Article pubs.acs.org/ac
Ring-Oven Washing Technique Integrated Paper-based Immunodevice for Sensitive Detection of Cancer Biomarker Wei Liu,* Yumei Guo, Mei Zhao, Huifang Li, and Zhujun Zhang Key Laboratory of Analytical Chemistry for Life Science of Shaanxi Province, School of Chemistry and Chemical Engineering, Shaanxi Normal University, Xi’an 710062, China ABSTRACT: A paper-based microfluidic immunodevice has recently attracted considerable interest for point-of-care testing (POCT) and a washing procedure was used as a standard procedure in immunoassay to eliminate the nonspecific binding protein from a paper surface. However, the traditional washing method cannot get rid of the nonspecific binding protein more completely to get a lower background. In this work, a novel washing strategy with a ring-oven technique integrated on a paper-based immunodevice was presented, which can effectively wash a nonspecific binding protein and enable a low background for sensitive detection of the carcinoembryonic antigen (CEA). By immobilizing the antibody on the detection area and incorporating the temperature-controlled ring-oven under the paper-based device, the continuous washing solution can carry the nonspecific binding protein to the waste area freely by capillary force and then the waste area dried quickly by heating. The paper device, which is matched to the size of the ring-oven, is composed of eight microfluidic channels by the simple and rapid paper-cutting fabrication method. With the HRP-catalyzed 3,3′,5,5′-tetramethylbenzidine (TMB)−H2O2 colorimetric detection method, a lower detection limit of 0.03 ng/mL CEA can be obtained by enzyme-linked immunosorbent assay (ELISA). The washing efficiency for the nonspecific binding protein was improved a lot compared to the traditional washing methods, and the established paper-based device can be used in the determination of CEA in human serum with high sensitivity. The paper-based device provides a new washing strategy for sensitive immunoassay and point-of-care diagnostics. aper,1 generally made of cellulose fibers, has been functionalized as a substrate to construct microfluidic devices which have the advantages of low-cost, high abundance, biodegradability, and excellent chemical compatibility with many applications. Since the microfluidic paper-based analytical devices (μPADs) were first demonstrated by Whitesides and co-workers,2 lab-on-a-paper systems have been promising technologies for point-of-care testing (POCT), public health, and environmental fields. Nowadays, much work is focused on the POCT field, and many papers were published on the immunoassay of biomarkers on μPADs2−6 such as enzymelinked immunosorbent assays (ELISA). ELISA7−10 is one part of the immunoassay methods which combine the specificity of antibodies with the effective catalytic properties of enzymes to provide a specific and sensitive method for some special analytes. The first work on paper-based ELISA (P-ELISA) was done by Whitesides’ group,11 which was performed in a 96microzone plate fabricated in paper. On the paper device, each test zone required only about 3 μL of sample, and an entire PELISA can be completed in less than an hour. As we all know, the washing procedure is the standard procedure which should be done many times in the ELISA immunoassay procedure.11,12 Actually, no matter what kind of detection method was adopted in the paper-based device, the tedious manual washing procedures13 would take much more time than the analysis itself. The washing effect can directly
P
© XXXX American Chemical Society
influence the sensitivity and precision of the detection results. Until now, the traditional washing procedure was carried out by using blotting paper to absorb the washing solution from the corner or the other side of the paper-based device.11 But the main problem is that we do not know if the nonspecific binding protein was cleaned entirely from the paper surface. Few papers14 have investigated the relationship between the washing method and cleaning effect on the paper-based chip. Nowadays, a new washing pattern was adopted in Yu’s work14 by using a folder on the 3D microfluidic paper-based immunodevice to achieve a better washing result. With the folder placed above the folding waste tab, washing buffer was added to the immunozones first, and the waste tab would be torn off from the paper-based device. The washing results with the minimum signal were shown in their work without the blank signal. But this did not mean that the nonspecific binding protein was cleaned totally from the paper surface. Another main problem is that the washing pattern will influence the precision in POCT, which may cause a false result to the researchers.15 In this work, the ring-oven technique16 was for the first time revisited as the heater to assist the washing procedure in the paper-based device. Nonspecific binding protein can be washed Received: May 14, 2015 Accepted: July 3, 2015
A
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 1. Schematic diagram and assay procedure on the paper-based chip. (A) Size and shape of this paper-based chip. (B) ELISA immunoassay procedure: (1) cutting paper-based chip, (2) immobilization and detection zone, (3) chitosan−glutaraldehyde modified paper zone, (4) capture antibodies, (5) BSA blocked paper zone, (6) incubation with antigen solution on paper zone, (7) incubation with HRP-labeled antibody, and (8) color assay with TMB and H2O2.
used cell-surface tumor markers. The measurement of CEA is mainly used to identify recurrences after surgical resection or localize cancer spread through dosage of biological fluids.23,24 Using the same ELISA kits for CEA detection, the detection limit of the novel washing method was 10 times lower than the conventional washing method. Chitosan coating and the glutaraldehyde cross-linking method were used to modify the paper-based immunodevice to covalently immobilize antibodies on it.25 Different concentrations of CEA can result in the variation of color intensity on the paper. A good linear relationship between the color intensity and the CEA concentration was obtained in the range of 0.1−20.0 ng/mL. This paper-based chip had been applied successfully for the determination of CEA in human serum. To our knowledge, this study is the first report to apply the ring-oven washing technique in the paper-based immunoassay system.
from the detection area, and a lower detection limit of CEA can be achieved. The ring-oven technique was proposed and developed in 1954 as a microanalytical technique by Herbert Weisz.17 It employed a simple oven in which a disc of filter paper was adapted. The washing process carried the analytes to the border of the circular wetted area, and the residues of the solvent were evaporated from the filter paper. The process generated a thin ring-like region on the filter paper where the analyte was concentrated.16 Most of the published work combining the ring-oven techinique was on the microdetermination of metal ions in the 1970s.18−20 In our work, the ring-oven technique was used as a novel washing technique on the paper-based chip, and nonspecific binding protein can be washed away from the detection area. After the washing buffer was dropped on the middle part of the paper-based chip, it can carry the nonspecific binding protein to spread to the surrounding detection areas. The heated ring-oven can make the excess washing buffer dry quickly in the waste area. The ring-oven technique can achieve less time for washing, and all the detection areas can be washed at the same time. Compared to the tedious manual washing procedure, the whole analysis time had been shortened greatly by this new washing pattern. In this work, CEA was used as a model for the ELISA detection method. Since CEA21,22 can be used to estimate curative and prognosis for cancer, it is one of the most widely
■
EXPERIMENTAL SECTION Reagents and Materials. Mouse monoclonal capture CEA antibodies and ELISA kits for CEA were purchased from Zhengzhou Biocell Biotechnology Co., Ltd. (China). Bovine serum albumin (BSA), NaH2PO4, Na2HPO4, chitosan (90% deacetylation and a relative molecular mass of 400 000 g/mol), and glutaraldehyde were purchased from Sinopharm Group Chemical Reagent Company (Shanghai, China). Western blot B
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry 3,3′,5,5′-tetramethyl benzidine (TMB) was obtained from Huzhou InnoReagents Co., Ltd. (China). Coupling buffer for antibody immobilization was 0.01 mol/L phosphate buffer solution (PBS) with the pH value of 7.4. Blocking buffer for the residual reactive sites on the antibody immobilized paper was PBS containing 0.5% BSA. A total of 0.25 mg/mL of the chitosan solution was prepared by dissolving chitosan in 3% acetic acid. All chemicals and reagents were of analytical grade and used without further purification. Deionized water was used in all assays and solutions. Whatman chromatography paper 3 MM was purchased from Sigma (U.S.A.). The clinical serum samples were from Shaanxi Normal University Hospital. Preparation of Paper-based Chip and Immunoassay Procedure. The fabrication procedure of the paper-based chip is shown in Figure 1. As illustrated in Figure 1A, the pattern of the device, which consisted of eight channels, was first designed by CorelDraw X6. The design was then exported as a DXF file into the controller software of a craft cutter, ROBO Master-Pro (Graphtec Corporation). Using a carrier sheet, the paper was cut by the cutting plotter (Graphtec Craft Robo-S, Graphtec Corporation). According to the ring-oven’s dimension, the diameter of the paper-based chip is 50 mm. The details of the dimension is shown in Figure 1A. Then, the chips were prepared for later use. The method to immobilize antibodies on paper through chitosan coating and glutaraldehyde cross-linking was referred to the literature.25,26 A total of 2.5 μL of 0.25 mg/mL chitosan solution was added into each detection zone (Figure 1B). After it dried at room temperature, 2.5 μL of 2.5% glutaraldehyde was dropped into each detection zone to activate for 2 h to cross-link the aldehyde group on paper surface. Then, the paper device was put and aligned on the top of the ring-oven. The syringe pump was turned on and the paper device was washed for 5 min and then dried. A total of 2.5 μL of 20 μg/mL antibodies was added to the detection area and incubated for 30 min at room temperature. Then, excess antibodies, which were not immobilized on paper, were washed for 5 min. A total of 2.5 μL of 0.5% BSA was dropped onto the detection area to block the residual reactive sites and was incubated for 15 min. Afterward, the paper-based device was washed again. A total of 2.5 μL of different concentrations of CEA was dropped on each paper detection zone. According to the literature,26 a short incubation time of 210 s was used here. After washing, the same volume of HRP-labeled antibody (in ELISA kits) was added to each detection zone, incubated for another 210 s, and washed again. Finally, 2.5 μL of TMB-H2O2 (in ELISA kits) solution was added to each detection zone. The intensity of the blue color was scanned by the scanner first and analyzed by ImageJ. Washing Procedure on the Paper-based Chip. Figure 2 depicts the modified ring-oven system which was made by our group and used for the washing step on the paper-based device. The ring oven is constituted basically of two parts. The base part can control the temperature very well. The upper cylinder is made of aluminum. The upper opening has a ring on it, and the inner and outer diameter of the ring is just fitted to the dimension of the waste zone. A paper-based chip with a diameter of 50 mm was placed on the top of the ring. After putting the cutting paper on the ring, the oven was turned on. Then, the PTFE tubing with a 0.25 mm inner diameter, which was connected with a syringe pump, was positioned over the center of the filter paper at a distance of about 3 mm upper to the paper surface.
Figure 2. Schematic diagram for washing procedure: (1) syringe pump, (2) PTFE tubing, (3) paper-based chip, (4) top heating ring to put the paper-based chip on, and (5) oven base with temperature control system.
The washing step with the ring-oven in this work is shown in Figure 2. The steps for the washing procedure in the immunoassay procedure are as follows: First, the cutting paper device was put on the ring-oven surface and the ringoven was turned on. Second, the PBS buffer solution was loaded in the injector, which was connected to the PTFE tubing. When the temperature of the ring part was 80 ± 0.5 °C, the syringe pump was turned on. Finally, the PBS solution was delivered by the syringe pump and dropped to the center zone of the paper-based chip. The washing buffer would spread to every detection and waste zone through the capillary force of the paper cellulose. After the last drop dried in the waste zone of the paper, the filter paper was removed and placed in a desiccator for later use. Color Intensity Measurement. To quantify the results, the images of the paper devices were acquired using a scanner (HP M175nw). ImageJ was used for analyzing the color images. The color can be changed to grayscale, and the grayscale value can then be used. The “window” command was used to equalize the color of the background to the paper strip. The integral density of the detection areas was then measured using ImageJ. The grayscale value was inversely proportional to the concentration of CEA.
■
RESULTS AND DISCUSSION Design and Fabrication Method of the Paper-Based Chip. In this work, the paper-based device should be designed to be coupled to the ring-oven. Different zones including the washing buffer injection zone, detection zone, and waste zone were needed on the device. The waste zone must match the size of the ring-oven and be separated from the detection area to eliminate contamination inferences. From Figure 2, we can see that the waste zone was the outside part of the paper-based chip. The heated ring was just below the paper-based chip. The outer diameter of the ring-oven was the same as the diameter of the paper-based chip. With the ring-oven turned on, the washing buffer can move to the waste zone while carrying the nonspecific binding protein. Then, the edge of the paper was dried because of the underlying oven. So the washing procedure can be done many times until the nonspecific binding protein is washed away from the detection area. Before we fabricated the microfluidic paper-based chip, we did some tests to choose the appropriate type of paper to meet our experiment demands. Different kinds of paper have been studied, such as filter paper, Whatman chromatography paper 3 MM, Whatman chromatography paper #1, quantitative filter C
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry paper, and qualified filter paper. The reagent migration rate was changeable and the repeatability was poor on quantitative filter paper, filter paper, and qualified filter paper. The reason is that the thickness of these kinds of paper was not always the same, which was in the range of 10−250 μm. The thickness of the paper affects the optical path length, scattering, assay sensitivity, and volume of fluid required for an assay. The repeatability of Whatman chromatography paper 3 MM and Whatman chromatography paper #1 was much better than the above three kinds of filter paper. But after the washing procedure was done many times on these two kinds of paper, we found that Whatman chromatography paper #1 deformed more than Whatman chromatography paper 3 MM. The deformation would influence the next steps in our work. At last, we chose Whatman chromatography paper 3 MM for later use. Two different fabrication methods were used here. The waxprinting method was tried here first for the fabrication of the paper-based chip. The hydrophobic edge of the microchannel was formed by the melted wax. With the same shape of the paper-based chip, the whole immbolization and washing procedure was done on the paper. But the hydrophobic edge of the microchannel was not clear after just several times of washing. The washing buffer began to penetrate into the hydrophobic wall of the paper, and it cannot migrate directly to the waste zone. The repeatability was poor for the detection. Then, the cutting fabrication method was tried with the unused hydrophobic part cut down. If there was no hydrophobic part on the paper, the washing buffer can migrate directly to the waste zone. So the simple cutting fabrication method was chosen here for later use. Evaluation of the Washing Protocol on Paper-Based Immunodevice. In this work, the heated oven was used for the washing procedure. The paper-based chip was put on the top of the ring-oven. The inner diameter of the ring-oven was 40 mm, and only the waste zone of the paper-based chip was aligned to the heated ring. The distance of the detection zone to the heated part was 10 mm. By using a precise infrared thermometer, we calibrated the temperature of the ring-oven, and the actual temperature of the ring part is 80 ± 0.5 °C. Then, we measured the temperature in different areas of the paper-based chip. With the washing solution dropped on the paper surface, the temperature of the waste zone rose to 64 ± 0.5 °C and the temperature of the detection zone was 34 ± 0.5 °C. Because the temperature of the detection zone was not high, the activity of HRP cannot be influenced in the detection area, and the nonspecific binding protein can be washed to the waste zone. The temperature of the waste zone was beneficial for the washing solution to be evaporated from the paper surface. But the protein may denature or stick into the paper fabrics in the waste zone. While, for the bigger area of the waste zone, the nonspecific binding protein can still migrate freely in the whole channel. So the ring-oven washing technique can be fulfilled successfully on the paper-based immunodevice The immunoassay has been the focus of recent research as a platform for the paper-based devices. Washing processes are essential in most heterogeneous labeled assays. The washing processes allow for low background levels by removing unbound labels and interfering species during the assays. The traditional washing method used blotting paper to absorb the washing solution from the paper surface.11 Recently, the washing method was improved by using the 3D origami-based paper device.14 But nobody investigated if the nonspecific binding protein was washed totally from the paper surface. In
our work, the excess reagents or nonspecific binding protein should be effectively washed away because the washing result would affect the later experiment. The below experiment was done to validate the washing result (Figure 3). After 2.5 μL of
Figure 3. (A) Picture of pipetting 2.5 μL of fluorescein (1.0 mg/mL) to the detection zone. (B) Picture of the washing result after 5 min washing. (C) Fluorescence signals of fluorescein (1.0 mg/mL) with different washing times.
fluorescein (1.0 mg/mL) was pipetted into every detection zone (Figure 3A), the paper-based chip was put onto the ringoven. The oven was turned on, and the washing buffer was continuously dropped and flowed to the detection area by a syringe pump with a flow rate of 30 μL/min. As shown in Figure 3B, after 5 min of washing, the fluorescein added in the detection zones was entirely washed off and dried in the waste area. Furthermore, the fluorescent signal of fluorescein was detected on the detection zones by a homemade laser-induced fluorescence spectrometer,27 which used a semiconductor laser (405 nm, 1 mW) as the laser light source and two silica optical fibers to transfer the light signal. From Figure 3C, we can see that the fluorescent intensity decreased with the washing time. After washing for 5 min, the fluorescent intensity was the same as the blank signal, which indicated that the fluorescein was washed completely. From the above experiment, we can see that the small molecule could be washed totally from the paper surface. But in this work, the antibody should be immobilized first and cannot be affected by the washing procedure. The nonspecific binding protein still should be removed from the paper surface. Compared to fluorescein, an antibody is a biomacromolecule, and we can see that the blank signal still had a light color change after the washing procedure in Figure 4. This meant that there still remained some specific binding effect on the paper surface. After antibody immobilization, BSA was used to block the residual reactive sites. But the reactive sites cannot be D
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
for CEA was 0.5 ng/mL, while the detection limit was 0.03 ng/ mL (Figure 6) by using the ring-oven washing technique. Assay Condition Optimization. After we fabricate the microfluidic paper-based chip, some parameters such as the proper reagent volume on the detection zones and the syringe pump rate should be chosen in this work. As shown in Figure 5A, different volumes of fluorescein were pipetted onto the detection zones. Without reagent diffusion, 2.5 μL of reagent can be sufficient to fill the detection zone. Thus, a volume of 2.5 μL for all reagents was selected as the optimal condition to pipette onto the detection zones for subsequent experiments. The syringe pump was used in this work to load the washing solution and drop the washing solution to the paper surface. Its rate would influence the washing time and the washing result. The flow rate was investigated in the range of 10−50 μL/min. When the flow rate of the syringe pump was faster than 40 μL/ min, the washing solution would accumulate in the whole channel and the detection zone. The second drop of the washing solution dropped and cannot wait for the first drop to migrate and evaporate from the waste zone. The nonspecific binding protein cannot be washed well, and the fast rate of the syringe pump will influence the washing result. But if the flow rate was slower than 15 μL/min, the time was longer for the washing solution to migrate to the waste zone. At last, 30 μL/ min was chosen because there is no accumulated solution in the whole channel, and this rate was proper for the washing time. Parameters such as the incubation time and the concentration of antibody which would affect the immunoassay results were chosen in this work. The short incubation time in the detection zone can be attributed to the high surface-to-volume ratio and the compact porous structure. The antigens, antibodies, and CL reagents have only short distances over which they must diffuse to react with each other. Short incubation time had been used in our former work, and it works well.26 So in this work, 210 s for the incubation time was still used. We also studied the concentration of the antibody which was immobilized on the paper surface. With 10 ng/mL of CEA and 20.0 μg/mL of HRP-labeled signal antibody, the concentration of the antibody was evaluated in the range from 0.01 to 30 μg/mL (Figure 5B). On the basis of the results, 20.0 μg/mL was chosen as the optimal antibody concentration.
Figure 4. Image results with different washing techniques. (A) Images of the ELISA results with CEA concentration of 7.5 ng/mL by ringoven washing. The washing time was from 0 to 6 min, which corresponds to images a−g. The upper row corresponds to the sample signal, and the lower row corresponds to the blank signals. (B) Images of the ELISA results with CEA concentration of 7.5 ng/mL by the traditional washing technique with the blotting paper. The upper row corresponds to the sample signal, and the lower row corresponds to the blank signals. (a−d) Images of CEA detection results with no washing and washing one, two, and three times.
totally blocked by BSA and HRP-labeled antibody had the possibility to be immobilized on the paper surface. This can induce a color change for the blank signal. Furthermore, the HRP-labeled antibody may have the chance to bind with the first immobilized antibody on the paper surface. This is another reason that the blank signal had a light color change. The next experiment was done to show the washing result with the antibody immobilized on the paper surface. By using 7.5 ng/mL of CEA, the sample signal and the blank signal with different washing times were obtained, which are shown in Figure 4. With the colorimetric readout, the blank signal did not change after 5 min of washing. From Figure 4A, we can see that the blank signal decreased a lot more than the sample signal. After washing the detection zone for 5 min, we can get the best S/N value. The conventional washing results were also carried here by using the blotting paper for washing (Figure 4B). The waxprinting method was used here for the fabrication of the paperbased chip. The zones can be washed by adding the washing buffer on the top of the zone. Then, the washing buffer was absorbed by a piece of blotting paper on the other side of the zone. The washing buffer went through the test zone vertically and into the blotting paper and carried unbound reagents with it. The results with different washing times and the blank signal are shown in Figure 4B. Compared with our washing method, the S/N value was lower and the detection limit was higher. Using the conventional washing technique, the detection limit
Figure 5. (A) Influence of pipetting volume on the detection zone. (B) Effects of the antibody concentration on ELISA immunoassay. E
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 6. (A) Picture of different concentrations of CEA. From (1) to (8), the concentrations of CEA were 0, 0.1, 1.0, 5.0, 7.5, 10.0, 15.0, and 20.0 ng/mL. (B) Calibration curve for CEA analysis using the established paper-based device. Each data point represents an average of five repeat experiments, and the error bars indicate one standard deviation.
Analytical Performance. Different concentrations of CEA were tested on this paper-based device. The intensity of the color increased with the CEA concentration, which can be clearly observed by the naked eye. To obtain quantitative data, the optical images of the filter papers were taken and then analyzed by ImageJ, which demonstrated a calibration curve between the color intensity and the CEA concentration from 0.1 to 20.0 ng/mL (Figure 6). The linear-regression equation was I = 4.04 C + 103.26, where I was the color intensity. The detection limit at a signal-to-noise ratio of 3 was 0.03 ng/mL. The relative standard deviations for the determination of 10.0 ng/mL of CEA were 3% in eight replicate measurements. The cutoff value of the CEA in clinical diagnostics is below 5 ng/ mL.28 Thus, the sensitivity of this proposed paper-based device with the ring-oven washing technique was enough for practical point-of-care testing. Application of the Paper-Based Device. The performance of this paper-based device with the novel washing technique for the determination of CEA was evaluated in human serum samples. The human serum from negative samples was obtained from Shaanxi Normal University Hospital. The results were compared with a reference to a commercialized ELISA kit and are shown in Table 1. Different concentration of CEA was added into the human serum samples. The relative difference indicated that there was no significant difference between these two methods. The proposed method was good at quantitative analysis of the content of CEA in the practical samples. In addition, the results
in comparison with other reported immunoassay methods of CEA are shown in Table 2. As a simple and rapid detection immunodevice, this method could be useful for rural populations where clinical laboratory facilities are limited. Table 2. Properties of this Work Compared with Others Reported detection method this work chemiluminescent immunoassay electrochemical immunoassay nanomaterials-based immunoassay colorimetric enzyme immunoassay CEA ELISA kita a
sample sample 1
sample 2
add (ng/mL)
1.81 1.81 1.81 2.04 2.04 2.04
2.00 4.00 6.00 2.00 4.00 6.00
measured ± SD (ng/mL) proposed method 3.89 5.79 7.92 4.14 6.12 7.91
± ± ± ± ± ±
0.11 0.18 0.21 0.12 0.16 0.17
± ± ± ± ± ±
reference
0.1−20 0.2−30
0.03 0.08
29
0.2−120
0.06
30
0.5−60
0.1
31
0.05−50
0.048
32
0.68
From Bio-Quant (San Diego, CA, U.S.A.; www.bio-quant.com).
CONCLUSION The ring-oven technique was revisited to exploit its novel washing characteristics on a paper-based device. The synergy between the ring-oven washing technique and paper-based immunodevice had been successfully demonstrated in the determination of CEA in human serum. With the waste zone separated from the detection zone, the nonspecific binding protein can be totally washed away from the detection zone. Compared to the conventional washing method, this effective washing method can achieve a lower detection limit. This proposed paper-based device with a high-throughput, rapid, sensitive, stable response to CEA determination in human serum samples will be very useful when the level of the analyte in the real sample is very important for simple, rapid, low-cost, point-of-care testing in developing countries. In addition, this novel washing technique with a paper-based chip can be readily adapted for the immunoassay of other biomarkers. This platform is especially useful for point-of-care
measured ± SD (ng/mL) ELISA 3.83 5.85 8.02 4.20 5.89 8.13
detection limit (ng/mL)
■
Table 1. Assay Results of CEA in Human Serum by the Proposed and Reference Method initial concentration (ng/mL)
linear response range (ng/mL)
0.12 0.21 0.24 0.09 0.23 0.12 F
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
(24) Tas, F.; Karabulut, S.; Ciftci, R.; Sen, F.; Sakar, B.; Disci, R.; Duranyildiz, D. Cancer Chemother. Pharmacol. 2014, 73, 1163−1171. (25) Wang, S. M.; Ge, L.; Song, X. R.; Yu, J. H.; Ge, S. G.; Huang, J. D.; Zeng, F. Biosens. Bioelectron. 2012, 31, 212−218. (26) Liu, W.; Cassano, C. L.; Xu, X.; Fan, Z. H. Anal. Chem. 2013, 85, 10270−10276. (27) Zhang, X. M.; Song, C. J.; Chen, L. L.; Zhang, K.; Fu, A. H.; Jin, B. Q.; Zhang, Z. J.; Yang, K. Biosens. Bioelectron. 2011, 26, 3958−3961. (28) Li, T. X. Modern Clinical Immunoassay; Military Medical Science Press: Beijing, 2001, 178−209. (29) Fu, Z. H.; Yan, F.; Liu, H.; Lin, J. H.; Ju, H. X. Biosens. Bioelectron. 2008, 23, 1422−1428. (30) He, X. L.; Yuan, R.; Chai, Y. Q.; Shi, Y. T. J. Biochem. Biophys. Methods 2008, 70, 823−829. (31) Chen, X.; Jia, X. L.; Han, J. N.; Ma, J.; Ma, Z. F. Biosens. Bioelectron. 2013, 50, 356−361. (32) Liu, M. Y.; Jia, C. P.; Jin, Q. H.; Lou, X. H.; Yao, S. H.; Xiang, J. Q.; Zhao, J. L. Talanta 2010, 81, 1625−1629.
and medical diagnostics in remote regions and resource-poor countries.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We are grateful to the National Natural Science Foundation of China (No. 21005048) for funding this work. The authors also thank the Fundamental Research Funds for the Central Universities (No. GK201402047) and Program for Innovative Research Team in Shaanxi Province (No. 2014KCT-28) for supporting this work.
■
REFERENCES
(1) Yetisen, A. K.; Akram, M. S.; Lowe, C. R. Lab Chip 2013, 13, 2210−2251. (2) Martinez, A. W.; Phillips, S. T.; Butte, M. J.; Whitesides, G. M. Angew. Chem., Int. Ed. 2007, 46, 1318−1320. (3) Yan, J. X.; Ge, L.; Song, X. R.; Yan, M.; Ge, S. G.; Yu, J. H. Chem. - Eur. J. 2012, 18, 4938−4945. (4) Wu, Y. F.; Xue, P.; Hui, K. M.; Kang, Y. J. Biosens. Bioelectron. 2014, 52, 180−187. (5) Ge, L.; Yan, J. X.; Song, X. R.; Yan, M.; Ge, S. G.; Yu, J. H. Biomaterials 2012, 33, 1024−1031. (6) Murdock, R. C.; Shen, L.; Griffin, D. K.; Kelley-Loughnane, N.; Papautsky, I.; Hagen, J. A. Anal. Chem. 2013, 85, 11634−11642. (7) Laing, S.; Hernandez-Santana, A.; Sassmannshausen, J.; Asquith, D. L.; McInnes, I. B.; Faulds, K.; Graham, D. Anal. Chem. 2011, 83, 297−302. (8) Ihara, M.; Yoshikawa, A.; Wu, Y. S.; Takahashi, H.; Mawatari, K.; Shimura, K.; Sato, K.; Kitamori, T.; Ueda, H. Lab Chip 2010, 10, 92− 100. (9) Parween, S.; Nahar, P. Biosens. Bioelectron. 2013, 48, 287−292. (10) Rajasekhar, R.; Parimal, R. J. Virol. Methods 2014, 207, 121−127. (11) Cheng, C. M.; Martinez, A. W.; Gong, J. L.; Mace, C. R.; Phillips, S. T.; Carrilho, E.; Mirica, K. A.; Whitesides, G. M. Angew. Chem., Int. Ed. 2010, 49, 4771−4774. (12) Watanabe, E.; Miyake, S.; Yogo, Y. J. Agric. Food Chem. 2013, 61, 12459−12472. (13) Mu, X.; Zhang, L.; Chang, S. Y.; Cui, W.; Zheng, Z. Anal. Chem. 2014, 86, 5338−5344. (14) Ge, L.; Wang, S. M.; Song, X. R.; Ge, S. G.; Yu, J. H. Lab Chip 2012, 12, 3150−3158. (15) Hu, J.; Wang, S. Q.; Wang, L.; Li, F.; Pingguan-Murphy, B.; Lu, T. J.; Xu, F. Biosens. Bioelectron. 2014, 54, 585−597. (16) Cortez, J.; Pasquini, C. Anal. Chem. 2013, 85, 1547−1554. (17) Weisz, H. Microanalysis by Ring-Oven Technique, 2nd ed.; Pergamon: New York, 1970. (18) Skuric, Z.; Valic, F.; Prpic-Marecic, J. Anal. Chim. Acta 1974, 73, 213−215. (19) McDaniel, M.; West, P. W. Anal. Chim. Acta 1974, 70, 482−484. (20) Dharmarajan, V.; West, P. W. Anal. Chim. Acta 1971, 57, 469− 472. (21) Denk, H.; Tappeiner, G.; Eckerstorfer, R.; Holzner, J. H. Int. J. Cancer 1972, 10, 262−272. (22) Shibutani, M.; Maeda, K.; Nagahara, H.; Ohtani, H.; Sakurai, K.; Toyokawa, T.; Kubo, N.; Tanaka, H.; Muguruma, K.; Ohira, M.; Hirakawa, K. Anticancer Res. 2014, 34, 3753−3758. (23) Silva, J.; Garcia, V.; Rodriguez, M.; Compte, M.; Cisneros, E.; Veguillas, P.; Garcia, J. M.; Dominguez, G.; Campos-Martin, Y.; Cuevas, J.; Peña, C.; Herrera, M.; Diaz, R.; Mohammed, N.; Bonilla, F. Genes, Chromosomes Cancer 2012, 51, 409−418. G
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Ring-Oven Washing Technique Integrated Paper-based Immunodevice for Sensitive Detection of Cancer Biomarker Wei Liu,* Yumei Guo, Mei Zhao, Huifang Li, and Zhujun Zhang Key Laboratory of Analytical Chemistry for Life Science of Shaanxi Province, School of Chemistry and Chemical Engineering, Shaanxi Normal University, Xi’an 710062, China ABSTRACT: A paper-based microfluidic immunodevice has recently attracted considerable interest for point-of-care testing (POCT) and a washing procedure was used as a standard procedure in immunoassay to eliminate the nonspecific binding protein from a paper surface. However, the traditional washing method cannot get rid of the nonspecific binding protein more completely to get a lower background. In this work, a novel washing strategy with a ring-oven technique integrated on a paper-based immunodevice was presented, which can effectively wash a nonspecific binding protein and enable a low background for sensitive detection of the carcinoembryonic antigen (CEA). By immobilizing the antibody on the detection area and incorporating the temperature-controlled ring-oven under the paper-based device, the continuous washing solution can carry the nonspecific binding protein to the waste area freely by capillary force and then the waste area dried quickly by heating. The paper device, which is matched to the size of the ring-oven, is composed of eight microfluidic channels by the simple and rapid paper-cutting fabrication method. With the HRP-catalyzed 3,3′,5,5′-tetramethylbenzidine (TMB)−H2O2 colorimetric detection method, a lower detection limit of 0.03 ng/mL CEA can be obtained by enzyme-linked immunosorbent assay (ELISA). The washing efficiency for the nonspecific binding protein was improved a lot compared to the traditional washing methods, and the established paper-based device can be used in the determination of CEA in human serum with high sensitivity. The paper-based device provides a new washing strategy for sensitive immunoassay and point-of-care diagnostics. aper,1 generally made of cellulose fibers, has been functionalized as a substrate to construct microfluidic devices which have the advantages of low-cost, high abundance, biodegradability, and excellent chemical compatibility with many applications. Since the microfluidic paper-based analytical devices (μPADs) were first demonstrated by Whitesides and co-workers,2 lab-on-a-paper systems have been promising technologies for point-of-care testing (POCT), public health, and environmental fields. Nowadays, much work is focused on the POCT field, and many papers were published on the immunoassay of biomarkers on μPADs2−6 such as enzymelinked immunosorbent assays (ELISA). ELISA7−10 is one part of the immunoassay methods which combine the specificity of antibodies with the effective catalytic properties of enzymes to provide a specific and sensitive method for some special analytes. The first work on paper-based ELISA (P-ELISA) was done by Whitesides’ group,11 which was performed in a 96microzone plate fabricated in paper. On the paper device, each test zone required only about 3 μL of sample, and an entire PELISA can be completed in less than an hour. As we all know, the washing procedure is the standard procedure which should be done many times in the ELISA immunoassay procedure.11,12 Actually, no matter what kind of detection method was adopted in the paper-based device, the tedious manual washing procedures13 would take much more time than the analysis itself. The washing effect can directly
P
© XXXX American Chemical Society
influence the sensitivity and precision of the detection results. Until now, the traditional washing procedure was carried out by using blotting paper to absorb the washing solution from the corner or the other side of the paper-based device.11 But the main problem is that we do not know if the nonspecific binding protein was cleaned entirely from the paper surface. Few papers14 have investigated the relationship between the washing method and cleaning effect on the paper-based chip. Nowadays, a new washing pattern was adopted in Yu’s work14 by using a folder on the 3D microfluidic paper-based immunodevice to achieve a better washing result. With the folder placed above the folding waste tab, washing buffer was added to the immunozones first, and the waste tab would be torn off from the paper-based device. The washing results with the minimum signal were shown in their work without the blank signal. But this did not mean that the nonspecific binding protein was cleaned totally from the paper surface. Another main problem is that the washing pattern will influence the precision in POCT, which may cause a false result to the researchers.15 In this work, the ring-oven technique16 was for the first time revisited as the heater to assist the washing procedure in the paper-based device. Nonspecific binding protein can be washed Received: May 14, 2015 Accepted: July 3, 2015
A
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 1. Schematic diagram and assay procedure on the paper-based chip. (A) Size and shape of this paper-based chip. (B) ELISA immunoassay procedure: (1) cutting paper-based chip, (2) immobilization and detection zone, (3) chitosan−glutaraldehyde modified paper zone, (4) capture antibodies, (5) BSA blocked paper zone, (6) incubation with antigen solution on paper zone, (7) incubation with HRP-labeled antibody, and (8) color assay with TMB and H2O2.
used cell-surface tumor markers. The measurement of CEA is mainly used to identify recurrences after surgical resection or localize cancer spread through dosage of biological fluids.23,24 Using the same ELISA kits for CEA detection, the detection limit of the novel washing method was 10 times lower than the conventional washing method. Chitosan coating and the glutaraldehyde cross-linking method were used to modify the paper-based immunodevice to covalently immobilize antibodies on it.25 Different concentrations of CEA can result in the variation of color intensity on the paper. A good linear relationship between the color intensity and the CEA concentration was obtained in the range of 0.1−20.0 ng/mL. This paper-based chip had been applied successfully for the determination of CEA in human serum. To our knowledge, this study is the first report to apply the ring-oven washing technique in the paper-based immunoassay system.
from the detection area, and a lower detection limit of CEA can be achieved. The ring-oven technique was proposed and developed in 1954 as a microanalytical technique by Herbert Weisz.17 It employed a simple oven in which a disc of filter paper was adapted. The washing process carried the analytes to the border of the circular wetted area, and the residues of the solvent were evaporated from the filter paper. The process generated a thin ring-like region on the filter paper where the analyte was concentrated.16 Most of the published work combining the ring-oven techinique was on the microdetermination of metal ions in the 1970s.18−20 In our work, the ring-oven technique was used as a novel washing technique on the paper-based chip, and nonspecific binding protein can be washed away from the detection area. After the washing buffer was dropped on the middle part of the paper-based chip, it can carry the nonspecific binding protein to spread to the surrounding detection areas. The heated ring-oven can make the excess washing buffer dry quickly in the waste area. The ring-oven technique can achieve less time for washing, and all the detection areas can be washed at the same time. Compared to the tedious manual washing procedure, the whole analysis time had been shortened greatly by this new washing pattern. In this work, CEA was used as a model for the ELISA detection method. Since CEA21,22 can be used to estimate curative and prognosis for cancer, it is one of the most widely
■
EXPERIMENTAL SECTION Reagents and Materials. Mouse monoclonal capture CEA antibodies and ELISA kits for CEA were purchased from Zhengzhou Biocell Biotechnology Co., Ltd. (China). Bovine serum albumin (BSA), NaH2PO4, Na2HPO4, chitosan (90% deacetylation and a relative molecular mass of 400 000 g/mol), and glutaraldehyde were purchased from Sinopharm Group Chemical Reagent Company (Shanghai, China). Western blot B
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry 3,3′,5,5′-tetramethyl benzidine (TMB) was obtained from Huzhou InnoReagents Co., Ltd. (China). Coupling buffer for antibody immobilization was 0.01 mol/L phosphate buffer solution (PBS) with the pH value of 7.4. Blocking buffer for the residual reactive sites on the antibody immobilized paper was PBS containing 0.5% BSA. A total of 0.25 mg/mL of the chitosan solution was prepared by dissolving chitosan in 3% acetic acid. All chemicals and reagents were of analytical grade and used without further purification. Deionized water was used in all assays and solutions. Whatman chromatography paper 3 MM was purchased from Sigma (U.S.A.). The clinical serum samples were from Shaanxi Normal University Hospital. Preparation of Paper-based Chip and Immunoassay Procedure. The fabrication procedure of the paper-based chip is shown in Figure 1. As illustrated in Figure 1A, the pattern of the device, which consisted of eight channels, was first designed by CorelDraw X6. The design was then exported as a DXF file into the controller software of a craft cutter, ROBO Master-Pro (Graphtec Corporation). Using a carrier sheet, the paper was cut by the cutting plotter (Graphtec Craft Robo-S, Graphtec Corporation). According to the ring-oven’s dimension, the diameter of the paper-based chip is 50 mm. The details of the dimension is shown in Figure 1A. Then, the chips were prepared for later use. The method to immobilize antibodies on paper through chitosan coating and glutaraldehyde cross-linking was referred to the literature.25,26 A total of 2.5 μL of 0.25 mg/mL chitosan solution was added into each detection zone (Figure 1B). After it dried at room temperature, 2.5 μL of 2.5% glutaraldehyde was dropped into each detection zone to activate for 2 h to cross-link the aldehyde group on paper surface. Then, the paper device was put and aligned on the top of the ring-oven. The syringe pump was turned on and the paper device was washed for 5 min and then dried. A total of 2.5 μL of 20 μg/mL antibodies was added to the detection area and incubated for 30 min at room temperature. Then, excess antibodies, which were not immobilized on paper, were washed for 5 min. A total of 2.5 μL of 0.5% BSA was dropped onto the detection area to block the residual reactive sites and was incubated for 15 min. Afterward, the paper-based device was washed again. A total of 2.5 μL of different concentrations of CEA was dropped on each paper detection zone. According to the literature,26 a short incubation time of 210 s was used here. After washing, the same volume of HRP-labeled antibody (in ELISA kits) was added to each detection zone, incubated for another 210 s, and washed again. Finally, 2.5 μL of TMB-H2O2 (in ELISA kits) solution was added to each detection zone. The intensity of the blue color was scanned by the scanner first and analyzed by ImageJ. Washing Procedure on the Paper-based Chip. Figure 2 depicts the modified ring-oven system which was made by our group and used for the washing step on the paper-based device. The ring oven is constituted basically of two parts. The base part can control the temperature very well. The upper cylinder is made of aluminum. The upper opening has a ring on it, and the inner and outer diameter of the ring is just fitted to the dimension of the waste zone. A paper-based chip with a diameter of 50 mm was placed on the top of the ring. After putting the cutting paper on the ring, the oven was turned on. Then, the PTFE tubing with a 0.25 mm inner diameter, which was connected with a syringe pump, was positioned over the center of the filter paper at a distance of about 3 mm upper to the paper surface.
Figure 2. Schematic diagram for washing procedure: (1) syringe pump, (2) PTFE tubing, (3) paper-based chip, (4) top heating ring to put the paper-based chip on, and (5) oven base with temperature control system.
The washing step with the ring-oven in this work is shown in Figure 2. The steps for the washing procedure in the immunoassay procedure are as follows: First, the cutting paper device was put on the ring-oven surface and the ringoven was turned on. Second, the PBS buffer solution was loaded in the injector, which was connected to the PTFE tubing. When the temperature of the ring part was 80 ± 0.5 °C, the syringe pump was turned on. Finally, the PBS solution was delivered by the syringe pump and dropped to the center zone of the paper-based chip. The washing buffer would spread to every detection and waste zone through the capillary force of the paper cellulose. After the last drop dried in the waste zone of the paper, the filter paper was removed and placed in a desiccator for later use. Color Intensity Measurement. To quantify the results, the images of the paper devices were acquired using a scanner (HP M175nw). ImageJ was used for analyzing the color images. The color can be changed to grayscale, and the grayscale value can then be used. The “window” command was used to equalize the color of the background to the paper strip. The integral density of the detection areas was then measured using ImageJ. The grayscale value was inversely proportional to the concentration of CEA.
■
RESULTS AND DISCUSSION Design and Fabrication Method of the Paper-Based Chip. In this work, the paper-based device should be designed to be coupled to the ring-oven. Different zones including the washing buffer injection zone, detection zone, and waste zone were needed on the device. The waste zone must match the size of the ring-oven and be separated from the detection area to eliminate contamination inferences. From Figure 2, we can see that the waste zone was the outside part of the paper-based chip. The heated ring was just below the paper-based chip. The outer diameter of the ring-oven was the same as the diameter of the paper-based chip. With the ring-oven turned on, the washing buffer can move to the waste zone while carrying the nonspecific binding protein. Then, the edge of the paper was dried because of the underlying oven. So the washing procedure can be done many times until the nonspecific binding protein is washed away from the detection area. Before we fabricated the microfluidic paper-based chip, we did some tests to choose the appropriate type of paper to meet our experiment demands. Different kinds of paper have been studied, such as filter paper, Whatman chromatography paper 3 MM, Whatman chromatography paper #1, quantitative filter C
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry paper, and qualified filter paper. The reagent migration rate was changeable and the repeatability was poor on quantitative filter paper, filter paper, and qualified filter paper. The reason is that the thickness of these kinds of paper was not always the same, which was in the range of 10−250 μm. The thickness of the paper affects the optical path length, scattering, assay sensitivity, and volume of fluid required for an assay. The repeatability of Whatman chromatography paper 3 MM and Whatman chromatography paper #1 was much better than the above three kinds of filter paper. But after the washing procedure was done many times on these two kinds of paper, we found that Whatman chromatography paper #1 deformed more than Whatman chromatography paper 3 MM. The deformation would influence the next steps in our work. At last, we chose Whatman chromatography paper 3 MM for later use. Two different fabrication methods were used here. The waxprinting method was tried here first for the fabrication of the paper-based chip. The hydrophobic edge of the microchannel was formed by the melted wax. With the same shape of the paper-based chip, the whole immbolization and washing procedure was done on the paper. But the hydrophobic edge of the microchannel was not clear after just several times of washing. The washing buffer began to penetrate into the hydrophobic wall of the paper, and it cannot migrate directly to the waste zone. The repeatability was poor for the detection. Then, the cutting fabrication method was tried with the unused hydrophobic part cut down. If there was no hydrophobic part on the paper, the washing buffer can migrate directly to the waste zone. So the simple cutting fabrication method was chosen here for later use. Evaluation of the Washing Protocol on Paper-Based Immunodevice. In this work, the heated oven was used for the washing procedure. The paper-based chip was put on the top of the ring-oven. The inner diameter of the ring-oven was 40 mm, and only the waste zone of the paper-based chip was aligned to the heated ring. The distance of the detection zone to the heated part was 10 mm. By using a precise infrared thermometer, we calibrated the temperature of the ring-oven, and the actual temperature of the ring part is 80 ± 0.5 °C. Then, we measured the temperature in different areas of the paper-based chip. With the washing solution dropped on the paper surface, the temperature of the waste zone rose to 64 ± 0.5 °C and the temperature of the detection zone was 34 ± 0.5 °C. Because the temperature of the detection zone was not high, the activity of HRP cannot be influenced in the detection area, and the nonspecific binding protein can be washed to the waste zone. The temperature of the waste zone was beneficial for the washing solution to be evaporated from the paper surface. But the protein may denature or stick into the paper fabrics in the waste zone. While, for the bigger area of the waste zone, the nonspecific binding protein can still migrate freely in the whole channel. So the ring-oven washing technique can be fulfilled successfully on the paper-based immunodevice The immunoassay has been the focus of recent research as a platform for the paper-based devices. Washing processes are essential in most heterogeneous labeled assays. The washing processes allow for low background levels by removing unbound labels and interfering species during the assays. The traditional washing method used blotting paper to absorb the washing solution from the paper surface.11 Recently, the washing method was improved by using the 3D origami-based paper device.14 But nobody investigated if the nonspecific binding protein was washed totally from the paper surface. In
our work, the excess reagents or nonspecific binding protein should be effectively washed away because the washing result would affect the later experiment. The below experiment was done to validate the washing result (Figure 3). After 2.5 μL of
Figure 3. (A) Picture of pipetting 2.5 μL of fluorescein (1.0 mg/mL) to the detection zone. (B) Picture of the washing result after 5 min washing. (C) Fluorescence signals of fluorescein (1.0 mg/mL) with different washing times.
fluorescein (1.0 mg/mL) was pipetted into every detection zone (Figure 3A), the paper-based chip was put onto the ringoven. The oven was turned on, and the washing buffer was continuously dropped and flowed to the detection area by a syringe pump with a flow rate of 30 μL/min. As shown in Figure 3B, after 5 min of washing, the fluorescein added in the detection zones was entirely washed off and dried in the waste area. Furthermore, the fluorescent signal of fluorescein was detected on the detection zones by a homemade laser-induced fluorescence spectrometer,27 which used a semiconductor laser (405 nm, 1 mW) as the laser light source and two silica optical fibers to transfer the light signal. From Figure 3C, we can see that the fluorescent intensity decreased with the washing time. After washing for 5 min, the fluorescent intensity was the same as the blank signal, which indicated that the fluorescein was washed completely. From the above experiment, we can see that the small molecule could be washed totally from the paper surface. But in this work, the antibody should be immobilized first and cannot be affected by the washing procedure. The nonspecific binding protein still should be removed from the paper surface. Compared to fluorescein, an antibody is a biomacromolecule, and we can see that the blank signal still had a light color change after the washing procedure in Figure 4. This meant that there still remained some specific binding effect on the paper surface. After antibody immobilization, BSA was used to block the residual reactive sites. But the reactive sites cannot be D
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
for CEA was 0.5 ng/mL, while the detection limit was 0.03 ng/ mL (Figure 6) by using the ring-oven washing technique. Assay Condition Optimization. After we fabricate the microfluidic paper-based chip, some parameters such as the proper reagent volume on the detection zones and the syringe pump rate should be chosen in this work. As shown in Figure 5A, different volumes of fluorescein were pipetted onto the detection zones. Without reagent diffusion, 2.5 μL of reagent can be sufficient to fill the detection zone. Thus, a volume of 2.5 μL for all reagents was selected as the optimal condition to pipette onto the detection zones for subsequent experiments. The syringe pump was used in this work to load the washing solution and drop the washing solution to the paper surface. Its rate would influence the washing time and the washing result. The flow rate was investigated in the range of 10−50 μL/min. When the flow rate of the syringe pump was faster than 40 μL/ min, the washing solution would accumulate in the whole channel and the detection zone. The second drop of the washing solution dropped and cannot wait for the first drop to migrate and evaporate from the waste zone. The nonspecific binding protein cannot be washed well, and the fast rate of the syringe pump will influence the washing result. But if the flow rate was slower than 15 μL/min, the time was longer for the washing solution to migrate to the waste zone. At last, 30 μL/ min was chosen because there is no accumulated solution in the whole channel, and this rate was proper for the washing time. Parameters such as the incubation time and the concentration of antibody which would affect the immunoassay results were chosen in this work. The short incubation time in the detection zone can be attributed to the high surface-to-volume ratio and the compact porous structure. The antigens, antibodies, and CL reagents have only short distances over which they must diffuse to react with each other. Short incubation time had been used in our former work, and it works well.26 So in this work, 210 s for the incubation time was still used. We also studied the concentration of the antibody which was immobilized on the paper surface. With 10 ng/mL of CEA and 20.0 μg/mL of HRP-labeled signal antibody, the concentration of the antibody was evaluated in the range from 0.01 to 30 μg/mL (Figure 5B). On the basis of the results, 20.0 μg/mL was chosen as the optimal antibody concentration.
Figure 4. Image results with different washing techniques. (A) Images of the ELISA results with CEA concentration of 7.5 ng/mL by ringoven washing. The washing time was from 0 to 6 min, which corresponds to images a−g. The upper row corresponds to the sample signal, and the lower row corresponds to the blank signals. (B) Images of the ELISA results with CEA concentration of 7.5 ng/mL by the traditional washing technique with the blotting paper. The upper row corresponds to the sample signal, and the lower row corresponds to the blank signals. (a−d) Images of CEA detection results with no washing and washing one, two, and three times.
totally blocked by BSA and HRP-labeled antibody had the possibility to be immobilized on the paper surface. This can induce a color change for the blank signal. Furthermore, the HRP-labeled antibody may have the chance to bind with the first immobilized antibody on the paper surface. This is another reason that the blank signal had a light color change. The next experiment was done to show the washing result with the antibody immobilized on the paper surface. By using 7.5 ng/mL of CEA, the sample signal and the blank signal with different washing times were obtained, which are shown in Figure 4. With the colorimetric readout, the blank signal did not change after 5 min of washing. From Figure 4A, we can see that the blank signal decreased a lot more than the sample signal. After washing the detection zone for 5 min, we can get the best S/N value. The conventional washing results were also carried here by using the blotting paper for washing (Figure 4B). The waxprinting method was used here for the fabrication of the paperbased chip. The zones can be washed by adding the washing buffer on the top of the zone. Then, the washing buffer was absorbed by a piece of blotting paper on the other side of the zone. The washing buffer went through the test zone vertically and into the blotting paper and carried unbound reagents with it. The results with different washing times and the blank signal are shown in Figure 4B. Compared with our washing method, the S/N value was lower and the detection limit was higher. Using the conventional washing technique, the detection limit
Figure 5. (A) Influence of pipetting volume on the detection zone. (B) Effects of the antibody concentration on ELISA immunoassay. E
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 6. (A) Picture of different concentrations of CEA. From (1) to (8), the concentrations of CEA were 0, 0.1, 1.0, 5.0, 7.5, 10.0, 15.0, and 20.0 ng/mL. (B) Calibration curve for CEA analysis using the established paper-based device. Each data point represents an average of five repeat experiments, and the error bars indicate one standard deviation.
Analytical Performance. Different concentrations of CEA were tested on this paper-based device. The intensity of the color increased with the CEA concentration, which can be clearly observed by the naked eye. To obtain quantitative data, the optical images of the filter papers were taken and then analyzed by ImageJ, which demonstrated a calibration curve between the color intensity and the CEA concentration from 0.1 to 20.0 ng/mL (Figure 6). The linear-regression equation was I = 4.04 C + 103.26, where I was the color intensity. The detection limit at a signal-to-noise ratio of 3 was 0.03 ng/mL. The relative standard deviations for the determination of 10.0 ng/mL of CEA were 3% in eight replicate measurements. The cutoff value of the CEA in clinical diagnostics is below 5 ng/ mL.28 Thus, the sensitivity of this proposed paper-based device with the ring-oven washing technique was enough for practical point-of-care testing. Application of the Paper-Based Device. The performance of this paper-based device with the novel washing technique for the determination of CEA was evaluated in human serum samples. The human serum from negative samples was obtained from Shaanxi Normal University Hospital. The results were compared with a reference to a commercialized ELISA kit and are shown in Table 1. Different concentration of CEA was added into the human serum samples. The relative difference indicated that there was no significant difference between these two methods. The proposed method was good at quantitative analysis of the content of CEA in the practical samples. In addition, the results
in comparison with other reported immunoassay methods of CEA are shown in Table 2. As a simple and rapid detection immunodevice, this method could be useful for rural populations where clinical laboratory facilities are limited. Table 2. Properties of this Work Compared with Others Reported detection method this work chemiluminescent immunoassay electrochemical immunoassay nanomaterials-based immunoassay colorimetric enzyme immunoassay CEA ELISA kita a
sample sample 1
sample 2
add (ng/mL)
1.81 1.81 1.81 2.04 2.04 2.04
2.00 4.00 6.00 2.00 4.00 6.00
measured ± SD (ng/mL) proposed method 3.89 5.79 7.92 4.14 6.12 7.91
± ± ± ± ± ±
0.11 0.18 0.21 0.12 0.16 0.17
± ± ± ± ± ±
reference
0.1−20 0.2−30
0.03 0.08
29
0.2−120
0.06
30
0.5−60
0.1
31
0.05−50
0.048
32
0.68
From Bio-Quant (San Diego, CA, U.S.A.; www.bio-quant.com).
CONCLUSION The ring-oven technique was revisited to exploit its novel washing characteristics on a paper-based device. The synergy between the ring-oven washing technique and paper-based immunodevice had been successfully demonstrated in the determination of CEA in human serum. With the waste zone separated from the detection zone, the nonspecific binding protein can be totally washed away from the detection zone. Compared to the conventional washing method, this effective washing method can achieve a lower detection limit. This proposed paper-based device with a high-throughput, rapid, sensitive, stable response to CEA determination in human serum samples will be very useful when the level of the analyte in the real sample is very important for simple, rapid, low-cost, point-of-care testing in developing countries. In addition, this novel washing technique with a paper-based chip can be readily adapted for the immunoassay of other biomarkers. This platform is especially useful for point-of-care
measured ± SD (ng/mL) ELISA 3.83 5.85 8.02 4.20 5.89 8.13
detection limit (ng/mL)
■
Table 1. Assay Results of CEA in Human Serum by the Proposed and Reference Method initial concentration (ng/mL)
linear response range (ng/mL)
0.12 0.21 0.24 0.09 0.23 0.12 F
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
(24) Tas, F.; Karabulut, S.; Ciftci, R.; Sen, F.; Sakar, B.; Disci, R.; Duranyildiz, D. Cancer Chemother. Pharmacol. 2014, 73, 1163−1171. (25) Wang, S. M.; Ge, L.; Song, X. R.; Yu, J. H.; Ge, S. G.; Huang, J. D.; Zeng, F. Biosens. Bioelectron. 2012, 31, 212−218. (26) Liu, W.; Cassano, C. L.; Xu, X.; Fan, Z. H. Anal. Chem. 2013, 85, 10270−10276. (27) Zhang, X. M.; Song, C. J.; Chen, L. L.; Zhang, K.; Fu, A. H.; Jin, B. Q.; Zhang, Z. J.; Yang, K. Biosens. Bioelectron. 2011, 26, 3958−3961. (28) Li, T. X. Modern Clinical Immunoassay; Military Medical Science Press: Beijing, 2001, 178−209. (29) Fu, Z. H.; Yan, F.; Liu, H.; Lin, J. H.; Ju, H. X. Biosens. Bioelectron. 2008, 23, 1422−1428. (30) He, X. L.; Yuan, R.; Chai, Y. Q.; Shi, Y. T. J. Biochem. Biophys. Methods 2008, 70, 823−829. (31) Chen, X.; Jia, X. L.; Han, J. N.; Ma, J.; Ma, Z. F. Biosens. Bioelectron. 2013, 50, 356−361. (32) Liu, M. Y.; Jia, C. P.; Jin, Q. H.; Lou, X. H.; Yao, S. H.; Xiang, J. Q.; Zhao, J. L. Talanta 2010, 81, 1625−1629.
and medical diagnostics in remote regions and resource-poor countries.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS We are grateful to the National Natural Science Foundation of China (No. 21005048) for funding this work. The authors also thank the Fundamental Research Funds for the Central Universities (No. GK201402047) and Program for Innovative Research Team in Shaanxi Province (No. 2014KCT-28) for supporting this work.
■
REFERENCES
(1) Yetisen, A. K.; Akram, M. S.; Lowe, C. R. Lab Chip 2013, 13, 2210−2251. (2) Martinez, A. W.; Phillips, S. T.; Butte, M. J.; Whitesides, G. M. Angew. Chem., Int. Ed. 2007, 46, 1318−1320. (3) Yan, J. X.; Ge, L.; Song, X. R.; Yan, M.; Ge, S. G.; Yu, J. H. Chem. - Eur. J. 2012, 18, 4938−4945. (4) Wu, Y. F.; Xue, P.; Hui, K. M.; Kang, Y. J. Biosens. Bioelectron. 2014, 52, 180−187. (5) Ge, L.; Yan, J. X.; Song, X. R.; Yan, M.; Ge, S. G.; Yu, J. H. Biomaterials 2012, 33, 1024−1031. (6) Murdock, R. C.; Shen, L.; Griffin, D. K.; Kelley-Loughnane, N.; Papautsky, I.; Hagen, J. A. Anal. Chem. 2013, 85, 11634−11642. (7) Laing, S.; Hernandez-Santana, A.; Sassmannshausen, J.; Asquith, D. L.; McInnes, I. B.; Faulds, K.; Graham, D. Anal. Chem. 2011, 83, 297−302. (8) Ihara, M.; Yoshikawa, A.; Wu, Y. S.; Takahashi, H.; Mawatari, K.; Shimura, K.; Sato, K.; Kitamori, T.; Ueda, H. Lab Chip 2010, 10, 92− 100. (9) Parween, S.; Nahar, P. Biosens. Bioelectron. 2013, 48, 287−292. (10) Rajasekhar, R.; Parimal, R. J. Virol. Methods 2014, 207, 121−127. (11) Cheng, C. M.; Martinez, A. W.; Gong, J. L.; Mace, C. R.; Phillips, S. T.; Carrilho, E.; Mirica, K. A.; Whitesides, G. M. Angew. Chem., Int. Ed. 2010, 49, 4771−4774. (12) Watanabe, E.; Miyake, S.; Yogo, Y. J. Agric. Food Chem. 2013, 61, 12459−12472. (13) Mu, X.; Zhang, L.; Chang, S. Y.; Cui, W.; Zheng, Z. Anal. Chem. 2014, 86, 5338−5344. (14) Ge, L.; Wang, S. M.; Song, X. R.; Ge, S. G.; Yu, J. H. Lab Chip 2012, 12, 3150−3158. (15) Hu, J.; Wang, S. Q.; Wang, L.; Li, F.; Pingguan-Murphy, B.; Lu, T. J.; Xu, F. Biosens. Bioelectron. 2014, 54, 585−597. (16) Cortez, J.; Pasquini, C. Anal. Chem. 2013, 85, 1547−1554. (17) Weisz, H. Microanalysis by Ring-Oven Technique, 2nd ed.; Pergamon: New York, 1970. (18) Skuric, Z.; Valic, F.; Prpic-Marecic, J. Anal. Chim. Acta 1974, 73, 213−215. (19) McDaniel, M.; West, P. W. Anal. Chim. Acta 1974, 70, 482−484. (20) Dharmarajan, V.; West, P. W. Anal. Chim. Acta 1971, 57, 469− 472. (21) Denk, H.; Tappeiner, G.; Eckerstorfer, R.; Holzner, J. H. Int. J. Cancer 1972, 10, 262−272. (22) Shibutani, M.; Maeda, K.; Nagahara, H.; Ohtani, H.; Sakurai, K.; Toyokawa, T.; Kubo, N.; Tanaka, H.; Muguruma, K.; Ohira, M.; Hirakawa, K. Anticancer Res. 2014, 34, 3753−3758. (23) Silva, J.; Garcia, V.; Rodriguez, M.; Compte, M.; Cisneros, E.; Veguillas, P.; Garcia, J. M.; Dominguez, G.; Campos-Martin, Y.; Cuevas, J.; Peña, C.; Herrera, M.; Diaz, R.; Mohammed, N.; Bonilla, F. Genes, Chromosomes Cancer 2012, 51, 409−418. G
DOI: 10.1021/acs.analchem.5b01814 Anal. Chem. XXXX, XXX, XXX−XXX
