The idea that mitotic defects that result in tetraploidy promote oncogenesis was proposed over 100 years ago by Theodor Boveri, and although numerous subsequent studies have supported this concept, proving it has been a challenge. In a study published in Nature, David Pellman and colleagues developed a system to directly compare p53-null diploid and tetraploid cells, and have shown that tetraploid cells have greater tumorigenic potential.

To generate genetically matched tetraploid and diploid cells, Pellman's group started out with p53-null mouse mammary epithelial cells (MMECs). Loss of p53 has been previously associated with tetraploidy in various tumour cell types. Cytokinesis was blocked in primary MMECs that were isolated from Trp53+/+ or Trp53−/− mice by treatment with the cytoskeletal disrupting agent dihydrocytochalasin B. Diploid (40 chromosomes) and tetraploid (80 chromosomes) cells were then isolated by fluorescence-activated cell sorting.

The authors observed that the diploid and tetraploid cultures of p53-null cells proliferated at similar rates in culture, but the Trp53+/+ cells did not form tetraploids that could be propagated. With further analysis of the p53-null cells, Pellman's group also found that although whole-chromosome aneuploidy arose in both the diploid and tetraploid populations, a higher percentage of the tetraploid-derived cells developed aneuploidy. Many of these cells also had gross chromosomal rearrangements.

What is the impact of tetraploidy on transformation? Neither the diploid nor the tetraploid p53-null cells were initially able to undergo anchorage-independent growth. However, when both cell types were treated with the mutagen 7,12-dimethylbenz[a]anthracene (DMBA), followed by exposure to the tumour promoter 12-O-tetradecanoylphorbol-13-acetate (TPA), the tetraploid cells alone developed the ability to grow in soft agar. These cells went on to form tumours within two weeks when injected into nude mice. In fact, the p53-null tetraploid cells were able to form tumours in vivo without being exposed to DMBA or TPA — out of 39 animals injected with untreated tetraploid cells, 10 developed tumours at the site of injection. By contrast, no tumours formed at sites where p53-null diploid cells were injected.

Genetic analysis of the tetraploid-derived tumour cells revealed extra centrosomes, numerical and structural chromosome aberrations, as well as abundant non-reciprocal translocations and dicentric chromosomes. Interestingly, each of the 9 tumours examined had amplifications in regions of chromosome 9 that contains a cluster of 10 matrix metalloproteinase genes, which are commonly overexpressed in human breast tumours.

The authors conclude that a transient block in cytokinesis, leading to the formation of genetically unstable tetraploid cells, promotes tumorigenesis in MMECs. However, further experiments are required to determine how often tetraploidy occurs in normal cells and what causes it (see further reading).