Novel pathway of centrosome amplification that does not require DNA lesions

Centrosome amplification (also known as centrosome overduplication) is common in cancer cells and can be induced by DNA damaging agents. However, the mechanism and significance of centrosome amplification during carcinogenesis or after DNA damage are not clear. Previously, we showed that centrosome amplification could be induced by 3-aminobenzamide (3-AB), an inhibitor of poly(ADP-ribose) polymerases (PARPs) in mouse embryonic fibroblasts. In this paper, we determined if the effect of 3-AB on centrosome amplification was dependent on DNA damage in CHO-K1 cells. We used the well-known mutagen/carcinogen N-methyl-N'-nitro-N-nitrosoguanidine (MNNG). Ten micromolar MNNG and 10 mM 3-AB induced significant centrosome amplification in 18.1 ± 1.1% and 19.4 ± 1.8% of CHO-K1 cells, respectively, compared to 7.0 ± 0.5% of untreated CHO-K1 cells. AG14361, another potent inhibitor of PARPs, also induced centrosome amplification. We then used γ-H2AX analysis and alkaline comet assays to show that 10 μM MNNG induced a significant number of DNA lesions and cell cycle arrest at the G(2) /M phase. However, 10 mM 3-AB neither induced DNA lesions nor altered cell cycle progression. In the umu test, 10 μM MNNG was mutagenic, but 10 mM 3-AB was not. In addition, 10 μM MNNG induced significant accumulation of ataxia telangiectasia mutated protein in the nuclei, but 10 mM 3-AB did not. Thus, we found no association between apparent DNA lesions and centrosome amplification after 3-AB treatment. Therefore, we propose the presence of a novel pathway for centrosome amplification that does not necessarily require DNA lesions but involves regulation of epigenetic changes or post-translational modifications including polyADP-ribosylation.