Báo cáo khoa học: Methanoferrodoxin represents a new class of superoxide reductase containing an iron–sulfur cluster docx

Methanoferrodoxin represents a new class of superoxide reductase containing an iron–sulfur cluster Christian Kra ¨ tzer 1 , Cornelia Welte 1 , Katerina Do ¨ rner 2 , Thorsten Friedrich 2 and Uwe Deppenmeier 1 1 Institut fu ¨ r Mikrobiologie und Biotechnologie, Universita ¨ t Bonn, Germany 2 Institut fu ¨ r Organische Chemie und Biochemie, Albert-Ludwigs-Universita ¨ t, Freiburg, Germany Introduction Oxidative stress is caused by reactive oxygen species such as hydrogen peroxide (H 2 O 2 ), the hydroxyl radi- cal (OH ) ) and the superoxide anion radical (O 2 ) ), which are generated by the partial reduction of oxygen [1]. Bacteria deal with oxidative stress with a set of detoxifying enzymes. Superoxide dismutases (SODs) were the first enzymes known to eliminate superoxide by disproportionation to hydrogen peroxide and diox- ygen [1]. Superoxide reductases (SORs) are a new fam- ily of enzymes that were discovered in sulfate-reducing bacteria of the Desulfovibrio genus [2,3], and catalyze the reduction of superoxide to peroxide. SORs are pre- dominantly found in anaerobic or microaerophilic bacteria such as Desulfovibrio desulfuricans [2] and Clostridium acetobutylicum [4], or anaerobic archaeons such as Archaeoglobus fulgidus [5] and Pyrococcus furiosus [6]. In the past decade, SORs from these organisms and others have been studied in detail, and a considerable amount of biochemical, crystallographic and spectroscopic information has been reported [7]. Methanosarcina mazei is one of the methanogenic archaeons, which are characterized by the ability to generate methane as the major end product of energy metabolism [8]. Many Methanosarcina strains are able to utilize H 2 +CO 2 , methylated C 1 compounds or acetate as energy and carbon sources, and are essential for closing the cycle of organic matter on earth in anaerobic environments. Methanogens are generally considered to be sensitive towards aeration with oxygen. However, it has been reported that some methanogens are surprisingly oxygen-stable, and can survive exposure to air for several hours, with the Keywords detoxification; iron–sulfur protein; methanogenic archaea; oxygen radicals; superoxide dismutase Correspondence U. Deppenmeier, Institut fu ¨ r Mikrobiologie und Biotechnologie, Universita ¨ t Bonn, Meckenheimer Allee 168, 53115 Bonn, Germany Fax: +49 228737576 Tel: +49 228735590 E-mail: udeppen@uni-bonn.de (Received 5 October 2010, revised 8 November 2010, accepted 17 November 2010) doi:10.1111/j.1742-4658.2010.07964.x Protein MM0632 from the methanogenic archaeon Methanosarcina mazei showed strong superoxide reductase activity and rapidly decomposed superoxide radicals to peroxides. The superoxide reductase activity of the heterologously produced enzyme was determined by a cytochrome c assay and in a test system with NADPH, ferredoxin:NADP + reductase, and rubredoxin. Furthermore, EPR spectroscopy showed that MM0632 is the first superoxide reductase that possesses an iron–sulfur cluster instead of a second mononuclear iron center. We propose the name methanoferrodoxin for this new class of superoxide reductase with an [Fe(NHis) 4 (SCys)] site as the catalytic center and a [4Fe–4S] cluster as second prosthetic group that is probably involved in electron transfer to the catalytic center. Methanosarcina mazei grows only under anaerobic conditions, but is one of the most aerotolerant methanogens. It is tempting to speculate that methanoferrodoxin contributes to the protection of cells from oxygen radicals formed by flavoproteins during periodic exposure to oxygen in nat- ural environments. Abbreviations NROR, NADH-rubredoxin oxidoreductase; SOD, superoxide dismutase; SOR, superoxide reductase. 442 FEBS Journal 278 (2011) 442–451 ª 2010 The Authors Journal compilation ª 2010 FEBS Methanosarcina species appearing to be the most aero- tolerant methanogens [9]. The ability to cope with oxi- dative stress is consistent with the finding that methanogens are widespread in habitats that are peri- odically exposed to oxygen, such as paddy soils. In this article, we report that the protein MM0632 from M. mazei reductively decomposes superoxide to peroxide. This SOR activity of the heterologously pro- duced enzyme was determined with a cytochrome c and an NADPH-dependent assay, respectively [4,6,10,11]. EPR spectroscopy revealed that MM0632 possesses the typical catalytic nonheme [Fe(N- His) 4 (SCys)] center and an iron–sulfur cluster instead of a second mononuclear iron center. We propose the name methanoferrodoxin for this SOR, and, together with homologous proteins from other methanogenic archaeons, this enzyme should be classified as a class IV SOR. Results One mechanism protecting oxygen-sensitive bacteria and archaeons from toxic oxygen reduction products involves reduction of superoxide, rather than the clas- sical disproportionation for its removal that occurs in aerobic microorganisms [19,20]. A close inspection of the genome of M. mazei [8] revealed the presence of an ORF (MM0632) with significant similarities to genes encoding SORs. Figure 1 shows alignments of selected SORs from dif- ferent prokaryotes in comparison with the amino acid sequence of MM0632 from M. mazei. Different classes of SORs can be distinguished. Desulfoferrodoxins or class I SORs (also named 2Fe-SORs), represented by enzymes from D. desulfuricans and A. fulgidus (Aful2), contain a small N-terminal desulforedoxin-type domain (domain I) with a rubredoxin-type [Fe(SCys) 4 ] mono- nuclear iron center, and a larger C-terminal domain similar to neelaredoxin (domain II), with an [Fe(N- His) 4 (SCys)] mononuclear iron center. Treponema palli- dum contains a variant of desulfoferrodoxin (class III SOR), composed of the C-terminal domain and a N-terminal domain that does not contain an [Fe(SCys) 4 ] center. As evident from the alignment, the critical Cys residues for binding iron center I (Fig. 1, asterisks) are absent from MM0632. In contrast, the four nitrogen ligands of His residues and the sulfur of one Cys are conserved in MM0632 (Fig. 1, black boxes), indicating that the neelaredoxin-type iron center is present. This hypothesis is supported by the align- ment of neelaredoxins (class II SORs or 1-Fe-SORs) from different prokaryotes and MM0632 (Fig. 1). Again the mononuclear iron-coordinating residues for the [Fe(NHis) 4 (SCys)] center are present in all proteins. Closer inspection of the alignment revealed that an inser- tion in the middle of the neelaredoxin-like domain II of MM0632 occurred that contains three Cys residues [C(x) 7 CxxC motif]. Furthermore, there was an extension of the C-terminal end of MM0632 containing one addi- tional Cys (Fig. 1). These Cys residues are the only ones present besides the Cys that coordinates the [Fe(N- His) 4 (SCys)] center. Homologous insertions and exten- sion were identified in proteins from the methanogenic archaeons Methanococcoides burtonii, Methanohalophi- lus mahii and Methanohalobium evestigatum (Fig. 1). This interesting finding prompted the question of whether a second metallocenter could be coordinated by the four Cys residues next to the catalytic [Fe(N- His 4 )(SCys)] center. A combined phi-psi-blast [21] search indicated that the C(x) 7 CxxC motif is also pres- ent in some ferredoxin-type proteins and iron–sulfur binding domain proteins [e.g. YP_002890971 (NapF) from Thauera sp. and YP_002990434 from Desulfovib- rio salexigens]. It was also evident from the alignment (Fig. 1) that the sequences containing the iron-binding motifs were highly conserved, whereas the remaining parts of the proteins showed only a low degree of simi- larity. This was especially true for the C-terminal ends of the proteins, with MM0632 being characterized by an extension that contained a high percentage of charged amino acids (Glu, Arg and Lys). Protein properties The gene encoding MM0632 from M. mazei was cloned into pASK-IBA3 and heterologously overex- pressed in Escherichia coli. The produced protein was purified to apparent homogeneity in a single step with a Strep-Tactin affinity matrix. A molecular mass of 20 kDa was found by SDS ⁄ PAGE, consistent with the expected mass of 19.2 kDa of the protein monomer (not shown). The native enzyme had a molecular mass of 19 kDa when analyzed by gel filtration, indicating a monomeric structure. Small amounts of dimers and tri- mers were also observed, but contributed to < 10% of the total protein (Fig. S1). When MM0632 was overproduced aerobically in E. coli, spectrometric analysis of the protein did not show any prosthetic groups. However, after anaerobic overproduction and purification, the protein prepara- tion had a red color when hydrogen peroxide was added to obtain the fully oxidized state (Fig. 2A). The spec- trum showed a broad increase in absorption between 420 and 550 nm, with a peak at 470 nm. This absorption vanished upon reduction with dithionate or ascorbate (Fig. 2A and inset). After reconstitution of the protein C. Kra ¨ tzer et al. Superoxide reductase from Methanosarcina mazei FEBS Journal 278 (2011) 442–451 ª 2010 The Authors Journal compilation ª 2010 FEBS 443 with Fe 3+ , sulfide and dithiothreitol, the protein eluted from the affinity column with a dark brown color. The UV–visible spectrum of the reconstituted fully oxidized protein revealed increased absorbance between 400 and 600 nm (Fig. 2B), with a broad peak at 420 nm and a shoulder at 470 nm. This absorption disappeared by reducing the preparation with sodium dithionite (Fig. 2B and inset). Reduction with ascorbate led to only a small decrease in absorbance between 400 and 600 nm (Fig. 2B). We also investigated whether the reconstitution led to unspecific binding of iron ions. For this purpose, the reconstitution was performed in the absence of sulfide, but artificial incorporation of Fe 3+ was not detected (not shown). The presence of iron–sulfur clusters was investigated by quantification of nonheme iron and acid-labile sul- fur after desalting and overnight dialysis against buf- fer W with 5 mm EDTA and 1 mm dithioerythritol to remove any possible contamination with nonenzyme bound ions. The determination of iron and sulfur yielded 5.7 ± 0.4 mol iron per mol enzyme and 4.5 ± 1.2 mol sulfur per mol enzyme, indicating the presence of a mononuclear iron center and a [4Fe–4S] cluster. In contrast, in the nonreconstituted protein, 1.3 ± 0.5 mol iron per mol enzyme was detected, and acid-labile sulfide was not found. EPR spectroscopy of MM0632 A representative EPR spectrum of reconstituted MM0632 is shown in Fig. 3. The EPR spectrum of the protein as isolated minus the spectrum of the dithio- nite-reduced sample recorded at 6 K (Fig. 3A) showed the typical EPR signal of a high-spin (S =5⁄ 2) ferric site. This signal at g = 4.3 is attributed to the [Fe(N- His) 4 (SCys)] center, and has also been detected in the SORs from P. furiosus and Desulfovibrio vulgaris, as well as in the D. desulfuricans desulfoferrodoxin [22–24]. At higher fields, the spectrum of the dithionite- reduced sample showed a distinct signal at 13 K (Fig. 3B) that was barely detectable at 40 K and was absent in the spectrum of the oxidized sample (data not shown). The axial signal with g^ = 1.93 and gk = 2.047 was assigned to a [4Fe–4S] cluster. These findings, together with those of the UV–visible spec- troscopy and iron–sulfur quantification, indicated that the recombinant MM0632 from M. mazei was success- fully produced in E. coli with a correctly incorporated [Fe(NHis) 4 (SCys)] center and the [4Fe-4S] metallocen- ter being reconstituted with Fe 3+ and sulfide. The axial signal mentioned above was disturbed by the contribution of another, as yet unknown, compound Fig. 1. Alignment of SOR sequences. Alignment of amino acid sequences was performed with CLUSTALW [40]. The GenBank accession num- bers of the proteins are as follows. Mma, M. mazei, methanoferrodoxin, MM0632: AAM30328. Mcc, Me. burtonii, methanoferrodoxin: YP_565539. Mhal, Met. mahii, methanoferrodoxin: YP_003542283. Mhalo, Meth. evestigatum, methanoferrodoxin: YP_003727000. Mac, M. acetivorans, neelaredoxin-like protein: NP_618610.1. Dgiga, Desulfovibrio gigas, neelaredoxin: O50258. Aful1, A. fulgidus, neelaredoxin- like protein: O29903. Tocea, Thermosediminibacter oceani DSM 16646, neelaredoxin-like protein: YP_003826213. Ddesulf, D. desulfuricans, desulfoferrodoxin: YP_002480584. Aful2, A. fulgidus, desulfoferrodoxin: NP_069667.1. Tpa1, T. pallidum, SOR type III: ADD72914.1. Amino acids forming the [Fe(NHis)4(SCys)] are indicated by black boxes. Cys residues involved in the coordination of the [Fe(SCys) 4 ] center are indi- cated by asterisks. The insertion and extensions of the methanoferrodoxin-like proteins are boxed. Cys residues predicted to be involved in the coordination of the [4Fe–4S] cluster are indicated by arrows. Superoxide reductase from Methanosarcina mazei C. Kra ¨ tzer et al. 444 FEBS Journal 278 (2011) 442–451 ª 2010 The Authors Journal compilation ª 2010 FEBS at g = 2.03. This signal most likely derives from a low-spin S = ½Fe 3+ species, probably at the [Fe(N- His 4 )(SCys)] site. SOR activity of MM0632 The purified MM0632 product was tested for SOR activity with the cytochrome c reduction assay. In this test system, cytochrome c was reduced by superoxide, which was provided continuously by xanthine, oxygen and xanthine oxidase. The reduction of cytochrome c was determined by the increase in absorbance at 550 nm. SOR functioned as a cytochrome c oxidase, and withdrew electrons from cytochrome c to reduce superoxide to peroxide (Fig. 4, inset). Cytochrome c reduction decreased with increasing amounts of MM0632. In contrast to the normal SODs, an excess of the protein from M. mazei caused reoxidation of reduced cytochrome c (Fig. 4). The enzymic activity of MM0632 was calculated on the basis that 1 U of SOR activity is defined by the amount of protein required to inhibit the rate of cytochrome c reduction by 50% [16]. In our test system, 72 ± 9 ng of protein led to 50% inhibition, representing an activity of 1 U. Hence, the enzyme was highly active with 13 900 ± 1700 U mg )1 protein, which is in the same order of 0 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 350 400 450 500 550 600 650 700 0 0.02 0.04 0.06 0.08 0.10 0.12 400 450 500 550 600 650 0 0.02 0.04 0.06 0.08 0.10 0.12 A B 1 3 2 1 2 3 Absorbance Absorbance Wavelength (nm) 350 400 450 500 550 600 650 700 Wavelength (nm) 400 450 500 550 600 650 Fig. 2. UV–visible spectra of M. mazei MM0632. (A) Nonreconsti- tuted protein (0.2 mg mL )1 ): (1) H 2 O 2 -oxidized; (2) dithionite- reduced; (3) ascorbate-reduced. Inset: oxidized–reduced spectrum. (B) Reconstituted protein (0.1 mg mL )1 ): (1) H 2 O 2 -oxidized; (2) ascorbate-reduced; (3) dithionite-reduced. Inset: oxidized–reduced spectrum. A Magnetic field [mT] 4.3 110 120 130 140 150 160 170 180 190 200 280 300 320 340 360 380 400 B Magnetic field [mT] 1.93 2.047 2.03 Fig. 3. X-band EPR spectra of MM0632. (A) difference spectrum obtained by subtracting the spectrum of the dithionite-reduced sample from that of the as-isolated enzyme. The spectra were recorded at 6 K and 10 mW, and five scans were accumulated for each spectrum. Other EPR conditions were as follows: microwave frequency, 9.46 GHz; modulation amplitude, 1.0 mT; time constant, 0.164 s; scan rate, 17.9 mT min )1 . (B) Spectrum of the dithionite- reduced sample recorded at 13 K and 10 mW. Other EPR condi- tions were as follows: microwave frequency, 9.46 GHz; modulation amplitude, 0.6 mT; time constant, 0.164 s; scan rate, 17.9 mT min )1 . The g-values are indicated. C. Kra ¨ tzer et al. Superoxide reductase from Methanosarcina mazei FEBS Journal 278 (2011) 442–451 ª 2010 The Authors Journal compilation ª 2010 FEBS 445 magnitude as the specific activity of the SORs from A. fulgidus [25]. In contrast, the SORs from C. acetobu- tylicum and Desulfoarculus baarsii had lower activities (160 and 53 U mg )1 protein, respectively) [4,17]. The reaction was dependent on the production of super- oxide by xanthine and xanthine oxidase (not shown). When cytochrome c was chemically reduced by sodium dithionite and used in the assay in the absence of xan- thine oxidase, no reaction was observed, even in the presence of oxygen (not shown). These experiments clearly indicated that oxygen cannot function as an electron acceptor of MM0632. Thus, the M. mazei protein functions as a cytochrome c–superoxide oxido- reductase. It has been shown that the activities of SORs signifi- cantly decrease when acetylated cytochrome is used as substrate [25]. Interestingly, the SOR from M. mazei showed no inhibition when the acetylated form of cytochrome c was used as electron donor in compari- son with the nonacetylated form of cytochrome c (not shown). The activity of nonreconstituted MM0632 missing the [4Fe–4S] cluster was also tested in the cytochrome c assay. The protein containing only [Fe(NHis) 4 (SCys)] showed an activity of 13 900 ± 2500 U mg )1 protein (not shown), which is in the same range as the activity of the reconstituted pro- tein. Hence, it is obvious that the [4Fe–4S] cluster is not necessary for cytochrome c-dependent superoxide reduction. It has been shown that SOD is able to inhibit the reduction of cytochrome c by dismutation of superox- ide to hydrogen peroxide [10]. Therefore, this enzyme could compete with SOR and horse heart cyto- chrome c for superoxide [16]. Indeed, the addition of bovine SOD showed a clear effect on the catalytic effi- ciency of SOR, because the former enzyme signifi- cantly decreased the concentration of superoxide that functions as an electron acceptor for MM0632 (Fig. 4). The standard photometric Nitro Blue tetrazo- lium-dependent SOD assay indicated that MM0632 had a slow SOD activity of 25 U mg )1 protein [18]. In C. acetobutylicum, an SOR acts as the terminal component of a superoxide detoxification system that transfers electrons from NADH to superoxide [4]. The short electron transfer chain involves NADH-rubred- oxin oxidoreductase, and a low molecular mass electron transfer protein named rubredoxin. Rubredoxin has a single [Fe(SCys) 4 ] center as active site, and is known to participate in electron transfer to SORs in various bac- teria [6,26]. This electron pathway was reconstituted in an in vitro assay with ferredoxin:NADP + reductase (FNR) from spinach as a replacement for NADH- rubredoxin oxidoreductase (Fig. 5). In our test system, FNR, which uses NADPH as substrate, could donate electrons to rubredoxin from C. acetobutylicum. Reduced rubredoxin then functioned as an electron donor for MM0632, which reduced superoxide gener- ated by xanthine ⁄ xanthine oxidase. As shown in Fig. 5, NADPH consumption was trig- gered by addition of rubredoxin. The SOR activity of MM0632 strongly depends on the presence of FNR, xan- thine oxidase and rubredoxin. No activity was observed when one of the proteins was omitted. FNR could not 0.46 0.48 0.5 0.52 0.54 0 50 100 150 200 250 1 2 3 4 5 6 Time (s) Absorbance (550 nm) Xanthine O 2 . O 2 XO SOR + O 2 O 2 . O 2 2– Cyt c red Cyt c ox Fig. 4. SOR activity of MM0632. Cytochrome c was reduced by superoxide generated by xanthine and xanthine oxidase (inset). The arrow indicates addition of MM0632 with subsequent oxidation of cytochrome c: (1) 31 ng; (2) 62 ng; (3) 93 ng; (4) 124 ng; (5) 470 ng; (6) 470 ng plus SOD (40 U). 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0 100 200 300 400 500 600 700 800 Time (s) Absorbance (340 nm) (2) (3) (1) (4) Fig. 5. Superoxide reduction as catalyzed by MM0632 with rubre- doxin as electron donor. The reaction was monitored by measuring the NADPH consumption spectrometrically at 340 nm. Addition of: (1) NADPH, FNR, xanthine and xanthine oxidase; (2) 2 l M MM0632; (3) 6 lM rubredoxin; (4) 60 U of SOD. Superoxide reductase from Methanosarcina mazei C. Kra ¨ tzer et al. 446 FEBS Journal 278 (2011) 442–451 ª 2010 The Authors Journal compilation ª 2010 FEBS transfer electrons directly to MM0632, indicating that MM0632 is a rubredoxin oxidase. The addition of SOD to the assay resulted in decreased activity, owing to the competitive consumption of superoxide (Fig. 5). Discussion SORs such as desulfoferrodoxin and neelaredoxin are produced by anaerobic or microaerophilic prokaryotes, and are widespread among the bacterial and archaeal domains. Desulfoferrodoxin from sulfate reducers (e.g. D. desulfuricans, Desulfoarculus baarsii and A. fuldigus) is a protein containing a small N-terminal desulfore- doxin-type domain with the mononuclear center [Fe(SCys) 4 ], and a larger C-terminal domain containing an [Fe(NHis) 4 (SCys)] center. The [Fe(N- His) 4 (SCys)] center is composed of a pentacoordinated Fe 2+ with four equatorial His residues and one axial Cys in a square pyramidal geometry [27,28]. The other axial position is either coordinated to a Glu (oxidized metal state), resulting in an octahedral geometry, or is vacant when the metal is reduced, and it is probably the superoxide-binding site [28]. The additional desulforedoxin-like iron center con- sists of iron tetrahedrally coordinated to four Cys resi- dues [24]. The [Fe(SCys) 4 ] center is obviously not involved in superoxide reduction [11]. In comparison, neelaredoxins are much smaller, containing a single iron site with (NHis) 4 (SCys) coordination, identical to what occurs in the C-terminal domain of desulfoferro- doxin [4,29–31]. We have shown that MM0632 from M. mazei func- tions as an SOR and interacts with reduced rubredoxin from C. acetobutylicum as an electron donor. Further- more, EPR spectroscopy and sequence comparison clearly revealed that the protein contains a mononuclear iron center, which is most probably coordinated by four His residues (His24, His51, His57 and His134), one Cys (Cys131) and, depending on the oxidation state, one Glu (Glu21) [28]. Hence, the center represents the typical neelaredoxin [Fe(N- His) 4 (SCys)] center. Interestingly, the EPR data showed the presence of a [4Fe–4S] cluster, indicated by the axial signal with g^ = 1.93 and g k = 2.047. To our knowledge, MM0632 is the first SOR containing an [Fe(NHis) 4 (SCys)] center and an iron–sulfur cluster described to date, and thus represents a new family of SORs. By analogy with desulfoferrodoxin from sul- fate-reducing bacteria, we propose to refer to MM0632 as methanoferrrodoxin, which is found in several methanogenic archaeons (see below). Sequence alignments indicated that methanoferro- doxin is homologous with desulfoferrodoxins and neelaredoxin, as well as with uncharacterized SORs from hyperthermophilic archaeons and bacteria such as Ther- motoga maritima, P. furiosus and A. fulgidus [5,28]. In addition, homologs were found in close relatives of M. mazei, such as Methanosarcina acetivorans and Met- hanococcus voltae. The residues that bind the [Fe(N- His) 4 (SCys)] center are highly conserved among all sequences analyzed. In contrast to all other SOR sequences mentioned previously, we observed an inser- tion in the amino acid sequence (position 104–126) of MM0632 containing three Cys residues and an extended highly charged C-terminus with an additional Cys (Cys157). It is tempting to speculate that the additional Cys residues in methanoferrodoxin are responsible for the coordination of the [4Fe–4S] cluster. The same inser- tion and extended C-terminal domain was found in the predicted SORs from the methylotrophic methanogens Me. burtonii (YP_565539), Met. mahii (YP_003542283) and Meth. evestigatum (YP_003727000), indicating that these organisms are also able to form SORs that belong to the methanoferrodoxin protein family. Possible physiological function of methanoferrodoxin In the cellular environment, superoxide can be gener- ated from oxygen by electron leakage from b-type cytochromes, which are present in the electron trans- port chain of M. mazei, or by spontaneous oxidation of reduced flavoproteins [32,33]. The flavin protein AfpA from A. fulgidus was reported to generate super- oxide as a byproduct of interaction with oxygen [34]. AfpA is similar to MM0635 from M. mazei, whose coding region is located upstream of the mm0632 gene [8]. Also, M. mazei contains flavoproteins, which are essential for methanogenesis. Examples are the F 420 H 2 dehydrogenase, a key enzyme in membrane-bound electron transport [35], and the F 420 -reducing hydroge- nase, which is responsible for the H 2 -dependent reduc- tion of the central electron carrier coenyzme F 420 [36]. These enzymes may be able to transfer electrons to oxygen, forming superoxide, when oxygen enters the natural environment of this organism. Under these conditions, MM0632 could protect the cell from super- oxide damage. Rubredoxin is the only known electron donor for SOR. It mediates electron transfer between an oxidoreductase and the catalytic [Fe(NHis) 4 (SCys)] center of the SOR. In this article, evidence is presented that MM0632 also has rubredoxin oxidase activity, which is probably also dependent on the [Fe(N- His) 4 (SCys)] iron center. However, rubredoxin cannot be the native electron donor in M. mazei, because the C. Kra ¨ tzer et al. Superoxide reductase from Methanosarcina mazei FEBS Journal 278 (2011) 442–451 ª 2010 The Authors Journal compilation ª 2010 FEBS 447 genome of M. mazei does not code for any rubredoxin [8]. Thus, it is likely that an alternative electron donor system is present in M. mazei. It is tempting to specu- late that the [4Fe–4S] cluster of methanoferrodoxin may be necessary for acceptance of electrons from the undiscovered native electron donor in M. mazei. Potential candidates that could transfer electrons to methanoferrodoxin are ferredoxins and reduced coenzyme F 420 [37]. In addition to the gene mm0632, which encodes methanoferrodoxin, the M. mazei genome contains two genes coding for a catalase (NP_634581, NP_633974) and one gene coding for a SOD (NP_634447)[8]. Two homologous proteins, a catalase (YP_304371) and a SOD (CAB82579), from the close relative Methano- sarcina barkeri are described in the literature [38,39]. Consequently, it is likely that they are also functional in M. mazei. Therefore, the question arises of why M. mazei possesses two systems to detoxify superoxide, whereas M. barkeri is equipped only with a SOD. For anaerobes in general, SODs have the disadvantage of oxygen generation, which is circumvented by the alternative reaction mechanism of SORs. Hence, methanogens containing methanoferrodoxin, such as M. mazei and Me. burtonii, may survive better under oxygen stress conditions. Experimental procedures Reagents and proteins Xanthine, xanthine oxidase (bovine milk), catalase and SOD (bovine liver), FNR (spinach leaves), cytochrome c and acetylated cytochrome c were purchased from Sigma- Aldrich (Munich, Germany). Cloning, expression and purification The mm0632 gene was amplified by PCR, with chromosomal DNA of M. mazei as template and the following primers: mm0632for, 5¢-ATGGTAGGTCTCAAATGATAGGAA ATGAAGAAAAAATAAATAAGC-3¢; and mm0632rev, 5¢-ATGGTAGGTCTCAGCGCTGGCTTTCCAGACGCA TTTTTTGC-3¢. The gene mm0632 was cloned via BsaI restriction sites in plasmid pASK-IBA3 (IBA GmbH, Go ¨ ttingen, Germany), resulting in a C-terminal strep-tag fusion. The coding region of C. acetobutylicum rubredoxin (NP_349382) was cloned in a pT vector, using the BamHI and the XmaI restriction sites [12]. Overproduction of proteins was per- formed in E. coli DH5a. Cells were grown on modified maximal induction medium [MI; 3.2% (w ⁄ v) tryptone and 2% (w ⁄ v) yeast extract, with additions of M9 salts, 1 mm CaCl 2 and 1 mm MgSO 4 ]. Ampicillin (100 lgmL )1 ) was added for plasmid maintenance. For overproduction of rubredoxin, 40 lgmL )1 FeNH 4 citrate was added to the cultures, which were grown aerobi- cally at 30 °C for 16 h [4]. Cells were harvested by centrifu- gation (8000 g, 10 min) and lysed by sonication. Protein purification was performed aerobically according to the manufacturer’s instructions (IBA GmbH, Go ¨ ttingen, Ger- many). For the production of MM0632, the growth med- ium was supplemented with l-cysteine (1 mm), FeNH 4 citrate (0.1 mg mL )1 ) and FeSO 4 .7H 2 O (0.1 mg mL )1 ) [13]. After induction of protein production, growth proceeded under anaerobic conditions for 16 h at 28 °C. Cells were harvested at 8000 g for 10 min under anaerobic conditions, and all subsequent purification steps were performed in an anaerobic chamber (Coy Laboratory Products, Grass Lake, Michigan, USA) containing an atmosphere of 97% N 2 and 3% H 2 . Cells were lysed with B-Per (Pierce, Rockford, IL, USA), and detergent and cell debris were removed by cen- trifugation at 12 000 g for 20 min. MM0632 was reconsti- tuted by addition of 1 mm FeCl 3 ,1mm Na 2 S and 10 mm dithiothreitol to the cleared lysate, and incubated for 30 min. Insoluble components were removed by centrifugation at 12 000 g for 20 min. Protein purification was performed anaerobically according to the manufac- turer’s instructions (http://www.iba-go.com), except that washing buffer W (50 mm Tris, 150 mm NaCl, pH 8) and elution buffer E (50 mm Tris, 150 mm NaCl, 2.5 mm des- thiobiotin, pH 8) were purged with N 2 . Aliquots of the purified protein (50 lL) were diluted with 250 lL of buf- fer W, and concentrated on Vivaspin ultrafiltration spin columns (cut-off 3 kDa; Sartorius Stedim, Goettingen, Germany) in an anaerobic chamber. This procedure was repeated twice, and the final protein concentration was adjusted to 1.5–2 mg mL )1 . The protein was stored at ) 70 °C under an atmosphere of N 2 . Nonreconstituted pro- tein was purified as described above, with the exception that FeCl 3 ,Na 2 S and 10 mm dithiothreitol were not added to the cleared lysate. Molecular sieve chromatography A Superdex 75 chromatography column (Amersham Biosci- enes, Piscataway, NJ, USA), with reference proteins cyto- chrome c (12.4 kDa), myoglobin (17.8 kDa), chymotrypsin (25 kDa) and albumin (43 kDa), was used to determine the native mass of MM0632. Quantification of iron and acid-labile sulfide Purified protein was desalted on High-trap desalting col- umns (GE Healthcare, Munich, Germany) or by dialysis against buffer W containing 2.5 mm dithioerythritol and 5mm EDTA under anaerobic conditions Nonheme iron was quantified as described by Landers & Sak [14]. Superoxide reductase from Methanosarcina mazei C. Kra ¨ tzer et al. 448 FEBS Journal 278 (2011) 442–451 ª 2010 The Authors Journal compilation ª 2010 FEBS Acid-labile sulfide was quantified photometrically at 670 nm by measuring the formation of methylene blue after the addition of N,N-dimethyl-p-phenylenediamine, with Na 2 S as standard [15]. UV–visible spectroscopy UV–visible spectra were recorded on a spectrophotometer (TIDAS; J&M Analytik AG, Germany) from 275 to 700 nm, with a 0.5-cm quartz cuvette. The spectrum of a preparation of MM0632 (1 mg mL )1 ) in buffer W was recorded after titration with hydrogen peroxide, and referred to as the oxidized state. A few grains of sodium dithionite were added, and the spectrum of the reduced protein was recorded. Protein concentrations were determined with the BCA assay (Merck, Darmstadt, Germany). EPR spectroscopy EPR measurements were conducted with a Bruker EMX 1 ⁄ 6 spectrometeroperatingatX-band. Thesampletemperaturewas controlled with an Oxfordinstrument ESR-9 helium flow cryo- stat.Themagneticfieldwascalibratedbyuseofastrongoraweak pitch standard. The sample (300 lL; 10 mg protein mL )1 ) was either measured as isolated or after reduction by a few grains of sodiumdithionite. Detection of SOR activity of MM0632 in a cytochrome c assay The SOR activity of MM0632 was measured with a stan- dard cytochrome c reduction assay [4,16,17]. With this assay, SOD and SOR activities were detected. SOD and SOR compete with horse heart cytochrome c for the superoxide anion, which is generated continuously by xan- thine and xanthine oxidase under aerobic conditions. SOD inhibits superoxide-dependent reduction of cytochrome c, whereas SOR functions as a cytochrome c oxidase and reduces superoxide with electrons derived from cyto- chrome c. The assay was performed in 1.5-mL cuvettes under aerobic conditions in 1 mL of buffer W. Catalase (250 U mL )1 ) was added to prevent inhibition by perox- ides. Addition of xanthine (0.2 mm) and xanthine oxidase led to the generation of superoxide anions and reduction of cytochrome c (40 lm), resulting in an increase in absor- bance at 550 nm. The amount of xanthine oxidase was adjusted to an increase in cytochrome c reduction of 0.025 ± 0.001 min )1 at 550 nm [17]. Addition of MM0632 led to oxidation of cytochrome c, and a decrease in absor- bance at 550 nm. Enzyme amounts between 0 and 100 ng were used for the test, and the activities were linearly pro- portional within this range of protein content. Higher amounts of protein led to a decrease in specific activity, indicating suboptimal concentrations of reduced cyto- chrome c and ⁄ or superoxide. SOD, which competes for superoxide consumption, was used as a control in this assay. All enzymatic assays were performed on a Jasco V-550 spectrometer. SOD activity assay SOD activity was quantified with the standard aerobic xan- thine ⁄ xanthine oxidase assay in the presence of Nitro Blue tetrazolium [18]. Superoxide generated by xanthine oxidase reduces Nitro Blue tetrazolium to blue formazan, which was detected at 560 nm. The assay (3 mL) was performed in 50 mm KH 2 PO 4 (pH 7.6). One unit of activity was defined as the amount of enzyme needed to inhibit 50% of the reduction of Nitro Blue tetrazolium. SOR–rubredoxin interaction assay The aim of this assay was to test the interaction of MM0632 with rubredoxin from C. acetobutylicum as elec- tron donor. The activity was followed spectrometrically at 340 nm as an FNR-dependent decrease in the absorbance of NADPH (e 340 = 6.22 mm )1 cm )1 ). NADPH is oxidized by FNR, and electrons are transferred to rubredoxin, which serves as an electron donor for MM0632. Generation of superoxide was performed in the same way as described for the cytochrome c assay. The initial reaction mixture included 100 lm NADPH, 500 U mL )1 catalase, 0.06 lm FNR and 0.2 mm xanthine in a buffer of 50 mm Mops and 0.1 mm EDTA at pH 7.5. Xanthine oxidase (5 lgmL )1 ), rubredoxin (6 lm) from C. acetobutylicum and MM0632 (2 lm) were added to this premix. SOD was added as a control to show that the activity depended on superoxide. Acknowledgements We thank E. Schwab for technical assistance and P. Schweiger for critical reading of the manuscript. We also thank O. Riebe for providing plasmid pTrd. This work was supported by the Deutsche Forschungsgeme- inschaft (grant De488 ⁄ 9-1). References 1 Fridovich I (1995) Superoxide radical and superoxide dismutases. Annu Rev Biochem 64, 97–112. 2 Moura I, Tavares P, Moura JJ, Ravi N, Huynh BH, Liu MY & LeGall J (1990) Purification and characterization of desulfoferrodoxin. A novel protein from Desulfovibrio desulfuricans (ATCC 27774) and from Desulfovibrio vulgaris (strain Hildenborough) that contains a distorted rubredoxin center and a mononuclear ferrous center. J Biol Chem 265, 21596–21602. C. Kra ¨ tzer et al. Superoxide reductase from Methanosarcina mazei FEBS Journal 278 (2011) 442–451 ª 2010 The Authors Journal compilation ª 2010 FEBS 449 3 Chen L, Sharma P, Le Gall J, Mariano AM, Teixeira M & Xavier AV (1994) A blue non-heme iron protein from Desulfovibrio gigas. Eur J Biochem 226, 613–618. 4 >Riebe O, Fischer RJ & Bahl H (2007) Desulfoferro- doxin of Clostridium acetobutylicum functions as a superoxide reductase. FEBS Lett 581, 5605–5610. 5 Abreu IA, Saraiva LM, Carita J, Huber H, Stetter KO, Cabelli D & Teixeira M (2000) Oxygen detoxification in the strict anaerobic archaeon Archaeoglobus fulgidus: superoxide scavenging by neelaredoxin. Mol Microbiol 38, 322–334. 6 GrundenAM,JenneyFEJr,MaK,JiM,WeinbergMV& AdamsMW(2005)In vitroreconstitutionofanNADPH- dependentsuperoxidereductionpathwayfromPyrococcus furiosus.ApplEnvironMicrobiol71,1522–1530. 7 Pinto AF, Rodrigues JV & Teixeira M (2010) Reduc- tive elimination of superoxide: structure and mecha- nism of superoxide reductases. Biochim Biophys Acta 1804, 285–297. 8 Deppenmeier U, Johann A, Hartsch T, Merkl R, Sch- mitz RA, Martinez-Arias R, Henne A, Wiezer A, Bau- mer S, Jacobi C et al. (2002) The genome of Methanosarcina mazei: evidence for lateral gene transfer between bacteria and archaea. J Mol Microbiol Biotechnol 4, 453–461. 9 Tholen A, Pester M & Brune A (2007) Simultaneous methanogenesis and oxygen reduction by Methanobrevibacter cuticularis at low oxygen fluxes. FEMS Microbiol Ecol 62, 303–312. 10 McCord JM & Fridovich I (1970) The utility of superoxide dismutase in studying free radical reactions. II. The mechanism of the mediation of cytochrome c reduction by a variety of electron carriers. J Biol Chem 245, 1374–1377. 11 Emerson JP, Cabelli DE & Kurtz DM Jr (2003) An engineered two-iron superoxide reductase lacking the [Fe(SCys)(4)] site retains its catalytic properties in vitro and in vivo. Proc Natl Acad Sci USA 100, 3802–3807. 12 Girbal L, Mortier-Barriere I, Raynaud F, Rouanet C, Croux C & Soucaille P (2003) Development of a sensitive gene expression reporter system and an inducible promoter-repressor system for Clostridium acetobutylicum. Appl Environ Microbiol 69, 4985– 4988. 13 Jaganaman S, Pinto A, Tarasev M & Ballou DP (2007) High levels of expression of the iron–sulfur proteins phthalate dioxygenase and phthalate dioxygenase reductase in Escherichia coli. Protein Expr Purif 52, 273–279. 14 Landers JW & Zak B (1958) Determination of serum copper and iron in single small sample. Tech Bull Regist Med Technol 28, 98–100. 15 Beinert H (1983) Semi-micro methods for analysis of labile sulfide and of labile sulfide plus sulfane sulfur in unusually stable iron–sulfur proteins. Anal Biochem 131, 373–378. 16 McCord JM & Fridovich I (1968) The reduction of cytochrome c by milk xanthine oxidase. J Biol Chem 243, 5753–5760. 17 Lombard M, Fontecave M, Touati D & Niviere V (2000) Reaction of the desulfoferrodoxin from Desulfoarculus baarsii with superoxide anion. Evidence for a superoxide reductase activity. J Biol Chem 275, 115–121. 18 Beauchamp C & Fridovich I (1971) Superoxide dismu- tase: improved assays and an assay applicable to acryl- amide gels. Anal Biochem 44, 276–287. 19 Abreu IA, Xavier AV, LeGall J, Cabelli DE & Teixeira M (2002) Superoxide scavenging by neelaredoxin: dis- mutation and reduction activities in anaerobes. J Biol Inorg Chem 7 , 668–674. 20 Auchere F & Rusnak F (2002) What is the ultimate fate of superoxide anion in vivo? J Biol Inorg Chem 7, 664– 667. 21 Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W & Lipman DJ (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25, 3389–3402. 22 Clay MD, Jenney FE Jr, Hagedoorn PL, George GN, Adams MW & Johnson MK (2002) Spectroscopic stud- ies of Pyrococcus furiosus superoxide reductase: implica- tions for active-site structures and the catalytic mechanism. J Am Chem Soc 124, 788–805. 23 Clay MD, Emerson JP, Coulter ED, Kurtz DM Jr & Johnson MK (2003) Spectroscopic characterization of the [Fe(His)(4)(Cys)] site in 2Fe-superoxide reductase from Desulfovibrio vulgaris. J Biol Inorg Chem 8, 671– 682. 24 Tavares P, Ravi N, Moura JJ, LeGall J, Huang YH, Crouse BR, Johnson MK, Huynh BH & Moura I (1994) Spectroscopic properties of desulfoferrodoxin from Desulfovibrio desulfuricans (ATCC 27774). J Biol Chem 269, 10504–10510. 25 Jenney FE Jr, Verhagen MF, Cui X & Adams MW (1999) Anaerobic microbes: oxygen detoxification with- out superoxide dismutase. Science 286, 306–309. 26 Riebe O, Fischer RJ, Wampler DA, Kurtz DM Jr & Bahl H (2009) Pathway for H 2 O 2 and O 2 detoxification in Clostridium acetobutylicum. Microbiology 155, 16–24. 27 Adam V, Royant A, Niviere V, Molina-Heredia FP & Bourgeois D (2004) Structure of superoxide reductase bound to ferrocyanide and active site expansion upon X-ray-induced photo-reduction. Structure 12, 1729–1740. 28 Yeh AP, Hu Y, Jenney FE Jr, Adams MW & Rees DC (2000) Structures of the superoxide reductase from Py- rococcus furiosus in the oxidized and reduced states. Biochemistry 39, 2499–2508. 29 Berthomieu C, Dupeyrat F, Fontecave M, Vermeglio A & Niviere V (2002) Redox-dependent structural changes in the superoxide reductase from Desulfoarculus baarsii Superoxide reductase from Methanosarcina mazei C. Kra ¨ tzer et al. 450 FEBS Journal 278 (2011) 442–451 ª 2010 The Authors Journal compilation ª 2010 FEBS and Treponema pallidum: a FTIR study. Biochemistry 41, 10360–10368. 30 Jovanovic T, Ascenso C, Hazlett KR, Sikkink R, Krebs C, Litwiller R, Benson LM, Moura I, Moura JJ, Radolf JD et al. (2000) Neelaredoxin, an iron-binding protein from the syphilis spirochete, Treponema pallidum,isa superoxide reductase. J Biol Chem 275, 28439–28448. 31 Silva G, Oliveira S, Gomes CM, Pacheco I, Liu MY, Xavier AV, Teixeira M, Legall J & Rodrigues-pousada C (1999) Desulfovibrio gigas neelaredoxin. A novel superoxide dismutase integrated in a putative oxygen sensory operon of an anaerobe. Eur J Biochem 259, 235–243. 32 Kamlage B & Blaut M (1992) Characterization of cyto- chromes from Methanosarcina strain Gol and their involvement in electron transport during growth on methanol. J Bacteriol 174, 3921–3927. 33 Imlay JA (2002) How oxygen damages microbes: oxy- gen tolerance and obligate anaerobiosis. Adv Microb Physiol 46, 111–153. 34 Zhao T, Cruz F & Ferry JG (2001) Iron–sulfur flavo- protein (Isf) from Methanosarcina thermophila is the prototype of a widely distributed family. J Bacteriol 183, 6225–6233. 35 Baumer S, Ide T, Jacobi C, Johann A, Gottschalk G & Deppenmeier U (2000) The F 420 H 2 dehydrogenase from Methanosarcina mazei is a redox-driven proton pump closely related to NADH dehydrogenases. J Biol Chem 275, 17968–17973. 36 Kulkarni G, Kridelbaugh DM, Guss AM & Metcalf WW (2009) Hydrogen is a preferred intermediate in the energy-conserving electron transport chain of Methano- sarcina barkeri. Proc Natl Acad Sci USA 106, 15915– 15920. 37 Deppenmeier U (2002) The unique biochemistry of methanogenesis. Prog Nucleic Acid Res Mol Biol 71, 223–283. 38 Shima S, Netrusov A, Sordel M, Wicke M, Hartmann GC & Thauer RK (1999) Purification, characterization, and primary structure of a monofunctional catalase from Methanosarcina barkeri. Arch Microbiol 171, 317–323. 39 Brioukhanov A, Netrusov A, Sordel M, Thauer RK & Shima S (2000) Protection of Methanosarcina barkeri against oxidative stress: identification and characteriza- tion of an iron superoxide dismutase. Arch Microbiol 174, 213–216. 40 Larkin MA, Blackshields G, Brown NP, Chenna R, McGettigan PA, McWilliam H, Valentin F, Wallace IM, Wilm A, Lopez R et al. (2007) Clustal W and Clus- tal X version 2.0. Bioinformatics 23, 2947–2948. Supporting information The following supplementary material is available: Fig. S1. SDS ⁄ PAGE analysis of the protein peaks from gel filtration. This supplementary material can be found in the online version of this article. Please note: As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. C. Kra ¨ tzer et al. Superoxide reductase from Methanosarcina mazei FEBS Journal 278 (2011) 442–451 ª 2010 The Authors Journal compilation ª 2010 FEBS 451 . [7]. Methanosarcina mazei is one of the methanogenic archaeons, which are characterized by the ability to generate methane as the major end product of energy metabolism [8]. Many Methanosarcina strains. 5¢-ATGGTAGGTCTCAAATGATAGGAA ATGAAGAAAAAATAAATAAGC-3¢; and mm0632rev, 5¢-ATGGTAGGTCTCAGCGCTGGCTTTCCAGACGCA TTTTTTGC-3¢. The gene mm0632 was cloned via BsaI restriction sites in plasmid pASK-IBA3 (IBA GmbH, Go ¨ ttingen,. Methanoferrodoxin represents a new class of superoxide reductase containing an iron–sulfur cluster Christian Kra ¨ tzer 1 , Cornelia Welte 1 , Katerina Do ¨ rner 2 , Thorsten Friedrich 2 and