Modulating the DNA strand-displacement kinetics with the one-sided remote toehold design for differentiation of single-base mismatched DNA

Abstract

A one-sided remote toehold design was proposed to provide the fine control over strand-displacement reaction kinetics with simplicity and versatility. The significantly improved single-base mismatch differentiating ability of reactions was achieved for strand-displacement reaction systems with the rate-saturated toehold length.

---

DNA is a type of versatile material for the construction of a variety of nanoscale structures.^1,2 The predictability of Watson–Crick binding thermodynamics and the specificity of base pairing allow the rational engineering of diversified static DNA nanostructures and dynamic DNA devices, including DNA assembly,^3,4 molecular computation,⁵ DNA circuits^6–8 and DNA nanomachines.⁹ DNA strand-displacement reactions provide the kinetic control over the reaction pathways in the construction and operation of most dynamic DNA structures.¹⁰ As the component for controlling the kinetics, toehold domains have been designed to meet assorted needs. Particularly, the modulation of a “remote toehold” which includes a spacer between the toehold and branch migration domains provides a delicate control over hybridization kinetics.¹¹ This remote-toehold strategy for operation of displacement reactions has been applied to the kinetic proofreading and concentration-robust regimes.¹¹ The hybridization based “associative toehold activation” strategy has also been proposed, in which the toeholds and branch migration domains are connected through hybridization of auxiliary domains instead of being “hard-wired” during DNA synthesis.¹² In this way, the toeholds and branch migration domains can be conveniently created or changed. Additionally, rational designs of the toehold length^13,14 or toehold sequences for response to environmental stimuli, including metallic and nonmetallic ions,^15,16 small molecules^17,18 and light,¹⁹ have also allowed regulation of DNA strand-displacement reactions.

The specificity of strand-displacement reaction, as an essential property of nucleic acids which ensures the accurate transport of messages encoded in the DNA strand sequences, is one of the cornerstones for constructing and operating dynamic DNA devices and enabling diverse biotechnological reactions and functions.^20–22 Although DNA probes with high specificity have been designed,^22–24 a simple and general design with minimal modifications to the reaction system is still urgently needed. As has been reported, the rate of toehold-mediated, strand-displacement reactions varies roughly exponentially with the increment of the toehold length.^25,26 One mismatch in the toehold domain may impede the reaction by at least one order of magnitude;²⁷ however, the rate of DNA strand-displacement reactions generally saturates when the toehold length exceeds 6–10 nt.¹¹ The reactions in these cases can barely discriminate the strand with single-base mismatched site from the complementary strand,^28,29 which jeopardizes the perfect balance between celerity and accuracy for dynamic construction and operation of DNA advices, limits the room for the design of the toehold domains, and reduces the possibility for using toehold domains as the address tags in DNA networks.¹¹ To this end, the control over the DNA strand-displacement reaction kinetics involving rate-saturated toehold length need to be improved to enlarge the reaction rate difference between the complementary and mismatched strands. Furthermore, it is most desirable to provide more toehold designs for controlling the kinetics to diversify the operation of dynamic devices.

To improve the specificity of DNA strand-displacement reactions while keeping a reasonable implementation time and further expand the rule set of dynamic DNA systems, we, inspired by the remote toehold design, proposed a “one-sided remote toehold” strategy. As illustrated in Scheme 1, the strand-displacement reaction mediated by the one-sided toehold occurs between a pre-hybridized double helix, denoted as SR, and a spacer region with the latter located between the toehold and the double helix, and strand R is fluorescently labelled with TET at the 5′-end and TAMRA at the 3′-end. Strand I, carrying no spacer itself, is complementary to the toehold and branch migration region of strand S. With the occurrence of the strand-displacement reaction, strand R is liberated from the duplex to produce a reduced separation between the donor and acceptor fluorophores due to the formation of the randomly coiled configuration, leading to efficient Förster resonant energy transfer (FRET).

Scheme 1 Schematic illustration of the principle of the one-sided remote toehold-mediated DNA strand-displacement reaction.

The one-sided remote toehold design is proposed based on the toehold-mediated docking and the internal diffusion driven branch migration mechanism suggested by Turberfield¹¹ and D. Y. Zhang.²⁶ First, strand I docked to strand S through toehold binding. Then the branch migration domains in strands S and I contact through the internal diffusion to facilitate branch migration. The product duplex SI is a stable structure with the spacer domain forming a small bulge which has minimal influence on the stability of the duplex. Compared with the proximal strand-displacement, an additional internal diffusion step is introduced by inserting the spacer region.¹¹ Thus the one-sided remote toehold increases the energy threshold at which the reaction rate saturates with the toehold binding strength.¹¹ By increasing the spacer length, the rate of single-base mismatched reaction can be greatly reduced while that of the complementary reaction changed little. In this way, the one-sided remote toehold strategy significantly improves the specificity of strand-displacement reactions with rate-saturated toeholds. Consequently, the discrimination between complementary and single-base mismatched DNA strands with rate-saturated toeholds can be sensitized, which facilitates the design of more diversified dynamic DNA structures.

The “one-sided remote toehold” design differs the “remote toehold” design proposed by Turberfield et al. in several aspects: first, in remote toehold design, the spacers are mostly inserted in both strand S and strand I, whereas in the one-sided remote toehold design, strand I does not carry a spacer, thus the experimental design is simpler and the reaction can be maintained at a reasonably fast rate to facilitate convenient kinetic measurements. Second, the main purpose of the remote toehold design is to provide additional and fine control of the kinetics of hybridization, thus little work on reaction specificity has been explored; while the one-sided remote toehold design is focused on enhancing the specificity of DNA strand-displacement reactions by modulating the reaction kinetics.

To achieve a fine control over the strand-displacement reaction rate and the desired specificity of the reaction with the minimal compromise of the experimental implementation time, short single-stranded oligonucleotides are adopted as the spacer to elevate the energy threshold. The spacer effect on the reaction rate and the kinetic single-base mismatch differentiating ability are evaluated with the rate-saturated toehold lengths, including 8 nt, 10 nt, 12 nt and 14 nt.

By examining the reaction rates at the same temperature level, it was found that for the same toehold length the reaction rate decreased with the increased spacer length. The reaction rate of 8 nt toehold design decreased most significantly with the longer spacer length, while the reaction rates of 10, 12, and 14 nt toehold designs showed moderate decrease, which is attributed to the leveraging effect of toehold with the length at which the strand-displacement reaction rate saturates (Fig. 1A). These results suggested that in these reaction systems 8 nt may be the critical toehold length for the reaction rate to reach saturation. The insertion of spacer added a barrier to the reaction, which exerted more profound effects on the reaction system with the critical toehold length. Second, the reaction rate with the same spacer length (e.g. 3 nt) increased with elevated temperature, and the temperature effect became more substantial for reaction systems with the longer toehold domain (Fig. 1B). The kinetic curves of reaction with the 8 nt toehold length did not response significantly to the change of temperature, which may be attributed to the incomplete hybridization of the short toehold region at the elevated temperature (Table S2†). Third, for reaction systems with the same toehold length (e.g. 12 nt), reaction rates increased with temperature for each spacer length, and the temperature effect became more profound for longer spacer, possibly due to the fact that the internal diffusion rate increased at the elevated temperature (Fig. 1C). As a result, the activation energies of the above strand-displacement reactions increased with the increment of the toehold length and the spacer length, respectively (Fig. S5†). These results confirmed that the additional internal diffusion with the introduction of the spacer makes a difference in tuning the strand-displacement kinetics.

Fig. 1 The reaction rate changing trends as a function of the spacer length with four different toehold length designs at 25 °C (A); the reaction rate changing trends as a function of temperature taking the 3 nt spacer design as an example (B). The reaction rate changing trends as a function of temperature taking the 12 nt toehold design as an example (C). For convenience, the reaction rate change is denoted in natural logarithm of the reaction rate constants (Δln k).

Based on the above results, it is expected that the hybridization kinetic differences between complementary and single-base mismatched DNA may be sensitized to allow the single-base mismatch differentiation by tuning the spacer to an appropriate length matching the corresponding toehold length. In light of this, the time-dependent fluorescence signals in the first 90 seconds of complementary and single-base mismatched reaction systems with four different toehold lengths (8 nt, 10 nt, 12 nt and 14 nt) and associated varying spacer lengths were recorded, and the corresponding discrimination factors were calculated by the initial reaction rate detection method (Fig. 2).²⁹

Fig. 2 The discrimination factor as a function of the spacer length and the mismatched site. Toehold lengths were 8 nt for (A), 10 nt for (B), 12 nt for (C) and 14 nt for (D). The mismatched site on the toehold domain started from the site “a” which is located next to the spacer domain. * The discrimination factor at this site is 771.

The discrimination factors varied with mismatched sites and were found to be enhanced for most mismatched sites compared with the proximal strand-displacement design which contains no spacer. For the 8 nt and 10 nt toehold systems, discrimination factors of quite a few sites exceeded 100 with some even greater than 200, which cannot be achieved by most non-enzymatic reaction systems. Mismatched sites located in the middle region of the toehold domain are more likely to be differentiated. With the 1 nt spacer for the 8 nt toehold domain design, the mismatch at the first site which is next to the spacer domain can be differentiated with a significantly improved factor. For longer toehold designs, the first significantly improved differentiation of mismatch was the second nucleotide for 10 nt toehold design and the fourth nucleotide for 12 nt toehold design, respectively. The differentiating ability varied with increment of the spacer length, and the best differentiating effects were achieved at the spacer length of 1 nt for the 8 nt toehold design, 2 nt for the 10 nt toehold design, and 3 nt for the 12 nt toehold design. However, few sites of the 14 nt toehold design achieved improved differentiation possibly because the added internal diffusion by the spacer is not enough to provide rate differentiation, which further demonstrated the dependence of differentiating ability on both toehold and spacer lengths. However, the dependence of differentiating abilities on mismatched sites, which was also encountered in other works,^22,30 remains to be elucidated.

To assure that the one-sided remote toehold design can be adopted as a general strategy for single-base mismatch differentiation, the discrimination factors for the 10 nt toehold design without or with a 2 nt spacer were obtained at two different toehold sequences (Fig. S6†). Although the discrimination factors for each site of the two systems were different, both one-sided remote toehold-mediated reactions showed significantly improved differentiating abilities compared with the proximal toehold-mediated reactions, which proved that the one-sided remote toehold design to enhance reaction specificity is applicable to different toehold sequence designs. Taking the 10 nt toehold with 2 nt spacer design as an example, the discrimination factor with the spacer sequence variation (TT, CT, CC, TC) was evaluated (Fig. S7†). The differentiating abilities with all the four tested spacer sequences were greatly improved compared with the proximal toehold design, although the factor at the same mismatched position differed with the spacer sequences. The cause of variation of differentiating ability in mismatched positions should be associated with the free energy change in hybridization reactions as well as the kinetics, though the mechanism may be complicated.

The single-base mismatch effect was compared amongst the proximal toehold, the symmetrical remote toehold, and the proposed one-sided remote toehold designs (Fig. S8†). Taking the 10 nt toehold domain design as an example, the symmetrical remote toehold design can increase the energy threshold at which the reaction rate saturates, thus presented better specificities than the proximal toehold design. The one-sided remote toehold strategy presented the faster strand-displacement rate and showed further improved differentiating abilities than the symmetrical remote toehold strategy. In addition to the better specificity and faster reaction rate which is friendly for experimental implementation, special design is not needed for the invader strand (strand I), which may simplify the sequence design for applications in dynamic DNA nanotechnologies.

In summary, we proposed a one-sided remote toehold strategy for modulating the strand-displacement reaction kinetics, which can substantially enhance the hybridization specificity of reactions with rate-saturated toehold lengths. The design is very effective for 10 nt and 12 nt toehold lengths with which discrimination factors of most sites were significantly improved, and quite a few factors exceeded 100 by simply tuning the spacer length. This work investigated the effect of one-sided remote toehold on the kinetic control and activation energy of the strand-displacement, which provided additional perspectives on modulating strand-displacement reaction kinetics by modifying the toehold domain. As a type of non-traditional base-pairing approach, the one-sided remote toehold design developed in this work should expand the room for designing DNA circuits and structures as well as the breadth of applications.^31,32

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 21475004, 21275011, and 21535006).

Notes and references

1. M. R. Jones, N. C. Seeman and C. A. Mirkin, Science, 2015, 347, 1260901.
2. B. A. Baker, G. Mahmoudabadi and V. T. Milam, Soft Matter, 2013, 9, 11160–11172.
3. L. Qiu, T. Chen, I. Oecsoy, E. Yasun, C. Wu, G. Zhu, M. You, D. Han, J. Jiang, R. Yu and W. Tan, Nano Lett., 2015, 15, 457–463.
4. P. Yin, H. M. T. Choi, C. R. Calvert and N. A. Pierce, Nature, 2008, 451, 318–322.
5. L. Qian, E. Winfree and J. Bruck, Nature, 2011, 475, 368–372.
6. S. M. Chirieleison, P. B. Allen, Z. B. Simpson, A. D. Ellington and X. Chen, Nat. Chem., 2013, 5, 1000–1005.
7. D. Y. Zhang and E. Winfree, Nucleic Acids Res., 2010, 38, 4182–4197.
8. X. Xu, J. Zhang, F. Yang and X. Yang, Chem. Commun., 2011, 47, 9435–9437.
9. H. Liu and D. Liu, Chem. Commun., 2009, 2625–2636,  10.1039/b822719e.
10. D. Y. Zhang and G. Seelig, Nat. Chem., 2011, 3, 103–113.
11. A. J. Genot, D. Y. Zhang, J. Bath and A. J. Turberfield, J. Am. Chem. Soc., 2011, 133, 2177–2182.
12. X. Chen, J. Am. Chem. Soc., 2012, 134, 263–271.
13. D. Yao, B. Wang, S. Xiao, T. Song, F. Huang and H. Liang, Langmuir, 2015, 31, 7055–7061.
14. B. A. Baker and V. T. Milam, Nucleic Acids Res., 2011, 39, e99.
15. W. Ding, W. Deng, H. Zhu and H. Liang, Chem. Commun., 2013, 49, 9953–9955.
16. W. Tang, H. Wang, D. Wang, Y. Zhao, N. Li and F. Liu, J. Am. Chem. Soc., 2013, 135, 13628–13631.
17. Y. Xing, Z. Yang and D. Liu, Angew. Chem., Int. Ed., 2011, 50, 11934–11936.
18. X. Xu and X. Yang, Chem. Commun., 2014, 50, 805–807.
19. F. Huang, M. You, D. Han, X. Xiong, H. Liang and W. Tan, J. Am. Chem. Soc., 2013, 135, 7967–7973.
20. D. A. Khodakov, A. S. Khodakova, A. Linacre and A. V. Ellis, J. Am. Chem. Soc., 2013, 135, 5612–5619.
21. H. Li, S. Xiao, D. Yao, M. H.-W. Lam and H. Liang, Chem. Commun., 2015, 51, 4670–4673.
22. D. Y. Zhang, S. X. Chen and P. Yin, Nat. Chem., 2012, 4, 208–214.
23. S. X. Chen, D. Y. Zhang and G. Seelig, Nat. Chem., 2013, 5, 782–789.
24. J. S. Wang and D. Y. Zhang, Nat. Chem., 2015, 7, 545–553.
25. B. Yurke and A. P. Mills Jr, Genetic Programming and Evolvable Machines, 2003, 4, 111–122.
26. D. Y. Zhang and E. Winfree, J. Am. Chem. Soc., 2009, 131, 17303–17314.
27. I. G. Panyutin and P. Hsieh, FASEB J., 1994, 8, A1395.
28. Q. Li, G. Luan, Q. Guo and J. Liang, Nucleic Acids Res., 2002, 30, e5.
29. C. Li, Y. Li, X. Xu, X. Wang, Y. Chen, X. Yang, F. Liu and N. Li, Biosens. Bioelectron., 2014, 60, 57–63.
30. Y. Xiao, K. J. I. Plakos, X. Lou, R. J. White, J. Qian, K. W. Plaxco and H. T. Soh, Angew. Chem., Int. Ed., 2009, 48, 4354–4358.
31. Y. S. Jiang, S. Bhadra, B. Li and A. D. Ellington, Angew. Chem., Int. Ed., 2014, 53, 1845–1848.
32. R. R. F. Machinek, T. E. Ouldridge, N. E. C. Haley, J. Bath and A. J. Turberfield, Nat. Commun., 2014, 5, 5324.

---

Footnote

† Electronic supplementary information (ESI) available: The kinetic curves, supplementary figures and the experimental details. See DOI: 10.1039/c6ra17006d

---