Most human melanomas have mutations in the serine/threonine kinase BRAF, but the role of this enzyme in tumorigenesis is not well understood, partly because there has been no animal model with the relevant mutations. E. Elizabeth Patton and colleagues have used zebrafish to show that BRAF interacts genetically with the p53 pathway to cause melanomas.

The zebrafish's characteristic stripes are created by melanin, and the fish's melanin-producing cells are remarkably similar to melanocytes in mammals — they derive from the same embryonic lineage (the neural crest) and at least one of the genes controlling melanocyte differentiation is homologous in mice and zebrafish. Patton and colleagues surmised that if the normal growth and development of melanocytes is evolutionarily conserved among vertebrates, then their dysfunctional growth might be too.

They microinjected zebrafish embryos at the one-cell stage with the human BRAF gene containing the most common activating mutation (V600E) under the control of the melanocyte-specific mitfa promoter. About 10% of the resulting fish developed mole-like foci of proliferating melanocytes, termed f-nevi. The melanocytes in the f-nevi were well differentiated and appeared not to invade the surrounding tissue. So although the activating BRAFV600E mutation is sufficient to cause expansion of the melanocytes, further mutations seem to be required for progression to melanoma.

Human melanomas often lack the CDKN2A locus, in addition to mutation of BRAF. This locus encodes the tumour-suppressor protein ARF, which functions in the p53 pathway. Although p53 mutations are rare in melanomas, inactivation of this pathway has been linked to their formation. Patton and colleagues speculated that a deficiency in the p53 pathway might be the extra push that drives BRAF-mutated melanocytes into forming invasive tumours.

Injecting mitfa-BRAFV600E into embryos homozygous for a mutation in TP53 resulted in about 6% of the fish developing malignant melanomas by 4 months of age. The melanoma cells expressed melanin and were morphologically similar to human melanoma cells. The cells spread rapidly following transplant into wild-type zebrafish (unlike cells from the f-nevi), and 89% of them showed chromosomal abnormalities similar to those seen in human melanomas.

With this zebrafish model of melanoma, Patton et al. demonstrate that it is feasible to use these animals to explore the genetic interactions that cause malignant phenotypes, and provide evidence that BRAF mutations do indeed underlie this disease.