Peptide Array Analyses to Identify Protein Protein Interaction Sites

The introduction of peptide array technology has greatly increased the speed at which sites of protein-protein interaction can be mapped (Frank 2002; Bolger et al. 2006; Baillie et al. 2007). In this technique, the sequence of one of the proteins of interest is used to generate a library of overlapping 25-mer peptides, each shifted by five amino acids, across the entire sequence of the protein. The peptides are then immobilized onto nitrocellulose and probed with a tagged, recombinant form of the interacting protein. Positive spots identify peptides that may contribute to the protein-protein binding site, with the intensity of the spots being a measure of strength of interaction. Interacting peptides can then be used as a template to design progeny peptides, each with a single amino acid substitution, usually to alanine. Substitutions that lead to loss of binding indicate those amino acids that are critical to the interaction. This information can then be used to generate mutations in the full-length proteins to confirm that the identified residues are also important in the context of the actual protein-protein interaction. Peptide array is a powerful technique that removes the guesswork from the time-consuming process of making mutants and so greatly facilitates the mapping process, which can be of particular importance when there is no structural information on which to base the generation of mutants. However, when structural information is present, then peptide array data coupled to structural information can rapidly facilitate further investigations using mutagene-sis to define sites and to facilitate the generation of peptidomimetics.

This technique was recently utilized to define the binding sites for the signaling scaffold proteins, P-arrestin and RACK1, on PDE4D5 (Bolger et al. 2006) (Fig. 4). Conventional mapping methods had already shown that there was a RACK1 interaction domain (RAID1) in the unique N-terminal region of RACK1 (Yarwood et al. 1999; Bolger et al. 2002), whose structure has recently been delineated by 1H-NMR (Smith et al. 2007). Reassuringly, the peptide array technique identified the same N-terminal region, but also identified a second RACK1-binding site in the catalytic region of PDE4D5 that was later confirmed by two-hybrid analysis. Two known binding sites were also confirmed for P-arrestin, and use of scanning alanine arrays extended previous studies by pinpointing the critical residues. Simultaneous overlay of the PDE4D5 peptide array with both proteins, in conjunction with dual labeling and detection using the Odyssey system, revealed that RACK1 and P-arrestin do indeed compete for their overlapping binding sites on PDE4D5 (Bolger et al. 2006). Thus, the use of peptide arrays has allowed the elucidation of a mechanism whereby the availability of RACK1 has important implications for the regulation of the P2AR by P-arrestin and PDE4D5 (see Sect. 2.5.6).

A peptide array approach was also used to define the binding sites for PDE4D5 on P-arrestin2 (Baillie et al. 2007). A comparison of P-arrestin2 peptide arrays overlain with either PDE4D3 or PDE4D5 revealed a region in the N-domain of P-arrestin2 that interacts with the common region of PDE4D and two sites in the C-domain that mediate specific interaction with PDE4D5. Subsequent alanine scanning peptide array analysis identified the key interacting residues at each site. Transfection of wild-type P-arrestin2 into P-arrestin1/2 knockout mouse embryonic fibroblasts was shown to result in a marked decrease in agonist-mediated PKA phos-phorylation of the P2AR. However, transfection of P-arrestin2 carrying mutations in any of the critical residues identified by peptide array failed to reduce P2AR phospho-rylation, despite their recruitment to the P2AR upon agonist stimulation. This peptide array-based approach has therefore revealed that PDE4D5 has to interact with both the N- and C- domains of P-arrestin2 in order to effect regulation of the P2AR.

0 0

Post a comment