An axiom in regulatory sciences is that there are no thresholds for genotoxicity of chemicals. It leads to another default assumption that genotoxic carcinogens impose cancer risk on humans without thresholds, i.e., a linear non-threshold model. Therefore, no acceptable daily intake (ADI) is set for food additives, pesticides and veterinary drugs when they have genotoxic and carcinogenic activities. However, humans possess a number of defense mechanisms such as metabolic inactivation, DNA repair, error-free translesion DNA synthesis and so on. These mechanisms may constitute practical thresholds for genotoxicity. Error-free translesion DNA synthesis is a process where DNA polymerases bypass lesions in DNA by insertion of correct bases opposite the lesion and continue replication of whole chromosomes. These mechanisms might have been evolved because organisms from bacteria to humans are exposed to endogenous as well as exogenous genotoxic compounds. In fact, levels of spontaneous mutagenesis are strongly influenced by ability of DNA repair and translesion DNA synthesis of the host cells. Here, I show evidence that DNA repair and translesion DNA synthesis play roles in practical genotoxic thresholds in Salmonella typhimurium used for bacterial mutation assays, and discuss future directions of the research on genotoxic thresholds in vivo.
In Europe, there has been a scientific discussion on possible thresholds in chemical carcinogens since the late 1990s. Based on this discussion, the Scientific Committee on Occupational Exposure Limits (SCOEL) of the European Union has discussed a number of chemical carcinogens and has issued recommendations. For some carcinogens, health-based Occupational Exposure Limits (OELs) were recommended, while quantitative assessments of carcinogenic risks were performed for others. For purposes of setting OELs the following groups of carcinogens were adopted: (A) Non-threshold genotoxic carcinogens; for low-dose assessment of risk, the linear non-threshold (LNT) model appears appropriate. For these chemicals, the risk management may be based on the ALARA principle (”as low as reasonably achievable”), technical feasibility, and other socio-political considerations. (B) Genotoxic carcinogens, for which the existence of a threshold cannot be sufficiently supported at present. In these cases, the LNT model may be used as a default assumption, based on the scientific uncertainty, and the ALARA principle may be applied as well. (C) Genotoxic carcinogens with a practical threshold is supported by studies on mechanisms and/or toxicokinetics; health-based exposure limits may be based on an established no-observed adverse effect level (NOAEL). (D) Non-genotoxic carcinogens and non DNA-reactive carcinogens; for these compounds a true (”perfect”) threshold is associated with a clearly founded NOAEL. The mechanisms shown by tumor promoters, spindle poisons, topoisomerase II poisons and hormones are typical examples of this category. Health-based OELs are derived for carcinogens of Groups C and D, while a risk assessment is carried out for carcinogens of Groups A and B. In order to highlight the most important differentiation between Groups B and C, the basic reasoning is given for the six compounds formaldehyde, vinyl acetate, acrylonitrile, acrylamide, trichloroethylene and methylene chloride.
Genotoxic carcinogens are chemicals or factors which not only induce neoplastic lesions in animal bioassays but also test positive in genotoxicity assays in vitro or in vivo. However, it is actually difficult to discriminate genotoxic and non-genotoxic carcinogens because both assays are basically independent each other, which raises a simple query as to how much the detected genotoxic potential can consequently contribute to carcinogenicity. To clarify this critical issue, we have studied the mechanisms of action of carcinogens in transgenic rats or mice carrying reporter genes, which are expected as powerful tools for the simultaneous evaluation of both genotoxicity and carcinogenicity at the same organ level. A number of studies of genotoxic carcinogens using these transgenic rodents have revealed good correlations between genotoxicity and carcinogenicity in terms of mechanism of action. On the other hand, a known non-genotoxic carcinogen dicyclanil increased in vivo genotoxicity as well as oxidative DNA damage in female mice, consistently with the sex specificity of its carcinogenicity, albeit without clear evidence of direct DNA reactivity. In contrast, a genotoxic chlorinated water by-product MX failed to exert in vivo genotoxicity and carcinogenicity in mice. We also confirmed that such reporter gene-carrying rodents are not susceptible or resistant to carcinogenicity as compared with intact counterparts. These results thus indicate that understanding of the detailed mechanism of carcinogenic action could be crucial for more precise risk assessment, and bioassay systems using transgenic rodents carrying reporter genes would be extremely useful for that purpose.
Recent findings have indicated that there may be a practical threshold or an ineffective dose range for the carcinogenicity of genotoxic carcinogens. In male Fischer 344 rats given a 16-week chronic feeding administration of 0.0001-1 ppm of N-nitrosodiethylamine (DEN), glutathione S-transferase placental form (GST-P)-positive liver preneoplasias developed at 0.1 ppm or higher, but hepatic level of 8-oxoguanine (8-oxoG), an oxidative DNA damage, was not elevated even at 1 ppm. In contrast, hepatic 8-oxoG level was elevated by a single intraperitoneal administration of 0.001-100 mg/kg body weight of DEN within 6 h and remained high within 72 h, in a clear dose-dependent manner without any ineffective doses, and GST-P-positive preneoplasias correspondingly developed through the selection procedure. The 8-oxoG level was elevated also in extrahepatic organs within 6 h but returned to the normal level within 72 h. In a separate experiment, hepatic 8-oxoG level remained high even 18 weeks after 2 weekly intraperitoneal administrations of 100 mg/kg body weight of DEN. The early prolonged elevation of 8-oxoG level in target organ DNA was similarly induced by heterocyclic amines and dimethylarsinic acid in association with the down-regulation of the Ogg1 gene encoding an 8-oxoG-specific repair enzyme. Taken together, it is suggested that adaptation mechanisms may be involved in the achievement of an ineffective dose range for the carcinogenicity of genotoxic carcinogens during their continuing exposure at sufficiently low level doses.
Cancer is due to multiple alterations to DNA. Chemicals can increase the cancer risk by directly damaging DNA (DNA reactivity) or by increasing cell proliferation (DNA replications), increasing the number of opportunities for spontaneous DNA damage. Genotoxicity is a more comprehensive term than DNA reactivity. Many of the mechanisms of genotoxicity, such as clastogenicity, inhibition of DNA repair, or damage to the mitotic apparatus, produce DNA damage indirectly. These non-DNA reactive mechanisms involve interactions with proteins and mechanistically are threshold phenomena. 2-Acetylaminofluorene (AAF) is DNA reactive. Its dose response for urinary bladder DNA adduct formation is linear, whereas the tumor response is non-linear. Non-linearity is at the dose at which increased cell proliferation occurs, related to the threshold phenomenon of cytotoxicity. Non-linearity for DNA reactive carcinogens can also be produced by changes in metabolic processes of activation and/or deactivation due to saturable kinetics. Arsenic produces bladder cancer with a non-linear dose response in animal models and humans. Genotoxicity of arsenic occurs secondarily to indirect mechanisms, not DNA reactivity, it has a non-linear dose response, and the genotoxic mechanism appears to have a threshold, occurring only at doses in excess of toxic concentrations. Numerous non-genotoxic agents have been identified as bladder carcinogens in rodent models, most acting by inducing cytotoxicity with regenerative proliferation. Cytotoxicity can be produced by formation of urinary solids or by urinary reactive chemicals. Urinary solids are a defined threshold phenomenon based on the physical-chemical property of solubility. Likewise, chemical induction of cytotoxicity is a known threshold phenomenon. Non-genotoxic chemicals have a threshold dose response with respect to carcinogenesis, as do most genotoxic agents. DNA reactive chemicals have a non-linear dose response.
A mechanistic understanding of genotoxicity is important for the risk assessment of the exposure of human populations to chemicals. The nature of the dose response relationship at low doses is valuable information in the evaluation of the biological importance of such exposures. A range of mathematical and statistical approaches have been used to try to characterize responses at these low doses. Methods include mathematical models which do or do not include thresholds and statistical methods which try to identify No-observable effect levels (NOELs). It is important to appreciate that determination of an NOEL is not evidence for a threshold. There is an increasing appreciation of the potential to identify ‘pragmatic’ thresholds using experimental systems with a range of biomarkers. The accurate characterization and estimation of these dose-response relationships requires careful experimental design which can improve the accuracy of the estimates of the response while avoiding the introduction of artifactual effects. Statistical approaches such as Design of Experiment (DoE) methodology, which builds on the traditional factorial design, can provide efficient approaches for the description and estimation of dose-response relationships of both individual and combinations of agents. Estimation approaches such as the benchmark dose methodology and the concept of thresholds of toxicological concern provide practical methods for addressing the threshold problem.
The determination and utilization of the actual low dose-response relationship for chemical carcinogens has long interested toxicologists, experimental pathologists, modelers and risk assessors. To date, no unequivocal examples of carcinogenic thresholds in humans are known. However, at least 5 examples of thresholds of preneoplastic foci or tumors have been observed in animals. The two largest dose-response studies utilized 20,880 mice (2-acetylaminofluorene) and 7,200 rainbow trout fry (aflatoxins). In both of these studies linear relationships were observed for DNA adducts and for liver tumors. A threshold relationship was observed for 2-acetylaminofluorene induced mouse urinary bladder cancer. Other comprehensive dose-response studies have examined the chemicals 2-amino-3,8-dimethylimidazo[4,5-f]-quinoxaline, 2-amino-1-methyl-6-phenolimidazo[4,5-b]pyridine and diethylnitrosamine. Taken collectively, the DNA adduct data for these 6 well studied chemicals are fairly linear. The foci and tumor data show either supralinear, linear or threshold curves, making it difficult to generalize. All the 6 studied chemicals cause multiple biological effects including genotoxicity, cytotoxicity and cell proliferation in complex dose and time dependent patterns that are not fully understood. We do know that there are multiple possible biological defenses (at least 7 pharmacokinetic and 7 pharmacodynamic) against the development of cancer. Currently, we have limited scientific and regulatory understanding of chemicals that act simultaneously or sequentially via both linear and nonlinear carcinogenic pathways (genotoxic and nongenotoxic). If an 100% experimental approach is used to elucidate the dose-response of chemicals of dual carcinogenic dose-response properties (linear and non linear), this would require studying 2 or more such chemicals in a large scale coordinated fashion employing at least 1,000 animals, 5 different treatment groups, 7 different study parameters and 8 different scientific disciplines.
The presence or absence of a threshold in carcinogenesis for genotoxic carcinogens was reevaluated. The ED01 study of 2-acetylaminofluorene, performed in the U.S. using more than 24,000 mice, provides us with information about the practical limits of an attainable experimental approach for determining carcinogenesis thresholds. The data indicated that the dose response was highly non-linear and an apparent threshold existed for bladder carcinogenesis, but that it was linear-no-threshold for liver carcinogenesis in the same animals. Despite smaller study sizes, we attempted to evaluate the carcinogenesis dose response to 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP) and 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx) in rats, using published data. Mammary tumors were induced in female F344 rats by PhIP in a linear-no-threshold dose response, as was lymphocytic leukemia in male and female rats. However, colon tumors were induced in a non-linear dose response, possibly with a threshold, in the same male animals. Liver tumors were induced in male F344 rats by MeIQx, and preneoplastic changes in the liver were induced in non-linear dose response, possibly with a threshold. From these findings, it can be deduced that linear or non-linear dose response with or without thresholds changes depending upon the exposing chemical, species and target organs. Considering heterogeneity of humans there would be no appropriate animal models to evaluate threshold in humans for carcinogenicity of chemicals. Some types of genotoxic carcinogens, such as methyl methanesulfonate, ethyl methanesulfonate and N-methyl-N'-nitro-N-nitrosoguanidine, show highly non-linear dose response with an apparent threshold in mammalian cells or bacteria in vitro. Involvement of repair mechanisms strongly supports the presence of a threshold. Although it is necessary to confirm non-linear dose response for animal carcinogenesis of a compound showing a threshold for in vivo genotoxicity, it is expected that such compounds exhibit thresholds for human carcinogenesis. Dose response studies of genotoxic carcinogens will provide information on using valuable chemicals safely.