Existence of a Threshold for Induction of Aberrant Crypt Foci in the Rat Colon with Low Doses of 2-Amino-1-methyl-6-phenolimidazo[4,5-b]pyridine

Abstract

The possible existence of a dose threshold for chemical carcinogenicity is of great importance in the regulatory science field. It has been generally considered that genotoxic carcinogens have no threshold in exerting carcinogenic potential (Preussmann, 1980; Tomatis et al., 1997), because classically carcinogens are mutagenic, interacting with DNA to produce irreversible genetic changes in target organ cells. This is based on acceptance of a linear curve down to zero at low doses for risk assessment of exposure to humans with chemicals found to be carcinogenic in animal studies. While there are only limited data available for estimation of cancer risk assessment in humans exposed to genotoxic carcinogens (Gaylor, 1979; Littlefield et al., 1979; Peto et al., 1991), it has been argued that the nonthreshold theory can be challenged. This is because life forms possess biological responses that can ameliorate genotoxic activity. It is very important to resolve this point from the viewpoint of cancer risk control and management, and recently we showed the existence of a threshold for hepatocarcinogenicity of both 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx) and diethylnitrosamine (DEN) in rats (Fukushima et al., 2002, 2003). In addition, Williams et al. (1998, 2000) and Yoshino et al. (2002) also recently provided evidence of a threshold for 2-acetylaminofluorene and diethylnitrosamine hepatocarcinogenicity.

There are many genotoxic carcinogens occurring naturally in our environment, including the large group of heterocyclic amine mutagens (Sugimura et al., 1995, 2000). The human daily intake of 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), one of these food-derived agents, is estimated to be 0.1–13.8 μg/person (Wakabayashi et al., 1993). PhIP can be detected in the urine of healthy volunteers after eating cooked meat (Donald et al., 1995; Ushiyama et al., 1991; Wakabayashi et al., 1993) and in rats, it causes DNA adduct formation in the colon (Fretland et al., 2001; Kaderlik et al., 1994) and treatment at high doses induces carcinomas in the colon, breast, and prostate (Hasegawa et al., 1993; Ito et al., 1991; Shirai et al., 1997).

Recently in vivo medium-term bioassays for carcinogens have been accepted as possible alternatives to long term carcinogenicity tests (Ito et al., 1988) and appropriate for assessment of low dose effects because of their high sensitivity. Aberrant crypt foci (ACF) are established preneoplastic markers in the colon of rats (Bird, 1987; Tudek et al., 1989) and their ready detectability underlies their acceptance as end-point lesions to assess carcinogenic responses in medium-term bioassays. PhIP has been shown to induce ACF in a dose-related fashion over a range of doses (Nakagawa et al., 2002; Tudek et al., 1989). In the present study, for clarification of human risk assessment of genotoxic carcinogens, we examined low dose carcinogenicity of PhIP in the rat colon in detail using a medium-term bioassay, with the primary aim of determining whether the response curve is indeed linear near zero.

DNA adduct formation is considered to be an important factor in carcinogenesis with heterocyclic amines and PhIP-DNA adducts are formed in rat colon (Ochiai et al., 1996). Generation of oxygen free radicals is also a key step in carcinogenesis, again induced by various heterocyclic amines (Maeda et al., 1995). 8-Hydroxy-2′-deoxyguanosine (8-OHdG) is the most abundant species of adduct associated with oxidative stress, resulting in DNA damage and specific types of mutation (Kasai et al., 1987). Therefore, levels of PhIP-DNA adducts and 8-OHdG were also examined in the present study to cast further light on mechanistic aspects of PhIP carcinogenicity at low doses in the colon.

MATERIALS AND METHODS

Animals and chemicals. A total of 1835 male five-week-old F344 rats were obtained from Charles River Japan, Inc. (Atsugi, Kanagawa, Japan), and housed in rooms maintained on a 12-h light/dark cycle, at constant temperature and humidity, and observed daily. PhIP (purity, 99.8%) was purchased from the Nard Institute, Nishinomiya, Japan.

Experimental procedures. The experiment was started when the animals were six weeks of age. They received PhIP at doses of 0 (group 1, a control), 0.001 (group 2), 0.01 (group 3), 0.1 (group 4), 1 (group 5), 10 (group 6), 50 (group 7), 100 (group 8), and 400 ppm (group 9) in powdered basal diet (Oriental MF, Oriental Yeast Co., Tokyo, Japan) for 16 weeks, continuously. The lowest level, 0.001 ppm of PhIP was established as equivalent to the daily intake of this carcinogen in humans (Wakabayashi et al., 1993). Numbers of rats were 240 in group 1, 242 in group 2, 241 in group 3, 243 in group 4, 244 in group 5, 212 in group 6, 214 in group 7, 62 in group 8, and 61 in group 9. The rats were killed at the end of week 16 under ether anesthesia for examination of ACF (61 to 244 rats) in the colon. Additional rats in groups 1 to 9 were given diets containing PhIP and killed at week 4 for examination of PhIP-DNA adducts (three or four rats) and 8-OHdG (five rats each) in the colon. The animals were carefully observed during the course of the experiment, body weights, water intake, and food consumption were measured every week. Calculations to achieve precise mole-per-rat of total PhIP ingested in every initiated group was estimated.

ACF counts. Colons were quickly excised, flushed with saline, and inflated by intraluminal injection of 10% phosphate-buffered formalin solution, slit open along the longitudinal median axis from the cecum to anus, and fixed flat between two pieces of filter paper in 10% phosphate-buffered formalin. After fixation for at least 24 h at 4°C, the colons were all stained with 0.2% methylene blue (in H₂O) for 3–5 min, and then examined for ACF by light microscopy at 40× and 100× magnification using the following criteria for identification: (1) increased size as compared to normal crypts, (2) enlarged pericryptal zone, (3) slight elevation above the surrounding mucosa, and (4) frequently more elongated shape of the luminal opening.

PhIP-DNA adducts and 8-OHdG formation. The colons were excised and flushed in saline, and the mucosa was scraped off to obtain samples, which were frozen in liquid nitrogen and stored at −80°C until the levels of PhIP-DNA adducts in the colon were measured by the ³²p-postlabeling method as described previously (Uehara et al., 1996). Measurement of 8-OHdG levels in colon DNA was performed with the method of Nakae et al. (1997).

Statistical analyses. Statistical analysis of our data was performed using the StatView-J 5.0 program (Abacus Concepts, Inc., Berkeley, CA). Differences from the control values were evaluated for significance by the Dunnet-test.

RESULTS

General Findings

All the rats survived in good condition until the scheduled sacrifices. No adverse effects on average body weight gain were observed in rats treated with PhIP except at 400 ppm. Final average body weights were significantly lower in group 9 (400 ppm PhIP-treated group) than those of group 1 (a control). There were no significant differences in liver and kidney weights among the group except group 9 (Table 1). Average total PhIP intake in each group was dose-dependent. No macroscopic lesions were apparent in any organs, including the colon.

Induction of ACF in the Colon

After 16 weeks treatment with PhIP at various doses in the diet, total numbers of ACF foci in the rat colon of groups receiving 0.001–10 ppm of the carcinogen did not differ from the control value (Table 2 and Figure 1), in contrast to the dose-dependent significant increase observed with 50 ppm and above. Numbers of ACF comprising one and two crypts in the groups given 0.001–10 ppm PhIP were also not different from the control values, while those with 50 ppm PhIP and over were significantly increased (Table 2). Numbers of ACF with three crypts and > four crypts were significantly increased only with PhIP at doses of 100 and 400 ppm.

FIG. 1.

Induction of aberrant crypt foci in the colons of rats treated with PhIP at various doses for 16 weeks. Significant differences from the 0 ppm group at ^*p < 0.05 or ^**p < 0.01. Bars, SD.

Open in new tabDownload slide

Formation of PhIP-DNA Adducts and 8-OHdG

At week 4, there was a linear relationship between the various doses (0.01–400 ppm) of PhIP and the levels of PhIP-DNA adducts at 0.01 ppm and above (Table 3 and Figure 2). However, no significant increase was evident at 0.001 ppm dose.

FIG. 2.

PhIP-DNA adducts in the colons of rats fed diets containing PhIP for four weeks. Significant difference from the 0 ppm group at ^*p < 0.01. Bars, SD. Values in the figure are shown in logarithmic scale.

Open in new tabDownload slide

Concerning the 8-OHdG levels in the colon DNA at week 4, no significant differences were apparent among the groups receiving PhIP from 0.001–100 ppm and the control group and only 400 ppm PhIP caused a significant increase (Table 3 and Figure 3).

FIG. 3.

The 8-OHdG formation levels in the colons of rats treated with PhIP for four weeks. Significant difference from 0 ppm group at ^*p < 0.05. Bars, SD.

Open in new tabDownload slide

DISCUSSION

The present results clearly indicate that the curve for induction of ACF in the rat colon by PhIP is not linear down to zero. Similarly, no-response levels were evident for both PhIP-DNA adducts and 8-OHdG formation, indicating that there are threshold for carcinogenesis-related parameters with PhIP colon carcinogenicity.

Recently we found that the hepatocarcinogens, MeIQx and DEN, do not induce the preneoplastic lesions, glutathione-S-tranceferase placented form (GST-P) positive foci, in rat liver at very low doses (Fukushima et al., 2002, 2003). Morever, MeIQx-DNA adducts and particularly 8-OHdG levels demonstrated no-observed effect levels. Our findings thus indicated the existence of a threshold for carcinogenicity with genotoxic agents.

The present results for ACF, DNA-adducts, and 8-OHdG in the colons of rats treated with PhIP at various doses point to the same conclusion (Fig. 4). On the other hand, it is noteworthy to mention that the no-response level for adduct formation (below 0.01 ppm; about 21 adducts/cell), and for induction of ACF (below 50 ppm; about 60 adducts/cell), supports the notion that rather large threshold number of adducts must be exceeded in order to induce formation of ACF. i.e., here, the dose required to initiate ACF is approximately 5000 times higher than that for adduct formation. Maeda et al. (1995) reported that carcinogenic heterocyclic amines generate oxyradicals at differing levels, MeIQx giving the highest values as judged by electron-spin resonance (ESR) spin trapping and PhIP being much less active in this regard. Two different metabolic pathways have been indicated for activation of PhIP. One is the hepatic pathway, involving N-hydroxylation by CYP1A2 and O-acetylation by N-acetyltransferase-2 and the other is extrahepatic, rendering free-radical metabolites. On ESR examination, generation of free radicals was greater with 2-amino-3-methylimidazo[4,5-f] quinoline (IQ) than PhIP although DNA analysis showed adduct formation to be similar with the two carcinogens (Moonen et al., 2002). From this evidence, participation of 8-OHdG to colon carcinogenesis due to PhIP may not be of direct importance, although this parameter also demonstrated a threshold in the present study.

FIG. 4.

Summarized relationships among carcinogenesis-related biomarkers in the colon of rats treated with PhIP.

Open in new tabDownload slide

Biological adaptive responses, resulting in physiological protection of cells against toxic agents, have recently become accepted for radiation carcinogenesis at low dose (Wollff et al., 1998). This concept might also be useful for understanding low dose effects of chemical carcinogenesis, since adaptation might be expected to occur in response to low doses of all types of DNA damaging agents (Kleczkowska and Althaus, 1996; Olivieri et al., 1984). It has been reported that extremely low doses of chemical carcinogens actually decrease the degree of DNA damage in treated animals, although the authors of the article in question hesitated to draw firm conclusions (Kitchin and Brown, 1994). In addition, the importance of toxicokinetics of chemicals for carcinogenicity has recently been stressed. Absorption of carcinogens into the body, distribution to target organs, metabolism to active ultimate forms which react with DNA, induction of detoxifying enzymes, formation of polar metabolites and excretion, all influence DNA damage. Moreover, various factors such as stimulation of the immune response, induction of different detoxification and repair enzymes, and upregulation of tumor suppressor genes could result in beneficial effects with low dose exposure to carcinogens.

In conclusion, a threshold may exist for the colon carcinogenic potential of PhIP, and by analogy, probably also for other colon-genotoxic carcinogens. Recently Waddell (2003) stressed the existence of a threshold for DEN carcinogenicity in the lever and esophagus of rodent in his review article. Williams et al. (2000) postulated that mechanisms differ between low and high exposures, and reflect thresholds for hepatocellular initiating effects by low dose genotoxic carcinogens. Previously we provided evidence that genotoxic hepatocarcinogens may exhibit a threshold (Fukushima et al., 2002). The present findings provide a new basis for extrapolation from animal carcinogenicity data to human risk assessment.

The authors would also like to acknowledge the encouragement of Dr. N. Ito (Emeritus Prof., Nagoya City University Medical School, Nagoya) and Dr. T. Kitagawa (Director, Cancer Institute, Tokyo). This research was supported by a grant from the Japan Science and Technology Corporation, included in the Project of Core Research for Evolutional Science and Technology (CREST) and a Grant-in-Aid for Specially Promoted Research from the Ministry of Education, Science, Sports, Culture and Technology of Japan.

REFERENCES

Author notes

*Department of Pathology, Osaka City University Medical School, 1-4-3 Asahi-machi, Abeno-ku, Osaka 545-8585, Japan; †Department of Pathology, Sasaki Institute, 2-2 Kandasurugadai, Chiyoda-ku, Tokyo 101-0062, Japan; ‡Department of Oncological Pathology, Cancer Center, Nara Medical University, 840 Sizyo-cho, Kashihara-shi, Nara 634-8521, Japan; §Experimental Pathology and Chemotherapy Division, National Cancer Center Research Institute, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan; ¶Department of Pathology, Kagawa Medical University, 1750-1 Ikenobe, Miki-cho, Kida-gun, Kagawa 761-0793, Japan; ∥Department of Pathology, Nagoya City University Medical School, 1 Kawasumi, Mizuho-cho, Mizuho-ku, Nagoya 467-8601, Japan; ∥|Division of Oncological Pathology, Aichi Cancer Center Research Institute, 1-1 Kanokoden, Chikusa, Nagoya 464-8681, Japan; and ∥∥Division of Pathology, National Institute of Health Sciences, 1-18-1 Kamiyoga, Setagaya-ku, Tokyo 158-8501, Japan