With the depletion of crude oil reserves, middle and heavy petroleum feedstocks containing high levels of nitrogen impurities such as carbazole, are becoming more important as precursors for lighter feedstocks. However, the combustion of carbazole contaminated fuel releases nitrogen oxides, which results in the formation of acid rain. In addition, carbazole inhibits refining processes, such as hydrodesulfurization and denitrogenation and leads to fuel instability in storage via gum formation. Hence, removal of the nitrogen contaminants is highly desirable.

To achieve this goal, we propose a novel enzymatic carbazole denitrogenation pathway that removes nitrogen contaminants and preserves the carbon content of the fuel at the same time. This pathway is based upon the combined action of two enzymes, carbazole-1,9a-dioxygenase (CarA) and aniline dioxygenase (AtdA). In this pathway, carbazole is first dioxygenated into 2'-aminobiphenyl-2,3-diol (2'-ABPD) by CarA. The amine group from 2'-ABPD is then removed by AtdA. However, it was found that AtdA does not accept aromatic amines with 2-position substituents larger than 2-ethylaniline. Through a stepwise directed evolution strategy, we have widened the substrate specificity of AtdA to accept larger aniline homologues, such as 2-isopropylaniline, on which the wild type was not active, and further improved the activity of the enzyme on aniline and 2,4-xylidine.

To the authors' knowledge this is the first application of protein engineering on the five-component AtdA enzyme. This study not only brings the application of enzymes to carbazole denitrogenation a step closer to fruition, but also sheds light on the role played by the α-subunit of AtdA in determining its substrate specificity. The ability to degrade a wider range of aniline homologues would also make the AtdA mutant a practical and valuable biocatalyst for the remediation of harmful aromatic amine contaminants from other industries such as pharmaceutical and dye manufacturing.