Xanthine dehydrogenase electrocatalysis: Autocatalysis and novel activity

Article abstract of DOI:10.1021/jp111809f

The enzyme xanthine dehydrogenase (XDH) from the purple photosynthetic bacterium Rhodobacter capsulatus catalyzes the oxidation of hypoxanthine to xanthine and xanthine to uric acid as part of purine metabolism. The native electron acceptor is NAD⁺ but herein we show that uric acid in its 2-electron oxidized form is able to act as an artificial electron acceptor from XDH in an electrochemically driven catalytic system. Hypoxanthine oxidation is also observed with the novel production of uric acid in a series of two consecutive 2-electron oxidation reactions via xanthine. XDH exhibits native activity in terms of its pH optimum and inhibition by allopurinol.

Full text of DOI:10.1021/jp111809f

ARTICLE
pubs.acs.org/JPCB
Xanthine Dehydrogenase Electrocatalysis: Autocatalysis and Novel
Activity
Palraj Kalimuthu,† Silke Leimk€uhler,‡ and Paul V. Bernhardt†,*
†Centre for Metals in Biology, School of Chemistry and Molecular Biosciences, University of Queensland, Brisbane, 4072, Australia and
‡Institut f€ur Biochemie und Biologie Universit€at Potsdam, 14476 Potsdam, Germany
S Supporting Information
b
ABSTRACT: The enzyme xanthine dehydrogenase (XDH) from the purple photo-
synthetic bacterium Rhodobacter capsulatus catalyzes the oxidation of hypoxanthine to
xanthine and xanthine to uric acid as part of purine metabolism. The native electron
acceptor is NADþ but herein we show that uric acid in its 2-electron oxidized form is
able to act as an artificial electron acceptor from XDH in an electrochemically driven
catalytic system. Hypoxanthine oxidation is also observed with the novel production of
uric acid in a series of two consecutive 2-electron oxidation reactions via xanthine.
XDH exhibits native activity in terms of its pH optimum and inhibition by allopurinol.
-
3
’ INTRODUCTION
Like many other oxidase enzymes, H2O2 and O2 are pro-
ducts of substrate turnover in XO when dioxygen is the cosub-
strate.10 The electroactivity of H2O2 enables its voltammetric
detection and provides a method for monitoring XO turnover
without requiring direct or mediated electron transfer with the
enzyme itself. A wide variety of electrode systems have been
described that utilize immobilized xanthine oxidase to produce
H2O2 as an electrochemically detectable product or alternatively
to monitor the depletion of cosubstrate dioxygen.10-12
Xanthine dehydrogenase (XDH) and xanthine oxidase (XO)
are complex molybdo/iron-sulfur/flavoproteins that catalyze
the hydroxylation of purines, pyrimidines, pterins, and aldehyde
substrates using NADþ (XDH) or molecular oxygen (XO) as
electron acceptors, respectively.1-3 XDH and XO are intercon-
vertible in mammalian systems including human, bovine, and
rat.3 The primary substrate is xanthine, which is oxidized to uric
acid in a two-electron, O-atom transfer reaction. In humans,
XDH/XO activity is associated with a number of medical condi-
tions including xanthinuria, hyperuricemia, and ischemic reper-
fusion injury.4 Xanthine dehydrogenase inhibitors are used in the
treatment of gout.
All XDH/XO enzymes contain an active site comprising a Mo
ions coordinated to a bidentate molybdopterin dithiolene ligand,
an equatorial terminal sulfido, an axial oxido, and an equa-
torial hydroxo/aqua ligand depending on pH. The mechanism
of conversion of hypoxanthine to xanthine or xanthine to uric
acid at the Mo active site is illustrated in Scheme 1. A glutamate
residue (E730, not shown) accepts the proton released by the
nucleophilic hydroxido ligand in the first step.
A better approach is to employ electrochemically active elec-
tron transfer mediators (instead of dioxygen) to accept electrons
from XO/XDH or to achieve direct (unmediated) electron
transfer with the enzyme. Many groups have claimed to have
achieved direct electrochemistry with XO/XDH13-18 but in each
case there has been doubt whether the native catalytically active
form of the enzyme has been maintained or whether direct
electron transfer has actually been achieved with the enzyme.
The FAD cofactor is noncovalently bonded to XO/XDH and
may dissociate from the enzyme to give free FAD (which is also
electroactive). A thorough and careful study revealed denaturation
of bovine XO at a mercury electrode coupled with dissociation of the
FAD cofactor.13 Dissociation of FAD should deactivate XO/XDH
so proof of catalytic activity is important but is often overlooked.14,15
Of equal concern are reports of XO/XDH catalytic electrochemistry
with unnatural (and unprecedented) substrates,17,19,20 which again
Three additional redox cofactors are present in all XDH/XO
enzymes comprising two spectroscopically distinct [2Fe-2S]
clusters (FeSI and FeSII) and a flavin adenine dinucleotide (FAD)
cofactor as shown in the crystal structure of the Rhodobacter
capsulatus XDH heterotetramer (Figure 1).1,5,6 Reactivation of
XDH/XO is achieved by a two electron oxidation at the FAD site
by NADþ (or O2) with the intervening Fe-S clusters acting as
electron relays (Figure 1).7-9
Received: December 13, 2010
Revised:
January 28, 2011
Published: March 01, 2011
r
2011 American Chemical Society
2655
dx.doi.org/10.1021/jp111809f J. Phys. Chem. B 2011, 115, 2655–2662
|

The Journal of Physical Chemistry B
ARTICLE
Scheme 1
Figure 1. Crystal structure of xanthine dehydrogenase5 (R. capsulatus) with its cofactors in space filling representation and the only the peptide
backbone of the protein shown. The substrate (xanthine) and cosubstrate (NADþ) are also included at their sites of reaction.
cast doubt on whether a native and functional enzyme is being
studied.
We have had some success in using cyclic voltammetry in
identifying nonturnover redox responses from the Mo, FAD, and
2656
dx.doi.org/10.1021/jp111809f |J. Phys. Chem. B 2011, 115, 2655–2662

The Journal of Physical Chemistry B
ARTICLE
Scheme 2
Fe-S cofactors of XDH18 and XO16 in the absence of substrate
but catalytic voltammetry (in the presence of xanthine) has been
unusual in both cases. The most curious observation is the
catalytic electrochemical potential (ca. þ200 mV vs Ag/AgCl),
which is more than 600 mV higher than the highest potential
cofactor in the enzyme. Normally, the catalytic potential is in the
vicinity of the redox potential of the cofactor in communication
with the electrode.
The purine substrates and products of XDH/XO, hypox-
anthine, xanthine, and uric acid are related by 2-electron, 2-pro-
ton, O-atom transfer reactions so each is electrochemically
active.21 The approximate electrochemically determined redox
potentials at pH 8 of these purines (from various studies)21-26
are shown in Scheme 2 in addition to related half reactions of the
clinically important xanthine oxidase inhibitor allopurinol (an
isomer of hypoxanthine). In each case, the oxidation of each
molecule is electrochemically irreversible or quasi-reversible and
requires very high overpotentials relative to the true thermo-
dynamic redox potentials.
Of most significance to this study, the product of XDH/XO
activity, uric acid, is itself able to be oxidized to a transiently stable
imine that is electrochemically detectable.26 There have been
many reports of uric acid electrochemistry using a variety of
working electrodes.22,27-35 Uric acid imine undergoes a slower
decomposition ultimately to allantoin at neutral pH (not shown
here).26 In this article, we reconcile previously unexplained elec-
trochemical behavior of XDH by showing that the uric acid
imine, on the voltammetric time scale, is capable of acting as an
NADþ mimic in accepting electrons from XDH following turn-
over in an unusual enzyme electro-autocatalytic system.
Electrochemical Measurements and Electrode Cleaning.
Cyclic voltammetry (CV) was carried out with a BAS 100B/W
electrochemical workstation using three electrode systems con-
sisting of an edge-plane pyrolytic graphite (EPG) working
electrode, a platinum wire counter electrode, and a Ag/AgCl
reference electrode. Experiments were carried out in Ar purged
solutions. The EPG electrode surface was cleaned by cleaving
several 1 μm layers from the face of the electrode with a
microtome followed by sonication in Milli-Q water. No abrasives
were used. No other electrode surface conditioning was neces-
sary and no promoters were used.
Enzyme Electrode Preparation. A 5 μL droplet of wild-type
XDH (77 μM) in 50 mM Tris buffer (pH 8.0) was carefully
deposited on a freshly prepared, inverted EPG electrode and this
was allowed to dry to a film at 4 °C. To prevent protein loss to the
bulk solution the electrode surface was carefully covered with a
permselective dialysis membrane presoaked in water. The dia-
lysis membrane (ca. 1 cm2) was carefully placed over the elec-
trode surface and fastened with a rubber O-ring to prevent
leakage of the internal membrane solution. The resulting mod-
ified electrode was stored at 4 °C in 50 mM Tris buffer solution
(pH 8.0) when not in use. The pH dependence of the catalytic
current was modeled by eq 1,36 which is applicable for an active
form of the enzyme that is deactivated reversibly by either a
deprotonation (at pKa1) or protonation (at pKa2).
iopt
ilimðpHÞ ¼
ð1Þ
1 þ 10ðpH - pK Þ þ 10ðpK
- pHÞ
a1
a2
’ RESULTS AND DISCUSSION
Enzymatic Xanthine Oxidation. No faradaic response is seen
from the EPG/XDH working electrode over the potential range
-200 to þ500 mV in the absence of xanthine. In previous
investigations of XDH18 and XO16 in the absence of xanthine,
weak ‘nonturnover’ responses emerge from the enzyme cofactors
(Mo active site, two Fe-S clusters and FAD) at lower potentials
in the range -750 to -500 mV versus Ag/AgCl at pH 8. These
signals are not consistently observed from past experience. The
FAD signal is the most prominent (-550 mV versus Ag/AgCl)
and this is also shown in Figure S1 of the Supporting Informa-
tion. The nature of this redox couple is explored in greater detail
later. Regardless, xanthine is electroinactive within the potential
range shown in part A of Figure 2 (Figure S2 of the Supporting
Information).22,25,27,34,37
’ EXPERIMENTAL SECTION
Materials. R. capsulatus XDH was purified from a heteroge-
neous expression system in Escherichia coli as previously described.6
Xanthine, hypoxanthine, uric acid and allopurinol were
purchased from Aldrich and were used as received. All other
reagents used were of analytical grade purity and used without
any further purification. All solutions were prepared in deionized
water (Millipore, resistivity 18.2 MΩ.cm). Tris (acetate) buffer
(50 mM, pH 8) was the supporting electrolyte. For experiments
conducted in the pH range of 5 to 10, a buffer mixture containing
both bis-tris propane and 2-amino-2-methylpropan-1-ol
(50 mM) was used as supporting electrolyte, titrated with acetic
acid to give the desired pH.
2657
dx.doi.org/10.1021/jp111809f |J. Phys. Chem. B 2011, 115, 2655–2662

The Journal of Physical Chemistry B
ARTICLE
Figure 2. (A) Cyclic voltammetry of the EPG/XDH electrode (sweep rate 5 mV s-1, pH 8) in the presence of: xanthine concentrations of (a) 20, (b)
40, (c) 60, (d) 80, (e) 100, (f) 120, (g) 140, (h) 160, (i) 180, (j) 200, (k) 400 and (l) 800 μM; and (B) currents (baseline subtracted) at þ200 mV
(squares) and þ250 mV (circles) vs Ag/AgCl.
Figure 3. (A) Cyclic voltammetry of the EPG/XDH electrode (sweep rate 5 mV s-1, pH 8) in the presence of uric acid concentrations of (a) 20, (b) 40,
(c) 60, (d) 80, (e) 100, (f) 120, (g) 140, (h) 160, (i) 180, (j) 200, (k) 300, (l) 400 and (m) 800 μM; and (B) currents (baseline subtracted) at þ200 mV
(squares) and þ250 mV (circles) vs Ag/AgCl. Note different vertical axis scales compared with Figure 1.
Upon addition of xanthine to the electrochemical cell (buf-
fered at pH 8), a sharp anodic peak appears at ca. þ200 mV
followed by a tailing wave around þ250 mV (part A of Figure 2)
and these currents increase with the concentration of xanthine
(part B of Figure 2). The sharp anodic peak at þ200 mV is
indicative of oxidation of an adsorbed molecule38 and its
potential coincides with oxidation of EPG-adsorbed uric acid at
pH 8.22,23,27,39 The higher potential tailing wave is characteristic
of a diffusion limited process38 and resembles the voltammetry of
uric acid at working electrodes where adsorption is suppressed
such as glassy carbon or boron-doped diamond.40,41
Proof that the response is of uric acid was obtained by cyclic
voltammetry at the EPG/XDH electrode in the presence of
increasing concentrations of uric acid but without xanthine (part
A of Figure 3). Under these conditions, no catalysis is possible,
instead direct uric acid electrochemistry is observed at a compar-
able electrode surface (as seen in part A of Figure 2). Although a
similar profile is obtained, there are some important differences
between the data in part A of Figure 2 and part A of Figure 3.
The presence of XDH adsorbed on the EPG electrode has
an important influence on the waveform. In its absence, the
voltammetry of uric acid alone is simpler (Figure S3 of the
Supporting Information) and comprises a single anodic peak of
the adsorbed compound. Evidently, XDH acts to block some of
surface sites from uric acid adsorption and a mixture of diffusional
and adsorbed voltammetry is seen.
The origin of the uric acid oxidation peaks in part A of Figure 2
must be from XDH catalysis. However, to account for the
continual increase in uric acid concentration (as a function of
added xanthine), an electron acceptor must be present to account
for the electrons generated by xanthine oxidation. There are two
notable differences between the CVs in part A of Figure 2
(xanthine addition) and part A of Figure 3 (uric acid addition).
First, in part A of Figure 2 the magnitude of the higher potential
diffusional wave increases substantially with the concentration of
xanthine. Second the magnitude of the peak-shaped adsorption
prewave at 200 mV in part A of Figure 2 appears to reach a
plateau at ca. 200 μM xanthine. Accurate determination of the
ratio of the two waves is complicated by the proximity of the two
waves and indeed the 200 mV anodic peak is enveloped by the
higher potential wave when the xanthine concentration exceeds
400 μM.
2658
dx.doi.org/10.1021/jp111809f |J. Phys. Chem. B 2011, 115, 2655–2662

The Journal of Physical Chemistry B
ARTICLE
Scheme 3
Figure 4. Cyclic voltammograms (sweep rate 5 mV s-1, pH 8)
measured with the same EPG/XDH working electrode used in part A
of Figure 3, which had been transferred to a fresh 50 mM Tris buffer
solution (no xanthine) (a) 1st cycle and (b) 2nd cycle.
To deconvolute the adsorptive and diffusional components of
the uric acid voltammetry in Figure 2 the EPG/XDH electrode
was removed from the electrochemical cell after the experiment
and transferred to a cell containing only Tris buffer solution (no
xanthine). A well-defined oxidation peak was observed at þ200
mV in the first cycle but no following tailing wave (curve a of
Figure 4). In the second cycle, this peak was absent altogether
(curve b of Figure 4) confirming that (i) the þ200 mV peak is
due to adsorbed uric acid and (ii) the tailing wave at þ250 mV
(absent from both cycles) is linked with a catalytic process
requiring the presence of xanthine (as a source of uric acid)
and XDH.
For the direct uric acid oxidation experiment (Figure 3), the
peak currents at þ200 mV and þ250 mV increase with uric acid
concentration but at about the same rate i.e. the ratio of the peak
currents is always close to 1:1 throughout. This is in contrast to
part B of Figure 2 where the þ250 mV wave grows markedly
in intensity relative to the þ200 mV peak beyond about 100 μM
xanthine. This is further evidence that the tailing wave at þ250 mV
is linked to catalysis. Second the uric acid anodic current in the
presence of 800 μM uric acid is 5 times smaller than in the pre-
sence of 800 μM xanthine. Recalling that uric acid voltammetry is
being observed in both cases, this is a remarkable observation.
The XDH/xanthine (800 μM) reaction is producing a higher
than 800 μM local concentration of uric acid at the electrode
surface.
The observation of an amplified uric acid oxidation wave upon
sequential addition of xanthine (Figure 2) reflects an autocata-
lytic mechanism (Scheme 3). Uric acid generated enzymatically
is oxidized at the electrode, producing uric acid imine in the
diffusion layer. Uric acid imine, a strong oxidant, is immediately
reduced by XDHred (mimicking the native electron acceptor
NADþ) thus regenerating XDHox for another turnover. In other
words both forms of the enzyme (XDHred and XDHox) react to
produce uric acid whose concentration at the electrode rises
nonlinearly. To initiate the electrocatalytic reaction, only a small
amount of adsorbed uric acid is required and this may be gen-
erated by one or two nonelectrochemical XDH turnovers (three
molecules of uric acid are able to be produced from the six-
electron oxidized enzyme). Uric acid then adsorbs to the graphite
electrode where it is electrochemically oxidized (the sharp peak
at þ200 mV). From this point, the uric acid imine (in solution)
Figure 5. pH Dependence of the electrocatalytic anodic peak current
(baseline corrected) in the presence of 100 μM xanthine at EPG/XDH
electrode at a scan rate of 5 mV s-1. The solid line is the fit of the
experimental data to eq 1 with pKa1 = 6.1 and pKa2 = 8.1.
mediates catalysis. A cascade of uric acid is produced through two
pathways (XDHred þ uric acid imine and XDHox þ xanthine) so
the reaction does not follow conventional Michaelis Menten
kinetics.
pH Dependence and Catalytic Inhibition of XDH. Catalysis
of a natively functioning enzyme should be subject to the same
pH dependence and inhibitor effects as seen in solution chemical
assays. In Figure 5, the catalytic peak current each pH value was
taken as a measure of activity with all other conditions the same.
The bell-shaped pH profile was modeled with eq 1 where pro-
tonation (pKa 6.1) of an essential proton accepting glutamate
residue (E730) near the active site or deprotonation (pKa 8.1)
of xanthine lead to a loss of activity. XDH exhibits an optimal
activity of ∼pH 7, which is consistent with solution assays.6
It is well-known that allopurinol, a XDH/XO inhibitor, is
oxidized to oxypurinol (Scheme 2), which then binds irreversibly
to the MoIV form of XDH thus blocking access by other sub-
strates.5,42 Upon addition of 100 μM xanthine to the electro-
chemical cell, the EPG/XDH electrode produces the characteristic
combination of a sharp peak at þ200 mV and tailing diffusional
2659
dx.doi.org/10.1021/jp111809f |J. Phys. Chem. B 2011, 115, 2655–2662

The Journal of Physical Chemistry B
ARTICLE
(parts a-e of Figure 7) until the hypoxanthine concentration
reaches about 100 μM (parts f-j of Figure 7).
A most important mechanistic difference here is that the
product of the XDH/hypoxanthine reaction (xanthine) is elec-
tro-inactive within the potential range studied (Figure S2 of the
Supporting Information) and is unable to mediate catalysis.
Notwithstanding, uric acid is generated in this experiment as
evident from a catalytic current in the region þ200-400 mV
versus Ag/AgCl; this demands two sequential XDH-catalyzed
2-electron oxidations (Scheme 2) on the same original hypox-
anthine bound at the active site to produce uric acid. The
sequential oxidation of hypoxanthine to xanthine and then xan-
thine to uric acid within the same XDH assay is atypical. In
solution enzyme assays, the high dilution conditions (particularly
of enzyme and product) means that re-entry of the product
(xanthine) is uncompetitive with binding of substrate (hypo-
xanthine present in large excess) so the enzyme reaction never
proceeds further than the first oxidation step in the assay. In the
electrochemical system, XDH, hypoxanthine, and xanthine are
concentrated within the reaction layer under the membrane facil-
itating bimolecular reactions between XDHox and xanthine.
Looking at the structures in Scheme 2 and the mechanism in
Scheme 1, it is evident that the substrate (hypoxanthine or
xanthine) undergoes hydroxylation at two different positions and
must bind next to the Mo active site in a different orientation. So
the initial xanthine product remains in proximity to the active
site, rotates by almost 180° to be positioned for a second hydrox-
ylation at C8 once the MoVI ion is restored electrochemically.
From that point onward the uric acid concentration slowly rises
(parts a-e of Figure 7) to the point where there is enough
present for autocatalysis to take over and the current then rises
rapidly (parts f-j of Figure 7) comprising both the XDH/
hypoxanthine and the XDH/xanthine reactions. This is an even
more complicated mechanism than that shown in Scheme 3 but it
follows the same principles, albeit with two sequential oxidation
reactions occurring at the active site.
Figure 6. Cyclic voltammetry (sweep rate 5 mV s-1, pH 8) of EPG/
XDH (a) in the presence of 100 μM xanthine and (b) 30 min after
addition of 2 mM allopurinol to the same solution.
Figure 7. Cyclic voltammetry of the EPG/XDH electrode (5 mV s-1
,
Nature of the EPG-XDH interaction. It is certain that the
catalytically active form of XDH in this study is weakly and
reversibly adsorbed to the EPG electrode. Conducting parallel
experiments without a membrane to entrap XDH led to similar
catalytic currents as seen in the parallel experiments carried out
with a membrane. However, the catalytic current was lost (Figure
S5 of the Supporting Information) at high concentrations of
xanthine (>600 μM) or hypoxanthine (>200 μM). Evidently the
two substrates (which both adsorb to the EPG electrode)
compete for surface sites on the electrode and ultimately displace
the enzyme altogether from the electrode and into the bulk
solution; effectively an irreversible process given the miniscule
amounts of enzyme (pmols) concerned. Therefore, the mem-
brane is critical and ensures that if enzyme desorption occurs it is
reversible; XDH is still trapped within the small volume close to
the electrode and can readsorb if the competitively adsorbing
substrates (hypoxanthine or xanthine, both membrane perme-
able) are removed by dialysis.
Nonturnover Redox Responses from the Cofactors. The
observation of nonturnover responses from XDH in the regions
expected for the various cofactors18 indicates that direct electron
transfer with the enzyme is possible yet none of the redox re-
sponses assigned to the cofactors is amplified in the presence of
substrate. The most prominent nonturnover response is that of
the putative FAD cofactor. This response is consistently ob-
served in experiments conducted at the EPG electrode as a pair of
pH 8) in the presence of hypoxanthine concentrations of (insert) (a) 10,
(b) 20, (c) 40, (d) 60, (e) 80 and (main figure) (f) 100, (g) 200, (h) 400,
(i) 800, and (j) 1600 μM (sweep rate 5 mV s-1, pH 8).
wave around þ250 mV (at pH 8). After adding 2 mM allopurinol
to the same cell and incubation for 30 min the voltammogram in
part b of Figure 6 was obtained where the catalytic current is sup-
pressed. The competitive inhibition of allopurinol blocks binding
of xanthine, thus preventing uric acid from being generated.
Enzymatic Hypoxanthine Oxidation. As hypoxanthine is
also a substrate of XDH a similar voltammetric experiment was
conducted with the EPG/XDH electrode in the presence of
increasing concentrations of hypoxanthine (Figure 7). As a con-
trol, in the absence of XDH, hypoxanthine exhibits an irreversible
anodic peak at þ950 mV (pH 8) at an EPG working electrode
(Figure S4 of the Supporting Information) as reported by
others.21,25 This is well outside the potential window in Figure 7.
It is immediately apparent from the profile in Figure 7 that uric
acid is again generated at the electrode as the hypoxanthine
concentration is raised although there are notable differences in
the waveform. One feature is hysteresis as the hypoxanthine
concentration increases that is the voltammograms cross over on
the reverse cathodic sweep at high concentrations (curves i and j
of Figure 7. Second, the anodic peak potential shifts to higher
potential. Also, there is an initial lag in the catalytic current
2660
dx.doi.org/10.1021/jp111809f |J. Phys. Chem. B 2011, 115, 2655–2662

The Journal of Physical Chemistry B
ARTICLE
sharp 2-electron peaks centered at -550 mV versus Ag/AgCl
(Figure S1 of the Supporting Information). EPR redox potentio-
metric titrations18 of XDH At pH 8 resolved two separate redox
does not follow Michaelis-Menten kinetics. XDH is saturated
with xanthine (KM,xanthine 64 μM)46 at most concentrations
shown in Figure 2 and essentially is in a steady state throughout.
However, conventional depletion of uric acid from the electrode
surface occurs as the subsequent electrochemical oxidation to
uric acid imine occurs at a rate that exceeds its replenishment by
the combined XDHox/xanthine and XDHred/uric acid imine
reactions. The present study and mechanism proposed is also
consistent with the catalytic electrochemistry we have reported
previously for XDH18 and also bovine XO.16 Adsorption of XDH
on an EPG electrode leads to significant loss of the FAD cofactor
with free FAD being observed as a clearly define reversible
2-electron response. This is part explains why direct electro-
chemical catalysis is problematic under these conditions with the
enzyme having partially lost its most accessible redox cofactor;
the site of electron egress.
processes (FAD (quinone)/FADH (semiquinone) -606 mV
3
vs Ag/AgCl and FADH /FADH2 (hydroquinone) -680 mV vs
3
Ag/AgCl). The discrepancy between the voltammetry and redox
titration results both in terms of the potentials and electron
stoichiometry (two single electron transitions in the EPR redox
titration as opposed to a single two electron process in the
voltammogram) has never been explained.
Addition of free FAD to the EPG/XDH electrode in increas-
ing concentrations led to an amplification of the -550 mV redox
couple (Figure S6 of the Supporting Information). This response
is evidently not due to enzyme-bound FAD but instead to free
FAD, which is known to undergo 2-electron reversible redox
reactions while adsorbed on graphite electrodes43,44 and is
unable to stabilize a single electron reduced form. The conclu-
sion of this is that some of the XDH enzyme adsorbed on the
EPG electrode has lost its FAD cofactor; the -550 mV redox
response being due to dissociated FAD. Integration of the FAD
redox peaks using the model for a 2-electron oxidation of an
adsorbed molecule45 (curve a of Figure S6 of the Supporting
Information) leads to an approximate amount of 100 pmol of free
FAD on the electrode. Conversely, 100 of the original 385 pmol
(5 μL of a 77 μM solution) of holoenzyme under the membrane
has lost its FAD cofactor (∼25%). The actual amount of free
FAD (or presumably inactive enzyme) may be even higher as
the CV response from FAD only reveals the dissociated flavin
adsorbed to the electrode. More FAD may escape through the
membrane as the molecular weight cutoff (3500 Da) is much
higher than that of FAD. Nevertheless catalytic activity is still
observed. It is most likely that the remaining ∼75% of putatively
active XDH is responsible for the uric acid generated during the
xanthine and hypoxanthine oxidation experiments. However, we
cannot exclude the possibility that the FAD-free XDH is still able
to undergo electrochemically mediated catalysis. In principle the
Fe-S clusters or the Mo active site may donate electrons directly
to the uric acid imine mediator.
The fact that at least some of XDH trapped under the mem-
brane exhibits normal activity was demonstrated by coadsorbing
XDH and NADþ (the native electron acceptor) on the EPG
electrode (covered by a membrane) and standing the electrode
(at open circuit) in a solution of xanthine (100 μM) under anaer-
obic conditions. No current flows under these conditions and the
system functions as in a conventional assay with xanthine the
electron/hydride donor and NADþ the acceptor. The results
(Figure S7 of the Supporting Information) show an increase in
the uric acid concentration under the membrane over a period of
30 min as anticipated. It should be emphasized that this experi-
ment only proves that some of the enzyme under the membrane
is functional and is consistent with the observation of only partial
(25%) loss of the FAD cofactor from the sample.
’ ASSOCIATED CONTENT
S
Supporting Information. Cyclic voltammograms of
b
XDH (non-turnover and catalytic) as well as the purines
xanthine, uric acid and hypoxanthine. This material is available
free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: p.bernhardt@uq.edu.au.
’ ACKNOWLEDGMENT
Financial support of the Australian Research Council to P.V.B.
(DP0880288) and from the Deutsche Forschungsgemeinschaft
Cluster of Excellence “Unifying concepts in catalysis” (to S.L.) is
gratefully acknowledged.
’ REFERENCES
(1) Hille, R. Chem. Rev. 1996, 96, 2757–2816.
(2) Choi, E.-Y.; Stockert, A. L.; Leimk€uhler, S.; Hille, R. J. Inorg.
Biochem. 2004, 98, 841–848.
(3) Enroth, C.; Eger, B. T.; Okamoto, K.; Nishino, T.; Nishino, T.;
Pai, E. F. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 10723–10728.
(4) Harrison, R. Free Radical Biol. Med. 2002, 33, 774–797.
(5) Truglio, J. J.; Theis, K.; Leimk€uhler, S.; Rappa, R.; Rajagopalan,
K. V.; Kisker, C. Structure 2002, 10, 115–125.
(6) Leimk€uhler, S.; Hodson, R.; George, G. N.; Rajagopalan, K. V.
J. Biol. Chem. 2003, 278, 20802–20811.
(7) Bray, R. C.; Palmer, G.; Beinert, H. J. Biol. Chem. 1964,
239, 2667–2676.
(8) Nishino, T. J. Biochem. 1994, 116, 1–6.
(9) Komai, H.; Massey, V.; Palmer, G. J. Biol. Chem. 1969,
244, 1692–1700.
(10) Mao, L.; Xu, F.; Xu, Q.; Jin, L. Anal. Biochem. 2001,
292, 94–101.
(11) Hoshi, T.; Noguchi, T.; Anzai, J.-i. Mater. Sci. Eng., C 2006,
26, 100–103.
’ CONCLUSIONS
(12) Bernhardt, P. V. In Engineering the Bioelectronic Interface:
Applications to Analyte Biosensing and Protein Detection; Davis, J. J., Ed.;
Royal Society of Chemistry, 2009.
(13) Rodrigues, C. G.; Wedd, A. G.; Bond, A. M. J. Electroanal. Chem.
Interfacial Electrochem. 1991, 312, 131–140.
(14) Wang, L.; Yuan, Z. Anal. Sci. 2004, 20, 635–638.
(15) Liu, X.; Peng, W.; Xiao, H.; Li, G. Electrochem. Commun. 2005,
7, 562–566.
This study has shown that the catalytic electrochemistry of R.
capsulatus XDH is very complicated. At electrochemical poten-
tials above þ200 mV versus Ag/AgCl the product of XDH
catalysis, uric acid, is electroactive and serves as an indicator of
turnover. Uric acid plays a dual role as an electron acceptor from
the reduced enzyme (replacing NADþ in its native role) in an
autocatalytic process and the current due to uric acid oxidation
2661
dx.doi.org/10.1021/jp111809f |J. Phys. Chem. B 2011, 115, 2655–2662

The Journal of Physical Chemistry B
ARTICLE
(16) Bernhardt, P. V.; Honeychurch, M. J.; McEwan, A. G. Electro-
chem. Commun. 2006, 8, 257–261.
(17) Zhou, H.; Xu, Y.; Chen, T.; Suzuki, I.; Li, G. Anal. Sci. 2006,
22, 337–340.
(18) Aguey-Zinsou, K. F.; Bernhardt, P. V.; Leimk€uhler, S. J. Am.
Chem. Soc. 2003, 125, 15352–15358.
(19) Wu, Y.; Hu, S. Anal. Chim. Acta 2007, 602, 181–186.
(20) Shan, D.; Wang, Y.-N.; Xue, H.-G.; Cosnier, S.; Ding, S.-N.
Biosens. Bioelectron. 2009, 24, 3556–3561.
(21) Hason, S.; Stepankova, S.; Kourilova, A.; Vetterl, V.; Lata, J.;
Fojta, M.; Jelen, F. Anal. Chem. 2009, 81, 4302–4307.
(22) Dryhurst, G.; Hansen, B. H. J. Electroanal. Chem. Interfacial
Electrochem. 1971, 30, 417–426.
(23) Dryhurst, G. Bioelectrochem. Bioenerg. 1974, 1, 49–66.
(24) Dryhurst, G. J. Electroanal. Chem. Interfacial Electrochem. 1976,
70, 171–197.
(25) Yao, T.; Taniguchi, Y.; Wasa, T.; Musha, S. Bull. Chem. Soc. Jpn.
1978, 51, 2937–2941.
(26) Dryhurst, G. Bioelectrochem. Bioenerg. 1985, 14, 251–274.
(27) Owens, J. L.; Marsh, H. A., Jr.; Dryhurst, G. J. Electroanal. Chem.
Interfacial Electrochem. 1978, 91, 231–247.
(28) Gilmartin, M. A. T.; Hart, J. P.; Birch, B. J. Analyst 1994,
119, 243–252.
(29) Fei, J.; Wu, K.; Wu, Y.; Hu, S. J. Solid State Electrochem. 2004,
8, 316–321.
(30) Ardakani, M. M.; Akrami, Z.; Kazemian, H.; Zare, H. R.
J. Electroanal. Chem. 2006, 586, 31–38.
(31) Huang, X.-J.; Im, H.-S.; Yarimaga, O.; Kim, J.-H.; Lee, D.-H.;
Kim, H.-S.; Choi, Y.-K. J. Phys. Chem. B 2006, 110, 21850–21856.
(32) Zhao, Y.; Ye, T.; Liu, H.; Kou, Y.; Li, M.; Shao, Y.; Zhu, Z.;
Zhuang, Q. Front. Biosci. 2006, 11, 2976–2982.
(33) Ndamanisha, J. C.; Guo, L. Biosens. Bioelectron. 2008,
23, 1680–1685.
(34) Kalimuthu, P.; John, S. A. Anal. Chim. Acta 2009, 647, 97–103.
(35) Tehrani, R. M. A.; Ab Ghani, S. Sens. Actuators, B 2010,
B145, 20–24.
(36) Brody, M. S.; Hille, R. Biochemistry 1999, 38, 6668–6677.
(37) Kathiwala, M.; Affum, A. O.; Perry, J.; Brajter-Toth, A. Analyst
2008, 133, 810–816.
(38) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Funda-
mentals and Applications, 2nd ed.; Wiley: New York, 2001.
(39) Goyal, R. N.; Mittal, A.; Agarwal, D. Can. J. Chem. 1994,
72, 1668–1674.
(40) Kachoosangi, R.; Banks, C.; Compton, R. Electroanalysis 2006,
18, 741–747.
(41) Yan, Z.; Zhang, J.-R.; Fang, H.-Q. Anal. Lett. 1999, 32, 223–234.
(42) Massey, V.; Komai, H.; Palmer, G.; Elion, G. B. J. Biol. Chem.
1970, 245, 2837–2844.
(43) Gorton, L.; Johansson, G. J. Electroanal. Chem. 1980,
113, 151–158.
(44) Huck, H. Z. Phys. Chem. 1980, 120, 39–49.
(45) Laviron, E. J. Electroanal. Chem. Interfacial Electrochem. 1974,
52, 355–393.
(46) Leimk€uhler, S.; Stockert, A. L.; Igarashi, K.; Nishino, T.; Hille,
R. J. Biol. Chem. 2004, 279, 40437–40444.
2662
dx.doi.org/10.1021/jp111809f |J. Phys. Chem. B 2011, 115, 2655–2662