Image: David Haussler
Timothy Archibald for HHMI
David Haussler, a Howard Hughes Medical Institute investigator at the University of California at Santa Cruz, uses computer software to analyze similarities and differences in the genetic codes for humans and other mammals.
By Alan Boyle Science editor
updated 11/30/2004 8:06:12 PM ET 2004-12-01T01:06:12

Researchers have reconstructed a long string of genetic code for what they believe is the common ancestor of placental mammals — a shrewlike creature that lived in Asia more than 75 million years ago.

The leader of the research team, David Haussler of the University of California at Santa Cruz, admits that the concept may sound like the plot line for a "Jurassic Park" sequel. But he says the real point of the research is to learn how the evolutionary paths of different species diverged over time — and make it easier to decipher present-day genetic code.

"It's driven by curiosity about our own origins," Haussler told Tuesday. "If we really want to understand in detail, at the molecular level, how evolution works, we have to reconstruct the evolutionary history of every base in the genome. It's only then that we'll see the key events."

Exercise in paleogenomics
The supposed ancestor is long gone, of course, and scientists don't plan to bring it back to life. Rather, this experiment served as an exercise in "paleogenomics" — the application of genetics to scientific questions about the ancient past.

To come up with the ancestral genetic code, scientists looked at a closely studied 10-gene region of the mammalian genome known as the CFTR locus. That region has been linked to the incidence of cystic fibrosis, so medical researchers have spent a lot of time compiling the code not only for humans, but for many other animals as well.

The code string consists of more than 1 million base pairs, or "letters" that read like this: GTCACAATT. A sophisticated computer program compared the equivalent strings for humans and 18 other placental mammals — that is, mammals who carry their young to birth within a placenta, as opposed to a pouch.

Image: Code comparison
Howard Hughes Medical Institute
This example shows how the genetic codes from various species were compared to come up with the code for a predicted ancestor ("Pred. Anc."). In addition to the 18 animals listed here, researchers also analyzed genetic coding from pigs.
Each of the 19 species exhibited its own additions, subtractions and changes in the genetic code. The computer looked beyond those differences to develop a consensus version of the code, using a mathematical formula weighted to account for evolutionary theory.

The scientists concluded that their results reflected the code for the common ancestor to an accuracy of 98 percent, based on repeated simulations and comparisons with the genetic codes for animals outside the original group, such as chickens and opossums.

Expedition into DNA
Just looking at the reconstructed code, you couldn't tell whether this common ancestor was a mouse or an armadillo. However, previous research has pointed to a shrew-sized animal known as Eomaia scansoria as one of the earliest placental mammals. Haussler said the genetic code generated by his team probably described a similar but more recent creature that existed more than 75 million years ago, when dinosaurs walked the earth.

Image: Eomaia scansoria
Mark A. Klingler  /  CMNH
This artistic reconstruction shows the earliest-known placental mammal, Eomaia scansoria, as it might have looked 125 million years ago. The rendering is based on fossils found in China.
"It's like a paleontological expedition, but into the DNA, not literally into the rocks," he said.

Haussler said his team is confident about the accuracy of the results in part because the mammalian branches spread out so quickly from the roots of the evolutionary family tree.

"You had a very rapid radiation of different lineages ... that fanned out like a starburst," he explained. "It's like there was an ancient text, and 20 different copies of this text were stored away in different places, and each one underwent independent and separate types of decay processes. You have a better chance of reconstructing the actual ancient text if the different changes on the different copies were independent."

Ironically, the technique doesn't work nearly as well to find the common ancestral code for cats and dogs, which are relatively close on the evolutionary tree.

Haussler, a Howard Hughes Medical Institute investigator, was joined in the research by Mathieu Blanchette of McGill University. Eric Green of the National Human Genome Research Institute and Webb Miller of Penn State University. Their findings were published in the December issue of the journal Genome Research.

Our genetic ancestors?
Cornell University geneticist Rasmus Nielsen said the newly published study would likely spark more work in the field. "Previously it was thought that we would never really know what our ancestors looked like at the genetic level, but now it appears that we'll be able to tell," he said.

Haussler said much more genetic data would be required to focus in on the full ancestral genome. Although the 1.1 million base pairs of the CFTR locus sound like a huge number, the full genome amounts to a whopping 3 billion base pairs.

Nevertheless, Haussler believes the technique he and his colleagues used could someday produce new insights into evolutionary history. "It's possible that some of these evolutionary changes will be studied in the laboratory to see the effect of a gene," he said.

For example, he suggested that particularly intriguing portions of ancestral DNA could be synthesized and inserted into mouse genes for testing. Such experiments could in turn yield new medical advances.

"We haven't even scratched the surface in terms of understanding the changes that actually occurred on a molecular level during our evolution," he said. "It's a completely wide-open field at this point, and I think there's much to discover."

© 2013 Reprints


Discussion comments


Most active discussions

  1. votes comments
  2. votes comments
  3. votes comments
  4. votes comments