Mechanisms of mutagenesis in mammalian cells. Application to human thyroid tumours

Mutations are defined as stable and irreversible modifications of the normal genetic message due to small changes in the number or type of bases, or to large modifications of the genome such as deletions, insertions or chromosome rearrangements. These lesions are due to either polymerase errors during normal DNA replication or unrepaired DNA lesions, which will give rise to mutations through a mutagenic pathway. The molecular process leading to mutagenesis depends largely on the type of DNA lesions. Base modifications, such as 8-oxo-guanine or thymine glycol, both induced by ionizing radiations (IR), are readily replicated leading to direct mutations, usually base-pair substitutions. The 8-oxo-G gives rise predominantly to G to T transversions, the type of mutations found in ras or p53 gene from IR-induced tumors. Bulky adducts produced by chemical carcinogens or UV-irradiation are usually repaired by the nucleotide excision repair (NER) pathway which is able to detect structural distortion in the normal double-strand DNA backbone. These lesions represent a blockage to DNA and RNA polymerases as well as some signal for p53 accumulation in the damaged cell. In the absence of repair, these lesions could be eventually replicated owing to the induction of specific proteins at least in bacteria during the SOS process. The precise nature of the error-prone replication across an unexcised DNA lesion in the template is not fully understood in detailed biochemical terms, in mammalian cells. IR basically produce a very large number of DNA lesions from unique base modifications to single- or double-strand breaks and even complex DNA lesions due to the passage of very high energy particles or to a local re-emission of numerous radicals. The breakage of the double-helix is a difficult lesion to repair. Either it will result in cell death or, after an incorrect recombinational pathway, it will induce frameshifts, large deletions or chromosomal rearrangements. Most of the IR-induced mutations are recessive ones, requiring therefore a second genetic event in order to exhibit any harmful effect and a long latency period before the development of a radiation-induced tumor. The fact that IR essentially induced deletions and chromosomal translocations renders very difficult the use of the p53 gene as a marker for mutation analysis. In agreement with the type of lesions induced by IR, it is interesting to point out that the presence has been observed, in a vast majority of radiation-induced papillary thyroid carcinomas (PTC), of an activated ret proto-oncogene originated by the fusion of the tyrosine kinase 3' domain of this gene with the 5' domain of four different genes. These ret chimeric genes which are due to intra- or inter-chromosomal translocations, were called RET/PTC1 to PTC5. The RET/PTC rearrangements were found in PTC from children contaminated by the Chernobyl fall-out as well as in tumours from patients with a history of therapeutic external radiation, with a frequency of 60-84%. This frequency was only 15% in 'spontaneous' PTC. The type of ret chimeric gene predominantly originated by the accidental or therapeutic IR was different. Indeed, PTC1 was present in 75% of the tumours linked to a therapeutic radiation and PTC3 in 75% of the Chernobyl ones. The other forms of RET/PTC were observed in only a minority of the post-Chernobyl PTC (< 20%). The difference in the frequency of PTC1 and PTC3 in both types of PTC, is statistically significant (P < 10(-5), Fischer's exact test). In two of the post-therapeutic radiation PTC, RET/PTC1 and PTC3 were simultaneously present. A PTC1 gene was also observed in 45% of the adenomas appearing after therapeutic radiation. The long-period of latency between exposure to IR and the appearance of thyroid tumours is probably due to the conversion of a heterozygote genotype of IR-induced mutations to a homozygote one. It will be interesting to use this time lag in accidental or therapeutic-irradiated p