Over the last couple of weeks, Michael Behe, the luminary of intelligent design creationism, has been trying to persuade the scientific community (that part of it that wanders into the Discovery Institute’s website anyway) that a paper from Joe Thornton’s lab, which appeared at the end of September in the journal Nature, somehow validates his ideas.
Thornton’s group has worked out, in exquisite detail, the evolution of the glucocorticoid receptor (GR) of vertebrates, whose job is to bind to the hormone cortisol. They predict that the modern GR evolved from an ancestral protein (AncGR1) capable of binding to other hormones, but barely able to bind to cortisol, present in the last common ancestor of cartilaginous and bony fishes some 450 million years ago. AncGR1 changed over 40 million years into the GR possessed by the last common ancestor of tetrapods and ray-finned fish (AncGR2), which binds to cortisol strongly and specifically. AncGR2 and AncGR1 differ in 37 amino acids. By painstakingly synthesizing proteins with different combinations of those mutations, Thornton’s group concluded that introducing 7 of those 37 amino acid changes to AncGR1 is sufficient to give it the cortisol-binding properties of AncGR2. However, when they tried to reverse those 7 changes starting from the AncGR2 receptor, they discovered that they obtained nonfunctional GRs. For evolution to work in reverse, they had to reverse 5 other changes that did not seem important for the original transition. In other words, evolution does not reverse itself easily, a beautiful molecular example of something the Belgian paleontologist Louis Dollo had proposed over a century ago. Evolutionary biology doesn’t get much cooler than that!
Behe appears to misunderstand some very important aspects of this research, such as neutral mutations and genetic drift.. There are 37 amino acid changes between the two forms of the receptor. That can take the ancestral form and put those 7 changes i, creating a molecule with the activity of the second protein.
But when they took the second protein and altered those 7 amino acids back to the more ancestral, first protein, they lost activity. This simply shows that evolution does not easily do backwards. There are still 30 amino acids difference and most likely several of these changes (5 in this case) are necessary to keep the activity.
Protein structure and function is determined by amino acid choices. Changing some of these could gain function, but only because the structure supported it. Obviously, going ‘backwards’ in a straight line did not work. But then, evolution did not work in a straight line going forwards, either.
Genetic drift and neutral mutations are an important part of evolution. In both, mutations exist that do not result in any real alteration in protein activity, at least not enough to be selected for or against. There are changes that can slightly alter structure or function which, while having little effect on current biological activity, can set the stage for an important change that can alter activity.
A common misunderstanding is that EVERY change is there for a reason. In fact, many mutations have little observable effect and thus are invisible to selection. At least until the point when a critical new mutation sets a new activity in place. Then some of this intermediate mutations can serve as barriers to reversing evolution.
They create a contingent brick wall that usually means evolution works in one direction.
But it is the response of Thornton to the misunderstanding of his work that is so excellent. Be sure to read it at Zimmer’s blog. It is pretty devastating, in my opinion.
Technorati Tags: Science