- Email: [email protected]
A novel microchip electrophoresis-based chemiluminescence immunoassay for the detection of alpha-fetoprotein in human serum
Talanta 165 (2017) 107–111
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
A novel microchip electrophoresis-based chemiluminescence immunoassay for the detection of alpha-fetoprotein in human serum ⁎
Jingwen Liu, Jingjin Zhao , Shuting Li, Liangliang Zhang, Yong Huang, Shulin Zhao
MARK
⁎
State Key Laboratory for the Chemistry and Molecular Engineering of Medicinal Resources, Key Laboratory of Ecology of Rare and Endangered Species and Environmental Protection of Ministry Education, Guangxi Normal University, Guilin 541004, China
A R T I C L E I N F O
A BS T RAC T
Keywords: Microchip electrophoresis Chemiluminescence detection Immunoassays Tumor marker Alpha-fetoprotein
A sensitive immunoassay method based on microchip electrophoresis chemiluminescence (MCE-CL) detection technology was developed for the detection of tumor marker alpha-fetoprotein (AFP). This method adopts the non-competitive immunoassay mode, and was conducted after AFP reacted with excessive horseradish peroxidase (HRP) labeled monoclonal antibody. The extreme pH value was adopted in the electrophoresis buffer solution. The use of brij 35 as an additive of electrophoresis buffer increased dramatically the resolution (Rs) and the reproducibility of the analysis. Under the optimized experimental conditions, effective separation of the immune complex Ag-Ab* and free Ab* was achieved within 60 s. The peak height of the immune complex Ag-Ab* was taken as quantification of AFP. Good linearity was observed within AFP concentrations ranging from 10 ng/mL to 150 ng/mL, and the detection limit was found to be 7.2 ng/mL (1.0×10−10 M). The present method was successfully applied for the detection of AFP in human serum from both healthy and cancer patients, and the AFP levels in the both were found be in the range of 16.5–23.4 ng/mL and 416.2–825.4 ng/ mL, respectively.
1. Introduction Alpha-fetoprotein (AFP), an oncofetal glycoprotein, is commonly used as a marker for hepatocellular carcinoma (HCC). Under healthy conditions, AFP comes from embryonic hepatocytes and its concentration begins to decline in fetal blood approximately two weeks after birth. In adult serum, the normal content of AFP is low than 25 ng/mL [1]. An elevated AFP concentration in serum may be an early indication of HCC, hepatoblastoma, and germ cell tumors [2,3]. Therefore, AFP is considered to be a specific clinical marker for the diagnosis of primary liver cancer. Similarly, AFP concentrations increase with disease progression in patients with pancreatic cancer, lung cancer, and hepatic cirrhosis [4,5]. Periodic detection of AFP concentration in human bodily fluids for the clinical analysis of treatment outcomes, prognosis assessment, prediction of recurrence and metastasis is very important. Currently, the commonly used AFP detection methods include radioimmuno-assay [6], enzyme-linked immunoassay (ELISA) [7], chemiluminescence immune- assay (CL-IA) [8], capillary electrophoresis immunoassay (CE-IA) [9–11] and immunosensors [12,13]. These methods usually require long analysis times and complex liquid handling procedures. Additionally, these methods utilize expensive
⁎
antibody reagents, which results in high costs, ultimately restricting their widespread application. In order to improve the sensitivity of AFP detection methods, a variety of nanomaterials including grapheme [14], carbon nanotubes [15], gold nanoparticles [16], PdNi nanoparticles [17] and mesoporous silica [18] have been utilized for signal amplification in the methods. In above mentioned methods, highly sensitivity was obtained, however, most of these methods require tedious surface modification of electrode or liquid handling procedures, which make the assays time-consuming. Therefore, it is necessary to develop a new immunoassay method with fast detection and high accuracy for clinical AFP detection. The immunoassay, which is a detection method based on specific responses that arise from the binding of an antibody with an antigen, has high selectivity and sensitivity, and is therefore extensively used in clinical diagnoses and biochemical analyses [19]. However, routine immunoassay requires extensive amounts of expensive reagents, long analysis time and complex operation procedures. Microchip electrophoresis (MCE) is a new separation assay technique developed after capillary electrophoresis (CE). MCE has the advantages of high efficiency, fast analysis, less sample and reagent consumption, and high automation and integration degree; therefore, it has been widely adopted in chemical analyses [20].
Correspondence to: College of Chemistry and Pharmacy, Guangxi Normal University, Guilin 541004, China. E-mail addresses: [email protected] (J. Zhao), [email protected] (S. Zhao).
http://dx.doi.org/10.1016/j.talanta.2016.12.038 Received 20 September 2016; Received in revised form 13 December 2016; Accepted 18 December 2016 Available online 19 December 2016 0039-9140/ © 2016 Elsevier B.V. All rights reserved.
Talanta 165 (2017) 107–111
J. Liu et al.
2.3. Immunoreaction procedure
Chemiluminescence (CL) detection is a highly sensitive detection system for MCE. When MCE-CL is used in conjunction with immunoassay, the resulting analyses have high selectivity of immunoassay and high sensitivity of CL detection. Therefore, it has been shown that the combination of microfluidic systems and immunoassay is a highly effective separation assay technique [21]. However, no study has reported the use of MCE-CL for AFP detection. In the present study, a new non-competitive immunoassay method based on the MCE-CL technology was established for AFP detection. Following the reaction of AFP antigen (Ag) with an excess of HRP labeled anti-AFP antibody (Ab*), MCE-CL method was used to conduct AFP detection. Free Ab* and the immune complex (Ag-Ab*) were separated within 60 s. The method has been successfully used for the detection of trace AFP in human serum, confirming the utility of the assay.
The non-competitive mode was adopted for the immunoassay. The principle binding process was as follows:
Ag + Ab*(excess)—Ag − Ab* + Ab* First, 10 μL of AFP antigens with different concentrations or the serum sample were mixed with 10 μL of 0.25 μg/mL HRP labeled mouse anti-AFP monoclonal antibody solution in a micro-centrifuge tube. The mixture was then diluted to 50 μL using the phosphate buffer (pH 7.4) and incubated at 37 °C for 40 min. The solution was then used for MCE-CL analysis.
2.4. MCE procedure Prior to and between any two test of electrophoreses, all chip channels were rinsed thoroughly with 0.1 M NaOH, water and the buffer solution for 10 min each. All the channels were filled with the buffer solution under a vacuum negative pressure of ∼50 mm Hg, and then, each corresponding reservoir was filled with different solutions for separation. Samples were injected in pinched mode. A voltage of 650 V was applied to S, while SW was connected to the ground. Meanwhile, B and BW were applied with 250 V and 400 V, respectively. The injection time was 15 s. During separation, 2600 V was applied to B, and BW was connected to the ground. Meanwhile, 1550 V was applied to S and SW to avoid samples leaking into the channels during separation. R was applied with 500 V. The sample components migrated to the Y-intersection, where they mixed with oxidizer solution to produce CL signal, which was then collected with an object lens (placed at the Y-intersection of the channels) and transferred to a photomultiplier tube (PMT). Finally, the signal was recorded via a HW2000 chromatography work station (Zhejiang University Star Information Technology, Hangzhou, China).
2. Experimental 2.1. Reagents and solutions The AFP immunoassay kit and the horseradish peroxidase (HRP) labeled mouse anti-AFP monoclonal antibody were purchased from Zhengzhou Bioassay Biotechnical Co., Ltd. (Zhengzhou, China). Luminol was purchased from Fluka (Buchs, Switzerland). Brij 35, iodophenol (PIP) and H2O2 were purchased from Shanghai Chemical Company (Shanghai, China). All other chemicals were of analytical reagent grade and used without further purification. Water was purified by employing a Milli-Q plus 185 equip from Millipore (Bedford, MA), and used throughout the work. AFP and HRP-mouse anti-AFP monoclonal antibody were diluted using 20 mM phosphate buffer (pH 7.2). The electrophoresis buffer solution was 10 mM Na3PO4 (pH 10.2) and contained 0.004% (w/w) Brij 35 and 1.0 mM luminol. The oxidizer solution was 45 mM NaHCO3 (pH 9.0) and contained 90 mM H2O2 and 1.2 mM PIP.
2.2. Experimental devices
3. Results and discussion
The MCE-CL detection system was self-constructed in the lab [22]. The glass/polydimethylsiloxane (PDMS) microchip with an expanded Y-shape CL detection pool was designed according to the literature [23] (Scheme 1). The chip had an area of 9.5 cm by 2 cm. The top width of the channel measured 65 µm (except at R and BW, where the distance was 250 µm) and the depth measured 25 µm. The effective length of the separation channel was 60 mm. The distance from the buffer reservoir (B) to the T-intersection was 5 mm. The distance between the separation channel and the sample reservoir (S) or the sample waste reservoir (SW) was 5 mm. The two T-intersections were 60 µm. The sampling volume was calculated to be about 190 pL. The distance from the oxidizing agent reservoir (R) to the Y-intersection was 1.5 cm, and the Y intersection was 1.2 cm from the buffer waste reservoir (BW).
3.1. Optimization of the CL conditions The pH of the oxidizer solution is a major factor affecting CL intensity. In this experiment, 45 mM NaHCO3 was adopted as the buffer to study the influence of varying pH values (pH 8.0–10.0) on the CL intensity. The results are shown in Fig. 1a. It can be seen that the CL intensity first increased and then decreased with continuously increasing pH. When the pH was 9.0, the highest CL intensity was observed. Therefore, 9.0 was selected as optimal pH value of NaHCO3 solution. H2O2 was used as the oxidizing agent in the experiment; its concentration also plays a major role in CL intensity. The influence of the H2O2 concentration on the CL intensity was studied within the range of 50–120 mM. The results indicate that the CL intensity increased with the increase of H2O2 concentration between 50 and 90 mM. When the concentration was above 90 mM, CL intensity showed a downward trend as the H2O2 concentration increased. Therefore, 90 mM was taken as the ideal H2O2 concentration. The PIP is a sensitizing agent that is extensively used in the luminol-H2O2-HRP CL system. PIP greatly increases the CL intensity of the system; therefore, its concentration has notable influence on the luminescence intensity. The influence of the PIP concentration on the CL intensity was investigated between 0.6 and 1.4 mM; the results are shown in Fig. 1b. As can be seen, the CL intensity increased gradually with increasing PIP concentration from 0 to 1.2 mM. After the concentrations are above 1.2 mM, the CL intensity decreased with increasing PIP concentration. Hence, 1.2 mM was selected as the optimal PIP concentration.
Scheme 1. Schematic diagram of the layout of the glass/PDMS microchip. S: sample reservoir; B: buffer reservoir; SW: sample waste reservoir; BW: buffer waste reservoir; R: oxidizing agent reservoir.
108
Talanta 165 (2017) 107–111
J. Liu et al.
Fig. 1. Effects of pH value for oxidizer solution (a), PIP concentration (b) and incubation time (c) on the CL intensity; the effect of brij 35 concentration on the Rs (d). The electrophoresis buffer was 10 mM phosphate buffer (pH 10.2) containing 0.002–0.006% brij 35 and 1.0 mM luminol. The oxidizer solution was 45 mM NaHCO3 (pH 8.0–10.0) solution containing 90 mM H2O2 and 0.6–1.6 mM PIP. The solution was incubated at 37 °C from 25 min to 60 min. The sampling volume was about 190 pL.
excessive Joule heat resulted in reduced separation efficiency. Additionally, high pH levels may cause decomposition of the immune complex. Hence, 10.2 was chosen as the optimal pH value for the buffer solution. The methods for inhibiting protein adsorption usually include permanent coating and dynamic coating. In order to further increase the separation efficiency, the nonionic surfactant brij 35 was used as a buffer additive. Its influence on the separation was investigated within the concentration range of 0.001–0.006% (w/w), and the results indicate that the separation efficiency was improved dramatically with concentration increased (Fig. 1d); however, the retention time gradually increased. Considering the Rs and analysis time, 0.04% (w/w) was selected as the optimal concentration of brij 35. A phosphate solution was chosen in the experiment as the electrophoresis buffer solution. The influence of the buffer concentration on Rs was investigated within the range of 4–14 mM, and the results indicate that the Rs increased with the increasing phosphate concentration. When the concentration exceeded 10 mM, the Rs started to drop; this observation may be due to Joule heating, which results from the high phosphate concentration. Therefore, 10 mM was selected as the optimal phosphate concentration. Additionally, the influence of the separation voltage was also studied. With rising separation voltage, the retention time of the labeled antibody and the immune complex was shortened, resulting in quicker analysis time and narrow peaks. When the voltage was above 2600 V, the Rs decreased rapidly. By considering the analysis speed and Rs, 2600 V was selected as the optimal separation voltage.
3.2. Optimization of the immunoreaction time The immunoreaction time is an essential factor of MCE-CL analysis as it directly determines the sensitivity of the method. The peak height of the immune complex (CL intensity) was taken as the reference to optimize the immunoreaction time of the labeled antibody and antigen; the results are shown in Fig. 1c. It can be seen that with prolonged reaction time, the CL intensity increased rapidly. When the reaction time was more than 40 min, the CL intensity reached a plateau, indicating a balanced condition for the antibody and antigen combination. Therefore, the optimal reaction time was set at 40 min. 3.3. Selection of electrophoresis separation conditions To achieve accurate AFP detection, the factors that affect electrophoresis separation were optimized, including the concentration and the pH of the electrophoresis buffer solution, the concentration of brij 35, and the separation voltage. The pH of the electrophoresis buffer solution is a key factor for electrophoresis separation. In the experiment, the extreme pH value was adopted to perform electrophoresis separation. A 10 mM phosphate solution was used as the buffer. Different pH values ranging from 9.0 to 11 were used to investigate the influence on separation of the immune complex and the free labeled antibody. The results indicate that the resolution (Rs) increased with increasing pH at pH < 10.2. Conversely, at pH > 10.2, the resolution decreased as the pH increased. The possible reason for this outcome was that the high pH levels led to enhanced ionic strength; thus, the large electrophoresis current and 109
Talanta 165 (2017) 107–111
J. Liu et al.
Fig. 2. Electropherograms obtained from separating the immunoreactions solution: (a) Ab* solution, (b) a mixture solution containing Ab* and 80 ng/mL of AFP, (c) a normal human serum sample, (d) a cancer patient serum sample (diluted 10 times). Peak 1 derived from free Ab*, peak 2 derived from the Ab*-Ag complex. The electrophoresis buffer was 10 mM phosphate buffer (pH 10.2) containing 1.0 mM luminol and 0.004% brij 35. The oxidizer solution was 45 mM NaHCO3 (pH 9.0) containing 90 mM H2O2 and 1.2 mM PIP. The sampling volume was about 190 pL.
3.4. MCE-CL noncompetitive immunoassay of AFP Under the optimal experimental conditions, various concentrations of AFP solutions were tested. The typical electropherograms are shown in Fig. 2. In Fig. 2a, the concentration of AFP antigen was 0, therefore only showed one peak (1), which corresponds to Ab*. In Fig. 2b, 80 ng/ mL of the AFP antigen was added; it thus shows two peaks 1 and 2, corresponding to Ab* and Ab*-Ag complex, respectively. The height of peak 2 (CL intensity) increased with increasing AFP antigen concentration. Therefore, the height of peak 2 was used to quantification of AFP. The results showed that AFP had good linearity within the concentration range of 10–150 ng/mL (Fig. 3), with the linear correlation coefficient (R2) being 0.9973. The linear regression equation was:
H = 0.0405 C + 0.2686, R2 = 0.9973 In this equation, H is the CL intensity (mV) and C is the concentration (ng/mL) of the AFP. Taking the signal to noise ratio equal to 3 (S/N=3), the detection limit was found to be 7.2 ng/mL (1.0×10−10 M). To ensure the reproducibility, the MCE-CL analysis was repeated seven times for the immune reaction mixture, and the peak heights and retention times were recorded. The reproducibility of the method was assessed by relative standard deviations (RSDs) of the peak heights and the retention times. The results showed that the RSDs of the peak heights and the retention times were lower than 2.7% and 1.8%, respectively. The results indicated a good reproducibility of the MCE-CL immunoassay method. Thus, it can be applied for the
Fig. 3. The calibration curve for AFP detection. Experiment conditions were as in Fig. 2.
detection of AFP in human serum. 3.5. Specificity for AFP detection To evaluate the specificity of proposed MCE-CL method for AFP detection, we challenged the system with AFP and several nonspecific 110
Talanta 165 (2017) 107–111
J. Liu et al.
monoclonal antibody. The extreme pH value was adopted in the experiment. The use of brij 35 as an additive of electrophoresis buffer dramatically increased the Rs and the reproducibility of the analysis. The proposed method has been used for AFP detection in human serum for both healthy people and cancer patients. This method has several important features. First, in contrast to the conventional immunoassay technique, the MCE-CL immunoassay is simple and does not need complex liquid handling procedures. Second, the use of MCE technology causes a rapid analysis, in addition to the time needed for the immunoreaction, the separation and detection can be finished within 1 min. Third, the proposed MCE-CL immunoassay exhibits an appropriate sensitivity for the determination of AFP content in human serum, which decreases the detection error from excessive concentration or dilution of sample. Thus, we anticipate that this method can be applied readily to detect AFP content in human serum for early diagnosis of cancers.
Table 1 Detection results and recovery of AFP in human serum samples. Sample
Found (ng/mL)
RSD (%, n=5)
Added (ng/mL)
Total found (ng/mL)
Recovery (%)
Normal 1 Normal 2 Normal 3 Normal 4 Patient 1 Patient 2 Patient 3
19.6 23.4 18.8 16.5 416.2 825.4 762.8
4.6 3.8 4.4 2.8 3.2 3.6 3.1
20.0 20.0 20.0 20.0 40.0 80.0 80.0
38.72 44.25 38.07 36.18 80.42 163.86 152.53
95.6 104.2 96.4 98.4 97.0 101.7 95.3
proteins such as immunoglobulin G (IgG), thrombin (Tb) and bovine serum albumin (BSA). The results indicate that only the AFP could lead to appearance of peak 2, no obvious peak 2 was observed in presence of other nonspecific proteins, and the CL intensity of peak 2 was same with that of blank, which shown a high specificity of proposed method for AFP detection.
Acknowledgments This work was supported by the National Natural Science Foundations of China (Nos. 21327007 and 21465005), Natural Science Foundations of Guangxi Province (Nos. 2014GXNSFBA118041), IRT1225, as well as BAGUI Scholar Program.
3.6. Determination of AFP content in human serum Under the above experimental conditions, the human serum samples collected from both liver cancer patient and healthy volunteer were analyzed. Fig. 2 shows typical electropherograms for undiluted serum of healthy volunteer (Fig. 2c) and 10 times diluted serum of liver cancer patient (Fig. 2d). The AFP contents in the serum of the healthy volunteer were in the range 16.5–23.4 ng/mL, which was consistent with the reported reference value (25 ng/mL) for normal people [1]. In contrast, the AFP content in serum of cancer patients varied from 416.2 ng/mL to 825.4 ng/mL. Thus, it is apparent that the AFP content in the serum of cancer patients is significantly higher than that of healthy people, which was in good agreement with previous studies [3,10]. Because the concentration of AFP in the serum of healthy people was two orders of magnitude lower than that of cancer patients, the sample solution needs to be diluted 10 times when detecting the concentration of AFP in cancer patient serum. Although large dilution of sample could lead to a source of error in measurements, this error can be ignored in trace analysis. Hence, the proposed method is feasible for clinical diagnosis of cancers. In addition, in order to determine the reliability of the method, the recovery rate and reproducibility were calculated. The results are listed in Table 1. The RSDs of 5 measurements of human serum samples ranged from 3.1% to 4.6%, and the recoveries of AFP in the human serum samples were in the range 95.3–105.5%.
References [1] Z.F. Fu, C. Hao, X.Q. Fei, H.X. Ju, J. Immunol. Methods 312 (2006) 61–67. [2] Y. Tsuchida, Y. Endo, S. Saito, M. Kaneko, K. Shiraki, K. Ohmi, J. Pediatr. Surg. 13 (1978) 155–156. [3] J.T. Wu, L.D. Book, K. Sudar, Pediatr. Res. 15 (1981) 50–52. [4] C.G. Bergstrand, Acta Paediatr. Scand. 75 (1986) 1–9. [5] D. Bader, A. Riskin, O. Vafsi, A. Tamir, B. Peskin, N. Israel, R. Merksamer, H. Dar, M. David, Clin. Chim. Acta 349 (2004) 15–23. [6] J.E. Fowler, I.A. Sesterhenn, F.K. Mostofi, Urology 33 (1989) 74–77. [7] J.A. Brochot, I.W. Siddiqi, Anal. Chim. Acta 224 (1989) 329–337. [8] Z.J. Yang, H. Liu, C. Zong, F. Yan, H.X. Ju, Anal. Chem. 81 (2009) 5484–5489. [9] R.Y. Wang, X.N. Lu, W.Y. Ma, J. Chromatogr. B 779 (2002) 157–162. [10] Y.M. Liu, H.B. Mu, Y.L. Zheng, C.Q. Wang, Y.H. Chen, F.R. Li, J.H. Wang, J.K. Cheng, J. Chromatogr. B 855 (2007) 280–285. [11] X.M. Li, F. Zhang, S.S. Zhang, J. Sep. Sci. 31 (2008) 336–340. [12] Z. Dai, S.N. Serban, H.X. Ju, N.E. Murr, Biosens. Bioelectron. 22 (2007) 1700–1706. [13] W.C. Tsai, I.C. Lin, Sens. Actuators B 106 (2005) 455–460. [14] R.Z. Zhang, B.G. Pan, H.N. Wang, J.M. Dan, C.L. Hong, H.L. Li, RSC Adv. 5 (2015) 38176–38182. [15] H. Yang, Z. Li, X. Wei, R. Huang, H. Qi, Q. Gao, C. Li, C. Zhang, Talanta 111 (2013) 62–68. [16] N. Hui, X. Sun, Z. Song, S. Niu, X. Luo, Biosens. Bioelectron. 86 (2016) 143–149. [17] N. Li, H. Ma, W. Cao, D. Wu, T. Yan, B. Du, Q. Wei, Biosens. Bioelectron. 74 (2015) 78–791. [18] Y. Wang, X. Li, W. Cao, Y. Li, H. Li, B. Du, Q. Wei, Talanta 129 (2014) 411–416. [19] A.C. Moser, D., S. Hage, Electrophoresis 29 (2008) 3279–3295. [20] A.J. Gaudry, Y.H. Nai, R.M. Guijt, M.C. Breadmore, Anal. Chem. 86 (2014) 3380–3388. [21] S. Gotz, U. Karst, Anal. Bioanal. Chem. 387 (2007) 183–192. [22] S. Zhao, X. Li, Y.M. Liu, Anal. Chem. 81 (2009) 3873–3878. [23] S. Zhao, Y. Huang, M. Shi, Y.M. Liu, J. Chromatogr. A 1216 (2009) 5155–5159.
4. Conclusions In this study, a novel MCE-CL immunoassay method was established for the detection of tumor marker AFP. The MCE-CL immunoassay was conducted after AFP reacted with excessive HRP labeled
111
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
A novel microchip electrophoresis-based chemiluminescence immunoassay for the detection of alpha-fetoprotein in human serum ⁎
Jingwen Liu, Jingjin Zhao , Shuting Li, Liangliang Zhang, Yong Huang, Shulin Zhao
MARK
⁎
State Key Laboratory for the Chemistry and Molecular Engineering of Medicinal Resources, Key Laboratory of Ecology of Rare and Endangered Species and Environmental Protection of Ministry Education, Guangxi Normal University, Guilin 541004, China
A R T I C L E I N F O
A BS T RAC T
Keywords: Microchip electrophoresis Chemiluminescence detection Immunoassays Tumor marker Alpha-fetoprotein
A sensitive immunoassay method based on microchip electrophoresis chemiluminescence (MCE-CL) detection technology was developed for the detection of tumor marker alpha-fetoprotein (AFP). This method adopts the non-competitive immunoassay mode, and was conducted after AFP reacted with excessive horseradish peroxidase (HRP) labeled monoclonal antibody. The extreme pH value was adopted in the electrophoresis buffer solution. The use of brij 35 as an additive of electrophoresis buffer increased dramatically the resolution (Rs) and the reproducibility of the analysis. Under the optimized experimental conditions, effective separation of the immune complex Ag-Ab* and free Ab* was achieved within 60 s. The peak height of the immune complex Ag-Ab* was taken as quantification of AFP. Good linearity was observed within AFP concentrations ranging from 10 ng/mL to 150 ng/mL, and the detection limit was found to be 7.2 ng/mL (1.0×10−10 M). The present method was successfully applied for the detection of AFP in human serum from both healthy and cancer patients, and the AFP levels in the both were found be in the range of 16.5–23.4 ng/mL and 416.2–825.4 ng/ mL, respectively.
1. Introduction Alpha-fetoprotein (AFP), an oncofetal glycoprotein, is commonly used as a marker for hepatocellular carcinoma (HCC). Under healthy conditions, AFP comes from embryonic hepatocytes and its concentration begins to decline in fetal blood approximately two weeks after birth. In adult serum, the normal content of AFP is low than 25 ng/mL [1]. An elevated AFP concentration in serum may be an early indication of HCC, hepatoblastoma, and germ cell tumors [2,3]. Therefore, AFP is considered to be a specific clinical marker for the diagnosis of primary liver cancer. Similarly, AFP concentrations increase with disease progression in patients with pancreatic cancer, lung cancer, and hepatic cirrhosis [4,5]. Periodic detection of AFP concentration in human bodily fluids for the clinical analysis of treatment outcomes, prognosis assessment, prediction of recurrence and metastasis is very important. Currently, the commonly used AFP detection methods include radioimmuno-assay [6], enzyme-linked immunoassay (ELISA) [7], chemiluminescence immune- assay (CL-IA) [8], capillary electrophoresis immunoassay (CE-IA) [9–11] and immunosensors [12,13]. These methods usually require long analysis times and complex liquid handling procedures. Additionally, these methods utilize expensive
⁎
antibody reagents, which results in high costs, ultimately restricting their widespread application. In order to improve the sensitivity of AFP detection methods, a variety of nanomaterials including grapheme [14], carbon nanotubes [15], gold nanoparticles [16], PdNi nanoparticles [17] and mesoporous silica [18] have been utilized for signal amplification in the methods. In above mentioned methods, highly sensitivity was obtained, however, most of these methods require tedious surface modification of electrode or liquid handling procedures, which make the assays time-consuming. Therefore, it is necessary to develop a new immunoassay method with fast detection and high accuracy for clinical AFP detection. The immunoassay, which is a detection method based on specific responses that arise from the binding of an antibody with an antigen, has high selectivity and sensitivity, and is therefore extensively used in clinical diagnoses and biochemical analyses [19]. However, routine immunoassay requires extensive amounts of expensive reagents, long analysis time and complex operation procedures. Microchip electrophoresis (MCE) is a new separation assay technique developed after capillary electrophoresis (CE). MCE has the advantages of high efficiency, fast analysis, less sample and reagent consumption, and high automation and integration degree; therefore, it has been widely adopted in chemical analyses [20].
Correspondence to: College of Chemistry and Pharmacy, Guangxi Normal University, Guilin 541004, China. E-mail addresses: [email protected] (J. Zhao), [email protected] (S. Zhao).
http://dx.doi.org/10.1016/j.talanta.2016.12.038 Received 20 September 2016; Received in revised form 13 December 2016; Accepted 18 December 2016 Available online 19 December 2016 0039-9140/ © 2016 Elsevier B.V. All rights reserved.
Talanta 165 (2017) 107–111
J. Liu et al.
2.3. Immunoreaction procedure
Chemiluminescence (CL) detection is a highly sensitive detection system for MCE. When MCE-CL is used in conjunction with immunoassay, the resulting analyses have high selectivity of immunoassay and high sensitivity of CL detection. Therefore, it has been shown that the combination of microfluidic systems and immunoassay is a highly effective separation assay technique [21]. However, no study has reported the use of MCE-CL for AFP detection. In the present study, a new non-competitive immunoassay method based on the MCE-CL technology was established for AFP detection. Following the reaction of AFP antigen (Ag) with an excess of HRP labeled anti-AFP antibody (Ab*), MCE-CL method was used to conduct AFP detection. Free Ab* and the immune complex (Ag-Ab*) were separated within 60 s. The method has been successfully used for the detection of trace AFP in human serum, confirming the utility of the assay.
The non-competitive mode was adopted for the immunoassay. The principle binding process was as follows:
Ag + Ab*(excess)—Ag − Ab* + Ab* First, 10 μL of AFP antigens with different concentrations or the serum sample were mixed with 10 μL of 0.25 μg/mL HRP labeled mouse anti-AFP monoclonal antibody solution in a micro-centrifuge tube. The mixture was then diluted to 50 μL using the phosphate buffer (pH 7.4) and incubated at 37 °C for 40 min. The solution was then used for MCE-CL analysis.
2.4. MCE procedure Prior to and between any two test of electrophoreses, all chip channels were rinsed thoroughly with 0.1 M NaOH, water and the buffer solution for 10 min each. All the channels were filled with the buffer solution under a vacuum negative pressure of ∼50 mm Hg, and then, each corresponding reservoir was filled with different solutions for separation. Samples were injected in pinched mode. A voltage of 650 V was applied to S, while SW was connected to the ground. Meanwhile, B and BW were applied with 250 V and 400 V, respectively. The injection time was 15 s. During separation, 2600 V was applied to B, and BW was connected to the ground. Meanwhile, 1550 V was applied to S and SW to avoid samples leaking into the channels during separation. R was applied with 500 V. The sample components migrated to the Y-intersection, where they mixed with oxidizer solution to produce CL signal, which was then collected with an object lens (placed at the Y-intersection of the channels) and transferred to a photomultiplier tube (PMT). Finally, the signal was recorded via a HW2000 chromatography work station (Zhejiang University Star Information Technology, Hangzhou, China).
2. Experimental 2.1. Reagents and solutions The AFP immunoassay kit and the horseradish peroxidase (HRP) labeled mouse anti-AFP monoclonal antibody were purchased from Zhengzhou Bioassay Biotechnical Co., Ltd. (Zhengzhou, China). Luminol was purchased from Fluka (Buchs, Switzerland). Brij 35, iodophenol (PIP) and H2O2 were purchased from Shanghai Chemical Company (Shanghai, China). All other chemicals were of analytical reagent grade and used without further purification. Water was purified by employing a Milli-Q plus 185 equip from Millipore (Bedford, MA), and used throughout the work. AFP and HRP-mouse anti-AFP monoclonal antibody were diluted using 20 mM phosphate buffer (pH 7.2). The electrophoresis buffer solution was 10 mM Na3PO4 (pH 10.2) and contained 0.004% (w/w) Brij 35 and 1.0 mM luminol. The oxidizer solution was 45 mM NaHCO3 (pH 9.0) and contained 90 mM H2O2 and 1.2 mM PIP.
2.2. Experimental devices
3. Results and discussion
The MCE-CL detection system was self-constructed in the lab [22]. The glass/polydimethylsiloxane (PDMS) microchip with an expanded Y-shape CL detection pool was designed according to the literature [23] (Scheme 1). The chip had an area of 9.5 cm by 2 cm. The top width of the channel measured 65 µm (except at R and BW, where the distance was 250 µm) and the depth measured 25 µm. The effective length of the separation channel was 60 mm. The distance from the buffer reservoir (B) to the T-intersection was 5 mm. The distance between the separation channel and the sample reservoir (S) or the sample waste reservoir (SW) was 5 mm. The two T-intersections were 60 µm. The sampling volume was calculated to be about 190 pL. The distance from the oxidizing agent reservoir (R) to the Y-intersection was 1.5 cm, and the Y intersection was 1.2 cm from the buffer waste reservoir (BW).
3.1. Optimization of the CL conditions The pH of the oxidizer solution is a major factor affecting CL intensity. In this experiment, 45 mM NaHCO3 was adopted as the buffer to study the influence of varying pH values (pH 8.0–10.0) on the CL intensity. The results are shown in Fig. 1a. It can be seen that the CL intensity first increased and then decreased with continuously increasing pH. When the pH was 9.0, the highest CL intensity was observed. Therefore, 9.0 was selected as optimal pH value of NaHCO3 solution. H2O2 was used as the oxidizing agent in the experiment; its concentration also plays a major role in CL intensity. The influence of the H2O2 concentration on the CL intensity was studied within the range of 50–120 mM. The results indicate that the CL intensity increased with the increase of H2O2 concentration between 50 and 90 mM. When the concentration was above 90 mM, CL intensity showed a downward trend as the H2O2 concentration increased. Therefore, 90 mM was taken as the ideal H2O2 concentration. The PIP is a sensitizing agent that is extensively used in the luminol-H2O2-HRP CL system. PIP greatly increases the CL intensity of the system; therefore, its concentration has notable influence on the luminescence intensity. The influence of the PIP concentration on the CL intensity was investigated between 0.6 and 1.4 mM; the results are shown in Fig. 1b. As can be seen, the CL intensity increased gradually with increasing PIP concentration from 0 to 1.2 mM. After the concentrations are above 1.2 mM, the CL intensity decreased with increasing PIP concentration. Hence, 1.2 mM was selected as the optimal PIP concentration.
Scheme 1. Schematic diagram of the layout of the glass/PDMS microchip. S: sample reservoir; B: buffer reservoir; SW: sample waste reservoir; BW: buffer waste reservoir; R: oxidizing agent reservoir.
108
Talanta 165 (2017) 107–111
J. Liu et al.
Fig. 1. Effects of pH value for oxidizer solution (a), PIP concentration (b) and incubation time (c) on the CL intensity; the effect of brij 35 concentration on the Rs (d). The electrophoresis buffer was 10 mM phosphate buffer (pH 10.2) containing 0.002–0.006% brij 35 and 1.0 mM luminol. The oxidizer solution was 45 mM NaHCO3 (pH 8.0–10.0) solution containing 90 mM H2O2 and 0.6–1.6 mM PIP. The solution was incubated at 37 °C from 25 min to 60 min. The sampling volume was about 190 pL.
excessive Joule heat resulted in reduced separation efficiency. Additionally, high pH levels may cause decomposition of the immune complex. Hence, 10.2 was chosen as the optimal pH value for the buffer solution. The methods for inhibiting protein adsorption usually include permanent coating and dynamic coating. In order to further increase the separation efficiency, the nonionic surfactant brij 35 was used as a buffer additive. Its influence on the separation was investigated within the concentration range of 0.001–0.006% (w/w), and the results indicate that the separation efficiency was improved dramatically with concentration increased (Fig. 1d); however, the retention time gradually increased. Considering the Rs and analysis time, 0.04% (w/w) was selected as the optimal concentration of brij 35. A phosphate solution was chosen in the experiment as the electrophoresis buffer solution. The influence of the buffer concentration on Rs was investigated within the range of 4–14 mM, and the results indicate that the Rs increased with the increasing phosphate concentration. When the concentration exceeded 10 mM, the Rs started to drop; this observation may be due to Joule heating, which results from the high phosphate concentration. Therefore, 10 mM was selected as the optimal phosphate concentration. Additionally, the influence of the separation voltage was also studied. With rising separation voltage, the retention time of the labeled antibody and the immune complex was shortened, resulting in quicker analysis time and narrow peaks. When the voltage was above 2600 V, the Rs decreased rapidly. By considering the analysis speed and Rs, 2600 V was selected as the optimal separation voltage.
3.2. Optimization of the immunoreaction time The immunoreaction time is an essential factor of MCE-CL analysis as it directly determines the sensitivity of the method. The peak height of the immune complex (CL intensity) was taken as the reference to optimize the immunoreaction time of the labeled antibody and antigen; the results are shown in Fig. 1c. It can be seen that with prolonged reaction time, the CL intensity increased rapidly. When the reaction time was more than 40 min, the CL intensity reached a plateau, indicating a balanced condition for the antibody and antigen combination. Therefore, the optimal reaction time was set at 40 min. 3.3. Selection of electrophoresis separation conditions To achieve accurate AFP detection, the factors that affect electrophoresis separation were optimized, including the concentration and the pH of the electrophoresis buffer solution, the concentration of brij 35, and the separation voltage. The pH of the electrophoresis buffer solution is a key factor for electrophoresis separation. In the experiment, the extreme pH value was adopted to perform electrophoresis separation. A 10 mM phosphate solution was used as the buffer. Different pH values ranging from 9.0 to 11 were used to investigate the influence on separation of the immune complex and the free labeled antibody. The results indicate that the resolution (Rs) increased with increasing pH at pH < 10.2. Conversely, at pH > 10.2, the resolution decreased as the pH increased. The possible reason for this outcome was that the high pH levels led to enhanced ionic strength; thus, the large electrophoresis current and 109
Talanta 165 (2017) 107–111
J. Liu et al.
Fig. 2. Electropherograms obtained from separating the immunoreactions solution: (a) Ab* solution, (b) a mixture solution containing Ab* and 80 ng/mL of AFP, (c) a normal human serum sample, (d) a cancer patient serum sample (diluted 10 times). Peak 1 derived from free Ab*, peak 2 derived from the Ab*-Ag complex. The electrophoresis buffer was 10 mM phosphate buffer (pH 10.2) containing 1.0 mM luminol and 0.004% brij 35. The oxidizer solution was 45 mM NaHCO3 (pH 9.0) containing 90 mM H2O2 and 1.2 mM PIP. The sampling volume was about 190 pL.
3.4. MCE-CL noncompetitive immunoassay of AFP Under the optimal experimental conditions, various concentrations of AFP solutions were tested. The typical electropherograms are shown in Fig. 2. In Fig. 2a, the concentration of AFP antigen was 0, therefore only showed one peak (1), which corresponds to Ab*. In Fig. 2b, 80 ng/ mL of the AFP antigen was added; it thus shows two peaks 1 and 2, corresponding to Ab* and Ab*-Ag complex, respectively. The height of peak 2 (CL intensity) increased with increasing AFP antigen concentration. Therefore, the height of peak 2 was used to quantification of AFP. The results showed that AFP had good linearity within the concentration range of 10–150 ng/mL (Fig. 3), with the linear correlation coefficient (R2) being 0.9973. The linear regression equation was:
H = 0.0405 C + 0.2686, R2 = 0.9973 In this equation, H is the CL intensity (mV) and C is the concentration (ng/mL) of the AFP. Taking the signal to noise ratio equal to 3 (S/N=3), the detection limit was found to be 7.2 ng/mL (1.0×10−10 M). To ensure the reproducibility, the MCE-CL analysis was repeated seven times for the immune reaction mixture, and the peak heights and retention times were recorded. The reproducibility of the method was assessed by relative standard deviations (RSDs) of the peak heights and the retention times. The results showed that the RSDs of the peak heights and the retention times were lower than 2.7% and 1.8%, respectively. The results indicated a good reproducibility of the MCE-CL immunoassay method. Thus, it can be applied for the
Fig. 3. The calibration curve for AFP detection. Experiment conditions were as in Fig. 2.
detection of AFP in human serum. 3.5. Specificity for AFP detection To evaluate the specificity of proposed MCE-CL method for AFP detection, we challenged the system with AFP and several nonspecific 110
Talanta 165 (2017) 107–111
J. Liu et al.
monoclonal antibody. The extreme pH value was adopted in the experiment. The use of brij 35 as an additive of electrophoresis buffer dramatically increased the Rs and the reproducibility of the analysis. The proposed method has been used for AFP detection in human serum for both healthy people and cancer patients. This method has several important features. First, in contrast to the conventional immunoassay technique, the MCE-CL immunoassay is simple and does not need complex liquid handling procedures. Second, the use of MCE technology causes a rapid analysis, in addition to the time needed for the immunoreaction, the separation and detection can be finished within 1 min. Third, the proposed MCE-CL immunoassay exhibits an appropriate sensitivity for the determination of AFP content in human serum, which decreases the detection error from excessive concentration or dilution of sample. Thus, we anticipate that this method can be applied readily to detect AFP content in human serum for early diagnosis of cancers.
Table 1 Detection results and recovery of AFP in human serum samples. Sample
Found (ng/mL)
RSD (%, n=5)
Added (ng/mL)
Total found (ng/mL)
Recovery (%)
Normal 1 Normal 2 Normal 3 Normal 4 Patient 1 Patient 2 Patient 3
19.6 23.4 18.8 16.5 416.2 825.4 762.8
4.6 3.8 4.4 2.8 3.2 3.6 3.1
20.0 20.0 20.0 20.0 40.0 80.0 80.0
38.72 44.25 38.07 36.18 80.42 163.86 152.53
95.6 104.2 96.4 98.4 97.0 101.7 95.3
proteins such as immunoglobulin G (IgG), thrombin (Tb) and bovine serum albumin (BSA). The results indicate that only the AFP could lead to appearance of peak 2, no obvious peak 2 was observed in presence of other nonspecific proteins, and the CL intensity of peak 2 was same with that of blank, which shown a high specificity of proposed method for AFP detection.
Acknowledgments This work was supported by the National Natural Science Foundations of China (Nos. 21327007 and 21465005), Natural Science Foundations of Guangxi Province (Nos. 2014GXNSFBA118041), IRT1225, as well as BAGUI Scholar Program.
3.6. Determination of AFP content in human serum Under the above experimental conditions, the human serum samples collected from both liver cancer patient and healthy volunteer were analyzed. Fig. 2 shows typical electropherograms for undiluted serum of healthy volunteer (Fig. 2c) and 10 times diluted serum of liver cancer patient (Fig. 2d). The AFP contents in the serum of the healthy volunteer were in the range 16.5–23.4 ng/mL, which was consistent with the reported reference value (25 ng/mL) for normal people [1]. In contrast, the AFP content in serum of cancer patients varied from 416.2 ng/mL to 825.4 ng/mL. Thus, it is apparent that the AFP content in the serum of cancer patients is significantly higher than that of healthy people, which was in good agreement with previous studies [3,10]. Because the concentration of AFP in the serum of healthy people was two orders of magnitude lower than that of cancer patients, the sample solution needs to be diluted 10 times when detecting the concentration of AFP in cancer patient serum. Although large dilution of sample could lead to a source of error in measurements, this error can be ignored in trace analysis. Hence, the proposed method is feasible for clinical diagnosis of cancers. In addition, in order to determine the reliability of the method, the recovery rate and reproducibility were calculated. The results are listed in Table 1. The RSDs of 5 measurements of human serum samples ranged from 3.1% to 4.6%, and the recoveries of AFP in the human serum samples were in the range 95.3–105.5%.
References [1] Z.F. Fu, C. Hao, X.Q. Fei, H.X. Ju, J. Immunol. Methods 312 (2006) 61–67. [2] Y. Tsuchida, Y. Endo, S. Saito, M. Kaneko, K. Shiraki, K. Ohmi, J. Pediatr. Surg. 13 (1978) 155–156. [3] J.T. Wu, L.D. Book, K. Sudar, Pediatr. Res. 15 (1981) 50–52. [4] C.G. Bergstrand, Acta Paediatr. Scand. 75 (1986) 1–9. [5] D. Bader, A. Riskin, O. Vafsi, A. Tamir, B. Peskin, N. Israel, R. Merksamer, H. Dar, M. David, Clin. Chim. Acta 349 (2004) 15–23. [6] J.E. Fowler, I.A. Sesterhenn, F.K. Mostofi, Urology 33 (1989) 74–77. [7] J.A. Brochot, I.W. Siddiqi, Anal. Chim. Acta 224 (1989) 329–337. [8] Z.J. Yang, H. Liu, C. Zong, F. Yan, H.X. Ju, Anal. Chem. 81 (2009) 5484–5489. [9] R.Y. Wang, X.N. Lu, W.Y. Ma, J. Chromatogr. B 779 (2002) 157–162. [10] Y.M. Liu, H.B. Mu, Y.L. Zheng, C.Q. Wang, Y.H. Chen, F.R. Li, J.H. Wang, J.K. Cheng, J. Chromatogr. B 855 (2007) 280–285. [11] X.M. Li, F. Zhang, S.S. Zhang, J. Sep. Sci. 31 (2008) 336–340. [12] Z. Dai, S.N. Serban, H.X. Ju, N.E. Murr, Biosens. Bioelectron. 22 (2007) 1700–1706. [13] W.C. Tsai, I.C. Lin, Sens. Actuators B 106 (2005) 455–460. [14] R.Z. Zhang, B.G. Pan, H.N. Wang, J.M. Dan, C.L. Hong, H.L. Li, RSC Adv. 5 (2015) 38176–38182. [15] H. Yang, Z. Li, X. Wei, R. Huang, H. Qi, Q. Gao, C. Li, C. Zhang, Talanta 111 (2013) 62–68. [16] N. Hui, X. Sun, Z. Song, S. Niu, X. Luo, Biosens. Bioelectron. 86 (2016) 143–149. [17] N. Li, H. Ma, W. Cao, D. Wu, T. Yan, B. Du, Q. Wei, Biosens. Bioelectron. 74 (2015) 78–791. [18] Y. Wang, X. Li, W. Cao, Y. Li, H. Li, B. Du, Q. Wei, Talanta 129 (2014) 411–416. [19] A.C. Moser, D., S. Hage, Electrophoresis 29 (2008) 3279–3295. [20] A.J. Gaudry, Y.H. Nai, R.M. Guijt, M.C. Breadmore, Anal. Chem. 86 (2014) 3380–3388. [21] S. Gotz, U. Karst, Anal. Bioanal. Chem. 387 (2007) 183–192. [22] S. Zhao, X. Li, Y.M. Liu, Anal. Chem. 81 (2009) 3873–3878. [23] S. Zhao, Y. Huang, M. Shi, Y.M. Liu, J. Chromatogr. A 1216 (2009) 5155–5159.
4. Conclusions In this study, a novel MCE-CL immunoassay method was established for the detection of tumor marker AFP. The MCE-CL immunoassay was conducted after AFP reacted with excessive HRP labeled
111