- Email: [email protected]
Electrochemical sensor for immunoassay of carcinoembryonic antigen based on thionine monolayer modified gold electrode
Cancer Detection and Prevention 29 (2005) 233–240 www.elsevier.com/locate/cdp
Electrochemical sensor for immunoassay of carcinoembryonic antigen based on thionine monolayer modified gold electrode Zong Dai PhDa, Jin Chen a, Feng Yan b, Huangxian Ju MD, PhDa,* a
Department of Chemistry, Institute of Chemical Biology, Key Laboratory of Analytical Chemistry for Life Science, Nanjing University, Nanjing 210093, China b Department of Medical Examination, Cancer Hospital and Institute of Jiangsu Province, Nanjing 210009, China Accepted 7 December 2004
Abstract A sensor based on thionine monolayer modified gold electrode for determination of carcinoembryonic antigen (CEA) in human serum is proposed. The sensor is prepared by covalently binding thionine to a cysteamine self-assembled monolayer with p-phthaloyl chloride as a linkage, which gives a surface coverage of 8.97 3.28 l0 12 mol/cm2 for thionine. The electrochemistry of the immobilized thionine displays a surface-controlled electrode process with an average electron transfer rate constant of 1.47 0.84 s 1. Based on an electrochemical enzyme-linked immunoassay by using the immobilized thionine as an electron transfer mediator between the electrode and the horseradish peroxidase (HRP) labeled anti-CEA antibody, a calibration curve with two linear ranges from 0.6 to 17 and 17 to 200 ng/mL and a detection limit of 0.2 ng/mL for CEA determination is obtained in pH 4.2 PBS containing 2.0 mmol/L H2O2 and 0.5 mol/L NaCl. The sensor shows a good accuracy. The precision and reproducibility are acceptable with the intra-assay CVof 4.9% and 5.9% at 10 and 100 ng/mL CEA concentrations, respectively, and the inter-assays CV of 7.8% at 100 ng/mL CEA. The response of thionine modified electrode shows only 1.6% decrease after 100 replicate measurements and the storage stability is acceptable in a pH 7.0 PBS at 4 8C for 1 week. The method avoids the addition of electron transfer mediator to the solution, thus is much simpler. The proposed method would be valuable for the diagnosis and monitoring of carcinoma and its metastasis. # 2005 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. Keywords: Sensors; CEA; Thionine; Gold electrode; Self-assembled monolayer; HRP; ELISA; Immunoassay
1. Introduction Carcinoembryonic antigen (CEA), an acidic glycoprotein with a molecular weight of about 200 kDa, is one of the most widely used tumor makers. It was first described in 1965 as a marker of colon cancer. Up to now, it has been found to exist in the serum of most tumor patients. The CEA level in serum is related to lung cancer [1–3], ovarian carcinoma [4,5], colon cancer [6–8], breast cancer [9–11], cystadenocarcinoma [5] and others [12–15]. Thus, the determination of CEA level is very helpful to clinical tumor diagnoses. On the other hand, the CEA level in serum is also related to the state of tumor, so it can be directly used for evaluating curative effect, judging recrudescence or metastasis as a marker. * Corresponding author. Tel.: +86 25 83593593; fax: +86 25 83593593. E-mail address: [email protected] (H. Ju).
The techniques usually used for quantitative determination of tumor markers are immunological methods, which have become the predominant analytical techniques in the fields of clinical diagnoses and biochemical studies [16]. Immunoassay methods possess extremely high selectivity and sensitivity [17]. The selectivity results from the capacity of antibody recognizing the corresponding antigen. A number of immunoassay methods for the determination of CEA level in serum have been reported [18–28]. These methods with commercially available assay kits include radioimmunoassay [18], chemiluminescence immunoassay [19,20], enzyme-linked immunosorbent assay [21,22], fluoroimmunoassay [23] and liposome immunoassay [24]. Several years ago, radioimmunoassay was the most commonly used method for antigen detection. However, radioactive labels are harmful to the operators. Thus, test kits of enzyme-linked immunosorbent assay and chemilumines-
0361-090X/$30.00 # 2005 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.cdp.2004.12.003
234
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
cence immunoassay have currently been used extensively for CEA determination in clinical diagnoses. Although these traditional methods have detection limits of 0.1–0.5 ng/mL for CEA [25,26] and are suitable enough for the determination of CEA in sera, they are usually time and sample consuming because of several incubation and washing steps needed in detection procedure. Moreover, implement point-of-care (POC) testing is difficult with conventional immunoassay because rather large equipment is necessary. In order to overcome these drawbacks, some works on microchip-based immunoassays have been reported [27–31]. These systems were based on the separation of the free form of antigen and the complex of antigen and antibody. They were not suitable for every analyte contained in biological samples due to the precipitation of the antigen–antibody complex [17,28]. Sato et al. structured a bead-bed immunoassay system on a microchip for the determination of CEA [17]. Anti-CEA molecules were firstly precoated on polystyrene beads, which were introduced into a microchannel. The serum sample containing CEA, the primary antibody and the secondary antibody conjugated with colloid gold were then reacted successively. The resulting complex was finally detected using a laser-induced thermal lens microscope (TLM) and gave a detection limit of 0.03 ng/mL that was lower than those of the conventional ELISA. But the TLM is obviously more expensive than the equipment required usually for ELISA. In this paper, we prepared a thionine monolayer modified gold electrode as a sensor to develop a novel fast and cheap thionine–H2O2–horseradish peroxidase (HRP) solid-phase immunoassay for CEA determination. The immobilized thionine was used as an electron transfer mediator between the electrode and horseradish peroxidase labeled anti-CEA antibody. This method avoids the addition of electron transfer mediator to the solution; thus, it needs less incubation and washing steps and is simpler than traditional ELISA methods. Moreover, the equipment required for this method is very cheap. In comparison with other immunosensors prepared with graphite [32–36], the presented sensor is steady, avoids being contaminated by the addition of mediator and can be repeatedly used without further pretreatment. At the same time, it is easily miniaturized and does not contaminate the sample. This method is more sensitive than traditional RIA and almost has the same detection limit as that of chemiluminescence immunoassay and fluoroimmunoassay, and thus, has a good application perspective for continuous determination of CEA.
different concentrations from 0 to 500 ng/mL, and a solution of horseradish peroxidase conjugated monoclonal anti-CEA antibody. CNBr-activated Sepharose 6 MB was purchased from Amersham Pharmacia Biotech. Thionine, cysteamine, p-phthaloyl chloride and H2O2 with analytical grade were from the Shanghai Biochemical Reagent Company (China). All other reagents were of analytical grade. Double distilled water was used for all experiments. The eluent buffer was a solution of 0.1 mol/L pH 4.2 PBS containing 0.5 mol/L NaCl. All reagents were brought to room temperature (ca. 20 8C) before use. 2.2. Preparation of thionine modified gold electrode Gold electrodes, made by sealing gold wire (0.5 mm diameter) in a glass tube of 3 mm exterior diameter, were polished to a mirror-like surface with 0.3 and 0.05 mm alumina slurry on microcloth pads (Buehler Co., USA), followed by being rinsed with water and ethanol and sonicated in double distilled water for 2 min. After cleaning, the electrode was immersed in mercury for about 30 min for amalgamation, the amalgamated gold electrode was immersed in 1.0 mol/L HNO3 till the amalgam was dissolved and the gold surface was exposed. The roughness factor of the electrode surface was 6.59, which was obtained from the geometric area of 1.05 l0 2 cm2 and the real area of 6.90 10 2 cm2. The real area was obtained by measuring the cathodic peak area of cyclic voltammogram formed during the re-reduction of superficial gold oxide in 1.0 mol/L H2SO4 [37]. Modification of the gold electrode was performed by soaking a previously pretreated electrode in a deoxygenated 20 mmol/L cysteamine solution for more than 20 h. The resulting monolayer-modified electrode was thoroughly rinsed with water to remove physically adsorbed cysteamine and dried in a jet of nitrogen. The cysteamine-modified electrode was then immersed into a pphthaloyl chloride saturated solution of toluene for 5 h and rinsed with toluene. Finally, the modified electrode was transferred into the acetonitrile solution of 2 mmol/L thionine for 10 h and rinsed with acetonitrile thoroughly to yield a thionine monolayer modified electrode, which was washed with double distilled water and soaked in pH 7.0 PBS at 4 8C for prior to use [38]. Scheme 1 shows the procedure for the preparation of thionine modified gold electrode. 2.3. Apparatus
2. Materials and methods 2.1. Reagents CEA-ELISA and IRMA kits, including 96 assay tubes precoated with murine monoclonal anti-CEA, were purchased from Diagnostic Products (DPC, USA). The ELISA kits consisted of a series of CEA standard solutions with
Electrochemical measurements were performed with a three-electrode system. A platinum wire with 1 mm diameter was used as auxiliary electrode, a ‘‘homemade’’ Ag/AgCl electrode with 2 mm exterior diameter acted as reference and a thionine monolayer modified gold electrode performed as working electrode. The three electrodes were connected with a BAS-100B electrochemical analyzer
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
235
Scheme 1. The procedure of thionine covalently binding to a cysteamine self-assembled monolayer with p-phthaloyl chloride for preparation of thionine modified gold electrode.
(Bioanalytical Systems Inc., USA) and directly fitted into the microtiter plate wells. IRMA procedure was carried out with a FMJ-182 Immunoradiometric Gamma Counter (China) according to the instruction and assay procedure in the operator’s manual. The chemiluminescent detections were carried out with an IFFM-D Luminescence Analyzer (Remex Electronic Instrument Limited Co., Xi’an, China). Polytetrafluoroethylene (PTFE) tubing (0.25 mm i.d., 0.8 mm o.d.) was used to connect all components in the flow system. 2.4. Procedure The analytical procedure for immunoassay is schematically depicted in Scheme 2. The principle of the sandwich assay was based on one-step immunoreactions. Each microtiter well precoated with primary anti-CEA antibody had 50 mL of CEA sample solution and 100 mL of HRP labeled anti-CEA antibody added. The incubation mixture was thoroughly mixed for 10 s and incubated at 23 8C for 60 min. After the HRP labeled anti-CEA antibody molecules were adsorbed on the microtiter plate wells through immune complex formation, the incubation mixture was
removed by emptying the plate contents into a waste container, the microtiter wells were rinsed and emptied five times with double distilled water. The residual water droplets were removed by striking the plate sharply onto absorbent paper. A 200 mL deoxygenated eluent buffer was then added into the wells and left at 25 8C for 10 min to elute the HRP labeled anti-CEA antibody from the wells. In order to increase the elution efficiency, the eluent buffer contained 0.5 mol/L NaCl. After dispensing 2 mL 0.2 mol/L H2O2 into the well, the electrochemical detection was performed by inserting the three-electrode system into the microtiter well. Differential pulse voltammetric (DPV) measurements were performed from 150 to 250 mV with a pulse amplitude of
Scheme 2. Schematic diagram of FIA system: (P) peristaltic pump, (V) eight-way valve, (C) affinity column, (PMT) photo multiplier, (D) detector and (PC) computer.
236
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
50 mV, a pulse width of 60 ms, a scan rate of 20 mV/s, a pulse period of 200 ms and a sampling width of 4 mV.
3. Results 3.1. Electrochemical behavior of thionine modified gold electrode Fig. 1 shows the cyclic voltammograms of different electrodes in 0.1 mol/L pH 4.2 PBS. No response was observed with bare, amalgamated and cysteamine modified gold electrodes in the working potential window (curves a– c). After binding thionine to the cysteamine self-assembled monolayer with p-phthaloyl chloride, a pair of stable and well-defined redox peaks was observed. The anodic and cathodic peak potentials of thionine monolayer were 32 and +28 mV at 50 mV/s, respectively (curve d). Both the anodic and cathodic peak currents increased with the increasing scan rate and were proportional to the scan rate in the range from 5 to 50 mV/s (Fig. 2), displaying a surfacecontrolled electrode process. The surface coverage of thionine was calculated to be 8.97 3.28 l0 12 mol/ cm2. When the scan rate was higher than 50 mV/s the electrode reaction showed a diffusion-controlled process due to the kinetic limitation. 3.2. Cyclic voltammetric response of enzymatic reaction product It is known that HRP can catalyze the oxidation reaction of thionine by H2O2. In the absence of CEA antigen in the incubation solution, the thionine monolayer modified gold electrode showed a pair of symmetrical redox peaks in pH 4.2 PBS containing 2.0 mmol/L H2O2 (Fig. 3a). When the incubation solution contained 120 ng/mL CEA the oxidation
Fig. 1. Cyclic voltammograms of different electrodes in pH 4.2 PBS at 45 mV/s: (a) unpretreated bare, (b) amalgamatic bare, (c) cysteamine modified and (d) thionine modified gold electrodes.
Fig. 2. Cyclic voltammograms of thionine modified electrode in pH 4.2 PBS at 5, 10, 15, 20, 25, 30, 35, 40, 45 and 50 mV/s (from lowest to highest peak currents). Inset: plots of peak current vs. scan rate.
peak current decreased and the reduction peak current increased greatly and the reduction peak potential shifted slightly to more negative value (Fig. 3b), showing an obvious enzymatic reaction process. Because DPV method can eliminate the background current of voltammetric response, the following experiments were done by DPV method for getting more sensitive results. When thionine modified electrode was scanned in pH 4.2 PBS using DPV, a reduction peak of the immobilized thionine was found at 46 mV (curve a, inset in Fig. 3). The reduction peak current slightly increased upon addition of
Fig. 3. Cyclic voltammetric response of enzymatic reaction on thionine modified electrode in pH 4.2 PBS containing 2.0 mmol/L H2O2 for the incubation processes using (a) 0 ng/mL CEA and (b) 120 ng/mL CEA. Inset: differential pulse voltammograms of the thionine modified electrode in (a) pH 4.2 PBS, (b) (a) + 2.0 mmol/L H2O2 and (c) (b) + 40 ng/mL HRP.
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
237
2.0 mmol/L H2O2 (curve b, inset in Fig. 3). After 40 ng/mL HRP was added into above solution, the reduction peak current increased greatly with a slight negative shift of the reduction peak potential (curve c, inset in Fig. 3), indicating an obvious enzymatic reaction process. 3.3. Optimization of CEA determination conditions The important factors for the elution of HRP labeled antiCEA antibody from the primary anti-CEA antibody/CEA immunocomplex included pH value of eluent buffer and elution time. The effect of the pH of the eluent buffer on the DPV peak current was studied in the pH range from 2.5 to 6.9 at a 500 ng/mL CEA concentration. From the results shown in Fig. 4A, the maximum response occurred at pH 4.2. The effect of elution time on the elution efficiency was studied using flow injection chemiluminescence (CL) immunoassay (Scheme 2). CEA molecules were immobilized on CNBr-activated Sepharose 6 MB and introduced into an affinity column according to the manufacturer’s instructions. After the affinity, column was incubated with HRP labeled anti-CEA antibody and the uncomplexed HRP labeled anti-CEA antibody was washed out, the eluent buffer was pumped through the column for 25 s and then the luminol and H2O2 were added each 60 s. The elution process was evaluated by the decrease of CL intensity of luminol. With increasing elution time, the CL intensity of luminol decreased greatly (Fig. 5), indicating a decrease in the amount of HRP in the affinity column. In Fig. 5 the elution efficiency was calculated by 100 DI/(Ii I0), in which Ii was the initial luminol CL intensity without being eluted, I0 was the blank value and DI was the decrease of the CL intensity of luminol relative to Ii(It Ii). After elution for about 400 s, the CL intensity was as low as that of the blank, and the elution efficiency tended to be 100%, indicating all the HRP labeled anti-CEA antibodies completed with CEA in the affinity column had been eluted. Thus, the optimal elution time of 10 min in the well was selected for CEA determination. The effect of H2O2 concentration on the enzymatic reaction was examined using DPV. As shown in Fig. 4B, with increasing H2O2 concentration, the DPV response displayed a Michaelis–Menten’s shaped curve. At H2O2 concentrations less than 2.0 mmol/L, the peak current
Fig. 5. Effect of elution time on luminolCL intensity (a) for a flow elution process and (b) an affinity column without incubation with HRP labeled CEA antibody. Inset: effect of elution time on the elution efficiency.
increased linearly with increasing H2O2 concentration. When H2O2 concentration was more than 2.0 mmol/L, the reduction peak current trended to a constant value. Therefore, the optimal H2O2 concentration was 2.0 mmol/L. 3.4. Calibration for CEA determination With thionine monolayer modified electrode as a sensor, the calibration graph for the determination of CEA was obtained under optimal conditions (Fig. 6). Following the standard immunometric sandwich and the enzymatic reaction procedures, the DPV peak current increased with increasing CEA concentration in two linear ranges from 0.6 to 17 and 17 to 200 ng/mL, the linear slopes were 9.5 and 0.55 nA/(ng mL), and the correlation coefficients were 0.9995 and 0.9880, respectively. The detection limit was 0.2 ng/mL at 3s. 3.5. Precision, reproducibility and stability The typical intra-assay and inter-assay precision for CEA assay were also estimated. The variation coefficients (CV) of
Fig. 4. Effects of pH value of eluent buffer (A) and H2O2 concentration (B) on DPV peak current for CEA determination.
238
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
Fig. 6. Plot of peak current vs. CEA concentration under optimal conditions. Inset: differential pulse voltammograms of the thionine modified electrode in pH 4.2 PBS containing 2.0 mmol/L H2O2 for 0, 0.6, 1, 5, 10, 12, 15, 30, 60, 100, 120, 200 and 300 ng/mL CEA (from lowest to highest peak currents).
intra-assay, obtained by three repeated determinations with the same modified electrode, were 4.9% and 5.9% at CEA concentrations of 10 and 100 ng/mL, respectively, and the CV value for inter-assay variation was 7.8% at 100 ng/mL. Moreover, when the peak currents of the thionine modified gold electrode in pH 4.2 PBS were determined 100 times, the response showed only a 1.6% decrease, indicating a good reproducibility of this sensor. The stability of this sensor was acceptable in pH 7.0 PBS at 4 8C for 1 week. 3.6. Accuracy and clinic application The CEA levels in two sera were quantified by using standard addition method. The CEA concentrations determined with this method and the IRMA method are shown in Table 1. These results by both methods were in good agreement. Thus, the proposed method could be satisfactorily applied to the clinical determination of CEA in human serum.
4. Discussion The amalgamated bare gold electrode shows a charging current much larger than that of bare gold electrode (curves a
and b in Fig. 1), indicating a rougher surface is obtained by amalgamation. The formation of a cysteamine monolayer results in a smaller charging current (curve c in Fig. 1) than that of amalgamated bare gold electrode. The surface is clean and smooth. This is also an effective way to provide the gold electrode with abundant amino groups for covalent immobilization of thionine to form thionine modified gold electrode. From the peak-to-peak separation of the thionine modified gold electrode at different scan rates, an average electron transfer rate constant of 1.47 0.84 s 1 was obtained for the surface-controlled redox process. When the incubation solution contains CEA, the HRP labeled anti-CEA antibody molecules are eluted into the wells after a deoxygenated eluent buffer is added into the wells. The presence of HRP results in an enzymatic reaction process upon addition of H2O2. The oxidation reaction of thionine by H2O2 under the enzymatic catalysis of HRP increases the concentration of thionine in its oxidized state at the electrode surface, thus, increasing greatly the reduction peak current of thionine modified gold electrode. The immunoassay procedure and the enzymatic reaction mechanism can be expressed as shown in Schemes 3 and 4 [39], respectively. Both the elution efficiency and the enzyme activity are related to the pH value of eluent buffer. High elution efficiency is usually obtained at low pH. However, most enzymes display their activity in a limited pH range. The maximum activity of HRP is in the range from pH 6.5 to 7.0. With increasing solution acidity its activity decreases greatly. The integrated action of solution acidity results in that the maximum response occurs at pH 4.2, which is selected for elution of HRP labeled anti-CEA antibody and immunoassay of CEA. The optimal pH is higher than those reported in flow immunoassay [40–42]. The elution efficiency also depends on the elution time, which affects the amount of HRP introduced into the detection solution. The decrease of the CL intensity of luminol is due to the reducing amount of HRP in the affinity column, which is a mark of elution efficiency. From the analysis shown in Fig. 5 the optimal elution time in the well is 10 min. With the optimal elution time and pH value of eluent buffer the amperometric immunoassay of CEA in serum samples can be performed at the proposed thionine self-assembled gold electrode at a H2O2 concentration of 2.0 mmol/L. The detection limit of CEA is 0.2 ng/mL, which is more sensitive than traditional RIA
Table 1 Correlation between two sera quantified by this method and IRMA method Serum samples
This method (ng/mL)
IRMA (ng/mL)
Relative deviation (%)
1
5.3 5.2 5.7
5.1
3.9 2.0 21
2
34.1 32.6 32.9
35.8
4.7 8.9 8.1
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
239
Scheme 3. The immunoassay procedure.
Scheme 4. The mechanism of enzymatic and electrode reactions.
and is almost the same as that of chemiluminescence immunoassay and fluoroimmunoassay. Considering its acceptable intra-assay and inter-assay precision and the good reproducibility and storage stability of the sensor, the presented method could be used in CEA determination for clinical diagnosis.
addition of electron transfer mediator to the solution, thus is much simpler and cheaper than traditional assays. Furthermore, both electrochemical analyzer and the sensor can be easily miniaturized; thus, this method is particularly well suited for point-of-care testing and application in emergency rooms, ambulances, bedside, or doctor’s office. At the same time, it can be readily extended toward detection of other clinically important antigens.
5. Conclusions A sensor prepared by immobilizing thionine monolayer on gold electrode can be used for determination of serum CEA levels. Under the optimal conditions, a detection limit of 0.2 ng/mL CEA and two linear ranges from 0.6 to 17 and 17 to 200 ng/mL for CEA determination were obtained. The sensor shows good accuracy and acceptable precision, reproducibility and storage stability. It can be repeatedly used without further pretreatment and seems to be rather suitable for clinic immunoassay. The method avoids the
Acknowledgments This project was supported by the Distinguished Young Scholar Fund to HX Ju (20325518), the National Natural Science Foundation of China (No. 20275017), the Science Foundation of Jiangsu (No. BS2001063), the Specialized Research Fund for Excellent Young Teachers from Ministry of Education of China, and the Key Research Project of Cancer Hospital and Institute of Jiangsu Province.
240
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
References [1] Moro D, Villemain D, Vuillez JP, et al. CEA, CYFEA21-1 and SCC in non-small cell lung cancer. Lung Cancer 1995;13:169–76. [2] Iwazawa T, Kanoh T, Matsui S, et al. Diagnosis of lung cancer metastasis with CEA extracted from the dissected regional lymph nodes. Lung Cancer 2000;29:254. [3] Nisman B, Lafair J, Peretz T, et al. Cyfra 21-1 and TPS-New markers in lung cancer. Eur J Cancer 1995;31:s263. [4] Gould DA, Moscoso GJ, Young MPA, et al. Human first trimester fetal ovaries express oncofetal antigens and steroid receptors. J Soc Gynecol Investig 2000;7:131–8. [5] Kazuya K, Yoshihiro K, Tsunekazu K, et al. Preoperative determination of several serum tumor markers in patients with primary epithelial ovarian carcinoma. Gynecol Obstet Invest 1999;47:52–7. [6] Eppler E, Horig H, Kaufman HL, et al. Carcinoembryonic antigen (CEA) presentation and specific T cell-priming by human dendritic cells transfected with CEA-mRNA. Eur J Cancer 2002;38:184–93. [7] Ford CHJ, Osborne PA, Rego BG, et al. Bispecific antibody targeting of doxorubicin to carcinoembryonic antigen-expressing colon cancer cell lines in vitro and in vivo. Int J Cancer 2001;92: 851–5. [8] Vogel I, Francksen H, Soeth E, et al. The carcinoembryonic antigen and its prognostic impact on immunocytologically detected intraperitoneal colorectal cancer cells. Am J Surg 2001;181:188–93. [9] Kramer S, Jager W, Lang N. Early treatment of metastatic breast cancer patients after increase of CEA and CA 15-3 serum levels. Eur J Cancer A 1998;34:s42. [10] Cameiro C, Costa L, Melo M, et al. Serum tumor markers in metastatic breast cancer comparative study between CEA, CA 15-3 and MCA. Eur J Cancer A 1998;34:s43. [11] Bremer K, Nicus S, Bremer G. CEA, CA 15-3 and MCA: comparative clinical relevance in breast cancer. Eur J Cancer 1995;31:s262. [12] Tachibana M, Takemoto Y, Nakashima Y, et al. Serum carcinoembryonic antigen as a prognostic factor in resectable gastric cancer. J Am College Surg 1998;187:64–8. [13] Kau SY, Shyr YM, Su CH, et al. Diagnostic and prognostic values of CA 19-9 and CEA in periampullary cancers. J Am College Surg 1999;188:415–20. [14] Lechner P, Lind P, Goldenberg DM. Can postoperative surveillance with serial CEA immunoscintigraphy detect respectable rectal cancer recurrence and potentially improve tumor-tree survival? J Am College Surg 2000;191:511–8. [15] Takamori S, Yoshimoto J, Kojima K, et al. Prognostic implication and intratumoral heterogeneity of p53, Ki-67, CEA and CA 19-9 expressions in cholangiocellular carcinoma. Hepatol Res 1997;8:81–94. [16] Karsten S, Andreas K, Dieter T, et al. Disposable optical sensor chip for medical diagnostics: new ways in bioanalysis. Anal Chem 1999;71:5430–5. [17] Sato K, Tokeshi M, Kimura H, et al. Determination of carcinoembryonic antigen in human sera by integrated bead-bed immunoassay in a microchip for cancer diagnosis. Anal Chem 2001;73:1213–8. [18] Cioffi M, Vietri MT, Gazzerro P, et al. Serum anti-p53 antibodies in lung cancer: comparison with established tumor markers. Lung Cancer 2001;33:163–9. [19] Forest JC, Masse J, Lane AA. Evaluation of the analytical performance of the boehringer mannheim elecsys. 2010 Immunoanalyzer. Clin Biochem 1998;31:81–8. [20] Rainer F, Steffen A, Barbara R, et al. Chemiluminometric determination of tissue polypeptide antigen (TPA), cancer antigen 15-3 (CA 153), carcinoembryonic antigen (CEA) in comparison with vascular endothelial growth factor (VEGF) in follow-up of breast cancer. Luminescence 2000;15:283–9.
[21] Mekler VM, Bystryak SM. Application of o-phenylediamine as a fluorogenic substrate in peroxidase-medical enzyme-linked immunoserbent assay. Anal Chim Acta 1992;264:359–63. [22] Spona J, Kessler AC, Deeg R. Multi-center study of a new CEA enzyme immunoassay using monoclonal antibodies. J Clin Chem Clin Biochem 1987;25:53–9. [23] Kazuko M, Yuan JL, Wang GL, et al. Simultaneous determination of a-fetoprotein and carcinoembryonic antigen in human serum by timeresolved fluoroimmunoassay. Anal Biochem 1999;276:81–7. [24] Haga M, Hoshino S, Okada H, et al. An improved chemiluminescencebased liposome immunoassay involving apoenzyme. Chem Pharm Bull 1990;38:252–4. [25] Mathieu F, Mareschal JC, Collet-Cassart D. Latex particle immunoassay of carcinoembryonic antigen. J Immunol Methods 1989;120:65–9. [26] Liu YS, Tobias RJ, Zurawski Jr VR. A more specific, simpler radioimmunoassay for carcinoembryonic antigen, with use of monoclonal antibodies. Clin Chem 1985;31:191–5. [27] Koutny LB, Schmalzing D, Taylor TA, et al. Microchip electrophoretic immunoassay for serum cortisol. Anal Chem 1996;68:18–22. [28] von Heeren F, Verpoorte E, Manz A, et al. Micellar electrokinetic chromatography separations and analyses of biological samples on a cyclic planar microstructure. Anal Chem 1996;68:2044–53. [29] Chiem N, Harrison DJ. Microchip-based capillary electrophoresis for immunoassays: analysis of monoclonal antibodies and theophylline. Anal Chem 1997;69:373–8. [30] Chiem N, Harrison DJ. Monoclonal antibody binding affinity determination by microchip-based capillary electrophoresis. Electrophoresis 1998;19:3040–4. [31] Chiem N, Harrison DJ. Microchip systems for immunoassay: an integrated immunoreactor with electrophoretic separation for serum theophylline determination. Clin Chem 1998;44:591–8. [32] Dutra RF, Coelho GD, Silva VL, et al. A reusable amperometric biosensor based on a novel silver-epoxy electrode for immunoglobulin detection. Biotechnol Lett 2000;22:579–83. [33] Fernandez-Sanchez C, Costa-Garcia A. Competitive enzyme immunosensor developed on a renewable carbon paste electrode support. Anal Chim Acta 1999;402:119–27. [34] Zhou YM, Liu GD, Wu ZY, et al. An amperometric immunosensor based on a conducting immunocomposite electrode for the determination of schistosoma japonicum antigen. Anal Sci 2002;18:155–9. [35] Liu GD, Yan JT, Shen GL, et al. Renewable amperometric immunosensor for complement 3 (C3) assay in human serum. Sens Actuators B Chem 2001;B80:95–100. [36] Santandreu M, Sole S, Fabregas E, et al. Development of electrochemical immunosensing systems with renewable surfaces. Biosens Bioelectron 1998;13:7–17. [37] Trasatti S, Petrii OA. Real surface area measurements in electrochemistry. J Electroanal Chem 1992;327:353–76. [38] Xu JJ, Fang HQ, Chen HY. The electrochemical characteristics of gold electrode modified by thionine covalently bound to self-assembled cysteamine monolayer and its electrocatalytic oxidation for ascorbic acid. Chem J Chin Univ 1997;18:706–10. [39] Ruan CM, Yang F, Lei CH, et al. Thionine covalently tethered to multiplayer horseradish peroxidase in a self-assembled monolayer as an electron-transfer mediator. Anal Chem 1998;70:1721–5. [40] Nakamura N, Lim TK, Jeong JM, et al. Flow immunoassay for detection of human chorionic gonadotrophin using a cation exchange resin packed capillary column. Anal Chim Acta 2001;439:125–30. [41] Lim TK, Matsunaga T. Construction of electrochemical flow immunoassay system using capillary columns and ferrocene conjugated immunoglobulin G for detection of human chorionic gonadotrophin. Biosens Bioelectron 2001;16:1063–9. [42] Reinecke M, Scheper T. Fast on-line flow injection analysis for IgG monitoring in bioprocesses. J Biotechnol 1997;59:145–53.
Electrochemical sensor for immunoassay of carcinoembryonic antigen based on thionine monolayer modified gold electrode Zong Dai PhDa, Jin Chen a, Feng Yan b, Huangxian Ju MD, PhDa,* a
Department of Chemistry, Institute of Chemical Biology, Key Laboratory of Analytical Chemistry for Life Science, Nanjing University, Nanjing 210093, China b Department of Medical Examination, Cancer Hospital and Institute of Jiangsu Province, Nanjing 210009, China Accepted 7 December 2004
Abstract A sensor based on thionine monolayer modified gold electrode for determination of carcinoembryonic antigen (CEA) in human serum is proposed. The sensor is prepared by covalently binding thionine to a cysteamine self-assembled monolayer with p-phthaloyl chloride as a linkage, which gives a surface coverage of 8.97 3.28 l0 12 mol/cm2 for thionine. The electrochemistry of the immobilized thionine displays a surface-controlled electrode process with an average electron transfer rate constant of 1.47 0.84 s 1. Based on an electrochemical enzyme-linked immunoassay by using the immobilized thionine as an electron transfer mediator between the electrode and the horseradish peroxidase (HRP) labeled anti-CEA antibody, a calibration curve with two linear ranges from 0.6 to 17 and 17 to 200 ng/mL and a detection limit of 0.2 ng/mL for CEA determination is obtained in pH 4.2 PBS containing 2.0 mmol/L H2O2 and 0.5 mol/L NaCl. The sensor shows a good accuracy. The precision and reproducibility are acceptable with the intra-assay CVof 4.9% and 5.9% at 10 and 100 ng/mL CEA concentrations, respectively, and the inter-assays CV of 7.8% at 100 ng/mL CEA. The response of thionine modified electrode shows only 1.6% decrease after 100 replicate measurements and the storage stability is acceptable in a pH 7.0 PBS at 4 8C for 1 week. The method avoids the addition of electron transfer mediator to the solution, thus is much simpler. The proposed method would be valuable for the diagnosis and monitoring of carcinoma and its metastasis. # 2005 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. Keywords: Sensors; CEA; Thionine; Gold electrode; Self-assembled monolayer; HRP; ELISA; Immunoassay
1. Introduction Carcinoembryonic antigen (CEA), an acidic glycoprotein with a molecular weight of about 200 kDa, is one of the most widely used tumor makers. It was first described in 1965 as a marker of colon cancer. Up to now, it has been found to exist in the serum of most tumor patients. The CEA level in serum is related to lung cancer [1–3], ovarian carcinoma [4,5], colon cancer [6–8], breast cancer [9–11], cystadenocarcinoma [5] and others [12–15]. Thus, the determination of CEA level is very helpful to clinical tumor diagnoses. On the other hand, the CEA level in serum is also related to the state of tumor, so it can be directly used for evaluating curative effect, judging recrudescence or metastasis as a marker. * Corresponding author. Tel.: +86 25 83593593; fax: +86 25 83593593. E-mail address: [email protected] (H. Ju).
The techniques usually used for quantitative determination of tumor markers are immunological methods, which have become the predominant analytical techniques in the fields of clinical diagnoses and biochemical studies [16]. Immunoassay methods possess extremely high selectivity and sensitivity [17]. The selectivity results from the capacity of antibody recognizing the corresponding antigen. A number of immunoassay methods for the determination of CEA level in serum have been reported [18–28]. These methods with commercially available assay kits include radioimmunoassay [18], chemiluminescence immunoassay [19,20], enzyme-linked immunosorbent assay [21,22], fluoroimmunoassay [23] and liposome immunoassay [24]. Several years ago, radioimmunoassay was the most commonly used method for antigen detection. However, radioactive labels are harmful to the operators. Thus, test kits of enzyme-linked immunosorbent assay and chemilumines-
0361-090X/$30.00 # 2005 International Society for Preventive Oncology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.cdp.2004.12.003
234
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
cence immunoassay have currently been used extensively for CEA determination in clinical diagnoses. Although these traditional methods have detection limits of 0.1–0.5 ng/mL for CEA [25,26] and are suitable enough for the determination of CEA in sera, they are usually time and sample consuming because of several incubation and washing steps needed in detection procedure. Moreover, implement point-of-care (POC) testing is difficult with conventional immunoassay because rather large equipment is necessary. In order to overcome these drawbacks, some works on microchip-based immunoassays have been reported [27–31]. These systems were based on the separation of the free form of antigen and the complex of antigen and antibody. They were not suitable for every analyte contained in biological samples due to the precipitation of the antigen–antibody complex [17,28]. Sato et al. structured a bead-bed immunoassay system on a microchip for the determination of CEA [17]. Anti-CEA molecules were firstly precoated on polystyrene beads, which were introduced into a microchannel. The serum sample containing CEA, the primary antibody and the secondary antibody conjugated with colloid gold were then reacted successively. The resulting complex was finally detected using a laser-induced thermal lens microscope (TLM) and gave a detection limit of 0.03 ng/mL that was lower than those of the conventional ELISA. But the TLM is obviously more expensive than the equipment required usually for ELISA. In this paper, we prepared a thionine monolayer modified gold electrode as a sensor to develop a novel fast and cheap thionine–H2O2–horseradish peroxidase (HRP) solid-phase immunoassay for CEA determination. The immobilized thionine was used as an electron transfer mediator between the electrode and horseradish peroxidase labeled anti-CEA antibody. This method avoids the addition of electron transfer mediator to the solution; thus, it needs less incubation and washing steps and is simpler than traditional ELISA methods. Moreover, the equipment required for this method is very cheap. In comparison with other immunosensors prepared with graphite [32–36], the presented sensor is steady, avoids being contaminated by the addition of mediator and can be repeatedly used without further pretreatment. At the same time, it is easily miniaturized and does not contaminate the sample. This method is more sensitive than traditional RIA and almost has the same detection limit as that of chemiluminescence immunoassay and fluoroimmunoassay, and thus, has a good application perspective for continuous determination of CEA.
different concentrations from 0 to 500 ng/mL, and a solution of horseradish peroxidase conjugated monoclonal anti-CEA antibody. CNBr-activated Sepharose 6 MB was purchased from Amersham Pharmacia Biotech. Thionine, cysteamine, p-phthaloyl chloride and H2O2 with analytical grade were from the Shanghai Biochemical Reagent Company (China). All other reagents were of analytical grade. Double distilled water was used for all experiments. The eluent buffer was a solution of 0.1 mol/L pH 4.2 PBS containing 0.5 mol/L NaCl. All reagents were brought to room temperature (ca. 20 8C) before use. 2.2. Preparation of thionine modified gold electrode Gold electrodes, made by sealing gold wire (0.5 mm diameter) in a glass tube of 3 mm exterior diameter, were polished to a mirror-like surface with 0.3 and 0.05 mm alumina slurry on microcloth pads (Buehler Co., USA), followed by being rinsed with water and ethanol and sonicated in double distilled water for 2 min. After cleaning, the electrode was immersed in mercury for about 30 min for amalgamation, the amalgamated gold electrode was immersed in 1.0 mol/L HNO3 till the amalgam was dissolved and the gold surface was exposed. The roughness factor of the electrode surface was 6.59, which was obtained from the geometric area of 1.05 l0 2 cm2 and the real area of 6.90 10 2 cm2. The real area was obtained by measuring the cathodic peak area of cyclic voltammogram formed during the re-reduction of superficial gold oxide in 1.0 mol/L H2SO4 [37]. Modification of the gold electrode was performed by soaking a previously pretreated electrode in a deoxygenated 20 mmol/L cysteamine solution for more than 20 h. The resulting monolayer-modified electrode was thoroughly rinsed with water to remove physically adsorbed cysteamine and dried in a jet of nitrogen. The cysteamine-modified electrode was then immersed into a pphthaloyl chloride saturated solution of toluene for 5 h and rinsed with toluene. Finally, the modified electrode was transferred into the acetonitrile solution of 2 mmol/L thionine for 10 h and rinsed with acetonitrile thoroughly to yield a thionine monolayer modified electrode, which was washed with double distilled water and soaked in pH 7.0 PBS at 4 8C for prior to use [38]. Scheme 1 shows the procedure for the preparation of thionine modified gold electrode. 2.3. Apparatus
2. Materials and methods 2.1. Reagents CEA-ELISA and IRMA kits, including 96 assay tubes precoated with murine monoclonal anti-CEA, were purchased from Diagnostic Products (DPC, USA). The ELISA kits consisted of a series of CEA standard solutions with
Electrochemical measurements were performed with a three-electrode system. A platinum wire with 1 mm diameter was used as auxiliary electrode, a ‘‘homemade’’ Ag/AgCl electrode with 2 mm exterior diameter acted as reference and a thionine monolayer modified gold electrode performed as working electrode. The three electrodes were connected with a BAS-100B electrochemical analyzer
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
235
Scheme 1. The procedure of thionine covalently binding to a cysteamine self-assembled monolayer with p-phthaloyl chloride for preparation of thionine modified gold electrode.
(Bioanalytical Systems Inc., USA) and directly fitted into the microtiter plate wells. IRMA procedure was carried out with a FMJ-182 Immunoradiometric Gamma Counter (China) according to the instruction and assay procedure in the operator’s manual. The chemiluminescent detections were carried out with an IFFM-D Luminescence Analyzer (Remex Electronic Instrument Limited Co., Xi’an, China). Polytetrafluoroethylene (PTFE) tubing (0.25 mm i.d., 0.8 mm o.d.) was used to connect all components in the flow system. 2.4. Procedure The analytical procedure for immunoassay is schematically depicted in Scheme 2. The principle of the sandwich assay was based on one-step immunoreactions. Each microtiter well precoated with primary anti-CEA antibody had 50 mL of CEA sample solution and 100 mL of HRP labeled anti-CEA antibody added. The incubation mixture was thoroughly mixed for 10 s and incubated at 23 8C for 60 min. After the HRP labeled anti-CEA antibody molecules were adsorbed on the microtiter plate wells through immune complex formation, the incubation mixture was
removed by emptying the plate contents into a waste container, the microtiter wells were rinsed and emptied five times with double distilled water. The residual water droplets were removed by striking the plate sharply onto absorbent paper. A 200 mL deoxygenated eluent buffer was then added into the wells and left at 25 8C for 10 min to elute the HRP labeled anti-CEA antibody from the wells. In order to increase the elution efficiency, the eluent buffer contained 0.5 mol/L NaCl. After dispensing 2 mL 0.2 mol/L H2O2 into the well, the electrochemical detection was performed by inserting the three-electrode system into the microtiter well. Differential pulse voltammetric (DPV) measurements were performed from 150 to 250 mV with a pulse amplitude of
Scheme 2. Schematic diagram of FIA system: (P) peristaltic pump, (V) eight-way valve, (C) affinity column, (PMT) photo multiplier, (D) detector and (PC) computer.
236
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
50 mV, a pulse width of 60 ms, a scan rate of 20 mV/s, a pulse period of 200 ms and a sampling width of 4 mV.
3. Results 3.1. Electrochemical behavior of thionine modified gold electrode Fig. 1 shows the cyclic voltammograms of different electrodes in 0.1 mol/L pH 4.2 PBS. No response was observed with bare, amalgamated and cysteamine modified gold electrodes in the working potential window (curves a– c). After binding thionine to the cysteamine self-assembled monolayer with p-phthaloyl chloride, a pair of stable and well-defined redox peaks was observed. The anodic and cathodic peak potentials of thionine monolayer were 32 and +28 mV at 50 mV/s, respectively (curve d). Both the anodic and cathodic peak currents increased with the increasing scan rate and were proportional to the scan rate in the range from 5 to 50 mV/s (Fig. 2), displaying a surfacecontrolled electrode process. The surface coverage of thionine was calculated to be 8.97 3.28 l0 12 mol/ cm2. When the scan rate was higher than 50 mV/s the electrode reaction showed a diffusion-controlled process due to the kinetic limitation. 3.2. Cyclic voltammetric response of enzymatic reaction product It is known that HRP can catalyze the oxidation reaction of thionine by H2O2. In the absence of CEA antigen in the incubation solution, the thionine monolayer modified gold electrode showed a pair of symmetrical redox peaks in pH 4.2 PBS containing 2.0 mmol/L H2O2 (Fig. 3a). When the incubation solution contained 120 ng/mL CEA the oxidation
Fig. 1. Cyclic voltammograms of different electrodes in pH 4.2 PBS at 45 mV/s: (a) unpretreated bare, (b) amalgamatic bare, (c) cysteamine modified and (d) thionine modified gold electrodes.
Fig. 2. Cyclic voltammograms of thionine modified electrode in pH 4.2 PBS at 5, 10, 15, 20, 25, 30, 35, 40, 45 and 50 mV/s (from lowest to highest peak currents). Inset: plots of peak current vs. scan rate.
peak current decreased and the reduction peak current increased greatly and the reduction peak potential shifted slightly to more negative value (Fig. 3b), showing an obvious enzymatic reaction process. Because DPV method can eliminate the background current of voltammetric response, the following experiments were done by DPV method for getting more sensitive results. When thionine modified electrode was scanned in pH 4.2 PBS using DPV, a reduction peak of the immobilized thionine was found at 46 mV (curve a, inset in Fig. 3). The reduction peak current slightly increased upon addition of
Fig. 3. Cyclic voltammetric response of enzymatic reaction on thionine modified electrode in pH 4.2 PBS containing 2.0 mmol/L H2O2 for the incubation processes using (a) 0 ng/mL CEA and (b) 120 ng/mL CEA. Inset: differential pulse voltammograms of the thionine modified electrode in (a) pH 4.2 PBS, (b) (a) + 2.0 mmol/L H2O2 and (c) (b) + 40 ng/mL HRP.
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
237
2.0 mmol/L H2O2 (curve b, inset in Fig. 3). After 40 ng/mL HRP was added into above solution, the reduction peak current increased greatly with a slight negative shift of the reduction peak potential (curve c, inset in Fig. 3), indicating an obvious enzymatic reaction process. 3.3. Optimization of CEA determination conditions The important factors for the elution of HRP labeled antiCEA antibody from the primary anti-CEA antibody/CEA immunocomplex included pH value of eluent buffer and elution time. The effect of the pH of the eluent buffer on the DPV peak current was studied in the pH range from 2.5 to 6.9 at a 500 ng/mL CEA concentration. From the results shown in Fig. 4A, the maximum response occurred at pH 4.2. The effect of elution time on the elution efficiency was studied using flow injection chemiluminescence (CL) immunoassay (Scheme 2). CEA molecules were immobilized on CNBr-activated Sepharose 6 MB and introduced into an affinity column according to the manufacturer’s instructions. After the affinity, column was incubated with HRP labeled anti-CEA antibody and the uncomplexed HRP labeled anti-CEA antibody was washed out, the eluent buffer was pumped through the column for 25 s and then the luminol and H2O2 were added each 60 s. The elution process was evaluated by the decrease of CL intensity of luminol. With increasing elution time, the CL intensity of luminol decreased greatly (Fig. 5), indicating a decrease in the amount of HRP in the affinity column. In Fig. 5 the elution efficiency was calculated by 100 DI/(Ii I0), in which Ii was the initial luminol CL intensity without being eluted, I0 was the blank value and DI was the decrease of the CL intensity of luminol relative to Ii(It Ii). After elution for about 400 s, the CL intensity was as low as that of the blank, and the elution efficiency tended to be 100%, indicating all the HRP labeled anti-CEA antibodies completed with CEA in the affinity column had been eluted. Thus, the optimal elution time of 10 min in the well was selected for CEA determination. The effect of H2O2 concentration on the enzymatic reaction was examined using DPV. As shown in Fig. 4B, with increasing H2O2 concentration, the DPV response displayed a Michaelis–Menten’s shaped curve. At H2O2 concentrations less than 2.0 mmol/L, the peak current
Fig. 5. Effect of elution time on luminolCL intensity (a) for a flow elution process and (b) an affinity column without incubation with HRP labeled CEA antibody. Inset: effect of elution time on the elution efficiency.
increased linearly with increasing H2O2 concentration. When H2O2 concentration was more than 2.0 mmol/L, the reduction peak current trended to a constant value. Therefore, the optimal H2O2 concentration was 2.0 mmol/L. 3.4. Calibration for CEA determination With thionine monolayer modified electrode as a sensor, the calibration graph for the determination of CEA was obtained under optimal conditions (Fig. 6). Following the standard immunometric sandwich and the enzymatic reaction procedures, the DPV peak current increased with increasing CEA concentration in two linear ranges from 0.6 to 17 and 17 to 200 ng/mL, the linear slopes were 9.5 and 0.55 nA/(ng mL), and the correlation coefficients were 0.9995 and 0.9880, respectively. The detection limit was 0.2 ng/mL at 3s. 3.5. Precision, reproducibility and stability The typical intra-assay and inter-assay precision for CEA assay were also estimated. The variation coefficients (CV) of
Fig. 4. Effects of pH value of eluent buffer (A) and H2O2 concentration (B) on DPV peak current for CEA determination.
238
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
Fig. 6. Plot of peak current vs. CEA concentration under optimal conditions. Inset: differential pulse voltammograms of the thionine modified electrode in pH 4.2 PBS containing 2.0 mmol/L H2O2 for 0, 0.6, 1, 5, 10, 12, 15, 30, 60, 100, 120, 200 and 300 ng/mL CEA (from lowest to highest peak currents).
intra-assay, obtained by three repeated determinations with the same modified electrode, were 4.9% and 5.9% at CEA concentrations of 10 and 100 ng/mL, respectively, and the CV value for inter-assay variation was 7.8% at 100 ng/mL. Moreover, when the peak currents of the thionine modified gold electrode in pH 4.2 PBS were determined 100 times, the response showed only a 1.6% decrease, indicating a good reproducibility of this sensor. The stability of this sensor was acceptable in pH 7.0 PBS at 4 8C for 1 week. 3.6. Accuracy and clinic application The CEA levels in two sera were quantified by using standard addition method. The CEA concentrations determined with this method and the IRMA method are shown in Table 1. These results by both methods were in good agreement. Thus, the proposed method could be satisfactorily applied to the clinical determination of CEA in human serum.
4. Discussion The amalgamated bare gold electrode shows a charging current much larger than that of bare gold electrode (curves a
and b in Fig. 1), indicating a rougher surface is obtained by amalgamation. The formation of a cysteamine monolayer results in a smaller charging current (curve c in Fig. 1) than that of amalgamated bare gold electrode. The surface is clean and smooth. This is also an effective way to provide the gold electrode with abundant amino groups for covalent immobilization of thionine to form thionine modified gold electrode. From the peak-to-peak separation of the thionine modified gold electrode at different scan rates, an average electron transfer rate constant of 1.47 0.84 s 1 was obtained for the surface-controlled redox process. When the incubation solution contains CEA, the HRP labeled anti-CEA antibody molecules are eluted into the wells after a deoxygenated eluent buffer is added into the wells. The presence of HRP results in an enzymatic reaction process upon addition of H2O2. The oxidation reaction of thionine by H2O2 under the enzymatic catalysis of HRP increases the concentration of thionine in its oxidized state at the electrode surface, thus, increasing greatly the reduction peak current of thionine modified gold electrode. The immunoassay procedure and the enzymatic reaction mechanism can be expressed as shown in Schemes 3 and 4 [39], respectively. Both the elution efficiency and the enzyme activity are related to the pH value of eluent buffer. High elution efficiency is usually obtained at low pH. However, most enzymes display their activity in a limited pH range. The maximum activity of HRP is in the range from pH 6.5 to 7.0. With increasing solution acidity its activity decreases greatly. The integrated action of solution acidity results in that the maximum response occurs at pH 4.2, which is selected for elution of HRP labeled anti-CEA antibody and immunoassay of CEA. The optimal pH is higher than those reported in flow immunoassay [40–42]. The elution efficiency also depends on the elution time, which affects the amount of HRP introduced into the detection solution. The decrease of the CL intensity of luminol is due to the reducing amount of HRP in the affinity column, which is a mark of elution efficiency. From the analysis shown in Fig. 5 the optimal elution time in the well is 10 min. With the optimal elution time and pH value of eluent buffer the amperometric immunoassay of CEA in serum samples can be performed at the proposed thionine self-assembled gold electrode at a H2O2 concentration of 2.0 mmol/L. The detection limit of CEA is 0.2 ng/mL, which is more sensitive than traditional RIA
Table 1 Correlation between two sera quantified by this method and IRMA method Serum samples
This method (ng/mL)
IRMA (ng/mL)
Relative deviation (%)
1
5.3 5.2 5.7
5.1
3.9 2.0 21
2
34.1 32.6 32.9
35.8
4.7 8.9 8.1
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
239
Scheme 3. The immunoassay procedure.
Scheme 4. The mechanism of enzymatic and electrode reactions.
and is almost the same as that of chemiluminescence immunoassay and fluoroimmunoassay. Considering its acceptable intra-assay and inter-assay precision and the good reproducibility and storage stability of the sensor, the presented method could be used in CEA determination for clinical diagnosis.
addition of electron transfer mediator to the solution, thus is much simpler and cheaper than traditional assays. Furthermore, both electrochemical analyzer and the sensor can be easily miniaturized; thus, this method is particularly well suited for point-of-care testing and application in emergency rooms, ambulances, bedside, or doctor’s office. At the same time, it can be readily extended toward detection of other clinically important antigens.
5. Conclusions A sensor prepared by immobilizing thionine monolayer on gold electrode can be used for determination of serum CEA levels. Under the optimal conditions, a detection limit of 0.2 ng/mL CEA and two linear ranges from 0.6 to 17 and 17 to 200 ng/mL for CEA determination were obtained. The sensor shows good accuracy and acceptable precision, reproducibility and storage stability. It can be repeatedly used without further pretreatment and seems to be rather suitable for clinic immunoassay. The method avoids the
Acknowledgments This project was supported by the Distinguished Young Scholar Fund to HX Ju (20325518), the National Natural Science Foundation of China (No. 20275017), the Science Foundation of Jiangsu (No. BS2001063), the Specialized Research Fund for Excellent Young Teachers from Ministry of Education of China, and the Key Research Project of Cancer Hospital and Institute of Jiangsu Province.
240
Z. Dai et al. / Cancer Detection and Prevention 29 (2005) 233–240
References [1] Moro D, Villemain D, Vuillez JP, et al. CEA, CYFEA21-1 and SCC in non-small cell lung cancer. Lung Cancer 1995;13:169–76. [2] Iwazawa T, Kanoh T, Matsui S, et al. Diagnosis of lung cancer metastasis with CEA extracted from the dissected regional lymph nodes. Lung Cancer 2000;29:254. [3] Nisman B, Lafair J, Peretz T, et al. Cyfra 21-1 and TPS-New markers in lung cancer. Eur J Cancer 1995;31:s263. [4] Gould DA, Moscoso GJ, Young MPA, et al. Human first trimester fetal ovaries express oncofetal antigens and steroid receptors. J Soc Gynecol Investig 2000;7:131–8. [5] Kazuya K, Yoshihiro K, Tsunekazu K, et al. Preoperative determination of several serum tumor markers in patients with primary epithelial ovarian carcinoma. Gynecol Obstet Invest 1999;47:52–7. [6] Eppler E, Horig H, Kaufman HL, et al. Carcinoembryonic antigen (CEA) presentation and specific T cell-priming by human dendritic cells transfected with CEA-mRNA. Eur J Cancer 2002;38:184–93. [7] Ford CHJ, Osborne PA, Rego BG, et al. Bispecific antibody targeting of doxorubicin to carcinoembryonic antigen-expressing colon cancer cell lines in vitro and in vivo. Int J Cancer 2001;92: 851–5. [8] Vogel I, Francksen H, Soeth E, et al. The carcinoembryonic antigen and its prognostic impact on immunocytologically detected intraperitoneal colorectal cancer cells. Am J Surg 2001;181:188–93. [9] Kramer S, Jager W, Lang N. Early treatment of metastatic breast cancer patients after increase of CEA and CA 15-3 serum levels. Eur J Cancer A 1998;34:s42. [10] Cameiro C, Costa L, Melo M, et al. Serum tumor markers in metastatic breast cancer comparative study between CEA, CA 15-3 and MCA. Eur J Cancer A 1998;34:s43. [11] Bremer K, Nicus S, Bremer G. CEA, CA 15-3 and MCA: comparative clinical relevance in breast cancer. Eur J Cancer 1995;31:s262. [12] Tachibana M, Takemoto Y, Nakashima Y, et al. Serum carcinoembryonic antigen as a prognostic factor in resectable gastric cancer. J Am College Surg 1998;187:64–8. [13] Kau SY, Shyr YM, Su CH, et al. Diagnostic and prognostic values of CA 19-9 and CEA in periampullary cancers. J Am College Surg 1999;188:415–20. [14] Lechner P, Lind P, Goldenberg DM. Can postoperative surveillance with serial CEA immunoscintigraphy detect respectable rectal cancer recurrence and potentially improve tumor-tree survival? J Am College Surg 2000;191:511–8. [15] Takamori S, Yoshimoto J, Kojima K, et al. Prognostic implication and intratumoral heterogeneity of p53, Ki-67, CEA and CA 19-9 expressions in cholangiocellular carcinoma. Hepatol Res 1997;8:81–94. [16] Karsten S, Andreas K, Dieter T, et al. Disposable optical sensor chip for medical diagnostics: new ways in bioanalysis. Anal Chem 1999;71:5430–5. [17] Sato K, Tokeshi M, Kimura H, et al. Determination of carcinoembryonic antigen in human sera by integrated bead-bed immunoassay in a microchip for cancer diagnosis. Anal Chem 2001;73:1213–8. [18] Cioffi M, Vietri MT, Gazzerro P, et al. Serum anti-p53 antibodies in lung cancer: comparison with established tumor markers. Lung Cancer 2001;33:163–9. [19] Forest JC, Masse J, Lane AA. Evaluation of the analytical performance of the boehringer mannheim elecsys. 2010 Immunoanalyzer. Clin Biochem 1998;31:81–8. [20] Rainer F, Steffen A, Barbara R, et al. Chemiluminometric determination of tissue polypeptide antigen (TPA), cancer antigen 15-3 (CA 153), carcinoembryonic antigen (CEA) in comparison with vascular endothelial growth factor (VEGF) in follow-up of breast cancer. Luminescence 2000;15:283–9.
[21] Mekler VM, Bystryak SM. Application of o-phenylediamine as a fluorogenic substrate in peroxidase-medical enzyme-linked immunoserbent assay. Anal Chim Acta 1992;264:359–63. [22] Spona J, Kessler AC, Deeg R. Multi-center study of a new CEA enzyme immunoassay using monoclonal antibodies. J Clin Chem Clin Biochem 1987;25:53–9. [23] Kazuko M, Yuan JL, Wang GL, et al. Simultaneous determination of a-fetoprotein and carcinoembryonic antigen in human serum by timeresolved fluoroimmunoassay. Anal Biochem 1999;276:81–7. [24] Haga M, Hoshino S, Okada H, et al. An improved chemiluminescencebased liposome immunoassay involving apoenzyme. Chem Pharm Bull 1990;38:252–4. [25] Mathieu F, Mareschal JC, Collet-Cassart D. Latex particle immunoassay of carcinoembryonic antigen. J Immunol Methods 1989;120:65–9. [26] Liu YS, Tobias RJ, Zurawski Jr VR. A more specific, simpler radioimmunoassay for carcinoembryonic antigen, with use of monoclonal antibodies. Clin Chem 1985;31:191–5. [27] Koutny LB, Schmalzing D, Taylor TA, et al. Microchip electrophoretic immunoassay for serum cortisol. Anal Chem 1996;68:18–22. [28] von Heeren F, Verpoorte E, Manz A, et al. Micellar electrokinetic chromatography separations and analyses of biological samples on a cyclic planar microstructure. Anal Chem 1996;68:2044–53. [29] Chiem N, Harrison DJ. Microchip-based capillary electrophoresis for immunoassays: analysis of monoclonal antibodies and theophylline. Anal Chem 1997;69:373–8. [30] Chiem N, Harrison DJ. Monoclonal antibody binding affinity determination by microchip-based capillary electrophoresis. Electrophoresis 1998;19:3040–4. [31] Chiem N, Harrison DJ. Microchip systems for immunoassay: an integrated immunoreactor with electrophoretic separation for serum theophylline determination. Clin Chem 1998;44:591–8. [32] Dutra RF, Coelho GD, Silva VL, et al. A reusable amperometric biosensor based on a novel silver-epoxy electrode for immunoglobulin detection. Biotechnol Lett 2000;22:579–83. [33] Fernandez-Sanchez C, Costa-Garcia A. Competitive enzyme immunosensor developed on a renewable carbon paste electrode support. Anal Chim Acta 1999;402:119–27. [34] Zhou YM, Liu GD, Wu ZY, et al. An amperometric immunosensor based on a conducting immunocomposite electrode for the determination of schistosoma japonicum antigen. Anal Sci 2002;18:155–9. [35] Liu GD, Yan JT, Shen GL, et al. Renewable amperometric immunosensor for complement 3 (C3) assay in human serum. Sens Actuators B Chem 2001;B80:95–100. [36] Santandreu M, Sole S, Fabregas E, et al. Development of electrochemical immunosensing systems with renewable surfaces. Biosens Bioelectron 1998;13:7–17. [37] Trasatti S, Petrii OA. Real surface area measurements in electrochemistry. J Electroanal Chem 1992;327:353–76. [38] Xu JJ, Fang HQ, Chen HY. The electrochemical characteristics of gold electrode modified by thionine covalently bound to self-assembled cysteamine monolayer and its electrocatalytic oxidation for ascorbic acid. Chem J Chin Univ 1997;18:706–10. [39] Ruan CM, Yang F, Lei CH, et al. Thionine covalently tethered to multiplayer horseradish peroxidase in a self-assembled monolayer as an electron-transfer mediator. Anal Chem 1998;70:1721–5. [40] Nakamura N, Lim TK, Jeong JM, et al. Flow immunoassay for detection of human chorionic gonadotrophin using a cation exchange resin packed capillary column. Anal Chim Acta 2001;439:125–30. [41] Lim TK, Matsunaga T. Construction of electrochemical flow immunoassay system using capillary columns and ferrocene conjugated immunoglobulin G for detection of human chorionic gonadotrophin. Biosens Bioelectron 2001;16:1063–9. [42] Reinecke M, Scheper T. Fast on-line flow injection analysis for IgG monitoring in bioprocesses. J Biotechnol 1997;59:145–53.