Scientists from the University of Colorado Anschutz Medical Campus and University College London have established a new theory of molecular evolution that allows the process to be analyzed using statistical mechanics, offering fresh insight into how proteins evolve.
The approach treats proteins as integrated systems, where interactions between different parts of the molecule influence subsequent changes. "Too often we ignore interactions between different parts of a protein, but we know that changes in one part of the protein affect subsequent changes in other parts," said Richard Goldstein, co-author of the paper. "It turns out this is really important for understanding why these molecules evolve the way they do."
Protein structure, function, and stability determine whether mutations become fixed or eliminated. Amino acid interactions throughout the protein cause evolution at one site to affect the likelihood of evolution elsewhere. The study found that it is possible to predict how often evolution will occur based on certain biochemical properties.
"This was a real surprise," explained co-author David Pollock. "Our theory accounts for well-known population genetics effects such as strength of selection and effective population size, but they drop out of the final equations that predict the rate of molecular evolution."
Molecular convergence patterns were seen to vary over time, suggesting that constraints in various areas of proteins were fluctuating. "This flips around the usual idea that the amino acids will adjust to the requirements of the rest of the protein," Goldstein said. "But we couldn’t explain exactly why this happened, or whether there was any regularity to the process."
Using concepts from statistical mechanics, the researchers established that amino acid entrenchment plays a key role in evolutionary divergence. The sequence entropy of folding—the randomness of this process—was found to balance the strength of selection in protein evolution.
The findings could help solve long-standing problems in reconstructing major evolutionary events, which standard models have struggled with. The research may enable new studies into how molecules change over long periods, how genes operate in the human body, how to predict the rate of evolutionary divergence, and the fundamentals of harmful mutations.
Editorial Notes