Fluorescence detection of single nucleotide polymorphisms using a universal molecular beacon

Abstract

We present a simple and novel assay—employing a universal molecular beacon (MB) in the presence of Hg²⁺—for the detection of single nucleotide polymorphisms (SNPs) based on Hg²⁺–DNA complexes inducing a conformational change in the MB. The MB (T₇-MB) contains a 19-mer loop and a stem of a pair of seven thymidine (T) bases, a carboxyfluorescein (FAM) unit at the 5′-end, and a 4-([4-(dimethylamino)phenyl]azo)benzoic acid (DABCYL) unit at the 3′-end. Upon formation of Hg²⁺–T₇-MB complexes through T–Hg²⁺–T bonding, the conformation of T₇-MB changes from a random coil to a folded structure, leading to a decreased distance between the FAM and DABCYL units and, hence, increased efficiency of fluorescence resonance energy transfer (FRET) between the FAM and DABCYL units, resulting in decreased fluorescence intensity of the MB. In the presence of complementary DNA, double-stranded DNA complexes form (instead of the Hg²⁺–T₇-MB complexes), with FRET between the FAM and DABCYL units occurring to a lesser extent than in the folded structure. Under the optimal conditions (20 nM T₇-MB, 20 mM NaCl, 1.0 μM Hg²⁺, 5.0 mM phosphate buffer solution, pH 7.4), the linear plot of the fluorescence intensity against the concentration of perfectly matched DNA was linear over the range 2–30 nM (R² = 0.991), with a limit of detection of 0.5 nM at a signal-to-noise ratio of 3. This new probe provides higher selectivity toward DNA than that exhibited by conventional MBs.

INTRODUCTION

The past decade has witnessed the development of many advanced biomolecular recognition probes for highly sensitive and selective detection of DNA molecules (genes) of interest (1–6). One such set of promising probes are single-stranded DNA molecular beacons (DNA-MB) that form hairpin-shaped structures to recognize targeted DNA molecules. To allow the monitoring conformation changes in DNA-MB upon reactions with targeted DNA, a fluorophore and a quencher are covalently conjugated at the termini of each DNA-MB strand. DNA-MBs act as fluorescence resonance energy transfer (FRET)-based switches that are normally in the closed or ‘fluorescence off’ state, but switch to the open or ‘fluorescence on’ state in the presence of target (complimentary) DNA strands (7).

When DNA-MBs are used for the detection of single nucleotide polymorphisms (SNPs), problems associated with their nonspecific binding to DNA-binding proteins and endogenous nuclease degradation occur, leading to false-positive signals and their limited applicability in complex biological samples (8–10). MBs containing nuclease-resistant backbone residues, such as negatively charged phosphorothioates and neutral peptide nucleic acids, have been developed, but they sometimes exhibit toxicity, self-aggregation and nonspecific binding to single-stranded DNA (ss-DNA)-binding protein (SSB) (11–13). To provide high sensitivity and fast hybridization kinetics, hybrid molecular probes consisting of two ss-DNA sequences tethered to two ends of a poly(ethylene glycol) chain have been developed (14). The two ss-DNA sequences are complementary to adjacent areas of a target sequence in such a way that hybridization of the probe with the target brings the 5′- and 3′-ends of the probe in close proximity. Nevertheless, hybrid molecular probes are more difficult to prepare and are more expensive than conventional DNA-MBs.

Probes based on the Hg²⁺-induced conformational change of a DNA molecule through thymidine (T)–Hg²⁺–T coordination have been realized for the detection of Hg²⁺ ions (15–18). A DNA sensor has been employed for the detection of Hg²⁺ through the enhanced efficiency of FRET as a result of formation of T–Hg²⁺–T complexes (15). Recently, we presented a simple and rapid colourimetric assay—employing poly-T_n and 13 nm-diameter Au NPs in the presence of salt—for the detection of Hg²⁺ ions based on Hg²⁺–DNA complexes inducing the aggregation of Au NPs (17).

In this article, we present a simple and novel assay—employing T₇-MB in the presence of salt and Hg²⁺—for the detection of SNPs based on Hg²⁺–DNA complexes inducing a conformational change in T₇-MB. The T₇-MB contains a stem of a pair of 7-mer T bases that interact with Hg²⁺ and a loop of 19-mer DNA bases that recognize targeted DNA. According to our previous study (18), for obtaining stable DNA–Hg complexes that allow selective detection of target DNA, the minimum number of T is 14. Therefore, 7-mer bp of Ts in the stem region are necessary in the stem region for providing a proper function. The T₇-MB probe contains a donor of carboxyfluorescein (FAM) at the 5′-end, and a quencher of 4-([4-(dimethylamino)phenyl]azo)benzoic acid (DABCYL) at the 3′-end (the sequence of the MB listed in Table 1). The T₇-MB is a random-coil structure that changes into a folded structure in the presence of Hg²⁺ ions through T–Hg²⁺–T bonding (19–21). As a result of the decreased distance between the donor and quencher, the fluorescence of FAM in the Hg²⁺–T₇-MB complexes becomes weaker because of FRET occurring between the FAM and DABCYL units. When the DNA loop of T₇-MB interacts with a targeted DNA more strongly than do the T₇ units in the stem with Hg²⁺, a double-stranded DNA forms, rather than the folded structure. In this case, the FAM and DABCYL units are separated far apart, resulting in FAM fluorescing strongly, as depicted in Scheme 1. We investigated the effect of the Hg²⁺ concentration on the sensitivity and selectivity of the T₇-MB probe, and compared its sensing performance toward SNPs with that of conventional DNA-MBs.

Table 1.

DNA sequences of MBs and trget DNA

Name	Sequence (5′–3′)
T7-MB	FAM-TTTTTTTTCTAAATCACTATGGTCGCTTTTT TT-DABCYL
DNA-MB1	FAM-ACTTAGTTCTAAATCACTATGGTCGCACTA AGT-DABCYL
DNA-MB2	FAM-ACCTAGCTCTAAATCACTATGGTCGCGCTA GGT-DABCYL
DNA-MB3	FAM-GCCGAGCTCTAAATCACTATGGTCGCGCTC GGC-DABCYL
DNApm	GCGACCATAGTGATTTAGA
DNAmm1	GCGACCATAATGATTTAGA
DNAmm2	GCGACCATACTGATTTAGA
DNAmm3	GCGACCATATTGATTTAGA
DNAmm4	GCGACCATAGAGATTTAGA
DNAmm5	GCGACCATAGCGATTTAGA

Bold and underlined letters indicate the mutant bases.

Scheme 1.

Schematic representation of the working principles of the T₇-MB.

MATERIALS AND METHODS

Chemicals

Mercury(II) chloride (HgCl₂) and magnesium(II) chloride (MgCl₂) used in this study were purchased from Aldrich (Milwaukee, WI, USA). Sodium phosphate dibasic anhydrous and sodium phosphate monobasic monohydrate, obtained from J. T. Baker (Phillipsburg, NJ, USA), were used to prepare the phosphate buffer (5.0 mM, pH 7.4). The T₇-MB, DNA-MB_x (x = 1−3), perfectly matched DNA (DNA_pm) and mismatched DNA (DNA_mmx) (see Table 1 for sequences) were purchased from Integrated DNA Technology, Inc. (Coralville, IA, USA). The sequences in T₇-MB and DNA-MB_x that do not have any biological targets were randomly designed to provide optimum selectivity toward the target DNAs and hybridization kinetics (4). Milli-Q ultrapure water was used in all experiments.

Analysis of samples

Aliquots (400 μl) of 5.0 mM phosphate buffer (pH 7.4) containing NaCl (0–250 mM) and MB (20 nM) were maintained at ambient temperature for 10 min. Aliquots (50 μl) of tested DNA (1.0 μM) were added to the solutions, which were then incubated for 30 min. The final ratio of the concentrations of the MB and the tested DNA was 1:5. An aliquot (50 μl) of Hg²⁺ (0–1.5 μM) was added to each solution, which was then incubated for 2 h prior to fluorescence measurements (Cary Eclipse; Varian, CA, USA) at various temperatures (10–90°C). To evaluate the resistance to endogenous nuclease degradation, aliquots (450 μl) of 5.0 mM phosphate buffer (pH 7.4) containing NaCl (20 mM), MgCl₂ (5.0 mM), T₇-MB or DNA-MB (20 nM) and Hg²⁺ (1.0 μM) were maintained at ambient temperature for 2 h. An aliquot (50 μl) of DNase I (final concentration: 5.0 μg/ml) was added to each solution and then the mixtures were subjected to fluorescence measurements after certain periods of time, as indicated in the Results and discussion section. To evaluate the nonspecific binding to SSB, 5.0 mM phosphate buffer (pH 7.4, 450 μl) solutions containing NaCl (20 mM), SSB (100 nM) and T₇-MB or DNA-MB (20 nM) were maintained at ambient temperature for 30 min. An aliquot (50 μl) of Hg²⁺ (1.0 μM) was added to each solution, which was then incubated for 2 h prior to fluorescence measurement.

RESULTS AND DISCUSSION

Sensing behavior

Two aliquots of the T₇-MB (20 nM) were separately added to 5.0 mM phosphate buffers containing 20 mM NaCl solution (pH 7.4) in the absence and presence of targeted DNA (DNA_pm; 100 nM) and then the mixtures were equilibrated for 30 min at ambient temperature. Two aliquots of Hg²⁺ (final concentration: 1.0 μM) were then added separately to the two mixtures. In the absence of the target DNA, the fluorescence of FAM (excitation wavelength: 475 nm) was low, as indicated in Figure 1 (spectrum a). In the presence of the targeted DNA, the fluorescence (spectrum b) of FAM was higher than that in the absence of the target DNA. These results support the sensing mechanism illustrated in Scheme 1. When using a single base mismatched DNA (DNA_mm1) having the sequence listed in Table 1 as a control, the fluorescence of FAM (spectrum c) was only slightly higher than that in the absence of the targeted DNA, suggesting that the T₇-MB probe has high specificity toward DNA_pm. In addition, the selectivity of T₇-MB toward DNA_pm to DNA_mm1 increased upon increasing in the ratio of the targeted DNA to T₇-MB, and achieved a maximum when the targeted DNA was used in 5-fold excess (inset of Figure 1). The selectivity values of T₇-MB (20 nM) toward DNA_pm over DNA_mm1 were 3.0, 3.0, 4.2, 9.5 and 46 when the molar ratios of the target DNA to T₇-MB were 0.1, 0.5, 1, 2 and 5, respectively. The selectivity increased upon increasing the concentration of the targeted DNA, because the hybrid structure of T₇-MB with DNA_pm is more stable than that with DNA_mm1. The use of four other single base mismatched DNA strands (DNA_mm2–5) provided similar results to those obtained using DNA_mm1. Furthermore, when using a random DNA sequence (5′-ACCTGGAAGAGTATTGCAA-3′) as a control to test the specificity of our T₇-MB, we did not observe any change in the fluorescence. The highly specific nature of our T₇-MB probe suggested that it would have great potential for use in SNPs studies.

Figure 1.

Fluorescence spectra of the T₇-MB (20 nM) in (a) the absence of target DNA and (b, c) the presence of (b) DNA_pm (100 nM) and (c) DNA_mm1 (100 nM). Inset: the values of (I_F1–I_F0)/(I_F2–I_F0) of T₇-MB in the presence of DNA_pm (I_F1) and DNA_mm1 (I_F2), as functions of concentration ratio of DNA_target to T₇-MB. The solution contained 5 mM phosphate buffer (pH 7.4), 1.0 μM Hg²⁺ and 20 mM NaCl.

Effect of Hg²⁺ concentration

The sensing capability of our T₇-MB probe for DNA depends on the interplay of the complexes formed between T₇ and Hg²⁺ and between the DNA sequence in the loop and the tested DNA. Thus, we expected that the specificity and sensitivity of our T₇-MB probe would depend on the concentration of Hg²⁺, because it affects the amount of Hg²⁺–T₇-MB complex formed. We investigated the effect of Hg²⁺ at various concentrations (0–1.5 μM) on the fluorescence of the FAM unit in the T₇-MB in the absence of tested DNA. Upon increasing concentration of Hg²⁺ in the presence of 20 nM T₇-MB (Figure 2A, closed square), the fluorescence of FAM initially decreased rapidly (from 0 to 0.5 μM) and then decreased more gradually (from 0.5 to 1.5 μM). This result suggests that the folded DNA structure was more stable in the presence of higher concentrations of Hg²⁺. To support this hypothesis, we conducted melting temperature measurements; here, we define T_m as the temperature at which the fluorescence of FAM reaches 50% of its original value. Upon increasing the temperature, the fluorescence intensity increased as a result of breaking the T–Hg²⁺–T bonds (Figure 2B). Upon increasing the Hg²⁺ concentration, the value of T_m increased, reaching a plateau at the concentration of Hg²⁺ of 1.0 μM (inset to Figure 2B).

Figure 2.

(A) Plots of (closed square) the fluorescence intensity at 518 nm of T₇-MB (20 nM) and (open square) the values of (I_F1– I_F0)/(I_F2–I_F0) of T₇-MB in the presence of DNA_pm (I_F1) and DNA_mm1 (I_F2), both as functions of the concentration of Hg²⁺ (0–1.5 μM). (B) Fluorescence spectra of the T₇-MB (20 nM) recorded a various temperatures. Inset: plot of the value of T_m of T₇-MB as a function of the concentration of Hg²⁺ (0–1.5 μM). Other conditions were the same as those described inFigure 1.

The results in Figure 2 suggest that the concentration of Hg²⁺ is an important factor determining the specificity of the T₇-MB. Thus, to determine the optimal Hg²⁺ concentration under the tested conditions, we plotted (I_F1 – I_F0)/(I_F2 – I_F0) against the Hg²⁺ concentration, where I_F0, I_F1, and I_F2 are the fluorescence intensities of the FAM unit in T₇-MB in the absence of the targeted DNA and in the presence of DNA_pm and DNA_mm1, respectively. A higher value of this ratio indicates better specificity of the T₇-MB probe toward DNA_pm over DNA_mm. Figure 2A (open square) indicates that the ratio was maximized at an Hg²⁺ concentration of 1.0 μM; at higher concentrations (e.g. 10 μM), the T₇-MB prefers to complex with Hg²⁺, reducing its ability to recognize its target DNA. In addition, the temperature also affected the specificity of the T₇-MB. The specificity of the T₇-MB probe toward DNA_pm over DNA_mm achieved a plateau at ambient temperature (25–30°C). At higher temperature, the T–Hg²⁺–T bonds were broken as a result of decreasing the specificity (Figure S1). Thus, the optimal conditions—providing the highest specificity of the T₇-MB toward its target DNA—involved the use of 20 nM T₇-MB in 5.0 mM phosphate buffer (pH 7.4) containing 1.0 μM Hg²⁺ and 20 mM NaCl at ambient temperature.

Next, we separately investigated the kinetics of forming folded structures of the T₇-MB with and without targeted DNA in the presence of Hg²⁺. The fluorescence intensity of the T₇-MB decreased immediately once Hg²⁺ was added. However, the fluorescence intensities took 1.5 and 2.0 h to achieve constant values in the presence of DNA_pm and DNA_mm1, respectively (Figure S2). Figure S2 reveals that the folded rate of the T₇-MB with DNA_mm1 was slower than that with DNA_pm. The kinetics of this probe is slow, because some undesired Hg-oligonucleotide complexes may be kinetically preferred formed, especially in the case of DNA_mm1 (20). Based on these kinetics, we employed an equilibrium time of 2.0 h in the following experiments.

Sensitivity and specificity

We investigated the sensitivity of the T₇-MB at different concentrations toward DNA_pm. Figure 3 indicates that the fluorescence intensity increased upon increasing the concentration of DNA_pm when using 20 nM T₇-MB. We obtained a linear response (R² = 0.991) of the fluorescence intensity against the concentration of DNA_pm over the range 2–30 nM, (inset to Figure 3), with a limit of detection of 0.5 nM at a signal-to-noise ratio of 3. The LODs of DNA_pm by using T₇-MB at the concentrations of 10.0 and 50.0 nM were 0.48 and 1.20 nM, respectively. High concentration of T₇-MB probe produced high background fluorescence intensity, leading to decreases in the sensitivity. When using low concentrations (<20 nM) of T₇-MB, poor selectivity toward DNA_pm is problematic. Relative to other existing methods for the detection of DNA using DNA-MBs (the optimum conditions as shown in Figure S3), the T₇-MB probe provides at least a 3-fold improvement in sensitivity. The relative standard deviation for quantitation of DNA using the T₇-MB probe was <0.8%.

Figure 3.

Fluorescence spectra of T₇-MB (20 nM) recorded at various concentrations of DNA_pm. Inset: plot of the fluorescence intensity at 518 nm of T₇-MB (20 nM) as a function of the concentration of DNA_pm. Other conditions were the same as those described in Figure 1.

To compare the present system to a conventional DNA-MB probe for the study of SNPs, we employed the two systems separately for the detection of DNA_pm and five mismatched strands DNA_mm1–5. Because the stability of DNA-MB_x probe depends on the GC content in the stem, three different DNA-MB_x probe (x = 1–3; no Hg²⁺) as listed in Table 1 were chosen. The performances of the four MB probes were evaluated according to the values of (I_F–I_F0)/I_F0, where I_F0 is the florescence intensity of the FAM in T₇-MB or DNA-MB in the absence of target DNA and I_F values are those in the presence of DNA_pm or DNA_mm1–5, respectively. Figure 4A reveals that our T₇-MB probe exhibits enhanced specificity over the conventional DNA-MB_x under the optimal conditions (20 nM T₇-MB in the presence of 1.0 μM Hg²⁺ or DNA-MB_x (x = 1–3), 20 mM NaCl and 5.0 mM phosphate buffer solution, pH 7.4 at 35°C). We further conducted similar experiments under physiological conditions (150 mM NaCl, 5.0 mM KCl, 1.0 mM MgCl₂, 1.0 mM CaCl₂ and 25 mM Tris–HCl buffer solution, pH 7.4). The specificity values of T₇-MB and DNA-MB_x (x = 1–3) toward DNA_pm over DNA_mm1 were 69-fold for the T₇-MB probe (20 nM in the presence of 100 μM Hg²⁺), and 1.0-, 1.1- and 1.2-fold for DNA-MB_x (20 nM; x = 1–3), respectively. We also compared the stabilities of the T₇-MB and DNA-MB₂ probes in the presence of the endonuclease DNase I (Figure 4B). The DNA-MB₂ degraded rapidly once DNase I was added, whereas the T₇-MB remained unaffected for at least 20 min under otherwise identical conditions. After 2 h, at least 50% of the T₇-MB in the presence of Hg²⁺ remained in its folded structure, based on changes in the fluorescence intensity. This behavior arose mainly because the folded structure of the T₇-MB is more stable than the random-coil structure of the DNA-MB_x. We finally compared the resistance of the T₇-MB and DNA-MB₂ probes toward nonspecific binding proteins. DNA-MB_x are subjected to nonspecific binding to SSB. Binding of the DNA-MB₂ to SSB caused it to remain in a randomly coiled structure, leading to a false-positive signal (Figure 4C). For simplicity, we normalized the fluorescence intensities of the two MBs in the presence of SSB to their respective values in the absence of SSB. Interestingly, our results reveal that the T₇-MB was barely affected after the addition of excess SSB, indicating that this probe is superior to conventional MBs for detecting target DNA strands within biological samples containing high amounts of SSB. Table 2 compares our present approach with four popular approaches [conventional DNA-MB, locked nucleic acid (LNA)-MB, superquenchers-MB and hybrid-MB] to SNPs study with respect to detection limit, specificity and resistance to SSB and nuclease digestion. The specificity of our method is superior to the other four methods. The sensitivity of our approach is comparable to those of superquenchers-MB and hybrid-MB approaches, and is better than those of conventional DNA-MB and LNA-MB approaches. Like our approach, LNA-MB and hybrid-MB resist to the binding of SSB and nuclease digestion. However, the LNA-MB and hybrid-MB are more difficult and expensive to prepare. Nevertheless, the use of toxic Hg²⁺ ions, albeit in small amounts, in our probe system is a disadvantageous feature. This disadvantage can be overcome by using different DNA sequences that respond to the presence of lower-toxicity metal ions such as Ag⁺ and K⁺ ions (22–27).

Figure 4.

(A) Fluorescence enhancements of T₇-MB and DNA-MB_x (20 nM) in the presence of DNA_mm1, DNA_mm2, DNA_mm3, DNA_mm4, DNA_mm5 and DNA_pm. The final concentration ratios of the T₇-MB and DNA-MB_x to the tested DNA were 1:5. The fluorescence measurements of T₇-MB and DNA-MB_x were at ambient temperature and 35°C, respectively. (B) Digestion of (a) T₇-MB and (b) DNA-MB₂ (20 nM) by DNase I (5.0 μg/ml) in the presence of 5.0 mM MgCl₂. (C) Responses of the two MBs toward the presence of SSB. The final ratio of the concentrations of MB and SSB was 1:5. Other conditions were the same as those described in Figure 1.

Table 2.

Comparison of SNPs studies using T₇-MB and other four different approaches

Type of MB	Detection limit of the DNApm (nM)	Specificitya	Resistance
			SSB	Nuclease	Reference
Conventional DNA-MB	1.5	7.0	No	No	In this study
T7-MB	0.5	69	Yes	Yes	In this study
LNA-MB	10	10	Yes	Yes	Wang, L., et al. (6)
Superquenchers- MB	0.1	30	No	No	Yang, C.J., et al.(7)
Hybrid-MB	0.8	25	Yes	Yes	Yang, C.J., et al. (14)

^aSpecificity: (I_F1–I_F0)/(I_F2–I_F0) where I_F0, I_F1 and I_F2 are the fluorescence intensities of the fluorophore units in the MBs without the targeted DNA and with DNA_pm and DNA_mm, respectively.

CONCLUSIONS

We have developed a new sensing strategy for SNPs study using T₇-MB probe in the presence of Hg²⁺. This new approach is simple, sensitive, selective and cost-effective for studying SNPs. The T₇-MB probe in the presence of Hg²⁺ has greater resistance toward nuclease digestion and undergoes less nonspecific binding with SSB. When compared with the conventional MB approaches, the T₇-MB probe provides a greater specificity toward perfect-matched DNA over mismatched DNA and is more stable in the presence of high concentrations of salt. When SNPs study under physiological conditions is needed, the stability and specificity of the T₇-MB probe can be further improved by carefully controlling Hg²⁺ concentrations and/or the stem length. The superior characteristics of the T₇-MB probe show its great potential for use in SNPs studies.

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online.

FUNDING

National Science Council of Taiwan (NSC 96-2627-M-002-013 and NSC 96-2627-M-002-014); National Taiwan University for PDF support (96R0066-37 to Y.-W.L.). Funding for open access publication charge: National Science Council of Taiwan (NSC 96-2627-M-002-013 and NSC 96-2627-M-002-014).

Conflict of interest statement. None declared.