Design to Data for mutants of β-Glucosidase B from Paenibacillus polymyxa: I45K, A357S, I20A, I20V, and I20E

Abstract

The relatively small size and scope of most current datasets of biophysical mutation effects in enzymes limit the ability to develop data-driven algorithms enabling accurate generative modeling tools for designing novel enzyme function. Here, the Michaelis-Menten constants (kcat, KM, and kcat/KM) and thermal stability (TM) of five new mutations of {beta}-glucosidase B from Paenibacillus polymyxa (BglB) are characterized. Foldit software was used to create molecular models of the mutants, for which synthetic genes were constructed and the corresponding proteins produced and purified from E. coli. It was found that mutations that disrupted pre-existing hydrogen bonds near the active site had reduced expression in contrast to mutations at the same site that did not affect native hydrogen bonding. This is consistent with previous results showing the relationship between hydrogen bonding and enzyme functionality. These mutants contribute to a growing data set of >100 mutants that have been characterized for expression, kinetic, and thermal properties