The function of this region of DNA is still unknown, but it appears to be directly involved in the development of the human brain. "It's very exciting to use evolution to look at regions of our genome that haven't been explored yet," said Howard Hughes Medical Institute investigator David Haussler, the leader of the team that included scientists from the University of California, Santa Cruz, the University of California, Davis, the University of Brussels, and Université Claude Bernard in France.
Their article will be published in an advance online publication on August 16, 2006, in the journal Nature.
Haussler's group found the DNA region using a technique developed by Katherine Pollard, a former postdoctoral fellow in Haussler's lab who is now an assistant professor at the University of California, Davis. Pollard compared the DNA sequences of chimps, mice, and rats to find the regions that had remained largely unchanged over the 80 million years or so since the common ancestor of those organisms. She then examined the same regions in humans to identify those that had changed markedly in the 6 million years since humans and chimps diverged from a common ancestor.
"Some DNA regions have hardly changed at all over many millions of years in most species," said Pollard. "My twist was to look for the subset of these regions that have changed just in humans."
Forty-nine regions, which the team called human accelerated regions (HARs), rose to the top of the list. Surprisingly, only two of these regions code for proteins. Instead, the majority of the regions tend to be located near genes that are involved in regulating the function of genes. Furthermore, 12 of the regions are adjacent to genes involved in the development of the brain.
The Nature paper looks in depth at the region that has undergone the most change in the human lineage, which the researchers called HAR1 (for human accelerated region 1). Only two of the region's 118 bases changed in the 310 million years separating the evolutionary lineages of the chicken and the chimp. Incredibly, since the human lineage separated from that of the chimp, 18 of the 118 nucleotides have changed. This region "stood out," said Pollard.
But what does it do? To find out, Pollard began working with the wet lab, led by Sofie Salama. Haussler established the wet lab following his appointment as an HHMI investigator. After months of work, Salama and her lab mates determined that HAR1 is part of a larger DNA that is transcribed into RNA in the brain.
Then Salama got lucky. Pierre Vanderhaegen, a neuroscientist at the University of Brussels, was visiting Santa Cruz because he knew Salama's husband, who is also a neuroscientist. "I learned that Pierre was setting up to do in-situ hybridizations [at his lab in Brussels] to look at gene expression patterns in human embryonic brain samples," said Salama. "So I gave him a DNA probe from the HAR1 region and said, 'Try this.'"
A few months later Vanderhaegen e-mailed Salama with exciting news. He had discovered that RNA including the HAR1 region is first produced between the 7th and 9th weeks of gestation in human embryos. Furthermore, the RNA was produced by a Cajal-Retzius neuron, a particular type of cell that plays a critical role in creating the six layers of neurons in the human cortex.
Salama then determined that HAR1 actually lies in the region of overlap of two RNA genes that are transcribed in opposite directions along the DNA. Both genes appear to make RNAs that are not translated into proteins. The UC Santa Cruz team showed that these RNAs fold into particular shapes characterized by several helices. The changes to HAR1 during human evolution seem to have altered the length and configuration of some of these helices. "It's a brand new structure, unique," said Salama. "The downside is that we don't have many good clues as to how it functions."
Haussler's team is now following up on the clues that they do have. Other DNA regions produce stable RNA structures that have a variety of functions such as gene regulation or controlling the action of proteins. Furthermore, the same cells that express the HAR1 genes also produce a protein called reelin, which helps establish the architecture of the brain.
According to Haussler, the possibility that the HAR1 regions may play a role in the function of reelin is especially interesting since defects in reelin expression have been associated with schizophrenia and other mental disorders. "We still can't say much about the function [of the DNA containing HAR1]," said Haussler, "but it's a very exciting finding because it is expressed in cells that have a fundamental role in the design and development of the mammalian cortex."
And beyond HAR1 lie HAR2, HAR3, and so on through the 49 regions Pollard identified with her DNA screen. "We've only studied one of these regions carefully," said Haussler. "Now we have to go through the other 48."