Brain tumor study reveals why treatment efforts fail in genetic disorder

Drugs used to treat the tumors common in people with a disorder called neurofibromatosis 1 rarely work, and scientists now know why. The chemotherapy drugs target a group of related proteins, call RAS proteins, which are thought to be responsible for these tumors. But researchers at Washington University School of Medicine in St. Louis found that the disease affects only one member of the protein family, and it happens to be the one form of RAS that does not respond well to these particular treatments.

The study, which will appear in the Jan. 1 issue of the journal Cancer Research, suggests where researchers should now look for more promising approaches to treating neurofibromatosis tumors, and may help scientists understand other cancers related to RAS. “The downside is our study proves we’re not using the right therapies for this particular problem,” says principal investigator David H. Gutmann, M.D., Ph.D., the Donald O. Schnuck Family Professor of Neurology and professor of genetics and of pediatrics. “But there’s a chance to make lemonade out of this lemon: We now have a rational reason for why these drugs aren’t working, so we should be able to explore new, more effective treatment options.”

About one in 4,000 newborns has neurofibromatosis 1, in which every cell in the body has one normal and one mutated copy of a gene called NF1. If a cell’s normal copy also is mutated, tumors can form. Children with neurofibromatosis 1 are therefore predisposed to developing a variety of serious complications as they grow older, including skin, spine and brain cancers.

Scientists previously found that RAS proteins become overly active when both copies of the NF1 gene are abnormal in tumors from patients with neurofibromatosis 1. So physicians have tried treating these tumors with drugs that prevent RAS activity. Unfortunately, the results have been disappointing.

To understand why, Gutmann’s team examined whether all forms of RAS are overly active in mouse cells lacking both copies of the Nf1 gene. They specifically examined support cells in the brain called astrocytes, which often are affected by neurofibromatosis 1. Surprisingly, only one member of the protein family, K-RAS, was significantly affected, suggesting it is an important factor in this disease.

Moreover, when the team activated K-RAS in normal astrocytes, the cells developed many of the same characteristics and activities as those lacking Nf1. For example, both types of abnormal astrocytes were round and dense, grew and multiplied at a similar rate and moved around more than normal. They also discovered they could reverse abnormalities in cells without Nf1 by decreasing K-RAS activity.

K-RAS activation also mimicked Nf1 loss in live mice. Gutmann’s team previously discovered that mice without Nf1 genes in their astrocytes grow an abnormally large number of astrocytes in their brains, but they don’t develop tumors unless all other brain cells are missing at least one copy of the gene. In this study, the researchers found that K-RAS follows a similar pattern: When the protein was overly active in astrocytes of mice with two normal copies of Nf1, the cells multiplied but did not develop into tumors; however, tumors did form when K-RAS was activated in astrocytes of mice lacking one copy of Nf1 in all cells.

Another form of RAS previously suspected to be linked to neurofibromatosis, called H-RAS, did not mimic loss of the Nf1 gene in tissue culture or in live animals. “Collectively, these results suggest that K-RAS activation, specifically, is the biological equivalent of Nf1 loss in astrocytes,” Gutmann says. “If we can understand what K-RAS does that’s unique, we should be able to develop targeted therapies.”

The research team already has made progress toward that goal. Too much RAS and too little Nf1 are both known to result in a cascade of events, including activation of another protein called Rac1, which in turn activates LIM kinase. Gutmann and his colleagues found that that effect could be mimicked in normal astrocytes by selectively activating K-RAS. Activating H-RAS did not trigger the cascade. “Though K-RAS doesn’t respond well to available chemotherapy drugs, one of the proteins it interacts with might,” Gutmann says. “By showing that K-RAS activates a pathway that is unique from the pathways activated by other RAS molecules, our findings may lead us to a variety of better treatment targets.”