Electrochemical measurement for hydroxyl radical / H. Tatsumi et al. / Anal. Biochem. 467 (2014) 22–27
23
Å
strate, and pyrroloquinoline quinone-dependent glucose dehydro-
genase (PQQ–GDH) as a catalyst, the anodic current signal of the
trapping adducts (catechol and hydroquinone) produced by the
hydroxylation of phenol was able to be amplified and detected sen-
sitively. In the current study, we improved this method and
as ascorbic acid and uric acid. To study the OH generation induced
by the xanthine–XO system, care must be taken to avoid the inter-
ference from uric acid (or urate) that is produced in the test
solution:
O ! Uric Acid þ 2Oꢀ þ 2Hþ:
ꢁ
ð1Þ
Å
Xanthine þ 2O
2
þ H
2
applied it to the study of OH generation induced by the xan-
2
Å
thine–XO system. The reaction rates of the OH generation in the
In our previous study, we used phenol as a trapping agent, and
the major hydroxylated adduct was catechol [23]. The applied
potential for the detection of catechol was 0.28 V, which was posi-
tive enough to oxidize urate; thus, this method could not be
employed for the xanthine–XO system. We needed to use a trap-
ping agent that gave a hydroxylated adduct with a more negative
formal potential.
presence and absence of various Fe(III) complexes and proteins
were compared.
Materials and methods
Reagents
In the current study, we chose 2,6-xylenol as a trapping agent.
Its hydroxylated adduct, 2,6-dimethylhydroquinone (DMHQ), can
be oxidized at approximately ꢁ0.1 V, as shown in cyclic voltammo-
gram b of Fig. 1. This is because methyl substitution stabilizes the
oxidized (quinone) form of the quinone/hydroquinone redox cou-
ple, and shifts its formal potential to more negative potential.
Moreover, the two methyl groups can prevent the substitution at
the ortho positions. Although hydroxylation may occur at the meta
positions, the resorcinol is not oxidized in the potential range
examined. Therefore, we can expect that DMHQ is the only detect-
able hydroxylated adduct, which leads to a more accurate determi-
Xanthine oxidase (EC 1.17.3.2) and pyrroloquinoline quinone-
dependent glucose dehydrogenase (EC 1.1.5.2) were obtained from
Toyobo and used without further purification. 2,6-Dimethylhydro-
quinone was purchased from Tokyo Chemical Industry. Fe–ethyl-
enediaminetetraacetate (Fe–EDTA) was purchased from Dojindo.
Cytochrome c (bovine heart), ferritin (equine spleen), hemin
(
bovine), and hemoglobin (bovine blood) were purchased from
Sigma. In aqueous solutions, the Fe atom of these complexes and
proteins had the oxidation number +3. The Fe content of ferritin
was determined spectrophotometrically by Drysdale and Munro’s
method [24]. Other reagents were obtained from Nacalai Tesque.
These chemicals were used as received.
nation. Curve c in Fig. 1 shows the voltammogram of 20 lM DMHQ
in a pH 7.0 phosphate buffer containing 0.1 M glucose obtained
with a PQQ–GDH-modified PFC electrode. The amplified anodic
current indicates that the product of the electrode reaction, 2,6-
dimethylbenzoquinone (DMBQ), was reduced catalytically by
PQQ–GDH in the presence of glucose, and DMHQ was regenerated
efficiently, as shown in Scheme 1.
Fig. 2A shows a typical I–t curve obtained with the PQQ–GDH-
modified PFC electrode at E = 0 V for the successive addition of
DMHQ in the buffer containing 0.1 M glucose. After each addition
of DMHQ, a catalytic steady-state current (Iss) was observed within
Preparation of PQQ–GDH-modified electrode
A plastic-formed carbon (PFC) disk electrode (BAS, cat. no.
2
0
02408, geometrical area 0.071 cm ) was polished successively
by a lapping film sheet (3M, no. 15000) and 0.05
lm alumina
slurry (Refine Tec, Yokohama, Japan) with a polishing cloth before
ꢁ1
use. A 5-
l
l aliquot of 0.1 U
ll
PQQ–GDH solution was dropped
onto the surface of the PFC electrode. The solvent was allowed to
evaporate, and then the electrode surface was covered with a dial-
ysis membrane (Wako, cutoff molecular weight of 12,000–14,000,
1
min. The Iss was proportional to the concentration of DMHQ (C),
as shown in Fig. 2B. The limit of detection (LOD) was estimated
from the regression line by taking signal/noise [S/N] = 3, with N
being regarded as the standard deviation of the residuals [25],
and an LOD of 1 nM was obtained. The improvement of the LOD
as compared with our previous result for catechol (8 nM) [23]
20-lm thick in the dry state). The electrode was covered by a nylon
net to give it physical strength. When not in use, the PQQ–GDH-
modified PFC electrode was kept immersed in 50 mM phosphate
buffer (pH 7.0) at 4 °C.
Electrochemical measurements
Voltammograms and current–time (I–t) curves at the PQQ–
GDH-modified electrode were recorded with a potentiostat (HECS
9
72, Husou Seisakusyo, Kawasaki, Japan). The measurements were
carried out with a three-electrode system. A platinum coil and an
Ag/AgCl (0.1 M KCl) electrode were used as the counter and refer-
ence electrodes, respectively. In this article, working electrode
potentials (E) are referred to as the Ag/AgCl (0.1 M KCl) electrode.
–
1
The XO reaction was initiated by adding 1.0 to 1.3 U ml XO
into an air-saturated test solution (normally 5 ml) containing 1.9
to 70
.19 mM 2,6-xylenol, and 0.1 M glucose. During the measure-
ments, the test solution was stirred by a magnetic stirrer at
lM xanthine, 50 mM phosphate buffer (pH 7.0), 0.18 to
0
5
3
00 rpm. The temperature of the test solution was maintained at
7 ± 1 °C by circulating thermostated water in an outer jacket of
a glass cell.
Results and discussion
Amperometric detection of hydroxylated adducts
Fig.1. Cyclic voltammograms: (a) 50 mM phosphate buffer (pH 7.0) containing
0
(
.1 M glucose at a PFC electrode covered by dialysis membrane without PQQ–GDH;
b) (a) +20 M DMHQ; (c) same as (b) except that the electrode was replaced by a
PQQ–GDH-modified PFC electrode. The scan rate was 5 mV s
It is well known that the electrochemical measurements of bio-
logical systems are often interfered by oxidative substances such
l
ꢁ
1
.