TEXT

Top Abstract Text References

Hydrogenases are enzymes catalyzing both molecular hydrogen production and uptake: H₂  2H⁺ + 2e. Hydrogen metabolism plays a central role in the energy-generating mechanisms of various microorganisms. For the genus Desulfovibrio, hydrogen cycling with vectorial electron transfer has been proposed as a general energy-coupling mechanism when the cells are grown on organic substrates with sulfate (20). However, since the discovery and description by Pfennig and Biebl of the sulfur reducer Desulfuromonas acetoxidans in 1976, it has been accepted that this specific bacterium lacked hydrogenase activity (24).

D. acetoxidans, a true sulfur-reducing bacterium, is strictly anaerobic, gram negative, flagellated, and rod shaped. It acquires its energy from sulfur respiration and completely oxidizes acetate with S⁰ to carbon dioxide via the citric acid cycle. Reduction of S⁰ produces hydrogen sulfide (H₂S). Instead of sulfur, many strains can use the nonselective electron acceptors L-malate and fumarate, which are quantitatively reduced to succinate (24). D. acetoxidans is unable to ferment organic compounds.

Several redox proteins have been purified from D. acetoxidans. This bacterium, like the Desulfovibrio species, contains abundant c-type cytochromes that appear to be involved in the electron transport to elementary sulfur. The best-characterized cytochrome, cytochrome c₇, is highly homologous to the tetraheme cytochrome c₃ isolated from sulfate-reducing bacteria (6, 17). Recently, three other c-type cytochromes have been characterized (10, 22): monoheme cytochrome c (M_r, 10,000), the first with a bishistidinyl axial coordination and a low redox potential; and cytochrome c (M_r, 50,000) and cytochrome c (M_r, 65,000) containing, respectively, four and eight hemes covering a wide range of reduction potentials (10, 22).

In this report we demonstrate, for the first time, a hydrogenase activity in D. acetoxidans. The enzyme was purified and partially characterized as a [NiFe] hydrogenase. In some Desulfovibrio species, hydrogenase transfers electrons to multiheme cytochromes (4, 9, 15, 18, 19, 27), and a key role of the hydrogenase-cytochrome c₃ complex in the metabolic pathway has recently been proposed (5, 23). To determine the possible physiological function of the novel hydrogenase, reactivity with various multiheme D. acetoxidans cytochromes was tested.

D. acetoxidans DSM 1675 was grown at 30°C on malate-ethanol medium containing, per liter, 1 g of KH₂PO₄, 0.3 g of NH₄Cl, 1 g of MgSO₄ · 7H₂O, 2 g of MgCl₂ · 6H₂O, 20 g of NaCl, 0.1 g of CaCl₂ · 2H₂O, 1.77 g of Na₂SO₄, 2.66 g of disodium DL-malate, 0.5 g of yeast extract, 1.85 g of NaHCO₃, 0.3 g of Na₂S · 9H₂O, 0.3 ml of ethanol, and 10 ml of trace element solution (24). The bacteria were grown on a large scale in a 300-liter anaerobic fermenter, and cells were collected after 18 h (optical density at 600 nm of 0.13). To optimize hydrogenase production, cultures were also grown anaerobically in 20-liter flasks at 30°C in malate-ethanol medium. One-liter aliquots were taken at various times, and the cells were collected by centrifugation (15 min at 5,500 × g).

Except for the extraction step, all purification procedures were performed at pH 7.6 and 4°C with argon-saturated buffers. Hydrogenase activity (hydrogen consumption assay) was measured continually during the purification.

The cells, resuspended in 20 mM Tris-HCl-20 mM NaCl-5 mM EDTA-0.1% Zwittergent 3-12-0.5 mM phenylmethylsulfonyl fluoride, were broken in a French press and centrifuged at 200,000 × g for 1 h. The supernatant was applied to a DEAE-52 column (5 by 20 cm; Whatman) equilibrated with 20 mM Tris-HCl-20 mM NaCl-5 mM EDTA, and a linear gradient of NaCl (20 to 500 mM) in the same buffer was applied. The hydrogenase fraction, eluted between 100 and 200 mM NaCl, was subsequently loaded onto a hydroxylapatite Bio-Gel-HTP (1.5 by 10 cm) column equilibrated in 20 mM Tris-HCl-200 mM NaCl. The hydrogenase fraction was eluted at 200 mM phosphate, dialyzed overnight against 20 mM Tris-HCl, then loaded onto a Q-Sepharose column (fast protein liquid chromatography [FPLC]) equilibrated in the same buffer, and eluted with a gradient of NaCl (0 to 500 mM) (flow rate, 2 ml/min). The column yielded hydrogenase preparation, which was further characterized.

After being activated with sodium dithionite under H₂ for 20 min, hydrogenase activity was routinely determined at 25°C by the hydrogen consumption assay with methyl viologen as the electron acceptor and hydrogen as the electron donor (9). Hydrogen production from dithionite-reduced methyl viologen was also monitored spectrophotometrically. The anaerobic cuvettes, filled with the oxygen-scavenging system, 50 µM methyl viologen in 50 mM HEPES (pH 7), were flushed for 10 min with argon before addition of 75 µM sodium dithionite. Oxidation of methyl viologen was monitored at 604 nm after the enzyme was added. One unit of hydrogenase activity is the amount of enzyme that catalyzes the reduction of 1 µmol of methyl viologen per min, in the presence of an excess of hydrogen, or the production of 1 µmol of H₂/min.

Cytochrome c₇ and cytochrome c (M_r, 50,000) from D. acetoxidans were purified as previously reported (10, 25). For cytochrome reduction by hydrogenase, anaerobic cuvettes were filled with the oxygen-scavenging system and either cytochrome c₇ or cytochrome c (M_r, 50,000) at a concentration of 5 µM in 50 mM HEPES (pH 7), and the cytochrome reduction rate was measured (9).

Electrophoreses were carried out under native or reducing conditions on a Pharmacia Phast system with 12% polyacrylamide Phast gels and native or sodium dodecylsulfate (SDS) buffer strips. Hydrogenase activity was revealed on the gel after activation by sodium dithionite (1). After electrophoresis under denaturing conditions, the proteins were transferred onto a nitrocellulose membrane (Pharmacia). Immunodetection was performed with a rabbit anti-Desulfovibrio fructosovorans [NiFe] hydrogenase antiserum (26).

The N-terminal amino acid sequences of the hydrogenase subunits were determined from the enzyme preparation after the subunits were separated by SDS-polyacrylamide gel electrophoresis (PAGE) (9).

Electronic paramagnetic resonance (EPR) spectra were recorded at cryogenic temperatures with a Bruker ESP 300e X-band spectrometer fitted with an Oxford Instrument He-cryostat and temperature control system.

D. acetoxidans can grow on a medium containing malate as the electron acceptor and ethanol as both carbon and energy sources. Cell yields obtained with this medium are about fourfold higher than with sulfur as the electron acceptor (24). To determine the growth phase yielding the highest levels of hydrogenase, D. acetoxidans was grown anaerobically in 20-liter flasks, and samples were taken at various time intervals. Periplasmic proteins were extracted, and the maximal level of the enzyme was found at the beginning of the exponential phase (Fig. 1). After 24 h of growth, no hydrogenase activity was detected.

View larger version (15K): [in this window] [in a new window]   FIG. 1.   Growth curve of D. acetoxidans in DSM medium () and activity of hydrogenase from D. acetoxidans in DSM medium (). OD, optical density.

The purified enzyme had a UV-visible absorption spectrum typical of an iron-sulfur protein with an A₃₉₀/A₂₈₀ of 0.22. The yield was 11% (Table 1).

The D. acetoxidans enzyme cross-reacted with the antibodies raised against the purified [NiFe] hydrogenase from D. fructosovorans (data not shown). On the other hand, the [Fe] hydrogenase from Desulfovibrio vulgaris Hildenborough did not cross-react with the same antibodies in the same conditions.

Only one band was observed when the enriched hydrogenase sample was run on native PAGE and revealed with activity stain (Fig. 2A). Two bands of 66 and 27.5 kDa were revealed by SDS-PAGE (Fig. 2B) and Western blotting (Fig. 2C).

View larger version (23K): [in this window] [in a new window]   FIG. 2.   Visualization of hydrogenase of D. acetoxidans enzyme by native PAGE (A), SDS-PAGE (B), and Western blotting (C). (A) Lane 1, activity staining of purified hydrogenase from D. norvegicum; lane 2, activity staining of partly purified hydrogenase from D. acetoxidans. (B) Lane 1, partly purified hydrogenase from D. acetoxidans in denaturing conditions; lane 2, molecular weight standard protein in denaturing conditions. (C) Lane 1, Western blotting of the [NiFe] hydrogenase from D. fructosovorans; lane 2, partly purified hydrogenase extract from D. acetoxidans with antibodies against [NiFe] hydrogenase from D. fructosovorans.

The N-terminal sequences of the two subunits were determined after transfer on a Glassybond membrane. The sequences for the large and the small subunits are RRTA(X)IERRTAETTIDLKLDLDG and DINGGGATLPQKLYLTPNVLTAGFAPYIGVG, respectively. The two subunits showed no clear homologies with N-terminal sequences of previously described hydrogenases.

The hydrogenase from D. acetoxidans was active in the assay of H₂ uptake. A methyl viologen reduction by activated enzyme was observed only in the presence of hydrogen. When the enzyme was tested without prior activation, the lag period and the reduction phase appeared considerably longer, indicating reactivation under reducing conditions (at least 60 min for the lag phase). The enzyme was also active in the H₂ evolution assay with dithionite and methyl viologen as donors but at a level about threefold lower, and the lag phase was absent. A slight inhibition of the methyl viologen reduction was observed when 4 µM CO was added to the reaction mixture; 63% of the initial activity remained. [Fe], [NiFe], and [NiFeSe] hydrogenases from D. vulgaris Hildenborough, D. fructosovorans, and D. norvegicum, respectively, were also tested for CO inhibition, and the fractions of residual activity were 10.7, 67, and 85%, respectively.

As described above, only insignificant quantities of hydrogenase were obtained unless the starting material (membranes) was treated with the detergent Zwittergent. Moreover, we detected a hydrogenase in the membrane fraction with anti-D. fructosovorans hydrogenase antibodies (data not shown), indicating that a hydrogenase is associated or at least displays a high affinity for the cytoplasmic membrane. This observation was corroborated by EPR spectra on sedimented (i.e., highly concentrated) membrane fragments not treated with detergent. EPR spectra in different reduction states showed the presence of both cytochromes and typical signals from Ni centers in [NiFe] hydrogenases at g = 2.39, 2.35, 2.21, and 2.17 (3, 29), as well as signals from reduced Fe/S centers (g_y = 1.92 and 1.94; g_x = 1.89 and 1.90).

The reactivity of D. acetoxidans hydrogenase with two periplasmic multiheme cytochromes from the same bacterium was investigated. Figure 3 presents the time course of reduction of these cytochromes by the enzyme preparation. The reduction of cytochrome c (M_r, 50,000) was inefficient (0.6 µmol of reduced cytochrome/min/mg of protein). Cytochrome c₇ reduction was more efficient, with a rate of 30 µmol of reduced cytochrome/min/mg of protein, in the same conditions.

View larger version (13K): [in this window] [in a new window]   FIG. 3.   Reduction of polyheme cytochromes from D. acetoxidans by hydrogenase from D. acetoxidans. The enzyme was preactivated with hydrogen and dithionite. , Reduction of 5 µM cytochrome c7 from D. acetoxidans; -----, reduction of 5 µM cytochrome c (Mr, 50,000) from D. acetoxidans.

It has been proposed that hydrogenases play a key role in the energetic pathway in various bacteria (reference 2 and references therein), but no hydrogenase has been found in Desulfuromonas species. Here we report, for the first time, the purification and characterization of a hydrogenase from the sulfur-reducing bacterium D. acetoxidans.

We examined the expression of the hydrogenase in cells grown on malate-ethanol medium. In contrast to the results reported for the [NiFe] hydrogenase (28) of various Desulfovibrio strains, the activity level of hydrogenase from D. acetoxidans peaks during the exponential growth phase.

Similar to typical [NiFe] hydrogenases, the enzyme from D. acetoxidans has two subunits with apparent molecular masses of 66 and 27.5 kDa (11), an activation under reducing conditions for at least 20 min is required to obtain an active enzyme (11), and only a slight inhibition by 4 µM carbon monoxide was observed (7). Moreover, the D. acetoxidans enzyme cross-reacted with the anti-[NiFe] hydrogenase antibodies from D. fructosovorans, while the [Fe] hydrogenase from D. vulgaris Hildenborough was not revealed by the same antibodies. A hydrogenase from D. acetoxidans was detected in the membrane fraction by EPR and Western blotting. The evidenced hydrogenase may be associated with the periplasmic face of the membrane, but we cannot exclude the presence of another membrane-bound hydrogenase.

Several attempts were made to purify the enzyme to homogeneity, but the low detectable amounts and/or the high sensitivity to O₂ precluded an extensive purification. The low amount and stability of the enzyme may explain the recovery loss during the last purification step, which is nonetheless essential.

In Desulfovibrio species, soluble and membrane-bound hydrogenases catalyze the reversible hydrogen oxidation with low-potential multiheme c-type cytochromes, in particular cytochrome c₃ (M_r, 13,000). The striking homologies between the tetraheme cytochrome c₃ (M_r, 13,000) and the triheme cytochrome c₇ from D. acetoxidans suggest that the two cytochromes may be functionally similar. The cytochrome c₇ was indeed reduced by [NiFe] hydrogenase from D. acetoxidans. In vitro, in the presence of D. vulgaris Hildenborough [Fe] hydrogenase, the cytochrome c₇ has a polysulfide reductase activity (22). It has been proposed that this cytochrome is the terminal reductase, although the physiological electron donor remains to be identified. The hydrogenase that we characterized may be the physiological electron donor for the cytochrome. In addition to cytochrome c₇, the 50-kDa cytochrome c and the 65-kDa cytochrome c (10, 22) are present in this bacterium. Under our conditions, the reduction of the 50-kDa cytochrome c by the hydrogenase from D. acetoxidans is very inefficient in comparison to the reduction of cytochrome c₇ and indicates another role for the cytochrome. Further studies will determine whether the cytochrome c (M_r, 10,000), the cytochrome c (M_r, 50,000), and the cytochrome c (M_r, 65,000) could be physiological partners of hydrogenases. As in the case in the sulfate-reducing bacteria, the role in D. acetoxidans of several c-type multiheme cytochromes with low redox potential is not understood.

Nonetheless, numerous bacteria with a hydrogenase activity do not possess multiheme cytochromes and can reduce other redox proteins. Multiheme cytochromes as obligatory electron acceptors from hydrogenases are specific to the genus Desulfovibrio. We thus hesitate to exclude electron acceptors other than the multiheme cytochromes for electrons resulting from oxidation of molecular hydrogen by hydrogenase from D. acetoxidans. There are several examples of electron transfer from a membrane-associated hydrogenase to a b-type cytochrome (8, 12, 13, 16). The sulfur-reducing microaerophilic bacterium Spirillum strain 5175 can grow anaerobically with sulfur as the electron acceptor as D. acetoxidans and several Campylobacter species do, and it contains two cytochromes b and a [NiFe] hydrogenase in the membrane (29). These bacteria do not possess multiheme cytochromes. It has been concluded that the b-type cytochromes are involved in the electron transfer from hydrogenase to H₂S. Under certain growth conditions, the presence of a menaquinone and cytochrome b has been demonstrated in D. acetoxidans (14, 21, 24). We have not detected any b-type subunit during purification of the hydrogenase, but experiments are in progress in our laboratory to determine whether the hydrogenase from D. acetoxidans reduces endogenous cytochrome b and to clone the D. acetoxidans hydrogenase gene. The cloning will indicate whether the hydrogenase gene is in an operon with the gene of a b-type cytochrome, as is the case for several membrane-bound [NiFe] hydrogenases. Further physiological, genetic, and structural studies on the multiheme cytochromes c, the b-type cytochrome(s), and the hydrogenase are required to elucidate the electron transport pathways involving the hydrogenase and its role in the metabolism of this sulfur-reducing bacterium, which had been previously described as not using H₂ as an electron donor.