Reconstruction of Protein Structure Function Relations

Most evolutionary models concede some form of relationship between protein structure and function.125,126 Understanding this relationship is central not only to evolutionary biology, but also to structural genomics.127 However, despite many efforts, the establishment of a clear relationship between structure and function has been elusive. This is in part due to the fact that neither structural nor functional relationships are well defined. The structure-function relationship between proteins can be understood in light of an evolutionary prospective. Two views on protein evolution have previously been suggested to account for the uneven distribution of sequences in fold space: that of convergent evolution17'128'129 and divergent evolution.

Convergent evolution posits that different folds evolved independendy and the same ("most popular") protein structures are recycled many times by proteins having different functions.17 According to this model, new proteins may not be related by evolution to their orthologues as new proteins with similar function are rediscovered anew in many organisms. New proteins spawn by chance, but some structures are more populated than others because they are suggested to be more advantageous (thermodynamically, kinetically, evolutionarily). Such a scenario would suggest little relationship between structure and function.130

An alternative scenario is that of divergent evolution that suggests that a single or a few progenitor proteins give rise to many different, perhaps even unrelated offspring via processes of gene duplication and mutation.131 These offspring can differ significantly from each other, either in sequence or structure, and can perform a varied array of functions many generations later.132 It was shown recently that divergent evolution scenario implies important, observable structural relationships between domains: namely a scale-free organization of the protein universe that relays the history of how proteins are related to each other and would suggest a strong relationship between structure and function.8

It is important to note that divergent evolution implies a structure-function relationship that mirrors the structural hierarchy of PDUG. As protein structures diverged from progenitor proteins, so did functions. This relationship is necessitated by the requirement that the protein domain and all its descendants remain both functional133 and structurally stable during the progression of evolution. Since evolution of function is similar in spirit and timeframe to that of structure, the structure-function relationship can also be observed in the context of a hierarchical functional annotation that allows comparison of protein functions at various levels of specificity of description. The level of hierarchical description is important, as it is the focal lens of functional evolution. Such hierarchical functional description is provided to the bioinformatics community by the Gene Ontology (GO) consortium.134

The main result of reference 135 is a striking finding that the corollary relationship between structural evolution and acquisition of new function by protein domains necessitated by a divergent evolution scenario can be quantitatively observed on the PDUG. Looking at PDUG through a hierarchical description of structural comparisons we find that we can characterize different clusters by the "functional fingerprint" that they display. A functional fingerprint is the distribution of functions within a particular cluster. We find that this distribution is quite unique to a given fold family at certain levels of functional annotation provided by GO. If we relax the Zm-m threshold, we can also see an influx of protein domains into structurally similar clusters. These newly joined domains do not destroy the functional fingerprint of these clusters. This preservation of unique functional fingerprints through evolutionary dynamics further highlights the close relationship between structure and function necessitated by divergent evolution.

0 0

Post a comment