Enhanced surface plasmon resonance immunosensing using a streptavidin–biotinylated protein complex

Abstract

In this paper, we present a novel strategy for improving the sensitivity of surface plasmon resonance immunosensing using a streptavidin–biotinylated protein complex. This amplification strategy is based on the construction of a molecular complex between streptavidin and biotin labeled protein. The complex can be formed in a cross-linking network of molecules so that the amplification of the response signal will be realized due to the big molecular size of the complex. The results show that the amplification strategy causes a dramatic improvement of the detection sensitivity. hIgG protein could be detected in the range of 0.005–10 µg ml⁻¹.

---

Introduction

The immunosensor based on surface plasmon resonance (SPR) has been receiving increasing attention in recent years, due to its potential as a label-free, real time, rapid, high selectivity immunoassay technique.^1,2 However, its major disadvantage for bioanalytical applications is that it is difficult to use for detection of low concentration or low molecular mass analytes. The detection limit is ca. 1–10 nM for a 20-kDa molecule and is even higher for smaller molecules.³ Many researchers used commercial sensor chips with an extended coupling dextran matrix to increase reasonably the surface loading of biomolecules, compared to a monolayer of proteins immobilized directly on a gold surface.^4,5 However, despite the signal enhancement afforded by the dextran matrix, the sensitivity of SPR was still finite and low molecular mass analytes could not be detected even using this kind of sensor chips. Other strategies have been proposed to enhance the response signal by using a latex particle,⁶ colloidal Au⁷ and liposome.⁸ Here, we present a novel strategy for improving the sensitivity of SPR immunosensing using a streptavidin–biotinylated protein complex.

Experimental

The BIAcore 1000 system, sensor chip C1 (a flat CO₂H layer on the surface), and HBS buffer (pH 7.4, consisting of 10 mM 4-[2-hydroxyethyl]piperazine-1-ethanesulfonic acid, 150 mM sodium chloride, 3.4 mM EDTA, 0.005% (v/v) surfactant P-20), were obtained from Pharmacia Biosensor AB (Uppsala, Sweden). Human immunoglobulin G (hIgG) was obtained from the Institute of Microbiology, Chinese Academy of Military Medicine Science (Beijing, China). Purified goat anti-hIgG antibody was purchased from Sino-Am Biotech. Co. (Beijing, China). Bovine serum albumin (BSA) was purchased from Sigma (St. Louis, USA). The biotinylated goat anti-hIgG antibody was prepared according to the procedure reported.⁹ All other reagents and solvents were analytical reagent grade and MilliQ-grade water was used. All solutions prepared were filtered (0.22 µm) and thoroughly degassed prior to use.

A constant flow rate of 5 µl min⁻¹ and 25 °C temperature were chosen for the immobilization procedure and other operations subsequently. Antibody immobilization was performed according to the general procedure described elsewhere.¹⁰ The immobilized antibody surface was used to determine the hIgG concentration. The sample (hIgG) was diluted with HBS and flowed over the immobilized surface by injecting a 40 µl volume. 40 µl of 0.1 mg ml⁻¹ biotinylated goat anti-hIgG antibody (bio-Ab) was subsequently injected, followed by 40 µl of streptavidin–biotinylated antibody complex (SABC) to enhance the response signal.

Results and discussion

Commercial sensor chip pioneer C1 (a flat COOH layer on the surface) was used for antibody immobilization by the amine coupling method in this study. Amine coupling introduces N-hydroxysuccinimide esters into the surface by modifying carboxymethyl groups with a mixture of NHS and EDC. These esters then react spontaneously with amines of antibody protein concentrated by electrostatic adsorption. The antibody was immobilized from a 0.1 mg ml⁻¹ solution in a 10 mM acetate buffer at pH 5.4. The immobilized amount of anti-hIgG antibody is about 1800 RU after a complete washing. It is said that the surface concentration of the immobilized antibody on the sensor surface is 0.68 × 10¹⁰ molecules mm² (the transformation 1 kRU = 1 ng mm² is applied according to the work on the correlation of the SPR response with the surface protein concentration by Stenberg et al.¹¹). The theoretical coupling amount of IgG on a flat surface to form a dense stack monolayer is about 0.71 × 10¹⁰ molecules mm⁻² (assuming a flat lying orientation as reported¹²) according to X-ray crystallography data of IgG (10 × 14 × 5 nm).¹³ In this experiment, the surface coverage of the immobilized antibody is about 95.8%. It is very clear that the antibody molecule can be immobilized on the sensing surface to form a dense stack monolayer by the amine coupling method in the optimal conditions.

The process of enhanced immunoassay of the analyte (antigen) is represented in Scheme 1. A capture antibody is firstly immobilized on the sensor chip surface by the amine couple method. Then, injections of an antigen sample and a biotinylated detecting antibody (bio-Ab) are completed, respectively. The bound amount of the detecting antibody is related to the analyte. Finally, the streptavidin–biotinylated protein complex (SABC) is added to amplify considerably the response signal and improve dramatically the detection sensitivity. This amplification strategy is based on the high-affinity interaction of biotin and streptavidin.^14,15 Each streptavidin has four equivalent sites for biotin. The tetravalency of streptavidin for biotin allows the construction of a molecular complex between streptavidin and biotin labeled proteins when one molecule of this protein is labeled by several molecules of biotin.¹⁶ This complex can be formed in a cross-linking network of molecules so that only a few binding events of the analyte at the sensing surface may lead to a detectable surface mass due to the large molecular size of the complex.

After the immobilization of goat anti-hIgG antibody, hIgG of 5 μg ml⁻¹ flows over the sensor surface (Fig. 1). The immune reaction between the immobilized antibody and the analyte leads to 99 RU response change. Then, 0.1 mg ml⁻¹ biotinylated goat anti-hIgG antibody is injected and leads to 197 RU response change. Finally, after adding the streptavidin–biotinylated protein complex, the response signal is about 1378 RU due to the binding reaction between the biotin molecules of the detecting antibody and the available streptavidin sites of SABC. The total response signal of the enhanced immunoassay based on the specific binding of hIgG is 1674 RU, about 17 times higher than the primary response (99 RU). It should be noted that the enhanced effect is influenced by the mole ratio of streptavidin and biotin labeled protein for the preparation and formation of the complex. We studied the effect of three different mole ratios by preparing the complex SABC1 (composing of 0.1 mg ml⁻¹ streptavidin and 0.02 mg ml⁻¹ bio-Ab), SABC2 (0.1 mg ml⁻¹ streptavidin and 0.08 mg ml⁻¹ bio-Ab) and SABC3 (0.1 mg ml⁻¹ streptavidin and 0.2 mg ml⁻¹ bio-Ab). The results show that the complex SABC2 has a better amplification effect than the complexes SABC1 and SABC3. The complex SABC3 leads to a very small response change (only 154 RU). It can be understood that the SABC could not bind onto the surface of biotinylated detecting antibody when biotin labeled protein occupies four binding sites of streptavidin completely due to the high mole ratio of the biotin labeled protein during preparation of the complex. On the other hand, if the mole ratio of biotin labeled protein is low, it is unfavorable for the formation of a large cross-linking network between biotin conjugate and streptavidin. So selecting the optimal mole ratio of biotin labeled protein to streptavidin for preparation of SABC is very important for a better amplified response. The other factor, the formation time of the complex, also affects the amplified effect.

Fig. 1 Sensorgrams showing the procedure and response of hIgG enhanced assay (BSA as blank). The captions indicate injections: (a) 40 µl of 5 µg ml⁻¹ hIgG, (b) 40 µl of 1 mg ml⁻¹ BSA, (c) 40 µl of 0.1 mg ml⁻¹ biotinylated detecting antibody, (d) 40 µl of streptavidin–biotin labeled antibody complex (SABC, 0.1 mg ml⁻¹ streptavidin, 0.08 mg ml⁻¹ bio-Ab). A constant flow of HBS of 5 µl min⁻¹ was maintained. The sensing surface has been immobilized with a layer of goat anti-hIgG antibody.

Control experiments were performed to reveal high selectivity and specificity of the enhanced immunoassay. Treatment of the antibody immobilized surface with the biotinylated detecting antibody, and then with SABC, leads to a very small change (only 68 RU) in the SPR response. Experimental work also started with an injection of 1 mg ml⁻¹ of bovine serum albumin (BSA) instead of hIgG (shown in Fig.  1). The same enhanced operations were applied to amplify the non-specific binding of BSA. No obvious changes occur in the SPR response of the anti-hIgG antibody immobilized surface with biotinylated detecting antibody or SABC. The SPR response is about 17 RU due to non-specific adsorption of BSA onto the antibody immobilized surface. After the biotinylated detecting antibody and streptavidin–biotinylated protein complex, the response increases by about 108 RU, much smaller than the enhanced response of hIgG. The results indicate that the amplified immunoassay of the specific immobilized surface is not affected by the non-specific adsorption of other proteins.

Biotin/avidin chemistry has been used extensively for protein immobilization and enzyme capture.^14–16 Our experiments started first by avidin-biotinylated protein complex (ABC) to enhance the response signal. However, the results reveal that non-specific adsorption of ABC to the sensing surface is larger than SABC. The carbohydrate moieties in combination with the high isoelectric point of avidin have been reported to result in non-specific binding in certain applications.¹⁴ However, streptavidin lacks the glycoprotein constituent and has a lower isoelectric point than avidin so that it is more preferable for enhanced immunoassay than avidin.

Fig. 2 shows that the SPR response of hIgG upon the enhanced immunoassay at different concentrations of the analyte hIgG. It is evident that as the hIgG concentration is higher, the observed primary and total SPR response increase, respectively. This is consistent with the fact that higher hIgG concentration yields a higher coverage of hIgG, then a higher coverage of biotinylated detecting antibody and a higher coverage of SABC on the sensing surface. As shown in Fig. 2, hIgG protein could be detected in the range of ca. 0.005–10 µg ml⁻¹ when streptavidin–biotin conjugate complex was introduced to amplify the response signal. However, the immobilized antibody surface has a good response to hIgG in the range of 1–10 µg ml⁻¹ without an amplified operation. It is shown that when streptavidin–biotinylated protein complex is used, a substantial improvement of the sensitivity is achieved and the detection limit is improved about 200-fold by this amplification strategy.

Fig. 2 Relative response of human IgG as a function of the different concentrations of hIgG. The plots are attained from the total response of enhanced immunoassay (a) and the primary response (b).

Scheme 1 Schematic diagram of enhanced immunoassay of hIgG.

In conclusion, we have demonstrated a novel enhanced immunoassay for the sensitive and selective detection of hIgG by means of the surface plasmon resonance technique. It is proven that the amplification strategy, using streptavidin–biotinylated protein complex, causes a dramatic improvement of the detection limit. The method can be generally applicable to the enhanced assay of other biomolecules and other transduction means, such as the quartz crystal microbalance or impedance spectroscopy. Further studies are in progress to detect other biomolecules using the enhanced strategy by means of quartz crystal microbalance, surface plasmon resonance and impedance spectroscopy, simultaneously.

Acknowledgements

This research is supported by the National Natural Science Foundation of China.

References

1. F. Deckert and F. Legay, Anal. Biochem., 1999, 274, 81.
2. T. P. Vikinge, A. Askendal, B. Liedberg, T. Lindahl and P. Tengvall, Biosens. Bioelectron., 1998, 13, 1257.
3. R. P. H. Kooyman, H. Kolkman, J. Van Gendt and J. Greve, Anal. Chim. Acta, 1988, 213, 35.
4. W. L. DeLano, M. H. Ultsch, A. M. Vos and J. A. Wells, Science (Washington, D.C.), 2000, 287, 1279.
5. M. A. Cooper, A. Hansson, S. Lofas and D. H. Williams, Anal. Biochem., 2000, 277, 196.
6. A. H. Severs and R. B. M. Schasfoort, Biosens. Bioelectron., 1993, 8, 365.
7. L. A. Lyon, M. D. Musick and M. J. Natan, Anal. Chem., 1998, 70, 5177.
8. T. Wink, S. J. Van Zuilen, A. Bult and W. P. Van Bennekom, Anal. Chem., 1998, 70, 827.
9. Y. Yang and X. Yin, Pract. Immunol., Changchun Press, Changchun, China, 1994, p. 503..
10. S. Lofas, M. Malmqvist, I. Ronnberg, E. Stenberg, B. Liedberg and I. Lundstrom, Sens. Actuators, B, 1991, 5, 79.
11. E. Stenberg, B. Persson, H. Roos and C. Urbaniczky, J. Colloid Interface Sci., 1991, 143, 513.
12. B. Walivaara, P. Warkentin, I. Lundstrom and P. Tengvall, J. Colloid Interface Sci., 1995, 174, 53.
13. A. G. Amit, R. A. Mariuzza, S. E. Phillip and R. J. Polja, Science (Washington, D.C.), 1986, 233, 747.
14. M. Wilchek and E. A. Bayer, Anal. Biochem., 1988, 171, 1.
15. S. Ghafouri and M. Thompson, Langmuir, 1999, 15, 564.
16. A. Erisir and C. Aoki, J. Neurosci. Methods, 1998, 81, 189.

---