Chromosomal events in carcinogenic initiation and promotion: implications for carcinogenicity testing and cancer prevention strategies.

We have provided experimental evidence in favour of the hypothesis that carcinogenesis is triggered by at least two chromosomal events, which must occur in a single diploid somatic cell in a specific time sequence: (i) specific recessive mutational or epigenetic chromosomal change(s) resulting in a heterozygous (m/+), latently premalignant state (initiation); this must be followed by (ii) a chromosomal rearrangement involving the affected locus, and leading to homozygosity (m/m) or hemizygosity (m/o), and subsequent expression of the recessive malignant character (promotion). The complete carcinogen, MNNG, induced mutations (6-thioguanine-resistance), chromosomal rearrangements and SCEs in V79 Chinese hamster cells. TPA, a potent tumour promoter, induced only SCEs and specific chromosomal effects. Antipain, a protease inhibitor and a known inhibitor of both carcinogenesis and tumour promotion, inhibited only the MMNG-induced chromosomal rearrangements (but not mutagenesis and SCEs) and the TPA-induced chromosomal events. These results suggest that (1) both TPA-induced and MNNG-induced chromosomal rearrangements are caused by the activation or induction of mitotic recombination and hence appear to be preventable; (2) chromosomal rearrangement is a rate-limiting step in carcinogenesis; and (3) if mutagenesis is involved in carcinogenesis, it is probably not sufficient. The existence of the human cancer-prone syndromes, Bloom's, Fanconi's anaemia and ataxia telangiectasia, which involve spontaneous chromosomal rearrangements analogous to those induced by carcinogens in normal cells, strongly supports our hypothesis that carcinogenesis involves two chromosomal events. We discuss the implications of this work to carcinogenicity testing and cancer prevention strategies.