Date of Award

Spring 1-1-2015

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Chemical & Biochemical Engineering

First Advisor

Joel L. Kaar

Second Advisor

Theodore W. Randolph

Third Advisor

John Carpenter

Fourth Advisor

Theodore Randolph

Fifth Advisor

Michael E. Himmel

Abstract

As media for biocatalysis, imidazolium based ionic liquids (ILs) have many applications, including improving enzyme refolding, selectivity, and replacing organic solvents as either the bulk media or cosolvent for biocatalysis. Despite their potential as media for biocatalysis, however, ILs have received underwhelming results due to the broad inactivation of enzymes in this media. Activity profiles of enzymes in increasing concentrations of the ionic liquid 1-butyl-3-methylimidazolium chloride ([BMIM][Cl]) show increasing deactivation over time. The deactivation appears unrelated to protein solubility losses but correlated with unfolding based on ultra-violet spectroscopy and fluorescence measurements, respectively. Additionally, equilibrium dialysis measurements show increased IL-enzyme binding at higher concentrations where the enzyme is thought to be unfolded. Increased binding to the denatured state will favor unfolding. Crystallographic studies further shed light on enzyme-IL binding by finding the binding modes of [BMIM][Cl] with enzymes tend to be based on cation-π stacking and hydrophobic interactions. Aromatic and hydrophobic groups are typically in the core of proteins. Unfolding will therefore expose these binding sites allowing for increased binding, as seen in the equilibrium dialysis results.

Our approach to stabilize enzymes in ILs was to mediate preferential IL-enzyme binding. There is a non-specific and site-specific mechanism by which enzymes are proposed to be stabilized. The non-specific approach is based on mutating residues to negatively charged amino acids. The negative charge is thought to repel the anion, [Cl], more than it attracts the cation, [BMIM], resulting in a net preferential exclusion of the IL. This non-specific preferential exclusion will, hypothetically, drive native state formation based on surface tension effects around the negative charge much like trehalose or other non-specifically excluded osmolytes. Alternatively, specific "high-affinity" sites where [BMIM][Cl] binds may be targeted such as tyrosines. If a tyrosine is exposed in the native state, it can likely still only bind one [BMIM] molecule from the exposed side. In the unfolded state, the tyrosine can now bind two [BMIM] molecules (one on either side). In theory, mutation of this high affinity residue will result in stabilization by eliminating one binding site in the folded protein and two binding sites in the unfolded protein.

Non-specific chemical modification approaches to model enzymes found that, in four cases, removing positively charged residues for a neutral moiety improved enzyme activity retention. Meanwhile, in all four cases, removing negatively charged residues for a neutral moiety was deleterious for enzyme stability. Certainly, this observation is a result of the total stability of the enzyme, which is the sum of its intrinsic stability (i.e. in buffer alone) and its ability to favorably mediate IL-solvent interactions. However, it was found that the intrinsic stability, measured via urea unfolding midpoint in buffer, was lower for all variants than the wild-type for two enzymes (the other two were not measured). While enzyme stabilization in ionic liquids may not be achieved via this chemical modification with every enzyme, it provides support for surface charge as a tool to non-specifically alter enzyme stability in ILs.

A site-specific approach to mediating binding interactions was employed using 2D HSQC NMR. [BMIM][Cl] was titrated into the enzyme sample, and perturbations in the resonance values of various peaks were observed. These peaks were interpreted as binding sites. In support of this, soaking enzyme crystals in [BMIM][Cl] resulted in resolution of [BMIM] and [Cl] molecules surrounding the residues with the largest resonance perturbations. Mutation to a glutamic acid at the location with the largest perturbations resulted in decreased chemical shift perturbations of the amino acids surrounding the mutation upon titration of the mutant enzyme with [BMIM][Cl]. Mutagenesis also resulted in an enhanced activity retention profile in [BMIM][Cl]. Moreover, two highly exposed positive charges on the enzyme were mutated to a glutamic acid in attempt to non-specifically exclude [BMIM][Cl]. Both mutations resulted in enhanced activity retention of the enzyme in [BMIM][Cl], although to varying degrees. Combination of stabilizing mutations around the surface of the enzyme provided an additive effect on stability. The mutations did not, however, improve the melting temperature of the enzyme in the absence of IL, suggesting the intrinsic stability of the enzyme in buffer was not enhanced. Moreover, activity retention profiles in [BMIM][Cl] show an equilibrium being achieved which is higher for the mutant enzyme than the wild-type, suggesting thermodynamic stabilization, and not purely a kinetic effect based on altering the unfolding barrier. Crystallographic studies of the mutant and wild-type enzymes showed a [BMIM] molecule resolved at tyrosine 49 for the wild-type enzyme, but not at glutamic acid 49 in the mutant. Also, equilibrium dialysis suggested decreased binding of [BMIM][Cl] to the mutant relative to the wild-type enzyme, consistent with the proposed mechanism of stabilization via reducing IL-enzyme binding, which would be stronger (and therefore more reduced by mutation) in the denatured state.

Share

COinS