The Function of DNA Form
According to a new computational analysis of DNA structure, variations in DNA shape—along the grooves of the double helix—may play an important role in defining how the genome works. The analysis revealed that six percent of the DNA ladder’s shape is conserved across a range of different mammals—even though the sequences that produce those conserved shapes could vary.
“We’ve found a new way that evolutionary selection is working in the human genome, beyond just preserving the strict sequence of nucleotides,” says Tom Tullius, PhD, chemistry professor at Boston University and one of the authors of the report, published April 17 in the journal Science. “I hope that this finding will open up some new ways of thinking about how the genome works. It’s more than just a collection of letters.”
A 2007 study by the ENCODE (Encyclopedia Of DNA Elements) research consortium hinted that something other than nucleotide sequence was at play in determining genome function. Looking at one percent of the human genome, the researchers found that only about half of the known functional regions (for example, sections of DNA where proteins bind) showed sequence conservation across a range of mammals (from mouse to human). “We were struck by the fact that you may not be looking at the complete story if you only look at sequence conservation to define function,” Tullius says.
Tullius and his colleagues wondered if shape might be a factor. They had previously discovered, experimentally, that different DNA sequences can have similar structures. Using the reactive hydroxyl radical molecule, they had probed for subtle differences in DNA shape. Small variations in the radical’s accessibility to the DNA yield a detailed structural map. These variations are often in the DNA’s minor groove width, which can range from four angstroms at the narrowest to 11 at the widest, Tullius says. This finding led them to wonder whether sequences could diverge through evolution while form remained the same.
To answer that question, Tullius, Elliott Margulies, PhD, of the National Institutes of Health, Steve Parker of Boston University and their colleagues created a computer program called Chai. The program compares computational predictions of DNA shapes from the same one percent of the human genome studied by ENCODE, and other mammalian genomes. They found that certain parts of the genome are conserved solely by structure, not sequence. Moreover, the combination of sequence and shape conservation almost entirely covers the functional sites identified by the ENCODE study. Tullius and his colleagues also found that polymorphisms associated with disease are more likely to cause structural changes in DNA than neutral polymorphisms—meaning that these shape changes could be disrupting the binding of some essential protein.
In a 2007 report in the journal Cell, Barry Honig, PhD, of Columbia University, had concluded that DNA shape influenced the binding of a homeodomain protein to developmental genes. “The combination of these two studies makes it clear that DNA shape is important in function,” Honig says. “This gives us a new avenue to study how DNA functions that we didn’t have before.”