. The two other non-liver carcinogens sodium saccharin and bucetin, also did not exert positive effects in the system. The present data are consistent with findings in previous medium-term bioassays using PH. Our results thus confirm that the present bioassay protocol with repeated administration of DGA instead of PH may offer a new and sensitive non-invasive method to screen large numbers of environmental carcinogens. Keywords:

u-Galactosamine; Medium-term bioassay; Glutathione S-transferase placental form positive foci: Rats

_---

1. Introduction ’ I’orrcsponding author. Tel +X2 47 8607454; fax: +X2 4:! 8607488. ’ Vresentaddress:Toxicology ResearchCenter, Korea Research insmute of Chemical Technology. 100 Jangdong, Klsongku. I)ac~on KF343. South Korea

Avoiding human exposure to chemical carcinogens is a practical approach to cancer prevention, for most neoplasia in man is considered to be caused by environmental carcinogenic agents 191. However, the

ir?O4-3X3S/Yh/~12.00 C 19Y6Elsevler Science Ireland Ltd. All rights reserved 1’11 S!f704-~X~S~96)04~~6-X

86

H.-C. Kim et al. /Cuncer

large number of compounds that have been introduced into our environment present a problem for routine long-term carcinogenicity tests because of the costs involved. Various in vitro genotoxicity tests have therefore been introduced for the purpose of mass screening of compounds for mutagenicity [1.5, 171 but with accumulating results, it has become clear that carcinogenic potential does not always equate with mutagenicity. Thus prediction of longterm carcinogenicity in rodents suffers from high numbers of false positives and false negatives [36]. Therefore, in order to fill the gap between mutagenicity tests and conventional long-term tests, several in vivo medium-term assay systems for liver carcinogens and tumor promoters have been developed, based on the two-step concept of carcinogenesis [24,26.27]. Of the medium-term assay systems, the so-called diethylnitrosamine-partial hepatectomy (DEN-PH) model system is now considered to be the most practical for rapid and economic screening of the carcinogenic or anti-carcinogenic potential of environmental chemicals [14,29,34]. In the DEN-PH model system, rats are treated with diethylnitrosamine (DEN) for the initiation step followed by administration of various test compounds combined with two-thirds partial hepatectomy (PH), and the preneoplastic liver lesions are scored in terms of the numbers and areas of glutathione S-transferase placental form positive (GST-P+) foci [24]. However, the use of PH requires skillful surgery under anesthesia, which is time- and labor-consuming, and occasionally causes fatalities to experimental animals due to technical difficulties [ 191. Therefore, we designed a non-surgical bioassay protocol using repeated administration of D-galactosamine (DGA) instead of PH, and showed that the number and area of GST-P+ liver cell foci per cm* in rats initiated by DEN and promoted by 2-acetylaminofluorene were larger with our new protocol than those using PH, the difference in the numbers being statistically significant [ 191. In the present study, we studied the validity of our non-surgical protocol for assay of carcinogenic potential by testing a series of known carcinogens and non-carcinogens, namely aflatoxin B,, benzo[a]pyrene, diethylstilbestrol, urethane, sodium saccharin, bucetin, D-mannitol and sodium chloride.

Letters

104 (1996)

85-90

2. Materials and methods 2.1. Animals and chemicals A total of 90 male F344 rats, 4 weeks old, were obtained from the Department of Laboratory Animal Breeding, Korea Research Institute of Chemical Technology, Daejon, Korea. They were acclimatized for 2 weeks prior to the experimentation under standard conditions (temperature, 23 + 3°C; relative humidity, 50 + 10%; illumination cycle, 12 h/12 h light/dark), were housed in stainless-steel wire cages, two rats per cage. Diet (Jeil Feed Co., Ltd., Daejon, Korea) and water were available ad libitum throughout the experiment. DEN and DGA were purchased from Sigma Chemical Co. (St. Louis, MO, USA). Rabbit anti-GST-P was purchased from Medical & Biological Laboratories (Nagoya, Japan) and a Vectastain ABC Elite kit from Vector Laboratories Inc. (Burlingame, USA). Aflatoxin B,, benzo[a]pyrene, diethylstilbestrol, urethane, sodium saccharin, bucetin, D-mannitol and sodium chloride were commercially available samples from Sigma Chemical Co. (St. Louis, MO, USA). These chemicals were incorporated into powdered basal diet using a mixer without adding oil and the prepared diets were stored in a cold room until use. 2.2. Experimental protocol (Fig. 1) The animals were divided into 9 groups, each consisting of 10 rats, and given an i.p. injection of DEN (200 mg/kg) dissolved in 0.9% NaCl solution for the initiation of hepatocarcinogenesis. After

Fig, 1. Experimental protocol of the present medium-term bioassay system.

H.-C. Kim PI (11./Cunwr

Letters

2 weeks on basal diet, animals were twice given i.p. in,jections of DGA (300 mg/kg) dissolved in 0.9% NaCl, at the end of weeks 2 and 5 to stimulate liver cell replication, and fed one of the test chemicals mixed with the basal diet or basal diet alone as the control filr weeks 3-8. The doses of the test chemicals were 2 ppm atlatoxin B,, 200 ppm benzo[a]pyrene, 5 ppm diethylstilbestrol, 3000 ppm urethane, 50 000 ppm sodium saccharin, 15 000 ppm bucetin, 50 000 ppm D-mannitol and 50 000 ppm for sodium chloride. The doses of DEN and DGA were based on the previous studies [ 18,30,32,33]. The doses of test compounds were selected according to the previous data f 181 or the maximum feasible dose level 50 000 ppm tn rats. All animals were killed at the end of week 8, their livers immediately excised and slices 3 mm thick cut from the left lateral lobe with a razor blade. The slices were fixed in 10% phosphate-buffered formalin solution for routine embedding in paraffin and immunohistochemical identification of GST-P excomplex pression. The avidin-biotin-peroxidase (ABC) method described by Hsu et al. [ 161 was used to demonstrate GST-P+ liver foci. After deparaffinization. liver sections were treated sequentially with normal goat serum, rabbit anti-GST-P (1:300), biotin-labeled goat anti-rabbit IgG (1:lOO) and ABC. The peroxidase binding sites were visualized by the diaminobenzidine method. Sections were then counter-stained with hematoxylin. The numbers and areas of GST-P+ foci larger than 0.1 mm in diameter and the total areas of liver sections examined were

104 (1996)

85-90

87

measured using an image analyzer LA-555 (PIAS Co., Japan). Statistical significance was determined using Student’s t-test.

3. Results 3.1. Body and liver weights

Data for final body and liver weights are summarized in Table 1. Body weights were significantly decreased in rats treated with aflatoxin Bi, diethylstilbestrol, urethane, sodium saccharin and bucetin as compared to the controi value (P < 0.05 or 0.01). Liver weights were significantly decreased in rats treated with diethylstilbestrol and sodium saccharin (P < 0.01). Liver/body weight ratios were significantly increased in rats treated with aflatoxin Bi, benzo[a]pyrene, diethylstilbestrol and urethane (P < 0.05 or 0.01). 3.2. Analysis of GST-P+ liver cell,foci

Data for numbers (no./cm*) and areas (mm2/cm2) of GST-P+ foci per unit area of liver section are summarized in Table 2. The mean numbers of GSTP+ foci in the livers of rats receiving aflatoxin B,, benzo[a]pyrene, diethylstilbestrol and urethane were 179.5, 19.6, 9.8 and 36.1/cm2, respectively, which were all significantly (P < 0.01 or 0.001) increased compared to the control value (3.0/cm2). The mean areas of GST-P+ foci in rat livers given aflatoxin Bi,

T&k 1 Body and hv~r weight data Final body

A&toxin H t Benzo[a]pyrene Diethylstilbestrol ?JrethCllle Sodium sacchm

weight(g)

Absolute liver weight(g)

Liver/body weight ratio (%)

12.4 + 2.4 12.5 + 0.8 8.5 f 0.6** 13.3 kO.8 9.1 f. 1.4** 1 1.o f I .o 1 I .6 k I .6 10.7 ix 0.5 Il.82 1.0

5.0 4.4 4.9 4.9 3.7 4.2 4.1 3.9 4.0

u-Mannito! Sodium chloride Control

250.1 t 15.4** 282.5 c 13.7 174.4 + 14.8** 271.6 + 17.0* 259.0 * 16.7** 260.9 + 17.6** 282.6 + 22.3 275.6 zk 15.2 295.6 + 16.3

Each v.lfuc represents a mean f SD. Significantly

different from the control group (*P < 0.05, **P < 0.01).

Kucctinr

k 0.8** 2 0.2* + 0.3** k 0.2** + 0.5 f 0.2 f 0.3 k 0.1 * 0. I

88

H.-C. Kim et (11./Cancer

Table 2 Numbers and areas of GST-P positive liver cell foci Compound

Aflatoxin Bt Benzo[a]pyrene Diethylstilbtstrol Urethane Sodium saccharin Bucetine D-Mannitol Sodium chloride Control

No. of rats

GST-P positive foci No./cm2

Area(mm*/cm*)

10 10 10 10 7 10 10 10 10

179.46 + 15.17*** 19.55 + 2.91*** 9.81 + 1.80** 36.07 f 6.1 I*** 0.43 + 0.43 0.33 e 0.33 3.10 + 0.92 7.42 k 3.00 3.00 + 0.78

32.92 + 7.67** 0.65 + 0.16** 0.26 e 0.07* 1.89 + 0.62* 0.02 * 0.02 0.01 + 0.01 0.07 f 0.02 0.24+0.12 0.06 + 0.02

Each value represents a mean + SE. *Significantly different from the corresponding control value (*P < 0.05, **P < 0.01, ***P < 0.001).

benzo[a]pyrene, diethylstilbestrol and urethane were 32.92, 0.65, 0.26 and 1.89 mm2/cm2, respectively, which were also significantly (P < 0.05 or 0.01) increased compared to the control value (0.06 mm2/ cm2). Thus judging from the results for both parameters, aflatoxin B, was the most potent carcinogen, followed by urethane, benzo[a]pyrene and diethylstilbestrol under the present experimental conditions. Sodium chloride slightly, but not significantly, increased the numbers and areas of the foci. In contrast sodium saccharin and bucetin slightly, but not significantly, decreased the numbers and areas of induced foci. 4. Discussion In the present study using repeated treatment with DGA as a non-surgical approach for medium-term carcinogenicity assay, positive results regarding numbers and areas of GST-P+ liver cell foci were obtained in rats treated with the liver carcinogens aflatoxin B,, diethylstilbestrol and urethane, and with the non-liver carcinogen benzo[a]pyrene. The main target of aflatoxin Bi is known to be the liver [23,35], while diethylstilbestrol causes mammary and liver tumors [ 131, and urethane induces lesions in the lung and liver [22]. Benzo[a]pyrene is not normally a hepatocarcinogen [3] but does cause mammary tumors 1211. The negative results found for the noncarcinogens D-mannitol and sodium chloride were

Letters

104 (1996) X5-90

unequivocal in the present assay system. Therefore, evaluation of these six chemicals with our new protocol gave good agreement with the outcome of previous long-term carcinogenicity studies. With regard to the two non-liver carcinogens, sodium saccharin and bucetin, the negative results highlight difficulties in drawing conclusions without an understanding of mechanisms. Sodium saccharin is known to cause urinary bladder tumors [ 1,111, but is not metabolically activated to a reactive electrophile, does not react with DNA, and is not mutagenic in multiple short-term screening systems [ 1I]. Although a multitude of variables which increase or decrease the effects of sodium saccharin on the rat bladder have been identified, the formation of silicate precipitates and/or microcrystals in the urine appears to play a role in the increased urothelial proliferation and raised incidence of bladder cancer in rats [6,11,25]. It may be a rat specific phenomenon, not occurring in mice, hamsters or monkeys [8] and it is also unlikely that man demonstrates either a bladder proliferative [2] or neoplastic [lO,l l] response to sodium saccharin. In fact, the generation of precipitates and microcrystals has been regarded as a significant obstacle in interpreting the carcinogenicity of various sodium salts such as sodium saccharin, ascorbate, glutamate, bicarbonate and chloride [5]. The other ‘false negative’ in our system, bucetin has been reported to cause kidney and urinary bladder tumors in mice [31] but since there is no report of any carcinogenicity of this compound in rats, it is difficult to determine whether the present negative result is indeed false. Direct or indirect possible interactions between DGA and test chemicals might be considered in some cases. Independent of these questions, as shown in Table 3, the carcinogenic potentials of the eight chemicals estimated in this study are completely consistent with earlier data obtained using the DEN-PH model by Ito et al. [ 181. The control value for areas of GST-P+ foci in the present study was 0.06 mm* per cm*, from examination of the left lateral lobe. In previous experiments, quantitative values for GST-P+ foci in the control groups given DEN followed by basal diet were 0.181.23 mm* per cm2 for the DEN-PH model [ 181 and 0.98 mm* per cm2 with our non-surgical model using repeated treatment with DGA [ 191, from examination of the right lateral and caudate lobes. The discrep-

89

Table 3 Comparison of results from the DEN-DGA model, the DEN-PH model and long-term carcinogenicity i’ompound

Aflatoxin H1 Henzo[ri]pyrcnr Diethylstilbestrol (!rethanc Sodium saccharin i3ucetine !>-Mannitol Sodium chloride

DEN-DGA model” dose (ppm) 2 200 5 3000 5oooo I5000 50000 soooa

DEN-PH modelb resultd

: f ? + + -, -+

dose (ppm) 2 200 5 2000 SO000 I5000 50000 I ooooo

testing

CC resultd f A * t + t -+ -+

+ t + t t t

.* Dicthylnitrosamine-D-galactosamine model. 5 Diethylnitrosamine-partial hepatectomy model. Data from Ito et al.. [2X]. ’ Results of carcinogenicity testing; +, positive; -, negative. il Significant increased (t) or no change (+) in GST-P+ liver cell foci development at P < 0.01 or P < 0.05.

ancy in data for areas of GST-P+ foci in control groups might largely be due to interlobe differences [20.28]. In the present study, the four chemicals that increased the liver/body weight ratios all gave rise to increased numbers and areas of GST-P+ foci. Ele\*ated relative liver/body weights could reflect enhanced hepatocyte proliferation and it is well documented that cell proliferation plays a key role during cancer induction by genotoxic or non-genotoxic chemicals in many organ systems, including the liver !4.?,121. in conclusion, it is proposed that the present nonsurgical bioassay protocol using repeated administration of DGA has advantages as a medium-term screening method for environmental chemicals thar may exert hepatocarcinogenicity. Moreover, we conclude that our system may find application in assessing non-liver carcinogens. Confirmatory studies are now under way. .4cknowledgements The authors are grateful for the assistance of Mr. Ju-Hyun Hae. This work was financed by the Korea Research Institute of Chemical Technology, Daejon. tiorea. References II/

Am~ltl. 111,. Moodie. C‘.A.. &ice. H.C., Charbonneau, S hl Sinvric. B.. C‘ollins. 15.7‘. McGuire. P I:., Zawidzka,

Z.Z. and Munro, I.C. (1980) Long-term toxicity of orthotoluenesulfonamide and sodium saccharin in the rat. Toxicol. Appl. Pharmacol., 52, 113-152. 121 Auerbach, 0. and Garfinkel, L. (1989) Histologic changes in the urinary bladder in relation to cigarette smoking and use of artificial sweeteners. Cancer, 64,983-987. 131 Cameron, R.G. (1989) Identification of the putative first cellular step of chemical hepatocarcinogenesis. Cancer Lett., 47. 163. 167. L41 Clayson, D.B., Nera, E.A. and Lok, E. (1989) The potential for the use of cell proliferation studies in carcinogen risk assessment. Reg. Toxicol. Pharmacol., 9, 284-295. ISI Cohen, S.M., Cano, M., Garland, E.M., St. John, M. and Arnold, L. (1995) Urinary and urothelial effects of sodium salts in male rats. Carcinogenesis, 16, 343-348. 161 Cohen, S.M. and Ellwein, L.B. (1990) Cell proliferation in carcinogenesis. Science (Washington, DC), 249, 1007-1011. 171 Cohen, S.M. and Ellwein, L.B. (1991) Genetic errors, cell proliferation and carcinogenesis. Cancer Res., 5 1, 6493~6505. 181Cohen. S.M. and Lawson, T.A. (1995) Rodent bladder tumors do not always predict for humans. Cancer Lett., 93, Y16. (91 Doll, R. and Peto, R. (1981) The causes of cancer: quantitative estimates of avoidable risks of cancer in United States today. J. Natl. Cancer Inst., 66, 1192-1308. II01 Ellwein, L.B. and Cohen, S.M. (1988) A cellular dynamics model of experimental bladder cancer: analysis of the effect of sodium saccharin in the rat. Risk Anal., 8, 215-221. I1 11 Ellwcin, L.B. and Cohen, S.M. (1990) The health risks of saccharin revisited. Crit. Rev. Toxicol., 20. 31 l-326. [I21 Farber. E. (1991) Hepatocyte proliferation in stepwisc development of experimental liver cell cancer. Digest. 1%. Sci., 36, 973-978. 1131Gibson, J.P., Bewbeme, J.W., Kuhn, W.L. and Elsea, J.R. (1967) Comparative chronic toxicity of three oral estrogens in rats. Toxicol. Appl. Pharmacol., 11, 489-510.

90

H.-C. Kim et al. /Cancer Letters 104 (1996) 85-90

[14] Hasegawa, R., Shirai, T., Hakoi, K., Takaba, K., Iwasaki, S., Hoshiya, T., Ito, N., Nagao, M. and Sugimura, T. (1991) Synergistic enhancement of glutathione S-transfemse placental form-positive hepatic foci development in diethylnitrosamine-treated rats by combined administration of five heterocyclic amines at low doses. Jpn. J. Cancer Res., 82, 1378-1384. [15] Haworth, S., Lawlor, T., Morlelmans, K., Spesk, W. and Zeiger, E. (1983) Sulmonella mutagenicity test results for 250 chemicals. Environ. Mutagen., 5 (Suppl. I), 3-142. [16] Hsu, S.M., Raine, L. and Fanger, H. (1981) Use of avidinbiotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J. Histochem. Cytochem., 29, 577580. [17] lshidate, M., Jr., Sofuni, T. and Yoshikawa, K. (1981) Chromosomal aberration test in vitro as a primary screening tool for environmental mutagens and/or carcinogens. In: Mutation, Promotion and Transformation in vitro. Gann Monogr. Cancer Res., 27,95-108. [ 181 lto, N., Tsuda, H., Tatematsu, M., lnoue. T., Tagawa, Y., Aoki, T., Uwagawa, S., Kagawa, M., Ogiso, T., Masui, T., lmaida, K., Fukushima, S. and Asamoto, M. (1988) Enhancing effect of various hepatocarcinogens on induction of preneoplastic glutathione S-transferase placental form positive foci in rats - an approach for a new medium-term bioassay system. Carcinogenesis, 9.387-394. [19] Kim, H.C., Lee, Y.S., Furukawa, F., Nishikawa, A. and Takahashi, M. (1995) En’lancement of GST-P positive liver cell foci development hv a medium-term carcinogenicity bioassay using repeated administration of D-galactosamine. Cancer Len., 98, 71-76. [20] Lawson, T.A. and Pound, A.W. (1974) The different susceptibility of rat liver lobes to carbon tetrachloride and dimethylnitrosamine. Br. J. Exp. Pathol., 55, 583-588. [21] McCormick, D.L., Burns, F.J. and Albert, R.E. (1981) lnhibition of benzo[a]pyrene-induced mammary carcinogenesis by retinyl acetace. J. Natl. Cancer Inst., 66, 559-564. [22] Mirvish, S.S. (1968) The carcinogenic action and metabolism of urethan and N-hydroxyurethane. In: Advances in Cancer Research, Vol. 11, pp. l-42. Editors: Haddow and Weinhouse. Academic Press, New York. [23] Nixon, J.E., Hendrick, J.D., Pawlowski, N.E., Loveland, P.M. and Sinnhuber, R.O. (1981) Carcinogenicity of aflatoxin in Fischer 344 rats. J. Natl. Cancer Inst., 66, 11591163. [24] Ogiso, T., Tatematsu, M., Tamano, S., Tsuda, H. and lto, N. (1985) Comparative effects of carcinogens on the induction of placental glutathione S-transferase-positive liver nodules in a short-term assay and of hepatocellular carcinomas in a long-term assay. Toxicol. Pathol., 13, 257-265.

[25] Olson, M.J., Johnson, J.T., and Reidy, C.A. (1990) A comparison of male rat and human urinary proteins: implications for human resistance to hyaline droplet nephropathy. Toxicol. Appl. Pharmacol., 102.524-536. [26] Peraino, C., Fry, R., Staffeldt, E. and Christopher, J.P. (1975) Comparative enhancing effect of phenobarbital, diphenylhydantoin, and dichorodiphenyl-trichloroethane on 2-acetyl-aminofluorene induced hepatic tumorigenesis in the rat. Cancer Res., 35, 28842890. [27] Pitot, H.C., Goldsworthy, T., Campbell, H.A. and Poland, A. (1980) Quantitative evaluation of the promotion by 2.3.78 tetra-chlorodibenzo-p-dioxin of hepatocarcinogenesis from diethylnitrosamine. Cancer Res., 40, 3616-3620. [28] Richardson, F.C., Morgan, P.H. Boucheron, J.A., Deal, F.H. and Swenberg, J.A. (1988) Hepatocyte initiation during continuous administration of diethylnitrosamine and 1,2sym-dimethylhydrazine. Cancer Res., 48.988-992. [29] Shirai, T., Kim, D.J., Hakoi, K., Thamavit, W., Pairojkul, C., Hoshiya, T., Hasegawa, R. and lto, N. (1991) Promotion of rat hepatocarcinogenesis by praziquantel. Jpn. J. Cancer Res., 82, 1085-1088. [30] Solt, D.B., Medline, A. and Farber, E. (1977) Rapid emergence of carcinogen-induced hyperplastic lesions in a new model for the sequential analysis of liver carcinogenesis. Am. J. Pathol., 88,595-609. [31] Togei, K., Sane, N., Maeda, T., Shibata, M. and Otsuka, H. (1987) Carcinogenicity of bucetin in (C57BU6xC3H) Ft mice. J. Nat]. Cancer Inst., 79, 1151-l 158. [32] Tsuda, H., Masui, T., lkawa, E., lmaida, K. and lto, N. (1987) Compared promoting potential of D-&ICtOSaIkIC, carbon tetrachloride and partial hepatectomy in rapid induction of preneoplastic liver lesions in the rat. Cancer Lett., 37, 163-171. [33] Tsuda, H., Matsumoto, K., Ogino, H., lto, M., Hirono, I., Nagao, M., Sato, K., Cabral, R. and Bartsch, H. (1993) Demonstration of initiation potential of carcinogens by induction of preneoplastic glutathione S-nansferase p-formpositive liver cell foci: possible in vivo assay system for environmental carcinogens. Jpn. J. Cancer Res., 84, 230-236. [34] Tsuda, T., Oz.&, K., Uwagawa, S., Yamaguchi, S., Hakoi, K., Aoki, T., Kato, T., Sate, K. and lto, N. (1992) Effects of modifying agents on conformity of enzyme phenotype and proliferative potential in focal preneoplastic and neoplastic liver cell lesions in rats. Jpn. J. Cancer Res., 83, 1154-l 165. [35] Wogan, G.N. and Newberne, P.M. (1967) Dose-response characteristics of aflatoxin Bj carcinogenesis. Cancer Res., 27,2370-2376. [36] Zeiger, E. (1987) Carcinogenicity of mutagens: predictive capability of the Salmonellu mutagenesis assay for rodent carcinogenicity. Cancer Res., 47, 1287-1296.