UC Berkeley

Abstract

Engineering and Identification of Ionic Liquid-Tolerant Cellulases for Biofuels

Production

by

Paul William Wolski

Doctor of Philosophy in Comparative Biochemistry

University of California, Berkeley

Professor Douglas S. Clark, Chair

Cellulose for biofuels production presents both a great opportunity, in that cellulose is the world's most abundant source of organic material, but also a great challenge in that cellulosic material is highly crystalline and very recalcitrant to degradation. Cellulose is composed of glucose, and this glucose can serve as fermentation feedstock to biofuels in processes that are very well developed. This glucose could also be converted to hydrocarbons similar to diesel for easier commercial adoption.

Before cellulase enzymes can degrade the cellulosic biomass, the biomass generally must be preteated to make the cellulose more accessible to the enzymes (i.e. less crystalline). Typically this involves acid or base treatment that only moderately affects the cellulose. Ionic liquids (ILs), which are organic salts that are liquid at or near room temperature, have the ability to dissolve cellulose by disrupting the hydrogen-bonding network that makes cellulose so strong. Ideally the cellulase enzymes would be active against dissolved cellulose.

The theme of this research has been to combine the pretreatment step with the enzymatic hydrolysis step. The design of the research was as follows: Identify ionic liquids that can dissolve cellulose, while still supporting enzymatic activity and enzymes from nature that can withstand high ionic liquid concentrations. Then, use directed evolution to enhance the ionic liquid tolerance of cellulases and screen for variants that were indeed more IL-tolerant.

Additionally, it was of interest to determine what happens to the enzymes, when inactivated by the ionic liquid. Do they unfold? Does the ionic liquid block the active site?

All of these main objectives were achieved, to varying degrees, in this work. First, using GFP as a reporter protein for quickly measuring protein stability by GFP fluorescence, we identified the ionic liquid 1,3-Dimethylimidazolium dimethylphosphate (Mmim DMP) to support greater cellulase activity than other ionic liquids, including the more commonly used 1-Ethyl-3-methylimidazolium (Emim) acetate.

Then, we found cellulases from hyperthermophiles, such as Pyrococcus furiosus , to be more stable in aqueous ionic liquid than cellulases from mesophilic organisms. A cellulase from this organism was active in up to 70% (w/w) Mmim DMP.

Using DNA shuffling we generated a library of chimeric cellulase (cellobiohydrolase I or Cel7A) genes from several homologous genes. After screening a library of over 1200 variants, we identified two variants that were more stable than the native enzyme from Talaromyces emersonii. However, the degree of increase in stability was much less after both the wild type and variant enzyme were treated with exogenous glutamine cyclase to convert the N terminal glutamine to pyroglutamate. This post-translational modification occurs in T. emersonii. However, when the wild type and variant enzymes were expressed in Saccharomyces cerevisiae, this modification did not occur, and was confirmed via differential scanning calorimetry.

We also used differential scanning calorimetry to determine that the ionic liquid Mmim DMP lowers the melting temperature of the enzyme, in some cases, below the assay temperature. For this reason, we concluded that the ionic liquid, in conjunction with the assay temperature, is working to inactivate the cellulase via global unfolding of the enzyme.

In this work we showed that although enzymatic hydrolysis against dissolved cellulose, was not achieved, we successfully enhanced ionic liquid tolerance of cellulases via directed evolution and by selection of cellulases from extremophiles. We demonstrated a strong correlation between ionic liquid tolerance and thermotolerance. Finally, we confirmed that in directed evolution the winning variants are based directly off the screen used. In this case, evolving cellulases in a system that does not perform the native post-translational modification will not necessarily produce the same results in one that does.

A future study using the same screening system thus should involve either the exogenous addition of the glutamine cyclase enzyme or endogenous production of the glutamine cyclase to make the screen as close to the production of the native host as possible.