Label-free immunosensor for prostate-specific antigen based on single-walled carbon nanotube array-modified microelectrodes

Biosensors and Bioelectronics 22 (2007) 2377–2381

Short communication

Label-free immunosensor for prostate-specific antigen based on single-walled carbon nanotube array-modified microelectrodes Jun Okuno a , Kenzo Maehashi a,∗ , Kagan Kerman b , Yuzuru Takamura b,c , Kazuhiko Matsumoto a , Eiichi Tamiya b,∗ a

Institute of Scientific and Industrial Research, Osaka University, 8-1 Mihogaoka, Ibaraki, Osaka 567-0047, Japan b School of Materials Science, Japan Advanced Institute of Science and Technology (JAIST), Asahidai, Nomi City, Ishikawa 923-1292, Japan c Precursory Research for Embryonic Science and Technology (PRESTO), Japan Science and Technology Agency (JST), 4-1-8 Honcho Kawaguchi, Saitama, Japan Received 25 April 2006; received in revised form 26 August 2006; accepted 1 September 2006 Available online 15 November 2006

Abstract We have fabricated a label-free electrochemical immunosensor using microelectrode arrays modified with single-walled carbon nanotubes (SWNTs). Label-free detection of a cancer marker, total prostate-specific antigen (T-PSA), was carried out using differential pulse voltammetry (DPV). The current signals, derived from the oxidation of tyrosine (Tyr), and tryptophan (Trp) residues, increased with the interaction between T-PSA on T-PSA-mAb covalently immobilized on SWNTs. The selectivity of our biosensor was challenged using bovine serum albumin (BSA) as the target protein. The detection limit for T-PSA was determined as 0.25 ng/mL. Since the cut-off limit of T-PSA between prostate hyperplasia and cancer is 4 ng/mL, the performance of our label-free electrochemical immunosensor seems promising for further clinical applications. © 2006 Elsevier B.V. All rights reserved. Keywords: Electrochemical biosensors; Single-walled carbon nanotubes; Arrayed electrodes; Prostate-specific antigen; Immunosensor

1. Introduction The detection of biomolecules has been under intense investigation in many diverse fields; such as genomics, forensic analysis, and environmental monitoring. Fundamentally, biosensors rely on the transduction of the molecular recognition event into an analytical output signal. Among the molecular recognition biomolecules, antibodies (Abs) have been the main focus of intensive research and development, due to their capability to bind various important antigens with high affinity and specificity (Luppa et al., 2001). The majority of the analytical approaches require adequate transducing elements; such as enzymes, fluorescent dyes, etc., to generate a physically readable signal from this recognition event. However, these methods involve timeconsuming labeling procedures or sophisticated experimental

∗

Corresponding authors. E-mail addresses: [email protected] (K. Maehashi), [email protected] (E. Tamiya). 0956-5663/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2006.09.038

techniques. Labeling with various agents may even compromise the binding efficiency of the Abs. Therefore, there is still a great demand for sensitive and disposable label-free immunosensors (Bange et al., 2005). Electrochemical detection for biomolecules is of great interest owing to its high sensitivity and compatibility for miniaturization and mass-fabrication (Drummond et al., 2003; Lucarelli et al., 2004). Especially, label-free electrochemistry can be performed using miniaturized biosensors, because of their simple handling and procedure that does not require a complicated labeling process (Kerman et al., 2004a). Single-walled carbon nanotubes (SWNTs) are one of the most promising candidates for the development of nanoscale biosensors due to their unique and well-defined electrical and mechanical properties (Dresselhaus et al., 1996; Maehashi et al., 2004a; Kaminishi et al., 2005). SWNTs were reported to have the high ability to promote electron-transfer reactions in electrochemical measurements (Wang, 2004). SWNTs have a high aspect ratio that the total surface area for electrodes becomes significantly larger on the same site than the surface of a bare

2378

J. Okuno et al. / Biosensors and Bioelectronics 22 (2007) 2377–2381

electrode (Dresselhaus et al., 1996). Electrochemical biosensors based on carbon nanotube array (CNT)-modified electrodes have been reported intensively (Kerman et al., 2004b, 2005a,b). Moreover, the direct growth of SWNTs on the surface of Ptarrayed microelectrodes has enabled a promising platform for highly sensitive measurements of biomolecules (Wildgoose et al., 2006; Sinha et al., 2006). In this paper, we performed the detection of a protein cancer marker, prostate-specific antigen (PSA) using microelectrodes modified with SWNTs. PSA is an androgen-regulated serine protease (Balk et al., 2003), and it exists in numerous forms (Lilja et al., 1991). PSA that enters the circulatory system, is rapidly bound by protease inhibitors, primarily ␣1 -antichymotrypsin (ACT), although a fraction is inactivated in the lumen by proteolysis and circulates as free PSA (f-PSA) (Lilja et al., 1991). Total PSA (T-PSA) refers to the sum of f-PSA and PSA/ACT complex in serum. T-PSA level significantly increases in serum during prostate cancer (Constantinou and Feneley, 2006). T-PSA screening has dramatically improved the diagnosis and management of prostate cancer and the prostatic diseases (Chodak and Warren, 2006). The cut-off limit of T-PSA between prostate hyperplasia and cancer is 4 ng/mL (Balk et al., 2003). The successful integration of a one-step lateral flow immunoassay format and capacitance-based detection of fPSA and T-PSA using an electrochemical transducer coated with a pH-sensitive polymer layer has recently been reported by Fern´andez-S´anchez et al. (2004). Li et al. (2005) reported the immunological detection of T-PSA using n-type In2 O3 nanowires and p-type carbon nanotubes on a field-effect transistor (FET)-based device. Recently, Lieber and co-workers described the label-free and multiplexed electrical detection of cancer markers using silicon nanowire-based FET devices with a detection limit of 75 fg/mL or ∼2 fM PSA in undi-

luted serum samples (Zheng et al., 2005). Recently, Kerman et al. (2006) reported the label-free electrochemical detection of human chorionic gonadotropin hormone in urine samples. In this report, T-PSA was detected for the first time by means of its intrinsic current signal on SWCNT-modified microelectrode arrays. 2. Experimental 2.1. Reagents Monoclonal antibody to total prostate-specific antigen (TPSA-mAb) was purchased from Scripps Laboratories (San Diego, CA) and Rohto Pharmaceutical Co., Ltd. (Osaka, Japan). The purified protein of T-PSA was obtained from Chemicon International (Temecula, CA). All other proteins and chemical reagents were supplied by Wako Pure Chemicals Co. (Tokyo, Japan) and were used as received. Ultra-pure water, obtained from Millipore Milli Q purification system (Millipore, Bedford, MA, USA), was used for the preparation of all solutions. 2.2. Instruments Differential pulse voltammetry (DPV) was performed with a ␮Autolab PGSTAT electrochemical analysis system (Eco Chemie, The Netherlands) in connection with its General Purpose Electrochemical System (GPES) software. The miniaturized reference electrode with 2 mm i.d. (RE, Ag/AgCl) was obtained from Cypress Systems (KS, USA). Fig. 1 shows the schematic structure of the experimental set-up for the microelectrodes modified with SWNTs. The working electrodes were surrounded by a silicon chamber attached on the substrate. The

Fig. 1. (A) Illustration for the experimental set-up with single-walled carbon nanotube (SWNT)-modified Pt microelectrode as the working electrode, Pt wire as the counter electrode (CE), and the miniaturized reference electrode (RE, Ag/AgCl) with the scanning electron microscopy (SEM) images of a SWNT-modified microelectrode; (B) illustration for the label-free electrochemical immunosensor design. Monoclonal antibodies against total prostate-specific antigen (T-PSA-mAb) were covalently anchored onto the SWNTs using 1-pyrenebutanoic acid succinimidyl ester (Linker). The peak current for the intrinsic oxidation of proteins, deriving from their electro-active amino acids, increases as the antigen–antibody complex is formed.

J. Okuno et al. / Biosensors and Bioelectronics 22 (2007) 2377–2381

counter electrode (CE, Pt wire) and RE were immersed into the analyte solution in the silicon chamber. Images of the microelectrodes and SWNTs in Fig. 1, were taken using a Hitachi S4800 (Japan) scanning electron microscope (SEM) at an acceleration voltage of 1 kV.

2379

rinsed with PBS, and the electrochemical measurement was taken in blank PBS. The electrochemical measurements were performed for five times for each condition (n = 5) except otherwise stated. The results show the average of five measurements with the error bars indicating the relative standard deviation (R.S.D.).

2.3. Device fabrication Si wafer (p+ -type) with a thermally oxidized SiO2 film (150 nm) was used as the substrate. Pt and Ti were patterned on the substrate as the base for working electrodes using conventional photolithography and metal lift-off processes. Then, SiO2 passivation film (100 nm) was formed on the patterned substrate by plasma CVD at 200 ◦ C. The film, which was located above the both ends of the working electrodes, was partially etched with CF4 using photolithography. The geometric surface area of a single electrode was 40,000 ␮m2 as shown in the SEM image in Fig. 1. Finally, SWNTs were synthesized on Pt electrodes by thermal CVD method using a catalyst, which contained Fe(NO3 )3 ·9H2 O, MoO2 (acac)2 and alumina nanoparticles in the liquid phase. Afterwards, the substrate was heated up to 900 ◦ C under Ar, and then ethanol vapor was supplied for 10 min (Maehashi et al., 2004b). The diameters of SWNTs were estimated to be 1–2 nm by Raman scattering measurements (Maehashi et al., 2004b). In this way, an array of 30 microelectrodes with SWNT modification was arrayed on a single substrate. 2.4. Immunosensor preparation SWNT-modified microelectrodes were incubated with 5 mM 1-pyrenebutanoic acid succinimidyl ester (Aldrich, Japan) in dry dimethylformamide solution (Linker) for 1 h, followed by rinsing with 50 mM phosphate buffer solution (PBS, pH 7.4). Next, in order to covalently immobilize T-PSA-mAb on the SWNTs, the microelectrodes were exposed to 12 ␮g/mL T-PSA-mAb in 50 mM phosphate buffer solution (pH 7.4) overnight. After rinsing with blank PBS, to deactivate and block the excess reactive groups remaining on the surface, 100 mM ethanolamine was added onto the resulting electrodes, and incubated for 30 min. After rinsing with blank PBS, T-PSA was introduced onto the microelectrodes, and incubated for 1 h. Then, the arrays were

3. Results and discussion The label-free electrochemical method relies on the detection of the intrinsic oxidation signal of proteins deriving from their electro-active amino acid residues, tyrosine (Tyr) and tryptophan (Trp). T-PSA contains 13 Tyr and 11 Trp (Lilja et al., 1991; Leinonen et al., 2002). Fig. 2A shows the electrochemical signals of SWNT-modified electrodes recorded using DPV. The peak current signal was obtained at +0.5 V from only T-PSAmAb as shown in Fig. 2A-b and B-b. After the incubation with 1 ng/mL T-PSA on the T-PSA-mAb-modified SWNT electrodes, the electrochemical current signal significantly increased as shown in Fig. 2A-a and B-a. When T-PSA-mAb-modified microelectrodes were exposed to a high concentration (100 ng/mL) of non-specific antigen, bovine serum albumin (BSA), no significant increase in the peak current signal was observed (Fig. 2A-c). As shown in Fig. 2B, only a slight decrease in the response was recorded, which was within the relative standard deviation as indicated by error bars on the columns (Fig. 2B-c). When 1 ng/mL T-PSA was mixed with 100 ng/mL BSA, we could observe the increase in the current response again. However, the response (Fig. 2B-d) was slightly lower than the one obtained in the absence of an interfering antigen (Fig. 2B-a). The electrochemical signals after T-PSA incubation with SWNT-modified and bare Pt microelectrodes were also monitored (data not shown). SWNTs were not grown on bare Pt microelectrodes. High peak current signals were clearly observed from the SWNT-modified microelectrodes, whereas no signal was obtained from the bare Pt ones. This result indicated that T-PSA and T-PSA-mAb were not non-specifically adsorbed on the bare Pt microelectrodes, because the linker molecule that we utilized to covalently attach T-PSA-mAb, had no affinity towards Pt surface. Therefore, the complementary binding between T-PSA and T-PSA-mAb, and the electrochemical signal resulted only from the SWNT-modified microelectrodes.

Fig. 2. (A) Differential pulse voltammograms obtained from the SWNT-modified microelectrodes covalently attached with 12 ␮g/mL T-PSA-mAb after the incubation with: (a) 1 ng/mL T-PSA, (b) blank 50 mM phosphate buffer solution (PBS, pH 7.4), and (c) 100 ng/mL BSA; (B) columns for the average of peak current signals (n = 5) obtained from the SWNT-modified microelectrodes covalently attached with 12 ␮g/mL T-PSA-mAb after the incubation with: (a) 1 ng/mL T-PSA, (b) blank PBS, (c) 100 ng/mL BSA, and (d) 1 ng/mL T-PSA and 100 ng/mL BSA together in PBS with error bars showing the relative standard deviation.

2380

J. Okuno et al. / Biosensors and Bioelectronics 22 (2007) 2377–2381

Fig. 3. (A) The dependence of the electrochemical signals on T-PSA concentration with error bars showing the relative standard deviation and (B) the dependence of the peak potentials on pH.

Since the large aromatic pyrene ring of the linker molecule had high affinity towards the graphite structure of SWNTs, the covalent attachment of T-PSA-mAbs was achieved only on SWNT-modified microelectrodes. The electrochemical current signals after the T-PSA binding were plotted as a function of T-PSA concentration, as shown in Fig. 3A, where the electrochemical intensities were normalized by the current signal of T-PSA-mAb alone. The error bars indicate the R.S.D. of five replicative measurements (n = 5). The electrochemical evaluation revealed that the peak current intensities linearly increased with increasing T-PSA concentration. The detection limit of T-PSA was determined as 0.25 ng/mL with a signal-to-noise (S/N) ratio of 3. Since the cut-off limit of T-PSA between prostate hyperplasia and cancer is 4 ng/mL, the performance of our biosensor seems promising for further applications with clinical samples. Two mechanisms are suggested for the label-free detection of Ab–Ag reactions in this report. The increase in current might be arising from the conformational changes of T-PSA 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 positively charged amino acids on T-PSA should be noted as it may account for the observed modulation of current. Additionally, we observed the dependence of T-PSA responses on pH of the buffer solution. In the final step of the experiments, T-PSA/T-PSA-mAb complex modified microelectrodes were exposed to blank buffer solutions at various pH conditions. We used 0.1 M HCl (pH 1 and 2), 0.5 M acetate buffer solution (pH 3, 4, and 4.5), PBS (pH 7.4), and 20 mM Tris buffer solution (pH 9). A reverse relationship between the peak potential of the intrinsic protein oxidation signal and pH was observed. As shown in Fig. 3B, a linear increase in the peak potential value was recorded as pH decreased. High sensitivity of proteins 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 T-PSA. 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.

4. Conclusions We have fabricated a label-free immunosensor with SWNTmodified microelectrodes. The electrochemical label-free detection of a cancer marker, T-PSA, was carried out using DPV. Label-free electrochemical immunosensor in connection with a linker molecule that has a high affinity for graphite structure, exhibited high selectivity with successful suppression of the non-specific adsorption. We suggest that SWNT-arrayed microelectrodes are promising devices for applications in numerous biosensor schemes. Acknowledgments This research was partially supported by Core Research for Evolutional Science and Technology (CREST), and proposaloriented research promotion program, Japan Science and Technology Corporation (JST), and the New Energy and Industrial Technology Development Organization (NEDO). References Bange, A., Halsall, H.B., Heineman, W.R., 2005. Biosens. Bioelectron. 20, 2488–2503. Balk, S.P., Ko, Y.-J., Bubley, G.J.J., 2003. Oncology 21, 383–391. Constantinou, J., Feneley, M.R., 2006. Prostate Cancer Prostatic Dis. 9, 6–13. Chodak, G.W., Warren, K.S., 2006. Prostate Cancer Prostatic Dis. 9, 25–29. Dresselhaus, M.S., Dresselhaus, G., Eklund, P.C., 1996. Science of Fullerenes and Carbon Nanotubes: Their Properties and Applications. Academic Press, New York. Drummond, T.G., Hill, M.G., Barton, J.K., 2003. Nature Biotechnol. 21, 1192–1199. Fern´andez-S´anchez, C., McNeil, C.J., Rawson, K., Nilsson, O., 2004. Anal. Chem. 76, 5649–5656. Kaminishi, D., Ozaki, H., Ohno, Y., Maehashi, K., Inoue, K., Matsumoto, K., Seri, Y., Masuda, A., Matsumura, H., 2005. Appl. Phys. Lett. 86, 113115. Kerman, K., Morita, Y., Takamura, Y., Tamiya, E., Maehashi, K., Matsumoto, K., 2005a. Nanobiotechnology 1, 65–70. Kerman, K., Morita, Y., Takamura, Y., Tamiya, E., 2005b. Anal. Bioanal. Chem. 381, 1114–1121. Kerman, K., Kobayashi, M., Tamiya, E., 2004a. Meas. Sci. Technol. 15, R1–R11. Kerman, K., Morita, Y., Takamura, Y., Ozsoz, M., Tamiya, E., 2004b. Electroanalysis 16, 1667–1672. Kerman, K., Nagatani, N., Chikae, M., Yuhi, T., Takamura, Y., Tamiya, E., 2006. Anal. Chem. 78, 5612–5616. Leinonen, J., Wu, P., Stenman, U.-H., 2002. Clin. Chem. 48, 2208–2216.

J. Okuno et al. / Biosensors and Bioelectronics 22 (2007) 2377–2381 Li, C., Curreli, M., Lin, H., Lei, B., Ishikawa, F.N., Datar, R., Cote, R.J., Thompson, M.E., Zhou, C.J., 2005. J. Am. Chem. Soc. 127, 12484–12485. Lilja, H., Christensson, A., Dahlen, U., Matikainen, M.T., Nilsson, O., Pettersson, K., Lovgren, T., 1991. Clin. Chem. 37, 1618–1625. Lucarelli, F., Marrazza, G., Turner, A.P.F., Mascini, M., 2004. Biosens. Bioelectron. 19, 515–530. Luppa, P.B., Sokoll, L.J., Chan, D.W., 2001. Clin. Chim. Acta 314, 1–26. Maehashi, K., Matsumoto, K., Kerman, K., Takamura, Y., Tamiya, E., 2004a. Jpn. J. Appl. Phys. 43, L1558–L1560.

2381

Maehashi, K., Ohno, Y., Inoue, K., Matsumoto, K., 2004b. Appl. Phys. Lett. 85, 858. Sinha, N., Ma, J., Yeow, J.T., 2006. J. Nanosci. Nanotechnol. 6, 573–590. Wang, J., 2004. Electroanalysis 17, 7–14. Wildgoose, G.G., Banks, C.E., Leventis, H.C., Compton, R.G., 2006. Microchim. Acta 152, 187–214. Zheng, G., Patolsky, F., Cui, Y., Wang, W.U., Lieber, C.M., 2005. Nature Biotechnol. 23, 1294–1301.