Structural determinants stabilizing helical distortions related to proline

The distortions related to proline in a alpha-helix can be accommodated by different structural elements. To obtain an exhaustive view of these distortions, we data mined a non-redundant subset of the Protein Data Bank in search of proline residues included either within contiguous helices or within structural motifs in which two helices are joined by a few linker residues with backbone dihedral angles in the generous alpha region. The distortions correspond to "typical" and "non-typical" proline distortions, with relative ratio of 0.65 and 0.35, respectively. Analysis of "non-typical" proline distortions indicates that most linkers have one (75%) or two residues (20%) and that proline is preferentially located at the second or third position of the second helix (95%). The dihedral angles of the linker residues are located in two areas of the generous alpha region. Structures with linker(s) in the alpha1 area, which is characterized by very negative phi values, possess i to i-5 H-bonds and correspond to pi bulges. Structures with linker(s) in the alpha2 area, which links the alpha and beta regions, possess i to i-3 H-bonds and correspond to tight turns. Further classification of bulges and turns as a function of the linker length and proline position yields five canonical structures, representing 85% of "non-typical" proline distortions. These structures are characterized by distinct H-bonding patterns and structural determinants and correspond to different classes of pi bulges and tight turns. This hierarchical approach provides a straightforward and robust classification of proline-related helical distortions.