A painstakingly constructed network map of yeast’s molecular responses to DNA damage is providing a glimpse into how cells might thwart cancer. Since most cancer is caused by the failure of cells to repair DNA damage due to aging, toxins, radiation, and other exposures, researchers can now turn to this new map to help pinpoint which cancer networks to target for therapeutics.
The map, which documents interactions between genes and regulatory proteins involved in repairing DNA, was prepared by Trey Ideker, assistant professor of bioengineering at the University of California, San Diego (see “Comparative Interactomics”). Ideker’s group exposed yeast cells to a DNA-damaging substance and carefully mapped the ensuing interactions between regulatory proteins and genes.
[For a diagram of the complex reactions of yeast cells to DNA damage, click here.]
The map represents an early success for the young field of interactomics, a branch of systems biology in which researchers attempt to understand cellular processes in all their complexity by examining the networks of interaction between genes, proteins, RNA, and other molecules.
Systems biology “tries to look at the complexity of diseases not in a reductive way, but by encompassing all components to get a comprehensive understanding [of their molecular basis],” says Dan Gallahan, chief of the National Cancer Institute’s Integrative Cancer Biology Program. “Cancer is so complex, and we’ve learned the hard way that we have to have a systems approach.”
Ideker and colleagues at MIT and the Whitehead Institute started with 30 proteins whose activity seemed to be associated with DNA damage response. These proteins were transcription factors – regulatory proteins that can bind to genes and turn them on and off. Ideker used a new protein-analysis technology called ChIP-chip to uncover which genes the transcription factors bound to after being exposed to a mutagen. This analysis turned up more than 5,200 candidate genes.
Most large-scale systems biology studies would stop at this point. But “the problem with these technologies when you’re looking at very many genes is that 50 percent of the time they’re wrong,” says Ideker. “Even when the proteins really bind, that doesn’t mean anything happens,” he says. “It could just be molecules bumping around. One molecule will kiss another and it’s completely inconsequential.” Ideker knew that not all 5,272 genes bound by the transcription factors would go through the transcription process and generate proteins – so he spent four years narrowing the field. His final wiring diagram shows a web of interactions between the proteins and 82 genes (see accompanying illustration).
Ideker’s study confirmed some of what was suspected about DNA damage: for example, that it activates genes not directly involved in DNA repair themselves. More importantly, though, says Gallahan, “by doing this kind of analysis you actually raise questions, you don’t answer them. This proposes a lot of new hypotheses that can be tested.” Gallahan was particularly struck by the involvement of genes that create and break down lipid molecules. The involvement of these pathways “would not have been identified by serendipity. They came forth because of this careful work,” says Gallahan.