Reduction of dioxygen catalyzed by pyrene-wired heme domain cytochrome P450 BM3 electrodes

Article abstract of DOI:10.1021/ja0466560

We have electronically wired the cytochrome P450 BM3 heme domain to a graphite electrode with the use of a pyrene-terminated tether. AFM images clearly reveal that pyrene-wired enzyme molecules are adsorbed to the electrode surface. The enzyme-electrode system undergoes rapid and reversible electron transfer, displaying a standard rate constant higher than that of any other P450-electrode system. We also show that the graphite-pyrene-BM3 system catalyzes the four-electron reduction of dioxygen to water. Copyright

Full text of DOI:10.1021/ja0466560

Published on Web 08/03/2004
Reduction of Dioxygen Catalyzed by Pyrene-Wired Heme Domain Cytochrome
P450 BM3 Electrodes
Andrew K. Udit,† Michael G. Hill,‡ V. Garrett Bittner,† Frances H. Arnold,† and Harry B. Gray*,†
DiVision of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California 91125,
and Department of Chemistry, Occidental College, Los Angeles, California 90041
Received June 7, 2004; E-mail: hbgray@caltech.edu
Cytochromes P450 (P450s) catalyze oxygenations of inert
substrates under physiological conditions.1 Exploiting this activity
in vitro would be greatly facilitated if reductants other than NADPH
could be found. While an electrode is perhaps the most attractive
source of electrons, direct electrochemistry of P450 has been
elusive, owing to poor electronic coupling to the deeply buried heme
and inactivation through surface adsorption. Investigations of
electrochemical reduction of P450 have led to clever techniques
for effecting electron transfer (ET). These methods include confin-
ing the protein within surfactant2,3 or polyelectrolyte4,5 films,
Figure 1. Cyclic voltammogram of the Py-hBM3 conjugate on BPG (0.07
modifying the electrode surface covalently (mercaptans on gold6)
or through adsorption (clay on carbon7) and modifying the enzyme
with molecular electronic relays.8
cm2) at 200 mV/s in 50 mM KPi/20 mM KCl/pH 7.
We are working on electrochemical methods for reduction of
the fatty acid hydroxylase flavocytochrome P450 from Bacillus
megaterium (BM3). BM3’s high turnover rates and broad substrate
specificity have stimulated interest in developing a bioelectronic
system for this enzyme, including studies of both fundamental ET
and biocatalysis.9,10 Notably, Sevrioukova et al. achieved rapid heme
reduction photochemically (2.5 × 106 s-1 and 4.6 × 105 s-1 with
and without substrate) by covalently tethering a ruthenium diimine
to an engineered cysteine (N387C) on the heme domain of BM3
(hBM3).11 The positioning of the Ru complex was meant to mimic
the interaction between hBM3 and its reductase: indeed, the rapid
rates suggest that the complex was attached at a position that was
well coupled to the heme. It occurred to us that “wiring” the N387C
hBM3 mutant to an electrode through the engineered cysteine could
also yield high electron tunneling rates. Previously, Katz utilized
N-(1-pyrene)iodoacetamide (Py) (thiol specific) to anchor and
electronically connect a photosynthetic reaction center to a basal
plane graphite (BPG) electrode.12 Thus, we made the hBM3 single
surface cysteine mutant at position 387, attached Py to the cysteine,
and successfully achieved rapid ET with the use of a BPG electrode.
Wiring the enzyme in this way creates a system where the electrode
mimics the reductase, leaving the active site accessible to molecules
in solution.
Protein integrity after labeling with Py was confirmed by
observing the Soret band of the reduced heme at 448 nm in CO-
saturated buffer. Labeled protein (Py-hBM3) was verified by
observing fluorescence of Py (∼50% labeled). The protein film
was prepared by suspending a BPG electrode in a ∼20 µM Py-
hBM3 solution. Cyclic voltammetry on the resulting film (Figure
1) revealed a couple centered at -340 mV. Neither unmodified
enzyme nor Py alone produced a similar couple. The observed E1/2
is assigned to the FeIII/II redox couple of the heme.5,7 Compared to
the native enzyme in its resting state (six-coordinate heme, low
spin) as measured by redox titration, this potential is shifted
approximately +230 mV.9 As previously suggested, local electro-
Figure 2. AFM images (800 nm × 800 nm) of HOPG soaked in (a) Py-
hBM3 and (b) hBM3.
static effects (e.g., solvation, surface interactions) likely contribute
to the altered potential.5
The cathodic to anodic peak-current ratio in Figure 1 is
approximately 1.05, indicating a chemically reversible system.13
A plot of the cathodic peak current versus the scan rate is linear,
characteristic of a surface-confined species.13 This plot also indicates
the number of electrons transferred: the slope of the line divided
by the area under the voltammogram at any sweep rate (39 nC) is
equal to nF/4RT.14 Performing this operation yields n ) 1.2 ( 0.1,
fully consistent with one-electron transfer.
Voltammetry in CO-saturated buffer shifted E1/2 by +35 mV,
as found for other P450 electrochemical systems (+45 to +80
mV).2,3,7 The E1/2 was also found to vary linearly with pH according
to E1/2 ) 56 mV - 58 mV/pH, indicating proton-coupled electron
transfer.4,15
To characterize the surface, protein films were cast onto highly
oriented pyrolytic graphite (HOPG) and imaged using atomic force
microscopy (AFM) in tapping mode. Figure 2a shows a section of
HOPG (800 nm × 800 nm) soaked in a Py-hBM3 solution,
revealing a series of small islands (dark spots) ranging from 2 to 5
nm in height. Given that hBM3 is ∼65 Å in diameter, it can be
inferred that these islands represent protein clusters on the surface.
Figure 2b shows the corresponding image of HOPG soaked in
unlabeled hBM3. Clearly, no surface features are visible; this image
is identical with HOPG soaked in buffer alone and implies that
only the Py-hBM3 conjugate adsorbs to the surface. Regarding
† California Institute of Technology.
‡ Occidental College.
9
10218
J. AM. CHEM. SOC. 2004, 126, 10218-10219
10.1021/ja0466560 CCC: $27.50 © 2004 American Chemical Society

C O MMU N I C A T I O N S
dioxygen primarily to water. Our proposal of an efficient four-
electron reduction pathway is further supported by results from an
Amplex Red fluorescence assay for hydrogen peroxide, which
revealed that only a small fraction of the current (<17%) was used
to generate the two-electron reduction product. This is in stark
contrast to other P450 electrochemical systems, where peroxide is
the primary product of dioxygen reduction.2,4 Conceivably, dioxygen
reduction to water can occur if ET is fast enough such that, after
initial reaction to form a peroxy complex, the final two electrons
arrive at the active site before peroxide dissociation. Precedent for
this can be found in previous work with ruthenium-modified cobalt
porphyrins:19 π-back-bonding by the ruthenium ligands increased
the ET rate, creating a catalyst that reduced dioxygen primarily to
water. For the BPG-Py-hBM3 system, the estimated k° is so high
that applying a potential of -0.5 V apparently leads to rapid
reduction of dioxygen to water.
Figure 3. Cyclic voltammograms at 200 mV/s of Py-hBM3 on BPG in
the presence of increasing amounts of dioxygen: (a) 0, (b) 42, (c) 71, and
(d) 94 µM.
Acknowledgment. We thank J. S. Magyar and J. R. Winkler
(Caltech), T. L. Poulos (U.C. Irvine), and E. M. Spain (Occidental
College) for helpful discussions; C. P. Collier (Caltech) for
assistance with the AFM; NSF (H.B.G.), NSERC (Canada)
(A.K.U.), and David and Lucile Packard Foundation (M.G.H.) for
research support.
Figure 4. Solid lines represent Levich plots derived for the one-, two-,
and four-electron reduction of dioxygen. The points represent the limiting
current at 400, 600, and 700 rpm for Py-hBM3 films on BPG-RDE in the
presence of dioxygen (250 µM).
Supporting Information Available: Details of electrode prepara-
tion, mutagenesis, protein purification, protein labeling, electron-transfer
rate calculations, Amplex Red peroxide assay, voltammetry, and RDE
methods. This material is available free of charge via the Internet at
http://pubs.acs.org.
surface coverage, Figure 2a suggests that there is submonolayer
coverage. Cyclic voltammetry experiments on HOPG (0.25 cm2)
with a Py-hBM3 film (hBM3 monolayer ) 1.4 × 10-12 mol)
confirm this finding: integrating under the cathodic peak yielded
6.2 × 10-13 mol of electroactive protein, or ∼44% surface coverage.
The standard rate constant (k°, ∆G° ) 0) for the BPG-Py-hBM3
system was evaluated using Laviron’s theory,16 yielding a value of
650 ( 50 s-1, which is the fastest electrode kinetics reported for
any P450 system (cf. 221 s-1 for hBM3 in DDAB films).2 ET rates
for photochemical reduction of the FeIII heme11 were used to
estimate rates at zero driving force (∆G° ) 0) for substrate-free
and substrate-bound hBM3: these were found to be 280 and 3300
References
(1) Cytochrome P450: Structure, Mechanism, and Biochemistry, 2nd ed.;
Ortiz de Montellano, P. R., Ed.; Plenum Press: New York, 1995.
(2) Fleming, B. D.; Tian, Y.; Bell, S. G.; Wong, L.; Urlacher, V.; Hill, H. A.
O. Eur. J. Biochem. 2003, 270, 4082-4088.
(3) Zhang, Z.; Nassar, A.-E.; Lu, Z.; Schenkman, J. B.; Rusling, J. F. J. Chem.
Soc., Faraday Trans. 1997, 93, 1769-1774.
(4) Munge, B.; Estavillo, C.; Schenkman, J. B.; Rusling, J. F. ChemBioChem
2003, 4, 82-89.
(5) Lvov, Y. M.; Lu, Z.; Schenkman, J. B.; Zu, X.; Rusling, J. F. J. Am.
Chem. Soc. 1998, 120, 4073-4080.
(6) Fantuzzi, A.; Fairhead, M.; Gilardi, G. J. Am. Chem. Soc. 2004, 126,
5040-5041.
s-1 17
As can be seen, k° from our electrochemical experiments
.
(7) Lei, C.; Wollenberger, U.; Jung, C.; Scheller, F. W. Biochem. Biophys.
Res. Commun. 2000, 268, 740-744.
falls within this range.
(8) Shumyantseva, V.; Bulko, T.; Bachmann, T.; Bilitewski, U.; Schmid, R.;
Archakov, A. Arch. Biochem. Biophys. 2000, 377, 43-48.
(9) Daff, S.; Chapman, S.; Turner, K.; Holt, R.; Govindaraj, S.; Poulos, T.;
Munro, A. Biochemistry 1997, 36, 13816-13823.
The BPG-Py-hBM3 system is an excellent catalyst for dioxygen
reduction (Figure 3). Negative controls (Py or hBM3 on BPG only)
reveal slow dioxygen reduction at more negative potentials. Possible
fates of dioxygen are reduction to peroxide (directly or through
superoxide decomposition) or to water. To determine the number
of electrons transferred to dioxygen, Py-hBM3 films were cast onto
a BPG rotated-disk electrode (RDE). Using the Levich equation
for a RDE,13
(10) Schneider, S.; Wubbolts, M.; Sanglard, D.; Witholt, B. Appl. EnViron.
Microbiol. 1998, 64, 3784-3790.
(11) Sevrioukova, I. F.; Immoos, C. E.; Poulos, T. L.; Farmer, P. Isr. J. Chem.
2000, 40, 47-53.
(12) Katz, E. J. Electroanal. Chem. 1994, 365, 157-164.
(13) Bard, A. J.; Faulkner, L. R. Electrochemical Methods; 2nd ed.; John Wiley
& Sons: New York, 2001.
(14) Brown, A. P.; Koval, C.; Anson, F. C. J. Electroanal. Chem. 1976, 72,
379-387.
(15) Aguey-Zinsou, K.; Bernhardt, P. V.; Voss, J. J. D.; Slessor, K. E. Chem.
Commun. 2003, 418-419.
iL ) 0.62nFADo2/3ω1/2v-1/6
C
(16) Laviron, E. J. Electroanal. Chem. 1979, 101, 19-28.
(17) See Supporting Information for details on the calculation.
(18) The variables for air saturated water are: limiting current (iL, A), electrode
area (A, cm2), diffusion coefficient (Do, 1.7 × 10-5 cm2/s for O2 in water),
angular velocity (ω, s-1), kinematic viscosity (v, 0.01 cm2/s for water),
bulk concentration (C, 2.5 × 10-7 mol/cm3 for O2 in air saturated buffer
at 25 °C).
theoretical lines for the one-, two-, and four-electron reduction of
dioxygen were generated (Figure 4).18 RDE experiments were
conducted by performing electrolysis at -0.5 V and determining
the limiting current for each rotation rate. The results of these
experiments (solid points, Figure 4) scatter around the theoretical
line for n ) 4, suggesting that the BPG-Py-hBM3 system converts
(19) Anson, F. C.; Shi, C.; Steiger, B. Acc. Chem. Res. 1997, 30, 437-444.
JA0466560
9
J. AM. CHEM. SOC. VOL. 126, NO. 33, 2004 10219