THE IN OVO CARCINOGENEICITY ASSAY (IOCA): A REVIEW OF AN EXPERIMENTAL APPROACH FOR RESEARCH ON CARCINOGENESIS AND CARCINOGENICITY TESTING

Harald Enzmann¹ and Klaus D. Brunnemann²

1Bayer AG, Institute of Toxicology, 42096 Wuppertal, Germany and ²American Health Foundation, Valhalla, NY 10595, USA

Received 7/21/97 Accepted 11/14/97

5. Induction of mtDNA damage in ovo

5.1 mtDNA as target for genotoxic chemicals

It is becoming clear that alterations of mitochondrial DNA (mtDNA) may be an alternative and perhaps a superior indicator for genotoxic effects (76). DNA damage may be the result of a direct reaction between a genotoxic chemical and the target DNA molecule. It may also be mediated by free radicals (77). Consequently, mtDNA seems more likely to be a target for the effect of these chemicals than the nuclear DNA (nDNA). In addition, several differences between nDNA and mtDNA may contribute to the higher sensitivity of mtDNA to genotoxic effects. This may, to some extent be due to the close association of nDNA with histone and non-histone proteins that may offer some protection, whereas the mtDNA appears much more exposed to genotoxic effects (78). Damage to nuclear DNA may be efficiently removed by repair systems, but mtDNA has a significantly lower DNA repair capacity (79). Furthermore, the D-loop structure of mtDNA is unique insofar as it contains a single-stranded DNA structure which controls both transcription and replication (80). Since single-stranded DNA is more easily damaged by various genotoxic mechanisms, the D-loop may be a natural hot spot for experimentally induced DNA damage (81). In addition, damage to multiple sites on nDNA may, to some extent, be self-limiting, since the enzymes for DNA replication are coded by nDNA. Severe damage to these genes may result in inactivation of enzymes and the inability of damaged cells to replicate. A similar mechanism may not occur after damage to the mtDNA since the enzymes required for the replication of mtDNA are coded in the nuclear genome (78). In contrast, mutated mitochondria may accumulate in cells and eventually dominate in certain daughter cells of a dividing tissue due to the non-mendelian inheritance of mtDNA (82,83). Another advantage of examining the effect of genotoxic chemicals on mtDNA is its invariability from 16 kb. The supercoiled, relaxed or linearized forms of mtDNA are all detectable by gel electrophoresis (84). Presence of DNA fragments of different sizes most likely reflect the induction of DNA damage. In contrast, during preparation the nDNA (one DNA molecule per chromosome!) usually break, unless very sophisticated methods are used. Consequently, the molecular size of nDNA observed depends on the isolation procedure rather than being reflective of DNA damage.

5.2 Studies of mtDNA using carcinogen-exposed embryonic turkey liver

A disadvantage of mtDNA is that only 0.1% to 1% of total cellular DNA is mtDNA. If methods established for investigations on nuclear DNA are to be used, DNA of about 100 to 1000 fold greater samples are required for experiments with mtDNA as compared with experiments with nuclear DNA. Tissue samples from the IOCA offer a sample size much greater than samples from cell culture systems. The purification of mtDNA from embryonic turkey liver according to the protocol of Welter and coworkers (85) gave mtDNA yields of about 0.2 to 0.4 ng per mg tissue wet weight or up to 0.5 µg mtDNA per liver. The physiologically supercoiled structure of mtDNA was well preserved during the isolation procedure and formed a distinct band in gel electrophoresis. Only small amounts of the mtDNA were of the relaxed form (86).

Separation of native as well as ribonuclease treated samples of mtDNA by gel electrophoresis showed a discrete smear starting at the band of the relaxed form of mtDNA (86).

The electrophoretic separation of mtDNA from turkey embryos previously exposed to hepatocarcinogenic nitrosamines revealed smaller size fragments that could be quantitated by densitometric measurement as smear-band-quotient (SBQ). In experiments with diethylnitrosamine, clear-cut damage to mtDNA was induced over a dose range from 1.24 to 12.4 mmol/kg (86) similar to the concentration range in other in vitro tests for genotoxicity (87,88). Table 4 illustrates the effect of the hepatocarcinogenic nitrosamine N-nitrosomorpholine (NNM) on mtDNA in embryonic turkey liver.