A methodology for creating thermostabilized mutants of G‐protein coupled receptors by combining statistical thermodynamics and evolutionary molecular engineering

We constructed a methodology for thermostabilizing a G-protein coupled receptor (GPCR) in the inactive state whose wild-type (WT) structure is unknown solely by multiple amino-acid mutations without the ligand binding. It is a combination of our recently developed theory based on statistical thermodynamics and site-directed saturation mutagenesis, a method often employed in evolutionary molecular engineering. First, the WT structure is predicted using the homology modeling. Second, a key residue is determined by our statistical-thermodynamics theory using suitably modeled mutant structures. Many of 19 different single mutations for the key residue are expected to produce significantly higher stabilization. Third, we undertake to mutate not only the key residue but also a few more residues whose side chains are close to the side chain of the key residue. The whole mutational space is then efficiently explored by introducing site-directed saturation mutations, and a gene (mutant) library is constructed using the small-intelligent and fully automatic single-tube recombination methods. Each mutant is expressed in Escherichia coli cells, and highly stabilized mutants are sorted out using a fluorescence-screening technique. The methodology was illustrated for the serotonin 2A receptor, 5-HT2AR, for stabilizing its inactive state. We could identify a double mutant whose apparent midpoint temperature of thermal denaturation is higher than that of a thermostabilized double mutant previously reported by similar to 8.9 degrees C and that of the WT by over 15 degrees C. Moreover, it exhibits higher binding affinity for spiperone, an antagonist which was previously proved to stabilize 5-HT2AR in the inactive state.