Quantitative Fluorescence Quenching on Antibody-conjugated Graphene Oxide as a Platform for Protein Sensing

Abstract

We created an immunosensing platform for the detection of proteins in a buffer solution. Our sensing platform relies on graphene oxide (GO) nanosheets conjugated with antibodies to provide quantitative binding sites for analyte proteins. When analyte proteins and standard fluorescein-labelled proteins are competing for the binding sites, the assay exhibits quantitative fluorescence quenching by GO for the fluorescein-labelled proteins as determined by the analyte protein concentration. Because of this mechanism, measured fluorescence intensity from unquenched fluorescein-labelled protein was shown to increase with an increasing analyte protein concentration. As an alternative to the conventional enzyme-linked immunosorbent assay (ELISA), our method does not require an enzyme-linked second antibody for protein recognition and the enzyme for optical signal measurement. Thus, it is beneficial with its low cost and fewer systematic errors caused by the series of antigen-antibody recognition steps in ELISA. Immune globulin G (IgG) was introduced as a model protein to test our method and our results showed that the limit of detection for IgG was 4.67 pmol mL^-1 in the buffer solution. This sensing mechanism could be developed into a promising biosensor for the detection of proteins, which would broaden the spectrum of GO applications in both analytical biochemistry and clinical diagnosis.

Figure 1. Graphene oxide-based fluorescence quenching for the detection of IgG proteins.

The more analyte IgGs were adsorbed by the antibody-conjugated graphene oxide, the fewer IgG-FITCs were quenched. The fluorescence signal from free IgG-FITCs in the buffer solution correlates with the analyte IgG concentration.

Figure 2

Surface morphologies of GO by AFM: (a) bare GO (−1 nm), (b) EDC-NHS activated GO (−4 nm), (c) antibody conjugated GO (−10 nm) on freshly cleaved mica, and (d) height profiles of the line scans (arrows) in (a), (b), and (c). Typical surface feature heights are provided in parenthesis. Image sizes: (a) 1 μm × 1 μm; (b) 1.05 μm × 1.05 μm; and (c) 2 μm × 2 μm.

Figure 3

Fluorescence (FL) intensity profile of (a) 1 μg mL⁻¹ of IgG-FITC reacting with an increasing concentration of bare GO; (b) 1 μg mL-1 of IgG -FITC reacting with an increasing concentration of antibody-modified GO; and (c) 1 μg mL⁻¹ of IgG-FITC reacting with an increasing concentration of BSA-blocked GO. All of the GO concentrations varied; 0, 2, 4, 8, 16, 32, 64, 100, 200, and 400 μg mL⁻¹. The relationship of the GO concentration and the FL intensity of (a), (b), and (c) are shown in (d) for comparison.

Figure 4

(a) The average fluorescence (FL) intensity change (ΔI) of 1 μg mL⁻¹ of IgG-FITC as the function of the increase of the sample IgG in the concentrations of 0, 0.1, 0.2, 0.4, 0.8, 1.6, 3.2, 6.4, 10, and 20 μg mL⁻¹ (experiments were performed 4 times). (b) The linear relationship between FL intensity change (ΔI) and low IgG concentrations of 0, 0.4, 0.8, 1.2, 1.6, 2.0, 2.4, 2.8, 3.2, and 3.6 μg mL⁻¹ (experiments were performed 4 times) for the detection of the LOD. The linear fitting uses the following: ΔI = 0.04157c (concentration of IgG), R² is 0.994.

Figure 5. Selectivity.

The concentrations of HSA and BSA were both 10 μg mL⁻¹, the IgG-FITC was 1 μg mL⁻¹, and the antibody-conjugated GO was 100 μg mL⁻¹ for each experiment. The blank sample did not contain analyte IgG or an interfering molecule.