Next top model

By Elie Dolgin Next top model A mouse lung riddled with tumors after inhaling Cre recombinase. Courtesy of David Dankort David Dankort was 4 years into his postdoc at the University of California, San Francisco, without a single paper to show for his work since his PhD. His first two major projects had failed, and if his third experiment didn’t pan out, he was ready to kiss his academic career goodbye. In a last-ditch effort, Dankort had constructed tra

Elie Dolgin
Sep 1, 2009

Next top model

A mouse lung riddled with tumors after inhaling Cre recombinase.
Courtesy of David Dankort

David Dankort was 4 years into his postdoc at the University of California, San Francisco, without a single paper to show for his work since his PhD. His first two major projects had failed, and if his third experiment didn’t pan out, he was ready to kiss his academic career goodbye. In a last-ditch effort, Dankort had constructed transgenic mice that could be induced to form cancer by activating a particular enzyme. In early 2005, as part of the final test of the experiment, he sedated his mice, placed a solution containing a virus expressing the enzyme under their noses, and waited a couple of weeks. “I’m waiting, I’m waiting, and I start to think: ‘Am I imagining or are these mice ill?’” Dankort recalls. He weighed the mice, and found that they...

Dankort, now an assistant professor at McGill University in Montreal, started working on his model in 2003 in the hopes of recapitulating the most common cause of lethal human skin cancer: malignant melanoma. Dankort didn’t want to build any old mouse model, though. He wanted to ensure that his transgenic mice were hit with cancer-causing mutations in the same way as people in the real world. “When you’re trying to model human cancers in mice you should at least stack the deck so that things are set up as closely as possible to the human condition,” Dankort says.

Dankort focused on a single base-pair substitution in codon 15 of the then-understudied oncogene BRAF. He built an inducible mutant that maintained BRAF’s normal sequence in early life, but could adopt an alternate, mutated form later in life, which in turn altered the BRAF protein. (Dankort swapped the normal and mutant versions by activating an enzyme called Cre-recombinase, which catalyzed site-specific recombination of DNA.) Although his primary interest was melanoma, as a first test, Dankort tried inducing lung cancer in his mice. “I needed to see if they worked,” he says.

The BRAF-deficient mice all produced white, lumpy lesions on their lungs, yet intriguingly, these tumors were all benign (Genes Dev, 21:379–84, 2007). When Dankort knocked out either of two tumor suppressor genes, however, the mice developed “massive and quick” lung cancers, he says, indicating that BRAF mutations alone were insufficient to cause full-blown cancer.

Dankort then turned his attention to melanoma. He modified his mice so that he could activate the Cre-recombinase in melanocytes simply by painting the mice’s skin with a particular chemical. Again, the mice produced benign lesions—little, black moles—that only progressed to melanoma when combined with tumor suppressor gene silencing (Nat Genet, 41:544–52, 2009).

Dankort wasn’t the only one to model BRAF ’s role in melanoma, however. Within a span of 6 weeks, two other groups published reports also modeling the exact same mutation. In one paper, researchers at Tufts Medical Center in Boston, Mass., reported a more traditional model, with the mutated BRAF gene behind a foreign promoter (Oncogene, 28:2289–98, 2009). The other study, led by Richard Marais of the University of London’s Institute of Cancer Research, used Cre-recombinase in “more or less exactly the same” way as Dankort did, Marais says (Cancer Cell, 15:294–303, 2009). “This is the best way to build a model and I guess we each independently came up with it.” But unlike Dankort’s mice, Marais’ BRAF mutants developed locally invasive melanomas even without the loss of tumor suppressor genes. Marais chalks up this difference to genetic background effects, whereas Dankort says that discrepancies in treatment regimes and knock-in constructs are also potential explanations.

Dankort and Marais’s models together provide “a huge advance in the field of melanoma preclinical science,” says David Tuveson of the Cancer Research UK Cambridge Research Institute. “However, there is much work to be done to clarify which model is going to be the most informative about melanoma or if they are both going to be useful for different reasons.”

Interested in reading more?

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?