Enzyme immobilization strategies and electropolymerization conditions to control sensitivity and selectivity parameters of a polymer-enzyme composite glucose biosensor

Abstract

In an ongoing programme to develop characterization strategies relevant to biosensors for in-vivo monitoring, glucose biosensors were fabricated by immobilizing the enzyme glucose oxidase (GOx) on 125 μm diameter Pt cylinder wire electrodes (Pt(C)), using three different methods: before, after or during the amperometric electrosynthesis of poly(ortho-phenylenediamine), PoPD, which also served as a permselective membrane. These electrodes were calibrated with H(2)O(2) (the biosensor enzyme signal molecule), glucose, and the archetypal interference compound ascorbic acid (AA) to determine the relevant polymer permeabilities and the apparent Michaelis-Menten parameters for glucose. A number of selectivity parameters were used to identify the most successful design in terms of the balance between substrate sensitivity and interference blocking. For biosensors electrosynthesized in neutral buffer under the present conditions, entrapment of the GOx within the PoPD layer produced the design (Pt(C)/PoPD-GOx) with the highest linear sensitivity to glucose (5.0 ± 0.4 μA cm(-2) mM(-1)), good linear range (K(M) = 16 ± 2 mM) and response time (< 2 s), and the greatest AA blocking (99.8% for 1 mM AA). Further optimization showed that fabrication of Pt(C)/PoPD-GOx in the absence of added background electrolyte (i.e., electropolymerization in unbuffered enzyme-monomer solution) enhanced glucose selectivity 3-fold for this one-pot fabrication protocol which provided AA-rejection levels at least equal to recent multi-step polymer bilayer biosensor designs. Interestingly, the presence of enzyme protein in the polymer layer had opposite effects on permselectivity for low and high concentrations of AA, emphasizing the value of studying the concentration dependence of interference effects which is rarely reported in the literature.

Keywords: amperometry; ascorbic acid interference; brain monitoring; enzyme-modified electrode; hydrogen peroxide; polyphenylenediamine.

Figure 1.

Sample steady-state calibration data and nonlinear regression analysis for the biosensor design, Pt_C/PoPD-GOx [Equation (4), R² = 0.998, n = 8; left], illustrating the graphical significance of the Michaelis-Menten constants, J_max and K_M. The linear region slope (LRS) was obtained using linear regression up to 10 mM glucose (R² = 0.996, n = 8; left inset), and represents the most suitable measure of analytical sensitivity of each biosensor design to enzyme substrate (see Table 1). Schematic representation of the PEC configuration for the same Pt_C/PoPD-GOx design (right), illustrating trapped GOx (∼8 nm diameter) in the PoPD layer deposited by the precipitation of insoluble chains formed during the electropolymerization of monomer solution containing the enzyme.

Figure 2.

Averaged steady-state AA calibrations for Pt_C/PoPD and Pt_C/PoPD-GOx electrosynthesized in 300 mM oPD solution made with either PBS*, PBS + GOx (1 mg mL⁻¹, n = 4), PBS+GOx (5 mg mL⁻¹, n = 8), distilled water*, or water+GOx (1 mg mL⁻¹, n = 6). The concentration of GOx for the bottom graph was 1 mg mL⁻¹. *The GOx-free data were taken from the literature [38] for comparison.

Figure 3.

Effect of different background electrolytes and GOx concentrations in the monomer solution on the subsequent P(AA)% values determined for Pt_C/PoPD electrodes electrosynthesized from 300 mM oPD. Top: 150 mM of either KCl (n = 12), NaCl (n = 8) or LiCl (n = 4) plotted against the hydrodynamic radius of the cations. Bottom (left to right): no added background electrolyte (i.e., distilled water, n = 28); distilled water containing 1 mg mL⁻¹ GOx (n = 6), 150 mM NaCl (n = 8); phosphate buffered 150 mM NaCl (PBS, n = 19); PBS containing 1 mg mL⁻¹ GOx (n = 4); or PBS containing 5 mg mL⁻¹ GOx (n = 8).