Tumour cells are characterized by many genomic aberrations, so finding those that are instrumental to tumorigenesis is difficult. Lynda Chin and colleagues have compared the amplifications in different mouse melanoma lines to narrow down the search, and have identified Nedd9 (neural precursor cell expressed, developmentally downregulated 9) as a metastasis gene.

The authors generated two metastatic variants from a non-metastatic mouse model of melanoma. Profiling the genomic DNA of these variants showed that they both had amplifications on chromosome 13 relative to the parental control. The shared region, which was amplified in both lines, was 850 kb and contained 8 genes.

To identify the key gene in the amplified region, expression patterns of these 8 genes were studied in a range of metastatic melanoma and non-transformed melanocyte cultures. Nedd9 was the only gene that was consistently upregulated in the melanomas but not in the melanocytes. This pattern was also found in human melanoma samples.

To validate Nedd9 as a candidate metastasis gene, RNA interference was used to knock down Nedd9 expression in the original melanoma lines. A 90% reduction in expression resulted in up to a 75% reduction in metastatic potential. Conversely, in an HRAS-overexpressing non-metastatic background, overexpression of NEDD9 led to metastasis.

So, how does NEDD9 cause this effect on metastasis? The authors looked at the effect of NEDD9 overexpression on the activity of several known mediators of invasion and metastasis. Only FAK (focal adhesion kinase) showed significant and consistent activation by NEDD9. RNA interference against FAK was sufficient to reverse the effect of NEDD9 overexpression on metastasis, and both proteins were shown to co-localize to dynamic focal adhesion structures.

This research has uncovered an important new gene in metastatic melanoma — Nedd9. It has also demonstrated the power of comparative oncogenomics for finding the genomic aberrations that matter.