In a new study released, researchers have shown that a specific protein plays an important role in inhibiting the development and spread of melanoma tumours in mouse and in human skin models. Increased expression of the protein, SOX9, may also decrease the resistance of melanoma cells to retinoic acid, which is used to treat several other types of cancer. The ability to increase sensitivity to retinoic acid by stimulating SOX9 expression could lead to new approaches for treating melanoma and other cancers. The study, led by researchers at the National Cancer Institute (NCI), part of the National Institutes of Health, appeared in The Journal of Clinical Investigation.

Melanoma, which begins in cells called melanocytes, is the most deadly form of skin cancer. Current approaches to treating melanoma include the use of conventional therapies, such as radiation therapy and chemotherapy, or stimulating the immune system to inhibit cancer growth.

"Unfortunately, in most cases, those approaches fail to cure patients," said lead author Thierry Passeron of NCI's Center for Cancer Research. "For some blood cancers, forms of retinoic acid have provided effective treatments. However, most solid cancers, including melanoma, are resistant to retinoic acid."

The SOX9 protein is a transcription factor expressed in various adult tissues, including the brain, heart, and kidneys. Transcription factors control the expression of genes and, therefore, function as key regulators of important biological processes. In previous work, the researchers demonstrated that SOX9 protein plays a role in regulating the differentiation, or acquisition of mature characteristics, of normal melanocytes, as well as in inhibiting the proliferation of human melanoma cells. Other work with laboratory-grown cancer cells suggests that retinoic acid may play a role in the antiproliferative effects of SOX9.

In the study, the researchers first looked at the expression of SOX9 protein in normal human skin samples and in samples of nevi (precancerous tissue), primary tumours, and metastatic melanoma tumours that had spread to other tissues. They found that the expression of SOX9 is highest in normal cells and is progressively reduced as cells transition from the precancerous state to the most advanced stages of cancer. Next, the researchers inserted the SOX9 gene into human melanoma cells. Using reconstructed skin samples, which contained the human melanoma cells that either had or did not have the inserted gene, the researchers found that cells without the inserted SOX9 gene formed tumours, whereas cells containing the added gene did not. Similar findings were observed in mice that were injected underneath the skin with melanoma cells that either had or did not have the inserted SOX9 gene.

The researchers next determined whether melanoma cells with the inserted SOX9 gene were sensitive to retinoic acid. Indeed, when exposed to retinoic acid, these cells showed a dramatic decrease in proliferation compared to cells without the inserted gene. The sensitivity to retinoic acid appeared to be a consequence of reduced expression of a protein called PRAME, which is a known repressor of the receptor that binds retinoic acid in melanoma cells.

The researchers next investigated whether the SOX9 protein could be activated or its expression increased in melanoma cells. Prostaglandin D2 (PGD2), a naturally occurring molecule in the body, had previously been shown to activate SOX9 protein in human testicular cancer cells, and the researchers found that melanoma cells exposed to PGD2 also had enhanced SOX9 activity. They treated mice with induced melanoma tumours with a substance that simulates the PGD2 biochemical pathway alone, retinoic acid alone, or a combination of the two agents. They found that the combination treatment decreased the size of tumours in the mice to a greater extent than treatment with either single agent alone.

"This study brings new insight into the cellular changes that occur during the development of melanoma and suggests a new therapeutic approach," said senior author Vincent Hearing of NCI's Center for Cancer Research. "We will also use our models to determine the optimal combinations of agents and forms of retinoic acid to stimulate SOX9. The next clinical steps will be phase-I and phase-II studies in humans, as our ultimate goal is to treat people affected by this public health burden."