3-D microarray and its microfabrication-free fluidic immunoassay device

Analytica Chimica Acta 889 (2015) 187e193

Contents lists available at ScienceDirect

Analytica Chimica Acta journal homepage: www.elsevier.com/locate/aca

3-D microarray and its microfabrication-free fluidic immunoassay device Yingshuai Liu a, b, c, **, Yuanyuan Zhang a, b, c, Zhisong Lu a, b, c, Chang Ming Li a, b, c, * a

Institute for Clean Energy & Advanced Materials, Faculty of Materials & Energy, Southwest University, Chongqing 400715, China Chongqing Engineering Research Center for Rapid Diagnosis of Dread Disease, Southwest University, Chongqing 400715, China c Chongqing Key Laboratory for Advanced Materials and Technologies of Clean Energies, Southwest University, Chongqing 400715, China b

h i g h l i g h t s

g r a p h i c a l a b s t r a c t


A new conceptual 3-D array fluidic immunoassay device is developed.
The device is made by simply coupling a glass cuboid into a circular tube without microfabrication process.
The 3-D cuboid substrate offers fourfold effective surface for more sensing spots.
The device greatly enhances the mass transport for rapid immunoassay.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 14 May 2015 Received in revised form 13 July 2015 Accepted 16 July 2015 Available online 10 August 2015

Conventional 2-D microarray is known to have high-throughput detection capability; however, the sensing spots density is significantly hindered by the spot-to-spot distance (gap) requirement for eliminating cross-talks between adjacent spots. Herein a new conceptual 3-D microarray device is proposed to significantly improve the spots density. To demonstrate advantages of the 3-D array, a microfabrication-free fluidic immunoassay device is further made by simply coupling an antibodiesarrayed glass cuboid into a circular glass tube. Rapid, sensitive and high-throughput flow-through immunoassays were accomplished with the 3-D array-based device for detection limits of 10e100 pg mL1 and wide dynamic range over 4e5 orders of magnitude in human serum with cancer biomarkers afetoprotein (AFP) and carcinoembryonic antigen (CEA) as model targets, holding great promise for practical clinical applications. The 3-D microarray device not only significantly increases the density of sensing spots, but also greatly enhances the mass transport for rapid immunoassay when using in a flowthrough device. © 2015 Elsevier B.V. All rights reserved.

Keywords: 3-Dimensional microarray High-throughput Microfabrication-free Flow-through immunoassay Cancer biomarkers

1. Introduction

* Corresponding author. Institute for Clean Energy & Advanced Materials, Faculty of Materials & Energy, Southwest University, Chongqing 400715, China. ** Corresponding author. Institute for Clean Energy & Advanced Materials, Faculty of Materials & Energy, Southwest University, Chongqing 400715, China. E-mail addresses: [email protected] (Y. Liu), [email protected] (C.M. Li). http://dx.doi.org/10.1016/j.aca.2015.07.044 0003-2670/© 2015 Elsevier B.V. All rights reserved.

Planar 2-D protein microarray has emerged as a popular tool to detect a large number of proteins in parallel and has been widely used in various areas such as fundamental bioscience [1,2], drug discovery [3,4] and clinical diagnostics [5e8]. Further enhancement of its assay capability is still demanded for high-throughput analysis of a huge number of proteins. One of the most straightforward

188

Y. Liu et al. / Analytica Chimica Acta 889 (2015) 187e193

way is to reduce the size of array spots through employment of hydrophobic surface and advanced printing techniques. However, as a consequence of decreasing of spot sizes, microarray printer and scanner with high resolution must be developed. Thus the cost of fabrication instruments will drastically increase. At the same time, noise from electronics becomes larger and larger with the increase of scan resolution [9]. Further improvement of the spots density on conventional 2-D substrates is also significantly hindered by the spot-to-spot distance (gap) requirement for eliminating diffusion cross-talks between adjacent spots. Nanoarrays have been developed by Mirkin's group with dippen nanolithography to greatly improve the spots density [10,11]. Images of the nanoarrays were captured by atomic force microscopy (AFM) and quantitative analysis was then performed by measuring the height difference between testing and control nanospots. Nevertheless, the technology requires very complicated and highly expensive manufacturing process and thus it is very difficult for practical applications. Most importantly, the number of probe molecules in a nanosized spot is reduced to a very low level, the sensing dynamic range must be significantly impeded [9]. In addition, the limitation of the spots density still exists. In this work, a new conceptual 3-D microarray is fabricated to significantly improve the spots density. To demonstrate its significance and application, a 3-D array-based flow-through immunoassay device is developed without use of expensive and complicated microfabrication processes. Most microfluidic-based multiplexed immunoassays employ fluidic network to isolate different capture molecules [12e15] for high-throughput detections but are difficult to acquire signals when performing detections in many independent microfluidic channels. In addition, they also need expensive and complicated microfabrication. The microfluidic device in this work is assembled with a rectangular glass cuboid and a circular capillary glass tube. The former acts as a 3-D microarray substrate, while the latter is employed as a tubular channel to guide a fluid during the assays. Analytes-targeted antibodies are deposited on four side surfaces of the glass cuboid in an array format, thus converting a 2-D format substrate to a 3-D one and fourfold effective surface area can be achieved theoretically (Scheme 1). The 3-D arrays-based microfluidic device in a flow stream can also greatly enhance the mass transport rate onto to the surface array spots due to suddenly reduced passage area for faster even turbulence flow in comparison to the 2-D format thus for reducing the assay time. This 3-D format is also completely compatible with well-developed commercial available array chip writers and fluorescence array scanners for great feasibility of highthroughput flow-through immunoassays. By simple integration with an automatic pump system, sample loading and washing steps are easily performed automatically. With this device multiple tumor biomarkers can be simultaneously analyzed in serum in

40e60 min. Thus the developed flow-through 3-D array immunoassay device demonstrates a great capability for miniaturized, rapid, high-throughput, and sensitive detection of tumor biomarkers. 2. Experimental 2.1. Materials and chemicals Monoclonal anti-rabbit IgG, rabbit IgG, Cy3-labeled rabbit antigoat IgG (rabbit IgG), Cy3-labeled streptavidin, tris buffered saline (TBS, 0.05 M, pH 8.0) with 0.05% Tween 20, phosphate buffered saline (PBS, 0.01 M, pH 7.4), (3-glycidoxypropyl)-trimethoxysilane (GPTMS, 98%), glycidyl methacrylate (GMA, 97%), poly(ethylene glycol) methacrylate (PEGMA, Mn ¼ 360), (3-aminopropyl) triethoxysilane (APTES, 99%), triethylamine (TEA, 99.5%), L-ascorbic acid, 2-bromoisobutyryl bromide (BIB, 98%), 2, 20 -bipyridyl (Bipy, 99%) and copper (II) bromide (CuBr2, 99.999%) were ordered from SigmaeAldrich (Shanghai, China). Monoclonal mouse anticarcinoembryonic antigen (CEA, clone 057e10009) was received from Abnova (Beijing, China). CEA full length protein, monoclonal anti-a-fetoprotein [M803209] (AFP), and AFP were purchased from Abcam (Shanghai, China). Cy3-labeled mouse anti-CEA (B5) and Cy3-labeled mouse anti-AFP (A2) were received from Bioss (Beijing, China). 2.2. Preparation of GPTMS-Glass tube Rectangular glass cuboids with dimensions of 1 mm  1 mm  55 mm puchased from Beijing Zhong Cheng Quartz glass Co. Ltd. were cleaned with alcohol under sonication for 10 min and rinsed with deionized (DI) water. The precleaned glass cuboids were immersed into 1 M potassium hydroxide (KOH) for overnight to remove organic residues and to promote hydroxylation, followed by rinsing with DI water. The hydroxylated glass cuboids were silanized with 3% (V/V) GPTMS alcohol solution for 2 h at room temperature. Subsequently, the glass cuboids were thoroughly rinsed with alcohol and DI water. After drying with N2 stream, they were placed into a vacuum oven and baked for 150 min at 110  C. 2.3. Preparation P (GMA-co-PEGMA)-glass cuboid P(GMA-co-PEGMA)-glass cuboids were prepared according to procedures reported in our previous work (Scheme 1(a)e(d)) [16,17]. Briefly, the hydroxylated glass cuboids were silanized with 3%(V/V) APTES to introduce amine group, which was then utilized to covalently conjugate initiator 2-bromoisobutyryl bromide. The initiator-activated glass cuboids were immersed in 25 mL of 1:1 (V/ V) methanol/water mixture containing CuBr2 (3.35 mg mL1), Bipy (4.6 mg mL1), GMA (V/V, 2%), and PEGMA (V/V, 20%). Then 500 mL of 119 mg mL1 ascorbic acid was rapidly added to initiate the polymerization. The reaction was kept at room temperature for 6 h, followed by intensive rinsing with ethanol and DI water (Scheme 2a). 2.4. Protein microarray fabrication on sidefaces of a glass cuboid

Scheme 1. Surface area of 2-D and 3-D substrates (a); effective surface area of rectangular glass cuboid (b).

0.01 M PBS with 0.003% triton X-100 was employed as a printing buffer in all experiments. Antibodies solution with a concentration of 100 mg mL1 was deposited on sidefaces of the glass cuboids by a robotic microarray spotter (PersonalArrayer™ 16, CapitalBio Corporation, China). The deposited volume was ~0.3 nL per spot calculated from contact angle and diameter of the microspots and thus the amount of deposited antibody was 30 pg per spot. To

Y. Liu et al. / Analytica Chimica Acta 889 (2015) 187e193

189

different velocities of 5, 15, and 30 mL min1. For comparison, the interaction was also studied under static incubation for 5 min to 120 min. In sandwich immunoassays, the mixture of AFP and CEA at concentrations ranging from 0 pg mL1 to 10 mg mL1 in 10% human serum was injected into each integrated device at the velocity of 15 mL min1 for 30 min, followed by wash with TBS at 50 mL min1 for 5 min. Then the mixture of Cy3-labeled anti-AFP and Cy3labeled anti-CEA at 1 mg mL1 each was pumped into all devices at 15 mL min1 for 30 min. 2.6. Signal acquisition and data analysis After completion of immunoassays, the glass cuboids were intensively washed with TBS and DI water. The glass cuboids were then dried with N2 flow and put into the lab-made holder for signal acquisition with a microarray scanner (LuxScan™ 10 K, CapitalBio Corporation, China). Similar to the array printing, sidefaces were scanned one-by-one by rotating the cuboid as shown in Scheme 2c. The signal intensity of each microspot was obtained by Luxscan 3.0 analysis software. Local background-subtracted fluorescence intensity of all replicated spots were averaged and used for further statistical analysis and plots. 3. Results and discussion 3.1. Characterization of P(GMA-co-PEGMA) functionalization

Scheme 2. 3-D glass cuboid surface functionalization (a), flow-through microarray device assembly (b), schematic illustration(c) and photographs(d) of laboratory-made glass cuboid holder.

conduct 3-D array printing, glass cuboids were in parallel placed into a laboratory-made holder for supporting, fixing and positioning. The top-side faces of all cuboids were first printed and other side faces were then arrayed one-by-one via simply rotating the cuboids by 90 per rotation. The schematic illustration and photographs of the holder are presented in Scheme 2c & d to clearly show the printing process. The printed microarrays were incubated for 2 h in a vacuum oven at room temperature to promote antibody immobilization. Remaining functional groups were passivated by TBS (0.05 M, pH ¼ 8.0 containing 0.05% Tween-20). 2  3 microspots were redundantly printed in each subarray to assure the assay reproducibility and reliability. The redundant design and the mean value obtained from subsequent data analysis greatly improve the assay accuracy. 2.5. Flow-through microarray immunoassay To conduct flow-through 3-D microarray immunoassays, the antibody arrayed P(GMA-co-PEGMA)-glass cuboid was placed into a circular glass tube (Scheme 2b). TBS was pumped into the integrated assay device with a flow rate of 50 mL min1 to wash away the unbound probe antibodies and block unreacted active sites at the same time. In direct immunoassays, Cy3-labeled rabbit IgG at 100 ng mL1 was flowed into each device. The antibody-antigen (AbeAg) interaction in 10e80 min was investigated under three

Surface topographies of pristine and brush modified glass cuboids were examined with atomic force microscopy (AFM, Dimension Icon, Bruker) to confirm functionalization of P(GMA-coPEGMA) copolymer brush. For sample preparation, the pristine and P(GMA-co-PEGMA)-modified glass cuboids were intensively cleaned with ethanol and DI water, followed by drying in a vacuum oven for 4 h. The samples were then tested by tapping mode in atmospheric environment. Fig. 1 shows the topographical images of surfaces of pristine and copolymer functionalized glass cuboids. A flat and featureless morphology is observed from surface of the pristine glass cuboid (Fig. 1a), while many worm-like nanoparticles indicated by the red arrow in Fig. 1b are densely and homogeneously distributed on the surface of the P(GMA-co-PEGMA) copolymer brush-modified glass cuboid. The worm-like nanostructures are obviously attributed to aggregation of the copolymer brush chains during sample preparation [16]. The corresponding section profiles (Fig. 1c and d) show that depths of the featured nanostructures (between top and bottom) are around 1 nm for pristine surface while about 3.5 nm for the brush-coated surface. The experimental results prove that P(GMA-co-PEGMA) copolymer brush is successfully prepared on surface of the glass cuboid. 3.2. P(GMA-co-PEGMA)-glass cuboid as a microarray substrate Antibody immobilization on P(GMA-co-PEGMA)-substrate was investigated by directly printing Cy3-labeled rabbit anti-goat IgG at 50 mg mL1. As comparison, same experiment was carried out on GPTMS-substrate, which is functionalized with an epoxy silane. After post-printing incubation and thorough clean, fluorescence signal was acquired and analyzed as described in Section 2.6. As shown in Fig. 2a (red), the observed fluorescence intensity from P(GMA-co-PEGMA)-substrate is much higher than that from GPTMS-substrate, demonstrating higher antibody loading capacity. The improved antibody loading is apparently ascribed to high density of epoxy groups and large specific surface area provided by the P(GMA-co-PEGMA) copolymer brush, which transfers the planar surface to a 3-D function layer [5,17]. However, GPTMS, a

190

Y. Liu et al. / Analytica Chimica Acta 889 (2015) 187e193

Fig. 1. AFM images and their corresponding section profiles of pristine (a, c) and P(GMA-co-PEGMA)-modified (b, d) glass cuboid surfaces. Red arrows indicate the grafted copolymer brush. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2. Antibody loading capacity and AbeAg binding efficiency (left) and signal-to-noise ratio (right) on GPTMS-substrate and P(GMA-co-PEGMA)-substrate.

small molecule, only introduces limited functional groups and specific surface area. The results further confirm the successful graft of P(GMA-co-PEGMA) and demonstrates its higher antibody loading capacity. AbeAg interaction efficiency was further evaluated on P(GMAco-PEGMA)-substrate and GPTMS-substrate via direct immunoassay. The AbeAg binding efficiency is defined to be a ratio of the signal measured from probe antibody captured Cy3-labeled rabbit IgG (antigen) to that directly obtained from substrate-immobilized Cy3-labeled rabbit IgG, which refers to the immobilized antibody. In this case, monoclonal anti-rabbit IgG at 100 mg mL1 was deposited on P(GMA-co-PEGMA)-glass cuboid and GPTMS-glass cuboid, followed by incubation for 2 h in a vacuum oven at room temperature. After wash and passivation, Cy3-labeled rabbit IgG at 1 mg mL1 was applied to each array and incubated for 2 h at room temperature, followed by rinse with TBS and DI water. As shown in Fig. 2a (blue), the fluorescence intensity obtained from P(GMA-coPEGMA)-substrate is significantly higher than that from GPTMSsubstrate, indicating higher AbeAg binding efficiency is achieved

on P(GMA-co-PEGMA)-glass cuboid. The higher binding efficiency is attributed to the brush structure, which as a floating and flexible spacer is more favorable for the immobilized antibody to efficiently capture the target molecules [16,17]. Signal-to-noise ratio is concerned as another significant parameter for microarray immunoassays. The signal-to-noise ratio on GPTMS-substrate and P(GMA-co-PEGMA)-substrate was calculated by dividing intensity of fluorescence signal from microspots by intensity from background around the microspots. As shown in Fig. 2b, the P(GMA-co-PEGMA)-substrate provides much higher signal-to-noise ratio than that on GPTMS-substrate. P(GMA-coPEGMA) copolymer brush significantly reduces the background signal and dramatically enhances signal-to-noise ratio of the immunoassay, thus improved sensitivity and specificity will be reasonably expected. Theoretically, strong nonspecific adsorption during immunoassays generates higher background signal and in turn leads to a poorer signal-to-noise ratio. In this case, introduction of P(GMA-co-PEGMA) brush significantly reduces nonspecific adsorption because of the good nonfouling property of PEG

Y. Liu et al. / Analytica Chimica Acta 889 (2015) 187e193

component. At the same time, the brush provides high density of antibody immobilization and efficient AbeAg binding efficiency, thus resulting in higher fluorescence signal intensity [16,17]. The improved signal-to-noise ratio is attributed to the lower nonspecific adsorption and higher signal intensity on P(GMA-co-PEGMA)cuboid compared with GPTMS-cuboid, on which higher nonspecific adsorption and lower probe immobilization capacity are observed. 3.3. AbeAg interaction under static and dynamic incubation AbeAg interaction under flow-through and static incubation was investigated with the integrated 3-D microfluidic device. Monoclonal anti-rabbit IgG at 100 mg mL1 was spotted on 4 sidefaces of a P(GMA-co-PEGMA)-glass cuboid with 4 subarrays (2  6 microspots per subarray) on each sideface. For static immunoassays, 1 mg mL1 Cy3-labeled rabbit IgG was applied to the device and incubated for different time durations over 0e120 min under static condition. In flow-through experiments, 1 mg mL1 Cy3-labeled rabbit IgG was pumped into the reaction tubes with different velocities at 5 mL min1, 15 mL min1, and 30 mL min1 for 10e80 min. As shown in Fig. 3, the fluorescence intensity increases much faster with the extension of incubation time under all pumping velocities compared with static incubation. It is worth noting that the higher velocity is employed, the shorter incubation time is needed to get the plateau since high flow velocity can continuously provide replenishment of targets during AbeAg interaction [18,19]. Fluorescence intensity rapidly reaches an equilibrium status in ~30e50 min under dynamic incubation with three different velocities, while the signal intensity

191

continuously increases within 120 min under static condition (Fig. 3, inset). In addition, at same incubation time the fluorescence intensity under dynamic conditions are much higher than that under static conditions. The results prove that the integrated microfluidic 3-D array device significantly speeds up AbeAg interaction because of fast mass transport under dynamic incubation. Furthermore, the micro-scale flow-through tube reduces diffusion distance of targets to surface-immobilized antibodies, leading to an additional decrease of incubation time [20,21]. Consideration of incubation time and sample consumption, flow rate at 15 mL min1 and incubation time for 30 min are selected as compromissary parameter for further experiments. The short incubation time is mainly ascribed to the accelerated flow speed and the flow turbulent in the integrated immunoassay device as shown in Fig. 3 b & c. The flow velocity on surface of the glass cuboid is about 2.75 times of that on surface of 2-D substrate when the pump flow rate is fixed (Fig. 3b). The higher flow velocity accelerates the refreshment of anlalytes and tracer molecules (Cy3labeled antibodies in this case) at substrate surface, thus reducing the incubation time for AbeAg binding. In addition, turbulent flow is formed at the end and surface of glass cuboid, while laminar flow is observed at the 2-D substrate surface. The turbulent flow makes the sample mix well and speed up the mass transport of biomolecules in vertical direction, further shortening the AbeAg reaction time. 3.4. Multi-detection specificity and selectivity To evaluate multi-detection specificity and selectivity, monoclonal anti-AFP and monoclonal anti-CEA at 100 mg mL1 were printed on P(GMA-co-PEGMA)-glass cuboid surface in array format. Cy3-labeled streptavidin was employed as positive control. After post printing incubation and blocking, AFP and CEA at 100 ng mL1 in PBS were injected into tube A and tube B, respectively. The mixtures of AFP and CEA with same concentration of 100 ng mL1 in PBS and serum were pumped into tube C and tube D, respectively. Subsequently, mixture of Cy3-labeled anti-AFP and Cy3labeled anti-CEA in PBS was pumped into tube A, tube B, and tube C at 15 mL min1 for 30 min, while the mixture prepared in human serum was injected into tube D at the same rate. Fig. 4a shows the fluorescence images of immunoassay microarray panels. From top to bottom, every three rows were immobilized with Cy3labeled streptavidin (positive control), monoclonal anti-AFP, and monoclonal anti-CEA, respectively. In tube A and B, strong fluorescence signal is only observed from corresponding microspots where monoclonal anti-AFP and monoclonal anti-CEA are immobilized as probe. However, positive responses are obtained from all microspots in tube C treated with the mixture of AFP and CEA prepared in PBS. As shown in Fig. 4b, quantitative bar plots illustrate that the fluorescence intensities of all positive responses from specific AbeAg binding are significantly higher than those of negative responses. Moreover, the signal intensity observed from tube D treated by analytes in human serum is comparable with that from tube C treated by analyates in PBS. Thus experimental results demonstrate that the integrated microfluidic 3-D array immunoassay device possesses excellent specificity for multi-detection in both PBS and human serum. 3.5. Integrated microfluidic 3-D array immunoassay device for biomarker detections in serum

Fig. 3. AbeAg binding under dynamic and static incubations, inset: fluorescence intensity vs. AbeAg reaction time (a); flow rate (b) and flow types (c) on surfaces of 3-D and 2-D substrate in glass tubes.

Detection of cancer biomarkers is one of the most effective way for clinical diagnosis and prognosis of various cancers [22,23]. In this study, well-developed cancer biomarkers AFP and CEA were employed as model analytes to investigate potential application of

192

Y. Liu et al. / Analytica Chimica Acta 889 (2015) 187e193

Fig. 4. Multi-detection configuration and specificity test of the integrated microfluidic 3-D array immunoassay device. (a) Fluorescent images; (b) Quantitative bar plots.

Fig. 5. Integrated microfluidic 3-D array immunoassay device for detection of AFP and CEA. (a) Fluorescent images; (b) Dose-dependent curve.

the microfluidic 3-D array immunoassay device in practical diagnosis. Sandwich immunoassays were performed by the developed flow-through 3-D array device for parallel detection of AFP and CEA in 10% human serum. Fig. 5(a) and (b) show the fluorescence images and the dose-dependent curve with double logarithmic axes over a concentration range from 0.1 pg mL1 to 10 mg mL1. The signal intensity increases with the increase of analyte concentration up to 1 mg mL1 for both analytes. LODs are determined to be 10 pg mL1 for AFP and 100 pg mL1 for CEA according the wellknown definition of Mbþ3Sd, where Mb and Sd is mean signal and standard deviation of negative control, respectively [24]. From the typical sigmoidal dose-dependent curves, it can be seen that the dynamic ranges are 10 pg ml1 to 1 mg mL1 for AFP and 100 pg mL1 to 1 mg mL1 for CEA. The performance in terms of LOD and dynamic range is better than or comparable with those of microfluidic flow-through immunoassays performed in microchannels and merged microfluidic 2-D array devices [17,25,26]. Therefore, a microfabrication-free, high-throughput, and sensitive microfluidic 3-D array device has been demonstrated for multidetection of cancer biomarkers in 10% human serum, demonstrating its great potential application in practical diagnosis. The developed 3-D microarray significantly increases the spotting surface area and in turn array density without loss of its compatibility with commercialized microarray printer and scanner. The 3-D array immunoassays are performed in a micro-scale flow-through tube, resulting in less sample and reagents consumption and decreased assay time. Furthermore, the fluidic 3-D array device is easily fabricated by assembly of arrayed glass cuboid with a glass tube, eliminating complicated and expensive microfabrication processes [20,26]. Over all, this new conceptual 3-D array device offers prominent advantages over tranditional 2-D microarray and

demonstrates a great potential for practical applications. 4. Conclusion A new conceptual 3-D array was fabricated and its flow-through immunoassay device was developed by simple coupling a glass cuboid to a circular glass tube without use of any microfabrication techniques. The 3-D microarray immunoassay device significantly improved density of the sensing spots and enhanced the mass transport for rapid immunoassay. Rapid, sensitive, and highthroughput flow-through immunoassays of cancer biomarkers AFP and CEA in human serum have been accomplished with a detection limit of 10e100 pg mL1, demonstrating its great feasibility for practical clinical applications. In a word, this work demonstrates a rapid, sensitive and high-throughput flow-through 3-D microarray chip with a great potential for screening of tumor biomarkers. Acknowledgments This work is financially supported by the National Natural Science Foundation of China (Grant 31200604, Grant 21475106), Natural Science Foundation of Chongqing (Grant cstc2012jjA10152), Chongqing Key Laboratory for Advanced Materials and Technologies of Clean Energies (Grant cstc2011pt), and Chongqing Engineering Research Center for Rapid diagnosis of Dread Disease. References [1] Y.Y. Lin, J.Y. Lu, J. Zhang, W. Walter, W. Dang, J. Wan, S.C. Tao, J. Qian, Y. Zhao,

Y. Liu et al. / Analytica Chimica Acta 889 (2015) 187e193

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9] [10] [11]

[12] [13]

J.D. Boeke, Protein acetylation microarray reveals that NuA4 controls key metabolic target regulating gluconeogenesis, Cell 136 (2009) 1073e1084. H. Zhu, M. Bilgin, R. Bangham, D. Hall, A. Casamayor, P. Bertone, N. Lan, R. Jansen, S. Bidlingmaier, T. Houfek, Global analysis of protein activities using proteome chips, Science 293 (2001) 2101e2105. M.I. Sereni, M. Pierobon, R. Angioli, E.F. Petricoin, M.J. Frederick, Reverse phase protein microarrays and their utility in drug development, Methods Mol. Biol. 986 (2013) 187e214. A. Gupta, A.S. Heimann, I. Gomes, L.A. Devi, Antibodies against G-protein coupled receptors: novel uses in screening and drug development, Comb. Chem. High. Throughput Screen 11 (2008) 463e467. Y.S. Liu, W.H. Hu, Z.S. Lu, C.M. Li, Photografted poly(methyl methacrylate)based high performance protein microarray for hepatitis B virus biomarker detection in human serum, Med. Chem. Commun. 1 (2010) 132e135. S. Lee, S. Kim, J. Malm, O.C. Jeong, H. Lilja, T. Laurell, Improved porous silicon microarray based prostate specific antigen immunoassay by optimized surface density of the capture antibody, Anal. Chim. Acta 796 (2013) 108e114. D.H. Kong, J.W. Jung, K. Na, S.K. Jeong, Y.K. Paik, S.H. Jung, I.B. Suh, Y.M. Kim, K.S. Ha, Normalization using a tagged-internal standard assay for analysis of antibody arrays and the evaluation of serological biomarkers for liver disease, Anal. Chim. Acta 718 (2012) 92e98. L. Zhu, K. Wang, J. Cui, H. Liu, X.L. Bu, H.L. Ma, W.Z. Wang, H. Gong, C. Lausted, L. Hood, G. Yang, Z.Y. Hu, Label-free quantitative detection of tumor-derived exosomes through surface Plasmon resonance imaging, Anal. Chem. 86 (2014) 8857e8864. M. Dufva, Fabrication of high quality microarrays, Biomol. Eng. 22 (2005) 173e184. K.B. Lee, S.J. Park, C.A. Mirkin, J.C. Smith, M. Mrksich, Protein nanoarrays generated By dip-Pen nanolithography, Science 295 (2002) 1702e1705. K.B. Lee, E.Y. Kim, C.A. Mirkin, S.M. Wolinsky, The use of nanoarrays for highly sensitive and selective detection of human immunodeficiency virus type 1 in plasma, Nano Lett. 4 (2004) 1869e1872. Q.D. Zhu, K.L. Yang, Microfluidic immunoassay with plug-in liquid crystal for optical detection of antibody, Anal. Chim. Acta 853 (2015) 696e701. rez, Y.H. Jin, J. Auerswald, S. Berchtold, H.F. Knapp, J.M. Diserens, G. Sua Y. Leterrier, J.A.E. Månson, G. Voirin, Lab-on-a-chip for multiplexed biosensing of residual antibiotics in milk, Lab Chip 9 (2009) 1625e1630.

193

[14] L. Yu, C.M. Li, Y.S. Liu, J. Gao, W. Wang, Y. Gan, Flow-through functionalized PDMS microfluidic channels with dextran derivative for ELISAs, Lab Chip 9 (2009) 1243e1247. [15] M. Herrmann, T. Veres, M. Tabrizian, Enzymatically-generated fluorescent detection in micro-channels with internal magnetic mixing for the development of parallel microfluidic ELISA, Lab Chip 6 (2006) 555e560. [16] W. Hu, Y. Liu, Z. Lu, C.M. Li, Poly [oligo (ethylene glycol) methacrylate-coglycidyl methacrylate] brush substrate for sensitive surface Plasmon resonance imaging protein arrays, Adv. Funct. Mater. 20 (2010) 3497e3503. [17] Y.S. Liu, W. Wang, W. Hu, Z. Lu, X. Zhou, C.M. Li, Highly sensitive poly [glycidyl methacrylate-co-poly (ethylene glycol) methacrylate] brush-based flowthrough microarray immunoassay device, Biomed. Microdevices 13 (2011) 769e777. [18] C. Situma, M. Hashimoto, S.A. Soper, Merging microfluidics with microarraybased bioassays, Biomol. Eng. 23 (2006) 213e231. [19] P.K. Ajikumar, J.K. Ng, Y.C. Tang, J.Y. Lee, G. Stephanopoulos, H.P. Too, Carboxyl-terminated dendrimer-coated bioactive interface for protein microarray: High-sensitivity detection of antigen in complex biological samples, Langmuir 23 (2007) 5670e5677. [20] M. Cretich, G. Di Carlo, C. Giudici, S. Pokoj, I. Lauer, S. Scheurer, M. Chiari, Detection of allergen specific immunoglobulins by microarrays coupled to microfluidics, Proteomics 9 (2009) 2098e2107. [21] M. Zhao, X.F. Wang, D. Nolte, Mass-transport limitations in spot-based microarrays, Biomed. Opt. Express 1 (2010) 983e997. [22] M.J. Duffy, Carcinoembryonic antigen as a marker for colorectal cancer: is it clinically useful? Clin. Chem. 47 (2001) 624e630. [23] A.L. Washburn, L.C. Gunn, R.C. Bailey, Label-free quantitation of a cancer biomarker in complex media using silicon photonic microring resonators, Anal. Chem. 81 (2009) 9499e9506. [24] G.L. Long, J.D. Winefordner, A closer look at the IUPAC definition, Anal. Chem. 55 (1983) 712Ae724A. [25] W.H. Hu, Z.S. Lu, Y.S. Liu, T. Chen, X.Q. Zhou, C.M. Li, A portable flow-through fluorescent immunoassay lab-on-a-chip device using ZnO nanorod-decorated glass capillaries, Lab Chip 13 (2013) 1797e1802. [26] K.A. Heyries, M.G. Loughran, D. Hoffmann, A. Homsy, L.J. Blum, C.A. Marquette, Microfluidic biochip for chemiluminescent detection of allergen-specific antibodies, Biosens. Bioelectron. 23 (2008) 1812e1818.