Enzymatic activity of glucose oxidase covalently wired via viologen to electrically conductive polypyrrole films

Biosensors and Bioelectronics 19 (2004) 823–834

Enzymatic activity of glucose oxidase covalently wired via viologen to electrically conductive polypyrrole films X. Liu, K.G. Neoh∗ , Lian Cen, E.T. Kang Department of Chemical and Environmental Engineering, National University of Singapore, 10 Kent Ridge Crescent, Singapore 119260, Singapore Received 11 March 2003; received in revised form 6 August 2003; accepted 19 August 2003

Abstract The surface functionalization of an electrically conductive polypyrrole film (PPY) with a viologen, (N-(2-carboxyl-ethyl)-N -(4-vinylbenzyl)-4,4 -bipyridinium dichloride, or CVV) for the covalent immobilization of glucose oxidase (GOD) has been carried out. The viologen was first synthesized and graft polymerized on PPY film. It then served as an anchor via its carboxyl groups for the covalent immobilization of GOD. The surface composition of the as-functionalized substrates was characterized by X-ray photoelectron spectroscopy (XPS). The effects of the CVV monomer concentration on the CVV-graft polymer concentration and the amount of GOD immobilized on the surface were investigated. The activity of the immobilized GOD was compared with that of free GOD and the kinetic effects were also obtained. The cyclic voltammetric (CV) response of the GOD-functionalized PPY substrates was studied in a phosphate buffer solution under an argon atmosphere. The CV results support the mechanism in which CVV acts as a mediator to transfer electron between the electrode and enzyme, and hence regenerating the enzyme in the enzymatic reaction with glucose. High sensitivity and linear response of the enzyme electrode was observed with glucose concentration ranging from 0 to 20 mM. © 2003 Elsevier B.V. All rights reserved. Keywords: Polypyrrole; Glucose oxidase; Viologen; Covalent immobilization; Electrical wiring of enzyme

1. Introduction Since the first publication on a glucose biosensor (Clark and Lyons, 1962), the detection of glucose has attracted a high degree of interest due to its biological importance. Enzyme redox reactions (for example with glucose oxidase (GOD) in Scheme 1a) operating in biosensors are particularly amenable to interfacing with electrochemical transducers because electron exchange is a key step in the catalytic process (Schultz, 1991). A great deal of work on the development of glucose sensors has been carried out by Heller’s group (1996). Several techniques such as electrochemical (Gavrilov et al., 1993), Langmuir–Blodgett (Hodak et al.,

Abbreviations: AAc, acrylic acid; CV, cyclic voltammetry; CVV, N-(2-carboxyl-ethyl)-N -(4-vinylbenzyl)-4,4 -bipyridinium dichloride; GOD, glucose oxidase; PBS, phosphate buffer solution; PPY, polypyrrole; TSA, toluene-4-sulfonic acid; VBC, 4-vinylbenzyl chloride; WSC, water-soluble-1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride; XPS, X-ray photoelectron spectroscopy ∗ Corresponding author. Tel.: +65-68742186; fax: +65-67791936. E-mail address: [email protected] (K.G. Neoh). 0956-5663/$ – see front matter © 2003 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2003.08.010

1997), layer-by-layer (Onda et al., 1996; Kenausis et al., 1997) and covalent binding (Okahata et al., 1989) techniques have been developed for the immobilization of enzymes on various matrices, and electrochemical adsorption of enzymes has been reported for the construction of enzyme electrodes (Sadik and Wallace, 1993; Compagnone et al., 1995). The incorporation of the enzyme into the growing conducting polymer during electropolymerization has recently been used as a convenient and simple method to achieve the immobilization. Various conducting polymers have been extensively considered as the material for immobilization of enzymes, such as polyacetylene, polythiophene, polypyrrole (PPY), polyindole and polyaniline (Pandey, 1988; Shinohara et al., 1988; Bartlett and Cooper, 1993). Of these conducting polymers, PPY is well characterized and is probably one of the most suitable polymers for biosensor applications because it has good environmental stability and biocompatibility. However, the mechanism by which enzyme molecules are trapped within the PPY network is uncertain (Cho et al., 1996), and the efficiency of the current response obtained from such electrodes due to successful reaction glucose with GOD is lower than the

824

X. Liu et al. / Biosensors and Bioelectronics 19 (2004) 823–834

expected values (Ramanathan et al., 1996). Among previous studies on the direct electrochemical oxidation of the enzyme by PPY, only very small amounts of mediator species can be efficiently cycled within the film. A detailed study (Belanger et al., 1989) investigating GOD immobilized in PPY films has shown that the conductivity of the film is destroyed due to over-oxidation of the polymer by the enzymically produced H2 O2 , making direct electron transport either along or across the polymer matrix improbable. While the response of the enzyme film will increase as the quantity of the enzyme at the electrode increases, an increase in the thickness of the films can result in a decrease in response due to the fact that bulk analyte entering the film will react with the enzyme at the front surface of the film. As a result, the co-product is more easily lost to the bulk solution, instead of diffusing through the film to be detected at the underlying electrode (Cooper and Bloor, 1993). To improve the electron transfer, redox mediators have permitted the coupling of enzymatic and electrochemical reactions. (Scheme 1b). The mediator (MED) participates in the transfer of electrons between the electrode and the enzyme, and thus the regeneration/recycling of the enzyme during a bio-reaction can be accomplished (Fisher et al., 2000). In present work, we report a technique for the covalent binding of GOD onto an electrically conductive PPY surface via a viologen (N-(2-carboxyl-ethyl)-N -(4-vinylbenzyl)-4,4 bipyridinium dichloride, or CVV). This viologen served as an anchor for the immobilization of GOD on the PPY surface as well as an electron mediator between the modified electrode and the enzyme. The direct linkage of GOD via an electron mediator to the electrode is expected to increase

the efficiency of electron transport, and hence the sensitivity of the GOD modified electrode. Our interest in using PPY as the electrode is due to its high electrical conductivity, biocompatibility and the possibility of fabricating small, lightweight and flexible electrodes from this material.

2. Experimental 2.1. Materials 4,4 -Bipyridine, 3-chloropropionic acid, 4-vinylbenzyl chloride (VBC) and pyrrole (99%) were obtained from Aldrich. The pyrrole was distilled before use. Water-soluble1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride (WSC) was purchased from Dojindo Chemical Co. (Japan) and was used as received. Glucose oxidase (GOD, Type II, 15,500 U g−1 from Aspergillus niger) was purchased from Sigma. Dulbecco’s phosphate buffer solution (PBS) (containing 81 mM anhydrous disodium phosphate, 19 mM anhydrous mono-patassium phosphate, 137 mM sodium chloride and 3 mM potassium chloride in water, pH = 7.4), used for the enzyme immobilization work, was freshly prepared. BioRad dye reagent for protein assay (Catalog No. 500-0006) was obtained from BioRad (USA). Toluene-4-sulfonic acid (TSA) was from Fluka. Glucose stock solutions were allowed to mutarotate overnight. The solvents and other reagents were of analytical grade and were used without further purification. The synthesis of N-(2-carboxyl-ethyl)-N -(4-vinylbenzyl)4,4 -bipyridinium dichloride was carried out according

(a) Reaction scheme for the oxidation of glucose in the presence of O 2

H C HO

CH2OH C H OH C H

O H C OH

OH C + O2 H

Glucose oxidase

H C HO

CH2OH C H OH C H

O H C OH

C

O + H 2O 2

Gluconolactone

Glucose

Electrode

(b) Reaction scheme for the oxidation of glucose in the absence of O2 using a redox mediator MEDred -e-

GO(FAD) +e

MEDox

-

-e-

Glucose +e

GO(FADH2)

-

- e-

Gluconolactone

Scheme 1. (a) Reaction scheme for the oxidation of glucose in the presence of O2 . (b) Reaction scheme for the oxidation of glucose in the absence of O2 using a redox mediator.

X. Liu et al. / Biosensors and Bioelectronics 19 (2004) 823–834

825

(a) Synthesis of N-(2-carboxyl-ethyl) N'-(4-vinylbenzyl)-4, 4' bipyridinium dichloride (CVV) HOOC

CH2 CH2 Cl

N

+

N

Reflux in acetonitrile, 60h -

Cl HOOC

CH2 CH2

N

o

Methanol, 75 C, 24h

Cl

CH2 CH2

CH2

CH

CH2

-

-

HOOC

(I)

N

Cl

Cl

N

N

CH2

CH

(CVV)

CH2

(b) Schematic presentation of graft polymerization of CVV on PPY film, preactivation with WSC and glucose oxidase immobilization C N R' CVV

O

N R COOH

UV-induced graft polymerization

C

O

C HN

Preactivation with WSC

N R' R

PPY

H2N

O C

NH

GOD

GOD

Scheme 2. (a) Synthesis of N-(2-carboxyl-ethyl) N -(4-vinylbenzyl)-4,4 -bipyridinium dichloride (CVV). (b) Schematic presentation of graft polymerization of CVV on PPY film, preactivation with WSC and glucose oxidase immobilization.

to standard procedures for viologen synthesis (Monk, 1998) and is shown in Scheme 2a. The reaction between 3-chloropropionic acid and 4,4 -bipyridine in a 0.9:1 molar ratio was carried out at reflux in acetonitrile for 60 h. The precipitate (I), (N-(2-carboxyl-ethyl)-4,4 -bipyridinium chloride, C13 H13 N2 O2 Cl), was collected, washed with acetone three times to remove the unreacted starting materials, and dried under vacuum. The elemental analysis of (I) gave the following weight percentages: C, 58.63; H, 5.23; N, 10.72. These values are close to the theoretical values expected for C13 H13 N2 O2 Cl (C, 58.93; H, 4.91; N, 10.5). 4-Vinyl benzyl chloride was then reacted with (I) at a 1.1:1 molar ratio in methanol at 75 ◦ C. After 24 h, the reaction was stopped and the solvent was removed by rotary evaporation. The product (CVV, C22 H22 N2 O2 Cl2 ) was washed with

excess toluene (which is a good solvent for VBC and (I)). The results of the C, H, N (wt.%) analysis of the product were as follow: C, 62.78; H, 5.55; N, 7.14. The results are again consistent with those expected of C22 H22 N2 O2 Cl2 : C, 63.25; H, 5.27; N, 6.70. Both products ((I) and CVV) were also characterized with XPS as discussed in a later section. Electrochemical polymerization of pyrrole was carried out in an Autolab-PGSTAT30 (Metrohm-Schmidt Ltd.) at 0 ◦ C under an argon atmosphere in a conventional three-electrode cell. An electrolyte solution of 0.1 M pyrrole and 0.1 M TSA in acetonitrile containing 1 vol.% water was used. A highly polished stainless steel plate served as the working electrode (anode) while a platinum wire gauze served as the counter electrode (cathode), and an Ag/AgCl elec-

826

X. Liu et al. / Biosensors and Bioelectronics 19 (2004) 823–834

trode was used as the reference electrode. PPY was electrodeposited on the stainless steel plate by applying a constant voltage of 8 V for 40 min to obtain doped PPY films of approximately 50 ␮m in thickness (Choi and Tachikawa, 1990). 2.2. Enzyme electrode preparation PPY film strips of about 2 cm × 2 cm were used in all experiments. To achieve graft polymerization with CVV, an aqueous CVV solution of a predetermined concentration was placed on the surface of the PPY film which was then sandwiched between two pieces of quartz plates. Following the procedure described in a previous study (Liu et al., 2002a), the assembly was exposed to near-UV irradiation in a Riko rotary photochemical reactor (RH400-10W) for 1.5 h. The reactor was equipped with a 1000 W high-pressure Hg lamp and a constant temperature water bath. All the UV-induced graft polymerization experiments were carried out at a constant temperature of 28 ◦ C. The graft polymerized substrate was extracted from the quartz plates after prolonged immersion in water and then subjected again to thorough washing with water to remove the viologen which was not graft polymerized on the substrates. The viologen grafted films (CVV-g-PPY) were then dried by pumping under reduced pressure and stored in the dark until further use. The covalent immobilization of GOD onto the CVV-graft polymerized film was facilitated by the activation of the carboxylic groups grafted on the film surface. The COOH groups were preactivated for 1 h with WSC at 4 ◦ C in PBS, containing 5 mg/ml of WSC. The films were then transferred to the PBS solution with 0.02 M CaCl2 added (PBS(+)) containing GOD at a concentration of 4 mg/ml. The immobilization was allowed to proceed at 4 ◦ C for 24 h. After that, the reversibly bound GOD was desorbed in copious amounts of PBS(+) for 1 h at 25 ◦ C (Kulik et al., 1993). The process for the immobilization of GOD is summarized in Scheme 2b. The GOD immobilized PPY films will be denoted as GOD-CVV-g-PPY in the following discussion. The GOD-CVV-g-PPY film was used as an enzyme electrode to test its response to glucose solutions. The enzyme electrode served as the anode in the electrochemical set-up mentioned earlier. The glucose solution was first thoroughly degassed by bubbling with pure argon. The cyclic voltammetric (CV) experiments were then carried out at room temperature under an argon atmosphere. The amperometric response of the enzyme electrode in glucose solution of different concentration was investigated. For each reading, at least five different enzyme electrodes were tested and an average value was reported. 2.3. Testing and characterization Surface compositions were monitored with X-ray photoelectron spectroscopy (XPS) on an AXIS HSi spectrometer

(Kratos Analytical Ltd.) using the monochromatized Al K␣ X-ray source (1486.6 eV photons) at a constant dwell time of 100 ms and a pass energy of 40 eV. The anode voltage was 15 kV and the anode current was 10 mA. The pressure in the analysis chamber was maintained at 6.7 × 10−6 Pa or lower during each measurement. The polymer films were mounted on standard sample studs by means of double-sided adhesive tape. The core-level signals were obtained at a photoelectron take-off angle of 90◦ (with respect to the sample surface). To compensate for surface charging effect, all core-level spectra were referenced to the C 1s hydrocarbon peak at 284.6 eV. The peak area ratios for the various elements were corrected using experimentally determined instrumental sensitivity factors. The amount of GOD immobilized on PPY film was determined by the modified dye interaction methods (Bonde et al., 1992; Kang et al., 1993), using the BioRad protein dye reagent. For the preparation of the dye solution, the BioRad stock dye solution was diluted five times with doubly-distilled water. GOD solution (100 ␮l) of known concentration was added to 5 ml of the dye solution. The GOD–dye solution was kept for 3 h and centrifuged at 5000 rev/min for 15 min. In the latter process, the GOD–dye complexes were precipitated and the free dye remained in the upper layer. The absorbance of the supernatant at 465 nm was used for the standard calibration. For the quantitative determination of immobilized GOD, the dye solution (5 ml) was added to a test tube and the GOD-functionalized PPY film (2 cm × 2 cm) was immersed into the dye solution. After 3 h of reaction, the film was removed and the absorbance of the dye solution was measured at 465 nm. The amount of GOD immobilized on the surface of PPY film was calculated on the basis of the standard calibration. The CVV-g-PPY (without GOD) was also tested to determine the extent of dye interaction. The amount of dye bound to this film was not significant compared to the corresponding amount on the GOD-CVV-g-PPY film. 2.4. Assay of GOD activity For the investigation of the activity of immobilized GOD, 5 ml ␤-d(+)-glucose solution (18 wt.%) was used as the assay mixture and the activity was calculated by measuring the difference in the concentrations of the ␤-d(+)-glucose solution before and after a GOD-CVV-g-PPY film was immersed for 30 min. The experiment was carried out in an open beaker at 25 ◦ C. The concentration of the ␤-d(+)-glucose solutions were measured by the YSI Model 2700 SELECT Biochemistry Analyzer (YSI Incorporated, USA). The activity of the immobilized GOD in units is defined as the number of ␮mol of ␤-d-glucose oxidized to d-gluconolactone per minute. The relative activity (RA) is defined as the ratio of the observed surface enzyme activity over the activity obtained from an equivalent amount of the free enzyme. The activity of the free enzyme was determined with a similar glucose solution using the YSI Analyzer.

X. Liu et al. / Biosensors and Bioelectronics 19 (2004) 823–834

827

3. Results and discussions

3.2. Graft polymerization of CVV

3.1. XPS analysis of viologen monomer

The success of the UV-induced surface graft polymerization of CVV on the PPY film can be ascertained by comparing the XPS spectra before and after grafting process. The XPS C 1s core-level spectrum of the pristine PPY (Fig. 2a) indicates a small amount of oxidized C, primarily C–O (286.2 eV) and C=O (287.6 eV). The presence of these species may have resulted from surface oxidation or charge transfer complexing with oxygen. These species are desirable for the subsequent surface grafting process. The N 1s core-level spectrum of the pristine PPY (Fig. 2b) shows a predominant peak at 399.4 eV assigned with the –NH– species and a high binding energy tail above 400 eV attributable to N+ . The N+ /N ratio of 0.32 is consistent with the doping level of PPY reported in a previous study (Zhang et al., 1996). After CVV-graft polymerization on the PPY film, a peak at about 288.7 eV associated with the carboxylic acid group is observed in the C 1s spectrum (Fig. 2c). In Fig. 2d, the distinct peak at 401.7 eV indicates that N+ has become the dominant species and this is attributed to CVV. The peak at 399.4 eV could be associated with the –NH– from PPY as well as the N• formed during X-ray excitation of CVV in the analysis chamber. The presence of the Cl− signal in the Cl 2p core-level spectrum (not shown) of the CVV-PPY surface could be taken as further evidence of the presence of CVV on the PPY film since this signal is absent from the pristine PPY. A qualitative indication of the extent of surface grafting of CVV can be obtained from the [Cl]/[N] mole ratio, where the Cl is attributed to the Cl− of CVV, while the total N is due to the PPY film and CVV. In Fig. 3, the change in

The results of the XPS analysis of the as-synthesized mono-quaternized (I) and diquaternized bipyridine (CVV) are shown in Fig. 1. The respective C 1s core-level spectra (Fig. 1a and c) can be curve-fitted with four peak components, with binding energies at 284.6, 285.5, 286.2 and 288.7 eV assigned to the C–H, C–N, C–O and O–C=O species, respectively. The presence of the carboxyl group (O–C=O) is clearly associated with the 3-chloropropionic acid. The Cl 2p core-level spectra of both (I) and (CVV) (not shown) show that the Cl exists fully in the Cl− state (Cl 2p3/2 peak at a binding energy of 197.1 eV) which suggests there is no halide remaining in either product. In Fig. 1b, the N 1s core-level spectrum of (I) can be fitted with three peaks. The peak at 401.7 eV is assigned to the positively charged nitrogen (N+ ), that at 399.5 eV is attributed to the viologen radical cation (N• ) formed during X-ray excitation in the analysis chamber, and finally the peak at 398.6 eV is assigned to the unreacted imine nitrogen (–N=) of the pyridine rings. The –N=/total N ([–N=]/[N]) ratio of 0.49 indicates that the pyridine rings in (I) have been mono-quaternized. The total Cl/total N ([Cl]/[N]) ratio of 0.48 and [Cl− ]/[N+ ] ratio of 1.03 provide further confirmation. After the further reaction of (I) with VBC, the unreacted imine nitrogen component is absent and the positively charged nitrogen (N+ ) becomes the dominant species in the N 1s spectrum (Fig. 1d). The [Cl]/[N] and [Cl− ]/[N+ ] ratios are almost equal to 1.0, implying the success of the di-quaternization of the bipyridine.

(a)

C 1s

(b)

N 1s

Intensity (arb. units)

C-C, C-H N

C-N C-O

(c)

N+

-N= O-C=O

(d)

284

286

288

398

400

402

Binding Energy (eV) Fig. 1. XPS C 1s and N 1s core-level spectra of as-synthesized mono-quaternized (I) bipyridine (a and b), and diquaternized bipyridine (CVV) (c and d).

828

X. Liu et al. / Biosensors and Bioelectronics 19 (2004) 823–834

(a)

(b)

C 1s

N 1s

C-C, C-H N+

-NHC-O

Intensity (arb. units)

C=O

(c)

(d)

C-N O-C=O

(e)

(f)

NHC=O

284

286

288

398

400

402

Binding Energy (eV) Fig. 2. XPS C 1s and N 1s core-level spectra of pristine PPY surface (a and b), CVV-g-PPY surface (c and d) and GOD-CVV-g-PPY surface (e and f). Graft polymerization was carried out with 40 wt.% CVV in water under 90 min of UV-irradiation.

[Cl]/[N] ratio with CVV monomer concentration from 40 to 60 wt.% is shown. The lowest CVV monomer concentration shown in Fig. 3 is 40 wt.% since at concentrations below this value, the graft layer lacks homogeneity. On the other hand, the CVV monomer exhibits poor solubility at concentrations >60 wt.%. From Fig. 3, it can be seen that the [Cl]/[N] ratio increases almost linearly from 0.55 to 0.78. The theoretical [Cl]/[N] ratio expected for the CVV monomer is 1:1. This indicates that the PPY surface is covered to a large extent by CVV up to the probing depth of the XPS technique (∼7.5 nm). The ability to control the graft concentration by varying the monomer concentration is important since the extent of immobilization (see below) of GOD is dependent on the number of carboxyl groups in the CVV on the PPY surface. 3.3. Immobilization of GOD The XPS C 1s and N 1s core-level spectra of CVV-g-PPY surface after GOD immobilization are shown in Fig. 2e and f, respectively. The peak envelopes of both the C 1s and N 1s signals are significantly different from those of the

CVV-g-PPY before GOD immobilization (Fig. 2c and d). In Fig. 2e, an additional peak is present at 287.8 eV. This peak is attributed to the formation of the peptide linkage (HNC=O) between CVV and the GOD as well as the peptide bonds in GOD itself. The N 1s spectrum (Fig. 2f) also confirms the presence of the peptide linkages with gives rise to a strong signal at 399.4 eV. The surface concentration of immobilized GOD on the CVV-g-PPY film is expressed as the weight of immobilized GOD per surface area of the films, as determined using the protein–dye interaction method. When the CVV-graft concentration increases, the amount of GOD immobilized via the covalent coupling between the grafted CVV through the WSC intermediate is expected to increase. This is illustrated in Fig. 3 where the amount of covalently immobilized GOD is plotted against the CVV monomer concentration used in the graft polymerization. The enzyme activity of the GOD-CVV-g-PPY as a function of the CVV monomer concentration used in the graft polymerization is shown in Fig. 4. As CVV concentration increases from 40 to 60 wt.%, the observed enzyme activity increases. This increase in the observed activity must be

X. Liu et al. / Biosensors and Bioelectronics 19 (2004) 823–834

829

0.8 0.14

0.7 0.12

0.6

0.10

40

45

50

55

Immobilized GOD (mg/cm2)

Graft Concentration ([Cl]/[N])

0.16

60

CVV Monomer Concentration (wt. %) Fig. 3. [Cl]/[N] ratio as determined by XPS and the amount of GOD immobilized on the CVV-g-PPY surface as a function of the concentration of the aqueous CVV solution used in the graft polymerization process.

associated with the increase in the amount of surface immobilized enzyme as shown in Fig. 3. Although the observed activity increases with increase in the CVV monomer concentration, the relative activity of the immobilized enzyme

gradually decreases (Fig. 4). The decrease in the activity of the immobilized enzyme compared with that of the free enzyme is a common phenomenon observed with covalently immobilized enzyme. A probable reason is the analyte

0.6 35 0.5

30 0.4

Relative Activity (%)

Enzyme Activity ( mol/min.cm2)

40

25

0.3 40

45

50

55

60

CVV Monomer Concentration (wt. %) Fig. 4. Observed enzymatic activity and relative activity of the covalently immobilized GOD on CVV-g-PPY film as a function of CVV concentration used in the graft polymerization process.

830

X. Liu et al. / Biosensors and Bioelectronics 19 (2004) 823–834

approach to the active site of the enzyme is hindered since the enzyme is attached onto the surface. Another possibility is the minor modification in the enzyme tertiary structure that may be reflected in the distortion of amino acid residues involved in catalysis (Caliceti et al., 1993). Further inhibition of enzyme activity at high CVV concentration may also arise from the increase in spatial hindrance. Further experiments on the kinetic effect of immobilization and the electrochemical response were carried out with the GOD-CVV-g-PPY films prepared using the 40 wt.% of CVV monomer solution. 3.4. Kinetic effect of immobilization

Inverse Reaction Velocity, 1/ (min/mM)

In order to study the influence of the substrate on the catalytic activity of the immobilized enzyme on the CVV-g-PPY films, and to evaluate the kinetic effect of the immobilization, the rates of glucose oxidation reaction by the free and immobilized GOD were measured at various glucose concentrations. Fig. 5 shows the Lineweaver–Burk plots for the rates of glucose oxidation by the free and immobilized GOD in the glucose concentration ranging from 5 to 50 mM at pH of 7.4 and 30 ◦ C. In this experiment, 0.42 mg of free enzyme or a 2 cm ×2 cm GOD-CVV-g-PPY film, containing 0.105 mg/cm2 of immobilized GOD, was used. The plots give rise to straight lines which conform to the Michaelis–Menten equation (ν = νmax S/(Km + S), where ν is the velocity of the reaction (as measured by the change in the glucose concentration per unit time) and S is

20

Table 1 Kinetic parameters for the free and immobilized GOD GOD

Relative activity of the immobilized GOD (%)

Free Immobilized

– 36.0

Kinetic parameters νmax (mM/min)

Km (mM)

1.49 0.65

30.68 65.85

glucose concentration) for the enzyme-catalyzed reaction (Rosevear et al., 1987). The maximum reaction rate, νmax , and the apparent Michaelis constant, Km , determined from the linear regression of each plot in Fig. 5 are presented in Table 1. The Km value for the immobilized GOD on the CVV-g-PPY films is higher than that for the free GOD (P < 0.001). The difference in Km values between the free and immobilized GOD can be attributed to the limited accessibility of the active sites of the immobilized GOD to the glucose molecules and the conformational changes of GOD molecules caused by the covalent immobilization (Rosevear et al., 1987). From Table 1, it can be seen that the decrease in νmax , which is the limiting value of ν when all the active sites are occupied, is also significant (P < 0.001) as a result of immobilization. 3.5. Electrochemical behavior of the enzyme electrode The PPY film remains conductive after CVV functionalization and GOD immobilization (electrical conductivity of

Immobilized GOD Free GOD

15

10

5

0 0.05

0.10

0.15

0.20

Inverse Analyte Conc., 1/[S] (1/mM) Fig. 5. Lineweaver–Burk plots for immobilized GOD on CVV-g-PPY film and an equivalent amount of free GOD in pH 7.4 PBS solution at room temperature. Graft polymerization was carried out with 40 wt.% CVV in water under 90 min of UV-irradiation.

X. Liu et al. / Biosensors and Bioelectronics 19 (2004) 823–834

Current ( A/cm2)

the GOD-CVV-g-PPY film prepared using the 40% CVV monomer solution is 10 ± 0.2 S/cm). Hence, it can serve as an electrode, and the possible pathway for electron transfer from the glucose analyte to the electrode can be visualized as in Scheme 1b. The electrochemical response of the as-functionalized GOD-CVV-g-PPY electrode is discussed below. The cyclic voltammogram of a pristine PPY film in 0.1 M KCl at a scan rate of 0.1 V/s between 0 and −1.0 V versus Ag/AgCl is shown in Fig. 6a. No obvious redox peaks are present. Fig. 6b shows the CV for CVV-g-PPY in 0.1 M KCl at a scan rate of 0.1 V/s. The response of CVV-g-PPY is similar to that of a previous study in which the electrochemical characteristics of a viologen modified ITO electrode was investigated (Liu et al., 2002b). The typical redox peaks (reduction peak at −0.8 V, oxidation peak at −0.48 V) signify that the CVV-graft polymerized PPY films are electrochemically active. With a view to understand the glucose oxidation reaction in the presence of GOD immobilized on the PPY electrode, CV experiments were carried out with different concentrations of glucose in the buffer solution. The CVs in Fig. 7a were obtained with the GOD-CVV-g-PPY electrode in glucose solutions of different concentrations ranging from 0 to 1.2 mM in steps of 0.2 mM. The change in the shape of the CVs as the glucose concentration changes substantiates the

(a)

831

5 mA

(a)

1 mA

(b)

-0.2

0

-0.4

-0.6

-0.8

-1.0

Potential (V) vs. Ag/AgCl Fig. 6. Cyclic voltammogram of (a) pristine PPY film, and (b) CVV-g-PPY film in 0.1 M KCl. Scan rate = 0.1 V/s.

fact that GOD is active on the electrode and the CVV is able to mediate the GOD-glucose reaction. The peak currents at −0.8 V of the cyclic voltammograms shown in Fig. 7a were plotted as a function of glucose concentration in Fig. 7b. The peak current increases linearly with increasing glucose concentration indicating that the electrode kinetics is the

500

(b)

300 400 200 0

0.2

0.4

0.6

0.8

1.0

Glucose Concentration (mM)

Increasing glucose concentration

1 mA

0

-0.2

-0.4

-0.6

-0.8

-1.0

Potential (V) vs. Ag/AgCl Fig. 7. (a) Cyclic voltammograms of GOD-CVV-g-PPY film in glucose solution of 0, 0.2, 0.4, 0.6, 0.8, 1.0, 1.2 mM glucose. Scan rate = 0.1 V/s. PBS buffer solution with 0.1 M KCl was used as supporting electrolyte. (b) Peak currents at −0.8 V as a function of glucose concentration.

832

X. Liu et al. / Biosensors and Bioelectronics 19 (2004) 823–834

rate-limiting step. As a control, experiments were also conducted with a CVV-g-PPY film (without GOD) serving as the working electrode. In this case, the CVs obtained do not change with different glucose solutions. Further comparisons were also made with a GOD-AAcg-PPY electrode (acrylic acid (AAc) used instead of CVV as the anchor for GOD) made according to the previous method (Cen et al., 2003). In the preparation of the GOD-AAc-g-PPY film, the maximum AAc concentration which can be used without extensive formation of the poly(acrylic acid) homopolymer is 10 wt.%. At this AAc concentration, the amount of GOD loading is 0.03 mg/cm2 , with an activity of 0.11 ␮mol/min cm2 . In contrast, for the GOD-CVV-g-PPY, as mentioned earlier, a high CVV concentration (>40 wt.%) is desired so as to obtain a homogenous layer. Hence, the GOD loading in the GOD-CVV-g-PPY film is significantly higher than in the case of the GOD-AAc-g-PPY film. For the former, with a GOD loading of 0.10 mg/cm2 , an activity of 0.35 ␮mol/min cm2 was obtained. As a first approximation, the observed activity is roughly proportional to the amount of GOD immobilized, and the enzymatic activity of the GOD anchored via AAc or CVV normalized by the amount immobilized appears to be equivalent. This can be expected since in both cases the linkage is between the amine groups of GOD and the COO− groups of either AAc or CVV. However, the electrochemical response of the GOD-AAc-g-PPY and GOD-CVV-g-PPY electrodes in glucose solution is very different. In the former, the peak current response with a glucose concentration of 10 mM is <6 ␮A/cm2 , whereas for the latter, the peak current at the same glucose concentration

is more than 350 times higher. This enhancement in the amperometric response of the GOD-CVV-g-PPY electrode is likely to be due to the CVV acting as a mediator for efficient electron transfer between the GOD and the PPY electrode. The amperometric response of the GOD-CVV-g-PPY electrode as a function of glucose concentration from 0 to 50 mM is shown in Fig. 8a. A linear relationship was observed up to 20 mM glucose with a regression coefficient of 0.997. To test the stability of the GOD-CVV-g-PPY electrode, such an electrode was stored in the buffer solution at 4 ◦ C, and its composition and characteristics were tested after 10 days. The surface composition (monitored by XPS analysis) shows no substantial change, but a 10% loss in the amount of GOD and 20% loss of GOD activity are observed after the 10 days storage period. The amperometric response of this electrode is compared with that of the freshly made electrode in Fig. 8. Compared with the freshly made electrode, there is a 40% loss of sensitivity in the amperometric response after 10 days. The loss of GOD may be due to hydrolysis reaction (Briggs and Chandler, 1992) and the remaining GOD on the electrode may also be denatured to some degree during the long term storage in buffer solution (Wu et al., 1999). In comparison, PPY-based glucose biosensors made by electropolymerization of pyrrole in the presence of GOD (without a redox mediator) (Kojima et al., 1995; Trojanowicz and Hitchman, 1996; Tian and Zhu, 2002) were reported to give a peak current response (obtained with the freshly made biosensor) which is 100 times lower than those in the present work. In addition, one of the studies (Trojanowicz and Hitchman, 1996) reported that a 50% loss of sensitivity in the amperometric

5

Current (mA/cm2)

(a)

4 (b)

3

2

1

0 10

20

30

40

50

Glucose Concentration (mM) Fig. 8. Peak currents (measured at −0.8 V) as a function of glucose concentration using (a) GOD-CVV-g-PPY film, and (b) a similar film as in (a) after storage in buffer solution at 4 ◦ C for 10 days.

X. Liu et al. / Biosensors and Bioelectronics 19 (2004) 823–834

response was observed after 3 days. It can be concluded that the GOD adsorbed and trapped in PPY film is less efficient for electrochemical glucose sensing than GOD covalently wired via a viologen to the PPY film. Since the enzymatic and electrochemical reactions are coupled together by the redox mediator in the present work, the change in bulk glucose solution concentration can be detected by the electrode more directly with the electron mediator on the surface of the electrode, hence increasing the current response. The GOD trapped in the PPY matrix as reported in the earlier work (Trojanowicz and Hitchman, 1996) is also more easily lost by diffusion into the buffer solution.

4. Conclusions This study shows that PPY functionalized with viologen possessing appropriate functional groups can serve not only as a matrix for enzyme immobilization, but also as the electrode material for glucose sensing. The as-synthesized N-(2-carboxyl-ethyl)-N -(4-vinylbenzyl)-4,4 -bipyridinium dichloride monomer was graft polymerized on PPY film, and the graft concentration can be controlled by varying the monomer concentration. GOD was then covalently immobilized on the PPY surface via the carboxyl groups at the end of CVV chains. A high graft concentration of CVV will lead to more GOD immobilized but the relative activity of the immobilized GOD gradually decreases probably due to changes in the enzyme structure as a result of immobilization, as well as spatial hindrance. The CVV acts as an electron mediator to transport electron between the active sites of GOD and the PPY electrode. The electrochemical response of the as-functionalized enzyme electrode changes linearly in the range of 0–20 mM of glucose in solution and is two orders of magnitude higher than those obtained in previous studies. The present technique for immobilization of GOD is unique in terms of sensitivity, but further work is needed to optimize the device performance.

References Bartlett, P.N., Cooper, J.M., 1993. A review of the immobilization of enzymes in electropolymerized films. J. Electroanal. Chem. 362, 1–12. Belanger, D., Nadreau, J., Fortier, G., 1989. Electrochemistry of the polypyrrole glucose-oxidase electrode. J. Electroanal. Chem. 274, 143– 155. Bonde, M., Pontoppidan, H., Pepper, D.S., 1992. Direct dye binding—a quantitative assay for solid-phase immobilized protein. Anal. Biochem. 200, 195–198. Briggs, T., Chandler, A.M., 1992. Biochemistry, second ed. Springer, New York, pp. 19–31. Caliceti, P., Schiavon, O., Sartore, L., Monfardini, C., Veronese, F.M., 1993. Active-site protection of proteolytic-enzymes by poly(ethylene glycol) surface modification. J. Bioact. Compat. Pol. 8, 41–50. Cen, L., Neoh, K.G., Kang, E.T., 2003. Surface functionalization of polypyrrole film with glucose oxidase and viologen. Biosens. Bioelectron. 18, 363–374.

833

Cho, J.H., Shin, M.C., Kim, H.S., 1996. Electrochemical adsorption of glucose oxidase onto polypyrrole film for the construction of a glucose biosensor. Sens. Actuators, B 30, 137–141. Choi, C.S., Tachikawa, H., 1990. Electrochemical-behavior and characterization of polypyrrole copper phthalocyanine tetrasulfonate thin-film: cyclic voltammetry and in situ Raman-spectroscopic investigation. J. Am. Chem. Soc. 112, 1757–1768. Clark, L.C., Lyons, C., 1962. Electrode systems for continuous monitoring in cardiovascular surgery. Ann. New York Acad. Sci. 102, 29–45. Compagnone, D., Federici, G., Bannister, J.V., 1995. A new conducting polymer glucose sensor based on polythianaphthene. Electroanalysis 7, 1151–1155. Cooper, J.M., Bloor, D., 1993. Evidence for the functional mechanism of a polypyrrole glucose oxidase electrode. Electroanalysis 5, 883–886. Fisher, R.J., Fenton, J.M., Iranmahboob, J., 2000. Electro-enzymatic synthesis of lactate using electron transfer chain biomimetic membranes. J. Membr. Sci. 177, 17–24. Gavrilov, A.B., Zueva, A.F., Efimov, O.N., Bogdanovskaya, V.A., Tarasvitch, M.R., 1993. New enzyme biosensor for determination of glucose. Synthetic Met. 60, 159–161. Heller, A., 1996. Amperometric biosensors. Curr. Opin. Biotechnol. 7, 50–54. Hodak, J., Etchenique, R., Calvo, E.J., Singhal, K., Bartlett, P.N., 1997. Layer-by-layer self-assembly of glucose oxidase with a poly(allylamine)ferrocene redox mediator. Langmuir 13, 2708–2716. Kang, I.K., Kwon, B.K., Lee, J.H., Lee, H.B., 1993. Immobilization of proteins on poly(methyl methacrylate) films. Biomaterials 14, 787– 792. Kenausis, G., Chen, Q., Heller, A., 1997. Electrochemical glucose and lactate sensors based on “wired” thermostable soybean peroxidase operating continuously and stably at 37 ◦ C. Anal. Chem. 37, 1054– 1060. Kojima, K., Nasu, H., Shimomura, M., Miyauchi, S., 1995. An interfering factor in the glucose oxidase membrane. Synthetic Met. 71, 2254– 2264. Kulik, E.A., Kato, K., Ivanchenko, M.I., Ikada, Y., 1993. Trypsin immobilization on to polymer surface through grafted layer and its reaction with inhibitors. Biomaterials 14, 763–769. Liu, X., Neoh, K.G., Zhao, L., Kang, E.T., 2002a. Surface functionalization of glass and polymeric substrates via graft copolymerization of viologen in an aqueous medium. Langmuir 18, 2914–2921. Liu, X., Neoh, K.G., Kang, E.T., 2002b. Viologen-functionalized conductive surfaces: physicochemical and electrochemical characteristics, and stability. Langmuir 18, 9041–9047. Monk, P.M.S., 1998. The Viologens: Physicochemical Properties, Synthesis and Applications of the Salts of 4,4 -Bipyridine. Wiley, Chichester, UK, p. 38. Okahata, Y., Tsuruta, T., Ijiro, K., Agira, K., 1989. Preparations of Langmuir–Blodgett films of enzyme lipid complexes-a glucose sensor membrane. Thin Solid Films 180, 65–72. Onda, M., Lvov, Y., Ariga, K., Kunitake, T., 1996. Sequential actions of glucose oxidase and peroxidase in molecular films assembled by layer-by-layer alternate adsorption. Biotechnol. Bioeng. 51, 163–167. Pandey, P.C., 1988. A new conducting polymer-coated glucose sensor. J. Chem. Soc. Farad. Trans. I 84, 2259–2265. Ramanathan, K., Ram, M.K., Verghese, M.M., Malhotra, B.D., 1996. Dielectric spectroscopic studies on polypyrrole glucose oxidase films. J. Appl. Polym. Sci. 60, 2309–2316. Rosevear, A., Kennedy, J.F., Cabral, J.M.S., 1987. Immobilized Enzymes and Cells. IOP, Philadelphia. Sadik, O.A., Wallace, G.G., 1993. Pulsed amperometric detection of proteins using antibody containing conducting polymers. Anal. Chim. Acta 279, 209–212. Schultz, J.S., 1991. Biosensors. Sci. Am. 265, 64–69. Shinohara, H., Chiba, T., Aizawa, M., 1988. Enzyme microsensor for glucose with an electro-chemically sunthesized enzyme polyaniline film. Sens. Actuators 13, 79–86.

834

X. Liu et al. / Biosensors and Bioelectronics 19 (2004) 823–834

Tian, F., Zhu, G., 2002. Bienzymatic amperometric biosensor for glucose based on polypyrrole/ceramic carbon as electrode material. Anal. Chim. Acta 451, 251–258. Trojanowicz, M., Hitchman, M.L., 1996. A potentiometric polypyrrolebased glucose biosensor. Electroanalysis 8, 263–267.

Wu, J., Suls, J., Sansen, W., 1999. Amperometric glucose sensor with enzyme covalently immobilized by sol–gel technology. Anal. Sci. 15, 1029–1032. Zhang, X., Kang, E.T., Neoh, K.G., 1996. Surface studies of pristine and surface-modified polypyrrole films. J. Appl. Polym. Sci. 60, 625–636.