U. Deppenmeier, Institut für Mikrobiologie und Biotechnologie, University of Bonn, Meckenheimer Allee 168, 53115 Bonn, Germany Fax: +49 228 737576 Tel: +49 228 735590 E-mail: firstname.lastname@example.org
Methanosarcina mazei belongs to the group of aceticlastic methanogens and converts acetate into the potent greenhouse gases CO2 and CH4. The aceticlastic respiratory chain involved in methane formation comprises the three transmembrane proteins Ech hydrogenase, F420 nonreducing hydrogenase and heterodisulfide reductase. It has been shown that the latter two contribute to the proton motive force. The data presented here clearly demonstrate that Ech hydrogenase is also involved in energy conservation. ATP synthesis was observed in a cytoplasm-free vesicular system of Ms. mazei that was dependent on the oxidation of reduced ferredoxin and the formation of molecular hydrogen (as catalysed by Ech hydrogenase). Such an ATP formation was not observed in a Δech mutant strain. The protonophore 3,5-di-tert-butyl-4-hydroxybenzylidene-malononitrile (SF6847) led to complete inhibition of ATP formation in the Ms. mazei wild-type without inhibiting hydrogen production by Ech hydrogenase, whereas the sodium ion ionophore ETH157 did not affect ATP formation in this system. Thus, we conclude that Ech hydrogenase acts as primary proton pump in a ferredoxin-dependent electron transport system.
Biological methanogenesis from acetate is one of the most important processes for the maintenance of the carbon cycle on Earth. The products of methanogenesis from acetate, CH4 and CO2 are released from anaerobic habitats and large amounts of these greenhouse gases reach the atmosphere. Therefore, the process of biological methane formation is of great interest for global ecology [1,2]. Moreover, the process of methanogenesis creates a combustible gas that can be used as an energy source. Only the genera Methanosarcina and Methanosaeta are able to use the aceticlastic pathway of methanogenesis, and Methanosarcinamazei strain Gö1 (hereafter referred to as Ms. mazei) is one of the important model organisms . In Ms.mazei, acetate is activated by phosphorylation and exchange of inorganic phosphate with CoA. The resulting acetyl-CoA is cleaved by the CO dehydrogenase/acetyl-CoA synthase (CODH/ACS). In the course of the reaction, enzyme-bound CO is oxidized to CO2 and the electrons are used for ferredoxin (Fd) reduction. The methyl group of acetate is transferred to tetrahydrosarcinapterin. The resulting methyl-tetrahydrosarcinapterin is converted to methane by the catalytic activities of a Na+-translocating methyl-CoM methyltransferase [forming methyl-2-mercaptoethanesulfonate (methyl-S-CoM)] and the methyl-S-CoM reductase, which uses N-7-mercaptoheptanoyl-l-threonine phosphate (HS-CoB) as the electron donor to reduce the methyl group to CH4. An additional product of this reaction is the heterodisulfide of 2-mercaptoethanesulfonate (HS-CoM) and HS-CoB (CoM-S-S-CoB), which serves as a terminal electron acceptor in the methanogenic respiratory chain (for a review see ).
The intermediates of the aceticlastic pathway, CoM-S-S-CoB and reduced ferredoxin (Fdred), are recycled by a membrane-bound electron transport system that can be defined as Fd:heterodisulfide oxidoreductase . In most Methanosarcina species (e.g. Ms. mazei and Ms. barkeri) the oxidation of Fdred is catalysed by Ech hydrogenase, resulting in the release of molecular hydrogen , which is then reoxidized by the F420 nonreducing hydrogenase and the electrons are channelled via methanophenazine to the heterodisulfide reductase . Some Methanosarcina species, e.g. Ms. acetivorans, lack Ech hydrogenase and must possess an alternative route for oxidation of Fdred. It has been shown that the F420 nonreducing hydrogenase and the heterodisulfide reductase are key elements in membrane-bound electron transport and are essential to generate the proton motive force , whereas the methyl-CoM methyltransferase generates a Na+ ion gradient [8,9]. Furthermore, it was suggested that Ech hydrogenase also contributes to the electrochemical ion gradient  because of homologies to certain subunits of ion-translocating oxidoreductases  and indirect evidence from experiments with resting cells of Ms. barkeri [11,12]. However, direct experimental evidence for this hypothesis is lacking. In this study, we present the first biochemical proof that Ech hydrogenase is indeed an ion-translocating enzyme, and thus represents an additional energy-conserving coupling site in methanogenic metabolism. Inhibitor studies clearly indicate that H+ and not Na+ is the coupling ion. Thus, the proton gradient can directly be used for ATP synthesis via A1AO ATP synthase .
To investigate Fd-mediated electron transport, we took advantage of washed inverted vesicle preparations of Ms. mazei, which contain all essential membrane-bound proteins involved in energy conservation and which are suitable for the generation of electrochemical ion gradients . These vesicles do not contain enzymatic activities that would produce Fdred. Therefore, Fd from Clostridium pasteurianum was used as the electron donor, which was reduced by the CODH/ACS from Moorella thermoacetica with CO as the initial substrate.
When the oxidation of Fdred in the absence of CoM-S-S-CoB was analysed in the washed vesicle preparation, the rate of H2 production was 32.8 nmol·min−1·mg protein−1 (Table 1) and was constant over a time period of 60 min. The reaction was coupled to the phosphorylation of ADP, as indicated by a rapid increase in ATP content upon the start of the reaction (Fig. 1). The rate of ATP production was 1.5 nmol ATP min−1·mg protein−1, which is comparable with ATP synthesis rates observed in the process of methanogenesis from methyl-S-CoM + H2 . In the absence of Fd or CO, H2 production was < 0.1 nmol·min−1·mg protein−1 (Table 1) and ATP synthesis was not observed (Fig. 1). However, ATP synthesis (Fig. 1) and H2 formation (not shown) were fully restored when Fd was subsequently added to the reaction mixture.
Table 1. Hydrogen formation by Fdred-dependent proton reduction. Test vials contained 5% CO/95% N2 in the headspace, 500 μg inverted membrane vesicles, 33.5 μg Fd, 20 μg CODH/ACS, 150 nmol AMP, 300 nmol ADP. The addition or exclusion of single components is indicated.
H2 production rate (%)
a Most active vesicle preparations showed a specific activity of 32.8 nmol min−1·mg protein−1.
+ 10 μm ETH157
+ 10 μm SF6847
+ 400 μm DCCD
To analyse this process in more detail, we used washed vesicle preparations of a Ms. mazeiΔech mutant and subjected these vesicles to the standard assay (described in Experimental Procedures under ‘Determination of ATP formation’ and ‘Determination of H2’). As expected, H2 formation from Fdred was not observed in this mutant (Table 1), whereas the activities of all other Fd-independent parts of the respiratory chain (F420 nonreducing hydrogenase, heterodisulfide reductase and F420H2 dehydrogenase) remained unaffected (not shown). As evident from Fig. 2, inverted membrane vesicles from the Δech mutant did not form ATP when incubated with Fdred in the absence of heterodisulfide. As a control, ATP formation associated with H2:heterodisulfide oxidoreductase activity was examined and the rate of ATP formation (1.9 nmol ATP min−1·mg protein−1) in vesicle preparations was the same for the mutant and the wild-type with H2 and CoM-S-S-CoB as substrates (Fig. 2). This process was independent of Ech hydrogenase because the F420 nonreducing hydrogenase oxidizes H2 and electrons are transferred via methanophenazine to heterodisulfide reductase. This process is coupled to proton translocation over the cytoplasmic membrane . In summary, these results clearly indicated that Ech hydrogenase is necessary to generate an electrochemical ion gradient when Fd is the only reducing equivalent and heterodisulfide is absent.
To rule out the possibility of substrate-level phosphorylation, independent of an ion gradient, N,N′-dicyclo-hexylcarbodiimide (DCCD) was added to the reaction. This compound specifically inhibits the ATP synthase from Ms. mazei , and 400 μm DCCD fully inhibited ATP synthesis (Fig. 3), whereas H2 evolution, an indicator for Ech hydrogenase activity, was not affected (Table 1). Taken together, these data indicated that energy is conserved by Ech hydrogenase by the generation of an ion gradient and ATP synthesis by the catalytic activity of the A1AO ATP synthase. However, the nature of the ion translocated over the cytoplasmic membrane still remained unclear. Protons and sodium ions are proposed as coupling ions , but biochemical evidence for either is missing. Therefore, inhibitor studies were performed to identify the translocated ion. It has already been shown that the Na+ ionophore ETH157 effectively dissipates Na+ gradients in vesicular systems of Ms. mazei . As evident from Fig. 3, the addition of ETH157 did not show any effect on the rate of ATP synthesis in the washed membrane vesicle system or on H2 formation (Table 1), indicating that Na+ is not the coupling ion of Ech hydrogenase. In contrast, 10 μm 3,5-di-tert-butyl-4-hydroxybenzylidene-malononitrile (SF6847), a potent protonophore , fully inhibited Fdred-dependent ATP formation. To ensure that SF6847 only abolished the formation of an H+ gradient used for ATP synthesis and not Ech hydrogenase activity, H2 evolution rates were measured (Table 1). Samples containing 10 μm of the protonophore SF6847 exhibited H2 evolving rates of 42 nmol H2 min−1·mg protein−1 and were higher than the control assay without the uncoupler. The effect of SF6847 on the rate of electron transport resembles the phenomenon of respiratory control that was observed previously in the Ms. mazei vesicle system when ATP synthesis was analysed by proton translocation coupled to the H2:heterodisulfide oxidroreductase system .
The energy-conserving transmembrane enzyme system used in the aceticlastic pathway of methanogenesis has been referred to as Fd:heterodisulfide oxidoreductase. The electron flow from Fdred to heterodisulfide reductase in Ms. mazei has been reconstructed in recent years (Fig. 4). Fdred is oxidized by Ech hydrogenase, which produces H2 by proton reduction . The F420 nonreducing hydrogenase oxidizes H2 on the outside of the cytoplasmic membrane , thereby releasing two protons. The electrons and two H+ from the cytoplasm are used for the reduction of methanophenazine, which is a membrane-integral electron carrier in Methanosarcina species . Reduced methanophenazine transfers electrons to heterodisulfide reductase (Fig. 4). The respective protons are released into the extracellular space , thereby generating an electrochemical proton gradient, which is used for ATP synthesis by the A1AO ATP synthase. Energy conservation for Ech hydrogenase based on growth data and experiments on resting cells and cell suspensions has been proposed in several studies [6,12,18–20], but ATP production or generation of an H+ or Na+ gradient directly by Ech hydrogenase has not been reported. The data presented here clearly demonstrate a direct involvement of Ech hydrogenase in energy conservation: (a) ATP synthesis was observed in the Ms. mazei vesicular system that was dependent on the oxidation of Fdred (catalysed by Ech hydrogenase); (b) the Ms. mazeiΔech mutant showed no formation of ATP in the presence of Fdred. In contrast, ATP synthesis from H2 + CoM-S-S-CoB was identical to wild-type levels, indicating that the Δech vesicle preparation was able to establish an ion gradient and that the ATP synthase was active; (c) the addition of protonophore SF6847 led to complete cessation of ATP formation without inhibiting Ech hydrogenase, whereas the sodium ion ionophore ETH157 did not affect ATP formation in this system. Therefore, protons are clearly used as coupling ions.
Proton translocation by Ech hydrogenase is similar to studies performed on the related Mbh hydrogenase from Pyrococcus furiosus , which also translocates protons in the process of Fdred oxidation. Both proteins belong to a small subset of multisubunit [NiFe] hydrogenases within the large group of [NiFe] hydrogenases that use Fdred or polyferredoxin as an electron donor . Members of this group are thought to couple hydrogen formation to energy conservation, primarily based on their homology to the proton pumping NADH:ubiquinone oxidoreductase (complex I). Biochemical evidence of proton translocation has so far only been presented for the Mbh [NiFe] hydrogenase from P. furiosus . Other members of this group are the Coo [NiFe] hydrogenases from Rhodospirillum rubrum  and Carboxydothermus hydrogenoformans , and the Hyc and Hyf [NiFe] hydrogenases from Escherichia coli [24–26]. Ech hydrogenase is now another member of the group of energy-conserving multisubunit [NiFe] hydrogenases that an energy-conserving function can be assigned due to biochemical data and not solely based on sequence similarity to complex I or Mbh hydrogenase of P. furiosus.
It is evident that the proton gradients generated by the Ech hydrogenase from Ms. mazei and the Mbh hydrogenase from P. furiosus are used for ATP synthesis catalysed by A1Ao-type ATP synthases. It has been shown that the enzyme from Ms. mazei has high sequence similarities to the Na+ translocating A1Ao ATPase from P. furiosus, but experimental data clearly show that the enzyme is H+-dependent . In contrast, the ATP synthase from P. furiosus uses the sodium ion gradient for ATP synthesis . Directly adjacent to the Mbh hydrogenase a gene encoding a Na+/H+ antiporter was found. Hence, the electrochemical proton gradient across the cytoplasmic membrane could be converted to a sodium ion potential by action of the Na+/H+ antiporter.
Under standard conditions, the CO-dependent H2 evolution is coupled to a change of free energy of −19.3 kJ·mol−1 (ΔE0` = 0.1 V). According to the equation n =2ΔEh/Δp (with n = number of translocated protons, ΔEh = redox potential difference, Δp = electrochemical potential, which is ∼ 0.15 V in methanogens ), Ech hydrogenase is able to translocate about one proton per hydrogen molecule formed. In many living cells, three protons are needed for the phosphorylation of ADP as catalysed by ATP synthases . Assuming that Ech hydrogenase translocates one proton per hydrogen molecule, the ratio of ATP synthesis and H2 production should be in the range of 0.33. The results presented showed rates of 1.5 nmol ATP·min−1·mg−1 and 32.8 nmol H2·min−1·mg−1, resulting in a ATP/H2 stoichiometry of 0.05 in the vesicular system of Ms. mazei. The apparent discrepancy is most probably due to disintegrated membrane vesicles in the vesicle preparations, which catalyse H2 formation in the process of Fdred oxidation, but do not allow the establishment of an ion gradient . Furthermore, it is possible that part of the A1 subcomplex of the ATP synthase was separated from the Ao subcomplex during the preparation of vesicles, leading to proton flux without ATP synthesis. Hence, the in vivo quotient of ADP phosphorylation over H2 formation is most probably much higher than the experimentally observed ATP/H2 ratio.
Fd is an important cytoplasmic electron carrier in Methanosarcina species. The redox active protein is involved in the process of methanogenesis from H2 + CO2 (carboxymethanofuran reduction ), methylated compounds such as methanol and methylamines (oxidation of formylmethanofuran ) and from acetate (oxidation of CO-bound to CODH/ACS ). The importance of Fd in the metabolism is evident from the finding that the genome of Ms. mazei contains approximately 20 genes encoding these electron transport proteins . Unfortunately, it is unknown which Fd is the natural electron acceptor of CODH/ACS. A couple of heterologously produced Fd were tested for their ability to transfer electrons from CODH/ACS to Ech hydrogenase, but the electron transfer rates were low (not shown). Therefore, the Fd from Clostridium pasteurianum was used in the experiments presented.
The free energy change associated with methane formation from 1 mol acetate is only −36 kJ·mol−1, which allows for the synthesis of less than 1 mol ATP. Thus, the loss of Ech hydrogenase as a proton-translocating enzyme will have a dramatic effect on energy metabolism, as these methanogens already live close to the thermodynamic limit. A severe impact can indeed be observed in Methanosarcina mutants lacking Ech hydrogenase. The Ms. mazeiΔech mutant and the Ms. barkeriΔech mutant are unable to grow on acetate as the sole energy source [20,33]. Growth on trimethylamine as the energy source is still possible for the Ms. mazeiΔech mutant (ΔGo′ = −76 kJ·mol−1 CH4), but with slower growth, less biomass and accelerated substrate consumption . These results underline the importance of Fdred oxidation by Ech hydrogenase in methanogenic pathways. In this context, it is important to mention that Ms. acetivorans, a close relative of Ms. mazei, does not contain an Ech hydrogenase, but is able to grow on acetate. Because Fdred is an essential intermediate in acetate metabolism, Ms. acetivorans must possess an alternative pathway for the utilization of this electron donor. It was suggested that in this organism the Rnf complex could substitute for the Ech hydrogenase .
By taking these data together, a new model of the Fd:heterodisulfide oxidoreductase system in Ms. mazei can be devised (Fig. 4) and the long discussed hypothesis of ion translocation by Ech hydrogenase can be confirmed. The results presented here not only indicate that Ech hydrogenase acts as an additional energy coupling site in methanogenesis from acetate, but also identify the translocated ion as H+. Both H+ and Na+ were feasible possibilities, but the results discussed above clearly exclude the involvement of Na+ in energy conservation by Ech hydrogenase. Instead, the data strongly support the model of proton translocation by Ech hydrogenase, leading to a direct contribution to proton motive force. Thus, Ech hydrogenase acts as primary proton pump in Fdred-dependent electron transport.
Preparation of inverted membrane vesicles, proteins and reagents
All experiments presented here were performed with Ms. mazei strain Gö1 (DSM 7222). Washed inverted membrane vesicles from Ms. mazei and Ms. mazeiΔech  were prepared as described previously . The strains were grown in 1 L glass bottles with 50 mm trimethylamine as the substrate. The preparations were tested for the absence of enzyme activity with the cytoplasmic marker CODH/ACS to ensure the complete removal of cytoplasm from the membrane vesicles. Activity was tested by measuring the change in absorbance at 604 nm with 8.3 mm methylviologen, 5% CO/95% N2 in the gas phase and 300–500 μg vesicle preparation in 40 mm potassium phosphate buffer (including 5 mm dithioerythritol, 1 μg·mL−1 resazurin, pH 7.0) in a total volume of 1 mL. Fd from Clostridium pasteurianum was isolated as described previously . with replacement of the last two steps (dialysation, crystallization) by ultrafiltration. Moorella thermoacetica CODH/ACS was isolated as described previously  with the modifications specified in . Synthesis of CoM-S-S-CoB was carried out as described previously .
Determination of ATP formation
ATP, ADP and AMP were supplied by Serva (Heidelberg, Germany). The inhibitors ETH157, DCCD and SF6847 and firefly lantern extract were supplied by Sigma-Aldrich (Schnelldorf, Germany). ETH157, DCCD and SF6847 were dissolved in 100% ethanol and used at final concentrations of 10–30 μm for ETH157 and SF6847, and 400 μm for DCCD.
To determine ATP formation, rubber stoppered glass vials were filled with 500 μL buffer A (20 mm potassium phosphate, 20 mm MgSO4, 500 mm sucrose, 10 mm dithioerythritol, 1 μg·mL−1 resazurin, pH 7.0), 5% CO/95% N2 in the 1.5 mL headspace, 500–700 μg washed inverted membrane vesicles, 33.5 μg Fd, 150 nmol AMP and 300 nmol ADP. Before starting the reaction by the addition of 20 μg CODH/ACS, the reaction mixture was preincubated for 5 min at 37 °C in a shaking water bath to inhibit the membrane-bound adenylate kinase. This enzyme catalyses the formation of ATP and AMP from two ADP and can be fully inhibited by high concentrations of AMP  present in the reaction mixture. Upon the start of the reaction, 10 μL samples were taken every 2.5 min. ATP detection was performed according to . The samples were mixed with 700 μL 20 mm glycylglycine buffer, pH 8.0, containing 4 mm MgSO4, and 100 μL firefly lantern extract. Emitted light was quantified after 10 s by a luminescence spectrometer LS50B (Perkin Elmer, Boston, MA, USA) at 560 nm and the values compared with a standard curve.
Determination of H2
To determine H2 production rates, rubber stoppered glass vials were filled with 500 μL buffer A, 5% CO/95% N2 in the 1.5 mL headspace, 500–700 μg washed inverted membrane vesicles, 33.5 μg Fd, 20 μg CODH/ACS, 150 nmol AMP and 300 nmol ADP. At various reaction time points, 10 μL of the headspace was injected into a gas chromatograph (GC-14A, Shimadzu, Kyoto, Japan) with argon as the carrier gas. Molecular hydrogen was analysed by a thermal conductivity detector and quantified by comparison with a standard curve.
We thank Elisabeth Schwab for technical assistance and Paul Schweiger for critical reading of the manuscript. Many thanks also go to Gunes Bender and Steve Ragsdale, Department of Biological Chemistry, University of Michigan Medical School for providing the CODH/ACS from Moorella thermoacetica. This work was supported by the Deutsche Forschungsgemeinschaft (grant De488/9-1).