3.5a).
This dynamic feature revealed by HS‐AFM provided a clue that allowed a chequerboard‐like pattern to be derived from the lattice structure via sequential self‐assembly (Figure 3.5b). To realize this pattern derivation, a two‐dimensional lattice wherein every other cavity has polyT strands was first self‐assembled from two types of cross‐shaped DNA origamis. Then, the square origami carrying polyA strands at its four corners was loaded onto the preassembled lattice. The polyA‐modified squares that entered correct positions (cavities with 8T strands) could be docked in the cavity by sticky‐ended cohesion despite insufficient adsorption onto the membrane surface, whereas those that entered false positions (cavities without 8T linkers) could desorb from the cavity. Hence, the square origami would be finally incorporated only in the correct positions to make a chequerboard‐like pattern as demonstrated in HS‐AFM images (Figure 3.5c).
Figure 3.5 Placing square‐shaped DNA origami into lipid bilayer‐supported 2D lattices. (a) Dynamic docking events of square‐shaped origamis into the 2D lattice cavities recorded at a scan rate of 0.2 fps. Scale bar = 200 nm. (b) Schematic representation of the sequential self‐assembly and directed placement of square‐shaped origamis to produce a chequerboard‐like pattern. The streptavidin‐modified cross A and cross B were self‐assembled into a framework structure. Topographic AFM images of the assembled chequerboard‐like pattern is also shown. (c) Dynamic rearrangement of square‐shaped origami on the false positions (dashed circles) and docking into a correct position (dashed square). Edge reorganization was also occurred in the area enclosed in orange dotted lines. Scale bar = 200 nm.
Source: Suzuki et al. [67]/with permission from John Wiley & Sons, Inc.
3.7 Photostimulated Assembly and Disassembly
Similar to the construction of mechanical DNA nanodevices, the assembly into and disassembly from higher order structures can also be regulated by employing strand displacement reactions [68] or stimuli‐responsive oligonucleotides [69–71]. In the example shown in (Figure 3.6), photoresponsive oligonucleotides were used to control the dimerization of DNA origami [72]. The dimer consists of two hexagonal origami, Hx1 and Hx2, each of which carries four azobenzene‐modified oligonucleotides (Azo‐ODN1 for Hx1 and Azo‐ODN2 for Hx2) at one of its outer edges (d‐edge) and cholesterol moieties at its bottom face, allowing monomerization and dimerization on the bilayer membrane under UV and Vis light, respectively (Figure 3.6a,b).
Figure 3.6 Photoresponsive reversible assembly of a hexagonal DNA origami dimer. (a) Schematic of the hexagonal DNA origami (top view). Four azobenzene‐modified oligonucleotides (Azo‐ODNs) are introduced into the outer d‐edge. Each domain of the hexagon (a–f) carries one cholesterol‐TEG‐modified staple strand (red filled circle). Relative orientation of the hexagonal units in the linked structure was distinguishable by the positions of the small rectangle at the inner side of the e‐domain. (b) Photoresponsive monomerization and dimerization. Two monomers, Hx1 and Hx2, carry four Azo‐ODN1 strands and four Azo‐ODN2 strands, respectively, at their d‐edges. (c–e) Time‐lapsed AFM images of the photo‐controlled disassembly and assembly obtained at 1.0 frame/second. The area indicated by the dashed line in the first frame is enlarged and shown in the lower panels. (c) Disassembly of the hexagonal dimer upon UV irradiation. (d) Reassembly of the hexagonal dimer under visible light irradiation. (e) Disassembly event imaged upon second UV irradiation. Image size = 755 nm × 573 nm (upper panels) and 360 nm × 300 nm (lower panels).
Source: Suzuki et al [72]/with permission from American Chemical Society.
Figure 3.6c–e shows consecutive HS‐AFM images demonstrating the hexagonal DNA’s reversible photo‐dependent assembly and disassembly. The irradiation of the imaging area with UV (4 seconds) resulted in the disassembly of the dimer (Figure 3.6c). The left hexagon dissociated at 6 seconds and diffused away from the imaging area, while the right hexagon in the face‐up orientation stayed at almost the same position, suggesting it was held to the bilayer by cholesterol‐mediated anchoring. Reassembly was induced by subsequently irradiating the same area with visible light (Figure 3.6d). The diffusing monomer indicated by the arrow came into the imaging area at 53 seconds and made contact with the remaining hexagon. From 54 to 56 seconds, the two hexagons were seen to make contact each other. In the subsequent few frames, the angle between the two d‐edges was getting narrower. Then, at 60 seconds, the two hexagons were observed to be connected at the d‐edges, forming a dimer in which components are facing opposite directions, demonstrating the photo‐induced dimerization. Since collision of two hexagons at respective d‐edges is a prerequisite for this dimerization, imaging of the reassembly process required a longer observation time to catch such an occasional event.
The second round of the UV irradiation from 64 seconds again resulted in the dissociation into two monomers (Figure 3.6e). Similar to the result of the first UV irradiation, the right hexagon (hexagon in the face‐up orientation) stayed at the almost same position, while the other one (in the face‐down orientation) was relatively mobile on the surface and diffused away. This set of results clearly demonstrate photo‐controlled monomerization and dimerization at the single‐molecule level.
3.8 Conclusions and Perspectives
This chapter provided an overview of how time‐lapse AFM techniques have been applied to study mechanically functional DNA origami nanodevices and 2D self‐assembly of DNA origami arrays. Most nanodevices introduced here are designed to respond ions and photostimuli. In addition to the development of these stimuli‐responsive DNA origami structures, attempts to drive DNA nanodevices by electric [73, 74] or magnetic fields [75] are also progressing. Many of these devices are composed of a stator part that is fixed onto a glass surface and a movable arm whose orientation or angle against the stator is controlled by electric or magnetic fields. Rotational or hinge‐like movements of arms are generally monitored using a single‐molecule fluorescence imaging technique, such as a total internal reflection fluorescence (TIRF) microscopy. However, the observed behavior of the fluorescent spot does not always provide direct information on how the entire single nanodevices actually behave. Therefore, the next advancement of this technology would grow out of the integration of HS‐AFM and fluorescence imaging techniques. This direction has now progressed from possibility to actuality thanks to the emergence of HS‐AFM combined with various fluorescence microscopies, such as inverted fluorescence microscopy [76], confocal laser scanning microscopy [77], and TIRF microscopy [78]. It is hoped that both structural changes in individual nanodevices or morphological changes of self‐assembly systems will be correlated with nano‐to‐meso scale dynamics of their components. The combination of HS‐AFM with microfluidic devices and those that exploit electric magnetic manipulation is also a fascinating
