Protein Engineering of Escherichia coli β-glucuronidase

Abstract

This thesis describes engineering studies with the Escherichia coli β-glucuronidase enzyme (E. coli β-GUS) that catalyzes the hydrolysis of D-glucuronic acids (glycone) that are conjugated through a β-O-glycosidic linkage to an aglycone. The enzyme is specific for the glucuronic acid component and will tolerate a variety of aglycones. A point mutation was known to convert β-GUS into a glucuronylsynthase, that is an enzyme capable of the synthesis of glucuronide conjugates. The long-term aim of the present research was to change the donor sugar specificity of the glucuronylsynthase from a glucuronyl donor to a glucosyl donor allowing the synthesis of glucosides. The approach taken to achieve the long-term aim was to first alter the specificity of β-GUS and to then convert the variants to a synthase. There were two approaches taken to altering the substrate specificity of β-GUS. The first approached was to use site saturation mutagenesis to alter key residues that are located at the active site. A more effective approach involved using directed evolution to generate variants with altered specificity. A selection of these variants were converted to (putative) synthetic enzymes and tested for activity. The structure-guided site saturation mutagenesis of 9 sites was carried out in an attempt to alter the substrate specificity of β-GUS. The choice of the residues to be altered was made with the aid of the structure of β-GUS with a bound substrate analogue. Nine codons in the β-GUS gene were randomised to create libraries that contained all possible amino acid of residues in and / or near the active site. Mutants were assayed using substrates presenting 5 different glycones. Of the positions randomised, most glycosyl binding residues were found not to tolerate amino acid substitutions, suggesting they are essential for β-GUS function while majority of non-glycosyl binding residues were found to tolerate amino acid substitutions – but not with good activity One of the common dogmas of directed evolution is the idea that evolvability is related to stability. We set out to test this idea while evolving substrate specificity. Other workers had generated a more thermostable variant of β-GUS. In parallel, we evolved 1) the native enzyme (GUS-WT) and 2) the thermostable (GUS-TR3337) variant of β-GUS. Mutant libraries of both GUS-WT and GUS-TR3337 were created under identical conditions and had the same distributions of mutations. After five rounds of evolution, the catalytic efficiency (kcat/Km) of the best mutant of the wild type parent for pNP-glucoside was increased ~307-fold while the best mutant of the thermophilic parent demonstrates a ~4-fold increased kcat/Km over the best mutant of wild type parent. Selected mutants from both libraries were characterised with regard to conformational properties and stability and these investigations, combined with kinetic data, provided valuable information about how thermostability promotes the ease of protein evolvability. The initial characterisation of β-GUS variants was done with crude lysates. It was observed that the GUS-WT and GUS-TR3337 variants lost their newly evolved activity after purification. It was eventually determined that the BugBusterTM reagent used to lyse cells prior to screening, had affected the directed evolution campaign. The n-octyl β-D-thioglucopyranoside (OTG) presents in the BugBusterTM reagent, was very similar in structure to pNP-glucoside and was identified as a competitive inhibitor that suppresses glucosidase activity of wild-type β-GUS, indicating that it can be bound in the active site. In response to the addition of OTG in the screening assay, OTG enhanced the glucosidase activity of selected mutants. This observation highlights the potential pitfall in the use of commercial reagents to lyse cells for enzymes with glycosidase activities. However, the evolution in the presence of OTG gives some insight into how an enzyme might evolve to be regulated by an effector molecule – OTG, in this case. Finally, improved cell lysis variants were converted to glucuronylsynthase variants by introducing the site specific mutation. Their glycosynthase ability was tested using a similar protocol developed by McLeod Group for assaying glucuronylsynthase activity. Unfortunately, glycosynthase activity was not observed with α-D-glucuronyl fluoride donor and steroid accepters. Time constraints did not allow other substrate to be tested.