Kinetics of Proximity-Induced Intramolecular DNA Strand Displacement

Jul 25, 2016 - The Logic Circuit Design of Fire Alarm System Device by DNA Strand Displacement. Yanfeng Wang , Jixiang Li , Chun Huang , Junwei Sun...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Kinetics of Proximity-Induced Intramolecular DNA Strand Displacement Feng Li,*,† Yanan Tang,‡ Sarah M. Traynor,† Xing-Fang Li,‡ and X. Chris Le*,‡ †

Department of Chemistry and Center for Biotechnology, Brock University, St. Catharines, Ontario L2S 3A1, Canada Department of Laboratory Medicine and Pathology, University of Alberta, Edmonton, Alberta T6G 2G3, Canada

‡

S Supporting Information *

ABSTRACT: Proximity-induced intramolecular DNA strand displacement (PiDSD) is one of the key mechanisms involved in many DNA-mediated proximity assays and protein-responsive DNA devices. However, the kinetic profile of PiDSD has never been systematically examined before. Herein, we report a systematic study to explore the kinetics of PiDSD by combining the uses of three DNA strand displacement techniques, including a binding-induced DNA strand displacement to generate PiDSD, an intermolecular DNA strand-exchange strategy to measure a set of key kinetic parameters for PiDSD, and a toehold-mediated DNA strand displacement to generate fluorescence signals for the real-time monitoring of PiDSD. By using this approach, we have successfully revealed the kinetic profiles of PiDSD, determined the enhanced local effective concentrations of DNA probes that are involved in PiDSD, and identified a number of key factors that influence the kinetics of PiDSD. Our study on PiDSD establishes knowledge and strategies that can be used to guide the design and operation of various DNA-mediated proximity assays and protein-triggered DNA devices.

D

biomarker validation, to early stage disease diagnostics, and to cell imaging.11−30 Although DNA-mediated proximity assays, including PLA,10−12 proximity extension assays,13,14 molecular pincer assays,15−17 electrochemical proximity assays,18−20 bindinginduced DNA hybridization assays,21−23 and binding-induced DNA strand displacement assays,24−30 vary largely at signal transduction and detection schemes, they all share one common feature: multiple DNA sequences are brought into proximity through affinity interactions to generate detectable DNA sequences or signals. During this process, proximityinduced intramolecular DNA strand displacement (PiDSD) has been frequently adopted as a key mechanism to generate targetdependent detection signals and/or to reduce target-independent backgrounds.10−12,20−30 Moreover, PiDSD is also a key component to the development of protein-responsive dynamic DNA devices.25−28 For example, by using the principle of

NA has been well-recognized as the genetic material for almost all living organisms, and it is also a unique class of biopolymers finding extensive applications in material and medical sciences.1−4 Due to the ease to be programmed through Watson−Crick base paring and manipulated by diverse enzymes, DNA serves as a powerful tool for the development of sensors and assays to achieve ultrasensitive protein analysis.3−8 In 1992, Sano et al. described the first example of DNAmediated ultrasensitive protein assay, immunopolymerase chain reaction (immuno-PCR), demonstrating the possibility to convert the challenging protein detection into the wellestablished DNA amplification.9 However, immuno-PCR and other separation-based approaches require tedious capturing and washing steps to minimize target-independent DNA amplification. To address this challenge, Fredriksson et al. developed proximity ligation assay (PLA) in 2002, enabling the ultrasensitive detection of proteins in homogeneous solutions.10 Since then, continuous efforts have been made to expand the concept and strategies of DNA-mediated proximity assays for ultrasensitive protein analysis with applications ranging from © XXXX American Chemical Society

Received: May 16, 2016 Accepted: July 25, 2016

A

DOI: 10.1021/acs.analchem.6b01900 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

estimated by plotting the value of ln(1 − [O]t/[O]max) at each time point against the reaction time t, and a least-squares linear regression was used to determine the observed rate constant kobs. Linear fittings were performed using OriginPro 9, and confidence intervals were determined using the York method. For a typical toehold-mediated strand displacement, the reaction mixture contained 10 nM OD, 10 nM C, 20 nM FQ, 50 nM ROX, 1 μM polydT oligo, and TE−Mg buffer. To monitor the kinetic process of the toehold-mediated strand displacement reaction using FQ, the same experimental protocol was used as for the study of PiDSD.

PiDSD, we have previously developed a number of proteinresponsive DNA devices, including binding-induced molecular translator,25 binding-induced DNA three-way junction,26,27 and protein-responsive catalytic hairpin assemblies.28 Despite its wide applications to biosensors and DNA device designs, the kinetic profile of PiDSD has never been systematically examined before. To our knowledge, the only relevant study is the design and study of the remote DNA toehold,31 where DNA hybridization is used to induce the internal strand displacement. While previous studies on the kinetics of intraand intermolecular DNA strand displacement are essential to understanding the kinetic behavior of PiDSD, the nature of DNA hybridization is very different from that of affinity interactions; therefore, knowledge obtained from previous studies cannot be readily applied to PiDSDs that are involved in proximity protein assays and protein-responsive DNA devices. To fill this knowledge gap, we designed a novel strategy by integrating three DNA strand displacement techniques, including binding-induced DNA strand displacement, intermolecular strand exchange, and toehold-mediated DNA strand displacement, to systematically study the kinetics of PiDSD.

■

RESULTS AND DISCUSSION Theoretical Consideration and Experimental Design. Kinetics of intermolecular DNA strand exchange (Scheme 1A) Scheme 1. Schematic Illustrations of (A) the Intermolecular DNA Strand Exchange between DNA Duplex OT and Competing DNA C, (B) Intramolecular DNA Strand Displacement That Is Involved in a Binding-Induced DNA Strand Displacement Reaction, and (C) the ToeholdMediated DNA Strand Displacement between O and a Strand Displacement Beacon FQa

■

EXPERIMENTAL SECTION Materials. All DNA samples were purchased from Integrated DNA Technologies (Coralville, IA) and purified by HPLC. The DNA sequences and modifications are listed in Table S1. Streptavidin from Streptomyces avidinii, biotin, magnesium chloride hexahydrate (MgCl2·6H2O), and 100× Tris−EDTA (TE, pH 7.4) buffer were purchased from Sigma. Platelet-derived growth factor BB (PDGF-BB) was purchased from R&D Systems. NANOpure H2O (>18.0 MΩ), purified using an Ultrapure Milli-Q water system, was used for all experiments. Prepare DNA Probes for PiDSD and the Displacement Beacon. DNA probe OT was prepared at a final concentration of 5 μM by mixing 20 μL of 50 μM template DNA T molecules with an equal amount of the output DNA O in 160 μL of TE− Mg buffer (1× TE, 10 mM MgCl2, 0.05% Tween20), heating to 90 °C for 5 min, and allowing the solution to cool down to 25 °C slowly in a period of 3 h. Displacement beacon FQ was prepared the same way as for the OT probe. Measure the Kinetics of PiDSD Using the Displacement Beacon. For a typical PiDSD reaction, the reaction mixture contained 10 nM OT, 10 nM competing DNA C, 20 nM FQ, 50 nM ROX reference dye, 1 μM polydT oligo, varying concentrations of streptavidin, and TE−Mg buffer. The reaction mixture was incubated at 25 °C for 150 min in a 96well microplate. Fluorescence was measured immediately after adding the mixture to the well of microplate and kept being measured every 1.5 min for the first 30 min and then every 5 min for another 2 h using a multimode microplate reader (DX880, Beckman Coulter) with both excitation/emission at 485/515 nm for the displacement beacon and excitation/ emission at 535/595 nm for ROX as a reference dye. The measured fluorescence signal was normalized so that 1 n.u. (normalized unit) of fluorescence corresponded to fluorescence signal generated by 1 nM O. This normalization was achieved using a positive control containing 10 nM O, 20 nM FQ, 1 μM polyT oligo, and 50 nM ROX in TE−Mg buffer, and a negative control containing identical reagents in positive control except that there was no O added. All samples were tested in triplicate to ensure assay reproducibility. Reaction rate constants were

a

In panel B, OT and C are each conjugated with an affinity ligand that binds to the same target protein but at different epitopes. The sequence of probe C is identical to the complementary part of the output DNA O; thus, the rate of intermolecular strand exchange between OT and C is very slow in the absence of the target protein. Once both DNA probes bind to the same target molecule through affinity interactions, the intermolecular strand displacement is converted to PiDSD; thus, the rate of strand exchange is greatly accelerated. As a consequence, output DNA O is released from OT duplex. In panel C, the toehold-mediated DNA strand displacement between O and strand displacement beacon FQ is able to turn on the fluorescence signals by separating the fluorophore F from the quencher Q and thus can be used to monitor the released O from intermolecular strand exchange (panel A) or PiDSD (panel B) in real time.

has been well-established previously.32 The observed rate constant kobs of this reaction can be described using the following equation: kobs = k1 + k 2[C]

(1)

where k1 is the dissociation rate constant of duplex OT, k2 is the rate constant of the sequential displacement through a branch migration between C and OT, and [C] is the effective concentration of the competing DNA probe C. The rate constant k1 corresponds to the dissociation of OT and thus can B

DOI: 10.1021/acs.analchem.6b01900 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry be modeled as a random walk of the ends at the double-helical region32,33 and expressed as k1 = 2k f (N − n + 1) exp(ΔH1(T − Tm)/(RTTm))

(2)

where kf is the rate constant for the formation of one base pair (bp) at the border of the helical region, N is the length of OT duplex, n is the minimum length of the stable duplex, ΔH1 is the formation enthalpy of OT duplex, and Tm is the melting temperature of OT. The rate constant k2 corresponds to the sequential displacement between OT and C through a branch migration,32 and can be expressed as k2 =

2ka

(

(N − 2n + 2) exp

ΔH2(T − Tm) RTTm

)

(3)

where ka is the association rate constant for OT duplex formation and ΔH2 is the enthalpy for opening n base pairs at one end of OT duplex. As PiDSD also involves DNA strand exchanges that are initiated and accelerated by forming the sandwiched affinity binding complexes (Scheme 1B), previous knowledge that describes the kinetics of intermolecular DNA strand exchange can also be applied to PiDSD.32 To experimentally explore the kinetics of PiDSD, it is essential to keep the affinity binding step effectively irreversible and sufficiently rapid so that it is not rate-limiting. To meet this requirement, we chose streptavidin as a model target protein and biotin molecules as affinity ligands because of their extremely low Kd value (thus can be considered to be irreversible in our experimental settings).34 The rapidness of this affinity interaction was confirmed by a binding-induced fluorescence quenching assay (details in the Supporting Information and Figure S1). Another important consideration when designing experiments to study the kinetics of PiDSD is the rapid system response time to allow the realtime monitoring of the reactions. To achieve this goal, we designed a strand displacement beacon FQ based on the principle of toehold-mediated DNA strand displacement (Scheme 1C).35 The DNA toehold, a short single-stranded DNA segment juxtaposing the branch migration domain, is a powerful strategy to accelerate the rates of strand exchange reactions to over 106-fold.35−37 To be used as a signal reporter, a pair of fluorophore and quencher was conjugated to DNA probes F and Q, respectively, so that when PiDSD occurs, the released output DNA O can turn on the fluorescence of FQ by separating the fluorophore-labeled F from the quencher-labeled Q through a toehold-mediated strand displacement. Once the model system is established, we aim to explore three important aspects in regard to PiDSD, including (1) the kinetic profile of PiDSD, (2) the driving force for PiDSD, and (3) factors that influence the kinetics of PiDSD. Kinetic Profile of PiDSD. To explore the kinetics of PiDSD, we monitored the released output DNA O in real time over a period of 2 h at 25 °C by using the strand displacement beacon FQ (Figure 1). Since the affinity binding between streptavidin and biotin completes almost instantaneously (Figure S1B), the PiDSD can then be considered as a dissociation reaction: OTC → TC + O (the overall reaction process is proposed in Scheme 2). Therefore, the rate V of PiDSD can be expressed as V = −d[OTC]/dt = k[OTC]x

Figure 1. (A) Schematic illustration of the strategy for real-time monitoring the PiDSD that is triggered by streptavidin−biotin interaction. Binding of streptavidin to the two biotin-containing DNA probes results in PiDSD and the release of the output DNA O. The subsequent toehold-mediated strand displacement reaction between O and FQ releases the fluorescently labeled strand F and turns on fluorescence. (B) Kinetic profile of PiDSD in the presence of varying concentrations of OTC showing normalized fluorescence signal over time. The concentration of OTC is determined by using the equation [OTC] = 1.25[streptavidin]. The fluorescence intensity was normalized such that 1 normalized unit (n.u.) corresponds to 1 nM O. The sample (target) solution contained varying concentrations of streptavidin, 20 nM OT, 20 nM C, and 20 nM FQ in TE−Mg buffer. In the negative control (N.C.), all reagents were the same as in the sample, except that there was no streptavidin. (C) Determining the order of PiDSD reaction by plotting the logarithmic values of the initial rates V against the logarithmic values of the concentrations of OTC. The slope was determined to be 1, suggesting that the PiDSD is a first-order reaction. The rate constant kobs was determined from the intercept of the plot to be 4.90 × 10−4 s−1.

Scheme 2. Schematic Illustration of the Proposed Internal Diffusion and Strand Exchange Processes Involved in PiDSD

trations of OTC in real time by using the displacement beacon FQ (Figure 1). We then measured the reaction rate V for each PiDSD reaction based on the obtained kinetic curve. The concentration of OTC is determined by using the following equation: [OTC] = 1.25[streptavidin]. The constant 1.25 was determined based on our previous estimation that each streptavidin molecule can on average lead to the formation of 1.25 OTC complexes.24 We then determined the order of PiDSD reaction x by plotting the logarithmic values of V against the logarithmic values of [OTC]. As shown in Figure 1C, the slope was determined to be 1 in a linear regression, suggesting that the PiDSD is a first-order reaction (x = 1). The rate constant kobs was then calculated to be 4.90 × 10−4 s−1 by using the intercept of the plot. To confirm that this observed rate constant truly reflects the kinetics of PiDSD, we further estimated the rate constant for the subsequent toeholdmediated DNA strand displacement between O and FQ as a

(4)

To determine the order of the PiDSD reaction, we monitored PiDSDs that were initiated by varying concenC

DOI: 10.1021/acs.analchem.6b01900 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

biotin interaction is equivalent to a 5-nt long DNA toehold regarding to its ability to accelerate DNA strand exchange reactions. Driving Force of PiDSD. The primary driving force for PiDSD is the affinity interactions that bring the competing DNA probe C into a close proximity to the target duplex probe OT. It has been frequently proposed that such proximity could drastically increase the local effective concentrations of DNA probes10−30 and hence promote the strand exchange between OT and C likely via the acceleration of the internal strand diffusion. Although the enhancement on the local effective concentration has been predicted by simple theoretical models previously,6 it has never been quantitatively determined in experimental settings. Since eq 1 establishes a linear relationship between the observed rate constant kobs and the concentration of the competing DNA ([C]), we proposed that the enhanced local effective concentration can be determined by measuring kobs, k1, and k2 experimentally and calculated by using the eq 1. As shown in Figure 1, the observed rate constant kobs of PiDSD can be determined by measuring the released O using the displacement beacon FQ, and thus the remaining question is how to estimate the rate constants k1 and k2 under the same condition. To achieve this goal, we made use of the intermolecular strand exchange reaction between OT and C (Figure 3A), where we fixed the concentration of OT at

positive control (P.C. in Figure 1B). The obtained rate constant (6.15 × 10−3 s−1) was over 10 times higher than that with a PiDSD component, confirming that PiDSD is the rate-limiting step in the overall strand displacement reaction. Once determining PiDSD to be a first-order reaction, we can then calculate the rate constant for each individual reaction by using the following equation: ln(1 − [O]t /[O]max ) = kobst

(5)

where [O]t is the amount of released O at time point t and [O]max is the maximum amount of O that can be released by a given concentration of streptavidin. As shown in Figure S2, the observed rate constant for a PiDSD between a 12-nt C (10 nM) and a 12-bp OT (10 nM) in the presence of 2.5 nM streptavidin was determined to be (5.56 ± 0.06) × 10−4 s−1. The first-order kinetic profile also indicates that a slow internal strand diffusion step was involved in PiDSD (Scheme 2), a feature similar to that of a remote toehold-mediated DNA strand displacement.31 To quantitatively understand the displacement rate of PiDSD, we further compared PiDSD to a series of toeholdmediated DNA strand displacement reactions that are of the same duplex sequence as OT (termed as OD) but with varying toehold lengths from 2 to 6 nt (Figure 2A). As the same output

Figure 2. Comparison between PiDSD that is triggered by streptavidin−biotin interaction with toehold-mediated DNA strand displacements. (A) Schematic illustration of the strategy to monitor the toehold-mediated DNA strand displacement between OD and C in real time using the displacement beacon FQ. (B) Monitoring the fluorescence increases as a function of time. (C) The logarithmic values of rate constants as a function of toehold length for toeholdmediated DNA strand displacement reactions. By fitting the rate constant of PiDSD under the same condition into this plot, we were able to determine the streptavidin−biotin interaction was equivalent to a 5-nt long DNA toehold regarding to its ability to accelerate the DNA strand exchange.

Figure 3. (A) Schematic illustration of the strategy to monitor the intermolecular strand exchange between OT and C using the displacement beacon FQ. (B) Determining the rate constant kobs under different concentrations of C by plotting the data according to the eq 4. (C) The values of the observed rate constants as a function of concentrations of C. Rate constants k1 and k2 were then estimated from the linear regression of the plot.

DNA O was generated by the duplex OD in the toeholdmediated displacement reaction, we were able to use the same FQ beacon to monitor the reaction in real time. By monitoring the fluorescence increase over a period of 150 min, we were able to determine the rate constants of toehold-mediated DNA strand displacements at different toehold lengths (Figure 2B). Consistent with previous observations,35−37 the logarithmic values of the rate constants showed a linear relationship with the toehold lengths (Figure 2C). By fitting the measured rate constant of PiDSD under the exact same experimental conditions into the plot, we were able to determine an effective toehold length to be 5 nt, suggesting that the streptavidin−

10 nM and varied the concentration of C from 0 nM to 2.5 μM (Figure 3B). We further confirmed that increasing concentrations of C did not affect the subsequent toehold-mediated DNA strand displacement reaction between O and FQ (Figure S3). Therefore, FQ was used to quantify the amount of O that was generated from the strand exchange between OT and C. By fitting the obtained data into eq 1, we were able to determine k1 and k2 for reactions with 12-nt DNA probes to be (2.51 ± 0.39) × 10−5 s−1 and 6.8 ± 1.3 M−1 s−1, respectively. By fitting the determined k1 and k2 with kobs for PiDSD, we then determined an average local effective concentration of C to be 65.0 μM for PiDSD with 12-nt DNA probes. When compared to its bulk D

DOI: 10.1021/acs.analchem.6b01900 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry concentration (10 nM), the effective concentration of C was enhanced by 6500-fold upon assembling to the OT duplex through streptavidin−biotin bindings, confirming the critical role of local effective concentration of DNA probes to PiDSD. The determined local effective concentration of C was consistent with the estimated value in PLA,12 but considerably lower than predicted values from the theoretical model by Zhang et al. (∼400 μM under the same condition),6 suggesting that other factors may also influence the kinetics of PiDSD. Factors That Influence the Kinetics of PiDSD. It is of good practical value to understand how various designs in DNA probes influence the kinetics of PiDSD, because such knowledge can readily be used to guide the design and operation of various DNA-mediated proximity assays and protein-responsive DNA devices. Herein, we investigated a number of key factors that can influence the kinetics of PiDSD, including the sequence designs of both DNA probes and the single-stranded DNA spacers that link the affinity ligands to the DNA probes. Equations 2 and 3 predict that the rate constants k1 and k2 of a DNA strand exchange reaction depend largely on the length of the DNA duplex and corresponding competing DNA. To experimentally examine the length effect of DNA probes on k1 and k2, we designed a set of DNA duplexes and competing DNA probes with lengths including 12, 14, and 16 nt. We then determined the rate constants k1 and k2 at different probe lengths by using the intermolecular strand exchange reactions (Figure S4). As shown in Table 1, both k1 and k2 decrease as we

Figure 4. Effect of the spacer sequence on the kinetics of PiDSD. (A) Schematic illustration of a PiDSD using streptavidin−biotin interaction with a competing DNA that contains a 6-bp (C-H6) or 4-bp (C-H4) stem−loop DNA hairpin in the spacer region. (B) Realtime monitoring the fluorescence increase as a function of time using the strand displacement beacon. The sample solution contained 5 nM streptavidin, 10 nM OT, 10 nM C (C12, C-H4, or C-H6), and 20 nM FQ in TE−Mg buffer.

the PiDSD, whereas a less stable DNA hairpin (estimated Tm = 46.5 °C) has no effect at all (Figure 4B). Such a sharp transition indicates that it is possible to design allosteric controls to protein-responsive DNA devices via the manipulation of spacer sequences. A more general concern when designing spacers for PiDSD is the choice of the spacer length. Previous study on remote DNA toehold has demonstrated that shorter spacers resulted in faster internal strand displacements.31 However, longer spacers were found to be more effective in designing DNA-mediated proximity assays for the homodimer of PDGFBB in our study (details in Figure S5). This result suggests that when applying to PiDSD to various DNA-mediated assays or devices, sufficient spacer flexibility shall be a primary concern to overcome the steric hindrance arising from the large molecular sizes of both target proteins and affinity ligands.

Table 1. Effect of DNA Probe Length (Duplex and Corresponding Competing DNA) on the Kinetics of PiDSD length of DNA probes

k1 (10−6 s−1)

k2 (M−1 s−1)

kcal (10−6 s−1)

kobs (10−6 s−1)

12

25.1 ± 3.9

6.8 ± 1.3

467 ± 22

14 16

4.9 ± 1.6 4.7 ± 0.7

3.1 ± 0.7 1.5 ± 0.4

not applicable 206 ± 47 102 ± 27

146 ± 6 96 ± 10

increase the probe length from 12 to 16 nt. This observation is consistent with the previous study on strand exchange reactions.32 Because adding two or four oligonucleotides to DNA probes would not significantly change their local effective concentrations upon affinity bindings, the concentration [C] in eq 1 remains a constant when probe lengths vary from 12 to 16 nt. We then estimated the rate constant kcal of PiDSD by fitting previously calculated [C] and newly determined k1 and k2 into eq 1. We also measured the observed rate constants kobs of PiDSD experimentally for the same set of DNA probes by using the displacement beacon FQ. As shown in Table 1, the calculated rate constants decreased sharply as we increased the lengths of DNA probes. This trend of calculated rate constants agrees well with that of the measured values, suggesting the significant roles of probe length on the kinetics of PiDSD. This result is also consist with many DNA-mediated proximity assay designs, where optimal duplex lengths involved in PiDSD are generally between 10 and 12 nt (e.g., 10 nt for PLA,10−12 11 nt for binding-induced DNA assembly assays,21 11 nt for bindinginduced molecular translator,25 and 12 nt for binding-induced strand displacement assays24). The design of the DNA spacers is another important factor that can influence the kinetics of PiDSD. As shown in Figure 4, incorporation of a stable DNA hairpin (estimated Tm = 58.7 °C) into the linear flexible polydT spacer completely turns off

■

CONCLUSIONS In this study, we have successfully established a strategy that allowed us to systematically study the kinetics of proximityinduced intramolecular DNA strand displacement (PiDSD), a widely used signal transduction mechanism in many DNAmediated proximity protein assays and protein-responsive DNA devices. Our studies revealed the kinetic profile of PiDSD and identified a number of key factors that influence the rates of PiDSD. We anticipate that the knowledge obtained from our study can readily be used to help the design and operation of various DNA-mediated proximity assays for sensitive protein analysis. Our study may also lead to the development of new strategies for the design and regulation of novel proteinresponsive DNA devices.

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.6b01900. E

DOI: 10.1021/acs.analchem.6b01900 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

■

(26) Li, F.; Lin, Y.; Le, X. C. Anal. Chem. 2013, 85, 10835−10841. (27) Tang, Y.; Wang, Z.; Yang, X.; Chen, J.; Liu, L.; Zhao, W.; Le, X. C.; Li, F. Chem. Sci. 2015, 6, 5729−5733. (28) Tang, Y.; Lin, Y.; Yang, X.; Wang, Z.; Le, X. C.; Li, F. Anal. Chem. 2015, 87, 8063−8066. (29) Zong, C.; Wu, J.; Liu, M.; Yan, F.; Ju, H. Chem. Sci. 2015, 6, 2602−2607. (30) Zhang, L.; Zhang, K.; Liu, G.; Liu, M.; Liu, Y.; Li, J. Anal. Chem. 2015, 87, 5677−5682. (31) Genot, A. J.; Zhang, D. Y.; Bath, J.; Turberfield, A. J. J. Am. Chem. Soc. 2011, 133, 2177−2182. (32) Reynaldo, L. P.; Vologodskii, A. V.; Neri, B. P.; Lyamichev, V. I. J. Mol. Biol. 2000, 297, 511−520. (33) Anshelevich, V. V.; Vologodskii, A. V.; Lukashin, A. V.; FrankKamenetskii, M. D. Biopolymers 1984, 23, 39−58. (34) Weber, P. C.; Ohlendorf, D. H.; Wendoloski, J. J.; Salemme, F. R. Science 1989, 243, 85−88. (35) Zhang, D. Y.; Seelig, G. Nat. Chem. 2011, 3, 103−113. (36) Zhang, D. Y.; Winfree, E. J. Am. Chem. Soc. 2009, 131, 17303− 17314. (37) Srinivas, N.; Ouldridge, T. E.; Sulc, P.; Schaeffer, J. M.; Yurke, B.; Louis, A. A.; Doye, J. P. K.; Winfree, E. Nucleic Acids Res. 2013, 41, 10641−10658.

Figures S1−S5, DNA sequences and modifications, and detailed experimental methods (PDF)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: fl[email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank the Natural Sciences and Engineering Research Council of Canada, the Canadian Institutes of Health Research, the Canada Research Chairs Program, and Alberta Health for financial support.

■

REFERENCES

(1) Jones, M. R.; Seeman, N. C.; Mirkin, C. A. Science 2015, 347, 1260901. (2) Aldaye, F. A.; Palmer, A. L.; Sleiman, H. F. Science 2008, 321, 1795−1799. (3) Liu, J.; Cao, Z.; Lu, Y. Chem. Rev. 2009, 109, 1948−1998. (4) Zhang, H. Q.; Li, F.; Dever, B.; Li, X.-F.; Le, X. C. Chem. Rev. 2013, 113, 2812−2841. (5) Landegren, U.; Vanelid, J.; Hammond, M.; Nong, R. Y.; Wu, D.; Ulleras̊, E.; Kamali-Moghaddam, M. Anal. Chem. 2012, 84, 1824− 1830. (6) Zhang, H.; Li, F.; Dever, B.; Wang, C.; Li, X.-F.; Le, X. C. Angew. Chem., Int. Ed. 2013, 52, 10698−10750. (7) Yang, X.; Tang, Y.; Alt, R. R.; Xie, X.; Li, F. Analyst 2016, 141, 3473. (8) Li, F.; Zhang, H.; Wang, Z.; Newbigging, M. A.; Reid, S. M.; Li, X.-F.; Le, X. C. Anal. Chem. 2015, 87, 274−292. (9) Sano, T.; Smith, C. L.; Cantor, C. R. Science 1992, 258, 120−122. (10) Fredriksson, S.; Gullberg, M.; Jarvius, J.; Olsson, C.; Pietras, K.; Gustafsdottir, S. M.; Ostman, A.; Landegren, U. Nat. Biotechnol. 2002, 20, 473−477. (11) Gullberg, M.; Gustafsdottir, S. M.; Schallmeiner, E.; Jarvius, J.; Bjarnegard, M.; Betsholtz, C.; Landegren, U.; Fredriksson, S. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 8420−8424. (12) Schallmeiner, E.; Oksanen, E.; Ericsson, O.; Spangberg, L.; Eriksson, S.; Stenman, U.-H.; Pettersson, K.; Landegren, U. Nat. Methods 2007, 4, 135−137. (13) Lundberg, M.; Eriksson, A.; Tran, B.; Assarsson, E.; Fredriksson, S. Nucleic Acids Res. 2011, 39, e102. (14) McGregor, L. M.; Gorin, D. J.; Dumelin, C. E.; Liu, D. R. J. Am. Chem. Soc. 2010, 132, 15522−15524. (15) Heyduk, E.; Dummit, B.; Chang, Y. H.; Heyduk, T. Anal. Chem. 2008, 80, 5152−5159. (16) Lass-Napiorkowska, A.; Heyduk, E.; Tian, L.; Heyduk, T. Anal. Chem. 2012, 84, 3382−3389. (17) Heyduk, E.; Heyduk, T. Anal. Chem. 2005, 77, 1147−1156. (18) Hu, J.; Wang, T.; Kim, J.; Shannon, C.; Easley, C. J. J. Am. Chem. Soc. 2012, 134, 7066−7072. (19) Hu, J.; Yu, Y.; Brooks, J. C.; Godwin, L. A.; Somasundaram, S.; Torabinejad, F.; Kim, J.; Shannon, C.; Easley, C. J. J. Am. Chem. Soc. 2014, 136, 8467−8474. (20) Ren, K.; Wu, J.; Yan, F.; Ju, H. Sci. Rep. 2014, 4, 4360. (21) Zhang, H.; Li, X.-F.; Le, X. C. Anal. Chem. 2012, 84, 877−884. (22) Zhang, H.; Li, F.; Li, X.-F.; Le, X. C. Methods 2013, 64, 322− 330. (23) Chen, J.; Deng, B.; Wu, P.; Li, F.; Li, X.-F.; Le, X. C.; Zhang, H.; Hou, X. Chem. Commun. 2016, 52, 1816−1819. (24) Li, F.; Zhang, H.; Wang, Z.; Li, X.; Li, X.-F.; Le, X. C. J. Am. Chem. Soc. 2013, 135, 2443−2446. (25) Li, F.; Zhang, H.; Lai, C.; Li, X.-F.; Le, X. C. Angew. Chem., Int. Ed. 2012, 51, 9317−9320. F

DOI: 10.1021/acs.analchem.6b01900 Anal. Chem. XXXX, XXX, XXX−XXX