Clinica Chimica Acta 431 (2014) 113–117
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Chemiluminescence immunoassay based on microfluidic chips for α-fetoprotein Fei Fan a,1, Haiying Shen b,1, Guojun Zhang a, Xingyu Jiang b,⁎, Xixiong Kang a,⁎ a b
Laboratory Diagnosis Center, Beijing Tiantan Hospital, Capital Medical University, Beijing 100050, China CAS Key Lab for Biological Effects of Nanomaterials and Nanosafety, National Center for Nanoscience and Technology, Beijing 100190, China
a r t i c l e
i n f o
Article history: Received 7 August 2013 Received in revised form 22 January 2014 Accepted 4 February 2014 Available online 12 February 2014 Keywords: Microfluidic chips Chemiluminescence Sandwich immunoassay α-Fetoprotein
a b s t r a c t Background: Conventional immunoassays are labor intensive, time consuming, expensive and require large pieces of equipment for detection. In an effort to overcome these shortcomings, this study established an immunoassay method of alpha fetoprotein (AFP) in serum in combination with the microfluidic chip technology. Methods: A sandwich immunoassay approach was applied to detect AFP based on microfluidic chips and the chemiluminescence as detection signal. The chip used in this method was composed of a polydimethylsiloxane (PDMS) microchannel layer over a PDMS base layer. Result: AFP concentration and chemiluminescence intensity were linearly correlated over the concentration ranging from 12.5 to 200 ng/ml, and a detection limit as low as 1.5 ng/ml using this method. The coefficients of variation were 9.91% and 11.4% for the within- and between-run assays, respectively. More than 50 clinical samples were tested and the results obtained for this method strongly correlated with Roche's electrochemiluminescence (ECL) kit. Conclusions: The proposed method offers a reliable, simple, reagent safe and inexpensive analytical platform for the determination of AFP in serum, and promotes the development of high throughput screening and point-ofcare testing (POCT) diagnostics in clinical practice. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Alpha fetoprotein (AFP) is a 70 kDa glycoprotein, containing approximately 4% carbohydrate, and is produced primarily by fetal liver during gestation [1]. Normally, AFP serum concentration falls sharply after birth and its synthesis in adult life is repressed. However, N70% of hepatocellular carcinoma patients have a high AFP serum concentration because of tumor excretion [2]. AFP is considered a tumor marker of hepatocellular carcinoma, but elevated serum levels of AFP also occur in patients with liver disease (hepatitis, cirrhosis) or in normal pregnancy. Therefore, the detection of serum AFP plays a very important role in early screening, clinical diagnosis, and prognosis of disease, particularly for liver cancer [3,4]. Since the first quantitative serum assay for AFP was established by Ruoslahti
⁎ Corresponding authors. Tel.: +86 10 67096875. E-mail address: [email protected] (X. Kang). 1 Fan, F. and Shen, H. contributed equally to this work. 0009-8981/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cca.2014.02.003
et al. [5], testing for AFP has become commonplace in clinical and physical examinations in recent years. At present, the most commonly used methods for the quantitative determination of AFP are the enzyme-linked immunosorbent assay (ELISA) [6,7], radioimmunoassay [8], and time-resolved fluoroimmunoassay [9]. Although the assays are performed solely within the wells of microtiter plates, the assays can be complicated, time consuming, expensive, wasteful, and require large pieces of equipment for detection. In addition, radioimmunoassays require the disposal of radiolabels, and the lanthanide labels used for fluoroimmunoassays are expensive and susceptible to outside interference [10]. The area of miniaturized or microfluidic analysis systems, also called “micro total analysis systems (μTAS) or lab-on-a-chip (LOC)”, has gained increased popularity [11,12]. In the mid-1990s, a new analysis platform was established on the microfluidic chip [13], and to date it has been widely applied in the field of biological research, including cell culture [14,15], single cell detection [16], genetic analysis [17,18], and immunoassays [19,20]. Several review papers on microfluidic bioanalysis discussed the necessity to miniaturize ELISA and other types of immunoassays for biological research and clinical diagnosis [21–23]. Because of the advantages of microminiaturization, microfluidic technology can be utilized to achieve these goals. In the current study,
114
F. Fan et al. / Clinica Chimica Acta 431 (2014) 113–117
we combine microfluidics, chemiluminescence and immunoassay technology to devise a novel method for measuring AFP concentration in serum.
2. Materials and methods 2.1. Reagents and samples AFP calibrators, AFP monoclonal antibodies (Ab 1 ) and HRPlabeled AFP monoclonal antibodies (Ab 2 ) were from Beijing Keybiotech Co., Ltd. Ab1 was diluted in 0.1 mol/l phosphate buffered saline (PBS, pH 7.4). The calibrators and Ab2 were diluted in 3 g/dl bovine serum albumin solution (BSA; Merck), which was used as the blocking agent. Chemiluminescent substrate reagent kits were obtained from Millipore. A Roche ECL kit was from Roche Diagnostics Shanghai Co. Ltd. PBS was used as the wash solution. All human serum used in this study which were from Beijing Tiantan hospital had been approved by IRB, and we have obtained informed consents from all the research objects. The 56 serum objects included 38 health individuals (which include 32 pregnant women), 11 liver disease subjects and 7 other disease subjects. All serum samples were analyzed directly without any pretreatment.
2.2. Methods 2.2.1. Preparation of microfluidic chips The microfluidic chip has 2 layers: an upper microchannel layer and a lower PDMS layer used as a base for protein immobilization. Polydimethylsiloxane (PDMS) [24,25] was used to make the microfluidic chip. The microchannel layer contained seven channels and was prepared as follows. First, the PDMS base and curing agent (Sylgard 184, Dow Corning Inc.) were mixed 10:1, and the mixture degassed to remove bubbles. Then, the mixture was poured into the mold and hardened in an oven (Shanghai Yiheng Technical Co. Ltd.) at 80 °C for 25 min [26]. Once cooled the curing PDMS layer was removed from the mold and inlet and outlet portals made using a syringe needle. The height and width of the channels were maintained at 500 μm.
2.2.2. Procedures This method consists of four stages: (i) Ab1 was delivered via microchannels for immobilization on the PDMS base, as shown in Fig. 1a. (ii) The microchannel immobilized on the PDMS base was peeled off and another one was placed perpendicularly to the immobilized antibody strips. Then, nonspecific binding sites were blocked with BSA. (iii) The antigen, either an AFP calibrator or test sample, was added, and reacted with an excess of Ab1 . The antigen–antibody (Ab1–AFP) complexes were formed at the intersection of the two microchannels. Then we washed the PDMS base via the microchannels 3 times with 20 μl PBS buffer each time with the procedure of inserting PBS in one side of portal and sucking it out on the other side, so unbound antibody and antigen removed by washing. (iv) Ab2 was introduced into the microchannels and the double sandwich immunocomplex (Ab1–AFP–Ab2) formed (Fig. 1c). Unbound reagents were removed by washing with PBS and the chemiluminescent substrate was injected into the chip. The chemiluminescence images were captured using an imager designed by the National Center for Nanoscience and Technology (Beijing, China), and the images analyzed using ImageJ software [27]. Because AFP concentration and chemiluminescence intensity were linearly correlated, the concentration of serum AFP was calculated from the calibration curve or regressive equation.
Fig. 1. Schematic of the microfluidic AFP immunoassay. (a) Ab1 immobilization: The 2-layer microfluidic chip comprises a PDMS microchannel (layer one) on a PDMS base (layer 2). Ab1 was introduced via microchannels, resulting in (b) the formation of parallel antibody strips on the PDMS base. (c) Immune reaction: a second microchannel was placed perpendicular to the immobilized antibody strips, and the antigen and Ab 2 introduced sequentially. The ECL substrate was added and the chemiluminescent signal measured (d).
2.3. Statistical analysis All statistical analyses were performed using SPSS 13.0 software. The sera AFP concentrations in healthy controls were expressed as the mean ± standard deviation (mean ± SD). 3. Results and discussion 3.1. Effect of Ab1 concentration on chemiluminescence intensity At varying Ab1 concentrations, the chemiluminescence intensity was different. The other experiment conditions were controlled, and Ab1 was serially diluted from 10 to 160 μg/ml; when the Ab1 concentration was between 40 and 160 μg/ml, the detection range of AFP fulfilled the requirements for use in a clinical immunoassay (Fig. 2a). By taking into consideration (i) the linear range of width of the proposed method,
F. Fan et al. / Clinica Chimica Acta 431 (2014) 113–117
115
3.2. Effect of the Ab2 dilution order on chemiluminescence intensity At a constant AFP concentration of 6.25 ng/ml, we found that if the Ab2 dilution order is greater than 1:200, the chemiluminescence intensity was difficult to detect. In consideration of reagent consumption, a 1:200 dilution ratio for Ab2 was suitable for the microfluidic AFP immunoassay developed in this study (Fig. 3b). 3.3. Standard curve and detection limit The standard curve was prepared by using a working standard solution of different concentrations of AFP. The concentration series of each standard was made at 12.5, 25, 50, 100 and 200 ng/ml, respectively, because AFP concentration and chemiluminescence intensity were linearly correlated at this range. The lower limit of detection (LOD) for the AFP assay was determined by detecting a sample signal three times higher than the background noise, and the LOD value in the tests was 1.5 ng/ml. 3.4. Sample detection
Fig. 2. Optimization of Ab1 and Ab2 concentrations. (a) Ab1 was serially diluted from 160 to 10 μg/ml. The AFP calibrator was diluted from 400 to 12.5 ng/ml. (b) Two-fold serial dilutions (1:50 to 1:800) of Ab2 were prepared, and the immunoassay performed using a constant AFP calibrator concentration of 6.25 ng/ml.
(ii) the practicality for clinical use, and (iii) reagent consumption, an Ab1 concentration of 40 μg/ml was determined to be the most favorable for this assay.
The clinical samples and serially diluted AFP calibrators (12.5 to 200 ng/ml) were introduced into parallel microchannels as shown in Fig. 3a. Chemiluminescent signals; uniform and square in shape were observed. No cross-contamination between different microchannels was detected. A blank control (last channel) was included in the assay for the monitoring of non-specific adsorption; no signal was detected. Fig. 3b represents the corresponding calibration curve for the immunoassay in Fig. 3a. Log2 AFP concentration/12.5 was presented on the X-axis, and the corresponding chemiluminescence intensity of the AFP calibrators on the Y-axis. A strong linear relationship existed between Log2 concentration/12.5 ng/ml and grayscale values, and the linear regression equation was: Y = 994.30213 + 1445.4315X; R2 = 0.99403. Therefore, the sample in Fig. 3a was determined as 21.84 ng/ml according to the formula mentioned above. 3.5. Within- and between-run coefficients of variation (CVs) Within- and between-run coefficients of variation (CVs) were determined to evaluate the precision and repeatability, respectively of the assay. The same sample was simultaneously analyzed 49 times for within-chip assay and 2 times each day for 11 consecutive working
Fig. 3. Immunoassay results for AFP detection using microfluidic chips. (a) Different concentrations of AFP calibrators were introduced into the first five microchannels; the clinical sample and blank control were injected into the two remaining microchannels. The gray spots on the microchip demonstrated the immune reactive signal for AFP. (b) Calibration curve of AFP concentration from 12.5 to 200 ng/ml. The chemiluminescence intensity across three spots for each AFP concentration was determined. The plotted values are the mean ± SD of the triplicate measurements. Chemiluminescence intensity of AFP spots and AFP concentration exhibited a significant linear relationship (R2 = 0.99403).
116
F. Fan et al. / Clinica Chimica Acta 431 (2014) 113–117
days for between-chip assays. The CVs were 9.91% for within-chip and 11.4% for between-chip assays, respectively. 3.6. Interference studies Different concentrations of hemoglobin and bilirubin in serum were added in serum to observe its effect on the gray value of AFP. Based on a relative error of b 5%, we found that AFP detection in serum by this approach was free of the interference of hemolytic and jaundice, when hemoglobin concentration was less than 1.5 g/dl and bilirubin concentration was below 60 mg/dl. 3.7. Comparison of methods The proposed method has been applied to evaluate AFP in 56 human serum samples, and the results were compared with those obtained by the commercial Roche method, a technique commonly used in hospitals. As can be seen in Fig. 4, the serum AFP concentrations measured using microfluidic chips strongly correlated with those obtained using Roche's electrochemiluminescence (ECL) kit; Y = −4.85987 + 0.92643X, and the coefficient was 0.9812. Then, we tested a sample 3 times with this method and 3 times with the Roche ECL, and the results showed that the bias of microfluidic chip immunoassay was 3.9%, which was lower than the Roche ECL. 3.8. Reference value We measured AFP concentration in sera from 30 fasting, apparently healthy subjects, and the mean concentration of AFP was 2.51 ± 1.51 ng/ml. 3.9. The characteristics and advantages of this novel AFP microfluidic immunoassay With the introduction of microfluidic chips, this immunoassay provides significant advantages over conventional immunoassays, including (i) ease of operation: the operating procedure is simple, does not require an experienced technician, nor does it require external power input; (ii) reduced assay time: the total detection time was 40 min; most commonly used immune diagnostic methods can detect only 1 protein/hormone in an individual assay and usually require a minimum of 2–3 h to complete; (iii) reduced sampling size: this immunoassay requires at most 17 μl of sample, and it is possible to detect multi-markers for a sample simultaneously. One of the practical
Fig. 4. Comparison of analytical parameters of the microfluidic immunoassay and Roche ECL methods. The AFP concentration in 56 human serum samples was determined using each of these methods.
advantages of this method is that if different antibodies were delivered into the parallel microchannels, then this approach can be used to detect multi-biomarkers simultaneously. Therefore, for some precious samples (i.e., neonatal blood and spinal fluid), sampling size is not only reduced, but the number of test items is increased for the same sample volume used in traditional assays; (iv) reduced cost: miniaturization dramatically reduces the consumption of expensive reagents, and the materials can be reused; the reagent costs for the microfluidic immunoassay is only 5% that of an equivalent ELISA; (v) simple instrumentation: the instrument used in this study was small and easily manipulated, the results rapidly obtained, and the system can be adjusted to meet the requirements for POCT and primary hospital diagnostic laboratories. 4. Conclusions In brief, we have developed a reliable, simple, reagent safe and cheap immunoassay method based on microfluidic chips, and applied it to the detection of AFP in serum. Therefore, we consider this technique to be suitable for use in clinical settings. We are committed to improving the assay, and expanding its application into other programs, such as for use with other biomarkers, as well as achieving the synchronous determination of multi-markers for a single sample. Acknowledgements This work was supported by the Instrument Developing Project of the Chinese Academy of Sciences, grant no. YZ201114. References [1] Ruoslahti E, Engvall E. Alpha-fetoprotein. Scand J Immunol 1978;7:1–17. [2] Zhou L, Liu J, Luo F. Serum tumor markers for detection of hepatocellular carcinoma. World J Gastroenterol 2006;12:1175–81. [3] Wright LM, Kreikemeier JT, Fimmel CJ. A concise review of serum markers for hepatocellular cancer. Cancer Detect Prev 2007;31:35–44. [4] Tamura Y, Igarashi M, Kawai H, Suda T, Satomura S, Aoyagi Y. Clinical advantage of highly sensitive on-chip immunoassay for fucosylated fraction of alpha-fetoprotein in patients with hepatocellular carcinoma. Dig Dis Sci 2010;55:3576–83. [5] Ruoslahti E, Seppälä M. Studies of carcino-fetal proteins. 3. Development of a radioimmunoassay for -fetoprotein. Demonstration of -fetoprotein in serum of healthy human adults. Int J Cancer (Journal international du cancer) 1971;8:374–83. [6] Zhang QY, Chen H, Lin Z, et al. Comparison of chemiluminescence enzyme immunoassay based on magnetic microparticles with traditional colorimetric ELISA for the detection of serum α-fetoprotein. JPA 2012;2:130–5. [7] Jin HJ, Gan N, Hou JG, et al. Signal amplification of electrochemical ELISA for the detection of alpha-fetoprotein using core–shell Fe3O4Au nanoparticles as labels. Sens Lett 2012;10:886–93. [8] Silver HK, Gold P, Feder S, Freedman SO, Shuster J. Radioimmunoassay for human alpha 1-fetoprotein. Proc Natl Acad Sci U S A 1973;70:526–30. [9] Lin G, Liu T, Zou L, Hou J, Wu Y. Development of a dual-label time-resolved fluoroimmunoassay for the detection of α-fetoprotein and hepatitis B virus surface antigen. Luminescence 2013;28:401–6. [10] Wang X, Zhang QY, Li ZJ, et al. Development of high-performance magnetic chemiluminescence enzyme immunoassay for α-fetoprotein in human serum. Clin Chim Acta 2008;393:90–4. [11] Gomez R, Bashir R, Sarikaya A, et al. Microfluidic biochip for impedance spectroscopy of biological species. Biomed Microdevices 2001;3:201–9. [12] Arora A, Simone G, Salieb-Beugelaar GB, Kim JT, Manz A. Latest developments in micro total analysis systems. Anal Chem 2010;82:4830–47. [13] Huang H, Zheng X, Zheng J, Pan J, Pu X. Rapid analysis of alpha-fetoprotein by chemiluminescence microfluidic immunoassay system based on super-paramagnetic microbeads. Biomed Microdevices 2009;11:213–6. [14] Tourovskaia A, Figueroa-Masot X, Folch A. Differentiation-on-a-chip: a microfluidic platform for long-term cell culture studies. Lab Chip 2005;5:14–9. [15] Ziółkowska K, Stelmachowska A, Kwapiszewski R, et al. Long-term threedimensional cell culture and anticancer drug activity evaluation in a microfluidic chip. Biosens Bioelectron 2013;40:68–74. [16] Hou S, Zhao L, Shen Q, et al. Polymer nanofiber-embedded microchips for detection, isolation, and molecular analysis of single circulating melanoma cells. Angew Chem Int Ed 2013;52:3379–83. [17] Ferguson BS, Buchsbaum SF, Wu T, et al. Genetic analysis of H1N1 influenza virus from throat swab samples in a microfluidic system for point-of-care diagnostics. J Am Chem Soc 2011;133:9129–35. [18] Pan X, Jiang L, Liu K, et al. A microfluidic device integrated with multichamber polymerase chain reaction and multichannel separation for genetic analysis. Anal Chim Acta 2010;674:110–5.
F. Fan et al. / Clinica Chimica Acta 431 (2014) 113–117 [19] Yang D, Niu X, Liu Y, et al. Electrospun nanofibrous membranes: a novel solid substrate for microfluidic immunoassays for HIV. Adv Mater 2008;20:4770–5. [20] Ng AC, Uddayasankar U, Wheeler A. Immunoassays in microfluidic systems. Anal Bioanal Chem 2010;397:991–1007. [21] Pan W, Chen W, Jiang X. Microfluidic Western blot. Anal Chem 2010;82:3974–6. [22] Whitesides GM. The origins and the future of microfluidics. Nature 2006;442:368–73. [23] Yager P, Edwards T, Fu E, et al. Microfluidic diagnostic technologies for global public health. Nature 2006;442:412–8.
117
[24] Cheng YQ, Yang L, Guo CL, et al. Research progress of materials and fabrication technologies of microfluidic chip. Adv Mater Res 2012;542–543:891–4. [25] Stanton MM, Ducker RE, MacDonald JC, et al. Super-hydrophobic, highly adhesive, polydimethylsiloxane (PDMS) surfaces. J Colloid Interface Sci 2012;367:502–8. [26] Liu Y, Yu J, Do M, et al. Accelerating microfluidic immunoassays on filter membranes by applying vacuum. Biomed Microdevices 2012;14:17–23. [27] Rasband WS. ImageJ. Bethesda, MD: US National Institutes of Health. 1997 [http:// rsb.info.nih.gov/ij(1997): 1997–2007].
Contents lists available at ScienceDirect
Clinica Chimica Acta journal homepage: www.elsevier.com/locate/clinchim
Chemiluminescence immunoassay based on microfluidic chips for α-fetoprotein Fei Fan a,1, Haiying Shen b,1, Guojun Zhang a, Xingyu Jiang b,⁎, Xixiong Kang a,⁎ a b
Laboratory Diagnosis Center, Beijing Tiantan Hospital, Capital Medical University, Beijing 100050, China CAS Key Lab for Biological Effects of Nanomaterials and Nanosafety, National Center for Nanoscience and Technology, Beijing 100190, China
a r t i c l e
i n f o
Article history: Received 7 August 2013 Received in revised form 22 January 2014 Accepted 4 February 2014 Available online 12 February 2014 Keywords: Microfluidic chips Chemiluminescence Sandwich immunoassay α-Fetoprotein
a b s t r a c t Background: Conventional immunoassays are labor intensive, time consuming, expensive and require large pieces of equipment for detection. In an effort to overcome these shortcomings, this study established an immunoassay method of alpha fetoprotein (AFP) in serum in combination with the microfluidic chip technology. Methods: A sandwich immunoassay approach was applied to detect AFP based on microfluidic chips and the chemiluminescence as detection signal. The chip used in this method was composed of a polydimethylsiloxane (PDMS) microchannel layer over a PDMS base layer. Result: AFP concentration and chemiluminescence intensity were linearly correlated over the concentration ranging from 12.5 to 200 ng/ml, and a detection limit as low as 1.5 ng/ml using this method. The coefficients of variation were 9.91% and 11.4% for the within- and between-run assays, respectively. More than 50 clinical samples were tested and the results obtained for this method strongly correlated with Roche's electrochemiluminescence (ECL) kit. Conclusions: The proposed method offers a reliable, simple, reagent safe and inexpensive analytical platform for the determination of AFP in serum, and promotes the development of high throughput screening and point-ofcare testing (POCT) diagnostics in clinical practice. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Alpha fetoprotein (AFP) is a 70 kDa glycoprotein, containing approximately 4% carbohydrate, and is produced primarily by fetal liver during gestation [1]. Normally, AFP serum concentration falls sharply after birth and its synthesis in adult life is repressed. However, N70% of hepatocellular carcinoma patients have a high AFP serum concentration because of tumor excretion [2]. AFP is considered a tumor marker of hepatocellular carcinoma, but elevated serum levels of AFP also occur in patients with liver disease (hepatitis, cirrhosis) or in normal pregnancy. Therefore, the detection of serum AFP plays a very important role in early screening, clinical diagnosis, and prognosis of disease, particularly for liver cancer [3,4]. Since the first quantitative serum assay for AFP was established by Ruoslahti
⁎ Corresponding authors. Tel.: +86 10 67096875. E-mail address: [email protected] (X. Kang). 1 Fan, F. and Shen, H. contributed equally to this work. 0009-8981/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.cca.2014.02.003
et al. [5], testing for AFP has become commonplace in clinical and physical examinations in recent years. At present, the most commonly used methods for the quantitative determination of AFP are the enzyme-linked immunosorbent assay (ELISA) [6,7], radioimmunoassay [8], and time-resolved fluoroimmunoassay [9]. Although the assays are performed solely within the wells of microtiter plates, the assays can be complicated, time consuming, expensive, wasteful, and require large pieces of equipment for detection. In addition, radioimmunoassays require the disposal of radiolabels, and the lanthanide labels used for fluoroimmunoassays are expensive and susceptible to outside interference [10]. The area of miniaturized or microfluidic analysis systems, also called “micro total analysis systems (μTAS) or lab-on-a-chip (LOC)”, has gained increased popularity [11,12]. In the mid-1990s, a new analysis platform was established on the microfluidic chip [13], and to date it has been widely applied in the field of biological research, including cell culture [14,15], single cell detection [16], genetic analysis [17,18], and immunoassays [19,20]. Several review papers on microfluidic bioanalysis discussed the necessity to miniaturize ELISA and other types of immunoassays for biological research and clinical diagnosis [21–23]. Because of the advantages of microminiaturization, microfluidic technology can be utilized to achieve these goals. In the current study,
114
F. Fan et al. / Clinica Chimica Acta 431 (2014) 113–117
we combine microfluidics, chemiluminescence and immunoassay technology to devise a novel method for measuring AFP concentration in serum.
2. Materials and methods 2.1. Reagents and samples AFP calibrators, AFP monoclonal antibodies (Ab 1 ) and HRPlabeled AFP monoclonal antibodies (Ab 2 ) were from Beijing Keybiotech Co., Ltd. Ab1 was diluted in 0.1 mol/l phosphate buffered saline (PBS, pH 7.4). The calibrators and Ab2 were diluted in 3 g/dl bovine serum albumin solution (BSA; Merck), which was used as the blocking agent. Chemiluminescent substrate reagent kits were obtained from Millipore. A Roche ECL kit was from Roche Diagnostics Shanghai Co. Ltd. PBS was used as the wash solution. All human serum used in this study which were from Beijing Tiantan hospital had been approved by IRB, and we have obtained informed consents from all the research objects. The 56 serum objects included 38 health individuals (which include 32 pregnant women), 11 liver disease subjects and 7 other disease subjects. All serum samples were analyzed directly without any pretreatment.
2.2. Methods 2.2.1. Preparation of microfluidic chips The microfluidic chip has 2 layers: an upper microchannel layer and a lower PDMS layer used as a base for protein immobilization. Polydimethylsiloxane (PDMS) [24,25] was used to make the microfluidic chip. The microchannel layer contained seven channels and was prepared as follows. First, the PDMS base and curing agent (Sylgard 184, Dow Corning Inc.) were mixed 10:1, and the mixture degassed to remove bubbles. Then, the mixture was poured into the mold and hardened in an oven (Shanghai Yiheng Technical Co. Ltd.) at 80 °C for 25 min [26]. Once cooled the curing PDMS layer was removed from the mold and inlet and outlet portals made using a syringe needle. The height and width of the channels were maintained at 500 μm.
2.2.2. Procedures This method consists of four stages: (i) Ab1 was delivered via microchannels for immobilization on the PDMS base, as shown in Fig. 1a. (ii) The microchannel immobilized on the PDMS base was peeled off and another one was placed perpendicularly to the immobilized antibody strips. Then, nonspecific binding sites were blocked with BSA. (iii) The antigen, either an AFP calibrator or test sample, was added, and reacted with an excess of Ab1 . The antigen–antibody (Ab1–AFP) complexes were formed at the intersection of the two microchannels. Then we washed the PDMS base via the microchannels 3 times with 20 μl PBS buffer each time with the procedure of inserting PBS in one side of portal and sucking it out on the other side, so unbound antibody and antigen removed by washing. (iv) Ab2 was introduced into the microchannels and the double sandwich immunocomplex (Ab1–AFP–Ab2) formed (Fig. 1c). Unbound reagents were removed by washing with PBS and the chemiluminescent substrate was injected into the chip. The chemiluminescence images were captured using an imager designed by the National Center for Nanoscience and Technology (Beijing, China), and the images analyzed using ImageJ software [27]. Because AFP concentration and chemiluminescence intensity were linearly correlated, the concentration of serum AFP was calculated from the calibration curve or regressive equation.
Fig. 1. Schematic of the microfluidic AFP immunoassay. (a) Ab1 immobilization: The 2-layer microfluidic chip comprises a PDMS microchannel (layer one) on a PDMS base (layer 2). Ab1 was introduced via microchannels, resulting in (b) the formation of parallel antibody strips on the PDMS base. (c) Immune reaction: a second microchannel was placed perpendicular to the immobilized antibody strips, and the antigen and Ab 2 introduced sequentially. The ECL substrate was added and the chemiluminescent signal measured (d).
2.3. Statistical analysis All statistical analyses were performed using SPSS 13.0 software. The sera AFP concentrations in healthy controls were expressed as the mean ± standard deviation (mean ± SD). 3. Results and discussion 3.1. Effect of Ab1 concentration on chemiluminescence intensity At varying Ab1 concentrations, the chemiluminescence intensity was different. The other experiment conditions were controlled, and Ab1 was serially diluted from 10 to 160 μg/ml; when the Ab1 concentration was between 40 and 160 μg/ml, the detection range of AFP fulfilled the requirements for use in a clinical immunoassay (Fig. 2a). By taking into consideration (i) the linear range of width of the proposed method,
F. Fan et al. / Clinica Chimica Acta 431 (2014) 113–117
115
3.2. Effect of the Ab2 dilution order on chemiluminescence intensity At a constant AFP concentration of 6.25 ng/ml, we found that if the Ab2 dilution order is greater than 1:200, the chemiluminescence intensity was difficult to detect. In consideration of reagent consumption, a 1:200 dilution ratio for Ab2 was suitable for the microfluidic AFP immunoassay developed in this study (Fig. 3b). 3.3. Standard curve and detection limit The standard curve was prepared by using a working standard solution of different concentrations of AFP. The concentration series of each standard was made at 12.5, 25, 50, 100 and 200 ng/ml, respectively, because AFP concentration and chemiluminescence intensity were linearly correlated at this range. The lower limit of detection (LOD) for the AFP assay was determined by detecting a sample signal three times higher than the background noise, and the LOD value in the tests was 1.5 ng/ml. 3.4. Sample detection
Fig. 2. Optimization of Ab1 and Ab2 concentrations. (a) Ab1 was serially diluted from 160 to 10 μg/ml. The AFP calibrator was diluted from 400 to 12.5 ng/ml. (b) Two-fold serial dilutions (1:50 to 1:800) of Ab2 were prepared, and the immunoassay performed using a constant AFP calibrator concentration of 6.25 ng/ml.
(ii) the practicality for clinical use, and (iii) reagent consumption, an Ab1 concentration of 40 μg/ml was determined to be the most favorable for this assay.
The clinical samples and serially diluted AFP calibrators (12.5 to 200 ng/ml) were introduced into parallel microchannels as shown in Fig. 3a. Chemiluminescent signals; uniform and square in shape were observed. No cross-contamination between different microchannels was detected. A blank control (last channel) was included in the assay for the monitoring of non-specific adsorption; no signal was detected. Fig. 3b represents the corresponding calibration curve for the immunoassay in Fig. 3a. Log2 AFP concentration/12.5 was presented on the X-axis, and the corresponding chemiluminescence intensity of the AFP calibrators on the Y-axis. A strong linear relationship existed between Log2 concentration/12.5 ng/ml and grayscale values, and the linear regression equation was: Y = 994.30213 + 1445.4315X; R2 = 0.99403. Therefore, the sample in Fig. 3a was determined as 21.84 ng/ml according to the formula mentioned above. 3.5. Within- and between-run coefficients of variation (CVs) Within- and between-run coefficients of variation (CVs) were determined to evaluate the precision and repeatability, respectively of the assay. The same sample was simultaneously analyzed 49 times for within-chip assay and 2 times each day for 11 consecutive working
Fig. 3. Immunoassay results for AFP detection using microfluidic chips. (a) Different concentrations of AFP calibrators were introduced into the first five microchannels; the clinical sample and blank control were injected into the two remaining microchannels. The gray spots on the microchip demonstrated the immune reactive signal for AFP. (b) Calibration curve of AFP concentration from 12.5 to 200 ng/ml. The chemiluminescence intensity across three spots for each AFP concentration was determined. The plotted values are the mean ± SD of the triplicate measurements. Chemiluminescence intensity of AFP spots and AFP concentration exhibited a significant linear relationship (R2 = 0.99403).
116
F. Fan et al. / Clinica Chimica Acta 431 (2014) 113–117
days for between-chip assays. The CVs were 9.91% for within-chip and 11.4% for between-chip assays, respectively. 3.6. Interference studies Different concentrations of hemoglobin and bilirubin in serum were added in serum to observe its effect on the gray value of AFP. Based on a relative error of b 5%, we found that AFP detection in serum by this approach was free of the interference of hemolytic and jaundice, when hemoglobin concentration was less than 1.5 g/dl and bilirubin concentration was below 60 mg/dl. 3.7. Comparison of methods The proposed method has been applied to evaluate AFP in 56 human serum samples, and the results were compared with those obtained by the commercial Roche method, a technique commonly used in hospitals. As can be seen in Fig. 4, the serum AFP concentrations measured using microfluidic chips strongly correlated with those obtained using Roche's electrochemiluminescence (ECL) kit; Y = −4.85987 + 0.92643X, and the coefficient was 0.9812. Then, we tested a sample 3 times with this method and 3 times with the Roche ECL, and the results showed that the bias of microfluidic chip immunoassay was 3.9%, which was lower than the Roche ECL. 3.8. Reference value We measured AFP concentration in sera from 30 fasting, apparently healthy subjects, and the mean concentration of AFP was 2.51 ± 1.51 ng/ml. 3.9. The characteristics and advantages of this novel AFP microfluidic immunoassay With the introduction of microfluidic chips, this immunoassay provides significant advantages over conventional immunoassays, including (i) ease of operation: the operating procedure is simple, does not require an experienced technician, nor does it require external power input; (ii) reduced assay time: the total detection time was 40 min; most commonly used immune diagnostic methods can detect only 1 protein/hormone in an individual assay and usually require a minimum of 2–3 h to complete; (iii) reduced sampling size: this immunoassay requires at most 17 μl of sample, and it is possible to detect multi-markers for a sample simultaneously. One of the practical
Fig. 4. Comparison of analytical parameters of the microfluidic immunoassay and Roche ECL methods. The AFP concentration in 56 human serum samples was determined using each of these methods.
advantages of this method is that if different antibodies were delivered into the parallel microchannels, then this approach can be used to detect multi-biomarkers simultaneously. Therefore, for some precious samples (i.e., neonatal blood and spinal fluid), sampling size is not only reduced, but the number of test items is increased for the same sample volume used in traditional assays; (iv) reduced cost: miniaturization dramatically reduces the consumption of expensive reagents, and the materials can be reused; the reagent costs for the microfluidic immunoassay is only 5% that of an equivalent ELISA; (v) simple instrumentation: the instrument used in this study was small and easily manipulated, the results rapidly obtained, and the system can be adjusted to meet the requirements for POCT and primary hospital diagnostic laboratories. 4. Conclusions In brief, we have developed a reliable, simple, reagent safe and cheap immunoassay method based on microfluidic chips, and applied it to the detection of AFP in serum. Therefore, we consider this technique to be suitable for use in clinical settings. We are committed to improving the assay, and expanding its application into other programs, such as for use with other biomarkers, as well as achieving the synchronous determination of multi-markers for a single sample. Acknowledgements This work was supported by the Instrument Developing Project of the Chinese Academy of Sciences, grant no. YZ201114. References [1] Ruoslahti E, Engvall E. Alpha-fetoprotein. Scand J Immunol 1978;7:1–17. [2] Zhou L, Liu J, Luo F. Serum tumor markers for detection of hepatocellular carcinoma. World J Gastroenterol 2006;12:1175–81. [3] Wright LM, Kreikemeier JT, Fimmel CJ. A concise review of serum markers for hepatocellular cancer. Cancer Detect Prev 2007;31:35–44. [4] Tamura Y, Igarashi M, Kawai H, Suda T, Satomura S, Aoyagi Y. Clinical advantage of highly sensitive on-chip immunoassay for fucosylated fraction of alpha-fetoprotein in patients with hepatocellular carcinoma. Dig Dis Sci 2010;55:3576–83. [5] Ruoslahti E, Seppälä M. Studies of carcino-fetal proteins. 3. Development of a radioimmunoassay for -fetoprotein. Demonstration of -fetoprotein in serum of healthy human adults. Int J Cancer (Journal international du cancer) 1971;8:374–83. [6] Zhang QY, Chen H, Lin Z, et al. Comparison of chemiluminescence enzyme immunoassay based on magnetic microparticles with traditional colorimetric ELISA for the detection of serum α-fetoprotein. JPA 2012;2:130–5. [7] Jin HJ, Gan N, Hou JG, et al. Signal amplification of electrochemical ELISA for the detection of alpha-fetoprotein using core–shell Fe3O4Au nanoparticles as labels. Sens Lett 2012;10:886–93. [8] Silver HK, Gold P, Feder S, Freedman SO, Shuster J. Radioimmunoassay for human alpha 1-fetoprotein. Proc Natl Acad Sci U S A 1973;70:526–30. [9] Lin G, Liu T, Zou L, Hou J, Wu Y. Development of a dual-label time-resolved fluoroimmunoassay for the detection of α-fetoprotein and hepatitis B virus surface antigen. Luminescence 2013;28:401–6. [10] Wang X, Zhang QY, Li ZJ, et al. Development of high-performance magnetic chemiluminescence enzyme immunoassay for α-fetoprotein in human serum. Clin Chim Acta 2008;393:90–4. [11] Gomez R, Bashir R, Sarikaya A, et al. Microfluidic biochip for impedance spectroscopy of biological species. Biomed Microdevices 2001;3:201–9. [12] Arora A, Simone G, Salieb-Beugelaar GB, Kim JT, Manz A. Latest developments in micro total analysis systems. Anal Chem 2010;82:4830–47. [13] Huang H, Zheng X, Zheng J, Pan J, Pu X. Rapid analysis of alpha-fetoprotein by chemiluminescence microfluidic immunoassay system based on super-paramagnetic microbeads. Biomed Microdevices 2009;11:213–6. [14] Tourovskaia A, Figueroa-Masot X, Folch A. Differentiation-on-a-chip: a microfluidic platform for long-term cell culture studies. Lab Chip 2005;5:14–9. [15] Ziółkowska K, Stelmachowska A, Kwapiszewski R, et al. Long-term threedimensional cell culture and anticancer drug activity evaluation in a microfluidic chip. Biosens Bioelectron 2013;40:68–74. [16] Hou S, Zhao L, Shen Q, et al. Polymer nanofiber-embedded microchips for detection, isolation, and molecular analysis of single circulating melanoma cells. Angew Chem Int Ed 2013;52:3379–83. [17] Ferguson BS, Buchsbaum SF, Wu T, et al. Genetic analysis of H1N1 influenza virus from throat swab samples in a microfluidic system for point-of-care diagnostics. J Am Chem Soc 2011;133:9129–35. [18] Pan X, Jiang L, Liu K, et al. A microfluidic device integrated with multichamber polymerase chain reaction and multichannel separation for genetic analysis. Anal Chim Acta 2010;674:110–5.
F. Fan et al. / Clinica Chimica Acta 431 (2014) 113–117 [19] Yang D, Niu X, Liu Y, et al. Electrospun nanofibrous membranes: a novel solid substrate for microfluidic immunoassays for HIV. Adv Mater 2008;20:4770–5. [20] Ng AC, Uddayasankar U, Wheeler A. Immunoassays in microfluidic systems. Anal Bioanal Chem 2010;397:991–1007. [21] Pan W, Chen W, Jiang X. Microfluidic Western blot. Anal Chem 2010;82:3974–6. [22] Whitesides GM. The origins and the future of microfluidics. Nature 2006;442:368–73. [23] Yager P, Edwards T, Fu E, et al. Microfluidic diagnostic technologies for global public health. Nature 2006;442:412–8.
117
[24] Cheng YQ, Yang L, Guo CL, et al. Research progress of materials and fabrication technologies of microfluidic chip. Adv Mater Res 2012;542–543:891–4. [25] Stanton MM, Ducker RE, MacDonald JC, et al. Super-hydrophobic, highly adhesive, polydimethylsiloxane (PDMS) surfaces. J Colloid Interface Sci 2012;367:502–8. [26] Liu Y, Yu J, Do M, et al. Accelerating microfluidic immunoassays on filter membranes by applying vacuum. Biomed Microdevices 2012;14:17–23. [27] Rasband WS. ImageJ. Bethesda, MD: US National Institutes of Health. 1997 [http:// rsb.info.nih.gov/ij(1997): 1997–2007].
