Recognition of proline-rich sequences plays an important role for the assembly of multi-protein complexes during the course of eukaryotic signal transduction and is mediated by a set of protein folds that share characteristic features.
Wei Gu and Mazen Ahmad
Intracellular protein domains recognizing proline-rich sequences (PRS) play a pivotal role in biological processes that require the coordinated assembly of multi-protein complexes. One example are Src kinases, where the terminal SH3 domain recognizes a long proline-rich stretch linking the nearby SH2 domain with the catalytic kinase domain. In vertebrate genomes, PRS are predicted to be among the most abundantly expressed amino acid sequence motifs and this corresponds to an increasing number of proteins that acquired PRS-recognition domains during the course of evolution. Up to now, the super-family of proline-rich sequence recognition domains consists of profilin, the SH3, the WW, the EVH1, the GYF, the UEV and probably the ligand binding domain of prolyl-4-hydroxylase. For each of these domains a set of conserved aromatic amino acid residues is important for peptide binding (see Figure 1).
Figure 1: NMR structure of GYF domain with wild-type peptide. The GYF domain is represented by its molecular surface; the peptide atoms are drawn as sticks. Residues forming the binding pocket are coloured in dark grey and labelled by their one-letter codes and sequence numbers. The four proline residues are coloured in green.
Among the 20 naturally occurring amino acids, proline is the only one in which the side chain atoms form a pyrrolidine ring with the backbone atoms (see Figure 2). This cyclic structure leads to some distinguished properties of proline: it induces conformational constraints among the atoms in the pyrrolidine ring, and it is the reason for the slow isomerization between cis/trans conformations and for the secondary structure preferences of proline-rich sequences (see below). Remarkably, the cis/trans preference of proline-X (where X is any amino acid) peptide bonds are different in different solvent environments.
Figure 2: Structure of (a) the cis and trans proline residues; (b) the PPI and PPII helices; (c) the PPII helix viewed along the helical axis. Molecules are shown as sticks in Corey-Pauling-Koltun colors.
We have previously studied the binding of wild type and some mutated PRS binding to the GYF adaptor domain by a combined theoretical (molecular dynamics simulations) and experimental (NMR and phage display) approach. The polyproline peptides considered in this study were found to be already folded into a PPII helix conformation in the unbound state and bind constitutively to the GYF domain. Obviously, this binding scenario is entropically more favorable than binding of unstructured peptides. The rigid PPII helix conformation of the unbound peptides studied is apparently intrinsically stable in solution and is also favorable for its specific binding motif.
In the study of PRS peptides with GYF domain, a register shift motion of the peptides was found for wild type and mutated complexes (see Figure 3): Pro5 and Pro6 of the peptide inserted into the binding pocket instead of Pro6 and Pro7 in the original binding modes. Although all four prolines in the peptide are rotated clockwise when viewed from the C to the N terminus, interestingly, the orientations of the remaining residues were kept and show only a slight translation toward the C terminus. Therefore, all interactions between the peptides and the domain (e.g. electrostatic attractions, hydrophobic interactions and intermolecular hydrogen bonds) were kept. This observation indicates an additional alternative binding mode of the peptides due to their three-fold symmetry around the helical axis (rather than the two-fold rotational symmetry along the helical axis).
Figure 3: The translation and rotation motions of the peptide between the two binding modes of the PRS ligand (shown as sticks) binding to the GYF domain (shown as ribbons).