By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
R. Hedderich, Max-Planck-Institut für terrestrische Mikrobiologie, Karl-von-Frisch-Straße, D-35043 Marburg, Germany. Fax: + 49 6421178299, Tel.: + 49 6421178230, E-mail: firstname.lastname@example.org
Methanosarcina barkeri has recently been shown to produce a multisubunit membrane-bound [NiFe] hydrogenase designated Ech (Escherichia coli hydrogenase 3) hydrogenase. In the present study Ech hydrogenase was purified to apparent homogeneity in a high yield. The enzyme preparation obtained only contained the six polypeptides which had previously been shown to be encoded by the ech operon. The purified enzyme was found to contain 0.9 mol of Ni, 11.3 mol of nonheme-iron and 10.8 mol of acid-labile sulfur per mol of enzyme. Using the purified enzyme the kinetic parameters were determined. The enzyme catalyzed the H2 dependent reduction of a M. barkeri 2[4Fe-4S] ferredoxin with a specific activity of 50 U·mg protein−1 at pH 7.0 and exhibited an apparent Km for the ferredoxin of 1 µm. The enzyme also catalyzed hydrogen formation with the reduced ferredoxin as electron donor at a rate of 90 U·mg protein−1 at pH 7.0. The apparent Km for the reduced ferredoxin was 7.5 µm. Reduction or oxidation of the ferredoxin proceeded at similar rates as the reduction or oxidation of oxidized or reduced methylviologen, respectively. The apparent Km for H2 was 5 µm. The kinetic data strongly indicate that the ferredoxin is the physiological electron donor or acceptor of Ech hydrogenase. Ech hydrogenase amounts to about 3% of the total cell protein in acetate-grown, methanol-grown or H2/CO2-grown cells of M. barkeri, as calculated from quantitative Western blot experiments. The function of Ech hydrogenase is ascribed to ferredoxin-linked H2 production coupled to the oxidation of the carbonyl-group of acetyl-CoA to CO2 during growth on acetate, and to ferredoxin-linked H2 uptake coupled to the reduction of CO2 to the redox state of CO during growth on H2/CO2 or methanol.
E. coli hydrogenase 3 type hydrogenase of Methanosarcina barkeri
H-S-CoM or 2-mercaptoethanesulfonate
heterodisulfide of coenzyme M and coenzyme B
Methanosarcina barkeri is a methanogenic archaeon that can utilize H2/CO2, methanol or methylamines as energy substrates (reviewed in [1–3]). Three different [NiFe] hydrogenases have been characterized from Methanosarcina species: F420-reducing hydrogenase, F420-nonreducing hydrogenase and Ech hydrogenase.
F420-reducing hydrogenase has been purified and characterized from M. barkeri[4,5]. The enzyme catalyzes the reduction of coenzyme F420 which plays an important role as redox carrier in methanogenic archaea. The genome of M. barkeri contains two gene clusters (frh and fre) encoding two related F420-reducing hydrogenases indicating the presence of two isoenzymes. Both operons were transcribed during growth on H2/CO2, methanol or trimethylamine . In acetate-grown cells no transcripts of these operons were detectable although these cells contain small amounts of F420-reducing hydrogenase activity (about 5% of the activity detectable in methanol-grown cells; Meuer and Hedderich, unpublished results).
F420-nonreducing hydrogenase has been purified from M. barkeri and from Methanosarcina mazei. The purified enzyme was found to be composed of two different subunits [7,8]. An analysis of the operon encoding these subunits in M. mazei indicated the presence of a third subunit, not present in the purified enzyme. This subunit was shown to be a b-type cytochrome, which is assumed to function as a membrane anchor of this hydrogenase in vivo. It was also shown that M. mazei encodes two closely related F420-nonreducing hydrogenases, designated Vho and Vht. The vho genes were found to be expressed constitutively, whereas the vht genes were only expressed during growth of the organism on H2/CO2 or methanol, rather than on acetate .
Ech hydrogenase is a multisubunit membrane-bound [NiFe] hydrogenase that has only recently been described . This hydrogenase was identified during the purification of heterodisulfide reductase from acetate-grown cells of M. barkeri. Heterodisulfide reductase and Ech hydrogenase were found to coelute on different chromatography columns. The preparations obtained contained eight polypeptides with apparent molecular masses of 55, 46, 39, 24, 23, 20, 16 and 15 kDa. The 46-kDa polypeptide and the 23-kDa polypeptide were shown to be subunits of heterodisulfide reductase . The remaining polypeptides were shown to be encoded by one operon, the echABCDEF operon . A sequence analysis of the different polypeptides indicated that the enzyme, which was designated as Ech hydrogenase, belongs to a small group of multisubunit membrane-bound [NiFe] hydrogenases. The list of enzymes includes hydrogenase 3 from Escherichia coli[13,14], hydrogenase 4 from E. coli, and carbon monoxide-induced hydrogenase from Rhodospirillum rubrum[16,17]. Two integral membrane proteins and three hydrophilic polypeptides of these multisubunit enzymes are highly conserved. A fourth hydrophilic protein present in these enzymes only shows a low sequence similarity. Via sequence analysis two of the hydrophilic polypeptides were identified as the so called ‘hydrogenase large subunit’ and the ‘hydrogenase small subunit’ which are found in all [NiFe] hydrogenases. The hydrogenase large subunit contains the four conserved cysteine residues acting as ligands of the binuclear [NiFe] active site. The hydrogenase small subunit of these membrane-bound hydrogenases is considerably smaller than that of other [NiFe] hydrogenases and contains only the cysteine ligands for the proximal [4Fe-4S] cluster. A sequence analysis of both the hydrogenase large and the hydrogenase small subunit indicates that these proteins are more closely related to subunits of the proton-pumping NADH:quinone oxidoreductase (complex I) from various organisms than to subunits of other [NiFe] hydrogenases [11,13,16,18]. The third hydrophilic subunit present in these enzymes contains two binding motifs for [4Fe-4S] clusters. A homologue of this subunit is also found in complex I. The two conserved integral membrane subunits also have closely related homologues in complex I. To differentiate these multisubunit membrane-bound hydrogenases from ‘standard’[NiFe] hydrogenases we have designated this group of enzymes as E. coli hydrogenase 3-type hydrogenases .
E. coli hydrogenase 3 and CO-induced hydrogenase from R. rubrum catalyze in vivo hydrogen formation. E. coli hydrogenase 3 is part of the formate–hydrogen lyase complex which catalyzes reaction a.
(a)HCOO-+H+⇌CO2+H2 ΔG°′=–3 kJ·mol-1
Andrews et al.  have proposed that E. coli hydrogenase 4 is part of a second formate–hydrogen lyase complex. Under natural conditions, the formate–hydrogen lyase reaction is exergonic, and the authors proposed that the reaction is coupled to energy conservation . The CO-induced hydrogenase from R. rubrum catalyzes, together with carbon monoxide dehydrogenase, reaction b.
(b)CO+H2O⇌CO2+H2 ΔG°′=–20 kJ·mol-1
Since R. rubrum can utilize CO as sole energy source, this reaction should be coupled to energy conservation. The hydrogenase has been proposed to be the coupling site [16,17].
Ech hydrogenase from M. barkeri has been partially purified from acetate-grown cells. The physiological function of this enzyme has not yet been elucidated. In this communication we describe the purification and the catalytic properties of Ech hydrogenase. The data presented strongly indicate that a ferredoxin is the physiological redox partner of this enzyme.
Materials and methods
M. barkeri strain Fusaro (DSM 804) and M. thermophila strain TM-1 (DSM 1825) were from the Deutsche Sammlung für Mikroorganismen und Zellkulturen (Braunschweig). M. barkeri was grown on methanol, acetate or H2/CO2 at 37°C as described in . M. thermophila was cultivated on acetate as described in . Dodecyl-β-d-maltoside and diphenyleneiodonium were from Fluka (Neu-Ulm). 2-Hydroxyphenazine was synthesized as described in . All other chemicals were from Merck (Darmstadt) or Sigma (Deisenhofen).
Preparation of cell extracts
Cell extracts were routinely prepared from 25 g of cells which were suspended in 150 mL of 50 mm Mops/NaOH pH 7.0, containing 2 mm dithiothreitol (buffer A). Cells were disrupted by sonication at 4°C in intervals of 3 × 5 min using an energy output of 200 W (Bandelin sonicator) under a gas phase of 5% H2/95% N2. Undisrupted cells and cell debris were removed by centrifugation at 10 000 g for 30 min.
Purification of Ech hydrogenase
All purification steps were performed under strictly anaerobic conditions under an atmosphere of 5% H2/95% N2. The crude membrane fraction was isolated from cell extracts by centrifugation at 150 000 g for 2 h. The 150 000 g supernatant was stored for the isolation of ferredoxin, carbon monoxide dehydrogenase/acetyl-CoA synthase and F420-reducing hydrogenase (see below). The crude membrane fraction was used for the isolation of Ech hydrogenase. Crude membranes were homogenized in 20 mL of buffer A using a Teflon potter and then suspended in a total volume of 150 mL of buffer A containing 15 mm dodecyl-β-d-maltoside (1.5 mg of protein/mL). The suspension was incubated for 12 h at 4°C under slight swirling. After centrifugation at 150 000 g for 30 min the solubilized membrane proteins present in the supernatant were loaded on a Q-Sepharose HiLoad column (2.6 × 10 cm) equilibrated with buffer A containing 4 mm dodecyl-β-d-maltoside (= buffer A + detergent). The column was washed with 50 mL of buffer A + detergent and proteins were eluted with NaCl in buffer A + detergent in a linear gradient from 0.1 m to 0.18 m NaCl (200 mL) and in two steps of 0.4 m NaCl (70 mL) and 1 m NaCl (70 mL). Ech hydrogenase was recovered in the fractions eluting with 0.13 m NaCl. Most of the heterodisulfide reductase was recovered in the fractions eluting with 0.11 m NaCl. The F420-nonreducing hydrogenase was present in the 1 m NaCl fraction (see below). Ech hydrogenase that still contained some minor protein contaminants was further purified by chromatography on hydroxyapatite. The column (1 × 15 cm) was equilibrated with 30 mm potassium phosphate buffer pH 7.0 containing 4 mm dodecyl-β-d-maltoside. Protein was eluted using a potassium phosphate step gradient: 75 mm (50 mL), 150 mm (50 mL), 265 mm (50 mL), 380 mm (50 mL) and 500 mm (100 mL). Ech hydrogenase was recovered in the fractions eluting with 500 mm potassium phosphate. The protein was concentrated by ultrafiltration (Molecular/Por cellulose ester ultrafiltration membranes, cut-off 100 kDa; Spectrum, Houston, TX, USA). Protein was stored in buffer A + detergent at protein concentrations of 3 mg·mL−1 and 4°C or −20°C.
Purification of ferredoxin, carbon monoxide dehydrogenase/acetyl-CoA synthase, F420-reducing hydrogenase and F420-nonreducing hydrogenase
The M. barkeri ferredoxin was purified under anaerobic conditions from the 150 000 g supernatant by chromatography on DEAE-Sephacel, Q-Sepharose HiLoad and Superose 12 as previously described . The isolated oxidized ferredoxin showed absorption maxima at 280 and 390 nm, and a A390/A280 ratio of 0.8. An ε390 value of 12.8 mm−1·cm−1 was used for the determination of ferredoxin concentrations.
Carbon monoxide dehydrogenase/acetyl-CoA synthase was partially purified from the 150 000 g supernatant by chromatography on DEAE-Sephacel and Q-Sepharose HiLoad. The preparation obtained contained five major polypeptides corresponding to the subunits of carbon monoxide dehydrogenase/acetyl-CoA synthase as shown previously . The preparation catalyzed the ferredoxin dependent reduction of metronidazole by CO at a rate of 13 U·mg protein−1.
F420-nonreducing hydrogenase present in the 1 m NaCl eluate of the Q-Sepharose HiLoad column, obtained during the purification of Ech hydrogenase (see above), was further purified by chromatography on hydroxyapatite using the conditions described for the purification of Ech hydrogenase. The enzyme eluted at a concentration of 75 mm potassium phosphate. An SDS/PAGE of the fractions containing F420-nonreducing hydrogenase activity indicated the presence of two major polypeptides corresponding to apparent molecular masses of 57 kDa and 35 kDa. These polypeptides had been previously shown to be subunits of F420-nonreducing hydrogenase of M. barkeri.
F420-reducing hydrogenase present in the 150 000 g supernatant of cell extract was partially purified and separated from Ech hydrogenase by chromatography on DEAE-Sephacel and hydroxyapatite. The preparation catalyzed the reduction of coenzyme F420 at a rate of 3 U·mg protein−1.
Determination of enzyme activities
The assays were routinely performed at 37 °C either in 8-mL serum bottles or in 1.5-mL cuvettes under anaerobic conditions. (1 U of enzyme activity corresponds to 1 µmol of H2 formed or consumed per min.)
Hydrogen uptake-activity with dyes as electron acceptors was determined by following the reduction of methylviologen, benzylviologen or methylene blue at 578 nm or the reduction of 2-hydroxyphenazine at 425 nm. The 0.8-mL assays contained 50 mm Tris/HCl pH 7.6, 2 mm methylviologen, 2 mm benzylviologen, 0.2 mm methylene blue or 0.2 mm 2-hydroxyphenazine. In standard assays, cuvettes were allowed to equilibrate with a 100% H2 headspace (1.2 × 105 Pa). For kinetic studies, varying amounts of H2 were injected into a N2 headspace of the assay buffer and the solution was equilibrated by vigorous shaking for 10 min. The dissolved H2 was calculated from solubility data . The inhibition of 2-hydroxyphenazine reduction by diphenyleneiodonium (DPI) was tested by adding DPI to the assay at end concentrations of 10 µm to 150 µm.
The H2 formation activity with reduced methylviologen as electron donor was measured by following the oxidation of reduced methylviologen at 578 nm. The standard assay contained 50 mm Tris/HCl pH 7.6, 2 mm methylviologen, which was reduced with sodium dithionite to a ΔA578 of 2 and N2 (1.2 × 105 Pa) as the gas phase. The reaction was started by addition of protein. Enzyme inhibition by CO was determined in the H2 formation assay by adding CO to the headspace of the assay cuvettes before addition of Ech hydrogenase to yield from 0 to 100% CO in the headspace.
The ferredoxin dependent reduction of metronidazole by H2 was measured by following the decrease in absorbance at 320 nm (εmetronidazole = 9.3 mm−1·cm−1, corresponding to a six-electron reduction) . The 0.8-mL assay contained 50 mm Mops/NaOH pH 7.0, 4 mm dodecyl-β-d-maltoside, 0.2 mm metronidazole, 10 µm ferredoxin, protein (purified Ech hydrogenase or fractions obtained during the purification) and 1.2 × 105 Pa of 100% H2 as the gas phase. To ensure completely anaerobic conditions 0.1 mm sodium dithionite was added.
The ferredoxin dependent H2 formation with sodium dithionite as electron donor was followed by determining the H2 concentration in the gas phase by gas chromatography. The 1-mL assay in 8-mL serum bottles contained 50 mm Mops/NaOH pH 7.0, 10 mm sodium dithionite, 20 µm ferredoxin (or as indicated) and 8 µg of purified Ech hydrogenase under N2 as gas phase (1.2 × 105 Pa). The solution was stirred vigorously with a magnetic bar. At time intervals of 1 min samples from the gas phase were withdrawn and H2 was quantified after separation by gas chromatography.
The ferredoxin dependent H2 formation from CO was measured by following H2 formation via gas chromatography as described above. The 1-mL assay mixture in 8-mL serum bottles contained 100 mm potassium phosphate buffer pH 7.0, 240 µg of carbon monoxide dehydrogenase/acetyl-CoA synthase, 20 µm ferredoxin and 12 or 24 µg of Ech hydrogenase. The gas phase was N2 with 17% CO at 1.2 × 105 Pa.
Nonheme iron was quantified colorimetrically with neocuproin (2,9-dimethyl-1,10-phenanthroline) and ferrozine[3-(2-pyridyl)- 5,6-bis(4-phenyl sulfonate)-1,2,4-triazine] as described by Fish . Acid-labile sulfur was analyzed as methylene blue .
Nickel was determined by atomic absorption spectroscopy on a 3030 Perkin Elmer atomic absorption spectrometer fitted with a HGA-600 graphite furnace assembly and an AS-60 autosampler.
Protein was determined by the method of Bradford  or by the bicinchoninic acid method .
Determination of the intracellular concentration of Ech hydrogenase
A chemiluminescent Western blotting system (ECF, Amersham Pharmacia, Freiburg), was used to quantify the Ech hydrogenase contents of M. barkeri cells grown on H2/CO2, methanol or acetate. Protein in cell extracts or purified Ech hydrogenase were denatured by incubation in 62 mm Tris/HCl pH 6.8 that contained 2% SDS and 16 mm dithiothreitol for 30 min at 20°C. Polypeptides were separated by SDS/PAGE and transferred to nitrocellulose membranes by electroblotting. Immunodetection was performed using the protocol for the ECF detection kit provided by Amersham Pharmacia with 1 : 1000 dilutions of a rabbit anti-Ech serum and 1 : 10 000 dilutions of alkaline phosphatase-conjugated anti-(rabbit IgG) antibody. After incubation with the ECF detection reagent, a phosphor storage screen of a Phosphor Imager (Storm 860, Molecular Dynamics, Krefeld) was exposed to the blots. Signals were analyzed using the ImageQuant software (Molecular Dynamics). A calibration curve using known amounts of Ech hydrogenase (60–320 ng of protein) was generated and was used to calculate the Ech content in cell extracts.
The anti-(Ech hydrogenase) serum was generated by Eurogentec (Seraing) from a rabbit immunized with a protein mixture of Ech hydrogenase and heterodisulfide reductase.
Purification and molecular properties of Ech hydrogenase
Using the purification procedure described previously , enzyme preparations were obtained which contained both heterodisulfide reductase and Ech hydrogenase. An SDS/PAGE of this preparation indicated the presence of eight polypeptides corresponding to the two subunits of heterodisulfide reductase and to the six polypeptides encoded by the ech operon. From 25 g of cells (wet mass) approximately 2–3 mg of this protein mixture was obtained. In addition to Ech hydrogenase, acetate-grown cells of M. barkeri contain the F420-nonreducing hydrogenase which is present in large amounts in these cells and the F420-reducing hydrogenase which is present in small amounts in these cells. It was therefore not possible to follow the purification of Ech hydrogenase by enzyme activity measurements using the standard viologen dye assay. Therefore the enzyme preparation containing heterodisulfide reductase and Ech hydrogenase was used to generate anti-Ech/anti-Hdr sera. An immunodetection was then used to follow the purification of Ech hydrogenase. It was found that more than 50% of Ech hydrogenase present in the membrane fraction was not solubilized from the membrane with Chaps, which had been used as detergent. Dodecyl-β-d-maltoside was found to be much more effective, because more than 95% of the enzyme was released from the cytoplasmic membrane fraction using this detergent. The enzyme was further purified by anion-exchange chromatography and hydroxyapatite chromatography in the presence of dodecyl-β-d-maltoside. Using this protocol about 12–16 mg of highly purified Ech hydrogenase was obtained from 25 g of cells (1700 mg of protein). This enzyme preparation only contained the six subunits which had previously been shown to be encoded by the ech operon (Fig. 1). Ech hydrogenase and heterodisulfide reductase, which were both present in the solubilized membrane fraction, were separated to more than 90% by anion-exchange chromatography on Q-Sepharose, and were completely separated by hydroxyapatite chromatography. Using this purification procedure also heterodisulfide reductase was obtained in high purity and high yield (20–25 mg of protein). The coelution of both enzymes observed previously  may be due to an insufficient solubilization of both enzymes which may result in large lipid–protein aggregates.
Later it was found that a M. barkeri 2[4Fe-4S] ferredoxin functions a as substrate of Ech hydrogenase, in contrast to the two other hydrogenases of M. barkeri, which do not catalyze the reduction of this ferredoxin (see below). Therefore, the hydrogen dependent reduction of the ferredoxin by Ech hydrogenase was used to follow Ech hydrogenase activity during the purification from the membrane fraction (Table 1). The 150 000 g supernatant of the cell extract was also found to contain substantial amounts of Ech hydrogenase as revealed by an immunoblot and by activity measurements. Up to 40% of the enzyme was found in this fraction. When this fraction was loaded on a DEAE-Sephacel column or other anion-exchange columns, the enzyme was completely found in the void volume of these columns (exclusion volume 1 × 106 Da). The fractions were turbid, indicating the presence of small membrane particles. Hence, Ech hydrogenase is considered to be also tightly membrane bound in the 150 000 g supernatant. The enzyme was not purified from this fraction.
Table 1. Purification of Ech hydrogenase from M. barkeri strain Fusaro. The enzyme was purified from 25 g cells (1700 mg of protein). 1 U of activity is defined as the amount of enzyme that catalyzes the ferredoxin dependent reduction of metronidazole with 1 µmol of H2 min−1. The yield refers to the recovery of activity units.
Total activity (U)
Total protein (mg)
Specific activity (U·mg−1)
Solubilized membrane proteins
The purified Ech hydrogenase was analyzed for its content on Ni, acid-labile sulfur and nonheme iron. Based on a calculated molecular mass of the enzyme of 184 kDa and based on a protein determination using the bicinchoninic acid method, 0.9 mol of Ni, 10.8 mol of acid-labile sulfur and 11.3 mol of nonheme iron were found. Per mol of Ni the enzyme thus contains 12.5 mol of Fe and 12 mol of acid-labile sulfur. From the primary structure, the enzyme is predicted to bind three [4Fe-4S] clusters and one binuclear [Ni-Fe] center, corresponding to 13 mol of nonheme iron and 12 mol of acid-labile sulfur.
The activity of the enzyme, purified under an atmosphere of 5% H2/95% N2 with 2 mm dithiothreitol in all buffers, could not be further increased by incubation under 100% H2 for 60 min. The enzyme was stable at 4°C under 100% H2 for several days.
The enzyme was rapidly inactivated by oxygen. After 30 min exposure to air only 10% enzyme activity was retained. After a short exposure to air (< 30 min) the enzyme could be fully reactivated under reducing conditions using reduced methylviologen and H2.
Catalytic properties of Ech hydrogenase
Ech hydrogenase catalyzed hydrogen uptake with several artificial electron acceptors. Highest activities were obtained with benzylviologen (210 U·mg protein−1), methylene blue (123 U·mg protein−1) and methylviologen (53 U·mg protein−1) at pH 7.6. The apparent Km for H2 was determined to be 5 µm using methylviologen as electron acceptor.
Hydrogen formation was assayed with reduced methylviologen as an electron donor. The pH dependence of the hydrogen uptake and hydrogen formation with methylviologen as electron donor or acceptor is shown in Fig. 2. At pH 7 both reactions proceed at a rate of 35–40 U·mg protein−1.
The activity of Ech hydrogenase, measured by following H2 formation from reduced methylviologen, was found to be inhibited by CO. A 50% inhibition was observed with 7.5% CO in the gas phase (≈ 65 µm CO in solution). Hence, Ech hydrogenase is more sensitive to CO than the closely related CO-induced hydrogenase from R. rubrum, which is inhibited to 50% at 300 µm CO . The CO sensitivity of Ech hydrogenase seems to be more similar to that of the standard [NiFe] hydrogenase from Desulfovibrio gigas.
Acetate-grown cells of M. barkeri contain a 2[4Fe-4S] ferredoxin which has been shown to be essential for CO-dependent H2 formation catalyzed by cell extracts of M. barkeri. It was therefore tested if Ech hydrogenase is able to catalyze the reduction of this ferredoxin with hydrogen as electron donor or to catalyze H2 formation with reduced ferredoxin as electron donor. To determine the ferredoxin reduction rate, the metronidazole assay was used . Ech hydrogenase catalyzed the reduction of metronidazole by H2 only in the presence of the ferredoxin, which indicates that the ferredoxin is a direct electron acceptor of the enzyme (Fig. 3A). Reduced ferredoxin is oxidized in a fast chemical reaction by metronidazole. The metronidazole reduction rate increased linearly with increasing Ech hydrogenase concentrations (Fig. 3B). Ech hydrogenase catalyzed the reduction of metronidazole and thus the reduction of the ferredoxin at a Vmax of 50 U·mg−1 of enzyme. The enzyme exhibited an apparent Km for the ferredoxin of 1 µm. Assuming a molecular mass for Ech hydrogenase of 184 kDa, a catalytic efficiency coefficient (kcat/Km) of 1.5 × 108m−1·s−1 was calculated.
Reduction of the ferredoxin was reversible. Ech hydrogenase catalyzed hydrogen formation with sodium dithionite-reduced ferredoxin as electron donor at an apparent Vmax of 90 U·mg protein−1. The direct reduction of the enzyme by sodium dithionite proceeded only at very low rates as in the absence of ferredoxin the rate of hydrogen formation was only 2 U·mg−1 of Ech hydrogenase (Fig. 4A). The rate of hydrogen formation from reduced ferredoxin increased linearly with increasing Ech hydrogenase concentrations (Fig. 4B). Using this assay an apparent Km for reduced ferredoxin of 7.5 µm was determined. The catalytic efficiency coefficient was calculated to be 3.7 × 107m−1·s−1.
In control experiments the two standard [NiFe] hydrogenases of M. barkeri were tested for their ability to catalyze the reduction of the M. barkeri ferredoxin. F420-nonreducing hydrogenase, partially purified from M. barkeri, did not catalyze ferredoxin reduction, which is in accord with the results obtained by Kemner and Zeikus . F420-reducing hydrogenase, partially purified from M. barkeri, also did not catalyze ferredoxin reduction, which is in line with the findings of Grahame & DeMoll . Kemner and Zeikus  had reported that F420-reducing hydrogenase, present in acetate-grown cells of M. barkeri, catalyzes ferredoxin reduction at high rates. These experiments were, however, performed with a crude enzyme preparation which most probably contained Ech hydrogenase and F420-reducing hydrogenase.
Ech hydrogenase is related to energy conserving NADH: quinone oxidoreductases. One of the two integral membrane subunits of Ech hydrogenase is a homologue of a subunit which is implicated in the binding of the quinone in complex I . It may therefore be concluded that quinone-like compounds act as a physiological electron acceptor of Ech hydrogenase. Methanogens do not contain quinones but from M. mazei a membrane-bound electron carrier designated methanophenazine was characterized which is a 2-hydroxyphenazine linked to an isoprenoid side chain via an ether linkage . Methanophenazine is thought to have a quinone-like function in the electron transport chain of Methanosarcina species. Because methanophenazine is highly hydrophobic the water soluble 2-hydroxyphenazine has been used in enzyme assays . Ech hydrogenase was found to catalyze the reduction of 2-hydroxyphenazine with an apparent Vmax of 9 U·mg protein−1 and was found to exhibit an apparent Km for 2-hydroxyphenazine of 0.2 mm. Do these data indicate that methanophenazine is a physiological electron acceptor of Ech hydrogenase or does the water soluble 2-hydroxyphenazine function as an artificial electron acceptor of Ech hydrogenase? Control experiments were performed with two nonmembrane-bound hydrogenases, the F420-reducing hydrogenase and the F420-nonreducing hydrogenase from Methanobacterium thermoautotrophicum which were available in a highly purified state. It was found that these enzymes also catalyze the reduction of 2-hydroxyphenazine. The F420-reducing hydrogenase catalyzed the reduction of 2-hydroxyphenazine at a rate of 25 U·mg protein−1 and the reduction of coenzyme F420 at a rate of 5 U·mg protein−1. The F420-nonreducing hydrogenase catalyzed the reduction of 2-hydroxyphenazine at a rate of 1.5 U·mg protein−1 and the reduction of benzylviologen at a rate of 300 U·mg protein−1. At least for the F420-reducing hydrogenase, a physiological function as methanophenazine-reducing hydrogenase can be excluded because the role of this enzyme in reduction of coenzyme F420 is well established [4,35]. It may therefore be concluded that 2-hydroxyphenazine functions as artificial electron acceptor of these hydrogenases. To further address this question, it was tested if the reduction of 2-hydroxyphenazine catalyzed by Ech hydrogenase is inhibited by DPI. DPI has been reported to be an inhibitor of M. mazei F420H2 dehydrogenase which catalyzes the F420H2 dependent reduction of 2-hydroxyphenazine. The inhibition is competitive with respect to 2-hydroxyphenazine (Ki = 0.45 µm) . DPI was also found to inhibit the reduction of 2-hydroxyphenazine by H2 catalyzed by the membrane fraction of M. mazei. The latter reaction is assumed to be catalyzed by the F420-nonreducing hydrogenase (Vho) present in the membrane fraction of M. mazei. In this work it was found that 2-hydroxyphenazine reduction catalyzed by Ech hydrogenase was not inhibited by DPI at concentrations up to 100 µm, which indicates that the binding sites of Ech hydrogenase and F420H2 dehydrogenase for 2-hydroxyphenazine are different, and that Ech hydrogenase probably binds 2-hydroxyphenazine at an artificial, low affinity site.
Reconstitution of a CO-oxidizing:H2-generating system
Cell suspensions of M. barkeri have previously been shown to catalyze the conversion of CO (5% CO in the gas phase) to CO2 and H2 at a rate of 80–120 mU·mg−1 of cell protein. The rate increased up to 200 mU·mg protein−1 in the presence of the uncoupler tetrachlorosalicylanilide. These cell suspensions catalyzed methane formation from acetate at a rate of 100–200 mU·mg protein−1. These in vivo rates were compared with the rates obtained with an in vitro system composed of purified M. barkeri Ech hydrogenase, M. barkeri ferredoxin and M. barkeri carbon monoxide dehydrogenase/acetyl-CoA synthase. Ech hydrogenase was present in rate-limiting amounts in the assay mixture. This system catalyzed the conversion of CO to CO2 and H2 at a rate of 15 U·mg−1 of Ech hydrogenase, calculated for a concentration of 5% CO in the gas phase (Fig. 5). Electron transfer between carbon monoxide dehydrogenase and Ech hydrogenase was strictly dependent on the presence of the ferredoxin, excluding a direct electron transfer between both enzymes. Based on an Ech hydrogenase content of about 3% of the total cell protein (see below) the in vivo activity should be in the order of 0.45 U·mg−1 of cell protein, provided the oxidation of the ferredoxin by Ech hydrogenase is the rate limiting step in vivo. Hence, the in vivo and the in vitro activities are in the same order which indicates that Ech hydrogenase is responsible for the CO-dependent H2 formation catalyzed by whole cells.
Presence of Ech hydrogenase in methanol and H2/CO2-grown cells of M. barkeri and in acetate-grown cells of M. thermophila
Immunoblotting experiments with anti-Ech sera showed that Ech hydrogenase is not only present in acetate-grown cells of M. barkeri, from which the enzyme has been purified, but also in methanol-grown and H2/CO2-grown cells. A quantitative immunoblot analysis showed that the content of Ech hydrogenase in acetate-grown cells is 3 ± 1% of the total cell protein. Similar amounts of Ech hydrogenase were detected in methanol-grown and H2/CO2-grown cells. This correlates with the specific activities of Ech hydrogenase determined in cell extracts. Acetate-grown cells catalyzed ferredoxin dependent H2 uptake at a rate of 1.6 U·mg protein−1, methanol-grown cells at a rate of 2 U·mg protein−1 and H2/CO2-grown cells at a rate of 0.7 U·mg protein−1.
The anti-Ech serum was used to detect Ech hydrogenase in cell extracts of acetate-grown cells of M. thermophila, a Methanosarcina species that is not able to grow on H2/CO2 as energy substrates . M. thermophila cell extract was found to cross-react with the anti-(M. barkeri Ech) serum (Fig. 6). A 55-kDa polypeptide and a 39-kDa polypeptide corresponding to the molecular masses of the subunits EchA and EchE of M. barkeri Ech hydrogenase gave the strongest reaction in the immunoblot with M. thermophila cell extract. The subunits EchB, EchC, EchD and EchF of M. barkeri Ech hydrogenase only exhibit a very weak reaction with the antiserum. Polypeptides corresponding to the molecular masses of these subunits were also detected in M. thermophila cell extract. The presence of Ech hydrogenase in acetate-grown cells of M. thermophila cannot be due to the constitutive expression of an enzyme only needed in H2/CO2 metabolism. Therefore these results can be taken as proof that the enzyme plays an essential function in the acetate metabolism of Methanosarcina species.
Ech hydrogenase was found to be a major membrane protein in acetate-grown, methanol-grown or H2/CO2-grown cells of M. barkeri. The data provided show that the enzyme efficiently mediates electron transfer to a 2[4Fe-4S] ferredoxin and catalyzes H2 formation with reduced ferredoxin as electron donor. In the following the physiological role of these reactions in the different energy metabolic pathways of M. barkeri will be discussed.
During growth on acetate, cleavage of the acetate molecule is catalyzed by carbon monoxide dehydrogenase/acetyl-CoA synthase (reviewed in ). This reaction generates enzyme-bound CO and an enzyme-bound methyl group. The methyl group is transferred to coenzyme M (H-S-CoM or 2-mercaptoethanesulfonate) via tetrahydrosarcinapterin. The methyl group of methyl-coenzyme M is subsequently reduced by coenzyme B (H-S-CoB or N-7-mercaptoheptanoylthreonine phosphate) to CH4, thereby forming the heterodisulfide (CoM-S-S-CoB). The CO bound to carbon monoxide dehydrogenase/acetyl-CoA synthase is oxidized to CO2, and the reducing equivalents are used for the reduction of CoM-S-S-CoB. In M. barkeri, a 2[4Fe-4S] ferredoxin with a midpoint potential (E°′) of −420 mV has been shown to be the direct electron acceptor of carbon monoxide dehydrogenase/acetyl-CoA synthase [23,31]. The electron transfer pathway coupling ferredoxin oxidation with CoM-S-S-CoB reduction has not yet been elucidated. Two possible pathways will be discussed.
The first pathway assumes that H2 is an obligatory intermediate of this electron transport chain. This ‘intraspecies hydrogen cycling’ pathway is based on the finding that cultures of Methanosarcina species produce significant levels of H2 during growth on acetate [38–40]. Furthermore, it was found that cell suspensions and cell extracts of M. barkeri catalyze the conversion of CO to CO2 and H2[22,37,41]. A CO-oxidizing:H2-generating activity was reconstituted with purified carbon monoxide dehydrogenase/acetyl-CoA synthase, ferredoxin and the membrane fraction from M. thermophila. It is assumed that in vivo the carbonyl group of acetyl-CoA instead of free CO is converted to CO2 and H2. The hydrogen thus formed has to be recaptured by the cell to generate reducing equivalents for the reduction of the heterodisulfide (CoM-S-S-CoB). This model requires the presence of two hydrogenases, a hydrogen-evolving and a hydrogen-uptake hydrogenase.
The data provided in this work support this model. They suggest that Ech hydrogenase catalyzes H2 formation with reduced ferredoxin as electron donor (Fig. 7). The catalytic efficiency coefficient determined is in agreement with the function of Ech hydrogenase as a physiological redox partner of the ferredoxin. In contrast to Ech hydrogenase, the two other [NiFe] hydrogenases present in M. barkeri cannot utilize the ferredoxin as substrate. In addition to Ech hydrogenase, acetate-grown cells of M. barkeri and M. mazei produce another membrane-bound hydrogenase, the F420-nonreducing hydrogenase (Vho) [7–9]. The enzyme contains a b-type cytochrome as membrane anchor . The vhoG gene encodes a ‘twin arginine’ leader peptide typical for enzymes translocated across the cytoplasmic membrane . It therefore can be assumed that the catalytic subunit of the enzyme is exposed to the extracellular side of the cytoplasmic membrane. The enzyme has been proposed to be part of an electron transport chain with H2 as electron donor and the heterodisulfide (CoM-S-S-CoB) as electron acceptor [34,36]. Hence, the H2 generated by Ech hydrogenase may diffuse to the extracellular side of the cytoplasmic membrane where it is recaptured by Vho hydrogenase. The reduction of CoM-S-S-CoB by H2 in acetate metabolism could involve the same electron transport chain as in H2/CO2 metabolism. The growth of Methanosarcina in cell aggregates would have enormous benefit for the genus because hydrogen lost by one cell could be regained by an adjacent cell.
An alternative model suggests that Methanosarcina species contain an electron transport chain that directly channels electrons from carbon monoxide dehydrogenase/acetyl-CoA synthase via the ferredoxin to heterodisulfide reductase. Such a CO:heterodisulfide oxidoreductase activity has been reconstituted with purified carbon monoxide dehydrogenase/acetyl-CoA synthase, ferredoxin, washed membranes and partially purified heterodisulfide reductase . Because this in vitro system still contained the membrane fraction and thus both membrane-bound hydrogenases, Vho and Ech, it cannot be excluded that H2 is an intermediate in this system. A similar reconstitution experiment was performed later on by Simianu et al.  with purified carbon monoxide dehydrogenase/acetyl-CoA synthase, ferredoxin and purified heterodisulfide reductase. This system catalyzed reduction of the heterodisulfide by CO but the catalytic efficiency coefficient was calculated to be 2.0 × 105m−1·s− 1, which led to the conclusion that the ferredoxin is not a physiological electron donor of heterodisulfide reductase . The kcat/Km of Ech hydrogenase for ferredoxin is about 800-fold higher in the H2 uptake assay and about 175-fold higher in the H2 formation assay than that of heterodisulfide reductase, which supports our assumption that the ferredoxin is the physiological electron donor or acceptor of Ech hydrogenase.
The ‘intraspecies H2 cycling model’ might offer the advantage of having two energy coupling sites in the electron transfer chain between the ferredoxin and the heterodisulfide (CoM-S-S-CoB). The reduction of the heterodisulfide CoM-S-S-CoB with H2 is a well established energy coupling step in the metabolism of Methanosarcina. Based on its similarity to complex I and based on the finding that the conversion of CO to CO2 and H2, as catalyzed by cell suspensions of M. barkeri, is coupled to the generation of a proton-motive force [37,41], it is tempting to speculate that Ech hydrogenase functions as proton pump and forms a second site of energy conservation in this electron transport chain.
The generation of H2 from reducing equivalents formed in the oxidation of the carbonyl group of acetate seems to be not the only function of Ech hydrogenase because the enzyme is produced in substantial amounts in methanol and H2/CO2-grown cells. It has been previously shown that H2 is an efficient electron donor for acetyl-CoA synthesis from methyl-tetrahydrosarcinapterin, CoA and CO2. For this reaction acetyl-CoA synthase, ferredoxin and a hydrogenase present in a high molecular mass fraction of a gel filtration column were necessary . The data presented here indicate that Ech is the hydrogenase involved in this reaction. At the H2 partial pressures found in the natural habitats (1–10 Pa) synthesis of acetyl-CoA becomes endergonic with H2 as electron donor. Reversed electron transport might be used to overcome this thermodynamic barrier. Therefore we propose that during growth of M. barkeri under autotrophic conditions Ech hydrogenase catalyzes the H2 dependent reduction of the ferredoxin driven by the proton-motive force. This is also supported by the finding of Bott and Thauer that the synthesis of CO from CO2 and H2, catalyzed by cell suspensions of M. barkeri and other methanogens, is driven by the proton-motive force .
We have recently shown, that in Methanobacterium thermoautotrophicum, a methanogen that can only utilize H2/CO2 as energy substrates, two gene clusters are expressed that encode enzymes similar to Ech hydrogenase from M. barkeri. For these enzymes a role in endergonic reductions, such as the synthesis of acetyl-CoA, has been proposed .
The members of the E. coli hydrogenase 3-type hydrogenases are related to energy conserving NADH:quinone oxidoreductase (complex I). E. coli complex I can be cleaved into three distinct fragments, the NADH-dehydrogenase fragment, the connecting fragment and the membrane fragment, which are thought to represent modules with distinct functions . The NADH-dehydrogenase fragment is formed by three subunits and mediates electron transfer from NADH to various artificial electron acceptors. This module is thought to function as the isopotential electron-input device and is thought to reduce the iron–sulfur clusters of the connecting fragment. Subunits with sequence similarity to subunits of this module are not found in Ech hydrogenase and related hydrogenases.
In contrast, the four subunits which form the connecting fragment of E. coli complex I have homologues in the membrane-bound hydrogenases. These are the four conserved hydrophilic subunits found in Ech hydrogenase and related hydrogenases. The connecting fragment of complex I harbors iron–sulfur cluster(s) N-2, which have a pH dependent midpoint redox potential and have been suggested to be part of a proton-pumping redox device [49,50].
Two of the seven subunits which form the membrane fragment of E. coli complex I are also conserved in the membrane-bound hydrogenases.
The data presented in this work indicate that the ferredoxin functions as the electron input-device of Ech hydrogenase and thus might have a similar function as the NADH-dehydrogenase module of complex I. Electron transfer proceeds most probably via the two [4Fe-4S] clusters of subunit EchF (the homologue of subunit NuoI of the E. coli complex I) and via the iron–sulfur cluster of subunit EchC (the homologue of NuoB) to the [NiFe] active site present on subunit EchE (the homologue of NuoD), where two H+ are reduced to H2. The enzyme can also operate in the reverse direction and can reduce the ferredoxin with H2 as electron donor.
For Ech hydrogenase a quinone can be excluded as electron acceptor and also the methanogenic coenzyme methanophenazine seems not to function as electron acceptor of this enzyme. This is probably the most pronounced difference to NADH: quinone oxidoreductase (complex I), where quinones function as the physiological electron acceptor. If Ech hydrogenase is a proton pump, as we propose, proton translocation cannot involve a quinone cycle mechanism. A conformational change in the hydrophilic part of the enzyme has to be coupled to vectorial translocation of H+ across the cytoplasmic membrane.
CO-induced hydrogenase from R. rubrum, which has been partially purified, catalyzes hydrogen formation with reducing equivalents generated in the oxidation of CO to CO2, catalyzed by carbon monoxide dehydrogenase. The electron carrier(s) mediating electron transfer between both enzymes have not yet been identified; probably also a ferredoxin is involved. Hydrogen formation is the physiological reaction catalyzed by this enzyme. This is also reflected by the kinetic properties of the enzyme because it catalyzes hydrogen formation at 22-fold higher rates than hydrogen uptake . There is no obvious function for a quinone to participate in this reaction or to function as electron acceptor of CO-induced hydrogenase.
E. coli hydrogenase 3 (Hyc) under physiological conditions also functions as hydrogen forming hydrogenase with reducing equivalents generated in the oxidation of formate to CO2. Cell suspensions of E. coli grown under fermenting conditions were also found to catalyze the reverse reaction, the formation of formate from CO2 and H2. The hyc operon encodes a 4[4Fe-4S] ferredoxin (HycB) which might mediate electron transfer between formate dehydrogenase and hydrogenase [13,14]. There is also no obvious function for a quinone to participate in this reaction or to function as electron acceptor of Hyc.
From the present data it may be concluded that the three hydrogenases, Ech, Hyc and Coo, catalyze related reactions. Reducing equivalents generated in the oxidation of a low potential electron donor (formate, carbon monoxide or the carbonyl group of acetate) are passed to the respective hydrogenase where protons are reduced to H2. Ech hydrogenase might also be involved in the catalysis of the reverse reaction, reduction of a low potential electron acceptor by H2. Ech hydrogenase, the only member of this hydrogenase family that has been purified, will be used in future studies for a characterization of the redox centers of this enzyme which might allow a more detailed comparison to complex I and to standard [NiFe] hydrogenases.
This work was supported by the Max-Planck-Gesellschaft, by the Deutsche Forschungsgemeinschaft, and by the Fonds der Chemischen Industrie.