Amperometric Immunosensors Based on Protein A Coupled

The stock solution of protein A with a concentration of 5.0 mg mL-1 was prepared ... Aldrich Nafion 117 [equivalent weight 1100, dissolved in ethanolâ...
3 downloads 0 Views 55KB Size
Anal. Chem. 2000, 72, 2925-2929

Amperometric Immunosensors Based on Protein A Coupled Polyaniline-Perfluorosulfonated Ionomer Composite Electrodes Chih-Hung Liu, Kuo-Tang Liao, and Hsuan-Jung Huang*

Department of Chemistry, National Sun Yat-sen University, Kaohsiung, 80424 Taiwan

A very sensitive immunosensor based on polyaniline/ Nafion/protein A (PA/NF/PrA) composite electrodes has been developed for the amperometric immunoanalysis with urease-labeled immunoreagents. The use of urease conjugated goat anti-RIgG (GaRIgG-Ur) as the labeled antibody and urea as the substrate with an amperometric detection at -200 mV (vs Ag/AgCl) resulted in a dynamic range of 50-2000 ng mL-1 and a low detection limit of 10 ng/mL (64 pM) for the immunoanalysis of rabbit immunoglobulin G (RIgG). Because of the special affinity between protein A and RIgG, the PA/NF/PrA electrode can be regenerated repetitively by changing the pH of the buffer solutions. Characteristics of the PA/NF/PrA/RIgG immunosensor and optimal conditions for the competitive immunoanalysis of RIgG with FIA were studied. There has been great interest in the development of new, simple, sensitive, and specific immunoassays for the quantitative determination of analytes of clinical or biological importance recently. Besides the use of radioactive labels,1 nonradioactive labels, such as enzymes,2-9 fluorescent molecules,10-13 bio- and chem-luminogenic reagents,14-16 for immunoassay have been developed. Because of its good sensitivity, selectivity, and ease * Corresponding author. E-mail: [email protected]. Fax: 886-75253919. (1) Edwards, R. In Principles and Practice of Immunoassay, 2nd ed.; Price, C. P., Newman, D. J., Eds.; Stockton: New York, NY, 1997; pp 325-48. (2) Lasalle, A. L.; Limoges, B.; Degrand, C.; Brossier, P. Anal. Chem. 1995, 67, 1245. (3) Santandreu, M.; Cespedes, F.; Alegret, S.; Martinez-Fabregas, E. Anal. Chem. 1997, 69, 2080. (4) Wang, J.; Pamidi, P. V. A.; Rogers, K. R. Anal. Chem. 1998, 70, 1171. (5) Sole, S.; Alegret, S.; Cespedes, F.; Martinez-Fabregas, E.; Diez-Caballero, T. Anal. Chem. 1998, 70, 1462. (6) Lu, B.; Iwuoha, E. I.; Smyth, M. R.; O’ Kennedy, R. Anal. Chim. Acta 1997, 345, 59. (7) Dou, X.; Takama, T.; Yamaguchi, Y.; Yamamoto, H.; Ozaki, Y. Anal. Chem. 1997, 69, 1492. (8) Okamoto, Y.; Murakami, H.; Nishida, M. Endocr. J. (Tokyo) 1997, 4, 22. (9) Kokado, A.; Tsuji, A.; Maeda, M. Anal. Chim. Acta 1997, 337, 335. (10) Sauer, M.; Zander, C.; Muller, R.; Ullrich, B.; Drexhage, K. H.; Kaul, S.; Wolfrum, J. Appl. Phys. B: Lasers Opt. 1997, B65, 427. (11) Loescher, F.; Boehme, S.; Martin, J.; Seeger, S. Anal. Chem. 1998, 70, 3202. (12) Matveeva, E. G.; Savitski, A. P.; Gomez-Hens, A. Anal. Chim. Acta 1998, 361, 27. (13) Yuan, J.; Wang, G.; Kimura, H.; Matsmoto, K. Anal. Biochem. 1997, 254, 283. (14) Murakami, S.; Ito, K.; Goto, T.; Kamadi, S.; Maeda, M. Anal. Chim. Acta 1998, 361, 19. 10.1021/ac9914317 CCC: $19.00 Published on Web 05/06/2000

© 2000 American Chemical Society

in use, the enzyme-linked immunosorbent assay (ELISA) becomes one of the most frequently used methods for immunoassay. Although spectrophotometric methods are widely used for the detection of enzymatic products resulting from the antigenantibody reactions in ELISA, the electrochemical methods can provide capabilities of in vivo monitoring, free from color and turbid interferences and which are relatively inexpensive, that the spectrophotometric methods cannot compete with.17,18 Species such as nitrophenol, H2O2, and NH3 that can be determined electrochemically are the substrates or enzymatic products of alkaline phosphatase, horseradish peroxidase, and urease, generally labeled on immunoreagents.2-9 Among these, because ammonia is electroinactive and can only be detected by an ammonia gas-sensing electrode, potentiometry is the only choice for the urease-labeled immunoassay. As the sensitivity and detection limit for potentiometry are inherently worse than that for amperometry, applications of the urease-labeled immunoassay are less reported despite the fact that urease offers several advantages over other commonly used enzymes (e.g., peroxidase, alkaline phosphatase, β-galactosidase) and has been widely used as a label in other biosensing systems.19-22 Urease was reported to be characterized with (1) a higher turnover rate, (2) a nontoxic and stable substrate, (3) a linear reaction time course, (4) a stable ligand-enzyme conjugate, (5) an absence of it in the mammalian system, and (6) being easy to work with.23,24 To demonstrate the applicability of an amperometric immunoassay with a urease-labeled immunoreagent, a very sensitive immunosensor based on PA/NF/PrA composite electrodes was developed for the analysis of rabbit immunoglobulin G. By taking advantage of the special affinity between protein A and RIgG, the PA/NF/PrA electrode can be regenerated repetitively. Characteristics of the PA/NF/PrA/RIgG (15) Brown, R. C.; Weeks, I.; Fisher, M.; Harbron, S.; Taylorson, C. J.; Woodhead, J. S. Anal. Biochem. 1998, 259, 142. (16) Rollag, J. G.; Liu, T.; Hage, D. S. J. Chromatogr., A 1997, 765, 145. (17) Skladal, P. Electroanalysis 1997, 9, 737. (18) Ngo, T. T., Ed. Electrochemical Sensors in Immunological Analysis, Plenum Press: New York, NY, 1987. (19) McNeil, C. J.; Athey, D.; Ball, M.; Ho, W.; Krause, S.; Amstrong, R. D.; Wright, J. D.; Rawson, K. Anal. Chem. 1995, 67, 3928. (20) Walcerz, I.; Glab, S.; Koncki, R. Anal. Chim. Acta 1998, 369, 129. (21) Starodub, N. F.; Kanjuk, N. I.; Kukla, A. L.; Shirshov, Y. M. Anal. Chim. Acta 1999, 385, 461. (22) Su, X. L.; Tan, H. W., Bao, L. L.; Wei, W. Z.; Yao, S. Z. Instrum. Sci., Technol. 1999, 127, 31. (23) Meyerhoff, M. E.; Rechnitz, G. A. Anal. Biochem. 1979, 95, 483. (24) Chandler, H. M.; Cox, J. C.; Healy, K.; Macgregor, A.; Premier, P. R.; Hurell, J. G. R. J. Immunol. Methods 1982, 53, 187.

Analytical Chemistry, Vol. 72, No. 13, July 1, 2000 2925

immunosensor and optimal conditions for the competitive immunoanalysis of RIgG with FIA were studied. EXPERIMENTAL SECTION Reagents and Apparatus. Rabbit immunoglobulin G (RIgG, I-5006), urease-conjugated goat anti-rabbit immunoglobulin G (GaRIgG-Ur, U-1397), protein-A (PrA, from Staphylococcus aureus), and bovine serum albumin (BSA, 96-99%) were obtained from Sigma Chemical Co. (St. Louis, MO). Urea, glutaraldehyde (GA, 25% solution), and Nafion (5 wt %) were obtained from Aldrich (Milwaukee, WI). All chemicals and solvents used were of analytical grade and were used as received. The pH 7.5 phosphate buffer (PB) solutions were prepared by the addition of an appropriate amount of 5 M H3PO4 solution to the 0.10 M disodium hydrogen phosphate solution. Standard RIgG solutions were prepared by dissolving 10 mg of RIgG in 1.0 mL of 0.6 w/v% BSA, pH 7.5 PB solution. Solutions prepared by dissolving 50 µL of GaRIgG-Ur (0.73 mg mL-1) in 5.0 mL of 0.6 w/v% BSA solution were used as the GaRIgG-Ur stock solutions. The stock solution of protein A with a concentration of 5.0 mg mL-1 was prepared by dissolving 5.0 mg of PrA in 1.0 mL of pure water. Solutions of RIgG, GaRIgG-Ur, and PrA were stored in an environment at 4 °C when not in use. The deionized-RO water prepared from a Milli-Q system (Millipore) with a resistance of 18 MΩ cm was used for solution preparation. For electrodeposition and electrochemical studies of PA, a three-electrode system was used. A piece of Pt wire and an Ag/ AgCl (with 3 M KCl solution) electrode were used, respectively, as the counter and reference electrodes. The electrodeposition was performed with a PAR 175 universal programmer coupled with a PAR 179 coulometer. For cyclic voltammetric and amperometric measurements, the system included a flow-through thinlayer electrochemical cell (BAS LC-17A) connected with a BAS 100B electrochemical analyzer. The FIA measurements were conducted with a Gilson Minipuls 3 peristaltic pump, a Rheodyne 7125 injector with a 20-µL sample loop, and the immunoglobulin immobilized electrode. Preparation of the PA/NF/PrA/RIgG Electrodes. Electrochemical preparation of PA/Nafion film was carried out according to the previous report.25 A 4.0-µL 2.0 wt. % Aldrich Nafion 117 [equivalent weight 1100, dissolved in ethanol-water (9:1)] solution was applied onto a glassy carbon disk electrode (3-mm dia.) and air-dried to allow the solvent to evaporate. The polyaniline film was deposited by immersing the electrode in a solution containing 0.1 M aniline and 1.0 M sulfuric acid and sweeping the potential between -200 and 800 mV (vs Ag/AgCl reference electrode) with a sweeping rate of 20 mV s-1 for about 40 min (24 cycles). The amount of charges for the polyaniline deposited was controlled at about 2.4 × 10-3 C. The weights of Nafion and polyaniline deposited on the glassy carbon electrode were estimated to be 7.12 × 10-5 and 1.12 × 10-6 g, respectively. The finished electrode shows a very smooth surface and shines with a light green color. The PA/Nafion coated electrode was conditioned in a pH 7.5 phosphate buffer (PB) solution by applying a potential of -200 mV for 30 min in an FIA system before further processing. A 4.0-µL solution containing protein A (5.0 mg mL-1), Glutaraldehyde (1.0%), and H2O in a volume ratio of 2.5:0.5:1.0 was (25) Sung, J. Y.; Huang, H. J. Anal. Chim. Acta 1991, 246, 275.

2926

Analytical Chemistry, Vol. 72, No. 13, July 1, 2000

applied on the PA/NF electrode for cross-linking protein A on the electrode. An appropriate amount (unless otherwise specified, 20 µL was used) of the standard RIgG solution was applied on the PA/NF/PrA electrode and there was a 30-min wait for the RIgG to associate with protein A on the electrode surface. After washing off the unbound RIgG with PB solution, a RIgG immunoelectrode (PA/NF/PrA/RIgG) was formed. The finished PA/ NF/PrA/RIgG electrode was stored at 4 °C in a pH 7.5 phosphate buffer solution when not in use. Procedures for RIgG Analysis. As the activity of urease conjugated on the GaRIgG is not affected by the association of GaRIgG-Ur to RIgG, procedures for heterogeneous competitive immunoassay were adopted in this experiment. An appropriate volume (e.g., 10 µL) of RIgG standard or sample solution was added to 3.0 mL of GaRIgG-Ur (7.3 µg mL-1) solution for preincubation for 30 min. The PA/NF/PrA/RIgG immunoelectrode was then immersed in the preincubated solution for 30 min. The incubation allowed the RIgG immobilized on the electrode surface to compete with the RIgG in solution for GaRIgG-Ur. After incubation, the amount of GaRIgG-Ur bound to RIgG on the PA/ NF/PrA/RIgG electrode was analyzed by passing 10 mM urea solutions over the electrode with an FIA system. With an applied potential at -200 mV, the current resulting from the plug of NH4+ decomposed from urea was measured. The magnitude of currents obtained depends on the concentration of RIgG in the standard or sample solutions. The current responses obtained from different immunoelectrodes were normalized by comparing their responses with that obtained by the injection of 0.1 M NH4+ solution. Regeneration of the PA/NF/PrA Electrode. To regenerate the PA/NF/Pr-A electrode, the analyzed immunoelectrode was immersed in a stirred pH 2.1 PB solution for 90 s to remove the RIgG and RIgG-GaRIgG-Ur complex from the bound protein A. The regenerated PA/NF/PrA electrode was rinsed and conditioned in a pH 7.5 PB solution before the next application. RESULTS AND DISCUSSION Characterization of the PA/NF/PrA/RIgG Immunoelectrode. The characteristics and functions of the PA/NF composite electrode have been reported in the literature.25-28 Detection of the amount of GaRIgG-Ur bound to RIgG on the electrode requires the urea molecules to diffuse to the electrode surface and undergo the enzymatic reaction to yield NH4+ and HCO3- ions. The NH4+ diffuses further to the PA/Nafion film and triggers the reduction of polyaniline on the electrode. As ammonium and bicarbonate ions are the products of the reaction of urease in a phosphate media,29,30 the enzymatic reaction which proceeded at the PA/ NF/PrA/RIgG/GaRIgG-Ur electrode can be expressed as GaRIgG-Ur

urea 9 8 NH4+ + HCO3PB solution

(1)

NH4+ + PA+.RSO3- + e- ) PA.NH4+.RSO3-

(2)

where PA+ and PA represent the oxidized and reduced forms of (26) Fan, F.-R. F.; Bard, A. J. Electrochem. Soc. 1986, 133, 301. (27) Chou, W. J.; Huang, H. J. Anal. Chem. 1998, 70, 3946. (28) Jespersen, N. D. J. Am. Chem. Soc. 1975, 97, 1662. (29) Blanchard, G. C.; Taylor, C. G.; Busey, B. R.; Williamson, M. L. J. Immunol. Methods 1990, 130, 263. (30) Jespersen, N. D. J. Am. Chem. Soc. 1975, 97, 1662.

Figure 2. Dependence of current responses on the concentration of RIgG immobilized on the PA/NF/PrA electrode. The prepared PA/ NF/PrA/RIgG electrodes were incubated in a GaRIgG-Ur (7.3 µg mL-1) solution for 30 min before running the FIA with 10 mM urea solutions. Other experimental conditions are the same as those specified in Figure 1.

Figure 1. Current responses of the PA/NF/PrA/RIgG electrode before (a) and after (b) incubation with GaRIgG-Ur (7.3 µg mL-1) solutions. Solutions of 10 mM urea and 0.1 mM NH4+, respectively, were injected with a flow rate of 0.4 mL min-1. Potential applied at the immunoelectrode was -200 mV (vs Ag/AgCl). A solution of 0.1 M PB with pH 7.50 was used as the carrier in the FIA system.

polyaniline, respectively, and RSO3- represents the skeleton of Nafion with the immobilized sulfonate groups. From eq 2, whenever NH4+ ions were produced and diffused into the PA/ Nafion film, flow of reduction current occurred that depended on the amount of GaRIgG-Ur bound on the electrode. The responses of the PA/NF/PrA/RIgG electrode to the injection of 0.1 mM NH4+ and 10 mM urea solutions before and after its incubation with the GaRIgG-Ur solution were shown in Figure 1. From Figure 1a, besides the responses to NH4+, small currents resulting from the hydrolysis of urea to urenium ions were also found. The much larger responses to the injection of urea shown in Figure 1b came from the decomposition of urea to NH4+ that proceeded at the incubated PA/NF/PrA/RIgG electrode. The rather pronounced responses found confirmed the feasibility of the immunoelectrode to RIgG analysis. Compared

with Figure 1a, the current response to the injections of NH4+ were found to be smaller in Figure 1b. It should be correlated to the bulkier RIgG-GaRIgG-Ur complex formed on the electrode surface that hindered the diffusion of NH4+ to the PA/NF film. The interference due to the presence of urenium ions in the urea solutions was corrected by subtracting the background current from the current responses obtained in the followed immunoanalysis. Optimization for the Immunoanalysis of RIgG. By varying the concentration of RIgG (from 2.5 to 10.0 mg mL-1) in the immobilization solution, the change in sensitivity of the PA/NF/ PrA/RIgG immunoelectrode was studied. The prepared immunoelectrodes were allowed to incubate with the GaRIgG-Ur solution (without addition of RIgG), and their current responses to the injection of urea solution were measured. Figure 2 shows the results. The current responses increased with the increment of RIgG concentration on the electrode and started to level off when the concentration of RIgG became larger than 7.5 mg mL-1. To ensure a large enough RIgG concentration for its immunoreaction with GaRIgG-Ur, solutions of 10 mg mL-1 RIgG were adopted for electrode immobilization. The effect of flow rate on the immunoanalysis was studied by varying the flow rate in a range of 0.20-1.20 mL min-1. The current responses to the injection of urea solutions decreased gradually as the flow rate increased. By considering the extreme condition at which the amount of GaRIgG-Ur attached to the PA/ NF/PrA/RIgG electrode was very small, a value of 0.40 mL min-1 was selected optimally for the flow injection analysis. The optimal concentration of GaRIgG-Ur used for incubation was also studied. By incubating the RIgG electrode in solutions of various GaRIgG-Ur concentrations (from 0 to 10.2 µg mL-1), it was found that the current responses increased with the increase of GaRIgG-Ur concentration in incubation solutions but leveled off as the concentration of GaRIgG-Ur in solution was larger than Analytical Chemistry, Vol. 72, No. 13, July 1, 2000

2927

Figure 3. Dependence of current responses of the incubated PA/ NF/PrA/RIgG electrodes on the incubation time. Solutions of 10 mM urea were injected with a FIA system.

Figure 4. Calibration graph for the PA/NF/PrA/RIgG immunoelectrode with RIgG concentration expressed in logarithmic scale. Concentrations of RIgG studied are in a range of 0-20 µg mL-1. Procedures of incubation and experimental conditions are the same as those shown in Figure 1.

5.8 µg mL-1. Solutions containing 7.3 µg mL-1 GaRIgG-Ur were adopted for the competitive immunoanalysis in this experiment. The relationship between the current responses and incubation time was studied and shown in Figure 3. The current responses increased almost linearly with the increment of incubation time up to 30 min and started to level off for longer incubation times. To shorten the analysis time, an incubation time of 30 min was adopted for this experiment. Sensitivity and Detection Limit. Figure 4 shows the standard calibration graph for RIgG analysis. An inverse sigmoid relationship between the current responses and the logarithm of RIgG concentrations was obtained. A dynamic range of 50-2000 ng mL-1 was found. The detection limit for RIgG analysis with the 2928 Analytical Chemistry, Vol. 72, No. 13, July 1, 2000

developed immunosensor was estimated to be 10 ng mL-1 or 64 pM. The detection limit and dynamic range obtained are comparable with 5 ng mL-1 and 50 ng mL-1-5 µg mL-1 obtained with sol-gel thick film immunosensors4 and 1.6 ng mL-1 obtained with a flow-through high surface area immunosensor40 and are lower and wider than 5 and 5-25 µg mL-1 obtained with graphite-epoxyimmunoreagents3. Compared with values of about 1.2 and 1.2320 µg mL-1 obtained using a magnetoimmunosensor system5 whereby a urease conjugated anti-RIgG was used and the enzymatic product detected by a pH sensor (ISFET), the performance of our developed sensor is far superior and applaudable. Regeneration and Reproducibility of the PA/NF/PrA Electrode. Regeneration of immunosensors is of interest to the immunoanalysts. Although the antibody-antigen linkage can be broken under drastic conditions (e.g., in alkalinic or acidic soltuions or with chaotropic agents), the immobilized immunoreagents could also suffer from the functional damage or even be released from the immunosorbents.17,31-34 In this experiment, instead of trying to regenerate a PA/NF/RIgG immunoelectrode, a procedure for the regeneration of the PA/NF electrode was developed. From the literature, many immunoreagents were found to bind reversibly to protein A. Because of the specific affinity of protein A to IgG, they are bound in such a manner that the recognition sites of Fab on IgG face the solution. The IgG molecules bound at protein A are believed to fully retain their binding activities toward anti-IgG,35-38 and their binding activities are found to be higher than that bound directly on the electrode surface or on the BSA-modified surface.39 A thin layer of protein A was thus coated on top of the PA/NF electrode to facilitate the immobilization and removal of RIgG for immunoanalysis. After regeneration, the removal of RIgG-GaRIgG complex from the electrode surface was confirmed by injection of urea solutions. The reproducibility of the regenerated PA/NF/PrA electrode for RIgG analysis was determined by incubation with the GaRIgGUr (7.3 µg mL-1) solutions and analysis with urea solutions. A relative standard deviation of 2.11% obtained from 13 successive assays (three analyses for each assay) confirms the renewability of the PA/NF/PrA electrode for RIgG analysis. CONCLUSIONS The performance of the developed PA/NF/PrA/RIgG electrode is as good as or superior to that of the similar amperometric immunosensors reported in the literature. A higher sensitivity and (31) Buhl, S. N.; Jackson, K. Y.; Lubinski, R.; Vanderlinde, R. E. Clin. Chem. (Washington, D.C.) 1976, 22, 1872. (32) Boitieux, J. L.; Groshemy, R.; Thomas, D. Anal. Chim. Acta 1987, 197, 229. (33) Sibley, D. E. T.; Ramsay, G.; Lubrano, G. J.; Guilbault, G. G. Anal. Lett. 1993, 26, 1623. (34) Bright, F. B.; Betts, T. A.; Litwiler, K. S. Anal. Chem. 1990, 62, 1065. (35) Owaku, K.; Goto, M.; Ikariyame, Y.; Aizawa, M. Anal. Chem. 1995, 67, 1613. (36) Stanley, C. J.; Cox, R. B.; Cardosi, M. F.; Turner, A. P. F. J. Immunol. Methods 1988, 112, 153. (37) Nakamura, R. M.; Robbins, B. A. J. Clin. Lab. Anal. 1988, 2, 51. (38) Cardosi, M. F.; Birch, S. W. Int. Ind. Biotechnol. 1988, 8, 6. (39) Lu, B.; Smyth, M. R.; O’Kennedy, R. Anal. Chim. Acta 1996, 331, 97. (40) Abdel-Hamid, I.; Atanasov, P.; Ghindilis, A. L.; Wilkins, E. Sens. Actuators, B 1998, 49, 202.

a lower detection limit for RIgG analysis with the developed immunosensor can be achieved by lengthening the incubation time and decreasing the GaRIgG-Ur concentration in the incubation solution. Shortening of the analysis time might be achieved by adopting an automatic or semiautomatic flow-through immunoassay system. Because of the special characteristics of protein A, a direct immunoanalysis can be developed by using a sandwich scheme that should faciliate a more efficient and sensitive immunoanalysis.

ACKNOWLEDGMENT The authors thank the National Science Council of ROC for financial support of this work (Contract NSC 88-2113-M-110-002). Received for review December 13, 1999. Accepted March 10, 2000. AC9914317

Analytical Chemistry, Vol. 72, No. 13, July 1, 2000

2929