Effect of chloroform on dichloroacetic acid and trichloroacetic acid-induced hypomethylation and expression of the c-myc gene and on their promotion of liver and kidney tumors in mice

Abstract

Chloroform, dichloroacetic acid (DCA) and trichloroacetic acid (TCA) are mouse liver carcinogens that are chlorine disinfection by-products found in drinking water. The effect of chloroform on DCA and TCA-induced hypomethylation and expression of the c-myc gene and on their promotion of liver and kidney tumors was determined. B6C3F1 mice were administered 0, 400, 800 and 1600 mg/l chloroform in drinking water and 500 mg/kg DCA or TCA-administered daily by gavage. DCA, TCA and to a lesser extent chloroform decreased the methylation and increased the mRNA expression of the c-myc gene. Co-administering chloroform prevented only DCA and not TCA-induced hypomethylation and increased mRNA expression of the gene. The effect of chloroform on tumor promotion by DCA and TCA was determined in female and male B6C3F1 mice initiated on day 15 of age with N-methyl-N-nitrosourea. Starting at 5 weeks of age, the mice received in their drinking water DCA (3.2 g/l) or TCA (4.0 g/l) with 0, 800 or 1600 mg/l chloroform until they were killed at 36 weeks. Liver tumors promoted by DCA and TCA were predominantly basophilic except for DCA-treated female mice that were eosinophilic. Only DCA promoted foci of altered hepatocytes and they were eosinophilic in both sexes. Chloroform prevented DCA, but not TCA promotion of liver foci and tumors. In male mice, TCA promoted kidney tumors while DCA promoted kidney tumors only when co-administered with chloroform. Hence, chloroform prevented the hypomethylation and increased mRNA expression of the c-myc gene and the promotion of liver tumors by DCA, while enhancing DCA-promotion of kidney tumors. Thus, the concurrent exposure to two carcinogens, chloroform and DCA resulted in less than additive activity in one organ and synergism in another organ.

Article

Introduction

Chloroform, dichloroacetic acid (DCA) and trichloroacetic acid (TCA) are mouse liver carcinogens (1,2) that are found in finished drinking water being formed as by-products of chlorine disinfection (3,4). This results in humans being exposed simultaneously to the three carcinogens in water. Although chloroform was carcinogenic in the liver of mice when administered by gavage (1), when administered in drinking water it was not carcinogenic and did not promote liver tumors (57). In contrast to chloroform, DCA and TCA when administered in drinking water were carcinogenic in mouse liver (2,812) and promoted N-methyl-N-nitrosourea (MNU)-initiated mouse liver tumors (13,14). The kidney is another potential target organ for the three carcinogens. Chloroform has been reported to induce kidney epithelial tumors in one of four strains of male mice (15) and in male Osborne-Mendel and F 344 rats (1,5). Although, DCA and TCA have not been sufficiently evaluated in 2-year carcinogenesis bioassays to eliminate the kidney as a target organ, the limited bioassays of them have not reported kidney tumors (814). However, both DCA and TCA are metabolites of trichloroethylene, a kidney carcinogen in mice (16) suggesting that at least one of the chloroacetic acids might be carcinogenic in this organ. As the liver and possibly the kidney are common target organs for chloroform and the chloroacetic acids, the potential exists for them to interact with respect to their carcinogenic activity.

5-Methylcytosine (5-MeC) is a natural modification of DNA. The formation of 5-MeC in DNA is catalyzed by DNA methyltransferase (DNA MTase), uses S-adenosylmethionine (SAM) as the methyl donor and yields S-adenosylhomocysteine (SAH). Decrease in the content of 5-MeC, i.e. DNA hypomethylation, is an early event in most human and animal cancers (1720). DNA methylation regulates the expression of genes, so that hypomethylation in the promoter region of some genes, including protooncogenes has been associated with increased expression (17,2123). We have demonstrated that chloroform, DCA and TCA decreased the methylation of the c-myc gene and that DCA and TCA increased the expression of the mRNA of the gene (2427). c-Myc is one of the immediate-early protooncogenes associated with increased cell proliferation in the liver (28,29). The c-myc protein is required for efficient progression through the cell cycle (30,31). It forms a heteromeric complex with the murine Myn protein (32) and the human homolog, Max protein (33). This activates transcription of genes including cdc25A and cyclins A, D1 and E that are associated with cell proliferation (34). The involvement of DNA methylation in the regulation of the expression of protooncogenes would suggest a potential for decreased methylation to up-regulate cell proliferation. This potential to enhance cell proliferation plus the DNA hypomethylation found in most cancers has lead to DNA hypomethylation being proposed as a mechanism of non-genotoxic carcinogens (35).

We report here the effect of co-administering chloroform on the ability of DCA and TCA to hypomethylate and increase the mRNA expression of the c-myc gene and to promote liver and kidney tumors. Tumor promotion was determined using an initiation–promotion assay consisting of initiation with MNU-administered to mice on day 15 of age followed after weaning by drinking water exposure to DCA or TCA with/without chloroform. We have used previously this assay to demonstrate the tumor promoting activity of DCA, TCA and mixtures containing them (13,14). MNU was chosen as the initiator because when administered to newborn mice it initiated both liver and kidney tumors (36,37). This allowed us to evaluate the effect of chloroform on the tumor promotion by DCA and TCA in both the liver and kidney.

Materials and methods

Chemicals and DNA probes

Chloroform, proteinase K and ribonuclease A type III-A were purchased from Sigma Chemical Co. (St Louis, MO), DCA and TCA were from Aldrich Chemical Company (Milwaukee, WI) and HpaII was from New England BioLabs (Beverly, MA). Hybond™-N+ nylon membranes, [α-32P]dCTP (6000 Ci/mmol), [γ-32P]ATP (5000 Ci/mmol), enhanced chemiluminescence reagents and T4 polynucleotide kinase were obtained from Amersham (Arlington Heights, IL). Prime-a-Gene Labeling System was obtained from Promega (Madison, WI). All other chemicals were of electrophoresis grade or of the highest purity available.

Animals

The mice were maintained at the AAALAC accredited laboratory animal faculty of the Medical College of Ohio and in accordance with the US Public Health Service `Guide for the Care and Use of Laboratory Animals'. They were housed in solid bottom polycarbonate cages with stainless steel wire-bar lids. The bedding consisted of Bed-o-Cob (Andersons, Toledo, OH). The animals were provided ad libitum with Laboratory Rodent Diet 5001 (J & B Feed, Toledo, OH) and deionized and filtered (0.2 mm) drinking water. The animal rooms were maintained at 64–76°F and 55±15% relative humidity. The light cycle consisted of 12 h each of light and dark.

Experiment 1: methylation and expression of the c-myc protooncogene

Female B6C3F1 mice at 7–8 weeks of age were exposed to 0, 400, 800 and 1600 mg/l chloroform in their drinking water for a total of 17 days. The high concentration of chloroform was chosen because it was the maximum concentration that did affect the body weight and drinking water consumption of the mice (26,38). Daily, during the last 5 days of exposure to chloroform, the mice were also administered by oral gavage 500 mg/kg of either DCA or TCA in 4.0 ml/kg of water neutralized with sodium hydroxide. We have demonstrated previously that these dose levels of DCA and TCA decreased the methylation and increased the expression of the c-myc protooncogene (24,27). Vehicle-control mice received 4.0 ml/kg of saline. All the treatment groups contained six mice. The mice were killed by carbon dioxide asphyxiation. They were killed at 100 min after the last dose of the chloroacetic acids or vehicle because we have shown it to result in peak expression of the mRNA for the c-myc gene (24). At necropsy, the liver was rapidly excised, weighed and frozen in liquid nitrogen. The frozen liver was stored at –75°C.

Methylation of the promoter region of the c-myc gene.

Methylation of the promoter region of the c-myc gene was evaluated using HpaII restriction enzyme digestion followed by Southern blot analysis as described previously (2427). HpaII does not cut CCGG sites when the internal cytosine is methylated. Briefly, isolated DNA was digested overnight with HpaII (10 U/mg DNA) at 37°C. The digested DNA was electrophoresed on a 1% agarose gel. To control for variation among gels, each gel contained the same number of samples from mice exposed to 0, 400, 800 and 1600 mg/l chloroform. Equal loading of the gel was indicated by equal ethidium bromide fluorescence. The gels were washed with SSC (saline-sodium citrate buffer) and transferred to Hybond™-N+ nylon membranes using a Model 785 vacuum blotter (Bio-Rad Laboratories, Hercules, CA). The DNA was cross-linked by UV irradiation. The membranes were then pre-hybridized at 42°C for 1 h in 20 ml pre-hybridization solution (50% formamide, Dehardt's Reagent, SSPE (saline-sodium phosphate–EDTA buffer), 10% dextran sulfate, 1% SDS and 100 mg/ml denatured non-homologous DNA). Random 32P-labeled c-myc probe (65 ng) was added to the pre-hybridization solution and hybridization continued for 12 h at 42°C. After hybridization, the membranes at 20 min intervals were stringently washed five times with SSC containing 0.5% SDS at 65°C, three times with SSC containing 0.5% SDS at 37°C and finally once with SSC at 37°C. They were then dried, sealed in plastic bags and processed at –70°C with Kodak Biomax MR X-ray film, Kodak intensifying screens and a Kodak M35A automatic film processor. Optical density of the autoradiograms was measured with a Scion Image Analysis System (Scion, Frederick, MD).

The c-myc probe was designed from the GeneBank database (GeneBank accession number M1234) to contain the 1–1315 bp in the promoter region of the gene. The probe was produced by PCR amplification of mouse liver DNA using sense 5′-TCTAGAACCAATGCA CAGAGCAAAAG-3′ and antisense 5′-GCCTCAGCCCGCAGTCCAGTACTCC-3′ primers.

Analysis for mRNA expression of the c-myc protooncogene.

Expression of the mRNA for the c-myc protooncogene was evaluated by reverse transcription–polymerase chain reaction (RT–PCR). cDNA was synthesized from 1 μg of total RNA using 15 U AMV (avian myeloblastosis virus) reverse transcriptase in 20 μl of reaction mixture containing 0.5 μg oligo(dT)15 primer, 5 mM MgCl2, 1 U/μl recombinant RNasin ribonuclease inhibitor, reverse transcription buffer [10 mM Tris–HCl (pH 9.0), 50 mM KCl and 0.1%Triton X-100], and 1 mM of each deoxynucleoside triphosphate. The reaction was at 42°C for 40 min, followed by 99°C for 5 min and 4°C for 5 min. The cDNA samples were diluted 5-fold with nuclease-free water and used for PCR amplification. c-myc and the housekeeping gene, hypoxanthine phosphoribosyl-transferase (HPRT) were co-amplified in 50 μl of reverse transcription buffer containing 10 ng/μl first strand cDNA, 100 μM each deoxynucleoside triphosphate, 1 mM MgCl2, 1.25 U of Taq DNA polymerase, and 25 pmol of upstream and downstream primers for c-myc and HPRT. Primer sequences for c-myc (GenBank database accession number Z38066) were upstream: 5′-TGACGAGACCTTCGTGAAGA-3′ (453–472 bp) and downstream: 5′-ATTGATGTTATTTACACTTAAGGGT-3′ (821–845 bp). For HPRT (GenBank database accession number J00423) the upstream primer sequence was 5′-GCTGGTGAAAAGGACCTCT-3′ (576–594 bp) and the downstream primer sequence was 5′-CACAGGACTAGAACACCTGC-3′ (805–824 bp). The incubation underwent 35 cycles of 94°C for 60 s, 57°C for 60 s and 72°C for 60 s, followed by 72°C for 10 min. The PCR product was electrophoresed in 1% agarose gel containing 0.5 μg/ml ethidium bromide in 0.5× TBE buffer. After electrophoresis, the gels were photographed under UV-irradiation and the optical density measured with the Scion Image Analysis System (Scion). The optical density of c-myc mRNA was standardized using the density of the HPRT gene.

Experiment 2: initiation–promotion bioassay

Female B6C3F1 mice with litters were purchased from Charles River Laboratories (Frederick, MD) and arrived at the laboratory animal facility of the Medical College of Ohio with pups at 7–8 days of age. On day 15 of age all the pups received by i.p. injection 30 mg/kg MNU in saline (4.0 ml/kg). The pups were weaned at 4 weeks of age and randomly assigned to the different treatment groups with littermates dispersed among the groups. One week later, they started to receive in their drinking water 0, 3.2 g/l DCA (25 mmol/l) or 4.0 g/l TCA (25 mmol/l) with 0, 800 or 1600 mg/l chloroform. The drinking water was neutralized with sodium hydroxide. The mice received, therefore, equal molar concentrations of DCA and TCA that we have shown previously to be well tolerated and to promote liver tumors (13,14). The concentrations of chloroform were chosen because in experiment 1 they prevented DCA-induced hypomethylation and increased expression of the c-myc gene suggesting that they might also prevent DCA-promotion of liver tumors. The mice were killed at 36 weeks of age by carbon dioxide asphyxiation.

The body, liver and kidney weights were obtained at necropsy and the two organs evaluated for tumors. The liver was cut into 3–4-mm blocks and along with the right kidney cut transversally and the left kidney cut longitudinally were fixed in buffered formalin and embedded in paraffin. Paraffin sections (5 mm) were obtained from all the liver blocks and from the two blocks of each kidney. The sections were stained with hematoxylin and eosin for histopathologic evaluation. The liver sections were evaluated for foci of altered hepatocytes, hepatocellular adenomas and hepatocellular adenocarcinomas. The liver foci and tumors were characterized as being either basophilic or eosinophilic. Foci of altered hepatocytes in this study contained six or more cells and hepatocellular adenomas were distinguished from foci by the occurrence of compression at >75% of their border. Hepatocellular adenocarcinomas had irregular borders and nuclear pleomorphism and atypia. The kidneys were evaluated for tumors. Renal adenomas were defined as being greater than five proximal convoluted tubules in size and most often compressing adjacent tissue. Morphologically they were characterized as cystic adenomas and papillary cystic adenomas. Adenocarcinomas in the kidney either markedly compressed or infiltrated adjacent tissue and contained prominent nuclear pleomorphism.

Statistical evaluation

SigmaStat software, version 2.03 (Jandel, San Rafeal, CA) was used to perform the statistical analysis. The results were analyzed for statistical significance by a one-way ANOVA followed by the Tukey test with a P-value < 0.05.

Results

Experiment 1. Methylation and expression of the c-myc gene

DCA and TCA with/without concurrent exposure to chloroform did not significantly affect the body weight of the mice (data not presented). The liver/body weight ratio was increased by DCA and TCA from 4.81 ± 0.04 to 7.07 ± 0.12 and 6.19 ± 0.12, respectively (P-value < 0.001). Hence, DCA increased the ratio more than TCA. Co-administering chloroform with DCA or TCA did not significantly affect their ability to increase the liver/body weight ratio.

Effect of chloroform on DCA- and TCA-induced hypomethylation of c-myc gene

The DCA, TCA and to a lesser extent 1600 mg/l chloroform decreased the methylation in the promoter region of the c-myc gene (Figure 1). This was evident after HpaII digestion of the DNA by the appearance of three bands, i.e. 0.5, 1.0 and 2.2 kb, that were absent in digested DNA from control mice. We have demonstrated previously that these bands resulted from HpaII digestion of the promoter region of the c-myc gene (23,24). Thus, when DNA from control or chloroacetic acid-treated mice was digested by both the methyl insensitive XbaI and Eco010901 enzymes only a single band was present and it was of the expected 1.4 kb. Digestion with only one of the two restriction enzymes did not result in this band. Furthermore, we have demonstrated that these bands resulted from hypomethylation of CCGG sites with at least one of the sites being within the promoter region of the c-myc gene and for the 2.2 kb band, the other site being upstream of this region. When DNA from control mice and from DCA, TCA and chloroform-treated mice was digested with the methylation insensitive MspI and probed for c-myc, numerous smaller bands were found between 100 and 600 bp. Apparently, the smear resulted from MspI cutting the large number of CCGG sites, i.e. 12 sites in the area of the gene probed. Thus using MspI, the DNA was demonstrated to be susceptible to digestion. Furthermore, while treatment with DCA, TCA and chloroform resulted in some of the sites becoming unmethylated, many were still methylated as indicated by the occurrence of distinct and larger bands after HpaII digestion.

The intensity of the three bands in Hpa II-digested DNA from mice exposed to DCA or TCA with/without chloroform is presented in Figure 2A and B. Chloroform (800 and 1600 mg/l) significantly reduced the intensity of the three bands in HpaII-digested DNA from DCA-administered mice, indicating that chloroform prevented DCA-induced hypomethylation of the c-myc gene (Figure 2A). In contrast, the intensity of the three bands in HpaII-digested DNA from TCA-treated mice was not affected by chloroform (Figure 2B). Thus, chloroform prevented the ability of DCA, but not TCA to decrease the methylation of the c-myc gene.

Effect of chloroform on DCA- and TCA-induced mRNA expression of the c-myc gene

Northern blot analysis for the c-myc gene indicated virtually undetectable levels of its mRNA in liver from vehicle-treated control mice with increased levels in liver from mice treated with DCA, TCA and chloroform (data not presented). The probes and procedure for the northern blot analysis have been published previously by us (24,25). Although, the high concentration of chloroform increased the expression of the mRNA for the c-myc gene, it was less effective than either DCA or TCA. Furthermore, northern blot analysis indicated that chloroform reduced the ability of DCA but not of TCA to increase the expression of the mRNA for c-myc. To confirm and better quantify this effect of chloroform, the expression of the mRNA for c-myc was determined by RT–PCR analysis (Figure 3). Similar to northern blot analysis DCA, TCA and to a lesser extent chloroform increased the expression of the mRNA. The expression of the mRNA for c-myc and HPRT (housekeeping gene) was quantified and the ratio of their optical density presented in Figure 4. Chloroform caused a dose-dependent reduction in the ability of DCA but not of TCA to increase the expression of the mRNA for c-myc that was statistically significant for 800 and 1600 mg/l chloroform.

Experiment 2: initiation–promotion bioassay

DCA and TCA with/without chloroform did not significantly affect the body weight of the mice (data not presented). The body weight of the mice at the terminal sacrifice is presented in Table I. DCA increased the liver/body weight ratio by the same extent in both sexes (8.86 ± 0.41 and 9.39 ± 0.35 for male and female mice, respectively), while TCA was more efficacious in male (7.64 ± 0.45) than in female mice (5.90 ± 0.34; P-value < 0.01). The ratio in control mice did not differ for the sexes and was 4.36 ± 0.16. In female mice, DCA increased the ratio more than TCA, while in male mice they were equally effective. Co-administering chloroform with DCA or TCA did not significantly affect their ability to increase the liver/body weight ratio. There was no significant treatment-related effect on kidney weight by either chloroacetic acid with/without chloroform (data not presented).

Liver tumors and foci of altered hepatocytes

The effect of DCA and TCA on the incidence of MNU-initiated mice with hepatocellular adenomas, adenocarcinomas and liver tumors (adenomas + adenocarcinomas) and on the multiplicity of liver tumors is presented in Table I and Figure 5A and B. The tumor multiplicity in MNU-initiated control mice was 0.07 ± 0.04 and 0.25 ± 0.16 in female and male mice, respectively. DCA increased the incidence of mice with tumors and the tumor multiplicity in both sexes. Hence, the 3.17 ± 0.76 and 3.92 ± 0.54 liver tumors/mouse in DCA-treated female and male mice, respectively, were significantly greater than the multiplicity in control mice (P-value < 0.001). In female mice, all of the tumors promoted by DCA were adenomas, while in male mice, 22% of the tumors were adenocarcinomas, i.e. 0.86 adenocarcinomas/mouse. Thus, although the multiplicity of liver tumors promoted by DCA was similar in both sexes, adenocarcinomas were found only in male mice.

TCA significantly increased in male but not female mice the incidence of mice with adenocarcinomas and with adenomas + adenocarcinomas (Table I) and the multiplicity of adenomas + adenocarcinomas (Figure 5B). Hence, only the 3.87 ± 0.82 tumors/mouse in male mice (P-value < 0.001) and not the 0.64 ± 0.22 tumors/mouse in female mice was significantly greater than the multiplicity in control mice. Sixty percent of the tumors from TCA-treated male mice were adenocarcinomas.

In female mice, co-administering chloroform decreased the incidence and multiplicity of liver adenomas (all the tumors were adenomas) promoted by DCA (Table I and Figure 5A), i.e. 1600 mg/l of chloroform decreased the multiplicity of adenomas from 3.17 ± 0.76 to zero adenomas/mouse (P-value < 0.01). In male mice, 1600 mg/l of chloroform reduced the multiplicity of DCA-promoted liver tumors from 3.92 ± 0.54 to 1.08 ± 0.37 (P-value < 0.05, Figure 5B). In both sexes, chloroform did not affect the incidence or multiplicity of liver tumors promoted by TCA (Table I and Figure 5A and B). No tumors were found in MNU-initiated mice that received only 1600 mg/kg chloroform (Table I).

The effect of chloroform on DCA and TCA-promotion of foci of altered hepatocyte is presented in Figure 6A and B. No foci were found in MNU-initiated control mice of either sex. DCA promoted a greater yield of foci in female than in male mice, i.e. 5.6 and 1.5 foci/mouse, respectively. Chloroform decreased the yield of foci promoted by DCA in both sexes, i.e. 1600 mg/l chloroform reduced the yield of foci to zero and 0.17 ± 0.11 foci/mouse in females and males, respectively. TCA did not significantly increase the yield of foci in either sex and co-administering chloroform did not alter the low yield of foci. No foci were found in MNU-initiated mice administered only 1600 mg/l chloroform.

In female mice, both the foci and tumors promoted by DCA were over 95% eosinophilic, while in male only the foci were eosinophilic (89%). The tumors in DCA-promoted male mice were basophilic (91%). In male mice, co-administering chloroform increased the percentage of foci and tumors that were basophilic from 11 to 75% and from 91 to 100%, respectively. In both sexes, all the tumors and the few foci promoted by TCA were over 97% basophilic, irrespective of whether chloroform was co-administered.

Kidney tumors

Renal tumors of tubular origin were found in male mice. The majority (>70%) were papillary cystic adenomas with the rest being cystic adenomas and to a lesser extent tubular cell carcinomas (~5%). TCA increased the incidence and the multiplicity of MNU-initiated kidney tumors from (0/8) to 87.5% (14/16) and 1.68 tumors/mouse (Figure 7). Chloroform did not alter the promotion of kidney tumors by TCA, i.e. incidence of mice with tumors was 71.4–87.5% and multiplicity was 1.00–1.68 kidney tumors/mouse for mice administered TCA with 0, 800 or 1600 mg/l chloroform. DCA did not significantly increase the incidence or multiplicity of kidney tumors, i.e. 24% (6/25) with tumors and 0.28 ± 0.11 tumors/mouse compared with zero tumors in eight control mice. However, co-administering 1600 mg/l of chloroform with DCA significantly increased the incidence of male mice with kidney tumors to 100% (12/12) and the multiplicity to 1.75 ± 0.39 tumors/mouse (P-value < 0.01). No kidney tumors were found in MNU-initiated male mice administered only 1600 mg/l chloroform. Thus, although chloroform itself did not increase the yield of kidney tumors, it did increase the yield when co-administered with DCA.

In female mice, a low incidence and multiplicity of kidney tumors was found in mice administered DCA or TCA with/without chloroform. The incidence of kidney tumors among all the treatment groups of female mice ranged from to 28.6% and the multiplicity ranged from to 0.29 tumors/mouse. Thus, in female mice neither DCA nor TCA with/without co-administered chloroform significantly increased the incidence or multiplicity of kidney tumors.

Discussion

Chloroform, DCA and TCA are mouse liver carcinogens for which a non-genotoxic mechanism involving the enhancement of cell proliferation has been proposed (12,39,40). Increased expression of c-myc has been associated with increased cell proliferation (30,31). Thus, chloroform, DCA and TCA have been reported to increase the expression of the mRNA of this gene (24,25,41). The mechanism of non-genotoxic carcinogens has also been proposed to involve decreased methylation of DNA and of protooncogenes including c-myc (35). Chloroform, DCA and TCA appear to cause hypomethylation of DNA and of the c-myc gene by preventing the methylation of hemimethylated DNA formed when DNA is replicated (2427). However, the liver has a very low level of DNA replication, so that DCA and TCA would have to increase cell proliferation and DNA replication for there to be hemimethylated sites requiring methylation. Thus, we have demonstrated that DCA, TCA and Wy-14,643, another peroxisome proliferator, increased cell proliferation and DNA replication before they induced hypomethylation of the c-myc gene (27,42). Prevention of the methylation of hemimethylated sites could result from: (i) decreased activity of DNA MTase; (ii) decreased availability of the methyl donor, SAM; (iii) increased concentration of SAH, an inhibitor of DNA MTase; and (iv) blockage of the access of DNA MTase to the hemimethylated sites in DNA (23). The first three possibilities are unlikely as the hypomethylation induced by DCA, TCA and Wy-14,643 was not associated with an alteration in liver concentration of either SAM or SAH or with decreased DNA MTase activity (27,42). Furthermore, the occurrence of only a few, three bands after HpaII digestion of DNA from mice treated with DCA and TCA would indicate that only a very few of the 12 CCGG sites in the probed region of the c-myc gene were hypomethylated. Thus, digestion of the DNA with MspI that also recognizes CCGG site but is not sensitive to methylation at the internal cytosine resulted in numerous smaller bands of 100–600 bp. This indicates that many of the 12 CCGG sites in the promoter region of the c-myc gene remained methylated in the liver of DCA and TCA-treated mice. The specificity of which CCGG sites become hypomethylated could result from DCA and TCA either directly or through a receptor interacting with the chromatin around CCGG sites to block methylation by DNA MTase.

Chloroform, when administered in the drinking water was less effective than either DCA or TCA in causing hypomethylation and increasing the mRNA expression of the c-myc gene. We have reported previously that chloroform was also less efficacious in decreasing the methylation of the c-myc gene when administered in drinking water than by gavage (26). This correlated with chloroform being carcinogenic in mouse liver when administered by gavage (1) but not when administered in drinking water (5). Chloroform administered in drinking water also did not promote carcinogen-initiated liver tumors in mice (6,7). In contrast, concentrations in drinking water of DCA and TCA that caused hypomethylation and increased expression of the c-myc gene, also induced and promoted liver tumors (814). Co-administering chloroform prevented both DCA-induced hypomethylation and increased expression of the c-myc gene and DCA-promotion of liver tumors. In contrast, chloroform did not alter the ability of TCA to hypomethylate and increase the expression of the c-myc gene and to promote liver tumors. Hence, the effect of chloroform on DCA and TCA-induced hypomethylation and increased expression of the c-myc gene correlated with its effect on tumor promotion by the two chloroacetic acids. Therefore, the effect of chloroform, DCA and TCA on the methylation and expression of the c-myc gene correlated with their effect on tumor promotion.

The mechanism by which chloroform prevented DCA but not by TCA-promotion of liver tumors and hypomethylation and increased expression of the c-myc gene could be to decrease the metabolism of DCA in the liver. DCA is metabolized to a much greater extent than TCA. Hence, only 2% of an administered dose of DCA has been reported to be eliminated unchanged in the urine in contrast to approximately half of a dose of TCA (43). In further contrast to TCA, DCA is metabolized by glutathione S-transferase-zeta (GST-zeta) (4446). Chloroform has been shown to decrease GSH concentration in the liver (4749). Thus, chloroform by decreasing GSH levels could prevent the metabolism of DCA by GST-zeta. As TCA is not metabolized by GST-zeta, this would explain the lack of an affect of chloroform on TCA hypomethylation and increased expression of the c-myc gene and on its promotion of liver and kidney tumors. Decreased hepatic metabolism of DCA should also increase its concentration in the kidney and thus could be the mechanism for the increase in kidney tumors when chloroform was co-administered. Furthermore, the lack of an affect of chloroform on TCA-promotion of kidney could be a result of a significant proportion of a dose of TCA reaching the kidney, as indicated by urinary excretion of TCA that had not been metabolized (43).

The sex sensitivity and pathogenesis of the liver tumors promoted by DCA and TCA were different. Male mice were equally sensitive to the promotion of liver tumors by DCA and TCA, i.e. 3.92 and 3.87 tumors/mouse, while female mice were more sensitive to DCA than TCA, 3.17 and 0.64 tumors/mouse, respectively. This was consistent with previous reports that DCA and TCA possessed similar carcinogenic activity in male mice (8), while DCA induced more tumors than TCA in female mice (1214). Although in female mice, DCA induced a greater yield of tumors (adenomas + adenocarcinomas), TCA induced more adenocarcinomas; this was consistent with previous reports (1214). In both sexes DCA promoted a greater yield of foci of altered hepatocytes (5.6 and 1.5 foci/mouse, respectively) than TCA, which promoted a yield of <0.14 foci/mouse. Hence, in both sexes DCA promoted the greater yield of foci, while TCA promoted the greater yield of adenocarcinomas. Furthermore, male mice were more sensitive than female to the promotion of adenocarcinomas by both chloroacetic acids, while female mice were more sensitive to the promotion by DCA of foci of altered hepatocytes.

The ability of TCA to promote more adenocarcinomas than DCA and the greater sensitivity of female mice to DCA-promotion of foci appeared to correlate with the histological characteristics of the adenocarcinomas and foci. All the adenocarcinomas promoted by TCA and DCA were basophilic. In contrast, >95% of the foci and tumors, all of which were adenomas promoted by DCA in female mice were eosinophilic, while in male mice 89% of the foci were eosinophilic and 91% of the tumors were basophilic. In female mice, foci and adenomas induced by DCA have been reported to be eosinophilic and GST-p positive, while the tumors including adenocarcinomas induced by TCA have been basophilic and GST-p negative (1214,50). In male mice, it would appear that DCA promotes two types of proliferative lesions. One lesion consists of eosinophilic foci and adenomas that have a low propensity to progress to adenocarcinomas and the other consists of basophilic lesions that rapidly progress to adenocarcinomas as indicated by the almost complete lack of basophilic foci. TCA also promoted basophilic adenocarcinomas with a very low yield of basophilic foci in both sexes. Thus, the promotion by TCA and DCA (male mice) of basophilic adenocarcinomas in the almost complete absence of basophilic foci would indicate that the basophilic lesions progressed more rapidly to cancer than eosinophilic foci and adenomas promoted by DCA.

TCA but not DCA increased the yield of MNU-initiated kidney tumors in male mice, while in female mice neither chlorinated acetic acid significantly increased the yield. This demonstrated for the first time the tumor promoting activity of TCA in the male mouse kidney. The ability of TCA but not DCA to promote kidney tumors could result from a greater amount of TCA reaching the kidney as indicated by the excretion of un-metabolized TCA in the urine (43). Co-administering chloroform with DCA, but not with TCA, significantly increased the yield of kidney tumors. In the kidney, DCA and TCA have been reported to cause hypomethylation of the c-myc gene (27) indicating that hypomethylation might also be involved in their renal carcinogenicity. Furthermore, we have unpublished data indicating that co-administering chloroform increased the ability of DCA but not TCA to hypomethylate the c-myc gene in the kidney. Thus, similar to the liver, hypomethylation of the c-myc gene correlated with the renal carcinogenic activity of DCA and TCA. However, unlike in the liver where co-administering chloroform prevented the promotion of tumors by DCA, in the kidney it increased the promotion of tumors and the hypomethylation of the c-myc gene induced by DCA. The proposed decrease by chloroform of DCA metabolism in the liver should allow more DCA to reach the kidney increasing its promotion of tumors. In contrast, the greater amount of an administered dose of TCA that reaches the kidney un-metabolized would make it less sensitive to the effects of chloroform on metabolism in the liver.

DCA and TCA are metabolites of the mouse kidney carcinogen, trichloroethylene (16). It has been proposed that the metabolism of trichloroethylene to S-(1,2-dichlorovinyl)-l-cysteine is involved in its carcinogenic activity in the kidney (51,52). However, the promotion of kidney tumors by TCA and possibly by DCA would suggest that they too could contribute to the carcinogenic activity of trichloroethylene in the kidney.

In summary, the ability of chloroform, DCA and TCA to hypomethylate and increase the mRNA expression of the c-myc gene in the liver correlated with their ability to promote liver tumors. Furthermore, the prevention by chloroform of DCA but not TCA-induced hypomethylation and increased expression of the gene correlated with its prevention of only DCA-promotion of liver tumors. Hence, our results are consistent with hypomethylation and increased expression of c-myc gene being involved in the promotion of liver tumors by DCA and TCA. Furthermore, our results indicate that TCA as well as DCA when chloroform is co-administered can promote kidney tumors in male mice.

Acknowledgments

Although the research described was funded in part by the US Environmental Protection Agency's (STAR) program through grants R-825384 and R-828083, it was not subjected to the Agency's review and does not necessarily reflect the views of the Agency.

Footnotes

References