- Email: [email protected]
Real-time study of a DNA strand displacement reaction using dual polarization interferometry
Biosensors and Bioelectronics 41 (2013) 505–510
Contents lists available at SciVerse ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Real-time study of a DNA strand displacement reaction using dual polarization interferometry Pingping Xu a, Fujian Huang a, Haojun Liang a,b,n a b
CAS Key Laboratory of Soft Matter Chemistry, Department of Polymer Science and Engineering, University of Science and Technology of China, Hefei, Anhui 230026, PR China Hefei National Laboratory for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei, Anhui 230026, PR China
a r t i c l e i n f o
abstract
Article history: Received 10 July 2012 Received in revised form 29 August 2012 Accepted 9 September 2012 Available online 15 September 2012
A DNA strand displacement reaction on a solid–liquid interface was investigated using dual polarization interferometry. This effective analytical technique allows the real-time, simultaneous determination of the thickness, density, and mass of a biological layer. The displacement process was examined, and the changes in thickness, density, and mass were determined. Injection of the displacement DNA resulted in an increase in density and a decrease in mass and thickness, which indicated that a portion of the target DNA was displaced from the double-stranded DNA (dsDNA). The effects of the displacement DNA concentration and toehold length on the displacement efficiency were also examined. Increasing the displacement DNA concentration and the toehold length increased the changes in mass and the displacement efficiency. At the concentration of 0.2 mM, the toeholds with 4, 5, 6, and 7 bases had displacement percentages of 24.54%, 25.99%, 30.16%, and 70.41%, respectively. At displacement DNA concentrations exceeding that of the dsDNA, the displacement percentage was not concentration-dependent. Above a certain concentration, the percentage remained stable with increasing concentration. Comparison using different toehold sequences showed that the displacement efficiency increases with increasing bonding force between the base pairs. & 2012 Published by Elsevier B.V.
Keywords: DNA strand displacement Displacement efficiency Toehold Dual polarization interferometry
1. Introduction In the past few decades, DNA has attracted increasing attention as an ideal construction material for self-assembly with static DNA structures and dynamic DNA devices at the nanometer-scale. Important breakthroughs in the use of nucleic acid as nanoscale engineering materials can be attributed to the specificity of interactions as well as predictability of their double-helical structure and Watson–Crick binding thermodynamics. Structural DNA nanotechnology has enabled the construction of triple- and quadruple-branched ‘‘junctions’’ (Kallenbach et al., 1983), as well as two- and three-dimensional DNA crystals. Structural DNA nanotechnology has also allowed for the development of a molecular to a macroscopic material via a ‘‘bottomup’’ approach (Rothemund, 2006; Rothemund et al., 2004; Winfree et al., 1998; Zheng et al., 2009). On the other hand, dynamic DNA nanotechnology is focused on logic gates (Lederman et al., 2006; Stojanovic and Stefanovic, 2003; Win and Smolke, 2008), motors (Dirks and Pierce, 2004; Green et al., 2008; Pei et al., 2006; Yurke
n Corresponding author at: CAS Key Laboratory of Soft Matter Chemistry, Department of Polymer Science and Engineering and Hefei National Laboratory for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei, Anhui 230026, PR China. Tel./fax: 865513607824. E-mail address: [email protected] (H. Liang).
0956-5663/$ - see front matter & 2012 Published by Elsevier B.V. http://dx.doi.org/10.1016/j.bios.2012.09.008
et al., 2000), DNA origami boxes (Andersen et al., 2009), and circuits (Levy and Ellington, 2003; Seelig et al., 2006; Zhang et al., 2007; Zhang and Winfree, 2008). These devices have been constructed by designing kinetic pathways based on DNA hybridization, branch migration, and dissociation processes (Green et al., 2008; Yurke et al., 2000; Zhang et al., 2007). Studies have focused on the kinetics of DNA hybridization (Morrison and Stols, 1993), branch migration (Panyutin and Hsieh, 1993, 1994), and dissociation processes (Green and Tibbetts, 1981), as well as to improve the understanding of the kinetics of nucleic acid reactions (Zhang and Winfree, 2009). Various methods have been employed to elucidate, measure, and monitor the oligonucleotide immobilization as well as hybridization through ellipsometry (Elhadj et al., 2004; Gray et al., 1997), surface plasmon resonance (Green et al., 2000; Striebel et al., 1994; Wolf et al., 2004), neutron reflectivity (Levicky et al., 1998), and quartz crystal microbalance (Okahata et al., 1998). However, most of these methods were studied in solution and lacked dynamic process information regarding DNA strand displacement reactions. Dual polarization interferometry (DPI) is an effective analytical approach for real-time, label-free, quantitative measurement, and has been applied in various research areas, including polyelectrolyte assemblies (Lane et al., 2008; Shovsky et al., 2011), proteins (Armstrong et al., 2004; Biehle et al., 2004; Lin et al., 2006b; Sonesson et al., 2008), antibody–antigen interaction (Lin
506
P. Xu et al. / Biosensors and Bioelectronics 41 (2013) 505–510
et al., 2006a), DNA–small-molecule interaction (Wang et al., 2009), biomimetic membranes (Aulin et al., 2008), and so on. DPI can also be used to monitor the cationic polymer-mediated DNA immobilization (Zhao et al., 2008) and hybridization of oligonucleotides on a silanized surface (Berney and Oliver, 2005; Lillis et al., 2006) in real time. All these studies indicate the potential and advantage of DPI for studying DNA structural changes in real time. The present work aims to utilize DPI to monitor a DNA strand displacement reaction through toehold initiation, branch migration, and dissociation processes. The effects of the different DNA concentrations were examined and the toehold modulated functions on the reaction process were discussed. We show the possibility of completing the DNA strand displacement reaction on an amino surface as well as of supplying real-time changes in the layer thickness, density, and mass. The accurate displacement percentage can be obtained by comparing the variations in mass. The kinetic process under different DNA concentrations and different toeholds was also elucidated.
2. Methods and materials 2.1. DPI analysis All experiments were monitored in real time using the AnaLight Bio200 DPI system (Farfield Group Ltd., Crewe, UK). The instrument consists of a helium–neon laser emitting light at 632.8 nm, a controller to select plane polarized light, transverse magnetic (TM) mode, and a transverse electric (TE) mode, a sensor constructed with two optical waveguides stacked on top of each other, and an array photodiode. When polarized light (TM) is introduced to the end of the stack, the well-known pattern of Young’s interference fringes is formed on an array photodiode. The fringe positions represent the relative phase positions of both sensing and reference waveguides. When the sensing waveguide surface is exposed to a biomolecule that adsorbs onto the surface or interacts with molecules already immobilized on the surface, the evanescent field is affected. The effective penetration depth of the evanescent field is approximately 100 nm. This results in a change in the effective index of the sensing waveguide and the phase changes in the sensing light output. The effective index of the reference waveguide does not change. The phase changes can be directly measured by monitoring the relative phase position of the fringe pattern through a Fourier transformation. A second polarization of light (TE) is introduced at right angles to the first into the stack. This responds differently to biomolecule adsorption or desorption, and provides a second independent measurement. For classical optical theory, it is possible to interpret the two measurements in terms of thickness and refractive index of the adsorbed layer. However, this technique allows separate measurements of both thickness and refractive index of the adsorbed biomolecule layer. The mass of the adsorbed layer can then be calculated. A schematic of the DPI system was shown in Supplementary material, Fig. S1. 2.2. DNA sequences All oligonucleotides used in the DPI analysis are outlined in Supplementary material, Table S1. The annealing process of probe DNA and target DNA was conducted using a PCR instrument (TC5000, TECHNE, UK). The course took 90 min to lower the temperature of the samples from 95 1C to 20 1C at a constant rate of 0.02 1C s 1. Displacement DNA that was fully complementary to the target DNA was employed to displace the target DNA through a branch migration reaction.
2.3. Double-stranded DNA (dsDNA) immobilization procedure All experiments were performed at 20 70.002 1C using a temperature controller built into the AnaLight Bio200 instrument. An amino-modified sensor chip, FB 100 Amine (Farfield Sensors) was mounted in the instrument. The sensor chip and the running buffer were calibrated as follows. Phosphate-buffered (PBS, pH 7.4) passed through the sensor chip surface at a flow rate of 50 mL min 1 until the baseline was stable. An ethanol/water mixture (80:20 w/w) and ultrapure water with known RIs were sequentially poured over the two channels. When the baseline restabilized, TE/Mg2 þ buffer (10 mM, pH 8.0) was injected for 2 min at 50 mL min 1. The data obtained were analyzed using the software of the AnaLight Bio200 instrument for all subsequent calculations. After calibration, PBS was introduced to both channels at a flow rate of 50 mL min 1 for 5 min to ensure baseline stabilization. Then, the flow rate was changed to 25 mL min 1. Glutaraldehyde solution (4 mg mL 1 in PBS) and streptavidin (SA) solution (0.5 mg mL 1 in PBS) were then injected for 6 min. Before the various DNA injections were performed, the flow was stopped, and the buffer was changed from PBS to TE/Mg2 þ buffer (10 mM, pH 8.0). The buffer flowed through both channels for 15 min to ensure baseline stabilization. The flow rate was then reduced to 15 mL min 1 for 2 min. DNA solution at 0.5 mM was prepared in TE/Mg2 þ buffer for subsequent experiments. The DNA solution was then injected into the surface at a rate of 15 mL min 1 for 10 min before returning to TE/Mg2 þ buffer. The immobilization process was illustrated in Supplementary material, Fig. S2. 2.4. Single-stranded DNA (ssDNA) displacement procedure After the dsDNA was immobilized and stabilized on the chip surface, the flow rate was reduced to 10 mL min 1. ssDNA solutions with different concentrations (1, 0.5, and 0.2 mM) prepared in TE/Mg2 þ buffer were passed through both channels for 15 min.
3. Results and discussion DPI was used to monitor a DNA strand displacement reaction on an amino surface, and to examine the effect of toehold length and sequence on the displacement efficiency. The thickness and refractive index (RI) data were directly resolved from the measured TM and TE phase values. The adsorbed mass was calculated based on Eq. (1). mL ¼ tL
rDNA ðnL nB Þ ðnDNA nB Þ
ð1Þ
where mL is the layer mass per unit area (ng mm 2), tL is the layer thickness (nm), rL is the layer density (g cm 3), rDNA is the DNA density, nL is the refractive index of the layer resolved from the phase data, nB is the refractive index of the bulk electrolyte solution calculated from the calibration, and nDNA is the refractive index of DNA. To immobilize the probe molecules onto an amino-functionalized sensor chip, glutaraldehyde was used as a linker via a condensation reaction between amine and aldehyde groups. After the SA layer formation, 50 -biotinylated dsDNA can be successfully immobilized on the surface through the specific interaction between SA and biotin. 3.1. DNA immobilization and displacement To investigate DNA adsorption and subsequent displacement processes, 50 -biotinylated dsDNA with a toehold of 6 bases was adsorbed onto an amine-functionalized chip surface. DPI was
P. Xu et al. / Biosensors and Bioelectronics 41 (2013) 505–510
employed to monitor the process in real time. A complementary 26-mer oligonucleotide single strand was introduced to the sensor chip. The spectral changes in thickness, density, and mass in the process were shown in Fig. 1. As can be seen, the thickness of the immobilized dsDNA layer increased by 4.14 nm. Theoretically, the distance between the base pairs is 0.34 nm and each 10 base pairs form a revolution. Thus, the dsDNA is 3.4 nm per 10 base pairs and 2 nm in width (Peterlinz et al., 1997). The thickness of a 20 base pair DNA is expected to be 6.8 nm if the immobilized DNA duplex is fully extended. Hence, the tilt angle of the immobilized dsDNA can be calculated as 531, which is slightly higher than the 461 previously reported (Berney and Oliver, 2005). The toeholds suspended at both ends of the dsDNA may explain this condition. ssDNA is more flexible and tends to absorb with the long axis parallel, rather than perpendicular, to the surface (Walker and Grant, 1996). The DNA probe adsorbed onto an amino chip had a measured thickness of 0.7 nm (Berney and Oliver, 2005), which was much smaller than that of dsDNA. Therefore, the ssDNA on both ends of the dsDNA slightly contribute to the increase in thickness. The density decreased accordingly with increasing dsDNA layer thickness. Compared with the SA layer, the immobilized dsDNA layer was likely relatively loose because of the rigidity of dsDNA. Upon returning to running buffer, the thickness and mass of the immobilized dsDNA layer slightly decreased, indicating the powerful combination between SA and biotin. During the process, a slight probe loss and rearrangement of the adsorbed layer occurred on the chip surface. Upon the addition of the displacement DNA (Fig. 1: 830 s), the mass decreased by 0.4 ng mm 2, which indicated that the target DNA detached from the sensing surface. The measured thickness value of the final DNA layer was approximately 2 nm, which is much thicker than the ssDNA probe adsorbed onto an amino chip (Berney and Oliver, 2005). This phenomenon indicated that only a part of the dsDNA was replaced and the final DNA layer contained both dsDNA and ssDNA. The displacement DNA can completely combine with the target DNA by toehold initiation, branch migration, and dissociation. The DNA molecules remaining on the chip were mostly 50 biotinylated ssDNA, and the changes in density can further demonstrate the result. Upon injection of the displacement DNA, the density drastically increased because the ssDNA is more random and flexible than dsDNA. Most of the reactions were completed within 180 s and then almost remained unchanged, demonstrating that the toehold-initiated reaction was very rapid. These results showed that the DNA strand displacement reaction occurred when the displacement DNA was added.
3.2. Effect of toehold length The influence of displacement DNA on the displacement reaction was investigated. The effect of different toehold lengths on the displacement efficiency was examined by comparing the response in the change in the DNA layer mass with different lengths of 4, 5, 6, and 7 bases (Fig. 2). Upon addition of the displacement DNA to the sensor chip, a plateau occurred before the displacement reaction took place. It was likely that the micro-fluid chamber was full of TE/Mg2 þ buffer before the injection of the ssDNA solution and a few seconds were needed for the solution to flow through the entire chip surface. At a displacement DNA concentration of 1 mM, which was twice as much as the dsDNA concentration, the displacement process rate was very rapid. At low concentrations (0.5 and 0.2 mM), the reaction was relatively slow. The lower the concentration of the displacement DNA, the lesser was the hybridization with the target DNA. With increasing concentration, both the reaction rate and the mass that dissociated from the dsDNA increased. At a toehold length of 4 bases, the change in mass was relatively moderate compared with those with longer toeholds. Approximately 1000 s was needed to complete the displacement reaction even at high concentrations. Given that the combination of the 4 base pairs was unstable, the initiation of the replacement reaction through a toehold was not that rapid and forceful. With 5, 6, and 7 bases in the toehold, the reaction rate clearly increased. For the toehold with 6 bases, most target DNA molecules were replaced within 180 s at 1 mM and 220 s at 0.5 mM. For the toehold with 5 bases, the corresponding displacement process was completed within 245 s and 310 s, which is slower than the DNA reaction with 6 bases in the toehold. Therefore, the binding of the 6 base pairs is more stable. For the toehold with 7 bases, most target DNA was replaced within 220 s at 1 mM and 300 s at 0.5 mM. The reaction was still slower than with 6 bases in the toehold, but the change in mass was more obvious. Although the binding of 7 base pairs was more stable, longer chains exhibited increasingly complex conformations and required more time to maintain a steady flow on the chip surface. Based on the changes in mass of the injected dsDNA and displacement DNA, the efficiency of the DNA strand displacement reaction and the mass of the target DNA that disassociated from the immobilized dsDNA can be calculated. Adding dsDNA to the sensor surface resulted in the immobilization of PCRprehybridized probe DNA and target DNA on the chip surface through SA-biotin interaction. Therefore, the increase in mass upon dsDNA injection reflected the total quality of the probe DNA and target DNA adsorbed on the surface. The addition of displacement DNA resulted in a combination with the target DNA through the toehold, which expended on the 50 end of the target DNA strand. Therefore, the decreased mass upon injection of displacement DNA injection reflected the quality of the target DNA. To obtain the replacement percentage, the percentage of target DNA in the dsDNA was calculated based on the molecular weight. The molecular weight percentage was calculated using the following equation: MðtargetÞ MðprobeÞ þMðtargetÞ
Fig. 1. Measurement of the thickness (’), mass (), and density (m) values on a sensor chip upon (a) injection of dsDNA (0.5 mM; toehold: 6 bases) and (b) injection of displacement DNA (1 mM).
507
ð2Þ
where M(probe) is the molecular weight of probe DNA and M(target) is the molecular weight of the target DNA. The mass percentages can be obtained through the mass changes in mass upon injection of the dsDNA and the displacement DNA. The displacement percentages of all the other reactions with different toeholds were calculated based on Eq. (2) and the changes in mass. The results were shown in Fig. 3.
508
P. Xu et al. / Biosensors and Bioelectronics 41 (2013) 505–510
Fig. 2. Measured changes in mass of a sensor chip modified with dsDNA upon addition of the displacement DNA with the following toehold lengths: 4 (A), 5 (B), 6 (C) and 7 (D) bases.
varied widely as the concentration decreased lower than 0.5 mM (the dsDNA concentration immobilized on the chip surface). At concentrations higher than 0.5 mM, the percentage did not increase as much with the concentration, which suggests that the displacement percentage is not concentration-dependent. The slight increase in percentage also indicated that the displacement percentage remained almost unchanged above a critical concentration. At high concentrations, the conformation of the DNA molecules in solution was folded and wound. When the displacement DNA solution flowed through the chip surface, the time to get them extended was insufficient. For the toehold with 7 bases, the critical concentration was 0.5 mM, suggesting that the critical concentration lowered with longer toehold length. Fig. 3. Displacement percentages at the four different toehold lengths.
3.3. Effect of toehold sequence At 1, 0.5, and 0.2 mM, the displacement percentages increased with increasing toehold length. At 0.2 mM, the toeholds with 4, 5, and 6 bases, had displacement percentages of 24.54%, 25.99%, and 30.16% respectively, which demonstrates that during the initiation stage, the toehold length greatly influenced the branch migration and displacement reaction. However, for the DNA with 7 bases in the toehold, the percentage reached 70.41%, which is much higher than with 6 bases. In this case, the binding force is much stronger than with shorter toeholds and can replace most of the target DNA molecules, even at lower concentrations. At 1 mM, the displacement percentages using toeholds with 4, 5, 6 and 7 bases were 57.57%, 63.63%, 75.4%, and 83.45%, respectively. The DNA molecules were mostly replaced rather than entirely, even when the displacement DNA concentration was higher than the dsDNA concentration and the toehold combination was powerful. The DNA molecules cannot be entirely replaced. The conformation was probably limited because the DNA strands were immobilized on a surface. For the four toehold lengths, the displacement percentage
The sequences of the toehold with 6 bases in target DNA and displacement DNA were 50 -TGGAGA-30 and 50 -TCTCCA-30 , respectively. The sequences consisted of a mixture of G/C/T/A bases and contain equal numbers of G/C as well as A/T base pairs (three C/G bases and three A/T bases). Adenine (A) pairs with thymine (T) through two hydrogen bonds, whereas three hydrogen bonds form between guanine (G) and cytosine (C), implying that the G/C combination is stronger than the A/T combination. To study the effect of the toehold strength on the displacement reaction, ‘‘weak’’ (all six bases were A and T) and ‘‘strong’’ (all six bases were C and G) toeholds were introduced. Based on a similar immobilization method, the changes in mass upon injection of the displacement DNA were measured (Fig. 4). The reaction rate was relatively slow with the weak toehold (Fig. 4A). Most displacement reactions were accomplished within 400 s at 1 mM. At lower concentrations, the displacement reactions took longer to plateau. With strong toeholds (Fig. 4B), the mass changes
P. Xu et al. / Biosensors and Bioelectronics 41 (2013) 505–510
509
Fig. 4. Measured changes in mass of a sensor chip modified with dsDNA upon the addition of displacement DNA with different toehold sequences: weak (A) and strong (B) toeholds.
combination process. At 1 mM, the displacement percentage was 98.4%, indicating that the target DNA was almost completely replaced and further demonstrating the strong G–C interaction. For the three DNA strands, the displacement percentages increased with increasing concentrations. At the same concentration, the displacement DNA with the strong toehold dissociated more target DNA molecules from the dsDNA. The changes in mass were minimal for the weak toehold. This finding indicates that the strength of the toehold has an important function in the strand displacement reaction.
4. Conclusions Fig. 5. Displacement percentages of three different toehold sequences with different base pair strengths.
at all concentrations were slightly different compared with the medium-strength (AT/CG; Fig. 2C) and the weak toeholds. At 0.5 mM, the mass curve was sigmoidal with two slow stages and a slightly fast stage. The strong toehold initiated the strand displacement reaction more quickly. However, given that it contains more C/G bases, more time was needed to stabilize the conformation when the displacement DNA reached the chip surface. At 0.2 mM, the reaction was slow during the first 500 s and only a few DNA molecules were replaced. After 500 s, the slope changed and the reaction during this period accelerated. The change in mass was almost the same at 0.5 mM, which indicates that the strong binding force between C and G plays an important function in the DNA strand displacement reaction. The mass of the target DNA that disassociated from the dsDNA and the efficiency of the DNA strand displacement reaction were also calculated. The displacement percentages of the three different toehold sequences are shown in Fig. 5. For the weak toehold at 0.2 mM, the displacement percentage was 26%, which was slightly lower than 30.16% because of the weak A–T interaction. When the concentration of the displacement DNA was lower than that of the dsDNA, the percentage increased sharply with increasing concentration. By contrast, the percentage change was relatively moderate, indicating that the percentage did not significantly depend on the concentration. Above a critical concentration, the percentage remained stable with increasing concentration. The condition was slightly different for the strong toehold. At 0.2 mM, the percentage reached 59.9%, which was higher than the weak toehold (26%) and the medium-strength toehold (30.16%). On one hand, this result may be explained by the strong and stable interaction between G and C. With six G–C base pairs in the DNA strand, some conformational adjustments occurred during the
This paper explored the DNA strand displacement reaction through toehold initiation and branch migration on an aminefunctionalized sensor surface. The effects of different toehold lengths and different toehold sequences on the displacement efficiency were also investigated using DPI. This method is useful because the instrument allows the real-time detection of layer thickness and refractive index, from which the density and mass can be calculated. The displacement efficiency can be obtained based on changes in mass upon the addition of different DNA strands onto the sensor surface. Through specific interaction, biotinylated dsDNA was immobilized onto the SA-covered surface of the sensor chip. Upon the addition of the displacement DNA, the mass and the layer thickness decreased, indicating the replacement of several DNA molecules, as confirmed by the increased density. The effect of the toehold length on the displacement efficiency was determined by comparing toeholds with 4, 5, 6, and 7 bases. The mass changes increased with increasing toehold length. When the concentration of the target DNA exceeded that of the dsDNA, the displacement percentage was not concentrationdependent. Above a critical concentration, the displacement percentage remained unchanged. For the toehold with 7 bases, the critical concentration was 0.5 mM. DNAs with weak, medium-strength, and strong toeholds were compared. Considering the reaction was initiated through a toehold, the strength of the toehold combination was highly significant. At 0.5 mM, the displacement percentage of the medium-strength toehold was 65.09% compared with 54.4% for the weak toehold and 74.88% for the strong toeholds. This phenomenon was mainly because the binding affinity between C and G is much stronger than between A and T. All the results show that the displacement efficiency increased with increasing base pair binding affinity.
510
P. Xu et al. / Biosensors and Bioelectronics 41 (2013) 505–510
Acknowledgments This work was supported by the National Natural Science Foundation of China (Grant nos. 20934004 and 91127046); the National Basic Research Program of China (NBR-PC) (Grant nos. 2012CB821500 and 2010CB934500).
Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2012.09. 008.
References Andersen, E.S., Dong, M., Nielsen, M.M., Jahn, K., Subramani, R., Mamdouh, W., Golas, M.M., Sander, B., Stark, H., Oliveira, C.L., Pedersen, J.S., Birkedal, V., Besenbacher, F., Gothelf, K.V., Kjems, J., 2009. Nature 459 (7243), 73–76. Armstrong, J., Salacinski, H.J., Mu, Q., Seifalian, A.M., Peel, L., Freeman, N., Holt, C.M., Lu, J.R., 2004. Journal of Physics: Condensed Matter 16 (26), S2483–S2491. Aulin, C., Varga, I., Claessont, P.M., Wagberg, L., Lindstrom, T., 2008. Langmuir 24 (6), 2509–2518. Berney, H., Oliver, K., 2005. Biosensors and Bioelectronics 21 (4), 618–626. Biehle, S.J., Carrozzella, J., Shukla, R., Popplewell, J., Swann, M., Freeman, N., Clark, J.F., 2004. Biochimica et Biophysica Acta 1689 (3), 244–251. Dirks, R.M., Pierce, N.A., 2004. Proceedings of the National Academy of Sciences of the United States of America 101 (43), 15275–15278. Elhadj, S., Singh, G., Saraf, R.F., 2004. Langmuir 20 (13), 5539–5543. Gray, D.E., CaseGreen, S.C., Fell, T.S., Dobson, P.J., Southern, E.M., 1997. Langmuir 13 (10), 2833–2842. Green, C., Tibbetts, C., 1981. Nucleic Acids Research 9 (8), 1905–1918. Green, R.J., Frazier, R.A., Shakesheff, K.M., Davies, M.C., Roberts, C.J., Tendler, S.J.B., 2000. Biomaterials 21 (18), 1823–1835. Green, S., Bath, J., Turberfield, A., 2008. Physical Review Letters 101 (23), 1–73. Kallenbach, N.R., Ma, R.I., Seeman, N.C., 1983. Nature 305 (5937), 829–831. Lane, T.J., Fletcher, W.R., Gormally, M.V., Johal, M.S., 2008. Langmuir 24 (19), 10633–10636. Lederman, H., Macdonald, J., Stefanovic, D., Stojanovic, M.N., 2006. Biochemistry 45 (4), 1194–1199. Levicky, R., Herne, T.M., Tarlov, M.J., Satija, S.K., 1998. Journal of the American Chemical Society 120 (38), 9787–9792.
Levy, M., Ellington, A.D., 2003. Proceedings of the National Academy of Sciences of the United States of America 100 (11), 6416–6421. Lillis, B., Manning, M., Berney, H., Hurley, E., Mathewson, A., Sheehan, M.M., 2006. Biosensors and Bioelectronics 21 (8), 1459–1467. Lin, S., Lee, C.K., Lin, Y.H., Lee, S.Y., Sheu, B.C., Tsai, J.C., Hsu, S.M., 2006a. Biosensors and Bioelectronics 22 (5), 715–721. Lin, S., Lee, C.K., Wang, Y.M., Huang, L.S., Lin, Y.H., Lee, S.Y., Sheu, B.C., Hsu, S.M., 2006b. Biosensors and Bioelectronics 22 (2), 323–327. Morrison, L.E., Stols, L.M., 1993. Biochemistry 32 (12), 3095–3104. Okahata, Y., Kawase, M., Niikura, K., Ohtake, F., Furusawa, H., Ebara, Y., 1998. Analytical Chemistry 70 (7), 1288–1296. Panyutin, I.G., Hsieh, P., 1993. Journal of Molecular Biology 230 (2), 413–424. Panyutin, I.G., Hsieh, P., 1994. Proceedings of the National Academy of Sciences of the United States of America 91 (6), 2021. Pei, R., Taylor, S.K., Stefanovic, D., Rudchenko, S., Mitchell, T.E., Stojanovic, M.N., 2006. Journal of the American Chemical Society 128 (39), 12693–12699. Peterlinz, K.A., Georgiadis, R.M., Herne, T.M., Tarlov, M.J., 1997. Journal of the American Chemical Society 119 (14), 3401–3402. Rothemund, P.W., 2006. Nature 440 (7082), 297–302. Rothemund, P.W., Papadakis, N., Winfree, E., 2004. PLoS Biology 2 (12), e424. Seelig, G., Yurke, B., Winfree, E., 2006. Journal of the American Chemical Society 128 (37), 12211–12220. Shovsky, A., Bijelic, G., Varga, I., Makuska, R., Claesson, P.M., 2011. Langmuir 27 (3), 1044–1050. Sonesson, A.W., Callisen, T.H., Brismar, H., Elofsson, U.M., 2008. Colloids and Surfaces B: Biointerfaces 61 (2), 208–215. Stojanovic, M.N., Stefanovic, D., 2003. Nature Biotechnology 21 (9), 1069–1074. Striebel, C., Brecht, A., Gauglitz, G., 1994. Biosensors and Bioelectronics 9 (2), 139–146. Walker, H.W., Grant, S.B., 1996. Langmuir 12 (13), 3151–3156. Wang, J., Xu, X., Zhang, Z., Yang, F., Yang, X., 2009. Analytical Chemistry 81 (12), 4914–4921. Win, M.N., Smolke, C.D., 2008. Science 322 (5900), 456–460. Winfree, E., Liu, F.R., Wenzler, L.A., Seeman, N.C., 1998. Nature 394 (6693), 539–544. Wolf, L.K., Gao, Y., Georgiadis, R.M., 2004. Langmuir 20 (8), 3357–3361. Yurke, B., Turberfield, A.J., Mills, A.P., Simmel, F.C., Neumann, J.L., 2000. Nature 406 (6796), 605–608. Zhang, D.Y., Turberfield, A.J., Yurke, B., Winfree, E., 2007. Science 318 (5853), 1121–1125. Zhang, D.Y., Winfree, E., 2008. Journal of the American Chemical Society 130 (42), 13921–13926. Zhang, D.Y., Winfree, E., 2009. Journal of the American Chemical Society 131 (47), 17303–17314. Zhao, X., Pan, F., Coffey, P., Lu, J.R., 2008. Langmuir 24 (23), 13556–13564. Zheng, J., Birktoft, J.J., Chen, Y., Wang, T., Sha, R., Constantinou, P.E., Ginell, S.L., Mao, C., Seeman, N.C., 2009. Nature 461 (7260), 74–77.
Contents lists available at SciVerse ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Real-time study of a DNA strand displacement reaction using dual polarization interferometry Pingping Xu a, Fujian Huang a, Haojun Liang a,b,n a b
CAS Key Laboratory of Soft Matter Chemistry, Department of Polymer Science and Engineering, University of Science and Technology of China, Hefei, Anhui 230026, PR China Hefei National Laboratory for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei, Anhui 230026, PR China
a r t i c l e i n f o
abstract
Article history: Received 10 July 2012 Received in revised form 29 August 2012 Accepted 9 September 2012 Available online 15 September 2012
A DNA strand displacement reaction on a solid–liquid interface was investigated using dual polarization interferometry. This effective analytical technique allows the real-time, simultaneous determination of the thickness, density, and mass of a biological layer. The displacement process was examined, and the changes in thickness, density, and mass were determined. Injection of the displacement DNA resulted in an increase in density and a decrease in mass and thickness, which indicated that a portion of the target DNA was displaced from the double-stranded DNA (dsDNA). The effects of the displacement DNA concentration and toehold length on the displacement efficiency were also examined. Increasing the displacement DNA concentration and the toehold length increased the changes in mass and the displacement efficiency. At the concentration of 0.2 mM, the toeholds with 4, 5, 6, and 7 bases had displacement percentages of 24.54%, 25.99%, 30.16%, and 70.41%, respectively. At displacement DNA concentrations exceeding that of the dsDNA, the displacement percentage was not concentration-dependent. Above a certain concentration, the percentage remained stable with increasing concentration. Comparison using different toehold sequences showed that the displacement efficiency increases with increasing bonding force between the base pairs. & 2012 Published by Elsevier B.V.
Keywords: DNA strand displacement Displacement efficiency Toehold Dual polarization interferometry
1. Introduction In the past few decades, DNA has attracted increasing attention as an ideal construction material for self-assembly with static DNA structures and dynamic DNA devices at the nanometer-scale. Important breakthroughs in the use of nucleic acid as nanoscale engineering materials can be attributed to the specificity of interactions as well as predictability of their double-helical structure and Watson–Crick binding thermodynamics. Structural DNA nanotechnology has enabled the construction of triple- and quadruple-branched ‘‘junctions’’ (Kallenbach et al., 1983), as well as two- and three-dimensional DNA crystals. Structural DNA nanotechnology has also allowed for the development of a molecular to a macroscopic material via a ‘‘bottomup’’ approach (Rothemund, 2006; Rothemund et al., 2004; Winfree et al., 1998; Zheng et al., 2009). On the other hand, dynamic DNA nanotechnology is focused on logic gates (Lederman et al., 2006; Stojanovic and Stefanovic, 2003; Win and Smolke, 2008), motors (Dirks and Pierce, 2004; Green et al., 2008; Pei et al., 2006; Yurke
n Corresponding author at: CAS Key Laboratory of Soft Matter Chemistry, Department of Polymer Science and Engineering and Hefei National Laboratory for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei, Anhui 230026, PR China. Tel./fax: 865513607824. E-mail address: [email protected] (H. Liang).
0956-5663/$ - see front matter & 2012 Published by Elsevier B.V. http://dx.doi.org/10.1016/j.bios.2012.09.008
et al., 2000), DNA origami boxes (Andersen et al., 2009), and circuits (Levy and Ellington, 2003; Seelig et al., 2006; Zhang et al., 2007; Zhang and Winfree, 2008). These devices have been constructed by designing kinetic pathways based on DNA hybridization, branch migration, and dissociation processes (Green et al., 2008; Yurke et al., 2000; Zhang et al., 2007). Studies have focused on the kinetics of DNA hybridization (Morrison and Stols, 1993), branch migration (Panyutin and Hsieh, 1993, 1994), and dissociation processes (Green and Tibbetts, 1981), as well as to improve the understanding of the kinetics of nucleic acid reactions (Zhang and Winfree, 2009). Various methods have been employed to elucidate, measure, and monitor the oligonucleotide immobilization as well as hybridization through ellipsometry (Elhadj et al., 2004; Gray et al., 1997), surface plasmon resonance (Green et al., 2000; Striebel et al., 1994; Wolf et al., 2004), neutron reflectivity (Levicky et al., 1998), and quartz crystal microbalance (Okahata et al., 1998). However, most of these methods were studied in solution and lacked dynamic process information regarding DNA strand displacement reactions. Dual polarization interferometry (DPI) is an effective analytical approach for real-time, label-free, quantitative measurement, and has been applied in various research areas, including polyelectrolyte assemblies (Lane et al., 2008; Shovsky et al., 2011), proteins (Armstrong et al., 2004; Biehle et al., 2004; Lin et al., 2006b; Sonesson et al., 2008), antibody–antigen interaction (Lin
506
P. Xu et al. / Biosensors and Bioelectronics 41 (2013) 505–510
et al., 2006a), DNA–small-molecule interaction (Wang et al., 2009), biomimetic membranes (Aulin et al., 2008), and so on. DPI can also be used to monitor the cationic polymer-mediated DNA immobilization (Zhao et al., 2008) and hybridization of oligonucleotides on a silanized surface (Berney and Oliver, 2005; Lillis et al., 2006) in real time. All these studies indicate the potential and advantage of DPI for studying DNA structural changes in real time. The present work aims to utilize DPI to monitor a DNA strand displacement reaction through toehold initiation, branch migration, and dissociation processes. The effects of the different DNA concentrations were examined and the toehold modulated functions on the reaction process were discussed. We show the possibility of completing the DNA strand displacement reaction on an amino surface as well as of supplying real-time changes in the layer thickness, density, and mass. The accurate displacement percentage can be obtained by comparing the variations in mass. The kinetic process under different DNA concentrations and different toeholds was also elucidated.
2. Methods and materials 2.1. DPI analysis All experiments were monitored in real time using the AnaLight Bio200 DPI system (Farfield Group Ltd., Crewe, UK). The instrument consists of a helium–neon laser emitting light at 632.8 nm, a controller to select plane polarized light, transverse magnetic (TM) mode, and a transverse electric (TE) mode, a sensor constructed with two optical waveguides stacked on top of each other, and an array photodiode. When polarized light (TM) is introduced to the end of the stack, the well-known pattern of Young’s interference fringes is formed on an array photodiode. The fringe positions represent the relative phase positions of both sensing and reference waveguides. When the sensing waveguide surface is exposed to a biomolecule that adsorbs onto the surface or interacts with molecules already immobilized on the surface, the evanescent field is affected. The effective penetration depth of the evanescent field is approximately 100 nm. This results in a change in the effective index of the sensing waveguide and the phase changes in the sensing light output. The effective index of the reference waveguide does not change. The phase changes can be directly measured by monitoring the relative phase position of the fringe pattern through a Fourier transformation. A second polarization of light (TE) is introduced at right angles to the first into the stack. This responds differently to biomolecule adsorption or desorption, and provides a second independent measurement. For classical optical theory, it is possible to interpret the two measurements in terms of thickness and refractive index of the adsorbed layer. However, this technique allows separate measurements of both thickness and refractive index of the adsorbed biomolecule layer. The mass of the adsorbed layer can then be calculated. A schematic of the DPI system was shown in Supplementary material, Fig. S1. 2.2. DNA sequences All oligonucleotides used in the DPI analysis are outlined in Supplementary material, Table S1. The annealing process of probe DNA and target DNA was conducted using a PCR instrument (TC5000, TECHNE, UK). The course took 90 min to lower the temperature of the samples from 95 1C to 20 1C at a constant rate of 0.02 1C s 1. Displacement DNA that was fully complementary to the target DNA was employed to displace the target DNA through a branch migration reaction.
2.3. Double-stranded DNA (dsDNA) immobilization procedure All experiments were performed at 20 70.002 1C using a temperature controller built into the AnaLight Bio200 instrument. An amino-modified sensor chip, FB 100 Amine (Farfield Sensors) was mounted in the instrument. The sensor chip and the running buffer were calibrated as follows. Phosphate-buffered (PBS, pH 7.4) passed through the sensor chip surface at a flow rate of 50 mL min 1 until the baseline was stable. An ethanol/water mixture (80:20 w/w) and ultrapure water with known RIs were sequentially poured over the two channels. When the baseline restabilized, TE/Mg2 þ buffer (10 mM, pH 8.0) was injected for 2 min at 50 mL min 1. The data obtained were analyzed using the software of the AnaLight Bio200 instrument for all subsequent calculations. After calibration, PBS was introduced to both channels at a flow rate of 50 mL min 1 for 5 min to ensure baseline stabilization. Then, the flow rate was changed to 25 mL min 1. Glutaraldehyde solution (4 mg mL 1 in PBS) and streptavidin (SA) solution (0.5 mg mL 1 in PBS) were then injected for 6 min. Before the various DNA injections were performed, the flow was stopped, and the buffer was changed from PBS to TE/Mg2 þ buffer (10 mM, pH 8.0). The buffer flowed through both channels for 15 min to ensure baseline stabilization. The flow rate was then reduced to 15 mL min 1 for 2 min. DNA solution at 0.5 mM was prepared in TE/Mg2 þ buffer for subsequent experiments. The DNA solution was then injected into the surface at a rate of 15 mL min 1 for 10 min before returning to TE/Mg2 þ buffer. The immobilization process was illustrated in Supplementary material, Fig. S2. 2.4. Single-stranded DNA (ssDNA) displacement procedure After the dsDNA was immobilized and stabilized on the chip surface, the flow rate was reduced to 10 mL min 1. ssDNA solutions with different concentrations (1, 0.5, and 0.2 mM) prepared in TE/Mg2 þ buffer were passed through both channels for 15 min.
3. Results and discussion DPI was used to monitor a DNA strand displacement reaction on an amino surface, and to examine the effect of toehold length and sequence on the displacement efficiency. The thickness and refractive index (RI) data were directly resolved from the measured TM and TE phase values. The adsorbed mass was calculated based on Eq. (1). mL ¼ tL
rDNA ðnL nB Þ ðnDNA nB Þ
ð1Þ
where mL is the layer mass per unit area (ng mm 2), tL is the layer thickness (nm), rL is the layer density (g cm 3), rDNA is the DNA density, nL is the refractive index of the layer resolved from the phase data, nB is the refractive index of the bulk electrolyte solution calculated from the calibration, and nDNA is the refractive index of DNA. To immobilize the probe molecules onto an amino-functionalized sensor chip, glutaraldehyde was used as a linker via a condensation reaction between amine and aldehyde groups. After the SA layer formation, 50 -biotinylated dsDNA can be successfully immobilized on the surface through the specific interaction between SA and biotin. 3.1. DNA immobilization and displacement To investigate DNA adsorption and subsequent displacement processes, 50 -biotinylated dsDNA with a toehold of 6 bases was adsorbed onto an amine-functionalized chip surface. DPI was
P. Xu et al. / Biosensors and Bioelectronics 41 (2013) 505–510
employed to monitor the process in real time. A complementary 26-mer oligonucleotide single strand was introduced to the sensor chip. The spectral changes in thickness, density, and mass in the process were shown in Fig. 1. As can be seen, the thickness of the immobilized dsDNA layer increased by 4.14 nm. Theoretically, the distance between the base pairs is 0.34 nm and each 10 base pairs form a revolution. Thus, the dsDNA is 3.4 nm per 10 base pairs and 2 nm in width (Peterlinz et al., 1997). The thickness of a 20 base pair DNA is expected to be 6.8 nm if the immobilized DNA duplex is fully extended. Hence, the tilt angle of the immobilized dsDNA can be calculated as 531, which is slightly higher than the 461 previously reported (Berney and Oliver, 2005). The toeholds suspended at both ends of the dsDNA may explain this condition. ssDNA is more flexible and tends to absorb with the long axis parallel, rather than perpendicular, to the surface (Walker and Grant, 1996). The DNA probe adsorbed onto an amino chip had a measured thickness of 0.7 nm (Berney and Oliver, 2005), which was much smaller than that of dsDNA. Therefore, the ssDNA on both ends of the dsDNA slightly contribute to the increase in thickness. The density decreased accordingly with increasing dsDNA layer thickness. Compared with the SA layer, the immobilized dsDNA layer was likely relatively loose because of the rigidity of dsDNA. Upon returning to running buffer, the thickness and mass of the immobilized dsDNA layer slightly decreased, indicating the powerful combination between SA and biotin. During the process, a slight probe loss and rearrangement of the adsorbed layer occurred on the chip surface. Upon the addition of the displacement DNA (Fig. 1: 830 s), the mass decreased by 0.4 ng mm 2, which indicated that the target DNA detached from the sensing surface. The measured thickness value of the final DNA layer was approximately 2 nm, which is much thicker than the ssDNA probe adsorbed onto an amino chip (Berney and Oliver, 2005). This phenomenon indicated that only a part of the dsDNA was replaced and the final DNA layer contained both dsDNA and ssDNA. The displacement DNA can completely combine with the target DNA by toehold initiation, branch migration, and dissociation. The DNA molecules remaining on the chip were mostly 50 biotinylated ssDNA, and the changes in density can further demonstrate the result. Upon injection of the displacement DNA, the density drastically increased because the ssDNA is more random and flexible than dsDNA. Most of the reactions were completed within 180 s and then almost remained unchanged, demonstrating that the toehold-initiated reaction was very rapid. These results showed that the DNA strand displacement reaction occurred when the displacement DNA was added.
3.2. Effect of toehold length The influence of displacement DNA on the displacement reaction was investigated. The effect of different toehold lengths on the displacement efficiency was examined by comparing the response in the change in the DNA layer mass with different lengths of 4, 5, 6, and 7 bases (Fig. 2). Upon addition of the displacement DNA to the sensor chip, a plateau occurred before the displacement reaction took place. It was likely that the micro-fluid chamber was full of TE/Mg2 þ buffer before the injection of the ssDNA solution and a few seconds were needed for the solution to flow through the entire chip surface. At a displacement DNA concentration of 1 mM, which was twice as much as the dsDNA concentration, the displacement process rate was very rapid. At low concentrations (0.5 and 0.2 mM), the reaction was relatively slow. The lower the concentration of the displacement DNA, the lesser was the hybridization with the target DNA. With increasing concentration, both the reaction rate and the mass that dissociated from the dsDNA increased. At a toehold length of 4 bases, the change in mass was relatively moderate compared with those with longer toeholds. Approximately 1000 s was needed to complete the displacement reaction even at high concentrations. Given that the combination of the 4 base pairs was unstable, the initiation of the replacement reaction through a toehold was not that rapid and forceful. With 5, 6, and 7 bases in the toehold, the reaction rate clearly increased. For the toehold with 6 bases, most target DNA molecules were replaced within 180 s at 1 mM and 220 s at 0.5 mM. For the toehold with 5 bases, the corresponding displacement process was completed within 245 s and 310 s, which is slower than the DNA reaction with 6 bases in the toehold. Therefore, the binding of the 6 base pairs is more stable. For the toehold with 7 bases, most target DNA was replaced within 220 s at 1 mM and 300 s at 0.5 mM. The reaction was still slower than with 6 bases in the toehold, but the change in mass was more obvious. Although the binding of 7 base pairs was more stable, longer chains exhibited increasingly complex conformations and required more time to maintain a steady flow on the chip surface. Based on the changes in mass of the injected dsDNA and displacement DNA, the efficiency of the DNA strand displacement reaction and the mass of the target DNA that disassociated from the immobilized dsDNA can be calculated. Adding dsDNA to the sensor surface resulted in the immobilization of PCRprehybridized probe DNA and target DNA on the chip surface through SA-biotin interaction. Therefore, the increase in mass upon dsDNA injection reflected the total quality of the probe DNA and target DNA adsorbed on the surface. The addition of displacement DNA resulted in a combination with the target DNA through the toehold, which expended on the 50 end of the target DNA strand. Therefore, the decreased mass upon injection of displacement DNA injection reflected the quality of the target DNA. To obtain the replacement percentage, the percentage of target DNA in the dsDNA was calculated based on the molecular weight. The molecular weight percentage was calculated using the following equation: MðtargetÞ MðprobeÞ þMðtargetÞ
Fig. 1. Measurement of the thickness (’), mass (), and density (m) values on a sensor chip upon (a) injection of dsDNA (0.5 mM; toehold: 6 bases) and (b) injection of displacement DNA (1 mM).
507
ð2Þ
where M(probe) is the molecular weight of probe DNA and M(target) is the molecular weight of the target DNA. The mass percentages can be obtained through the mass changes in mass upon injection of the dsDNA and the displacement DNA. The displacement percentages of all the other reactions with different toeholds were calculated based on Eq. (2) and the changes in mass. The results were shown in Fig. 3.
508
P. Xu et al. / Biosensors and Bioelectronics 41 (2013) 505–510
Fig. 2. Measured changes in mass of a sensor chip modified with dsDNA upon addition of the displacement DNA with the following toehold lengths: 4 (A), 5 (B), 6 (C) and 7 (D) bases.
varied widely as the concentration decreased lower than 0.5 mM (the dsDNA concentration immobilized on the chip surface). At concentrations higher than 0.5 mM, the percentage did not increase as much with the concentration, which suggests that the displacement percentage is not concentration-dependent. The slight increase in percentage also indicated that the displacement percentage remained almost unchanged above a critical concentration. At high concentrations, the conformation of the DNA molecules in solution was folded and wound. When the displacement DNA solution flowed through the chip surface, the time to get them extended was insufficient. For the toehold with 7 bases, the critical concentration was 0.5 mM, suggesting that the critical concentration lowered with longer toehold length. Fig. 3. Displacement percentages at the four different toehold lengths.
3.3. Effect of toehold sequence At 1, 0.5, and 0.2 mM, the displacement percentages increased with increasing toehold length. At 0.2 mM, the toeholds with 4, 5, and 6 bases, had displacement percentages of 24.54%, 25.99%, and 30.16% respectively, which demonstrates that during the initiation stage, the toehold length greatly influenced the branch migration and displacement reaction. However, for the DNA with 7 bases in the toehold, the percentage reached 70.41%, which is much higher than with 6 bases. In this case, the binding force is much stronger than with shorter toeholds and can replace most of the target DNA molecules, even at lower concentrations. At 1 mM, the displacement percentages using toeholds with 4, 5, 6 and 7 bases were 57.57%, 63.63%, 75.4%, and 83.45%, respectively. The DNA molecules were mostly replaced rather than entirely, even when the displacement DNA concentration was higher than the dsDNA concentration and the toehold combination was powerful. The DNA molecules cannot be entirely replaced. The conformation was probably limited because the DNA strands were immobilized on a surface. For the four toehold lengths, the displacement percentage
The sequences of the toehold with 6 bases in target DNA and displacement DNA were 50 -TGGAGA-30 and 50 -TCTCCA-30 , respectively. The sequences consisted of a mixture of G/C/T/A bases and contain equal numbers of G/C as well as A/T base pairs (three C/G bases and three A/T bases). Adenine (A) pairs with thymine (T) through two hydrogen bonds, whereas three hydrogen bonds form between guanine (G) and cytosine (C), implying that the G/C combination is stronger than the A/T combination. To study the effect of the toehold strength on the displacement reaction, ‘‘weak’’ (all six bases were A and T) and ‘‘strong’’ (all six bases were C and G) toeholds were introduced. Based on a similar immobilization method, the changes in mass upon injection of the displacement DNA were measured (Fig. 4). The reaction rate was relatively slow with the weak toehold (Fig. 4A). Most displacement reactions were accomplished within 400 s at 1 mM. At lower concentrations, the displacement reactions took longer to plateau. With strong toeholds (Fig. 4B), the mass changes
P. Xu et al. / Biosensors and Bioelectronics 41 (2013) 505–510
509
Fig. 4. Measured changes in mass of a sensor chip modified with dsDNA upon the addition of displacement DNA with different toehold sequences: weak (A) and strong (B) toeholds.
combination process. At 1 mM, the displacement percentage was 98.4%, indicating that the target DNA was almost completely replaced and further demonstrating the strong G–C interaction. For the three DNA strands, the displacement percentages increased with increasing concentrations. At the same concentration, the displacement DNA with the strong toehold dissociated more target DNA molecules from the dsDNA. The changes in mass were minimal for the weak toehold. This finding indicates that the strength of the toehold has an important function in the strand displacement reaction.
4. Conclusions Fig. 5. Displacement percentages of three different toehold sequences with different base pair strengths.
at all concentrations were slightly different compared with the medium-strength (AT/CG; Fig. 2C) and the weak toeholds. At 0.5 mM, the mass curve was sigmoidal with two slow stages and a slightly fast stage. The strong toehold initiated the strand displacement reaction more quickly. However, given that it contains more C/G bases, more time was needed to stabilize the conformation when the displacement DNA reached the chip surface. At 0.2 mM, the reaction was slow during the first 500 s and only a few DNA molecules were replaced. After 500 s, the slope changed and the reaction during this period accelerated. The change in mass was almost the same at 0.5 mM, which indicates that the strong binding force between C and G plays an important function in the DNA strand displacement reaction. The mass of the target DNA that disassociated from the dsDNA and the efficiency of the DNA strand displacement reaction were also calculated. The displacement percentages of the three different toehold sequences are shown in Fig. 5. For the weak toehold at 0.2 mM, the displacement percentage was 26%, which was slightly lower than 30.16% because of the weak A–T interaction. When the concentration of the displacement DNA was lower than that of the dsDNA, the percentage increased sharply with increasing concentration. By contrast, the percentage change was relatively moderate, indicating that the percentage did not significantly depend on the concentration. Above a critical concentration, the percentage remained stable with increasing concentration. The condition was slightly different for the strong toehold. At 0.2 mM, the percentage reached 59.9%, which was higher than the weak toehold (26%) and the medium-strength toehold (30.16%). On one hand, this result may be explained by the strong and stable interaction between G and C. With six G–C base pairs in the DNA strand, some conformational adjustments occurred during the
This paper explored the DNA strand displacement reaction through toehold initiation and branch migration on an aminefunctionalized sensor surface. The effects of different toehold lengths and different toehold sequences on the displacement efficiency were also investigated using DPI. This method is useful because the instrument allows the real-time detection of layer thickness and refractive index, from which the density and mass can be calculated. The displacement efficiency can be obtained based on changes in mass upon the addition of different DNA strands onto the sensor surface. Through specific interaction, biotinylated dsDNA was immobilized onto the SA-covered surface of the sensor chip. Upon the addition of the displacement DNA, the mass and the layer thickness decreased, indicating the replacement of several DNA molecules, as confirmed by the increased density. The effect of the toehold length on the displacement efficiency was determined by comparing toeholds with 4, 5, 6, and 7 bases. The mass changes increased with increasing toehold length. When the concentration of the target DNA exceeded that of the dsDNA, the displacement percentage was not concentrationdependent. Above a critical concentration, the displacement percentage remained unchanged. For the toehold with 7 bases, the critical concentration was 0.5 mM. DNAs with weak, medium-strength, and strong toeholds were compared. Considering the reaction was initiated through a toehold, the strength of the toehold combination was highly significant. At 0.5 mM, the displacement percentage of the medium-strength toehold was 65.09% compared with 54.4% for the weak toehold and 74.88% for the strong toeholds. This phenomenon was mainly because the binding affinity between C and G is much stronger than between A and T. All the results show that the displacement efficiency increased with increasing base pair binding affinity.
510
P. Xu et al. / Biosensors and Bioelectronics 41 (2013) 505–510
Acknowledgments This work was supported by the National Natural Science Foundation of China (Grant nos. 20934004 and 91127046); the National Basic Research Program of China (NBR-PC) (Grant nos. 2012CB821500 and 2010CB934500).
Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2012.09. 008.
References Andersen, E.S., Dong, M., Nielsen, M.M., Jahn, K., Subramani, R., Mamdouh, W., Golas, M.M., Sander, B., Stark, H., Oliveira, C.L., Pedersen, J.S., Birkedal, V., Besenbacher, F., Gothelf, K.V., Kjems, J., 2009. Nature 459 (7243), 73–76. Armstrong, J., Salacinski, H.J., Mu, Q., Seifalian, A.M., Peel, L., Freeman, N., Holt, C.M., Lu, J.R., 2004. Journal of Physics: Condensed Matter 16 (26), S2483–S2491. Aulin, C., Varga, I., Claessont, P.M., Wagberg, L., Lindstrom, T., 2008. Langmuir 24 (6), 2509–2518. Berney, H., Oliver, K., 2005. Biosensors and Bioelectronics 21 (4), 618–626. Biehle, S.J., Carrozzella, J., Shukla, R., Popplewell, J., Swann, M., Freeman, N., Clark, J.F., 2004. Biochimica et Biophysica Acta 1689 (3), 244–251. Dirks, R.M., Pierce, N.A., 2004. Proceedings of the National Academy of Sciences of the United States of America 101 (43), 15275–15278. Elhadj, S., Singh, G., Saraf, R.F., 2004. Langmuir 20 (13), 5539–5543. Gray, D.E., CaseGreen, S.C., Fell, T.S., Dobson, P.J., Southern, E.M., 1997. Langmuir 13 (10), 2833–2842. Green, C., Tibbetts, C., 1981. Nucleic Acids Research 9 (8), 1905–1918. Green, R.J., Frazier, R.A., Shakesheff, K.M., Davies, M.C., Roberts, C.J., Tendler, S.J.B., 2000. Biomaterials 21 (18), 1823–1835. Green, S., Bath, J., Turberfield, A., 2008. Physical Review Letters 101 (23), 1–73. Kallenbach, N.R., Ma, R.I., Seeman, N.C., 1983. Nature 305 (5937), 829–831. Lane, T.J., Fletcher, W.R., Gormally, M.V., Johal, M.S., 2008. Langmuir 24 (19), 10633–10636. Lederman, H., Macdonald, J., Stefanovic, D., Stojanovic, M.N., 2006. Biochemistry 45 (4), 1194–1199. Levicky, R., Herne, T.M., Tarlov, M.J., Satija, S.K., 1998. Journal of the American Chemical Society 120 (38), 9787–9792.
Levy, M., Ellington, A.D., 2003. Proceedings of the National Academy of Sciences of the United States of America 100 (11), 6416–6421. Lillis, B., Manning, M., Berney, H., Hurley, E., Mathewson, A., Sheehan, M.M., 2006. Biosensors and Bioelectronics 21 (8), 1459–1467. Lin, S., Lee, C.K., Lin, Y.H., Lee, S.Y., Sheu, B.C., Tsai, J.C., Hsu, S.M., 2006a. Biosensors and Bioelectronics 22 (5), 715–721. Lin, S., Lee, C.K., Wang, Y.M., Huang, L.S., Lin, Y.H., Lee, S.Y., Sheu, B.C., Hsu, S.M., 2006b. Biosensors and Bioelectronics 22 (2), 323–327. Morrison, L.E., Stols, L.M., 1993. Biochemistry 32 (12), 3095–3104. Okahata, Y., Kawase, M., Niikura, K., Ohtake, F., Furusawa, H., Ebara, Y., 1998. Analytical Chemistry 70 (7), 1288–1296. Panyutin, I.G., Hsieh, P., 1993. Journal of Molecular Biology 230 (2), 413–424. Panyutin, I.G., Hsieh, P., 1994. Proceedings of the National Academy of Sciences of the United States of America 91 (6), 2021. Pei, R., Taylor, S.K., Stefanovic, D., Rudchenko, S., Mitchell, T.E., Stojanovic, M.N., 2006. Journal of the American Chemical Society 128 (39), 12693–12699. Peterlinz, K.A., Georgiadis, R.M., Herne, T.M., Tarlov, M.J., 1997. Journal of the American Chemical Society 119 (14), 3401–3402. Rothemund, P.W., 2006. Nature 440 (7082), 297–302. Rothemund, P.W., Papadakis, N., Winfree, E., 2004. PLoS Biology 2 (12), e424. Seelig, G., Yurke, B., Winfree, E., 2006. Journal of the American Chemical Society 128 (37), 12211–12220. Shovsky, A., Bijelic, G., Varga, I., Makuska, R., Claesson, P.M., 2011. Langmuir 27 (3), 1044–1050. Sonesson, A.W., Callisen, T.H., Brismar, H., Elofsson, U.M., 2008. Colloids and Surfaces B: Biointerfaces 61 (2), 208–215. Stojanovic, M.N., Stefanovic, D., 2003. Nature Biotechnology 21 (9), 1069–1074. Striebel, C., Brecht, A., Gauglitz, G., 1994. Biosensors and Bioelectronics 9 (2), 139–146. Walker, H.W., Grant, S.B., 1996. Langmuir 12 (13), 3151–3156. Wang, J., Xu, X., Zhang, Z., Yang, F., Yang, X., 2009. Analytical Chemistry 81 (12), 4914–4921. Win, M.N., Smolke, C.D., 2008. Science 322 (5900), 456–460. Winfree, E., Liu, F.R., Wenzler, L.A., Seeman, N.C., 1998. Nature 394 (6693), 539–544. Wolf, L.K., Gao, Y., Georgiadis, R.M., 2004. Langmuir 20 (8), 3357–3361. Yurke, B., Turberfield, A.J., Mills, A.P., Simmel, F.C., Neumann, J.L., 2000. Nature 406 (6796), 605–608. Zhang, D.Y., Turberfield, A.J., Yurke, B., Winfree, E., 2007. Science 318 (5853), 1121–1125. Zhang, D.Y., Winfree, E., 2008. Journal of the American Chemical Society 130 (42), 13921–13926. Zhang, D.Y., Winfree, E., 2009. Journal of the American Chemical Society 131 (47), 17303–17314. Zhao, X., Pan, F., Coffey, P., Lu, J.R., 2008. Langmuir 24 (23), 13556–13564. Zheng, J., Birktoft, J.J., Chen, Y., Wang, T., Sha, R., Constantinou, P.E., Ginell, S.L., Mao, C., Seeman, N.C., 2009. Nature 461 (7260), 74–77.