Moles, Melanomas and Mouse Models

By Jeffrey Norris on April 22, 2009
A melanoma may start out looking just a little different from a normal mole. But it may end up being every bit as deadly as cancers that first arise in vital organs. When melanoma is diagnosed late, the prognosis is bad. Better treatment is a must. A recent study completed in the lab of UCSF researcher Martin McMahon, PhD, published online in March in Nature Genetics, may lead to finding a smoother route for promising new drug candidates to treat melanomas. Many genes can be abnormal in cancers. It is not always clear which are vital to tumor growth. While some of these genes already are being targeted with new drugs in clinical trials, the value of targeting other abnormal genes in particular cancers awaits better evidence. McMahon and Marcus Bosenberg, MD, PhD, of Yale University now have shown the value of targeting two abnormal genes that are found together in about one-third of melanomas. They proved that these genetic abnormalities – loss of a gene called PTEN and excess activity of another gene called BRAF – really do cause melanoma in mice. “This nails it down,” McMahon says. “These are the relevant targets that are driving a subset of human melanomas. The mice are telling us something important about the pathogenesis of the disease.”

Unique Drug Combo Shrinks Melanomas

The researchers showed that preventing the abnormalities prevents melanoma. They also showed that treating the mice with existing drugs – never before tested in combination – is better than treatment with either drug alone in reducing tumor size. McMahon and Bosenberg have achieved a kind of tour de force of mouse engineering. They did so with some clever genetic manipulations to turn key genes on and off at will. McMahon suggests that success in the fight against cancer will require a public sector-private sector partnership. His melanoma research has been sponsored by just such a partnership – between Genentech, UCSF and a state funding mechanism called UC Discovery. For their part, academic scientists have traditionally made breakthrough discoveries about basic mechanisms of disease. They examine the interplay among the misbehaving molecules involved. Sometimes they experiment on cells growing in a lab dish to gather evidence for why a molecule ought to be a good target for drugs to treat human disease. There are even pharmaceutical chemists in academia who have made drug prototypes, as well as researchers who find out what experimental prototype drugs will do in an animal. Most of the time, that animal is a lab mouse. But the expense of testing drugs in animals, even small mice, mounts quickly. The researchers may run out of research money or return to other, more basic research without attracting much interest from companies that develop drugs for human treatments. Traditionally, cancer drug testing is conducted by injecting human cancer cells into mice. The mice that are used have a suppressed immune system to prevent rejection of the human cells by the mouse immune cells. The injected human cancer cells form a tumor, permitting responses to cancer drugs to be readily measured. Although this is a convenient way to conduct cancer drug testing, results obtained by this technique have been poorly predictive of the future success of cancer drugs when tested in human patients, McMahon says. Pioneering work by UCSF colleague Doug Hanahan, PhD, convinced McMahon that the more sophisticated, genetically engineered mouse models of human cancer would be a better way to understand cancer initiation, progression and, most important, therapy. Making these mouse models is still extraordinarily time-consuming – McMahon and Bosenberg spent five years engineering exactly what they wanted into this uniquely programmed strain. Moreover, mice are expensive. And just like humans, they are not all stricken with cancer at the same time. One has to wait. But studying this strain of mice will allow McMahon to learn more about what makes tumors grow, provide more convincing evidence for the validity of drug targets and allow more efficient testing of new candidates to be the next targeted cancer drugs.

Turning a Tumor Gene On and Off at Will

BRAF is very commonly overactive in other cancers, including colon, thyroid and lung cancer. McMahon and his collaborators are using his genetically engineered mice to learn more about drug targets in these cancers too. With genetic switches in the mice activated by administering a hormonal drug, he says, “We can turn on oncogenic BRAF wherever and whenever we want.” David Dankort, PhD, a former postdoctoral fellow in McMahon’s lab who is now on the faculty of McGill University, was a major contributor to the Nature Genetics study.

BrafV600E Cooperates with Pten Loss to Induce Metastatic Melanoma

David Dankort, David P. Curley, Robert A. Cartlidge, Betsy Nelson, Anthony N. Karnezis, William E. Damsky Jr, Mingjian J. You, Ronald A. DePinho, Martin McMahon and Marcus Bosenberg

Nature Genetics (Published online March 12, 2009)

Related Links