Enhancing the catalytic activity of solubilized enzymes in organic solvents

The goal of this project is to utilize the methods of biophysical spectroscopy with contemporary enzyme engineering so as to craft proteins to be more catalytically active in non-native environments. My work to date falls into two categories: I examined several properties of the enzyme subtilisin's active site when the enzyme was dissolved in organic solvents and at various hydration levels. These active-site specific studies helped me determine that active-site polarity and motion on a particular timescale are important for catalysis and that hydration of the enzyme promotes both. These findings were consistent with an observed large entropy of activation for solubilized enzyme preparations relative to insoluble preparations. The other pan of my project utilizes 2D NMR methods to interrogate the backbone of a homologous enzyme in aqueous/organic to determine which amino acids are most affected by the presence of polar organic co-solvents, and therefore likely responsible for the large drop in observed catalytic activity. These spectra guide us in our engineering approach by suggesting which residues may be fruitful targets for mutagenesis in order to improve catalytic activity in organic solvents.;I used several 19F NMR experiments on a fluorinated active-site inhibitor to examine the enzyme subtilisin's active-site polarity, dynamics and structure when dissolved in two organic solvents (isooctane and THF). By changing solvent type and water content I observed a range of catalytic efficiencies (kcat/KM) that spanned three orders of magnitude. Hydration increased the catalytic turnover in both solvents, 11-fold in isooctane and 50-fold in THF. I calculated the active-site dielectric constant from the inhibitor 19F NMR chemical shift using an empirical relationship, and found that the dielectric constant increased with hydration in both solvents, from epsilonas ≈ 8 to epsilonas ≈ 13, indicating that the more polar active site of hydrated enzyme preparations increases affinity for the charged transition state and increases reaction rates. Using NMR methods, I quantified active-site sub-microsecond (fast) and supra-microsecond dynamical processes. I found that fast motions at the active site were accelerated at moderate hydration levels, but these increases did not correlate with catalysis. The NMR-derived rate constants for slow motions also increased with hydration and strongly correlate with catalytic activity. These results reveal that hydration promotes favorable dynamic and electrostatic effects which both contribute to accelerated catalysis by solubilized subtilisin Carlsberg in organic solvents.;In the second part of my thesis work I used two-dimensional NMR to examine the chemical shifts of multiple sites on the protein backbone of a homologous protease as the solvents DMF and THE were introduced to the aqueous phase. The addition of these solvents reduced the catalytic activity of this protease and was correlated with dramatic NMR peak shifts in several amino acids. Residues that were most susceptible to polar solvents, as evidenced by large peak shifts, were subjected to saturation mutagenesis and screened for activity in high concentrations of organic solvent. Using this engineering approach, an enzyme with 5 mutations was constructed which exhibited 130-fold higher catalytic activity in 40% DMF than the wild type enzyme. This work demonstrates the feasibility of re-designing enzymes to function better in organic, and in particular, polar solvents.;I also include here initial studies in which we attempt to combine the success of active-site specific NMR in anhydrous organic solvents and the power of the 1H-15N HSQC experiment by acquiring site-specific NMR data on a fully isotopically labelled enzyme dissolved in anhydrous organic solvents. While the NMR resonances are very broad, qualitative peak-narrowing of protein backbone and surfactant head-group resonances occurs upon protein hydration, suggesting increased backbone mobility which coincides with an increase in catalytic activity. These initial studies indicate that simplified multi-nuclear NMR spectra can be useful in developing a biophysical model to explain the observed rate dependence on solvent and hydration level.