Principal component analysis of conformational states within the EF-hand protein superfamily

The EFh domain is comprised of a pair of calcium binding EFh motifs, tethered together by a linker of variable length, each motif having a helix-loop-helix structure (Figure). It is one of the most successful metal binding domains in the whole proteome, with more than six hundred EFh domains annotated in the human genome alone. The variety of functions which EFh proteins are involved in is reflected in almost a continuum of conformational states occupied by the domain, under various conditions of calcium or peptide binding. Looking for recognizable structural patterns, derived by reducing the essential features to simple descriptors, permits a classification of the types of conformations (or conformational changes) that are significant for the functional role associated to calcium binding. We find that the two most meaningful linear combinations (PC1 and PC2 scores) of the six interhelical angles, obtained using a standard Principal Component Analysis (PCA) approach, are able to describe the system retaining 80% of the ability to describe the structure and the structural changes within the EFh protein superfamily. Members of different protein families are identified by their characteristic movements in the PC1-PC2 plane upon calcium or peptide binding. New structures can be easily assigned to specific families by their PC1-PC2 scores under various conditions.