Chenxiang Lin

Designer DNA Nanoarchitectures

Naturally existing biological systems, from the simplest unicellular diatom to the most sophisticated organ such as the human brain, are functional self-assembled architectures. Scientists have long been dreaming about building artificial nanostructures that can mimic such elegance in nature. Structural DNA nanotechnology, which uses DNA as a blueprint and building material to organize matter with nanometer precision, represents an appealing solution to this challenge. On the basis of the knowledge of helical DNA structure and Watson-Crick base pairing rules, scientists have constructed a number of DNA nanoarchitectures with a large variety of geometries, topologies, and periodicities with considerably high yields. Modified by functional groups, those DNA nanostructures can serve as scaffolds to control the positioning of other molecular species, which opens opportunities to study intermolecular synergies, such as protein-protein interactions, as well as to build artificial multicomponent nanomachines. In this review, we summarize the principle of DNA self-assembly, describe the exciting progress of structural DNA nanotechnology in recent years, and discuss the current frontier.

DNA‐Tile‐Directed Self‐Assembly of Quantum Dots into Two‐Dimensional Nanopatterns

Organizing nanoparticles (NPs) into rationally designed ensemble structures is of great scientific interest because architecturally defined collective properties from multiple NPs could lead to applications such as photonic antennas and controlled plasmonic interactions.[1] Recently, structural DNA nanotechnology has opened a new avenue for directed self-assembly of NPs [2] and other molecular species [3] into patterned nanostructures, taking advantage of the exciting progress in design and construction of artificial nanostructures with complex geometry or patterns via DNA self-assembly.[4] Among these, success of using DNA tile based nanostructure to organize NPs has only been limited to metallic gold NPs. To our knowledge, there has been no report demonstrating DNA tile directed self-assembly of semiconducting NPs (QDs) into rationally designed architectures, partly may be due to the significantly different surface properties of QDs and gold NPs. The difficulty of making QDs compatible to DNA tile based nanostructure has prohibited many interesting studies of multi-component NP photonic systems, e.g. distance dependent plasmonic quenching or enhancement between metallic NPs and QDs. Herein, we worked out a strategy to use two-dimensional DNA tile arrays to direct the assembly of streptavidin conjugated CdSe/ZnS core/shell QDs into well-defined periodic patterns. We anticipate that this first example of DNA tile based QD assembly would pave the way for controlling more sophisticated nanopatterns of QDs and beyond. The strategy and schematic process of the DNA tile directed QD assembly is illustrated in Figure 1. We used a set of four double crossover (DX) molecules, named the ABCD tile system,[5] as scaffolds for QD assembly. Each different DX tile (DX-A, DX-B, DX-C and DX-D) is shown by a different color in Fig. 1. The ‘A’ tile contains a short DNA stem protruding out of the tile plane that carries a biotin group (illustrated as yellow star) at the end (see supporting information for sequence information). Upon self-assembly, the four-tile system gives 2D arrays displaying parallel lines of biotin groups, with a periodic distance between two neighboring biotin lines ~ 64 nm, and a distance between two biotins within the line about 4–5 nm. After adding streptavidin coated CdSe/ZnS QDs (Invitrogen, QdotR 545 ITK streptavidin conjugate, or STV-QD) to the DNA array, streptavidin molecules (illustrated as yellow blocks) specifically bind to the biotin groups so that the QDs (illustrated as black ball) will be organized onto the DX tile arrays (see supporting information for detailed methods). Figure 1 Schematic showing the process of DNA tile directed self-assembly of QD arrays. AFM images and cross-section profiles shown in Figure 2 clearly demonstrate that STV-QDs bind specifically to the DX array and get organized into periodical stripes of QD arrays with the designed distances between the parallel lines. Due to the short stem on A tile carrying the biotin group, the patterned 2D array of the ABCD tile system alone (left image and blue trace) show a small height change across the line of the biotin groups, ~ 0.5 nm. When STV-QDs bind to the DX array (middle image and green trace), the average height across the biotin sites increases to ~ 9 nm. A control experiment by adding streptavidin to the same type of biotin modified DX arrays shows a height change of only ~ 2 nm (right image and red trace). The height change resulting from the STV-QD binding is significantly higher than that of streptavidin binding to the array, the line widths are also much wider, owning to much larger sizes of the STV-QDs (diameter estimated ~ 10 nm including the surface polymer and protein coating), compared to that of the protein molecules alone ~ 2–3 nm. Figure 2 a) From left to right: AFM images of DX-ABCD array alone with each A tile bearing a biotin; the biotinylated DX-ABCD array incubated with STV-QD conjugate; and same array incubated with streptavidin only; b) Cross section analysis of the AFM images. Each ... TEM imaging further reveals the patterning of QD arrays templated by the DX tiles. TEM images shown in Figure 3 represent the periodic alignment of QDs into parallel stripes with measured periodicity of ~ 64 nm between the lines, matched well with the designed parameters. The diameter of each individual QD particle is measured ~ 4 nm, corresponds to the size of CdSe/ZnS QDs with green emission. The protein and polymer coating on the surface of the QDs are not visible due to low electron density thus low contrast in the TEM image. It is notable the QDs within each stripe sometimes slightly shift out of the line, which may be due to the orientational flexibility of the short protruding DNA stems bearing the biotin group and the tendency of neighboring QDs to avoid steric crowdedness. The center to center distance between QDs within the same line measures from 7 to 15 nm, which is larger than the distance between neighboring biotin groups, 4–5 nm in the DX array. This can be explained by the large size of the protein coated QDs causing steric hindrance within the line. In addition, the multiple streptavidin molecules conjugated on each QD and the multivalency of streptavidin-biotin binding may also contribute to this effect. It is possible that multiple streptavidin molecules on a single QD may obtain orientations that allow for binding of two or three neighboring biotin groups in the same line. Taking the size of the QD with the protein and polymer coating on surface into account, this 7–15 nm distance is close to the highest possible packing density of the protein coated QDs in the line. More TEM and AFM images of larger sample areas are also included in the supplemental materials, demonstrating the fidelity of the successful assembly of the STV-QDs using the DNA tile scaffolds. Figure 3 a) TEM images of the periodic patterns of the organized QD arrays. b) High resolution TEM images with an insert in the right corner reveal the crystalline structure of the QDs. c) EDX spectrum verifies the composition of the CdSe/ZnS QDs. High resolution TEM images (insert in Fig. 3b) clearly reveals the crystalline structure of the QDs and the energy dispersive X-ray (EDX) data supports that the NPs in the image are composed of CdSe/ZnS, as shown in Figure 3c. In addition to AFM and TEM imaging, we further used laser fluorescence imaging and photo-bleaching experiments to demonstrate that the QDs were assembled onto the DX array. In this experiment, a DNA strand in the B tile of the DX array was modified with an organic fluorophore with red emission, Cy5 (λem = 648 nm). The DX arrays carrying both Cy5 and STV-QD of green emission (λem = 545 nm) was imaged by fluorescence imaging (Figure 4a), revealing the co-localization of red Cy5 and green QDs on the DNA array (see superimposed fluorescent image in Fig. 4a, rightmost panel). It is well-known that QDs have higher photostability than organic fluorophores. A rectangular shaped region of 11×15 μm2 was selected from the imaged area to be photobleached (Fig. 4b). This region was constantly irradiated by a focused 405 nm laser beam at the power of 0.9 mW for 81 s. Images were taken using the same sequential scanning set-up with 9 s intervals during the photobleaching process. The change of the relative intensity of the red and green fluorescence in the bleached region was also plotted in Figure 4b. It is clear that the organic fluorophore was photobleached with a 90% drop of intensity within 30 s, while the emission intensity of the green QDs still persists after 80 s. This experiment further indicates that QDs are successfully organized onto the DNA tile arrays. Figure 4 a) Confocal fluorescent microscope images and b) photobleaching on the DNA arrays. Scale bar: 20 μM. In summary, we have constructed well-aligned two-dimensional arrays of QDs with controlled periodicity by coupling DNA self-assembly with streptavidin coated QDs. As an elegant bottom-up method, DNA self-assembly has the inherent advantage in generating programmable nanostructures with rationally designed functionality and nanometer precision in addressability. Our work demonstrates the capability to direct QDs into designer nanoarchitectures, this opens up opportunities to construct discrete nanostructures of multi-component NP systems for energy and biosensing applications. It is worthy to point out that both the surface and bioconjugation chemistry of QDs are much more complicated than gold nanoparticles, this led us to optimize many important experimental parameters to achieve successful QD assembly on DNA tile arrays (see supporting information for further comments). We acknowledge that more robust bioconjugation chemistry on QDs to obtain DNA sequence coded QDs (desirable to have discrete number of DNA oligos displayed on the QD surface) is needed for sequence addressable organization of QDs into more sophisticated architectures. Combining QD assembly strategies with previous success of metallic NP organization, we expect that many new properties of the well-controlled multi-component nanophotonic structures will be revealed. Indeed, the organizational power of structural DNA nanotechnology demonstrated so far is only at its horizon.

Self-Assembled Combinatorial Encoding Nanoarrays for Multiplexed Biosensing

Multiplexed and sensitive detection of nucleic acids, proteins, or other molecules from a single solution and a small amount of sample is of great demand in biomarker profiling and disease diagnostics. Here we describe a new concept using combinatorial self-assembly of DNA nanotiles into micrometer-sized two-dimensional arrays that carry nucleic acid probes and barcoded fluorescent dyes to achieve multiplexed detection. We demonstrated the specificity and sensitivity of the arrays by detecting multiple DNA sequences and aptamer binding molecules. This DNA tile-array-based sensor platform can be constructed through DNA self-assembly. The attachment of different molecular probes can be achieved by simple DNA hybridization so bioconjugation is not necessary for the labeling. Accurate control of the interprobe distances and solution-based binding reactions ensures fast target binding kinetics.

Knitting complex weaves with DNA origami.

The past three decades have witnessed steady growth in our ability to harness DNA branched junctions as building blocks for programmable self-assembly of diverse supramolecular architectures. The DNA-origami method, which exploits the availability of long DNA sequences to template sophisticated nanostructures, has played a major role in extending this trend through the past few years. Today, two-dimensional and three-dimensional custom-shaped nanostructures comparable in mass to a small virus can be designed, assembled, and characterized with a prototyping cycle on the order of a couple of weeks.

In vivo cloning of artificial DNA nanostructures

Mimicking nature is both a key goal and a difficult challenge for the scientific enterprise. DNA, well known as the genetic-information carrier in nature, can be replicated efficiently in living cells. Today, despite the dramatic evolution of DNA nanotechnology, a versatile method that replicates artificial DNA nanostructures with complex secondary structures remains an appealing target. Previous success in replicating DNA nanostructures enzymatically in vitro suggests that a possible solution could be cloning these nanostructures by using viruses. Here, we report a system where a single-stranded DNA nanostructure (Holliday junction or paranemic cross-over DNA) is inserted into a phagemid, transformed into XL1-Blue cells and amplified in vivo in the presence of helper phages. High copy numbers of cloned nanostructures can be obtained readily by using standard molecular biology techniques. Correct replication is verified by a number of assays including nondenaturing PAGE, Ferguson analysis, endonuclease VII digestion, and hydroxyl radical autofootprinting. The simplicity, efficiency, and fidelity of nature are fully reflected in this system. UV-induced psoralen cross-linking is used to probe the secondary structure of the inserted junction in infected cells. Our data suggest the possible formation of the immobile four-arm junction in vivo.

A replicable tetrahedral nanostructure self-assembled from a single DNA strand.

We report the design and construction of a nanometer-sized tetrahedron from a single strand of DNA that is 286 nucleotides long. The formation of the tetrahedron was verified by restriction enzyme digestion, Ferguson analysis, and atomic force microscopy (AFM) imaging. We further demonstrate that synthesis of the tetrahedron can be easily scaled up through in vivo replication using standard molecular cloning techniques. We found that the in vivo replication efficiency of the tetrahedron is significantly higher in comparison to in vitro replication using rolling-circle amplification (RCA). Our results suggest that it is now possible to design and replicate increasingly complex, single-stranded DNA nanostructures in vivo.

Recovery of intact DNA nanostructures after agarose gel-based separation

To the editor: Molecular self-assembly using DNA as a structural building block has proven an efficient route for construction of nanoscale objects and arrays of ever increasing complexity1. An important catalyst for advancing the field in recent years has been the “scaffolded DNA origami” strategy, in which a long “scaffold strand” derived from a viral genome (M13) can be folded with hundreds of short synthetic “staple strands” into a variety of custom two- and three-dimensional shapes2,3. This technology is being used to develop molecular tools for applications in fields such as structural biology4, single-molecule biophysics, and drug delivery. Many of these applications require a homogenous sample of properly folded nanostructures greatly enriched over the misfolded intermediates and large aggregates characteristic of multilayer DNA-origami self-assembly. Agarose-gel electrophoresis currently provides the most effective method available for high-resolution separation of well-folded objects on this size scale, however extraction of DNA nanostructures intact with high yield from the agarose matrix is problematic. Existing methods rely on thermal, chemical, and/or mechanical destruction of the agarose gel, or else electroelution of the DNA to a solid support, leading to problems of low yield, damage to structures, and/or contamination with residual agarose. We modified a DNA electroelution method for recovery of DNA from a standard horizontal agarose-gel–electrophoresis apparatus to optimize it for efficient, high resolution, and scalable recovery of large and complex intact DNA nanostructures5,6. Initial attempts to purify our DNA nanostructures by electroelution revealed the need for a well sealed elution bed to eliminate high-conductivity buffer paths that served as escape routes for the nanostructures. To address this problem, we poured a 1–2% agarose resolving gel on top of a thinner and more rigid basement layer of 4% agarose previously set within the gel-casting tray (Supplementary Fig. 1 and Supplementary Methods). Once the sample had been sufficiently resolved on our dual-layer agarose system, an elution well was cut into the resolving gel directly in front of the band of interest and filled with a viscous solution of 30–50% sucrose. The elution well is simple to cut down to the interface with the 4% layer due to the difference in rigidity of the layers, and the seal between the layers adjacent to the elution well is not disturbed. To eliminate high conductivity paths in buffer above the gel we maintained the running buffer level even with, or below, the surface of the resolving gel. Elution of the band was achieved by electrophoresis of the sample into the sucrose bed where movement of the DNA is slowed enough to allow efficient recovery by ultraviolet detection and micropipetting. The identity of the elution buffer has profound consequences for the efficacy of purification. Using a 400 nm-long six-helix bundle nanostructure as a model to assess purification performance (Fig. 1a and Supplementary Table 1), we screened three solutes at varying concentrations. Use of glycerol or polyethylene glycol resulted in retarded migration of the DNA band and a slow elution time of 1–3 hours, with inconsistent recovery yields between 20% and 60% (Supplementary Fig. 2). The most efficient yields were obtained with solutions of 30%–50% sucrose (Supplementary Fig. 3). ImageJ analysis of the purified six-helix bundles indicated 71±3% of the well-folded structure could be recovered from the agarose matrix versus 15±5% by the pellet-pestle homogenization method7. Our analysis by negative-stain transmission electron microscopy (TEM) also indicated a strong enrichment of the properly folded structures. Figure 1 Agarose-gel analysis and transmission electron microscopy (TEM) of various DNA origami after gel purification To evaluate the compatibility of this purification method with other 3D nanostructures, we folded and purified three objects that reflect the range of complexity and the fragility of more elaborate shapes as well as a high level of heterogeneity for unpurified samples. The first shape was a twelve-helix bundle (Fig. 1b, Supplementary Fig. 4, and Supplementary Table 1), whose folding yields more aggregates than a six-helix bundle. Analysis of the purified twelve-helix bundles on agarose gel indicated that it was not possible to resolve the well-folded structure from aggregates via ion-exchange chromatography (Supplementary Fig. 5), however this separation was successful using agarose-gel–based separation. The second shape was a six-helix bundle bent into a circle (Fig. 1c, Supplementary Fig. 6, and Supplementary Table 1)8. The final object was a “tensegrity” structure (Fig. 1d and Supplementary Table 1)9. Purification and analysis of each structure by TEM and agarose-gel electrophoresis indicated enrichment of the properly folded structures and yields of 70%, 50% and 45% for the twelve-helix bundle, ring and tensegrity structure, respectively—values up to four-fold greater than achieved by the pellet-pestle homogenization method7. The use of 800 nm six-helix bundle heterodimers as an alignment medium for membrane-protein NMR experiments7 requires a relatively high degree of purity and nanotube integrity to achieve a liquid crystalline state. When purified via our agarose-gel method (Supplementary Fig. 7), the six-helix bundles not only dimerize appropriately, but also form high-quality liquid crystals (assayed using birefringence) indicating that the structures retained a high degree of structural integrity. A continued challenge in the field is the hierarchical construction of larger objects from individual nanostructure building blocks. Because individual components often fold with misfolded intermediates in the mixture, the probability of assembling a multimer free from defects becomes very low without prior purification of the components. Using a twelve-helix bundle designed to assemble into a tetramer, we demonstrated that if the individual components of a larger oligomerized structure are purified before super-assembly, then that super-assembly can proceed with minimal production of large aggregates (Supplementary Fig. 8). Previously reported methods of purification are incompatible with more fragile structures that span larger areas or volumes. For example, recovery and TEM detection of a double-cross tensegrity structure has been achieved only by application of our method (communication from Tim Liedl, data unpublished). In addition, we found in a few cases that the structural integrity of DNA nanostructures was better preserved when they were extracted using our electrophoresis method instead of the pellet-pestle homogenization method7 (Supplementary Fig. 9). With the method presented here for purifying and oligomerizing larger structures, more sophisticated three-dimensional DNA nanostructures and DNA liquid crystals should be achievable.

Purification of DNA-origami nanostructures by rate-zonal centrifugation

Most previously reported methods for purifying DNA-origami nanostructures rely on agarose-gel electrophoresis (AGE) for separation. Although AGE is routinely used to yield 0.1–1 µg purified DNA nanostructures, obtaining >100 µg of purified DNA-origami structure through AGE is typically laborious because of the post-electrophoresis extraction, desalting and concentration steps. Here, we present a readily scalable purification approach utilizing rate-zonal centrifugation, which provides comparable separation resolution as AGE. The DNA nanostructures remain in aqueous solution throughout the purification process. Therefore, the desired products are easily recovered with consistently high yield (40–80%) and without contaminants such as residual agarose gel or DNA intercalating dyes. Seven distinct three-dimensional DNA-origami constructs were purified at the scale of 0.1–100 µg (final yield) per centrifuge tube, showing the versatility of this method. Given the commercially available equipment for gradient mixing and fraction collection, this method should be amenable to automation and further scale up for preparation of larger amounts (e.g. milligram quantities) of DNA nanostructures.

Self-assembly of size-controlled liposomes on DNA nanotemplates

Artificial lipid-bilayer membranes are valuable tools for the study of membrane structure and dynamics. For applications such as studying vesicular transport and drug delivery, there is a pressing need for artificial vesicles with controlled size. However, controlling vesicle size and shape with nanometer precision is challenging and approaches to achieve this can be heavily affected by lipid composition. Here we present a bio-inspired templating method to generate highly monodispersed sub-100nm unilamellar vesicles, where liposome self-assembly was nucleated and confined inside rigid DNA nanotemplates. Using this method we produced homogenous liposomes with four distinct pre-defined sizes. We also show that the method can be used with a variety of lipid compositions and probed the mechanism of the templated liposome formation by capturing key intermediates during membrane self-assembly. The DNA nanotemplating strategy represents a conceptually novel way to guide the lipid bilayer formation, and could be generalized to engineer complex membrane/protein structures with nanoscale precision.

Mirror image DNA nanostructures for chiral supramolecular assemblies.

L-DNA, the mirror image of natural D-DNA, can be readily self-assembled into designer discrete or periodic nanostructures. The assembly products are characterized by polyacrylamide gel electrophoresis, circular dichroism spectrum, atomic force microscope, and fluorescence microscope. We found that the use of enantiomer DNA as building material leads to the formation of DNA supramolecules with opposite chirality. Therefore, the L-DNA self-assembly is a substantial complement to the structural DNA nanotechnology. Moreover, the L-DNA architectures feature superior nuclease resistance thus are appealing for in vivo medical applications.