Label-free electrochemical impedance biosensor to detect human interleukin-8 in serum with sub-pg/ml sensitivity

Abstract

Biosensors with high sensitivity and short time-to-result that are capable of detecting biomarkers in body fluids such as serum are an important prerequisite for early diagnostics in modern healthcare provision. Here, we report the development of an electrochemical impedance-based sensor for the detection in serum of human interleukin-8 (IL-8), a pro-angiogenic chemokine implicated in a wide range of inflammatory diseases. The sensor employs a small and robust synthetic non-antibody capture protein based on a cystatin scaffold that displays high affinity for human IL-8 with a KD of 35 ± 10 nM and excellent ligand specificity. The change in the phase of the electrochemical impedance from the serum baseline, ∆θ(ƒ), measured at 0.1 Hz, was used as the measure for quantifying IL-8 concentration in the fluid. Optimal sensor signal was observed after 15 min incubation, and the sensor exhibited a linear response versus logarithm of IL-8 concentration from 900 fg/ml to 900 ng/ml. A detection limit of around 90 fg/ml, which is significantly lower than the basal clinical levels of 5-10 pg/ml, was observed. Our results are significant for the development of point-of-care and early diagnostics where high sensitivity and short time-to-results are essential.

Keywords: Antibody mimetic protein; CXCL8; Electrochemical impedance spectroscopy; Interleukin-8; Label-free biosensor; Point-of-care diagnostics.

Fig. 1

(a) Binding response at equilibrium of human IL-8 binding to non-antibody capture molecule immobilised onto the sensor surface detected using bio-layer interferometry. The solid line shows the least-square fit of the Langmuir binding isotherm to the linearised data. Inset: linearised form of binding data where the y-axis R/C corresponds to the sensor response at equilibrium (R) divided by human IL-8 concentration (C), and the x-axis to the sensor response at equilibrium (R). (b) Sensogram showing change in the SPR angle of the sensor functionalised with the binding protein to both human IL-8 and BSA from the 100 mM phosphate buffer pH 7.4 baseline.

Fig. 2

Schematic illustration of the biosensor assembly. A monothiol-alkane-PEG acid SAM is first assembled on a gold electrode. The carboxylic acid groups of the SAM are then activated with EDC/NHS to which the non-antibody capture proteins are covalently attached. Following deactivation of residual activated acid sites with ethanolamine, the sensor surface is challenged with fluids containing human IL-8 protein and the electrochemical impedance of the sensor is monitored.

Fig. 3

EIS Bode plots showing (a) the phase θ(ƒ) after the formation of the monothiol-alkane-PEG acid SAM on the gold electrode at 0 mV and +80 mV dc potential, and (b) the phase θ(ƒ) after immobilisation of the non-antibody capture molecules on the SAM in comparison to the SAM only at a dc offset potential of +80 mV vs Ag/AgCl. The EIS measurements were conducted in 100 mM phosphate buffer at pH 7 and the data shown represent the average θ(ƒ) of five EIS scans. Corresponding Nyquist plots are shown in Fig. S7.

Fig. 4

(a) EIS sensogram showing the change in phase from the baseline at 0.1 Hz, ∆θ(ƒ)_0.1 Hz, of the sensor response vs IL-8 concentration between 9 fg/ml and 900 ng/ml (a representative Nyquist plot for one of the devices is shown in Fig. S8). Inset: effect of IL-8 incubation time on ∆θ(ƒ)_0.1 Hz from the baseline, with measurements taken immediately and after 15 min of incubation. (b) EIS sensogram showing the response of the sensor when exposed to IL-8 free serum and BSA-spiked serum. The inset shows the variance of the sensor signal over time. All EIS scans were performed at a dc offset of +100 mV vs Ag/AgCl.