A sensitive impedimetric thrombin aptasensor based on polyamidoamine dendrimer

Talanta 78 (2009) 1240–1245

Contents lists available at ScienceDirect

Talanta journal homepage: www.elsevier.com/locate/talanta

A sensitive impedimetric thrombin aptasensor based on polyamidoamine dendrimer Zhanxia Zhang a,b , Wen Yang a,b , Juan Wang a,b , Cheng Yang a,b , Fan Yang a , Xiurong Yang a,∗ a

State Key Laboratory of Electroanalytical Chemistry, Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Renmin Street 5625, Changchun, Jilin 130022, China Graduate School of the Chinese Academy of Sciences, Beijing 100039, China

b

a r t i c l e

i n f o

Article history: Received 28 October 2008 Received in revised form 14 January 2009 Accepted 18 January 2009 Available online 24 January 2009 Keywords: Aptamer Thrombin Polyamidoamine dendrimer Electrochemical impedance spectroscopy

a b s t r a c t A label-free and highly sensitive impedimetric aptasensor based on a polyamidoamine dendrimer modified gold electrode was developed for the determination of thrombin. Amino-terminated polyamidoamine dendrimer was firstly covalently attached to the cysteine functionalized gold electrode through glutaraldehyde coupling. Subsequently, the dendrimer was activated with glutaraldehyde, and aminomodified thrombin aptamer probe was immobilized onto the activated dendrimer monolayer film. The layer-by-layer assembly process was traced by surface plasmon resonance and electrochemical impedance spectroscopy. After electrode preparation, the detection of thrombin was investigated in the presence of the reversible [Fe(CN)6 ]3−/4− redox couple using impedance technique. The results showed that the charge-transfer resistance (Rct ) value had a linear relationship with the concentrations of thrombin in the range of 1–50 nM, and the detection limit (S/N = 3) as low as 0.01 nM was obtained. The covalent immobilization of dendrimer on the electrode surface not only improved the immobilization capacity of probe molecules but also magnified the response signal. The aptasensor exhibited favorable regeneration ability, selectivity and stability. It also showed the detectability in biological fluid. © 2009 Elsevier B.V. All rights reserved.

1. Introduction Aptamers are a kind of single-stranded DNA or RNA sequences generated by in vitro selection techniques from a pool of DNA or RNA by repetitive binding of the target molecules. They possess high recognition ability towards specific molecular targets ranging from small inorganic and organic substances to proteins and even cells [1,2]. Aptamers have numerous advantageous characteristics over traditional recognition elements such as antibodies, including ease of synthesis, thermal stability and lack of immunogenicity [3]. Based on their biomolecular recognition ability, a number of aptamer biosensors (aptasensors) have been developed over the past decade [4]. Up to now, aptasensors mainly include optical aptasensors [5–8] and electrochemical aptasensors [9–12]. The field of electrochemical aptasensors has developed rapidly in recent years because they can provide fast, simple and inexpensive detection capabilities for biological binding events [13,14]. In order to amplify signals, many electrochemical aptasensors have been constructed mainly based on the aptamers labeled with electroactive materials, such as ferrocene and methylene blue [15–18]. Such labeling bears the

∗ Corresponding author. Tel.: +86 431 85262056; fax: +86 431 85689278. E-mail address: [email protected] (X. Yang). 0039-9140/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2009.01.034

disadvantages that the labeling process not only makes the experiment comparatively complicated and expensive but also affects the binding affinity between the targets and their aptamers to a certain degree [19]. Therefore, it is very necessary to develop a kind of label-free and highly sensitive electrochemical aptasensors. Electrochemical impedance spectroscopy (EIS) has been proved as one of the most sensitive tools for the analysis of interfacial properties because of its unique properties of high sensitivity, low cost, convenience and being label-free [20–23]. Recently, Zhang’s group has reported an EIS thrombin aptasensor with the detection limit of 0.06 nM by using the signal enhancement of gold nanoparticles, which were electrodeposited onto a glassy carbon electrode [24]. In the paper, a label-free impedimetric aptasensor for the determination of thrombin based on a dendrimer-functionalized gold electrode was constructed, and the dendrimer employed here was expected to improve the response signals. Dendrimers are a new class of synthetic macromolecules. They possess regularly branched treelike spherical morphologies and monodisperse sizes depending on the number of generations. The unique structural properties of dendrimers, such as structural homogeneity, integrity, controlled composition and biocompatibility, extend their use in biosensing applications [25,26]. Many approaches adopting dendrimers as biosensor materials have been reported [27,28]. A DNA biosensor using a polyamidoamine (PAMAM) dendrimer modified electrode had been previously

Z. Zhang et al. / Talanta 78 (2009) 1240–1245

1241

Scheme 1. The schematic illustration for the fabrication procedure on a gold surface. (1) 10 mM CH, 10 h; (2) 5% GA in 0.01 M PBS (pH 7.4), 4 h; (3) 5% PAMAM dendrimer methanolic solution (16 h) and 5 mM NaBH4 (30 min); (4) 5% GA in 0.01 M PBS (pH 7.4), 4 h; (5) 10 ␮M probe TBA in 0.01 M PBS (pH 7.4) (16 h) and 5 mM NaBH4 (30 min); (6) interaction with target thrombin, 1 h. CH, cysteamine hydrochloride; GA, glutaraldehyde; PAMAM, G4 -NH2 PAMAM dendrimer; TBA, thrombin-binding-aptamer.

developed by our group [29]. The sensor exhibited good assembly capacity for probe DNA with favorable sensitivity and stability. Thrombin was selected as target molecules for this study because it is a major stimulus of both procoagulant and anticoagulant reactions, and thus is a key element in various pathogenesis, including leukemia, arterial thrombosis and liver disease, etc. [30,31]. The most extensively investigated prototype thrombinbinding-aptamer (TBA) is a 15-mer single-stranded DNA. When binding to thrombin, TBA forms an intermolecular quadruplex structure and restricts the activity of thrombin [32]. So the study of thrombin and its aptamer interaction plays essential roles in fundamental research and clinical application, such as detection and quantification of thrombin in plasma, regulation of blood clotting in surgery, etc. [33]. In this paper, amino-terminated PAMAM dendrimer was firstly covalently attached to the cysteine self-assembly gold electrode surface through glutaraldehyde coupling. Then the dendrimer was activated with glutaraldehyde, and amino-modified TBA probe was immobilized onto the activated dendrimer monolayer film. The whole assembly process was characterized by SPR and EIS, and recognition events of the aptasensor for thrombin were monitored by impedance technique. 2. Experimental 2.1. Chemical and reagents Cysteamine hydrochloride (CH) was purchased from Fluka. Glutaraldehyde (GA, 25% aqueous solution) was supplied by Acros. G4 -NH2 polyamidoamine dendrimer (MW 14,215, 10% methanol solution) was obtained from Aldrich. ␣-Thrombin from bovine plasma (1000 ␮) was provided by Sigma. Bovine serum albumin (BSA) was obtained from Beijing Like biochemical technology and trade Co., Ltd. Guanidine hydrochloride and bovine serum was purchased from Beijing Dingguo biotechnology Co., Ltd. 15-mer amino-terminal TBA was provided by Shanghai Sangon biological engineering technology and services Co., Ltd. 15-mer TBA: 5 -NH2 (CH2 )6 -GGTTGGTGTGGTTGG-3 . The immobilization buffer was 0.01 M phosphate-buffered saline (PBS, pH 7.4). The redox couple solution was 5 mM equimolar mixture of K3 [Fe(CN)6 ] and K4 [Fe(CN)6 ] in 0.01 M PBS (pH 7.4) containing 0.1 M KCl. And 20 mM Tris–HCl buffer (pH 7.4) containing 140 mM NaCl, 5 mM KCl and 5 mM MgCl2 was used as thrombin-

binding buffer. All other chemicals were of analytical grade and used without further purification. Doubly distilled water was used throughout. 2.2. Apparatus 2.2.1. SPR system SPR was used to characterize the fabrication procedure of the aptasensor because of its gold reaction surface. Angle-resolved SPR measurements were carried out with a cuvette-based Autolab SPR instrument (Eco Chemie BV, Netherlands). A cleaned gold disk was attached to the half-cylinder prism with a refractive-indexmatching oil, and then the SPR cuvette was mounted on the gold substrate. Before sample injection, the cuvette cell was balanced with 0.01 M PBS (pH 7.4) for at least 30 min. The cell and the disk were washed three times with water after each reaction process and the SPR data were recorded in 0.01 M PBS (pH 7.4). For this instrument, the measured  values correspond to the amount of adsorbed substances with a mass sensitivity coefficient of 120 millidegree per 100 ng/cm2 . 2.2.2. EIS system EIS experiments were carried out using a conventional threeelectrode system with an Autolab PGSTAT30 electrochemical analyzer system (Eco Chemie BV, Netherlands), controlled by frequency response analyzer (FRA) 4.9 software. The modified gold electrode was used as working electrode, a platinum wire as counter electrode and an Ag/AgCl electrode with saturated KCl solution as reference electrode. All impedance measurements were performed with a 0.23 V alternating current potential and a 5 mV voltage amplitude in a frequency range from 100 kHz to 0.1 Hz. And the supporting electrolyte was 5 mM [Fe(CN)6 ]3−/4− in 0.01 M PBS (pH 7.4) containing 0.1 M KCl. The frequency interval was divided into 61 logarithmically equidistant measuring points (10 points pre decade). The electrochemical cell was housed in a specially shielded cage to reduce stray electrical noise during measurements. 2.3. Modification procedure A gold electrode (A = 6.4 mm2 ) was polished carefully with 1, 0.3, 0.05 ␮m alumina slurries and washed ultrasonically with water. Then it was electrochemically cleaned in 0.1 M H2 SO4 by cyclic potential scanning between −0.2 and 1.55 V until a standard cyclic voltammogram of gold electrode was obtained. Subsequently, the

1242

Z. Zhang et al. / Talanta 78 (2009) 1240–1245

gold electrode was washed ultrasonically with water and absolute ethanol, respectively, and dried in a nitrogen stream. Scheme 1 provides an overview of the modification procedure of the aptasensor. Firstly, the cleaned gold electrode was immersed into a 10 mM cysteamine hydrochloride aqueous solution for about 10 h. Then the CH modified electrode was placed in 5% glutaraldehyde (GA) in 0.01 M PBS (pH 7.4) for 4 h. Secondly, the gold electrode was incubated in a 5% G4 -NH2 PAMAM dendrimer methanolic solution for about 16 h, washed with methanol and dried under flowing nitrogen gas. The covalent immobilization of G4 -NH2 PAMAM dendrimer is based on the fact that each aldehyde group of GA allows the introduction of one amino group, either from the cysteamine or from the dendrimer to form Schiff base configuration. The double bond in Schiff base was reduced by 5 mM NaBH4 for 30 min and formed a more stable single bond structure. Finally, for the immobilization of probe TBA, the dendrimer modified electrode was activated with 5% GA for 4 h. Then the electrode was immersed in 0.01 M PBS (pH 7.4) containing 10 ␮M TBA for about 16 h at ambient temperature, and reduced by 5 mM NaBH4 for 30 min. The interaction time of thrombin and its aptamer ranged from less than 20 min to about 1 h according to the previous articles [16,29,34]. In order to make the reaction completely, 1 h was adopted in our experiments. The detection of thrombin was carried out by immersing the probe TBA functionalized gold electrode into thrombin-binding buffer containing a certain concentration of target thrombin for 1 h at room temperature. Before characterizing with EIS, the electrode was rinsed separately with thrombin-binding buffer and water, dried with nitrogen gas. The thrombin-binding electrode was regenerated with 6 M guanidine hydrochloride for 5 min after each thrombin measurement. 3. Results and discussion 3.1. SPR measurements of assembly process SPR experiments were carried out to track the assembly process of various organic layers on the surface of gold substrate. Fig. 1 showed the superimposed angle-resolved SPR curves for the modification process. After formation of the CH monolayer film on the bare gold surface, the SPR angle shift () was about 150 millidegree. When the GA layer and the dendrimer layer were linked to the CH modified electrode, the  values were approximately 150 and 400 millidegree, respectively. The  values for the sequential deposition of the second GA layer and the TBA layer were about 200 and 300 millidegree, respectively. Because the mass sensitiv-

Fig. 1. Angle-resolved SPR curves for the layer-by-layer assembly procedure on a gold substrate. (a) Bare Au disk; (b) CH; (c) GA; (d) den + NaBH4 ; (e) GA; (f) TBA + NaBH4 . den, G4 -NH2 PAMAM dendrimer.

Fig. 2. (A) The Randle modified equivalent circuit model for the impedance spectra. (B) The Nyquist impedance spectra for the modification procedure on a gold electrode in the presence of 5 mM [Fe(CN)6 ]3−/4− in 0.01 M PBS (pH 7.4) containing 0.1 M KCl. (a) Bare Au electrode; (b) CH; (c) GA + den + NaBH4 + GA; (d) TBA + NaBH4 ; (e) after incubation in 20 nM thrombin. AC potential, 0.23 V; frequency range, 100 kHz to 0.1 Hz; voltage amplitude, 5 mV.

ity coefficient of the measured  value is 120 millidegree per 100 ng/cm2 . Therefore 300 millidegree corresponds to 250 ng/cm2 . The molecular weight of TBA is 4726.03 ␮g/␮mol. So the surface coverage of TBA molecules was calculated to be about 3.2 × 1013 molecules/cm2 , this value is higher than related literature [19]. It is likely that TBA is a 15-mer short-chain DNA molecule and this method improves the assembly capacity of probe molecule greatly. 3.2. EIS characterization of modified electrode As shown in Fig. 2A, Randle modified equivalent circuit model [3] was used to fit impedance data. The parameters in the equivalent circuit included the solution resistance (Rs ), the Warburg impedance (Zw ) resulting from the diffusion of the redox-probe, the double layer capacitance (Cd ) is substituted by the constant phase element (Q) when taking into account electrode roughness, and the charge-transfer resistance (Rct ). The latter two components (Q and Rct ) represent interfacial properties of the electrode, which are highly sensitive to the surface modification. Impedance responses of the functionalized gold electrode accompanying the stepwise modification process were depicted in Fig. 2B. The bare gold electrode exhibited a very small semicircle domain, corresponding to the Rct value of approximately 250 . The CH deposited electrode resulted in an almost straight line, which is characteristic of a diffusional limiting step due to the extremely fast charge-transfer process. This is attributed to the electrostatic attraction between the positive charges of CH molecules and the negatively charged [Fe(CN)6 ]3−/4− redox couple. When the GA layer, the dendrimer layer, and the second GA layer were further grafted onto the CH modified electrode, the Rct value increased to about 2 k, due to the compact structure of fabricated multilayer films and the neutral property of GA molecules. The assembly of the TBA monolayer film onto the electrode surface led to a further increase in the Rct value by approximately 0.8 k, because there are a lot of negative charges on the TBA sugar-phosphate back-

Z. Zhang et al. / Talanta 78 (2009) 1240–1245

Fig. 3. The impedance spectra of a non-dendrimer modified gold electrode. (a) Au/CH; (b) Au/CH/GA/TBA + NaBH4 ; (c) the modified electrode (b) after being immersed in 20 nM thrombin. Other conditions as in Fig. 2.

bone. After incubation in 20 nM thrombin, the formation of the TBA–thrombin complex on the electrode surface contributed to a significant increase in the Rct value to about 4.4 k. This is consistent with the fact that thrombin and its aptamer complex insulates the electron transfer between the electrode surface and the electrolyte solution [20,24,34]. Both the SPR and the EIS data indicated that various organic layers were successfully immobilized onto the gold substrate. To investigate whether the application of the dendrimer improved the immobilization capacity of probe molecules and magnified the response signals of the aptasensor, a non-dendrimer modified electrode was fabricated as control. As depicted in Fig. 3, when both GA and TBA molecules were immobilized on the CH functionalized gold electrode surface, the Rct value changed from about 20–110 . After the non-dendrimer modified electrode was immersed in 20 nM thrombin, the Rct value increased to about 260 . Fig. 4 showed the changes of Rct (Rct ) value for the probebinding capacity and the response signals of the two methods. It was observed that the Rct values of the immobilization capacity of TBA molecules and the response signals of 20 nM thrombin of the dendrimer modified electrode were approximately 8-fold and at least 10 times higher than those of the non-dendrimer modi-

1243

Fig. 5. EIS responses of the dendrimer modified electrode to different concentrations of thrombin, n = 3. Inset: the calibration curve of the aptasensor between the Rct values and the concentrations of thrombin in the range of 1–50 nM; n = 5.

fied electrode. This may be due to the fact that each molecular CH can only combine with one TBA molecule, while an individual dendrimer molecule can bind several TBA molecules due to 64 amino groups on its surface. Therefore, the dendrimer layer is not only able to remarkably improve the immobilization capacity of probe molecules but also to magnify the response signals greatly. 3.3. Aptasensor responses to target thrombin For thrombin detection, the Au/CH/den/TBA-modified electrode was incubated in the thrombin solution with different concentrations for 1 h, and washed with thrombin-binding buffer and water, respectively. Fig. 5 indicated that the Rct value increased with increasing the concentration of thrombin from 0.01 to 1000 nM. When the concentration of thrombin exceeded 500 nM, the change of the Rct value could be neglected as a result of the saturated adsorption of target molecules. As also seen in Fig. 5, the Rct value had a positive linear relationship with the concentration of thrombin in the range of 1–50 nM. The regression equation was Y = 4.029 + 0.015X (Y: the Rct value, k; X: the concentration of thrombin, nM), and the correlation coefficient (R) was 0.978. The detection limit (S/N = 3) for thrombin as low as 0.01 nM was obtained, which was more sensitive than most available impedimetric thrombin aptasensors [20,34]. It was reported that thrombin has two positively charged sites termed Exosite I (the fibrinogenrecognition exosite) and II (the heparin-binding exosite) on the opposite sides of the protein [24]. The 15-mer TBA used has specificity to the Exosite I, while a 29-mer TBA is easy to bind with the Exosite II [35]. So the affinity of the 15mer TBA to the fibrinogenrecognition exosite should be higher than to the heparin-binding exosite of thrombin. 3.4. Aptasensor regeneration

Fig. 4. The histogram of the change of Rct value (Rct = Rct(2) − Rct(1) ) for different methods of electrode preparation. The dark gray part corresponds to the Au/CH/GA/den/GA/TBA-modified electrode, while the light gray part represents the Au/CH/GA/TBA prepared electrode. (a) the Rct value of TBA immobilization; (b) the Rct value of the detection of 20 nM thrombin; n = 3.

After target detection, the sensing interface could be regenerated with acid, alkali or salt to remove adsorptive target molecules for the second measurement [17,29]. And the regeneration of sensing interface is still a challenge for most existing biosensors [19]. In this paper, a mild regeneration reagent guanidine hydrochloride was used to renew the sensing interface. As shown in Fig. 6, when the thrombin-binding electrode was immersed in 6 M guanidine hydrochloride for 5 min, the Rct value was almost recovered to the original value of newly prepared electrode, suggesting that the treated electrode was reusable. When this electrode was incubated

1244

Z. Zhang et al. / Talanta 78 (2009) 1240–1245

Fig. 6. Thrombin detection and regeneration of the sensing interface: (a) newly prepared electrode; (b) after being immersed in 20 nM thrombin; (c) regeneration with 6 M guanidine hydrochloride for 5 min; (d) detection of 20 nM thrombin for the second time; n = 3.

in 20 nM thrombin for the second time, the Rct value was approximately the same value of the first detection of 20 nM thrombin. After being challenged with 20 nM thrombin and regenerated for at least six cycles, the electrode still possessed 85% sensing activity. This may be because the stable covalent immobilization strategy and the insusceptible TBA molecules are used in this aptasensor. 3.5. Aptasensor selectivity and stability To investigate the specificity of the aptasensor, bovine serum albumin was adopted in the control experiments. BSA normally constitutes about 60% of plasma protein and is the most abundant protein in blood plasma [36]. As depicted in Fig. 7, when the Au/CH/den/TBA-functionalized electrode was exposed to 100 nM BSA for 1.5 h, the Rct value increased to about 280 . However, after the electrode was incubated in 20 nM thrombin for 1 h, the Rct value increased to approximately 1.6 k, which was 5–6-fold higher than that of 100 nM BSA. The results indicated that the aptasensor had considerable selectivity to its target thrombin. As for the reproducibility of the electrode modification procedure, using the same electrode, after five processes of electrode construction under identical experimental conditions, the measured Rct value of 20 nM thrombin was 4.35 ± 0.16 k and the relative standard deviation (R.S.D.) was 2.7%. The life stability of

Fig. 8. EIS responses of the aptasensor to thrombin at different concentrations (1–50 nM) in 1% bovine plasma; n = 3.

the aptasensor was also studied. A newly prepared Au/CH/den/TBAmodified electrode was dried in a nitrogen stream and covered with a plastic cap to protect the gold surface. The modified electrode was stored at 4 ◦ C and measured at intervals of one week. It remained about 85% of its original response after one month, indicating that the modified electrode was very stable. 3.6. Application of the aptasensor in biological assay For further study the potential application of the modified electrode, the determinations of thrombin in real samples were performed. As shown in Fig. 8, the EIS responses at the sensing interface to the different concentrations of thrombin in 1% bovine plasma exhibited the same phenomena to those in blank buffer but were less sensitive. The Rct value increased with increasing the concentrations of thrombin in the range of 1–50 nM. The results definitely illuminate the potential application of this aptasensor in real samples. 4. Conclusions We have described a reusable aptasensor based on the Au/CH/den/TBA-modified electrode. The dendrimer immobilization on the electrode surface not only improved the probe-binding capacity but also magnified the response signals of the aptasensor. The immobilization process of the aptasensor and recognition events changed the charge-transfer kinetics of [Fe(CN)6 ]3−/4− redox couple at the electrode interface, which had been characterized by EIS. The dendrimer-functionalized aptasensor exhibited high sensitivity, favorable specificity and stability in the detection of thrombin. Although the preparation process of this aptasensor is a bit complex, the sensitivity is satisfactory. In a word, this work establishes a methodology for the developing of biosensors with good analytical properties. Acknowledgements This work was supported by the National Natural Science Foundation of China (No. 90713022), the National Key Basic Research Development Project of China (No. 2007CB714500) and the Project of Chinese Academy of Sciences (No. KJCX2-YW-H11). References

Fig. 7. Control experiments of the aptasensor. (a) 100 nM BSA, 1.5 h; (b) 20 nM thrombin, 1 h; n = 3.

[1] A.D. Ellington, J.W. Szostak, Nature 346 (1990) 818. [2] C. Tuerk, L. Gold, Science 249 (1990) 505.

Z. Zhang et al. / Talanta 78 (2009) 1240–1245 [3] J.A. Lee, S. Hwang, J. Kwak, S.I. Park, S.S. Lee, K.C. Lee, Sens. Actuators B: Chem. 129 (2008) 372. [4] C.K. O’Sullivan, Anal. Bioanal. Chem. 372 (2002) 44. [5] V. Pavlov, B. Shlyahovsky, I. Willner, J. Am. Chem. Soc. 127 (2005) 6522. [6] X.H. Fang, Z.H. Cao, T. Beck, W.H. Tan, Anal. Chem. 73 (2001) 5752. [7] H. Wei, B.L. Li, J. Li, E.K. Wang, S.J. Dong, Chem. Commun. (2007) 3735. [8] Y. Li, H.J. Lee, R.M. Corn, Nucleic Acids Res. 34 (2006) 6416. [9] R. Polsky, R. Gill, L. Kaganovsky, I. Willner, Anal. Chem. 78 (2006) 2268. [10] Y. Xiao, B.D. Piorek, K.W. Plaxco, A.J. Heeger, J. Am. Chem. Soc. 127 (2005) 17990. [11] K. Ikebukuro, C. Kiyohara, K. Sode, Biosens. Bioelectron. 20 (2005) 2168. [12] F.L. Floch, H.A. Ho, M. Leclerc, Anal. Chem. 78 (2006) 4727. [13] E. Bakker, Y. Qin, Anal. Chem. 78 (2006) 3965. [14] M. Zayats, Y. Huang, R. Gill, C.A. Ma, I. Willner, J. Am. Chem. Soc. 128 (2006) 13666. [15] Y. Xiao, A.A. Lubin, A.J. Heeger, K.W. Plaxco, Angew. Chem. 117 (2005) 5592. [16] Y. Lu, X.C. Li, L.M. Zhang, L.Q. Mao, Anal. Chem. 80 (2008) 1883. [17] A.E. Radi, J.L.A. Sánchez, E. Baldrich, C.K. O’sullivan, J. Am. Chem. Soc. 128 (2006) 117. [18] G.S. Bang, S. Cho, B.G. Kim, Biosens. Bioelectron. 21 (2005) 863. [19] Y. Du, B.L. Li, H. Wei, Y.L. Wang, E.K. Wang, Anal. Chem. 80 (2008) 5110.

1245

[20] A.E. Radi, J.L.A. Sánchez, E. Baldrich, C.K. O’Sullivan, Anal. Chem. 77 (2005) 6320. [21] D.K. Xu, D.W. Xu, X.B. Yu, Z.H. Liu, W. He, Z.Q. Ma, Anal. Chem. 77 (2005) 5107. [22] M.C. Rodriguez, A.-N. Kawde, J. Wang, Chem. Commun. (2005) 4267. [23] C.Z. Li, Y.L. Liu, J.H.T. Luong, Anal. Chem. 77 (2005) 478. [24] X.X. Li, L.H. Shen, D.D. Zhang, C.X. Zhang, Biosens. Bioelectron. 23 (2008) 1624. [25] H.C. Yoon, H.S. Kim, Anal. Chem. 72 (2000) 922. [26] H.C. Yoon, M.Y. Hong, H.-S. Kim, Anal. Biochem. 282 (2000) 121. [27] Z.M. Liu, Y. Yang, H. Wang, Y.L. Liu, G.L. Shen, R.Q. Yu, Sens. Actuators B: Chem. 106 (2005) 394. [28] L. Shen, N.F. Hu, Biomacromolecules 6 (2005) 1475. [29] A.X. Li, F. Yang, Y. Ma, X.R. Yang, Biosens. Bioelectron. 22 (2007) 1716. [30] R. Kita, A. Takahashi, M. Kaibara, K. Kubota, Biomacromolecules 3 (2002) 1013. [31] J.A. Páramo, J. Rifón, J. Fernández, B. Cuesta, E. Rocha, Blood Coagul. Fibrin. 2 (1991) 227. [32] T. Hermann, D.J. Patel, Science 287 (2000) 820. [33] A. Joachimi, G. Mayer, J.S. Hartig, J. Am. Chem. Soc. 129 (2007) 3036. [34] H. Cai, T.M.H. Lee, I.M. Hsing, Sens. Actuator B: Chem. 114 (2006) 433. [35] Z.H. Cao, W.H. Tan, Chem. Eur. J. 11 (2005) 4502. [36] K.A. Mahmoud, S. Hrapovic, J.H.T. Luong, ACS Nano 2 (2008) 1051.