Label-Free Electrochemical Immunoassay for the ... - ACS Publications

Since hCG is found at highly elevated levels in pregnant female urine with the range of 30 000−200 000 mIU/mL (∼30−200 nM) by 8−10 weeks into ...
0 downloads 0 Views 253KB Size
Download PDF
Anal. Chem. 2006, 78, 5612-5616

Label-Free Electrochemical Immunoassay for the Detection of Human Chorionic Gonadotropin Hormone Kagan Kerman,*,† Naoki Nagatani,†,‡ Miyuki Chikae,† Teruko Yuhi,‡ Yuzuru Takamura,† and Eiichi Tamiya†

School of Materials Science, Japan Advanced Institute of Science and Technology (JAIST), Asahidai, Nomi City, Ishikawa, 923-1292 Japan, and Japan Science and Technology Agency (JST), Innovation Plaza, 2-13 Asahidai, Nomi City, Ishikawa 923-1211, Japan

Here we report on a new and rapid immunoassay for the label-free voltammetric detection of human chorionic gonadotropin hormone (hCG) in urine. Monitoring the changes in the current signals of antibodies (Abs) before and after the binding of the antigen (Ag) provides the basis for an immunoassay that is simple, rapid, and costeffective. Since hCG is found at highly elevated levels in pregnant female urine with the range of 30 000-200 000 mIU/mL (∼30-200 nM) by 8-10 weeks into pregnancy, its label-free electrochemical detection was achieved by using our method. The coverage of the electrode surface with the Ab and the incubation time with the target Ag were optimized for the detection of hCG. The limit of detection of our method was calculated to be 15 pM (n ) 3, ∼15 mIU/mL) in synthetic hCG samples and 20 pM (n ) 3, ∼20 mIU/mL) in human urine. The electrochemical results for the detection of hCG in the urine samples were in agreement with the results obtained using a reference system, enzyme-linked immunosorbent assay. Further research about the intrinsic electroactivity of Abs and their target molecules would surely provide new and sensitive screening assays, as well as extensive data regarding their interaction mechanisms. In clinical diagnostics, enzyme-linked immunosorbent assays (ELISAs) are reliable and useful means of detecting antigen (Ag) or antibody (Ab) concentrations. Since the report of Heineman and co-workers that the electrochemical detection of enzymebased immunoassays is possible using differential pulse polarography,1 a new approach has been defined, known as “electrochemical immunoassay”.2-5 The latest developments in the * Corresponding author: (e-mail) [email protected]. † Japan Advanced Institute of Science and Technology (JAIST). ‡ Japan Science and Technology Agency (JST). (1) Heineman, W. R.; Anderson, C. W.; Halsall, H. B. Science 1979, 204, 865866. (2) Wehmeyer, K. R.; Doyle, M. J.; Halsall, H. B.; Heineman, W. R. Methods Enzymol. 1983, 92, 432-444. (3) Thomas, J. H.; Kim, S. K.; Hesketh, P. J.; Halsall, H. B.; Heineman, W. R. Anal. Chem. 2004, 76, 2700-2707. (4) Farrell, S.; Ronkainen-Matsuno, N. J.; Halsall, H. B.; Heineman, W. R. Anal. Bioanal. Chem. 2004, 379, 358-367. (5) Boyaci, I. H.; Aguilar, Z. P.; Hossain, M.; Halsall, H. B.; Seliskar, C. J.; Heineman, W. R. Anal. Bioanal. Chem. 2005, 382, 1234-1241.

5612 Analytical Chemistry, Vol. 78, No. 15, August 1, 2006

microfluidic immunosensor systems have recently been reviewed by Heineman and co-workers.6 Label-free detection of the interaction between an Ag and Ab, that is, when a label, such as a fluorescent dye,7-9 is not employed, has become possible with optical, including surface plasmon resonance (SPR)10 and localized SPR,11,12 piezoelectrical, such as quartz crystal microbalance,13 and surface topology scanning, such as atomic force microscopy14 transducers. Recently, label-free voltammetric detection of proteins has become an important topic in bioanalysis.15-18 Our group has recently reported the electrochemical detection of amyloid peptides related to the Alzheimer’s disease by using their intrinsic electro-activity.19 Human chorionic gonadotropin (hCG) was chosen as the model protein in this report, because there are numerous commercially available monoclonal Abs, which provide excellent sensitivity and selectivity for hCG. hCG is a glycoprotein that is produced by the developing placenta shortly after a fertilized egg has been implanted in the uterine lining. hCG is composed of R(6) Bange, A.; Halsall, H. B.; Heineman, W. R. Biosens. Bioelectron. 2005, 20, 2488-2503. (7) Zhi, Z.-L.; Murakami, Y.; Morita, Y.; Hasan, H.; Tamiya, E. Anal. Biochem. 2003, 318, 236-243. (8) Murakami, Y.; Endo, T.; Yamamura, S.; Nagatani, N.; Takamura, Y.; Tamiya, E. Anal. Biochem. 2004, 334, 111-116. (9) Endo, T.; Okuyama, A.; Matsubara, Y.; Nishi, K.; Kobayashi, M.; Yamamura, S.; Morita, Y.; Takamura, Y.; Mizukami, H.; Tamiya, E. Anal. Chim. Acta 2005, 531, 7-13. (10) Yano, K.; Yoshino, T.; Shionoya, M.; Sawata, S. Y.; Ikebukuro, K.; Karube, I. Biosens. Bioelectron. 2003, 18, 1201-1207. (11) Endo, T.; Yamamura, S.; Nagatani, N.; Morita, Y.; Takamura, Y.; Tamiya, E. Sci. Technol. Adv. Mater. 2005, 6, 491-500. (12) Endo, T.; Kerman, K.; Nagatani, N.; Takamura, Y.; Tamiya, E. Anal. Chem. 2005, 77, 6976-6984. (13) Yano, K.; Bornscheuer, U. T.; Schmid, R. D.; Yoshitake, H.; Ji, H.-S.; Ikebukuro, K.; Masuda, Y.; Karube, I. Biosens. Bioelectron. 1998, 13, 397405. (14) Dong, Y.; Shannon, C. Anal. Chem. 2000, 72, 2371-2376. (15) Masarik, M.; Kizek, R.; Kramer, K. J.; Billova, S.; Brazdova, M.; Vacek, J.; Bailey, M.; Jelen, F.; Howard, J. A. Anal. Chem. 2003, 75, 2663-2669. (16) Masarik, M.; Stobiecka, A.; Kizek, R.; Jelen, F.; Pechan, Z.; Hoyer, W.; Jovin, T. M.; Subramaniam, V.; Palecek, E. Electroanalysis 2004, 16, 1172-1181. (17) Palecek, E.; Masarik, M.; Kizek, R.; Kuhlmeier, D.; Hassmann, J.; Schulein, J. Anal. Chem. 2004, 76, 5930-5936. (18) Palecek, E., Scheller, F., Wang, J., Eds. Electrochemistry of nucleic acids and proteins; Elsevier: Amsterdam, 2005. (19) Vestergaard, M.; Kerman, K.; Nagatani, N.; Takamura, Y.; Tamiya, E. J. Am. Chem. Soc. 2005, 127, 11892-11893. 10.1021/ac051762l CCC: $33.50

© 2006 American Chemical Society Published on Web 06/29/2006

and β-subunits.20 The R-subunit, a 116-amino acid sequence, is identical with that of other glycoprotein hormones, such as the luteinizing hormone and follicle and thyroid-stimulating hormones.21 The β-subunit, a 132-amino acid sequence, is unique to hCG and specific tests for it are not subject to hormonal crossreactivity.22 Besides, hCG can be found in highly elevated levels in the urine samples of pregnant females, which makes its labelfree detection an easy task. A woman normally produces 25 milliinternational units per milliliter (mIU/mL) of hCG 10 days after the conception. Generally, hCG levels doubles every two to three days after the conception. Accordingly, the concentration of hCG rises rapidly, frequently exceeding 100 mIU/mL by the first missed menstrual period and peaks in the range of 30 000-200 000 mIU/mL (∼30-200 nM) by 8-10 weeks into pregnancy. A hCG level of less than 5 mIU/mL (∼10 pM) generally indicates that one is not pregnant. A previously reported enzyme-based electrochemical immunoassay based on amperometry and magnetic beads23 in urine samples had an limit of detection (LOD) of 150 mIU/mL. Duan and Meyerhoff24 reported a separation-free, sandwich-type amperometric immunoassay using microporous gold electrodes with alkaline phosphatase. Chetcuti et al.25 reported an indirect perfluorosulfonated ionomer-coated electrochemical immunosensor for hCG detection in connection with horseradish peroxidase. Their detection limits was 11.2 mIU/mL. Lim and Matsunaga26 reported an electrochemical flow immunoassay system using capillary columns and ferrocene-conjugated immunoglobulin G for the detection of hCG. Recently, a hand-held amperometric sensor based on horseradish peroxidase has been reported to have an LOD of 1 mIU/mL.27 Electrochemical investigation of Ag-Ab interactions by differential pulse polarography on mercury electrodes was first reported by Wehmeyer et al.28 Then, Rodriguez-Flores29 reported the adsorptive stripping voltammetry of hCG and two of its specific Abs in solution using mercury electrodes. In this report, we achieved the detection of hCG in both synthetic and real samples at a carbon paste electrode by using label-free voltammetry. EXPERIMENTAL SECTION Apparatus. Adsorptive transfer square wave stripping voltammetry (SWSV) was performed with an Autolab PGSTAT 12 electrochemical analysis system (Eco Chemie, The Netherlands) in connection with its general purpose electrochemical system software. The three-electrode system consisted of a carbon paste electrode (CPE) as the working electrode, the reference electrode (20) Stenman, U.-H.; Alftan, H.; Hotakainen, K. Clin. Biochem. 2004, 37, 549561. (21) Filicori, M.; Fazleabas, A. T.; Huhtaniemi, I.; Licht, P.; Rao, C. V.; Tesarik, J.; Zygmunt, M. Fertil. Steril. 2005, 84, 275-284. (22) Snyder, J. A.; Haymond, S.; Parvin, C. A.; Gronowski, A. M.; Grenache, D. G. Clin. Chem. 2005, 51, 1830-1835. (23) Robinson, G. A.; Hill, H. A.; Philo, R. D.; Gear, J. M.; Rattle, S. J.; Forrest, G. C. Clin. Chem. 1985, 31, 1449-1452. (24) Duan, C.; Meyerhoff, M. E. Anal. Chem. 1994, 66, 1369-1377. (25) Chetcuti, A. F.; Wong, D. K.; Stuart, M. C. Anal. Chem. 1999, 71, 40884094. (26) Lim, T.-K.; Matsunaga, T. Biosens. Bioelectron. 2001, 16, 1063-1069. (27) Ivnitsky, D.; Sitdykov, R.; Ivnitsky, N. Anal. Chim. Acta 2004, 504, 265269. (28) Wehmeyer, K. R.; Halsall, H. B.; Heineman, W. R. Clin. Chem. 1982, 28, 1968-1972. (29) Rodriguez-Flores, J.; O’Kennedy, R.; Smyth, M. R. Biosensors 1989, 4, 1-13.

(Ag/AgCl), and a platinum wire as the auxiliary electrode. Carbon paste contained 75% (w/w) graphite powder and 25% (w/w) mineral oil (Sigma). The carbon paste was packed firmly into the electrode cavity (3 mm, i.d.) of a Teflon sleeve (Bioanalytical Systems Inc., Lafayette, IN) and polished to a smooth finish by gently rubbing over an ordinary weighing paper. The miniaturized reference electrode with 2-mm i.d. was obtained from Cypress Systems. Reagents. Recombinant hCG was obtained from Rohto Pharmaceutical Co., Ltd. (Osaka, Japan). Monoclonal antibody for the β-subunit of human chorionic gonadotropin hormone (β-hCGmAb) with an affinity constant of ∼1.5 × 1010 M-1 and monoclonal antibody to total prostate-specific antigen (T-PSA-mAb) were purchased from Japan Clinical Laboratories, Inc. (Kyoto, Japan). N-Hydroxysulfosuccinimide (NHS) and N-(3-dimethylamino)propyl)-N′-ethylcarbodiimide hydrochloride (EDC) were purchased from Aldrich. All other chemical reagents were supplied by Wako Pure Chemicals Co. (Tokyo, Japan) and were used as received. Ultrapure water, obtained from a Millipore Milli Q purification system (Millipore, Bedford, MA), was used for the preparation of all solutions and for cleaning of electrodes. Procedure. The present work with human samples was carried out in accordance to the ethical standard of the Helsinki Declaration of 1975, as revised in 1996. All electrodes were first rinsed with ultrapure water and blot dried before use. CPE was then electrochemically activated by applying 1.70 V for 1 min, which caused the formation of carboxylic groups on the carbon surface. Then, the electrodes were dipped in 50 mM phosphate buffer solution (PBS, pH 7.4) containing 5 mM EDC and 8 mM NHS for 1 h. After rinsing with ultrapure water, the electrode surface was immersed into 100 µL of 100 µg/mL protein A in acetate buffer (pH 4.8, 3 mM acetic acid and 7 mM sodium acetate) and left to react for 1 h at room temperature, to couple the lysine residues of protein A to the covalently activated electrodes. Protein A-modified electrodes were then extensively rinsed with water. No loss of IgG binding activity was observed over a three-day storage period. Then, the protein A-modified electrodes were immersed into an aliquot (100 µL) of a desired amount of mAb in PBS and allowed to interact with the Fc fragment of the Abs for 15 min to yield the sensing interfaces. The unbound Ab was washed away with blank PBS. The unreacted covalent-active surface groups were subsequently passivated by reaction with 1 mM ethanolamine under conditions similar to those reported by Lieber and co-workers30 for 1 h. For the direct capture assay, 20 µL of the synthetic hCG or real human urine sample was pipetted onto the inverted CPE surface modified with β-hCG-mAb. The urine sample was from the first urine of the morning, when hCG levels were usually the highest. The dilution rate of the freshly obtained urine was usually 1:2 with 50 mM PBS (pH 7.40), except as stated otherwise. The initial pH values of the real samples were found to be between 5.9 and 7.3. The real samples were measured by both enzymelinked immunosorbent assay (ELISA) and our electrical assay within 24 h of their receipt from the consenting donors. The preparation of “blank” electrodes was identical, except that there was no Ab or protein A immobilization step. After rinsing of the (30) Zheng, G.; Patolsky, F.; Cui, Y.; Wang, W. U.; Lieber, C. M. Nat. Biotechnol. 2005, 23, 1294-1301.

Analytical Chemistry, Vol. 78, No. 15, August 1, 2006

5613

Ab-Ag complex-modified electrode with PBS, it was transferred into 1 mL of blank PBS. The electrochemical measurement was performed using SWSV, while scanning the I/V plot with amplitude of 25 mV, a step potential of 5 mV, and 200 Hz frequency at 1 V/s. The raw voltammograms were treated using SavitzkyGolay smoothing (level 4). Three different measurements were performed for each concentration in three separately prepared solutions with three different electrodes (n ) 3), except as stated otherwise. Since the electrochemical protein oxidation signal was irreversible and only one measurement could be performed, disposable screen-printed carbon electrodes are suitable candidates for future applications of our method. RESULTS AND DISCUSSION The target protein, hCG, was first detected in synthetic samples by a standard additions calibration method of the known concentrations in blank PBS and negative male urine sample (Figure 1). The oxidation peak current of both β-hCG-mAb and hCG was observed at ∼0.60 V (vs Ag/AgCl). We obtained excellent recovery values in the case of negative male urine sample ranging between 88.9 and 97.4%. Figure 1 shows the SWSV data obtained from the spiking experiments with 20 pM hCG in PBS (Figure 1A-a) and in human male urine (Figure 1A-b). The current signal of the β-hCG-mAb/protein A-modified electrode is shown in Figure 1A-c. The protein A-modified electrode showed less current intensity (Figure 1A-d). The blocking reagent-coated electrode with no protein A and Ab modification showed no electrochemical signals (Figure 1A-e). Figure 1B shows the dependence of peak current intensities on the concentration of spiked hCG in PBS (Figure 1B-a) and in human male urine (Figure 1B-b). The calibration curve had a linear range from 15 to 300 pM. The LOD of our method was observed to be 15 pM (n ) 3, ∼15 mIU/mL) in synthetic hCG samples and 20 pM (n ) 3, ∼20 mIU/mL) in human male urine. ELISA provided quantitative data for the concentration of hCG in our samples, and our voltammetric results were in agreement with the ELISA results as shown in Figure 1C. First, we spiked known concentrations of hCG (100, 200, and 300 pM) into healthy male urine and compared the electrochemical and ELISA results. We also had two urine samples from 3-month pregnant females. Their electrochemical signals were above the linear range of our electrode, and according to ELISA, they contained 80 and 100 nM hCG. In the case of these concentration levels beyond our linear range (Figure 1C), we reached the conclusion that the sample was “positive”, because the main goal of this study was to be able to predict if a woman were pregnant or not by looking at the label-free electrochemical signals that we get from her urine sample. In other words, we determined that our biosensor was promising for “qualitative” determination of pregnancy, because the “quantitative” range was not wide enough to cover high hCG concentration levels. The coverage of CPE surface with Ab was controlled by monitoring the dependence of the Tyr/Trp oxidation signal of Ab on its concentration. The saturation of the signals was observed when 7.5 µg/mL β-hCG-mAb solution was immobilized (Figure 2A). According to our results with quantum dots,31 there was 7.5 ng/µm2 β-hCG-mAb on the electrode surface. We also determined (31) Kerman, K.; Endo, T.; Tsukamoto, M.; Chikae, M.; Takamura, Y.; Tamiya, E. Unpublished results.

5614 Analytical Chemistry, Vol. 78, No. 15, August 1, 2006

Figure 1. (A) Adsorptive transfer square wave stripping voltammetry for the spiked levels of synthetic hCG: (a) 20 pM hCG in PBS at β-hCG-mAb/protein A-modified CPE; (b) 20 pM hCG spiked in male human urine sample 2 at β-hCG-mAb/protein A-modified CPE; (c) blank PBS at β-hCG-mAb/protein A-modified CPE; (d) blank PBS at protein A-modified CPE; (e) blank male human urine 2 at blocking reagent-coated CPE. (B) Plot for the dependence of electrochemical signal obtained from spiked levels of synthetic hCG in blank 50 mM PBS (pH 7.4) (a, black line) and in human male serum sample 2 (b, gray line). Error bars indicate the standard deviations obtained in triplicates (n ) 3) as described in the Experimental Section. (C) Plot for the comparison of the hCG concentration results obtained from the electrochemical measurements and ELISA with three spiking experiments and two results from pregnant female urine samples 1 and 2.

that 85% ((10) of the fractions were active. Only a negligible amount, which was 0.50% ((0.25) Abs, eluted off during the incubation period with hCG and did not effect our results. The effect of incubation time (Figure 2B) on the electrical responses was also observed. We applied 20 pM hCG on our A-modified electrode surface and incubated for different periods of time. The increase in the electrochemical signal reached a

Figure 2. (A) Dependence of electrochemical signal on the concentration of β-hCG-mAb/protein A-modified CPE. Error bars indicate the standard deviations obtained in triplicates (n ) 3) as described in the Experimental Section. (B) Dependence of electrochemical signal on the incubation time obtained from 20 pM hCG at a β-hCG-mAb/protein A-modified CPE. The gray line indicates the moving average tendency of the curve. Error bars indicate the standard deviations obtained in triplicates (n ) 3) as described in the Experimental Section. The electrochemical measurement was performed as described in Figure 2A.

saturation level after 10 min. As a result, we employed 10 min for the incubation of hCG on the inverted electrode surface modified with 7.5 ng/µm2 β-hCG-mAb. We have tried to determine the reproducibility of a large number of separately prepared electrodes. We have obtained 11.8% reproducibility for 20 pM synthetic hCG in PBS by using 20 electrodes. Our bioelectronic immunoassay was challenged using human urine samples obtained from consenting healthy pregnant and nonpregnant donors. Figure 3A shows a representative voltammogram of the electrical measurements in the human urine samples. The current signal of 9-month pregnant female urine sample at the β-hCG-mAb/protein A-modified electrode is shown in Figure 3A-a. When the male urine sample was applied as the negative control, there was no significant change in the current response of the β-hCG-mAb/protein A-modified electrode (Figure 3A-b). We also used a different Ab (T-PSA-mAb) on the electrode. The urine sample of the 9-month pregnant female did not show any increase in the current response at T-PSA-mAb-modified electrode (Figure 3A-c). When we exposed the blocking reagentcoated bare CPE with no protein A and mAb modifications to the urine sample of the 9-month pregnant female, no electrochemical signals were observed, indicating that the suppression of the nonspecific adsorption of interfering molecules on the electrode surface could be achieved (Figure 3A-e). We were able to detect hCG in human urine samples of three females, sample 1 at 6 months (Figure 3A-a and B-b) and 2 and 3 both at 9 months of pregnancy (Figure 3B-c and d). We could also detect the hCG levels in the positive samples by using ELISA with the results in

Figure 3. (A) Adsorptive transfer SWSV for the detection of hCG in human urine samples: (a) 9-month pregnant female urine sample 1 at β-hCG-mAb/protein A-modified CPE; (b) male urine sample 1 at β-hCG-mAb/protein A-modified CPE; (c) 9-month pregnant female urine sample 1 at T-PSA-mAb-modified CPE; (d) 9-month pregnant female urine sample 1 at blocking reagent-coated CPE. (B) Current responses obtained from (a) β-hCG-mAb/protein A-modified CPE (b) after interaction with 6-month pregnant female urine sample, (c) after interaction with 9-month pregnant female urine sample 1, (d) after interaction with 9-month pregnant female urine sample 2, (e) after interaction with nonpregnant female urine sample 1, (f) after interaction with nonpregnant female urine sample 2, (g) after interaction with male urine sample 1, and (h) after interaction with male urine sample 2 with other conditions as described in the Experimental Section.

agreement with our voltammetric method. Urine samples from two nonpregnant females (Figure 3B-e and f) and two males (Figure 3B-g and h) were used as the negative controls. Accordingly, the hCG levels in these negative control samples were not detectable using ELISA. The true assay sensitivity is limited by the variations produced by different samples. This is usually established by measuring different negative samples that have been spiked with the same concentration of the analyte. Therefore, we have measured different negative samples that have been spiked with the same concentration of the analyte as shown in Figure 4. We spiked six male urine samples with 20 pM hCG and detected the electrochemical signals in triplicates. We also observed that the sampleto-sample variation was negligible. Uric acid and ascorbic acid exist together in biological fluids, such as blood and urine. The direct electrooxidation of uric acid (Figure 5a) and ascorbic acid (Figure 5b) at bare electrodes requires high peak potentials. The normal levels of uric acid in a 24-h collected urine sample range between 250 and 750 mg, which is approximately 1.5-4.5 mM. The voltammetric peak of 4.5 mM uric acid in PBS (pH 7) appeared at ∼0.46 V on a bare CPE Analytical Chemistry, Vol. 78, No. 15, August 1, 2006

5615

Figure 4. Current responses obtained from different negative samples that have been spiked with 20 pM hCG with other conditions as described in the Experimental Section.

Figure 5. Adsorptive transfer SWSV for the detection of (a) 4.5 mM uric acid at a bare CPE in PBS (pH 7.4), (b) 100 mg/dL ascorbic acid at a bare CPE in PBS (pH 7.4), and (c) 4.5 mM uric acid and 100 mg/dL ascorbic acid spiked in 9-month pregnant female urine sample 2 at β-hCG-mAb/protein A-modified CPE as described in the Experimental Section.

(Figure 5a), which is in agreement with the report of Wang et al.34 Moreover, when we spiked 4.5 mM uric acid into pregnant female urine sample, the oxidation peak potential that we obtained for hCG was at 0.60 V and no uric acid signal was observed at 0.46 V (Figure 5c). The oxidation of ascorbic acid (pKa ) 4.17) has been documented to be a one-proton and two-electron process at 0.42 V in neutral media.32 At a bare CPE, ascorbic acid shows a broad and irreversible oxidation peak at 0.42 V (Figure 5b). Brigden et al.33 examined 4379 routine urinalysis specimens and found that 22.8% were positive for ascorbic acid at a mean concentration level of (32) Raj, C. R.; Ohsaka, T. J. Electroanal. Chem. 2001, 496, 44-49. (33) Brigden, M. L.; Edgell, D.; McPherson, M.; Leadbeater, A.; Hoag, A. Clin. Chem. 1992, 38, 426-431. (34) Wang, Z.; Wang, Y.; Luo, G. Analyst 2002, 127, 1353-1358.

5616

Analytical Chemistry, Vol. 78, No. 15, August 1, 2006

over 37 mg/dL. This same study showed that even a modest 250mg dose of vitamin C produced a mean urinary ascorbate value of 31 mg/dL, and this increased to 62 mg/dL with a 500-mg dose.33 Therefore, we spiked 100 mg/dL ascorbic acid in the pregnant female urine sample and were successful to suppress the signal of ascorbic acid at 0.42 V and observe the signal of hCG at 0.60 V (Figure 5c). This significant peak potential difference between the common electroactive solutes and proteins was useful for us to discriminate their signals. Two mechanisms are proposed for the label-free detection of Ab-Ag reactions in this report. The increase in current might be arising from the conformational changes of hCG during the binding process, which in turn could affect the packing density of the Ab causing defects in the monolayer coverage. These defects could permit electron transfer to occur more readily through the electrolyte, which would have better contact with the carbon surface and be trapped within the matrix. The presence of an excess of nine positively charged amino acids on the protein should be noted as it may account for the observed modulation of current. The sensitivity of our assay to pH suggested that the assay response might actually be caused by changes in the local pH produced by the highly positive charge of the bound hCG. The variations in pH would alter the ionization states in the side chains of the amino acids, which would then change the protein charge distributions and hydrogen-bonding requirements. Thus, the modulations in the intrinsic oxidation current signals could be monitored. CONCLUSIONS We have used hCG, the pregnancy biomarker, as the model system in our bioelectronic immunoassay with both synthetic and real samples. Our protocol is readily transferable for application to other immunological tests, in which the highly elevated target protein levels would be utilized as a diagnostic and therapeutic indicator. The remarkable sensitivity and simplicity of our system is derived from utilizing the intrinsic electroactivity of Abs and Ags. Further research about the electroactivity of Abs and their target molecules would surely provide new and sensitive screening assays, as well as extensive data regarding their interaction mechanisms in vitro and, if applicable, in vivo. Our laboratory is currently working on the development of a miniaturized electrochemical biosensor for the parallel detection of a large number of clinically important proteins.

Received for review October 1, 2005. Accepted June 13, 2006. AC051762L