Computational characterization of a desulfinase for enhanced crude oil desulfurization

Increasingly stringent fuel quality regulations in combination with decreasing availability of low-sulfur crude oil present significant technical and economic challenges to petroleum refineries. Conventional means of addressing crude oil sulfur content are energy intensive and require significant capital investment. Furthermore, these treatments are often rather ineffective at removal of thiophenic sulfur, which is generally the most abundant form of sulfur in crude oil and its fractions. Downstream biodesulfurization represents a potential improvement upon conventional hydrodesulfurization, wherein a microbial cocktail of enzymes frees the covalently bound sulfur from catalytically recalcitrant organosulfur compounds without breaking carbon bonds, maintaining fuel heating values. However, key challenges to operational implementation of biodesulfurization remain, including optimization of the final rate-limiting step in the enzymatic degradation of dibenzothiophene, catalyzed by the 2’- hydroxybiphenyl-2-sulfinic acid desulfinase (dszB). Recent solution of the three-dimensional structure of dszB alongside biochemical characterization proffers preliminary details as to the structure-function relationship; though without comprehensive mutagenesis studies, fundamental mechanistic aspects governing substrate recognition, binding, and specificity and product inhibition remain unknown. Understanding the molecular-level contributions to dszB’s catalytic mechanism is critical to engineering enhanced activity for desulfurization of fossil fuels. The proposed research focuses on the challenge of improving desulfinase conversion rates by using molecular simulation methods to uncover the roles of key active site residues in catalytic turnover. This knowledge can be used to rationally design dszB variants for activity improvement leading to significant cost reductions in biodesulfurization implementation. The proposed work will address the question of how to design more active variants through three objectives. First, molecular dynamics simulations of dszB bound to set of known substrates and inhibitors will uncover dynamical features related to substrate recognition and binding and allow prediction of residues directly involved in product inhibition. Targeted molecular dynamics simulations of an experimentally observed conformational change will detail recruiting mechanisms within the active site. The second objective includes calculation of binding free energies of substrates and inhibitors using advanced sampling free energy methods. The calculations will be validated against experimentally measured affinity data, where available. The residues identified as participating in product inhibition in the first objective will be evaluated here confirming their role. The final objective will address the origins of product specificity by quantitatively characterizing a proposed catalytic reaction mechanism with hybrid quantum mechanical/molecular mechanics methods.