Scientists Reconstruct Ancient Mutational Steps That Led To Modern Protein Function

By Charla Lambert

"Any replay of the tape would lead evolution down a pathway radically different from the road actually taken."

This statement, made by the eminent naturalist Steven Jay Gould, highlights a central difficulty in evolutionary science. If evolution proceeds in small steps that are chiefly random, how does one formulate testable hypotheses that can be examined in a controlled laboratory setting? After all, an experiment attempting to recapitulate an evolutionary pathway isn't guaranteed to travel the same road that life actually took.

Despite this difficulty, biologists at the University of Oregon have done just that: They reconstructed a pathway leading to the present-day function of a hormone receptor common to all jawed vertebrates. In the September 14th issue of Science, Joseph Thornton and his colleagues describe a series of mutational events that took place millions of years ago in the receptor. When followed in succession, the mutations demonstrate how the receptor evolved its modern function.

Thornton's results are remarkable because they are backed by both statistical theory and experimental evidence. For much of the 20th century, evolutionary processes appeared to elude basic principles of the scientific method. Scientists could make qualified, statistical statements about the evolution of particular biological functions, but they couldn't test those statements in a laboratory setting. Each random step taken by evolution added a layer of complexity that was nearly impossible for scientists to disentangle using traditional experimental methods. Indeed, this "irreducible complexity" is cited by intelligent-design proponents as evidence that some biological systems could not have evolved on their own, and therefore must have been designed.

Thornton's work unravels one example of biological complexity that is seemingly irreducible. In reality, however, the complexity is a scientifically verifiable by-product of evolution that can be tested and examined in a laboratory.

The scientists began with the protein sequence ancestral to the modern glucocorticoid (GR) and mineralcorticoid (MR) receptors. These receptors bind hormones in order to regulate different biological processes. While GR binds to cortisol to regulate an organism's stress response, MR binds to aldosterone in humans and deoxycorticosterone (DOC) in fish to regulate processes in the kidney and colon.

The two receptors are related through a gene duplication event that took place more than 450 million years ago. Earlier work published in an April 2006 issue of Science documents how Thornton and his colleagues used statistical models to infer the protein sequence of the ancient corticoid receptor (AncCR) from a large database of present-day sequences. The 2006 article also shows how the scientists "resurrected" the ancient protein by synthesizing a gene encoding it, expressing the gene in cell culture, and measuring its sensitivity to cortisol, aldosterone, and DOC. The AncCR protein resembled the MR protein in the way it bound these hormones, suggesting the cortisol specificity of GR emerged through evolution.

For the current study, Thornton set out to unravel the evolutionary steps between AncCR and GR. First, he teamed up with chemists at the University of North Carolina to determine AncCR's crystal structure. With the ancient protein's structure in hand, the scientists were able to see exactly how AncCR differs from both GR and MR. Next, Thornton and a postdoctoral researcher, Jamie Bridgham, resurrected several proteins intermediate between AncCR and its modern-day GR counterpart using the methods they developed in 2006. The researchers measured the sensitivity of each intermediate protein to the three hormones, and determined a time interval in which the cortisol specificity of GR emerged. Thornton and Bridgham then introduced mutations that occurred in this time interval and observed how each one affected the protein's function.

The scientists found seven mutations that, when introduced in a particular order, create a protein that is activated by cortisol with a strength and specificity similar to GR's. Surprisingly, two of the seven mutations have no effect on the function of AncCR when they are considered in isolation. Instead, they set the stage for three other mutations, which destroy the receptor's function altogether in the absence of the two "permissive" mutations.

"If the history of life could be repeated, there is no reason to expect that those two apparently neutral mutations would occur again," says Thornton. "Without them, the path that molecular evolution actually took would remain inaccessible, and we might end up with an endocrine system very different from the one we have today."