Search

Rational redesign of the binding pocket of Cellular Retinoic Acid Binding Protein II (CRABPII) has provided a mutant that can bind retinal as a protonated Schiff base, mimicking the binding observed in rhodopsin. The reengineering was accomplished through a series of choreographed manipulations to ultimately orient the reactive species (the epsilon-amino group of Lys132 and the carbonyl of retinal) in the proper geometry for imine formation. The guiding principle was to achieve the appropriate Bürgi-Dunitz trajectory for the reaction to ensue. Through crystallographic analysis of protein mutants incapable of forming the requisite Schiff base, a highly ordered water molecule was identified as a key culprit in orienting retinal in a nonconstructive manner. Removal of the ordered water, along with placing reinforcing mutations to favor the desired orientation of retinal, led to a triple mutant CRABPII protein capable of nanomolar binding of retinal as a protonated Schiff base. The high-resolution crystal structure of all-trans-retinal bound to the CRABPII triple mutant (1.2 A resolution) unequivocally illustrates the imine formed between retinal and the protein.

Due to the difficulties in handling and manipulating membrane-bound proteins, such as rhodopsin, and the lack of crystallographic information on the cone opsins, we have opted to engineer a protein mimic of the transmembrane G-protein coupled receptor. Human cellular retinoic acid binding protein (CRABPII), a well studied and characterized protein, has been reengineered into a protein that now will bind retinal as a protonated Schiff base with high binding affinity (Kd = 2 nM) mimicking that of rhodopsin.