A new issue of Science is going to be coming out later today, but I'm going to take a quick look at a paper that appeared last week that is both interesting in and of itself, and received a typical (and unfortunate) treatment in the press. The paper attracted that interest in the first place because of its subject: it seems to tell us a little something about what makes us human.
The new study was based on a past analysis that identified sequence of interest in the human genome. This is harder than it might sound; line up the same sequences in a human and another primate, and well over 90 percent of the bases will be identical. What researchers have to do is figure out which sequences are important first by comparing primate genomes with more distantly related species, like mice and chickens. Once you have a set of sequences that are conserved in nonhuman primates and more distant species, you can go back and look at the human version—if it has a lot of differences in the human lineage, then there's a chance that those differences contribute to some of humanity's distinctive features.
Once the genomes are done, that analysis only needed computing power, so it was performed a few years back. Now, researchers are in the midst of doing the more challenging follow up work. The recent paper focuses on a sequence, called HACNS1, which is about 550 bases long. If it were picking up mutations at random, the human version would be expected to have four; instead, it has 13 differences with the chimp sequence.
HACNS1 lies outside of any known genes, suggesting it is regulatory DNA. The researchers hooked it up to a gene that couldn't otherwise be expressed and injected the DNA into mice. As expected, the regulatory function caused the gene to be expressed in a very specific pattern in the head and limbs. The key result came when the chimp version was hooked up to the same gene and injected into mice—the limb expression was severely reduced or absent.
The obvious inference here is that the sequence drives human-specific gene expression, specifically in the areas that form the hands and feet, which are obviously quite distinct in humans. The researchers also made two constructs that had only six of the 13 changes from chip to humans and found that these drove expression was somewhere in between the humans and the chimp sequences, which is exactly what you'd expect from a gradual, evolutionary change.
There's still work to be done, as HACNS1 lies about halfway between two different genes, and the researchers don't know which of the two it regulates. Still, it's a pretty exciting result, and worthy of the attention it received.
If only that attention was of a somewhat higher quality. The press has an annoying habit of calling any DNA that doesn't code for a protein "junk DNA." It's not; junk DNA is something different entirely. Regulatory DNA hasn't been considered junk for years; the concept is so old and thoroughly established that seminal work on the Lac operon's regulatory DNA appeared in my high school text book in the 1980s.
But the error is worse in this case, because HACNS1 was chosen specifically because it doesn't look like junk. The sequence in the area is conserved from humans to chickens, which shared a common ancestor that predates the dinosaurs. That conservation of sequence positively screams "this DNA is important." I realize headlines such as "junk makes us human" are hard to resist but, for the sake of a modicum of accuracy, more writers should have done so.
Science, 2008. DOI: 10.1126/science.1159974Posted on