Reversible interconversion of CO2 and formate by a molybdenum-containing formate dehydrogenase

Article abstract of DOI:10.1021/ja508647u

CO₂ and formate are rapidly, selectively, and efficiently interconverted by tungsten-containing formate dehydrogenases that surpass current synthetic catalysts. However, their mechanism of catalysis is unknown, and no tractable system is available for study. Here, we describe the catalytic properties of the molybdenum-containing formate dehydrogenase H from the model organism Escherichia coli (EcFDH-H). We use protein film voltammetry to demonstrate that EcFDH-H is a highly active, reversible electrocatalyst. In each voltammogram a single point of zero net current denotes the CO₂ reduction potential that varies with pH according to the Nernst equation. By quantifying formate production we show that electrocatalytic CO₂ reduction is specific. Our results reveal the capabilities of a Mo-containing catalyst for reversible CO₂ reduction and establish EcFDH-H as an attractive model system for mechanistic investigations and a template for the development of synthetic catalysts.

Full text of DOI:10.1021/ja508647u

Communication
pubs.acs.org/JACS
Reversible Interconversion of CO2 and Formate by a Molybdenum-
Containing Formate Dehydrogenase
Arnau Bassegoda,†,§ Christopher Madden,‡,§ David W. Wakerley,‡ Erwin Reisner,*,‡ and Judy Hirst*,†
†Medical Research Council Mitochondrial Biology Unit, Hills Road, Cambridge, CB2 0XY, United Kingdom
‡Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge, CB2 1EW, United Kingdom
S
* Supporting Information
enzyme; no genetic manipulation is possible; and the enzyme
ABSTRACT: CO2 and formate are rapidly, selectively,
contains an extensive cohort of iron−sulfur (FeS) centers to
and efficiently interconverted by tungsten-containing
formate dehydrogenases that surpass current synthetic
catalysts. However, their mechanism of catalysis is
unknown, and no tractable system is available for study.
Here, we describe the catalytic properties of the
molybdenum-containing formate dehydrogenase H from
the model organism Escherichia coli (EcFDH-H). We use
protein film voltammetry to demonstrate that EcFDH-H is
a highly active, reversible electrocatalyst. In each
voltammogram a single point of zero net current denotes
the CO2 reduction potential that varies with pH according
to the Nernst equation. By quantifying formate production
we show that electrocatalytic CO2 reduction is specific.
Our results reveal the capabilities of a Mo-containing
catalyst for reversible CO2 reduction and establish EcFDH-
H as an attractive model system for mechanistic
investigations and a template for the development of
synthetic catalysts.
transfer electrons to and from the active site, making
overexpression strategies untenable.10,12,13 Therefore, a more
versatile and robust experimental system is required.
Sf FDH1 is a member of the large class of prokaryotic
formate dehydrogenases that contain either Mo- or W-
cofactors; they also contain a second, independent active site
where quinone, protons, or NAD(P)+ react,14,15 which may be
replaced functionally by an electrode to produce an electro-
catalyst for CO2/formate interconversion. The W-containing
active site is known to be thermodynamically reversible10 but is
found exclusively in anaerobic bacteria such as Desulfovibrio16
and Eubacterium17 species. Conversely, the thermodynamic
reversibility of the more common molybdenum-containing
active site remains to be established. Several indications that
CO2 reduction is possible have been reported: a multisubunit
Mo-containing FDH from Desulfovibrio vulgaris Hildenborough
has been reported to reduce CO2 slowly in solution,18 and
production of formate from CO2 and H2 was observed in early
whole-cell experiments with Escherichia coli19 (suggesting the
formate hydrogenlyase can operate in reverse), and from a
putative Mo-containing FDH in the multisubunit H2-depend-
ent CO2 reductase of Acetobacterium woodii.20 Here, we define
the catalytic properties of the structurally defined Mo-
containing formate dehydrogenase H, a component of the
formate hydrogenlyase complex in the model organism E. coli
(EcFDH-H).21
EcFDH-H contains a Mo coordinated by a selenocysteine
(SeCys) residue and two molybdopterin guanine dinucleotides
(MGDs), and just one [4Fe-4S] cluster for electron transfer to
and from the active site (Figure 1).21 It was produced by
overexpression in E. coli under anaerobic growth conditions22,23
and purified by adapting a previously reported method24 (see
Supporting Information (SI) for details). All purification and
experimental procedures were performed under strictly
anaerobic conditions.
Figure 2 shows protein film voltammograms recorded at
different pH values with EcFDH-H adsorbed on the surface of a
graphite-epoxy rotating disk working electrode (geometric
surface area 0.07 cm2; see SI for details). EcFDH-H catalyzes
the reversible interconversion of CO2 and formate with
electrocatalytic characteristics similar to those observed
previously for Sf FDH1.10 Each set of voltammograms slices
he efficient reduction of carbon dioxide (CO2) to
Tgenerate reduced carbon compounds for use as fuels
and chemical feedstocks is an essential requirement for a
carbon-based sustainable energy economy.1 The electro-
chemical reduction of CO2, powered by carbon-neutral
electricity, would produce liquid fuels that are easier to store
and transport than hydrogen, but only limited progress has
been made in developing synthetic catalysts to overcome the
kinetic and thermodynamic challenges of CO2 activation.
Catalysts developed so far are inefficient and expensive, due to
their requirement for high overpotentials or their reliance on
noble metals.2−9
The rapid, reversible, and specific electrochemical reduction
of CO2 to formate by a tungsten-containing formate
dehydrogenase from the anaerobic bacterium Syntrophobacter
fumaroxidans (SfFDH1) provided the paradigm case for a
formate/CO2 catalyst.10 Sf FDH1 catalyzes the rapid inter-
conversion of CO2 and formate at the reduction potential for
the reaction, establishing it as a thermodynamically reversible
catalyst.11 Therefore, the catalytic mechanism of CO2 reduction
by the W-center in SfFDH1 is a valuable source of information
to aid the design of improved synthetic catalysts. However,
Sf FDH1 itself is intractable for mechanistic studies: cell
cultures of S. fumaroxidans take several months to achieve
low cell densities that provide only minuscule amounts of
Received: August 22, 2014
Published: October 17, 2014
© 2014 American Chemical Society
15473
dx.doi.org/10.1021/ja508647u | J. Am. Chem. Soc. 2014, 136, 15473−15476

Journal of the American Chemical Society
Communication
Figure 1. Structure of EcFDH-H showing the active site consisting of
Mo coordinated by a SeCys and two MGDs and the [4Fe-4S] cluster
that transports electrons to and from the active site. Figure generated
with PyMOL from 1AA6 pdb.21
Figure 3. Reduction potentials for the CO2/formate interconversion
measured using EcFDH-H, superimposed on a Pourbaix diagram to
show the most stable species present under each condition. The
reduction potentials (measured as illustrated in Figure 2) were
recorded in 10 mM formate and 10 mM CO225 at 23 °C and fitted to
the Nernst equation (solid line)10 using pKred1 = 3.75, pKox1 = 6.39,
pKox2 = 10.32,10,26 and E0′ = −0.075 V.
To confirm the selectivity of catalysis, bulk CO2 reduction
was carried out using a graphite-epoxy “pot” working electrode
(geometric surface area ∼5.3 cm2) with EcFDH-H adsorbed to
the surface. The current was recorded during electrolysis
periods of 1−2 h at −0.5 and −0.6 V vs SHE (∼110 mV and
210 mV overpotential, respectively), and the total charge
passed calculated by integration of the current over time (see
Figure S2). Following electrolysis, analysis by ion chromatog-
raphy revealed formate as the single observable product, formed
Figure 2. Cyclic voltammograms showing reversible CO2 reduction
and formate oxidation by EcFDH-H adsorbed on a graphite-epoxy
electrode, pH 6.0 (left), 6.8 (middle), and 8.0 (right). The points of
intersection (marked with crosses) define the reduction potentials for
the CO2/formate interconversion (vertical lines). First voltammetric
cycles are shown in black and subsequent cycles (2−4, 10, and 20) in
gray. Voltammograms recorded in the absence of substrates are shown
as dashed traces. Conditions: 10 mM CO2,25 10 mM formate, 25 mM
of each of four pH buffers (acetate, MES, HEPES, and TAPS),
voltammetric scan rate 25 mV s−1, electrode rotation rate 2000 rpm,
23 °C.
with a Faradaic efficiency of 101.7
2.0% (standard error
measurement, n = 3). These data confirm formate as the
quantitative product of the electrocatalytic reduction of CO2 to
formate by EcFDH-H.
Despite the Mo-containing EcFDH-H and the W-containing
Sf FDH1 being similarly capable of the electrocatalytic
interconversion of CO2 and formate, their abilities to catalyze
in solution-based assays are strikingly different. Standard FDH
solution assays utilize either benzyl viologen (BV2+) or methyl
viologen (MV2+) as the redox partner and monitor either
formation (reduction of MV2+) or consumption (oxidation of
MV+) of the blue radical-cation reduced viologen (MV+),
coupled to formate oxidation or CO2 reduction, respec-
tively.12,27 Table 1 compares solution assay and electrochemical
data from SfFDH1 and EcFDH-H.
Both EcFDH-H and Sf FDH1 catalyze the rapid oxidation of
formate coupled to the reduction of BV2+, an electron acceptor
with a reduction potential more positive than that of CO2
(−0.36 V vs SHE).28 For this reason, BV+ is unsuitable for CO2
reduction assays. Only Sf FDH1 is capable of rapid formate
oxidation coupled to the reduction of MV2+, an electron
acceptor with a lower reduction potential (−0.45 V vs SHE)28
that is comparable to that of CO2. Furthermore, solution
measurements of CO2 reduction using MV+ revealed rapid
formate production only by SfFDH1. Assays with EcFDH-H
yielded turnover numbers for CO2 reduction below 1 s−1, more
than 2 orders of magnitude slower than the rate of BV2+-linked
formate oxidation, or of the rates of both reactions catalyzed by
Sf FDH1. This result is likely the reason why CO2 reduction by
this Mo-containing FDH has not been observed previ-
cleanly through unique zero-current points, traced out as points
of intersection as the current decays during successive scans.
The zero-current points denote the thermodynamic reduction
potential for the CO2/formate interconversion (net formate
oxidation occurs at more positive potentials and net CO2
reduction at more negative potentials), demonstrating that
the electrocatalytic reaction is thermodynamically reversible.11
Both the oxidation and reduction currents increase rapidly as
the overpotential is increased but do not reach potential-
independent limiting currents within the accessible potential
range. This behavior is typical of highly catalytically active
enzymes, for which the electrocatalytic rate is limited by
interfacial electron transfer.11
Figure 3 shows how the measured CO2 reduction potentials
vary with pH, and that they are consistent with potentials
calculated using the Nernst equation and known pK values.10,26
They also match values measured previously using SfFDH1,10
with a small decrease in the predicted value of E0′ attributable
to the lower temperature used here. The data in Figure 3 are
key evidence in establishing Mo-containing EcFDH-H as a
catalyst specific for CO2/formate interconversion as well as a
demonstration that catalysis remains reversible over a wide
range of conditions.
15474
dx.doi.org/10.1021/ja508647u | J. Am. Chem. Soc. 2014, 136, 15473−15476

Journal of the American Chemical Society
Communication
are able to reduce CO2 electrochemically, but large over-
potentials are typically required and their efficiency is low.2−9 In
contrast, formate dehydrogenases catalyze the two electron
reduction of CO2 directly to energy-rich formic acid with high
selectivity, under mild conditions, with little overpotential
requirement, and elucidation of their catalytic mechanism may
inform the development of improved synthetic catalysts.
Tungsten-containing Sf FDH1 from S. fumaroxidans set an
important paradigm but is intractable for in-depth studies. Here
we have demonstrated that molybdenum-containing EcFDH-H
from E. coli is also capable of reversible, specific, and efficient
CO2 reduction, so the Mo-center is capable of reversible CO2
reduction and provides a new blueprint for synthetic catalyst
design. Based on its simplicity and relative ease of production
and manipulation, we establish EcFDH-H as a new model
system of choice for mechanistic investigations of enzymatic
CO2 reduction.
Table 1. Comparison of Catalysis by Sf FDH1 and EcFDH-H
in Solution Assays and Electrochemically (pH 7.5 0.1, 23
°C)
reaction
W-Sf FDH1
Mo-EcFDH-H
a
formate + MV2+
formate + BV2+
CO2 + MV+
1500 s−1
4 s−1
b
est. 1100 s−1
160 s−1
a
500 s−1
<1 s−1
180 μA cm−2
80 μA cm−2
c
a
formate + electrode
160 μA cm−2
5 μA cm−2
c
a
CO2 + electrode
a
b
Taken from ref 10. Value estimated from published values,12
c
supported by preliminary in-house data. 250 mV overpotential for
SfFDH1, 150 mV for EcFDH-H. Solution assays contained 1 mM
BV2+ or MV2+ for formate oxidation or 0.1 mM MV+ for CO2
reduction. All experiments used 10 mM formate or 10 mM CO2.25
ously,24,27,29 leading to the general concept that the Mo-
containing active site does not catalyze in a thermodynamically
reversible manner.
ASSOCIATED CONTENT
* Supporting Information
■
S
Electrochemically, both enzymes catalyze in both directions
with significant rates. In both cases formate oxidation increases,
relative to CO2 reduction, as the pH is increased, but at pH 7.5
EcFDH-H is biased more strongly toward CO2 reduction than
SfFDH1 (Table 1), suggesting that the “operating potential”11
of EcFDH-H is more negative than that of Sf FDH1. Although
this interesting suggestion is contrary to expectations that the
W-center should operate at a lower potential than the Mo-
center, intramolecular electron transfer to and from the active
site may also influence the catalytic bias.11
Experimental methods for protein preparation, solution assays,
and electrocatalysis experiments, and comparison with the
kinetic data of Axley and Grahame.27 This material is available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Authors
jh@mrc-mbu.cam.ac.uk
reisner@ch.cam.ac.uk
The reason why EcFDH-H, despite being such a good
electrochemical catalyst, is unable to catalyze CO2 reduction by
MV+ or formate oxidation by MV2+ is intriguing. We suggest
two possibilities. First, it may be purified in an inactive state
that cannot be recovered easily in solution-based assays.
Attempts to use different conditions and pretreatments to
reactivate the enzyme have failed to substantiate this
suggestion, but catalytic lag phases observed in formate
oxidation assays by Sf FDH1 (which can be avoided by
pretreatment with MV+) suggest that inactive states of FDH
enzymes can be formed. Second, the activity of EcFDH-H may
be dominated by the single [4Fe-4S] cluster that transfers
electrons to and from the active site. During formate oxidation
the Mo-center readily reduces the cluster, but the cluster
(having, we expect, a more positive reduction potential) is able
to pass its electron efficiently only to BV2+, not MV2+. Similarly,
for CO2 reduction, MV+ readily reduces the cluster, but the
electron tends to remain on the higher-potential cluster
(blocking further electron transfers from MV+), rather than
move to the active site. In contrast, on the electrode surface the
abundance of electrons with sufficient driving force overcomes
the FeS barrier to CO2 reduction by backfilling the oxidized
cluster immediately, when the electron moves otherwise
transiently to the active site (and similarly, for formate
oxidation it takes the electron from the cluster at the active
site potential). In contrast to EcFDH-H, SfFDH1 contains
around 10 FeS clusters10 to buffer electron supply and demand.
Our results highlight the problems of relying on inefficient and
slow redox mediators to report on catalysis by a rapidly
catalyzing, buried active site.
Author Contributions
§These authors contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by the Biotechnology and
Biological Sciences Research Council (grant nos. BB/
I026367/1 to J.H. and BB/J000124/1 to E.R.), the Medical
Research Council (grant no. U105663141 to J.H.), and the
Engineering and Physical Sciences Research Council (grant
number EP/H00338X/2 to E.R.). We thank Professor J. H.
Golbeck from Pennsylvania State University for providing E.
coli strain JG0205 and an initial expression plasmid for EcFDH-
H.
REFERENCES
■
(1) Appel, A. M.; Bercaw, J. E.; Bocarsly, A. B.; Dobbek, H.; DuBois,
D. L.; Dupuis, M.; Ferry, J. G.; Fujita, E.; Hille, R.; Kenis, P. J. A.;
Kerfeld, C. A.; Morris, R. H.; Peden, C. H. F.; Portis, A. R.; Ragsdale,
S. W.; Rauchfuss, T. B.; Reek, J. N. H.; Seefeldt, L. C.; Thauer, R. K.;
Waldrop, G. L. Chem. Rev. 2013, 113, 6621−6658.
(2) Kang, P.; Cheng, C.; Chen, Z.; Schauer, C. K.; Meyer, T. J.;
Brookhart, M. J. Am. Chem. Soc. 2012, 134, 5500−5503.
(3) Huff, C. A.; Sanford, M. S. ACS Catal. 2013, 3, 2412−2416.
(4) Hull, J. F.; Himeda, Y.; Wang, W.-H.; Hashiguchi, B.; Periana, R.;
Szalda, D. J.; Muckerman, J. T.; Fujita, E. Nat. Chem. 2012, 4, 383−
388.
(5) Kang, P.; Meyer, T. J.; Brookhart, M. Chem. Sci. 2013, 4, 3497−
3502.
(6) Costentin, C.; Drouet, S.; Robert, M.; Savea
338, 90−94.
(7) Sampson, M. D.; Nguyen, A. D.; Grice, K. A.; Moore, C. E.;
Rheingold, A. L.; Kubiak, C. P. J. Am. Chem. Soc. 2014, 136, 5460−
5471.
́
nt, J.-M. Science 2012,
The reduction of CO2 to liquid fuel products is currently of
much interest, and there is a strong requirement for catalysts
that operate efficiently, selectively and under mild conditions.
Some ruthenium, iron, manganese, and copper based catalysts
15475
dx.doi.org/10.1021/ja508647u | J. Am. Chem. Soc. 2014, 136, 15473−15476

Journal of the American Chemical Society
Communication
(8) Smieja, J. M.; Sampson, M. D.; Grice, K. A.; Benson, E. E.;
Froehlich, J. D.; Kubiak, C. P. Inorg. Chem. 2013, 52, 2484−2491.
(9) Haines, R. J.; Wittrig, R. E.; Kubiak, C. P. Inorg. Chem. 1994, 33,
4723−4728.
(10) Reda, T.; Plugge, C. M.; Abram, N. J.; Hirst, J. Proc. Natl. Acad.
Sci. U.S.A. 2008, 105, 10654−10658.
(11) Armstrong, F. A.; Hirst, J. Proc. Natl. Acad. Sci. U.S.A. 2011, 108,
14049−14054.
(12) de Bok, F. A. M.; Hagedoorn, P.; Silva, P. J.; Hagen, W. R.;
Schiltz, E.; Fritsche, K.; Stams, A. J. M. Eur. J. Biochem. 2003, 270,
2476−2485.
(13) de Bok, F. A. M.; Luijten, M. L. G. C.; Stams, A. J. M. Appl.
Environ. Microbiol. 2002, 68, 4247−4252.
(14) Hille, R.; Hall, J.; Basu, P. Chem. Rev. 2014, 114, 3963−4038.
(15) Johnson, M. K.; Rees, D. C.; Adams, M. W. W. Chem. Rev. 1996,
96, 2817−2840.
(16) Raaijmakers, H.; Macieira, S.; Dias, J. M.; Teixeira, S.; Bursakov,
S.; Huber, R.; Moura, J. J. G.; Moura, I.; Romao, M. J. Structure 2002,
̃
10, 1261−1272.
(17) Graentzdoerffer, A.; Rauh, D.; Pich, A.; Andreesen, J. R. Arch.
Microbiol. 2003, 179, 116−130.
(18) da Silva, S. M.; Pimentel, C.; Valente, F. M. A.; Rodrigues-
Pousada, C.; Pereira, I. A. C. J. Bacteriol. 2011, 193, 2909−2916.
(19) Woods, D. D. Biochem. J. 1936, 30, 515−527.
(20) Schuchmann, K.; Muller, V. Science 2013, 342, 1382−1385.
̈
(21) Boyington, J. C.; Gladyshev, V. N.; Khangulov, S. V.; Stadtman,
T. C.; Sun, P. D. Science 1997, 275, 1305−1308.
(22) Zhang, J. W.; Butland, G.; Greenblatt, J. F.; Emili, A.; Zamble,
D. B. J. Biol. Chem. 2005, 280, 4360−4366.
(23) Hopper, S.; Babst, M.; Schlensong, V.; Fischer, H.-M.;
Hennecke, H.; Bock, A. J. Biol. Chem. 1994, 269, 19597−19604.
̈
(24) Axley, M. J.; Grahame, D. A.; Stadtman, T. C. J. Biol. Chem.
1990, 265, 18213−18218.
(25) Note that we refer to ‘CO2 reduction’ but (due to the very
nonideal behavior of CO2 as a solute) have defined the total
concentrations of CO2 and carbonate species present by addition of
sodium carbonate.
(26) Palmer, D. A.; Van Eldik, R. Chem. Rev. 1983, 83, 651−731.
(27) Axley, M. J.; Grahame, D. A. J. Biol. Chem. 1991, 266, 13731−
13736.
(28) Wardman, P. J. Phys. Chem. Ref. Data 1989, 18, 1637−1755.
(29) Enoch, H. G.; Lester, R. L. J. Biol. Chem. 1975, 250, 6693−6705.
15476
dx.doi.org/10.1021/ja508647u | J. Am. Chem. Soc. 2014, 136, 15473−15476