Fax: + 49 6421 178299, Tel.: + 49 6421 178230, E-mail: hedderic@staff.uni-marburg.de Abbreviations: HdrABC, soluble flavin–iron–sulfur heterodisulfide reductase from Methanothermobacter spp
Trang 1Two distinct heterodisulfide reductase-like enzymes
Gerd J Mander1, Antonio J Pierik2, Harald Huber3and Reiner Hedderich1
1
Max-Planck-Institut for Terrestrial Microbiology, Marburg, Germany;2Laboratory for Microbiology, Department of Biology, Philipps University Marburg, Germany;3Department of Microbiology and Archaeenzentrum, University of Regensburg, Germany
Heterodisulfide reductase (Hdr) is a unique disulfide
reduc-tase that plays a key role in the energy metabolism of
methanogenic archaea Two types of Hdr have been
identi-fied and characterized from distantly related methanogens
Here we show that the sulfate-reducing archaeon
Archaeo-globus profundus cultivated on H2/sulfate forms enzymes
related to both types of Hdr From the membrane fraction of
A profundus, a two-subunit enzyme (HmeCD) composed of
a b-type cytochrome and a hydrophilic iron–sulfur protein
was isolated The amino-terminal sequences of these
sub-units revealed high sequence identities to subsub-units HmeC
and HmeD of the Hme complex from A fulgidus HmeC
and HmeD in turn are closely related to subunits HdrE and
HdrD of Hdr from Methanosarcina spp From the soluble
fraction of A profundus a six-subunit enzyme complex
(Mvh:Hdl) containing Ni, iron–sulfur clusters and FAD was isolated Via amino-terminal sequencing, the encoding genes were identified in the genome of the closely related species
A fulgidusin which these genes are clustered They encode a three-subunit [NiFe] hydrogenase with high sequence iden-tity to the F420-nonreducing hydrogenase from Methano-thermobacterspp while the remaining three polypeptides are related to the three-subunit heterodisulfide reductase from Methanothermobacterspp The oxidized enzyme exhibited
an unusual EPR spectrum with gxyz¼ 2.014, 1.939 and 1.895 similar to that observed for oxidized Hme and Hdr Upon reduction with H2this signal was no longer detectable Keywords: Archaeoglobus; heterodisulfide reductase; Hmc complex; iron-sulfur proteins; sulfate-reducing bacteria
Heterodisulfide reductase (Hdr) is a unique disulfide
reduc-tase, which has a key function in the energy metabolism of
methanogenic archaea The enzyme catalyses the reversible
reduction of the mixed disulfide (CoM–S–S–CoB) of the
two methanogenic thiol-coenzymes, called coenzyme M
(CoM-SH) and coenzyme B (CoB-SH) This disulfide is
generated in the final step of methanogenesis [1] Two types
of Hdr have been identified and characterized from distantly
related methanogens [2–6]
One type of Hdr, which was purified and characterized
from Methanothermobacter marburgensis, is a soluble iron–
sulfur flavoprotein composed of the three subunits HdrA,
HdrB and HdrC [2,3] For clarity this enzyme will be called HdrABC throughout this paper From sequence data it has been deduced that HdrA contains an FAD-binding motif and four binding motifs for [4Fe)4S] clusters HdrC was shown to contain two binding motifs for [4Fe)4S] clusters while in subunit HdrB no characteristic binding motif of any known cofactor could be identified However, this subunit contains 10 highly conserved cysteine residues present in two Cx31)38CCx33)34Cx2C motifs
The second type of Hdr, designated as HdrDE, is found
in Methanosarcina species [4,6] This enzyme is tightly membrane bound It is composed of two subunits, a mem-brane anchoring b-type cytochrome (HdrE) and a hydro-philic iron–sulfur protein (HdrD) The amino-terminal part of HdrD contains two characteristic binding motifs for [4Fe)4S] clusters also conserved in subunit HdrC of the Mt marburgensis enzyme The carboxy-terminal part
of HdrD harbours the two Cx31)38CCx33)34Cx2C motifs also present in HdrB Subunit HdrD of the Methanosarcina enzyme can be regarded as a hypothetical fusion protein of subunits HdrC and HdrB of Mt marburgensis Hdr [5] The catalytic centre must be located on Ms barkeri HdrD and Mt marburgensis HdrCB, which are conserved
in both enzymes [3,5] A detailed spectroscopic character-ization showed that the active site harbours a [4Fe)4S] cluster [7,8], which is most probably coordinated by some of the cysteine residues present in the Cx31)38CCx33)34Cx2C motifs With both HdrABC and HdrDE a reaction intermediate is trapped when only coenzyme M is added
to the oxidized enzyme (in the absence of coenzyme B) It
is characterized by a unique S¼ 1/2 EPR spectrum with
Correspondence to R Hedderich, Max-Planck-Institute for Terrestrial
Microbiology, Karl-von-Frisch Str., D-35043 Marburg, Germany.
Fax: + 49 6421 178299, Tel.: + 49 6421 178230,
E-mail: hedderic@staff.uni-marburg.de
Abbreviations: HdrABC, soluble flavin–iron–sulfur heterodisulfide
reductase from Methanothermobacter spp.; HdrDE, heme-containing
membrane-bound heterodisulfide reductase from Methanosarcina
spp.; Hdl, HdrABC-like enzyme from Archaeoglobus spp.; Hme,
HdrDE-like menaquinol-oxidizing enzyme from Archaeoglobus spp.;
Mvh, F 420 -nonreducing hydrogenase (methylviologen-reducing
hydrogenase) from Methanothermobacter spp.; Mvh:Hdl, F 420
-nonreducing hydrogenase:heterodisulfide reductase-like enzyme
complex; APS, adenosine 5¢-phosphosulfate; DMN,
2,3-dimethyl-1,4-naphtoquinone.
Enzyme: heterodisulfide reductase (EC 1.99.4.-).
(Received 24 September 2003, revised 10 December 2003,
accepted 26 January 2004)
Trang 2principal g values¼ 2.013, 1.991 and 1.938 for HdrDE
and gxyz¼ 2.011, 1.993, 1.944 for HdrABC observable at
temperatures below50 K [7] In this paramagnetic species,
which was designated as CoM–Hdr, coenzyme M was
shown to be directly bound to the cluster via its thiol group
[9] Hence, the active site iron–sulfur cluster is directly
involved in the disulfide cleavage reaction
The two types of Hdr differ with respect to their
physiological electron donor HdrDE receives reducing
equivalents from the reduced methanophenazine pool via
its b-type cytochrome subunit The enzyme is part of an
energy-conserving membrane-bound electron transport
chain w ith H2or reduced coenzyme F420as electron donor
and the heterodisulfide as terminal electron acceptor [10,11]
HdrABC forms a tight complex with the F420-nonreducing
hydrogenase (Mvh) This six-subunit complex catalyses the
reduction of the heterodisulfide by H2 After cell lysis, this
complex is almost completely localized in the soluble
fraction It is yet unknown how the exergonic reduction of
the heterodisulfide is coupled to energy conservation in
Mt marburgensis[12]
An interesting result obtained from the analysis of the
genome sequence of the sulfate-reducing archaeon
Archaeo-globus fulgiduswas the presence of several genes encoding
enzymes closely related to heterodisulfide reductase from
methanogens [13] The isolation of one of these enzymes,
called Hme, has recently been reported [14] Hme, when
purified, is composed of four subunits HmeACDE The
encoding gene cluster predicts the presence of a fifth subunit
(HmeB) One of the Hme subunits (HmeC) is a b-type
cytochrome, a second subunit (HmeD) is closely related to
subunit HdrD of Ms barkeri Hdr HmeD also contains one
copy of the Cx31)38CCx33)34Cx2C motif, which was found
to be characteristic for Hdr However, in HmeD this
cysteine-rich motif is composed of only four cysteine
residues, an aspartate residue replaces the last cysteine
residue
Oxidized Hme exhibited an unusual EPR spectrum with
g-values at 2.031, 1.995, and 1.951 The paramagnetic
species could be reduced in a one-electron transfer reaction,
but could not be oxidized further It thus shows EPR and
redox properties similar to a paramagnetic species formed
when the active-site iron–sulfur cluster of Hdr from
Ms barkeri or Mt marburgensis binds one of its thiol
substrates as an extra ligand during the catalytic cycle [9]
Based on the spectroscopic properties of Hme and based
on the presence of the Cx31)38CCx33)34Cx2C/D motif in
A fulgidus HmeD, this subunit was proposed to have a
catalytic site similar to that of Hdr [14]
Enzymes related to Hme have also been identified in
the sulfate-reducing bacterium Desulfovibrio vulgaris, the
green sulfur bacterium Chlorobium tepidum and the purple
sulfur bacterium Allochromatium vinosum [15–18] One of
the major open questions in understanding the energy
metabolic pathways of sulfate reducing bacteria and
archaea concerns the path of reducing equivalents
gener-ated in the oxidative branch of the metabolic pathway to
the enzymes of sulfate reduction This electron transfer
process is thought to be coupled with energy conservation
The A fulgidus Hme protein has been proposed to
participate in this electron transfer reaction Evidence
has been provided that this enzyme functions as a
menaquinol-acceptor oxidoreductase mediating the elec-tron transfer from the quinone pool to a yet unidentified electron carrier in the cytoplasm which in turn could function as an electron donor of the enzymes of sulfate reduction, adenosine 5¢-phosphosulfate (APS) reductase and sulfite reductase [14]
In this communication we address the question whether Hme or one of its homologues is also involved in sulfate reduction when H2 is the electron donor Although
A fulgidus has been reported to grow with H2 as sole electron donor, growth under these conditions is very poor Lactate-grown A fulgidus cells do not exhibit hydrogenase activity [19] Therefore, the hydrogenotrophic Archaeo-globusspecies, A profundus, was used in this study
Materials and methods
Materials Unless otherwise stated, chemicals were from Merck (Darmstadt, Germany) and chromatographic materials and columns were from Amersham Biosciences
Organism growth
A profundus(DSMZ 5631) was grown in a 300-L fermenter
at 85C as described previously [20] Cells were harvested after shock cooling to 4C in a continuous flowcentrifuge (Z61, Padberg Lahr, Germany) at 17 000 g; the pellet was frozen in liquid nitrogen and stored at)80 C prior to use Purification of HmeCD
All purification steps were carried out under strictly anaerobic conditions under an atmosphere of N2/H2 (95 : 5; v/v) at 18C Cells were lysed by sonication and then centrifuged at 6400 g for 1 h The supernatant was ultracentrifuged at 150 000 g for 2 h The pellet was resuspended in 50 mM Mops/KOH, pH 7.0 (buffer A) using a Teflon homogenizer Protein was solubilized from the membrane w ith 15 mM dodecyl-b-D-maltoside (2 mg dodecyl-b-D-maltosideÆmg)1 protein) at 4C for 12 h Proteins not solubilized after 12 h were removed by ultracentrifugation as described above Solubilized protein was loaded to a Q-Sepharose column (2.6· 10 cm) equil-ibrated with buffer A containing 2 mMdodecyl-b-D -malto-side (buffer A1) Protein was eluted in a stepwise NaCl gradient (80 mL each in buffer A1): 0 mM, 300 mM,
400 mM, 500 mM, 600 mM and 1M The fractions were checked for their heme-content by UV/visible spectroscopy The majority of the heme-containing proteins eluted at
600 mMNaCl These fractions were applied to a Superdex
200 gel-filtration column (2.6· 60 cm) equilibrated in buffer A1 with 100 mMNaCl Protein was eluted with the same buffer The only heme-containing fraction eluted after
180 mL (peak maximum) These fractions were loaded on a MonoQ column (1.0· 10 cm) equilibrated with buffer A1 Protein was eluted using a linear NaCl gradient (0–1M,
100 mL) Heme-containing protein(s) eluted at 600 mM
NaCl These fractions were pooled and concentrated by ultrafiltration (100-kDa cut off, Molecular/Por ultrafiltra-tion membranes, Houston) and stored in buffer A1 at 4C
Trang 3under N2 Protein was judged to be > 95% pure by SDS/
PAGE
Purification of the Mvh:Hdl enzyme complex
fromA profundus
All purification steps were performed as described above for
the purification of HmeCD The 150 000 g supernatant
of cell-free extracts was applied to a Q-Sepharose
(2.6· 10 cm) anion exchange column equilibrated with
buffer A Protein was eluted in a stepwise NaCl gradient in
buffer A (see above) The majority of the hydrogenase
activity eluted at 400 mMNaCl (Table 2) These fractions
were pooled, the buffer was changed to 10 mM
Na-phos-phate buffer pH 7.0 by ultrafiltration (50-kDa cut-off)
and protein was loaded on a hydroxyapatite column
(1.6· 10 cm) equilibrated with 10 mMNa-phosphate
buf-fer pH 7.0 Protein was eluted in a linear Na-phosphate
gradient (10 mM to 1M, 350 mL) The majority of the
hydrogenase activity eluted at 150–180 mMNa-phosphate
These fractions were pooled and the buffer was changed to
buffer A by ultrafiltration The resulting fraction was loaded
on a MonoQ column (1.0· 10 cm) Protein was eluted in a
linear NaCl gradient in buffer A (0–700 mM, 150 mL) The
majority of the hydrogenase activity eluted at 400 mM
NaCl These fractions were concentrated by ultrafiltration
and stored in buffer A at 4C under N2
Determination of enzyme activities
Enzyme assays were routinely carried out under anoxic
conditions in 1.5-mL quartz cuvettes at 65C One unit of
enzyme activity corresponds to 1 lmol H2consumedÆmin)1
Hydrogen uptake activity with benzylviologen as electron
acceptor was determined by following the reduction of
benzylviologen at 578 nm (e¼ 8.6 mM )1Æcm)1) The
0.8-mL assays contained 2 mM benzylviologen and 0.1 mM
sodium dithionite in 50 mMMops/KOH pH 7.0) One unit
of H2-oxidation activity is defined as the reduction of
2 lmol benzylviologenÆmin)1
UV/visible spectroscopy
Spectra of samples in 1.5-mL Quartz cuvettes in an
anaerobic chamber under N2/H2(95/5, v/v) were recorded
using a Zeiss Specord S10 diode array spectrophotometer
connected to a quartz photoconductor (Hellma Mu¨hlheim,
Germany) Sodium dithionite was added to an enzyme
solution (0.7 mg proteinÆmL)1 in buffer A) to obtain
the spectrum of the fully reduced enzyme The spectrum
of the oxidized enzyme was obtained after oxidation by air
The oxidation of the heme groups by DMN
(2,3-dimethyl-1,4-naphtoquinone) was followed
spectrophotometri-cally DMN was added to the enzyme solution (1 mg
proteinÆmL)1in 50 mMMops/KOH pH 7.0), 2 mM
dode-cyl-b-D-maltoside) to a final concentration of 150 lM and
spectra were recorded every 5 s
EPR spectroscopy measurements
EPR spectra at X-band (9.45 GHz) were obtained with a
Bruker EMX spectrometer All spectra were recorded with
a field modulation frequency of 100 kHz and a modulation amplitude of 0.6 mT The sample was cooled by an Oxford Instrument ESR 900 flowcryostat with an ITC4 tempera-ture controller Spin quantifications were carried out under nonsaturating conditions using copper perchlorate as standard (10 mMCuSO4, 2MNaClO4, 10 mMHCl) When EPR signals overlapped with other signals, e.g radical signals from flavins, the signals were simulated, and the simulations were double integrated to obtain the spin intensity Temperature dependence studies were carried out under nonsaturating conditions where possible For all signals, the peak amplitude was measured at different temperatures These values were used to obtain Curie plots describing the temperature behaviour of the respective signal EPR signals were simulated using noncommercial programs supplied by S.P.J Albracht based on formulas described previously [21]
Determination of amino-acid sequences For determination of amino-terminal amino acid sequences, polypeptides were separated by SDS/PAGE and blotted on to poly(vinylidene difluoride) membranes (Applied Biosystems) as described previously [5] Sequences were determined using an Applied Biosystems 4774 protein/peptide sequencer and the protocol given by the manufacturer
Analytical methods Iron was quantified colorimetrically with neocuproin (2,9-dimethyl-1,10-phenanthroline) and ferrozine[3-(2-pyr-idyl)-5,6-bis-(4-phenylsulfonate)-1,2,4-triazine] as described previously [22] Acid labile sulfur was analysed with the methylene blue method [23]
Protein concentration was routinely measured by the method of Bradford (Rotinanoquant; Roth, Karlsruhe, Germany) using BSA as standard [24]
Nickel was determined by atomic absorption scopy on a 3030 Perkin Elmer atomic absorption spectro-meter fitted with a HGA-600 graphite furnace assembly and
an AS-60 autosampler
For identification of the flavin and determination of the flavin content of the Mvh:Hdl complex, protein (200 lL, 8.9 mgÆmL)1) was denatured by exposure to 10% (m/v) trichloroacetic acid Denatured protein was removed by centrifugation, the resulting supernatant was adjusted to
pH 7 with K2HPO4and analysed by chromatography using
a reverse-phase HPLC column (LiChrospher 60 RP 18,
5 lm, 125· 4 mm, Merck, Germany) equilibrated with
50 mM ammonium formate containing 25% methanol Flavins were eluted isocratically with the equilibration buffer FAD and FMN standards were used to identify and quantify the flavin
Hemes were characterized by their pyridine hemochrome spectra [25] Protein (500 lL, 2 mgÆmL)1) was mixed with
500 lL of a stock solution of 200 mMNaOH in 40% (v/v) pyridine/H2O and 3 lL of 0.1MK3Fe(CN)6 in a 1.5-mL cuvette to record the oxidized spectrum Solid sodium dithionite was then added (2–5 mg) and several successive spectra of the reduced pyridine hemochromes were recorded
Trang 4Purification of a heme-containing protein
from the membrane fraction ofA profundus
To purify heme-containing enzymes possibly related to the
Hme complex from A fulgidus, the characteristic UV/
visible spectrum of heme proteins was followed throughout
the purification This resulted in the isolation of the major
heme-containing protein from the membrane fraction of
A profunduscells cultivated on H2/sulfate For analysis of
the protein by SDS/PAGE samples were either boiled for
5 min in SDS sample buffer or incubated in SDS sample
buffer for 30 min at room temperature (Fig 1) The sample
incubated at room temperature yielded two major
polypep-tide bands with apparent molecular masses of 53 kDa and
32 kDa (Fig 1; lane 1) In the boiled sample, the 32-kDa
polypeptide was undetectable in Coomassie Brilliant
blue-stained gels (Fig 1, lane 2) This could be due to
aggregation of the protein upon boiling, which is frequently observed for integral membrane proteins From 5 g of wet cell mass ( 500 mg protein) 5 mg of purified enzyme were obtained
Determination of the amino-terminal sequences and identification of homologous genes inA fulgidus The amino-terminal sequences of the two polypeptides were determined by Edman degradation (Table 1) Using these sequences, the genome of A fulgidus was searched for corresponding genes [13] The amino-terminal sequence of the 53-kDa polypeptide shows 45% sequence identity to the gene product of AF502, the amino-terminal sequence of the 32-kDa polypeptide shows 50% sequence identity to the polypeptide encoded by AF501 (Table 1) In A fulgidus both gene products are part of the Hme complex, which has recently been described [14] AF501 (HmeC) and AF502 (HmeD) were shown to share sequence identity with the two subunits HdrE and HdrD of Hdr from Methanosarcina species Based on its similarity to subunits of the A fulgidus Hme complex the A profundus enzyme was designated as HmeCD
Characterization of the heme groups by UV/visible and EPR spectroscopy
The enzyme purified under anaerobic conditions generally contained the heme groups in the reduced state Fig 2 shows the dithionite-reduced minus air-oxidized absorbance difference spectrum The absorbance maxima at 426 nm (c band), 530 nm (b band) and a split a band at 557 nm and
562 nm are characteristic for hemes of the b-type [26] A similar splitting of the a band has been observed for other heme proteins, for example the cytochrome bL of the cytochrome bc1complex form Rhodopseudomonas sphaero-ides GA [27] Pyridine hemochrome reduced–oxidized difference spectra showmaxima for the a and b band at
553 and 521 nm These values are blue-shifted by 4 nm relative to the published values for protoheme IX [28] The same result was obtained for the heme present in Hme from
A fulgidus[14] This suggests that Hme in both organisms contains a modified protoheme IX as prosthetic group Addition of DMN to the reduced enzyme resulted in a rapid oxidation of the heme groups present in the enzyme The rates were too rapid to be resolved
In oxidized HmeCD and at temperatures below10 K
a sharp absorption-shaped signal with g-values at 6.06 and 5.83 characteristic for ferric high-spin (S¼ 5/2,
Fig 1 SDS/PAGE of purified HmeCD Proteins were separated in a
14% slab gel (8 · 7 cm), which was subsequently stained with
Coo-massie Brilliant Blue R250 The molecular masses of marker proteins
are given on the right side, the apparent molecular masses of the
polypeptides in lanes 1 and 2 are given on the left side M,
Low-molecular-mass marker (Amersham Biosciences) Lane 1, 10 lg
A profundus HmeCD denatured for 30 min at room temperature in
SDS sample buffer (Laemmli buffer containing 5 m M dithiothreitol
and 2% SDS); lane 2, 10 lg Hme complex denatured for 5 min at
100 C in SDS sample buffer The polypeptide with an apparent
molecular mass of 32 kDa, identified as a b-type cytochrome by
N-terminal sequencing, is nondetectable in the boiled sample; it
probably forms aggregates that do not run into the gel (lane 2) This
behaviour is typical for integral membrane proteins.
Table 1 Amino-terminal sequences of the membrane-bound heme containing enzyme of A profundus Amino-terminal sequences of the A profundus enzyme were derived by Edman degradation, amino-terminal sequences of A fulgidus were derived from the genome sequence [13] In both sequences, identical amino acids are underlined X, No clear assignment to an amino acid could be made; –, gaps inserted to allowan alignment.
Amino-terminal sequences
Sequence identity (%)
Corresponding gene
A fulgidus
Trang 5E/D < 0.01) heme was observed as described previously
for Hdr from M thermophila [6] The third g-value (g 2
region) could not observed due to the presence of other
signals (see below) The oxidized enzyme at pH 7 showed a
additional signal with g-values at 2.87 and 2.28 which is
characteristic for low-spin (S¼ 1/2) ferric heme [29]
Oxi-dized HmeCD also showed an intense signal at a g-value of
4.3 characteristic for adventitiously bound high spin ferric
iron
Characterization of the iron-sulfur clusters of HmeCD
by EPR spectroscopy
HmeCD was shown to contain 107 ± 3 nmol nonheme
iron and 117 ± 3 nmol acid-labile sulfurÆmg)1 protein
From the SDS/PAGE an apparent molecular mass of the
enzyme of 85 kDa was calculated, this corresponds to
9.2 ± 0.2 mol nonheme iron per mol enzyme and
10 ± 0.3 mol acid-labile sulfur per mol enzyme indicating
the presence of two to three [4Fe)4S] clusters These clusters
were further characterized by EPR spectroscopy The
sodium dithionite reduced enzyme showed at temperatures
below15 K a broad featured spectrum around g¼ 1.93
indicative of spin–spin coupling between different
[4Fe)4S]1+clusters
In ferricyanide or duroquinone oxidized samples a
paramagnetic species was detected with gxyz values at
2.031, 1.995 and 1.948 (Fig 3) The total spin concentration
of this species was 2.8 lM, corresponding to 0.15 spinÆmol)1
enzyme This signal was detectable without significant
broadening from 15 to 35 K At temperatures below15 K
the signal was readily power saturated and at temperatures
higher than 35 K the signal started to broaden and was
broadened beyond detection at 60 K The EPR properties
of this paramagnetic species are very similar to that of the
paramagnetic species recently described for the oxidized
A fulgidusHme complex [14]
Purification of a six-subunit [NiFe] hydrogenase from the soluble fraction ofA profundus Starting from cell-free extracts of A profundus, hydro-genase was purified by following the hydrogen uptake activity using benzylviologen as artificial electron acceptor The majority (97%) of the hydrogenase activity was found
in the soluble fraction (Table 2) Further purification resulted in an enzyme preparation consisting of six major polypeptides with apparent molecular masses of 72, 50, 35,
31, 22 and 15 kDa (Fig 4) It exhibited a specific hydrogen uptake activity of 420 UÆmg)1 protein at
65C From 5 g wet cells ( 500 mg protein) 12 mg of the purified enzyme were obtained The amino-terminal sequences of the six polypeptides showed highest sequence identity to proteins encoded by the A fulgidus genome (Table 3) [13] These genes are organized in a putative transcriptional unit (AF1377 to AF1372) (Fig 5) Only the amino-terminal sequence of the AF1376 gene product
Fig 2 Room temperature reduced/oxidized difference spectrum of the
purified HmeCD from A profundus The spectrum of the reduced
enzyme was recorded after reduction of Hme (0.7 mg protein per mL
in 50 m M Mops/KOH, pH 7.0) with sodium dithionite The oxidized
spectrum was obtained after oxidation by air The arrows indicate the
maxima of the split a-band at 557 and 562 nm.
Fig 3 EPR spectrum of A profundus HmeCD EPR spectrum ob-tained after oxidation of HmeCD (2 mgÆmL)1with 3 m M K 3 Fe(CN) 6
(thin black line) EPR conditions: temperature, 20 K; microwave power, 2 mW; microwave frequency, 9.458 GHz; modulation ampli-tude, 0.6 mT; modulation frequency, 100 kHz The spin concentration was 0.15 spinÆmol)1enzyme as determined by double integration of the simulated EPR signal (thick grey line) Simulation parameters:
g 1,2,3 ¼ 1.948, 1.995 and 2.031; W 1,2,3 ¼ 1.25, 1.15 and 1.325 mT.
Table 2 Purification of Mvh:Hdl enzyme complex from A profundus Hydrogenase-uptake activity was measured after each chromato-graphic step as described in Materials and methods One unit of enzyme activity corresponds to the reduction of two lmol of benzyl-viologen per minute.
Trang 6did not correspond to one of the amino-terminal
sequences determined for the subunits of the purified
enzyme, however, its molecular mass corresponds to
the apparent molecular mass of the 22 kDa subunit of
the purified enzyme The amino-terminal sequence of the
22-kDa polypeptide did not showany significant sequence similarity to proteins in the databases The AF1374 to AF1372 gene products revealed high sequence identity to the three subunits of F420-nonreducing hydrogenase (Mvh) from Methanothermobacter spp and related methanogens [12,30], the AF1377 to AF1375 proteins showed high sequence identity to subunits HdrA, HdrB and HdrC of Hdr from Methanothermobacter species and related meth-anogens [3] Due to these sequence identities the proteins
of the A profundus enzyme complex were designated as MvhA (AF1372) MvhG (AF1373) and MvhD (AF1374), HdlA (AF1377), HdlC (AF1376) and HdlB (AF1375) Hdl stands for HdrABC-like
A detailed sequence analysis revealed the following data (Table 4) HdlA shows sequence similarity to subunit HdrA
of heterodisulfide reductase and shares four binding motifs for [4Fe)4S] clusters (Cx2Cx2Cx3C) and one binding motif for FAD [GxGx2Gx16)19(D/E)] with HdrA HdlC corres-ponds to subunit HdrC of Hdr and shares two binding motifs for [4Fe)4S] clusters with HdrC A multiple sequence alignment of various members of the HdrC protein family showed that the amino terminus of these proteins is poorly conserved This may explain why the determined amino terminus of the 22-kDa polypeptide could not be assigned to the AF1376 gene product HdlB shows sequence similarity to subunit HdrB of Hdr The two
CX31)39CCX35)36CX2C sequence motifs present in HdrB are also conserved in HdlB For Hdr it has been proposed that some of these cysteine residues ligate the active-site iron–sulfur cluster [8,9] MvhD from Mt marburgensis binds a [2Fe)2S] cluster [31] It contains five cysteine residues also conserved in the AF1374 gene product MvhG (AF1373) was identified as hydrogenase small subunit with highest sequence identity to MvhG of Mt thermoautotro-phicus This protein contains 14 cysteine residues, 12 of these are highly conserved among the hydrogenase small subunits
of several [NiFe] hydrogenases and are predicted to ligate three [4Fe)4S] clusters MvhA (AF1372) was identified as
Fig 4 SDS/PAGE of the Mvh:Hdl enzyme complex from A
profun-dus Proteins were separated in a 14% slab gel (8 · 7 cm), w hich w as
subsequently stained with Coomassie Brilliant Blue R250 Lane 1,
25 lg of purified Mvh:Hdl complex; M, low-molecular-mass marker
(Amersham Pharmacia Biotech) The molecular masses of the marker
peptides are given on the right side The apparent molecular masses of
the polypeptides of lane 1 are given on the left side.
Table 3 Amino-terminal sequences of the soluble hydrogenase of A profundus Amino-terminal sequences of the A profundus enzyme were derived
by Edman-degradation, amino-terminal sequences of A fulgidus were derived from the genome sequence [13] In both sequences, identical amino acids are underlined Annotations made are based on sequence identities of the respective polypeptides (see text) X, No clear assignment to an amino acid could be made; –, gaps inserted to allowan alignment.
Amino-terminal
Trang 7-hydrogenase large subunit carrying the four cysteine ligands
of the binuclear [Ni–Fe] centre
None of the polypeptides reported above has extended
hydrophobic regions, which could form
membrane-span-ning helices This agrees well with the finding that the
enzyme was purified from the soluble fraction
Cofactor analysis and characterization of the iron–sulfur
clusters by EPR spectroscopy
The enzyme preparation contained 3.7 ± 0.5 nmol NiÆ
mg)1protein, 214 ± 9 nmol acid-labile sulfurÆmg)1protein
and 207 ± 11 nmol ironÆmg)1protein As predicted from
the primary structure it contains a flavin identified as
FAD) 3.0 ± 0.2 nmol FADÆmg)1protein were found A
densitometric analysis of Coomassie Brilliant blue-stained
SDS gels indicated the presence of all subunits in
stoi-chiometric amounts in the complex From the genome
sequence of the close relative A fulgidus the molecular mass
of the complex was calculated to be 220 kDa Per mol
the enzyme complex thus contains 0.83 ± 0.12 mol Ni,
0.73 ± 0.05 mol FAD, 47 ± 2 mol acid-labile sulfur and
45 ± 2 mol nonheme iron From the primary sequence the
enzyme is predicted to harbour one FAD, one [Ni–Fe]
centre, one [2Fe)2S] cluster and nine [4Fe)4S] clusters and
one active-site [Fe–S] cluster in HdlB
A characterization of the iron–sulfur clusters by EPR
spectroscopy revealed the following results: the H2reduced
enzyme exhibited a spectrum dominated by a signal with
g-values at 2.036, 1.933 and 1.912 (Fig 6A) This signal was
detectable without significant broadening at temperatures
up to 80 K The g-values, temperature behaviour and redox
properties are reminiscent of a signal detected in purified
Mvh from Mt marburgensis where this signal was
attrib-uted to a [2Fe)2S]1+cluster [31] In the spectrum of the
H2-reduced enzyme a radical signal around g¼ 2 w as also visible The intensity of this signal increased upon further reduction of the enzyme by sodium dithionite (not shown) The line width of the radical signal is 1.5 mT as determined from a spectrum recorded at 120 K (data not shown) In the absorption spectrum there is no maximum around 600 nm, which would be indicative for a neutral (blue) semiquinone This all points to an anionic (red) flavinsemiquinone radical (line width 1.5 mT) [32]
At temperatures below20 K additional signals in the reduced enzyme were detectable as a shoulder of the much more intensive [2Fe)2S]1+ signal at g¼ 1.890 (Fig 6A) These signals are indicative of spin–spin coupling between the different [4Fe)4S]1+clusters in the enzyme [33,34] The duroquinone (2,3,5,6-tetramethyl-p-benzoquinone)-oxidized enzyme exhibited a rhombic EPR signal with gxyz
values at 2.014, 1.939 and 1.895 The line shape of this spectrum was similar to the spectrum observed for oxidized HmeCD (Fig 3), however, the g-values are shifted to lower values (Fig 6B) This paramagnetic species could be measured under nonsaturating conditions at 20 K and was detectable without significant broadening up to 60 K The signal broadened beyond detection at 110 K The total spin concentration was 13 lMcorresponding to 0.35 spinÆmol)1 enzyme When the oxidized sample was incubated under 100% H2the signal was no longer detectable and again the [2Fe)2S]+ signal described above was observed Experi-ments with heterodisulfide (CoM-S-S-CoB) as electron acceptor and hydrogen as electron donor showed that the complex has no detectable activity with these substrates (data not shown) When the enzyme was oxidized with
K3Fe(CN)6 a g¼ 2.02-EPR signal indicative of a [3Fe)4S]+ cluster was observed This cluster was most probably formed by the oxidative degradation of a [Fe)4S] cluster at high redox potentials
Table 4 Features of the subunits of the Mvh:Hdl complex from A profundus.
TIGR
annotation
Calculated molecular mass
Cofactor binding sites
Mt thermoautotrophicus
Sequence
Fig 5 Genomic organization of the genes encoding the subunits of the Mvh:Hdl enzyme complex and a putative membrane-bound hydrogenase in
A fulgidus The ORFs annotated by TIGR are given above the arrowrepresenting the genes and their direction of transcription The gene names of genes encoding the Mvh:Hdl complex are given belowthe gene symbols The genes have almost no intergenic regions or they overlap The AF1371 gene product was not found in the enzyme preparation It is predicted to encode a hydrogenase maturation protease The AF1381–AF1379 genes encode a putative membrane-bound hydrogenase closely related to the F 420 -nonreducing hydrogenases (Vho and Vht) from Mt mazei [37] AF1378 encodes a putative hydrogenase maturation protease.
Trang 8In addition, signals derived from the [NiFe] centre were observed with gxyz¼ 2.338, 2.174 and 2.007 in the duro-quinone oxidized enzyme (data not shown) This signal most probably corresponds to the Ni(III) ready form of the enzyme [35] The identification of this Ni(III) derived signal clearly shows that the enzyme is an oxidized state
Discussion
In the present study two Hdr-like enzymes, HmeCD and the Mvh:Hdl-complex, were isolated from H2/sulfate-grown cells of A profundus Each enzyme contains a subunit (HmeD or HdlB) with sequence similarity to the proposed catalytic subunit of Hdr (HdrD from Ms barkeri or HdrB from Mt marburgensis) The EPR signals observed for both, HmeCD and the Mvh:Hdl complex from A profun-dusare reminiscent to the CoM–Hdr signal However, the two enzymes from A profundus form this paramagnetic species already when oxidized with either ferricyanide
or duroquinone in the absence of any added thiol The same result was recently obtained with the A fulgidus HmeACDE complex [14] One possible explanation could
be that the enzymes contain substoichiometric amounts of a tightly bound thiol, which becomes ligated to the active-site [Fe–S] cluster upon oxidation This could also explain why the spin concentration obtained is much lower than 1 spin per molecule With the Mt marburgensis enzyme the CoM– Hdr EPR signal could also be obtained when oxidized in the presence of nonsubstrate thiols such as b-mercaptoethanol
or cysteine With these thiols the midpoint potential of the signal was, however, shifted to rather nonphysiological, high values [7] Also in the A profundus enzymes a nonsubstrate thiol might induce the paramagnetic species observed by EPR spectroscopy
The architecture of A profundus HmeCD resembles that
of HdrDE from Ms barkeri (Fig 7) [4,5] It contains a b-type cytochrome as membrane anchor and a hydrophilic iron–sulfur protein, presumably carrying the active-site for the reduction of a yet unidentified substrate In Ms barkeri Hdr together with a membrane-bound [NiFe] hydrogenase and the membrane-bound electron carrier methanophena-zine forms an electron transport chain catalysing the reduction of CoM-S-S-CoB by H2 This reaction is coupled
to the formation of a proton motive force [10,36] Two isoenzymes of the membrane-bound hydrogenase, called Vho and Vht, are present in Ms barkeri Both enzymes contain a membrane anchoring b-type cytochrome in addition to the hydrogenase large and small subunit The two latter subunits are predicted to extrude into the extracytoplasmic side of the membrane [37] Interestingly,
a closely related hydrogenase is encoded by the genome of
A fulgidus (AF1381–1379) [13] (Fig 5) The genes are directly upstream of the genes encoding the Mvh:Hdl complex However, only 3% of the hydrogenase activity present in cell extracts of A profundus were localized in the membrane fraction This could be due to the lability of the enzyme Also the Vho/Vht hydrogenases from Ms barkeri rapidly dissociate from their membrane anchor [37,38] It is thus reasonable to assume that the proposed membrane-bound hydrogenase of A profundus became detached from its membrane anchoring b-type cytochrome subunit upon cell lysis and thus was released into the soluble fraction
Fig 6 EPR spectra of the H 2 -reduced and duroquinone-oxidized
A profundus Mvh:Hdl enzyme complex (A) EPR spectra obtained
after reduction of the Mvh:Hdl complex (8.9 mg proteinÆmL)1 at
pH 7.0) with hydrogen (1.2 · 10 5 Pa) at 10 K (pow er ¼ 0.2 mW)
and 25 K (power ¼ 2 mW) The upper spectrum shows a
simula-tion of the [2Fe )2S] 1+
signal at 25 K Simulation parameters:
g 1,2,3 ¼ 1.912, 1.933 and 2.036; W 1,2,3 ¼ 2.0, 3.6 and 4.6 mT The
flavin radical signal is saturated under these conditions (B) EPR
spectrum obtained after oxidation of the Mvh:Hdl enzyme complex
(8.9 mg proteinÆmL)1) w ith 3 m M duroquinone (thin black line).
The total spin concentration was 0.35 spinÆmol)1 enzyme as
determined by double integration of the simulated EPR signal
(thick grey line) Simulation parameters: g 1,2,3 ¼ 1.895, 1.939 and
2.014; W 1,2,3 ¼ 2.54, 1.62 and 1.00 mT EPR conditions:
tempera-ture, 20 K; microw ave pow er, 2 mW; microw ave frequency,
9.458 GHz; modulation amplitude, 0.6 mT; modulation frequency,
100 kHz.
Trang 9During the purification of the Mvh:Hdl complex no other
major fraction with hydrogenase activity was detected
However, chromatography on hydroxyapatite resulted in a
significant loss of hydrogenase activity This could be due
to the inactivation or irreversible binding of this second
hydrogenase during chromatography on hydroxyapatite
During chromatography of solubilized membrane proteins
on Q-Sepharose, in addition to the HmeCD-containing
fraction further heme-containing fractions were observed
which contained 20% of the total heme present in the
membrane fraction These fractions, which might contain
the b-type cytochrome of the membrane-bound
hydro-genase, have not yet been analysed further This proposed
membrane-bound hydrogenase and HmeCD could be part
of an electron transport chain very similar to that present in
Ms barkeriwith the exception that methanophenazine is
replaced by the modified menaquinone described for
A fulgidus [39] Unlike HmeCD from A profundus the
A fulgidus HmeACDE complex contains the two
addi-tional subunits HmeA and HmeC Both subunits are
predicted to extrude into the extracytoplasmic side of the membrane These subunits have recently been proposed to form a distinct module, which may mediate the electron transfer from the menaquinone pool to alternative electron acceptors or oxidoreductases [14] We can currently not exclude that these subunits are also formed in A profundus and in vivo form a complex with HmeCD but are lost during the purification
The Mvh:Hdl complex from A profundus is closely related to the Mvh:Hdr complex from Methanothermo-bacter species (Fig 7) The sequences deduced from the AF1377–1372 (hdlACB and mvhAGD) genes of A fulgidus not only showhigh sequence identities to the corresponding subunits of the Mvh:Hdr complex from Methanothermo-bacter spp but also contain all cofactor binding sites present in the Mt marburgensis enzyme complex This was also confirmed by the biochemical characterization of the
A profundusenzyme complex A putative transcription unit encoding all six subunits was identified in the genome of
A fulgidus In the genome of Mt thermoautotrophicus the
Fig 7 Schematic presentation of (A) HmeCD from A profundus in comparison to HdrDE from Ms barkeri and the HmeABCDE complex from
A fulgidus and (B) the Mvh:Hdl complex from A profundus in comparison to the Mvh:Hdr complex from Methanothermobacter spp The scheme is based on the sequence analysis of the encoding genes from A fulgidus The physiological substrate of the A profundus enzyme is yet unknown, but based on the similarity to Hdr a disulfide is proposed MP, methanophenazine; MQ, menaquinone.
Trang 10genes encoding the six subunits of the Mvh:Hdr complex are
located at three different loci [40]
Also the sulfate-reducing bacterium D vulgaris contains a
putative six-gene operon (ORF2976–2967) encoding an
enzyme complex, designated as H2-heterodisulfide
oxido-reductase complex, closely related to the Mvh:Hdl complex
described here These genes are expressed in D vulgaris as
was shown by macroarray RNA hybridizations [41]
Expres-sion of these genes was found to be downregulated in a
Fe-only hydrogenase mutant strain (Dhyd) Downregulation
was also observed in a strain carrying a deletion of the adh
gene, encoding an alcohol dehydrogenase One possible
function discussed for the H2-heterodisulfide oxidoreductase
complex in D vulgaris is the formation of H2with reducing
equivalents generated by the oxidation of ethanol to acetate,
as part of a H2-cycling system [41] Thus far, the Mvh:Hdl
genetic organization is unique to sulfate reducers
The complete understanding of the process of
dissimila-tory sulfate reduction is hampered by the high complexity of
the systems studied thus far Many of the organisms studied
are able to utilize several electron donors for growth and
contain many different electron transfer components in
parallel In contrast, A profundus is obligatory
hydrogeno-trophic [20] For this organism H2is the ultimate electron
donor and hence reducing equivalents generated upon H2
oxidation have to be channelled to the enzymes of sulfate
reduction Our finding that a major hydrogenase present in
cell extracts of A profundus forms a tight complex with an
Hdr-like enzyme strongly supports previous assumptions
that Hdr-like enzymes play an essential role in the electron
transport chain(s) of sulfate reducing archaea and bacteria
The Mvh:Hdl enzyme complex accounts for at least 2.5% of
the total cell protein of A profundus indicating an important
catabolic function In analogy to the Mvh:Hdr complex
from Methanothermobacter species, the A profundus
enzyme complex is proposed to reduce an electron acceptor
which in turn could function as electron donor of the
enzymes of sulfate reduction
Several questions remain to be answered What is the
nature of the physiological electron acceptor of the two
Hdr-like enzymes present in A profundis? The similarity to Hdr
suggests a disulfide Do both enzymes reduce the same
electron acceptor or are different substrates used? If both
systems reduce the same substrate, why are two different
enzyme systems operating? Is the Mvh:Hdl enzyme complex
in vitrobound to the cytoplasmic membrane via additional
membrane subunits and is the reaction catalysed by this
complex coupled to energy conservation, as has been
proposed for the Mvh:Hdr complex from
Methanothermo-bacter marburgensis[12] In future studies the identification of
the substrate(s) of both enzymes will be attempted A possible
way this could be done is the addition of low molecular mass
fractions of cell extracts to purified HmeCD or Mvh:Hdl in
the presence of an oxidant This could result in higher spin
concentrations of the paramagnetic species, assuming that
the substrate binds to an iron–sulfur cluster in the active site
Acknowledgements
This work was supported by the Max-Planck-Gesellschaft, by the
Deutsche Forschungsgemeinschaft, and by the Fonds der Chemischen
Industrie We thank D Linder for amino-terminal sequencing.
References
1 Thauer, R.K (1998) Biochemistry of methanogenesis: a tribute to Marjory Stephenson Microbiology 144, 2377–2406.
2 Hedderich, R., Berkessel, A & Thauer, R.K (1990) Purification and properties of heterodisulfide reductase from Methanobacter-ium thermoautotrophicum (strain Marburg) Eur J Biochem 193, 255–261.
3 Hedderich, R., Koch, J., Linder, D & Thauer, R.K (1994) The heterodisulfide reductase from Methanobacterium thermo-autotrophicum contains sequence motifs characteristic of pyridine nucleotide-dependent thioredoxin reductases Eur J Biochem.
225, 253–261.
4 Heiden, S., Hedderich, R., Setzke, E & Thauer, R.K (1994) Purification of a two-subunit cytochrome-b-containing hetero-disulfide reductase from methanol-grown Methanosarcina barkeri Eur J Biochem 221, 855–861.
5 Ku¨nkel, A., Vaupel, M., Heim, S., Thauer, R.K & Hedderich, R (1997) Heterodisulfide reductase from methanol-grown cells of Methanosarcina barkeri is not a flavoenzyme Eur J Biochem 244, 226–234.
6 Simianu, M., Murakami, E., Brew er, J.M & Ragsdale, S.W (1998) Purification and properties of the heme- and iron-sulfur-containing heterodisulfide reductase from Methanosarcina thermophila Biochemistry 37, 10027–10039.
7 Madadi-Kahkesh, S., Duin, E.C., Heim, S., Albracht, S.P.J., Johnson, M.K & Hedderich, R (2001) A paramagnetic species with unique EPR characteristics in the active site of heterodisulfide reductase from methanogenic archaea Eur J Biochem 268, 2566–2577.
8 Duin, E.C., Madadi-Kahkesh, S., Hedderich, R., Clay, M.D & Johnson, M.K (2002) Heterodisulfide reductase from Methano-thermobacter marburgensis contains an active-site [4Fe-4S] cluster that is directly involved in mediating heterodisulfide reduction FEBS Lett 512, 263–268.
9 Duin, E.C., Bauer, C., Jaun, B & Hedderich, R (2003) Coenzyme
M binds to a [4Fe-4S] cluster in the active site of heterodisulfide reductase as deduced from EPR studies with the [33S]coenzyme M-treated enzyme FEBS Lett 538, 81–84.
10 Ide, T., Ba¨umer, S & Deppenmeier, U (1999) Energy conserva-tion by the H 2 : heterodisulfide oxidoreductase from Methano-sarcina mazei Go¨1: Identification of two proton-translocating segments J Bacteriol 181, 4076–4080.
11 Ba¨umer, S., Ide, T., Jacobi, C., Johann, A., Gottschalk, G & Deppenmeier, U (2000) The F 420 H 2 dehydrogenase from Methanosarcina mazei is a redox-driven proton pump closely related to NADH dehydrogenases J Biol Chem 275, 17968–17973.
12 Setzke, E., Hedderich, R., Heiden, S & Thauer, R.K (1994) H 2 : heterodisulfide oxidoreductase complex from Methanobacterium thermoautotrophicum Composition and properties Eur J Biochem 220, 139–148.
13 Klenk, H.P., Clayton, R.A., Tomb, J.F., White, O., Nelson, K.E., Ketchum, K.A., Dodson, R.J., Gwinn, M., Hickey, E.K., Peter-son, et al (1997) The complete genome sequence of the hyper-thermophilic, sulphate-reducing archaeon Archaeoglobus fulgidus Nature 390, 364–370.
14 Mander, G.J., Duin, E.C., Linder, D., Stetter, K.O & Hedderich,
R (2002) Purification and characterization of a membrane-bound enzyme complex from the sulfate-reducing archaeon Archaeoglobus fulgidus related to heterodisulfide reductase from methanogenic archaea Eur J Biochem 269, 1895–1904.
15 Valente, F.M., Saraiva, L.M., LeGall, J., Xavier, A.V., Teixeira,
M & Pereira, I.A (2001) A membrane-bound cytochrome c 3 : a type II cytochrome c 3 from Desulfovibrio vulgaris Hildenborough Chembiochemistry 2, 895–905.