Crystal structure of peroxiredoxin 3 from Vibrio vulnificus and its implications for scavenging peroxides and nitric oxide
Peroxiredoxins (Prxs) are ubiquitous cysteine-based peroxidase enzymes. Recently, a new type of Prx, VvPrx3, was identified in the pathogenic bacterium Vibrio vulnificus as being important for survival in macrophages. It employs only one catalytic cysteine residue to decompose peroxides. Here, crystal structures of VvPrx3 representing its reduced and oxidized states have been determined, together with an H2O2-bound structure, at high resolution. The crystal structure representing the reduced Prx3 showed a typical dimeric interface, called the A-type interface. However, VvPrx3 forms an oligomeric interface mediated by a disulfide bond between two catalytic cysteine residues from two adjacent dimers, which differs from the doughnut-like oligomers that appear in most Prxs. Subsequent biochemical studies showed that this disulfide bond was induced by treatment with nitric oxide (NO) as well as with peroxides. Consistently, NO treatment induced expression of the prx3 gene in V. vulnificus, and VvPrx3 was crucial for the survival of bacteria in the presence of NO. Taken together, the function and mechanism of VvPrx3 in scavenging peroxides and NO stress via oligomerization are proposed. These findings contribute to the understanding of the diverse functions of Prxs during pathogenic processes at the molecular level.
Most organisms have evolved efficient defence systems that rapidly detoxify potentially damaging reactive oxygen species (ROS). ROS are generated during aerobic respiration and in the host immune system (Temple et al., 2005; Cabiscol et al., 2000; Halliwell, 2006). Peroxiredoxins (Prxs) represent a superfamily of thiol-specific antioxidant peroxidases that use an invariant cysteine residue(s) (Engelman et al., 2013). They are highly conserved throughout evolution and are ubiquitously found from humans to bacteria. Prxs decompose peroxides such as hydrogen peroxide (H2O2) and alkyl hydroperoxides in the cytosol (Rhee et al., 2005; Wood et al., 2003; Engelman et al., 2013). Recent research has reported that Prxs are also involved in the detoxification of reactive nitrogen species (RNS), which are produced in phagocytes as a key component for killing invading bacteria (Wong et al., 2002; Bryk et al., 2000). Prxs have been found to detoxify peroxinitrite (ONOO−) and nitrosylated thiols (Fang, 2004; Beckman & Koppenol, 1996; Pfeiffer et al., 1997).
Prxs have a peroxidatic cysteine (CP) that reacts with peroxides in the first step of catalysis. The CP is oxidized to cysteine sulfenic acid (CP-SOH) by attacking the O—O bond of peroxides during the peroxidase reaction, resulting in the decomposition of peroxides (Wood et al., 2003). The CP-SOH is then reduced during the resolution step, which is needed to recycle the protein. Depending on the resolution step, Prxs are classified into three types, referred to as typical 2-Cys, atypical 2-Cys and 1-Cys Prxs (Hall et al., 2011). Both types of 2-Cys Prx have an additional conserved cysteine residue called the resolving cysteine (CR), which forms a stable intermolecular or intramolecular disulfide bond with the CP-SOH. While CR is located in a separate subunit in typical 2-Cys Prxs, it is located within the same subunit in atypical 2-Cys Prxs (Seo et al., 2000; Hall et al., 2009). The disulfide bond in 2-Cys Prxs can be broken into two free thiols by cellular reductants (Chae et al., 1999). In contrast, the CP-SOH of 1-Cys Prxs has been implicated to be directly reduced by cellular reductants such as glutaredoxin (Grx) and/or the reduced form of glutathione (GSH) (Noguera-Mazon et al., 2006; Kang et al., 1998; Rhee et al., 2005). In a yeast 1-Cys Prx, the CP-SOH has been proposed to form a disulfide bond with the CP residue of another molecule of Prx (Pedrajas et al., 2016).
Although the oligomeric states of Prxs differ depending on the type of Prx, most oligomeric Prxs are formed via two major distinct interfaces: B-type and A-type interfaces (Parsonage et al., 2005). In typical 2-Cys Prxs, dimers are primarily formed via the B-type interface, in which CP and CR are located in different protomers and disulfide formation between CP and CR is facilitated by peroxides (Echalier et al., 2005). In most 2-Cys Prxs, five dimers are further associated via the A-type interface to form a doughnut-like decamer (Wood et al., 2002; Parsonage et al., 2005). Most atypical 2-Cys Prxs and 1-Cys Prxs exhibit dimeric assembly via the A-type interface (Evrard et al., 2004; Li et al., 2005). Exceptionally, a non-A and non-B oligomeric interface between CP and CR was observed in human Prx5 (HsPrx5), which belongs to the atypical 2-Cys Prx family (Smeets et al., 2008) and showed redox-dependent structural changes in the CP-containing region. The structural change was distinguished from that in some typical 2-Cys Prxs (Wood et al., 2003; Parsonage et al., 2005). In the presence of excess peroxides the CP can be overoxidized to cysteine sulfinic acid or cysteine sulfonic acid (CP-SO2H or CP-SO3H, respectively). Interestingly, the protomers with overoxidized CP exhibit the same conformation as in Prxs in the reduced state (Mizohata et al., 2005; Nakamura et al., 2006).
Vibrio vulnificus is a highly virulent foodborne pathogen that can cause fatal septicaemia (Oliver, 2005; Horseman & Surani, 2011). The bacterium has three different types of Prx: Prx1, Prx2 and Prx3. Prx1 and Prx2 belong to the typical 2-Cys Prx family of proteins and their expression is controlled by the H2O2-activated transcription factor OxyR (Kim et al., 2014; Jo et al., 2015). V. vulnificus Prx3 (VvPrx3) belongs to the 1-Cys class of Prxs, with one conserved catalytic cysteine (Cys48), and has been shown to be involved in pathogenicity in a mouse model. According to the PeroxiRedoxin Classification Index (PREX) database (https://www.csb.wfu.edu/prex), it belongs to the Prx5 subfamily. VvPrx3 expression is controlled by IscR, which is activated by oxidative stress and iron starvation (Lim et al., 2014). In this study, we determined crystal structures of VvPrx3 in two different forms and in complex with H2O2. Based on the crystal structures, we present a molecular mechanism of 1-Cys Prxs that highlights the oligomer interface containing a disulfide bond between two peroxidatic cysteine residues. More importantly, we examined the expression of prx3 in V. vulnificus induced by nitric oxide gas (NO); consistently, VvPrx3 reacts with nitric oxide.
The VvPrx3 (C48D/C73S) and VvPrx3 (C73S) genes of V. vulnificus were cloned into the pET-21c vector (Invitrogen). To clone the VvPrx3 gene for the expression of VvPrx3 (C73S), two primers (forward, 5′-GGTCATATGATCGCTCAAGGCCAAACTTTACC-3′; reverse, 5′-CCTCTCGAGCGCGGCAAGAATCGTTTCAGC-3′) were designed, which contained NdeI and XhoI sites (underlined), respectively. Site-directed mutagenesis was performed in two subsequent PCR reactions for the expression of VvPrx3 (C48D/C73S) (Ho et al., 1989). The PCR products and pET-21c plasmid were digested by NdeI and XhoI, and the digested products were ligated with DNA ligase. The ligation products were transformed into Escherichia coli C43 (DE3) cells by heat shock. Finally, the recombinant plasmid was confirmed by DNA sequencing.
Recombinant proteins were expressed in LB medium containing 50 µg ml−1 ampicillin, with growth at 310 K to an OD600 of 0.5. Samples were induced with 0.5 mM isopropyl β-D-1-thiogalactopyranoside (IPTG) for 6 h at 303 K. The cells were harvested and resuspended in lysis buffer consisting of 20 mM Tris–HCl pH 8.0, 0.15 M NaCl, 2 mM β-mercaptoethanol. After disrupting the cells with a French press, the cell debris was removed by centrifugation. The supernatant was loaded onto Ni–NTA affinity resin (Qiagen, The Netherlands) that had been pre-incubated with lysis buffer. The target protein was eluted with lysis buffer supplemented with 250 mM imidazole. Eluted recombinant proteins were diluted threefold with 20 mM Tris–HCl pH 8.0 buffer containing 2 mM β-mercaptoethanol and loaded onto a HiTrap Q column (GE Healthcare, USA). A linear gradient of increasing NaCl concentration was applied to the column. The fractions containing the protein were pooled, concentrated and applied onto a size-exclusion chromatography column (HiLoad 16/600 Superdex 200 pg, GE Healthcare) pre-equilibrated with lysis buffer. The final protein samples were concentrated to 20 mg ml−1 using a centrifugal filter concentration device (Millipore; 10 kDa cutoff) and stored frozen at 193 K until use.
2.3. Crystallization, structure determination and refinement
Crystallization was performed at 287 K using the hanging-drop vapour-diffusion method. Crystals of Prx3 (C48D/C73S) were obtained using a precipitant solution consisting of 0.2 M NaCl, 0.8 M sodium citrate, 0.1 M Tris–HCl pH 6.5. The structure of reduced Prx3 (C48D/C73S) was determined using the molecular-replacement method with MOLREP in the CCP4 package (Winn et al., 2011) using a putative thioredoxin reductase from Burkholderia cenocepacia (PDB entry 4f82; Seattle Structural Genomics Center for Infectious Disease, unpublished work) as a search model. The crystal of Prx3 (C48D/C73S) obtained under reduced conditions [reduced Prx3 (C48D/C73S)] belonged to space group P3221, with unit-cell parameters a = 73.9, b = 73.9, c = 62.3 Å and one protein molecule in the asymmetric unit (Table 1). After 5 d, Prx3 (C73S) crystals appeared with a precipitant solution consisting of 0.1 M sodium citrate pH 5.5, 0.2 M ammonium acetate, 22%(w/v) PEG 4000. Crystals were exchanged into the appropriate mother liquor containing 20%(v/v) glycerol, mounted on cryo-loops and flash-cooled in a liquid-nitrogen stream at 100 K. X-ray diffraction data sets were collected on beamline 5C at the Pohang Accelerator Laboratory and were processed with the HKL-2000 package (Otwinowski & Minor, 1997). The crystal of Prx3 (C73S) obtained under oxidized conditions [oxidized Prx3 (C73S)] belonged to space group P212121, with unit-cell parameters a = 39.5, b = 57.4, c = 124 Å and two protein molecules in the asymmetric unit (Table 1). The oxidized Prx3 (C73S) crystals were flash-cooled using a crystallization solution with the same cryoprotectant as used for the reduced Prx3 (C48D/C73S) crystals in a nitrogen stream at 100 K. The initial model of oxidized Prx3 (C73S) was determined by molecular replacement using the structure of reduced Prx3 (C48D/C73S) as a model and was refined at 2.0 Å resolution (Table 1). The final structure of oxidized Prx3 (C73S) was refined at 1.48 Å resolution using PHENIX (Afonine et al., 2012). To obtain H2O2-bound Prx3 (C48D/C73S) crystals, 20 mM H2O2 was added to the reservoir solution and 500 µM H2O2 was added to the hanging-drop solution containing the crystals. After 10 d, the H2O2-bound Prx3 (C48D/C73S) crystals were flash-cooled using the same cryoprotectant as used for the reduced Prx3 (C48D/C73S) crystals in a nitrogen stream at 100 K. The crystals belonged to space group P1, with unit-cell parameters a = 75.08, b = 97.72, c = 97.49 Å and 12 protein molecules in the asymmetric unit. Statistics regarding data collection and processing are presented in Table 1.
To prepare the reduced VvPrx3 (C73S) protein, 10 mM DTT was added to the purified VvPrx3 (C73S) protein and was then removed using a HiPrep 26/10 Desalting column (GE Healthcare, USA) pre-equilibrated with 20 mM Tris–HCl pH 8.0 buffer containing 150 mM NaCl. The reduced VvPrx3 (C73S) mutant protein was reacted with various concentrations of H2O2 or tertiary butyl hydrogen peroxide (t-BOOH) for 30 min at 298 K in 1 ml 20 mM Tris–HCl buffer pH 8.0 containing 150 mM NaCl. The reaction mixture was treated with iodoacetamide (IAA) to stop the reaction and subjected to SDS–PAGE under nonreducing or reducing conditions. To analyze Prx3 oxidation during the time course, samples were treated with 50 µM H2O2 for 0–60 s and the reaction was quenched with 10 mM N-ethylmaleimide (NEM). For reaction with NO, reduced VvPrx3 (C73S) protein (10 nmol) was treated with 1 mg NO-releasing PLGA–PEI nanoparticles (NO/PPNPs) in 10 ml reaction mixture in 20 mM Tris pH 7.5, 150 mM NaCl. The amount of NO was calculated based on the release of 3.3 nmol NO per hour by 1 mg NO/PPNPs (Nurhasni et al., 2015).
To prepare the reduced VvPrx3 (C73S) protein, the purified protein was treated with 10 mM DTT for 30 min at room temperature. After reduction of the protein, residual DTT was removed using a HiPrep 26/10 Desalting (GE Healthcare, USA) column pre-equilibrated with 20 mM Tris–HCl pH 7.5 buffer containing 150 mM NaCl. Reaction mixtures containing 5 µM horseradish peroxidase (HRP; Sigma–Aldrich) and various concentrations (0–9.15 µM) of reduced VvPrx3 (C73S) were treated with 2.5 µM H2O2 at room temperature in reaction buffer consisting of 50 mM sodium phosphate pH 7.4, 150 mM NaCl. The concentration of HRP was calculated using the absorbance at 403 nm (∊403 = 1.02 × 105 M−1 cm−1) and the concentration of VvPrx3 (C73S) was determined at 280 nm (∊VvPrx3 = 8.48 × 103 M−1 cm−1). The competitive kinetics of VvPrx3 (C73S) were determined using a previously described procedure (Ogusucu et al., 2007; Winterbourn & Peskin, 2016; Cox et al., 2009). In brief, the ratio of inhibition of HRP oxidation [F/(1 − F)] was measured at 403 nm with a spectrophotometer using a 10 mm cuvette. The second-order rate constants (k1) of VvPrx3 (C73S) were determined from the slope of a plot of [F/(1 − F)]kHRP[HRP] against [VvPrx3 (C73S)] (kHRP = 1.7 × 107 M−1 s−1).
A Superose 6 HR 10/30 column coupled to an FPLC instrument (Biologic Duo Flow, Bio-Rad) was equilibrated with 20 mM Tris–HCl pH 8.0, 150 mM NaCl. Samples were incubated with 10–30 µM H2O2 for 30–60 min at room temperature before loading them onto the column to form the intermolecular disulfide bonds of VvPrx3 (C73S). Samples were subjected to chromatography, which was performed at a flow rate of 0.2 ml min−1. Several standard preparations were run to calibrate the column: ferritin (440 kDa), aldolase (158 kDa), conalbumin (75 kDa), ovalbumin (43 kDa), carbonic anhydrase (29 kDa) and ribonuclease A (13.7 kDa) (Supplementary Fig. S8). The absorbance at 280 nm was used to monitor the presence of the proteins.
The V. vulnificus cultures were grown to an A600 of 0.5 in LBS broth [LB broth supplemented with 2.0%(w/v) NaCl] at 30°C and the cells were harvested by centrifugation at 1500g for 7 min at room temperature. The cell pellets were resuspended in M9 minimal medium supplemented with 0.4%(w/v) glucose. The bacterial cells [∼4 × 107 colony-forming units (CFU)] were exposed to 0.1%(w/v) NO/PPNPs (Nurhasni et al., 2015) for 2 h and an aliquot was acquired every 20 min. The number of live bacterial cells in the aliquots was determined as the CFU on LBS agar plates. Bacterial strains and plasmids are described in Supplementary Table S1.
Wild-type V. vulnificus cells grown to an A600 of 0.5 in LBS broth were exposed to various concentrations of PLGA–PEI nanoparticles (PPNPs) and NO/PPNPs for 20 min and were harvested to isolate the total RNA. Total RNAs were isolated using an RNeasy Mini kit (Qiagen). For quantitative real-time PCR (qRT-PCR), the concentration of total RNA from the strains was measured using a NanoVue Plus spectrophotometer (GE Healthcare). The cDNA was synthesized from 1 µg total RNA using an iScript cDNA-synthesis kit (Bio-Rad). Real-time PCR amplification of cDNA was performed using a Chromo 4 real-time PCR detection system (Bio-Rad) with a pair of specific primers (Supplementary Table S2) as described previously (Jang et al., 2016). The relative expression levels of iscR and prx3 mRNA in the same amount of total RNA were calculated using the 16S rRNA expression level as an internal reference for normalization.
3.1. Structural determination of reduced Prx3 (C48D/C73S), oxidized Prx3 (C73S) and H2O2-bound Prx3 (C48D/C73S)
VvPrx3 contains two cysteine residues: Cys48 and Cys73. Cys48 is the CP residue that is crucial for the function of the protein. However, Cys73 is not involved in protein function because the enzymatic activity of VvPrx3 (C73S) was similar to that of wild-type VvPrx3 in the GSH/Grx3/GR system (Lim et al., 2014). To focus on the function of Cys48 and prevent unwanted disulfide-bond formation at Cys73 during SDS–PAGE analyses, Cys73 was replaced with a serine residue in this study. To investigate an oxidized structure of Prx3 containing a disulfide bond, the crystal of the oxidized VvPrx3 (C73S) protein was obtained under crystallization conditions without a reducing agent. The crystal structure of VvPrx3 (C73S) [named oxidized Prx (C73S) in this study] was determined at 1.5 Å resolution in space group P212121 using the molecular replacement method. The resulting structure was refined to an Rwork of 0.181 and an Rfree of 0.198 (Table 1). The final model comprises residues 1–157 and contains two protein molecules in the asymmetric unit (Table 1). To study the structure of Prx3 in the reduced state, crystals of a VvPrx3 (C48D/C73S) variant were also obtained. The CP Cys48 in VvPrx3 was replaced with a non-oxidizable Asp residue, which could also partly mimic an over-oxidized form (Cys-SO2H) of VvPrx3 (Jo et al., 2015). The crystal belonged to a different space group, P3221, and the VvPrx3 (C48D/C73S) structure [named reduced Prx (C48D/C73S) in this study] was determined by the molecular replacement approach. The final model, refined at 1.9 Å resolution (Rwork = 0.220 and Rfree = 0.258), contains one protomer (residues 1–18 and residues 22–157) in the asymmetric unit (Table 1). To examine the interactions between VvPrx3 and peroxides, the H2O2-bound Prx3 (C48D/C73S) structure was determined at 1.9 Å resolution. A high concentration (500 µM) of H2O2 was incubated with crystals of reduced Prx3 (C48D/C73S), which changed the space group and unit-cell parameters. The resulting structure was refined to an Rwork of 0.195 and an Rfree of 0.244. Electron-density maps were well defined for all 12 protomers in the asymmetric unit (Table 1).
All three crystal structures of VvPrx3 revealed a protomer exhibiting a compact and spherical structure without a C-terminal extension region, similar to the previously reported 1-Cys Prx AhpE from Mycobacterium tuberculosis (Supplementary Figs. S1 and S2; Li et al., 2005). The VvPrx3 protomer is formed based on a central five-stranded β-sheet (β5–β4–β3–β8–β9; Fig. 1a). α-Helices α2 and α5 are located on one side of the central β-sheet, while α-helix α4 and the β-hairpin (β1 and β2) are on the other side of the central β-sheet. Strands β4–β3–β8–β9 and helices α2, α4 and α5 comprise the typical Trx fold. In the peripheral region, a β4–α3–β5 structural motif is located in the A-type dimeric interface (Figs. 1a and 1b). CP Cys48 (or the mutated Asp48) is located at the N-terminus of the long, kinked α2, similar to other Prxs. The mutated Ser73 is buried in the Trx fold region (Fig. 1a). Structural superposition of reduced Prx (C48D/C73S) on human PrxV (PDB entry 1oc3; sequence identity 52%; Evrard et al., 2004) revealed high structural similarity, with an r.m.s.d. of 1.46 Å between the Cα positions of 247 matched amino acids (Supplementary Fig. S2).
All three crystal structures of VvPrx3 [reduced Prx3 (C48D/C73S), oxidized Prx3 (C73S) and H2O2-bound Prx3 (C48D/C73S)] showed a similar dimeric assembly in the crystals characterized by an A-type dimeric interface, similar to that in AhpE from M. tuberculosis (Supplementary Figs. S1 and S2; Li et al., 2005). In the reduced Prx3 (C48D/C73S) structure the Cp-containing α2 helix at the active site adopted the `fully folded' or reduced conformation usually observed in Prx structures in the reduced and overoxidized states (Fig. 2; Mizohata et al., 2005; Schröder et al., 2000).
Strikingly, a dimeric interface was found in the oxidized Prx3 (C73S) that was distinct from both the A-type and B-type interfaces (Figs. 1a and 1b). This dimeric interface was similar to an exceptional oligomeric interface observed in the oxidized structure of HsPrx5, although this structure contains the disulfide bond between CP and CR (Smeets et al., 2008), while VvPrx3 directly forms a dimeric interface through the intermolecular disulfide bond between the CP Cys48 residues. This interface was designated the `C-type' interface in this study since these oligomeric interfaces was induced by the disulfide bonds.
The C-type interface mainly consists of two CP-containing α2 helices from adjacent protomers (Fig. 1c). Hydrophobic interactions were observed in the C-type interface that appear to further stabilize this dimeric interface. The Phe145 residue in loop β9–α5 of one protomer forms a hydrophobic core with Phe117 in loop β6–β7, while Pro41 and Pro46 form a hydrophobic core in the loop connecting β3 and α2 of the other molecule (Fig. 2a). Owing to disulfide bonds and hydrophobic interactions, a significant conformational alteration was observed in the active site of helix α2 containing CP and nearby loops. In addition, loop β9–α5 moved outwards from the core and loop β3–α2 became ordered in the oxidized Prx3 (C73S) structure representing the oxidized structure, compared with the reduced Prx3 (C48D/C73S) structure representing the overoxidized or reduced conformation (Fig. 2b). In HsPrx5, the conformational change of loop β7–α6 corresponding to loop β9–α5 of VvPrx3 was induced by forming an intramolecular disulfide bond depending on the redox state (Supplementary Fig. S3).
3.3. Intermolecular disulfide formation of VvPrx3 reacted with peroxides, H2O2 and t-BOOH
To confirm whether the observed disulfide bond in the crystal structure is generated in the solution state, we performed the following biochemical experiment with H2O2 and t-BOOH, which are known substrates of VvPrx3 (Dubbs & Mongkolsuk, 2007; Beckman & Koppenol, 1996; Lim et al., 2014). Reduced VvPrx3 (C73S) protein was incubated with H2O2 or t-BOOH and analyzed by SDS–PAGE under nonreducing conditions to observe intermolecular disulfide-bond formation. While boiling the protein sample for SDS–PAGE, free cysteine residues appeared to form nonspecific disulfide bonds. To avoid these nonspecific disulfide bonds, the free thiol-specific alkylating agent iodoacetamide was added to the sample prior to boiling to block free cysteines. We observed that the protein bands were up-shifted to a dimeric size on the SDS–polyacrylamide gels on treatment with H2O2 or t-BOOH (Figs. 3a and 3b). The protein bands shifted back to a monomeric size when a reducing agent was added to cleave disulfide bonds during boiling. These results indicate that two VvPrx3 protomers were linked by an intermolecular disulfide bond between two Cys48 residues. Interestingly, the intensities of the dimeric protein bands decreased at higher concentrations of peroxide. Maximum band intensities were shown at 10 µM H2O2 or t-BOOH (Figs. 3a and 3b), likely owing to overoxidation of CP at high peroxide concentrations (Claiborne et al., 1999; Rhee et al., 2007). These findings indicate that the intermolecular disulfide bond between CP residues is induced by the peroxides H2O2 and t-BOOH, unlike the prevailing mechanisms for 1-Cys Prxs via the direct reduction of CP-SOH by Grx or GSH (Noguera-Mazon et al., 2006; Kang et al., 1998; Rhee et al., 2005).
The second-order reaction rate constant (k1) for H2O2 consumption by VvPrx3 was measured by employing competition kinetics with horseradish peroxidase (HRP; Winterbourn & Peskin, 2016; Cox et al., 2009; Ogusucu et al., 2007). The measured value was 5.6 × 107 M−1 s−1, which was of the same order as that for human Prx2 (Peskin et al., 2007; Fig. 3c). This finding indicates that VvPrx3 is as efficient as 2-Cys Prxs. To estimate the reaction rate of dimerization of VvPrx3, the amount of dimeric VvPrx3 was measured using the same SDS–PAGE method with different incubation times of VvPrx3 and H2O2. We observed that almost half of the VvPrx3 formed an intermolecular disulfide bond within 30 s (Fig. 3d), indicating that the reaction velocity for disulfide-bond formation between VvPrx3-SOH and VvPrx3-SH is comparable to a similar reaction with human Prx2 or Prx3 (Gupta & Carroll, 2014). Taken together, our findings suggest that dimerization of VvPrx3 could provide a rapid kinetic path to resolving CP and is central to the catalysis of the enzyme.
3.4. Redox-dependent oligomerization of VvPrx3
A protein oligomer representing the oxidized state was found in the oxidized Prx3 (C73S) crystals (Fig. 1c). The oligomer is generated in a linear arrangement by alternating connections between the protomers via A-type contacts and C-type contacts (Supplementary Fig. S4). As observed in Fig. 1(c), peroxides induce a disulfide bond between protomers through the C-type interface. The A-type dimer in the reduced state is further linked by the disulfide bond between the CP residues at the C-type interface, which can lead to the linear oligomer observed in oxidized Prx3 (C73S) crystals.
To investigate changes in the oligomeric states of VvPrx3 proteins induced by oxidative stress, the oligomeric state and disulfide-bond formation of VvPrx3 proteins in the presence/absence of H2O2 was analyzed by combining size-exclusion chromatography and SDS–PAGE. In these experiments, two Prx variants were used: C73S and C48D/C73S. To create a disulfide bond between the CP residues, VvPrx3 (C73S) protein was treated with H2O2 before loading it onto a size-exclusion chromatograpy column pre-equilibrated with lysis buffer devoid of reducing agent. The elution profile showed a broad early-eluting peak and shoulders before achieving the dimeric size of VvPrx3, reflecting mixed forms of higher oligomers. Decamers, hexamers, tetramers and dimers were observed on the size-exclusion chromatography column when the molecular sizes were estimated based on the elution volume. The higher oligomeric forms in solution were further confirmed by native gel electrophoresis (Supplementary Fig. S5). The higher oligomeric forms contained more VvPrx3 dimers that are linked by a disulfide bond between two CP residues, as judged by SDS–PAGE under nonreducing conditions (Fig. 4). These results indicate that the higher oligomeric forms of VvPrx3 were created by noncovalently associated A-type interfaces and disulfide-linked C-type interfaces, as observed in crystals of oxidized Prx3 (C73S).
Ovoid-shaped electron density indicating H2O2 was found near Asp48 and the active-site region in the H2O2-bound Prx3 (C48D/C73S) structure (Fig. 5 and Supplementary Fig. S6). The H2O2 binding site is fully accessible from the solvent since it is exposed to the external medium. The structure is analogous to that of the H2O2-bound archaeal 2-Cys Prx thioredoxin peroxidase from Aeropyrum pernix K1 (Nakamura et al., 2010), in which the corresponding cysteine residue was overoxidized and shares the H2O2 binding site (Supplementary Fig. S6d). Thus, it is likely that substrate binding near CP is important in the rapid catalysis of VvPrx3. In Prx structures only one O atom of H2O2 near Cys48 (or Asp48 in our structure) makes further interactions with VvPrx3, while the other O atom makess no interaction with H2O2-bound Prx3 (C48D/C73S) (Fig. 5). Both O atoms in H2O2 were extensively involved in polar interactions in OxyR, which can recognize only H2O2 and not alkyl hydrogen peroxides (Jo et al., 2015). This difference in H2O2 recognition might result in the different substrate specificities of Prx and OxyR proteins.
Increasing attention has been given to the NO-mediated reactions of host immune systems (Bogdan et al., 2000; Nathan & Shiloh, 2000). A large amount of NO is synthesized in macrophages or epithelial cells stimulated by pathogens or by substances secreted by pathogens (Beckman & Koppenol, 1996; Fang, 2004). To investigate the functional relevance of VvPrx3 and NO, we examined whether NO can induce the disulfide bonding of VvPrx3, similar to H2O2 (Engelman et al., 2013). Although Prxs are known to scavenge reactive nitrogen species such as the peroxinitrite that is rapidly generated by combining NO and superoxide ion (Fang, 2004; Beckman & Koppenol, 1996; Pfeiffer et al., 1997), direct treatment with NO has not been tested on Prxs. We treated V. vulnificus with NO-releasing polymeric nanoparticles (NO/PPNPs), which have recently been developed to allow the sustained release of NO in solution (Nurhasni et al., 2015), and then measured the mRNA levels of prx3 and its known transcription factor iscR. The results showed that the mRNA levels of prx3 and iscR significantly increased, implying that NO induces VvPrx3 in V. vulnificus via IscR (Fig. 6a).
We next tested whether NO directly induces disulfide bonds in the VvPrx3 protein. NO/PPNPs were treated with VvPrx3 protein, and the disulfide-bond formation of the protein was analyzed by SDS–PAGE and size-exclusion chromatography (Fig. 6b and Supplementary Fig. S7). Compared with air-oxidized VvPrx3 protein, significant protein bands representing the disulfide bond-linked VvPrx3 increased on the SDS–polyacrylamide gel under nonreducing conditions after treatment of the nanoparticles releasing ∼1 µM NO. The NO concentration in solution is much lower than the effective concentration (∼10 µM) of H2O2 for VvPrx3 (Fig. 6b). Thus, our results suggest that VvPrx3 could respond to NO stress through a similar mechanism as for H2O2 stress.
To evaluate the role of VvPrx3 in resistance to NO in bacteria, the survival rates of wild-type and prx3 gene-deleted V. vulnificus strains were measured. Compared with the wild-type strain, the Prx3 mutant strain exhibited substantially impaired growth in a minimal medium containing NO/PPNPs. When the functional prx3 gene (pJK1303) was added back to the mutant strain, the bacterial survival rate was restored (Fig. 6c). Taken together, our results suggest that VvPrx3 plays an important role in scavenging NO imposed on the bacteria.
Crystal structures of VvPrx3, a bacterial 1-Cys Prx, were determined in different redox states in this study. The H2O2 binding site near the CP residue was also shown. Although structural and sequence similarities to the 1-Cys Prx AhpE from M. tuberculosis were identified, especially in the dimeric assembly with the A-type interface, our structures provide unexpected and important insights into the function and molecular mechanism of Prxs. One striking feature of oxidized Prx3 (C73S) is a dimeric C-type interface with structural changes, which was previously observed in the oxidized form of HsPrx5 belonging to the atypical 2-Cys Prx family (Smeets et al., 2008). Since the C-type interface has been observed in two different kinds of Prxs, this oligomeric interface could be shared by many other Prx proteins. This study further observed that protein–protein interaction at the C-type interface was induced by both NO and peroxides. Owing to the oxidation-dependent C-type interface, VvPrx3 formed a higher order linear oligomer in response to peroxides and NO. This linear oligomeric structure is distinct from the circular decameric assembly of typical 2-Cys Prxs, which is generated by alternate A-type and B-type interfaces. Since the linear oligomer of VvPrx3 is generated only in the oxidized state, the redox-dependent oligomerization of VvPrx3 differs from that of typical 2-Cys Prxs. Interestingly, this linear oligomeric form was also predicted in HsPrx5 (Smeets et al., 2008).
According to the prevailing mechanisms, the CP in 1-Cys Prxs is itself oxidized to CP-SOH through attack of the substrate peroxide. CP-SOH is then reduced by the reductase Grx or GSH (Noguera-Mazon et al., 2006; Kang et al., 1998; Rhee et al., 2005). This study presents another pathway for the reduction of CP-SOH in 1-Cys Prxs, which might complement the direct reduction of CP-SOH by Grx or GSH under certain circumstances. In this pathway, the catalytic cysteine residue (CP) in 1-Cys Prx plays a dual role in the catalytic cycle. Not all VvPrx3 proteins are simultaneously oxidized by H2O2. Thus, remaining or unreacted VvPrx3 could act on other VvPrx3s with CP-SOH. The CP-SOH can be reduced and resolved by free CP in a different dimer of VvPrx3, in which the free CP acts as a CR. The complementary structural features at the C-type interface provide specific interactions with VvPrx3 proteins, facilitating the prompt formation of a disulfide bond between CP-SOH and the free CP of VvPrx3 proteins (Fig. 1c). This reaction could also occur successively, leading to the formation of the linear oligomers observed in the crystal structure (Supplementary Fig. S4). Finally, cellular GSH or Grx that is reduced by GSH reduces the disulfide bond in the oligomer, as proposed previously (Kim et al., 2014). NO can be held in the H2O2 binding site of the VvPrx3 active site to be readily attacked by CP. The free CP from another VvPrx3 molecule resolves the resulting nitrosylated cysteine (CP-NO), as depicted in Fig. 7. Many different members are present in the 1-Cys Prx family. Of these, 1-Cys Prxs from yeast have previously been characterized. As in VvPrx3, the intermolecular disulfide bond between CP was observed in yeast 1-Cys Prx on an SDS–polyacrylamide gel under nonreducing conditions and predicted a doughnut-like oligomer formation in the oxidized state (Pedrajas et al., 2016). We speculate that yeast 1-Cys Prx might share a catalytic cycle with VvPrx3, although further studies are required to elucidate its mechanism.
The pathogenic bacterium V. vulnificus can defend against NO stress to increase survival in this harsh host environment. A previous study showed that VvPrx3 is important in the pathogenicity of V. vulnificus in mice (Lim et al., 2014). In this study, we observed that NO induced prx3 together with iscR in V. vulnificus and that NO is a substrate of VvPrx3. These findings suggest that VvPrx3 is important for bacterial survival in NO-challenged environments and might account for the involvement of VvPrx3 in pathogenicity in the host.
Diverse peroxiredoxins exist in nature and play different roles in the scavenging of toxic radicals. We have proposed a molecular mechanism for bacterial 1-Cys Prxs and a new role in scavenging NO in this study, which may in part account for the pathogenesis of V. vulnificus. Our findings expand the understanding of the molecular mechanism of survival of pathogenic bacteria in the host environment by scavenging diverse oxidative and nitrosative stresses.
This study made use of beamlines 5C and 7A at Pohang Light Source, Pohang, Republic of Korea. All authors declare that there is no conflict of interest regarding this study. Author contributions: JA, KKJ, IJ, SHC and N-CH designed the research; JA, KKJ, HN and J-WY performed the research; JA, KKJ, IJ, SHC and N-CH analyzed the data; JA, KKJ and N-CH wrote the paper.
This work was supported by the Korea Institute of Planning and Evaluation for Technology in Food, Agriculture, Forestry (IPET) through Agriculture, Food and Rural Affairs Research Center Support Program (to SHC: 710012-03-1-SB110 and NCH: 710012-03-1-HD120), funded by Ministry of Agriculture, Food and Rural Affairs(MAFRA) and by Mid-career Researcher Program Grant 2015R1A2A1A13001654 (to SHC) through the National Research Foundation funded by the Ministry of Science, ICT and Future Planning, Republic of Korea.
Afonine, P. V., Grosse-Kunstleve, R. W., Echols, N., Headd, J. J., Moriarty, N. W., Mustyakimov, M., Terwilliger, T. C., Urzhumtsev, A., Zwart, P. H. & Adams, P. D. (2012). Acta Cryst. D68, 352–367. Web of Science CrossRef CAS IUCr Journals Google Scholar
Beckman, J. S. & Koppenol, W. H. (1996). Am. J. Physiol. 271, C1424–C1437. CAS PubMed Google Scholar
Bogdan, C., Röllinghoff, M. & Diefenbach, A. (2000). Curr. Opin. Immunol. 12, 64–76. CrossRef PubMed CAS Google Scholar
Bryk, R., Griffin, P. & Nathan, C. (2000). Nature (London), 407, 211–215. Web of Science PubMed CAS Google Scholar
Cabiscol, E., Tamarit, J. & Ros, J. (2000). Int. Microbiol. 3, 3–8. PubMed CAS Google Scholar
Chae, H. Z., Kang, S. W. & Rhee, S. G. (1999). Methods Enzymol. 300, 219–226. Web of Science CrossRef CAS PubMed Google Scholar
Claiborne, A., Yeh, J. I., Mallett, T. C., Luba, J., Crane, E. J. III, Charrier, V. & Parsonage, D. (1999). Biochemistry, 38, 15407–15416. Web of Science CrossRef PubMed CAS Google Scholar
Cox, A. G., Peskin, A. V., Paton, L. N., Winterbourn, C. C. & Hampton, M. B. (2009). Biochemistry, 48, 6495–6501. CrossRef PubMed CAS Google Scholar
Dubbs, J. M. & Mongkolsuk, S. (2007). Subcell. Biochem. 44, 143–193. CrossRef PubMed Google Scholar
Echalier, A., Trivelli, X., Corbier, C., Rouhier, N., Walker, O., Tsan, P., Jacquot, J.-P., Aubry, A., Krimm, I. & Lancelin, J.-M. (2005). Biochemistry, 44, 1755–1767. Web of Science CrossRef PubMed CAS Google Scholar
Engelman, R., Weisman-Shomer, P., Ziv, T., Xu, J., Arnér, E. S. & Benhar, M. (2013). J. Biol. Chem. 288, 11312–11324. CrossRef CAS PubMed Google Scholar
Evrard, C., Capron, A., Marchand, C., Clippe, A., Wattiez, R., Soumillion, P., Knoops, B. & Declercq, J.-P. (2004). J. Mol. Biol. 337, 1079–1090. CrossRef PubMed CAS Google Scholar
Fang, F. C. (2004). Nat. Rev. Microbiol. 2, 820–832. CrossRef PubMed CAS Google Scholar
Gupta, V. & Carroll, K. S. (2014). Biochim. Biophys. Acta, 1840, 847–875. CrossRef CAS PubMed Google Scholar
Hall, A., Karplus, P. A. & Poole, L. B. (2009). FEBS J. 276, 2469–2477. Web of Science CrossRef PubMed CAS Google Scholar
Hall, A., Nelson, K., Poole, L. B. & Karplus, P. A. (2011). Antioxid. Redox Signal. 15, 795–815. Web of Science CrossRef CAS PubMed Google Scholar
Halliwell, B. (2006). Plant Physiol. 141, 312–322. CrossRef PubMed CAS Google Scholar
Ho, S. N., Hunt, H. D., Horton, R. M., Pullen, J. K. & Pease, L. R. (1989). Gene, 77, 51–59. CrossRef CAS PubMed Web of Science Google Scholar
Horseman, M. A. & Surani, S. (2011). Int. J. Infect. Dis. 15, e157–e166. CrossRef PubMed Google Scholar
Jang, K. K., Gil, S. Y., Lim, J. G. & Choi, S. H. (2016). J. Biol. Chem. 291, 5774–5787. CrossRef CAS PubMed Google Scholar
Jo, I., Chung, I.-Y., Bae, H.-W., Kim, J.-S., Song, S., Cho, Y.-H. & Ha, N.-C. (2015). Proc. Natl Acad. Sci. USA, 112, 6443–6448. Web of Science CrossRef CAS PubMed Google Scholar
Kang, S. W., Baines, I. C. & Rhee, S. G. (1998). J. Biol. Chem. 273, 6303–6311. Web of Science CrossRef CAS PubMed Google Scholar
Kim, S., Bang, Y.-J., Kim, D., Lim, J. G., Oh, M. H. & Choi, S. H. (2014). Mol. Microbiol. 93, 992–1009. CrossRef CAS PubMed Google Scholar
Li, S., Peterson, N. A., Kim, M.-Y., Kim, C.-Y., Hung, L.-W., Yu, M., Lekin, T., Segelke, B. W., Lott, J. S. & Baker, E. N. (2005). J. Mol. Biol. 346, 1035–1046. CrossRef PubMed CAS Google Scholar
Lim, J. G., Bang, Y.-J. & Choi, S. H. (2014). J. Biol. Chem. 289, 36263–36274. CrossRef CAS PubMed Google Scholar
Mizohata, E., Sakai, H., Fusatomi, E., Terada, T., Murayama, K., Shirouzu, M. & Yokoyama, S. (2005). J. Mol. Biol. 354, 317–329. Web of Science CrossRef PubMed CAS Google Scholar
Nakamura, T., Kado, Y., Yamaguchi, T., Matsumura, H., Ishikawa, K. & Inoue, T. (2010). J. Biochem. 147, 109–115. Web of Science CrossRef PubMed CAS Google Scholar
Nakamura, T., Yamamoto, T., Inoue, T., Matsumura, H., Kobayashi, A., Hagihara, Y., Uegaki, K., Ataka, M., Kai, Y. & Ishikawa, K. (2006). Proteins, 62, 822–826. Web of Science CrossRef PubMed CAS Google Scholar
Nathan, C. & Shiloh, M. U. (2000). Proc. Natl Acad. Sci. USA, 97, 8841–8848. Web of Science CrossRef PubMed CAS Google Scholar
Noguera-Mazon, V., Lemoine, J., Walker, O., Rouhier, N., Salvador, A., Jacquot, J.-P., Lancelin, J.-M. & Krimm, I. (2006). J. Biol. Chem. 281, 31736–31742. PubMed CAS Google Scholar
Nurhasni, H., Cao, J., Choi, M., Kim, I., Lee, B. L., Jung, Y. & Yoo, J.-W. (2015). Int. J. Nanomedicine, 10, 3065–3080. CAS PubMed Google Scholar
Ogusucu, R., Rettori, D., Munhoz, D. C., Soares Netto, L. E. & Augusto, O. (2007). Free Radical Biol. Med. 42, 326–334. CrossRef CAS Google Scholar
Oliver, J. D. (2005). Epidemiol. Infect. 133, 383–391. CrossRef PubMed CAS Google Scholar
Otwinowski, Z. & Minor, W. (1997). Methods Enzymol. 276, 307–326. CrossRef CAS PubMed Web of Science Google Scholar
Parsonage, D., Youngblood, D. S., Sarma, G. N., Wood, Z. A., Karplus, P. A. & Poole, L. B. (2005). Biochemistry, 44, 10583–10592. Web of Science CrossRef PubMed CAS Google Scholar
Pedrajas, J. R., McDonagh, B., Hernandez-Torres, F., Miranda-Vizuete, A., Gonzalez-Ojeda, R., Martinez-Galisteo, E., Padilla, C. A. & Barcena, J. A. (2016). Antioxid. Redox Signal. 24, 115–128. CrossRef CAS PubMed Google Scholar
Peskin, A. V., Low, F. M., Paton, L. N., Maghzal, G. J., Hampton, M. B. & Winterbourn, C. C. (2007). J. Biol. Chem. 282, 11885–11892. CrossRef PubMed CAS Google Scholar
Pfeiffer, S., Gorren, A. C., Schmidt, K., Werner, E. R., Hansert, B., Bohle, D. S. & Mayer, B. (1997). J. Biol. Chem. 272, 3465–3470. CrossRef CAS PubMed Google Scholar
Rhee, S. G., Chae, H. Z. & Kim, K. (2005). Free Radical Biol. Med. 38, 1543–1552. CrossRef CAS Google Scholar
Rhee, S. G., Jeong, W., Chang, T. S. & Woo, H. A. (2007). Kidney Int. Suppl., pp. S3–S8. Google Scholar
Schröder, E., Littlechild, J. A., Lebedev, A. A., Errington, N., Vagin, A. A. & Isupov, M. N. (2000). Structure, 8, 605–615. Web of Science PubMed Google Scholar
Seo, M. S., Kang, S. W., Kim, K., Baines, I. C., Lee, T. H. & Rhee, S. G. (2000). J. Biol. Chem. 275, 20346–20354. Web of Science CrossRef PubMed CAS Google Scholar
Smeets, A., Marchand, C., Linard, D., Knoops, B. & Declercq, J.-P. (2008). Arch. Biochem. Biophys. 477, 98–104. CrossRef PubMed CAS Google Scholar
Temple, M. D., Perrone, G. G. & Dawes, I. W. (2005). Trends Cell Biol. 15, 319–326. CrossRef PubMed CAS Google Scholar
Winn, M. D. et al. (2011). Acta Cryst. D67, 235–242. Web of Science CrossRef CAS IUCr Journals Google Scholar
Winterbourn, C. C. & Peskin, A. V. (2016). Mol. Cells, 39, 26–30. CAS PubMed Google Scholar
Wong, C.-M., Zhou, Y., Ng, R. W. M., Kung, H.-F. & Jin, D.-Y. (2002). J. Biol. Chem. 277, 5385–5394. CrossRef PubMed CAS Google Scholar
Wood, Z. A., Poole, L. B., Hantgan, R. R. & Karplus, P. A. (2002). Biochemistry, 41, 5493–5504. Web of Science CrossRef PubMed CAS Google Scholar
Wood, Z. A., Schröder, E., Robin Harris, J. & Poole, L. B. (2003). Trends Biochem. Sci. 28, 32–40. CrossRef PubMed CAS Google Scholar
This is an open-access article distributed under the terms of the Creative Commons Attribution (CC-BY) Licence, which permits unrestricted use, distribution, and reproduction in any medium, provided the original authors and source are cited.