Smooth Change, Mechanistic Fluctuation: Thermodynamic System Drift in Protein Evolution

If you heat a protein up, the tertiary contacts that secure its three-dimensional shape weaken until eventually the protein unfolds, or ‘‘melts.’’ Not surprisingly, proteins from thermophilic organisms, such as the bacteria that live in hot springs, tend to have higher melting temperatures than their homologs from mesophiles like humans or the bacteria that live within them. The increased melting temperature of a thermophile protein reflects an increase in the underlying stability of the protein, which itself is the sum of multiple independent properties of both the folded and unfolded states, ultimately encoded in the amino acid sequence of the protein. While differences in melting temperature seem likely to be the product of natural selection, it is less clear that the underlying biophysical differences themselves are the direct result of selection; alternatively, they could arise as a byproduct of other selective events, from neutral sequence drift, or other evolutionary mechanisms. In this issue of PLOS Biology, Kathryn Hart, Michael Harms, and colleagues, together with their advisors, Susan Marqusee and Joseph Thornton, show that this alternative explanation is likely the case for at least one protein, a feat they accomplished by measuring the properties of ancient proteins they reconstructed from scratch using statistical evolutionary methods. The authors began by comparing the optimal growing temperatures of multiple extant bacteria to the melting temperatures of the enzyme RNase H1 (RNH) in each. As expected, an increase in the growing temperature correlated with an increase in the protein melting temperature, strongly suggesting that the melting temperature was selected in response to increasing environmental temperature. They then traced the evolution of thermostability by statistically reconstructing the protein ancestors of RNH in