Cross-linking of DNA induced by chloroethylnitrosourea is presented by O6-methylguanine-DNA methyltransferase

The DNA repair enzyme O6-methylguanine-DNA methyltransferase has been used as a reagent to analyse the initial reaction sites of alkylating agents such as chloroethylnitrosourea that cross-link DNA. The transferase can be employed for this purpose because it removes substituted ethyl groups from DNA, as shown by its ability to act on O6-hydroxyethylguanine residues in DNA. The enzyme counteracts the formation of interstrand cross-links induced by bis-chloroethylnitrosourea, but not those induced by nitrogen mustard. Once formed, chloroethylnitrosourea-induced cross-links are not broken by the enzyme. In agreement with deductions from experiments with living cells, it is concluded that chloroethylnitrosourea act by forming reactive monoadducts at the O6 position of guanine and/or the O4 position of thymine, which subsequently generate -CH2CH2- bridges to the complementary DNA strand. A new method for quantitating interstrand cross-links in DNA has been employed.

Properties of carcinogen altered mouse epidermal cells resistant to calcium-induced terminal differentiation

Eight cell lines exhibiting resistance to Ca2+ induced terminal differentiation were derived from primary mouse epidermal cultures and their properties analyzed. The lines developed either spontaneously (2 lines) or after exposure of primary cultures to carcinogens or carcinogens and tumor promoter. All but one of the lines were of epithelial or epitheloid morphology but 3 of the 8 lines lacked desmosomes, keratin filaments and immunoprecipitable keratin proteins, and thus could not be defined as keratinocytes. Two of the 5 keratinocyte lines were tumorigenic in syngeneic Balb/c newborns after 4 months in medium containing 1.2 mM Ca2+, and 3 lines remained non-tumorigenic even after 11 months in 1.2 mM Ca2+. All three of the non-keratinizing lines were tumorigenic. Tumorigenic potential of the 5 keratinocyte lines did not correlate with ploidy (as determined by DNA content), transglutaminase activity or growth in soft agar. However, the 2 tumorigenic keratinocyte lines contained cells which stained intensely red for gamma glutamyl transpeptidase activity, while the non-tumorigenic keratinocyte lines did not. Only those lines lacking desmosomes and keratin filaments grew in soft agar, but these lines were negative for gamma glutamyl transpeptidase activity. Ploidy and transglutaminase activity did not correlate with tumorigenicity in these non-keratinizing lines. These results show that cell lines derived from cultured mouse epidermal cells and selected on the basis of their resistance to Ca2+ induced terminal differentiation may be preneoplastic. Furthermore the association of additional markers with malignant change in these cell lines depended on whether or not the cells were keratinizing.

Induction of neoplastic transformation and DNA single-strand breaks in C3H/10T1/2 clone 8 cells by polycyclic hydrocarbons and alkylating agents

The standard C3H/10T1/2 clone 8 (C3H/10T1/2 CL8) cell transformation assay was tested for its ability to identify a variety of polycyclic hydrocarbons and alkylating agents. Dose-dependent morphologic transformation occurred with benzo[a]pyrene (BaP), 3-methylcholanthrene (MCA), 7,12-dimethylbenz[a]anthracene, BaP-7,8-dihydroxy-7,8-dihydrodiol (BaP-7,8-diol), as well as with the relatively weak in vivo carcinogen benzo[e]pyrene. Dibenz[a,h]anthracene yielded a relatively weak response, whereas anthracene and phenanthrene were negative. In contrast, treatment of C3H/10T1/2 CL8 cells with two directly acting alkylating agents, N-nitroso-N-methylnitroguanidine (MNNG) and styrene oxide, gave no transformation, whereas a third alkylating agent, ethyl methanesulfonate (EMS), gave a weak response. Treatment with MCA (2.5 micrograms/ml) yielded a reproducible positive response and, therefore, served as a positive control for routine use of the C3H/10T1/2 CL8 assay. When cells treated with the hydrocarbons BaP, BaP-7,8-diol, or MCA were analyzed for nonspecific DNA damage (single-strand breaks or alkaline-labile sites) by alkaline elution techniques, little if any DNA damage was observed. In contrast, the alkylating agents MNNG, styrene oxide, and EMS yielded substantial numbers of single-strand breaks.

Cross-resistance studies with V79 Chinese hamster cells adapted to the mutagenic or clastogenic effect of N-methyl-N'-nitro-N-nitrosoguanidine

When V79 cells are exposed to a single low dose of MNNG or MNU they acquire resistance to the mutagenic or to the clastogenic effect of the agents. Here the effect of MNNG pretreatment on mutagenesis (6-thioguanine resistance) and aberration formation in cells challenged with various mutagens/clastogens is reported. MNNG-adapted cells were resistant to the mutagenic effects of MNU and, to a lower extent, of EMS. No mutagenic adaptation was observed when MNNG-pretreated cells were challenged with MMS, ENU, MMC or UV. Cells pretreated with a dose of MNNG which makes them resistant to the clastogenic effect of this compound were also resistant to the clastogenic activity of other methylating agents (MNU, MMS), but not so with respect to ethylating agents (EMS, ENU). Cycloheximide abolished the aberration-reducing effect of pretreatment. However, when given before the challenge dose of MNNG, MNU or MMS, it drastically enhanced the aberration frequency in both pretreated and non-pretreated cells. No significant enhancement of aberration frequency by cycloheximide was found for ethylating agents. The results indicate that clastogenic adaptation is due to inducible cellular functions. It is concluded that mutagenic and clastogenic adaptation are probably caused by different adaptive repair pathways.

Cytotoxicity of monofunctional alkylating agents. Methyl methanesulfonate and methyl-N'-nitro-N-nitrosoguanidine have different mechanisms of toxicity for 10T1/2 cells

N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) and methyl methanesulfonate (MMS) are directly active alkylating agents that methylate cellular macromolecules by SN1 and SN2 mechanisms, respectively. These two chemicals produce similar types of alkylation products in DNA and a similar level of total alkylations on a molar basis, but strikingly different proportions of alkylations of ring oxygen atoms of purines and pyrimidines. Because of this attribute, they have been used in combination to attempt to determine which types of alkylation products are responsible for mutation, transformation, and toxicity. Studies have suggested that the mutation rates produced by these and similar chemicals in cells surviving toxicity correlate well with the number of methyl adducts at the O6 position of guanine, but that cytotoxicity (reduced colony-forming efficiency) does not correlate with any single adduct or with the total level of alkylation of DNA. In this study we have investigated the cytotoxic mechanisms of MNNG and MMS in synchronized 10T1/2 cells, using colony-forming ability as a measure of toxicity. Both MNNG and MMS cause dose-dependent reduction in the ability of 10T1/2 cells to produce colonies of more than 50 cells after 2 weeks in culture. MNNG is about 100-fold more toxic than MMS on a molar basis. As indicated by the inability of cells to exclude trypan blue, MMS kills a fraction of the population of treated 10T1/2 cells after a 30-min exposure; the fraction of cells that excludes trypan blue is correlated with dose of MMS and with colony-forming efficiency. Neither the fraction of cells that is permeable to trypan blue nor the relative colony-forming efficiency is affected by the phase of the cycle when 10T1/2 cells are treated with MMS. Furthermore, MMS toxicity for 10T1/2 cells is not potentiated by caffeine, MMS treatment does not delay progress of S phase, and cells that survive acute membrane toxicity complete the cell cycle without significant delay. In contrast, MNNG treatment produces toxicity that is maximal when 10T1/2 cells are exposed during the S phase and the effect is potentiated by caffeine. MNNG treatment delays DNA replication and this delay is reversed by caffeine. In sharp contrast to 10T1/2 cells treated with MMS, MNNG-treated cells are not made permeable to trypan blue, but are blocked in their ability to proliferate. These observations indicate that MNNG and MMs kill 10T1/2 cells by drastically different mechanisms, MNNG producing toxicity mainly by preventing chromosome replication and MMS producing toxicity mainly by damaging cell membranes.

Studies on adaptation of V79 Chinese hamster cells to low doses of methylating agents

Pretreatment of V79 cells with N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) or N-methyl-N-nitrosourea (MNU) significantly reduced the frequency of mutations (6-thioguanine resistance) induced by a challenge dose of these agents. Mutagenic adaptation can be observed after exposure to pretreatment doses which have no measurable toxic or cytogenetic effects, and which do not change DNA synthesis as measured at the moment of challenge. The influence of pretreatment on challenge dose-induced cell killing, chromosomal aberrations, and sister chromatid exchanges has also been studied. The effect of the challenge with respect to these end points was not in the same way dependent on the pretreatment dose as the challenge dose-induced mutation frequency. Therefore it is concluded that mutagenic adaptation and adaptation with respect to the other end points studied have no identical underlying causes.

Analysis of spontaneous, carcinogen-induced and promoter-induced chromosomal instability in patients with hereditary retinoblastoma

Skin fibroblasts from patients with hereditary retinoblastoma (RB cells) were examined since predisposition to the tumour might be expected to involve some degree of chromosomal instability, as has been noted for several cancer-prone conditions. Spontaneous and N-Methyl-N'-nitro-N-nitrosoguanidine (MNNG)-induced aberration frequencies were measured, the cytogenetic effects of long-term treatment with 12-O-tetradecanoyl-phorbol-13acetate (TPA) were examined and also the spontaneous and TPA-induced sister chromatid exchange (SCE) frequencies. In all the studies the RB cells behaved in a similar fashion to normal human skin fibroblasts.

Isolation and partial characterization of a cerulenin-sensitive mutant of Pseudomonas aeruginosa

Sensitivity of Pseudomonas aeruginosa to cerulenin was first tested. The result indicated that this bacterium is resistant to cerulenin. Cerulenin-sensitive mutants were isolated from P. aeruginosa PML 1552 by 1-methyl-3-nitro-1-nitrosoguanidine treatment and following carbenicillin plus D-cycloserine screening. Isolated mutants were designated CSM-1 to CSM-19, and some characters of CSM-19, which showed rapid growth almost as well as parent strain in the medium without cerulenin, were examined. The cell growth of CSM-19 was greatly inhibited by 50 micrograms/ml of cerulenin, but when the mixture of cellular fatty acids or both cis-vaccenic acid and palmitic acid were added to the medium, the growth was partially recovered. Incorporation of radioactivity into fatty acids from [1-14C]acetate was lowered by cerulenin. Those results mean that the fatty acid synthesis of CSM-19 was decreased by cerulenin. Although cellular fatty acid composition and amount were not notably different between CSM-19 and PML 1552, CSM-19 had less phosphatidylethanolamine, and more phosphatidylglycerol and cardiolipin than PML 1552. CSM-19 was also supersensitive to several other antibiotics, especially to carbenicillin and tetracycline, when compared with PML 1552, although both strains showed identical sensitivity to D-cycloserine, polymyxin B, and chloramphenicol.

The protective role of thiol reducing agents in the in vitro inhibition of rat liver DNA methylase by direct acting carcinogens

The direct-acting carcinogens N-acetoxy-N-acetyl-2-aminofluorene (AcAAF), methyl nitrosourea (MNU), and N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) were tested for their ability to inhibit highly purified, rat liver DNA methylase in vitro. Fifty percent inhibition of DNA methylase activity was achieved with 4.3 mM AcAAF, 47 mM MNU and 2.8 mM MNNG. When the enzyme was reassayed in the presence and absence of dithiothreitol, it was shown that DNA methylase was protected by increasing amounts of the thiol reducing agent. When other thiol reducing agents were tested for their ability to protect DNA methylase from carcinogen damage, a differential protective ability was observed. Dithiothreitol, beta-mercaptoethanol, and reduced glutathione were effective in protecting DNA methylase from carcinogen inhibition, while the effect of cysteine was intermediary and the effect of ergothioneine was minimal. These results may be related to the hypomethylation of DNA observed in several cancers, suggesting that the carcinogens achieve this effect at least in part by inhibiting crucial sulfhydryl group(s) in the methylase molecule. These data also suggest that various intracellular thiols may play an important role in protecting DNA-modifying enzymes from carcinogen damage.

Promotion by histamine of carcinogenesis in the forestomach and protection by histamine against carcinogenesis induced by N-nitroso-N-methylnitroguanidine in the glandular stomach in W rats

The effect of histamine on induction of tumors in the forestomach and the glandular stomach after N-nitroso-N-methylnitroguanidine (MNNG) administration was studied in male inbred W rats. Animals were given 50 micrograms MNNG solution/ml orally for 25 weeks and then 4 mg histamine dihydrochloride sc per day in depot form. Administration of histamine in depot form after MNNG significantly increased the incidence of tumors in the forestomach, but it significantly decreased the incidence of adenocarcinomas in the glandular stomach. All of the tumors induced in the forestomach were of the squamous cell type, and 50% of them were squamous cell carcinomas. Histamine alone had no apparent carcinogenicity in rats.

Potentiating effects of caffeine on teratogenicity of alkylating agents in mice

Teratogenic to subteratogenic doses of x-ray, mitomycin C, MNNG, thio-TEPA, cyclophosphamide, and chlorambucil were administered to pregnant ICR mice together with caffeine at doses of 12.5, 25, or 50 mg/kg on day 11 of gestation. Fetuses were examined for gross malformations on day 18 of gestation. The teratogenicity of mitomycin C was significantly potentiated by caffeine at a dose as low as 12.5 mg/kg. The teratogenicity of chlorambucil was also significantly potentiated by caffeine at 50 mg/kg, but similar potentiation was not observed for x-ray, MNNG, thio-TEPA, and cyclophosphamide.

Rapid changes in deoxynucleoside triphosphate pools in mammalian cells treated with mutagens

In this communication we describe the rapid increase in cellular deoxynucleoside triphosphate (dNTP) concentrations in Chinese Hamster cell line V79 after exposure to known mutagens. With this cell line an expansion of dATP and dTTP pools was detected; changes in dCTP were not large; changes in dGTP were either not significant or too low to quantitate. This situation may reflect the existence of imbalances in dNTP pools at the DNA replication fork. The expansion of dATP and dTTP pools occurred within 2 to 4 hours after exposure of cultured cells to N-methyl-N'-nitro-N-nitrosoguanidine (MNNG). Ultraviolet light (UV), mitomycin C, and cytosine arabinoside also caused similar dNTP pool changes.

[Histopathological studies of the acute and chronic toxic effects of 2 N-nitroso compounds on the blue mussel (Mytilus edulis)]

The acute and chronic effects of two N-Nitroso compounds (N-Nitrosodimethylamine) (DMN) and N-Methyl-N-Nitro-N-Nitrosoguanidine (MNNG) were studied by light microscopy after injection of the chemicals into the foot of the mussel, Mytilus edulis. Acute toxic effects: DMN produced dose-dependent (0.1 to 0.8 mg DMN/mussel) tissue alterations characterized by extensive inflammatory reactions in the Leydig cell tissue of the digestive gland, and necrosis in the germinal epithelium lining the genital ducts. MNNG produced dose-dependent (0.063 to 0.5 mg MNNG/mussel) tissue alterations characterized by extensive necrosis in the epithelial lining of the tubules of the digestive gland. Chronic toxic effects: 0.2 mg DMN/mussel given once a week for 8 weeks, and examined over a period of 30 weeks, revealed tissue alterations characterized by extensive collagenous scar tissue formation in between the tubules of the digestive gland, presence of numerous granulocytomas, and necrosis of the germinal epithelium of the genital ducts. 0.125 mg MNNG/mussel given once a week for 4 weeks, and examined over a period of 7 months, produced tissue alterations mainly characterized by encapsulation or replacement of damaged digestive tubules by collagenous scar tissue.

Effect of 3-aminobenzamide on the rate of ligation during repair of alkylated DNA in human fibroblasts

3-Aminobenzamide, an inhibitor of polyadenosine diphosphoribose polymerase, produced rapid reversible changes in single-strand break frequencies in DNA from primary human fibroblasts damaged by alkylating agents, but it did not cause such changes in the DNA of cells damaged by ultraviolet light. The increase in single-strand peak frequencies was not due to an accumulation of blocked repair sites, such as occurs with DNA polymerase inhibitors, but to a delay in the rejoining of induced breaks. 3-Aminobenzamide increases the net break frequency that results from a dynamic balance between excision and ligation. This balance appears to be regulated at the ligation step by adenosine diphosphate ribosylation, which is rapidly altered by addition or removal of 3-aminobenzamide. The rapidity with which strand break frequencies change in the presence of 3-aminobenzamide implies that individual strand breaks resulting from excision at any time after exposure have a lifetime of no more than about 30 min in the cell.

Weak promotion of C3H/10T1/2 cell transformation by repeated treatments with formaldehyde

C3H/10T 1/2 cells were treated with N-methyl-N'-nitro-N-nitrosoguanidine and then repeatedly exposed to formaldehyde (0.1 to 2.0 micrograms/ml). Exposure of N-methyl-N'-nitro-N-nitrosoguanidine-initiated cultures to formaldehyde concentrations of 0.5 or 1.0 micrograms/ml in a variety of treatment regimens resulted in focus formation in up to 9% of the treated dishes. Transformed foci were observed in 2% or less of the cultures treated with N-methyl-N'-nitro-N-nitrosoguanidine or formaldehyde alone. Formaldehyde thus appears to be only a weak tumor promoter for C3H/10T 1/2 cell transformation.

Alterations of human endometrial stromal cells produced by N-methyl-N'-nitro-N-nitrosoguanidine

Stromal cell cultures obtained from human endometrium were treated repetitively with N-methyl-N'-nitro-N-nitrosoguanidine in vitro at concentrations ranging from 0.5 to 4.0 micrograms/ml, and alterations in growth potential and morphology were analyzed. A single exposure to the carcinogen resulted in morphological evidence of toxicity and reductions in growth rates, plating efficiency, and saturation density as compared to solvent-treated control cells. Cytotoxicity was reduced after additional exposures to the carcinogen. Following repetitive treatments with N-methyl-N'-nitro-N-nitrosoguanidine, human endometrial stromal cells developed enhanced growth potential, the capacity to form macroscopic colonies in soft agar, and elevated gamma-glutamyltranspeptidase activity. Carcinogen-treated cells displayed atypical morphology characterized by irregularities in cell and nuclear size and shape, large bizarre nucleoli, increased nuclear:cytoplasmic ratios, and cellular crowding. Control cells did not display altered morphology or growth parameters even following multiple exposures to solvent and repetitive subculturing. These alterations in growth potential and morphology suggest that the cells are progressing towards preneoplastic and perhaps neoplastic transformation in vitro.

Enhancement of benzo[a]pyrene mutagenicity by chrysotile asbestos in rat liver epithelial cells

The co-mutagenicity of chrysotile asbestos and benzo[a]pyrene (B(a)P) was studied in adult rat liver (ARL-18) epithelial cells. Exposure to chrysotile alone did not increase the mutant incidence whereas B(a)P was mutagenic. Simultaneous exposure of the cells to chrysotile and B(a)P resulted in an enhanced mutant recovery compared to exposure to either of these substances alone. However, chrysotile did not enhance the mutagenicity of the carcinogen, N-methyl-N'-nitro-N-nitrosoguanidine.

Mutagenesis of amino acid pyrolysis products in Chinese hamster V79 cells

The mutagenic effects of Trp-P-1, Trp-P-2, Glu-P-1, Glu-P-2, Lys-P-1, MeA alpha C, A alpha C, MeIQ and IQ on Chinese hamster V79 cells with or without X-irradiated Syrian hamster embryo cells were examined. Of these products, only Trp-P-2 and Lys-P-1 induced ouabain-resistant mutants of Chinese hamster V79 cells in a dose-related fashion; X-irradiated Syrian hamster embryo cells were required for the mutagenicity of Lys-P-1, but not for that of Trp-P-2. None of the other pyrolysis products examined was mutagenic with or without X-irradiated Syrian hamster embryo cells.

Skin fibroblasts from humans predisposed to colon cancer are not abnormally sensitive to DNA damaging agents

We report here that skin fibroblasts from individuals with hereditary adenomatosis of the colon and rectum (ACR), an autosomal dominant trait, were not abnormally sensitive to x-rays, UV-light or MNNG. ACR cells were also competent in restoring x-rays and UV-light induced damage of SV40 T-antigen expression following infection of these cells by the irradiated virus. We concluded, therefore, that sensitivity to x-rays, UV-light and MNNG can not be used to identify gene-carriers dominantly predisposed to colon cancer.

DNA breaks induced by micromolar concentrations of dimethylnitrosamine in liver primary cell cultures from untreated and phenobarbital treated rats

Direct genotoxic effects of the alkylating agent dimethylnitrosamine (DMN) have been difficult to detect in several short-term tests. We simplified our method to detect DNA breaks induced by DMN in rat liver primary cell cultures, and increased its sensitivity about 150 times by changing the conditions of ultracentrifugation and exposure to DMN. Additionally we increased 4 times the sensitivity of the improved assay by isolating hepatocytes from rats treated with phenobarbital (PB). Treatment for 24 h with 60 microM and 13.5 microM DMN of hepatocytes isolated from untreated and PB-treated rats, respectively, decreased the molecular weight of DNA by 50%. After 24 h exposure to 13.5 microM [14C]DMN, hepatocytes from PB-treated rats incorporated 3 times more radioactivity into trichloroacetic acid precipitable material than hepatocytes from untreated rats. Also PB-treatment increased remarkably cytotoxic effects of DMN while it did not modify the cytotoxicity nor the genotoxicity of the direct-acting alkylating agent N-methyl-N'-nitro-N-nitrosoguanidine. These results show that DMN is more genotoxic for hepatocytes from PB-treated rats, and suggest that the enhanced genotoxicity is probably due to an augmented metabolism of DMN by these cultures. Our improved assay of DNA breaks as an indicator of DMN genotoxicity is now as sensitive but faster to perform than hepatocyte-mediated mutagenesis. It could be used to explore genotoxic effects of other alkylating agents and the action of microsomal enzyme modifiers on genotoxicity.

The effects of N-methyl-N'-nitro-N-nitrosoguanidine on KB cells. I. Growth inhibition and cell killing in relation to inhibition of macromolecular synthesis

Cytotoxic effects of N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) on KB cells were analyzed taking into account cell killing and growth inhibition. Attempts to correlate these effects with macromolecular synthesis inhibition were monitored. At low doses (1 X 10(-6) at 1 X 10(-5) M), MNNG acting for 1 h only inhibited cell proliferation in a dose dependent manner and specifically inhibited DNA synthesis. At intermediate doses (1.6 X 10(-5) at 3.2 X 10(-5) M), treatment resulted in detachment and death of the cells during a 20-h post-incubation period. This effect was related to a dose-dependent inhibition of RNA and protein synthesis. At high doses (greater than 3.2 X 10(-5) M) MNNG treatment resulted in cell killing which was distinct from that produced by lower doses, since the cells did not detach from the glass. At these high doses strong inhibition of RNA and protein synthesis was observed early. When cells prelabelled with radioactive uridine or amino acids were treated with MNNG, the release of labelling into the culture medium increased, especially with uridine prelabelled cells. This increase was very high with the intermediate doses of the drug, but more modest with the strong doses. Furthermore, with the moderate doses, cell size showed greater reduction compared to the control than with the high doses. Thus, cell size reduction and loss of cellular material varied parallel to cell detachment. We hypothesize that moderate doses of MNNG may induce cellular degradation, whereas high doses may prevent it. The significance of this finding is discussed.

Karyological analysis of N-methyl-N'-nitro-N-nitrosoguanidine-transformed Syrian hamster cell lines using high resolution G-banding technique

The karyotypes of two Syrian hamster cell lines transformed by N-methyl-N'-nitro-N-nitrosoguanidine have been analyzed using high resolution G-banding technique. The change of gene dosage appears to be the main feature of the chromosome aberrations. One of the cell lines (BHLB4) has been passaged in vitro for 68 times and has a pseudodiploid chromosome number of 44. The most common feature of aberrations of this cell line is monosomy of chromosome 15 and in addition trisomy of chromosome 8. The other cell line (HEC5) was newly transformed, having a modal chromosome number of 50. In spite of the occurrence of trisomy of some autosomes, monosomy of chromosome 15 was still evident. No obvious structural changes were found even in G-banded prophase chromosomes. Our observations support the previous report on the monosomy of chromosome 15, as the most common feature of virally- and chemically-transformed Syrian hamster cell lines.

Neoplastic transformation of primary tracheal epithelial cell cultures

Primary cultures of rat tracheal epithelial cells were treated with the chemical carcinogen N-methyl-N'-nitro-N-nitrosoguanidine (MNNG) to quantitatively study the early events during neoplastic transformation. Epithelial cells were dissociated from tracheas of specific-pathogen-free Fischer-344 rats and were plated on collagen-coated tissue culture dishes. To determine cytotoxicity, cells were exposed on day 1 to various concentrations of MNNG for 3 h and colony forming efficiency (CFE) was determined on day 7. MNNG at a concentration of 0.1 microgram/ml did not decrease CFE as compared to the control cultures, whereas 1 microgram/ml reduced the CEF by 75%. For transformation studies, primary cell cultures received single exposures to MNNG (0.1-0.6 microgram/ml) or multiple exposures to 0.1 microgram/ml of MNNG for 3 h between days 1 and 17. In carcinogen-exposed cultures, morphologically altered foci appeared on day 18, recognizable by high cell density. Transformation frequencies between 1 and 8% were observed depending on MNNG concentration. Cultures containing altered foci continued to grow during the third and fourth week when control cultures had ceased to proliferate and exfoliated from the dish. Over 40% of the cultures which received multiple exposures to MNNG acquired cell line status and could be subcultured greater than or equal to 20 times. None of the 30 control cultures became cell lines. Seventy per cent of MNNG-exposed cell lines showed the anchorage independent growth phenotype at passage 20 as judged by growth in agarose. Four of 10 cultures exposed either 6 or 8 times to MNNG formed invasive squamous cell carcinomas at passage 20 upon inoculation into nude mice. Based on these and previous studies, we feel that unrestricted cell replication is an early key event in carcinogen-exposed epithelial cell populations, preceding neoplastic transformation.