Local conserved sequence regions (motifs) often indicate structurally or functionally important regions of a protein. Blocks are local multiple alignments of protein motifs. CYRCA is a method to identify consistently aligned blocks - essentially multiple alignments of blocks themselves [Kunin et. al. JMB (2001) 307, 939-949]. The method identifies distantly related motifs whose similarity could not be detected by common sequence alignment methods (i.e. blast, psi-blast). Here we present examples of such motifs. One example is the Rosmann type folds where the conserved motif contains the ligand-binding region. Between some of the families the structural similarity is in the binding site alone, but between others the whole fold is similar. In another set sequence similarity across 17 amino acids defines the structural cores of two protein families. Proteins from these families have the same structural fold and similar sequence motifs but their function, position of the active sites and overall homo-oligomer structure are different. While the proteins have the same monomer fold, most probably originating from a common ancestor, this fold is utilized in dissimilar ways for different functions. We also identified another set with TIM barrel proteins, which catalyze diverse reactions and have extremely low sequence similarity between members from different families and are thus hard to identify. Here, again, a local sequence similarity identifies a global structural fold.
In the sets we studied conserved local motifs identified local structural similarities, or the similarity of whole folds. Motifs can be the core of a conserved structure, or the active site of a conserved function. Sometimes both structure and function is related to the short sequence motif. The described sets typically contain families with yet unknown structures together with families that have one or few members with determined structures. This is also true for the function assignment of protein families in CYRCA sets. Using the distant relations identified by CYRCA between protein motifs presents testable structural and functional predictions. Our examples show how proper analysis of short sequence motifs can provide valuable information on protein structure and function. |
