By Aaron J. Sender
Dutch biochemist Bas van Steensel credits the less-than-efficient transportation system in Vietnam for the idea that has shaped his academic career: a straightforward way to unearth genomic sequences that regulate gene expression.
“I needed a break,” says van Steensel. He had already earned a PhD from the University of Amsterdam and put in two years as a postdoc at Rockefeller University when he took a six-month trek through the Far East. “Particularly in Vietnam, the bus rides could be very long and boring,” says van Steensel. “So I just started thinking: What are the important questions that are out there in biology?”
While microarrays produce expression profiles that describe which genes are activated in various situations, it is the thousands of proteins called transcription factors that bind with DNA that actually turn the genes on or off. But how these transcription factors flip the switch is poorly understood, mostly because their target sequences in the genome were hard to find. “Wouldn’t it be great,” van Steensel thought during his journey, “if we could sort of attach a stamping machine to each of these proteins that would leave a mark on the DNA?”
The stamp would likely be an enzyme attached to the protein that would somehow modify the DNA only where the factor was bound. But did such an enzyme exist? And would it work? The possible implications were profound. It would mean looking beyond simply what genes are expressed in, say, cancer to the mechanism that actually instigates their expression. “But since I was on the bus in Vietnam I wasn’t able to get to the library and figure out which enzyme would be an efficient stamping machine to mark the DNA,” says van Steensel.
He took his idea to Steve Henikoff, a chromosome chromatin expert at the Fred Hutchinson Cancer Research Center in Seattle. The “stamping machine” they picked was Dam, an enzyme found in E. coli that adds methyl groups to adenine residues. They then fused Dam to a protein with a known genomic target — the centromere — to see if it worked. “The magical moment was when I looked into the microscope and saw those centromeres light up,” says van Steensel. “Then I knew I was in business.”
Further tests with other proteins confirmed they were onto something. Yet while a single transcription factor could bind to dozens or even thousands of targets throughout the genome, the method they came to call DamID could identify only one target at a time. Yet it’s the pattern of transcription factor binding that regulates gene expression. To solve this problem he decided to design DamID to work with microarrays. “That was the next big step,” says van Steensel.
This is how it works: Van Steensel and his colleagues tether each transcription factor with a Dam molecule and send it into a cell to find its corresponding DNA sequences. While the protein docks to its target, Dam marks the spot by methylating the adenines in the surrounding sequence. A restriction enzyme then cuts the DNA at the methylated sites. The researchers then fluorescently label the methyl groups and hybridize the fragments to a microarray. “The spots that light up correspond to sequences in the genome that were methylated,” explains van Steensel. “And therefore the protein must have bound there.”
So far, van Steensel, who now heads his own lab at the Netherlands Cancer Institute, has only tried DamID on Drosophila. Now he is optimizing the technique for mammalian cells and has plans for a public database of transcription factor binding sites genome-wide. Van Steensel is looking at two possible commercial applications: Licensing the method to reagent vendors eager to sell DamID, patented by the Hutch, as a kit or collaborating with pharmaceutical companies on particular disease areas.
Where to for his next lengthy idea-spawning trip? “It would be nice,” he says. “But now that I have my own lab, it’s not so easy.”