news20090924nn4

[naturenews] from [nature.com]

[naturenews]
Published online 23 September 2009 | Nature | doi:10.1038/news.2009.940
News
Protein burns its evolutionary bridges
Mutations can set genetic change on an irreversible path.
By Elie Dolgin

Time always marches forward — and so does evolution, according to a new study showing that protein changes that happened over the course of tens of millions of years can prevent molecular turnarounds and render evolution irreversible.

"This backwards pathway is not accessible to selection," says study author Joe Thornton, a molecular evolutionary biologist at the University of Oregon in Eugene. "It's very strongly supported evidence for the view that contingency plays a major role in evolution."

More than a century ago, the French–born Belgian palaeontologist Louis Dollo proposed that evolution cannot retrace its steps to restore a lost trait — an idea that has remained controversial. Dubbed Dollo's law, this proposition had a certain appeal to many evolutionary biologists. Whales and snakes never regained legs, for example; birds did not reacquire teeth. But more recently, studies have shown that silenced genes and dormant developmental programs can be reactivated, leading many researchers to believe that evolution can indeed double back on itself.

Evolution's arrow

Thornton and his colleagues decided to test Dollo's law at the molecular level. They focused on the glucocorticoid receptor (GR), a protein that binds the hormone cortisol to regulate the stress response and other functions in humans and other vertebrates. Thornton's team had previously showed that the first GR protein evolved more than 400 million years ago from an ancestral receptor that was activated by a trio of hormones: cortisol, aldosterone and deoxycorticosterone¹. Over the course of 40 million years, the ancestral protein acquired 37 amino-acid alterations, cortisol binding to produce the 'modern' GR protein².

{“The chances of reversibility ever happening are vanishingly small.”
Joe Thornton
University of Oregon}

The researchers reverse-engineered the protein to undo all seven mutations that affected cortisol binding, but the resulting receptor was non-functional and did not bind any hormone. To determine whether other mutations might be having an effect on receptor function, the team looked at the three-dimensional structure of the protein and pinpointed five additional mutations. These extra mutations did not affect the cortisol-binding specificity of the receptor but they did prevent it from doing its job properly. When the researchers reversed these mutations as well, the protein was transformed back into its ancestral functional state and could bind all three hormones.

Although it is technically possible for the molecular changes to backtrack, admits Thornton, such a shift couldn't be driven by natural selection. Undoing the five extra mutations in the absence of the other seven changes had negative or neutral consequences on protein function, and so would not be favoured by evolution, the researchers report in Nature³. "The chances of reversibility ever happening are vanishingly small," Thornton says. "It's virtually impossible."

Traces of time

The study reveals how interdependencies between amino acids can limit the scope of evolution, says Günter Wagner, an evolutionary developmental biologist at Yale University. "Because [the additional mutations] are only enabling but not themselves selective, they can mutate away, effectively burning the bridge you went over," he says.

Michael Rose, an evolutionary biologist at the University of California, Irvine, says that the experimental demonstration of irreversibility was "impressive" but not all that surprising. "That this phenomenon would apply on this timescale is what every evolutionary biologist would expect," he says. But in the short term, selection often goes back to its old ways. For example, earlier this year, Rose and his colleagues showed that fruit flies that had diverged over decades of selection in the lab returned to their ancestral state after just two years of reverse selection⁴. Thornton's study, by contrast, "is way, way macro in time frame," Rose says, adding, "It illustrates the importance of historicity in evolutionary biology."

But Fyodor Kondrashov, an evolutionary geneticist at the Centre for Genomic Regulation in Barcelona, Spain, disagrees. "There's absolutely nothing impossible about reversing everything that has happened," he says. Rather than focusing on the authors' "spin of reversibility", Kondrashov says, the paper highlights how the likelihood of particular evolutionary trajectories is contingent on a protein's genetics and structure. "We really need these kinds of examples", he adds.

References
1. Bridgham, J. T., Carroll, S. M. & Thornton, J. W. Science 312, 97-101 (2006).
2. Ortlund, E. A., Bridgham, J. T., Redinbo, M. R. & Thornton, J. W. Science 317, 1544-1548 (2007).
3. Bridgham, J. T., Ortlund, E. A. & Thornton, J. W. Nature 461, 515-519 (2009).
4. Teotónio, H., Chelo, I. M., Bradic, M., Rose, M. R. & Long, A. D. Nature Genetics 41, 251-257 (2009).