Hundreds of stretches of
DNA may be so critical to life's machinery that they have been “ultra-conserved” throughout hundreds of millions of years of evolution. Researchers have found precisely the same sequences in the genomes of humans, rats, and mice; sequences that are 95 to 99 percent identical to these can be found in the chicken and dog genomes, as well.
Most of these ultra-conserved regions do not appear to code for proteins, but may instead play a regulatory role. Evolutionary theory suggests these sequences may be so central to mammalian biology that even small changes in them would compromise the animal's fitness.
Led by Howard Hughes Medical Institute investigator David Haussler, at the University of California at Santa Cruz, the researchers published their findings online May 6, 2004, in Science Express, the Web counterpart of the journal Science. The lead author on the paper was Gill Bejerano in Haussler's laboratory. Also co-authoring the paper were John Mattick and his colleagues from the University of Queensland in Australia.
“It's extraordinarily exciting to think that there are these ultra-conserved elements, so many of which are near well-studied genes, that weren't noticed by the scientific community before because we didn't have the comparative data that highlighted these regions,” said Haussler. “The real credit goes to the prodigious efforts in sequencing these multiple genomes, which have given us this tremendous opportunity, opening our eyes to these very unusual genomic elements,” he said.
According to Haussler, the researchers were launched on their analysis when initial studies hinted at major regions of conserved DNA sequences. “When we had compared the human and mouse genomes, we found that about five percent of each of these showed some kind of evolutionary selection that partially preserved the sequence,” he said. “We got excited about this because only about 1.5 percent of the human genome codes for protein. So five percent was about three times as much as one might expect from the standard model of the genome, in which it basically codes for proteins, with a little bit of regulatory information on the side, and the rest is nonfunctional or “junk” DNA.
“These initial findings suggested that quite a lot of the genome was performing some kind of regulatory or structural role - doing something important other than coding for proteins,” said Haussler.
When the rat genome sequence became available, the researchers decided to search for the most extreme cases of conservation among the three mammalian species. They looked for long stretches of DNA, at least 200 base-pairs in a row, that were identical among humans, rats and mice. Statistically, the likelihood that a sequence of this length would appear unchanged among all three genomes by chance was infinitesimally small.
The results, said Haussler, were startling. The comparison of the three genomes revealed 481 such elements that they called “ultra-conserved.” “What really surprised us was that the regions of conservation stretched over so many bases. We found regions of up to nearly 800 bases where there were absolutely no changes among human, mouse and rat.”
Although 111 of these ultra-conserved elements overlapped with genes known to code for proteins, 256 showed no evidence that they overlap genes, and another 114 appeared inconclusively related to genes. In the 111 that overlapped genes, relatively small portions were actually in coding regions. Many were either in untranslated regions of the gene's messenger RNA transcript or in regions that are spliced out before the message is translated into protein.
Ultra-conserved regions were often found overlapping genes that specified proteins involved in binding RNA and regulating its splicing. “One of these genes is known to regulate its own splicing so as to either include or not include an ultra-conserved section, depending on conditions. There is also evidence for regulatory `crosstalk' with another member of the same gene family at this point. We may want to investigate further to see if these ultra-conserved elements that overlap RNA-processing genes are part self-regulating networks of RNA-processing activity,” said Haussler.
As to the function of the conserved regions that don't overlap genes, Haussler said, “there are hints that they may be involved in regulating transcription, but if so, it's a complete mystery how they work. What people find most interesting and exciting about these results is that they raise more questions than they answer.”