Nongenotoxic carcinogens: development of detection methods based on mechanisms: a European project.

While the accumulation of genetic changes in a somatic cell is considered essential for the genesis of a cancer, it has become clear that not all carcinogens are genotoxic, suggesting that some carcinogens indirectly participate in the generation of genetic changes during carcinogenesis. A European project funded by the European Community was thus conceived to study mechanisms of nongenotoxic aspects of carcinogenesis. Two main strategical approaches were adapted: (i) to study whether and how Syrian hamster embryo (SHE), Syrian hamster dermal (SHD) and BALB/c 3T3 cell transformation systems simulate in vivo carcinogenesis, and to examine whether they can detect nongenotoxic carcinogens; (ii) to study, refine and validate mechanisms-based end-points for detection of nongenotoxic carcinogens. For mechanisms-based research, the proposed end-points included gap junctional intercellular communication (GJIC) inhibition, altered expression of critical genes, immortalization and aberrant cell proliferation. We also selected model compounds commonly usable for various endpoints. Our major results can be summarized as follows: (1) SHE and BALB/c 3T3 transformation systems reflect both genotoxic and nongenotoxic carcinogenic events; they detect not only genotoxic but also many although not all, nongenotoxic carcinogens. This is further supported by the fact that both genotoxic and nongenotoxic carcinogens were able to immortalize SHD cells. (2) Many nongenotoxic carcinogens, although not all, inhibit GJIC in vitro as well as in vivo. Mechanistic studies suggest an important role of blocked GJIC in carcinogenesis and that different mechanisms are involved in inhibition of the communication by different agents used. However, inhibition of GJIC is not a prerequisite for the enhancement (or induction) of transformation of SHE or BALB/c 3T3 cells. (3) Among compounds examined, there was a good correlation between induction of micronuclei and cell transformation in SHE cells while no such correlation was found between the induction of cell transformation and ornithine decarboxylase activity. (4) Two transgenic mouse mutation assays (lacI and lacZ) were established and validated. The genotoxin dimethylnitrosamine was shown to be mutagenic to the liver in both assays. Ortho-anisidine, a bladder-specific carcinogen that was inactive in standard rodent genetic toxicity assays was uniquely mutagenic to the bladder of the transgenic mice. The peroxisome proliferator methyl clofenipate was established as nonmutagenic to the liver of both transgenic mice. That eliminated DNA damage as a cause of the liver tumours produced by this chemical and weakened the idea that induced cell division leads to mutation induction. (5) With an in vitro DNA replication model, it was found that DNA damage induced by genotoxic agents can be responsible for inhibition of DNA replication, while certain nongenotoxic agents such as phorbol esters increase DNA replication. (6) An attempt to use structure-activity relationship for subfamilies of nongenotoxic carcinogens, e.g., receptor-mediated carcinogens, has been initiated with some promising results. Our results support the idea that there are multiple nongenotoxic mechanisms in carcinogenesis, and that working hypothesis-oriented approaches are encouraged rather than simple screening of chemicals in developing test systems for the detection of nongenotoxic carcinogens.