Cancer genetics: Looking at the big picture

signaling update home

SOURCE:

A new, large-scale mutagenesis screen in different mouse backgrounds identifies candidate oncogenes and tumor suppressors, and reveals the networks that link them together.

We have plenty of lists of mutations that are associated with cancers, but little data on how these mutations interact with each other. A new, large-scale mutagenesis screen in different mouse backgrounds identifies candidate oncogenes and tumour suppressors, and reveals the networks that link them.

It is thought that the main function of the tumour suppressor ARF (encoded by Cdkn2a) is to activate p53, but there is evidence to suggest other functions. Therefore, the authors used retroviral insertional mutagenesis in Arf- and Trp53-deficient, as well as wild-type, mice to identify genes that interacted with one of these tumour suppressors but not the other. A key component of their strategy was ensuring that they identified a large proportion of the mutations in each tumour, rather than just a small fraction.

From the 510 tumours in 455 mice, the authors identified 10,806 independent insertions, and 346 loci in the genome were mutated significantly more often than expected by chance. These 'common insertion sites' included established oncogenes such as Myc, but also genes that had not previously been associated with cancer, such as the Notch receptor modulator Lunatic fringe. Of particular interest were the mutations that only occurred in one or two of the three genotypes. For example, several genes that are implicated in p53 signalling were mutated in Arf-deficient and wild-type backgrounds, but not in a Trp53-deficient one, implying that mutations in these genes are selected because they reduce p53 activity, or abrogate p53-dependent effects, something that is not necessary in a p53-null background.

In order to build a network of collaborative and mutually exclusive interactions, the authors looked at all possible pairs of their mutations and assessed whether they occurred together in the same tumour more or less often than expected by chance. The resulting network contained some interactions that confirmed previous associations in the literature and many that could be explained on the basis of the biochemical properties of the proteins involved. In addition, when the authors looked more closely at some individual genes such as Notch1, they found several different types of mutation, each of which affects gene function in a different way and is associated with a different set of mutations in other genes.

The authors have made their results available as an online resource (http://mutapedia.nki.nl). This will allow all researchers to take part in the process of validating these interactions and investigating specific ones in more detail. It is hoped that the number of such resources will grow.

Patrick Goymer

References

Uren, A. G. et al. Large-scale mutagenesis in p19^ARF- and p53-deficient mice identifies cancer genes and their collaborative networks. Cell 133, 727–741 (2008) | Article | PubMed |