A team of Columbia scientists have discovered two genes that, when simultaneously activated, are responsible for the most aggressive forms of human brain cancer [Nature Online, Dec. 23, 2009].

This finding was made possible by the assembly of the first comprehensive network of molecular interactions that determine the behavior of these cancer cells, a map so complex and elusive that, until now, it could not be constructed. The discovery may lead to completely novel strategies to diagnose and treat these incurable tumors.

The research was conducted by a team of Columbia scientists led by Antonio Iavarone, MD, associate professor of neurology in the Herbert Irving Comprehensive Cancer Center, and Andrea Califano, PhD, director of the Columbia Initiative in Systems Biology.

The researchers studied a type of human malignancy, called glioblastoma multiforme, which is among the most lethal because it rapidly invades the normal brain producing inoperable brain tumors. Recently, glioblastoma claimed the life of Senator Edward Kennedy only sixteen months after diagnosis.

Before this study, cancer researchers had little idea why glioblastoma is so aggressive. "We now know that two genes - C/EPB and Stat3 - are the disease's master 'control knobs'," said Dr. Iavarone. "When simultaneously activated, they work together to turn on hundreds of other genes that transform brain cells into highly aggressive, migratory cells."

The two genes are active in about 60 percent of all glioblastoma patients and help identify poor-prognosis patients. All patients in the study whose tumors showed activation of both factors died within 140 weeks after diagnosis, while one half of the patients without these factors were still alive.

"The finding means that suppressing both genes simultaneously, using a combination of drugs, may be a powerful therapeutic approach for these patients, for whom no satisfactory treatment exists," said Dr. Califano.

This approach, called combination therapy, is supported by this study since silencing both genes in human glioblastoma cells completely blocked their ability to form tumors when injected in a mouse. Based on these results, the Columbia scientists received a grant from the American Recovery and Reinvestment Act to develop drugs specifically aimed at these genes.

The Columbia team, which includes physicists and biologists, has for the first time assembled and experimentally validated such a cellular network for a glioblastoma cell, a hugely complex challenge that required several novel approaches drawn from the fields of information theory and computational biology.

"Armed with such a blueprint of the cell machinery, we can now ask pointed questions, such as which genes are responsible for the most deadly features of these tumors," said Dr. Iavarone.

From this blueprint, produced in Dr. Califano's lab, the scientists pinpointed the tumor's two master regulator genes. Experiments conducted by Dr. Iavarone in brain cancer cells and mice then confirmed the accuracy of the network and the importance of the two genes.

 "The identification of C/EPB and Stat3 came as a complete surprise to us, since these genes had never been implicated before in brain cancer," said Dr. Iavarone. Based on traditional approaches, their critical role may have eluded researchers for a long time.

"From a therapeutic perspective, it means we are no longer wasting time developing drugs against minor actors in brain cancer," he said. "We can now attack the major players."