Among the many interaction domains identified in the past decade, PDZ domains are one of the most frequently encountered. They are approximately 90 residues long and were first identified as regions of sequence homology in diverse signaling proteins. The name PDZ is derived from the first three proteins in which these domains were found: PSD-95 (a 95 kDa protein involved in signaling in the post-synaptic density), Dlg (the Drosophila discs large protein), and ZO1 (the zonula occludens 1 protein involved in maintaining epithelial cell polarity). The function of PDZ domains is to mediate protein-protein interactions by binding to the C-terminus of their target protein in a sequence-specific fashion (Fig. 1A). More recently, it has also been shown that some PDZ domains can bind other PDZ domains, either in a head-to-tail fashion, as is shown in Fig. 1B for the a-1 syntrophin / nNOS heterodimer, or in a back-to-back fashion that does not block either peptide-binding groove, as is shown in Fig. 1C for the homodimer of Grip1 PDZ6.
PDZ domains are often found in combination with other interaction modules and play a role in directing the specificity of receptor tyrosine kinase-mediated signaling, in establishing cell polarity, in directing protein trafficking, and in coordinating synaptic signaling. Their importance is underscored by the severe neuronal and developmental phenotypes observed in PDZ knockout mice and by their implication in human congenital diseases like Usher syndrome and Dejerine-Sottas neuropathy. The enormous diversity of PDZ function is manifest in their abundance; there are over 250 PDZ’s encoded in the mouse genome. To understand their individual roles, it is necessary first to define their recognition properties in a comprehensive and relevant fashion. Previous efforts to define PDZ selectivity have focused on only a few domains and have relied on collections of randomized peptides, rather than on physiological ligands. In order to provide a genome-wide understanding of PDZ function, we are investigating the selectivity of every PDZ encoded in the mouse genome relative to a large collection of genomically-encoded C-terminal peptides using protein microarray technology. We are also investigating their ability to bind to each other in a comprehensive and quantitative n x n screen. Our hope is that by defining the selectivity of PDZ domains in a genomically-informed way, and by integrating these data with current knowledge in a comprehensive database of PDZ function, we will enhance significantly our understanding of how individual PDZ’s contribute to complex biological processes, and how defects in PDZ proteins and their binding partners lead to human disease.
Figure 1. PDZ domain structures. (A) PDZ3 of PSD-95 (cyan), complexed with the C-terminal pentapeptide of CRIPT (KQTSV, yellow). (B) The PDZ domain of a-1 syntrophin (green), complexed with the PDZ domain of nNOS (blue). (C) Homodimer of Grip1 PDZ6 (pink and purple), complexed with the C-terminal octapeptide of Liprin (ATVRTYSC, yellow).