Nanostructure and Corresponding Quenching Efficiency of Fluorescent DNA Probes

Based on the fluorescence resonance energy transfer (FRET) mechanism, fluorescent DNA probes were prepared with a novel DNA hairpin template method, with SiO2 coated CdTe (CdTe/SiO2) core/shell nanoparticles used as the fluorescence energy donors and gold (Au) nanoparticles (AuNPs) as the energy acceptors. The nanostructure and energy donor/acceptor ratio in a probe were controlled with this method. The relationship between the nanostructure of the probes and FRET efficiency (quenching efficiency) were investigated. The results indicated that when the donor/acceptor ratios were 2:1, 1:1, and 1:2; the corresponding FRET efficiencies were about 33.6%, 57.5%, and 74.2%, respectively. The detection results indicated that the fluorescent recovery efficiency of the detecting system was linear when the concentration of the target DNA was about 0.0446–2.230 nmol/L. Moreover, the probes showed good sensitivity and stability in different buffer conditions with a low detection limit of about 0.106 nmol/L.


Introduction
Fluorescent DNA probes [1] are a kind of optical DNA biosensor [2] based on the fluorescence resonance energy transfer (FRET) mechanism [3]. FRET occurs when the electronic excitation energy of a donor chromophore is transferred to a nearby acceptor molecule via a through-space dipole-dipole interaction between the donor-acceptor pair within 1-10 nm distances. The energy donor molecule quickly jumps from the ground state to the electron excitation state. The field of fluorescent DNA probes improved considerably with the introduction of inorganic nanoparticles, such as quantum dots (QDs) [4,5] and gold (Au) nanoparticles (AuNPs) [6][7][8] as energy donors and acceptors [9], respectively. As a new kind of luminescent inorganic fluorophores, QDs are being widely used in chemical sensors [10], DNA detection [11], cell labeling [12], and imaging [13] because they have a broad and continuous excitation spectrum, a narrow size-tunable symmetric emission spectrum, and a high fluorescence quantum yield. AuNPs are widely used in gene delivery [14] and cell labeling [15] because they are bioinert, nontoxic, and readily synthesized and functionalized. In the field of fluorescent DNA biosensors, AuNPs are commonly used as energy acceptors because of their high extinction coefficient and wide absorption spectra.
The FRET-based DNA probe uses the hybridized DNA chains as scaffolds and detection union as a detection mechanism. The donor and acceptor are conjugated with complementary single-stranded
CdTe/SiO 2 fluorescent nanoparticles were synthesized according to the reverse microemulsion method [26]. A total of 2.5 mL of CdTe QDs µL aqueous solution and 1.25 mL of aqueous ammonia (25 wt %) were introduced into a liquid system containing 37.5 mL of cyclohexane, 9 mL of n-hexanol, and 8.85 mL of Triton X-100 under stirring for 30 min. Then, 100 µL of tetraethyl orthosilicate (TEOS) was added into the reaction system with vigorous stirring. The silica growth was completed in the dark at room temperature for 24 h. The resultant nanoparticles were isolated from the microemulsion using acetone and an ultracentrifuge, and the precipitate of the CdTe/ SiO 2 composite particles was sequentially washed with 1-butanol, iso-propanol, ethanol, and water to remove the Triton X-100 and unreacted molecules.
The CdTe/SiO 2 nanoparticles were modified by using a silane coupling agent 3-aminopropyl triethylsilane (APTES), and succinic anhydride to create carboxyl groups on the surface (CdTe/SiO 2 -COOH nanoparticles). A total of 0.2 g of CdTe/SiO 2 and 19 mL of ethanol were added into two 50-mL round bottom flasks under ultrasound for 1 h. Then, a certain amount of APTES was added into the reaction system with stirring at 25 • C for 5 h. Then, the purified nanoparticles, 30 mL of N, N-dimethylformamide (DMF), and 0.5 g of succinic anhydride were mixed under ultrasound for 1 h and then stirred at ambient temperature for 12 h. The CdTe/SiO 2 -COOH nanoparticles were washed three times with ethanol by centrifugation and dried under a vacuum.

Preparation of DNA Hairpin Templates
NH 2 -dsDNA hairpin templates were hybridized from ssDNA 1 , ssDNA 2 , ssDNA 3 , and ssDNA 4 . A total of 0.1 µmol of the above ssDNA were mixed with 1 mL of Tris-HCl buffer solution (0.05 M Tris-HCl containing 50 µL of BSPP (10 mg/mL) and 2 M of NaCl) when the molar ratio of DNA 1 :DNA 2 :DNA 3 :DNA 4 was 1:1:1:1 in a 90 • C water bath for 10 min, and then cooled down to 30 • C at a rate of −3 • C/3 min. The DNA hairpin template solution was added to 0.5 mL of 0.5× TBE buffer (45 mM Tris-HCl, 45 mM boric acid, 1 mM EDTA) containing 10 mg/mL BSPP and 50 mM NaCl, then stored and at 4 • C.
The HS-dsDNA hairpin templates were prepared with ssDNA 3 , ssDNA 4 , ssDNA 5 , and ssDNA 6 using the same method. For the dehybridization of the template on the CdTe/SiO 2 , the sediment was incubated at 40 • C for 15 min. The ssDNA 3 and ssDNA 4 in the dsDNA template were released from the surface of the CdTe/SiO 2 nanoparticles and removed from the suspension together with the supernatant after centrifugation under the same conditions described above. The dehybridization process was repeated three times. The resulting sediment, after the repetition of the dehybridization, was resuspended in a hybridization buffer (0.5 mL of 0.5× TBE buffer containing 0.25 mg/mL BSPP and 50 mM NaCl).

Conjugation of 40-nm AuNPs with ssDNA (AuNPs-ssDNA)
Fren's method was used to obtain AuNPs when the citrate was used as the reduction [24]. HAuCl 4 (1 mL, 1%) was added into 100 mL of double-deionized water and fluxed until boiling. Then, 1% trisodium citrate solution was quickly added into this mixture with strong stirring for 15 min. After the color of reaction system turned wine-red, the reaction was stopped and cooled to room temperature. AuNPs coated with BSPP had a high stability in buffer solutions [24]. The AuNP (500 µL, 16 or 40 nm) suspension was mixed with 100 µL of BSPP solution (50 mg/mL in water) in a microtube and incubated at 50 • C for 1 h. The suspension was then centrifuged, and washed with 0.5× TBE including BSPP (1 mg/mL) buffer solution for three times. The resulting BSPP stabilized AuNPs were resuspended in 1 mL of 0.5× TBE including BSPP (1 mg/mL) buffer solution and stored at 4 • C. The AuNP concentration in the suspensions was estimated using the ultraviolet (UV)-vis spectra of the AuNPs.
AuNPs-ssDNA conjugations were prepared using the traditional method. The AuNPs were conjugated with thiolated ssDNA (ssDNA 5 or ssDNA 6 ), by mixing 200 µL of BSPP-coated AuNPs, 8 µL of 0.5× TBE containing 1.4 M NaCl, 0.1 µmol of ssDNA in 0.5× TBE containing 1 mg/mL BSPP, and 10 mM NaCl. The mixture was incubated under shaking at 25 • C for 24 h. The ssDNA-conjugated AuNPs were separated using ultracentrifugation, and washed with the diluted BSPP solution three times. Finally, the sediment was resuspended in a 0.2 mg/mL BSPP solution.
The synthesis of the 40-nm AuNPs-ssDNA (Au40-ssDNA) conjugation using the mercapto-dsDNA hairpin template was the same as that described in Section 2.4.

Preparation and Detection of Fluorescent DNA Probes
The resultant suspensions prepared in Sections 2.4 and 2.5 were mixed at different ratios. CdTe/SiO 2 -ssDNA and AuNPs-ssDNA were hybridized in a buffer solution of 20 mM Tris-HCl, 50 mM NaCl, and 5 mM MgCl 2 (pH = 8.0) at 37 • C, for 12 h under shaking. For detection, an amount of target ssDNA was introduced into the detection solution under shaking for 12 h. The fluorescence spectra of the probe system were measured with an F-7000 Luminescence Spectrometer (Hitachi, Japan). According to Foster's theory, the FRET efficiency (quenching efficiency, Q e ) can be measured experimentally and is commonly defined as [16]: where F DA is the integrated fluorescence intensity of the donor in the presence of the acceptor(s) and F D is the integrated fluorescence intensity of the donor alone (no acceptors present). When the complementary DNA was present, the hybridized structure of probes opens and the fluorescent intensity of detection system recovers. The fluorescent recovery efficiency (F r ) is given as: where F 0 is the fluorescent intensity of the DNA probe, and F is the fluorescent intensity of the DNA probe with target ssDNA.

Synthesis and Modification of CdTe/SiO 2 Composite Nanoparticles
The aqueous synthesis of CdTe QDs resulted in a small particle size of about 2-3 nm [25]. The coat of SiO 2 increases the particle size for coverage by the DNA hairpin template. The SiO 2 -coated CdTe (CdTe/SiO 2 ) was prepared by the hydrolysis of tetraethyl orthosilicate (TEOS) in a reverse micro-emulsion, with ammonia and CdTe QDs in the aqueous phase, cyclohexane and n-hexanol in the oil phase, and the surfactant TritonX-100 as the emulsifier. For the reaction with the amino groups of the DNA template, the CdTe/SiO 2 nanoparticles were modified with carboxyl surface groups using APTES and succinic anhydride ( Figure 2).

Synthesis and Modification of CdTe/SiO2 Composite Nanoparticles
The aqueous synthesis of CdTe QDs resulted in a small particle size of about 2-3 nm [25]. The coat of SiO2 increases the particle size for coverage by the DNA hairpin template. The SiO2-coated CdTe (CdTe/SiO2) was prepared by the hydrolysis of tetraethyl orthosilicate (TEOS) in a reverse micro-emulsion, with ammonia and CdTe QDs in the aqueous phase, cyclohexane and n-hexanol in the oil phase, and the surfactant TritonX-100 as the emulsifier. For the reaction with the amino groups of the DNA template, the CdTe/SiO2 nanoparticles were modified with carboxyl surface groups using APTES and succinic anhydride ( Figure 2). The transmission electron microscopy (TEM) results for CdTe/SiO2 indicated that the CdTe/SiO2 fluorescent nanoparticles were about 50 nm in diameter with a round and smooth surface, and the particle size had no obvious changes after modification (Figure 3).  The Fourier transform infrared (FT-IR) spectra of CdTe/SiO2 particles before and after modification are shown in Figure 4. The original CdTe/SiO2 fluorescent nanoparticles had strong absorption peaks of Si-OH at 3427 and 945 cm −1 , as well as the Si-O-Si absorption peak at 1095 cm −1 (Figure 4a). The absorption peak at 1563 cm −1 indicated that the amino groups were introduced to the surface of the CdTe/SiO2 particle (Figure 4b), and the peak at 1729 cm −1 was the carboxyl groups of CdTe/SiO2-COOH particles (Figure 4c), indicating that the modification was successful. The transmission electron microscopy (TEM) results for CdTe/SiO 2 indicated that the CdTe/SiO 2 fluorescent nanoparticles were about 50 nm in diameter with a round and smooth surface, and the particle size had no obvious changes after modification (Figure 3).

Synthesis and Modification of CdTe/SiO2 Composite Nanoparticles
The aqueous synthesis of CdTe QDs resulted in a small particle size of about 2-3 nm [25]. The coat of SiO2 increases the particle size for coverage by the DNA hairpin template. The SiO2-coated CdTe (CdTe/SiO2) was prepared by the hydrolysis of tetraethyl orthosilicate (TEOS) in a reverse micro-emulsion, with ammonia and CdTe QDs in the aqueous phase, cyclohexane and n-hexanol in the oil phase, and the surfactant TritonX-100 as the emulsifier. For the reaction with the amino groups of the DNA template, the CdTe/SiO2 nanoparticles were modified with carboxyl surface groups using APTES and succinic anhydride ( Figure 2). The transmission electron microscopy (TEM) results for CdTe/SiO2 indicated that the CdTe/SiO2 fluorescent nanoparticles were about 50 nm in diameter with a round and smooth surface, and the particle size had no obvious changes after modification (Figure 3).  The Fourier transform infrared (FT-IR) spectra of CdTe/SiO2 particles before and after modification are shown in Figure 4. The original CdTe/SiO2 fluorescent nanoparticles had strong absorption peaks of Si-OH at 3427 and 945 cm −1 , as well as the Si-O-Si absorption peak at 1095 cm −1 (Figure 4a). The absorption peak at 1563 cm −1 indicated that the amino groups were introduced to the surface of the CdTe/SiO2 particle (Figure 4b), and the peak at 1729 cm −1 was the carboxyl groups of CdTe/SiO2-COOH particles (Figure 4c), indicating that the modification was successful.  (Figure 4a). The absorption peak at 1563 cm −1 indicated that the amino groups were introduced to the surface of the CdTe/SiO 2 particle (Figure 4b), and the peak at 1729 cm −1 was the carboxyl groups of CdTe/SiO 2 -COOH particles (Figure 4c), indicating that the modification was successful. modification are shown in Figure 4. The original CdTe/SiO2 fluorescent nanoparticles had strong absorption peaks of Si-OH at 3427 and 945 cm −1 , as well as the Si-O-Si absorption peak at 1095 cm −1 (Figure 4a). The absorption peak at 1563 cm −1 indicated that the amino groups were introduced to the surface of the CdTe/SiO2 particle (Figure 4b), and the peak at 1729 cm −1 was the carboxyl groups of CdTe/SiO2-COOH particles (Figure 4c), indicating that the modification was successful.

Carriable ssDNA Number on AuNPs and CdTe/SiO 2 Particles
If ssDNA is freely conjugated into inorganic particles in an aqueous solution using the traditional method, the number of ssDNA on a particle cannot be controlled [20,21]. Keating et al. reported that the curvature of a AuNP considerably influenced the possible number of ssDNA on its surface. Small diameter AuNPs carry a lower number of thiolated single-stranded ssDNA strands [27]. In this work, 16 and 40-nm AuNPs (Au 16 and Au 40 ) were used to investigate the possible ssDNA number on AuNPs when they were conjugated with two complementary ssDNA: ssDNA 7 and ssDNA 9 . The possible ssDNA numbers on AuNPs results are shown in Figure 5, indicating that thiol-terminated ssDNA (HS-DNA) could be easily self-assembled on the surface of Au 40 with a Au-S bond. Therefore, one Au 40 could carry about two to nine ssDNA 7 , while one Au 16 could carry only one ssDNA 9 (Figure 5a), which confirmed Keating's curvature theory of AuNPs [27]. Moreover, when the CdTe/SiO 2 nanoparticles (with ssDNA 8 ) and Au 16 (with ssDNA 9 ) were used to prepare fluorescent DNA probes using the same method, the results showed that the carriable ssDNA number was also one to three (Figure 5b), which indicated that the conjugation of NH 2 -ssDNA with CdTe/SiO 2 was more difficult than that of HS-ssDNA with AuNPs.

Carriable ssDNA Number on AuNPs and CdTe/SiO2 Particles
If ssDNA is freely conjugated into inorganic particles in an aqueous solution using the traditional method, the number of ssDNA on a particle cannot be controlled [20,21]. Keating et al. reported that the curvature of a AuNP considerably influenced the possible number of ssDNA on its surface. Small diameter AuNPs carry a lower number of thiolated single-stranded ssDNA strands [27]. In this work, 16 and 40-nm AuNPs (Au16 and Au40) were used to investigate the possible ssDNA number on AuNPs when they were conjugated with two complementary ssDNA: ssDNA7 and ssDNA9. The possible ssDNA numbers on AuNPs results are shown in Figure 5, indicating that thiol-terminated ssDNA (HS-DNA) could be easily self-assembled on the surface of Au40 with a Au-S bond. Therefore, one Au40 could carry about two to nine ssDNA7, while one Au16 could carry only one ssDNA9 (Figure 5a), which confirmed Keating's curvature theory of AuNPs [27]. Moreover, when the CdTe/SiO2 nanoparticles (with ssDNA8) and Au16 (with ssDNA9) were used to prepare fluorescent DNA probes using the same method, the results showed that the carriable ssDNA number was also one to three (Figure 5b), which indicated that the conjugation of NH2-ssDNA with CdTe/SiO2 was more difficult than that of HS-ssDNA with AuNPs.
(a) (b) Figure 5. Nanostructures of (a) Au nanoparticles (AuNPs) self-assemblies and (b) fluorescent DNA probes by the traditional method.

Controllable Nanostructures of DNA Probes by DNA Hairpin Template
For the investigation and control of the nanostructures of a DNA probe, CdTe/SiO2 nanoparticles, as energy donors, were conjugated with a NH2-dsDNA hairpin template (hybridization by ssDNA1, ssDNA2, ssDNA3, and ssDNA4). After the release of ssDNA3 and ssDNA4 from the conjugation, only one ssDNA1 and ssDNA2 remained on the particle surface ( Figure 1) because the maximum ssDNA number on CdTe/SiO2 nanoparticles is three. Therefore, because ssDNA5 cannot hybridize with ssDNA2, the CdTe/SiO2-ssDNA1-ssDNA2 conjugations could only hybridize with one ssDNA5 on Au40 and form the conjugations of CdTe/SiO2-dsDNA-Au40, the fluorescent DNA probes, into a 1:1 donor/acceptor (CdTe/SiO2:Au40) structure (Figure 6a). If the Au40-ssDNA6 conjugations were added into this preparation system, the fluorescent DNA probes could have a 1:2 donor/acceptor ratio in the nanostructure (Figure 6b) because ssDNA6 was only

Controllable Nanostructures of DNA Probes by DNA Hairpin Template
For the investigation and control of the nanostructures of a DNA probe, CdTe/SiO 2 nanoparticles, as energy donors, were conjugated with a NH 2 -dsDNA hairpin template (hybridization by ssDNA 1 , ssDNA 2 , ssDNA 3 , and ssDNA 4 ). After the release of ssDNA 3 and ssDNA 4 from the conjugation, only one ssDNA 1 and ssDNA 2 remained on the particle surface (Figure 1) because the maximum ssDNA number on CdTe/SiO 2 nanoparticles is three. Therefore, because ssDNA 5 cannot hybridize with ssDNA 2 , the CdTe/SiO 2 -ssDNA 1 -ssDNA 2 conjugations could only hybridize with one ssDNA 5 on Au 40 and form the conjugations of CdTe/SiO 2 -dsDNA-Au 40 , the fluorescent DNA probes, into a 1:1 donor/acceptor (CdTe/SiO 2 :Au 40 ) structure (Figure 6a). If the Au 40 -ssDNA 6 conjugations were added into this preparation system, the fluorescent DNA probes could have a 1:2 donor/acceptor ratio in the nanostructure (Figure 6b) because ssDNA 6 was only complementary to ssDNA 2 . Through the same process, when the HS-DNA hairpin template (hybridization by ssDNA 3 , ssDNA 4 , ssDNA 5 , and ssDNA 6 ) was used, the fluorescent DNA probes in a 2:1 donor/acceptor (CdTe/SiO 2 :Au 40 ) radio could be prepared (Figure 6c).
Due to the dsDNA hybridization of probes, the distance between the energy donor, CdTe/SiO 2 , and energy acceptor, AuNPs, is close enough to result in the FRET effect (quenching effect) and a decrease in the fluorescence intensity of the DNA probes. Though some investigations about nanostructures of DNA probes based on inorganic nanoparticles have been completed [21], research into the relationship between nanostructures and the corresponding quenching effect is limited. In this study, the nanostructures of CdTe/SiO 2 -dsDNA-Au 40 probes were controlled with hairpin scaffolds and the corresponding quenching efficiency (Q e ) was 33.6%, 57.5%, and 74.2% when the donor/acceptor ratios of the probes were 2:1, 1:1, and 1:2, respectively ( Figure 6d).

Detection for Target DNA
The detection of the DNA probe as a biosensor for completely complementary DNA (ssDNA10) as the target was investigated when the DNA probe had a 1:2 donor/acceptor ratio in the nanostructure (Figure 7). The fluorescent intensities of the detection system gradually increased with the increase in ssDNA10 concentration. When the ssDNA10 concentration range was 0.0446 to 2.230 nmol/L, the corresponding linear regression equation was Fr = 0.034 + 0.565c with an R 2 of 0.995, where Fr and c are the fluorescent increase efficiency and ssDNA10 concentration,

Detection for Target DNA
The detection of the DNA probe as a biosensor for completely complementary DNA (ssDNA 10 ) as the target was investigated when the DNA probe had a 1:2 donor/acceptor ratio in the nanostructure (Figure 7). The fluorescent intensities of the detection system gradually increased with the increase in ssDNA 10 concentration. When the ssDNA 10 concentration range was 0.0446 to 2.230 nmol/L, the corresponding linear regression equation was F r = 0.034 + 0.565c with an R 2 of 0.995, where F r and c are the fluorescent increase efficiency and ssDNA 10 concentration, respectively. The detection limit for completely complementary DNA was about 0.106 nmol/L.  (Figures 8b,c), respectively, when the DNA concentration was 1.115 nmol/L. Therefore, this DNA probe can be used to detect whether a single point mutation has occurred on the target DNA. Both the fluorescence of the donors and the hybridized structure of dsDNA in probes could change under different detection conditions [24,28]. Therefore, the stability of this fluorescent DNA probe system was investigated for different buffer conditions and pH values (Figure 9). The results indicated that the FRET recovery efficiency (Fr) of the detection system reached a maximum when the pH value was 8 and the ssDNA10 concentration was 1.115 nmol/L (Figure 9a). Therefore, the optimal pH for detection was about 8.0. The effect of NaCl on Fr in the binding buffer solution indicated that the Fr did not obviously change with different NaCl concentrations (Figure 9b) because the BSPP could improve the stability of inorganic nanoparticle-based DNA probes [24]. The specificity of the DNA probe was investigated when the completely non-complementary DNA (ssDNA 11 ), completely complementary DNA (ssDNA 10 ), and a mismatched base-pair DNA (ssDNA 12 ) were the target DNA ( Figure 8). The results indicated that when the completely non-complementary ssDNA 11 was the target, the fluorescent intensity slightly increased by about 8.6% (Figure 8a); whereas when the completely complementary ssDNA 10 and the mismatched base-pair ssDNA 12 were the targets, the increase in fluorescent intensities were 171.4% and 152.3% (Figure 8b,c), respectively, when the DNA concentration was 1.115 nmol/L. Therefore, this DNA probe can be used to detect whether a single point mutation has occurred on the target DNA. The specificity of the DNA probe was investigated when the completely non-complementary DNA (ssDNA11), completely complementary DNA (ssDNA10), and a mismatched base-pair DNA (ssDNA12) were the target DNA ( Figure 8). The results indicated that when the completely non-complementary ssDNA11 was the target, the fluorescent intensity slightly increased by about 8.6% ( Figure 8a); whereas when the completely complementary ssDNA10 and the mismatched base-pair ssDNA12 were the targets, the increase in fluorescent intensities were 171.4% and 152.3% (Figures 8b,c), respectively, when the DNA concentration was 1.115 nmol/L. Therefore, this DNA probe can be used to detect whether a single point mutation has occurred on the target DNA. Both the fluorescence of the donors and the hybridized structure of dsDNA in probes could change under different detection conditions [24,28]. Therefore, the stability of this fluorescent DNA probe system was investigated for different buffer conditions and pH values (Figure 9). The results indicated that the FRET recovery efficiency (Fr) of the detection system reached a maximum when the pH value was 8 and the ssDNA10 concentration was 1.115 nmol/L (Figure 9a). Therefore, the optimal pH for detection was about 8.0. The effect of NaCl on Fr in the binding buffer solution indicated that the Fr did not obviously change with different NaCl concentrations (Figure 9b) Both the fluorescence of the donors and the hybridized structure of dsDNA in probes could change under different detection conditions [24,28]. Therefore, the stability of this fluorescent DNA probe system was investigated for different buffer conditions and pH values (Figure 9). The results indicated that the FRET recovery efficiency (F r ) of the detection system reached a maximum when the pH value was 8 and the ssDNA 10 concentration was 1.115 nmol/L (Figure 9a). Therefore, the optimal pH for detection was about 8.0. The effect of NaCl on F r in the binding buffer solution indicated that the F r did not obviously change with different NaCl concentrations (Figure 9b) because the BSPP could improve the stability of inorganic nanoparticle-based DNA probes [24].

Conclusions
A DNA hairpin template method was presented in this paper for the preparation of fluorescent DNA probes and to control the nanostructure (energy donor/acceptor ratio) of probes, when CdTe/SiO2 fluorescent nanoparticles were the energy donors and AuNPs were the energy acceptors. A series of DNA probes with different donor/acceptor ratios, 2:1, 1:1, and 1:2, were obtained and the corresponding FRET efficiency of the probes were 33.6%, 57.5%, and 74.2%, respectively. Moreover, this probe showed good sensitivity and specificity under different buffer conditions.