Sequence-Modulated Interactions between Single Multivalent DNA

Apr 13, 2017 - From the binding dynamics analysis, it was found that the binding of nanoconjugates with DNA length longer than nine bases is kinetical...
0 downloads 13 Views 3MB Size
Download PDF
Article pubs.acs.org/ac

Sequence-Modulated Interactions between Single Multivalent DNAConjugated Gold Nanoparticles Chunyan Qiao,† Jia Wu,† Zhenrong Huang,† Xuan Cao,† Jiayu Liu,† Bin Xiong,*,† Yan He,*,†,‡ and Edward S. Yeung†,§ †

State Key Laboratory of Chemo/Biosensing and Chemometrics, College of Chemistry and Chemical Engineering, Hunan University, Changsha, 410082, P. R. China ‡ Department of Chemistry, Tsinghua University, Beijing, 100084, P. R. China § Department of Chemistry, Iowa State University, Ames, Iowa 50011, United States S Supporting Information *

ABSTRACT: DNA-conjugated gold nanoparticle (AuNP) is an attractive building block to construct elegant plasmonic nanomaterials by self-assembly but the complicated interaction between multivalent nanoconjugates governing the assembly process and the properties of assembled structures remains poorly understood. Herein, with an in situ kinetic single-particle imaging method, we report the dynamic interaction between single multivalent DNA-conjugated AuNPs quantitatively depends on the nucleic acid sequence in nanoconjugates. From the binding dynamics analysis, it was found that the binding of nanoconjugates with DNA length longer than nine bases is kinetically irreversible and the binding rate is dependent on both the sequence length and GC content, enabling us to predict the rational modulation of binding rates of individual building blocks for stepwise assembly. Moreover, the reversibility for the multivalent interaction between single nanoconjugates at constant temperature can be reinstated by adopting the DNA sequence with single-nucleotide mismatch and the lifetime for nanoconjugates at bound state can be tailored by changing the mismatch positions in DNA strands, providing new opportunity to create active nanostructures with controlled dynamic properties. All these findings provide new insights for understanding the multivalent interaction during the assembly process at the single-nanoconjugate level and predicting the programmable self-assembly of engineered nanoconjugates for the fabrication of dynamic nanomaterials. ntroduced by Mirkin’s group first, DNA-conjugated gold nanoparticles (AuNPs) have become versatile tools for broad applications from biochemical analysis to materials science.1−10 Especially, using DNA-conjugated AuNPs as building blocks, programmable self-assembly has shown great promise to construct sophisticated plasmonic nanomaterials at multiple dimensions.11,12 For these nanostructured materials, the assembly process and structural properties often depend on the multivalent interaction between individual nanoconjugates. Thus, illuminating the complicated interaction between single DNA-conjugated AuNPs is particularly important for understanding the mechanism governing the properties of assembled nanostructures and achieving controllable self-assembly of individual nanoconjugates. The kinetics and thermodynamics of DNA hybridization on the surface of AuNPs have been previously investigated,13−15 suggesting the 15-base DNA functionalized AuNPs approach the balance of optimized binding strength and sensitivity. The single-molecule imaging study also revealed the hybridization of free DNA molecules is reversible and the reaction equilibrium can be reached.16,17 These studies have greatly advanced the understanding and the exploitation of DNA hybridization for

I

© 2017 American Chemical Society

versatile applications but are unable to clarify the multivalent interaction between DNA-conjugated nanoparticles. For example, the hybridization/melting investigation suggested that both the thermodynamic stability and the structure evolution for the assembled nanoconjugates are dependent on the nucleic acid sequence.18−20 However, the question of how the multivalent interaction between single nanoconjugates are kinetically regulated by the nucleic acid sequence remains obscure, leading to great difficulty in rational modulation of the complicated interaction between polyvalent building blocks during programmable self-assembly. In this work, using an in situ real-time single-particle imaging strategy, we report that the dynamic multivalent interaction of DNA-conjugated AuNPs quantitatively correlates with the DNA sequences in individual nanoconjugates. By monitoring the binding process, it was found that the collective interaction makes the multivalent binding between DNA-conjugated Received: March 1, 2017 Accepted: April 13, 2017 Published: April 13, 2017 5592

DOI: 10.1021/acs.analchem.7b00763 Anal. Chem. 2017, 89, 5592−5597

Article

Analytical Chemistry

removing the AuNPs by centrifugation, the fluorescence intensity of the supernatant solution containing FAM-labeled DNA oligonucleotides was then measured. The loading number of DNA oligonucleotides on single AuNPs was calculated on the base of the concentration of DNA and AuNPs. In Situ Kinetic Imaging of the Interaction between DNA-Conjugated AuNPs. The binding of DNA-conjugated AuNPs was performed at room temperature with PBS as the hybridization buffer (0.125 M NaCl and 0.01 M phosphate buffer, pH 7.4). First, the probe DNA-functionalized AuNPs were adsorbed on the MPTMS-modified cover glass. Then a homemade flow channel was produced on this glass substrate. After the injection of the complementary nanoparticle-DNA conjugates into the channel, the interaction process was monitored with a darkfield microscope (Nikon 80i, Japan) that equipped with a darkfield condenser (NA, 1.20−1.43) and a 60× objective (with adjustable NA from 0.5 to 1.25). The binding kinetics was recorded with a color CCD camera (Olympus, DP72, Japan), and all the data in the darkfield images was extracted by using the ImageJ software. Theoretical Considerations for the Binding and Dissociation Dynamics. From the viewpoint of reaction kinetics, the binding process of two multivalent DNAconjugated AuNPs can be simplified as a second-order reaction,

AuNPs containing DNA strands longer than nine bases irreversible and the binding rate constant is dependent on both the sequence length and the GC content of DNA strands, allowing us to predict the rational modulation of binding rates for the stepwise assembly of building blocks. Moreover, the reversibility of the interaction between two multivalent nanoconjugates without heating can be reinstated by using DNA strands containing single-nucleotide mismatch and the lifetime for nanoconjugates at bound state can also be carefully regulated by designating the mismatch positions in DNA sequence, making it possible to create smart nanostructures with controlled dynamic properties. To the best of our knowledge, it is the first investigation on the quantitative dependence of binding and dissociation kinetics of multivalent interaction between individual DNA-conjugated AuNPs on the nucleic acid sequence. These findings could provide important information for understanding the multivalent interaction at the single-nanoparticle level and predicting the engineering of DNA-conjugated AuNPs in the fabrication of active plasmonic nanomaterials with controllable dynamic properties.

■

EXPERIMENTAL METHODS Chemicals and Materials. All the chemicals including HAuCl4·4H2O, Na3C6H5O7·2H2O, NaCl, NaH2PO4·2H2O, Na2HPO4·12H2O, mercaptoethanol, and hydroxylamine hydrochloride were purchased from Sinopharm Chemical (Shanghai, China). Bis(p-sulfonatophenyl)phenylphosphine (BSPP) and (3-mercaptopropyl)trimethoxysilane (MPTMS) were purchased from Sigma-Aldrich. All the DNA oligonucleotides with different length and GC contents were obtained from Sangong Biotech Co. Ltd. (Shanghai, China), and the sequences from 5′ to 3′ can be found in the Supporting Information (Tables S1−S4). Synthesis of AuNPs. The AuNPs used in this experiment were prepared via a seed mediated method.21 First, 18 nm AuNPs were synthesized according to the classical Fren’s method. Then, 1.432 mL of 18 nm AuNPs and 255 μL of hydroxylamine hydrochloride (400 mM) were gently mixed with 18 mL of water. Then, 386 μL of 24.28 mM HAuCl4 was gradually added into the mixture drop by drop in 30 min. The obtained AuNPs solution was stored in 4 °C for further use. The as-prepared AuNPs were characterized with UV−visible spectroscopy (Shimadzu, UV-1800 Japan) and transmission electron microscopy (JEM 1230, JEOL, Japan) Preparation and Characterization of DNA-Conjugated AuNPs. The as-prepared AuNPs (0.1 nM) were incubated with DNA at a DNA/AuNPs molar ratio of 6000 overnight at room temperature. Then the solution was mixed with 50 μL of 100 mM phosphate buffer (pH 7.4). After that, 7.6 μL of 2 M NaCl was added to the mixture every 30 min (5 times in total) until the final NaCl concentration reached 125 mM and the incubation was kept undisturbed overnight. The prepared DNA-conjugated AuNPs were then washed three times with 10 mM phosphate buffer (pH 7.4) and stored at 4 °C before use. The measurements by UV−visible spectroscopy and dynamic light scattering (Nano ZS, Malvern, U.K.) were performed for the characterization of DNA oligonucleotides modification. To measure the loading number of DNA oligonucleotides on single AuNPs, the carboxyfluorescein (FAM) labeled DNA oligonucleotides with different sequence lengths (recognition units from 9 bases to 15 bases), and GC contents (36−73%) were used to prepare the nanoconjugates. After incubating the obtained nanoconjugates with excessive mercaptoethanol and

dCAB = kon·(CA,0 − CAB) ·(C B,0 − CAB) dt

(1)

where C A,0 and C B,0 are the concentration of initial concentration of nanoconjugates on substrate and complement nanoconjugates in solution, CAB is the concentration of bound nanoconjugates at a given time (t), and kon is the apparent binding rate constant. By integrating with the boundary conditions, including CAB = 0 at t = 0 and CAB = CA,0 at the end of reaction, eq 1 gives the following results, C B,0

C B,0 − CAB CA,0 1 ·ln · = kon·t − CA,0 CA,0 − CAB C B,0

CAB =

(2)

C B,0·CA,0·(1 − exp[(C B,0 − CA,0) ·kon·t ]) CA,0 − C B,0·exp[(C B,0 − CA,0) ·kon·t ]

(3)

Thus, the binding ratio as a function of time can be described as, Pb =

C B,0 − C B,0·e(CB,0− CA,0)·kon·t CAB = CA,0 CA,0 − C B,0·e(CB,0− CA,0)·kon·t

(4)

By fitting the measured binding ratio curve with above equation, the binding rate constant can be extracted. By treating the dissociation of two bound nanoconjugates as a first-order reaction,

dCAB = −koff ·CAB dt

(5)

where the CAB is the concentration of nanoconjugates at bound state and koff is the dissociation rate constant. By using the initial concentration of bound nanoconjugates (CAB,0), the solution to the above equation can be given by CAB = CAB,0·e−koff ·t

(6)

Hence, the dissociation probability (Pd) as a function of time (t) can be described as 5593

DOI: 10.1021/acs.analchem.7b00763 Anal. Chem. 2017, 89, 5592−5597

Article

Analytical Chemistry Pd =

CAB = e−koff ·t CAB,0

induced signal changes separately under darkfield microscopy. The results showed that the nonspecific interaction of two DNA-conjugated AuNPs containing totally mismatched DNA strands are rather weak, only resulting in slight variations both in color and intensity for the scattering spots, whereas the specific binding of two DNA-conjugated AuNPs can induce notable changes in color and intensity, making them readily distinguishable (Figure 1a,b and Figures S3 and S4).

(7)

On the base of eq 7, the dissociation rate constant can be obtained by fitting to the measured time-dependent probability curve. Therefore, the mean lifetime (τ) for bound nanoconjugates can be calculated by τ = 1/koff.

■

RESULTS AND DISCUSSION Thanks to the near-filed interaction of two proximal plasmonic nanoparticles,22,23 the binding process of two DNA-conjugated AuNPs can be visualized under darkfield microscopy by tracking the plasmon coupling induced color and intensity changes. The prepared AuNPs were monodispersed with a diameter of ∼42 nm (Figure S1a). For all the DNA strands used to modify the prepared AuNPs in this study (Tables S1− S4), the first 10 nucleotides (A10) of the DNA strands serve as a linker that can make the oligonucleotides attached onto the surface of AuNPs, and the rest of the nucleotides act as the recognition unit to bind with the corresponding target sequence. Following a reported procedure,6 the DNAconjugated AuNPs were prepared and then characterized with UV−visible spectroscopy and DLS measurements (Figure S1b and Table S5), all of which indicate the success in the preparation of DNA-conjugated AuNPs. The loading number of DNA oligonucleotides on single AuNPs was ∼1200 for different types of nanoconjugates prepared in this experiment (Figure S2), suggesting that the surface density of DNA strands for all these multivalent nanoconjugates used are comparable and the DNA density induced difference in binding dynamics can be ignored in this study. Because of the rapid diffusion and uncontrollable binding of freely moving nanoconjugates, it is quite difficult to image the whole binding process of two DNA-conjugated AuNPs in bulk solutions. To address this problem, the DNA-conjugated AuNPs were first fixed onto the surface of a thiol-modified glass substrate. Then, the binding process can be investigated readily after adding the complementary DNA-conjugated AuNPs onto the surface of substrate with the decoration of DNA-conjugated AuNPs (Scheme 1). To study the interaction between individual nanoconjugates on the surface of a substrate by in situ single-particle imaging, the first issue is to discriminate the specific binding from nonspecific adsorption of DNA-conjugated AuNPs. To do this, we examined the specific binding and nonspecific absorption

Figure 1. In situ investigation of interaction of multivalent DNAconjugated AuNPs by monitoring the color and intensity changes under darkfield microscopy. (a and d) Nonspecific interaction between totally mismatched DNA-AuNP conjugates, (b and e) specific binding between complementary DNA-conjugated AuNPs, and (c and f) reversible interaction between DNA-conjugated AuNPs with singlenucleotide mismatch.

Interestingly, because of the collective interaction enhanced thermodynamic stability, two DNA-conjugated AuNPs become bound state irreversibly after effective collision (Movies S1 and S2), being different from the thermodynamic equilibrium approached by the hybridization of free DNA strands in the previous literature.16,17 Nevertheless, the reversibility of the multivalent interaction can be recovered by using DNAconjugated AuNPs containing single-nucleotide mismatch, where the binding and dissociation can be verified according to the clear color and intensity changes (Figure 1c,f and Figure S5). The remarkably different binding behaviors induced by slight sequence changes suggest that the multivalent interaction between individual DNA-conjugated AuNPs can be effectively regulated by tailoring the sequence of DNA strands in the nanoconjugates. Moreover, the binding behavior might be used as a promising indicator for the discrimination of single nucleotide mutations. Previous investigations suggested that the binding strength of DNA-nanoparticle conjugates often depends on the nucleic acid sequence in individual nanoconjugates.18,24 However, how the binding dynamics of two single nanoconjugates kinetically correlates with the sequence length as well as GC content has not been elucidated. By using DNA strands with a recognition unit from 9 nucleotides to 15 nucleotides (all these sequences with a constant GC content at ∼55%), we first studied the impact of sequence length on the binding kinetics of DNAconjugated AuNPs. From the representative darkfield images during the interaction processes (Figure 2a−d), the amount of bound nanoconjugates that can be detected under darkfield microscopy at a given time clearly decreased with the extension of recognition sequence in the DNA strands from 9 base to 15 base, suggesting the binding rate is strongly related to the sequence length. By counting the bound nanoconjugates at different time, the time-dependent binding ratio of nanoconjugates can be obtained (Figure 2e). From the viewpoint of reaction kinetics, the binding ratio for the nanoconjugates on a

Scheme 1. Schematic Illustration for in Situ Imaging the Irreversible and Reversible Interaction Kinetics of Multivalent DNA-Conjugated AuNPs under Darkfield Microscopy

5594

DOI: 10.1021/acs.analchem.7b00763 Anal. Chem. 2017, 89, 5592−5597

Article

Analytical Chemistry

Figure 3. GC-content dependent binding kinetics: (a) time-dependent binding ratio curves for nanoconjugates with different GC contents in recognition units and (b) plot of apparent reaction rate constant as a function of GC content in DNA strand.

is proportional to the GC-content, which was further validated by the investigations launched in bulk-solution measurements and using DNA strands with different lengths (Figures S8− S10). This correlation is reasonable because the intermediate state formed by the two DNA-conjugated AuNPs with a GCrich strand is more stable, facilitating the generation of bound nanoconjugates. On the basis of the dependence of binding kinetics on sequence length and GC content for DNAconjugated AuNPs, the engineering of binding rate during programmable self-assembly can be predicted, making it possible to achieve stepwise assembly of different building blocks for controlled fabrication of specific plasmonic nanostructures. The analysis of binding process for DNA-conjugated AuNPs in the presence of single-nucleotide mismatches shows the reversibility of multivalent binding between individual nanoconjugates can be retrieved without the requirement of temperature elevation. The reversible multivalent interaction offers the opportunity to construct active nanomaterials with dynamic properties.27 However, one of the key issues that should be addressed is how to regulate the kinetic properties of these active nanomaterials. To achieve the regulation of reversibility of multivalent interaction, we set out to study the dissociation kinetics of DNA-conjugated AuNPs with singlenucleotide mismatch. The deduction of the dissociation probability as a function of time suggested the occurrence of dissociation events should be exponentially decayed over time (eq 7), where the mean lifetime at a bound state is determined by the dissociation rate constant. Since the dissociation of bound nanoconjugates is attributed to their thermodynamic instability, the lifetime at the bound state might be dependent on the mismatch position in the DNA sequence. To test this assumption, the dissociation kinetics of DNAconjugated AuNPs containing a single-nucleotide mismatch at different points in the DNA strands from the middle to the distal end was examined. By calculating the time interval between binding and dissociation of individual nanoconjugates (Figures S11−S14), the exponentially decayed distributions of lifetime was obtained (Figure 4a−d), which is consistent with the theoretical consideration. On the basis of the lifetime distribution for each case, the corresponding dissociation probability as a function of time was obtained. By fitting the dissociation probability curves, the characteristic dissociation rate constants at different mismatch points in the DNA strands can be extracted (Figure 4e−h). It is evident that subtle variation in mismatch position (e.g., one nucleotide) can induce a remarkable alternation of dissociation rate constant,

Figure 2. Sequence length dependent binding kinetics: (a−d) representative darkfield images from the interaction process of single nanoconjugates, (e) time-dependent binding ratio for nanoconjugates with recognition nucleotides from 9 bases to 15 bases, (f) plot for the apparent reaction rate constant as a function of sequence length of DNA strands in individual nanoconjugates.

substrate at a given time was determined by the binding rate constant (kon, eq 4). Therefore, the characteristic binding rate constants for nanoconjugates containing DNA strands with different lengths can be extracted by fitting the time-dependent binding ratio curves with the above expression. Figure 2f shows the correlation between the characteristic binding rate constant and the sequence length of DNA strands in nanoconjugates, implying that the nanoconjugates containing short DNA strands have faster binding rate than those containing long DNA sequences. This surprising kinetic property was further confirmed by bulk solution measurements by UV−visible spectroscopy (Figure S6). Both the single-particle and bulksolution investigations suggested the binding rate of DNAconjugated nanoparticles is proportional to the reverse of the sequence length, displaying an opposite trend for the hybridization kinetics of free DNA strands according to the previous literature.25 This is possibly because short DNA strands in nanoconjugates have higher sequence contact efficiency and require less time in the base-pairing interaction process.26 Next, we investigated the dependence of binding kinetics of DNA-conjugated AuNPs on GC-content of DNA strands with GC-content from 36% to 73% at a constant sequence length. Both the representative darkfield images from the interaction process and the time-dependent binding ratio curves indicated that DNA-conjugated AuNPs with GC-rich strands have larger binding rates (Figure S7 and Figure 3a). The obtained characteristic binding rate constant can be plotted as a function of GC content (Figure 3b), implying the binging rate constant 5595

DOI: 10.1021/acs.analchem.7b00763 Anal. Chem. 2017, 89, 5592−5597

Article

Analytical Chemistry

position dependent dissociation kinetics for DNA-conjugated AuNPs provide a new avenue to construct dynamic plasmonic nanostructures with controllable kinetic properties.

■

CONCLUSION In summary, we have reported the quantitative dependence of the interaction of multivalent DNA-conjugated AuNPs on the nucleic acid sequence by using an in situ kinetic single-particle imaging method. From the binding dynamics analysis, the binding between multivalent DNA-conjugated AuNPs with length longer than nine bases was found to be kinetically irreversible and the binding rate is proportional to the sequence length as well as GC content of DNA strands simultaneously. The reversibility for the multivalent interaction can be granted by adopting DNA sequences with single-nucleotide mismatch and the lifetime at bound state can be regulated by tuning the mismatch positions in the DNA strands. These findings not only provide significant information for improving the understanding of multivalent interaction at the single-nanoconjugate level and constructing dynamic plasmonic nanomaterials via programmable self-assembly of engineered DNAconjugated AuNPs but also hold the potential for applications in other fields such as single-nucleotide mutation analysis.

■

Figure 4. Mismatch point dependent dissociation dynamics. The distributions of lifetime for individual dissociation events for DNAconjugated AuNPs containing single mismatches at different sites in the recognition unit of the DNA strands, site 8 (a), site 9 (b), site 10, (c) and site 11 (d), respectively. Insert: the representative time-course of intensity changes with different lifetime for the four cases. The measured dissociation probability distributions and the fitted dissociation probability curves as a function of time (e−h).

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.7b00763. Tables for the DNA sequences used in this study (Tables S1−S4); additional results for the characterization of DNA-conjugated AuNPs (Figures S1 and S2 and Table S5); and investigations on the interaction dynamics (Figures S3−S16) (PDF) Movie for the effective collisions of DNA-conjugated AuNPs (AVI) Movies for the ineffective collisions of DNA-conjugated AuNPs (AVI)

suggesting the mean lifetime of bound nanoconjugates can be sensitively mediated by changing the mismatch positions in the DNA strands of nanoconjugates. The mismatch-position dependent dissociation kinetics was further validated by using nanoconjugates with another group of DNA sequences with single-nucleotide mismatches (Figures S15 and S16). On the basis of the obtained correlation, the averaged lifetime for the dynamic component in a smart nanostructure can be predicted and modulated by tuning its dissociation rates (Figure 5). Meanwhile, the dynamic range of mean lifetime can be extended by using DNA strands with recognition units at different lengths for specific applications. The mismatch-

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. ORCID

Yan He: 0000-0001-7133-0558 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors are grateful for the support from National Natural Science Foundation of China with Grant Numbers of 91027037, 21127009, 21425519, 21221003, and 21605045, Natural Science Foundation of Hunan Province Grant 13JJ1015, Hunan University 985 Fund, the fund of SKLCBSC and Fundamental Research Funds for the Central Universities (Hunan University, Grant 531107040869).

■

REFERENCES

(1) Elghanian, R.; Storhoff, J. J.; Mucic, R. C.; Letsinger, R. L.; Mirkin, C. A. Science 1997, 277, 1078−1081. (2) Mirkin, C. A.; Letsinger, R. L.; Mucic, R. C.; Storhoff, J. J. Nature 1996, 382, 607−609.

Figure 5. Plots of dissociation rate for single nanoconjugates (a) and calculated mean lifetime for bound nanoconjugates (b) as a function of mismatch position in the DNA strands. 5596

DOI: 10.1021/acs.analchem.7b00763 Anal. Chem. 2017, 89, 5592−5597

Article

Analytical Chemistry (3) Zagorovsky, K.; Chou, L. Y. T.; Chan, W. C. W. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 13600−13605. (4) Li, D.; Wieckowska, A.; Willner, I. Angew. Chem., Int. Ed. 2008, 47, 3927−3931. (5) Xiao, L.; Wei, L.; He, Y.; Yeung, E. S. Anal. Chem. 2010, 82, 6308−6314. (6) Zhu, D.; Song, P.; Shen, J.; Su, S.; Chao, J.; Aldalbahi, A.; Zhou, Z.; Song, S.; Fan, C.; Zuo, X.; Tian, Y.; Wang, L.; Pei, H. Anal. Chem. 2016, 88, 4949−4954. (7) Lan, X.; Lu, X.; Shen, C.; Ke, Y.; Ni, W.; Wang, Q. J. Am. Chem. Soc. 2015, 137, 457−462. (8) Pal, S.; Deng, Z.; Ding, B.; Yan, H.; Liu, Y. Angew. Chem., Int. Ed. 2010, 49, 2700−2704. (9) Lim, I. I. S.; Chandrachud, U.; Wang, L.; Gal, S.; Zhong, C.-J. Anal. Chem. 2008, 80, 6038−6044. (10) Li, Y.; Liu, Z.; Yu, G.; Jiang, W.; Mao, C. J. Am. Chem. Soc. 2015, 137, 4320−4323. (11) Lan, X.; Wang, Q. NPG Asia Mater. 2014, 6, e97. (12) Jones, M. R.; Seeman, N. C.; Mirkin, C. A. Science 2015, 347, 1260901. (13) Lytton-Jean, A. K. R.; Mirkin, C. A. J. Am. Chem. Soc. 2005, 127, 12754−12755. (14) Xu, J.; Craig, S. L. J. Am. Chem. Soc. 2005, 127, 13227−13231. (15) Chen, C.; Wang, W.; Ge, J.; Zhao, X. S. Nucleic Acids Res. 2009, 37, 3756−3765. (16) Johnson-Buck, A.; Su, X.; Giraldez, M. D.; Zhao, M.; Tewari, M.; Walter, N. G. Nat. Biotechnol. 2015, 33, 730−732. (17) Johnson-Buck, A.; Nangreave, J.; Jiang, S.; Yan, H.; Walter, N. G. Nano Lett. 2013, 13, 2754−2759. (18) Thaner, R. V.; Eryazici, I.; Macfarlane, R. J.; Brown, K. A.; Lee, B.; Nguyen, S. T.; Mirkin, C. A. J. Am. Chem. Soc. 2016, 138, 6119− 6122. (19) Hill, H. D.; Hurst, S. J.; Mirkin, C. A. Nano Lett. 2009, 9, 317− 321. (20) Macfarlane, R. J.; O’Brien, M. N.; Petrosko, S. H.; Mirkin, C. A. Angew. Chem., Int. Ed. 2013, 52, 5688−5698. (21) Peng, L.; Cao, X.; Xiong, B.; He, Y.; Yeung, E. S. Chem. Commun. 2016, 52, 7616−7619. (22) Xiong, B.; Huang, Z.; Zou, H.; Qiao, C.; He, Y.; Yeung, E. S. ACS Nano 2017, 11, 541−548. (23) Zhao, Y.; Cao, L.; Ouyang, J.; Wang, M.; Wang, K.; Xia, X.-H. Anal. Chem. 2013, 85, 1053−1057. (24) Seifpour, A.; Dahl, S. R.; Lin, B.; Jayaraman, A. Mol. Simul. 2013, 39, 741−753. (25) Wetmur, J. G.; Davidson, N. J. Mol. Biol. 1968, 31, 349−370. (26) Wetmur, J. G. Annu. Rev. Biophys. Bioeng. 1976, 5, 337−361. (27) Li, Z.; Zhu, Z.; Liu, W.; Zhou, Y.; Han, B.; Gao, Y.; Tang, Z. J. Am. Chem. Soc. 2012, 134, 3322−3325.

5597

DOI: 10.1021/acs.analchem.7b00763 Anal. Chem. 2017, 89, 5592−5597