Structural snapshots of OxyR reveal the peroxidatic mechanism of H2O2 sensing

Abstract

Hydrogen peroxide (H₂O₂) is a strong oxidant capable of oxidizing cysteinyl thiolates, yet only a few cysteine-containing proteins have exceptional reactivity toward H₂O₂ One such example is the prokaryotic transcription factor OxyR, which controls the antioxidant response in bacteria, and which specifically and rapidly reduces H₂O₂ In this study, we present crystallographic evidence for the H₂O₂-sensing mechanism and H₂O₂-dependent structural transition of Corynebacterium glutamicum OxyR by capturing the reduced and H₂O₂-bound structures of a serine mutant of the peroxidatic cysteine, and the full-length crystal structure of disulfide-bonded oxidized OxyR. In the H₂O₂-bound structure, we pinpoint the key residues for the peroxidatic reduction of H₂O₂, and relate this to mutational assays showing that the conserved active-site residues T107 and R278 are critical for effective H₂O₂ reduction. Furthermore, we propose an allosteric mode of structural change, whereby a localized conformational change arising from H₂O₂-induced intramolecular disulfide formation drives a structural shift at the dimerization interface of OxyR, leading to overall changes in quaternary structure and an altered DNA-binding topology and affinity at the catalase promoter region. This study provides molecular insights into the overall OxyR transcription mechanism regulated by H₂O₂.

Keywords: X-ray structure; hydrogen peroxide sensor; redox regulation; transcription factor.

Fig. 1.

Cg-OxyR crystal structure shows a more symmetric subunit arrangement compared with the asymmetric relation between the RD and DBD homodimers of Pa-OxyR. (A) Crystal structure of tetrameric Pa-OxyR C199D [PDB ID code 4X6G (23)]. (B) Crystal structure of tetrameric Cg-OxyR_C206S. (C) Overview of the Cg-OxyR_C206S monomeric subunit, with the DBD and the RD highlighted.

Fig. 2.

Comparison of Cg-OxyR active-site pocket with previously published structures. (A) OxyR structures show conservation of the active-site region. The active-site architecture from the OxyR crystal structures of C. glutamicum C206S (present study), V. vulnificus E204G OxyR2 [PDB ID code 5B7D (25)], N. meningitidis [PDB ID code 3JV9 (24)]*, P. gingivalis C199S [PDB ID code 3T22 (26)], and P. aeruginosa [PDB ID code 4Y0M (23)] are shown. (B) Comparison of the different H₂O₂ binding modes of the H₂O₂-soaked crystal structures of C. glutamicum OxyR_C206S (present study) and P. aeruginosa C199D [PDB ID code 4X6G (23)]. Polar interactions between the residue side-chains and ligands (SO₄²⁻, H₂O₂) or water molecules (red spheres) are presented as dashed black lines. *The crystal structure of N. meningitidis OxyR contains two protein chains; in the active-site pocket of the alternate protein chain (not shown), the conserved water molecule which hydrogen bonds R270 is present but the conserved water molecule hydrogen bonding T129 is absent.

Fig. 3.

C206, T107, and R278 are the crucial residues for H₂O₂ catalysis. (A) In the crystal structure of Cg-OxyR_H2O2, an H₂O₂ molecule occupies the active-site pocket in two of the four chains present in the asymmetric unit. Displayed is one of the two the H₂O₂ binding environments, termed pocket A (see Crystallographic and Kinetic Evidence for H₂O₂ Binding and Reduction). Here, the side-chain of S206 is directed away from the active site, whereas in pocket B, the side-chain of S206 is directed into the active site. Both conformations are displayed here, with a double-headed arrow indicating the rotameric change. The hydrogen-bonding network between H₂O₂, two conserved water molecules, and residues of the active site is expressed in terms of interatomic distances. (B) Mutation of the active-site residues impairs H₂O₂ reduction and supports the crystallographic data. WT Cg-OxyR and mutant variants (C206S, T107V, T136V, H205A, R278Q) were mixed with an equimolar amount of H₂O₂, and the H₂O₂ concentration was monitored in function of time with the FOX assay reagent. The lines correspond to single exponential fittings of the H₂O₂ consumption (n = 3). The error bars correspond to SD.

Fig. 4.

Cg-OxyR disulfide formation reorganizes its tetrameric conformation. (A) Crystal structure of reduced tetrameric Cg-OxyR_C206S and a close-up view of the S206 and C215 of a single protomer (Right). (B) Crystal structure of the disulfide tetrameric Cg-OxyR_SS, generated by crystallographic symmetry, with a close-up of the disulfide-bonded C206-C215 of a single protomer (Right). For both structures, each protomer is colored separately, and the RD homodimer formed by the blue and light-blue protomers were aligned before figure preparation, thereby providing a common orientation for structural comparison. (C) Each panel is a focused view of the relative positions of S206/C206 and C215 of the three crystal structures reported in this study. A structural alignment of all three structures was performed to provide a common orientation. The Left panel depicts the S206 and C215 of Cg-OxyR_C206S, and is equivalent to the fully folded state of Cg-OxyR_H2O2. The Center panel shows the locally unfolded state of Cg-OxyR_H2O2, and the Right panel shows the C206-C215 disulfide of Cg-OxyR_SS.

Fig. 5.

Cg-OxyR shows rapid conformational changes upon oxidation. (A) Structural view of the tryptophan location and its side-chain flip upon oxidation. (B) The side-chain flip leads to a decrease in intrinsic fluorescence. Fluorescence spectroscopy of Cg-OxyR upon oxidation by H₂O₂ and rereduction by DTT. (C) Cg-OxyR requires both cysteines for the full conformational change. One micromolar WT and C206S/C215S variants were mixed with 5 μM H₂O₂, and fluorescence decrease was monitored over time in a stopped-flow mixing device. (D) The conformational change rate increases with H₂O₂ concentration in a hyperbolic manner. Increasing concentrations of H₂O₂ were added to Cg-OxyR. The progress curves were fitted to single exponential equation, and the observed rate constants plotted against the H₂O₂ concentration (n = 3). Inset shows the progress curves of Cg-OxyR fluorescence changes upon addition of 10, 20, or 30 μM H₂O₂.

Fig. 6.

Cg-OxyR oxidation decreases binding affinity and extension to the catalase promoter region. (A) DNase I footprint of the C. glutamicum catalase promoter/operator region in presence of WT or C206S/C215S CgOxyR, and under reducing (Left) or oxidizing (Right) conditions. A C/T and A/G sequencing ladder and a no Cg-OxyR control were added in each condition. Binding regions are highlighted in blue, with the binding affinity being proportional to the color intensity. The nucleotides that are hypersensitive to DNase I treatment in presence of OxyR are marked by a green dot. The transcription start point (+1) is highlighted in red, and the −10 region is highlighted in salmon. A summary of the binding experiments is shown below. (B) Fluorescence polarization experiments using a 6-FAM–labeled oligonucleotide containing the catalase binding region (highlighted in gray in A). The labeled oligonucleotide was mixed with increasing concentrations of WT/C206S Cg-OxyR, under reducing or oxidizing conditions, until reaching binding saturation (n = 3).

Fig. 7.

Cg-OxyR controls catalase transcription, recovery from lag phase, survival to H₂O₂ stress, and H₂O₂ scavenging rate. (A) katA mRNA levels under normal conditions in the different C. glutamicumstrains (n = 3). (B) katA mRNA levels in the different C. glutamicum strains after 10 or 30 min upon a 10 mM H₂O₂ challenge (n = 3). (C) Growth curves for the different C. glutamicum strains (n = 3). (D) In vivo resistance to H₂O_2. The different Cg-OxyR strains were spotted on agar plates containing increasing concentrations of H₂O₂, and growth was evaluated. (E) H₂O₂ consumption rates. An 8-mM (Left) or 40-mM (Right) H₂O₂ bolus was added to growing cultures of the Cg-OxyR strains, and the H₂O₂ concentration decrease was monitored in function of time using the FOX reagent (n = 3). Samples were statistically compared with Student’s t test when indicated (NS, not significant; *P < 0.1; **P < 0.05; ***P < 0.01; ^#P < 0.005).

Fig. 8.

Comparison of H₂O₂ reduction mechanism by OxyR and Prdx. Schematic drawing of the Prdx peroxide-bound active site (A) [reprinted from ref. . Copyright (2010), with permission from Elsevier] in comparison with the OxyR active site (B). The key hydrogen bonds (dashed lines) that allow for H₂O₂ reduction are highlighted.