Electrical communication between glucose oxidase and electrodes mediated by phenothiazine-labeled poly(ethylene oxide) bonded to lysine residues on the enzyme surface

Article abstract of DOI:10.1021/ac025872t

A series of glucose oxidase (GOx) hybrids (GOx-phenothiazine-labeled poly(ethylene oxide) (PT-PEO)) capable of direct electrical communication with electrodes is synthesized by covalently modifying PT-PEO to lysine residues on the enzyme surface. The length of the PEO chain and the number of PT groups are systematically altered. After the PT-PEO modification, all the hybrids maintain more than 50% of enzyme activity relative to that of native GOx, although loss of the activity becomes greater with increasing PEO chain length. The catalytic current, i_cat, is observed at a potential more positive than 0.55 V after the addition of glucose, due to the intramolecular electron transfer (ET) from reduced forms of flavin adenine dinucletide (FADH₂/FADH) to PT⁺ that are electrogenerated at the electrode. The i_cat value increases with the number of PT groups, indicating that most of the modified PT groups act as mediators. The magnitude of the i_cat increase depends on the PEO chain length and reveals a maximum for PT-PEO with the molecular weight of 3000. In contrast, the i_cat is almost constant for GOx-2-(10-phenothiazyl)propionic acid (PT-PA) hybrids with more than two PT groups synthesized by covalently modifying PT-PA to surface lysines, indicating that only a few key PT groups function as mediators. The maximum rate constant (130 s^-1) for the ET from FADH₂/FADH to PT⁺ is obtained for the GOx hybrid modified with five PT-PEO groups with the molecular weight of 3000.

Full text of DOI:10.1021/ac025872t

Anal. Chem. 2003, 75, 910-917
Electrical Communication between Glucose
Oxidase and Electrodes Mediated by
Phenothiazine-Labeled Poly(ethylene oxide)
Bonded to Lysine Residues on the Enzyme Surface
Kazumichi Ban, Takeshi Ueki, Yoshinori Tamada, Takahiro Saito, Shin-ichiro Imabayashi, and
Masayoshi Watanabe*
Department of Chemistry and Biotechnology, Yokohama National University, Yokohama 240-8501, Japan
frequently used to shuttle electrons between FAD and the
electrode under an oxygen-free condition.2-7
A series of glucose oxidase (GOx) hybrids (GOx-phe-
nothiazine-labeled poly(ethylene oxide) (P T-P EO)) ca-
pable of direct electrical communication with electrodes
is synthesized by covalently modifying P T-P EO to lysine
residues on the enzyme surface. The length of the P EO
chain and the number of P T groups are systematically
altered. After the P T-P EO modification, all the hybrids
maintain more than 5 0 % of enzyme activity relative to that
of native GOx, although loss of the activity becomes
greater with increasing P EO chain length. The catalytic
current, icat, is observed at a potential more positive than
0 .5 5 V after the addition of glucose, due to the intramo-
lecular electron transfer (ET) from reduced forms of flavin
adenine dinucletide (FADH2 / FADH) to P T+ that are
electrogenerated at the electrode. The icat value increases
with the number of P T groups, indicating that most of the
modified P T groups act as mediators. The magnitude of
the icat increase depends on the P EO chain length and
reveals a maximum for P T-P EO with the molecular weight
of 3 0 0 0 . In contrast, the icat is almost constant for GOx-
2 -(1 0 -phenothiazyl)propionic acid (P T-P A) hybrids with
more than two P T groups synthesized by covalently
modifying P T-P A to surface lysines, indicating that only a
few key P T groups function as mediators. The maximum
rate constant (1 3 0 s-1 ) for the ET from FADH2 / FADH to
P T+ is obtained for the GOx hybrid modified with five
P T-P EO groups with the molecular weight of 3 0 0 0 .
In terms of the application to enzymatic sensor systems8 and
the electrochemical control of an enzymatic function, enzymes
capable of a direct electron exchange with electrodes have been
arousing great interest. One possible way to provide electrochemi-
cal activity for enzymes is the covalent immobilization of an
electron relay to the FAD unit9,10 or on the surface of GOx.11-16
Heller et al.11 and Schuhmann12 reported the synthesis and
electrochemical properties of GOx with ferrocene (Fc) derivatives
attached to the sugar or acidic amino acid residues on its surface
via spacer chains of different lengths. These studies demonstrated
that redox mediators bound via long, flexible, and hydrophilic
spacer chains to the outer surface of GOx can transfer electrons
to the electrode surface according to the so-called “wipe mecha-
nism” with the enzyme-bound mediators swinging in and out of
the active site of the enzyme. Although the number and the
location of mediators and the length of spacer chains are decisive
factors to achieve efficient mediation properties, they have not
been systematically controlled in the previous studies and their
(2) Cass, A. E. G.; Davis, G.; Francis, G. D.; Hill, H. A. O.; Aston, W. J.; Higgins,
I. J.; Plotkin, E. V.; Scott, L. D.; Turner, A. P. F. Anal. Chem. 1 9 8 4 , 56,
667-671.
(3) Degani, Y.; Heller, A. J. Phys. Chem. 1 9 8 7 , 91, 1285-1289.
(4) Bartlett, P. N.; Bradford, V. Q.; Whitaker, R. G. Talanta 1 9 9 1 , 38, 57-63.
(5) Ikeda, T.; Hamada, H.; Senda, M. Agric. Biol. Chem. 1 9 8 6 , 50, 883-890.
(6) Taniguchi, I.; Miyamoto, S.; Tomimura, S.; Hawkridge, F. M. J. Electroanal.
Chem. 1 9 8 8 , 240, 333-339.
(7) Degani, Y.; Heller, A. J. Am. Chem. Soc. 1 9 8 8 , 110, 2615-2620.
(8) Schuhmann, W.; Wohlschl¨ager, H.; Lammert, R.; Schmidt, H.-L.; Lo¨ ffler,
V.; Wiemho¨ fer, H.-D.; Go¨ pel, W. Sens. Actuators, B 1 9 9 0 , 1, 571-574.
(9) Katz, E.; Riklin, A.; Heleg-Shabtai, V.; Willner, I.; Bu¨ ckmann, A. F. Anal.
Chim. Acta 1 9 9 9 , 385, 45-58.
(10) Willner, I.; Katz, E. Angew. Chem., Int. Ed. 2 0 0 0 , 39, 1180-1218.
(11) Schuhmann, W.; Ohara, T. J.; Schmidt, H.-L.; Heller, A. J. Am. Chem. Soc.
1 9 9 1 , 113, 1394-1397.
(12) Schuhmann, W. Biosens. Bioelectron. 1 9 9 5 , 10, 181-193.
(13) Badia, A.; Carlini, R.; Fernandez, A.; Battaglini, F.; Mikkelsen, S. R.; English,
A. M. J. Am. Chem. Soc. 1 9 9 3 , 115, 7053-7060.
(14) Ryabov, A. D.; Trushkin, A. M.; Baksheeva, L. I.; Gorbatova, R. K.;
Kubrakova, I. V.; Mozhaev, V. V.; Gnedenko, B. B.; Levashov, A. V. Angew.
Chem., Int. Ed. Engl. 1 9 9 2 , 31, 789-791.
Glucose oxidase (GOx; EC 1.1.3.4) is a dimeric glycoprotein
of 186 kDa containing two tightly bound flavin adenine dinucle-
otide (FAD) cofactors, which catalyze the electron transfer (ET)
from glucose to oxygen accompanying the production of glucono-
lactone and hydrogen peroxide.1 The direct electrical communica-
tion between the redox center (FAD) and electrodes is prevented
because the redox center is located too far from the outermost
surface. Therefore, freely diffusing redox mediators with a more
positive redox potential than that of the redox center have been
(15) Bartlett, P. N.; Booth, S.; Caruana, D. J.; Kilburn, J. D.; Santamaria, C. Anal.
* Corresponding author: (e-mail) mwatanab@ynu.ac.jp.
(1) Hecht, H. J.; Kalisz, H. M.; Hendle, J.; Schmid, R. D.; Schomburg, D. J.
Mol. Biol. 1 9 9 3 , 229, 153-172.
Chem. 1 9 9 7 , 69, 734-742.
(16) Battaglini, F.; Bartlett, P. N.; Wang, J. H. Anal. Chem. 2 0 0 0 , 72, 502-
509.
910 Analytical Chemistry, Vol. 75, No. 4, February 15, 2003
10.1021/ac025872t CCC: $25.00 © 2003 American Chemical Society
Published on Web 01/21/2003

effect on the mediation electrochemistry of GOx hybrids still
remains ambiguous.
Scheme 1. Synthetic Procedures for
GOx-(PT-PEO) and GOx-(PT-PA) Hybrids
We synthesized a series of new biopolymer conjugates (GOx-
phenothiazine-labeled poly(ethylene oxide) (PT-PEO)) by co-
valently bonding phenothiazine (PT) mediators to lysine residues
on the enzyme surface via PEO chains and electrochemically
investigated the ET from FADH2/ FADH to PT+ (oxidized PT
group) as a function of the number of attached PT groups and
the length of the PEO chain under a substrate-saturated and
diffusion-limited condition. PEO is a nontoxic, flexible, water-
soluble polymer, and PEO covalently attached to the protein
surface does not denature proteins or hinder the approach of other
small molecules in the case of low modification numbers.17 PT is
expected to have higher ability to mediate the ET between FAD
and electrodes than Fc due to its higher redox potential. We found
that all the PT groups bonded to lysine residues effectively mediate
the ET reaction and an optimum length of the PEO chain exists
in terms of the ET rate. We also synthesized GOx-2-(10-
phenothiazyl)propionic acid (PT-PA) hybrids, in which PT-PA is
directly bound to lysine residues on the GOx surface, and
compared its catalytic electrochemistry with that of GOx-(PT-
PEO) hybrids to gain further insight into the effect of the PEO
spacer on the catalytic ET reaction of GOx hybrids.
EXPERIMENTAL PROCEDURES
Materials. GOx from Aspergillus niger and horseradish per-
oxidase (HRP; EC 1.11.1.7.) were purchased from Toyobo;
succinic anhydride, N,N′-dicyclohexyl carbodiimide (DCC), so-
dium acetate trihydrate, piperazine-1,4-bis(2-ethansulfonic acid)
(PIPES), and phosphate buffer powder were from Wako Pure
Chemicals; diethylene glycol dimethyl ether, D-glucose, 1-ethyl-
3-(3-(dimethylamino)propyl)carbodiimide hydrochloride (EDC),
and phenothiazine were from Junsei Chemicals; 40% aqueous
solution of benzyltrimethylammonium hydroxide and â-D-glucose
were from Tokyo Kasei; N-hydroxysuccinimide (NHS) was from
Aldrich; o-dianisidine was from Sigma; N-hydroxysulfosuccinimide
(sulfo-NHS) was from Fluka; and N-ethyl-N-(2-hydroxy- 3-sulfo-
propyl)-m-toluidine (EHSPT) was from Dojindo Laboratories. All
the reagents were used without further purification. Disposable
ultrafiltration units (type, USY-5; cutoff MW, 50 000) and chromato
disks were purchased from Advantec and GL Science, respectively.
Sodium acetate buffer (0.05 mol dm-3) was prepared by
dissolving an appropriate amount of sodium acetate trihydrate into
deionized water and adjusting its pH to 5.1 by adding 0.1 mol
dm-3 HCl. PIPES buffer (0.15 mol dm-3) was prepared by
dissolving an appropriate amount of PIPES powder into deionized
water, and then its pH was adjusted to 7.4 by adding 0.1 mol dm-3
NaOH. Phosphate buffer (1/ 15 mol dm-3, pH 7.4) was prepared
by dissolving a bag of phosphate buffer powder into 1 dm3 of
deionized water.
equivalent molar ratio of ethylene oxide to phenothiazine was
admitted from a tank. While the reaction mixture was stirred at
110 °C for 2 h, the pressure inside the autoclave was reduced to
the initial value, indicating that 1 molar equiv of ethylene oxide
was added to phenothiazine. The appropriate amount of ethylene
oxide to obtain the desired molecular weight of PT-PEO was
gradually introduced into the autoclave as the inside pressure was
maintained at (1.47-1.96) × 105 Pa at 120 °C. After confirming
the end of the polymerization reaction indicated by the reduction
of the inside pressure, ethylene glycol dimethyl ether was
completely removed under reduced pressure.
Synthesis of P T-P EO (Scheme 1 a). Phenothiazine (0.1 mol)
was dissolved in dried ethylene glycol dimethyl ether (110 mL)
and then dried potassium hydroxide (0.064 mol) was added as a
catalyst. After this mixture was introduced into an autoclave, the
atmosphere in the autoclave was replaced with N2 by repeated
evacuation and N2 gas purge. Under the evacuated condition, an
(17) Harris, J. M. Poly(ethylene glycol) Chemistry: Biotechnical and Biomedical
Applications; Plenum: New York, 1992; Chapter 1.
Analytical Chemistry, Vol. 75, No. 4, February 15, 2003 911

atmosphere, the reaction mixture was evaporated and the remain-
ing solid was redissolved into pure water. After addition of diethyl
ether, the aqueous phase was collected and mixed with chloro-
form. The chloroform layer was collected, evaporated, and dried
in vacuo, yielding s-PT-PEO; yield, 75%: 1H NMR (δ from TMS
in CDCl3) 2.66 (4H, s, CH2COO), 3.50-3.80 (xH, m, CH2OCH2),
3.85 (2H, t, NCCH2O), 4.10 (2H, t, NCH2), 4.24 (2H, t, OCOCH2),
6.85-7.2 (8H, m, Harom). The number of protons at δ ) 3.50-
3.80, x, is 4(n - 1), where n is the degree of polymerization of
PEO.
Table 1. Number-Average Molecular Weight and
Hydrodynamic Properties of PT-PEO
number-average
mol wt
OH
titration
Db
rc
a
PT-PEO
EA
Mw/ Mn
(×106 cm2 s-1
)
(nm)
PT-PEO1000
PT-PEO2000
PT-PEO3000
PT-PEO4200
PT-PEO8000
1050
2050
2970
4170
7779
1050
2030
2860
4360
8750
1.20
1.09
1.05
1.58
1.20
2.4
1.0
1.4
1.8
2.6
3.4
1.7
1.3
0.93
0.71
s-PT-PEO (3.7 mmol) and N-hydroxysuccinimide (4.4 mmol,
NHS) were dissolved in 50 mL of dimethylformamide (DMF), and
the resulting solution was cooled to 0 °C. To this solution, a DMF
solution of DCC (4.4 mmol) was added dropwise at 0 °C under
stirring. The mixture was allowed to react at room temperature
for 24 h. After the reaction, the mixture was again cooled to 0 °C
and filtered to remove DCC-urea. The solid obtained by evapo-
rating the filtrate was dissolved in benzene, and insoluble impuri-
ties were removed by filtration. The filtrate was dropped into
petroleum ether. The precipitate was collected and dried under
reduced pressure, yielding a-PT-PEO; yield, 70%: 1H NMR (δ from
TMS in CDCl3) 2.70-3.0 (4H, t, CH2COO; 4H, m, HNHS), 3.50-
3.80 (xH, m, CH2OCH2), 3.86 (2H, t, NCCH2O), 4.10 (2H, t, NCH2),
4.27 (2H, t, OCOCH2), 6.86-7.20 (8H, m, Harom).
a
Measured by GPC using tetrahydrofuran as a carrier solvent and
calibrated by poly(ethylene glycol) standards. b Diffusion coefficient
at 25 °C in 0.05 mol dm-3 sodium acetate buffer (pH5.1) determined
by chronoamperometry. c Hydrodynamic radius calculated from the
Stokes-Einstein equation.
GOx and a-PT-PEO were dissolved in 1/ 15 mol dm-3 phosphate
buffer (pH 7.4) and kept at 25 °C for 24 h. The produced GOx-
(PT-PEO) hybrid was separated from excess PT-PEO by ultrafil-
tration. The yellow substances remaining on the ultrafiltration
membrane were dissolved into sodium acetate buffer (pH 5.1) and
filtered through a chromato disk to remove impurities. The
resulting sodium acetate solution of GOx-(PT-PEO) hybrid was
used for the electrochemical measurements.
P reparation of GOx-(P T-P A) Hybrid (Scheme 1 b). Cya-
noethylation of phenothiazine resulted in 2-(10-phenothiazyl)-
propionitrile,18 followed by its hydrolysis in an alkaline solution.
Repeated recrystallization of the obtained solid from ethanol
yielded PT-PA as a white solid; yield, 30%; mp, 164 °C: 1H NMR
(δ from TMS in DMSO-d6) 2.7 (2H, t, CH2CO), 4.14 (2H, t, NCH2),
6.9-7.3 (8H, m, Harom), 12.35 (1H, s, COOH).
An appropriate amount of PT-PA (7.4-129 µmol) was dissolved
in 0.15 M Na-PIPES buffer (pH 7.4) aided by sonication, and the
resulting solution was cooled to 0 °C. Equimolar quantities of EDC
and sulfo-NHS were added to the PT-PA solution to activate PT-
PA, and then GOx was added to the buffer solution 30 min after
the activation. The reaction mixture was kept at 25 °C for 24 h.
Unreacted PT-PA and impurities were removed from the reaction
mixture by ultrafiltration. The yellow substances remaining on
the ultrafiltration membrane, GOx-(PT-PA) hybrids, were dis-
solved in a sodium acetate buffer (pH 5.1), which was employed
for the electrochemical measurements.
Figure 1. UV-visible spectra of PT-PEO (a), native GOx (b), and
GOx-(PT-PEO) hybrid (c).
The reaction mixture remaining in the bottom of the autoclave
was taken and neutralized by adding sulfuric acid until its pH
became 6. To remove ionic impurities, the mixture was stirred
with an acid adsorbent and then with a base adsorbent for 30 min,
respectively, and then filtered. The filtrate was evaporated to
dryness at 100 °C for 1 h under reduced pressure. The structure
and number-average molecular weight of obtained PT-PEO were
characterized by a combination of 1H NMR, hydroxyl group
titration, gel permeation chromatography (GPC), and elemental
analysis (EA), as listed in Table 1. The number-average molecular
weight was calculated from the content of nitrogen in EA and the
end hydroxyl group titration, for both assuming the structure of
PT-PEO as shown in Figure 1. The reasonable agreement between
the values determined by the hydroxyl group titration and EA
reveals that PT-PEO molecules synthesized in this work have the
structure shown in Figure 1. The Mw/ Mn values close to unity
also support that PT-PEO can be prepared by the ring-opening
anionic polymerization.
Estimation of the Number of Modified Mediators and the
Relative Enzymatic Activity. Figure 1 shows typical UV-visible
absorption spectra of PT-PEO (a), native GOx (b), and GOx-
(PT-PEO) hybrid (c) recorded by a Shimadzu UV-160A spectro-
photometer. The concentration of the GOx hybrid can be
P reparation of GOx-(P T-P EO) Hybrids (Scheme 1 a). PT-
PEO (5 mmol) and succinic anhydride (25 mmol) were dissolved
in 1,2-dichloroethane (100 mL) containing a small amount of
pyridine (2 mL). After refluxing at 65 °C for 10 h under nitrogen
estimated using the extinction coefficient of native GOx (ꢀ452
)
21 600 M-1 cm-1 19 from the absorbance at 452 nm, where there
)
(18) Smith, N. L. J. Org. Chem. 1 9 5 0 , 15, 1125-1130.
912 Analytical Chemistry, Vol. 75, No. 4, February 15, 2003

is no absorption of PT-PEO or PT-PA. The difference in the
absorbance between the GOx hybrid and native GOx at 320 nm
corresponds to the absorption of PT-PEO or PT-PA groups
attached to the GOx surface and can be converted to their
concentration using their ꢀ320 values.20 The concentration ratio of
the PT groups to GOx corresponds to the average number of
modified mediators per GOx molecule. At present, we have no
information about the distribution of the number of attached PT
groups among GOx hybrids or about the location of lysine
residues on the GOx surface responsible for the binding. The
absorption arising from the PT groups was not observed when
the PT-PEO or PT-PA modification was carried out without
carbodiimide and NHS or its derivatives, confirming the formation
of amide bonding between lysine residues on the GOx surface
and a carboxylic acid group of PT-PEO or PT-PA.
The enzymatic activity relative to that of native GOx was
determined using the peroxidase-o-dianisidine assay21 for GOx-
(PT-PEO) hybrids and the 4AA-EHSPT assay22 for GOx-(PT-
PA) hybrids in 0.05 mol dm-3 sodium acetate buffer (pH 5.1) at
25 °C under O2 saturation. The relative activity equals the
proportion of active FAD to all FAD groups in the hybrids.
Electrochem ical Measurem ents. A conventional three-
electrode cell was used with a glassy carbon working electrode
(geometrical area, 0.071 cm2), a Ag|AgCl|3 M KCl reference
electrode, and a Pt wire auxiliary electrode. A glassy carbon
electrode was polished with alumina powder (0.05-µm diameter)
and sonicated in pure water prior to use. A solution of GOx hybrids
was introduced to the cell and gently deaerated by N2 purge for
20 min. Cyclic voltammograms (CVs) were recorded under N2
atmosphere between 0.3 and 0.7 V at a scan rate of 10 mV s-1 in
the absence and presence of 0.05 mol dm-3 glucose using a BAS-
CV-50W electrochemical analyzer. The differential pulse voltam-
mogram (DPV) was measured from 0.3 to 0.7 V at a scan rate of
1 mV s-1, a pulse amplitude of 50 mV, and a sampling time of 50
ms in the absence of glucose using the same electrochemical
analyzer.
Figure 2. Number of mediators attached per GOx molecule in
GOx-(PT-PEO) and GOx-(PT-PA) (0) hybrids as a function of the
molar ratio of activated PT-PEO or PT-PA to GOx in feed. The
molecular weights of modified PT-PEO groups are 1000 (9), 2000
(3), 3000 (O), 4200 (2), and 8000 (×).
Figure 3. Relationship between the number of PT-PEO or PT-PA
(0) groups attached per GOx molecule and the enzyme activity of
GOx hybrids relative to native GOx. The molecular weights of modified
PT-PEO groups are 1000 (9), 2000 (3), 3000 (O), 4200 (2), and
8000 (×).
The diffusion coefficient of PT-PEO was determined by
chronoamperometry (CA) in 0.05 mol dm-3 acetate buffer (pH
5.1) containing 1 mmol dm-3 PT-PEO. The electrode potential was
stepped up from 0.3 to 0.7 V. The electrochemical analyzer, the
cell configuration, and the electrodes used in CA were the same
as those in CV.
a-PT-PEO or a-PT-PA to GOx. In the case of GOx-(PT-PEO)
hybrids, the reactivity of a-PT-PEO to lysine residues tends to
decrease with increasing molecular weight except PT-PEO with
a molecular weight of 8000 (PT-PEO8000). Due to the steric
hindrance of a longer PEO chain, it is considered that the
probability of encountering lysine residues becomes lower for the
activated carboxylic acid group of a-PT-PEO with higher molecular
weight. The maximum number of modified PT-PEOs, which is 9,
was obtained for PT-PEO1000. This value is comparable to those
in the previous reports7,13 obtained for the lysine modification
without denaturants, which are used to modify amino acid residues
inside enzymes. Considering that some of the 30 lysine residues
per GOx molecule are not located on the surface,23 the maximum
modification number seems to be reasonable.
RESULTS AND DISCUSSION
Number of Modified Mediators and Relative Activity of
the GOx Hybrid. The number of mediators attached per GOx
molecule was controlled by varying the molar ratio of PT-PEO or
PT-PA to the GOx molecule in the reaction mixture. Figure 2
shows the relation between the number of bonded PT-PEO or
PT-PA groups per GOx and the molar ratio of a-PT-PEO or a-PT-
PA to GOx in the hybrid preparation. For all the hybrids, the
number of attached PT groups increases with the molar ratio of
(19) Nakamura, S.; Fujiki, S. J. Biochem. 1 9 6 8 , 63, 51-58.
(20) The ꢀ320 value measured for each PT-PEO ranges from 2390 to 2870 M-1
cm-1 and depends on the molecular weight of PT-PEO. The ꢀ320 value of
Figure 3 represents the dependence of the relative enzymatic
activity of hybrids to native GOx on the number and the molecular
weight of the attached PT-PEO groups. The relative enzymatic
PT-PA is 2900 M-1 cm-1
.
(21) Keesey, J. Biochemical Information; Boehringer Mannheim Biochemicals:
Indianapolis, IN, 1987; p 27.
(22) Enzyme Catalog: Toyobo Enzymes; Toyobo Co., Ltd.: Osaka, Japan, 1998; p
(23) Hecht, H. J.; Schomburg, D.; Kalisz, H.; Schmid, R. D. Biosens. Bioelectron.
1 9 9 5 , 8, 197-203.
112.
Analytical Chemistry, Vol. 75, No. 4, February 15, 2003 913

Figure 5. Catalytic current of GOx-(PT-PEO3000)4.0 as a function
of the hybrid concentration measured at 0.62 V at a glassy carbon
electrode in 0.05 mol dm-3 sodium acetate buffer (pH 5.1) containing
0.05 mol dm-3 glucose.
a-PT-PEO or a-PT-PA and lysine residues on GOx explains why
the high relative activity of GOx was maintained after the mediator
modification. Actually, SDS-PAGE measurements of the hybrids
revealed the absence of cross-linked GOx.
Electrochemical Measurements. Figure 4A shows CVs of
GOx-(PT-PEO3000)3.8 hybrid (the average number of modified
PT-PEO3000 is 3.8) in the absence (a) and presence (b) of 0.05
mol dm-3 glucose. Although no apparent redox peak appeared in
the CV in the absence of glucose, PT groups attached to the GOx
surface retained redox activity at 0.53 V, as shown in a DPV of
GOx-(PT-PEO3000)3.8 hybrid (Figure 4B). The GOx-(PT-PA)5.3
hybrid also exhibited a pair of small peaks around 0.55 V in the
absence of glucose, which is attributed to the redox of PT groups
attached to the GOx surface (Figure 4C(a)). On the other hand,
the oxidation current of GOx hybrids (Figure 4A(b) and C(b))
greatly increased at a potential higher than the redox potential of
PT groups in the presence of glucose, indicating the enzymatic
reduction of the electrochemically oxidized PT (PT+). This means
that PT groups bound to the GOx surface can function as electron
mediators between the electrode and the FAD center of GOx.
Despite a similar number of modified PT groups, the GOx-(PT-
PEO) hybrid showed greater catalytic current icat than that of the
GOx-(PT-PA) hybrid in the same potential range.
The icat measured at 0.62 V for GOx-(PT-PEO) or GOx-(PT-
PA) hybrid increased with the glucose concentration and leveled
off above 0.02 mol dm-3 glucose (data not shown). The icat linearly
increased with the concentration of GOx-(PT-PEO) hybrid up
to 20 µmol dm-3 as shown in Figure 5, indicating that molecular
interactions between hybrids are not significant and the intramo-
lecular ET from FADH2/ FADH to PT+ groups on the same GOx
molecule mainly occurs in the GOx-(PT-PEO) at the hybrid
concentration used (9 µmol dm-3).
Figure 4. Cyclic voltammograms of GOx-(PT-PEO3000)3.8 hybrid
(8.9 µmol dm-3) (A) and GOx-(PT-PA)5.3 hybrid (9.7 µmol dm-3) (C)
at a glassy carbon electrode in 0.05 mol dm-3 sodium acetate buffer
(pH 5.1) at a scan rate of 10 mV s-1 in the absence (a) and presence
(b) of 0.05 mol dm-3 glucose. Differential pulse voltammogram (B)
of GOx-(PT-PEO3000)3.8 hybrid (8.9 µmol dm-3) measured at a
glassy carbon electrode in the same buffer at a scan rate of 1 mV
s-1 in the absence of glucose is also shown.
activity of hybrids seems to significantly depend not only on the
molecular weight of PT-PEO but also on the number of attached
PT-PEO groups. The modification of PT-PEO with low molecular
weight little affects the relative enzymatic activity of hybrids.
Although the modification with a large number of PT-PEO groups
brought about a decline in the relative activity of the GOx hybrid,
all GOx hybrids used in this work retained more than 50% relative
activity. NHS esters have high reaction selectivity for lysine
residues and are effective in preventing cross-linking among the
GOx molecules.24 We think that the use of NHS or its derivative
for the carbodiimide-promoted amide bond formation between
Figure 6 represents the icat of the hybrids measured at 0.62 V
under a substrate-saturated condition (0.05 mol dm-3 glucose) as
functions of the number and the molecular weight of the modified
PT-PEO groups. For all the GOx-(PT-PEO) hybrids, the icat
increased with the number of attached PT groups, demonstrating
(24) Brinkley, M. Bioconjugate Chem. 1 9 9 2 , 3, 2-13.
914 Analytical Chemistry, Vol. 75, No. 4, February 15, 2003

shorter than those in the present work. The use of longer PEO
spacers resulted in the finding of optimum spacer length for the
direct electron exchange between the FAD center and mediators.
In contrast to GOx-(PT-PEO) hybrids, GOx-(PT-PA) hybrids
with more than two PT groups revealed an almost constant icat
(Figure 6). Only a few key PT groups function as mediators for
the ET reaction between electrodes and the FAD center in
GOx-(PT-PA) hybrids due to the short and hydrophobic PA
spacer.
Estimation of the Rate Constants of ET from FADH2 /
FADH to P T+ Groups. Under the oxygen-free environment, the
intramolecular ET reactions from FADH2/ FADH to PT+ in GOx
hybrids are described as follows:
GOx(FAD)-(PT)n + glucose f GOx(FADH2)-(PT)n +
Figure 6. Relationship between the number of mediators attached
per GOx molecule and the catalytic current for GOx hybrids modified
with PT-PEO of molecular weight 1000 (9), 2000 (3), 3000 (O), 4200
(2), and 8000 (×), and GOx-(PT-PA) hybrids (0) at 0.62 V at a
glassy carbon electrode in 0.05 mol dm-3 sodium acetate buffer (pH
5.1) containing 0.05 mol dm-3 glucose. The concentration of the
gluconolactone (1)
GOx(FADH2)-(PT)n f GOx(FADH2)-(PT)n-1(PT+) + e-
(2)
hybrids was 9 µmol dm-3
.
k1
9
GOx(FADH2)-(PT)n-1(PT+)
8 GOx(FADH)-(PT)n + H+
(3)
GOx(FADH)-(PT)n f GOx(FADH)-(PT)n-1(PT+) + e-
(4)
k2
9
GOx(FADH)-(PT)n-1(PT+)
8 GOx(FAD)-(PT)n + H+ (5)
Under a glucose-saturated condition, the diffusion-limited icat
for the intramolecular mediation reaction of GOx hybrid is given
by13
1/ 2
icat ) 2FA(DGOx-hybrid kobs
)
C
(6)
GOx-hybrid
Figure 7. Dependence of the catalytic current of GOx hybrids (9
µmol dm-3) on the molecular weight of modified PT-PEO. The number
of modified PT-PEO per GOx molecule is 3.6 for GOx-(PT-
PEO1000), 4.0 for GOx-(Pt-PEO2000), 4.1 for GOx-(PT-PEO3000),
3.2 for GOx-(PT-PEO4200), and 3.2 for GOx-(PT-PEO8000). The
catalytic current was measured at 0.62 V at a glassy carbon electrode
in 0.05 mol dm-3 sodium acetate buffer (pH 5.1) containing 0.05 mol
dm-3 glucose.
where icat is the catalytic current, F is the Faraday constant, A is
the electrode area, DGOx-hybrid is the diffusion coefficient of the
GOx hybrid, kobs is the average rate constant for the oxidation of
FADH2/ FADH by PT+, and CGOx-hybrid is the concentration of the
GOx hybrid. The kobs is expressed in the form of k1k2/ (k11/ 2
+
k21/ 2)2, where k1 and k2 are the rate constants for the oxidation of
FADH2 and FADH by PT+, respectively. As described by eq 6,
the icat is a function of DGOx-hybrid, kobs, and CGOx-hybrid. Because
preliminary light scattering measurements indicate that the
difference in D between native GOx and GOx hybrids is small,
the value for native GOx (4.1 × 10-7 cm2 s-1 at 25 °C)7 was used
for all GOx hybrids to estimate kobs values.
the participation of most of the PT groups bonded to the GOx
surface in the ET reaction between the electrode and FAD.
Further, it is interesting to note that the magnitude of the icat of
a hybrid with a similar number of attached PT depends on the
molecular weight of PT-PEO and reveals a maximum for GOx-
(PT-PEO3000) hybrids, as shown in Figure 7. This suggests the
existence of an optimum PEO chain length in terms of the ET
from FADH2/ FADH to PT+ groups in the GOx-(PT-PEO)
hybrids. It has been reported by some research groups that the
intramolecular ET rate between the redox center and mediators
was enhanced with an increase in the chain length linking
mediators to the enzyme.25 The length of linkers in those works
was, however, less than 15 methylene units and considerably
The obtained kobs values (Table 2) are more than 2 orders of
magnitude greater for GOx-(PT-PEO) hybrids than for GOx-
(PT-PA) hybrids. Similarly to the icat, kobs increases with the
number of PT groups in GOx-(PT-PEO) hybrids and is almost
independent of the number of PT groups in GOx-(PT-PA)
hybrids, indicating that the difference in the icat mainly comes
from the difference in the kobs. Since the through-space distance
from the FAD center is more than 2.3 nm for all the 30 lysine
residues per GOx, it is considered that the approach of the PT
groups close to FAD is not achieved by the short alkyl chain but
by the long and flexible PEO chain.
(25) Willner, I.; Lapidot, N.; Riklin, A.; Kasher, R.; Zahavy, E.; Katz, E. J. Am.
Chem. Soc. 1 9 9 4 , 116, 1428-1441.
Analytical Chemistry, Vol. 75, No. 4, February 15, 2003 915

The dependence of the kobs on the number of attached PT
groups includes the influence of the location of PT-PEO modifica-
tion with respect to the FAD center, since the reactivity to a-PT-
PEO and the distance from FAD are not equivalent among the
lysine residues. We examined the electrochemical properties of
GOx hybrids prepared by the systematic modification of PT-PEO
to surface glutamic acid and aspartic acid residues,27 some of
which are located closer to FAD than the lysine residues.1
Comparing the kobs values for glutamic acid- and aspartic acid-
modified hybrids with those for the lysine-modified hybrids at a
similar number of modified PT groups, greater values were
obtained for the former hybrids in the case of shorter PEO
chains.27 This suggests that the PT-PEO modification closer to
FAD is effective to achieve fast ET, and a long PEO spacer is
possible to compensate for the disadvantage in the location of PEO
modification.
The independence of the icat on the number of modified PT-
PA suggests that the location of modified mediators is critical in
GOx-(PT-PA) hybrids due to their short spacer chains. We
prepared GOx-(PT-PA) hybrids with 0.96 and 1.70 PT groups
by incubating GOx with 50- and 80-fold excess PT-PA at 25 °C for
3 h, respectively. Almost half of the icat for those of GOx-(PT-
PA) hybrids with more than two PT groups was observed for these
two hybrids. Taking into account that one GOx molecule consists
of two subunits with the same structure, the constant icat for GOx-
(PT-PA) hybrids with more than two PT groups might imply that
the PT groups bonded to the most reactive lysine residues in the
PT-PA modification became the most effective key mediators.
Table 2. Rate Constants for ET from FADH2/FADH to
PT+ in GOx-(PT-PEO) and GOx-(PT-PA) Hybrids with a
Different Number of Modified Mediators
b
GOx hybrids
na
kobs (s-1
)
GOx-(PT-PA)n
1.0
2.4
4.5
5.7
2.0
3.6
6.0
8.7
0.4
2.6
4.0
5.5
0.8
2.5
4.1
5.0
0.4
1.3
2.0
3.2
0.2
1.3
3.2
4.4
0.063
0.24
0.31
0.25
4.6
17
28
68
GOx-(PT-PEO1000)n
GOx-(PT-PEO2000)n
GOx-(PT-PEO3000)n
GOx-(PT-PEO4200)n
GOx-(PT-PEO8000)n
0.82
10
19
52
1.8
33
47
130
0.76
1.7
7.1
12
0.3
1.9
7.5
19
a
The number of PT-PEO or PT-PA groups attached per GOx hybrid
molecule. b Rate constants were calculated using the equation icat
2FA(DGOx-hybrid kobs
matic activity.
)
1/ 2
)
C
and corrected for the relative enzy-
GOx-hybrid
The kobs of GOx-(PT-PEO) alters from 0.3 to 130 s-1 depend-
ing on the molecular weight and number of PT-PEO groups. For
GOx with Fc attached to sugar or acidic amino acid residues on
the surface via long spacer chains, kobs values comparable to or
greater than the values for GOx-(PT-PEO) hybrids were ob-
served.11,12 However, direct comparison among these values is
difficult because of the difference in the modified positions and
the ambiguity of the number of modified mediators in the previous
reports. Although many factors participate in the presence of the
optimum PEO chain for the ET from FADH2/ FADH to PT+, the
motion of the PEO chains is one important factor. It is considered
that the end PT group of surface-tethered PT-PEO moves around
the bonding site in an arc. The hydrodynamic radius of PT-PEO,
which was calculated from its chronoamperometrically determined
diffusion coefficient using the Stokes-Einstein relation, increases
with the molecular weight. It ranges from 1.0 nm for PT-PEO1000
to 3.4 nm for PT-PEO8000 (Table 1). With an increase in the
molecular weight of PT-PEO in the hybrids, the possibility of
access of PT groups close to FAD increases and a faster ET rate
from FADH2/ FADH to PT+ can be achieved. A slight difference
in the distance between FADH2/ FADH and PT+ at the closest
approach may yield significant difference in kobs, since the ET rate
decays exponentially with the distance between a donor and an
acceptor.26 Factors in reducing the ET rate, however, must exist
in the case of too long PEO chains. Overlap of the accessible space
of PT groups on the GOx surface with each other and interference
in the access of the PT group to FAD from a bulky long PEO
seem to be included in the factors reducing the ET rate.
CONCLUSION
We have found here that PT groups bonded to lysine residues
on the GOx surface via the PEO chain effectively mediate fast
ET between the electrode and the FAD center. This fast ET is
primarily attributed to the fast oxidation of FADH2/ FADH by PT+
in GOx-(PT-PEO) hybrids. While the ET rate from FADH2/
FADH to PT+ increased with the number of PT-PEO groups, the
optimum molecular weight, that is, the optimum spacer length of
PT-PEO, was found with respect to the ET from FADH2/ FADH
to PT+. This is the first finding of the optimum spacer length for
direct electron exchange with electrodes in GOx with covalently
modified mediators via spacer chains. The optimum spacer length
would be determined also by many factors unexamined in this
work: the relative positions of mediators, entropic and enthalpic
factors associated with chain mobility and penetration, and
interactions of PT-PEO with glycan chains and other PT-PEOs.
Elucidation of the relative importance of these factors resulting
in the optimum spacer length is left for future studies. The
comparison of catalytic electrochemistry between GOx-(PT-PEO)
and GOx-(PT-PA) hybrids has shed light on the importance of a
long, hydrophilic and flexible PEO spacer for generating fast ET
from FADH2/ FADH to PT+. The fast ET enables GOx-(PT-
PEO3000)5.0 to exhibit a catalytic current comparable to that for
the corresponding freely diffusing PT-PEO3000 system despite
the smaller diffusion coefficient of mediators attached to the GOx
hybrid.28 Additionally, the effect of dioxygen on the catalytic
(27) Aoki, S.; Ishii, K.; Ueki, T.; Ban, K.; Imabayashi, S.; Watanabe, M. Chem.
Lett. 2 0 0 2 , 256-257.
(26) Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta. 1 9 8 5 , 811, 265-322.
916 Analytical Chemistry, Vol. 75, No. 4, February 15, 2003
(28) Imabayashi, S.; Ban, K.; Ueki, T.; Watanabe, M. J. Phys. Chem. B submitted.

current of this GOx hybrid was negligible. These results suggest
the possibility of synthesizing GOx hybrids that realize a specific
glucose sensing without perturbation by dioxygen.
Design of New Materials System” (404/ 11167234) from the
Ministry of Education, Culture, Sports, Science, and Technology,
Japan.
Received for review June 19, 2002. Accepted November
20, 2002.
ACKNOWLEDGMENT
This work was partly supported by a Grant-in-Aid for Scientific
Research on Priority Areas of “Molecular Synchronization for
AC025872T
Analytical Chemistry, Vol. 75, No. 4, February 15, 2003 917