Ohr Protects Corynebacterium glutamicum against Organic Hydroperoxide Induced Oxidative Stress

Ohr, a bacterial protein encoded by the Organic Hydroperoxide Resistance (ohr) gene, plays a critical role in resistance to organic hydroperoxides. In the present study, we show that the Cys-based thiol-dependent Ohr of Corynebacterium glutamicum decomposes organic hydroperoxides more efficiently than hydrogen peroxide. Replacement of either of the two Cys residues of Ohr by a Ser residue resulted in drastic loss of activity. The electron donors supporting regeneration of the peroxidase activity of the oxidized Ohr of C. glutamicum were principally lipoylated proteins (LpdA and Lpd/SucB). A Δohr mutant exhibited significantly decreased resistance to organic hydroperoxides and marked accumulation of reactive oxygen species (ROS) in vivo; protein carbonylation was also enhanced notably. The resistance to hydrogen peroxide also decreased, but protein carbonylation did not rise to any great extent. Together, the results unequivocally show that Ohr is essential for mediation of organic hydroperoxide resistance by C. glutamicum.


Introduction
Reactive oxygen species (ROS) are among the most potent threats to living organisms; ROS modulate the intrinsic balance between life and death [1]. ROS, including hydroxyl radicals, singlet oxygen, and hydrogen peroxide, are by-products generated via aerobic metabolic processes or upon stress caused by external agents [2]. When ROS levels are significant, the protective systems of living organisms are destroyed, and nucleic acids, proteins, carbohydrates, and lipids are damaged. In addition, pathogenic bacteria invading a host induce a burst of enzymatic ROS synthesis, and the host seeks to mount a defense [3]. Of the various types of ROS, organic hydroperoxides are particularly toxic, partly because they can generate free organic radicals, which in turn react with membranes and other macromolecules to promote ATCC 13032 was the wild-type background of all derivatives used in this study. Antibiotics were used at the following concentrations: kanamycin, 50 μg ml -1 for E. coli and 25 μg ml -1 for C. glutamicum; nalidixic acid, 40 μg ml -1 for C. glutamicum; chloroamphenicol, 20 μg ml -1 for E. coli and 10 μg ml -1 for C. glutamicum; ampicillin, 100 μg ml -1 for E. coli.

DNA manipulation and plasmid construction
General DNA manipulations, transformations and agarose gel electrophoresis were carried out by applying standard molecular techniques. Primers used in this study are listed in S2 Table. PCR was performed with EasyTaq or EasyPfu DNA polymerase (TransGen Biotech, Beijing, China). Plasmids were isolated with plasmid DNA miniprep spin columns (TIANGEN, Beijing, China), and DNA fragments were purified from agarose gels by EasyPure Quick Gel Extraction Kit (TransGen Biotech, Beijing, China). Genes encoding C. glutamicum Ohr (NCgl0023, GI:23308767), dihydrolipoamide dehydrogenases (LpdA, NCgl0658, GI:19551 918; Lpd, NCgl0355, GI:19551612), dihydrolipoamide acyltransferase (SucB, NCgl2126, GI:19553408), were amplified by PCR using C. glutamicum RES167 genomic DNA as template. These DNA fragments were digested by corresponding restriction enzymes and cloned into pET-28a and pET15b vectors to construct plasmid pET-28a-ohr, pET-28a-lpdA, pET-28a-sucB and pET15b-lpd, respectively. The plasmid pET-28a-ohr was used to generate the two Ohr site-directed mutants by overlap PCR [27] to replace active site Cys 60 and Cys 124 with Ser residue (Ohr:C60S and Ohr:C124S). Briefly, for C60S DNA construct, segments were amplified by PCR in two steps with mutagenic primers Ohr-F/Ohr-C60S-R and Ohr-C60S-F/Ohr-R (mutation sites are shown in bold in S2 Table) used to amplify segments 1 and 2, respectively. The second round of PCR was carried out by primer pair Ohr-F/Ohr-R using fragment 1 and 2 as templates to get ohr:C60S segment with desired mutation. The ohr:C124S DNA fragments were obtained by similar procedure with Ohr-F/Ohr-C124S-R and Ohr-C124S-F/Ohr-R primer pairs. ohr:C60S and ohr:C124S segments were digested and cloned into plasmid pET-28a to produce plasmids pET-28a-ohr::C60S and pET-28a-ohr::C124S, respectively. To construct the plasmid for ohr gene knock out, the 996-bp upstream PCR product and 980-bp downstream PCR product of ohr were amplified using primer pairs Dohr-F1/Dohr-R1 and Dohr-F2/Dohr-R2. Then, the upstream and downstream PCR fragments were fused together with primer pair Dohr-F1/Dohr-R2 by overlap PCR. The resulting DNA fragments were digested with BamHI/ SalI and inserted into suicide vector pK18mobsacB to create pK18mobsacB-Δohr. For complementation of ohr in the Δohr mutant, ohr DNA fragments were digested and cloned into pXMJ19 vector to yield pXMJ19-ohr.
To construct the lacZ fusion reporter vector pK18mobsacB-P ohr ::lacZ, overlap PCR was used to fuse the ohr promoter to the lacZY reporter gene [27]. Firstly, the 1,000 bp ohr promoter and the lacZY DNA fragment were amplified with the primer pair Pohr-F/Pohr-R and lacZYF/lacZY-R, respectively. Secondly, the two PCR products were used as the template with Pohr-F and lacZY-R as primers, and the resulting PCR fragment was inserted into vector pK18mobsacB to get the pK18mobsacB-P ohr ::lacZ. The fidelity of all constructs was confirmed by DNA sequencing (Sangon Biotech, Shanghai, China).

Construction and complementation of an ohr deletion mutant strain in C. glutamicum
To construct the C. glutamicum ohr in-frame deletion mutant (Δohr), the plasmid pK18mob-sacB-Δohr was transformed into C. glutamicum RES167 by electroporation, and chromosomal integration was selected by plating on LB agar plates supplemented with kanamycin. The Δohr deletion mutant was subsequently screened on LB agar plates containing 10% sucrose and confirmed by PCR and sequencing as previously described [28]. For complementation of ohr in the Δohr mutant, pXMJ19-ohr was transformed into the mutant strain by electroporation. The transformants were selected on LB agar plates supplemented with nalidixic acid and chloroamphenicol and ohr gene expression in C. glutamicum was induced by addition of 0.5 mM isopropyl-D-thiogalactopyranoside (IPTG) to the culture broth.

Over-expression and purification of recombinant proteins
To obtain purified Ohr, SucB, Lpd, and LpdA proteins, E. coli BL21(DE3) transformed with pET-28a and pET-15b derivatives (S1 Table) were used for recombinant protein expression and purification. The recombinant strains were grown at 37°C in LB-kanamycin broth (A 600 = 0.4-0.5), shifted to 22°C and induced by IPTG with final concentration of 0.5 mM. After grown for additional 12 h, the cells were harvested by centrifugation. The cell pellet was suspended in PBS and seven cycles of 30 s of sonication in ice were applied for cell disruption. The cell extract was centrifuged for 30 min to remove nucleic acid precipitates and final extract was purified with HisÁBind Ni-NTA resin (Novagen, Madison, WI) according to manufacturer's instructions. Recombinant thioredoxin (Trx), thioredoxin reductase (TrxR), Mrx1 and mycothione reductase (Mtr) proteins were prepared as reported previously [25]. Purified recombinant proteins were dialyzed against PBS at 4°C overnight and stored at -80°C until use.

Peroxidase activity assay
The catalytic properties of Ohr with H 2 O 2 and CHP as the substrate was determined as described previously [10] by monitoring the rate of NADPH oxidation. The reaction mixtures (300 μl) contained 50 mM sodium phosphate (pH 7.4), 50 mM NaCl, 1 mM DTPA (Diethylene triamine pentacetate acid, pH 7.4), 0.2 mM NADPH, 0.1 μM Ohr, an electron donor, and 1 mM peroxides(H 2 O 2 or CHP). The Trx system (Trx 0-120 μM + TrxR 5 xM), Mrx1 (Mrx1 0-120 μM + Mtr 5 μM + MSH 500 μM), Lpd/SucB system (Lpd 0-120 μM + SucB 5 μM) and LpdA system (0-120 μM) were used as the electron donor system in the assays, respectively. All reactions were performed at 37°C and were initiated by the addition of H 2 O 2 or CHP following 5 min pre-incubation, and oxidation of NADPH was monitored at 340 nm (A 340 ). The amount of NADPH oxidized was calculated as μM s -1 . Negative controls (without Mrx1, Trx, or peroxide) were run in parallel. The catalytic parameters for one substrate were obtained by varying its concentration at saturating concentrations of the other substrate (peroxide between 0 and 1 mM, or Trx, Mrx1, LpdA and Lpd 40 μM). The activity was determined after subtracting the spontaneous reduction rate observed in the absence of Ohr. Three independent experiments were performed at each substrate concentration.
Peroxidase activity was also analyzed by determining the consumption of peroxides with the ferrous iron xylenol orange (FOX) assay as previously described [29]. Reactions were initiated by the addition of thiol compounds and stopped at different intervals by addition of 20 μl HCl (1 M) into 100 μl reaction mixtures. The resulting mixture was mixed with 880 μl freshly prepared FOX reagent and incubated at 37°C for 20 min. The absorbance at 560 nm (A 560 ) of each sample was determined after the color reaction had reached equilibrium.

MALDI-TOF MS-MS analysis
Ohr was incubated with 10 mM DTT, 10 mM H 2 O 2 and 5 mM CHP for 30 min at room temperature. The resulting proteins were subjected to nonreducing SDS-PAGE, and Coomassie brilliant blue stained bands were excised, in-gel digested with trypsin, and analyzed by MAL-DI-TOF MS-MS (Voyager-DE STR; Applied Biosystems).

Ohr inactivation by NEM treatment
Recombinant His 6 -Ohr protein (2 mg/ml) was treated in 1 mM NEM for 1 h at room temperature and dialyzed against phosphate buffer (20 mM, pH 7.4) as described previously [12]. The concentration of His 6 -Ohr was determined by the Bradford Protein Assay Kit (Bio-Rad, Hercules, CA) with bovine serum albumin (BSA) as the standard.

Determination of sulfenic acid formation
Determination of sulfenic acid (R-SOH) in Ohr proteins was performed by the TNB anion method described before [30]. TNB was known to react with sulfenic acids in a 1:1 stoichiometry, generating a mixed disulfide between TNB and a cysteine residue. Ohr wild-type (10 μM) and variants (10 μM) preincubated with H 2 O 2 (50 μM) were treated with a 10-fold excess of TNB, prepared by incubation of an almost equimolar mixture of DTNB and DTT [12]. The amount of TNB remained (which was equal to the total TNB minus the consumed TNB) was determined spectrophotometrically.

Sensitivity assays for oxidative agents
Efficacy of various environmental stress conditions was tested on C. glutamicum strains. Exponentially-grown C. glutamicum cultures (LB medium at 30°C) were diluted 100-fold with LB medium and various concentrations of oxidants were added into diluted cells before shaking at 30°C for 30 min. After treatment, the cultures were serially diluted and plated onto LB agar plates and colonies were counted after 36 h growth at 30°C. Percentage survival was calculated as follows: (CFU ml -1 of stressed cells/CFU ml -1 of cells without stress) ×100. All the assays were performed in triplicate.

Measurement of intracellular ROS levels
2',7'-dichlorofluorescein diacetate (DCFH-DA)-based assay described previously [31] was used to quantify levels of ROS in vivo with minor modifications. Briefly, aerobically grown cells (A 600 = 1.6) were harvested by centrifugation, washed and resuspended in 50 mM PBS (pH 7.4) prior to pre-incubation with 2 μM DCFH-DA at 28°C for 20 min. Various concentration of stress inducers were added and incubated for another 30 min. Cells were washed twice with 50 mM PBS and resuspended in the same buffer. Fluorescence intensity was measured by Spectromax spectrofluorimeter (Molecular Devices, Sunnyvale, CA) with excitation at 502 nm and emission at 521 nm.

Determination of cellular levels of protein carbonylation
Protein carbonylation assay was performed based on the method described previously [32] with minor modifications. Overnight-grown C. glutamicum strains were treated with various oxidants for 30 min with shaking at 30°C. Harvested cells were washed and resuspended in 25 mM Tris-HCl (pH 8.0) containing protease inhibitor cocktails (Sigma-Aldrich, St. Louis, MO), and sonication was performed to obtain a clear cell lysate. The soluble protein fraction was collected by centrifugation and protein concentration was measured by using the Bradford assay according to the manufacturer's protocol (Bio-Rad, Hercules, CA). Protein carbonylation levels were detected with an OxyBlot Protein Oxidation Detection Kit (Millipore, Billerica, MA) based upon the manufacturer's instructions, which measures carbonyl groups of proteins generated by oxidative reactions. Carbonyl groups in proteins were derivatized with 2,4-dinitrophenyl hydrazine (DNPH), and 20 μg of each DNPH-derivatized protein were loaded and electrophoresis was conducted on a 15% SDS-PAGE gel. After electrophoresis, DNPH-derivatized proteins were electroblotted onto nitrocellulose membranes, and immunodetection of DNPH-derivatized proteins was done using a rabbit antidinitrophenyl antibody (1:500; Millipore, Billerica, MA).

MSH purification and determination
MSH was purified as previously described with thiopropyl sepharose 6B and Sephadex LH-20 chromatography and the concentration of purified MSH was determined with the thiol-specific fluorescent-labeling HPLC method using commercial GSH as the thiol standard [25].

Construction of chromosomal fusion reporter strains and βgalactosidase assay
The lacZ fusion reporter plasmid pK18mobsacB-P ohr ::lacZ was transformed into the wild-type C. glutamicum and the ΔsigH mutant by electroporation, and the chromosomal pK18mob-sacB-P ohr ::lacZ fusion reporter strains were selected by plating on LB agar plates supplemented with kanamycin [25]. β-galactosidase activities were assayed with ONPG as the substrate [33]. The β-galactosidase data represent the means of one representative assay performed in triplicate, and the error bars represent the standard deviation. Statistical analysis was carried out with Student's t-test.

C. glutamicum Ohr is a Cys-based thiol-dependent peroxidase
We identified the gene encoding a putative Ohr (NCgl0023, GI:23308767) by running a BLAST search and analyzing the genomic sequence. Ohr is located between base pairs 24,295 and 24,732 of the C. glutamicum genome (and is thus 438 nucleotides in length), and it encodes a protein of 145 amino acid residues with a theoretical molecular mass of 14.9 kDa. Ohr shares 53%, 46%, and 48% amino acid sequence identity with the Ohr proteins of Vibrio cholerae, X. fastidiosa, and Deinococcus radiodurans, respectively, and the ohr gene is present as a single copy (unlike the ohr genes of Bacillus subtilis [34] and Streptomyces coelicolor [35], both of which have 2-3 copies). Sequence homology analysis showed that C. glutamicum Ohr has two conserved Cys residues at positions 60 and 124, both of which are in domains highly conserved among Ohr homologs [12]. Cys 60 , bracketed by several hydrophobic residues, lies in hydrophobic domain 1, and Cys 124 is located in a VCP motif of hydrophilic domain 2 [12] (Fig 1).
To explore the biochemical activities of C. glutamicum Ohr, a recombinant protein was expressed in E. coli BL21 as an N-terminal His 6 -tagged fusion protein. The purified recombinant displayed as two bands on SDS-PAGE, both of which were approximately 17 kDa in size (Fig 2A, lane 1), consistent with the theoretical molecular mass. We hypothesized that the upper band might correspond to native Ohr protein and the lower to oxidized Ohr produced during purification. We confirmed this to be the case by showing that Ohr treated with H 2 O 2 (Fig 2A, lane 3) and CHP (Fig 2A, lane 4) migrated to the position of the lower band, which was completely absent after treatment with DTT (Fig 2A, lane 2). MALDI-TOF MS-MS analysis further confirmed this hypothesis (Fig 2B-2D). The presence of 3,252.4-Da and 4,852.3-Da peaks, with molecular masses of the 39-to-70 peptide (ALGGSGEGTNPEQLFAVGYAAC 60 FH MHSVAR) and the 85-to-131 peptide (VSIGPNGAGGFEIAVELEVSIPQLPQAEAQELA DAAHQVC 124 PYSNATRGNIS), respectively, indicated that both Cys 60 and Cys 124 were in the thiol state in DTT-treated Ohr (Fig 2B). These two peaks were absent with oxidized Ohr. Also, a novel peak of 8,102.3 Da (Fig 2C), thus 2.4 Da smaller in size than the sum of the sizes of the 39-to-70 peptide (3,252.4 Da) and the 85-to-131 peptide (4,852.3 Da), was observed with both H 2 O 2 - (Fig 2C) and CHP-treated Ohr (Fig 2D), indicating formation of an intramolecular disulfide bond between Cys 60 and Cys 124 . This rendered Ohr more compact, thus associating with a faster migration rate on nonreducing SDS-PAGE gels (Fig 2A).
We next explored whether Ohr decomposes peroxides (Fig 3). Each reaction was initiated by addition of DTT; Ohr attacked peroxides only in the presence of DTT [12]. The Ohr-specific activities were 7.9 and 0.62 μm/min/ng, respectively, when CHP and H 2 O 2 served as substrates. Thus, Ohr was approximately 12-fold more active against CHP than H 2 O 2 . The peroxidase activity was strongly inhibited by pretreatment with N-ethylmaleimide (NEM) for 1 h, indicating that the conserved Cys residues Cys 60 and Cys 124 are essential for peroxide decomposition.

Ohr peroxidase activity requires the Cys residues
To further explore the roles played by the two Cys residues, we constructed Cys 60 and Cys 124 variants. As shown in Fig 4A, the activity against CHP was completely abolished in the C60S and C124S strains. However, when H 2 O 2 served as the substrate, the C124S variant exhibited some peroxidase activity, but the C60S variant did not. Thus, both Cys residues play essential roles in CHP decomposition, and Cys 60 is also critical in terms of H 2 O 2 decomposition.
Both Cys 60 and Cys 124 are essential for intramolecular disulfide bond formation, as H 2 O 2and CHP-treated C60S and C124S variants migrated at the same position of DTT-treated wild-type protein, but slower than the oxidized wild-type protein, on nonreducing SDS-PAGE gels (Fig 4B). To identify the peroxidatic (C P ) Cys, TNB [2-nitro-5-thiobenzoic acid] was used to identify sulfenic acid intermediates (R-SOH) in wild-type, C60S, and C124S Ohr. As shown in Fig 4C, the C124S protein formed a sulfenic acid intermediate, but neither the wild-type nor C60S variant did so, indicating that Cys 60 is the peroxidatic Cys and is thus more reactive than Cys 124 (Fig 4C). These data suggest that, during catalysis, Cys 60 first reacts with a peroxide with concomitant formation of a sulfenic acid intermediate, which is then attacked by Cys 124 , triggering formation of an intramolecular disulfide bond between Cys 60 and Cys 124 .
The LpdA, Lpd/SucB and Trx/TrxR systems support the peroxidase activity of C. glutamicum Ohr As reported earlier and confirmed in the present study, Ohr-mediated reduction of organic peroxides to less toxic organic alcohols is associated with formation of oxidized Ohr containing an intramolecular disulfide bond [12]. To complete the catalytic cycle in vitro, various reducing agents can be used to regenerate the sulfhydryl groups at the active Cys sites. Recently, Ohr regeneration in vivo (in X. fastidiosa [10] and M. smegmatis [13]) was shown to be mediated by a dedicated reducing system featuring lipoylated proteins (Lpd and SucB). Thus, we explored whether a lipoyl-dependent reducing system supports Ohr regeneration in C. glutamicum. Also, we determined whether the classical reducing systems of C. glutamicum, thus the Trx/ TrxR and Mrx1/Mtr/MSH systems, support the peroxidase activity of Ohr. The Mrx1/Mtr/ MSH reducing system is unique to MSH-producing high-G+C Gram-positive Actinobacteria, being functionally equivalent to the widespread Grx/GR/GSH system of most Gram-negative bacteria [25,26]. The catalytic constants of Ohr in the presence of Trx, Mrx1, or LpdA and Lpd/ SucB as recycling reductants were determined under steady-state conditions at saturating concentrations of peroxides (1 mM) and different concentrations of the reductants (0-120 μM). As shown in Table 1, the k cat values of Ohr-mediated CHP reduction in the presence of the Trx, LpdA, and Lpd/SucB systems were 2.32±0.10 s -1 , 5.31±0.03 s -1 , and 4.72±0.19 s -1 , respectively; and the respective K m values were 24.51±0.83 μM, 12.13±0.51 μM, and 4.65±0.37 μΜ, respectively. The catalytic efficiencies were thus 9.50±0.74×10 4 M -1 s -1 , 43.83±1.57×10 4 M -1 s -1 , and 100.89±3.99×10 4 M -1 s -1 , respectively. However, all three reducing systems facilitated Ohr activity only poorly when H 2 O 2 was used as the substrate. Thus, the LpdA and Lpd/SucB systems more effectively support Ohr peroxidase activity than does the Trx system when CHP is the substrate; the Lpd/SucB system had the highest regenerative activity. Under all conditions evaluated, the Mrx1/Mtr/MSH reducing system failed to support Ohr activity when either CHP or H 2 O 2 was the substrate ( Table 1). Ohr activities were also measured in the presence of fixed concentrations (40 μM) of the Trx, LpdA, and Lpd/SucB reducing systems and different concentrations of peroxides (0-1 mM) ( Table 2). As expected, Ohr reduced CHP at least 10,000-fold more efficiently than H 2 O 2 when LpdA and Lpd/SucB provided the reducing power, and at least 1,700-fold more efficiently when Trx served to that end. Together, the data suggest that Ohr can be regenerated by both the lipoyl-dependent and the classic Trx/TrxR reducing systems of C. glutamicum, but the lipoylated proteins (LpdA and Lpd/SucB) are the prime electron donors.

Survival of the ohr mutant is compromised by organic peroxides
Ohr from other bacterial species plays an important role in the decomposition of organic hydroperoxides [5,11,17]. Therefore, we speculated that Ohr from C. glutamicum has the same function. To confirm a role of Ohr in mediating resistance to oxidative stress, we constructed a Δohr deletion mutant and assessed the sensitivity thereof to CHP and H 2 O 2 (Fig 5). The Δohr mutant was not significantly less resistant to H 2 O 2 than was the wild-type strain at even a high H 2 O 2 concentration (150 mM) (Fig 5A). However, upon addition of 6 mM CHP,   survival of the Δohr mutant was only approximately 70% that of the wild-type, and this decreased further to 25% with 10 mM CHP ( Fig 5B). Moreover, the hypersensitivity of the Δohr mutant to both H 2 O 2 and CHP was partially restored when the genomic mutation was complemented by the wild-type gene of plasmid pXMJ19-ohr (Fig 5). These results show that the Ohr of C. glutamicum plays an important role in the defense against organic peroxides, as does Ohr of other bacteria [5,11,36].

C. glutamicum Ohr reduces intracellular ROS levels in vivo
Of the various types of ROS, organic peroxides are particularly toxic, partly because these compounds engage in free radical reactions generating reactive organic radicals that in turn damage membranes and other macromolecules [2]. To explore the physiological role played by Ohr in removal of ROS when C. glutamicum is oxidatively stressed, intracellular ROS levels were measured using the fluorogenic probe 2',7'-dichlorodihydrofluorescein diacetate. As shown in Fig 6A, the wild-type strain had a significantly lower ROS level than that of the Δohr mutant Peroxidase assays were performed as described in after exposure to CHP for 30 min. However, the ROS level in the Δohr mutant was partially restored to that of the wild-type by introducing the complementation plasmid pXMJ19-ohr ( Fig 6A). The ROS levels of Δohr mutant also increased when treated with H 2 O 2 (100 mM and 150 mM) compared to the wild-type, although the increase is not as strong as induced by CHP treatment (Fig 6B). These findings suggest that C. glutamicum Ohr inhibits intracellular ROS accumulation triggered by peroxides. Cys thiol groups are especially susceptible to modification by ROS that elude antioxidant defense systems, causing irreversible formation of methionine sulfoxide and Cys disulfide bonds, eventually triggering protein carbonylation [37]. As Ohr inhibits intracellular ROS accumulation by C. glutamicum, we explored whether Ohr protects cells against protein carbonylation under conditions of oxidative stress. We prepared total protein suspensions of wild-type and Δohr mutant strains grown in the presence of CHP or H 2 O 2 . Protein carbonyl groups were derivatized with 2,4-dinitrophenyl hydrazine, and these derivatives were detected via Western blotting using an anti-dinitrophenyl antibody. As expected, the extent of protein carbonylation in the wild-type was significantly lower than that of the Δohr mutant after growth under CHP stress. However, this was not the case with H 2 O 2 stress (Fig 6C). Together, the data indicate that Ohr plays an important role in protection against oxidative stress induced by organic hydroperoxides; Ohr inhibits intracellular ROS accumulation and protein carbonylation.

CHP induces Ohr expression
As Ohr promoted C. glutamicum survival under conditions of oxidative stress, we measured ohr expression in the presence of CHP or H 2 O 2 using a chromosomal P ohr ::lacZ fusion reporter. As shown in Fig 7A, the wild-type P ohr promoter activity increased by 58% and 77% upon exposure to 6 mM and 10 mM CHP, respectively, compared with the untreated control. However, H 2 O 2 did not obviously induce ohr expression ( Fig 7B). These data show that the organic peroxide CHP specifically induced ohr expression, directly contributing to CHP tolerance.
Such specific induction of ohr by CHP suggests that a transcriptional regulator might be at play. In many bacteria, ohr expression is regulated by OhrR, a transcriptional repressor of the MarR family [34,38]. However, no ohrR homolog is evident in the C. glutamicum genome. Recently, several studies have found that the stress-responsive extracytoplasmic functionsigma (ECF-σ) factor SigH regulates the expression of many oxidative stress resistance genes in C. glutamicum [39,40]. We thus measured ohr-lacZ expression in a ΔsigH mutant. However, ohr expression did not obviously differ between the ΔsigH mutant and the wild-type strain exposed to either CHP or H 2 O 2 (Fig 7). Thus, ohr expression in C. glutamicum appears to be regulated by an unknown mechanism that responds specifically to organic peroxides.

Discussion
Our results show that ohr plays an essential role in organic hydroperoxide reduction in C. glutamicum. Ohr deletion significantly increased bacterial sensitivity to the organic hydroperoxide CHP, and such sensitivity was nearly restored to wild-type levels upon complementation with the ohr gene. The Δohr mutant was clearly more sensitive to CHP than to H 2 O 2 . Also, intracellular ROS accumulation and the associated protein carbonylation were significantly higher (compared with wild-type) in the Δohr mutant exposed to CHP, but only slight increases were evident upon exposure to H 2 O 2 , suggesting that Ohr plays a vital role in the decomposition of organic hydroperoxides. These results are in line with the roles played by Ohr in other organisms, specifically in imparting resistance to organic peroxides [5,11].
C. glutamicum Ohr contains two conserved Cys residues. Cys 60 , located within a very hydrophobic environment, is directly involved in peroxide reduction, whereas Cys 124 (located in a hydrophilic environment) is the resolving Cys [12]. During catalysis, Cys 60 first reacts with a peroxide, with concomitant formation of a sulfenic acid intermediate, which is then attacked by Cys 124 , leading to formation of an intramolecular disulfide bond between Cys 60 and Cys 124 and release of a molecule of water. The disulfide bond is directly reduced by either the Trx/ TrxR or the lipoyl-dependent reducing system to complete the catalytic cycle. Recently, Ohr of X. fastidiosa [10] and M. smegmatis [13] was shown to be regenerated by a special cellular reducing system featuring lipoylated proteins (Lpd and SucB). In the present study, we confirmed that oxidized C. glutamicum Ohr is also efficiently reduced by LpdA-and Lpd/SucB-coupled systems, suggesting that lipoylated protein-based reducing systems reduce the Ohr enzymes of various bacteria. Indeed, the Lpd and SucB reducing systems also serve as electron donors for OsmC and AhpC [10,41]. To further explore reducing systems donating electrons to Ohr, we determined whether the classical Trx/TrxR and Mrx1/Mtr/MSH systems supported Ohr peroxidase activity. Although earlier work indicated that neither Trx nor Grx reduced the Ohr of X. fastidiosa, it is important to note that this bacterium has at least three Grx-and four Trx-encoding genes. Therefore, the possibility that (an) alternative Trx and/or Grx donate(s) electrons to Ohr cannot be excluded. In fact, the Trx/TrxR system functioned weakly, but not negligibly, to reduce C. glutamicum Ohr; the LpdA and Lpd/SucB systems were much more efficient. The Mrx/Mtr/MSH system did not support the Ohr activity of M. smegmatis [13]. We showed here that this was also true of C. glutamicum ( Table 2).
Traditionally, Ohr is regulated by OhrR, a transcriptional repressor of the MarR family [42]. However, we could not identify an OhrR homolog in C. glutamicum, suggesting that ohr is regulated by a different mechanism. In B. subtilis, OhrR mutation eliminates regulation of ohrA, but not ohrB. Indeed, ohrB expression is regulated in a σ B -dependent manner, suggesting that ohr can also be regulated by sigma factors [34]. Recently, several studies found that the stressresponsive ECF-σ factor SigH regulates the expression of many oxidative stress resistance genes in C. glutamicum [39,40]. Interestingly, both the basal and inducible expression of sigB in C. glutamicum was reported to be regulated by SigH [43]. These findings prompted us to explore whether Ohr of C. glutamicum was also regulated by SigH. However, we found no obvious difference in ohr expression levels between the ΔsigH mutant and the wild-type strain exposed to either CHP or H 2 O 2 . Thus, ohr expression in C. glutamicum may be regulated by an unknown mechanism responding specifically to organic hydroperoxides.
In conclusion, we found that C. glutamicum Ohr was essential for conferring effective resistance to a highly toxic organic hydroperoxide, which it decomposed more efficiently than did hydrogen peroxide. These results improve our knowledge of stress resistance in C. glutamicum and indicate how future robust industrial strains may be engineered.
Supporting Information S1