Circular Permutation in Proteins and Its Identification, Characterization, and Prediction

Abstract

Circular permutation (CP) is a protein structural rearrangement phenomenon that can be visualized as if the original termini of a protein were relocated to another position in the structure. Since circular permutants usually retain native structures and functions with sometimes increased stabilities, activities or functional diversities, CP has been applied as a technique in many fields, such as elucidating protein folding, improving enzyme activity and designing novel bifunctional proteins. The limitation of applying CP lies in its technical complexity and high cost. Not every position in a protein can be used to create a viable circular permutant; thus, successful application of CP requires selection of an appropriate CP site — a process that is yet ill defined. We had previously developed an efficient method, namely CPSARST, for identifying circularly-permuted homologs from large protein structure databases. Utilizing CPSARST, the first circular permutation database, i.e., CPDB, was established. Recently, based on CPDB, the sequence, structure and dynamics preferences of CP have been extensively examined. These preferences are then used to develop a CP viability prediction system called CPred, which combines several machine learning techniques and achieves prediction performances of a practical level. These developments are believed helpful to facilitate the application of CP in research and biotechnology.