Electron transport during aceticlastic methanogenesis by Methanosarcina acetivorans involves a sodium-translocating Rnf complex

Authors


Correspondence

V. Müller, Molecular Microbiology and Bioenergetics, Institute of Molecular Biosciences, Johann Wolfgang Goethe University Frankfurt/Main, Max-von-Laue-Str. 9, 60438 Frankfurt, Germany

Fax: +49 69 7982 9306

Tel: +49 69 7982 9507

E-mail: vmueller@bio.uni-frankfurt.de

Abstract

The anaerobic methanogenic archaeon Methanosarcina acetivorans lives under extreme energy limitation. Methanogenesis from acetate as carried out by M. acetivorans involves an anaerobic electron transport chain with ferredoxin as electron donor and heterodisulfide as electron acceptor, and so far only the heterodisulfide reductase has been shown to translocate H+. Here, we describe a second Na+-translocating coupling site in this electron transport chain. Inside-out membrane vesicles of M. acetivorans catalyzed Na+ transport coupled to an electron transport catalyzed by the ferredoxin:heterodisulfide oxidoreductase activity. Ionophore studies revealed that Na+ transport was primary and electrogenic. A ∆rnf mutant was unable to grow on acetate and the ferredoxin:heterodisulfide oxidoreductase-coupled Na+ transport was abolished. These data are consistent with the hypothesis that the Rnf complex of M. acetivorans is an Na+-translocating coupling site and the entry point of electrons derived from reduced ferredoxin into the electron transport chain leading to the heterodisulfide.

Abbreviations
CODH/ACS

carbon monoxide dehydrogenase/acetyl-CoA synthase

CoB-SH

coenzyme B

CoM-SH

coenzyme M

CoM-S-S-CoB

heterodisulfide

Fho

ferredoxin:heterodisulfide oxidoreductase

Fno

ferredoxin:NAD+ oxidoreductase

IMV

inverted membrane vesicle

KPi

potassium phosphate

methyl-H4SPT

methyl-tetrahydrosarcinapterin

Mtr

methyl-H4SPT:CoM methyltransferase

pac

puromycine resistance

Introduction

Methanogenic archaea live at the thermodynamic edge of life: methanogenesis from H2 + CO2 has a free energy change (ΔG0′) of only −131 kJ·mol−1 which allows for ~ 2 mol of an ATP under standard conditions. However, at the low hydrogen partial pressures observed in nature, the ΔG value is even lower and allows for the synthesis of only a fraction of an ATP [1, 2]. Some methanogens can also grow on acetate according to

display math(1)
display math

Aceticlastic methanogenesis according to Eqn (1) has the lowest free energy change of any methanogenic reactions. Acetate metabolism involves activation of acetate to acetyl-CoA and a subsequent cleavage of the C–C bond by the CO dehydrogenase/acetyl-CoA synthase (CODH/ACS) yielding methyl-tetrahydrosarcinapterin (methyl-H4SPT), carbon monoxide and CoA. The methyl group of methyl-H4SPT is transferred to coenzyme M (CoM-SH) yielding methyl-CoM. In the next step, another coenzyme, coenzyme B (CoB-SH), attacks the methyl-CoM giving rise to a heterodisulfide of CoM and CoB (CoM-S-S-CoB) and also methane that is released into the environment. In the last step, the heterodisulfide is reduced to the respective thiols with electrons derived from the oxidation of CO to CO2 by the CODH/ACS [3, 4].

The overall enzymology involves only a few reactions and the question is how the carbon and/or electron flow is coupled to the generation of ATP. ATP is not synthesized by substrate level phosphorylation but ion gradient driven phosphorylation [1]. The methyl-H4SPT-CoM methyltransferase (Mtr) is a membrane-integral, six-subunit-containing enzyme complex that couples the methyl transfer reaction with the translocation of Na+ across the cytoplasmic membrane, thus establishing a sodium ion motive force across the membrane [5-7]. Reduction of the heterodisulfide involves an ion-translocating electron transport chain containing cytochromes as well as the membrane-permeable electron carrier methanophenazine [8-10]. The electron output module of this electron transport chain, the heterodisulfide reductase, receives electrons from reduced methanophenazine and this process is coupled to the translocation of protons across the membrane thus establishing a proton motive force [8, 11, 12]. Electrons for methanophenazine reduction ultimately stem from oxidation of CO to CO2. The redox potential difference of the donor (ΔE0′ CO/CO2 = −520 mV) and the acceptor methanophenazine (ΔE 0′ = −165 mV) allows for additional ion translocation. One possible scenario is a two-step process in which the electrons from CO are transferred via ferredoxin to the well established proton-translocating Ech hydrogenase that reduces protons to hydrogen [13]. In the second step, hydrogen is oxidized by the well known membrane-bound hydrogenase that funnels electrons into the membrane (methanophenazine) via a cytochrome b subunit [13-15]. However, although this is possible in Methanosarcina mazei, it is not in Methanosarcina acetivorans since it lacks hydrogenases. An alternative for M. acetivorans is the Rnf complex [16], a multi-subunit membrane-bound electron transport protein complex that is assumed to couple ferredoxin-dependent NAD+ reduction in bacteria with the translocation of ions across the cytoplasmic membrane [17]. Recently, we have shown a sodium motive ferredoxin:NAD+ oxidoreductase (Fno) in the acetogenic bacterium Acetobacterium woodii and inhibitor studies were consistent with the hypothesis that the Fno activity is catalyzed by the Rnf complex [18]. Rnf genes are also present in M. acetivorans and upregulated during growth on acetate [19]. We have addressed the question whether electron transfer from reduced ferredoxin to the heterodisulfide involves Rnf and is coupled to Na+ translocation. Using inverted membrane vesicles (IMVs) of M. acetivorans we were able to establish sodium ion transport coupled to ferredoxin-dependent heterodisulfide reduction and we provide genetic evidence that the Rnf complex of M. acetivorans is a sodium ion pump.

Results

A ferredoxin:heterodisulfide oxidoreductase activity in membrane preparations of M. acetivorans

IMVs of M. acetivorans were prepared from cells grown on acetate to exponential growth phase (up to two-thirds of maximal methane production). Cells were harvested by flow-through centrifugation under strictly anaerobic conditions, disrupted in a French press under reduced pressure, and membrane vesicles were collected by ultra centrifugation. To demonstrate that the IMVs were energetically coupled, an artificial proton gradient was imposed using an ammonium diffusion potential as described previously [20]. Crude vesicles were also able to synthesize ATP driven by the growth substrate of the cells used for vesicle preparation [20].

To test for ferredoxin:heterodisulfide oxidoreductase (Fho) activity, ferredoxin was reduced with CODH/ACS using CO as electron donor. Upon addition of reduced ferredoxin to membrane preparations incubated in the presence of 0.4 mm heterodisulfide, the heterodisulfide was reduced as was evident from the production of the corresponding thiols (Fig. 1). Heterodisulfide reduction was ferredoxin dependent and proceeded at a rate of 154 nmol heterodisulfide·min−1·mg protein−1.

Figure 1.

Ferredoxin-dependent heterodisulfide reduction in wild-type and ∆rnf membrane preparations. The assays contained 400 μm CoM-S-S-CoB, 20 μm ferredoxin, 70 μg CODH/ACS and 200 μg washed membrane preparation in a total volume of 250 μL: ■, wild-type membranes; □, membranes from Δrnf mutant. Wild-type and ∆rnf mutant were grown under the same conditions using methanol as carbon and energy source.

Fho activity is coupled to 22Na+ translocation

To measure sodium ion transport in the IMVs, they were incubated at 1 mm 22Na+. In the absence of a substrate, the internal 22Na+ equilibrated with the external Na+. IMVs incubated with CODH/ACS, ferredoxin and heterodisulfide under an N2 atmosphere did not accumulate Na+. However, upon reduction of the ferredoxin with CO, 22Na+ was rapidly accumulated inside the lumen of the IMVs at a rate of 0.3 nmol·min−1·mg protein−1. 22Na+ transport was dependent on ferredoxin or the heterodisulfide (Fig. 2A). CO + CODH/ACS could be substituted by H2 + ferredoxin-reducing hydrogenase from A. woodii [21].

Figure 2.

Ferredoxin:heterodisulfide oxidoreductase activity is coupled to primary Na+ transport. Washed membrane vesicles were diluted in 1 mL KPi buffer (40 mm), pH 7.0, containing 25 mm MgSO4, 0.4 m sucrose, 1 μg·mL−1 resazurin and 10 mm dithioerythritol. The assay was supplemented with 17 μm valinomycin, 100 mm KCl, 20 μm ferredoxin, 0.16 mm heterodisulfide and 19 μg CODH/ACS. The reaction was started by addition of CO (200 μL) (arrow). The experiment was repeated three times using independent IMV preparations. (A) Ferredoxin (♦) or heterodisulfide (▲) was omitted as control. The final protein concentration was 3.4 mg·mL−1. (B) One assay was pre-incubated with ETH157 (▲) and one with SF6847 (■). The final protein concentration was 0.7 mg·mL−1.

22Na+ transport coupled to Fho activity is primary and electrogenic

If transport of 22Na+ is electrogenic a membrane potential should be established that in turn should slow down 22Na+ transport. To test this, we used valinomycin in combination with KCl to dissipate the membrane potential, ΔΨ. KCl itself (100 mm) had no effect on Na+ transport. However, in combination with valinomycin 22Na+ transport was stimulated slightly (data not shown). Fho activity could be directly or indirectly (via a primary H+ gradient in combination with an Na+/H+ antiporter) coupled to Na+ transport. To discriminate between these possibilities, we used protonophores or sodium ionophores and analyzed their effect on 22Na+ transport. The protonophore SF6847 (10 μm) did not inhibit 22Na+ accumulation (Fig. 2B) but stimulated it to a value of 150%. This is due to the SF6847-mediated proton transfer out of the IMVs along the established electrical field that leads to a dissipation of ∆Ψ. The effectiveness of the uncoupler was verified by its ability to dissipate an artificial ΔpH created by an NH4+ diffusion potential. The sodium ionophore ETH157 (20 μm) completely inhibited 22Na+ accumulation (Fig. 2B). These data demonstrate that 22Na+ transport is directly linked to Fho activity and is electrogenic.

Generation of a ∆rnf mutant

The Fho activity may involve the Rnf complex as the entry site for electrons derived from reduced ferredoxin and as an energetic coupling site. To address the role of Rnf in Fho activity and Na+ transport, a ∆rnf mutant was generated. For this purpose, the rnf operon was exchanged against a puromycine resistance (pac) cassette by double homologous recombination. The correctness of the construct was verified by southern blotting. For this purpose, chromosomal DNA of the mutant and the wild-type was restricted and blotted onto a nylon membrane. A digoxigenin-labeled fragment complementary to the downstream region of the MA0665 (bases 781747–782872) or the pac cassette was used and blotted against the membrane-bound DNA. As can be seen in Fig. 3, the pac cassette did not hybridize against the wild-type but the mutant DNA. The apparent fragment size (4600 bp) corresponds nicely to the predicted size (4560 bp). When the downstream region was used as probe, it hybridized to only one fragment that was of the expected size in wild-type (3900 bp) and mutant (4550 bp). These data demonstrate that the pac cassette had integrated into the correct locus.

Figure 3.

Genotype of M. acetivoransrnf::pac. Genomic DNA of M. acetivorans C2A (line labeled 2) and M. acetivoransrnf::pac (line labeled 1) was digested with ClaI (A) or EcoRV (B). The fragments were separated via gel electrophoresis, blotted onto a nylon membrane and probed against a digoxigenin-labeled fragment of the pac cassette (A) or a 1000 bp downstream fragment of the rnf genes (B).

Phenotype of a ∆rnf mutant

We first analyzed the growth of the mutant with different carbon and energy sources. The mutant grew as well as the wild-type on trimethylamine, methanol or methanol + acetate or CO, but it did not grow on acetate alone. The doubling time of the ∆rnf mutant grown on trimethylamine (13.6 h), for example, increased only about 1.3-fold in comparison with the doubling time of the wild-type (9.9 h). Even on methanol + acetate the doubling time of mutant (19.8 h) and wild-type (13 h) did not differ significantly. The final optical density of about 1 was in both cases the same for wild-type and ∆rnf. The growth experiments indicate a role of the Rnf complex in the aceticlastic pathway.

Deletion of the rnf cluster results in a loss of Fho-coupled Na+ transport

To elucidate the role of the Rnf complex, the Fho activities of wild-type and ∆rnf mutant were analyzed. Electron transfer from reduced ferredoxin to the heterodisulfide was still possible, albeit with a reduced rate of 45% of that of wild-type membrane preparations (Fig. 1), explaining the growth phenotype described above. In contrast, the activity of respiratory enzymes not involved in ferredoxin oxidation was similar in the wild-type and the ∆rnf mutant: the F420H2 dehydrogenase exhibited activities of 37.7 ± 1.2 and 40.8 ± 1.7 mU·mg membrane protein−1, respectively. The heterodisulfide reductase that is part of the Fho system was also similar with 606 ± 128 mU·mg membrane protein−1 in the wild-type and 607 ± 94 mU·mg membrane protein−1 in the ∆rnf mutant.

For previous 22Na+ transport experiments, IMVs were prepared from acetate-grown cells since the production of Rnf was about six-fold lower in methanol- versus acetate-grown cells and thus Na+ transport was not only much lower but also difficult to reproduce. Since the ∆rnf mutant did not grow on acetate, we grew the mutant and the parental strain on methanol (150 mm) plus acetate (40 mm) to induce stable Rnf-mediated Na+ transport. Compared with methanol as substrate, the amount of rnfB transcript was two-fold higher on methanol + acetate and six-fold higher on acetate, compared with methanol alone. IMVs prepared from methanol + acetate-grown cells also catalyzed a primary Na+ transport (Fig. 4A) comparable with acetate-grown cells. Most important, the ∆rnf mutant completely lost the ability to translocate Na+. To ensure that the IMVs of the mutant are as impermeable to Na+ as the wild-type IMVs, we analyzed ATP-driven Na+ transport in IMVs from the wild-type and the mutant [20]. As can be seen in Fig. 4B, both preparations showed comparable Na+ transport rates and accumulation factors excluding altered Na+ permeability in the ∆rnf mutant. These data provide compelling evidence that the Rnf complex is the Na+ pump operative during ferredoxin-dependent heterodisulfide reduction. The observation that the Δrnf mutant still had considerable heterodisulfide reductase activity but lost Na+ transport is also consistent with the hypothesis that the heterodisulfide reductase in M. acetivorans is using H+ and not Na+ as coupling ion.

Figure 4.

A ∆rnf mutant is impaired in Fho-coupled Na+ transport. Washed membrane vesicles of wild-type (2.1 mg·mL−1) (■) and ∆rnf mutant (3.4 mg·mL−1) (▲) were diluted in 1 mL KPi buffer (40 mm), pH 7.0, containing 25 mm MgSO4, 0.4 m sucrose, 1 μg·mL−1 resazurin and 10 mm dithioerythritol. (A) The assay was supplemented with 17 μm valinomycin, 100 mm KCl, 20 μm ferredoxin, 0.16 mm heterodisulfide and 19 μg CODH/ACS. The reaction was started by addition of CO (200 μL) (arrow). One assay containing wild-type IMVs was pre-incubated with ETH157 (♦), another with SF6847 (●). (B) The same IMV preparations (wild-type ♦; ∆rnf ▼) were incubated in 1 mL KPi buffer (40 mm), pH 7.0, containing 25 mm MgSO4, 0.4 m sucrose, 1 μg·mL−1 resazurin and 10 mm dithioerythritol. The assay was supplemented with 17 μm valinomycin, 10 mm KHSO3 and 40 mm KCl. The reaction was started by the addition of 5 mm K2-ATP. All experiments were repeated three times using independent IMV preparations.

Discussion

Methanogenesis from acetate has the lowest free energy change of any methanogenic reactions [22]. Considering a phosphorylation potential of around 45 kJ·mol−1 under cellular conditions [23], this would allow for the synthesis of only around one mole ATP for one mole of acetate consumed. This theoretical consideration may lead to the speculation that efficient mechanisms are employed to conserve as much energy in the form of ATP as possible. Two coupling sites have been known before [7, 9, 24-26]; here, a third one was added.

After acetate has been activated at the expense of ATP hydrolysis to acetyl-phosphate, acetyl-CoA is formed and split by the CODH/ACS to enzyme-bound CO and an enzyme-bound methyl group (Fig. 5). The methyl group is transferred to H4SPT and the methyl group of methyl-H4SPT is then transferred via an Na+-translocating Mtr complex to CoM. This reaction allows for the translocation of 2Na+/CH3-H4SPT [27]. Enzyme-bound CO is oxidized to CO2 and the electrons are transferred to ferredoxin. Reduced ferredoxin is then the electron donor for a membrane-bound electron transport chain that involves the membrane-integral electron carrier methanophenazine and the final electron acceptor heterodisulfide. Electron transfer from methanophenazine to the heterodisulfide reductase involves a methanophenazine loop that drives one H+ per electron out of the cell [8]. The ways electrons are transferred from ferredoxin into the electron transport are obviously different in different methanogens.

Figure 5.

Proposed model of the aceticlastic pathway of methanogenesis in M. acetivorans. H4SPT, tetrahydrosarcinapterin; Fdred, reduced ferredoxin; Fdox, oxidized ferredoxin; Mtr, methyl-H4SPT:CoM methyltransferase; MPhe, methanophenazine; Hdr, heterodisulfide reductase; ATP, A1A0 ATP synthase. The cytochrome c is a subunit of the Rnf complex (Rnf X).

In M. mazei and Methanosarcina barkeri a membrane-bound multi-subunit (NiFe) hydrogenase, Ech, was described that oxidizes reduced ferredoxin and produces molecular hydrogen [28]. In the above-mentioned organisms this enzyme is part of the ferredoxin:heterodisulfide oxidoreductase system and participates in energy conservation by the translocation of one H+ in the course of hydrogen formation [15]. Molecular hydrogen is then oxidized by a second hydrogenase (Vho) at the outside of the cytoplasmic membrane. Two protons are released at the extracellular side when the H2 molecule is oxidized, and methanophenazine takes up two protons upon reduction, so a vectorial proton translocation with the stoichiometry of 2H+/2e can be observed [11]. In contrast to M. mazei and M. barkeri, M. acetivorans does not contain hydrogenases. Hence, electron transport from ferredoxin must be different.

Here, we have demonstrated that the Rnf complex is part of this electron transport chain in M. acetivorans; it catalyzes Na+ transport, most probably coupled to electron transport from reduced ferredoxin to methanophenazine. The coupling between Na+ and electron transport was apparently weak. However, increasing CODH concentrations and substrate concentrations increased the Na+/e ratio to 0.02 Na+/e. The weak coupling may also result from different electric pathways leading to heterodisulfide (Fig. 1). Anyway, the fact that the rnf mutant was no longer able to couple Fho activity to Na+ transport is consistent with the hypothesis that the Rnf complex is the Na+-translocating coupling site in this electron transport chain. Rnf genes were first discovered in a mutant of Rhodobacter capsulatus that was impaired in nitrogen fixation (Rhodobacter nitrogen fixation) [29]. These genes are widespread in bacteria where they encode a membrane-integral complex containing six subunits with FeS centres and flavins as electron carriers [17]. Recently, a sodium-motive Fno was demonstrated in the anaerobic bacterium A. woodii [18]. The Fno was enriched and identified as Rnf complex. A similar function is suggested for Clostridia such as Clostridium tetani [30, 31], Clostridium kluyveri [32] or Acidaminococcus fermentans [33]. Biochemical proof that the Rnf complex is indeed a Na+-motive enzyme has to await purification and biochemical characterization of the complex. However, here we present first genetic evidence that the Rnf complex is indeed a sodium-translocating enzyme: the ∆rnf mutant was impaired in Na+ transport.

Despite its wide distribution in bacteria the rnf genes are found in only a few archaea, notably methanogens such as M. acetivorans and Methanococcoides burtonii [17]. Their function is to link cellular ferredoxin pools to the heterodisulfide reduction, a link that is established in other methanogens by the combined action of the Ech hydrogenase and the membrane-bound hydrogenase. This would argue for a lateral transfer of the rnf genes from bacteria to the methanogens. However, the rnf gene cluster of methanogens (and many sulfate-reducing bacteria) differs from all others by the presence of an additional gene encoding a cytochrome c next to the rnf genes [34, 35]. Cytochrome c is reduced by reduced ferredoxin and reoxidized by the addition of heterodisulfide [36] and thus it is clearly part of the electron transport chain from reduced ferredoxin to heterodisulfide.

The overall free energy change of ferredoxin-coupled heterodisulfide reduction is 68 kJ·mol−1 that would allow for the translocation of ~ 4 Na+ per two electrons. All together this would gain eight ions translocated per mole of acetate consumed. This would be sufficient to synthesize two moles of ATP by the promiscuous H+- and Na+-translocating ATP synthase of M. acetivorans [20]. Taking into account the one mole ATP invested for acetate activation, the overall yield is one ATP per acetate by a chemiosmotic mechanism. Thus, the novel coupling site for the electron transport chain described here not only provides the missing link in the electron transport chain of M. acetivorans but is the major contributor to the overall energetics of methanogenesis from acetate in M. acetivorans. The elucidation of the structure and function of the Rnf complexes is still in its infancy [17] but it should be pointed out that the Rnf complex from M. acetivorans contains an unusual cytochrome c subunit [34-36]. Whether this subunit is an adaptor module to connect the enzyme to methanophenazine remains to be established in the future.

It is interesting to note that the Rnf complex as well as the Mtr complex use sodium ions as coupling ions, underlining the evolutionary primacy of sodium in bioenergetics. The use of sodium ions is advantageous for M. acetivorans. Acetate acts as a protonophore and, for an organism living at the thermodynamic edge in the presence of high acetate concentrations, the dissipation of the proton gradient by acetate would be detrimental. With the evolution of the methanophenazine-coupled, proton-translocating heterodisulfide reductase a mechanism to use both ion gradients for ATP synthesis had to be evolved. In M. acetivorans, both ion gradients are used simultaneously by the promiscuous ATP synthase [20].

Experimental procedures

Organisms and cultivation

M. acetivorans [37] WWM1 (C2A, Δhpt) [38] was grown in single-cell morphology [39] at 37 °C in high salt medium [40] containing either 50 mm trimethylamine, 150 mm methanol, 150 mm methanol + 40 mm acetate, 1.5 bar CO or 120 mm acetate as carbon and energy source. The Na2S × 9H2O concentration was raised to 0.15 g·L−1. Growth conditions were strictly anaerobic under an atmosphere of N2/CO2 (80 : 20, v/v) in 1250 mL glass bottles containing 500 mL medium or in glass bottles filled with 20 L medium (Ochs, Bovenden, Germany). The growth was determined by measuring the optical density at 578 nm (D578) using a Hitachi U-1800 spectrometer.

Generation of the deletion mutant M. acetivorans WWM1 ∆rnf::pac

The deletion mutant M. acetivorans WWM1 ∆rnf::pac was generated via double homologous recombination as described previously [41]. For the deletion, upstream and downstream regions of the rnf genes (MA0658–MA0665) were cloned into the multiple cloning site of pJK3, flanking a pac cassette. The cloned regions as well as the resistance cassette were cut at integrated restriction sites using AscI and transferred into M. acetivorans WWM1. The rnf genes were exchanged against the pac cassette. The mutants were selected for puromycine resistance and verified by southern blot analysis as described previously [42]. The primers used for deletion are given in Table 1.

Table 1. Primers used in this study
PrimerSequenceUse
MA0658upf ATGCACTAGTTCGCTTTCATTCGGAGCGTA Amplification of up region of MA0658
MA0658upr TGTCTGCAGGCGAGAAGGAACTGGCATGA Amplification of up region of MA0658
MA0665downf ATGCGATATCCCTGCAAGGAGAATTCCCACA Amplification of down region of MA0665
MA0665downr TTGTCTCGAGCACCGTAACCCTCGGAGGAA Amplification of down region of MA0665

Preparation of proteins and cofactors

Clostridium pasteurianum ferredoxin and Moorella thermoacetica CODH/ACS were prepared as described previously [43]. The cofactor F420 was purified from M. mazei cytoplasm as described in the same paper. F420 was reduced to F420H2 under N2 flow using NaBH4. Excess NaBH4 was eliminated by titrating with HCl to pH 0. CoM-S-S-CoB was synthesized according to Welte and Deppenmeier [43].

Preparation of washed membrane preparations

Membranes were prepared as described previously for M. mazei [43] with the exception that cells were disrupted by French pressure treatment (1000 psi outer cell pressure). The buffer used for this procedure contained 40 mm potassium phosphate (KPi), pH 7.0, 25 mm MgSO4, resazurin (1 μg·mL−1) and 5 mm dithiothreitol.

Preparation of washed inverted membrane vesicles

The preparation of IMVs was performed as described previously [20]. The procedure was done under strictly anaerobic conditions under an atmosphere of N2/H2 (95 : 5, v/v) using 40 mm KPi buffer, pH 7.0, containing 25 mm MgSO4, 0.4 m sucrose, resazurin (1 μg·mL−1) and 10 mm dithioerythritol. The vesicles were washed at least twice to remove remaining cytoplasm.

Measurement of electron transport

Electron transport from ferredoxin to the heterodisulfide was measured in rubber-stoppered glass vials containing 250 μL 40 mm KPi buffer, pH 7.0, 20 mm NaCl, 0.4 mm CoM-S-S-CoB, resazurin (1 μg·mL−1), 200 μg washed membrane preparation, 20 μm ferredoxin and 70 μg CODH/ACS under an atmosphere of 5% CO/95% N2 (v/v). The buffer was reduced with titanium(III) citrate (150 μm). Measurements were started by the addition of CODH/ACS. At different time points 10 μL samples were removed and subjected to a modified Ellman's test [43, 44]. F420H2 dehydrogenase was assayed in 600 μL 40 mm KPi, pH 7.0, resazurin (1 μg·mL−1), 5 mm dithioerythritol containing 16 μm F420H2, 40 μg membrane preparation, 0.5 mm metronidazole and 0.3 mm methyl viologen as mediator. The change of absorption at 420 nm was used to calculate enzyme activities (ε = 40 mm−1·cm−1). The heterodisulfide was assayed in the same buffer containing 180 μm CoM-S-S-CoB, 1.5 mm benzyl viologen reduced with Na dithionite to A575 ~ 2. Enzyme activities were calculated from the change of absorbance at 575 nm (ε = 8.9 mm−1·cm−1).

Measurement of Na+ translocation

The experiments were performed under anaerobic conditions in 40 mm KPi buffer, pH 7.0, containing 25 mm MgSO4, 0.4 m sucrose, resazurin (1 μg·mL−1) and 10 mm dithioerythritol at 37 °C in a water bath in a 3.5 mL glass vial; 1 mL of buffer was supplemented with 17 μm valinomycin, 100 mm KCl, 20 μm ferredoxin and 0.16 mm heterodisulfide and 20 μm ETH157 or 10 μm SF6847 was added as indicated. 22NaCl (final activity 0.5 μCi·mL−1, carrier-free) was added and incubated for 40 min to ensure Na+ equilibrium before the reaction was started, and 19 μg of Mthermoacetica CODH/ACS was added 10 min before the start of the sampling. During the experiment, 80 μL of sample was taken with a 100 μL syringe at the time points indicated. The reaction was started by addition of CO. The external 22Na+ was removed by passing the sample over a small column (0.5 × 3 cm) of DOWEX 50WX8 (Serva, Mannheim, Germany) [18]. The vesicles were eluted by 1 mL of 420 mm sucrose; 4 mL of Rotiscint eco plus (Roth, Karlsruhe, Germany) was added and the radioactivity was measured with a liquid scintillation counter (wallac 1409; Wallac, PerkinElmer Inc, Waltham, MA, USA) [18, 45].

The Na+ concentration of the assay was determined with an Na+-selective electrode (model 720A; Orion, Thermo Fisher Scientific, Waltham, MA, USA). SF6847, ETH157 and valinomycin were added as ethanolic solutions. Controls received the solvent only.

Acknowledgements

This work was supported by grants from the Deutsche Forschungsgemeinschaft to UD and VM and through the LOEWE funding program of Hesse's Ministry of Higher Education, Research and Arts.

Ancillary