- Email: [email protected]
Comparison by QCM and photometric enzymatic test of the biotin-avidin recognition on a biotinylated polypyrrole
Talanta 55 (2001) 981– 992 www.elsevier.com/locate/talanta
Comparison by QCM and photometric enzymatic test of the biotin-avidin recognition on a biotinylated polypyrrole A. Dupont-Filliard, M. Billon *, S. Guillerez, G. Bidan Laboratoire d’Electrochimie Mole´culaire et Structures des Interfaces UMR 5819 (CNRS-CEA-Uni6ersite´ J. Fourier), De´partement de Recherche Fondamentale sur la Matie`re Condense´e/CEA-Grenoble, 17, a6enue des Martyrs, 38054 Grenoble Cedex 9, France Received 19 February 2001; received in revised form 19 June 2001; accepted 27 June 2001
Abstract By gravimetric measurements using a quartz cristal microbalance (QCM), we have studied the immobilization of biotinylated glucose oxidase enzymes (B-GOx) bound through on an intermediate avidin layer to a biotinylated polypyrrole film. The aim is to assess the amount of B-GOx specifically anchored on the biotinylated polypyrrole/ avidin assembly thank to the biotin/avidin interaction between avidin and B-GOx. Indeed the estimated amount from the QCM measurement corresponds to the specific recognition of avidin/B-GOx added to a non-specific recognition (adsorption) of B-GOx. In order to discriminate these two phenomena, we have carried out a study by QCM of the anchoring of B-GOx on an avidin layer linked by adsorption to a polypyrrole free from biotin units. From QCM measurements we have deduced for the biotinylated polypyrrole/avidin assembly that the amount of B-GOx bound via the biotin/avidin interaction and those due to the avidin adsorption process correspond to 3.9 pmol cm − 2 (1.3 equivalent of B-Gox monolayer) and 1.4 pmol cm − 2 (0.46 equivalent of B-GOx monolayer) respectively. These values have been corroborated by measurements of the enzymatic activity of GOx. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Polypyrrole; Biotin; QCM; Enzymatic test
1. Introduction Over the last few years, the immobilization of biomolecules on an electrode surface has been the subject of numerous studies. Several methods have been used to anchor biomolecules (DNA, peptides or enzymes) such as adsorption [1], direct * Corresponding author. Tel.: +33-4-76884833; fax: + 334-76885145. E-mail address: [email protected] (S. Guillerez).
covalent binding [2], entrapment in a polymer matrix [3,4] or indirect binding by the use of intermediate systems [5–12]. A particularly attractive method consisted in the immobilization of biomolecules into electronic conducting polymers (ECPs) [13 –16]. The preparation of biosensors constituted by an ECP matrix incorporating biomolecules could be achieved in a single step by electropolymerization of monomers bearing biological entities at the end of a spacer arm. The biomolecules are irreversibly included at the sur-
0039-9140/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 0 3 9 - 9 1 4 0 ( 0 1 ) 0 0 4 9 0 - 8
982
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
face and in the bulk of the ECP matrix during its electrogeneration on the electrode. The advantage of the electrochemical synthesis is that it allows the control of the thickness of the ECP matrix and the manufacturing of the biosensor on a very small surface electrode, opening the possibility of developing miniaturized biosensors. In most cases, polypyrrole has been used because it could be electrosynthesized at a low oxidation potential in aqueous solutions which is compatible with most molecules of biological interest. Another way to immobilize biomolecules on a polypyrrole film consists in post-polymerization functionalization. This two-step process requires to synthesize first a precursor polymer (or copolymer) bearing reactive at the end of a spacer arm. In the second step, biological entities are covalently bound on the polymer by the chemical substitution of these reactive groups [17,18]. Recently, an alternative method has been developed based on the biotin/avidin affinity system [19– 22] where the biomolecules were anchored to a biotinylated polypyrrole film through an intermediate avidin layer. This presents the advantages of (1) incorporating biomolecules only at the surface of the polymer film and (2) avoiding the use of chemical reagents which may damage more fragile biological species. By comparison with the direct attachment of biomolecules to the polypyrrole matrix, this approach is also a versatile process for the immobilization of biological species since a large number of biosensors could be synthesized with the wide variety of commercially available biotin conjugates. For example, Cosnier et al. [21] have elaborated amperometric biosensors to glucose or catechol by the immobilization of biotinylated glucose oxidase (B-GOx) or polyphenol oxidase on the supramolecular scaffolding, biotinylated polypyrrole/avidine. These authors have shown that this immobilization method leads to an enhanced enzymatic activity compared to the classical situation where the enzymes were entrapped in the PCE matrix. By this post-functionalization mode, the biomolecules are anchored on the polymer matrix by specific recognition thank to the biotin/avidin interaction (specific recognition) and also by adsorption (non specific recognition). For this last
process, the weaker interactions between biomolecules and the molecular architecture could cause the desorption of biological species which will induced the decrease of the sensor activity with the time. That is why the aim of this study is to discriminate the amount of the B-GOx anchored specifically by affinity interaction of this one induced by non specific recognition via gravimetric measurements using a Quartz Crystal Microbalance (QCM). To this end, the amount of avidin linked by adsorption on a unbiotinylated polypyrrole film has been previously estimated. In principle, QCM may be a fast and very sensitive method to assess the recognition at the interface polymer/solution [23]. Indeed, this technique has proved to be an extremely sensitive mass sensor, able of measuring subnanogram levels [24] but experimental data need to be interpreted cautiously particularly in the case of complex systems [25,26] (multilayer, viscoelastic behavior…). Consequently, the quantity of active B-GOx enzymes has been also estimated from enzymatic tests and then compared with the values obtained from QCM data to ensure that the gravimetric detection is a suitable method for the characterization of the molecular architecture: biotinylated polypyrrole/avidin/B-GOx.
2. Experimental section
2.1. Material NaCN, LiAlH4, DCC (N,N%-dicyclohexylcarbodiimide), N-hydrosuccinimide were purchased from Aldrich. DMF, diethyl ether, chloroform, ethanol, dichloromethane, acetonitrile and hexane (HPLC grade) were obtained from Carlo Erba or Prolabo and used without purification. d-Biotin (Vitamin H), avidin (10.7 units/mg solid), glucose oxidase-biotinaminocaproyl labeled (B-GOx) (from Aspergillus niger, 138 units/mg), horseradish peroxydase (HRP) (type II, 200 U mg − 1), b-D-glucose, o-Tolidine were purchased from Sigma. Stock solutions of glucose were allowed to mutarotate at room temperature for 24 h before use and were kept refrigerated.
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
The 0.01 M phosphate buffer solution (PBS) used for avidin and B-GOx solutions was prepared by dissolution in 200 ml of purified water of one phosphate buffered saline tablet (Sigma). According to Sigma, this solution contained 0.01 M phosphate buffer, 2.7 mM KCl and 0.137 M NaCl. Its pH was ca. 7.4 without further adjustment. Concerning the phosphate buffer solution used for the reactional solution of enzymatic test, it was prepared in our laboratory and contained only 0.01 M of KH2PO4 and Na2HPO4 in distilled water (pH 6.0). Tetrabutylammonium hexafluorophosphate (TBAPF6) (Fluka purum) was dried under vacuum at 80 °C for two days and stored in a dessicator. 2.2. Instrumentation Proton NMR spectroscopy was performed at 200 MHz on a Bru¨ cker AM-200 spectrometer, using solvents as internal references. Chemical shifts (d) are reported in ppm down field from tetramethylsilane. Electrochemical experiments were performed with EG&G Princeton Applied Research 273 potentiostat controlled by computer with an EG&G software in a conventional three-electrode cell containing a Pt or Au disk (surface area, 0.2 cm2) as working electrode, a Pt counter electrode and an Ag/Ag+ (AgNO3 10 mM in CH3CN) reference electrode. An AT-cut quartz crystal of 9 MHz (EG&G Princeton Applied Research) coated with two identical Pt (or Au) electrodes (diameter, 5 mm) is mounted in an home-made Teflon cylindrical cell (Scheme 1). Only one side of the quartz crystal is in contact with the solution and the reagents added in the cell. The cell was kept open and its maximum volume was 1 ml. The quartz frequency was measured by an Quartz Crystal Analyzer (QCA 917) from Seiko EG&G. The sensitivity of the QCM was 1.1 ng Hz − 1 as determined by silver deposition method according to the Sauerbrey equation [27]. Before starting the experiments, the electrode surface was carefully cleaned in order to eliminate all impurities (with a solution 50:50 of aqueous hydrogen peroxide 30%: 1 M aqueous sulfuric acid).
983
Spectrophotometric assays were performed with a Bio-Logic modular UV-Vis spectrophotometer, model MOS-250.
2.3. Synthesis of the monomers All the reactions were carried out under an argon atmosphere.
2.3.1. Monomer I We have chosen to prepare this pyrrole monomer (Scheme 2) by another strategy than the one commonly described [20,21] and which presents the advantage of obtaining a more important yield. The synthesis of the biotinylated pyrrole (Scheme 3) was achieved by coupling an aminoalkylpyrrole and a biotin entity. The aminoalkylpyrrole synthesis was carried out using several substitution reactions starting from the pyrrolylpotassium salt. 12-(pyrrol-1-yl)dodecanol (2) and N-(v-tosyloxydodecyl)pyrrole (3) were synthesized by the method of Bidan et al. [28,29]. The procedure was repeated without modification and NMR data were in accordance with that reported.
Scheme 1. Cell used for QCM measurements.
984
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
Scheme 2. Structures of monomers [I] et [II].
12-(pyrrol-1-yl)dodecanitrile (4): To a solution of (3) (4.05 g, 10 mmol) in 100 ml of DMF was added sodium cyanide (1.47 g, 30 mmol) under vigorous stirring. The mixture was heated at 70 °C for 24 h. The red crude mixture was cooled,
poured into water and extracted with diethyl ether. The yellow organic phase was washed several times with distilled water, dried over anhydrous sodium sulfate and evaporated to dryness. The resulting oil was purified by chromatography on silica gel eluted with a solution of chloroform: ethanol (95:5) to provide 2.47 g (95%) of an yellowish oil. H1-NMR (CDCl3): l 6.64 (t, 2H); 6.12 (t, 2H); 3.85 (t, 2H); 2.32 (t, 2H); 1.75 (m, 2H); 1.62 (m, 2H); 1.26 (m, 16H). 13-(pyrrol-1-yl)tridecylamine (5): (4) (1.04 g, 4 mmol) dissolved in 10 ml of ether was added slowly to a stirred suspension of LiAlH4 (0.45 g, 12 mmol) in 10 ml of anhydrous ether. After 2 h of stirring, 3.2 ml of water, 3.2 ml of NaOH 15% and finally 10 ml of water were successively added to the reaction mixture. A gel was filtered off and washed twice with ether. The combined organics were washed several times with water, dried over
Scheme 3. Synthetic route for the functionalization of the monomer pyrrole biotin [I].
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
anhydrous sodium sulfate and the crude oil was purified by chromatography on silica gel with chloroform: ethanol (95:5) as eluent giving 0.76 g (72%) of a yellowish oil. 1H-NMR (CDCl3): l 6.62 (t, 2H); 6.11 (t, 2H); 3.83 (t, 2H); 2.65 (t, 2H); 1.73 (m, 2H); 1.25 (m, 2H); 1.24 (m, 18H). Activated ester of biotin (6): [30,31] DCC (3.92 g, 19 mmol) was added to a mixture of d-biotin (3.66 g, 15 mmol) and N-hydrosuccinimide (1.73 g, 15 mmol) in 100 mL of DMF. The reaction mixture was stirred at room temperature for 2 days. After this period, DCU (N,N%-dicyclohexylurea) which was formed, was eliminated by filtration and the filtrate was evaporated to dryness. The resulting oil was vigorously stirred with 500 ml of diethyl ether for 2 h. The resulting precipitate was washed with ether and recrystallized in isopropanol to give 4 g (78%) of a white powder. 1 H-NMR (DMSO): l 1.40-1.70 (m, 6H); 2.55 (s, 4H); 2.63 (t, 2H); 2.76 (d, 1H); 2.82 (d, 1H); 3.10 (m, 1H); 4.11 (m, 1H); 4.25 (m, 1H). Monomer [I]: A mixture of (5) (0.74 g, 2.8 mmol) and (6) (1.16 g, 3.4 mmol) in 20 ml of DMF was stirred at room temperature for 2 days. The solvent was eliminated by rotary evaporation and the resulting oil was purified by column chromatography on silica gel with dichloromethane: ethanol (80:20) as eluent giving a yellow powder which was dissolved in a minimum of CH2Cl2. Addition of 700 ml of hexane led to the precipitation of [I] as a white solid, 1.1 g (80%). FAB-MS (m/z) 337 (M+); 1H-NMR (CDCl3): l 1.18 (m, 22H); 1.36 (m, 2H); 1.47 (m, 1H); 1.65 (m, 1H); 2.03 (t, 2H); 2.74 (d, 1H); 2.80 (d, 1H); 2.95 (m, 1H); 3.05 (t, 2H); 3.82 (t, 2H); 4.11 (m, 1H); 4.29 (m, 1H); 5.94 (t, 2H); 6.32 (s, 1H); 6.38 (s, 1H); 6.70 (t, 2H).
2.3.2. Monomer II This pyrrole monomer (Scheme 2) was synthesized by the method of Bidan et al. [28,29] reacting the pyrrolylpotassium salt with the 1-bromododecane in DMSO. 2.4. Enzyme electrode preparation and QCM measurement The procedure for constructing the modified
985
electrodes is schematically shown in Scheme 4. The poly[I] and poly[II] films were elaborated on the quartz electrode placed in QCM cell by controlled potential oxidation at 0.8 V of 5 mM monomer (I and II) in CH2Cl2 + 0.2 M TBAPF6. The modified electrodes were washed with acetonitrile and then transfered in H2O+ 0.1 M LiClO4 in order to replace the initially incorporated cation (TBA+) by the electrolyte cation (Li+) by repeatedly scanning the electrode potential at 50 mV/s on the electroactivity domain of the polypyrrole matrix. Then, these electrodes were plunged in 700 ml of PBS (pH 7.4). After stabilization of the frequency, binding of avidin to the electrode was performed by injecting 35 ml of avidin solution (2.5 g l − 1). After 5 min without stirring, the resulting electrodes were carefully rinsed with PBS, immersed again in 700 ml of a fresh PBS and after stabilization of the frequency, 35 ml of B-GOx solution (2.5 g l − 1) were injected. The B-GOx grafted electrodes were obtained after 5 min of incubation followed by careful washing with PBS. Avidin and B-GOx immobilization steps were monitored by gravimetric QCM measurements. The enzyme modified electrodes were kept refrigerated in PBS for further enzymatic tests.
2.5. Enzyme acti6ity measurement In order to estimate the amount of active enzymes retained at the surface of the poly[I] and poly[II] films, an enzymatic activity test of these enzymes was carried out. GOx activity was measured by an assay procedure of hydrogen peroxide based on the increase of absorbance at 420 nm generated by the o-Tolidine/HRP system (Scheme 4) [32]. The modified electrode containing GOx enzymes at its surface was plunged into the reactional solution in an UV tank under stirring. The enzymatic activity of the enzyme was determined from the slope of the linear part of the absorbance versus time dependence, by comparison with the calibration curves obtained under the same conditions with a known quantity of free enzyme in solution.
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
986
Scheme 4. Principle of elaboration of a QCM bioaffinity sensor coupled with an enzymatic detection.
3. Results and discussion As shown in Scheme 4, the biotinylated polypyrrole/avidin/B-GOx sensor studied in this paper is performed in three steps. The first (Scheme 4, step 1) is the electrosynthesis of biotinylated polypyrrole (poly[I]) on the electrode surface resulting from the electro-oxydation of pyrrole monomer I bearing a biotin entity (Scheme 2). Then, the biotin entities linked to the polypyrrole network are used as anchoring points for the immobilization of avidin units (Scheme 4, step 2), thanks to the high affinity of the biotin/ avidin interaction (binding constant, 1015 M − 1)
[33]. Only the biotin units at the surface of the biotinylated polypyrrole film are able to take part to the specific interaction with avidin. As a matter of fact, the bulkiness of avidin prevents its diffusion into the film. The avidin having four sites for biotin recognition, three sites remain available for further interaction with biotinylated biomolecules. By the way, biotinylated glucose oxidase enzymes (B-GOx) could be grafted on the avidin layer (Scheme 4, step 3). Thus the B-GOx molecules are anchored onto the polypyrrole via a double affinity system (specific recognition) (type 1, Scheme 5B). Nevertheless, B-GOx molecules could be also immobilized thank to non-specific
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
987
Scheme 5. Synoptic scheme of works: the various ways of interaction biomolecules/sensor are shown at each step of functionnalization together with the number of immobilized monolayers obtained by QCM and/or enzymatic measurements.
988
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
recognition either by affinity binding with previously adsorbed avidin (type 2, Scheme 5B) or just by adsorption at the biotinylated polymer surface (type 3, Scheme 5B). The study of the non-specific recognition has been carried out with a poly[II] film electrosynthetised from the monomer II (Scheme 2) no bearing a biotin molecule at the end of the spacer arm unlike the monomer I. The poly[II] polymer presents certainly an identical structure than poly[I] since the pyrrole unit in the monomers I and II bears the same alkyl chain.
3.1. Elaboration and characterization of poly(pyrrole-biotin) (poly[I]) and poly(pyrrole-alkyl) film (poly[II]) The electrosynthesis of poly[I] and poly[II] films was realized by controlled-potential oxidation at 0.8 V to allow an efficient polymerization and to preserve the film conductivity. The synthesis charge was limited to 20 mC/cm2. This value was chosen to fit the requirements of QCM measurements, i.e. an efficient coverage of the electrode surface and a low roughness. After synthesis, the resulting electrodes were transferred, after thorough rinsing, onto an electrolytic solution (CH2Cl2 + 0.2 M TBAPF6) free of monomer. The voltamperometric behavior of these polymers showed a reversible and stable response with a shape characteristic of that observed with classical ECPs. The E1/2 value for the both polymers was around a potential of 0.3 V which is in good agreement with those reported for similar N-substitued pyrroles [22,34]. However these films are electroinactive when they are transferred in aqueous electrolytic media in opposite to previously studies carried out with a hydrophilic biotinylated polypyrrole [19,20].
the QCM response of the quartz modified with the poly[I] film before and after the injection of avidin solution. Just after the injection of avidin, a fast decrease of frequency was instantaneously observed of an average value Zft = 145 Hz. The values and the average of assays are reported in Table 1. It should be noted that it was necessary to obtain a total stability of the QCM frequency before the beginning of the experiment. Moreover, the injection of the solution must be made with the minimal perturbation in order to maintain this frequency stability. The QCM result suggests that the injection of the protein into the solution influences significantly the mass of the quartz. Then we deduce that avidin units have been deposited onto the poly[I] film leading to the molecular system poly[I]/avidin (poly[I]/Av). The frequency change corresponded to an increase of mass of Zmt = 797 ng cm − 2 (11.8 pmol cm − 2) on the poly[I] electrode. Taking into account that the maximum coverage corresponding to a full active avidin monolayer was estimated to be 5.5 pmol cm − 2 [35], we can deduce than approximately 2.1
3.2. QCM measurements 3.2.1. Determination of the a6idin amount immobilized on poly[I] and poly[II] films (Scheme 5A) At first, poly[I] was electrogenerated on one side of a Pt quartz electrode of the QCM and after was placed in a PBS solution. Fig. 1 shows
Fig. 1. Frequency response of a quartz modified by a biotinylated poly[I] film in aqueous solution before (a) and after (b) the injection of an avidin solution. The zero frequency was arbitrarily set.
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
989
Table 1 Assays of QCM measurements of the detection of avidin and B-Gox immobilized on poly[I] and poly[II] polypyrrole films No. assay
Zfrequency/Hz of avidin adsorption on poly[II] (Zfads )
Zfrequency/Hz of B-GOx adsorption on poly[II]/adsorbed avidin (Zf %ads)
Zfrequency/Hz of avidin Zfrequency/Hz of B-GOx recognition on poly[I] (Zft ) recognition on poly[I]/avidin (Zf %t )
1 2 3 4 Average
35 36 40 30 35
41 47 50 40 45
136 147 155 140 145
equivalent of avidin monolayer have been immobilized on this film. The same experiment has been realized with an unbiotinylated poly[II] film that does not contain biotin entities. The use of the model polymer poly[II] allowed us to study the adsorption of avidin that occurs on this type of polymer film. In this case, the injection of avidin induced a less important variation of frequency Zfads =35 Hz (Fig. 2). This value corresponds to an increase of mass of Zmads =192 ng cm − 2 (2.8 pmol cm − 2). This decrease of the frequency is attributed only to the non-specific adsorption of avidin on the unbiotinylated film surface. Consequently the modified electrode poly[II]/adsorbed avidin (poly[II]/Avads) was formed. We can deduce than approximately 0.5 equivalent of avidin monolayer has been immobilized on this film. The adsorption contribution of 35 Hz was subtracted to the value obtained with the biotinylated poly[I] film to evaluate the frequency variation due to the specific recognition, Zfsp =110 Hz. This value corresponded to an increase of mass of approximately Zmsp =600 ng cm − 2 corresponding to 9 pmol cm − 2. We conclude that the poly[I] can bind by affinity 1.6 equivalent of avidin monolayer. This high amount of bound avidin ( \ of one molecular layer) may be due to the morphology of the film and particularly to its roughness which could increase the effective binding surface. Another explanation for this high value may be the high hydration level of molecules since the protein layer is known to be hydrophile.
190 176 161 151 170
3.2.2. Determination of B-GOx enzymse quantity immobilized onto poly[I] /a6idin and poly[II] /absorbed a6idin system (Scheme 5B) A biotinylated glucose oxidase (B-GOx) was chosen as model biomolecule to be immobilized on the resulting modified electrode poly[I]/Av and poly[II]/Avads. The modified poly[I]/Av and poly[II]/Avads electrodes were placed in an aqueous solution to which was added the B-GOx solution.
Fig. 2. Frequency response of a quartz modified by an electrogenerated unbiotinylated film poly [II] in aqueous solution before (a) and after (b) the injection of an avidin solution (-) and then biotinylated GOx (-). The zero frequency was arbitrarily set.
990
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
Fig. 3. Frequency response of a quartz modified by a biotinylated poly[I]/avidin film in aqueous solution before (a) and after (b) the injection of a biotinylated GOx solution. The zero frequency was arbitrarily set.
Injection of this solution onto poly[II]/Avads induced an instantaneous decrease of the frequency attributed to the adsorption of B-GOx on the film (Fig. 2). The average frequency change was Zf %ads = 45 Hz leading to an increase of mass Zm%ads = 247 ng cm − 2 corresponding to 1.4 pmol cm − 2. Taking into account that a GOx molecule occupies an area of 56 nm2 [36,37], a compact enzyme monolayer corresponds to a coverage of 3 pmol cm − 2. Consequently, approximately 0.46 equivalent of B-GOx monolayer have been immobilized on this film (type 2 and 3 binding Scheme 5B). The same experiment was realized on the biotinylated film, poly[I]/avidin. The injection of B-GOx induces a steeper variation of the frequency of Zf %t =170 Hz (Fig. 3). This behavior is attributed to the immobilization of B-GOx on the film (Scheme 5B) by the type 1 binding (specific binding between avidin/biotin) in addition to the type 2 and 3 binding (non-specific adsorption of B-GOx) onto the film. From the QCM value, the increase of mass has been estimated at Zm%t =935 ng cm − 2 corresponding to 5.2 pmol cm − 2 or approximately 1.74 equivalent of B-GOx
monolayer. The adsorption signal of 45 Hz was subtracted from this value to evaluate the frequency variation due to specific recognition between avidin and biotinylated GOx only. The relative value of this specific recognition of avidin/biotin was of Zf %sp = 125 Hz corresponding to an increase of mass Zm%sp = 687 ng cm − 2 (3.9 pmol cm − 2). Consequently, the poly[I]/Av can bind by affinity 1.3 equivalent of B-GOx monolayer. From these QCM results, we can observe that not all of avidin units anchored in the polypyrrole film were engaged in specific interaction with biotinylated GOx enzyme since 1.3 Gox monolayers were formed onto 1.6 avidin monolayers. In addition it appears that the B-GOx/Avidine ratio is higher for a low rate of coverage i.e. onto poly[II] than poly[I], ratio equal to 0.92 and 0.83 respectively. This could be due to the fact that a GOx molecule has a larger size (56 nm2) than that of avidin (33 nm2) inducing a cluttered surface.
3.3. Enzymatic acti6ity measurements 3.3.1. Determination of B-Gox enzyme quantity immobilized in poly[I] /a6idin system The ability of the poly[I] film for the soft anchoring of biotinylated macromolecular biomolecules via avidin/biotin bridge was examined by enzymatic activity measurements. Enzymatic activity of immobilized enzymes can be easily accessed and consequently allows to evaluate the quantity of enzyme molecules anchored on the surface of the polymer provided that all of the immobilized enzyme remains active (Scheme 4). As for the QCM measurements, B-GOx solution (2.5 g l − 1) was added in the PBS in which the modified electrodes poly[I]/Av were immersed. In presence of dioxygen, the immobilized B-GOx catalyses the oxidation of glucose with the production of H2O2. As a consequence, the enzymatic activity of the modified electrodes was evaluated by soaking the electrodes in phosphate buffer containing 50 mM glucose and all reactants for measuring glucose oxidation (see experimental section). A fully active B-GOx monolayer should exhibit a theoretical enzymatic activity of approximately 52 munits cm − 2 [21]. An increase of the
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
absorbance was measured for a poly[I]/Av/B-GOx electrode corresponding to an enzymatic activity of at = 100 munits cm − 2. This result corresponds to 5.7 pmol cm − 2 (1.9 monolayer) of active GOx have been immobilized on the film. It should be noted that this value is similar to that previously reported for polypyrrole-GOx electrodes [38,39] and for biosensors based on B-GOx anchored to avidin-modified surface [21,22].
3.3.2. Determination of B-Gox enzymes quantity immobilized in poly[II] /adsorbed a6idin system To determine the value corresponding to the specific recognition of avidin/B-GOx, a control experiment was carried out with a poly[II]/Avads electrode with B-GOx in order to evaluate the non-specific recognition. The enzymatic activity measured for a poly[II]/Avads/B-GOx electrode was aads =30 munits cm − 2 which corresponds to 1.7 pmol cm − 2 of GOx have been immobilized in the film (0.56 monolayer). Consequently, the enzymatic activity of B-GOx immobilized on poly[I] by type 1 binding could be evaluated of asp =70 munits cm − 2 which corresponds to 4 pmol cm − 2 or 1.35 equivalent of B-GOx monolayer (Scheme 5B). 3.4. General The values obtained for the amount of immobilized B-GOx evaluated by QCM measurements (3.9 pmol cm − 2) are in good agreement with those obtained by the enzymatic measurements (4.0 pmol cm − 2). It must be stated that QCM constitutes a reliable technique to provide quantitative values of biomolecules binding even in the case of multilayer devices as experimented in the present work. Moreover, from the B-GOx quantity immobilized specifically on the poly[I] film (type 1 binding, Scheme 5B) and this one due to the single affinity (type 2 and 3 binding, Scheme 5B), we can deduce that the B-GOx quantity anchored by non-specific interaction corresponds to 36% of the total amount of B-GOx immobilized. In addition, the value of ratio, number of B-GOx monolayer assessed by enzymatic test by this one of B-GOx monolayer evaluated by QCM measurements, is
991
comparable for the poly[I] and poly[II] (1.1 and 1.2 respectively). This suggests that the activity of the enzyme is almost independent of the type of avidin immobilization. In particular the spacing arm between biotin and the polypyrrole support in the case of poly[I] does not play a specific role. At last, from the B-GOx monolayers/avidin monolayers ratio, it appears that the enzymatic coverage rate is higher for the poly[II] than the poly[I].
4. Conclusion In this paper, the comparison of the results obtained from microgravimetric and enzymatic measurements demonstrates the successful utilization of QCM for the estimation of biomolecule amounts deposited on a polypyrrole film. From the QCM measurements we have also discriminated the B-GOx amount specifically anchored on the biotinylated polypyrrole film from this one due to the non-specific recognition. The enzymatic activity of GOx is maintained after its immobilization onto the avidin layer. Thus the anchoring of biomolecules on the polypyrrole film by postfunctionalization via the biotin/avidin interaction is a suitable way for the elaboration of biosensors. Furthers studies are in progress to immobilize other biomolecules such as DNA strands.
Acknowledgements We thank Dr. S. Cosnier (LEOPR/UMR CNRS 5630, Universite´ Joseph Fourier, BP 53, 38041 Grenoble Cedex 9, France) and Dr. P. Mailley for helpful discussion. Enzymatic assays were performed with the assistance of Dr. E. Mintz at the UMR-CNRS-5090, Biophysique Mole´ culaire et Cellulaire, DBMS/CEA/ Grenoble.
References [1] W.J. Albery, P.N. Bartlett, D.H. Craston, J. Electroanal. Chem. 194 (1985) 223. [2] R.M. Ianniello, T.J. Linsay, A.M. Tacynych, Anal. Chem. 54 (1982) 1980.
992 [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992 A. Heller, J. Phys. Chem. 96 (1992) 258. J.E. Frew, A.O. Hill, Anal. Chem. 59 (1987) 933A. T. Hoshi, J.-I. Anzai, T. Osa, Anal. Chem. 67 (1995) 770. M. Vreeke, P. Rocca, A. Heller, Anal. Chem. 67 (1995) 303. J.D. Wright, K.M. Rawson, W.O. Ho, P. Athey, C.J. McNeil, Biosensors & Bioelectronics 10 (1995) 495. P. He, J. Ye, Y. Fang, J. Anzai, T. Osa, Talanta 44 (1997) 885. P.N. Bartlett, J.M. Cooper, J. Electroanal. Chem. 362 (1993) 1. S. Lee, J. Anzai, T. Osa, Sensors and Actuators B 12 (1993) 153. P. Pantano, W.G. Kuhr, Anal. Chem. 65 (1993) 623. N. Dontha, W.B. Nowall, W.G. Kuhr, Anal. Chem. 69 (1997) 2619. F. Garnier, H. Korri-Youssoufi, P. Srivastava, A. Yassar, J. Am. Chem. Soc. 116 (1994) 8813. A. Emge, P. Ba¨ uerle, Synth. Met. 84 (1997) 213. W. Schuhmann, Mikrochim. Acta 121 (1995) 1. S. Cosnier, Electroanalysis 9 (1997) 894. H. Korri-Houssoufi, F. Garnier, P. Srivastava, S. Godillot, A. Yassar, J. Am. Chem. Soc. 119 (1997) 7389. H. Ro¨ ckel, J. Huber, R. Gleiter, W. Schuhmann, Adv. Mater. 7 – 8 (1994) 568. L.M. Torres-Rodriguez, M. Billon, A. Roget, G. Bidan, Synth. Metals 102 (1999) 1328. Torres-Rodriguez L.M., Roget A., Billon M., Livache T., Bidan G. J. Chem. Soc. Chem. Commun. 1998 1993. S. Cosnier, M. Stoytcheva, A. Senillou, H. Perrot, R.P.M. Furriel, F.A. Leone, Anal. Chem. 71 (1999) 3692.
[22] S. Cosnier, A. Lepellec, Electrochim. Acta 44 (1999) 1833. [23] S. Tombelli, M. Mascini, Anal. Lett. 33 (2000) 2129. [24] R. Lucklum, B. Henning, P. Hauptmann, K.D. Schierbaum, S. Vaihinger, W. Gopel, Sensors and Actuators A25 (1991) 705. [25] J. Rickert, A. Brecht, W. Gopel, Anal. Chem. 69 (1997) 1441. [26] C.C. White, J.L. Schrag, J. Chem. Phys. 111 (1999) 11192. [27] G. Sauerbrey, Z. Phys. 155 (1959) 206. [28] G. Bidan, Tetrahedron Lett. 26 (1985) 735. [29] G. Bidan, M. Guglielmi, Synth. Met. 15 (1986) 49. [30] E.A. Bayer, M. Wilchek, Methods Biochem. Anal. 261 (1980) 1. [31] E.A. Bayer, M. Wilchek, Methods Enzymol. 34 (1974) 267. [32] L. Coche-Guerente, S. Cosnier, C. Innocent, P. Mailley, Anal. Chim. Acta 311 (1995) 23. [33] M. Wilchek, E.A. Bayer, Anal. Biochem. 171 (1988) 1. [34] A.F. Diaz, J.I. Castillo, K.K. Kanazawa, J.A. Logan, M. Salmon, O. Fajardo, J. Electroanal. Chem. 133 (1982) 233. [35] N.M. Green, M.A. Joynson, Biochem. J. 118 (1970) 71. [36] H.J. Hecht, D. Schomberg, H. Kalisz, R.D. Schmid, Biosensors & Bioelectronics 8 (1993) 197. [37] C. Bourdillon, C. Demaille, J. Gueris, J. Moiroux, J.-M. Save´ ant, J. Am. Chem. Soc. 115 (1993) 12264. [38] N.C. Foulds, C.R. Lowe, J. Chem. Soc. Faraday Trans. I 182 (1986) 1259. [39] Y. Kajiya, H. Sugai, C. Iwakura, H. Yoneyama, Anal. Chem. 63 (1991) 49.
Comparison by QCM and photometric enzymatic test of the biotin-avidin recognition on a biotinylated polypyrrole A. Dupont-Filliard, M. Billon *, S. Guillerez, G. Bidan Laboratoire d’Electrochimie Mole´culaire et Structures des Interfaces UMR 5819 (CNRS-CEA-Uni6ersite´ J. Fourier), De´partement de Recherche Fondamentale sur la Matie`re Condense´e/CEA-Grenoble, 17, a6enue des Martyrs, 38054 Grenoble Cedex 9, France Received 19 February 2001; received in revised form 19 June 2001; accepted 27 June 2001
Abstract By gravimetric measurements using a quartz cristal microbalance (QCM), we have studied the immobilization of biotinylated glucose oxidase enzymes (B-GOx) bound through on an intermediate avidin layer to a biotinylated polypyrrole film. The aim is to assess the amount of B-GOx specifically anchored on the biotinylated polypyrrole/ avidin assembly thank to the biotin/avidin interaction between avidin and B-GOx. Indeed the estimated amount from the QCM measurement corresponds to the specific recognition of avidin/B-GOx added to a non-specific recognition (adsorption) of B-GOx. In order to discriminate these two phenomena, we have carried out a study by QCM of the anchoring of B-GOx on an avidin layer linked by adsorption to a polypyrrole free from biotin units. From QCM measurements we have deduced for the biotinylated polypyrrole/avidin assembly that the amount of B-GOx bound via the biotin/avidin interaction and those due to the avidin adsorption process correspond to 3.9 pmol cm − 2 (1.3 equivalent of B-Gox monolayer) and 1.4 pmol cm − 2 (0.46 equivalent of B-GOx monolayer) respectively. These values have been corroborated by measurements of the enzymatic activity of GOx. © 2001 Elsevier Science B.V. All rights reserved. Keywords: Polypyrrole; Biotin; QCM; Enzymatic test
1. Introduction Over the last few years, the immobilization of biomolecules on an electrode surface has been the subject of numerous studies. Several methods have been used to anchor biomolecules (DNA, peptides or enzymes) such as adsorption [1], direct * Corresponding author. Tel.: +33-4-76884833; fax: + 334-76885145. E-mail address: [email protected] (S. Guillerez).
covalent binding [2], entrapment in a polymer matrix [3,4] or indirect binding by the use of intermediate systems [5–12]. A particularly attractive method consisted in the immobilization of biomolecules into electronic conducting polymers (ECPs) [13 –16]. The preparation of biosensors constituted by an ECP matrix incorporating biomolecules could be achieved in a single step by electropolymerization of monomers bearing biological entities at the end of a spacer arm. The biomolecules are irreversibly included at the sur-
0039-9140/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S 0 0 3 9 - 9 1 4 0 ( 0 1 ) 0 0 4 9 0 - 8
982
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
face and in the bulk of the ECP matrix during its electrogeneration on the electrode. The advantage of the electrochemical synthesis is that it allows the control of the thickness of the ECP matrix and the manufacturing of the biosensor on a very small surface electrode, opening the possibility of developing miniaturized biosensors. In most cases, polypyrrole has been used because it could be electrosynthesized at a low oxidation potential in aqueous solutions which is compatible with most molecules of biological interest. Another way to immobilize biomolecules on a polypyrrole film consists in post-polymerization functionalization. This two-step process requires to synthesize first a precursor polymer (or copolymer) bearing reactive at the end of a spacer arm. In the second step, biological entities are covalently bound on the polymer by the chemical substitution of these reactive groups [17,18]. Recently, an alternative method has been developed based on the biotin/avidin affinity system [19– 22] where the biomolecules were anchored to a biotinylated polypyrrole film through an intermediate avidin layer. This presents the advantages of (1) incorporating biomolecules only at the surface of the polymer film and (2) avoiding the use of chemical reagents which may damage more fragile biological species. By comparison with the direct attachment of biomolecules to the polypyrrole matrix, this approach is also a versatile process for the immobilization of biological species since a large number of biosensors could be synthesized with the wide variety of commercially available biotin conjugates. For example, Cosnier et al. [21] have elaborated amperometric biosensors to glucose or catechol by the immobilization of biotinylated glucose oxidase (B-GOx) or polyphenol oxidase on the supramolecular scaffolding, biotinylated polypyrrole/avidine. These authors have shown that this immobilization method leads to an enhanced enzymatic activity compared to the classical situation where the enzymes were entrapped in the PCE matrix. By this post-functionalization mode, the biomolecules are anchored on the polymer matrix by specific recognition thank to the biotin/avidin interaction (specific recognition) and also by adsorption (non specific recognition). For this last
process, the weaker interactions between biomolecules and the molecular architecture could cause the desorption of biological species which will induced the decrease of the sensor activity with the time. That is why the aim of this study is to discriminate the amount of the B-GOx anchored specifically by affinity interaction of this one induced by non specific recognition via gravimetric measurements using a Quartz Crystal Microbalance (QCM). To this end, the amount of avidin linked by adsorption on a unbiotinylated polypyrrole film has been previously estimated. In principle, QCM may be a fast and very sensitive method to assess the recognition at the interface polymer/solution [23]. Indeed, this technique has proved to be an extremely sensitive mass sensor, able of measuring subnanogram levels [24] but experimental data need to be interpreted cautiously particularly in the case of complex systems [25,26] (multilayer, viscoelastic behavior…). Consequently, the quantity of active B-GOx enzymes has been also estimated from enzymatic tests and then compared with the values obtained from QCM data to ensure that the gravimetric detection is a suitable method for the characterization of the molecular architecture: biotinylated polypyrrole/avidin/B-GOx.
2. Experimental section
2.1. Material NaCN, LiAlH4, DCC (N,N%-dicyclohexylcarbodiimide), N-hydrosuccinimide were purchased from Aldrich. DMF, diethyl ether, chloroform, ethanol, dichloromethane, acetonitrile and hexane (HPLC grade) were obtained from Carlo Erba or Prolabo and used without purification. d-Biotin (Vitamin H), avidin (10.7 units/mg solid), glucose oxidase-biotinaminocaproyl labeled (B-GOx) (from Aspergillus niger, 138 units/mg), horseradish peroxydase (HRP) (type II, 200 U mg − 1), b-D-glucose, o-Tolidine were purchased from Sigma. Stock solutions of glucose were allowed to mutarotate at room temperature for 24 h before use and were kept refrigerated.
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
The 0.01 M phosphate buffer solution (PBS) used for avidin and B-GOx solutions was prepared by dissolution in 200 ml of purified water of one phosphate buffered saline tablet (Sigma). According to Sigma, this solution contained 0.01 M phosphate buffer, 2.7 mM KCl and 0.137 M NaCl. Its pH was ca. 7.4 without further adjustment. Concerning the phosphate buffer solution used for the reactional solution of enzymatic test, it was prepared in our laboratory and contained only 0.01 M of KH2PO4 and Na2HPO4 in distilled water (pH 6.0). Tetrabutylammonium hexafluorophosphate (TBAPF6) (Fluka purum) was dried under vacuum at 80 °C for two days and stored in a dessicator. 2.2. Instrumentation Proton NMR spectroscopy was performed at 200 MHz on a Bru¨ cker AM-200 spectrometer, using solvents as internal references. Chemical shifts (d) are reported in ppm down field from tetramethylsilane. Electrochemical experiments were performed with EG&G Princeton Applied Research 273 potentiostat controlled by computer with an EG&G software in a conventional three-electrode cell containing a Pt or Au disk (surface area, 0.2 cm2) as working electrode, a Pt counter electrode and an Ag/Ag+ (AgNO3 10 mM in CH3CN) reference electrode. An AT-cut quartz crystal of 9 MHz (EG&G Princeton Applied Research) coated with two identical Pt (or Au) electrodes (diameter, 5 mm) is mounted in an home-made Teflon cylindrical cell (Scheme 1). Only one side of the quartz crystal is in contact with the solution and the reagents added in the cell. The cell was kept open and its maximum volume was 1 ml. The quartz frequency was measured by an Quartz Crystal Analyzer (QCA 917) from Seiko EG&G. The sensitivity of the QCM was 1.1 ng Hz − 1 as determined by silver deposition method according to the Sauerbrey equation [27]. Before starting the experiments, the electrode surface was carefully cleaned in order to eliminate all impurities (with a solution 50:50 of aqueous hydrogen peroxide 30%: 1 M aqueous sulfuric acid).
983
Spectrophotometric assays were performed with a Bio-Logic modular UV-Vis spectrophotometer, model MOS-250.
2.3. Synthesis of the monomers All the reactions were carried out under an argon atmosphere.
2.3.1. Monomer I We have chosen to prepare this pyrrole monomer (Scheme 2) by another strategy than the one commonly described [20,21] and which presents the advantage of obtaining a more important yield. The synthesis of the biotinylated pyrrole (Scheme 3) was achieved by coupling an aminoalkylpyrrole and a biotin entity. The aminoalkylpyrrole synthesis was carried out using several substitution reactions starting from the pyrrolylpotassium salt. 12-(pyrrol-1-yl)dodecanol (2) and N-(v-tosyloxydodecyl)pyrrole (3) were synthesized by the method of Bidan et al. [28,29]. The procedure was repeated without modification and NMR data were in accordance with that reported.
Scheme 1. Cell used for QCM measurements.
984
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
Scheme 2. Structures of monomers [I] et [II].
12-(pyrrol-1-yl)dodecanitrile (4): To a solution of (3) (4.05 g, 10 mmol) in 100 ml of DMF was added sodium cyanide (1.47 g, 30 mmol) under vigorous stirring. The mixture was heated at 70 °C for 24 h. The red crude mixture was cooled,
poured into water and extracted with diethyl ether. The yellow organic phase was washed several times with distilled water, dried over anhydrous sodium sulfate and evaporated to dryness. The resulting oil was purified by chromatography on silica gel eluted with a solution of chloroform: ethanol (95:5) to provide 2.47 g (95%) of an yellowish oil. H1-NMR (CDCl3): l 6.64 (t, 2H); 6.12 (t, 2H); 3.85 (t, 2H); 2.32 (t, 2H); 1.75 (m, 2H); 1.62 (m, 2H); 1.26 (m, 16H). 13-(pyrrol-1-yl)tridecylamine (5): (4) (1.04 g, 4 mmol) dissolved in 10 ml of ether was added slowly to a stirred suspension of LiAlH4 (0.45 g, 12 mmol) in 10 ml of anhydrous ether. After 2 h of stirring, 3.2 ml of water, 3.2 ml of NaOH 15% and finally 10 ml of water were successively added to the reaction mixture. A gel was filtered off and washed twice with ether. The combined organics were washed several times with water, dried over
Scheme 3. Synthetic route for the functionalization of the monomer pyrrole biotin [I].
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
anhydrous sodium sulfate and the crude oil was purified by chromatography on silica gel with chloroform: ethanol (95:5) as eluent giving 0.76 g (72%) of a yellowish oil. 1H-NMR (CDCl3): l 6.62 (t, 2H); 6.11 (t, 2H); 3.83 (t, 2H); 2.65 (t, 2H); 1.73 (m, 2H); 1.25 (m, 2H); 1.24 (m, 18H). Activated ester of biotin (6): [30,31] DCC (3.92 g, 19 mmol) was added to a mixture of d-biotin (3.66 g, 15 mmol) and N-hydrosuccinimide (1.73 g, 15 mmol) in 100 mL of DMF. The reaction mixture was stirred at room temperature for 2 days. After this period, DCU (N,N%-dicyclohexylurea) which was formed, was eliminated by filtration and the filtrate was evaporated to dryness. The resulting oil was vigorously stirred with 500 ml of diethyl ether for 2 h. The resulting precipitate was washed with ether and recrystallized in isopropanol to give 4 g (78%) of a white powder. 1 H-NMR (DMSO): l 1.40-1.70 (m, 6H); 2.55 (s, 4H); 2.63 (t, 2H); 2.76 (d, 1H); 2.82 (d, 1H); 3.10 (m, 1H); 4.11 (m, 1H); 4.25 (m, 1H). Monomer [I]: A mixture of (5) (0.74 g, 2.8 mmol) and (6) (1.16 g, 3.4 mmol) in 20 ml of DMF was stirred at room temperature for 2 days. The solvent was eliminated by rotary evaporation and the resulting oil was purified by column chromatography on silica gel with dichloromethane: ethanol (80:20) as eluent giving a yellow powder which was dissolved in a minimum of CH2Cl2. Addition of 700 ml of hexane led to the precipitation of [I] as a white solid, 1.1 g (80%). FAB-MS (m/z) 337 (M+); 1H-NMR (CDCl3): l 1.18 (m, 22H); 1.36 (m, 2H); 1.47 (m, 1H); 1.65 (m, 1H); 2.03 (t, 2H); 2.74 (d, 1H); 2.80 (d, 1H); 2.95 (m, 1H); 3.05 (t, 2H); 3.82 (t, 2H); 4.11 (m, 1H); 4.29 (m, 1H); 5.94 (t, 2H); 6.32 (s, 1H); 6.38 (s, 1H); 6.70 (t, 2H).
2.3.2. Monomer II This pyrrole monomer (Scheme 2) was synthesized by the method of Bidan et al. [28,29] reacting the pyrrolylpotassium salt with the 1-bromododecane in DMSO. 2.4. Enzyme electrode preparation and QCM measurement The procedure for constructing the modified
985
electrodes is schematically shown in Scheme 4. The poly[I] and poly[II] films were elaborated on the quartz electrode placed in QCM cell by controlled potential oxidation at 0.8 V of 5 mM monomer (I and II) in CH2Cl2 + 0.2 M TBAPF6. The modified electrodes were washed with acetonitrile and then transfered in H2O+ 0.1 M LiClO4 in order to replace the initially incorporated cation (TBA+) by the electrolyte cation (Li+) by repeatedly scanning the electrode potential at 50 mV/s on the electroactivity domain of the polypyrrole matrix. Then, these electrodes were plunged in 700 ml of PBS (pH 7.4). After stabilization of the frequency, binding of avidin to the electrode was performed by injecting 35 ml of avidin solution (2.5 g l − 1). After 5 min without stirring, the resulting electrodes were carefully rinsed with PBS, immersed again in 700 ml of a fresh PBS and after stabilization of the frequency, 35 ml of B-GOx solution (2.5 g l − 1) were injected. The B-GOx grafted electrodes were obtained after 5 min of incubation followed by careful washing with PBS. Avidin and B-GOx immobilization steps were monitored by gravimetric QCM measurements. The enzyme modified electrodes were kept refrigerated in PBS for further enzymatic tests.
2.5. Enzyme acti6ity measurement In order to estimate the amount of active enzymes retained at the surface of the poly[I] and poly[II] films, an enzymatic activity test of these enzymes was carried out. GOx activity was measured by an assay procedure of hydrogen peroxide based on the increase of absorbance at 420 nm generated by the o-Tolidine/HRP system (Scheme 4) [32]. The modified electrode containing GOx enzymes at its surface was plunged into the reactional solution in an UV tank under stirring. The enzymatic activity of the enzyme was determined from the slope of the linear part of the absorbance versus time dependence, by comparison with the calibration curves obtained under the same conditions with a known quantity of free enzyme in solution.
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
986
Scheme 4. Principle of elaboration of a QCM bioaffinity sensor coupled with an enzymatic detection.
3. Results and discussion As shown in Scheme 4, the biotinylated polypyrrole/avidin/B-GOx sensor studied in this paper is performed in three steps. The first (Scheme 4, step 1) is the electrosynthesis of biotinylated polypyrrole (poly[I]) on the electrode surface resulting from the electro-oxydation of pyrrole monomer I bearing a biotin entity (Scheme 2). Then, the biotin entities linked to the polypyrrole network are used as anchoring points for the immobilization of avidin units (Scheme 4, step 2), thanks to the high affinity of the biotin/ avidin interaction (binding constant, 1015 M − 1)
[33]. Only the biotin units at the surface of the biotinylated polypyrrole film are able to take part to the specific interaction with avidin. As a matter of fact, the bulkiness of avidin prevents its diffusion into the film. The avidin having four sites for biotin recognition, three sites remain available for further interaction with biotinylated biomolecules. By the way, biotinylated glucose oxidase enzymes (B-GOx) could be grafted on the avidin layer (Scheme 4, step 3). Thus the B-GOx molecules are anchored onto the polypyrrole via a double affinity system (specific recognition) (type 1, Scheme 5B). Nevertheless, B-GOx molecules could be also immobilized thank to non-specific
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
987
Scheme 5. Synoptic scheme of works: the various ways of interaction biomolecules/sensor are shown at each step of functionnalization together with the number of immobilized monolayers obtained by QCM and/or enzymatic measurements.
988
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
recognition either by affinity binding with previously adsorbed avidin (type 2, Scheme 5B) or just by adsorption at the biotinylated polymer surface (type 3, Scheme 5B). The study of the non-specific recognition has been carried out with a poly[II] film electrosynthetised from the monomer II (Scheme 2) no bearing a biotin molecule at the end of the spacer arm unlike the monomer I. The poly[II] polymer presents certainly an identical structure than poly[I] since the pyrrole unit in the monomers I and II bears the same alkyl chain.
3.1. Elaboration and characterization of poly(pyrrole-biotin) (poly[I]) and poly(pyrrole-alkyl) film (poly[II]) The electrosynthesis of poly[I] and poly[II] films was realized by controlled-potential oxidation at 0.8 V to allow an efficient polymerization and to preserve the film conductivity. The synthesis charge was limited to 20 mC/cm2. This value was chosen to fit the requirements of QCM measurements, i.e. an efficient coverage of the electrode surface and a low roughness. After synthesis, the resulting electrodes were transferred, after thorough rinsing, onto an electrolytic solution (CH2Cl2 + 0.2 M TBAPF6) free of monomer. The voltamperometric behavior of these polymers showed a reversible and stable response with a shape characteristic of that observed with classical ECPs. The E1/2 value for the both polymers was around a potential of 0.3 V which is in good agreement with those reported for similar N-substitued pyrroles [22,34]. However these films are electroinactive when they are transferred in aqueous electrolytic media in opposite to previously studies carried out with a hydrophilic biotinylated polypyrrole [19,20].
the QCM response of the quartz modified with the poly[I] film before and after the injection of avidin solution. Just after the injection of avidin, a fast decrease of frequency was instantaneously observed of an average value Zft = 145 Hz. The values and the average of assays are reported in Table 1. It should be noted that it was necessary to obtain a total stability of the QCM frequency before the beginning of the experiment. Moreover, the injection of the solution must be made with the minimal perturbation in order to maintain this frequency stability. The QCM result suggests that the injection of the protein into the solution influences significantly the mass of the quartz. Then we deduce that avidin units have been deposited onto the poly[I] film leading to the molecular system poly[I]/avidin (poly[I]/Av). The frequency change corresponded to an increase of mass of Zmt = 797 ng cm − 2 (11.8 pmol cm − 2) on the poly[I] electrode. Taking into account that the maximum coverage corresponding to a full active avidin monolayer was estimated to be 5.5 pmol cm − 2 [35], we can deduce than approximately 2.1
3.2. QCM measurements 3.2.1. Determination of the a6idin amount immobilized on poly[I] and poly[II] films (Scheme 5A) At first, poly[I] was electrogenerated on one side of a Pt quartz electrode of the QCM and after was placed in a PBS solution. Fig. 1 shows
Fig. 1. Frequency response of a quartz modified by a biotinylated poly[I] film in aqueous solution before (a) and after (b) the injection of an avidin solution. The zero frequency was arbitrarily set.
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
989
Table 1 Assays of QCM measurements of the detection of avidin and B-Gox immobilized on poly[I] and poly[II] polypyrrole films No. assay
Zfrequency/Hz of avidin adsorption on poly[II] (Zfads )
Zfrequency/Hz of B-GOx adsorption on poly[II]/adsorbed avidin (Zf %ads)
Zfrequency/Hz of avidin Zfrequency/Hz of B-GOx recognition on poly[I] (Zft ) recognition on poly[I]/avidin (Zf %t )
1 2 3 4 Average
35 36 40 30 35
41 47 50 40 45
136 147 155 140 145
equivalent of avidin monolayer have been immobilized on this film. The same experiment has been realized with an unbiotinylated poly[II] film that does not contain biotin entities. The use of the model polymer poly[II] allowed us to study the adsorption of avidin that occurs on this type of polymer film. In this case, the injection of avidin induced a less important variation of frequency Zfads =35 Hz (Fig. 2). This value corresponds to an increase of mass of Zmads =192 ng cm − 2 (2.8 pmol cm − 2). This decrease of the frequency is attributed only to the non-specific adsorption of avidin on the unbiotinylated film surface. Consequently the modified electrode poly[II]/adsorbed avidin (poly[II]/Avads) was formed. We can deduce than approximately 0.5 equivalent of avidin monolayer has been immobilized on this film. The adsorption contribution of 35 Hz was subtracted to the value obtained with the biotinylated poly[I] film to evaluate the frequency variation due to the specific recognition, Zfsp =110 Hz. This value corresponded to an increase of mass of approximately Zmsp =600 ng cm − 2 corresponding to 9 pmol cm − 2. We conclude that the poly[I] can bind by affinity 1.6 equivalent of avidin monolayer. This high amount of bound avidin ( \ of one molecular layer) may be due to the morphology of the film and particularly to its roughness which could increase the effective binding surface. Another explanation for this high value may be the high hydration level of molecules since the protein layer is known to be hydrophile.
190 176 161 151 170
3.2.2. Determination of B-GOx enzymse quantity immobilized onto poly[I] /a6idin and poly[II] /absorbed a6idin system (Scheme 5B) A biotinylated glucose oxidase (B-GOx) was chosen as model biomolecule to be immobilized on the resulting modified electrode poly[I]/Av and poly[II]/Avads. The modified poly[I]/Av and poly[II]/Avads electrodes were placed in an aqueous solution to which was added the B-GOx solution.
Fig. 2. Frequency response of a quartz modified by an electrogenerated unbiotinylated film poly [II] in aqueous solution before (a) and after (b) the injection of an avidin solution (-) and then biotinylated GOx (-). The zero frequency was arbitrarily set.
990
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
Fig. 3. Frequency response of a quartz modified by a biotinylated poly[I]/avidin film in aqueous solution before (a) and after (b) the injection of a biotinylated GOx solution. The zero frequency was arbitrarily set.
Injection of this solution onto poly[II]/Avads induced an instantaneous decrease of the frequency attributed to the adsorption of B-GOx on the film (Fig. 2). The average frequency change was Zf %ads = 45 Hz leading to an increase of mass Zm%ads = 247 ng cm − 2 corresponding to 1.4 pmol cm − 2. Taking into account that a GOx molecule occupies an area of 56 nm2 [36,37], a compact enzyme monolayer corresponds to a coverage of 3 pmol cm − 2. Consequently, approximately 0.46 equivalent of B-GOx monolayer have been immobilized on this film (type 2 and 3 binding Scheme 5B). The same experiment was realized on the biotinylated film, poly[I]/avidin. The injection of B-GOx induces a steeper variation of the frequency of Zf %t =170 Hz (Fig. 3). This behavior is attributed to the immobilization of B-GOx on the film (Scheme 5B) by the type 1 binding (specific binding between avidin/biotin) in addition to the type 2 and 3 binding (non-specific adsorption of B-GOx) onto the film. From the QCM value, the increase of mass has been estimated at Zm%t =935 ng cm − 2 corresponding to 5.2 pmol cm − 2 or approximately 1.74 equivalent of B-GOx
monolayer. The adsorption signal of 45 Hz was subtracted from this value to evaluate the frequency variation due to specific recognition between avidin and biotinylated GOx only. The relative value of this specific recognition of avidin/biotin was of Zf %sp = 125 Hz corresponding to an increase of mass Zm%sp = 687 ng cm − 2 (3.9 pmol cm − 2). Consequently, the poly[I]/Av can bind by affinity 1.3 equivalent of B-GOx monolayer. From these QCM results, we can observe that not all of avidin units anchored in the polypyrrole film were engaged in specific interaction with biotinylated GOx enzyme since 1.3 Gox monolayers were formed onto 1.6 avidin monolayers. In addition it appears that the B-GOx/Avidine ratio is higher for a low rate of coverage i.e. onto poly[II] than poly[I], ratio equal to 0.92 and 0.83 respectively. This could be due to the fact that a GOx molecule has a larger size (56 nm2) than that of avidin (33 nm2) inducing a cluttered surface.
3.3. Enzymatic acti6ity measurements 3.3.1. Determination of B-Gox enzyme quantity immobilized in poly[I] /a6idin system The ability of the poly[I] film for the soft anchoring of biotinylated macromolecular biomolecules via avidin/biotin bridge was examined by enzymatic activity measurements. Enzymatic activity of immobilized enzymes can be easily accessed and consequently allows to evaluate the quantity of enzyme molecules anchored on the surface of the polymer provided that all of the immobilized enzyme remains active (Scheme 4). As for the QCM measurements, B-GOx solution (2.5 g l − 1) was added in the PBS in which the modified electrodes poly[I]/Av were immersed. In presence of dioxygen, the immobilized B-GOx catalyses the oxidation of glucose with the production of H2O2. As a consequence, the enzymatic activity of the modified electrodes was evaluated by soaking the electrodes in phosphate buffer containing 50 mM glucose and all reactants for measuring glucose oxidation (see experimental section). A fully active B-GOx monolayer should exhibit a theoretical enzymatic activity of approximately 52 munits cm − 2 [21]. An increase of the
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992
absorbance was measured for a poly[I]/Av/B-GOx electrode corresponding to an enzymatic activity of at = 100 munits cm − 2. This result corresponds to 5.7 pmol cm − 2 (1.9 monolayer) of active GOx have been immobilized on the film. It should be noted that this value is similar to that previously reported for polypyrrole-GOx electrodes [38,39] and for biosensors based on B-GOx anchored to avidin-modified surface [21,22].
3.3.2. Determination of B-Gox enzymes quantity immobilized in poly[II] /adsorbed a6idin system To determine the value corresponding to the specific recognition of avidin/B-GOx, a control experiment was carried out with a poly[II]/Avads electrode with B-GOx in order to evaluate the non-specific recognition. The enzymatic activity measured for a poly[II]/Avads/B-GOx electrode was aads =30 munits cm − 2 which corresponds to 1.7 pmol cm − 2 of GOx have been immobilized in the film (0.56 monolayer). Consequently, the enzymatic activity of B-GOx immobilized on poly[I] by type 1 binding could be evaluated of asp =70 munits cm − 2 which corresponds to 4 pmol cm − 2 or 1.35 equivalent of B-GOx monolayer (Scheme 5B). 3.4. General The values obtained for the amount of immobilized B-GOx evaluated by QCM measurements (3.9 pmol cm − 2) are in good agreement with those obtained by the enzymatic measurements (4.0 pmol cm − 2). It must be stated that QCM constitutes a reliable technique to provide quantitative values of biomolecules binding even in the case of multilayer devices as experimented in the present work. Moreover, from the B-GOx quantity immobilized specifically on the poly[I] film (type 1 binding, Scheme 5B) and this one due to the single affinity (type 2 and 3 binding, Scheme 5B), we can deduce that the B-GOx quantity anchored by non-specific interaction corresponds to 36% of the total amount of B-GOx immobilized. In addition, the value of ratio, number of B-GOx monolayer assessed by enzymatic test by this one of B-GOx monolayer evaluated by QCM measurements, is
991
comparable for the poly[I] and poly[II] (1.1 and 1.2 respectively). This suggests that the activity of the enzyme is almost independent of the type of avidin immobilization. In particular the spacing arm between biotin and the polypyrrole support in the case of poly[I] does not play a specific role. At last, from the B-GOx monolayers/avidin monolayers ratio, it appears that the enzymatic coverage rate is higher for the poly[II] than the poly[I].
4. Conclusion In this paper, the comparison of the results obtained from microgravimetric and enzymatic measurements demonstrates the successful utilization of QCM for the estimation of biomolecule amounts deposited on a polypyrrole film. From the QCM measurements we have also discriminated the B-GOx amount specifically anchored on the biotinylated polypyrrole film from this one due to the non-specific recognition. The enzymatic activity of GOx is maintained after its immobilization onto the avidin layer. Thus the anchoring of biomolecules on the polypyrrole film by postfunctionalization via the biotin/avidin interaction is a suitable way for the elaboration of biosensors. Furthers studies are in progress to immobilize other biomolecules such as DNA strands.
Acknowledgements We thank Dr. S. Cosnier (LEOPR/UMR CNRS 5630, Universite´ Joseph Fourier, BP 53, 38041 Grenoble Cedex 9, France) and Dr. P. Mailley for helpful discussion. Enzymatic assays were performed with the assistance of Dr. E. Mintz at the UMR-CNRS-5090, Biophysique Mole´ culaire et Cellulaire, DBMS/CEA/ Grenoble.
References [1] W.J. Albery, P.N. Bartlett, D.H. Craston, J. Electroanal. Chem. 194 (1985) 223. [2] R.M. Ianniello, T.J. Linsay, A.M. Tacynych, Anal. Chem. 54 (1982) 1980.
992 [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]
A. Dupont-Filliard et al. / Talanta 55 (2001) 981–992 A. Heller, J. Phys. Chem. 96 (1992) 258. J.E. Frew, A.O. Hill, Anal. Chem. 59 (1987) 933A. T. Hoshi, J.-I. Anzai, T. Osa, Anal. Chem. 67 (1995) 770. M. Vreeke, P. Rocca, A. Heller, Anal. Chem. 67 (1995) 303. J.D. Wright, K.M. Rawson, W.O. Ho, P. Athey, C.J. McNeil, Biosensors & Bioelectronics 10 (1995) 495. P. He, J. Ye, Y. Fang, J. Anzai, T. Osa, Talanta 44 (1997) 885. P.N. Bartlett, J.M. Cooper, J. Electroanal. Chem. 362 (1993) 1. S. Lee, J. Anzai, T. Osa, Sensors and Actuators B 12 (1993) 153. P. Pantano, W.G. Kuhr, Anal. Chem. 65 (1993) 623. N. Dontha, W.B. Nowall, W.G. Kuhr, Anal. Chem. 69 (1997) 2619. F. Garnier, H. Korri-Youssoufi, P. Srivastava, A. Yassar, J. Am. Chem. Soc. 116 (1994) 8813. A. Emge, P. Ba¨ uerle, Synth. Met. 84 (1997) 213. W. Schuhmann, Mikrochim. Acta 121 (1995) 1. S. Cosnier, Electroanalysis 9 (1997) 894. H. Korri-Houssoufi, F. Garnier, P. Srivastava, S. Godillot, A. Yassar, J. Am. Chem. Soc. 119 (1997) 7389. H. Ro¨ ckel, J. Huber, R. Gleiter, W. Schuhmann, Adv. Mater. 7 – 8 (1994) 568. L.M. Torres-Rodriguez, M. Billon, A. Roget, G. Bidan, Synth. Metals 102 (1999) 1328. Torres-Rodriguez L.M., Roget A., Billon M., Livache T., Bidan G. J. Chem. Soc. Chem. Commun. 1998 1993. S. Cosnier, M. Stoytcheva, A. Senillou, H. Perrot, R.P.M. Furriel, F.A. Leone, Anal. Chem. 71 (1999) 3692.
[22] S. Cosnier, A. Lepellec, Electrochim. Acta 44 (1999) 1833. [23] S. Tombelli, M. Mascini, Anal. Lett. 33 (2000) 2129. [24] R. Lucklum, B. Henning, P. Hauptmann, K.D. Schierbaum, S. Vaihinger, W. Gopel, Sensors and Actuators A25 (1991) 705. [25] J. Rickert, A. Brecht, W. Gopel, Anal. Chem. 69 (1997) 1441. [26] C.C. White, J.L. Schrag, J. Chem. Phys. 111 (1999) 11192. [27] G. Sauerbrey, Z. Phys. 155 (1959) 206. [28] G. Bidan, Tetrahedron Lett. 26 (1985) 735. [29] G. Bidan, M. Guglielmi, Synth. Met. 15 (1986) 49. [30] E.A. Bayer, M. Wilchek, Methods Biochem. Anal. 261 (1980) 1. [31] E.A. Bayer, M. Wilchek, Methods Enzymol. 34 (1974) 267. [32] L. Coche-Guerente, S. Cosnier, C. Innocent, P. Mailley, Anal. Chim. Acta 311 (1995) 23. [33] M. Wilchek, E.A. Bayer, Anal. Biochem. 171 (1988) 1. [34] A.F. Diaz, J.I. Castillo, K.K. Kanazawa, J.A. Logan, M. Salmon, O. Fajardo, J. Electroanal. Chem. 133 (1982) 233. [35] N.M. Green, M.A. Joynson, Biochem. J. 118 (1970) 71. [36] H.J. Hecht, D. Schomberg, H. Kalisz, R.D. Schmid, Biosensors & Bioelectronics 8 (1993) 197. [37] C. Bourdillon, C. Demaille, J. Gueris, J. Moiroux, J.-M. Save´ ant, J. Am. Chem. Soc. 115 (1993) 12264. [38] N.C. Foulds, C.R. Lowe, J. Chem. Soc. Faraday Trans. I 182 (1986) 1259. [39] Y. Kajiya, H. Sugai, C. Iwakura, H. Yoneyama, Anal. Chem. 63 (1991) 49.