Urea Biosensors Based on Immobilization of Urease into Two

Tetraborate acts as a competitive inhibitor for urease in the two different .... The rotation speed of the electrode was controlled by a Tacussel EDI ...
0 downloads 0 Views 107KB Size
Download PDF
Anal. Chem. 2002, 74, 4037-4043

Urea Biosensors Based on Immobilization of Urease into Two Oppositely Charged Clays (Laponite and Zn-Al Layered Double Hydroxides) J. V. de Melo,† S. Cosnier,‡ C. Mousty,*,‡ C. Martelet,*,† and N. Jaffrezic-Renault†

Laboratoire d’Inge´ nierie et Fonctionnalisation des Surfaces, UMR CNRS 5621, Ecole Centrale de Lyon, BP 163, 69131 Ecully Cedex, France, and Laboratoire d’Electrochimie Organique et de Photochimie Re´ dox, UMR CNRS 5630, Universite´ Joseph Fourier, BP 53, 38041 Grenoble Cedex 9, France

Enzyme-based field effect transistors (ENFETs) for urea determination were developed based on the immobilization of urease within two different clay matrixes, one cationic (Laponite) and the other anionic (layered double hydroxide (LDH)), cross-linked with glutaraldehyde. The biosensor based on the enzyme immobilized in Laponite shows a greater sensitivity and smaller dynamic linear range, because the enzymatic reaction is protected from the effect of the buffer capacity of the outer medium. The apparent Michaelis-Menten constant, Kapp m , is quite similar for both biosensors. Inhibition of the enzyme by sodium tetraborate was investigated. Tetraborate acts as a competitive inhibitor for urease in the two different types of clay, the inhibitor effect being stronger for the LDH/ urease biosensor. In particular, the maximum limit of the dynamic linear range extends from 1.4 mM in the absence of the inhibitor to 12 mM in the presence of 0.5 mM tetraborate. The Kapp m values in the presence of 0.5 mM tetraborate for Laponite and LDH biomembranes were 10 and 62 mM, respectively. Comparison of the inhibition constant values, Ki 0.16 and 0.05 mM for Laponite and LDH biosensors, respectively, clearly indicates a stronger enzyme-inhibitor interaction in the LDH/urease biomembrane. Many clinical cases dealing with kidney and liver diseases require a fast and accurate urea measurement in urine or blood samples. Because they are highly selective and simple to use in complex media, biosensors have usually been used for urea determination. They are generally based on urease, an enzyme that catalyses the conversion of urea to hydrogenocarbonate and ammonium through the following reaction: urease

(NH2)2CO + 2H2O + H+ 98 2NH4+ + HCO3- (1) The increasing demand for diagnostic testing paved the way for the elaboration of microbiosensors. In particular, the develop* To whom correspondence should be addressed: (e-mail) Christine.Mousty@ ujf-grenoble.fr; (fax) 33.4.76.51.42.67. (e-mail) [email protected]; (fax) 33.4.78.33.11.40. † Ecole Centrale de Lyon. ‡ Universite ´ Joseph Fourier. 10.1021/ac025627+ CCC: $22.00 Published on Web 07/18/2002

© 2002 American Chemical Society

ment of biosensors based on pH-sensitive field effect transistors (pH-FETs) has been growing over the past decades due to their potential advantages, such as reduced size, sensitivity, integration of the information-processing circuit on the same chip, low cost, and the facility for creating multisensors.1-4 Microbiosensors for urea determination have thus been prepared via urease immobilization on the surface of pH-FETs.3,5,6 One of the great challenges for enzyme-pH-FETs (ENFETs) is how to immobilize enzymes on these transducer surfaces. Besides the conventional methods of enzyme deposition such as covalent binding, physical adsorption, or cross-linking, enzyme entrapment has been widely used, producing coatings with high enzymatic activity.7 Among the various host materials, the use of inorganic materials such as oxides,8 clays,9-14 zeolites,15,16 and sol-gel materials17 for enzyme immobilization has been, in recent years, the subject of increasing research efforts. Clays are an attractive material for electrode functionalization of transducer surfaces because of their thermal stability, chemical inertia, well-defined layered structure, ion-exchange properties, and low cost. Swelling cationic clays, known as smectite clays, have been used in the elaboration of modified electrodes, in particular enzyme electrodes.18 On the other hand, since 1987, (1) Starodub, N. F.; Kanjuk, N. I.; Kukle, A. L.; Shiroshov, Y. M. Anal. Chim. Acta 1999, 385, 461-466. (2) Volotovsky, V.; Nam, Y. J.; Kim, N. Sens. Actuators 1997, B42, 233-237. (3) Poyard, S.; Gorchkov, D.; Jdanova, A.; Jaffrezic-Renault, N.; Martelet, C.; Soldatkin, A. P.; El’Skaya. CRAS-Se´ r. III 1996, 319, 257-262. (4) Soldatkin, A. P.; Volotovsky, V.; El’skaya, A. V.; Jaffrezic-Renault, N.; Martelet, C. Anal. Chim. Acta 2000, 403, 25-29. (5) Soldatkin, A. P.; Burbriak, O. A.; Starodub, N. F.; El’skaya, A. E.; Sandrovskii, A. K.; Shul’ga, A. A.; Strikha, V. I. Russ. J. Electrochem. 1993, 29, 279-283. (6) Kharitonov, A. B.; Zayats, M.; Lichtenstien, A.; Katz, E.; Wilnner, I. Sens. Actuators, B 2000, 70, 222-231. (7) Eggins, B. Biosensors, an introduction; Wiley-Teubner Chemistry: Chichester, U.K., 1996. (8) Liu, B.; Hu, R.; Deng, J. Anal. Chim. Acta 1997, 341, 161-169. (9) Lei, C.; Deng, J. Anal. Chem. 1996, 68, 3344-3349. (10) Shyu, S. C.; Wang, C. M. J. Electrochem. Soc. 1998 145, 154-158. (11) Poyard, S.; Jaffrezic-Renault, N.; Martelet, C.; Cosnier, S.; Labbe´, P. Anal. Chim. Acta 1998, 3364, 165-172. (12) Coche-Guerent, L.; Desprez, V.; Labbe, P. J. Electroanal. Chem. 1998, 458, 73-86. (13) Cosnier, S.; Lambert, F.; Stoytcheva, M. Electroanalysis 2000, 12, 356360. (14) Mousty, C.; Cosnier, S.; Shan, D.; Mu, S. Anal. Chim. Acta 2001, 443, 1-8. (15) Walcarius, A. Electroanalysis 1996, 8, 971-985. (16) Liu, B.; Hu, R.; Deng, J. Anal. Chem. 1997, 69, 2343-2348. (17) Wang, J. Anal. Chim. Acta 1999, 399, 21-27.

Analytical Chemistry, Vol. 74, No. 16, August 15, 2002 4037

several papers have described the alternative use of synthetic anionic clays for the modification of electrode surfaces. The latter are of special interest because a layered double hydroxide (LDH)type structure has an anion-exchange property and can act as a passive filter allowing preconcentration of analytes or redox mediators.18-22 Although LDH presents a lamellar structure and intercalation properties similar to cationic clays, such kinds of materials have never been exploited for the immobilization of biological macromolecules. The aim of the present work was to use an anionic clay coating, for the first time, to elaborate sensitive urea ENFETs. A comparative study of urea ENFET performance, based on the immobilization of urease in two synthetic clays, one Laponite Si8[Mg5.5Li0.5H4O24]0.7-Na0.7+, a cationic clay, and the other LDH Zn3Al(OH)8Cl‚2H2O referred to as [Zn-Al-Cl], an anionic clay, highlights the advantages of each kind of clay coating. In addition, we report the use of sodium tetraborate, a strong competitive inhibitor of urease, as an alternative to modulate the dynamic range of the clay/urease biosensors. EXPERIMENTAL SECTION Materials. Urease (EC 3.5.1.5) from jack beans with an activity of 150 units/mg and bovine serum albumin (BSA) were obtained from Sigma, Laponite was from Rockwood Specialties, and glutaraldehyde was from Fluka. The layered double hydroxide, Zn3Al(OH)8Cl‚2H2O, was synthesized by the coprecipitation method developed by De Roy.23 All other reagents were of pure analytical grade. pH-FETs were produced by ESIEE (Noisy-legrand, France). Enzyme Immobilization. The Laponite colloidal suspension was prepared by following the method described elsewhere.13 The Laponite (2.5 g/L) was dispersed overnight in deionized water under stirring conditions. The LDH colloidal suspension was prepared by following the same procedure using deionized and decarbonated water. Biomembranes were deposited on pH-FETs using the following method. To improve adhesion of the membrane, the pH-FET surface was treated with sulfochromic solution in order to obtain a clean, hydrophilic surface. Then the sensor was thoroughly washed with bidistilled water and dried under a nitrogen flow. A drop of a homogeneous mixture (2 µL) containing a 1:1 Laponite/ urease (2.5 g/L) was deposited onto the sensitive area of the pHFET. The device was placed in saturated glutaraldehyde vapor for the cross-linking of the membrane. After exposure to glutaraldehyde, the biomembranes were dried at room temperature for ∼30 min. Before measurements, the ENFETs were stored overnight at 4 °C in a 5 mM phosphate buffer solution, 1 mM EDTA, pH 7.4, to equilibrate the membrane system. A reference pH-FET (Laponite/BSA) was prepared using the same procedure by replacing urease by BSA, an inert protein. (18) Macha, S. M.; Fich, A. Mikrochim Acta 1998, 128, 1-18. (19) Therias, S.; Mousty, C. Appl. Clay Sci. 1995, 11, 147-162. (20) Therias, S.; Mousty, C.; Forano, C.; Besse, J.-P. Langmuir 1996, 12, 49144920. (21) Therias, S.; Lacroix, B.; Scho ¨llhon, B.; Mousty, C.; Palvadeau, P. J. Electroanal. Chem. 1998, 454, 91-97. (22) Ballarin, B.; Seeber, B.; Tonelli, D.; Zanardi, C. Electroanalysis 2000, 12, 434-441. (23) De Roy, A.; Forano, C.; El Molki, K.; Besse, J. P. In Expended clays and other microporous solids; Occelli, M. L., Robson, H. E., Eds.; Van Nostrand Reinhold: New York, 1992; pp 108-169.

4038 Analytical Chemistry, Vol. 74, No. 16, August 15, 2002

Figure 1. Calibrations curves of urea in 5.0 mM PB solution (pH 7.4) at Laponite/urease (1:1, 1 µL) biosensors as a function of crosslinking time (a) 8, (b) 15, and (c) 25 min and (d) Laponite/urease (1:2, 1 µL) cross-linking time 8 min.

For the LDH/urease biomembranes, the aforementioned procedure was repeated but modified by adding 3.5% glycerol to the LDH/urease mixtures (2.5:2.5 or 2.5:5.0 g/L). The crosslinking time and the volume of the mixture coated on the surface of the ENFETs were the same as those used for the Laponite/ urease biomembranes. For the rotating disk electrode experiments, 8 µL of the clay/ enzyme mixture was deposited on a glassy carbon electrode disk (3 mm) and the enzyme macromolecules were cross-linked by glutaraldehyde vapor for 8 min. The resulting coating was then dried in air at room temperature. Measurements. The biosensor response was measured at a controlled temperature (27 ( 0.3 °C). The following conditions were used for the potentiometric measurements. The pH-FETs were connected to an amplifier system (ISFETmeter from IMT, Neuchaˆtel, Suisse) allowing the source voltage (Vs) to be measured while drain current (Id) and drain voltage (Vd) remained constant. Vd was equal to 0.5 V, and Id was equal to 100 µA. Illumination had practically no influence on the characteristics of these biosensors. The biosensor response as a function of temperature was studied by using a differential setup because pHFETs are also sensitive to temperature. In this case, an inert protein, BSA, replaced the urease in the reference ENFET as described above. The experiments were performed in 1.0 mM EDTA phosphate buffer solution (5.0 mM, pH 7.4). The linear sweep voltammetry was carried out in a conventional three-electrode cell, using a PAR 273 EG&G potentiostat coupled to a computer. The rotation speed of the electrode was controlled by a Tacussel EDI 101T/CTV 101T. The counter electrode was a Pt wire. For all experiments, the reference electrode used was an Ag/ AgCl/KClsat electrode. RESULTS AND DISCUSSION Optimization of the Laponite/Urease Biomembrane. pHFETs were modified by simple adsorption of a Laponite/urease 1:1 aqueous mixture (1 µL) followed by chemical cross-linking of the incorporated enzyme molecules by glutaraldehyde. The latter step was carried out at different time intervals. Figure 1 shows the calibration curves obtained for the potentiometric response to urea of the resulting ENFETs. A decrease in the ENFET’s sensitivity and maximum potential response is clearly observed with the increase in cross-linking time (Table 1). This may be

Table 1. Analytical Characteristics of ENFET Biosensors Based on Laponite for Urea Determination Laponite/ urease ratioa

timeb (min)

max response (mV)

sensitivity (V/M)

linear range (mM)

R

1:1 1:1 1:1 1:2

8 15 25 8

69.5 40.5 22.3 65.0

53.22 27.30 13.59 33.82

0.005-0.55 0.01-0.65 0.03-0.55 0.05-0.77

0.9979 0.9974 0.9991 0.9992

a Coating elaborated from 1 µL of aqueous Laponite solution (2.5 g/L) containing urease. b Exposure time to glutaraldehyde vapor.

Figure 2. Calibrations curves of urea in 5.0 mM PB solution (pH 7.4) for different biosensor configurations (cross-linking time 8 min): Laponite/urease (1:1) (a) 1, (b) 2, (c) 3, and (d) 4 µL and (e) Laponite/ BSA (1:1, 2 µL).

ascribed to a decrease in the activity of the immobilized urease or to an increase in steric hindrances in the coating related to the high reticulation degree of the entrapped enzymes. Although a 20-min cross-linking time was previously reported for the stabilization of enzyme/Laponite coatings,11 the latter was thereafter fixed at 8 min. We previously reported that the biosensor’s performance increases with the increase in the enzyme/Laponite ratio from 1.1 to 3.3.11 We therefore produced a similar biosensor arrangement with double the amount of immobilized urease. Figure 1d represents the calibration curve obtained with a 1:2 Laponite/ urease mixture under 8-min cross-linking. The comparison between calibration curves a and d indicates that a proportion of 1:1 Laponite/urease and 8 min of cross-linking result in the best composition for this biomembrane. Figure 2 shows the influence of the amount of Laponite/urease mixture deposited on a pH-FET surface on biosensor performance, the Laponite/urease ratio and the cross-linking time for all membranes being 1:1 and 8 min, respectively. The maximum response, at saturating substrate conditions, first increases with the increase in the amount of coating from 1 to 2 µL and hence with the amount of immobilized enzyme, and then it seems to reach a constant value for thicker coatings. On the other hand, the response time of the biosensors increases strongly with the amount of coating, varying from 1.5 to 4 min when the deposited mixture increases from 1 to 4 µL, respectively. For this reason, 2 µL was chosen as the optimum deposited mixture for further experiments. Several reference pH-FET sensors have been prepared. Figure 2e corresponds to the calibration curve obtained with a pH-FET modified by Laponite containing an inert protein (BSA) instead of urease. No response was observed in the presence of urea.

Figure 3. Calibrations curves of urea in 5.0 mM PB solution (pH 7.4) for different biosensor configurations (cross-linking time 8 min): (a) Laponite/urease (1:1, 2 µL), (b) LDH/urease (1:1, 2 µL), and (c) LDH/urease (1:2, 2 µL).

This result demonstrates that the potentiometric responses of ENFETs are due to the activity of urease and not to an accumulation of urea within the clay. Moreover, a Laponite/urease (1:1, 2 µL) biomembrane elaborated without the glutaraldehyde crosslinking step appears to be totally inert in the presence of urea. This may be due to the complete release of enzyme into the buffer solution since the biosensors were stored overnight in aqueous solution before measurements. Indeed, as previously observed with other enzyme/clay electrodes, the slow release of the enzyme, molecules incorporated within clay films, has implied the chemical cross-linking of the incorporated enzymes by glutaraldehyde.13 Likewise, depositing an equivalent amount of urease cross-linked with glutaraldehyde, without Laponite, produces a biosensor exhibiting a maximum response of ∼14% of that obtained with the equivalent biosensor containing Laponite. This confirms the useful role of Laponite as an entrapment matrix or as a hydrophilic additive.24,25 Comparison between LDH/Urease and Laponite/Urease Biosensor Performance. A study was also carried out on the exposure time to glutaraldehyde vapor for the cross-linking of LDH/urease biomembranes, and 8 min was considered as ideal. The calibration curves a and b in Figure 3 correspond to two biomembranes having the same configuration (1:1, 2 µL). The maximum response for the ENFET based on LDH is ∼50% lower than that obtained with the corresponding ENFET based on Laponite while a less sensitive detection limit (5 µM) was recorded, highlighting the role of the clay structure (Table 2). In contrast to Laponite, the optimum LDH/enzyme configuration was obtained for a clay/enzyme proportion of 1:2. The latter enhances the maximum response (+40% vs the initial LDH/urease configuration) and the detection limit, namely, 3.5 µM, but the performance remains lower than that obtained with the Laponite/ urease biosensor (Figure 3c). A similar evolution was observed for sensitivity values while different linear dynamic ranges were determined for both biomembranes (Table 2). Moreover, the response times of both biomembrane types are different too. For the ENFET based on Laponite, the response time varies between 1.5 and 4 min depending on the amount of Laponite/urease mixture deposited on the pH-FET surface. In contrast, for the ENFET based on LDH, it is only around 5-10 s whatever configuration is used. (24) Cosnier, S.; Fontecave, M.; Innocent, C.; Nivie`re, V. Electroanalysis 1997, 9, 685-688. (25) Besombes, J.-L.; Cosnier, S.; Labbe, P.. Talanta 1997, 44, 2209-2215.

Analytical Chemistry, Vol. 74, No. 16, August 15, 2002

4039

Table 2. Comparison of ENFETs Based on Laponite and LDH for the Determination of Urea

a

biosensor compositiona

max response (mV)

sensitivity (V/M)

detection limit (µM)

Laponite/urease (1:1) LDH/urease(1:1) LDH/urease (1:2)

93 47 72

110.6 18.2 37.26

2 5 3.5

linear range (mM) 0.005-0.5 0.06-1.18 0.04-1.37

R 0.9996 0.9989 0.9993

Coating elaborated from 2 µL of aqueous clay solution (2.5 g/L) containing urease; exposure time to glutaraldehyde vapor, 8 min.

It is possible that the porosity of the membrane, combined with the characteristics of each clay, is the reason for such differences. The particle size of the LDH is larger than that of Laponite, namely, 200 × 200 × 20 and 40 × 10 × 1 nm, respectively.18,23 Therefore, the LDH/urease biomembrane may be more porous than the Laponite/urease. Thus, the substrate as well as the products of the enzymatic reaction diffuses faster through the LDH/urease membrane, resulting in a decrease in the response time. Another large difference between biosensor responses is the signal/noise ratio. It appears that the biosensor response for ENFET based on Laponite is quite stable. In contrast, the noise grows as the substrate concentration increases for the ENFET based on LDH. Due to its high charge density, the LDH colloidal suspension is less stable and a slow precipitation can occur during the drying process resulting in a less homogeneous coating than with the well-delaminated Laponite gel. Porosity Study of the Biomembranes. The mass-transfer process through the clay/enzyme biomembranes onto the electrode surfaces can be investigated by using rotating disk electrode (RDE) voltammetry.26-30 To avoid any interference with the charges of the clay sheets, the rotating disk voltammograms were recorded from Laponite/urease- and LDH/urease-modified electrodes in an aqueous solution of a neutral molecule, hydroquinone, and compared to that obtained from a bare glassy carbon electrode (Figure 4A). By comparing the ilim, corresponding to hydroquinone oxidation, it can be seen that the Laponite/urease biomembrane presents a larger diffusional resistance to the substrate than the LDH/urease biomembrane. The permeability Pm of the biomembranes was estimated using RDE experiments at different rotation rates. The results were treated through eqs 2-4, introduced by Gough and Leypoldt,26-28

1/ilim ) 1/0.62nFADs2/3cov1/6ω1/2 + δ/nFACoKDm

(2)

Pm ) KDm/δ

(3)

1/ilim ) 1/0.62nFADs2/3cov1/6ω1/2 + 1/nFPmAco

(4)

that describe the variation of steady-state limiting current ilim with the mass transport for a rotating disk electrode coated with an electro-inactive membrane, where terms Ds and Dm are the diffusion coefficients for the substrate in the bulk solution and in the membrane, respectively, ν is the kinematic viscosity of the (26) Gough, D. A.; Leypoldt, J. K. Anal. Chem. 1979, 51, 439-444. (27) Gough, D. A.; Leypoldt, J. K. Anal. Chem. 1980, 52, 1126-1130. (28) Gough, D. A.; Leypoldt, J. K. J. Electrochem. Soc. 1980, 127, 1278-1286. (29) Oyama, N. I.; Anson, F. E. Anal. Chem. 1980, 52, 1192-1198. (30) Save´ant, J. M. J. Electroanal. Chem. 1981, 302, 91-101.

4040

Analytical Chemistry, Vol. 74, No. 16, August 15, 2002

Figure 4. (A) Linear sweep voltammograms of 2.0 mM hydroquinone in 0.1 M LiClO4 at (a) a bare glassy carbon electrode (3-mm diameter), (b) GCE coated with LDH/urease, and (c) GCE coated with Laponite/urease biomembranes. (v ) 20 mV s-1, ω ) 500 rpm). (B) Koutecky-Levich plot for 2 mM hydroquinone in 0.1 M LiClO4 at (a) a bare glassy carbon electrode, (b) GCE coated with a LDH/ urease, and (c) GCE coated with Laponite/urease biomembranes.

solution, ω the rotation rate of the RDE, δ the thickness of the polymer film, and K the partition equilibrium constant of the substrate between solution and film. Equation 2 is composed of two terms, where the first represents the current flow under the same conditions, in the absence of a membrane, and is therefore characteristic of the diffusion of the substrate in the bulk solution. The second term accounts for the diffusion of the substrate in the membrane and depends on the product of the partition equilibrium constant K and diffusion constant of the substrate in the membrane. Only the first term of the equation is dependent upon the rotation rate of the RDE. Therefore, a plot of 1/ilim versus 1/ω1/2 presents a linear behavior with the same slope as for a bare electrode with a positive intercept, whose value depends on

the permeability Pm of the biomembrane, (eq 4). Figure 4B shows the results obtained for a bare electrode and for an electrode coated with LDH/urease and Laponite/urease biomembranes. The estimated permeability of the LDH/urease biomembrane was 1.4 × 10-2 cm/s, therefore greater than that of the Laponite/ urease biomembrane (3.1 × 10-3 cm/s). This result is in agreement with the observed response time of the two biomembranes. Nevertheless, it is quite surprising that the less permeable membrane (Laponite) leads to the most sensitive biosensor. This may be attributed to the cationic-exchange properties of Laponite. Effectively, it has been previously reported that a negatively charged polymeric membrane (Nafion), which has similar properties, provides very sensitive urease ENFET due to a local pH effect.31 As this phenomenon was not observed when conductometric detection was used,32 it has been interpreted in terms of protecting the enzymatic reaction from the effect of the buffer capacity of the outer medium. Influence of pH on Biosensor Response. Because enzyme activity is known to be dependent upon pH,3,33 the pH effect on biosensor response was examined in a buffer solution for both clay/urease biosensors at a fixed urea concentration. A maximum activity for both biomembranes (Laponite/urease and LDH/ urease) was found at a pH of ∼7.5. These results are in agreement with previous ones reported for free enzymes16 or entrapped enzymes in different materials.3,5 Thermal Stability of the Biosensors. The effect of temperature on the biosensor response was also studied. The measurements for both biosensors were carried out between 13 and 52 °C (Figure 5). The ENFET based on Laponite presents a linear increase in sensitivity up to 35 °C and then begins to lose its activity irreversibly for higher temperatures. This result is in agreement with that obtained for urease in solution16 or entrapped in another matrix.33 The ENFET based on LDH presents a wider stability range, with maximum activity at ∼45 °C. The activation energy was estimated for both biomembranes, through the Arrhenius equation. Plotting ln k versus 1/T produces a straight line whose slope is -Ea/R, inside the temperature range where enzyme activity increases. The activation energy values found for the Laponite/urease and LDH/urease biomembranes were very similar, 24.9 and 24.6 kJ/mol, respectively. These results differ significantly from those reported for the free urease, 12 kJ/ mol, but are in agreement with those reported for this enzyme entrapped in different matrixes, ∼20 kJ/mol.34 Operational and Storage Stability of the Biosensors. To estimate the operational stability of the biosensors, a continuous measurement of voltage was carried out for 90 min in a 0.2 mM urea solution for the Laponite/urease biosensor and 0.6 mM urea for the LDH/urease biosensor. For both biosensors, a small increase of the biosensor response is observed with time. This may be related to the variation of the local pH with time due to OH- accumulation within the clay matrix. At the end of the experiment, the biosensors were washed with deionized water and (31) Gorchkov, D. V.; Soldatkin, A. P.; Poyard, S.; Jaffrezic-Renault, N.; Martelet, C. Mater. Sci. Eng. 1997, C5, 23-28. (32) Jdanova, A. S.; Poyard, S.; Soldatkin, A. P.; Jaffrezic-Renault, N.; Martelet, C. Anal. Chim. Acta 1996, 321, 35-40. (33) Liu, D.; Ge, K.; Chen, K.; Nie, L.; Yao S. Anal. Chim. Acta 1995, 307, 6169. (34) Eser, E. A.; Murat, E. Y. Artif. Cells, Blood Substitutes, Immobilization Biotechnol. 2000, 28, 95-111.

Figure 5. (A) Relationship between temperature and the biosensor response in 5.0 mM PB solution (pH 7.4): (a) Laponite/urease biomembrane (1:1, 2 µL) in the presence of 0.2 mM urea and (b) LDH/urease biomembrane (1:2, 2 µL) in the presence of 0.6 mM urea. (B) Arrhenius plot for (a) Laponite/urease biomembrane and (b) LDH/ urease biomembrane.

Figure 6. Storage stability test of the biosensors in 5.0 mM PB solution (pH 7.4): (a) LDH/urease (1:2, 2 µL) and (b) Laponite/urease (1:1, 2 µL) biosensors.

their response was again measured. The two biosensors presented the same response found during the first minutes of the previous experiments. These results show that both biosensors have good operational stability, indicating that they can be used in measurements, lasting a long time. The storage stability of the biosensors was monitored for several weeks. The measurements were carried out under the same experimental conditions described for the operational stability study. The relative biosensor responses as a function of storage time are shown in Figure 6. Each point of both curves corresponds to an arithmetic average of about 10 measurements, and they are plotted as a value relative to the initial response corresponding to the first day. After each measurement, the ENFETs were stored at 4 °C in a 1 mM EDTA phosphate buffer solution 5 mM. For the ENFET based on LDH, an increase of ∼30% in its response was recorded in the first days followed by a stabilization of the response for at least 40 days (Figure 6, curve a). Then, a Analytical Chemistry, Vol. 74, No. 16, August 15, 2002

4041

continuous decrease of the analytical signal was observed for the next 30 days. After 70 days, the biosensor presents only ∼54% of the initial biosensor response. By contrast, the ENFET based on Laponite shows a loss of ∼40% of its sensitivity in the first 15 days. After this period, the signal decreases slowly with time so that after 70 days, the biosensor response represents 43% of the initial signal (Figure 6, curve b). As urease is very sensitive to the presence of heavy metals, a chelating agent, 1 mM EDTA, was added to the storage solution.35,36 The weak initial storage stability of ENFET based on Laponite may therefore be due to the more efficient action of EDTA in LDH than in Laponite, which is less porous. In addition, the anion-exchange properties of LDH should facilitate the percolation of the anionic EDTA unlike Laponite, which presents cation-exchange properties. Effect of Sodium Tetraborate on the Kinetic Parameters of the Biosensors. One limitation of the application of urea ENFETs is their linear domain, generally limited to low concentrations. In human blood, urea concentration varies between 6 and 8 mM for a healthy person. For a person with kidney dysfunction, urea concentration is higher, up to 40 mM.5 To widen the linear range of the ENFET sensors, we have investigated the use of sodium tetraborate as an inhibitor of urease activity. For this purpose, we studied the effect of sodium tetraborate on the analytical characteristics of ENFETs based on urease immobilized in both oppositely charged clays. Figure 7A shows calibration curves for urea obtained with an ENFET based on Laponite in a phosphate buffer in the presence and absence of sodium tetraborate. The sensitivity strongly decreases with the sodium borate concentration from 0.125 to 0.5 mM, while the linear range remains almost constant (Table 3). The inhibition effect of tetraborate was also examined on the ENFET based on LDH. The results, presented in Table 3, show a greater influence of the tetraborate concentration on the linear dynamic range for the LDH/urease biosensor. Indeed, the LDH membrane presents two factors favorable to the increase of inhibitor concentration within the membrane: first its permeability, which is higher than that of the Laponite membrane, second its anion-exchange property, allowing the accumulation of tetraborate. Using the classical kinetic approach based on the MichaelisMenten model, parameters such as the maximum rate, Vmax, and the Michaelis constant, Km, can be evaluated from a LineweaverBurk plot of the calibration curves obtained with the Laponite/ urease and LDH/urease biosensors in the presence and absence of tetraborate. The Kapp m values for urea without any inhibitor determined for the Laponite/urease and LDH/urease biomembranes are quasi-identical (2.78 and 2.36 mM, respectively) (Table 3). These results are in agreement with those (1-5 mM) reported in the literature for free urease.33 In the presence of the inhibitor, the Lineweaver-Burk plots present the same Vmax values, reflecting a competitive inhibition of urease by sodium tetraborate (Figure 7B). However, for the ENFET based on Laponite, Kapp m values do not change noticeably with the inhibitor concentration (Table 3). In contrast, for the ENFET based on LDH, an important increase in Kapp m values in (35) Volotovsky, V.; Kim, N. Electroanalysis 1998, 10, 61-63. (36) Zhylyak, G. A.; Dsyadevich, S. V.; Korpan, Y. I.; Soldatkin, A. P.; El’skaya, A. V. Sens. Actuators 1995, B 24-25, 145-148.

4042 Analytical Chemistry, Vol. 74, No. 16, August 15, 2002

Figure 7. (A) Calibrations curves of urea in 5.0 mM PB solution (pH 7.4) for the Laponite/urease biomembrane without (a) and with 0.125 (b), 0.25 (c), and 0.5 mM sodium tetraborate (d). (B) Linear representation of Michaelis-Menten equation for the Laponite/urease biomembrane in 5.0 mM PB solution (pH 7.4) without (a) and with 0.125 (b), 0.25, and (c) 0.5 mM sodium tetraborate. (C) Plots of the inverse of the response versus tetraborate concentration in the presence of (a) 0.4, (b) 0.8, (c) 1.8, and (d) 3.6 mM urea. Table 3. Role of Tetraborate on the Laponite/Urease and LDH/Urease Biosensors Laponite/urease [tetraborate] (mM) 0.0 0.125 0.25 0.5

linear range (mM) 0.005-0.5 0.025-1.8 0.03-1.8 0.035-1.8

LDH/urease

Kapp m

linear range (mM)

2.8 4.4 6.1 10.0

0.035-1.4 0.045-7.2 0.06-9.5 0.1-12

(mM)

Kapp m (mM) 2.4 10.1 37.2 61.7

the presence of increasing inhibitor concentration from 0.125 to 0.5 mM tetraborate corroborates the competitive nature of this inhibition. An apparent inhibitor binding constant Ki was determined from different urea concentrations for each ENFET sensor (Figure 7C). The results obtained for the Laponite/urease and LDH/urease biomembranes were 0.16 and 0.051 mM, respec-

tively. These results indicate that the enzyme/inhibitor affinity is stronger in the LDH/urease than in the Laponite/urease biomembrane, highlighting the role of the clay matrix.

the LDH/urease biosensor greatly extends the dynamic range, which allows this biosensor to be used in routine analysis for urea assays in urine and blood.

CONCLUSIONS In this work, we have described a comparative study between the properties of two different urea biosensors based on the immobilization of the urease in two oppositely charged inorganic clay matrixes (Laponite and LDH) on the pH-FET surface. The ENFETs based on LDH showed remarkable properties such as thermal, operational, and storage stability and response time while the main advantage of the ENFET based on Laponite was its sensitivity. The permeability of the biomembrane and the role of the electrostatic interactions due to the charges of the clay greatly affected the biosensor performances. In particular, we have demonstrated that tetraborate acted as a competitive inhibitor for the two biosensors, even more strongly for the ENFET based on LDH. The action of the tetraborate on the calibration curve of

ACKNOWLEDGMENT The synthesis and characterization of LHD were performed in the Laboratoire des Mate´riaux Inorganiques (UMR 6002) of Blaise Pascal University of Clermont-Ferrand. C.M. thanks C. Forano in particular for providing the [Zn-Al-Cl] phase. J.V.d.M. thanks FACEPE for his air passage between Brazil and France. This work was supported by the French MENRT (Ministry of Research and National Education) through the postdoctoral stay for J.V.d.M. (DSU4 2).

Received for review March 14, 2002. Revised manuscript received May 31, 2002. Accepted June 18, 2002. AC025627+

Analytical Chemistry, Vol. 74, No. 16, August 15, 2002

4043