Letter Cite This: ACS Catal. 2019, 9, 253−258 pubs.acs.org/acscatalysis Mechanistic Studies of a Nonheme Iron Enzyme OvoA in Ovothiol Biosynthesis Using a Tyrosine Analogue, 2‑Amino-3-(4-hydroxy-3- (methoxyl) phenyl) Propanoic Acid (MeOTyr) † ‡ ⊥ ‡ ⊥ † ‡ ‡ § Li Chen, , , Nathchar Naowarojna, , Bin Chen, Meiling Xu, Melissa Quill, Jiangyun Wang, † † ‡ ‡ Zixin Deng, Changming Zhao,*, , and Pinghua Liu*, † Key Laboratory of Combinatory Biosynthesis and Drug Discovery, Ministry of Education, School of Pharmaceutical Sciences, Wuhan University, Hubei 430072, People’s Republic of China ‡ Department of Chemistry, Boston University, 590 Commonwealth Avenue, Boston, Massachusetts 02215, United States § Institute of Biophysics, Chinese Academy of Sciences, Beijing 100101, People’s Republic of China *S Supporting Information ABSTRACT: Ovothiols are thiol-histidines that play important roles in protecting cells against oxidative stresses. Because of challenges faced in their chemical synthesis, biosynthesis provides an alternative option. In ovothiol biosynthesis, a nonheme iron enzyme (OvoA) catalyzes a four- electron oxidative coupling between L-His and L-Cys. There are debates in the literature over whether oxidative C−S bond formation or sulfur oxidation is the first half of OvoA-catalysis. In this report, by incorporating a tyrosine analogue, 2-amino-3-(4-hydroxy-3-(methoxyl) phenyl) prop- anoic acid (MeOTyr), via an amber-suppressor method, we modulated the rate-limiting steps of OvoA-catalysis and observed an inverse 2 deuterium KIE for [U- H5]-His. In conjunction with the reported quantum mechanics/molecular mechanics (QM/MM) studies, our results suggest that Y417 plays redox roles in OvoA-catalysis and imply that oxidative C−S bond formation is most likely the first half of the OvoA-catalysis. KEYWORDS: ovothiol, sulfur-containing natural products, kinetic isotope effect, unnatural amino acid, biosynthesis ver the last few decades, studies on sulfur-containing synthases (OvoA, EgtB, and Egt1) are mononuclear nonheme vitamins/cofactors (e.g., thiamin pyrophosphate,1,2 bio- iron enzymes that catalyze an overall four-electron oxidation O− 1,3 6 7 8 9 tin, lipoic acid, molybdopterin, coenzyme B, coenzyme process using O2 as the oxidizing agent. Such a sulfur transfer M10,11) have provided the basis for a mechanistic under- strategy is distinct from all other well-characterized sys- standing of many sulfur-transfer processes that are occurring in tems.12,43 Downloaded via INST OF BIOPHYSICS on March 1, 2019 at 01:29:56 (UTC). cells. The explosion of genomic information led to the The crystal structure of Mycobacterium thermoresistibile EgtB discovery of many sulfur-containing natural product biosyn- was reported recently (Figure 1A).24 EgtB possesses a 3-His thetic pathways, including ovothiol A 4 and ergothioneine 8 − ligand environment instead of the initially proposed 2-His-1- See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles. 12 34 (Scheme 1). Under physiological conditions, ovothiol A carboxylate.17,20 Upon mutation of an EgtB active site tyrosine exists predominantly in its thiolate form and functions as a (Y377) to phenylalanine, the EgtB Y377F mutant exhibits potent radical and peroxide scavenger.35,36 Besides its roles 37,38 barely detectable sulfoxide synthase activity. Instead, this EgtB against oxidative stress, anticancer activities have also been ffi γ 39 Y377F mutant e ciently mediates the oxidation of -glutamyl- reported for ovothiols. Currently, industrial-scale production cysteine (γ-Glu-Cys) to γ-glutamyl-cysteine sulfinic acid.27 for both ergothioneine and ovothiol has been proven to be − Built upon this M. thermoresistibile EgtB crystal structure, three challenging.16,40 42 Therefore, biosynthesis was explored as an EgtB computational mechanistic studies were reported alternative option. In recent years, synergistic work between recently, using density function theory or quantum mechan- the Seebeck and Liu laboratories has led to the full elucidation 28,29,44 of the aerobic ergothioneine and ovothiol biosynthetic ics/molecular mechanics (QM/MM) methods. Con- 12−29 clusions from these reports differ in at least two aspects. First, pathways. In these two pathways, a sulfur atom is fi transferred from L-Cys to an L-His side-chain by a combination there are two possibilities for the rst half of this four-electron of two reactions: an oxidative C−S bond formation (OvoA in oxidation process: the sulfur oxidation suggested by Tian, Wei, ovothiol, Egt1/EgtB in ergothioneine biosynthesis, Scheme 1) and a reductive C−S lyase reaction (OvoB in ovothiol, and Received: September 27, 2018 EgtE/Egt2 in ergothioneine biosynthesis, Scheme 1). In Revised: November 12, 2018 ergothioneine and ovothiol biosynthesis, the sulfoxide Published: December 7, 2018 © 2018 American Chemical Society 253 DOI: 10.1021/acscatal.8b03903 ACS Catal. 2019, 9, 253−258 ACS Catalysis Letter Scheme 1. Ergothioneine and Ovothiol A Biosyntheses: (A) Owing to the lack of an OvoA crystal structure, we created Ovothiol A Biosynthetic Pathway; (B) Two Aerobic an OvoA homology model (Figure S1), which suggested that Ergothioneine Biosynthetic Pathways; and (C) Two OvoA Y417 is the counterpart of EgtB Y377 (Figure 1B). Potential Pathways for the Oxidative C−S Bond Formation Indeed, in our recent report,14 by replacing Y417 in OvoA with in OvoA-Catalysis 2-amino-3-(4-hydroxy-3-(methylthio) phenyl) propanoic acid (MtTyr, 14, Figure 1C) through amber-suppressor-mediated unnatural amino acid incorporation method, we modulated the two OvoA activities: oxidative coupling and sulfinic acid formation (eq 1, Figure 1D). However, because tyrosine and ff 45−48 MtTyr 14 di er in both pKa and reduction potentials, the role of Y417 as a Lewis acid/base29,44 or as part of redox chemistries remains obscure.28 To address this issue, we employed a tyrosine analogue, 3-methoxytyrosine (MeOTyr, 15), which has almost the same pKa, and a much lower reduction potential (by almost 200 mV) relative to that of tyrosine (Figure 1C).49 In addition, the presence of two competing pathways in OvoA-catalysis (eq 1, Figure 1D) enables the use of kinetic isotope effect (KIE) studies via two different approaches: kinetic isotope sensitive branching (eq 2, Figure 1D) to measure isotope effect at one of the branching − steps50 52 and substrate deuterium KIE on V/K (DV/K) using − intermolecular competition.53 55 We found that replacing Y417 in OvoA with MeOTyr alters the rate-limiting step of the OvoA reaction and MeOTyr OvoA-catalysis displays an inverse deuterium KIE when deuterium-labeled histidine was used as the substrate. These results in conjunction with the mechanistic models (Scheme 1C) suggest that Y417 plays a redox role in OvoA-catalysis and that the oxidative C−S bond formation precedes the sulfoxidation reaction in this four- electron oxidation process. The Methanococcus jannaschii tyrosyl amber suppressor Tyr tRNA (MjtRNA CUA)/tyrosyl-tRNA synthetase (MjTyrRS) pair has been developed to specifically recognize MeOTyr.45 MeOTyr was synthesized (Figures S2 and S3)and incorporated into OvoA to produce OvoAY417MeOTyr (MeOTyr-OvoA).14 The identity of MeOTyr-OvoA was further verified by tandem mass spectrometry (Figures S4 and S5), which confirmed the successful replacement of Y417 with MeOTyr. The iron content of the purified MeOTyr- OvoA was determined to be 1.14 ± 0.06 equiv of iron using atomic emission spectroscopy (Figure S4).56,57 The MeOTyr- OvoA was then characterized by oxygen consumption, 1H NMR, and 13C NMR assays.17,18 Based on the oxygen consumption rate measured by NeoFox oxygen electrode, ± the kinetic parameters for MeOTyr-OvoA were: kcat of 1.5 Figure 1. Key information related to the experimental design of the −1 ± μ ± μ 0.1 s and KM of 399 36 M for L-His, KM of 175 19 M present study. (A) Reported EgtB·γGC·hercynine complex (PDB: for L-Cys (Figure S6). The KMs of both L-His and L-Cys of the 4X8D). (B) Predicted OvoA structure (Figure S1 for more detail). MeOTyr OvoA variant are close to that of wild-type OvoA (C) Tyrosine-cysteine cross-linking in cysteine dioxygenase and two ± μ ± μ (KM of 420 31 M for L-His, and KM of 300 34 M for L- tyrosine analogues (MtTyr and MeOTyr). (D) Schematic represen- 19 ∼ tation of isotope sensitive branching method and equations for Cys) reported previously, while kcat is 6.6-fold lower than ± −1 1 calculating kinetic isotope effects (KIEs). that of wild type OvoA (kcat of 9.53 0.33 s ). In the H NMR assay, the spectrum of the reaction mixture confirmed the production of sulfoxide 2 (Figure S7). When [β-13C]-Cys was used as the substrate for MeOTyr-OvoA, 1H- and 13C and co-workers (pathway I, Scheme 1C)29,44 or the oxidative NMR spectra indicated that cysteine sulfinic acid 13 comprises C−S bond formation proposed by Faponle et al. (pathway II, up to ∼26% of the product mixture (Figures 2, S8 and S9). In Scheme 1C).28 Second, the EgtB Y377 could either function as wild type OvoA, cysteine sulfinic acid 13 is observed as less a Lewis acid/base as proposed by Tian et al.,29,44 or it could be than 10% of the product mixture.17 These results indicate that a part of a proton-coupled electron transfer (PCET) process as substitution of Y417 by MeOTyr in OvoA does not affect the suggested by Faponle et al.28 In the analogous OvoA-catalytic identities of the products from OvoA catalysis; rather, this process, based on these two EgtB-mechanistic models, OvoA mutation alters the partitioning between oxidative coupling reaction may proceed via either pathway I (intermediate 10)or and cysteine sulfinic acid formation pathways (eq 1 of Figure pathway II (intermediate 11 or 12, Scheme 1C). 1D). This result differs from the case of using hercynine as the 254 DOI: 10.1021/acscatal.8b03903 ACS Catal.