Exploring new UbiD (de)carboxylase functionalities using protein engineering

Abstract

The microbial UbiD enzyme family utilises the prenylated flavin (prFMN) cofactor to perform non-oxidative (de)carboxylation on a wide range of a,b-unsaturated acids. The fungal ferulic acid decarboxylases (Fdcs) represent a specific branch of the UbiD enzyme family that is the best studied and has been subject to multiple directed evolution campaigns. This has led to Fdc variants capable of the production of 1,3-butadiene, isobutene, and a range of vinyl or (hetero)aromatic compounds. Unactivated aromatic acids present the most formidable challenge, due to the inherent aromatic stability, and robust activity with the latter has yet to be conclusively demonstrated. We applied thermostability engineering and directed enzyme evolution to Aspergillus niger ferulic acid decarboxylase (AnFdc), following reports of the I327S variant being capable of weak 2-naphthoic acid decarboxylation. Thermostability engineering in the form of disulphide bond introduction led to the selection of an AnFdc variant (DB1) which displayed significant increases in Tm and Tagg measurements over the wild-type enzyme. Introduction of the active site mutation I327S to DB1 yielded substantially improved decarboxylation of 2-naphthoic acid at elevated temperatures. Following in vitro cofactor reconstitution, this allowed for kinetic characterisation of 2-naphthoic acid enzymatic decarboxylation at elevated temperatures, revealing this stabilised variant performs ~ 1 turnover per hour. Directed enzyme evolution aimed at generating improved 2-naphthoic acid decarboxylases used a substrate walking approach, as the AnFdc I327S variant (called AnNdcI - 1st naphthoic acid decarboxylase in this chapter) was not sufficiently active to allow product detection in clarified lysate assays. Using a screen for increased benzofuran-2-carboxylic acid decarboxylation, selected variants were re-screened for activity with 2-naphthoic acid. This resulted in the variant AnNdcII which when purified displayed an 11-fold increase in naphthalene production when compared to AnNdcI. AnNdcII was active enough to allow detection of naphthalene in clarified lysate plate assays, supporting a 2nd round of direct enzyme evolution on 2-naphthoic acid. This resulted in AnNdcIII which displayed a further 2-fold improvement in naphthalene production compared to AnNdcII. Crystal structures along the evolutionary trajectory reveal removal of active site ordered water molecules and minor alteration in relative substrate-prFMN orientation. Kinetic studies show the increase in naphthalene production is underpinned by a 4.5-fold increase in kobs from 0.064 h-1 for AnNdcI to 0.28 h-1 for AnNdcIII with enzymes co-expressed with UbiX (flavin prenyltransferase). AnNdcII and IIIs increase in activity against 2-naphthoic acid also unlocked previously unreported enzymatic decarboxylations of benzothiophene-3-carboxylic acid and 1-naphthoic acid. Finally, we provide biochemical and structural characterisation of a UbiD enzyme from Pyrococcus furiosus (PfUbiD). Unusually, this enzyme preferentially binds FMN in vivo when expressed in E. coli. While the enzyme is capable of binding prFMN, no activity could be observed for a wide panel of a,b-unsaturated acids. Following structural determination, mutations were made to make the PfUbiD active site more closely resemble that of AnFdc. The corresponding variant (PfAnUbiD) readily bound prFMN in vivo, although did not demonstrate any enzymatic activity. It however provides a thermostable prFMN-binding UbiD scaffold for future directed evolution campaigns.

Date of Award	30 Jan 2024
Original language	English
Awarding Institution	The University of Manchester
Supervisor	David Leys (Main Supervisor), Sam Hay (Co Supervisor) & Igor Larrosa (Co Supervisor)