Evidence for Cis Retinol Specificity in the Interphotoreceptor Retinoid–Binding Protein (IRBP)

Abstract

Abstract: : The cone visual cycle presents the difficulty of protecting and solubilizing the high flux of 11–cis and all–trans retinols between the cone and Müller cells. Mata et al. (Neuron 2002 36:69–80) proposed that IRBP may function in the extracellular transport of retinoids specific to the cone visual cycle. IRBP is an IPM glycolipoprotein consisting of multiple modules (4 in mammals; 2 in teleosts) each ∼300 amino acids. Our long–term goal is to evaluate IRBP’s involvement in this cycle. Purpose: To determine whether IRBP exhibits selectivity in its interactions with 9–cis and 11–cis retinols and retinaldehydes compared to all–trans retinol. Methods: Bovine IRBP was purified to homogeneity through concanavalin A, ion–exchange and size–exclusion chromatography. Full–length zebrafish, expressed as a soluble thioredoxin–fusion protein in E.coli, was purified by nickel–affinity, ion–exchange, and size–exclusion chromatography. Protein concentrations were determined by amino–acid analysis. Retinoid binding was studied by fluorescence spectroscopy, with titrations monitoring ligand enhancement, protein quenching, and energy transfer. Data were fitted by a ligand–binding equation that assumes equivalent non–interacting binding sites. Results:Fluorescence enhancement assays indicate that bovine and zebrafish IRBP exhibit similar affinities for the three retinols, all falling between 0.11µM and 0.16 µM. Quenching assays with the zebrafish IRBP reveal corroborating evidence, as all five isomers bind the protein similarly, all with an affinity around 0.15–0.20 µM. However, when applied to the bovine IRBP, quenching assays indicate that the cis retinols, which fall between 0.02 and 0.05 µM, bind the protein with a greater affinity than the cis aldehydes or all–trans retinol, which fall between 0.10 and 0.15 µM. Conclusions: Ligand enhancement assays probe a single site of bovine IRBP and reveal a binding affinity between the three retinols on both bovine and zebrafish IRBP. In contrast protein quenching assays, which probe two sites, indicate that bovine but not zebrafish IRBP exhibits tighter binding with the cis retinols relative to its binding with the cis retinaldehydes or all–trans retinol. Bovine IRBP may possess a particular site not found in the zebrafish that specifically accommodates the cis bend and requires a hydroxylated ligand for maximal interaction.

Keywords: protein structure/function • retinoids/retinoid binding proteins • photoreceptors