A label-free kissing complex-induced fluorescence sensor for DNA and RNA detection by using DNA-templated silver nanoclusters as a signal transducer

Abstract

Riboswitches are complex folded RNA domains that serve as receptors for specific metabolites which are widely used in biosensor design. They are comprised of a biosensor that includes the binding site for a small ligand and they respond to association with this ligand by undergoing a conformational change. In the present study, we report on the integration of silver nanoclusters (AgNCs) and riboswitches for the expanding of the application of a kissing complex-induced sensor (KCIS). This strategy allows for simple tethering of the specific oligonucleotides stabilizing the AgNCs to the nucleic acid probes. We specifically apply the tunable riboswitches properties of this strategy to demonstrate the multiplexes analysis of DNA and RNA. This is a new concept for oligonucleotides detection, and opens an opportunity for design of more novel biosensors based on the kissing complex-induced strategy.

---

Introduction

Riboswitches are RNA-based sensors that use an aptamer domain to bind to a target, change shape and alter gene expression levels by modulating translation initiation or transcriptional termination of mRNAs.^1–3 While natural riboswitches have evolved as exquisite sensors, it remains difficult to design riboswitch based non-natural biosensors that utilize different aptamers to detect and respond to targets of interest.^4,5 On the other hand, most aptamers have not been converted into functional riboswitches as existing methods have relied on qualitative design by experts, combinatorial library generation and high-throughput screening, which have limited the applications of riboswitch-based approaches. The first event leading to riboswitches dimerization is the formation at the dimerization initiation site (DIS) of a loop–loop complex, also called “kissing complex”,^6,7 by Watson–Crick pairing of a self-complementary sequence within an apical hairpin loop. Toulmé et al. reported a riboswitches based kissing complex for the detection of small ligands.⁸ They exploited the formation of kissing complexes for sensing the presence of an adenosine or guanosine triphosphate (GTP) that is specifically recognized by a hairpin aptamer. The binding of the target shifts the aptamer conformation from an unfolded to a folded shape, hence the name aptaswitch. The folded structure is then recognized by a second hairpin, named aptakiss, which is able to form a kissing complex with the aptaswitch via the “loop–loop recognition”. We also reported a kissing complexes-induced strategy for the detection of adenosine and adenosine deaminase.³ This new strategy is a new concept for aptamer based biosensors, and opens an opportunity for the design of more novel biosensors based on the kissing complexes-induced strategy. However, these methods strictly based on the aptamer recognition for the target which limits the field of application. So, it remains a significant challenge to rationally design sensors for other biomolecules (DNA and RNAs) of interest.

Molecular clusters of pure metals, such as gold nanoclusters (AuNCs) and silver nanoclusters (AgNCs), are a new class of fluorophores and have great potential for applications in bioassay.^9–11 Especially, AgNCs have desirable photophysical properties and low toxicity suitable for biological applications and have attracted special attention due to their facile synthesis, tunable fluorescence emission, and high photostability.^9,12,13 In recent years, AgNCs have been successfully employed to detect various biologically important biomolecules based on different signal-transducing mechanisms.^9,13–16 However, the scarcity of efficient and accurate strategy for modulating the fluorescence of DNA/AgNCs restricts their potential for general application in biomolecules analysis.

In the present study, we report on the integration of DNA–AgNCs and the formation of “kissing complex” for the development of an oligonucleotide sensor. We specifically apply the tunable riboswitches-like properties of this strategy to demonstrate the multiplexes analysis of DNA (Homo sapiens hemoglobin beta chain (HBB) gene) and RNA (Let-7a). In contrast to related riboswitches-based aptasensors, the present study allows for simple detection of DNA and RNA, which extends the application of kissing complex in biosensor design.

Experimental

Materials and chemicals

Human serum was purchased from Sigma-Aldrich (Shanghai, China). Nuclease inhibitor RNasin, and diethypyrocarbonate (DEPC) were ordered from Promega (Madison, USA). Oligonucleotides were purchased from Genscript Biotechnology Co., Ltd (Nanjing, China) and listed in Fig. 1. Silver nitrate (AgNO₃) were purchased from Sinopharm Chemical Reagent Company (Shanghai, China). Other chemicals were all of analytical grade. All solutions were prepared with Milli-Q (Branstad) purified double distilled water having specific resistance of >18 MΩ cm. All oligonucleotide samples were prepared with phosphate buffer (20 mM phosphate, 10 mM magnesium acetate, 1 unit per μL RNasin, 0.1% DEPC, pH 7.0). The tips and tubes are RNase-free and do not require pretreatment to inactivate RNases.

Fig. 1 Oligonucleotides used in this strategy. The colors of the sequences are the same as given in Scheme 1.

Preparation of DNA-templated silver nanoclusters (DNA–AgNCs)

DNA stabilized Ag nanoclusters were synthesized according to the modifications of a literature procedure.¹⁷ Briefly, 3 μM oligo1 or control oligonucleotides (oligo3, oligo5, oligo7, and oligo9) and 18 μM AgNO₃ were sequentially added and mixed with sodium phosphate buffer, and the reaction mixture was incubated at room temperature, in the dark, for 20 minutes. 18 μM NaBH₄ was then added and the reaction mixture was incubated at room temperature, in the dark, for 14 h. Following reduction of Ag⁺ ions, fluorescent DNA–AgNCs were produced with fluorescence emission at 570 nm (excitation at 465 nm).

Probe stability detection

Oligo2 template (3 μM) and 3 μM HBB were mixed in sodium phosphate buffer for 30 min. Then the oligo1–AgNCs complex solution was mixed with oligo2–target DNA duplex solution. After incubating at room temperature for 5 min, the fluorescence emission spectra were recorded along with incubation time.

DNA and RNA detection

Oligo2 template (3 μM) (oligo11 for Let-7a assay) or control oligonucleotides (oligo4, oligo6, oligo8, and oligo10) and appropriate concentrations of HBBs or Let-7as were sequentially added and mixed with sodium phosphate buffer, and the reaction mixture was incubated at room temperature for 30 min. Then the prepared oligo1–AgNCs complex solution was mixed with oligo2–HBB duplex or oligo11–Let-7a duplex solution immediately. After incubating at room temperature for 5 min, the fluorescence emission spectra of the obtained solution were directly recorded.

In order to evaluate the stability of the formed complex, the fluorescence intensity of oligo1–AgNCs/oligo2–HBB complex (1 μM HBB) complex was recorded at different times.

Gel electrophoresis

Denaturing polyacrylamide gel electrophoresis (PAGE) (15%) was carried out in 1 × TBE buffer at a 120 V constant voltage for 1.5 h. For oligo1 and oligo2 solutions preparation, the oligo1 and oligo2 was incubated with oligo2 for 5 min. And for oligo2/HBB duplex preparation, oligo2 and HBB were mixed with sodium phosphate buffer and incubated at room temperature for 30 min. In order to prepare oligo1/oligo2/HBB complex, oligo1 template and HBB were mixed with sodium phosphate buffer, and the reaction mixture was incubated at room temperature for 30 min. Then the oligo1 was mixed with oligo2–HBB duplex for 5 min. The gel was taken photograph under UV light after staining with EtBr for 15 min. The concentration of each DNA is 1 μM.

Results and discussion

Mechanism of this strategy

The analysis of HBB or Let-7a by the new strategy is depicted schematically in Scheme 1. We designed our strategy based on KC24-KG51, a RNA–RNA kissing complex previously identified by Toulmé's group.⁶ These hairpins potentially form a 6 bp loop–loop helix that includes one GU pair and five GC pairs. Oligokiss was modified by adding a stem sequence and an AgNC nucleation sequence at the 5 prime end, and oligoswitch was modified by adding a complementary stem sequence and a G-rich overhang at the 3 prime end. These complementary stem sequences are designed not to form a stable duplex when there is no kissing complex formation. In the absence of HBB, the oligoswitch retains the “free” state. In this state, the RNA loop helix is unstable, which results the formation of oligoswitch–oligokiss complex is prevented. So, the AgNC nucleation sequence and the G-rich overhang far apart from each other which results in the observing of weak AgNC fluorescence intensity. When the HBB or Let-7a is introduced into the system, the recognition part hybridizes the HBB or Let-7a, thereby generating oligoswitch. The oligoswitch can combine with oligo2–AgNCs to form the oligoswitch–oligokiss complex. The formation of oligoswitch–oligokiss complex then promotes hybridization between the complementary stem sequences attached to the oligoswitch and oligokiss. The hybridization brings the G-rich overhang to be close to AgNCs, enhancing the fluorescence of AgNCs.¹⁷ As a result, this assay is capable of effectively detecting DNA or RNA.

Scheme 1 Schematic illustration of the kissing complexes-induced oligosensor for the detection of DNA or RNA.

Feasibility study

The fluorescence signal to background ratio (F_sig/F_back) was highly sensitive to the length of poly T spacer. Oligoswitch–oligokiss complexes were formed using 3 μM HBB and a series of oligo-strands (oligoswitch and oligokiss) by changing the length of poly T spacer from 6 to 14 bases (Fig. 1). As shown in Fig. 2, the best fluorescence signal to background ratio was obtained when the length of poly T spacer is 10 bases. This interesting phenomenon may be caused due to the stability of the hybridization part and the increase of fluorescence intensity in the absence of HBB. The oligo3 (6 bases poly T) and oligo5 (8 bases poly T) may not hybridize with oligo4 (6 bases poly T) and oligo6 (8 bases poly T) effectively, respectively, which makes AgNC “free” in the solution, leading to the AgNCs far away from G-rich overhang. Oligo7 (12 bases poly T) and oligo9 (14 bases poly T) may hybridize with oligo8 (12 bases poly T) and oligo10 (14 bases poly T) effectively, respectively. However, the background in the absence of the HBB was increased with the prolongation of the length of poly T spacer. This phenomenon may be caused by the free hybrids between the hybridization parts in the absence of HBB.¹⁷ Therefore, we chose oligo1 and oligo2 (10 bases poly T each) for the detection in this study.

Fig. 2 Effect of the length of poly T spacer on the signal-to-background ratio for detecting adenosine. The concentration of HBB was 3 μM.

The stability of the oligo1–AgNCs/oligo2–HBB complex is important for the assay. Previous work shows that Ag nanoclusters can be oxidized over time.^18,19 However, the fluorescence intensity shows no obvious decrease in this work within 12 hours (Fig. S1†). So, the stability of the complex is satisfied for the oligonucleotides assay.

Gel electrophoresis characterization

The viability of proposed strategy was further investigated by gel electrophoresis (Fig. 3A). When oligo2 incubated with HBB, the product of oligo2/HBB duplex was formed (lane 5), confirming the recognition ability of oligo2. The results also showed that HBB can help the formation of oligo1/oligo2/HBB complex (lane 6) and without HBB, oligo1 and oligo2 are dispersed in the solution (lane 3).

Fig. 3 (A) Nondenaturing polyacrylamide gel electrophoresis (PAGE) (15%) of the products by the amplification method. Lane (M): marker. Lanes 1–6: (1) oligo1; (2) oligo2; (3) oligo1 and oligo2; (4) HBB; (5) oligo2/HBB duplex; and (6) oligo1/oligo2/HBB complex, respectively. (B) Emission spectra of the strategy for the assay of HBB at different concentrations (0, 10, 40, 50, 100, 200, 500, 1000, and 5000 nM, respectively, from a to i). (C) The relationship between the fluorescence intensity increase and the concentrations of HBB. The inset shows the linear relationship over the concentration range from 0 to 1000 nM. All the data are taken from independent experiments with repetition for at least three times, and the presented data are the results of averaging. (D) The assay readily differentiates between perfectly matched and mismatched HBB. (a) Perfectly matched HBB; (b) single-base mismatched HBB (HBB a in Fig. 1); (c) three-base mismatched HBB (HBB b in Fig. 1) and (d) without target DNA.

Sensitivity of the sensing system

In the presence of HBB, the fluorescence emission spectra of DNA/AgNC was found to be enhanced by HBB, which is attributed to the HBB-mediated formational alteration of the oligoswitch–oligokiss complex resulting in its self-hybridization to facilitate the proximity of the G-rich overhang region and the as-prepared AgNCs templated with the NC region. To further evaluate the analytical performance of the developed method for HBB detection based on the concept demonstrated above, the different concentrations of HBB from one stock solution were added to the oligo2 probes. The fluorescent probe possesses the ability to react with HBB in a dose-response manner (Fig. 3B), which can be utilized for the quantitation of the analyte. The fluorescence intensity gradually increases with the increasing concentration of HBB (Fig. 3C). The fluorescence intensity versus the HBB concentrations can be fitted to a linear regression equation from 0 to 1000 nM (Fig. 3C, inset) with an equation Y = 0.31X + 87.53 (R² = 0.990), where Y is the fluorescence intensity and X is the concentration of HBB. A detection limit was calculated to be 1.2 nM according to the responses of the blank tests plus 3 times the standard deviation (3σ method). The proposed method had a lower detection limit than some of the previously reported fluorescent dye-based sensing approaches.²⁰

The specificity detection

The sensing system is also specific. In order to evaluate this, we challenged our assay using single-base- and three-base-mismatched targets (the sequences are listed in Fig. 1) and found that it readily discriminates between single nucleotide polymorphisms (Fig. 3D). Three-base-mismatched DNA caused a slightly increased fluorescence intensity (curve c) than control (curve d) and only one base mutation could give a distinguished fluorescence intensity (curve b) with perfect match target (curve a). These results demonstrate the feasibility of identifying different types of single-nucleotide mutations using the highly sequence dependent fluorescent AgNC formation in DNA scaffolds. Moreover, we observed that the fluorescence intensity caused by three-base-mismatched DNA was much smaller than that for single-base mismatched targets. This high specificity with the mismatch discrimination ability was derived from the recognition part-DNA hybridization step, which was dominated by the perfect match and highly dependent upon the hybridization stability between the target and the recognition part.

Detection of HBB in complex biological system

A significant challenge for practical analyses is the ability to be applied in a complex biological system. To demonstrate the feasibility of the method in a complex biological system, HBB spiked in 10-fold diluted human serum was evaluated with this method (Table 1). HBB concentration recoveries of 98.1–102.2% were achieved. These results showed that the interference of human serum could overcome since the max acceptable range of recovery is 80–120%.²¹

Table 1 Results of the recovery test of HBB in 10-fold diluted human serum

Sample	HBB
	Added (nM)	Found (nM)	Recovery (%)
1	0	0.2
2	10	10.36	101.6 ± 2.6
3	50	51.31	102.2 ± 2.5
4	100	98.26	98.1 ± 3.2
5	200	202.26	101.0 ± 2.9

---

The assay of target RNA

To demonstrate further the feasibility and universality of present proof-of-concept, a homogeneous model was adopted for detection of RNA with just a little change in the oligoswitch sequence (oligo11). The formation of oligoswitch–oligokiss (oligo11 and oligo1) complex at the help of Let-7a then promotes hybridization between the complementary stem sequences attached to the oligoswitch and oligokiss. Therefore, the hybridization brings the G-rich overhang to be close to AgNCs, enhancing the fluorescence of AgNCs.

We further explored the fluorescence emission spectra of the AgNC in the presence of different concentrations of Let-7a (Fig. 4A). The results showed that as the RNA concentration increased, the fluorescence intensity increased accordingly. Fig. 4B shows the relationship between the fluorescence intensity and the Let-7a concentration, and the inset shows the calibration curve for quantitative analysis of RNA. The intensity was linearly dependent on the concentration of RNA over the range from 0 to 600 nM with an equation Y = 0.34X + 69.97 (R² = 0.998), where Y is the fluorescence intensity and X is the concentration of RNA, and a detection limit of 22 nM could be obtained according to the responses of the blank tests plus 3 times the standard deviation (3σ); this was lower than or comparable to the detection limits for RNA detection.^22,23 Similarly, to test the specificity of this sensing system, three members of let-7 family (let-7a, let-7b, and let-7c), which only one- or two-nucleotide differences out of 22 nucleotides between them, are selected as the detection model. Fig. 4C shows the fluorescence changes for the target RNA and the mismatched DNA strands. These results clearly demonstrate that the detection approach shows a high selectivity toward the target RNA. The recoveries (92.4–100.5%) of Let-7a in 10-time diluted human serum were obtained (Table 2). These results also showed that the interference of human serum could overcome.

Fig. 4 (A) Emission spectra of the strategy for the assay of Let-7a at different concentrations (0, 20, 50, 100, 200, 300, 500, 600, 2000 and 5000 nM from a to i, respectively). (B) The relationship between the fluorescence intensity and the concentrations of Let-7a. The inset shows the linear relationship over the concentration range from 0 to 600 nM. All the data are taken from independent experiments with repetition for at least three times, and the presented data are the results of averaging. (C) Selectivity of Let-7a analysis.

Table 2 Results of the recovery test of Let-7a in 10-fold diluted human serum

Sample	Let-7a
	Added (nM)	Found (nM)	Recovery (%)
1	20	18.9	94.5 ± 1.9
2	50	48.2	96.4 ± 2.8
3	100	92.36	92.4 ± 2.2
4	200	185.25	92.6 ± 2.8
5	500	502.36	100.5 ± 3.2

---

Conclusions

In summary, we have successfully developed a fluorescent biosensor based on the kissing complexes-induced oligosensor for label-free detection of DNA or RNA by using the AgNC as a signal transducer, which exhibits intrinsic attractive properties and expands the application of riboswitches. The identification based on this strategy was further extended to more general types of single-oligonucleotide mismatches. This is a new concept for oligonucleotides assay and opens an opportunity for design of more novel biosensors based on the kissing complexes-induced strategy.

Acknowledgements

This work was supported by grants from the National Natural Science Foundation (81300787), the Natural Science Foundation of Jiangsu Province (BK20141103), the Major Project of Wuxi Municipal Health Bureau (ZS201401, Z201508), the Project of Jiangsu Provincial Commission of Health and Family Planning (No. H201546), and the Project of Wuxi Municipal Science and Technology Bureau (CSE31N1520).

Notes and references

1. C. Brunel, R. Marquet, P. Romby and C. Ehresmann, Biochimie, 2002, 84, 925–944.
2. M. Mandal and R. R. Breaker, Nat. Rev. Mol. Cell Biol., 2004, 5, 451–463.
3. K. Zhang, K. Wang, X. Zhu and M. Xie, Biosens. Bioelectron., 2016, 78, 154–159.
4. A. Rentmeister, G. Mayer, N. Kuhn and M. Famulok, Biol. Chem., 2008, 389, 127–134.
5. A. Espah Borujeni, D. M. Mishler, J. Wang, W. Huso and H. M. Salis, Nucleic Acids Res., 2016, 44, 1–13.
6. E. Ennifar, P. Walter, B. Ehresmann, C. Ehresmann and P. Dumas, Nat. Struct. Mol. Biol., 2001, 8, 1064–1068.
7. F. Ducongé, R. I. C. Eacute and J.-J. Toulmé, RNA, 1999, 5, 1605–1614.
8. G. Durand, S. Lisi, C. Ravelet, E. Dausse, E. Peyrin and J.-J. Toulmé, Angew. Chem., Int. Ed., 2014, 53, 6942–6945.
9. Y.-Q. Liu, M. Zhang, B.-C. Yin and B.-C. Ye, Anal. Chem., 2012, 84, 5165–5169.
10. K. Zhang, K. Wang, M. Xie, X. Zhu, L. Xu, R. Yang, B. Huang and X. Zhu, Biosens. Bioelectron., 2014, 52, 124–128.
11. K. Zhang, T. Ren, K. Wang, X. Zhu, H. Wu and M. Xie, Chem. Commun., 2014, 50, 13342–13345.
12. Z. Huang, F. Pu, Y. Lin, J. Ren and X. Qu, Chem. Commun., 2011, 47, 3487–3489.
13. Z. Huang, Y. Tao, F. Pu, J. Ren and X. Qu, Chem.–Eur. J., 2012, 18, 6663–6669.
14. 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.
15. S. W. Yang and T. Vosch, Anal. Chem., 2011, 83, 6935–6939.
16. H.-C. Yeh, J. Sharma, I.-M. Shih, D. M. Vu, J. S. Martinez and J. H. Werner, J. Am. Chem. Soc., 2012, 134, 11550–11558.
17. J. Li, X. Zhong, H. Zhang, X. C. Le and J.-J. Zhu, Anal. Chem., 2012, 84, 5170–5174.
18. K. Morishita, J. L. MacLean, B. Liu, H. Jiang and J. Liu, Nanoscale, 2013, 5, 2840–2849.
19. S. Walczak, K. Morishita, M. Ahmed and J. Liu, Nanotechnology, 2014, 25, 155501.
20. X. Liu, F. Wang, R. Aizen, O. Yehezkeli and I. Willner, J. Am. Chem. Soc., 2013, 135, 11832–11839.
21. K. Zhang, K. Wang, X. Zhu, F. Xu and M. Xie, Biosens. Bioelectron., 2017, 87, 358–364.
22. L. Qiu, C. Wu, M. You, D. Han, T. Chen, G. Zhu, J. Jiang, R. Yu and W. Tan, J. Am. Chem. Soc., 2013, 135, 12952–12955.
23. K. A. Cissell, S. Shrestha and S. K. Deo, Anal. Chem., 2007, 79, 4754–4761.

---

Footnote

† Electronic supplementary information (ESI) available: Fig. S1. See DOI: 10.1039/c6ra22515b

---