Modulating Biophysical Properties of Insulin with Non-Canonical Mutagenesis at Position B28

Abstract

Non-canonical amino acids are tools for altering the chemical and physical properties of proteins, providing a facile strategy to engineer proteins with novel properties, especially where canonical amino acid mutagenesis has exhausted nearly all avenues for further optimization. Insulin, for example, is one of the most widely studied therapeutic proteins; however, non-canonical insulin engineering is a subfield that has been largely unexplored. To this end, my thesis research has focused on the use of non-canonical amino acids to understand and engineer the biophysical properties of insulin.

Protein structure and function are sensitive to the smallest of changes; even small differences such as a single atom substitution or change in stereo-orientation can cause large, unsuspected, and unpredictable global changes. Chapters 2 to 4 describe substitutions at the 4th position of proline at position 28 of insulin’s B chain (ProB28) in a manner analogous to the structure-activity-relationships that are widely used in medicinal chemistry. Canonical mutagenesis at ProB28 has led to the discovery of rapid-acting insulins (RAIs): a class of therapeutic insulins with enhanced pharmacokinetic properties. Therefore, we chose to incorporate proline analogs with substitutions such as hydroxyl and fluoro groups, and different ring compositions to assess their effects on the biophysical properties (i.e. stability, dissociation rates and oligomerizations states) of insulin. Chapter 2 describes the discovery of a hydroxyinsulin variant with faster hexamer dissociation rates and enhanced stability compared to wild-type insulin in vitro. We find crystallographic evidence of a novel hydrogen bond in the insulin dimer interface which we hypothesize stabilizes the insulin dimer state. To complement the findings in chapter 2, chapters 3 and 4 detail an investigation of the importance of hydrophobic and nonpolar interactions for modulating the biophysical properties of insulin.

Establishing structure-activity-relationships for insulin will create new opportunities for further engineering using non-canonical amino acids. In chapter 5, we describe progress towards a general, simple screening method to discover new aminoacyl-tRNA synthetases for the incorporation of non-canonical amino acids in E. coli. Implementing such a high-throughput screening system will allow scientists to perform medicinal chemistry on proteins and discover new or improved therapeutics to help manage human diseases.