We conjecture that the hierarchical organization of structurally similar proteins may be the result of the separation of the evolutionary time scales, shown schematically in Figure 2. On a time scale To, a set of mutations occur that do not affect those amino acids that play crucial thermodynamic, kinetic and/or functional roles. As a result, there is little variation in sequences at the important sites of proteins. If a mutation occurs at the thermodynamically, kinetically and/or functionally important sites, it usually substitutes amino acids with close physical properties so that core, nucleus and/or functional site are not disrupted and the protein folds into its family fold, is stable in this fold, and its function is preserved. At this time scale, a family of homologs is born.
Rarely, at time scale T, correlated mutations or larger-scale sequence rearrangements occur108"110 that modify several amino acids at the core, nucleus and/or functional site, so that the stability and kinetics of proteins are not altered. Such a set of mutations can drastically modify the sequence of the protein. However, within the time scale To, a family of homologs is born within which there is conservation of (already new) amino acids in the specific (important) sites of homologous proteins. Although there are alternations in the specific sites of the proteins at the time scale T, these sites are more preserved than the rest of the sequence. The proposed view of protein evolution is consistent with the observations of the hierarchical organization of structurally similar proteins in families of homologs. Sets of families of homologs are organized, in turn, in super-families of (possible) analogs. The evolutionary time in our analysis is associated with the number of mutations that accumulate in the course of evolution. Because the rates may vary between families and even proteins, the relation of evolutionary time to physical time is not straightforward. Evolutionary time can be rigorously defined statistically as the number of mutations that occur in a fold family, averaged over all family members. The real time for one family may be different from that of another. These considerations complicate interpretation of sequence-based approaches to organismic phylogeny and calls for more robust, structure based approaches to phylogeny (Deeds, Hennesey, Shakhnovich, in preparation).
Support for such a scenario comes from several studies reporting observations of correlated mutations in proteins in the course of evolution.108'109'111 In addition, Axe et al112 have demonstrated that random substitution of core residues in ribonuclease barnase by hydrophobic residues preserves the activity of barnase in a significant number of cases. They produced barnase mutants in which 12 of 13 hydrophobic core residues have together been randomly replaced by hydrophobic alternatives. A strikingly high proportion (23%) of mutants maintained structural integrity enough to support enzymatic activity of barnase.
Murzin33 proposed an elegant scenario of the evolution of protein architecture while maintaining its function. He argued that protein folding pathways may be altered by mutations. As a result, a local free energy minimum of the wild type protein may become a global free energy minimum of a mutant protein. The conformations at these states—global free energy minima of mutant and wild type proteins—may have no structural resemblance. However, these states may maintain the same function. As an example, Murzin argued that catalytic domain of the carboxypeptidase G2113 is structurally similar to aminopeptidase from Aeromonas Proteolytics114 However, these enzymes fold into two topologically different topoisomers.
Was this article helpful?