- Email: [email protected]
Nano-ELISA for highly sensitive protein detection
Biosensors and Bioelectronics 24 (2009) 2836–2841
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Nano-ELISA for highly sensitive protein detection Chun-Ping Jia a , Xiao-Qin Zhong a , Bao Hua a , Mei-Ying Liu a , Feng-Xiang Jing a , Xin-Hui Lou a , Shi-Hua Yao b , Jia-Qing Xiang b , Qing-Hui Jin a,∗ , Jian-Long Zhao a,∗∗ a b
Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Science, 865 Changning Road, Shanghai, 200050, PR China Shanghai Tumor Hospital, Shanghai, 200050, PR China
a r t i c l e
i n f o
Article history: Received 7 November 2008 Received in revised form 22 January 2009 Accepted 12 February 2009 Available online 6 March 2009 Keywords: Gold nanoparticles Immunoassay Nano-ELISA ELISA p53
a b s t r a c t Highly sensitive protein detection method based on nanoparticles and enzyme-linked immunosorbent assays (ELISAs), named Nano-ELISA, was introduced. In this method, the micro-magnetic beads were modified with monoclonal antibody of the target protein p53. Gold nanoparticles (AuNPs) were modified with another monoclonal detector antibody and Horseradish peroxidase (HRP, for signal amplification). The presence of target protein p53 causes the formation of the sandwich structures (magnetic beads–target protein–AuNP probes) through the interaction between the antibodies and the antigen p53. The HRP at the surface of AuNPs catalytically oxidize the substrate and generate optical signals that reflected the quantity of the target protein. Down to 5 pg mL−1 of protein was detected in less than 2 h with this method. The detection sensitivity of p53 classic ELISA kit is 0.125 ng mL−1 . This method is as simple as ELISA and has higher sensitivity than ELISA, which can potentially be exploited in clinic. This method can be used to detect protein markers of tumors, nervous system or other diseases for early diagnostics. © 2009 Elsevier B.V. All rights reserved.
1. Introduction ELISA is one of the most important bio-chemical techniques used mainly to detect the presence of antibodies or antigens in a sample based on antibody–antigen immunoreactions. Due to its simplicity, low-cost, easy reading, acceptability and safety (Engvall and Perlman, 1971; Lequin, 2005), ELISA is widely used for detection of cancer protein markers, pathogen, and other proteins relative to various diseases, with the detection limit from 0.1 ng to 1 g mL−1 (Engvall and Perlman, 1971; Koppelman et al., 2004). However, the level of relative tumor protein marker is very low at the early stage of cancers or other diseases, which is beyond the detection limit of ELISA. And, most of classic ELISA, such as sandwich ELISA (capture antibody and detector antibody involved), requires long time and multiple steps, which limits the real-life application in clinics. There is a need to improve the sensitivity of the current ELISA method for highly sensitive detection protein biomarker, which is important for early diagnosis of cancer, neurodegenerative and other diseases. Gold nanoparticles, having high surface areas and unique physicochemical properties, are used widely in developing biomarker platforms (Alivisators, 2004; Rosi and Mirkin, 2005; Jain et al., 2006; Castaneda et al., 2007). Gold nanoparticles can be con-
∗ Corresponding author. Tel.: +86 21 62511070x8703; fax: +86 21 62511070x8714. ∗∗ Corresponding author. Tel.: +86 21 62511070x8702; fax: +86 21 62511070x8714. E-mail addresses: [email protected] (Q.-H. Jin), [email protected] (J.-L. Zhao). 0956-5663/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2009.02.024
jugated with DNAs, antibodies, enzymes and other bio-molecules, which can afford them promising applications in signal enhancement of bio-chemical detection (Zhang et al., 2006; Ambrosi et al., 2007; Zhang et al., 2007; Cui et al., 2008; Selvaraju et al., 2008). The recently developed nanoparticle-based bio-bar-code assay (BCA) has low attomolar sensitivity for protein targets and high zeptomolar sensitivity for nucleic acid targets (Nam et al., 2003; Thaxton et al., 2005; Georganopoulou et al., 2005; Bao et al., 2006). This detection system utilizes short oligonucleotides immobilized on AuNPs with antibody as target identification strands and surrogate amplification units in protein detection. Although this method is extraordinarily sensitive, the cumbersome experimental steps limit its application in clinic. A double-codified gold nanolabels for enhanced immunoanalysis was developed recently by Ambrosi et al. (2007). In their method, AuNPs were modified with anti-human-IgG–HRP-conjugated antibody. HRP was used here for signal amplification, which could efficiently catalyze 10,000 substrate turnovers, which significantly improved the detection limit. HRP can oxidize various chromogenic substrates and produce visible product, which can be easily detected. The protein concentration was measured spectrophotometrically based on HRP label activity and quantified by stripping voltammetry based on the intrinsic electrochemical properties of gold nanoparticles. One major drawback of this method is that the detector antibody used in this method needs to be conjugated with HRP and the process of creating HRP-linked antibodies is very complex and expensive (Crowther, 2000; Gosling, 2000), which limits its application widely.
C.-P. Jia et al. / Biosensors and Bioelectronics 24 (2009) 2836–2841
The p53 gene is a cancer suppressor gene and the p53 protein is a transcriptional factor protein that plays important roles in cell cycles control, cell growth, DNA repair and cell apoptosis. The p53 gene mutation was found in over 50% of human cancers and the level of p53 protein in serum and tissues was increased in response to these mutations and the development of cancers. So, p53 protein can be a marker for cancer diagnostic (Porebska et al., 2006; Matakidou et al., 2007). The most common method for the detection of p53 protein is classic sandwich ELISA, which requires a long time with multiple steps and low sensitivity. So, more detection methods should be developed. Several reports have described the new detection methods of p53 protein and antibodies against p53 protein (Yan et al., 2004; Cressey et al., 2008; Wang et al., 2008). These antibodies are products of immunoresponse against p53 protein and can indirectly show the level of p53 protein in serum. In Wang’s work, double-stranded oligonucleotides containing the p53 protein binding sites are immobilized onto gold electrodes to capture p53 protein based on the highly specific and strong DNA-binding property of p53 protein. The use of DNA-modified electrodes to capture p53 and the follow-up amplification of the voltammetric signals using the gold nanoparticle afforded the high sensitivity and selectivity necessary for detecting p53. The detection limit of p53 protein was 2.2 pM (116 pg mL−1 ) (Wang et al., 2008). In this study, we demonstrated a very simple and more sensitive protein detection method, termed Nano-ELISA, which combines the nanotechnology with the classic sandwich ELISA method. In NanoELISA, the detector antibody mixed with HRP at certain ratio was immobilized on the gold nanoparticles and HRP was used here for signal amplification. As a proof of concept, we chose p53 protein as target antigen to demonstrate the feasibility of Nano-ELISA method. The un-optimized detection limit of this method is 5.7 pg mL−1 and the assay time was less than 2 h. 2. Experimental 2.1. Materials and instrumentation HAuCl4 ·3H2 O, trisodium citrate, sodium chloride (NaCl), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), 2-(N-morpholino)ethane sulfonic acid (MES), streptavidin–HRP, and 3,3 ,5,5 -tetramethylbenzidine (TMB)–H2 O2 were purchased from Sigma–Aldrich Corp. (USA). p53 antibody was got from Abcam Corp. (USA). Micromagnetic particles (1 m in diameter) (MMP) were purchased from Dynal Invitrogen Corp. (USA). p53 ELISA kit was purchased from Upstate Corp. (USA). All other reagents were in analytical grade. We used a Hitachi UV-3010 UV–vis spectrophotometer to characterize the optical properties of citrate, and antibody coated gold nanoparticles. The prepared gold nanoparticles were diluted with double distilled water before transferring them to the sample sell. The scanning wavelength was set from 200 to 700 nm for all the measurements of Au nanoparticle conjugates. Transmission electron microscopy (TEM) was performed with an H-600 transmission electron microscope. A typical sample was prepared by dropping 10 L of the nanoparticle solution onto a copper TEM grid with formvar/carbon supportive films, followed by staining with uranyl acetate and lead citrate. The grid was subsequently dried in the air and imaged. 2.2. Preparation of MMP probes The MMPs were functionalized with anti-p53 capture antibody according to the manufacture’s protocol. Briefly, the carboxylic acid groups on the surface of MMPs (300 L stock solution) were activated by incubating with EDC (150 L, 50 mg mL−1 ) and NHS (150 L, 50 mg mL−1 ) solution in 25 mM MES, pH 6 at room temper-
2837
ature for 30 min. After incubation, the activated beads were washed and 150 g anti-p53 antibody was then added and incubated at room temperature for 30 min with slow rotation. The MMPs were washed 3–4 times with the washing buffer (10 mM PBS, pH 7.2, 0.1% BSA, 0.05% Tween-20) on the magnet. The remaining active sites on MMPs were blocked by incubating the washed MMPs with 10 mM PBS buffer (pH 7.2, 2% BSA, 2.5% sucrose and 0.1% PEG20000 included) for 30 min at room temperature. The MMPs were finally reconstituted in 240 L washing buffer and stored at 4 ◦ C. 2.3. Preparation and characterization of gold nanoparticles probes Gold nanoparticles (10 and 15 nm in diameter) were prepared according to previously published protocols (Turkevich et al., 1951). In particular, 100 mL double distilled water and 1 mL 1% HAuCl4 was mixed and boiled. 3 or 2 mL 1% trisodium citrate solution for the preparation of 10 and 15 nm AuNPs, respectively, was quickly added to the refluxed HAuCl4 solution, resulting in a rapid color change from pale yellow to deep red that indicated the formation of gold nanoparticles. After a continuous reflux for an additional 15 min the solution was slowly cooled down to room temperature and the particle solution was filtered through a 0.22-m cellulose nitrate filter to remove any floating aggregates. The quality of the particles was monitored with UV–vis and TEM. All glassware and the stir bar used in the preparation was cleaned in aqua regia (3:1 HCl:HNO3 ), rinsed with Millipore water, and then dried in an oven. Gold nanoparticles–protein conjugations were prepared by following the published procedures (Roth, 1982; Hermanson, 1996; Ambrosi et al., 2007). The pH of gold nanoparticles solution was adjusted to 8.2–8.5 with 0.2 M K2 CO3 . Streptavidin–HRP stock solution (1 mg mL−1 ) and p53 detector monoclonal antibody stock solution (1 mg mL−1 ) were mixed at different ratios (0.5, 1, 2, 3, 4, and 5). 1 mL gold nanoparticles solution was centrifuged at 9000 rpm for 50 min and the supernatant was removed. The protein mixture (0.9, 1.2, 1.8, 2.4, 3, and 3.6 L) was then added into the centrifuge tube containing the gold nanoparticles pellet. The final volume was adjusted to 100 L by adding distilled water. The solution was then stirred for 10 min and stood for 2 h at room temperature without mixing, followed by addition of 5% (g/v) PEG20000 to a final concentration of 0.5% PEG20000. The PEG20000 can be used to stabilize and passivate the gold nanoparticles. Then, the AuNP probes were purified by centrifugation at 9000 rpm and 3–4 times washes with 1 mL of 10 mM PBS buffer (pH 7.2, 0.5% BSA, 2.5% sucrose and 0.1% PEG20000 included). The AuNP probes were redispersed finally in the same buffer (100 L) and were stored at 4 ◦ C. The successful immobilization of the antibody and streptavidin–HRP was confirmed with UV–Vis and TEM. And the immobilization of the streptavidin–HRP was also confirmed by adding a chromogenic enzymatic substrate to produce a visible signal. TMB is the specific enzymatic substrate for HRP. 0.2 L AuNP probes was added into 50 L TMB solution and the solution was incubated for 10–15 min avoid of light. After 10–15 min, the solution turns blue due to the soluble blue reaction product of peroxidase with TMB. The addition of an acidic stopping solution produces a yellow color that can be read at 450 nm and the intensity is proportional to the concentration of the enzyme label. The production of chromogenic products can be recorded with photos or the absorbance value was measured at 450 nm after blocking the reaction with 50 L 0.5 M H2 SO4 . 2.4. Quantification of HRP molecules immobilized on AuNP probes On the assumption that free and bound HRP molecules possessed similar activities, HRP molecules on AuNP probes were detected based on the standard spectrophotometric assay (Li et al.,
2838
C.-P. Jia et al. / Biosensors and Bioelectronics 24 (2009) 2836–2841
2008). 10 L diluted gold nanoparticle probes and a series of dilutions 10 L HRP solution (at concentration of 400, 200, 100, 50, 25, 12.5, 6.25, and 0 ng mL−1 ) were incubated in 50 L TMB–H2 O2 solution for 15 min avoid of light at 37 ◦ C. Then the absorbance value was measured at 450 nm after blocking the reaction with 50 L 0.5 M H2 SO4 solution. The activity of HRP immobilized on gold nanoparticle probes was detected and the amount of HRP molecules was determined based on the calibration curve from HRP of known concentrations. The concentration of AuNP probes (or the number of AuNPs) was determined from its absorbance at 528 nm. The number of HRP molecules on each gold nanoparticle was calculated by dividing the numbers of HRP molecules by the numbers of AuNPs. 2.5. Immunoassay The assay of the Nano-ELISA method includes these following steps. 1.5 L MMP probes (1 mg mL−1 ) modified with anti-p53 antibody was transferred into an eppendorf tube and mixed with 38.5 L p53 antigen with different concentrations at 2.5 × 105 , 1 × 105 , 2.5 × 104 , 6.25 × 103 , 1.5 × 103 , 370, 90, 22, 5, and 0 pg mL−1 diluted in assay buffer (10 mM PBS buffer containing 0.1% BSA). Two background proteins (500 ng mL−1 CEA and 500 ng mL−1 NSE) were included in the assay buffer. After 1 h incubation with gentle shaking, the eppendorf tube was played onto the magnet for
2–3 min. The mixtures of magnetic beads and target protein were collected and the supernatant was deleted. The mixtures were then resuspended with 200 L wash buffer (10 mM PBS, 0.1% BSA and 0.05% Tween-20 included) and collected again to wash away any excess proteins. This wash process should be done 3–4 times. After washing 3–4 times, the mixture was then resuspended with assay buffer and incubated with AuNPs probes for 30 min. Through the interaction between the p53 antibodies and p53 antigen, the sandwich structures (magnetic beads–target protein–AuNP probes) were formed and collected. After washing for 4 times, the mixture was incubated with 50 L TMB solution for 15 min avoid of light. All the incubation steps in this assay were performed at 37 ◦ C. Then the absorbance at 450 nm was measured after blocking the reaction with 50 L 0.5 M H2 SO4 . The assay of classic sandwich ELISA was done based on manufacture’s protocol. Briefly, 100 L p53 antigen with different concentrations was added into the wells of the plate. The plate was washed 4–5 times with 250 L 10 mM PBS buffer (0.1% BSA and 0.05% Tween-20 included) after incubation for 2 h. 100 L detector antibody was added into each well and given for the incubation for 1 h. After washing for 4–5 times, each well was added with 100 L anti-rabbit IgG (secondary antibody) HRP conjugate and incubated for 1 h. And, then the each well was incubated with 100 L TMB solution for 15 min avoid of light after washing for 4–5 times. All the incubation steps in this assay were also performed at 37 ◦ C. Then
Scheme 1. Schematic of preparation of AuNP probes and MMP probes (upper part) and Nano-ELISA procedure (lower part). MMPs were modified with p53 capture antibody and AuNP probes were modified with p53 detector antibody and streptavidin–HRP. First reaction step is the reaction of MMP probes and p53 protein and the second is the reaction of p53 protein and AuNP probes. The sandwich structures (magnetic beads–target protein–AuNP probes) were formed. After wash, the mixture was incubated with TMB solution and the absorbance value was measured.
C.-P. Jia et al. / Biosensors and Bioelectronics 24 (2009) 2836–2841
2839
the absorbance at 450 nm was measured after blocking the reaction with 100 L 0.5 M H2 SO4 . 3. Results and discussion The basic principle of this Nano-ELISA method is illustrated in Scheme 1. The capture monoclonal antibody is immobilized onto micro-magnetic particles. The detector monoclonal antibody and HRP are immobilized on the gold nanoparticles (AuNPs) (Scheme 1, part A). The detected antigen is sandwiched by capture antibody and detector antibody mobilized on MMPs and AuNPs, respectively (Scheme 1, part B). Unbound components are removed by washing in a magnetic field. The HRP molecules on AuNP probes are signal amplification probes that can catalyze the oxidation of TMB into colorful products to indicate the presence of the target antigen. The amount of the target antigen is proportional to the absorbance value of the formed colorful products. For classic indirect sandwich ELISA, the detector antibody modified with biotin can be detected through streptavidin coupled HRP, or can be detected by a secondary antibody linked with the HRP molecules (Crowther, 2000; Gosling, 2000). So, this classic indirect sandwich ELISA includes three reaction steps and four wash steps, which takes about 6 h. And this kind of detector antibody must be labeled with biotin, which limits the versatility of the test (Crowther, 2000; Gosling, 2000). For classic direct sandwich ELISA, similar with Nano-ELISA here, the detector antibody is conjugated with HRP directly through covalent bond. It includes two reaction steps and three wash steps, which takes about 4 h. Since a single enzyme-conjugated antibody is used and one antibody molecule can only eventually bind one HRP molecule, this direct sandwich system is limited to the sensitivity. While for the NanoELISA method, one detector antibody can be linked with many HRP molecules through nanoparticles, which can substantially increase the detection sensitivity. As shown in part B of Scheme 1, NanoELISA includes only two reaction steps and three wash steps, which can significantly simplify the whole detection assay and shorten the assay time. 3.1. Synthesis and characterization of AuNP probes In this method, gold nanoparticles were modified with detector antibody and streptavidin–HRP. The detector antibody is different from the capture antibody bound to micro-magnetic particles. The success of the immobilization was confirmed by TEM images and UV–vis measurements (Figure S1 of supporting information). The TEM images (Figure S1 B–D of supporting information) showed that antibodies and HRP molecules were conjugated to AuNPs. The small fuscous spots around the AuNPs could be iron metals present in the heme group of HRP molecules (Ambrosi et al., 2007). Furthermore, the optical spectra of the gold nanoparticles probes were recorded with an UV–vis spectrophotomer. Compared with bare gold nanoparticles, the absorption peak of modified gold nanoparticles (10 nm) shifted from 522 ± 0.76 to 532 ± 2.75 nm (not shown). This red shift of the plasmon band of gold is due to the coating of antibody and HRP (Ambrosi et al., 2007; Chen et al., 2008). And the immobilization of HRP was also confirmed by adding TMB solution to detect the absorbance at 450 nm. Figure S2 showed that only gold nanoparticles with HRP-binding had the activity of HRP. The absorbance at 450 nm reached to 3.5 after addition of TMB solution and the acidic stopping solution (Figure S2). 3.2. Effect of incubation time The incubation time of the first reaction step (the reaction between the capture antibody and target protein) was optimized by
Fig. 1. Detection results of human p53 protein (10 ng mL−1 ) using AuNP probes with different diameter (30, 15, and 10 nm). The concentration of AuNP probes was 500 nM. It can be seen that, the detection signal of antigen at same concentration with 10 nm AuNP probes was stronger than that with 15 and 30 nm probes.
varying the incubation time. More than 75% of the maximum signal was achieved after 30 min incubation. Further incubation only slightly increased the detection signal (data not shown). 30–60 min incubation was then practically used. The micro-magnetic particles as carriers can accelerate the immunoassay of protein or DNA hybridization significantly (Nam et al., 2003; Selvaraju et al., 2008). The incubation time of second reaction step was also optimized, which is the reaction time of detected protein and detector antibody immobilized onto gold nanoparticles. And this reaction can be completed in 30 min. Longer time, such as 60, 90, and 120 min, cannot increase the signal (data not shown). So, the entire Nano-ELISA procedure can be completed in 2 h (three wash steps included), which is 4 h shorter than that used in classic ELISA procedure. 3.3. Effect of the size of the gold nanoparticles probes The sizes of the AuNP probes, or the diameters of the AuNPs, can significantly affect the assay sensitivity. In this study, we compared three kinds of gold nanoparticles probes with 30, 15, and 10 nm in diameters. Fig. 1 shows the detection result of 10 ng mL−1 p53 protein with different AuNP probes. The absorbance with 30, 15, and 10 nm AuNP probes at 450 nm was 0.7365, 1.090, and 1.657, respectively (Fig. 1). The background signal (no p53 protein) was 0.076, 0.093, and 0.085. The results showed that the detection signal of antigen at same concentration with 10 nm AuNP probes was stronger than that with 15 and 30 nm probes (Fig. 1). This difference maybe caused by the variant of the specific surface area of gold nanoparticles. 10 nm gold nanoparticles have the higher surface-to-volume ratio than 15 and 30 nm nanoparticles. Furthermore, the size of most antibody is about 5–9 nm (Saleh and Sohn, 2003; Ambrosi et al., 2007) and the antibody on the 10 nm AuNP probes maybe more easy to overcome steric hindrance effect than 15 nm or larger AuNP probes (Kent et al., 1978; Jie et al., 2008). 3.4. Effect of different concentration of MMP probes and AuNP probes In this immunoassay, various parameters, such as concentrations of MMP probes and AuNP probes should were optimized. A small quantity of probes can break weak signal, as well as low detection sensitivity. Deficient MMP probes with capture antibody cannot capture all the detected protein, which can break weak signal or even lead to the false negative. Excess probes can lead to high nonspecific backgrounds. MMP probes were used with different con-
2840
C.-P. Jia et al. / Biosensors and Bioelectronics 24 (2009) 2836–2841
Fig. 2. Concentration optimization graph of AuNP probes. AuNP probes were used with different concentrations at 0, 100, 200, 300, 400, 500, and 600 nM. The concentration of p53 protein detected was 10 ng mL−1 and the concentration of MMP probes was 1.5 mg mL−1 . The best concentration of AuNP probes is 500 nM.
centrations at 0, 0.25, 0.5, 1, 1.5, and 2 mg mL−1 . The concentration of p53 protein detected was 10 ng mL−1 and the concentration of AuNP probes was 500 nM. The results showed us that MMP probes at concentration of 1.5 mg mL−1 were the best to be used (Figure S3). AuNP probes with different concentrations at 0, 100, 200, 300, 400, 500, and 600 nM were used (Fig. 2). The concentration of p53 protein detected was also 10 ng mL−1 and the concentration of MMP probes was 1.5 mg mL−1 . Fig. 4 showed the optimization of the AuNPs probes concentration and the best used concentration of AuNP probes was 500 nM for this immunoassay system. 3.5. Effect of different ratios of antibody to streptavidin–HRP molecules AuNP probes were immobilized with detector antibody and streptavidin–HRP. To confirm that the streptavidin–HRP cannot bind with antibody but with gold nanoparticles, the following experiments were devised. In process of Nano-ELISA, after the mixtures of magnetic beads and target p53 protein (10 ng mL−1 ) was washed, AuNP probes (immobilized with p53 detector antibody and streptavidin–HRP), p53 detector antibody, streptavidin–HRP, the mixtures of p53 detector antibody and streptavidin–HRP, were added, separately. After incubation and washing for 4 times, TMB solution was added. Only the group with AuNP probes (immobilized with p53 detector antibody and streptavidin–HRP) added had the chromogenic products and had the absorbance (1.6) at 450 nm after addition with stop solution, which confirmed that the p53 detector antibody could not bind with streptavidin–HRP. If the streptavidin–HRP can bind with p53 detector antibody unspecifically, the absorbance of the group with addition of the mixtures should be obvious. In fact, the absorbance was very low and was backgrounds signal, which also confirmed that the p53 detector antibody could not bind with streptavidin–HRP. The results were shown in Figure S4. Detector antibody immobilized onto AuNP probes is a “bridge” of detected protein and detection signal (HRP). Because the activity of HRP rather than the detected antibody is detected in this assay, different proportion of these kinds of molecules can affect the detection signal significantly. Fig. 3 showed the optimization result of ratios of HRP to antibody molecules when 10 nm AuNPs were immobilized with these two kinds of molecules. The ratio of these two molecules was 0.5, 1, 2, 3, 4, and 5. Results showed that the detection signal was the best when the ratio of these two molecules was about 3:1 (Fig. 3). And based on quantification of HRP molecules immobilized on the gold nanoparticles probes,
Fig. 3. Optimization graph of molecule ratio of HRP and antibody when 10 nm AuNPs were immobilized with these two kinds of molecules. The molecule ratio is 0.5, 1, 2, 3, 4, and 5. The concentration of p53 protein detected was 10 ng mL−1 and the concentration of MMP probes was 1.5 mg mL−1 . And the graph shows that the detection signal is the best when the ratio of these two molecules is about 3:1.
the number of HRP molecules was detected on each 10 nm gold nanoparticle immobilized with HRP and antibody at this ratio. The concentration of streptavidin–HRP immobilized onto AuNP probes was 1.325 × 10−2 g L−1 based on the activity of HRP and the calibration curve of streptavidin–HRP (Figure S5). The molecular weight of streptavidin–HRP is 100 KDa (Dalton). So, the concentration was 1.325 × 10−7 M. The concentration of AuNP probes was 0.238 × 10−7 M based on the absorption of AuNP probes at 528 nm. Therefore, the concentration ratio of streptavidin–HRP and gold nanoparticles was 5.6, which meant that there were about 5–6 HRP molecules on each 10 nm gold nanoparticle. The sandwich complex will show reactive activity relative to 5–6 HRP molecules once bound one gold nanoparticle probe through immunoassay, which can amplify the detection signal. The previous reports showed that there were 10 anti-humanIgG–HRP molecules on each 13 nm gold nanoparticle (Ambrosi et al., 2007). And 7 HRP molecules could be adsorbed onto one 15 nm gold nanoparticle immobilized together with DNA probes (Li et al., 2008). So, there is a good correspondence between our and previous results. And there are about 1–2 anti-p53 detector antibody molecules together with 5.6 HRP molecules on each 10 nm gold nanoparticle. 3.6. The preliminary application of Nano-ELISA for detecting p53 protein Fig. 4 showed the caliberation curve of Nano-ELISA for p53 protein with different concentrations of 2.5 × 105 , 1 × 105 , 2.5 × 104 , 6.25 × 103 , 1.5 × 103 , 370, 90, 22, and 5 pg mL−1 . The assay buffer containing two background proteins was also detected as the background signals. Low to 5 pg mL−1 p53 protein was detected (S/N ≥ 2). And the serum sample (50 ng mL−1 p53 protein) of lung cancer was diluted to 500, 50, 5 pg mL−1 and detected. The results were shown in Fig. 5. The p53 protein in these serum dilutions was also detected. The absorbance (450 nm) of 5 pg mL−1 p53 protein from the recombined p53 protein with only two background proteins and the serum dilutions, were 0.12 and 0.18, separately. So, this Nano-ELISA method can be used to detect the serum samples. The detection sensitivity of p53 classic ELISA kit sold by Upstate is 0.125 ng mL−1 (data not shown) (www.Upstate.com). And for another commercial p53 ELISA kit sold by Roche, the detection limit of p53 protein is 1.6 ng mL−1 (www.Roche.com). The detection limit obtained using Nano-ELISA method is much lower than that using classic ELISA method. And the two ELISA kits should
C.-P. Jia et al. / Biosensors and Bioelectronics 24 (2009) 2836–2841
2841
nanoparticles here were used as carrier and amplifier. Through immunoassay of antibody and antigen, the sandwich structures (magnetic beads–target protein–AuNP probes) were formed. And then, the magnetic field was used to collect the sandwich structures. Excess components were removed through washing for 2–3 times. Protein was detected based on absorbance value. Low to 5 pg mL−1 of protein was detected with this method. The assay time was less than 2 h. This method can be used to detect protein markers of tumors, nervous system or other diseases for early diagnostics. Acknowledgements
Fig. 4. Calibration curves of human p53 protein at concentration of 2.5 × 105 , 1 × 105 , 2.5 × 104 , 6.25 × 103 , 1.5 × 103 , 370, 90, 22, 5, and 0 pg mL−1 diluted in assay buffer (10 mM PBS buffer containing 0.1% BSA, 500 ng mL−1 CEA and 500 ng mL−1 NSE as background proteins) with 10 nm AuNP probes. The detection limit is 5 pg mL−1 with 10 nm AuNP probes.
This work was supported by the National High Technology Research and Development Program (“863” Program) of China (No. 2006AA03Z334), Ministry of Science and Technology (No. 2007CB936000) and Shanghai Municipal Commission for Science and Technology (No. 0752nm019; 0652nm016; 0752nm021). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bios.2009.02.024. References
Fig. 5. Detection results of p53 protein in serum sample dilutions. The serum sample (50 ng mL−1 p53 protein) of lung cancer was diluted to 500, 50, 5 pg mL−1 in assay buffer (10 mM PBS buffer containing 0.1% BSA, 500 ng mL−1 CEA and 500 ng mL−1 NSE as background proteins) and detected. It can be seen that low to 5 pg mL−1 p53 protein in serum sample can be detected out.
conduct many amplification steps from polyclonal antibodies and biotin–streptavidin system, including three reaction steps and four wash steps, which is tedious and take more time. So, use of gold nanoparticles as the carrier and amplifier can increase the sensitivity of immunoassay significantly and Nano-ELISA is more effective and time saving. The basic operation and instrument of Nano-ELISA is similar with ELISA, which can help this method to be used more widely in clinic and lab than other methods such as, electrochemistry. 4. Conclusions Nano-ELISA, a highly sensitive protein detection method, was introduced here. Gold nanoparticles (AuNPs) were modified with a monoclonal detector antibody and HRP (signal molecules). Gold
Alivisators, P., 2004. Nat. Biotechnol. 22 (1), 47–52. Ambrosi, A., Castaneda, M.T., Killard, A.J., Smyth, M.R., Alegret, S., Merkoci, A., 2007. Anal. Chem. 79 (14), 5232–5240. Bao, Y.P., Wei, T.F., Lefebvre, P.A., Hao An, He, L.X., Kunkel, G.T., Muller, U.R., 2006. Anal. Chem. 78 (6), 2055–2059. Castaneda, M.T., Alegret, S., Merkoci, A., 2007. Electroanalysis 19, 743–753. Chen, Y., Yu, C., Cheng, T., Tseng, W., 2008. Langmuir 24 (7), 3654–3660. Cressey, R., Pimpa, S., Chewaskulyong, B., Lertprasertsuke, N., Saeteng, S., Tayapiwatana, C., Kasinrerk, W., 2008. BMC Biotechnol. 8, 16–29. Crowther, J.R., 2000. The Elisa Guidebook. The International Atomic Energy Agency, Vienna. Cui, R., Huang, H., Yin, Z., Gao, D., Zhu, J., 2008. Biosens. Bioelectron. 23, 1666–1673. Engvall, E., Perlman, P., 1971. Immunochemistry 8 (9), 871–874. Georganopoulou, D.G., Chang, L., Nam, J.M., Thaxton, C.S., Mufson, E.J., William, L., Klein, W.L., Mirkin, C.A., 2005. Proc. Natl. Acad. Sci. 102 (7), 2273–2276. Gosling, J.P. (Ed.), 2000. Immunoassays: Practical Approach. Great Clarendon, Oxford. Hermanson, G.T., 1996. Bioconjugate Techniques, first ed. Academic Press, San Diego. Jain, P.K., Lee, K.S., El-Sayed, I.H., 2006. J. Phys. Chem. B 110 (14), 7238–7248. Jie, G., Huang, H., Sun, X., Zhu, J., 2008. Biosens. Bioelectron. 23, 1896–1899. Kent, S.P., Ryan, K.H., Siegel, A.L., 1978. J. Histochem. Cytochem. 26 (8), 618–621. Koppelman, S.J., Lakemond, C.M., Vlooswijk, R., Hefle, S.L., 2004. J. AOAC Int. 87 (6), 1398–1407. Lequin, R., 2005. Clin. Chem. 51 (12), 2415–2418. Li, J., Song, S., Liu, X., Wang, L., Pan, D., Huang, Q., Zhao, Y., Fan, C., 2008. Adv. Mater. 20, 497–500. Matakidou, A., El Galta, R., Webb, E.L., Rudd, M.F., Bridle, H., Eisen, T., Houlston, R.S., Consortium, G., 2007. Lung Cancer 57 (2), 207–212. Nam, J.M., Thaxton, C.S., Mirkin, C.A., 2003. Science 301, 1884–1886. ´ Porebska, I., Wyrodek, E., Kosacka, M., Adamiak, J., Jankowska, R., HarłozinskaSzmyrka, A., 2006. In Vivo 20 (5), 599–604. Rosi, N.L., Mirkin, R.C., 2005. Chem. Rev. 105 (4), 1547–1550. Roth, J., 1982. J. Histochem. Cytochem. 30, 691–696. Saleh, O.A., Sohn, L.L., 2003. Proc. Natl. Acad. Sci. 100 (3), 820–824. Selvaraju, T., Das, J., Han, S., Yang, H., 2008. Biosens. Bioelectron. 23, 932–938. Thaxton, C.S., Hill, H.D., Georganopoulou, D.G., Stoeva, S.I., Mirkin, C.A., 2005. Anal. Chem. 77 (24), 8174–8178. Turkevich, J., Stevenson, P., Hiller, J., 1951. Discuss. Faraday Soc. 11, 55–75. Wang, J., Zhu, X., Tu, Q., Guo, Q., Zarui, C.S., Momand, J., Sun, X., Zhou, F., 2008. Anal. Chem. 80 (3), 769–774. Yan, G., Xing, D., Tan, S., Chen, Q., 2004. J. Immunol. Methods 288 (1–2), 47–54. Zhang, J., Song, S., Zhang, L., Wang, L., Wu, H., Pan, D., Fan, C., 2006. J. Am. Chem. Soc. 128, 8575–8580. Zhang, J., Song, S., Wang, L., Pan, D., Fan, C., 2007. Nat. Protocols 2, 2888–2895.
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Nano-ELISA for highly sensitive protein detection Chun-Ping Jia a , Xiao-Qin Zhong a , Bao Hua a , Mei-Ying Liu a , Feng-Xiang Jing a , Xin-Hui Lou a , Shi-Hua Yao b , Jia-Qing Xiang b , Qing-Hui Jin a,∗ , Jian-Long Zhao a,∗∗ a b
Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Science, 865 Changning Road, Shanghai, 200050, PR China Shanghai Tumor Hospital, Shanghai, 200050, PR China
a r t i c l e
i n f o
Article history: Received 7 November 2008 Received in revised form 22 January 2009 Accepted 12 February 2009 Available online 6 March 2009 Keywords: Gold nanoparticles Immunoassay Nano-ELISA ELISA p53
a b s t r a c t Highly sensitive protein detection method based on nanoparticles and enzyme-linked immunosorbent assays (ELISAs), named Nano-ELISA, was introduced. In this method, the micro-magnetic beads were modified with monoclonal antibody of the target protein p53. Gold nanoparticles (AuNPs) were modified with another monoclonal detector antibody and Horseradish peroxidase (HRP, for signal amplification). The presence of target protein p53 causes the formation of the sandwich structures (magnetic beads–target protein–AuNP probes) through the interaction between the antibodies and the antigen p53. The HRP at the surface of AuNPs catalytically oxidize the substrate and generate optical signals that reflected the quantity of the target protein. Down to 5 pg mL−1 of protein was detected in less than 2 h with this method. The detection sensitivity of p53 classic ELISA kit is 0.125 ng mL−1 . This method is as simple as ELISA and has higher sensitivity than ELISA, which can potentially be exploited in clinic. This method can be used to detect protein markers of tumors, nervous system or other diseases for early diagnostics. © 2009 Elsevier B.V. All rights reserved.
1. Introduction ELISA is one of the most important bio-chemical techniques used mainly to detect the presence of antibodies or antigens in a sample based on antibody–antigen immunoreactions. Due to its simplicity, low-cost, easy reading, acceptability and safety (Engvall and Perlman, 1971; Lequin, 2005), ELISA is widely used for detection of cancer protein markers, pathogen, and other proteins relative to various diseases, with the detection limit from 0.1 ng to 1 g mL−1 (Engvall and Perlman, 1971; Koppelman et al., 2004). However, the level of relative tumor protein marker is very low at the early stage of cancers or other diseases, which is beyond the detection limit of ELISA. And, most of classic ELISA, such as sandwich ELISA (capture antibody and detector antibody involved), requires long time and multiple steps, which limits the real-life application in clinics. There is a need to improve the sensitivity of the current ELISA method for highly sensitive detection protein biomarker, which is important for early diagnosis of cancer, neurodegenerative and other diseases. Gold nanoparticles, having high surface areas and unique physicochemical properties, are used widely in developing biomarker platforms (Alivisators, 2004; Rosi and Mirkin, 2005; Jain et al., 2006; Castaneda et al., 2007). Gold nanoparticles can be con-
∗ Corresponding author. Tel.: +86 21 62511070x8703; fax: +86 21 62511070x8714. ∗∗ Corresponding author. Tel.: +86 21 62511070x8702; fax: +86 21 62511070x8714. E-mail addresses: [email protected] (Q.-H. Jin), [email protected] (J.-L. Zhao). 0956-5663/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.bios.2009.02.024
jugated with DNAs, antibodies, enzymes and other bio-molecules, which can afford them promising applications in signal enhancement of bio-chemical detection (Zhang et al., 2006; Ambrosi et al., 2007; Zhang et al., 2007; Cui et al., 2008; Selvaraju et al., 2008). The recently developed nanoparticle-based bio-bar-code assay (BCA) has low attomolar sensitivity for protein targets and high zeptomolar sensitivity for nucleic acid targets (Nam et al., 2003; Thaxton et al., 2005; Georganopoulou et al., 2005; Bao et al., 2006). This detection system utilizes short oligonucleotides immobilized on AuNPs with antibody as target identification strands and surrogate amplification units in protein detection. Although this method is extraordinarily sensitive, the cumbersome experimental steps limit its application in clinic. A double-codified gold nanolabels for enhanced immunoanalysis was developed recently by Ambrosi et al. (2007). In their method, AuNPs were modified with anti-human-IgG–HRP-conjugated antibody. HRP was used here for signal amplification, which could efficiently catalyze 10,000 substrate turnovers, which significantly improved the detection limit. HRP can oxidize various chromogenic substrates and produce visible product, which can be easily detected. The protein concentration was measured spectrophotometrically based on HRP label activity and quantified by stripping voltammetry based on the intrinsic electrochemical properties of gold nanoparticles. One major drawback of this method is that the detector antibody used in this method needs to be conjugated with HRP and the process of creating HRP-linked antibodies is very complex and expensive (Crowther, 2000; Gosling, 2000), which limits its application widely.
C.-P. Jia et al. / Biosensors and Bioelectronics 24 (2009) 2836–2841
The p53 gene is a cancer suppressor gene and the p53 protein is a transcriptional factor protein that plays important roles in cell cycles control, cell growth, DNA repair and cell apoptosis. The p53 gene mutation was found in over 50% of human cancers and the level of p53 protein in serum and tissues was increased in response to these mutations and the development of cancers. So, p53 protein can be a marker for cancer diagnostic (Porebska et al., 2006; Matakidou et al., 2007). The most common method for the detection of p53 protein is classic sandwich ELISA, which requires a long time with multiple steps and low sensitivity. So, more detection methods should be developed. Several reports have described the new detection methods of p53 protein and antibodies against p53 protein (Yan et al., 2004; Cressey et al., 2008; Wang et al., 2008). These antibodies are products of immunoresponse against p53 protein and can indirectly show the level of p53 protein in serum. In Wang’s work, double-stranded oligonucleotides containing the p53 protein binding sites are immobilized onto gold electrodes to capture p53 protein based on the highly specific and strong DNA-binding property of p53 protein. The use of DNA-modified electrodes to capture p53 and the follow-up amplification of the voltammetric signals using the gold nanoparticle afforded the high sensitivity and selectivity necessary for detecting p53. The detection limit of p53 protein was 2.2 pM (116 pg mL−1 ) (Wang et al., 2008). In this study, we demonstrated a very simple and more sensitive protein detection method, termed Nano-ELISA, which combines the nanotechnology with the classic sandwich ELISA method. In NanoELISA, the detector antibody mixed with HRP at certain ratio was immobilized on the gold nanoparticles and HRP was used here for signal amplification. As a proof of concept, we chose p53 protein as target antigen to demonstrate the feasibility of Nano-ELISA method. The un-optimized detection limit of this method is 5.7 pg mL−1 and the assay time was less than 2 h. 2. Experimental 2.1. Materials and instrumentation HAuCl4 ·3H2 O, trisodium citrate, sodium chloride (NaCl), 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), 2-(N-morpholino)ethane sulfonic acid (MES), streptavidin–HRP, and 3,3 ,5,5 -tetramethylbenzidine (TMB)–H2 O2 were purchased from Sigma–Aldrich Corp. (USA). p53 antibody was got from Abcam Corp. (USA). Micromagnetic particles (1 m in diameter) (MMP) were purchased from Dynal Invitrogen Corp. (USA). p53 ELISA kit was purchased from Upstate Corp. (USA). All other reagents were in analytical grade. We used a Hitachi UV-3010 UV–vis spectrophotometer to characterize the optical properties of citrate, and antibody coated gold nanoparticles. The prepared gold nanoparticles were diluted with double distilled water before transferring them to the sample sell. The scanning wavelength was set from 200 to 700 nm for all the measurements of Au nanoparticle conjugates. Transmission electron microscopy (TEM) was performed with an H-600 transmission electron microscope. A typical sample was prepared by dropping 10 L of the nanoparticle solution onto a copper TEM grid with formvar/carbon supportive films, followed by staining with uranyl acetate and lead citrate. The grid was subsequently dried in the air and imaged. 2.2. Preparation of MMP probes The MMPs were functionalized with anti-p53 capture antibody according to the manufacture’s protocol. Briefly, the carboxylic acid groups on the surface of MMPs (300 L stock solution) were activated by incubating with EDC (150 L, 50 mg mL−1 ) and NHS (150 L, 50 mg mL−1 ) solution in 25 mM MES, pH 6 at room temper-
2837
ature for 30 min. After incubation, the activated beads were washed and 150 g anti-p53 antibody was then added and incubated at room temperature for 30 min with slow rotation. The MMPs were washed 3–4 times with the washing buffer (10 mM PBS, pH 7.2, 0.1% BSA, 0.05% Tween-20) on the magnet. The remaining active sites on MMPs were blocked by incubating the washed MMPs with 10 mM PBS buffer (pH 7.2, 2% BSA, 2.5% sucrose and 0.1% PEG20000 included) for 30 min at room temperature. The MMPs were finally reconstituted in 240 L washing buffer and stored at 4 ◦ C. 2.3. Preparation and characterization of gold nanoparticles probes Gold nanoparticles (10 and 15 nm in diameter) were prepared according to previously published protocols (Turkevich et al., 1951). In particular, 100 mL double distilled water and 1 mL 1% HAuCl4 was mixed and boiled. 3 or 2 mL 1% trisodium citrate solution for the preparation of 10 and 15 nm AuNPs, respectively, was quickly added to the refluxed HAuCl4 solution, resulting in a rapid color change from pale yellow to deep red that indicated the formation of gold nanoparticles. After a continuous reflux for an additional 15 min the solution was slowly cooled down to room temperature and the particle solution was filtered through a 0.22-m cellulose nitrate filter to remove any floating aggregates. The quality of the particles was monitored with UV–vis and TEM. All glassware and the stir bar used in the preparation was cleaned in aqua regia (3:1 HCl:HNO3 ), rinsed with Millipore water, and then dried in an oven. Gold nanoparticles–protein conjugations were prepared by following the published procedures (Roth, 1982; Hermanson, 1996; Ambrosi et al., 2007). The pH of gold nanoparticles solution was adjusted to 8.2–8.5 with 0.2 M K2 CO3 . Streptavidin–HRP stock solution (1 mg mL−1 ) and p53 detector monoclonal antibody stock solution (1 mg mL−1 ) were mixed at different ratios (0.5, 1, 2, 3, 4, and 5). 1 mL gold nanoparticles solution was centrifuged at 9000 rpm for 50 min and the supernatant was removed. The protein mixture (0.9, 1.2, 1.8, 2.4, 3, and 3.6 L) was then added into the centrifuge tube containing the gold nanoparticles pellet. The final volume was adjusted to 100 L by adding distilled water. The solution was then stirred for 10 min and stood for 2 h at room temperature without mixing, followed by addition of 5% (g/v) PEG20000 to a final concentration of 0.5% PEG20000. The PEG20000 can be used to stabilize and passivate the gold nanoparticles. Then, the AuNP probes were purified by centrifugation at 9000 rpm and 3–4 times washes with 1 mL of 10 mM PBS buffer (pH 7.2, 0.5% BSA, 2.5% sucrose and 0.1% PEG20000 included). The AuNP probes were redispersed finally in the same buffer (100 L) and were stored at 4 ◦ C. The successful immobilization of the antibody and streptavidin–HRP was confirmed with UV–Vis and TEM. And the immobilization of the streptavidin–HRP was also confirmed by adding a chromogenic enzymatic substrate to produce a visible signal. TMB is the specific enzymatic substrate for HRP. 0.2 L AuNP probes was added into 50 L TMB solution and the solution was incubated for 10–15 min avoid of light. After 10–15 min, the solution turns blue due to the soluble blue reaction product of peroxidase with TMB. The addition of an acidic stopping solution produces a yellow color that can be read at 450 nm and the intensity is proportional to the concentration of the enzyme label. The production of chromogenic products can be recorded with photos or the absorbance value was measured at 450 nm after blocking the reaction with 50 L 0.5 M H2 SO4 . 2.4. Quantification of HRP molecules immobilized on AuNP probes On the assumption that free and bound HRP molecules possessed similar activities, HRP molecules on AuNP probes were detected based on the standard spectrophotometric assay (Li et al.,
2838
C.-P. Jia et al. / Biosensors and Bioelectronics 24 (2009) 2836–2841
2008). 10 L diluted gold nanoparticle probes and a series of dilutions 10 L HRP solution (at concentration of 400, 200, 100, 50, 25, 12.5, 6.25, and 0 ng mL−1 ) were incubated in 50 L TMB–H2 O2 solution for 15 min avoid of light at 37 ◦ C. Then the absorbance value was measured at 450 nm after blocking the reaction with 50 L 0.5 M H2 SO4 solution. The activity of HRP immobilized on gold nanoparticle probes was detected and the amount of HRP molecules was determined based on the calibration curve from HRP of known concentrations. The concentration of AuNP probes (or the number of AuNPs) was determined from its absorbance at 528 nm. The number of HRP molecules on each gold nanoparticle was calculated by dividing the numbers of HRP molecules by the numbers of AuNPs. 2.5. Immunoassay The assay of the Nano-ELISA method includes these following steps. 1.5 L MMP probes (1 mg mL−1 ) modified with anti-p53 antibody was transferred into an eppendorf tube and mixed with 38.5 L p53 antigen with different concentrations at 2.5 × 105 , 1 × 105 , 2.5 × 104 , 6.25 × 103 , 1.5 × 103 , 370, 90, 22, 5, and 0 pg mL−1 diluted in assay buffer (10 mM PBS buffer containing 0.1% BSA). Two background proteins (500 ng mL−1 CEA and 500 ng mL−1 NSE) were included in the assay buffer. After 1 h incubation with gentle shaking, the eppendorf tube was played onto the magnet for
2–3 min. The mixtures of magnetic beads and target protein were collected and the supernatant was deleted. The mixtures were then resuspended with 200 L wash buffer (10 mM PBS, 0.1% BSA and 0.05% Tween-20 included) and collected again to wash away any excess proteins. This wash process should be done 3–4 times. After washing 3–4 times, the mixture was then resuspended with assay buffer and incubated with AuNPs probes for 30 min. Through the interaction between the p53 antibodies and p53 antigen, the sandwich structures (magnetic beads–target protein–AuNP probes) were formed and collected. After washing for 4 times, the mixture was incubated with 50 L TMB solution for 15 min avoid of light. All the incubation steps in this assay were performed at 37 ◦ C. Then the absorbance at 450 nm was measured after blocking the reaction with 50 L 0.5 M H2 SO4 . The assay of classic sandwich ELISA was done based on manufacture’s protocol. Briefly, 100 L p53 antigen with different concentrations was added into the wells of the plate. The plate was washed 4–5 times with 250 L 10 mM PBS buffer (0.1% BSA and 0.05% Tween-20 included) after incubation for 2 h. 100 L detector antibody was added into each well and given for the incubation for 1 h. After washing for 4–5 times, each well was added with 100 L anti-rabbit IgG (secondary antibody) HRP conjugate and incubated for 1 h. And, then the each well was incubated with 100 L TMB solution for 15 min avoid of light after washing for 4–5 times. All the incubation steps in this assay were also performed at 37 ◦ C. Then
Scheme 1. Schematic of preparation of AuNP probes and MMP probes (upper part) and Nano-ELISA procedure (lower part). MMPs were modified with p53 capture antibody and AuNP probes were modified with p53 detector antibody and streptavidin–HRP. First reaction step is the reaction of MMP probes and p53 protein and the second is the reaction of p53 protein and AuNP probes. The sandwich structures (magnetic beads–target protein–AuNP probes) were formed. After wash, the mixture was incubated with TMB solution and the absorbance value was measured.
C.-P. Jia et al. / Biosensors and Bioelectronics 24 (2009) 2836–2841
2839
the absorbance at 450 nm was measured after blocking the reaction with 100 L 0.5 M H2 SO4 . 3. Results and discussion The basic principle of this Nano-ELISA method is illustrated in Scheme 1. The capture monoclonal antibody is immobilized onto micro-magnetic particles. The detector monoclonal antibody and HRP are immobilized on the gold nanoparticles (AuNPs) (Scheme 1, part A). The detected antigen is sandwiched by capture antibody and detector antibody mobilized on MMPs and AuNPs, respectively (Scheme 1, part B). Unbound components are removed by washing in a magnetic field. The HRP molecules on AuNP probes are signal amplification probes that can catalyze the oxidation of TMB into colorful products to indicate the presence of the target antigen. The amount of the target antigen is proportional to the absorbance value of the formed colorful products. For classic indirect sandwich ELISA, the detector antibody modified with biotin can be detected through streptavidin coupled HRP, or can be detected by a secondary antibody linked with the HRP molecules (Crowther, 2000; Gosling, 2000). So, this classic indirect sandwich ELISA includes three reaction steps and four wash steps, which takes about 6 h. And this kind of detector antibody must be labeled with biotin, which limits the versatility of the test (Crowther, 2000; Gosling, 2000). For classic direct sandwich ELISA, similar with Nano-ELISA here, the detector antibody is conjugated with HRP directly through covalent bond. It includes two reaction steps and three wash steps, which takes about 4 h. Since a single enzyme-conjugated antibody is used and one antibody molecule can only eventually bind one HRP molecule, this direct sandwich system is limited to the sensitivity. While for the NanoELISA method, one detector antibody can be linked with many HRP molecules through nanoparticles, which can substantially increase the detection sensitivity. As shown in part B of Scheme 1, NanoELISA includes only two reaction steps and three wash steps, which can significantly simplify the whole detection assay and shorten the assay time. 3.1. Synthesis and characterization of AuNP probes In this method, gold nanoparticles were modified with detector antibody and streptavidin–HRP. The detector antibody is different from the capture antibody bound to micro-magnetic particles. The success of the immobilization was confirmed by TEM images and UV–vis measurements (Figure S1 of supporting information). The TEM images (Figure S1 B–D of supporting information) showed that antibodies and HRP molecules were conjugated to AuNPs. The small fuscous spots around the AuNPs could be iron metals present in the heme group of HRP molecules (Ambrosi et al., 2007). Furthermore, the optical spectra of the gold nanoparticles probes were recorded with an UV–vis spectrophotomer. Compared with bare gold nanoparticles, the absorption peak of modified gold nanoparticles (10 nm) shifted from 522 ± 0.76 to 532 ± 2.75 nm (not shown). This red shift of the plasmon band of gold is due to the coating of antibody and HRP (Ambrosi et al., 2007; Chen et al., 2008). And the immobilization of HRP was also confirmed by adding TMB solution to detect the absorbance at 450 nm. Figure S2 showed that only gold nanoparticles with HRP-binding had the activity of HRP. The absorbance at 450 nm reached to 3.5 after addition of TMB solution and the acidic stopping solution (Figure S2). 3.2. Effect of incubation time The incubation time of the first reaction step (the reaction between the capture antibody and target protein) was optimized by
Fig. 1. Detection results of human p53 protein (10 ng mL−1 ) using AuNP probes with different diameter (30, 15, and 10 nm). The concentration of AuNP probes was 500 nM. It can be seen that, the detection signal of antigen at same concentration with 10 nm AuNP probes was stronger than that with 15 and 30 nm probes.
varying the incubation time. More than 75% of the maximum signal was achieved after 30 min incubation. Further incubation only slightly increased the detection signal (data not shown). 30–60 min incubation was then practically used. The micro-magnetic particles as carriers can accelerate the immunoassay of protein or DNA hybridization significantly (Nam et al., 2003; Selvaraju et al., 2008). The incubation time of second reaction step was also optimized, which is the reaction time of detected protein and detector antibody immobilized onto gold nanoparticles. And this reaction can be completed in 30 min. Longer time, such as 60, 90, and 120 min, cannot increase the signal (data not shown). So, the entire Nano-ELISA procedure can be completed in 2 h (three wash steps included), which is 4 h shorter than that used in classic ELISA procedure. 3.3. Effect of the size of the gold nanoparticles probes The sizes of the AuNP probes, or the diameters of the AuNPs, can significantly affect the assay sensitivity. In this study, we compared three kinds of gold nanoparticles probes with 30, 15, and 10 nm in diameters. Fig. 1 shows the detection result of 10 ng mL−1 p53 protein with different AuNP probes. The absorbance with 30, 15, and 10 nm AuNP probes at 450 nm was 0.7365, 1.090, and 1.657, respectively (Fig. 1). The background signal (no p53 protein) was 0.076, 0.093, and 0.085. The results showed that the detection signal of antigen at same concentration with 10 nm AuNP probes was stronger than that with 15 and 30 nm probes (Fig. 1). This difference maybe caused by the variant of the specific surface area of gold nanoparticles. 10 nm gold nanoparticles have the higher surface-to-volume ratio than 15 and 30 nm nanoparticles. Furthermore, the size of most antibody is about 5–9 nm (Saleh and Sohn, 2003; Ambrosi et al., 2007) and the antibody on the 10 nm AuNP probes maybe more easy to overcome steric hindrance effect than 15 nm or larger AuNP probes (Kent et al., 1978; Jie et al., 2008). 3.4. Effect of different concentration of MMP probes and AuNP probes In this immunoassay, various parameters, such as concentrations of MMP probes and AuNP probes should were optimized. A small quantity of probes can break weak signal, as well as low detection sensitivity. Deficient MMP probes with capture antibody cannot capture all the detected protein, which can break weak signal or even lead to the false negative. Excess probes can lead to high nonspecific backgrounds. MMP probes were used with different con-
2840
C.-P. Jia et al. / Biosensors and Bioelectronics 24 (2009) 2836–2841
Fig. 2. Concentration optimization graph of AuNP probes. AuNP probes were used with different concentrations at 0, 100, 200, 300, 400, 500, and 600 nM. The concentration of p53 protein detected was 10 ng mL−1 and the concentration of MMP probes was 1.5 mg mL−1 . The best concentration of AuNP probes is 500 nM.
centrations at 0, 0.25, 0.5, 1, 1.5, and 2 mg mL−1 . The concentration of p53 protein detected was 10 ng mL−1 and the concentration of AuNP probes was 500 nM. The results showed us that MMP probes at concentration of 1.5 mg mL−1 were the best to be used (Figure S3). AuNP probes with different concentrations at 0, 100, 200, 300, 400, 500, and 600 nM were used (Fig. 2). The concentration of p53 protein detected was also 10 ng mL−1 and the concentration of MMP probes was 1.5 mg mL−1 . Fig. 4 showed the optimization of the AuNPs probes concentration and the best used concentration of AuNP probes was 500 nM for this immunoassay system. 3.5. Effect of different ratios of antibody to streptavidin–HRP molecules AuNP probes were immobilized with detector antibody and streptavidin–HRP. To confirm that the streptavidin–HRP cannot bind with antibody but with gold nanoparticles, the following experiments were devised. In process of Nano-ELISA, after the mixtures of magnetic beads and target p53 protein (10 ng mL−1 ) was washed, AuNP probes (immobilized with p53 detector antibody and streptavidin–HRP), p53 detector antibody, streptavidin–HRP, the mixtures of p53 detector antibody and streptavidin–HRP, were added, separately. After incubation and washing for 4 times, TMB solution was added. Only the group with AuNP probes (immobilized with p53 detector antibody and streptavidin–HRP) added had the chromogenic products and had the absorbance (1.6) at 450 nm after addition with stop solution, which confirmed that the p53 detector antibody could not bind with streptavidin–HRP. If the streptavidin–HRP can bind with p53 detector antibody unspecifically, the absorbance of the group with addition of the mixtures should be obvious. In fact, the absorbance was very low and was backgrounds signal, which also confirmed that the p53 detector antibody could not bind with streptavidin–HRP. The results were shown in Figure S4. Detector antibody immobilized onto AuNP probes is a “bridge” of detected protein and detection signal (HRP). Because the activity of HRP rather than the detected antibody is detected in this assay, different proportion of these kinds of molecules can affect the detection signal significantly. Fig. 3 showed the optimization result of ratios of HRP to antibody molecules when 10 nm AuNPs were immobilized with these two kinds of molecules. The ratio of these two molecules was 0.5, 1, 2, 3, 4, and 5. Results showed that the detection signal was the best when the ratio of these two molecules was about 3:1 (Fig. 3). And based on quantification of HRP molecules immobilized on the gold nanoparticles probes,
Fig. 3. Optimization graph of molecule ratio of HRP and antibody when 10 nm AuNPs were immobilized with these two kinds of molecules. The molecule ratio is 0.5, 1, 2, 3, 4, and 5. The concentration of p53 protein detected was 10 ng mL−1 and the concentration of MMP probes was 1.5 mg mL−1 . And the graph shows that the detection signal is the best when the ratio of these two molecules is about 3:1.
the number of HRP molecules was detected on each 10 nm gold nanoparticle immobilized with HRP and antibody at this ratio. The concentration of streptavidin–HRP immobilized onto AuNP probes was 1.325 × 10−2 g L−1 based on the activity of HRP and the calibration curve of streptavidin–HRP (Figure S5). The molecular weight of streptavidin–HRP is 100 KDa (Dalton). So, the concentration was 1.325 × 10−7 M. The concentration of AuNP probes was 0.238 × 10−7 M based on the absorption of AuNP probes at 528 nm. Therefore, the concentration ratio of streptavidin–HRP and gold nanoparticles was 5.6, which meant that there were about 5–6 HRP molecules on each 10 nm gold nanoparticle. The sandwich complex will show reactive activity relative to 5–6 HRP molecules once bound one gold nanoparticle probe through immunoassay, which can amplify the detection signal. The previous reports showed that there were 10 anti-humanIgG–HRP molecules on each 13 nm gold nanoparticle (Ambrosi et al., 2007). And 7 HRP molecules could be adsorbed onto one 15 nm gold nanoparticle immobilized together with DNA probes (Li et al., 2008). So, there is a good correspondence between our and previous results. And there are about 1–2 anti-p53 detector antibody molecules together with 5.6 HRP molecules on each 10 nm gold nanoparticle. 3.6. The preliminary application of Nano-ELISA for detecting p53 protein Fig. 4 showed the caliberation curve of Nano-ELISA for p53 protein with different concentrations of 2.5 × 105 , 1 × 105 , 2.5 × 104 , 6.25 × 103 , 1.5 × 103 , 370, 90, 22, and 5 pg mL−1 . The assay buffer containing two background proteins was also detected as the background signals. Low to 5 pg mL−1 p53 protein was detected (S/N ≥ 2). And the serum sample (50 ng mL−1 p53 protein) of lung cancer was diluted to 500, 50, 5 pg mL−1 and detected. The results were shown in Fig. 5. The p53 protein in these serum dilutions was also detected. The absorbance (450 nm) of 5 pg mL−1 p53 protein from the recombined p53 protein with only two background proteins and the serum dilutions, were 0.12 and 0.18, separately. So, this Nano-ELISA method can be used to detect the serum samples. The detection sensitivity of p53 classic ELISA kit sold by Upstate is 0.125 ng mL−1 (data not shown) (www.Upstate.com). And for another commercial p53 ELISA kit sold by Roche, the detection limit of p53 protein is 1.6 ng mL−1 (www.Roche.com). The detection limit obtained using Nano-ELISA method is much lower than that using classic ELISA method. And the two ELISA kits should
C.-P. Jia et al. / Biosensors and Bioelectronics 24 (2009) 2836–2841
2841
nanoparticles here were used as carrier and amplifier. Through immunoassay of antibody and antigen, the sandwich structures (magnetic beads–target protein–AuNP probes) were formed. And then, the magnetic field was used to collect the sandwich structures. Excess components were removed through washing for 2–3 times. Protein was detected based on absorbance value. Low to 5 pg mL−1 of protein was detected with this method. The assay time was less than 2 h. This method can be used to detect protein markers of tumors, nervous system or other diseases for early diagnostics. Acknowledgements
Fig. 4. Calibration curves of human p53 protein at concentration of 2.5 × 105 , 1 × 105 , 2.5 × 104 , 6.25 × 103 , 1.5 × 103 , 370, 90, 22, 5, and 0 pg mL−1 diluted in assay buffer (10 mM PBS buffer containing 0.1% BSA, 500 ng mL−1 CEA and 500 ng mL−1 NSE as background proteins) with 10 nm AuNP probes. The detection limit is 5 pg mL−1 with 10 nm AuNP probes.
This work was supported by the National High Technology Research and Development Program (“863” Program) of China (No. 2006AA03Z334), Ministry of Science and Technology (No. 2007CB936000) and Shanghai Municipal Commission for Science and Technology (No. 0752nm019; 0652nm016; 0752nm021). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bios.2009.02.024. References
Fig. 5. Detection results of p53 protein in serum sample dilutions. The serum sample (50 ng mL−1 p53 protein) of lung cancer was diluted to 500, 50, 5 pg mL−1 in assay buffer (10 mM PBS buffer containing 0.1% BSA, 500 ng mL−1 CEA and 500 ng mL−1 NSE as background proteins) and detected. It can be seen that low to 5 pg mL−1 p53 protein in serum sample can be detected out.
conduct many amplification steps from polyclonal antibodies and biotin–streptavidin system, including three reaction steps and four wash steps, which is tedious and take more time. So, use of gold nanoparticles as the carrier and amplifier can increase the sensitivity of immunoassay significantly and Nano-ELISA is more effective and time saving. The basic operation and instrument of Nano-ELISA is similar with ELISA, which can help this method to be used more widely in clinic and lab than other methods such as, electrochemistry. 4. Conclusions Nano-ELISA, a highly sensitive protein detection method, was introduced here. Gold nanoparticles (AuNPs) were modified with a monoclonal detector antibody and HRP (signal molecules). Gold
Alivisators, P., 2004. Nat. Biotechnol. 22 (1), 47–52. Ambrosi, A., Castaneda, M.T., Killard, A.J., Smyth, M.R., Alegret, S., Merkoci, A., 2007. Anal. Chem. 79 (14), 5232–5240. Bao, Y.P., Wei, T.F., Lefebvre, P.A., Hao An, He, L.X., Kunkel, G.T., Muller, U.R., 2006. Anal. Chem. 78 (6), 2055–2059. Castaneda, M.T., Alegret, S., Merkoci, A., 2007. Electroanalysis 19, 743–753. Chen, Y., Yu, C., Cheng, T., Tseng, W., 2008. Langmuir 24 (7), 3654–3660. Cressey, R., Pimpa, S., Chewaskulyong, B., Lertprasertsuke, N., Saeteng, S., Tayapiwatana, C., Kasinrerk, W., 2008. BMC Biotechnol. 8, 16–29. Crowther, J.R., 2000. The Elisa Guidebook. The International Atomic Energy Agency, Vienna. Cui, R., Huang, H., Yin, Z., Gao, D., Zhu, J., 2008. Biosens. Bioelectron. 23, 1666–1673. Engvall, E., Perlman, P., 1971. Immunochemistry 8 (9), 871–874. Georganopoulou, D.G., Chang, L., Nam, J.M., Thaxton, C.S., Mufson, E.J., William, L., Klein, W.L., Mirkin, C.A., 2005. Proc. Natl. Acad. Sci. 102 (7), 2273–2276. Gosling, J.P. (Ed.), 2000. Immunoassays: Practical Approach. Great Clarendon, Oxford. Hermanson, G.T., 1996. Bioconjugate Techniques, first ed. Academic Press, San Diego. Jain, P.K., Lee, K.S., El-Sayed, I.H., 2006. J. Phys. Chem. B 110 (14), 7238–7248. Jie, G., Huang, H., Sun, X., Zhu, J., 2008. Biosens. Bioelectron. 23, 1896–1899. Kent, S.P., Ryan, K.H., Siegel, A.L., 1978. J. Histochem. Cytochem. 26 (8), 618–621. Koppelman, S.J., Lakemond, C.M., Vlooswijk, R., Hefle, S.L., 2004. J. AOAC Int. 87 (6), 1398–1407. Lequin, R., 2005. Clin. Chem. 51 (12), 2415–2418. Li, J., Song, S., Liu, X., Wang, L., Pan, D., Huang, Q., Zhao, Y., Fan, C., 2008. Adv. Mater. 20, 497–500. Matakidou, A., El Galta, R., Webb, E.L., Rudd, M.F., Bridle, H., Eisen, T., Houlston, R.S., Consortium, G., 2007. Lung Cancer 57 (2), 207–212. Nam, J.M., Thaxton, C.S., Mirkin, C.A., 2003. Science 301, 1884–1886. ´ Porebska, I., Wyrodek, E., Kosacka, M., Adamiak, J., Jankowska, R., HarłozinskaSzmyrka, A., 2006. In Vivo 20 (5), 599–604. Rosi, N.L., Mirkin, R.C., 2005. Chem. Rev. 105 (4), 1547–1550. Roth, J., 1982. J. Histochem. Cytochem. 30, 691–696. Saleh, O.A., Sohn, L.L., 2003. Proc. Natl. Acad. Sci. 100 (3), 820–824. Selvaraju, T., Das, J., Han, S., Yang, H., 2008. Biosens. Bioelectron. 23, 932–938. Thaxton, C.S., Hill, H.D., Georganopoulou, D.G., Stoeva, S.I., Mirkin, C.A., 2005. Anal. Chem. 77 (24), 8174–8178. Turkevich, J., Stevenson, P., Hiller, J., 1951. Discuss. Faraday Soc. 11, 55–75. Wang, J., Zhu, X., Tu, Q., Guo, Q., Zarui, C.S., Momand, J., Sun, X., Zhou, F., 2008. Anal. Chem. 80 (3), 769–774. Yan, G., Xing, D., Tan, S., Chen, Q., 2004. J. Immunol. Methods 288 (1–2), 47–54. Zhang, J., Song, S., Zhang, L., Wang, L., Wu, H., Pan, D., Fan, C., 2006. J. Am. Chem. Soc. 128, 8575–8580. Zhang, J., Song, S., Wang, L., Pan, D., Fan, C., 2007. Nat. Protocols 2, 2888–2895.