Turn-off colorimetric sensor for sequence-specific recognition of single-stranded DNA based upon Y-shaped DNA structure

A novel turn-off colorimetric sensor for sequence-specific recognition of single-stranded DNA (ssDNA) was established by combining Y-shaped DNA duplex and G-quadruplex-hemin DNAzyme. A G-rich single-stranded DNA (Oligo-1) displays peroxidase mimicking catalytic activity due to the specific binding with hemin in the presence of K+, which was able to catalyze the oxidation of colorless 2,2′-azinobis(3-ethylbenzothiazoline)-6-sulfonic acid (ABTS2−) by H2O2 to generate green ABTS•− radical for colorimetric assay. Oligonucleotide 2 (Oligo-2) was partly complementary with Oligo-1 and the target DNA. Upon addition of target DNA, Oligo-1, Oligo-2 and target DNA can hybridize with each other to form Y-shaped DNA duplex. The DNAzyme sequence of Oligo-1 was partly caged into Y-shaped DNA duplex, resulting in the inactivation of the DNAzyme and a sharp decrease of the absorbance of the oxidation product of ABTS2−. Under the optimum condition, the absorbance decreased linearly with the concentration of target DNA over the range of 1.0–250 nM and the detection limit was 0.95 nM (3σ/slope) Moreover, satisfied result was obtained for the discrimination of single-base or two-base mismatched DNA.

Herein, we developed a turn-off colorimetric sensor for the recognition of single nucleotide polymorphisms based upon Y-shaped DNA duplex and G-quadruplex-hemin DNAzyme. In the absence of target DNA, a G-rich single-stranded DNA (Oligo-1) did not hybridize with assistant DNA (Oligo-2), which induced the formation of DNAzyme in the presence of hemin and K + . When the probes were changed with target DNA (Oligo-3), Oligo-1, Oligo-2 and target DNA hybridized with each other to form Y-shaped DNA duplex, resulting in a sharp decrease of the absorbance of UV-Vis absorption spectrometry. The detection of single-stranded DNA can be realized by observing the optical signal change before and after the addition of target molecules. The assay could not only enhance the sensitivity for the catalytic activity of G-quadruplex-hemin DNAzyme, but also improve the selectivity for the structure of Y-shaped DNA duplex.

Materials and Methods
Materials. Dimethyl sulfoxide (DMSO) and hydrogen peroxide (30%) were purchased from Guangzhou Chemical Reagent Factory (Guangzhou, China). 3,6-dimethyl-2-(4-dimethylaminophenyl) benzo-thiazolium cation (Thioflavin T, ThT) and SYBR Green I (SGI) were purchased from Sangon Biotech Inc. (Shanghai, China). All oligonucleotides were obtained from Sangon Biotech Inc. (Shanghai, China) and the sequences were listed in Table 1. The stock solution of oligonucleotides was prepared with deionized water. The concentration of oligonucleotides was accurately quantified by UV-Vis absorption spectroscopy according to the extinction coefficients (ε 260nm , M −1 cm −1 ): A = 15400, G = 11500, T = 8700, C = 7400. Hemin was obtained from Aladdin Chemistry Co. Ltd. (Shanghai, China). The stock solution of 2.0 mM hemin was prepared in DMSO and stored in darkness at −20 °C. Tris-HCl buffer (pH 7.0, 150 mM KCl) was used in our experiments. All chemicals were used as received without further purification.
Apparatus. Centrifugation experiments were performed on an Anke GL-20G-II centrifuge (Anting Scientific Instrument Factory, China). The absorption spectra of radical anion ABTS• − was carried on a TU-1901 double-beam spectro-photometer (Beijing Purkingje General Instrument Co. Ltd, China). A JASCO Model J-810-150S spectropolarimeter (JASCO International CO. Ltd, Japan) was used for the measurement of circular dicroism (CD) spectroscopy. Fluorescence spectra data were performed with an RF-5301PC spectrophotometer (Shimadzu, Japan).

Results and Discussion
Scheme of the assay. The scheme of the assay is depicted in Fig. 1. Oligo-1 is signal probe that be composed of a G-quadruplex-hemin DNAzyme sequence and a recognition domain for part of the target DNA (Table 1).
Oligo-2 is used as assistant probe, which is partially complementary with Oligo-1 and target DNA. In the absence of target DNA, Oligo-1 does not hybridize with Oligo-2, and the free Oligo-1 can catalyze the oxidation of colorless 2,2′-azinobis(3-ethylbenzothiazoline)−6-sulfonic acid (ABTS 2− ) by H 2 O 2 into green ABTS• − radical. However, Y-shaped structure formed between Oligo-1, Oligo-2 and target DNA upon addition of target DNA, resulting in the inactivation of the DNAzyme and a sharp decrease of the absorbance of UV-Vis absorption spectroscopy. The detection of target DNA can be realized by observing the optical signal change.
Absorption spectrum. To evaluate the possibility of the assay, UV-Vis absorption spectroscopy of the mixture of ABTS 2− and H 2 O 2 was investigated under different conditions. As shown in Fig. 2, the absorbance of Oligo-1 was strong, which indicated that Oligo-1 can form DNAzyme and exhibited a very high catalytic activity towards H 2 O 2 −ABTS 2− system . There was almost no change after incubation with Oligo-2, which demonstrated that Oligo-1 could not hybridize with Oligo-2 solely. But the absorbance decreased dramatically with further addition of target DNA, which might be due to the change of the conformation of Oligo-1.

Validation of the conformational switches.
To explore the conformational switches of the mentioned phenomena, circular dicroism (CD) spectroscopy of DNA was investigated under different conditions. As shown in Fig. 3A, a positive peak at 265 nm and a negative peak at around 245 nm could be observed in the CD spectroscopy of Oligo-1, indicating the formation of G-quadruplex. There was no obvious change upon addition of Oligo-2, which indicated that Oligo-1 still kept G-quadruplex structure. The spectroscopy appeared a positive peak at around 275 nm and a negative Cotton effect of DNA helicity at 245 nm with further addition of Oligo-3, which revealed that the conformation of Oligo-1 changed from G-quadruplex structure to duplex structure 31 .
ThT had been demonstrated as highly fluorescent responsive for G-quadruplexe compared with single/ double-stranded [32][33][34] . Therefore, ThT was used to further observe the conformational switches. As shown in Fig. 3B, strong fluorescence was observed in the presence of Oligo-1, which indicated that Oligo-1 can form G-quadruplex and strongly bind with ThT to generate significantly enhanced fluorescent signal. Almost no remarkable change was observed after incubation with Oligo-2, which demonstrated that Oligo-1 could not hybridize with Oligo-2 and was able to fold into a G-quadruplex. But, the fluorescence intensity decreased dramatically with further addition of target DNA, accompany by a spontaneous conformational change from G-quadruplex to Y-shaped DNA duplex. These were corresponding exactly with the results of CD spectroscopy.  The concentration of Oligo-1 decided the number of DNAzyme. Therefore, the effect of Oligo-1 concentration was investigated. As shown in Fig. 4A, ΔA increased rapidly with the increase of Oligo-1 concentration over the range from 0.2 μM to 1.2 μM and reached a plateau over the range from 1.2 μM to 1.8 μM, so 1.5 μM of Oligo-1 was used for the research.
Oligo-2 can repress the formation of G-quadruplex structure of Oligo-1 in the presence of target DNA, so the effect of Oligo-2 concentration was measured (Fig. 4B). ΔA increased gradually with the concentration of Oligo-2 over the range of 0.3-0.75 μM, ΔA reached maximum and kept a plateau if the concentration of Oligo-2 was higher than 1.0 μM. Therefore, 1.5 μM of Oligo-2 was selected in the assay.
A G-rich sequence can form G-quardruplex with K + , so the concentration of K + was investigated. As shown in Fig. 4C, ΔA was found to be proportional to the KCl concentration over the range of 0-125 mM, then it reached a plateau when the concentration of KCl was higher than 125 mM. Therefore, 150 mM of KCl was selected for further research.
K + -stabilized G-quardruplex exhibited catalytic activity with hemin as the cofactor. Thereby, the concentration of hemin played an important role in the activity of DNAzyme. Herein, the effect of hemin concentration was investigated in Fig. 4D. ΔA increased rapidly with the increase of hemin concentration over the range of 0-12.0 μM, then it kept a plateau during the range of 12.0-20.0 μM. Therefore, 15.0 μM of hemin was selected for further research.    10 base pairs, respectively. As shown in Fig. 5, the change of absorbance (ΔA) was almost the same upon addition of Oligo-2 and Oligo-2b, which was higher than that of Oligo-2a. Therefore, Oligo-2 was used in later experiment.
Sensitivity of the assay. Under the optimum conditions, the relationship between the concentration of target DNA and absorbance was evaluated. As demonstrated in Fig. 6, the absorbance decreased linearly with the increase of the concentration of target DNA during the range of 1.0-250 nM. The linear regression equation was Y = −0.0050 C + 2.1 (C in nM, R = 0.9975) with the detection limit (DL) of 0.95 nM, which was obtained from the equation DL = 3σ/slope. Selectivity of the assay. Sequence selectivity was important to the assay, two control sequences containing 1-bp mismatch (Oligo-4) and 2-bp mismatch (Oligo-5) were designed to evaluate the selectivity of the assay. As shown in Fig. 7A, the change of absorbance (ΔA) of 1-bp mismatch and 2-bp mismatched DNA were lower than that of target DNA. It indicated that single base mismatched sequence could be discriminated well from the target DNA.
To verify the advantage for discrimination of single-base mutation by using Y-shaped DNA duplex, linear DNA duplex acted as the control DNA structure was designed for the fluorescent detection of single-stranded DNA by using the double-stranded DNA-binding dye SYBR Green I (SGI). Block DNA (5′-CTA GTC AGT GTG GAA AAT CTC TAG CCA G-3′) acted as the recognition probe was complementary with the target DNA (Oligo-3). As shown in Fig. 7B, the value of ΔA for one base and two bases mutated sequences were 93.10% and 81.85%  of that for perfect target DNA by using linear DNA duplex, which showed tiny change as compared to that of perfect target DNA. Meanwhile, ΔA of the DNA mismatched by one base and two bases were 35.80% and 32.55% of that for perfect target DNA by using Y-shaped DNA duplex, respectively (Fig. 7A). These results indicated that the selectivity of Y-shaped DNA duplex is better than that of linear DNA duplex.

Conclusions
A novel turn-off colorimetric sensor for the recognition of single-stranded DNA was established by combining Y-shaped DNA duplex and G-quadruplex-hemin DNAzyme. The assay had high sensitivity and selectivity for the use of G-quadruplex-hemin DNAzyme and Y-shaped DNA duplex structure. Moreover, no expensive and sophisticated instruments and no tedious DNA covalent labelling procedure was used in the assay, which reduce cost of the assay. Under the optimum conditions, the proposed sensor allowed the detection of target DNA over the range of 1.0-250 nM with a detection limit of 0.95 nM. Furthermore, single-base and two-base mismatched sequence could be discriminated well from the target DNA.