Identification of differentially diverged structural duplicates – Abstract

Amit Rosner and Ming-Jing Hwang

Institute of Biomedical Sciences, Academia Sinica, Taipei, Taiwan. In a project of our structural bioinformatics program, we attempted to identify proteins with different parts of it undergone different rates of evolution. A classic example of such proteins is antibody where a region of it, the antigen combining site, is hypervariable. OPAAS, with its ability to find alignments for different parts of a protein structure, allowed us to carry out a structure-based study of this problem. In an all-against-all comparison of a fold classification database of protein structures, we found, to our surprise, only a handful of protein families satisfy our criterion of exhibiting differential divergence: the percentage of sequence identity must differ more than 15% between the two parts of the protein that are aligned (to another protein, mainly another member of the same protein family). Moreover, all of these proteins are composed of structure duplicates, despite our search scheme was not confined to such proteins.

Although differential divergence and its functional manifestation for almost all of the protein families we have identified (immunoglobulin, fibronectin type III, crystallin, EF-hand, complement protein, leucine-rich repeat, and a few others) have been individually noted in the literature, our study is the first automated and large-scale search of such proteins. Our results also suggest that duplication is indeed an important mechanism used by evolution to expand a protein’s functional repertoire