The authors had previously found that the evolution of cortisol specificity in the ancestral glucocorticoid receptor (GR) protein was made possible through a series of amino acid mutations at residues that had no apparent effect on receptor function. The researchers speculate that these substitutions were necessary for GR to tolerate the larger mutations needed to shift specificity. These are referred to as "permissive mutations", and have been found in the evolution of a number of other ancestral proteins. Since these mutations are neutral, they are not selected against, and they are allowed to accumulate in the protein. But, these mutations are needed to stabilize function-altering mutations.
The catch here is that we currently don't know how many of these permissive mutations are needed to enable evolutionary transitions. If a lot of these mutations are needed, then there are many evolutionary paths that could facilitate alteration in function, and thus the outcome of protein evolution would only be weakly contingent on a specific historical event (in evolutionary biology, a historical event is a mutation). But, if only a few permissive mutations are needed, then that narrows the potential evolutionary paths, and the outcome of protein evolution would be more strongly contingent on a specific mutation event.
Using directed evolution, the authors introduced random mutations into every new copy of GR, mimicking the variation that evolution could have produced over a longer time scale. The idea was to create a number of potential "might-have-been" scenarios. To identify mutations, the library of mutated GR was introduced into a yeast strain only able to grow through a cortisol-dependent interaction, meaning that yeast growth required GR to specifically bind with cortisol. The researchers found 10 unique mutations that either completely or partially improved the yeast strain's cortisol sensitivity. Of these 10 mutations, one pair of mutations affected cortisol sensitivity similar to the historical set.
|The form of GR that existed 450|
million years ago is labeled here as
- they must stabilize specific local elements of the protein structure,
- maintain the correct energetic balance between functional conformations, and
- be compatible with the ancestral and derived structures.
Meeting these three criteria is challenging, and indicates why permissive mutations are rare. As the authors have shown, very few mutations can satisfy these narrow constraints. If evolutionary history were replayed from its ancestral starting point, it would be unlikely that the same mutations that allowed the shift in GR function would re-occur. On a larger scale, that means that the evolution of many other proteins could also be different, and we would end up with a radically different biology from the one we have now.
These results are a highly significant contribution to our understanding of how proteins can evolve new functions. While large-scale historical events, like mass extinctions and climate change, are generally the focus of evolutionary biology, smaller-scale, equally low-probability events like permissive mutations are also a factor in evolution and in the diversity of life.