The complexity of signaling pathways has led to an appreciation of the diversity and importance of protein-protein interactions, yet the methodologies for identification of novel interacting gene products are time consuming and limited in scope. The goal of this proposal is to develop a methodology for high throughput screening of multiple target proteins for binding partners, and rapid identification of the domain to which they bind. The methodology we are developing is based on libraries of cDNAs which are expressed at the capsid surface of bacteriophage. Phage bearing peptide sequences that bind directly to target proteins are isolated in batch. Overlapping synthetic peptides mimicking specific sequences in the target protein are then used as targets for phage binding to identify specific binding domains. This novel paradigm is made possible by two recent advances: 1. The development of an expression system in phage T7 which fuses the coding sequence for the major capsid protein with cDNA inserts of approximately 150 base pairs, and 2. the development of a semi-automated system in which peptides are synthesized covalently attached to a 96 """"""""pin"""""""" support, dramatically simplifying high throughput identification of target residues. This system is readily expandable to 384 """"""""pin""""""""or greater and is potentially fully automatable. The proposal has 2 steps or aims: First, perfection of the system for cloning and expressing optimal length cDNAs, as well as testing their ability to interact with defined protein and synthetic peptide targets. Second, testing the versatility of the system by identifying novel molecules that interact with myelin structural proteins for which partners have yet to be identified and verification of their expression pattern. We have chosen myelin specific proteins as experimental targets, as mutations in these proteins cause severe dysmyelinating phenotypes. Thus for the first time, these studies will permit powerful correlations to be made between single amino acid substitutions and the loss of effector binding in human disease states.
