Determination of carcinoembryonic antigen using a novel ...

Science in China Series B: Chemistry © 2007

Science in China Press Springer-Verlag

Determination of carcinoembryonic antigen using a novel amperometric enzyme-electrode based on layer-by-layer assembly of gold nanoparticles and thionine YUAN Ruo†, ZHUO Ying, CHAI YaQin, ZHANG Ying & SUN AiLi Chongqing Key Laboratory of Analytical Chemistry, College of Chemistry and Chemical Engineering, Southwest University, Chongqing 400715, China

Electrochemical sensing of carcinoembryonic antigen (CEA) on a gold electrode modified by the sequential incorporation of the mediator, thionine (Thi), and gold nanoparticles (nano-Au), through covalent linkage and electrostatic interactions onto a self-assembled monolayer configuration is described in this paper. The enzyme, horseradish peroxidase (HRP), was employed to block the possible remaining active sites of the nano-Au monolayer, avoid the non-specific adsorption, instead of bovine serum albumin (BSA), and amplify the response of the antigen-antibody reaction. Electrochemical experiments indicated highly efficient electron transfer by the imbedded Thi mediator and adsorbed nano-Au. The HRP kept its activity after immobilization, and the studied electrode showed sensitive response to CEA and high stability during a long period of storage. The working range for the system was 2.5 to 80.0 ng/mL with a detection limit of 0.90 ng/mL. The model membrane system in this work is a potential biosensor for mimicking the other immunosensor and enzyme sensor. immunosensor, gold nanoparticles (nano-Au), thionine (Thi), carcinoembryonic antigen (CEA), horseradish peroxidase (HRP), layer-by-layer (LBL)

Carcinoembryonic antigen (CEA), an acidic glycoprotein, is a kind of important tumor marker[1], associated with colon cancer, lung cancer, ovarian carcinoma and breast cancer[2]. A variety of methods and strategies have been reported for the determination of CEA, such as radioimmunoassay (RIA), enzyme immunoassay (ELISA) and immunohistochemical test (IHC). However, these methods are relative to radiation hazards, tedious assay time, qualified personnel and sophisticated instrumentation. As a result, alternative approaches to de－ tect CEA in human serum are desirable[3 5]. The amperometric enzyme immunoassay manipulations coupled with the intrinsic selectivity and sensitivity of enzyme labeled on immunological components have gained considerable attention due to their simple design, high sensitivity and low cost in the last two decades[6]. www.scichina.com

www.springerlink.com

However, most of amperometric immunoassay techniques rely on the label of either antigen or antibody, which requires highly qualified personnel, tedious assay time, or sophisticated instrumentation. Therefore, an increasing number of new enzyme immunosensors had been reported[7,8]. Here we developed a novel strategy, which employed the enzyme to block possible remaining active sites of the gold nanoparticles (nano-Au) monolayer instead of bovine serum albumin (BSA) and amplify the response of the antigen-antibody reaction at Received August 31, 2005; accepted June 12, 2006 doi: 10.1007/s11426-007-0017-9 † Corresponding author (email: [email protected]) Supported by the National Natural Science Foundation of China (Grant No. 20675064), the Natural Science Foundation of Chongqing City (Grant Nos. CSTC2004BB4149 and 2005BB4100) and the High Technology Project Foundation of Southwest University (XSGX02), China

Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 97-104

the same time, to construct a new amperometric enzyme immunosensor with good sensitivity and high stability. It was reported that Nafion (Nf), as a polyanionic perfluorosulfonated polymer, was able to absorb some phenothiazine dye (e.g. Thionine, abbreviated as Thi) molecules via the ion exchange adsorption to form a redox Nf/Thi composite film[9,10]. The present investigation tried to improve previously reported work and develop a more simple and sensitive strategy for the determination of a model analyte (CEA). The immunosensor was prepared by forming a nano-Au monolayer onto Nf/Thi composite film modified on a gold electrode. Immobilization of Thi was attributed to the electrostatic force between positively charged Thi and the negatively charged sulfonic acid groups in Nf polymer, whereas immobilization of nano-Au particles was attributed to the chemisorption of the amine groups － of the Thi and the opposite-charged adsorption[11 13]. Then the Thi and nano-Au was alternately adsorbed onto the modified electrode surface, which constructed the {nano-Au/Thi}n multilayer films by layer-by-layer (LBL) assembly, based on the electrostatic interaction between positively charged Thi and negatively charged nanoAu[14,15]. Finally, carcinoembryonic antibody (anti-CEA) was adsorbed onto the surface of the nano-Au layer and horseradish peroxidase (HRP) was used to block possible remaining active sites of the nano-Au monolayer instead of BSA and amplify the response of the antigen-antibody reaction at the same time. Compared with the conventional enzyme immunosensors, this proposed immunosensor has several novelties, such as to get rid of the laborious enzyme labeling process and the complicated competition reaction. Furthermore, the new enzyme immunosensors are able to detect the physical changes during the immune complex formation, based on the signal-generating amplification of HRP, with more sensitive and versatile detection modes. Tests performed with this immunosensor showed good linearity, sensitivity, specificity and high stability when it was evaluated on several standard serum samples. The electrochemical behaviors and factors influencing the performance of the resulting immunosensors were investigated in detail.

1 Experimental 1.1 Apparatus and reagent Cyclic voltammetric (CV) measurements were carried 98

out with a CHI 610A electrochemistry workstation (Shanghai CH Instruments, China). A three-compartment electrochemical cell contained a platinum wire auxiliary electrode, a saturated calomel reference electrode (SCE) and the modified gold electrode (Φ = 2 mm) as working electrode. The size of the nanoparticles colloid was estimated from transmission electron microscopy (TEM) (H600, Hitachi Instrument, Japan). The pH measurements were made with a pH meter (MP 230, Mettler-Toledo, Switzerland) and a digital ion analyzer (Model PHS-3C, Dazhong Instruments, Shanghai, China). CEA and anti-CEA were purchased from Biocell Company (Zhengzhou China). BSA(96%－99%), Thi, HRP, gold chloride (HAuCl4) and sodium citrate were obtained from Sigma Chemical Co. (St. Louis, MO, USA). All other materials used were of the highest quality available and purchased from regular sources. Double distilled water was used throughout this study. Phosphate buffered solutions (PBS) (pH = 7.4) were prepared using 0.01 mol·L−1 Na2HPO4, 0.01 mol·L−1 KH2PO4 and 0.1 mol·L−1 KCl. The working buffer was composed of 0.1 mol·L−1 HAc-NaAc buffer solution containing 0.1 mol·L−1 KCl of pH 5.5. The prepared solutions were kept at 4℃ before use. Nano-Au was produced by reducing gold chloride tetrahydrate with citric acid at 100℃ for half an hour[16]. The mean size of the prepared Au colloids was about 16 nm, which was estimated from transmission electron microscopy (the graph not shown). Silver nanoparticles (nano-Ag) and Ag@Au core shell nanoparticles (Ag@Au) were produced according to the reported refs. [17,18], respectively. 1.2 Fabrication of the immunosensor Gold electrode (Φ = 2 mm) was polished with 1.0, 0.3 and 0.05 μm alumina to obtain a mirror like surface. After rinsing with distilled water the electrode was chemically cleaned by immersing it into freshly prepared 2:1 mixture of H2SO4 and H2O2 for 30 s. After a short rinse with distilled water and cleaning by bi-distilled water and ethanol in an ultrasonic bath, the electrode was replaced into an electrochemical cell with 0.5 mol·L−1 H2SO4 until background signal stabilization for electrochemical cleaning, and then allowed to dry at room temperature. The cleaned electrode was coated with 5 μL Nf ethanol solution (v/v, 2%) firstly, and after the coated Nf

YUAN Ruo et al. Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 97-104

film dried, then soaked in 5 mL Thi solution (3 mmol·L−1) for about 15 min. Following that, it was thoroughly rinsed with distilled water and then immersed in 0.5 mL prepared gold colloids (10 mg/mL) for 4 h to form a nano-Au monolayer. Then {nano-Au/Thi}n multilayer films were grown by alternately dipping the modified electrode into the positively charged Thi solution and negatively charged nano-Au aqueous solution for 20 min, respectively. Subsequently, the resulting electrode was immersed in an anti-CEA solution (0.415 μg/mL, pH 5.5) at 4℃ over night. Finally, the modified immunosensor was incubated in 0.25% pH 6.5 HRP solution about 2 h in order to block possible remaining active sites of the nano-Au monolayer and avoid the non-specific adsorption. The finished immunosensor was stored at 4℃ when not in use. The schematic illustration of the immunosensor was shown in Figure 1. 1.3 Experimental measurements The electrochemical characteristics of the modified electrode were characterized by cyclic voltammetry (CV). Electrochemical experiments were performed in a conventional electrochemical cell containing a threeelectrode arrangement. The CV scan was taken from −0.6－0.2 V (vs. SCE) at 50 mV/s in working buffer at 25±0.5℃.

2 Results and discussion 2.1 Characteristics of electrochemistry on electrode surface The cyclic voltammograms of differently modified elec-

Figure 1

trodes (in pH 5.5 HAc-NaAc buffer) were shown in Figure 2. No obvious electrochemical peak is found in HAc-NaAc at the bare gold electrode (Figure 2(1)). A decrease in the current can be observed (Figure 2(2)) after being coated with Nf, as the Nf film can hinder the transmission of electrons toward the electrode surface. Then it was found that the peak currents and the stability increased regularly with the increasing number of {nano-Au/Thi}n layers, suggesting the nano-Au and Thi were adsorbed onto the Nf film in a LBL fashion. Till positively charged Thi and negatively charged nano-Au were assembled for 4 times, alternately, the increase of the peak currents began to level off. Figure 2(3) shows a quasi-reversible CV wave with ipa/ipc ≈ 1 at 50 mV/s, indicating a well performance of the {(nano-Au)3/(Thi)4} multilayer film[14,15]. After the modified electrode was soaked in gold colloids for the fourth time, a further increase in anodic and cathodic peaks could be seen (Figure 2(4)). The reason is that nanometer-sized nano-Au plays an important role similar to a conducting wire or electron-conducting tunnel, which makes it easier for the electron transfer to take place. However, after anti-CEA was adsorbed on the nano-Au monolayer (Figure 2(5)), the peak currents decreased, which can be contributed to the immobilized anti-CEA monolayer hindering the transmission of electrons. Finally, the modified immunosensor was blocked with HRP; a further decrease of the peak currents may indicate HRP has been immobilized on the electrode surface, and it can hinder the transmission of electrons toward the electrode surface (Figure 2(6)). EIS is an effective method to probe the interfacial

Schematic illustration of the proposed immunosensor.


99

Figure 2 CVs of the different electrodes in HAc-NaAc buffer (pH 5.5). 1, Bare gold electrode; 2, Nf; 3, {(nano-Au)3/(Thi)4}/Nf; 4, {nano-Au/ Thi}4/Nf; 5, anti-CEA/{nano-Au/Thi}4/Nf; 6, HRP/anti-CEA/{nano-Au/ Thi}4/Nf modified gold electrode. The scan rate was 50 mV/s.

excellent electrochemical activity of Thi, as well as the good conductivity of nano-Au. When the anti-CEA and HRP film were obtained in turn, the interfacial resistance increased, which are consistent with the fact that the hydrophobic layer of the protein insulates the conductive support and perturbs the interfacial electron transfer (Figure 3(4) and (5)). The CVs of the modified immunosensor in HAcNaAc buffer solution (pH 5.5) at different scan rates were shown in Figure 4. It can be found the anodic and cathodic peak currents were directly proportional to the potential scan rates in the range of 50－800 mV/s, as shown in the inset of Figure 4, suggesting a surface confined process. Furthermore, the reversible surface waves in cyclic voltammetry indicate that the {nano-Au/Thi}n multilayer modified electrode is quite stable.

properties of modified electrode. Figure 3 illustrates the EIS of the different electrodes in PBS (pH 7.4) + 0.1 mol·L−1 KCl +5.0 mmol·L−1 Fe(CN)64−/3−. It can be found that the insert in Figure 3 is almost a straight line, which implies the characteristic of a diffuse limiting step of the electrochemical process on a bare gold electrode. After being coated with Nf, Figure 3(1) shows a very large semicircle domain, clarifying a high resistance of the electrode interface, which demonstrated that the insulating Nf membrane obstructed electron-transfer of the electrochemical probe. The resistance decreased after the {nano-Au/Thi}n multilayer film was obtained (Figure 3(2) and (3)), which can be attributed to the Figure 4 CVs of the modified electrode at different scan rates (from inner to outer): 50, 100, 150, 200, 300, 400, 500, 600, 700, 800 mV/s in 5 mL 0.1 mol/L HAc-NaAc buffer solution (pH 5.5) under room temperature. All potentials are given vs. SCE. The inset shows the dependence of redox peak currents on the potential sweep rates.

2.2 The amplification properties of HRP

Figure 3 EIS of the different electrodes. 1, Nf; 2, {(nano-Au)3/(Thi)4}/ Nf; 3, {nano-Au/Thi}4/Nf; 4, anti-CEA/{nano-Au/Thi}4; 5, HRP/antiCEA/{nano-Au/Thi}4/Nf modified gold electrode. The inset shows the EIS of the bare gold electrode.

100

The HRP is widely labeled to the antigen or antibody, because of its advantages, such as small size (Mw = 44000), high turnover rate of the enzymatic reaction and cheapness. The employment of HRP here played the following double roles: first, it substituted for BSA to block possible remaining active sites of the nano-Au monolayer and avoid the non-specific adsorption; second, it amplified the current response with the innate amplification properties of enzymes. Figure 5 depicts the CVs obtained with the immunosensor with and


without H2O2. One couple of symmetrical oxidation-reduction peaks can be observed in the absence of H2O2 (Figure 5(1)). While a dramatic enhancement of the cathodic peak current and a concomitant decrease of the anodic peak current, as well as the negative shift of Epc from −0.21 to −0.25 V, can be found in the presence of 0.31 mmol·L−1 H2O2 (Figure 5(2)). According to this fact, we come to the conclusion that the HRP attached to the immunosensor surface has retained high enzymatic catalytic activity. Curve 3 in Figure 5 shows the CV after the proposed immunosensor reacts with 50 ng/mL CEA standard solution. The decrease of the cathodic peak current can be attributed to the changes of the electrode surface circumstance after the immunoreaction. For HRP, the active center may be hindered by the immunocompound, which lowered the high turnover rate of the enzymatic reaction, as a result the changes of the current signals were amplified by the immobilized HRP. In a word, the using of HRP instead of BSA to block possible remaining active sites to amplify the response was simple and efficient.

Figure 5 CVs of the immunosensor in working buffer (pH 5.5) without (1) and with 0.31 mmol·L-1 H2O2 (2). (3) shows the CV of the proposed sensor after incubation with 50 ng/mL CEA. The scan rate is 50 mV/s.

2.3 Optimization of the assay conditions 2.3.1 Effect of the different nanoparticles. Nano-Au, nano-Ag and Ag@Au were prepared according to the reported refs. [16－19], respectively. The insert in the top left corner of Figure 6 was the UV-Vis spectra of these different nanoparticles. The results of the absorption peak were consistent with what was reported previously. Figure 6 shows the CVs of the different nanopar-

ticles modified electrodes, which were tested in the same experimental conditions. Figure 6(1)(including curve 1 of the insert in the top right corner) is the CV of Thi/Nf modified electrode in HAc-NaAc. Figure 6(2) presents the CV of nano-Au/Thi/Nf modified electrode, the increase of the peaks can be found, as the nano-Au is able to make it easier for the electrons transfer to take place[14,15]. Curve 3 in the insert in the top right corner of Figure 6 is the CV of nano-Ag/Thi/Nf modified electrode. A new couple of obvious waves observed can be attributed to the redox of the nano-Ag[20]. Although the nano-Ag was absorbed onto the Thi/Nf modified electrode surface, there is little increase of the redox waves of Thi. Figure 6(4) was the CV of Ag@Au/Thi/Nf modified electrode, on the one hand, it can be found a little increase of the redox waves of Thi; on the other hand, the characteristic waves of the redox of nano-Ag can also be observed. Compared with the different nanoparticles modified Thi/Nf electrode, the nano-Au shows the most increase of the redox waves of Thi.

Figure 6 The CVs of different nanoparticles modified electrodes. 1, Thi/Nf; 2, nano-Au/Thi/Nf; 3, nano-Ag/Thi/Nf modified electrode (the insert in the top right corner); 4, Ag@Au/Thi/Nf in 0.1 mol·L−1 HAc-NaAc (pH = 5.5) respectively. The scan rate is 50 mV/s. The insert in the top left corner shows the UV-vis spectra of different nanoparticles.

2.3.2 Amount of hydrogen peroxide. Investigation of the influence of the amount of H2O2 on the response is of great importance, because the amplifying performance of the immobilized HRP is H2O2 dependent. As shown in Figure 7, it can be seen that the current response increases when H2O2 amount is in the range from 0.18 to 0.31 mmol·L−1, then the current rises slowly to show a saturation. The insert illustrates the chronoam-


101

peromeric curve of the biosensor for successive addition of different volume of H2O2. This indicated that the HRP immobilized on the electrode surface retained a relatively high enzymatic activity. As a result, 0.31 mmol·L−1 H2O2 were chosen for the test.

Figure 8 CVs of the inmunosensor in HAc-NaAc buffer at various pHs. 1, 7.0; 2, 6.5; 3, 6.0; 4, 5.5; 5, 5.0; 6, 4.5; 7, 4.0 at 50 mV/s scan rate under 25℃. All potentials are given vs. SCE.

Figure 7 Calibration plots of the current response vs. concentration of H2O2. Inset shows typical current-time response curve of the biosensor upon successive additions of different amount of H2O2 (44.6 mol·L−1) into 5 mL pH 5.5 HAc-NaAc at applied potential of −0.25 V.

2.3.3 Influence of the pH value. The current response of the immunosensor will be affected by the pH of the assay solution. It was found that the current response increased and cathodic (Epc), and anodic (Epa) potential shifted in a positive direction with a decrease of the pH in the range from 7.0 to 4.0 (Figure 8). Although a lower pH was favorable for getting the better current response as the Thi was easily protonized in the solution containing enough H+, considering the response and the lifetime of the immunosensor, a pH 5.5 of the working buffer was selected as a compromise. 2.3.4 Influence of the incubation time and temperature. The effect of the immunochemical incubation time on response current was also investigated. The immunosensor was incubated with 50 ng/mL CEA standard solution for 1, 2, 3, 5, 8, 10 and 15 min, and then was tested in 5 mL pH 5.5 HAc-NaAc containing 0.31 mmol·L−1 H2O2. The results show that the response current was rapidly down with the duration of incubation time in the first 10 min and then tended to level off, which indicated an equilibration state was reached. Therefore, the incubation time of 10 min was adopted in the subsequent work. 102

The influence of the temperature on the current response was investigated at different temperatures ranging from 10 to 50℃, under the same experimental conditions. It was found that the current response decreased with the increasing of the temperature till 37℃, and the trend became to level off above 37℃, indicating the immunocomplexes might be destroyed over 37 ℃ . However, we employed 25±0.5℃ as the assay temperature for the room temperature, which is commended to be used in order to prolong the life-time of the biosensor and acquire better stability of H2O2. 2.4 Performance of the immunosensor 2.4.1 Calibration curves for CEA. The calibration plot for CEA detection with the proposed immunosensor under optimal experimental conditions is illustrated in Figure 9. As expected, the response signal decreased with the increase of CEA concentration. The linear range covered from 2.5－80.0 ng/mL with a regression equation of the form y = 0.0863x − 7.7663 and correlation coefficient of 0.9972 (Figure 9(2)) with a detection limit of 0.90 ng/mL. Curve 1 in Figure 9 was the calibration plot of the BSA blocked immunosensor. It is observed that HRP blocked electrode has led to wider dynamic measurement range and higher sensitivity compared with that of the BSA blocked electrode. 2.4.2 Regeneration and selectivity of the immunosensor. The regeneration of the proposed immunosensor was tested. Different immunosensors were regenerated


by 0.1 mol·L−1 NaOH, 0.1 mol·L−1 HCl and 4 mol·L−1 urea, respectively. It was found that the urea solution was the most proper regeneration solution with 6 times regenerations and measurements, whereas the NaOH and HCl solution made the immunosensors regenerate only 2 times. This result indicates that the proposed immunosensor could provide reproducible determination of CEA by the urea solution.

ng/mL CEA containing the above interferences; and (3) a solution only containing the above interferences. The results showed that the current response of the immunosensor, which was incubated with the (3) solution, changed little before and after the incubation, demonstrating the elimination of the fake positive test. Furthermore, it was found that the peak current responses, after being incubated with the (1) and (2) solutions by the same immunosensor, showed less than 0.251 μA error with the average of Ipc = −3.47 μA. The test was repeated 10 times, and an average error of 0.178 μA was obtained. 2.4.3 Stability of the immunosensor. The stability of the successive assays was studied by 100 cycles CV measurements in HAc-NaAc buffer after being incubated with 50 ng/mL CEA. The difference of the Ipc between the first cycles and the last cycles was only about 0.125 μA, and the average value of Ipc was −3.644 μA with a RSD of 2.77%. The long-time stability of the immunosensor was also investigated over 60-day period. And the immunosensor was stored at 4℃ and measured intermittently (every 3－5 d). It was found that about 86.7% response of the first value was obtained at the last test during the 16 times in which tests were performed over a total of 45 days. The good stability may be due to the fact that the amount of nano-Au was consistent and protein molecules were attached firmly onto the surface of nano-Au monolayer.

Figure 9 Calibration plots of the cathodic peak current response vs. concentration of CEA with the different modified immunosensors under optimal conditions. The anti-CEA/{nano-Au/Thi}4/Nf modified electrode blocked with BSA (1) and HRP (2). The amperometric detection was performed by CVs in 5 mL 0.1 mol/L HAc-NaAc buffer solution (pH 5.5) containing 0.31 mmol/L H2O2 (except 1) at 25℃ after the immunosensor incubated with different concentration CEA solution for 10 min. All potentials are given vs. SCE and the scan rate was 50 mV/s.

2.4.4 Preliminary application of immunosensor in human serum. The feasibility of the proposed method for the detection of CEA was evaluated by applying it to the analysis of human serum. The results with 40 human serum samples obtained from the immunosensor were compared with those from an established ELISAs technique. A good correlation was found between the results of the two methods (Table 1).

The effect of substances that might interfere with the response of the immunosensor was studied. The interference substances included α-1-fetoprotein (25 ng/mL), hepatitis B surface antigen (20 ng/mL), hepatitis B core antigen(20 ng/mL), hepatitis B e antigen (20 ng/mL), ascorbic acid (10 ng/mL), L-cysteine (50 μmol·L−1), L-lysine (100 μmol·L−1), L-glutamate (100 μmol·L−1) and BSA (40 mg/mL). There are 3 test solutions, which include the following: (1) the CEA standard solution containing 50 ng/mL CEA; (2) standard solution of 50 Table 1

3 Conclusions A new strategy was described for developing an am-

Part of the results of different methods applied in clinic serum test Methods

The proposed Immunosensor ELISAs

I (μA) results(ng/mL)

1 −7.729 0.43 0.5

6 −5.742 23.46 24.0

13 −6.87 10.39 10.0

Sample numbers 18 24 −7.687 −7.762 0.92 0.05 1.0 0

30 −3.451 50.0 49.0


37 −5.993 20.55 20.0

40 −7.632 1.56 1.5

103

perometric enzyme immunosensor for the determination of CEA with good sensitivity and high stability based on the {nano-Au/Thi}n multilayer film by the LBL intercalation of Thi and nano-Au, and the nano-Au monolayer adsorbing anti-CEA and blocked by HRP. This strategy has several attractive advantages, such as high stability of {nano-Au/Thi}n multilayer film, easily adsorptive immobilization of antibody on nano-Au monolayer, ef1 2

Adam J K, Odhav B, Bhoola K D. Immune responses in cancer.

ficient activity retention of loading immunoreactants, and the use of HRP instead of BSA to block possible remaining active sites and amplify the current responses, as well as the simplicity of use and cost effectiveness. Although the strategy has only been applied to anti-CEA and CEA as a model system, it could be readily extended toward the determination of other clinically or environmentally interested biospecies. 11

chemistry, quantum-size-related properties, and applications toward

Hammarström S. The carcinoembryonic antigen CEA family: Struc-

biology, catalysis, and nanotechnology. Chem Rev, 2004, 104:

tures, suggested functions and expression in normal and malignant tissues. Semin Cancer Biol, 1999, 9: 67－81 3

293－346 12

Dai Z, Yan F, Yu H, et al. Novel amperometric immunosensor for Immunol Methods, 2004, 287: 13－20

B-Chem, 2005, 48(1): 49－57 13

Du D, Yan F, Liu S L, et al. Electrochemical immunosensor for car-

2003, 283: 67－75 5

Verhaar M J, Damen C A, Zonnenberg B A, et al. In vitro upregulation of carcinoembryonic antigen expression by combinations of cy-

nanoparticles. Chin Sci Bull, 2001, 46(13): 1074－1076 14

Maeda M. New label enzymes for bioluminescent enzyme immunoassay. J Pharm Biomed Anal, 2003, 30(6): 1725－1734

7

Calvo E J, Danilowicz C, Lagier C M, et al. Characterization of self-assembled redox polymer and antibody molecules on thiolated

15

Zacco E, Pividori M I, Llopis X, et al. Renewable Protein A modified graphite-epoxy composite for electrochemical immunosensing. J

Chem Commun, 2002, 1706－1707 16

nanocrystals. J Am Chem Soc, 1999, 121: 9932－9939 18

John S A, Ramaraj R. Microenvironment effects on the electrochemical and photoelectrochemical properties of thionine loaded Nafion® films. J Electroanal Chem, 2004, 561: 119－126

104

Shankar S S, Rai A, Ahmad A, et al. Rapid synthesis of Au, Ag, and bimetallic Au core-Ag shell nanoparticles using Neem (Azadirachta indica) leaf broth. J Colloid Interf Sci, 2004, 275(2): 496－502

19

Ji X H, Wang L Y, Yuan H, et al. Formation mechanism of core-shell Au/Ag nanoparticles. Acta Chim Sinica, 2003, 61(10): 1556－1560

film. Talanta, 2000, 51: 565－572 10

Michaels A M, Nirmal M, Brus L E. Surface enhanced Raman spectroscopy of individual rhodamine 6G molecules on large Ag

Wang B Q, Dong S J. Sol-gel-derived amperometric biosensor for hydrogen peroxide based on methylene green incorporated in Nafion

Frens G. Controlled nucleation for the regulation of the particle size in monodisperse gold suspensions. Nature-Phys Sci, 1973, 241: 20－22

17

Immunol Methods, 2004, 286: 35－46 9

Cheng W L, Jiang J G, Dong S J, et al. Alternate assemblies of thionine and Au-nanoparticles on an amino functionalized surface.

gold electrodes. Biosens Bioelectron, 2004, 19: 1219－1228 8

Ding Y H, Zhang X M, Liu X X, et al. Adsorption characteristics of thionine on gold nanoparticles. Langmuir, 2006, 22: 2292－2298

tokines. Cancer Lett, 1999, 139: 67－73 6

Zhao H Q, Ling L, Tang J A, et al. Enhancement of the immobilization and discrimination of DNA probe on a biosensor using gold

bohydrate antigen 19-9 using an electrochemical immunosensor and antigen immobilization in titania sol-gel matrix. J Immunol Methods,

Yuan R, Tang D P, Chai Y Q, et al. Highly sensitive potentiometric immunosensor for hepatitis B surface antigen diagnosis. Sci China Ser

rapid separation-free immunoassay of carcinoembryonic antigen. J 4

Daniel M C, Astruc D. Gold nanoparticles: Assembly, supramolecular

Pharmacol Therapeut, 2003, 99: 113－132

20

Zeng B, Ding X J, Pan D C, et al. Accumulation and stripping behavior of silver ions at DL-dithiothreitol self-assembled monolayer modified gold electrodes. Talanta, 2003, 59: 501－507
