Silver nanocluster-lightened hybridization chain reaction

Abstract

Herein we report for the first time a hybridization chain reaction (HCR) lightened by DNA-stabilized silver nanoclusters (AgNCs) as a label-free and turn on fluorescence platform for nucleic acid assays in a homogeneous format.

---

Introduction

Fluorescent metallic nanoclusters, especially nucleic-acid-stabilized silver nanoclusters (DNA–AgNCs), have attracted great research interest in recent years due to their facile synthesis, tunable fluorescence emission, and high photostability.^1,2 Due to these and their other features, including water solubility, subnanometer size and biocompatibility, AgNCs have been utilized as attractive materials for biosensing, bioimaging and so on.^3–5 In 2010, Werner et al. reported that dark DNA–AgNCs could be converted into emitters of bright red light with a 500-fold enhancement of fluorescence when placed in proximity to a guanine-rich sequence,⁶ and this discovery further enriched the applications of DNA–AgNCs in sensitive “turn on” assays.^7,8

The hybridization chain reaction (HCR)⁹ is an enzyme-free nucleic acid amplification technique, in which the amplification relies on successive toehold-mediated isothermal DNA strand displacement and hybridization events between two DNA hairpins to assemble nicked polymeric double helices upon the introduction of a triggering strand. It is a well-programmed and kinetically controlled assembling process of DNAs according to the Watson–Crick base-pairing principle, transforming a target-recognizing event into hundreds of simple repeated DNA sequences.^9,10 Being completely enzyme-free is the most advantageous feature of the HCR, and makes it possible to avoid some problems that enzyme-assisted nucleic acid amplification techniques may suffer from, such as easy contamination, high cost, susceptibility to false amplification, and tendency to have sequence mismatches.^9–12 Like most nucleic acid assay techniques, the HCR also needs signal transformation strategies to convert the nucleic acid products into other types of signals to make the assay more convenient and, at the same time, to improve its sensitivity.^13–15 Fluorescence, electrochemical and colorimetric detection are being extensively explored in HCR-based assays with the use of multiple labels, such as fluorophores,^15,16 small molecules¹⁷ or moieties,¹⁸ nanoparticles¹⁹ and so on. Nevertheless, label-dependent methods require multi-step procedures or delicately balanced affinities of interacting biomolecules in competitive assays. To avoid these disadvantages, Dong and co-workers²⁰ incorporated a sequence of G-quadruplex within one of the HCR hairpins for colorimetric and fluorescent assays. Their label-free HCR strategy showed great potential in single-nucleotide polymorphism (SNP) genotyping.²⁰

Here we report the development of a novel AgNC-lightened HCR as a label-free fluorescence platform for nucleic acid assays in a homogeneous format. As illustrated in Scheme 1, this method relies on an analyte-mediated formation of HCR products to capture dark DNA-stabilized AgNCs and then lighten the AgNCs, like a bright nucleic acid string of lights. When placed in proximity to guanine-rich DNA sequences, otherwise poorly fluorescent DNA-stabilized AgNCs have been shown to become transformed into bright red light-emitting clusters,⁶ and are hence denoted here as “AgNC-lightened”. To exploit such a unique feature of DNA-stabilized AgNCs for label-free fluorescence HCR assays, we designed an HCR hairpin probe (H1) to have a tail full of guanine bases at its 3′ end which could cause the DNA–AgNCs to emit light upon approach of the nanoclusters. Rather than applying the regular HCR, which uses only two hairpin probes to form repeats of the H1–H2 nucleic acid sequence,⁹ we used four hairpin probes for the HCR in order to avoid including the guanine-rich tail and the strand that captures the DNA–AgNC in the same hairpin probe and to induce non-specific fluorescence activation. With the use of four hairpin probes, i.e., H1–H4, the strand capturing the DNA–AgNC was designed as a tail of the hairpin probe H3 at its 5′ end. Therefore, the analyte-mediated HCR generated a long chain of H1–H4 repeats with the guanine-rich 3′ end of H1 being quite close to the 5′ end of H3. Consequently, loading dark DNA–AgNCs on the 5′ end of H3 of the HCR products enabled proximity-dependent fluorescence enhancement, which lit up the nucleic acid string because of the bright red light-emitting DNA–AgNCs. In contrast, without an analyte-mediated HCR, dark DNA–AgNCs captured by free probe H3 merely delivered a weak fluorescence signal because of the large distance between the free probes H1 and H3. Since the enhanced fluorescence signal could only be activated in response to the analyte-mediated formation of chain-like HCR products, the strategy we used afforded high sensitivity and selectivity for the enzyme-free fluorescence detection of nucleic acids. Therefore, in combination with the label-free, homogeneous format, which enables robust, cost-effective and easily automated assays of nucleic acid samples, the proposed AgNC-lightened HCR may create a useful platform for sensitive and specific nucleic acid assays.

Scheme 1 Illustration of using the transformation of dark AgNCs to light emitters in an HCR to detect a nucleic acid.

Experimental section

Reagents and materials

The hairpin probes (H1, H2, H3, and H4) and mRNAs (mRNA, mRNA-1, mRNA-2, mRNA-3, mRNA-4, and nonhomologous RNA) were synthesized and purified by Sangon Biotech Company, Ltd. (Shanghai, China). All the nucleic acids were in ultra-PAGE grade and their sequences are list in Table S1.† All other chemicals were of analytical grade and purchased from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China). Ultrapure water was obtained by using a Millipore Milli-Q water purification system (Billerica, MA, USA) with an electric resistance >18.2 MΩ.

Preparation of DNA-stabilized AgNCs

DNA–AgNCs were synthesized according to the protocol documented by Ritchie et al.²¹ Briefly, DNA–AgNCs were prepared by adding AgNO₃ (99.9%, Sigma-Aldrich) into a DNA solution, followed by reduction using freshly prepared NaBH₄ under vigorous shaking of five seconds in 20 mM phosphate buffer (PB, pH 6.6). The final concentrations were 90 μM, 15 μM and 90 μM for AgNO₃, template DNA and NaBH₄, respectively. Then, the mixture was kept in the dark at room temperature for 18 h to allow for the reaction to be completed before use.

AgNC-lightened HCR for mRNA detection

In typical assays, a sample of mRNA of a given concentrate was mixed with hairpin probes H1–H4 (400 nM each) in 50 mM PB (pH 6.8) supplemented with 500 mM NaNO₃ and 50 mM Mg(NO₃)₂. After incubating this mixture at 37 °C for 4 h, the prepared DNA–AgNCs (800 nM in final) were added into the mixture and allowed to react for 2 h at room temperature.

The fluorescence spectra were recorded at room temperature in a quartz cuvette on a FluoroMax-4 spectrofluorometer (HORIBA, NJ, USA). The excitation wavelength was 580 nm and the emission wavelengths were between 605 and 700 nm with both excitation and emission slits of 5 nm under a PMT voltage of 950 V.

Gel electrophoresis analysis of HCR

Agarose gels (3%) containing the coloring agents Goldview (0.5 μg mL⁻¹) and EB (0.5 μg mL⁻¹) in 1× TBE were loaded with 10 μL volumes of the samples and then run for 90 min at room temperature and 101 V. The gel was visualized using a Tocan 240 gel imaging system (Tocan Biotech. Co., Shanghai).

Atomic force microscopy (AFM) and high-response transmission electron microscopy (HRTEM)

A drop of 2 μL of the reaction product of 1.6 μM of DNA–AgNCs was deposited onto a freshly cleaved mica surface, as was one of the sample prepared without DNA–AgNCs. These samples were then gently rinsed with 200 μL of deionized H₂O and dried in air. AFM images were recorded on a Nanoscope IIIa (Veeco) in air under tapping mode. A drop of 10 μL of reaction product of 40 nM of DNA–AgNCs was deposited onto a carbon-coated grid, and dried in air. HRTEM images were recorded on a Hitachi H-7000 electron microscope (Tokyo, Japan).

Cell culture and RNA extraction

The three human cancer cell lines used were MCF-7 (human breast cancer), SKBr-3 (human breast cancer) and MCF-10A (human mammary epithelial). MCF-10A cells were cultured in DMEM containing 100 ng mL⁻¹ cholera toxin supplemented with 10% fetal bovine serum (FBS, GIBCO), 100 U mL⁻¹ streptomycin and 100 U mL⁻¹ penicillin. MCF-7 cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum, 100 U mL⁻¹ streptomycin and 100 U mL⁻¹ penicillin. SKBr-3 cells were cultured in McCoy's 5A medium (containing tricine) supplemented with 15% heat-inactivated fetal bovine serum. These cells were maintained at 37 °C and 5% CO₂ in a humidified atmosphere. Total RNA samples were isolated from each cell line by using the Trizol RNA purification kit (Beyotime Co., Ltd, Shanghai, China). The extract was used immediately for the mRNA assay or stored at −80 °C.

Results and discussion

We first used gel electrophoresis to investigate the feasibility of the four-hairpin based HCR. As shown in Fig. 1A, many bright gel electrophoresis bands corresponding to a maximum size of over 1500 base pairs were obtained from incubating the target mRNA with the four hairpin probes H1–H4. These bands provided evidence for the formation of chain-like duplex assemblies of H1–H4, indicating the potential of the HCR reactions for signal amplification in target detection. In the absence of target mRNA or any hairpin probes, no bands with large molecular weight were obtained. These results verified the formation of long nucleic acid chains from the four-hairpin probes and target mRNA. In addition, the DNA–AgNCs were inspected using HRTEM (Fig. S1 in ESI†). These images showed that the DNA-stabilized AgNCs (with diameters of ∼2.6 nm) were dispersed with a separation distance about 3 nm. After loading the DNA–AgNCs on the target mRNA-mediated HCR products, AFM revealed that the heights of the complex increased to ∼1.5 nm from ∼0.5 nm for the individual double-stranded HCR products (Fig. S2 in ESI†), suggesting the typical height of DNA–AgNCs exhibited by AFM to be ∼1 nm.²²

Fig. 1 (A) Agarose gel electrophoresis image of HCR products. Lane 1: H1; lane 2: H2; lane 3: H3; lane 4: H4; lane 5: H1–H4; lane 6: mRNA target + H1–H4; lane 7: mRNA target + H1; lane 8: mRNA target + H1 + H3; lane 9: mRNA target + H1–H3; lane M is the DNA size marker. (B) Fluorescence spectral responses obtained by H1–H4 (black), nonhomologous RNA with H1–H4 (pink), H2–H4 with mRNA target (blue), H1–H4 with mRNA target (red). The concentrations for different components are as follows: mRNA target, 100 nM; H1–H4, 400 nM each; DNA–AgNCs, 800 nM.

Based on the demonstrated four-hairpin HCR, a strategy involving a label-free fluorescent nucleic acid assay was developed by utilizing the unique optical properties of DNA–AgNCs. Since the fluorescence of dark DNA–AgNCs can only be enhanced when placing the nanoclusters in proximity to a guanine-rich sequence, in the reaction system where target mRNA was absent we observed only very weak signals with an average peak intensity of 3500 at 630 cm⁻¹ (Fig. 1B), about that of the background fluorescence of DNA–AgNCs. Otherwise, an intense fluorescence signal was obtained in the solution where target mRNA was incubated with probes H1–H4 followed by the addition of DNA–AgNCs. The peak intensity was increased up to 33 000, giving obvious evidence for the greatly enhanced fluorescence of DNA–AgNCs. To verify the specificity of the developed strategy, a control experiment was also conducted, in which 100 nM nonhomogeneous mRNA was incubated with the four-hairpin probes instead of the target mRNA. It was observed that the fluorescence did not show any appreciable increase, suggesting that non-specific mRNA was unable to activate the fluorescence. There was also no increase in the fluorescence signal when incubating target mRNA with all of the hairpin probes but no probe H2. These results implied that the fluorescence activation was specifically controlled by the interaction between the long HCR products and the DNA–AgNCs.

Since the HCR efficiency was greatly dependent upon the assay conditions such as the concentration of Mg²⁺ and the HCR reaction temperature, to optimize the performance of the developed strategy, mRNA assays were conducted with different concentrations of Mg²⁺ and reaction temperatures (Fig. S3 in ESI†). The peak shape of the fluorescence response was observed to depend on the concentration of Mg²⁺, with a maximum intensity achieved by using 50 mM Mg²⁺. Thus, this concentration of Mg²⁺ was used throughout the subsequent experiments. In order to optimize the reaction temperature, we chose the fluorescence intensity ratio (F/F₀), i.e., the signal from the reaction system with target mRNA divided by that without any target, as the measure to assess the fluorescence, because high temperature could help the HCR but at the same time increase the background signal. The peak shape of the fluorescence intensity ratio was also observed to vary with reaction temperature, with a maximum achieved when the reaction temperature was 37 °C, which was then used throughout the subsequent experiments.

Under the optimized conditions, the ability of the proposed method to quantitatively analyze the nucleic acids was further investigated by incubation of the HCR hairpin probes with target mRNA of varying concentrations followed by the addition of DNA–AgNCs (Fig. 2). The fluorescence intensity was found to increase gradually as the concentration of the target was increased from 10 pM to 0.1 μM. Plots of the fluorescence peak intensity at 630 cm⁻¹ versus mRNA concentration showed the fluorescence signal to be exponentially correlated with the target concentration in a low concentration range from 10 pM to 10 nM and linearly correlated in a high concentration range from 100 pM to 0.1 μM, respectively, with an estimated detection of 7 pM (Fig. 2B). Additionally, the developed method displayed a very desirable reproducibility. The relative standard deviations (RSDs) of the peak intensities at 630 cm⁻¹ were 1.5%, 2.8%, 1.7% and 2.3% for 10 pM, 1 nM, 10 nM and 0.1 μM of target mRNA, respectively, based on four repeats of the measurement at each concentration. Such excellent reproducibility seemed attributed to performing the assays in a homogeneous format, which contributed to a highly reproducible and well-controlled assembly of the nanoparticles. Additionally, compared with most HCR methods,^13–15,23 the AgNC-lightened HCR is a completely homogeneous, label-free fluorescence strategy, which can be performed in simple and cost-effective procedures. Moreover, as opposed to the AgNC-based HCR method directly using the poorly fluorescent AgNCs to generate signals,^24,25 the AgNC-lightened HCR showed improved sensitivity because of the use of the fluorescence enhancement strategy. These results suggested that the AgNC-lightened HCR offered a robust platform for quantitative detection of nucleic acids with good sensitivity and superior reproducibility.

Fig. 2 (A) Typical fluorescence spectra upon the addition of different concentrations of the mRNA target. (B) Correlation of fluorescence intensity (FL) to mRNA target concentration. Error bars are standard deviations of four repeats of the experiments.

The specificity of the AgNC-lightened HCR was also demonstrated. In the control experiments, single-base-mismatched mRNAs were incubated with the four hairpin probes instead of the specific mRNA. The value of the intensity of the signal from the system in which mRNA was present divided by that when the mRNA was absent was used as the measure of specificity. mRNA-1, mRNA-2, mRNA-3, and mRNA-4 with single-mismatched bases at different sites were found to give quite a bit lower fluorescence intensity ratio (F/F₀) values than did the complementary target mRNA (Fig. S4 in ESI†), and this result revealed the desirable selectivity of the AgNC-lightened HCR and its potential ability to discriminate a difference of a single base between mRNA family members.

To demonstrate the ability of AgNC-lightened HCR to detect mRNA in complex samples, target mRNA in total RNA extracts from three human cancer cell lines was analyzed using the developed strategy, including the human breast cancer cell line MCF-7, human breast cell line SKBr-3, and mammary epithelial cell line MCF-10A. The results showed that the AgNC-lightened HCR strategy gave quantitative data consistent with those obtained using qRT-PCR with a maximum relative deviation of 11%, implying the potential of the developed label-free fluorescence-based HCR for quantifying mRNA in real, complex samples (Fig. 3). In addition, the target mRNA showed different expression levels in the cancer cell lines, with the SKBr-3 cell line having the highest expression while MCF-10 cell lines having the lowest expression, which was in good agreement with the data previously reported.²⁶

Fig. 3 Detection of target mRNA expression levels in different cell lysates, as measured using the AgNC-lightened HCR and qRT-PCR method. Error bars are standard deviations, based on four repeats of each experiment.

Conclusions

We have developed a novel label-free and turn-on fluorescence-based strategy by designing an HCR using four hairpin probes and exploiting the unique optical properties of DNA-stabilized AgNCs, which realized homogeneous, enzyme-free detection of nucleic acids under mild conditions. Compared with the regular HCR, the design of an HCR with four hairpin probes avoided non-specific fluorescence activation of the DNA–AgNCs, and the introduction of a guanine-rich sequence as a tail of one of the hairpin probes dramatically enhanced the fluorescence of the dark DNA–AgNCs loaded on the HCR products. Hence, a selective and sensitive strategy for assaying nucleic acids was created. Due to its label-free, enzyme-free, homogeneous, and fluorescence-based detection format, this technique should be very robust, cost-efficient, and readily automatable, supporting its considerable potential in clinical applications. By releasing the initiating probe for the HCR through other molecular recognition events, for example, the protein/small molecule-induced structural switch of aptamers, the AgNC-lightened HCR may also be developed for label-free fluorescence detection of various types of molecules. Therefore, the developed strategy may create a useful platform for the sensitive and selective detection of a broader class of targets and for related studies.

Acknowledgements

This work was supported by NSFC (21527810, 21575036, 21190041, and 21521063), National Key Basic Research Program (2011CB911000), and Fundamental Research Funds for the Central Universities and Young Scholar Support Program of Hunan University.

Notes and references

1. J. Sharma, R. C. Rocha, M. L. Phipps, H. C. Yeh, K. A. Balatsky, D. M. Vu, A. P. Shreve, J. H. Werner and J. S. Martinez, Nanoscale, 2012, 4, 4107–4110.
2. J. T. Petty, J. Zheng, N. V. Hud and R. M. Dickson, J. Am. Chem. Soc., 2004, 126, 5207–5212.
3. T. Li, L. B. Zhang, J. Ai, S. J. Dong and E. K. Wang, ACS Nano, 2011, 5, 6334–6338.
4. H. C. Yeh, J. Sharma, I. Shih, D. M. Vu, J. S. Martinez and J. H. Werner, J. Am. Chem. Soc., 2012, 134, 11550–11558.
5. X. Q. Liu, F. Wang, R. Aizen, O. Yehezkeli and I. Willner, J. Am. Chem. Soc., 2013, 135, 11832–11839.
6. H. Yeh, J. Sharma, J. J. Han, J. S. Martinez and J. H. Werner, Nano Lett., 2010, 10, 3106–3110.
7. J. J. Yin, X. X. He, K. M. Wang, F. Z. Xu, J. F. Shangguan, D. G. He and H. Shi, Anal. Chem., 2013, 85, 12011–12019.
8. J. J. Li, X. Q. Zhong, H. Q. Zhang, X. C. Le and J. J. Zhu, Anal. Chem., 2012, 84, 5170–5174.
9. R. M. Dirks and N. A. Pierce, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 15275–15278.
10. H. M. T. Choi, V. A. Beck and N. A. Pierce, ACS Nano, 2014, 8, 4284–4294.
11. H. F. Dong, K. H. Hao, Y. P. Tian, S. Jin, H. T. Lu, S. F. Zhou and S. J. Zhang, Biosens. Bioelectron., 2014, 53, 377–383.
12. S. Bi, J. L. Zhang and S. S. Zhang, Chem. Commun., 2010, 46, 5509–5511.
13. Z. L. Ge, M. H. Lin, P. Wang, H. Pei, J. Yan, J. Y. Shi, Q. Huang, D. N. He, C. H. Fan and X. L. Zuo, Anal. Chem., 2014, 86, 2124–2130.
14. J. Ge, Z. M. Huang, Q. Xi, R. Q. Yu, J. H. Jiang and X. Chu, Chem. Commun., 2014, 50, 11879–11882.
15. B. Yang, X. B. Zhang, L. P. Kang, G. L. Shen, R. Q. Yu and W. H. Tan, Anal. Chem., 2013, 85, 11518–11523.
16. Z. Wu, G. Q. Liu, X. L. Yang and J. H. Jiang, J. Am. Chem. Soc., 2015, 137, 6829–6836.
17. J. Huang, Y. Wu, Y. Chen, Z. Zhu, X. Yang, C. J. Yang, K. Wang and W. H. Tan, Angew. Chem., Int. Ed., 2011, 50, 401–404.
18. C. X. Li, H. Y. Wang, J. Shen and B. Tang, Anal. Chem., 2015, 87, 4283–4291.
19. W. Q. Song, K. L. Zhu, Z. J. Cao, C. Lau and J. Z. Lu, Analyst, 2012, 137, 1396–1401.
20. J. Dong, X. Cui, Y. Deng and Z. Tang, Biosens. Bioelectron., 2012, 38, 258–263.
21. C. M. Ritchie, K. R. Johnsen, J. R. Kiser, Y. Antoku, R. M. Dickson and J. T. Petty, J. Phys. Chem. C, 2007, 111, 175–181.
22. H. K. Dong, K. H. Hao, Y. P. Tian, S. Jin, H. T. Lu, S. F. Zhou and X. J. Zhang, Biosens. Bioelectron., 2014, 53, 377–383.
23. L. Chen, L. Sha, Y. W. Qiu, G. F. Wang, H. Jiang and X. J. Zhang, Chem. Commun., 2015, 7, 3300–3308.
24. R. Orbach, W. W. Guo, F. Wang, O. Lioubashevski and I. Willner, Langmuir, 2013, 29, 13066–13071.
25. X. Qiu, P. Wang and Z. J. Cao, Biosens. Bioelectron., 2014, 60, 351–357.
26. X. H. Peng, Z. H. Cao, J. T. Xia, G. W. Carlson, M. M. Lewis, W. C. Wood and L. Yang, Cancer Res., 2005, 65, 1909–1917.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details and additional figures. See DOI: 10.1039/c6ra09337j

---