Malignant melanoma, which is both the most dangerous and the rarest type of skin cancer, is a tumour of melanocytes, or pigment-producing skin cells. The genetic causes of this cancer are only now being elucidated. In about 80% of patients, melanoma development is known to be associated with one particular mutation, V600E, in the kinase B-Raf. This mutation activates the kinase, dysregulating the mitogen-activated protein (MAP) kinase signalling pathway and increasing cell division. However, this mutation itself is found also in the tissue of benign naevi, or moles, and so cannot itself be sufficient for melanoma development. Two studies, both led by Leonard Zon of Howard Hughes Medical Institute, Harvard Medical School, Boston, Massachusetts, USA and published in the same issue of Nature, have now shed light on additional molecular mechanisms of melanoma development that interact with aberrant functioning of the B-Raf kinase.

Zon's group uses a transgenic zebrafish model of melanoma in which the B-Raf protein has the V600E mutation and the tumour suppressor protein p53 is missing to probe molecular changes in cancer cells [1]. The first study focused on human chromosome region 1p21, which is known to be often amplified in melanoma cells. The researchers identified genes in this region that are over-expressed in melanoma, cloned these and injected them into their transgenic zebrafish embryos. They found that injection of one gene in this region, SETDB1, significantly accelerated melanoma development. This gene encodes a histone methyltransferase enzyme that methylates one particular lysine residue (K9) on histone H3. Transgenic zebrafish with melanocytes expressing SETDB1 develop pigmentation in patches, rather than the stripes found in the wild type, and melanomas that are more locally invasive than are usual even in this strain.

The researchers then defined a gene set comprising human orthologs of 67 zebrafish genes that were found to be down-regulated in the SETDB1 zebrafish model, and used microarrays to test the relationship between these genes and SETDB1 expression in human melanoma cells. This gene set included genes in the HOX family and other transcriptional regulators; broadly, the expression of these genes was inversely correlated with that of SETDB1. This study identified SETDB1 as an oncogene and implicated it in melanoma development; furthermore, Zon and his co-workers speculate that this arises through changes to patterns of histone methylation.

Melanocytes, the cells in which melanoma arises, originate in the embryonic neural crest. The second study by Zon's group investigated the molecular processes involved in the transformation of neural crest cells into melanoma cells [2]. They found that embryos of the same transgenic zebrafish model as used in the first study, and the melanomas that developed in adult fish of this model, both had gene expression signatures enriched in genes known to be markers of multipotent neural crest stem cells. Many genes in this signature were found to be expressed in neural crest cells at developmental stages after their expression will have ceased in normal embryos, indicating that the BRAF(V600E) mutation maintains these cells in an undifferentiated, stem-like condition. Both zebrafish and human melanoma cells were also found to express some of these genes, which can be assumed to be markers for neural crest cell multipotency.

These results led Zon and his co-workers to propose that small molecule compounds that suppress neural crest progenitors may be effective in treating melanoma. To test this, they screened 2,000 compounds to identify those that inhibit proteins such as Crestin that are characteristic of the progenitor cells. The one non-toxic molecule identified in this screen, NSC210627, was found using chemoinformatics to be similar to brequinar, an inhibitor of the enzyme dihydroorotate dehydrogenase (DHODH). Leflunomide, an anti-arthritis drug that is also an inhibitor of this enzyme, was given to early zebrafish embryos and found to greatly reduce the numbers of all cell types deriving from neural crest progenitor cells.

The enzyme DHODH catalyses a step in the biosynthesis of pyrimidine nucleotides that are necessary for DNA synthesis. The researchers suggested that DHODH inhibitors may act by suppressing the transcriptional elongation of genes involved in normal neural crest development. This hypothesis was confirmed by observing accumulations of incomplete transcripts of these genes in leflunomide-treated neural crest cells. Leflunomide and other DHODH inhibitors were then found to inhibit melanoma growth in vitro and in a mouse model, both alone and in conjunction with a specific B-Raf inhibitor. These results not only shed light on melanoma development but suggest the enzyme DHODH as a novel melanoma target with known inhibitors already in clinical use.

Articles:

[1] Ceol, C.J., Houvras, Y., Jane-Valbuena, J, and 19 others (2011). The histone methyltransferase SETDB1 is recurrently amplified in melanoma and accelerates its onset. Nature 471, 513-515. doi:10.1038/nature09806

[2] White, R.M., Cech, J., Ratanasirintrawoot, S. and 17 others (2011). DHODH modulates transcriptional elongation in the neural crest and melanoma. Nature 471, 518-521. doi:10.1038/nature09882