Biosensors and Bioelectronics 65 (2015) 139–144

Contents lists available at ScienceDirect

Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios

Chemiluminescence resonance energy transfer imaging on magnetic particles for single-nucleotide polymorphism detection based on ligation chain reaction Sai Bi a,n, Zhipeng Zhang b, Ying Dong b, Zonghua Wang a,n a Laboratory of Fiber Materials and Modern Textiles, the Growing Base for State Key Laboratory, Shandong Sino-Japanese Center for Collaborative Research of Carbon Nanomaterials, Collaborative Innovation Center for Marine Biomass Fiber Materials and Textiles, Qingdao University, Qingdao 266071, PR China b Key Laboratory of Sensor Analysis of Tumor Marker, Ministry of Education, College of Chemistry and Molecular Engineering, Qingdao University of Science and Technology, Qingdao 266042, PR China

art ic l e i nf o

a b s t r a c t

Article history: Received 27 June 2014 Received in revised form 24 September 2014 Accepted 9 October 2014 Available online 17 October 2014

A novel ligation chain reaction (LCR) methodology for single-nucleotide polymorphism (SNP) detection was developed based on luminol–H2O2–horseradish peroxidase (HRP)-mimicking DNAzyme–fluorescein chemiluminescence resonance energy transfer (CRET) imaging on magnetic particles. For LCR, four unique target-complement probes (X and Xn, YG and Yn) for the amplification of K-ras (G12C) were designed by modifying G-quadruplex sequence at 3′-end of YG and fluorescein at 5′-end of Yn. After the LCR, the resulting products of XYG/XnYn with biotin-labeled Xn were captured onto streptavidin-coated magnetic particles (SA-MPs) via specific biotin-SA interaction, which stimulated the CRET reaction from hemin/G-quadruplex-catalyzed luminol–H2O2 CL system to fluorescein. By collecting signals by a cooled low-light CCD, a CRET imaging method was proposed for visual detection and quantitative analysis of SNP. As low as 0.86 fM mutant DNA was detected by this assay, and positive mutation detection was achieved with a wild-type to mutant ratio of 10,000:1. This high sensitivity and specificity could be attributed to not only the exponential amplification and excellent discrimination of LCR but also the employment of SA-MPs. SA-MPs ensured the feasibility of the proposed strategy, which also simplified the operations through magnetic separation and separated the reaction and detection procedures to improve sensitivity. The proposed LCR-CRET imaging strategy extends the application of signal amplification techniques to SNP detection, providing a promising platform for effective and high-throughput genetic diagnosis. & Elsevier B.V. All rights reserved.

Keywords: Chemiluminescence resonance energy transfer Ligation chain reaction Magnetic particles Single-nucleotide polymorphism

1. Introduction Single nucleotide polymorphism (SNP) is the most common variations in human genomes, which associates closely with many human diseases, such as cancer and mitochondrial diseases (Kwok, 2003). For example, K-ras (G12C), a common oncogenic mutant, has been found to be the most expressed mutation in lung cancer (Ostrem et al., 2013). In addition, different types of K-ras mutations could affect tumor behavior and drug sensitivity (Garassino et al., 2011). Thus, SNP analysis is of great significance in determining the genetic predisposition for inherited diseases, early diagnosis and risk assessment of malignancy, and drug response. The

n

Corresponding authors. Fax: þ 86 0532 85950873. E-mail addresses: [email protected] (S. Bi), [email protected] (Z. Wang). http://dx.doi.org/10.1016/j.bios.2014.10.025 0956-5663/& Elsevier B.V. All rights reserved.

development of effective assays for SNP screening is in urgent demand. So far, many methods have been developed for SNP detection, especially amplification techniques. For example, enzyme conjugates (Patolsky et al., 2001; Ermini et al., 2014) and nanoparticles (Abbaspour and Noori, 2012; Wang et al., 2013) have been used as amplifying labels for SNP genotyping. In addition, as the standard method for amplification of nucleic acids, polymerase chain reaction (PCR) is used to detect trace amount of samples from genomic DNA, which however could introduce errors into SNP detection during the exponential amplification process (Mhlanga and Malmberg, 2001). Alternatively, ligase chain reaction (LCR) has become a powerful and robust technique for SNP detection due to its simplicity and rapidity, high amplification efficiency, and good specificity (Shen et al., 2012; Zhang et al., 2013). A typical LCR system is consisted of two pairs of oligonucleotide probes. Upon the introduction of target DNA, the complementary adjacent probes can be ligated by thermophilic DNA ligase, such as

140

S. Bi et al. / Biosensors and Bioelectronics 65 (2015) 139–144

ampligase. The ligated products in turn serve as templates to ligate their correspondingly adjacent probes, resulting in an exponential amplification process by thermal cycles of annealing, ligation and melting. Due to the excellent discrimination capacity of ligation, LCR holds better specificity than primer extension reaction for SNP (Shen et al., 2013; Chen et al., 2012; Zhou et al., 2013). In general, the LCR products were detected by gel electrophoresis (Barany, 1991; Yan et al., 2010) or fluorescence techniques (Cheng et al., 2012). Recently, chemiluminescence (CL) has become an attractive tool for biochemical research (Créton and Jaffe, 2001). Due to the emission of light caused by a chemical reaction without an external excitation source, CL has virtually nonspecific signals and is extremely sensitive with simple instrumental set-up. Chemiluminescence resonance energy transfer (CRET) occurs via nonradiative dipole–dipole energy transfer between a CL donor and a suitable acceptor molecule that are in close proximity without the need of a light source (Lei et al., 2014; Chen and Li, 2014; Mun et al., 2014). The CRET signal can be easily readout by photometers or luminometers, especially by sensitive imaging photon detector of cooled low-light charge-coupled device (CCD) for chemiluminescence imaging assay (Bi et al., 2013; Zong et al., 2012). Herein, a LCR-CRET imaging strategy is proposed for SNP detection with high sensitivity and specificity. This method is based on the specific ligation of thermostable ligase to discriminate single-base variation and the exponential amplification of LCR to improve sensitivity. In addition, the probes that are designed with G-quadruplex sequence and fluorescein enable the ligated LCR products to perform the horseradish peroxidase (HRP)-mimicking DNAzyme-catalyzed luminol–H2O2 CL reaction in the presence of hemin, leading to the CRET to fluorescein. Moreover, streptavidincoated magnetic particles (SA-MPs) are employed to capture the biotin-tagged LCR products, which can significantly improve the sensitivity by removing the unintercalated hemin that could induce high background. This platform provides a rapid, specific, sensitive, and high-throughput technique for SNP detection.

2. Experimental section 2.1. Materials Oligonucleotides used in this study were synthesized by Shanghai Sangon Biotech Co., Ltd. (China) (sequences see Table S1). Ampligases enzyme and buffer were purchased from Epicenter Biotechnologies (Madison, WI). Streptavidin-coated magnetic particles (SA-MPs) (1 mm) were ordered from Zhengzhou Innosep Biosciences Co., Ltd. (China). Hemin, luminol, hydrogen peroxide, and 4-(2-hydroxyethyl)piperazine-1 ethanesulfonic acid sodium salt (HEPES) were obtained from Aladdin Chemistry Co. Ltd. (China). Magnetic separation rack for 96-well plate was purchased from Tiajin BaseLine Chromtech Research Centre (China). The chemicals were of analytical grade and used as received without further purification. Double-distilled, deionized ultrapure water was used in all of the experiments. 2.2. LCR process A 15 mL of reaction solution contains 2.5 mL of TE buffer (10 mM Tris–HCl, 1 mM EDTA, 12.5 mM MgCl2, pH 8.0), 1.5 mL of 10  ampligase reaction buffer, probes X, Xn, YG, and Yn with the final concentration of each component of 0.1 mM, and the various amounts of target DNA. After heating the mixture to 95 °C for 3 min, 1 mL of ampligase (5U) was added at 75 °C. 30 Thermal cycles for LCR were carried out on a Little Genius thermal cycler (Hangzhou Bioer, China) with each cycle consisting of a 1 min at

95 °C for denaturation and 1 min at 50 °C for annealing/ligation. After the thermal cycling, the mixture was further reaction at 37 °C for 30 min. 2.3. CRET imaging A 20 mL of SA-MPs was added to capture the biotin-labeled ligated products at room temperature for 30 min. After magnetic separation, 5.0  10  7 M hemin in HEPES was added and incubated at 25 °C for 20 min to form the HRP-mimicking DNAzymes. Subsequently, the SA-MPs were washed with HEPES buffer twice, followed by adding 0.02 M luminol and 0.2 M H2O2 to initiate the CRET reaction. The CRET images were recorded using an EC3 imaging system with a thermoelectrically cooled CCD camera (UVP, USA). The CRET intensity of each spot was calculated as the mean pixel within a circle of a given diameter around spot center. In specific experiments, the mixture of mutDNA and wtDNA at different ratios with a total concentration of 1.0  10  11 M was used as a target sample to carry out the LCR-CRET reaction as mentioned above.

3. Results and dicussion 3.1. Assay principle The principle of the proposed ligase chain reaction (LCR)-based chemiluminescence resonance energy transfer (CRET) imaging of single-nucleotide polymorphism (SNP) by using magnetic particles (MPs) to reduce background is illustrated in Scheme 1. The system is consisted of four short single-stranded DNA probes (X and Xn, YG and Yn; G represents the G-quadruplex nucleic acid structure), target DNA (mutDNA or wt DNA), and ampligase. The targets DNA are fragments of the common human oncogene K-ras, in which the wtDNA and mutDNA are the wide-type and mutant genes, respectively. The wtDNA is different from mutDNA with a single base of C-A transversion that causes G12C mutation in the K-ras gene. X and Y are designed the same as one half of the mutDNA target, while Xn and Yn are complementary to one half of the target. In the presence of mutDNA, probes Xn and Yn are hybridized to adjacent positions on the target template at 50 °C, which are subsequently covalently joined by thermostable ampligase to form the ligated product of XnYn, whereas the duplexes are simultaneously formed between X and Xn, Y and Yn. Upon heated to 95 °C, all duplexes are denatured, releasing target DNA, the ligated product XnYn, and partial probes. In this way, when the temperature is reduced to 50 °C, the DNA target is recycled to perform the target-recycled ligation, while the ligated XnYn serves as the new template for the probes X and YG to form the new ligation product of XYG that also serves as the secondary target for the subsequent annealing/ligation with probes Xn and Yn to form the ligation product of XnYn. It should be noted that excess probes over target DNA play an important role in minimizing the re-hybridization of the target to the ligated products. Thus, from the second thermal cycle, the amount of the ligated products will be doubled theoretically after each cycle, resulting in an exponential amplification of mutDNA to generate a large number of ligation probes (XnYn and XYG) through repeating the thermal cycling of annealing/ligation at 50 °C and denaturation at 95 °C. This process is named as ligation chain reaction (LCR). After LCR, the mixture is further reacted at 37 °C for 30 min to make the resulting duplex products stable with four types of mutDNA/XnYn, XYG/XnYn, X/Xn, and YG/Yn. Here probe Yn is designed with fluorescein label at its 5′-end, and probe Xn is modified with biotin at its 3′-end. Thus, the duplex products with Xn can be easily captured on streptavidin-coated magnetic particles

S. Bi et al. / Biosensors and Bioelectronics 65 (2015) 139–144

141

Scheme 1. (A) Schematic illustration of the LCR-based amplification strategy for SNP detection of G12C mutation in K-ras gene by the luminol–H2O2–HRP-mimicking DNAzyme–fluorescein CRET imaging. The resulting products of LCR are captured by SA-MPs via specific biotin-SA interaction. The superfluous hemin can be easily removed by magnetic separation, which efficiently reduces background induced by hemin itself. (B) Schematic illustration of the proposed strategy for wtDNA detection, in which the LCR cannot be performed and thus the CRET reaction cannot occur. Domains Xn and Yn are the complement of domains X and Y.

(SA-MPs) via specific biotin-SA recognition, and the unbound free species of YG/Yn can be easily removed after magnetic separation. Among them, the major ligated products of XYG/XnYn immobilized on microspheres are acted as the binary signaling probes based on luminol–H2O2–HRP-mimicking DNAzyme–fluorescein CRET system. In the proposed CRET reaction, luminol acts as both CL substrate and donor; HRP can continuously catalyze the luminol–H2O2 CL reaction; fluorescein can not only enhance the CL reaction but only act as the acceptor (Bi et al., 2012; Al-Ogaidi et al., 2014; Luo et al., 2012; Liu et al., 2013). As for the YG/Yn on SA-MPs, the G-quadruplex nucleic acid structure intercalates hemin to form HRP-mimicking DNAzymes that act as the peroxidase catalysts of the CRET system. Since the catalysts, the formed DNAzyme uints, are in close proximity to fluorescein, thus, an enhanced CL is observed because CRET occurs from DNAzyme–catalyzed luminol–H2O2 to fluorescein. The CRET signals are collected by a cooled low-light CCD to perform the CRET imaging, achieving visual detection and quantitative analysis of SNP. As shown in Scheme 1B, if the probes with a mutated locus anneal with a widetype target, no ligated template can be formed, and therefore no CRET imaging signal can be detected. The total time per assay in the present work is 2.5 h, consisting of LCR for 90 min, SA-MPs capture for 30 min, and hemin intercalation for 20 min. The CRET reaction occurs immediately after the addition of luminol and

H2O2 into the samples. Moreover, to achieve effective and high throughput screening, the samples are transferred to 96-well plate after LCR, followed by performing the subsequent procedures by using multichannel pipette and magnetic separation rack for 96well plate. It is noteworthy to point out that SA-MPs used in this assay have many advantages. On one hand, the feasibility in magnetic microsphere handling facilitates thorough washing. The unbound duplex YG/Yn and excess free hemin can be easily removed through magnetic separation, which can significantly simplify the experimental operation. On the other hand, the reaction systems, such as LCR buffer and CRET working buffer, can be separated by using SA-MPs to capture biotin-tagged ligation products. Thus, the experiments can be carried out under respective optimum conditions. More importantly, as shown in Scheme S1, if the experiments are carried out homogeneously without using SA-MPs, the CRET signals generated by ligated (XYG/XnYn) and unligated (YG/ Yn) products could not be distinguished due to the proximity of DNAzyme unit to fluorescein in either case. From the results of homogeneous reaction (Fig. S1), the CRET signals of the blank (in the absence of target) and the samples (in the presence of mutant DNA at different concentrations) before and after LCR are almost the same, which thus cannot be applied to quantitative detection of SNP. However, after introduction of SA-MPs into the system

142

S. Bi et al. / Biosensors and Bioelectronics 65 (2015) 139–144

4

CRET intensity (× 10 )

10

8

6

4

2

0 blank Fig. 1. Native PAGE (15% native gel) of the reaction pathways. Lane M: 500-base pair (bp) DNA ladder; lane 1: Xþ X*; lane 2: YG þ Y*; lane 3: mutDNA þX* þ Y*; lane 4: mutDNA þX* þ Y* þXþ YG; lane 5: lane 4þ ampligase (5U) to perform LCR after 10 thermal cycles. The concentrations of X, X*, YG, and Y* are 1.0 mM, and mutDNA is 1.0  10  8 M.

random DNA

wtDNA

mutDNA

Fig. 2. CRET images of the LCR products by using no mutDNA target (blank), random DNA, wtDNA, and mutDNA as target, respectively. The probe pairs are matched with mutDNA. The concentration of each target sample is 1.0  10  11 M. The error bars represent standard deviation of triplicates.

3.3. Optimization of conditions

The reaction mechanism of the proposed LCR strategy for SNP detection is directly verified by native polyacrylamide gel electrophoresis (PAGE) (Fig. 1). The duplexes of X/Xn, YG/Yn, and mutDNA/XnYn with different mobilities are shown in lanes 1–3, respectively. When the components of the reaction system (mutDNA, X, YG, Xn, and Yn) coexist without ampligase, three types of duplexes are formed, X/Xn, YG/Yn, and mutDNA/XnYn (lane 4). Whereas ampligase is added to initiate the LCR, a less mobile species is observed in lane 5, indicating the formation of the ligated duplex of XYG/XnYn. Although the XYG/XnYn has the same base pairs as the T/XnYn, an overhang of G-quadruplex nucleic acid in XYG/XnYn decreases its mobility. The PAGE results demonstrate the feasibility of the proposed LCR for SNP detection that is carried out as designed. The proposed reaction is further confirmed by CRET imaging that are obtained by the LCR products of mutDNA, wtDNA, random DNA and no mutDNA target (blank) after immobilizing on SA-MPs, treating with hemin, and separating through magnetic field. From Fig. 2, the CRET signal is only observed for mutDNA, while no signal is observed for random DNA and blank sample, even for wtDNA with one-base mismatched. The CRET intensity for 1.0  10  11 M mutDNA is five times higher than that for wtDNA at the same concentration. Evidently, the LCR is only achieved by using probe pairs matched with mutDNA which can further perform the CRET reaction on the surface of MPs, indicating the good selectivity of this method for SNP detection.

3.3.1. The thermal cycle number of LCR The LCR amplification efficiency is directly related to the thermal cycle number of LCR. As shown in Fig. S2A that compiles the CRET images and intensities of the LCR amplification during thermal cycling, the CRET intensity increases progressively as increasing the thermal cycle number, but the rate of increase gradually levels off from 30 cycles. This can be attributed to the depletion of partial probes. The highest CRET signal is occurred at 30 thermal cycles, which is therefore used in the following experiments. 3.3.2. The concentration of partial probes Fig. S2B shows the CRET images and intensities corresponding to different concentrations of partial probes (1.0  10  8–1.0  10  6 M) to perform the LCR. As increasing the concentration of probes from 1.0  10  8 to 1.0  10  7 M, the CRET intensities 8 4

3.2. Feasibility of the proposed strategy

The reaction conditions of thermal cycle number of LCR, partial probes concentration, SA-MP amounts, and hemin concentration are investigated and optimized.

relative CRET intensity (× 10 )

(Scheme 1), the duplex products (X/Xn and ligated products of XYG/XnYn) with biotin-modified Xn can be easily captured on SAMPs via specific biotin-SA recognition and the unligated species of YG/Yn can be easily removed via magnetic separation, which can effectively eliminate the negative signals produced by YG/Yn and achieve sensitive and specific detection of mutant DNA with a good reproducibility and validity. In addition, the specific recognition of SA and biotin offers high binding capacity, further enhancing the assay sensitivity.

7 0

6

10

-15

10

-14

10

-13

10

-12

10

-11

(M)

I = 1.60lgC + 24.42 (R = 0.9915)

5 4 3 2 1 0

10

-15

10

-14

10

-13

10

-12

10

-11

concentraion of mutDNA (M) Fig. 3. CRET images and corresponding calibration curve of the proposed LCR-CRET amplification strategy for mutDNA detection. The error bars represent standard deviation of triplicates. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)

S. Bi et al. / Biosensors and Bioelectronics 65 (2015) 139–144

3.3.3. The amounts of SA-MPs The effect of the SA-MP amounts on the capture of biotintagged ligated products and signal response is studied. As shown in Fig. S2C, for the detection of 2.0  10  12 M target mutDNA, the CRET intensities increase remarkably with increasing the amounts of SA-MPs, which saturates at 20 mL (black curve). In contrast, as for the blank sample that is no target mutDNA, quite small increase of CRET signal is observed with the amounts of SA-MPs from 10 to 50 mL (red curve). Thus, SA-MPs have little influence on the CRET reaction system. To obtain the maximum relative CRET signal between mutDNA target and blank, 20 mL of SA-MPs is selected in subsequent experiments. 3.3.4. The concentration of hemin The CRET response is directly related to the generated HRPmimicking DNAzymes on SA-MPs to perform the CRET to fluorescein. Fig. S2D shows the CRET signal is increased remarkably with an increase of hemin concentration from 5.0  10  9 to 5.0  10  7 M for the detection of 1.0  10  11 M mutDNA. Thus, the concentration of 5.0  10  7 M hemin is selected to be the optimum in this assay. 3.4. Sensitivity Under the optimized experimental conditions, the mutDNA at various concentrations are analyzed by the proposed LCR-CRET imaging strategy. From Fig. 3, a quite weak CRET signal is observed in the absence of mutDNA target. When the concentration of target is at 1.0  10  15 M, the sensing site can be observed a bright spot. The brightness intensifies with the increase of target concentration. A linear relationship between the relative CRET intensity and the logarithm of target concentration is achieved over a widely dynamic range from 1.0  10  15 to 1.0  10  11 M e xtending five orders of magnitude with a regression equation of ΔI ¼16452.3 lg C þ 249803.9 (a correlation coefficient of R¼ 0.98927), in which the relative CRET intensity is obtained after background correction (blank control background signal is subtracted from the sample signals). The detection limit is calculated to be 0.86 fM (3s), which is lower than or comparable to those reported by other ligation-based SNP assays (Table S2). Notably, the sensitivity of this protocol is far below the requirement for direct gene expression profiling, which is attributed to (i) the substantial increase of the amount of the amplied target/ secondary target during LCR, (ii) the high sensitivity of the luminol–H2O2–HRP-mimicking DNAzyme–FAM CRET imaging system, (iii) the collection of CRET signals by a sensitive cooled low-light CCD, and (iv) particularly the employment of SA-MPs. In addition, the relative standard deviation (RSD) is estimated to be 4.9% from five repetitive measurements of 10 fM mutDNA, indicating the good precision of this assay. The RSD of the detection for different times is 7.7%, which is acceptable for instrumental analysis. 3.5. Specificity To determine the specificity of this assay, the wtDNA and mutDNA are mixed at different ratios of 10,000: 1, 5000:1, 2000:1, 1000:1, 100:1, 10:1, and 0:1 with a total concentration of 1.0  10  11 M, subsequently serving as the target sample to react with partial probes that are complementary to the mutDNA to carry out the proposed LCR-CRET reaction. The blank sample

10 9

4

CRET intensity (× 10 )

produced by mutDNA increases obviously, which level off when probe concentrations are higher than 1.0  10  7 M. Thus, partial probes at the concentration of 1.0  10  7 M are chosen as the optimal concentration for the subsequent experiments.

143

8 7 6 5 4 3 2 1 0 nk 0:1 00:1 00:1 00:1 00:1 bla 000 1 10 20 50 1

10

:1

0:1

wtDNA: mutDNA (K-ra G12C) Fig. 4. Effect of the ratio of wtDNA to mutDNA on the detection of mutDNA. The error bars represent standard deviation of triplicates.

contains 1.0  10  11 M wtDNA only. The CRET images and integrated CRET counts for the imaged regions of bead samples prepared after LCR containing various dilutions of mutDNA into wtDNA are shown in Fig. 4. The CRET signal for the mutation decreases as the mutDNA content is reduced. The positive control signals could be discriminated from the blank when the concentration of mutDNA is as low as 1.0 fM (a mutant to wild-type ratio up to 1:10,000), indicating the high selectivity of our LCRCRET assay in analysis of a rare point mutation even in the presence of a large excess of wide types. It should be noted that the CERT images are recorded using an EC3 imaging system with a thermoelectrically cooled CCD camera (UVP, USA) that has excellent precision and accuracy in sample detection. Thus, the CRET signals of the wide-type and mutant DNA at ratios of 10,000:1, 5000:1 and 2000:1 with integrated CRET counts of 11,179, 12,804 and 12,758 can be considered as the detectable signals in comparison with that of blank (9627). To make it more clear, the enlarged image of wide-type to mutant DNA at ratios of 1:0, 10,000:1, 5000:1, 2000:1 and 1000:1 is shown in inset of Fig. 4. The excellent specificity of our LCR-based amplification assay can be greatly owed to the fidelity of the thermalstable ampligase to perform the thermal cycles, which further results in the fabrication of ligated probes on MPs to produce a CRET signal. 3.6. Real sample assay To demonstrate the practicability of the proposed methodology in the present work, the detection of the real sample using the method has been carried out by spiking different concentrations of mutDNA (10  15, 10  13, and 10  11 M) into human serum that was diluted 10 times and investigating the recovery to confirm the performances of the assay in complex matrixes. From Table 1, the recovery was calculated to be 101.276.8%, indicating the Table 1 Detection of mutDNA spiked in human serum by the proposed LCR-CERT strategy. Sample Blood serum (M) –a – –

1 2 3 a

No CRET response.

Added mutDNA (M)

Detected mutDNA (M)

Recovery (%)

10  15 10  13 10  11

1.07  10  15 9.82  10  14 9.55  10  12

107.0 98.2 95.5

144

S. Bi et al. / Biosensors and Bioelectronics 65 (2015) 139–144

potentiality of the proposed method for SNP detection in real sample assay.

4. Conclusion In conclusion, through the incorporation of LCR with CRET imaging, a specific ligation, exponential amplification and highthroughput strategy was proposed for SNP analysis. The developed methodology offered not only excellent target amplification but also easy operation by employing SA-MPs. Under the optimum conditions, we showed our approach can detect as low as 0.86 fM mutDNA target along with wide dynamic range from 1.0 fM to 10 pM. In addition, it can quantitate the rare mutDNA even in a large excess of coexisting wtDNA with a selectivity factor down to 1:10,000 (mutDNA:wtDNA). Therefore, this approach appears to be a powerful signal amplification strategy for sensitive and specific detection of SNP, which can be applied to early diagnosis and prognosis of SNP-related diseases clinically.

Acknowledgments This work was supported by the National Natural Science Foundation of China (21375056, 21105052, and 21275082), the Program for New Century Excellent Talents in University of Ministry of Education of China (NCET-12-1024), the China Postdoctoral Science Foundation (2012M510130 and 2013T60515).

Appendix A. Suplementary information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2014.10.025.

References Abbaspour, A., Noori, A., 2012. Biosens. Bioelectron. 37, 11–18. Al-Ogaidi, I., Gou, H., Aguilar, Z.P., Guo, S., Melconian, A.K., Al-kazaz, A.K.A., Meng, F., Wu, N., 2014. Chem. Commun. 50, 1344–1346. Barany, F., 1991. Proc. Natl. Acad. Sci. U. S. A. 88, 189–193. Bi, S., Ji, B., Zhang, Z., Zhang, S., 2013. Chem. Commun. 49, 3452–3454. Bi, S., Zhao, T., Luo, B., 2012. Chem. Commun. 48, 106–108. Chen, C., Li, B., 2014. Biosens. Bioelectron. 54, 48–54. Chen, X., Ying, A., Gao, Z., 2012. Biosens. Bioelectron. 36, 89–94. Cheng, Y., Du, Q., Wang, L., Jia, H., Li, Z., 2012. Anal. Chem. 84, 3739–3744. Créton, R., Jaffe, L.F., 2001. BioTech. 31, 1098–1105. Ermini, M.L., Mariani, S., Scarano, S., Minunni, M., 2014. Biosens. Bioelectron. 61, 28–37. Garassino, M.C., Marabese, M., Rusconi, P., Rulli, E., Martelli, O., Farina, G., Scanni, A., Broggini, M., 2011. Ann. Oncol. 22, 235–237. Kwok, P.-Y., 2003. Single Nucleotide Polymorphisms: Methods and Protocols. Humana Press Inc. Lei, J., Jing, T., Zhou, T., Zhou, Y., Wu, W., Mei, S., Zhou, Y., 2014. Biosens. Bioelectron. 54, 72–77. Liu, X., Niazov-Elkan, A., Wang, F., Willner, I., 2013. Nano Lett. 13, 219–225. Luo, M., Chen, X., Zhou, G., Xiang, X., Chen, L., Ji, X., He, Z., 2012. Chem. Commun. 48, 1126–1128. Mhlanga, M.M., Malmberg, L., 2001. Methods 25, 463–471. Mun, H., Jo, E.-J., Li, T., Joung, H.-A., Hong, D.-G., Shim, W.-B., Jung, C., Kim, M.-G., 2014. Biosens. Bioelectron. 58, 308–313. Ostrem, J.M., Peters, U., Sos, M.L., Wells, J.A., Shokat, K.M., 2013. Nature 503, 548–551. Patolsky, F., Lichtenstein, A., Willner, I., 2001. Nat. Biotechnol. 19, 253–257. Shen, W., Deng, H., Gao, Z., 2012. J. Am. Chem. Soc. 134, 14678–14681. Shen, W., Lim, C.L., Gao, Z., 2013. Chem. Commun. 49, 8114–8116. Wang, X., Zou, M., Huang, H., Ren, Y., Li, L., Yang, X., Li, N., 2013. Biosens. Bioelectron. 41, 569–575. Yan, J., Li, Z., Liu, C., Cheng, Y., 2010. Chem. Commun. 46, 2432–2434. Zhang, Y., Guo, Y., Quirke, P., Zhou, D., 2013. Nanoscale 5, 5027–5035. Zhou, L., Du, F., Zhao, Y., Yameen, A., Chen, H., Tang, Z., 2013. Biosens. Bioelectron. 45, 141–147. Zong, C., Wu, J., Wang, C., Ju, H., Yan, F., 2012. Anal. Chem. 84, 2410–2415.