Covalent Immobilization of and Glucose Oxidase On Carbon Electrodes

Covalent Immobilization of FAD and
Glucose Oxidase on Carbon Electrodes
H. M. Sonawat, Ratna S. Phadke, and G. Govil

Biochemical Fuel Cells Project, Chemical Physics Group, Tata Institute
of Fundamental Research, Homi Bhabha Road, Bombay 400 005, India
Accepted for Publication January 13, 1984
The effectiveness of attaching flavin adenine dinucleotide (FAD) via a C bridge to Teflon-bonded carbon
black (CB), and the subsequent immobilization of glucose oxidase on the FAD-modified electrodes has been
studied by cyclic voltammetry. When FAD alone is bound
to the electrode, it undergoes reduction and oxidation at
- 0.62 and - 0.5 V, respectively-values similar to those
obtained with free FAD. Compared to the free enzyme,
the reduction of FAD as part of the immobilized enzyme
is 200 mV more cathodic, while the oxidation potential remains the same in both cases.
INTRODUCTION
Glucose oxidase (EC 1.1.3.4) has been immobilized on
various matrices and carriers for a wide range of applications. I-5 Its importance in monitoring glucose in clinical
laboratories, food processing, and fermentation has created interest for its use in automated analysis. To investigate its use in biofuel cells, Yahiro and co-workersh employed native glucose oxidase in conjunction with oxygen
cathode. They reported that the system generates a potential
of 175-350 mV. Glucose oxidase immobilized on platinum mesh by entrapment in polyacrylamide gel or glutaraldehyde crosslinking was used by Lahoda and co-workers.
Localization of the enzyme on glass electrodes by semipermeable membranes has been the immobilization method
in commercial glucose s e n ~ o r s . ~
Low
. ~ current densities
observed in these cases were probably due to the diffusional barrier posed by the entrapping or crosslinking matrices or the membranes. Electroactivity at the electrode
surface and transport of electron-carrying species from
the enzyme active site to electrode might also be low. lo The
use of highly electroactive electron mediators has been investigated as an approach to circumvent this problem. II
Covalent linking of glucose oxidase directly to the electrode surface through a carbodiimide spacer arm was attempted by Bourdillon and co-workers, l 2 but no direct electron transfer from enzyme active site to the electrode was
reported. Direct electron transfer between the active site
and electrode was reported by Ianniello and YacynychI3
and Ianniello and co-workers14by immobilization of the
enzyme on cyanuric chloride-modified graphite electrodes.
'
Biotechnology and Bioengineering, Vol. XXVI, Pp. 1066-1070 (1984)

0 1984JohnWiley&Sons,lnc.
FAD-modified electrodes offer a convenient handle for

immobilizing glucose oxidase for direct electron transfer.
Wingard and GureckaI5 covalently immobilized the flavin
moiety on carbon electrodes through a conjugate doublebond system to its phenyl ring. Our efforts in this direction
started with charge-density calculations of the isoalloxazine ring of FAD during its redox reactions. I h These results suggest that the two carbonyl groups of the flavin
moiety are the most suitable positions for attachment to
the carbon electrodes. In this article, we present the results
of the covalent immobilization of FAD alone and glucose
oxidase on the surface of carbon electrodes as assessed by
cyclic voltammetry experiments.
MATERIALS AND METHODS

Materials
Teflon-bonded carbon black (CB; ca. 250 m2/g surface
area, 3% wt Teflon, ca. 77% porosity, a mean pore diameter of 0.25 pm, and a pore area of 225 m2/g) was a gift
from Energy Research Corporation. Glassy carbon (GC)
was from Atomergic Chemetal Corporation. Flavin adenine dinucleotide (FAD), Sephadex G-200, and DEAESephadex A-50 were from Sigma Chemical Co. (St. Louis,
MO). The Dee-0 concentrate (Miles Laboratories) was a
gift from Professor R. H. Sarma, Institute of Biomolecular
Stereodynamics (Albany, NY), and the glucose oxidase
was purified from the Dee-0 concentrate (up to 121
units/mg) by a sequence of steps involving ammonium sulfate fractionation, gel filtration on Sephadex G-200, and
ion exchange on DEAE-Sephadex A-50. The other chemicals are of analytical grade and are used without further
purification, unless ot henvise mentioned.
Procedures
Immobilization of FAD
A summary of the overall reaction sequence is presented
in Table I. The GC electrodes were cut in ca. 0.5 X 1.O-cm
pieces by a power-driven carborundum wheel. The edges
were smoothened on the same wheel. Increasingly fine
CCC 0006-3592/84/091066-05$04.00
Table I. Summary of the reaction steps for FAD immobilization on

carbon electrodes.
Step
Reaction
1
2
refluxing with acetone for 1 h and vacuum drying

concentrated HNO, and H,SO, treatment for 15 min and 1 h
at room temperature and 170"C, respectively
sodium/alcohol treatment for reduction
potassium bromide sulfuric acid treatment for bromination
refluxing in chloroform with triphenylphosphine for 3-4 h
Wittig condensation with FAD in the presence of sodium methoxide
3
4
5
6
physically adsorbed FAD was removed. This step was carried out in the dark to prevent the photoreaction of FAD.
Quantitation of the FAD coverage on the electrode could
not be carried out accurately because of the ambiguities
involved in determining the surface area of the electrode
exposed to the supporting electrolyte in CV experiments.
Assuming that the surface area characteristics of CB remain unaffected by the various modification treatments,
the bound FAD concentration on the electrode was estimated to be 0.4 X
pmol/cm2.
Immobilization of Glucose Oxidase

grade of emery paper was used to give the electrodes a mirror polish. The CB electrodes were much easier to handle.
They were also cut to the same size. Organic impurities
were removed by refluxing the electrodes with acetone for
about one hour and then drying in vacuum. Oxidation of
the electrode surface was carried out by the procedure described by Wingard and Gurecka. I5 The electrodes were
thoroughly washed with distilled water and vacuum dried.
Carboxyl groups were reduced by treating the electrodes
with sodium and alcohol. The electrodes were placed in a
round-bottomed flask fitted with a large-bore reflux condensor containing 30 mL absolute ethanol. Sodium (1.6 g)
was added through the condensor, rapidly enough to keep
up a vigorous reaction. The flask was shaken occasionally.
When the initial reaction had subsided, ca. 4 mL absolute
ethanol was added, and the flask was heated on a steam
bath until all the sodium had reacted. Then, 10 mL distilled water was added and the mixture refluxed for one
hour. The contents were allowed to cool, the electrodes
taken out, washed thoroughly with ether, and vacuum
dried.
Potassium bromide and sulfuric acid treatment was used
for bromination of the alcoholic groups on electrode surface. KBr (30 g) was dissolved in 50 mL distilled water, and
25 mL concentrated sulfuric acid was added slowly with
constant stirring and cooling. The electrodes were immersed in the filterate, and 15 mL concentrated sulfuric
acid was added and the contents were refluxed for 3-4 h.
The electrodes were taken out and washed first with water
and then with freshly distilled chloroform.
Triphenyl phosphonium derivative was obtained by refluxing the electrodes from the earlier step for 4 h in chloroform containing 1 g triphenylphosphine. The contents
were cooled to OOC. The electrodes were washed with cold
distilled methanol.
The in situ generation of ylide and its subsequent condensation with the carbonyl group of the isoalloxazine ring
of FAD was achieved in the following manner: Triphenylphosphonium-derivatized electrodes were kept in a 0.5
mg/mL solution of FAD in methanol. Sodium (0.12 g/10
mL) was added with stirring and left for ca. 10-12 h with
occasional stirring. The electrodes were washed thoroughly until the absorbance of the washings at 340 nm was
zero, then vacuum dried. In this manner, we made certain
that covalently linked FAD remained on the surface and
Glucose oxidase was immobilized by incubating the

FAD-modified electrodes with its apoenzyme, prepared
by the method of Swoboda, for about 2 h. The electrodes
were washed until the washings showed no absorbance at
280 nm. The FAD-modified electrodes and the immobilized enzyme electrodes were stored refrigerated in the
dark when not in use.
Enzyme Assay
Immobilized enzyme activity was detected by modification of the method described by Bergmeyer. l8 The reaction mixture (3.0 mL) consisted of o-dianisidine di-HC1
(0.08 mg/mL), peroxidase (0.06 mg/mL), and glucose
(0.10M) in 0.1M potassium phosphate buffer at pH 7.0.
The color of oxidized o-dianisidine was read spectrophotometrically at 450 nm.
Cyclic Voltammetry
All cyclic voltammetry experiments were done using an
instrument locally fabricated by Santhanam and Bhagat. l 9
A conventional three-electrode system was used in a cell
where platinum mesh was the counter electrode and a saturated calomel electrode was the reference. The supporting
electrolyte in all cases was 1.OM KCI at pH 7.0. The potential was swept at the rate of 150 mV/s. Analar-grade nitrogen was bubbled through the electrolyte for 30-60 min
prior to the experiments, which were performed in nitrogen atmosphere.
RESULTS AND DISCUSSION

FAD is known to undergo a redox cycle (Fig. I ) where
two protons and two electrons are released and taken up.
In order to understand the mechanism, charge-density
calculations were carried out using semiemperical quantum mechanical calculations (INDO) on the flavin ring of
FAD. Analysis of these results indicated that during its
redox cycle, maximum charge mobilization takes place at
the carbonyl groups. I h It was concluded, therefore, that a
conducting bridge between one of these two groups and
the electrode surface should lead to the conduction of the
electrons from the FAD molecule to the electrode.
The cyclic voltammetric behavior of FAD in solution is
shown in Figure 2(a). The experiment was performed in
the dark with CB as the working electrode. The supporting
SONAWAT, PHADKE, AND GOVIL: FAD AND GLUCOSE OXIDASE IMMOBILIZATION
1067
(REDUCED)
Figure 1. The reduction oxidation cycle of FAD. Only the isoalloxazine

ring is shown, with -CH, replacing rest of the molecule. In the presence
of the substrate, reduction occurs enzymatically. Electrons pass to the
electrode during the reoxidation of FADH,.
Figure 3(a) presents the cyclic voltammogram of FAD

covalently attached to a GC surface through Wittig condensation. Reduction peaks are observed at -0.35 and
0.68 V. The return sweep shows a peak at -0.60 V and
another broad peak centered at -0.25 V. It seems that the
peaks represent two redox couples. The triphenylphosphonium-derivatized GC does not show any of these characteristics [Fig. 3(b)]. Reduction peaks at -0.40 and
-0.68 V have also been reported by Wingard and GureckaI5
for the riboflavin-modified GC. Due to the absence of any
reoxidation peaks in the return sweep, they have concluded (under the experimental conditions used) the electrochemical reduction of immobilized riboflavin to be irreversible. From our experiments, we observe that the
reduction peaks are well separated from the corresponding oxidation peaks. It follows that the redox process is
chemically reversible. However, the peaks in question are
not very distinct and are represented merely by shoulders.
This may be due to low coverage of immobilized FAD on
the electrode. Repeated attempts at increasing the concentration by changing the amount of FAD for Wittig condensation in the last step and the amount of Ph3P in the penultimate step, showed little improvement in the voltammograms,
if any. Due to this, and also because of the unpredictability
of the source of the peaks, further experiments were carried out on CB only.
A cyclic voltammogram of FAD immobilized on CB in
supporting electrolyte is shown in Figure 3(c). Predominant reduction and oxidation peaks at -0.62 and -0.50
V, similar to those observed for FAD in solution, indicate
U
Figure 2. Cyclic voltammogram of FAD: (a) 1.0 mg/mL in supporting
electrolyte; (b) the same as in (a) after the electrode has been left in the
FAD solution for some time. The broken lines show a voltammogram of
the electrode in supporting electrolyte.
electrolyte was degassed as described in the Materials and

Methods section and enough FAD stock solution (in supporting electrolyte) added to a final concentration of 1.0
mg/mL. A well-defined peak centered at -0.62 V was observed upon reduction of FAD. In the return sweep, a peak
was seen at -0.50 V, corresponding to its sxidation. Earlier reports have established that the electrochemical reduction of FAD takes place by two indistinguishable oneelectron
Adsorption of FAD on the surface
of the electrode leads to its reduction at more cathodic potential. l 4 A major peak at -0.7 V [Fig. 2(b)] indicates excessive adsorption, while the shoulder represents solution
species. The reoxidation also shifts to a 0.06 V more
negative potential.
1068
SCE
vs
A
Figure 3. Cyclic voltammogram of: (a) FAD-modified glassy carbon

( G C ) electrode; (b) triphenylphosphonium-derivat!/ed GC electrode;
and (c) FAD-modified carbon black (CB)electrode. The broken lines
Thaw unmodified GC or CB in supporting electrolyte.
BIOTECHNOLOGY AND BIOENGINEERING, VOL. 26, SEPTEMBER 1984
positive FAD coupling to the electrode surface. The possibility of adsorption being the mode of attachment may be
ruled out because no shoulder is observed at potentials
more negative than -0.62 V. Moreover, the peaks for
both reduction and oxidation are sharper than those in
Figure 2(b) where adsorption is a dominant phenomenon.
Immobilization of glucose oxidase on FAD-modified
CB yields the cyclic voltammogram presented in Figure
4(a). It may be observed that while the reduction potential
is not affected, the peak potential for oxidation is shifted
0.11 V to the anodic side. The redox peaks are sharper
than the ones seen for free enzyme [Fig. 4(b)]. Also, the reduction peak is 0.20 V more cathodic than the one for the
enzyme in solution. The fact that we have been able to detect enzyme activity on the surface of the electrode indicates that the redox peaks observed for immobilized enzyme electrode were due to FAD as a component of the
apoenzyme-coenzyme (holoenzyme) system rather than
the immobilized FAD alone. Furthermore, the apoenzyme, when studied on an unmodified CB electrode, does
not show any redox peaks [Fig. 4(c)], which means that the
protein part of the enzyme does not contribute to the peaks
observed for immobilized enzyme. Upon repeated scanning of both FAD and glucose oxidase bound electrode in
CV experiments, no appreciable effect could be observed
on the peak current amplitudes. The coenzyme-bound
electrode retained its capacity to bind with the apoenzyme
and immobilized glucose oxidase retained its enzymatic

activity after the cyclic voltammetry experiments.
The low enzyme activity that we observed might be due
to several reasons. First, the low coverage of FAD on the
electrode surface will limit the number of molecules of enzyme being immobilized. Second, if it is assumed that the
coverage of FAD is sufficient, their spatial distribution on
the electrode surface may not be very favorable for the
apoenzyme to bind. Each molecule of glucose oxidase has
two molecules of FAD tightly bound.22 Hence, it is conceivable that the spontaneous latching of an apoenzyme
molecule would require its juxtaposition with two FAD
molecules which, in this case, are hooked on to the electrode surface. Third, if the enzyme active site is deeply
seated in the molecule and the FAD is attached too close to
the electrode surface, i.e., if the conducting bridge is short
(the C chain in our study is ca. 3 A), then steric hinderances will come into play. We are, at present, investigating
the role of the steric factor by increasing the length of the
chain which links FAD to the electrode surface. It will then
serve the dual purpose of a conducting bridge, between the
enzyme active site and the electrode, and a spacer arm.
The enzyme electrode could be an efficient bioanode in
biofuel cells and glucose sensors.
CONCLUSIONS
Covalent attachment of FAD on a carbon electrode
through the carbonyl group of the flavin ring structure
seems to favor the exchange of electrons between the FAD
molecule and the electrode. Immobilization of glucose oxidase on an electrode so modified apparently orients the
enzyme molecule with respect to the electrode and feads to
direct electron swapping between the enzyme active site
and the electrode.
The authors wish to thank Professor K. S. V. Santhanam for his help
with the cyclic voltammetric studies. Financial support of the project by
the Department of Science and Technology is gratefully acknowledged.
References
Figure 4. Cyclic voltammogram of: (a) glucose oxidase immobilized on

FAD-modified CB electrode; (b) glucose oxidase (0.33pM) in supporting
electrolyte: and (c) glucose oxidase apocnzyme (0.30pM) in supporting
electrolyte. The broken lines show unmodified CB in supporting elcctrolyte.
1. Y. Ishimori, 1. Karube, and S. Suzuki, Eur. J. AppL Microhio/. Biotechno/. , 13, 197 (1981).
2. J. Zemek, L. Kuniak, P. Gemeiner, J. Zaniocky, and S. Kucar, Elizyrne Microb. Techno/., 4, 233 (1982).
3. C. 0. Malikkides and R. H. Weiland, Biotechriol. Biorng., 24, 2419
(1982).
4. B. N. Alberti and A. M. Klibanov, Enzynie Microh. Technol.,4, 47
( 1982).
5. L. DAngiuro and P. Cremonesi, Biotechnol. Biorng., 24, 207
(1982).
6. A. T. Yahiro, S. M. Lee, and D. 0. Kimblc, Biochim. Biophys. Actcr,
88, 375 (1964).
7. E. J. Lahoda, C . C . Liu, and L. B. Wingard, Jr., Biotrchnol.
Bioeng., 17, 413 (1975).
8. J. Havas, E. Porjesz, G. Nagy, and E. Pungor, Hung. Sci. 112strumenfs, 49,53 (1980).
9. M. Koyama, J. Koezuka, and Y. Sato, Toshihci Rev., 132, 2 (1981).
10. L. B. Wingard, Jr., EflZJnW Eng., 5, 101 (1980).
SONAWAT, PHADKE, AND GOVIL: FAD AND GLUCOSE OXIDASE IMMOBILIZATION
1069
11. S. Varfolomeev, A. I. Yaropolov, I. A. Berezin, M. R. Tarasevich,

and V. A. Bogdanovskaya, Bioelectrochem. Bioenergy, 4, 314
(1977).
12. C. Boudillon, J . P. Bourgeois, and D. Thomas, J. Am. Chem. Soc.,
102,4231 (1980).
13. R. M. Iannielli and A. M. Yacynych, A n d . Chem., 53,2090 (1981).
14. R. M. lanniello, T. J. Lindsay, and A. M. Yacynych, A n d Chem.,
54, 1098 (1982).
15. L. 9. Wingard, Jr. and J. L. Gurecka, Jr., J. Mof. Cutal., 9, 209
( 1980).
16. R. S. Phadke, H. M. Sonawat, and G. Govil, hit. J. Qnuntuni
Chern.. Qiiurrfum B i d . Synip., 10, 251 (1983).
1070
17. B. E. P. Swoboda, Biochim. Biophys. Actu, 175,365 (1969).

18. H. U. Bergmeyer, Ed., Methods in Enzymutic Anulysis (Academic,
New York, 1974), Vol. 1, p. 457.
19. K. S. V. Santhanam and V. R. Bhagat, J. Sci. Ind. Res., 30, 235
(1971).
20. F. Sheller, G. Strnad, 9. Neumann, M. Kuhn, and W. Ostrowski,
Bioelectrochem. Bivenergy, 6 , 117 (1979).
21. 9. Janik and P. J. Elving,J. Chem. Rev., 68, 312 (1968).
22. H. J. Bright and D. J. T . Porter, The Enzymes, 3rd ed., P. D. Boyer,
Ed. (Academic, New York, 1975), Vol. 12, pp. 421-505.
BIOTECHNOLOGY AND BIOENGINEERING, VOL. 26, SEPTEMBER 1984

Covalent Immobilization of and Glucose Oxidase On Carbon Electrodes

Uploaded by

Copyright:

Available Formats

Covalent Immobilization of and Glucose Oxidase On Carbon Electrodes

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Covalent Immobilization of and Glucose Oxidase On Carbon Electrodes

Uploaded by

Copyright:

Available Formats

Covalent Immobilization of FAD and

Glucose Oxidase on Carbon Electrodes

H. M. Sonawat, Ratna S. Phadke, and G. Govil

Biotechnology and Bioengineering, Vol. XXVI, Pp. 1066-1070 (1984)

FAD-modified electrodes offer a convenient handle for

MATERIALS AND METHODS

Table I. Summary of the reaction steps for FAD immobilization on

refluxing with acetone for 1 h and vacuum drying

Immobilization of Glucose Oxidase

Glucose oxidase was immobilized by incubating the

RESULTS AND DISCUSSION

SONAWAT, PHADKE, AND GOVIL: FAD AND GLUCOSE OXIDASE IMMOBILIZATION

Figure 1. The reduction oxidation cycle of FAD. Only the isoalloxazine

Figure 3(a) presents the cyclic voltammogram of FAD

electrolyte was degassed as described in the Materials and

Figure 3. Cyclic voltammogram of: (a) FAD-modified glassy carbon

BIOTECHNOLOGY AND BIOENGINEERING, VOL. 26, SEPTEMBER 1984

and immobilized glucose oxidase retained its enzymatic

Figure 4. Cyclic voltammogram of: (a) glucose oxidase immobilized on

SONAWAT, PHADKE, AND GOVIL: FAD AND GLUCOSE OXIDASE IMMOBILIZATION

11. S. Varfolomeev, A. I. Yaropolov, I. A. Berezin, M. R. Tarasevich,

17. B. E. P. Swoboda, Biochim. Biophys. Actu, 175,365 (1969).

BIOTECHNOLOGY AND BIOENGINEERING, VOL. 26, SEPTEMBER 1984

You might also like