 |
INTRODUCTION |
The methanoarchaea are strictly anaerobic microbes that evolve
methane as a product of their energy-yielding metabolism and are
classified within the Archaea domain (1). These microbes utilize two
distinct pathways for most of the methane produced in the biosphere. In
the acetate fermentation pathway (Reaction 1), the methyl group is
reduced to methane with electrons derived from oxidation of the
carbonyl group to CO2. In the CO2 reduction pathway, CO2 is reduced to methane at the expense of
electrons provided by either hydrogen (Reaction 2) or formate (Reaction 3).
Energy is conserved in both pathways by an electron
transport-coupled phosphorylation of ADP. Although the understanding of
electron transport in both of these pathways is incomplete, flavoproteins are of major importance. A flavoprotein has been described that appears to function in electron transport pathways of
CO2-reducing methanoarchaea (2, 3). The oxidations of H2 and formate are catalyzed by the Fe-S flavoproteins,
hydrogenase, and formate dehydrogenase (4-7). The final electron
transfer step for both pathways (Reaction 4) is catalyzed by
heterodisulfide reductase
(HDR),1 which is a FAD/Fe-S
protein in the CO2 reducer, Methanobacterium thermoautotrophicum (8), and a heme/Fe-S in the acetate
fermenters, Methanosarcina thermophila (9) and
Methanosarcina barkeri (10). This reaction regenerates the
active sulfhydryl forms of the coenzymes, CoB and CoM (8). The
substrate for HDR, the heterodisulfide CoB-S-S-CoM, is the product of
the methane-forming reaction (Reaction 5), which is also common to both
pathways and catalyzed by methyl-CoM reductase.
Recently, a homodimeric Fe-S flavoprotein Isf
(iron-sulfur flavoprotein) was
identified from the acetate-fermenting methanoarchaeaon M. thermophila (11). The isf gene was cloned and
sequenced, and Isf was produced in Escherichia coli and
partially characterized. Comparisons of the deduced Isf sequence with
sequences in the available protein data bases suggest that Isf is a
novel Fe-S flavoprotein. Reconstitution experiments suggest Isf is a
required component of the electron transport chain in the pathway for
the fermentation of acetate (11). The UV-visible absorption spectrum and cofactor and elemental analyses indicate that Isf contains one FMN
and either one [4Fe-4S] or one [3Fe-4S] center per monomer. The
sequence of Isf contains several cysteine residues; however, none of
the typical cysteine motifs known to accommodate [4Fe-4S] or
[3Fe-4S] centers are recognizable. Thus, the identity of the Fe-S
center is uncertain. Here we report on EPR and Mössbauer studies
of Isf from M. thermophila, which demonstrate that the Fe-S
center is of the [4Fe-4S] type. Potentiometric titrations of the
[4Fe-4S] and FMN centers predict a role for Isf in the electron
transport pathway leading to the formation of CH4 in acetate-fermenting methanoarchaea. Additional results are presented that indicate Isf homologs also have a role in electron transport pathways of CO2-reducing methanoarchaea.
 |
EXPERIMENTAL PROCEDURES |
Materials--
M. thermophila Isf was produced in
E. coli strain BL21(DE3) and purified as described
previously (11). Isf was stored at
80 °C in 50 mM
potassium phosphate buffer at pH 7.0. For production of
57Fe-labeled Isf, E. coli cells were in medium
containing 20 µM 57Fe (Penwood Chemicals,
Inc.), which was prepared by dissolving the iron metal in 2 N H2SO4 at 60 °C for 1 week.
M. thermophila Fd (12) was purified as described. M. thermoautotrophicum strain Marburg was grown as described (13).
The following redox dyes were used: phenosafarin (Sigma), benzyl
viologen (Sigma), methyl viologen (Sigma),
1,1'-trimethylene-2,2'-bipyridyl, and 4,4'-dimethyl-2,2'-dipyridyl (Aldrich). All experiments utilized NANOpure deionized water. N2 (99.98%) and helium (99.998%) were obtained from
Linweld (Lincoln, NE). Helium was deoxygenated by passing through an
Oxisorb column and a heated column containing BASF catalyst.
Methods--
Protein manipulations were performed under strictly
anaerobic conditions at 20 °C in a Vacuum Atmospheres chamber
maintained between 1 and 5 ppm oxygen. The concentrations of the
oxidized Isf were determined spectrophotometrically with the extinction coefficients previously determined (11) and by the Rose Bengal assay
(14).
UV-visible Spectroscopy and Potentiometric
Titrations--
Potentiometric measurements were performed as
described previously (15, 16). All electrochemical potentials are
reported relative to the standard hydrogen electrode. Isf (3-4
µM) was titrated at 20 °C in 50 mM
potassium phosphate buffer (pH 7.0-7.05) in a solution containing
methyl viologen (0.1 mM) as the mediator dye with
phenosafarin (Em =
0.252 V, pH 7.0) (5 µM) and benzyl viologen (Em =
0.362
V, pH 7.0) (5 µM) as the indicator dyes. The pH measured
after the experiment was recorded as the pH for the titration. The
visible spectra in each experiment were obtained and stored on an
Olis-14 interfaced Cary spectrophotometer. The absorbance at 480 nm was
used to monitor the amount of oxidized and reduced FMN after correcting
the spectra for turbidity. The reduction potentials reported were
determined by potentiometric measurements in the reductive direction.
After each potentiometric titration of the FMN chromophore, the
iron-sulfur flavoprotein was reoxidized completely using ferrocyanide
(0.1 mM) as the mediator dye. Equilibrium of the system in
the UV-visible potentiometric measurements was considered to be
obtained when the measured potential drift was less than 1 mV in 5 min;
this was typically around 1-2 h. The midpoint potentials
(Em) and n values were calculated using
the Nernst equation (Equation 1),
|
(Eq. 1)
|
where E is the measured equilibrium potential at each
point in the titration, and n is the number of electrons.
The typical error in the reported reduction potential values was ±2-3
mV. All midpoint potential value determinations exhibited Nernstian behavior as indicated by their n values.
EPR Spectroscopy and Potentiometric Titrations--
The
EPR spectra in each experiment were recorded on a Bruker ESP 300E
spectrometer equipped with an Oxford ITC4 temperature controller, a
Hewlett Packard model 5340 automatic frequency counter, and Bruker
gaussmeter. The spectroscopic parameters are given in the figure
legends. Double integration of the EPR signals was performed with
copper perchlorate (1 mM) as the standard. Isf was frozen
in liquid nitrogen prior to EPR analyses.
For the power saturation studies, Isf (74 µM, pH 7.0) was
reduced in the presence of 50 mM methyl viologen with a
40-fold excess of sodium dithionite prepared freshly at pH 9.0. The
solution was immediately frozen in an EPR tube and stored in liquid
nitrogen. Spectra of the reduced [4Fe-4S] cluster were recorded at
powers varying from 0.1 to 200 milliwatts at five different
temperatures between 5 and 25 K. The power for half saturation
(P1/2) at each temperature was determined by a fit
to a plot of log
(S/P·e0.5)
versus log P using Equation 2, where S
is the signal amplitude, P is the microwave power incident
on the cavity, and b is the inhomogeneity parameter (17).
Best fits to the data were obtained by using a b value of
1.2.
|
(Eq. 2)
|
The zero field splitting constant (
) was determined by a
linear fit to a plot of ln P1/2 versus
1/T according to Equation 3, where T is the
temperature, P1/2 is the power for half saturation, k is the Boltzmann constant, and A is a
coefficient representative of the phonon spin-coupling properties of
the [4Fe-4S] cluster (17).
|
(Eq. 3)
|
Potentiometric measurements of the [4Fe-4S] cluster were
performed in an EPR spectroelectrochemical cell (18). Isf samples (80-160 µM) in 50 mM potassium phosphate
buffer (pH 7.0) were titrated at 20 °C in the presence of the
mediator dyes, 150 µM benzyl viologen
(Em =
0.362 V), 150 µM methyl
viologen (Em
0.440 V), 100 µM
1,1'-trimethylene-2,2'-bipyridyl (Em =
0.540 V),
and 100 µM 4,4'-dimethyl-2,2'-dipyridyl (Em =
0.586 V). The intensity of the EPR signal
with a g value of 1.93 was monitored to determine the redox state of the [4Fe-4S] cluster during the titration. Potentiometric
measurements were performed in the reductive and oxidative directions.
The system was considered to have reached equilibrium when the measured potential drift was less than 1 mV in 2 min.
Mössbauer Spectroscopy--
Mössbauer spectra were
recorded on a constant acceleration spectrometer, model MS-1200D from
Ranger Scientific, using a Janis SuperVaritemp cryostat (model 8DT), a
Lakeshore temperature controller (model 340), and a 57Co
source in a rhodium foil purchased from Isotope Products Laboratory. All isomer shifts are quoted relative to iron metal at room
temperature.
 |
RESULTS |
Sequence Analysis--
Resequencing of M. thermophila
genomic DNA revealed an error in the previously reported isf
sequence (11). The corrected isf and deduced Isf sequences
are shown in Fig. 1. The corrected Isf
contains 191 residues, 81 fewer than previously reported. Additionally,
the C-terminal residues 177KLCDVLELIQKNRDK191
in the corrected Isf sequence replace
177NSVMSWNLFRKIEIN191 in the previously
reported Isf. Resequencing revealed the correct isf sequence
in pML701 used for the heterologous production of Isf reported here and
previously (11).

View larger version (21K):
[in this window]
[in a new window]
|
Fig. 1.
Nucleic acid sequence and predicted amino
acid sequence of isf from M. thermophila.
The DNA is presented in the 5' to 3' direction. The predicted amino
acid sequence of Isf is shown in single-letter code directly
below the first base of each codon. *, initial base of
translation stop codon.
|
|
Evidence for Isf Homologs in Phylogenetically and Physiologically
Diverse Methanoarchaea--
M. thermophila is physiologically distinct
from Methanococcus jannaschii and M. thermoautotrophicum, which are unable to ferment acetate to
CH4 and instead oxidize H2 and reduce
CO2 to CH4. The genome of M. jannaschii (19, 20) contains two ORFs (MJ1083 and MJ0731) encoding
predicted proteins with 194 and 192 residues that have high identity to
the corrected M. thermophila Isf sequence (Fig.
2). Inspection of the genomic sequence of
M. thermoautotrophicum (21) identified three ORFs that also
share significant identity to Isf (Fig. 2). An N-terminal cysteine
motif in Isf from M. thermophila is strictly conserved in
the M. jannaschii and M. thermoautotrophicum sequences. These results suggest a previously unrecognized electron transfer role for Isf homologs in electron transport pathways of
CO2-reducing methanoarchaea. M. jannaschii, M. thermoautotrophicum, and M. thermophila represent all
three currently described families of methanoarchaea, suggesting that
Isf homologs are present in phylogenetically diverse methanoarchaea,
which further supports a general function for this electron
carrier.

View larger version (59K):
[in this window]
[in a new window]
|
Fig. 2.
Multiple amino acid sequence alignment of Isf
from M. thermophila with sequences deduced from open
reading frames identified in the genomic sequences of M. jannaschii and M. thermoautotrophicum. Accession
numbers for the M. jannaschii (2) and M. thermoautotrophicum (21) sequences and percentage identity (in
parentheses) with M. thermophila Isf are as follows: MTH1350
(41%), MTH1473 (34%), MTH1595 (30%), MJ1083 (49%), and MJ0731
(40%). Asterisks indicate conserved cysteine
residues.
|
|
Extracts of H2-CO2-grown M. thermoautotrophicum cells catalyzed the reduction of M. thermophila Isf with either H2 or CO as the electron
donor (Fig. 3). These results suggest
that Isf homologs are components of electron transport chains in
CO2-reducing methanoarchaea initiating with either
hydrogenase or CO dehydrogenase (CODH). The rate of Isf reduction with
CO as the electron donor was greater than with H2,
indicating that CO-dependent reduction of Isf does not
require prior conversion to H2 and CO2. The
addition of ferredoxin stimulated the rate of Isf reduction 10-fold by
CO and 1-fold by H2. These results suggest that 8Fe
ferredoxins are able to couple the oxidation of either
H2 or CO to the reduction of Isf; indeed, the sequence of
the M. thermoautotrophicum genome reveals several ORFs
encoding putative 8Fe ferredoxins (21). The ability of H2
or CO to serve as electron donors for the reduction of M. thermophila Isf in a CO2-reducing species is further
indicative of a function for Isf homologs in electron transport
pathways of phylogenetically and physiologically diverse
methanoarchaea. Additional research with purified proteins is necessary
to determine the precise role of Isf homologs in electron transport
pathways of CO2-reducing methanoarchaea.

View larger version (14K):
[in this window]
[in a new window]
|
Fig. 3.
Time course for reduction of Isf with extract
from M. thermoautotrophicum. The assay mixture (700 µl) contained cell extract (180 µg protein), M. thermophila ferredoxin (13.5 µg), 50 mM Tris (pH
7.6), and 2 mM dithiothreitol. Ferredoxin was omitted in
two of the assays ( , ). The assay mixtures were anaerobically
equilibrated with 1.0 atmosphere of CO ( ), H2
( , ), or N2 ( ) in a stoppered 1.0-ml cuvette
maintained at 35 °C. After a 10-min incubation, the reaction was
initiated by the addition of 180 µg of Isf.
|
|
EPR Spectroscopy of the Fe-S Center--
The EPR spectra of the
reduced Fe-S center (Fig. 4) exhibited g
values of 2.06 and 1.93 typical of [4Fe-4S]1+ clusters;
however, the signal in the region of gmin shows two, instead of one, negative absorption features with g values at 1.86 and
1.82. The relaxation properties also are indicative of a [4Fe-4S]
cluster, since the g = 1.93 signal was not observable above 25 K
(22). An average of 1.3 spins/mol of dimeric Isf was determined from
different samples by double integration of the EPR signal at 10 K
referenced to a copper perchlorate standard. The existence of two
negative features in the spectrum of the cluster is unexpected, since
rhombic spectra should exhibit only three g values. The extra features
did not derive from the viologen dyes, since reduction of Isf with
sodium dithionite alone yielded an EPR spectrum identical to that shown
in Fig. 4. In some Fe-S proteins, complex spectral features can be
simplified by incubation with low concentrations of urea; however,
dithionite reduction of Isf in the presence of 2 M urea had
no effect on the spectrum. The possibility that the two negative
absorption features derived from strong hyperfine interaction between
the unpaired electron on the cluster and a strongly coupled proton
(would produce a two-line splitting because it has a nuclear spin of
1/2) also was ruled out. When Isf was extensively exchanged with
D2O and reduced with dithionite, the spectrum was identical
to that shown in Fig. 4. Often, complex EPR spectra are observed when
two paramagnetic centers are close enough to undergo dipolar
interactions, including wings in the gmax and
gmin regions. In addition, there apparently is not another
paramagnet, since the flavin is diamagnetic in the oxidized and reduced
states. Therefore, the peculiar morphology of the EPR spectrum derives
from a unique characteristic of the reduced [4Fe-4S] cluster
tentatively attributed to microheterogeneity within the population of
Isf molecules. This hypothesis is supported by the Mössbauer
results described below.

View larger version (20K):
[in this window]
[in a new window]
|
Fig. 4.
EPR spectroscopy of the [4Fe-4S] cluster in
Isf poised at various redox potentials in 50 mM potassium
phosphate buffer (pH 7.0). Experimental conditions were as
follows: temperature, 10 K; microwave power, 1.26 milliwatts; microwave
frequency, 9.43 GHz; receiver gain, 2 × 104;
modulation amplitude, 10 G; modulation frequency, 100 kHz. The
derivative feature at g = 2.0 results from the mediator
dyes.
|
|
Power saturation studies of the cluster were performed at five
different temperatures. The power for half-saturation
(P1/2) ranged from 79 milliwatts (25 K) to 14.4 milliwatts (5 K). A linear dependence of the ln P1/2
on 1/T was observed at temperatures above 5 K (Fig.
5). The zero field splitting parameter
(
) was between 10 and 11.5 cm
1 (Fig. 5,
inset).

View larger version (16K):
[in this window]
[in a new window]
|
Fig. 5.
Semilogarithmic plot of
P1/2 versus 1/T, which
shows a linear relationship according to Equation 3. The slope
( /k = 1.5 ± 0.73) yields an estimate of 11.5 cm 1 for the the zero field splitting value ( ).
Inset, a nonlinear plot of P1/2
versus 1/T, which includes the data at 5 K. The
slope ( /k = 14.1 ± 1.8) yields an estimate
for of 9.8 cm 1.
|
|
Mössbauer Spectroscopy--
Mössbauer spectra of the
oxidized iron-sulfur center (Fig.
6A) exhibited a single broad
quadrupole doublet with an average isomer shift (
) of 0.45 mm/s and
an average quadrupole splitting (
EQ) of 1.22 mm/s
at 4.2 K, with the quadrupole splitting decreasing slightly
(
EQ = 1.12 mm/s) at 100 K. These parameters are
typical for [4Fe-4S] clusters in the 2+ oxidation state. The line
shape of the quadrupole doublet indicates either that the irons within
the cluster are not identical (one possible deconvolution:
EQ(1) = 1.6 mm/s,
EQ(2) = 1.3 mm/s,
EQ(3) = 1.16 mm/s,
EQ(4) = 0.85 mm/s) or that there is a
microheterogeneous population of iron-sulfur clusters in the Isf
molecules, each having slightly different quadrupole splittings.

View larger version (18K):
[in this window]
[in a new window]
|
Fig. 6.
Mössbauer spectra recorded at 100 K. A, oxidized Isf protein. The solid
line is a least-squares fit, assuming a single quadrupole
doublet and a Voigt line shape ( EQ = 1.12 mm/s,
= 0.43 mm/s). B, reduced Isf protein. The
solid line is a least-squares fit, assuming four
quadrupole doublets and a Voigt line shape ( EQ(1) = 0.74 mm/s, (1) = 0.52 mm/s; EQ(2) = 1.07 mm/s, (2) = 0.54 mm/s; EQ(3) = 1.47 mm/s,
(3) = 0.55 mm/s; and EQ(4) = 1.89 mm/s, (4) = 0.59 mm/s).
|
|
In the Mössbauer spectrum of the reduced Isf protein recorded at
100 K (Fig. 6B), a single, very broad, nonsymmetrical
quadrupole doublet is observed with an average isomer shift,
= 0.55 mm/s, and an average quadrupole splitting,
EQ = 1.30 mm/s. These parameters are typical of [4Fe4S] clusters in the 1+
oxidation state. The nonsymmetrical shape of the doublet, particularly
noticeable in the right component, indicates that the irons
within the cluster have distinct quadrupole splittings. The spectrum
was fit using four distinct
EQ values (see Fig.
6B), which gives a reasonable but not unique deconvolution,
since the quadrupole doublets of the four iron sites are not resolved.
The line shape of the doublet is Voigt (gaussian distribution of a
Lorentzian line shape) rather than Lorentzian, indicative of a
microheterogenous environment for the iron sites.
Above 40 K, the electronic spin of the iron-sulfur cluster relaxes fast
and only quadrupole interactions are observed in Mössbauer spectra; however, below 40 K paramagnetic hyperfine interactions also
are observed. Mössbauer spectra of the reduced Isf protein recorded at 4.2 K (Fig. 7) exhibit
paramagnetic hyperfine structure, as expected for a reduced [4Fe-4S]
cluster with S = 1/2. The shape of the spectra
and the derived hyperfine interactions in weak applied fields are
similar to those observed for the [4Fe-4S]1+ clusters of
Bacillus subtilis ferredoxin (23), E. coli
sulfite reductase (24), and Azobacter vinelandii ferredoxin
(25). The average of the derived hyperfine interactions for the reduced Isf protein (Aave(mixed valence pair) =
32
MHz, Aave(ferrous pair) = +12 MHz) are
comparable with those observed for the [4Fe-4S]1+
clusters of the ferredoxins from B. subtilis (
31 MHz, +16
MHz) (23) and A. vinelandii (
29 MHz, +16 MHz) (25) and
E. coli sulfite reductase (
33 MHz, +17 MHz) (24). A
detailed data analysis of the paramagnetic hyperfine interactions will
require a high field Mössbauer study.

View larger version (22K):
[in this window]
[in a new window]
|
Fig. 7.
Mössbauer spectra of reduced Isf
protein recorded at 4.2 K and 450 G applied parallel (A) or
450 G applied perpendicular (B) to the beam. The
solid line is a theoretical fit of an
S = 1/2 Hamiltonian using the parameters cited
in Fig. 6 and A1 = 27.4, 41.2, 32.9 MHz;
A2 = 27.4, 27.4, 35.0 MHz; A3 = +2.7, +6.2, or +17.8 MHz; A4 = +11.0, +27.4,
+2.7 MHz; 1 = 0, 2 = 0.9, 3 = 0.7, 4 = 0.9.
|
|
Potentiometric Titrations of FMN and the [4Fe-4S]
Cluster--
The visible spectrum of Isf was monitored between 300 and
700 nm during potentiometric titration (Fig.
8). The FMN absorption dominates the
spectrum of the oxidized protein. At potentials of
0.305 and
0.342
V, the flavin is almost completely reduced and the spectrum of the
oxidized [4Fe-4S] cluster is dominant. The ratio of the
oxidized:reduced flavin was determined by monitoring the FMN absorbance
peak at 480 nm, instead of 452 nm, to minimize possible spectral
interference from the [4Fe-4S] cluster. Since [4Fe-4S] clusters
have their maximum absorbance between 390 and 420 nm, the cluster
exhibits a negligible contribution to the 480 absorbance value. The
difference extinction coefficient is calculated to be 48 mM
1 cm
1 for the dimeric
protein. An Em value of
0.277 ± 0.003 V was
determined for the FMN/FMNH2 couple of Isf (Fig. 8,
inset). A Nernst plot gave a 34-mV slope for the
potentiometric measurement, which is near the theoretical value of 29 mV for a two-electron transfer. Neither the red nor blue forms of the
FMN semiquinone, which have characteristic spectra, were observed
during the titration. The FMN semiquinone also was not observed in
titrations of Isf with sodium dithionite (data not shown) (11). Thus,
the individual formal potential value of
E1o' (FMN/FMNH·) must be below
0.380 V, and the formal potential value of
E2o'
(FMNH·/FMNH2) must be above
0.175 V.

View larger version (27K):
[in this window]
[in a new window]
|
Fig. 8.
Potentiometric titration of the FMN in Isf
(3.2 µM) in 50 mM potassium phosphate buffer
(pH 7.0) at 20 °C (curves 1-7, fully oxidized, 0.262,
0.272, 0.281, 0.290, 0.305, and 0.342 V, respectively).
Inset, Nernst plot of the potentiometric data.
|
|
Although the flavin semiquinone is thermodynamically unstable during
the potentiometric titrations, the FMN
radical was formed when
Isf was reacted with CODH/ACS from M. thermophila and CO. Samples were frozen at different time points during the reaction and
monitored by EPR spectroscopy. Over the course of 50 min, the cluster
underwent reduction as the FMN was fully reduced to the hydroquinone
state. A radical signal at g = 2.00 (Fig.
9) that can be attributed to the flavin
semiquinone formed to a maximum of 2.5% of the total FMN present at 28 min and then decayed. The EPR signal yielded a line width of 16 gauss,
indicative of an anionic or red semiquinone species (1). Thus, although
formation of semiquinone is not thermodynamically favorable, these
results suggest that a transiently stable semiquinone can exist under physiological conditions during the electron transfer from CODH/ACS to
Isf.

View larger version (11K):
[in this window]
[in a new window]
|
Fig. 9.
EPR spectrum of Isf (170 µM
dimer) was recorded at 10 K following incubation for 17 min with CO and
CODH (45 µg) at 25 °C. The amount of FMN hydroquinone and
reduced FeS cluster at this time point were 46 and 3% (0.03 spin/mol),
respectively. After the sample was frozen in liquid nitrogen, the
spectrum was recorded using the conditions described in Fig. 4.
|
|
The midpoint potential determined for the 2+/1+ couple of the
[4Fe-4S] cluster was
0.394 V at pH 7.0 (Figs. 4 and
10). The slope of the log-linear plot
was 53 mV, which is close to the theoretical value of 58 mV for a
one-electron transfer. The redox reaction was fully reversible, since
the reduction was titrated in both the oxidative and reductive
directions. Thus, the midpoint potential of the [4Fe-4S] cluster is
more than 100 mV lower than that of the FMN/FMNH2 couple
(Figs. 8 and 10) and is similar to the value reported for other low
potential [4Fe-4S] clusters (26, 27).

View larger version (14K):
[in this window]
[in a new window]
|
Fig. 10.
A fit of the Isf midpoint potential data to
a theoretical curve generated from the Nernst equation for two
redox centers with reduction potentials of 0.277 V (n
= 2) and 0.394 V (n = 1).
|
|
 |
DISCUSSION |
The methanogenic fermentation of acetate by the
methanosarcina involves cleavage of acetyl-CoA into carbonyl and methyl
groups, the latter of which is reduced to CH4 with an
electron pair derived from oxidation of the former to CO2
(1). Cleavage of acetyl-CoA and oxidation of the carbonyl group is
catalyzed by a five-subunit CODH/ACS complex. The methyl group is
transferred to tetrahydrosarcinapterin and then to coenzyme M. Reductive demethylation of the methyl group of CH3-S-CoM to
CH4 requires the electron donor coenzyme B (HS-CoB), which
produces the heterodisulfide CoB-S-S-CoM as a second reaction product.
The heterodisulfide is reduced to the corresponding active sulfhydryl
forms of the cofactors by HDR. The electron pair for this reduction
originates from oxidation of the carbonyl group of acetyl-CoA by the
CODH/ACS complex. Oxidation of either exogenous CO or the carbonyl
group of acetyl-CoA is proposed to take place at the nickel/Fe-S center
(center C) in the CdhA subunit of the CODH/ACS complex (28, 29), which
also contains a low potential [4Fe-4S] center (28) that is proposed to shuttle electrons from center C to a low potential 8Fe ferredoxin (12). Electron transfer from ferredoxin to HDR apparently involves membrane-bound electron carriers consistent with the formation of a
transmembrane proton gradient (9, 30). Possible membrane-bound electron
carriers include methanophenazine (31) and/or b-type cytochromes. Cytochromes of the b-type are reduced by CO and
oxidized by CoB-S-S-CoM, which suggests that they could play a role in electron transport during the fermentation of acetate (30). Acetate-grown methanosarcina are reported to contain three
b-type cytochromes with midpoint potentials of
330,
250,
and
182 mV (32). In addition, the purified HDRs from M. thermophila (9) and M. barkeri (10, 33) contain two
b-type cytochromes.
Isf is reducible by ferredoxin, and Isf stimulates electron flow from
CO to CoB-S-S-CoM in a reconstituted system containing CODH/ACS,
ferredoxin, membranes, and HDR, indicating that Isf is a component of
the electron transport pathway of M. thermophila (11);
however, it was not possible to propose with confidence a specific role
for Isf in electron transport without further characterization of the
redox centers. The sequence of Isf reveals a four-cysteine spacing that
is perfectly conserved with Isf-like sequences from phylogenetically
distant methanoarchaea (Fig. 2); however, the cysteine spacing is
atypical of motifs coordinating [4Fe-4S] or [3Fe-4S] centers. The
EPR and Mössbauer spectroscopic results presented here
unequivocally demonstrate the presence of a [4Fe-4S] center in Isf.
Additional experiments are required to conclusively identify the
coordinating ligands to the [4Fe-4S] center. The presence of two
negative absorption features in the EPR spectra of the reduced cluster
is unusual. The results presented here indicate that microheterogeneity
within the population of Isf molecules accounts for this atypical
feature. This situation is similar to that of a class of corrinoid/Fe-S
proteins from methanoarchaea and homoacetogenic anaerobes from the
bacteria domain in which the reduced [4Fe-4S] cluster exhibits a
broad absorption feature in the same g value region (34-36). The
electrochemical properties of the [4Fe-4S] cluster in Isf are highly
similar to those of other low potential [4Fe-4S] clusters like the
clostridial ferredoxins (37). Since the Em of
M. thermophila ferredoxin is
0.407 V (12), reduced
ferredoxin would be expected to efficiently donate electrons to the Isf
[4Fe-4S] cluster (Fig. 11). It is
likely that both redox centers of Isf are involved in electron transfer reactions coupled to CO metabolism, since both the [4Fe-4S] cluster and the flavin of Isf are reduced by CO in the presence of ferredoxin and the CODH/ACS complex (11). The Em values
presented here suggest that intramolecular electron transfer is from
the [4Fe-4S] center to FMN. The Em values
presented here for Isf, combined with published Em
values for other electron transfer components, suggest that the
[4Fe-4S] and FMN redox centers in Isf mediate electron transfer
between ferredoxin and membrane-bound b-type
cytochromes.

View larger version (85K):
[in this window]
[in a new window]
|
Fig. 11.
Proposed electron transport pathway for
oxidation of CO or the carbonyl group of acetyl-CoA and reduction of
CoB-S-S-CoM. Cyt b, cytochrome b;
CdhA, subunit of the CO dehydrogenase/acetyl-CoA synthase;
FdxA, ferredoxin; CoB, coenzyme B;
CoM, coenzyme M; ?, postulated unknown electron
carrier. Midpoint potentials are shown in mV.
|
|
Most flavin/Fe-S proteins stabilize four redox states:
flavinox:FeSox, flavin
semiquinone:FeSox, flavin semiquinone:FeSred, and flavinred:FeSred (38, 39). Kinetic or
thermodynamic stabilization of the semiquinone allows the versatility
of mediating both one-electron and two-electron transfer reactions. In
contrast, the flavin semiquinone of bound FMN in Isf is
thermodynamically unstable. Therefore, the physiological electron
acceptor for Isf (which is unknown) could be a two-electron carrier;
Isf would then function as a one-electron/two-electron switch.
Involvement of the obligate two-electron carrier coenzyme
F420 in the electron transport chain has been excluded
(30); however, other two-electron carriers, like methanophenazine,
cannot be ruled out. Another possibility is that the semiquinone
radical, which is indeed transiently formed, does transfer electrons to
one-electron accepting b-type cytochromes. It is plausible
that the semiquinone state in Isf may become even more stabilized when
Isf is complexed with other physiological electron donors/acceptors.
For example, in nitrate reductase, the semiquinone is highly stabilized
upon reduction by its physiological electron donor NADH although the
amount of semiquinone species formed during potentiometric titrations
is just 1% of the FAD present (40). This allows one electron transfer
from the flavin to the b-type cytochrome in nitrate
reductase. Thus, despite the instability of the flavin semiquinone in
Isf during the potentiometric titrations, it is still feasible that the
flavin mediates one-electron transfers from the Fe-S cluster in
vivo.
The H2- or CO-dependent reduction of Isf by
extracts of M. thermoautotrophicum and the occurrence of Isf
homologs in the genomes of M. thermoautotrophicum and
M. jannaschii suggest that Isf functions in
CO2-reducing methanoarchaea. Cells of M. thermoautotrophicum grow and produce CH4 with CO as
the sole energy source (41), indicating a physiological role for CODH
in the energy metabolism. Furthermore, this methanoarchaeon involves a
CODH in the synthesis of CO for incorporation into acetyl-CoA for cell
carbon (42). The CODHs from either M. thermoautotrophicum or
M. jannaschii have not been purified, and therefore, the
electron acceptor is unknown. The results presented here are consistent
with ferredoxin as the electron acceptor; however, purification of the
CODH is necessary to prove this hypothesis. The gene organization in
M. jannaschii (20) provides additional support for diverse
functions of Isf homologs in the methanoarchaea. In this organism, ORF
MJ731 has a deduced sequence 40% identical to Isf from M. thermophila and is located in a gene cluster containing an ORF
(MJ728) identified as CODH (20). The sequence of this putative enzyme
has greatest identity to the CODH from Rhodospirillum
rubrum, which catalyzes the oxidation of CO but not acetyl-CoA
synthesis (43). MJ731 and MJ728 are also located near an ORF (MJ722)
identified as an 8Fe ferredoxin (20), consistent with a role for this
electron carrier as an electron donor to Isf. However, none of the
deduced sequences of ORFs with identity to Isf in the genome of
M. thermoautotrophicum (MTH1350, MTH1474, and MTH1595) are
organized near CODH or any other redox proteins (21).
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) U50189 (corrected nucleotide sequence of isf); MTH135,
MTH1473, and MTH1595 (for M. thermoautotrophicum); and MJ1083 and
MJ10731 (for M. jannaschii).