Purchase this article with an account.
K.P. Mitton, J.M. Johnson, M. Dowd; Finding Protein Binding Motifs of Human Retinal Disease Proteins Using a Random-peptide Yeast Two-hybrid / Bioinformatics Strategy . Invest. Ophthalmol. Vis. Sci. 2003;44(13):4558.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Purpose: Mutations in retinal transcription factors (ie. NRL, CRX) are associated with retinal disease, however we lack knowledge about the signal transduction networks of which they are part. Finding other proteins in these networks reveals disease mechanisms and finds new candidate disease genes based on function. We explored a novel approach to screen a random peptide library to define binding motifs. These motifs can be used to search genome databases for candidate interacting proteins. Methods: Criteria for bait domains: 1) protein regulates gene expression, 2) mutations in protein cause human retinal disease, 3) previous evidence that the domain interacts with another protein. Domains were cloned in the yeast vector pHybLex/Zeo, to express LexA-bait hybrid proteins. Bait strains were transformed with a synthetic library expressing Random-Peptide-Gal4-Activation Domain (RP-G4AD) hybrid proteins (16-amino acids long). Primary screening of 1-3 million library transformants for growth on minus-His medium was followed with secondary screening for ß-galactosidase activity. Putative interacting RP-G4AD clones were tertiary screened to remove any that interacted with LexA-laminin. Custom in-lab analysis of peptides was carried out using Clustal-W, BLAST on a Unix G-4 workstation. Results: Random peptides were collected for several LexA-bait hybrids. RP-G4AD clones were obtained after three levels of screening. As an example, LexA-NRL-leucine zipper interacting RP-G4AD clones were analyzed for conserved motifs and then for similarities to previously reported NRL-interacting proteins (CRX, Fiz1). A single interaction experiment could obtain several random peptides that show homology within a region of the Fiz1 protein that is known to interact with the NRL-leucine zipper domain. One interaction motif may share homology with regions of Fiz1 and CRX known to interact with NRL. Conclusions: Protocols were optimized for yeast two-hybrid screening with a synthetic random peptide library and successfully applied to domains of proteins involved in the control of gene expression. Interacting random peptides were used to define interaction motifs. These motifs are being used to study regulatory mechanisms at the molecular level. Motifs are also being used to search genome databases for novel candidate interacting-proteins.
This PDF is available to Subscribers Only