Triplex DNA: A new platform for polymerase chain reaction – based biosensor

Non - specific PCR amplification and DNA contamination usually accompany with PCR process, to overcome these problems, here we establish a sensor for thrombin by sequence - specific recognition of the PCR product with molecular beacon through triplex formation. Probe A and probe B were designed for the sensor, upon addition of thrombin, two probes hybridized to each other and the probe B was extended in the presence of Klenow Fragment polymerase and dNTPs. The PCR amplification occurred with further addition of Taq DNA Polymerase and two primers, the PCR product was recognized by molecular beacon through triplex formation. The fluorescence intensity increased with the logarithm of the concentration of thrombin over the range from 1.0 × 10−12 M to 1.0 × 10−7 M, with a detection limit of 261 fM. Moreover, the effect of DNA contamination and non - specific amplification could be ignored completely in the proposed strategy.

Scientific RepoRts | 5:13010 | DOi: 10.1038/srep13010 To overcome the drawback of PCR -based biosensor and explore new direction for the study of triplex DNA, here we developing a novel biosensor for protein by combining the amplification property of PCR and sequence -specific recognition ability of triplex formation for DNA duplex. As a proof of concept, here a sensitive fluorescent sensor for thrombin is being established with molecular beacon based upon triplex formation and PCR.

Results and Discussion
The scheme of the biosensor is shown in Fig. 1. Oligonucleotide 5′ -AAT ACC CGA TTG CAG TAC GAC TCT CCA CAA GCC TTT TTT TTT TTT TTT TTT TTT TTT TTT TTT TTG GTT GGT GTG GTT GG -3′ (probe A) and Oligonucleotide5′ -CGA GTC CGT GGT AGG GCA GGT TGG GGT GAC TAA AGA GAA GGA AGA GAA GAA GAA AGA AAA GAA AAA GTT TTT TTG GCT TGT G-3′ (probe B) are designed for the sensor, the underlined sequence are aptamers for thrombin, the T 32 in probe A and T 7 in probe B are designed as spacer. Oligonucleotide5′ -AAT ACC CGA TTG CAG TAC GAC TC-3′ (P1) and 5′ -GAG TCC GTG GTA GGG CAG GT-3′ (P2) are designed as primers. Upon addition of thrombin, the red and italic sequences of two probes hybridize to each other and the probe B is extended in the presence of Klenow Fragment polymerase and dNTPs. Then PCR amplification is carried out with further addition of dNTPs, Taq DNA Polymerase and two primers. Green sequence of Probe B is homopurine strand, it is designed to create homopurine•homopyrimidine double strand during the process of PCR amplification, which can be recognized by molecular beacon with the sequence of 5′ - 6-FAM-GTG  GAG TTT CTC TTC CTT CTC TTC TTC TTT CTT TTC TTT TTC CTC CAC-BHQ-1-3′ (MB) through triplex formation [49][50][51][52][53][54] . As shown in Fig. 2A, the fluorescence intensity of PCR product and molecular beacon was only about 60 in the absence of thrombin, it increased to about 155 upon addition of 1.0 × 10 −7 M thrombin. These results confirmed the probability of the proposed strategy.
The numbers of PCR cycle affect the fluorescence intensity of sensor. As shown in Fig. 2A, the fluorescence intensity increased with the numbers of PCR cycle, and reached a maximum at 45 cycles. Thus 45 cycles of PCR reaction cycle was used.
Circular dichroism (CD) spectroscopy was a powerful tool and was usually applied into the study of DNA structure 61 . It was reported that the negative peak of 210 nm was the marker for triplex DNA 62 , thereby CD spectroscopy was used to investigate the structure information of Oligonucleotide5′ -CGA GTC CGT GGT AGG GCA GGT TGG GGT GAC TAA AGA GAA GGA AGA GAA GAA GAA AGA AAA GAA AAA GTT TTT TTG GCT TGT GGA GAG TCG TAC TGC AAT CGG GTA TT-3′ (R1) and Oligonucleotide5′ -AAT ACC CGA TTG CAG TAC GAC TCT CCA CAA GCC AAA AAA ACT TTT TCT TTT CTT TCT TCT TCT CTT CCT TCT CTT TAG TCA CCC CAA CCT GCC CTA CCA CGG ACT CG-3′ (R2), which were two sequences of the PCR product. As shown in Fig. 2B, there were no (or little) negative peak at 210 nm for CD spectroscopy of R1 ·R2 duplex and MB respectively, while there occurred a strong negative peak at 210 nm for the mixture of R1 ·R2 and MB, which indicated that there was triplex formation between R1-R2 (PCR product) and MB. These results was in accordance with that of literature [49][50][51][52][53][54] .
The  the fluorescence intensity), with a detection limit of 261 fM, which was obtained from the equation of DL = 3б/slope. The comparison of the proposed method with others was listed in Table 1, it had merit of wide linear range.
Non -specific PCR amplification usually accompanied with the PCR amplification process. DNA contamination might easily occurred during PCR amplification process as well. Both DNA contamination and non -specific amplification PCR product were double -stranded DNA, which was similar to that of PCR product, so it was difficult to distinguish the signal for non -specific amplification and DNA contamination from that for PCR product, and then they usually affect the detecting result. To investigate the effect of non -specific PCR amplification and DNA contamination on the proposed strategy, the fluorescence intensity of the PCR product were measured with the proposed strategy and the common probe (SYRB -Green) for real -time PCR respectively. Here lambda -DNA was used to act as non -specific amplification product and DNA contamination due to its duplex structure. As shown in Fig. 4, the fluorescence intensity increased with concentration of lambda -DNA by using the probe of SYRB -Green, while the fluorescence intensity had little change when the concentration of lambda -DNA increased from 0 to 1.6 μ g/mL by using the proposed strategy. These results indicated that the proposed strategy could overcome the effect of non -specific PCR amplification and DNA contamination completely, and the effect of non -specific amplification could be ignored in the proposed method. These results mainly due to the stringent sequence -specific recognition between PCR product and MB through triplex formation [49][50][51][52][53][54] .
In order to test the selectivity of the proposed sensor, the effect of other possible interferences were investigated. 1.0 × 10 −7 M of lysozyme, hemoglobin and apo -transferrin human were used to replace the   1.0 × 10 −8 M of thrombin for PCR amplification respectively, As demonstrated in Fig. 5A, each of their fluorescence intensity was almost at the same level of the blank. Moreover, the existence of 10-folded lysozyme, hemoglobin and apo -transferrin human had little effect on the fluorescence intensity of thrombin. To further certificate the selectivity of the sensor, electrophoresis photograph of PCR product was investigated, as demonstrated in Fig. 5B, there were no bands with addition of lysozyme, hemoglobin and apo-transferrin human, while there was the correct size band in the thrombin, which indicated that these proteins had no effect on detection of thrombin. These results indicated that the proposed biosensor had good selectivity.
To assess the analytical application of the sensor, the method was used to detect thrombin in human sera. Since no thrombin was found from the human sera, addition and recovery experiment was performed to estimate the application of the assay in complex sample. As demonstrated in Table 2, 5.0 × 10 −11 -5.0 × 10 −8 M of thrombin was added into each human sera, the recovery ranges from 87.6% to 112.6%, and the relative standard deviation values were in the ranges of 4.3%-9.8%, which indicated that the method had good analytical application in human sera.
In summary, we have developed an ultrasensitive and selective sensor for thrombin by combining the PCR amplification and triplex formation. The fluorescence intensity was proportional to the logarithm of the concentration of thrombin over the range of 1.0 × 10 −12 M-1.0 × 10 −7 M, with a detection limit of 261 fM. The most important contribution is that the proposed strategy can overcome the effect of non -specific amplification and DNA contamination completely, which was realized by detecting PCR

Apparatus. The PCR amplification was performed on a ETC 811 PCR Instrument (Eastwin Life
Sciences, Inc., China). The fluorescence signal was measured on a RF-5301PC spectrofluorimeter (Shimadzu, Japan). The circular dichroism (CD) spectroscopy was obtained from a J-810-150S spectropolarimeter (JASCO International Co. Ltd., Japan).
Recognition of thrombin and preparation of PCR template. Probe A and probe B were denatured at 90 °C for 5 min, then cooled to 0 °C rapidly. 2 μ L of thrombin solution was mixed with 2 μ L 4.0 × 10 −10 M probe A and probe B in Tris-HAc buffer, kept at 37 °C for 30 min. After that, 0.5 U Klenow Fragment (exo-) polymerase and 0.9 μ L 2.5 mM dNTPs were added, and the total mixture volume reached 10 μ L, kept the mixture at 37 °C for 30 min, along with heating to 95 °C to inactivate the Klenow Fragment (exo-) polymerase. Then the inactivated mixture was acted as PCR template for following PCR amplification.
PCR amplification. The PCR amplification was carried out in a 200 μ L PCR tube that containing 5.0 μ L PCR template, 2.0 μ L 10 μ M of two primers, 5 μ L of 10 × PCR Buffer, 1 μ L 5 U/μ L of Taq DNA Polymerase, 3 μ L 25 mM of MgCl 2 , 1 μ L 10 mM of dNTPs and 31 μ L of Nanopure H 2 O. The thermal program was comprised of an initial denaturation at 95 °C for 10 min, 45 cycles of PCR amplification was carried out by using 30 s of denaturation at 94 °C, 30 s of annealing temperature at 60 °C, and 10 s of extending temperature at 72 °C. All the reactions were run in triplicates, and the control experiment was carried out with same reaction mixture in the absence of thrombin.
Measurement of fluorescence spectrum. 10 μ L of 10 μ Μ MB were mixed with 490 μ L PBS buffer solution which contained different amounts of PCR product. After 1.5 hours of incubation(25 °C), the fluorescence signal was measured with spectrofluorimeter. Slit widths were both 5.0 nm, and the excitation and emission wavelengths were set at 495 and 518 nm, respectively.
Measurement of CD spectroscopy. The circular dichroism (CD) spectroscopy was measured at room temperature and performed over the wavelength range from 200 to 300 nm in 0.1 cm path length cuvettes. The result was obtained by averaging 3 scans at the scanning rate of 100 nm per minute with a response time of 1.0 s and the bandwidth of 1.71 nm.