pubs.acs.org/Langmuir © 2009 American Chemical Society
Circular DNA Logic Gates with Strand Displacement Cheng Zhang,* Jing Yang,† and Jin Xu* Institute of Software, School of Electronics Engineering and Computer Science, Key Laboratory of High Confidence Software Technologies, Ministry of Education, Peking University, Beijing, China, 100871. †This author contributed equally with the first author. Received August 23, 2009. Revised Manuscript Received November 24, 2009 Circular DNA logic gates were constructed on the basis of DNA three-way branch migration. In this logic system, circular DNA was used as a basic work unit and linear single-strand DNA was used as input and output signals. Making use of the circular structure, most of the DNA-specific recognition regions were designed in a single DNA ring. Depending on accurate DNA sequence recognition and highly effective strand displacement, the logic gates yielded correct results. In addition, the positions of gold nanoparticles (AuNPs) were detected as an alternative approach to determine logic results. Thus, the accurate and tunable control of DNA/AuNPs may be applied widely in DNA nanotechnology.
1. Introduction Logic gates, which form the basis of computing and logic operations, are the core components in conventional siliconbased computers. Not only used in computing, logic gates are also applied to information processing in chemical molecular systems. Normally, molecular logic operations are stimulated by physical and chemical input signals. Therefore, molecular logic gates have been constructed from various materials, such as nucleic acids, enzymes, and other small molecules.1-8 Particularly, nucleic acids are promising materials for the construction of synthetic chemical logic gates. Recent attempts to construct reconfigurable DNA nanodevices have been undertaken on the basis of well-developed technology of DNA branch migration.9 Not only delicate DNA nanodevices such as tweezers, bipedal legs, and boxes, but also logic circuits have been established for detecting chemical signal processing in vitro and researching applications for gene diagnostics.10-16 In previous work, most of the enzyme-free nucleic acid logic gates consisted of *To whom correspondence should be addressed. Email: zhangcheng369@ gmail.com;
[email protected]. Tel: þ86-10-62752366.
(1) Chen, X.; Wang, Y. F.; Liu, Q.; Zhang, Z. Z.; Fan, C. H.; He, L. Angew. Chem., Int. Ed. 2006, 45, 1759. (2) Niazov, T.; Baron, R.; Katz, E.; Lioubashevski, O.; Willner, I. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 17160. (3) Privman, M.; Tam, T. K.; Pita, M.; Katz, E. J. Am. Chem. Soc. 2009, 131, 1314. (4) Mu, L.; Shi, W.; She, G.; Chang, J. C.; Lee, S.-T. Angew. Chem., Int. Ed. 2009, 48, 1. (5) Amir, L.; Tam, T. K.; Pita, M.; Meijler, M. M.; Alfonta, L.; Katz, E. J. Am. Chem. Soc. 2009, 131, 826. (6) Privman, M.; Tam, T. K.; Pita, M.; Katz, E. J. Am. Chem. Soc. 2009, 131, 1314. (7) Miyoshi, D.; Inoue, M.; Sugimoto, N. Angew. Chem., Int. Ed. 2006, 45, 7716. (8) Szaciowski, K.; Macyk, W.; Stochel, G. J. Am. Chem. Soc. 2006, 128, 4550. (9) Panyutin, I. G.; Hsieh, P. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 2021–5. (10) Yurke, B.; Turberfield, A. J.; Jr, A. P. M.; Simmel, F. C.; Neumann, J. L. Nature 2000, 406, 605. (11) Omabegho, T.; Sha, R.; Seeman, N. C. Science 2009, 324, 67. (12) Andersen, E. S.; Dong, M.; Nielsen, M. M.; Jahn, K.; Subramani, R.; Mamdouh, W.; Golas, M. M.; Sander, B.; Stark, H.; Cristiano, L. P.; et al. Nature 2009, 459, 73. (13) Seelig, G.; Soloveichik, D.; Zhang, D. Y.; Winfree, E. Science 2006, 314, 1585. (14) Seelig, G.; Yurke, B.; Winfree, E. J. Am. Chem. Soc. 2006, 128, 12211. (15) Frezza, B. M.; Cockroft, S. L.; Ghadiri, M. R. J. Am. Chem. Soc. 2007, 129, 14875. (16) Shin, J. S.; Pierce, N. A. Nano Lett. 2004, 4, 905.
1416 DOI: 10.1021/la903137f
linear DNA strands. Winfree and co-workers constructed molecular logic gates with the characteristics of logic, cascading, restoration, and modularity.13 In multilevel systems, DNA logic gates were isolated on spatially separated solid surfaces to avoid cross-talk error signals.15 Moreover, a one-dimensional DNA array was developed to support independent and reversible memory.16 While these strategies are elegant, there is still room for improvement. In particular, most of them used short linear DNAs as basic structures, which have relatively monotonous displacement styles, thus constraining the construction of complicated hierarchical DNA scaffolds by subunit self-assembly. In addition, the conventional detection methods for DNA computing, such as PAGE (polyacrylamide gel electrophoresis) and fluorescence spectroscopy, may not be appropriate because they are laborious, time-consuming, and unsuitable for directly detecting subtle structures.17-20 In contrast to previous reports, we present a DNA logic gate system, using circular DNAs as the basic subunits and AuNPs for detection. The design of circular DNA gates has several features. First, the unique branch migration activity of circular DNA was implemented in logic computation. Second, on the basis of one DNA ring, one long DNA strand displaced both DNA strands from both directions. Third, similar to the linear DNA assemblies, the cross-linking of different circular DNA sub-elements can provide well-ordered scaffolds for secondary functional binding.21 For such hierarchical DNA scaffolds, nanoparticles, biomolecules, and dyes can be fixed under precise control.22 On the other hand, as a highly stable means of detection, AuNPs have been exploited as detectors for nucleic acids, proteins, and other small molecules.23-26 DNA/AuNP conjugates are already used to (17) Chakraborty, B.; Sha, R.; Seeman, N. C. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 17245. (18) Gu, H.; Chao, J.; Xiao, S.; Seeman, N. C. Nature Nanotechnol. 2009, 4, 245. (19) Goodman, R. P.; Schaap, I. A. T.; Tardin, C. F.; Erben, C. M.; Berry, R. M.; Schmidt, C. F.; Turberfield, A. J. Science 2005, 310, 1661. (20) Goodman, R. P.; Heilemann, M.; Doose, S.; Erben, C. M.; Kapanidis, A. N.; Turberfield, A. J. Nature Nanotechnol. 2008, 3, 93. (21) Wang Z. G.; Wilner O. I.; Itamar, W. Nano Lett. Online Early Access August 31, 2009; DOI: 10.1021/n1902317p. (22) Weizmann, Y.; Braunschweig, A. B.; Wilner, O. I.; Cheglakov, Z.; Willner, I. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 5289. (23) Storhoff, J. J.; Lucas, A. D.; Garimella, V.; Bao, Y. P.; Muller, U. R. Nat. Biotechnol. 2004, 22, 883. (24) Hazarika, P.; Ceyhan, B.; Niemeyer, C. M. Small 2005, 1, 844.
Published on Web 12/03/2009
Langmuir 2010, 26(3), 1416–1419
Zhang et al.
Letter
Figure 1. Mechanisms of circular DNA logic gate YES operation. The initial state is on the left of the arrows and the report states are shown on the right. Thick and thin line segments of the same color are complementary to each other. Oligonucleotide strand A1 is 63-mer; strand B is 19-mer; strand C is 23-mer. (a) Input strand A1 displaces both output strands B and C. (b) The fluorescence intensities are listed in a truth table.
construct reconfigurable nanostructures in a wide range of molecular analytical detection systems.27-30 Through transmission electron microscopy (TEM), with the advantages of high resolution and easy sample preparation, it is convenient to observe the real states and structures of DNA molecules by detecting the specific positions, unique sizes, and angle relationships of AuNPs. In this system, ssDNA (single-strand DNA) was used as input data and selectively released other ssDNA strands as output information. Making use of the circular structure, several DNA recognition regions were designed in one circular DNA molecule. DNA displacements occurred among three DNA molecules, and the report gates were integrated with all logic gates. In addition to the conventional detection methods, an alternative approach was developed to determine the results of DNA logic gates by detecting the positions of AuNPs.
2. Experimental Methods Materials. DNA strands, modified with disulfide, were purified by high-performance liquid chromatography. Unmodified DNA strands were purified by denaturing PAGE. All DNA strands were from Shanghai Sangon Co. Ltd. AuNPs (5 nm) were from Ted Pella Inc. Bis(p-sulfonatophenyl)phenylphosphine dihydrate dipotassium salt (BSPP) was from J&K Chemical Ltd. T4 ligase was from Takara Bio Inc. The sequences of the oligonucleotides are listed in the Supporting Information. Logic Gate Preparation. Logic gates were formed in a course of slow annealing. To avoid defective gates (e.g., DNA polymers), the desired products were cut out directly from PAGE and purified (Supporting Information Figure S1). With careful crushing, the gel powders were incubated in 0.5 TBE buffer for 12-24 h. Supernatant was collected after centrifugation at 8000 rpm for 10 min. The gate solutions were measured by optical absorbance at 260 nm and stored at 4 °C. Displacement of Input DNA Strands. Logic gates were formed and displaced in 0.5 TBE buffer, with NaCl at a final concentration of 100 mM. The DNA gates and input DNA strands (at specific concentrations) were added together and reacted for more than 6 h at room temperature. Then, the branch migration products were stored at 4 °C for detection. Displacement of AuNP/DNA Conjugates. In recent years, a method of DNA self-assembly has been developed using TAE/ Mg2þ buffer.17,18 However, in the presence of Mg2þ, AuNPs tend (25) Liu, J.; Lu, Y. Anal. Chem. 2004, 76, 1627. (26) Nam, J. M.; Thaxton, C. S.; Mirkin, C. A. Science 2003, 301, 1884. (27) Sharma, J.; Chhabra, R.; Liu, Y.; Ke, Y.; Yan, H. Angew. Chem., Int. Ed. 2006, 45, 730. (28) Aldaye, F. A.; Sleiman, H. F. J. Am. Chem. Soc. 2007, 129, 4130. (29) Tian, Y.; Mao, C. J. Am. Chem. Soc. 2004, 126, 11410. (30) Lee, I.-H.; Yang, K.-A.; Lee, J.-H.; Park, J.-Y.; Chai, Y. G.; et al. Nanotechnology 2008, 19, 395103.
Langmuir 2010, 26(3), 1416–1419
to aggregate in a disorganized manner.31 To avoid this, 0.5 TBE buffer was applied, with 100 mM NaCl. Monoconjugated AuNP/ DNA particles were prepared by incubating 5 nm AuNPs and 50 -thiolated ssDNA together overnight.32-37 To implement displacement, the gate sample (0.25 μM) was incubated with AuNP/ DNA conjugates 2C-1 and 2F-1 (0.3 μM) in a final volume of 100 μL for 10 h at room temperature. Fluorescence Experiments. Each logic gate sample was diluted from 100 μL to a final volume of 500 μL for detection, with 80 pmol gate strand and 100 pmol input strand. After adding input strands to react for 3 h or other specific intervals, the samples were excited at 559 nm with emission readings at 580 nm on a spectrometer (Hitachi F-4500). DNA displacements were implemented at 22 °C, in TAE/Mg2þ buffer (0.04 M Tris acetate, 1 mM EDTA, 12.5 mM Mg acetate, pH 8.3). TEM Analysis. Two to four microliters of DNA branch migration products were deposited on carbon-coated grids (400 mesh, Ted Pella). Then, the excess sample liquid was wicked using a piece of filter paper. The grid was washed with water or directly air-dried. TEM images were obtained using a Tecnai G20 transmission electron microscope.
3. Results and Discussion Each DNA logic gate was constructed with circular DNA (Figures 1-3). Strand A was a 63-mer phosphorylated oligonucleotide, containing partial recognition regions that were complementary with input strands A1, B1, and C1. Strand F was totally complementary with input strand F1. Logic computation was initiated by adding input strands to the gate solutions. In fluorescence experiments, strand B was marked with TAMRA fluorophore at the 50 end, and C was marked with BHQ quencher at the 30 end. After separation of fluorophore and quencher, the fluorescence intensity in gate solutions increased accordingly. The YES gate, a single-input gate, is the simplest logic device. In the YES gate system, the result value depends on the following relationship: input 1 produces output 1; input 0 produces output 0. Here, we used strand A1 to displace strands B and C (Figure 1; DNA sequence of A1 is a combination of both input strands B1 and C1). In this displacement, one long DNA strand A1 displaced (31) Williams, S. C. Patterning nanocrystals using DNA; University of California: Berkeley, 2003. (32) Loweth, C. J.; Caldwell, W. B.; Peng, X.; Alivisatos, A. P.; Schultz, P. G. Angew. Chem., Int. Ed. 1999, 38, 1808. (33) Claridge, S. A.; Goh, S. L.; Frchet, J. M. J.; Williams, S. C.; Micheel, C. M.; Alivisatos, A. P. Chem. Mater. 2005, 17, 1628. (34) Claridge, S. A.; Liang, H. W.; Basu, S. R.; Frechet, J. M. J.; Alivisatos, A. P. Nano Lett. 2008, 8, 1202. (35) Zanchet, D.; Micheel, C. M.; Parak, W. J.; Gerion, D.; Alivisatos, A. P. Nano Lett. 2001, 1, 32. (36) Hill, H. D.; Millstone, J. E.; Banholzer, M. J.; Mirkin, C. A. ACS Nano 2009, 2, 418. (37) Hurst, S. J.; Lytton-Jean, A. K. R.; Mirkin, C. A. Anal. Chem. 2006, 78, 8313.
DOI: 10.1021/la903137f
1417
Letter
Zhang et al.
Figure 2. Mechanisms of circular DNA logic gate OR operation. Oligonucleotide strand B1 is 29-mer; strand C1 is 34-mer. (a) Adding input strand B1. (b) Adding input strand C1. (c) Adding both input strands B1 and C1. (d) The fluorescence intensities are listed in a truth table.
Figure 3. Mechanisms of circular DNA logic gate AND operation. Oligonucleotide strands F and F1 are both 40-mer. (a) Adding input strand F1. (b) Adding input strand B1. (c) Adding both input strands F1 and B1. (d) The fluorescence intensities are listed in a truth table.
both DNA strands from both directions, and yielded a significant fluorescence signal (Figure 1b). In the OR logic system, the true result is produced only if either or both of two input values are true. In the computing course, addition of either input strands B1 or C1 displaced either output strands B or C, and fluorescence signals were produced (Figure 2a,b,d). In addition, when both inputs B1 and C1 were present, an effective output signal was detectable (Figure 2c,d). In the AND logic system, the true result is produced only when both of the input values are true. The AND gate (Figure 3) was implemented in three ways. First, addition of input strand F1 displaced only strand F, leaving strands B and C anchored to the DNA ring (Figure 3a). Since there was no separation between fluorophore and quencher, no fluorescence signal was produced. Moreover, if input strand B1 alone was added, the gate state did not change, as evidenced by no significant fluorescence signal, because there was no exposed toehold region (Figure 3b,d). Therefore, adding either F1 or B1 input strands did not result in an increase of fluorescence intensity. Second, both input strands F1 and B1 were added at the same time (Figure 3c). Initially, strand F1 occupied strand F by strand displacement. Then, the toehold regions for strands B1 and C1 on DNA ring A were exposed. Subsequently, the B1 strand displaced the B strand and caused the separation of quencher and fluorophore. The fluorescence signal produced suggested that such multilevel displacements operated effectively in the circular DNA logic gate. 1418 DOI: 10.1021/la903137f
In this experiment, although the F strand occupied the toehold region of A, some leakage still occurred (Figure 3b,d). In addition, kinetic experiments were performed to with the OR and AND gates to examine the time course of operation (Supporting Information, Figure S2). Compared with previous research,38 the extremely slow displacement rates found here may be caused by the lower DNA concentrations (0.16-0.2 μM). In this experiment, the positions of AuNPs were used to further reveal the structures of the logic gates. The DNA strands 2C, 2F, and B were individually designed as 18-mer, 16-mer, and 19-mer in length, which were complementary to certain regions of DNA ring A (Figure 4a). According to the design principles, input strands 2F-1 and 2C-1 were modified with disulfide at the 50 -ends and had a T5-ssDNA arm for flexibility. Here, the lengths of the two designed thiolated oligonucleotides were 59 bases, including long poly-T tails. This was because, if the length was shorter than 50-mer, it would be hard to discriminate among discrete AuNP/ DNA conjugates by agarose gel.31-35 After adding input strands 2F-1 and 2C-1 (anchored to AuNPs), the exposed toehold region on DNA ring A allowed for its specific removal of the two original short DNA strands and migration with the fully complementary input strands. Thus, when DNA displacement was completed, a pair of AuNPs were attached to each DNA ring. The experimental result was detected by TEM, and many pairs of AuNP assembly products were readily observed in grids (Figure 4b-d). (38) Li, Q; Luan, G; Guo, Q; Liang, J. Nucleic Acids Res. 2002, 30, E5.
Langmuir 2010, 26(3), 1416–1419
Zhang et al.
Letter
Figure 4. Analysis of AuNP/DNA particles after strand displacement. (a) Experiment schema. (b-f) Transmission electron micrographs of dimer, trimer, and polycluster nanoparticles. Scale bars 10, 20, and 50 nm. (g) Percentage of AuNP structures (dimer, trimer, and cluster) observed in TEM images.
The TEM images of the AuNP groups showed several nanostructures in the sample solution. In the images, many pairs of AuNPs were easily detected and the locations were in accordance with expectation. There were still some single nanoparticles, because of extra addition of 2C-1 or 2F-1 AuNP/DNA monoconjugates (Supporting Information Figures S5,6). However, trimers and polyclusters were also produced (Figure 4b,e,f). In the TEM images, the percentage of AuNP structures was counted (Figure 4g). The reason these nonspecific structures formed was that not all of the AuNP/DNA conjugates were monoconjugates. In the course of purification, some impure conjugates, with more DNA strands connected to one gold particle, were also obtained. In such cases, the extra DNA strands extended as bridges from one nanoparticle to another DNA ring and induced the formation of nonspecific structures. To avoid such structures, when the monoconjugate band was run into a glass fiber, the running time was limited to 5-7 min. Although some nonspecific structures were produced, it was still easy to detect DNA structures by AuNP position.
4. Conclusion In summary, we have constructed a logic gate system based on circular DNA, and demonstrate that the logic results are easily detected by fluorescence signals and assemblies of AuNPs. Although there are some drawbacks, like gate leakage and cluster aggregation of AuNPs, most results of the logic gates are correct
Langmuir 2010, 26(3), 1416–1419
and reliable. Work like this provides researchers with a solid experimental basis for additional gate design. This work not only represents a delicate and controllable nanoparticle assembly system, but also provides a means of carrying out logic operations by the precise positioning of nanoparticles. In addition, this device can be envisioned as a carrier, by which the DNA strands and AuNPs are released or attached under control. Since DNAlabeled nanoparticles can be assembled in an orderly way with other DNA structures, the accurate control of AuNP/DNA may be applicable to a wide range of nanoparticle-based technologies and DNA nanotechnology. Acknowledgment. This research was supported by the National Natural Science Foundation of China (nos. 60533010, 30670540, 60874036, and 60503002), the 863 Program of China (Grant No. 2006AA01Z104), the Ph.D. Programs Foundation of the Ministry of Education of China (no. 20070001020), and the Postdoctoral Science Foundation of China (no. 20060400344). The authors acknowledge Dr. Iain C. Bruce (Zhejiang University School of Medicine), Yinlin Sha (Peking University School of Basic Medical Sciences), and Shudong Wang (Shandong University of Science and Technology) for helpful discussions and revisions. Supporting Information Available: Supporting figures and structures. This material is available free of charge via the Internet at http://pubs.acs.org.
DOI: 10.1021/la903137f
1419
