This article presents a detailed protocol for T4 ligation and denaturing PAGE purification of small circular DNA molecules, annealing and native PAGE analysis of circular tiles, assembling and AFM imaging of 1D and 2D DNA nanostructures, as well as agarose gel electrophoresis and centrifugation purification of finite DNA nanostructures.
This article presents a detailed protocol for synthesis of small circular DNA molecules, annealing of circular DNA motifs, and construction of 1D and 2D DNA nanostructures. Over decades, the rapid development of DNA nanotechnology is attributed to the use of linear DNAs as the source materials. For example, the DAO (double crossover, antiparallel, odd half-turns) tile is well-known as a building block for construction of 2D DNA lattices; the core structure of DAO is made from two linear single-stranded (ss) oligonucleotides, like two ropes making a right hand granny knot. Herein, a new type of DNA tiles called cDAO (coupled DAO) are built using a small circular ss-DNA of c64nt or c84nt (circular 64 or 84 nucleotides) as the scaffold strand and several linear ss-DNAs as the staple strands. Perfect 1D and 2D nanostructures are assembled from cDAO tiles: infinite nanowires, nanospirals, nanotubes, nanoribbons; and finite nano-rectangles. Detailed protocols are described: 1) preparation by T4 ligase and purification by denaturing PAGE (polyacrylamide gel electrophoresis) of small circular oligonucleotides, 2) annealing of stable circular tiles, followed by native PAGE analysis, 3) assembling of infinite 1D nanowires, nanorings, nanospirals, infinite 2D lattices of nanotubes and nanoribbons, and finite 2D nano-rectangles, followed by AFM (Atomic Force Microscopy) imaging. The method is simple, robust, and affordable for most labs.
DNA molecules have been used to build many kinds of nanostructures over decades. Typical motifs include DAE (double crossover, antiparallel, even half-turns) and DAO tiles1,2,3, star tiles4,5,6,7, single stranded (ss) tiles8,9,10, and DNA origami11,12,13. These DNA motifs and lattices are assembled from linear ss-DNAs. Recently, others and we have reported the use of circular ss-oligonucleotides as scaffolds to build motifs, 1D nanotubes, and 2D lattices14,15,16,17. By inserting a Holliday junction (HJ)18,19,20,21 at the center of c64nt, a pair of two coupled DAO tiles can be formed17. This new cDAO motif and its derivatives are stable and rigid enough to assemble 2D DNA lattices up to 3 × 5 µm2. In this paper, we use a term of "circular tile", which is defined as a stable DNA complex molecule constructed with one circular scaffold and other linear staples of ss-oligonucleotides, and another term of "linear tile", which is built from a full set of linear ss-oligonucleotides.
This protocol demonstrates how to construct five kinds of DNA nanostructures with small circular DNA molecules as scaffolds: 1) infinite 1D c64nt and c84nt nanowires, 2) infinite 2D cDAO-c64nt-O and cDAO-c64nt-E (-O represents an odd number of 5 half-turns and -E represents an even number of 4 half-turns) lattices, 3) infinite 2D cDAO-c84nt-O and cDAO-c84nt-E lattices, 4) finite 2D 5 × 6 cDAO-c64nt-O and 5 × 6 cDAO-c74&84nt-O rectangles, 5) infinite 1D acDAO-c64nt-E nanorings and nanospirals (please refer to Figure 3-5 for the schematic drawings and images of the above five kinds of DNA nanostructures). The 1D c64nt and c84nt nanowires are assembled from each c64nt and c84nt scaffold associated with two linear staples respectively. Each circular tile of cDAO-c64nt, acDAO-c64nt, cDAO-c74nt, or cDAO-c84nt is annealed from its corresponding scaffold of c64nt, c74nt, or c84nt with four linear staples respectively. The infinite 2D lattices are assembled from the same type of two circular tiles with different sequences. The two finite 2D rectangle lattices are assembled from two sets of 32 circular sub-tiles respectively. To save money, only one-sequenced c64nt, c74nt, and c84nt is used as the respective scaffold while different overhangs are used to anneal the 32 cDAO-c64nt, 12 cDAO-c74nt, and 20 cDAO-c84nt circular sub-tiles respectively in the first sub-tile annealing step, then mix the corresponding 32 circular sub-tiles together and apply the second lattice annealing step to assemble the finite 5 × 6 cDAO-c64nt-O and 5 × 6 cDAO-c74&84nt-O lattices, respectively. Definitely, differently-sequenced circular scaffolds can be adopted to assemble a variety of finite size nanostructures, however it will cost more money and labors. The infinite 1D acDAO-c64nt-E nanorings and nanospirals are annealed from one-sequenced asymmetric acDAO-c64nt tiles with linear connections of an even number of 4 half-turns. There are two approaches to assemble infinite 2D lattices from circular tiles of cDAO-c64nt and cDAO-c84nt, which are distinguished by the intertile distances of an even number of 4 and an odd number of 5 half-turns respectively. The former requires all tiles to be aligned identically; the latter requires alternation of the faces of two neighboring tiles along the helical axes. If the tile is rigid and planar, such as cDAO-c64nt, both approaches will generate planar nanoribbons; if the tile is curved towards one direction, such as cDAO-c84nt, the intertile connection of an even number of 4 half turns will generate nanotubes, whereas the intertile connection of an odd number of 5 half turns will produce planar nanoribbons due to elimination of curvature-biased growth by alternate alignment of curved tiles. The successful assembly of 1D and 2D DNA nanostructures from circular tiles indicates several advantages of this new approach: enforced stability and rigidity of circular tiles over linear tiles, chiral tiles for assembly of asymmetrical nanostructures such as nanorings and nanoribbons, new visions on understanding the DNA mechanics and molecular structures, etc.
1. Preparation of Circular DNAs
2. Annealing of Assembly Solutions
3. Native PAGE Analysis
4. Purification of Finite Lattices
5. AFM Imaging
The circular DNA moves slightly slower than its precursor linear DNA in denaturing PAGE (Figure 2) because the pore inside the circular DNA is penetrated and retarded by gel fibers23,24,25. The correct ligation reaction efficiency for oligo-monomer cyclization depends on the substrate sequence and concentration, reaction temperature, time, etc. As the concentration of a precursor linear DNA is high enough at around 3.5 µM, the cyclization products of c64nt (or c84nt) and the precursor reference in this protocol can be directly shadowed as bands in the TLC plate under UV light without dying. If the bands of circular DNA are vague or invisible, indicating a failure of the ligation reaction or a much low product yield. Sometimes there are two bands above the linear precursor band, referring an additional oligo-dimer ring except for the correct oligo-monomer ring. Just leave the higher bands alone and collect the lower ones. The purified circular DNA products can be seen as white powder in the tube after vacuum drying. Except for the denaturing PAGE, the DNA purity can also be measured by UV spectrometer. The absorption peak of DNA is at 260 nm. Two standard criteria for the DNA purity are the absorption ratio of 280 nm/260 nm at 1.8 and that of 280 nm/230 nm commonly in the range of 2.0-2.2. If the above two ratios deviate from the standard values, the remains should be extracted again by following steps 1.18-1.20. The yields of c64nt and c84nt for the correct oligo-monomer cyclization are measured at the range of 30-60% according to this protocol.
Native PAGE analysis provides a lot of information about the motif's stability, purity, rigidity, assembly mode of monomer or polymer, etc (Figure 3). The c64nt assembly families of c64bp, HJ-c64nt, aHJ-c64nt, cDAO-c64nt, and acDAO-c64nt have only one clear and clean band for each assembly, representing they are stable monomer motifs. While the c84nt assembly families of HJ-c84nt, tHJ-c84nt, and cDAO-c84nt tiles have smears around their main bands, indicating minor by-products except for the target monomer motifs. Regardless of the minor byproducts of incorrect associates, excellent cDAO-c84nt-O (E) lattices with high yields can be assembled. To get a clear and clean electrophoresis image, the loading volume should be no more than 10 µL and the quantity of DNA should be 0.01~0.02 µg/mL.
The success of experiments is finally evaluated by 1D and 2D DNA nanostructures imaged by AFM (Figure 4 and Figure 5). Each assembly has its own morphological features in the micrometer scale such as nanowires, nanotubes, nanospirals, nanoribbons, etc. Moreover, the detailed textures of the DNA assemblies in the nanometer scale correlated to their theoretical circular tile sizes and organization modes very well respectively are the key for verification of successful and correct assembly. Therefore, both panoramic and high-resolution AFM images in micrometer and nanometer scales must be obtained. Choices of AFM methods and probes are crucial for the high-quality AFM images. The scan force should be adjusted as small as 50 pN26. If the scan force is too big, it would damage the DNA nanostructural patterns. The environmental cleanliness is another key parameter to get a clean and beautiful high-resolution AFM image in fluid. All buffers must be filtered by 0.22 µm filter; the probe holder and tweezers must be washed by detergent and rinsed by deionized water. If the environment is polluted by particulate debris, the probe tip would be damaged or clung by particles in the buffer, thus affecting the quality of AFM images.
Figure 1. Synthesis of circular DNA. The flow diagram represents how a long 5'-phosphorylated linear oligonucleotide in blue evolves to a circular DNA molecule. The two short red strands represent the splint oligonucleotides. Please click here to view a larger version of this figure.
Figure 2. Denaturing PAGE photograph of DNA cyclization products under UV light without dying. A precursor linear 64nt DNA band is in the far-left lane and its cyclization products of circular 64nt (c64nt) DNA appear as 9 bands at the same horizontal level in other 9 lanes. The 9 bands of c64nt will be cut off for abstracting circular DNAs. Please click here to view a larger version of this figure.
Figure 3. Native PAGE photographs after dying and schematic double-helical models of circular tiles. Both polymers of c64nt nanowire and c84nt nanowire are represented by their simplest folding unit cells, in which the two aligned dots above and below each unit cell indicate the infinite alignment of unit cells vertically up and down, with equal distances between duplexes to form nanowires. Monomer circular tiles of c64bp, c84bp, HJ-c64nt, aHJ-c64nt, HJ-c84nt, and tHJ-c84nt in A) have no protruding overhangs out of their rings, whereas cDAO-c64nt, acDAO-c64nt, and cDAO-c84nt in B) have both blunt-ended 10 bp overhangs respectively. For sequences, please refer to the Table of DNA Sequences. This figure has been modified from a previously published figure17. Please click here to view a larger version of this figure.
Figure 4. AFM images of typical 1D and 2D infinite DNA assemblies scanned in air. A) c64nt nanowire, B) infinite acDAO-c64nt-E, C) infinite cDAO-c64nt-E, D) infinite cDAO-c64nt-O, E) infinite cDAO-c84nt-E, and F) infinite cDAO-c84nt-O are annealed by sticky end cohesion. All texture details in these AFM images are in line with the tile sizes and organization modes very well. For sequences, please refer to the Table of DNA Sequences. This figure has been modified from a previously published figure17. Please click here to view a larger version of this figure.
Figure 5. AFM images of finite rectangle assemblies scanned in fluid. The finite rectangle assembly of A) 5 × 6 cDAO-c64nt-O is composed of 32 cDAO-c64nt sub-tiles, and B) 5 × 6 cDAO-c74&84nt-O is composed of 12 cDAO-c74nt and 20 cDAO-c84nt sub-tiles. For sequences, please refer to the Table of DNA Sequences. This figure has been modified from a previously published figure17. Please click here to view a larger version of this figure.
Table of DNA Sequences. Please click here to download this file.
The protocols presented in this article focus on the synthesis of small circular DNA molecules and the assembly of DNA nanostructures. Most of randomly-sequenced DNA designs can be used in this protocol. The purity of circular DNAs is critical for the success of DNA assemblies. The production yield of cyclization can be improved by lowering the concentration of 5′-phosphorylated linear DNA; however, this will increase the workload to produce the same amounts of circular DNAs. The length of splint DNA strands also affects the correct ligation reaction, it is optimized to be around 20 nucleotides long for both c64nt and c84nt.
An appropriate concentration of magnesium cations (e.g., 12.5 mM) in the solution during and after assembly is very important for the formation and maintenance of DNA nanostructures. Thus, a magnesium cation concentration of 12.5 mM is always kept during the processes of annealing, native PAGE, agarose gel electrophoresis and PEG buffer centrifugation purifications, AFM imaging, etc.
For finite rectangle assemblies of 5 × 6 cDAO-c64nt-O and 5 × 6 cDAO-c74&84nt-O, the agarose gel purification does not affect the texture details, while the PEG buffer centrifugation purification harms the texture details of DNA nanostructures in the high-resolution AFM images.
Benefit from the stability and rigidity of circular tiles, it is much easier to produce well-organized and large-size single crystalline 2D lattices from circular modules than from linear tiles and scaffolded origami9,11, although extra work is needed to synthesize and purify the circular DNA molecules. With only one circular DNA as the same core structure, many circular modules can be generated with different overhangs; by means of specific sticky end cohesions of overhangs finite nanostructures can be built; this strategy reduces the workload and cost of finite nanostructures. One significant advantage of circular DNA nanostructures is the resolution of secondary and tertiary structures of DNA molecules and their key elements such as Holliday junction from the texture details of single crystalline lattices. The 1D, 2D and 3D nanostructures built from circular modules and their potential applications in biology, medicine, and nano-engineering will become a new family member of DNA nanotechnology in the future.
The authors have nothing to disclose.
We are grateful for financial support from the NSFC (grants no. 91753134 and 21571100), and the State Key Laboratory of Bioelectronics of Southeast University.
T4 ligase | TaKaRa | 2011A | |
T4 buffer | TaKaRa | 2011A | |
TE buffer | Sangon | B548106 | |
Thermo bottle | Thermos | SK-3000 | |
Thermo cycler | Bio Gener | GE4852T | |
Exonuclease I | TaKaRa | 2650A | |
Exonuclease I buffer | TaKaRa | 2650A | |
30% (w/v) Acryl/Bis solution (19:1) | Sangon | B546016 | |
TAE premix podwer | Sangon | B540023 | |
Mg(Ac)2·4H2O | Nanjing Chemical Reagent | C0190550223 | |
Urea | Sangon | A510907 | |
TEMED | BBI | A100761 | |
Ammonium Persulfate | Nanjing Chemical Reagent | 13041920295 | |
Power supply | Beijing Liuyi | DYY-8C | |
Water bath | Sumsung | DK-S12 | |
Formamide | BBI | A100314 | |
DNA Marker (25~500 bp) | Sangon | B600303 | |
DNA Marker (100~3000 bp) | Sangon | B500347 | |
Loading buffer | Sangon | B548313 | |
PAGE electrophoresis systerm | Beijing Liuyi | 24DN | |
Filter | ASD | 5010-2225 | 0.22 µM |
UV imaging System | Tanon | 2500R | |
n-butanol | Sangon | A501800 | |
Absolute Ethanol | SCR | 10009257 | |
NaOAc | Nanjing Chemical Reagent | 12032610459 | |
Centrifuge | eppendorf | Centrifuge 5424R | |
Vacuum concentrator | CHRIST | RVC 2-18 | |
Ultraviolet spectrum | Allsheng | Nano-100 | |
nucleic acid stain | Biotium | 16G1010 | GelRed |
Agarose | Biowest | G-10 | |
Agarose electrophoresis systerm | Beijing Liuyi | DYCP-31CN | |
Heating Plate | Jiangsu Jintan | DB-1 | |
TBE premix podwer | Sangon | B540024 | |
filter column | Bio-Rad | 7326165 | Freeze 'N Squeeze column |
AFM | Bruker | Dimension FastScan | |
PEG8000 | BBI | A100159 | |
Mica | Ted Pella | BP50 | |
triangular AFM probe in air | Bruker | FastScan-C | |
triangular AFM probe in fulid | Bruker | ScanAsyst-fluid+ | |
DNA strands | Sangon |