Constructing novel materials with DNA

ISSN:1748 0132 © Elsevier Ltd 2007

Constructing novel materials with DNADNA, apart from being a natural biological information carrier, has also been recognized as a useful building material in the field of nanotechnology. Its miniature scale, geometric properties, and molecular recognition capacity make DNA an appealing candidate for the construction of novel nanomaterials. Here we summarize the latest developments and describe the challenges and emerging applications of this field.

Thom H. LaBean1* and Hanying Li2

1*Departments of Computer Science and Chemistry, Duke University, Durham, NC 27708, USA

2*Department of Pathology, Duke University Medical Center, Durham, NC 27710, USA

E-mail: [email protected]

Although the detailed structure of DNA was revealed by Watson

and Crick1,2 back in 1953, even today we continue to discover

stunning and useful new structural modes for this versatile

macromolecule. Taking lessons from its in vivo role and aided by

technological advances, nanoengineers have begun to explore

novel and creative uses for DNA including: molecular detection3,

therapeutic regimens4, complex nanodevices5, nanomechanical

actuators and motors6-8, directed organic synthesis9,10, and

molecular computation11,12 . Excellent reviews of many of these

aspects of DNA can be found in this issue of Nano Today and

elsewhere10,12,13.

In this review, we will focus on the ‘materials’ side of DNA by

examining major architectural strategies (linear, branching, and

multibranched complexes) and application strategies (directed

organization of nanomaterials including biomolecules, templating

of inorganics, and approaches combining preformed and

templated materials) in which DNA nanotechnology plays a

starring role.

Architectural strategiesLinear DNA for conducting nanowiresThe use of DNA to form building blocks for nanoelectronic constructs

is quite promising, although for many years the conductivity of

bare DNA remained controversial. Electron transfer in neat DNA has

been observed in many cases14-17 but there are other experiments

suggesting that DNA might display semiconducting18,19, insulating20, or

even superconducting behavior21. Although coherent electron tunneling

and diffusive thermal hopping – the two most fundamental processes

for charge transfer – have been clearly demonstrated in DNA22,23,

the electron transfer behavior of bare DNA is typically insufficient for

nanoelectronic engineering purposes. However, DNA has been used as

a template upon which to organize more highly conductive materials

such as metals for electronic applications. Lee et al.24 described a new

form of DNA, M-DNA, in which the imino proton of the DNA base-

pairs is replaced by a Zn2+, Ni2+, or Co2+ ion (Fig. 1a). It has been

shown that M-DNA behaves like a molecular wire and has potential for

the development of future molecular electronics25-27.

APRIL 2007 | VOLUME 2 | NUMBER 226

NT202p26_35.indd 26NT202p26_35.indd 26 23/02/2007 08:38:2623/02/2007 08:38:26

mailto:[email protected]

Novel materials from DNA REVIEW

APRIL 2007 | VOLUME 2 | NUMBER 2 27

Other methods to enhance the conductivity of linear DNA have also

been explored. Braun et al.20 successfully constructed DNA templated

Ag nanowires by electroless deposition, producing nanowires ~100 nm

thick and 15 µm long. They attached two short oligonucleotides

to electrodes and introduced λ-phage DNA as a bridge (with ends

complementary to electrode-linked DNAs). Ag ions were then loaded

onto the DNA and reduced to form Ag nanoparticles (AgNPs) and

fine nanowires. The deposition of Pd28, Au29, and Pt30 on DNA have

also been investigated as a potential approach for creating conductive

nanowires. Monson et al.31 described the construction of DNA-

templated Cu nanowires. These wires were about 3 nm tall and may

prove useful in construction of single electron devices. Keren et al.32

recently reported protecting specific regions of DNA molecules from

metal deposition by associating proteins along selected sections of

the DNA. The ability to control metallization spatially provides an

important technological advantage for the assembly of functional

nanocircuits. Recently, Ag nanowires with widths down to 15 nm and

several microns in length (Fig. 1b) have been templated on various

DNA nanostructures and characterized electrically at room temperature

and low temperature33-36.

Linear DNA as smart glueSince the binding strength of DNA double-helices can be easily

controlled (by tuning helix length and base composition) and a huge

lexicon of unique sequences exist, linear DNA is extremely useful

as a structural linker for controlled aggregation of nanomaterials.

In back-to-back papers in 1996, the groups of Mirkin and Alivisatos

described two methods for assembling colloidal Au nanoparticles

(AuNPs) into aggregates using DNA as linkers37,38 (Fig. 2). The Mirkin

procedure uses particles labeled with multiple copies of the same

oligo, while the Alivisatos method requires an additional purification

step but produces defined conjugates with a single DNA oligo per

AuNP. Mirkin also constructed binary NP networks composed of 8 nm

and 31 nm Au particles by coating the two types of particles with

different 12-mer oligonucleotides (thiol labeled) via S-Au bonds39.

When a third DNA sequence (24-mer), complementary to both oligos

is added, hybridization leads to the association of particles. Different

types of assemblies can be formed by adjusting the ratio of small to

large particles (Fig. 2a). Electrical and optical properties of these DNA

AuNP assemblies were further elucidated in a series of subsequent

reports40,41 and various molecular detection applications have been

developed and reviewed recently42.

The single oligonucleotide label method, as described by Alivisatos

et al.38,43, has been used to produce AuNP dimers and trimers via

hybridization of NP-bound DNA molecules with complementary

template strands (Fig. 2b). This method allows the placement of Au

particles at specific locations on the template and greatly enhances

the control of assembly architecture. Two-dimensional assembly of

AuNPs on random DNA networks created from poly(dA-dT) has also

been reported44. DNA hybridization has been used to create a large

random network (up to 1 cm2) and also to bind AuNPs to the network.

As demonstrated by these and other studies, the molecular recognition

properties of DNA can be used to control interparticle distances,

connection strength, the size and chemical identity of the particles,

and other properties of the assemblies, making such DNA-NP

conjugates very promising building blocks for nanoscale materials

synthesis.

Fig. 1 (a) Schematic view of DNA double-helix and M-DNA, where metal

ions are bound and stacked into the DNA base pairs. (b) Scanning electron

microscopy images of λ-DNA and synthetic dsDNA templated Ag nanowires.

(Reprinted with permission from33. © 2006 American Institute of Physics.)

Fig. 2 DNA linked NP assemblies. (a) Binary AuNP network. In the presence

of complementary target DNA, two types of oligonucleotide-functionalized

AuNPs (8 nm and 30 nm) aggregate, as shown in the transmission electron

microscopy (TEM) image at the bottom. The inset shows a satellite structure

formed when the ratio of 8 nm to 30 nm is large. (b) Schematic illustration and

TEM images of 5 nm and 10 nm AuNP templated along template DNA strand.

[Part (a) reprinted with permission from116. © 2000 American Chemical

Society. Part (b) reprinted with permission from43. © 1999 Wiley-VCH.]

(b)

(a)

(b)(a)

NT202p26_35.indd 27NT202p26_35.indd 27 23/02/2007 08:38:3823/02/2007 08:38:38

REVIEW Novel materials from DNA


Branching DNA motifsLinear DNA can be used to assemble a range of different structures,

however, in order to make more diverse constructs, branching DNA

units are highly useful. Substantial progress has been made recently

in designing branching structures from DNA. In 1997, Shchepinov

et al.45 reported the synthesis of DNA dendrimers using synthon, a

novel phosphoramidite. By chemically connecting DNA to different

synthon molecules, dendrimers with a branching backbone and

varying numbers of DNA arms were synthesized46. The development

of DNA block copolymers was also described47,48, in which DNA was

covalently attached to an organic polymer backbone and provided

bridges for assembly and subsequent chemical modification. Recently,

Luo and coworkers49 demonstrated well-defined dendrimer-like

structures assembled via sticky end cohesion from DNA units with

branched secondary structure (Fig. 3a). This study demonstrates

that nearly-monodisperse dendrimeric DNA nanostructures can be

synthesized in a highly controlled fashion with relatively high yield

and purity. More recently, the construction of three-dimensional

hydrogels made entirely from flexible branched DNA building blocks

has been reported50. X-, Y-, and T-shaped DNA units were hybridized

and crosslinked with each other via ligase catalyzed assembly. The size

and shape of these large, three-dimensional hydrogels can be easily

controlled by using different molds (Fig. 3b). Novel soft materials

created from DNA hydrogels may find future applications in cell and

tissue culture, drug delivery, and cell-free protein synthesis.

Complex DNA motifs for structural building blocksApart from playing the role of smart glue to facilitate the assembly

of other molecules, DNA itself can be used to form rigid building

blocks for the construction of complex nanostructures. Seeman has

referred to this strategy as ‘bricks plus mortar’ because DNA itself

makes up the building blocks and the cement holding them together.

The obvious advantage is that it is easier to define solution conditions

in which DNA is soluble and well-behaved than conditions under

which both DNA and metallic NPs (for example) are soluble and

stable. Seeman and coworkers were the first to exploit DNA’s self-

complementarity for construction of novel nanostructures51. Since

the simple double-helix lacks the complexity needed for forming

tightly controlled two- and three-dimensional structures, they sought

to design more complex building blocks. They succeeded in making

branched junction motifs with four double-helical arms, which resemble

Holliday junctions, a natural conformation of DNA found in biological

homologous recombination complexes. In theory, these branched

junction units should assemble into a quadrilateral lattice by sticky end

cohesion52-54, yet in solution, the junctions (illustrated in Fig. 4) do

not assemble into a two-dimensional lattice because the structure is

Fig. 3 DNA dendrimer and hydrogel. (a) Dendrimer-like DNA (DL-DNA) formed by the ligation of Y-shaped DNA. The scheme in the middle shows higher generation

DL-DNA, which corresponds to the AFM image on the right. Scale bar corresponds to 100 nm. (b) Left and middle: schematic view of Y-shaped DNA monomer and

three-dimensional DNA hydrogels. (Right) DNA hydrogels built from X-shaped monomers patterned into different shapes. Scale bar corresponds to 1 cm. [Part (a)

reprinted with permission from49. © 2004 Nature Publishing Group. Part (b) reprinted with permission from50,117. © 2006 Nature Publishing Group.]

(b)

(a)

NT202p26_35.indd 28NT202p26_35.indd 28 23/02/2007 08:38:4023/02/2007 08:38:40



not stiff enough to hold the helical domains in the same plane and

instead they twist by ~60° (Fig. 4a). Seeman’s group had previously

reported the construction of a closed DNA cube55 and a truncated

octahedron56,57 (Fig. 4b) albeit at very low yields. Interestingly, the

four-arm branch motif can be stiffened when combined in pairs

and larger constructs. Mao et al.58 have fused four junctions into

a rhombus-like building block and successfully demonstrated the

further assembly into two-dimensional lattices (Fig. 5e). Turberfield

recently reported the use of RuvA, a Holliday junction binding protein,

to control the conformation and facilitate the assembly of square

lattices59.

Based on the idea of employing immobile DNA junctions, a large

number of distinct DNA building blocks (or tiles) have been designed

and experimentally implemented in the past two decades (Fig. 5).

Li et al.60 have reported the construction of double-crossover (DX)

complexes, which consist of two juxtaposed Holliday-junction-like

crossover motifs joined together by two double-helical domains.

Properly designed sticky ends further facilitate the assembly into

periodic one- and two-dimensional lattices (Fig. 5a,b)61,62. A related

motif combining a stem-loop hairpin with one of the duplex arms of

DX (known as DX+J) has also been reported61. The extra hairpin is

used as a topographical marker, visible by atomic force microscopy

(AFM)61,63. Seeman and coworkers also describe the generation of

paranemic crossovers (PX)64,65, which arise from fusion of two close,

parallel double-helices by reciprocal exchange at every possible contact

point. By controlling the interconversion between a PX junction and its

topoisomeric JX state, a robust DNA nanomechanical device has been

built and studied66,67.

A more complex planar building block was reported by LaBean

et al.68 in 2000. This triple crossover complex (TX) contained three

helices and four crossovers, and as in DX tiles, two adjacent helices

were connected by two four-arm junctions. Compared to DX tiles, TXs

provide larger space for gaps in two-dimensional arrays and further

extend the prototyping of useful branching building blocks (Fig. 5c).

Another DNA motif, the 4 x 4 cross-tile, consisting of four four-arm

junctions was reported in 2003 (Fig. 5e)34. Since the cross-tile has a

square aspect ratio and helix stacking in all four directions in the plane,

they can assemble into very large two-dimensional lattices. Structures

with triangular building blocks have also been developed. In 1998, DX

tiles were successfully ligated with DNA triangles, to create a unique

zigzag pattern69. Two triangle tile types have been prototyped, which

feature the formation of triangular and hexagonal patterns70,71.

These basic building blocks and their variants have been used

in the construction of self-assembled lattices. DX and TX lattices,

ribbons, and tubes have been observed. More complicated designs

including double-double crossover72, 4-, 8-, and 12-helix DNA tile

complexes have also been realized recently73. These tiles have also

been used for the assembly of planar and tubular structures. DNA tiles

that hold their helices in nonplanar domains have been prototyped

by several groups35,74,75, although attempts at using these tiles for

Fig. 4 (a) Schematic drawings of four-arm junctions and the proposed assembly of a two-dimensional lattice via sticky ends cohesion. The twist between the two

helical domains is illustrated in the bottom row. (b) DNA molecule with the connectivity of a cube and an octahedron. Each of the edges is composed of double-

helical DNA. [Part (a) adapted from54. Part (b) reprinted with permission from54. © 1998 Institute of Physics.]

(b)

(a)

NT202p26_35.indd 29NT202p26_35.indd 29 23/02/2007 08:38:4123/02/2007 08:38:41



three-dimensional lattice assembly have not yet succeeded. Uniform-

width nanoribbons, one-dimensional nanotracks, and two-dimensional

nanogrids displaying periodic square cavities have been synthesized

using a two-tile system (A and B cross-tiles)76. Stunning applications

of cross-tiles for the assembly of finite-size addressable arrays include

5 x 5 arrays77, molecular pegboards78, and 16-tile constructs79 (Fig. 6).

All these structures have a defined size and can be used to build up to

even more complex systems. The molecular-level control demonstrated

by these systems represents a major step toward developing DNA-

based controllable nanostructures. Potential uses include fixed-size

algorithmic assemblies for DNA computing and complex patterning of

nanomaterials for the fabrication of artificial bionanomachines, devices,

and sensors.

Despite the above examples, which demonstrate the great potential

of DNA-based self-assembly, important limitations exist, including

the fact that all component strands need to be in strict stoichiometric

ratios in order to achieve high assembly yields. Mao and coworkers

recently introduced the concept of ‘sequence symmetry’80. Instead of

using nine different strands, they used only three unique strands in the

4 x 4 cross-tile motif and successfully demonstrated corrugated two-

dimensional assembly. Symmetric DX tiles were also built81. Symmetric

designs have been shown to dramatically increase the yield and the

size of the fully-assembled structure by decreasing the proportion of

unproductive molecular collisions during both the tile assembly and the

lattice assembly stages of annealing.

Linear DNA scaffolds folded into complex motifs Scaffolded assembly (also known as nucleated assembly) uses a long

DNA strand as a molecular scaffold and many small staple strands to

bind and make crossovers on the scaffold, thus folding the scaffold

strand into an addressable shape that can display desired patterns on

its surface. The idea dates back to 1999, when scaffolded assembly was

first explored for computational purposes and implemented in small

prototypes82. Directed nucleation was demonstrated experimentally

again in 2003 when successful assembly of multiple DNA tiles around

a scaffold strand was shown within a patterned lattice displaying

barcode information (Fig. 7a)83. In 2004, Shih and colleagues showed

that a single strand of DNA, 1669 nucleotides long, could be driven

to self-assemble into a nanoscale octahedron by the addition of five

short DNA strands84. This work is notable in that it created a three-

dimensional object rather than a two-dimensional structure and

that many paranemic (rather than plectonemic) DNA associations

were used. The most stunning demonstration of scaffolded assembly

was recently presented by Rothemund85, where he used a 7.3 Kb

single-stranded viral genome as the scaffold and generated a variety

of beautiful and complex two-dimensional structures (Fig. 7b). This

scaffolded assembly method is easy, features high yields, and has

great potential for the implementation of even larger two- and three-

dimensional micro- and nanostructures.

The two- and three-dimensional nano-objects built from DNA

have greatly broadened the definition of ‘material’ in the field of

nanotechnology. Their finite size, unique structures, and variety

of shapes have made them ideal candidates for generating multi-

component nanoarchitectures. Their application potential will be fully

realized when any set of biomolecules and/or functional nanomaterials

can be organized into any desired pattern.

Application strategiesDNA-directed organization of nanomaterialsAs mentioned in the previous section, DNA has been used to

create a variety of two- and three-dimensional nanoarchitectures.

These objects are themselves interesting and novel synthetic

materials, but they can also serve as templates upon which other

functional groups and components can be organized. Both duplex

and complex DNA nanoarrays have been exploited for spatially

positioning other functional molecules with nanoscale precision and

programmability.

Fig. 5 Schematic diagrams of common building blocks used in DNA self-

assembly and some of the typical structures visualized by atomic force

microscopy. Scan sizes in (a) and (b) are 1.5 µm x 1.5 µm and the scale

bars in insets correspond to 300 nm. Scan sizes in (c), (d) and (e) are

1.4 µm x 1.4 µm, 1 µm x 1 µm and 1.56 µm x 1.56 µm respectively. From

left to right, the three images in (f) have scan sizes: 500 nm x 500 nm,

1 µm x 1 µm and 500 nm x 500 nm and the inset in the middle is a scan area

of 150 x 150 nm. (Reprinted with permission from34,35,58,61,68,76. © 2003

American Association for the Advancement of Science, © 2005, 1999,

American Chemical Society, © 1998 Nature Publishing Group, © 2000, 2005

American Chemical Society, respectively.)

(b)(a) (c)

(d) (e)

(f)

NT202p26_35.indd 30NT202p26_35.indd 30 23/02/2007 08:38:4223/02/2007 08:38:42



Fig. 6 Finite-sized and addressable DNA patterns. (a) DNA nanoarrays with increased complexity and defined sizes based on the cross-tile or 8-helix bundle. From

left to right: molecular pegboard, 4 x 4, symmetric 5 x 5, and 8-helix bundle based 5 x 5 arrays. (b) Full addressability of the ten tile system with an additional index

tile added to the pegboard design. (c) The letters ‘D’, ‘N’, and ‘A’ displayed on self-assembled 4 x 4 cross-tile arrays. [Part (a) reprinted with permission from77-79.

© 2005 American Chemical Society, © 2006 Wiley-VCH. Part (b) reprinted with permission from78. © 2005 American Chemical Society. Part (c) reprinted with

permission from79. © 2006 Wiley-VCH.]

Fig. 7 Nucleated DNA self-assembly and scaffolded origami. (a) Self-assembly of 01101 barcode lattice around scaffold DNA strand and corresponding atomic

force microscopy (AFM) visualization. (b) The left schematic shows an arbitrary shape formed when the scaffold strand (black) is folded by hybridization with staple

strands (colors). The inset in the left panel shows a staple strand with a protruding stem-loop, used to produce the raised (lighter) pixels in the next panels. The

middle and right panels show a schematic drawing and an AFM image, respectively, of structures formed by folding six copies of the 7.3 Kb M13 virus genomic DNA

into triangles then hexagons. [Part (a) reprinted with permission from83. © 2003 National Academy of Sciences. Part (b) reprinted with permission from85. © 2006

Nature Publishing Group.]

(b)

(a)

(c)

(b)

(a)

NT202p26_35.indd 31NT202p26_35.indd 31 23/02/2007 08:38:4323/02/2007 08:38:43



In the past, many strategies have been exploited for DNA-directed

NP assemblies and a wide variety of patterns have been built. First

of all, metal or semiconductor ions can directly adsorb to DNA

templates via electrostatic interactions, groove binding, or intercalation.

Reduction of these ions will facilitate the formation of NPs along the

DNA templates. Ag20, Au32, Pd28,86, Pt30, and semiconductor NPs87,88

have been successfully templated on DNA using this approach. These

DNA-NP arrays can serve as precursors for nanowires or nanoelectronic

devices, as previously discussed. However, it is hard to achieve

consistent interparticle spacing at long range by using this technique

and only coarse and irregular metallized structures can be built.

Le et al.89 demonstrated the self-assembly of AuNPs into two-

dimensional arrays by a process in which DNA-Au nanocomponents

are hybridized to a preassembled two-dimensional DNA template on

mica (Fig. 8a). In this study, 5 nm Au particles were functionalized with

multiple strands of 3’-thiolated (dT15) and hybridized to protruding

(dA15) sequences on one of the template strands. Using a similar

strategy, Zhang et al.90 reported the use of self-assembled nanogrids

to generate AuNP nanoarrays. The periodic square-like arrangement of

NPs was confirmed by AFM (Fig. 8b).

An alternative strategy for periodic NP display is to conjugate

the NP with one single-stranded DNA (ssDNA) and then use this

DNA-NP conjugate in the assembly of tiles and lattices. Yan and

coworkers91 reported the assembly of 5 nm AuNPs employing this

strategy. They showed that interparticle spacing can be controlled

through variation of the DNA-tile dimensions (Fig. 8c). A more complex

periodic pattern can be achieved by attaching AuNPs to strands in two

different triangle building blocks92 and a well-formed alternating two-

dimensional array of 5 nm and 10 nm AuNPs is produced (Fig. 8d). By

introducing unique sticky ends to the NP-bearing DNA building blocks,

this approach opens up great opportunities for building complex,

controllable systems.

Another interesting application is DNA-directed assembly of

carbon nanotubes (CNTs). CNTs are nanometer-scale materials with

superlative thermal, mechanical, and electronic properties, yet their

use typically requires tedious, exact positioning. DNA self-assembly

offers intriguing possibilities for efficient integration of CNT ‘building

blocks’ into multicomponent structures or devices. Xin et al.93 have

reported the localization of single-walled carbon nanotubes (SWNTs)

onto 1-pyrenemethylamine (PMA) treated lambda-DNA on a Si surface.

PMA interacts electrostatically with DNA and simultaneously with

SWNTs through π-stacking. Direct covalent coupling of 5’-amino DNA

to carboxyl groups on the nanotubes has also been reported94,95. DNA-

functionalized nanotubes can be annealed with complementary DNA

attached to NPs, thus demonstrating the formation of SWNT/AuNP

aggregates. Assembly of SWNTs between Au electrodes has been

achieved using a similar strategy96. The assembly of peptide nucleic

acid (PNA), an uncharged DNA analogue, coupled with SWNTs along

complementary DNA template has also been reported97.

DNA-programmed assembly of materials other than NPs and

nanotubes has also been reported, such as nanorods98, mesoscale

particles99,100, dendrimers101, and fullerenes102. Thus, it has been

well established that DNA’s molecular recognition properties can be

effectively harnessed to program the assembly of other nanomaterials.

DNA-programmed assembly of biomolecules Semisynthetic DNA-protein conjugates, generated by either covalent

or noncovalent coupling chemistry, have been widely used in

the development of bioanalytical procedures and nanomaterials

construction. A commonly used generic approach for templating

biomolecules onto DNA tile arrays is through the use of the strong,

noncovalent biotin/avidin interaction. Streptavidin is a tetravalent

protein endowed with a very high affinity for the small molecule biotin.

The protein can be either covalently conjugated to DNA103, or it can

Fig. 8 (a), (b) Patterned AuNPs displayed on two-dimensional DX tile lattice

and cross-tile nanogrids. In both cases, particles are functionalized with

multiple strands and then hybridized to preformed DNA nanostructures.

(c) Two-component two-dimensional triangle array displaying 5 nm and

10 nm NPs. (d) A, B tile system nanogrid with specific A tile AuNP attachment.

Top row: schematic representations of the DNA-templated assembly of

periodic AuNP nanoarrays. Bottom row: TEM or AFM illustration. Scan sizes

in (b) and (d) are 500 nm x 500 nm and 800 nm x 800 nm. [Part (a) and (b)

reprinted with permission from89,90. © 2004, 2006 American Chemical Society.

Part (c) reprinted with permission from92. © 2006 American Chemical Society.

Part (d) reprinted with permission from91. © 2006 Wiley-VCH.]

(b)(a)

(c) (d)

NT202p26_35.indd 32NT202p26_35.indd 32 23/02/2007 08:38:4523/02/2007 08:38:45



interact with biotinylated DNA104. The conjugates can be utilized as

biomolecular adapters for positioning biotinylated components along

nucleic acid backbones. Any type of biotinylated compound can be

arranged, such as peptides, antibodies, enzymes, and low molecular

weight components104,105. Because of the tetravalent nature of

streptavidin (STV), more complex DNA-STV networks can be built,

such as supramolecular nanocircles106 and supercoiling mediated STV

networks (Fig. 9a)107.

Besides linear double-stranded DNA (dsDNA), self-assembled

DNA tiling systems have been used to organize biomolecules into

patterns. For example, arrays of evenly spaced STV molecules

were assembled onto a TX tile lattice108 and 4 x 4 cross-tile lattice

(Fig. 9b)34. More complex patterns have been achieved by employing

two tile types for the 4 x 4 cross-tile system76 and the periodicity of

the displayed patterns was verified (Fig. 9c). In order to gain greater

control of spatial positioning of proteins, finite-sized arrays with

individual addressing capability have been developed, as described by

Lund et al.78, Park et al.79, etc.

Another strategy for DNA-templated protein display is to employ

specific DNA binding proteins. Liu and coworkers109 first demonstrated

the use of aptamer to direct the assembly of thrombin onto sites on

TX tile arrays. Thrombin binding aptamer sequence is introduced into

the TX tiles and acts as docking sites for the thrombin molecules.

Li et al.110 further extended this approach. They describe the selection

from phage display libraries of single-chain antibodies (scFv) for binding

to a specific DNA aptamer. Decoration of various DNA tile structures

with the aptamer and binding of the scFv was demonstrated (Fig. 9d).

The technology is highly modular and can be extended to assemble

virtually any protein, therapeutic molecules, or other nanomaterials of

interest.

Recently, Turberfield and coworkers demonstrated the binding

of RuvA, a Holliday junction binding protein, to a two-dimensional

Holliday junction lattice59. Interestingly, when RuvA binds to the

building blocks during the self-assembly process, the lattice shows a

square-planar configuration rather than the original kagome lattice.

This shows not only that DNA can be used to create ordered protein

arrays, but also that the protein molecules can play an active, decisive

role in dictating the shape of the DNA tile lattices.

Combination strategies – bringing it all togetherStrategies for fabrication of complex nanostructures that make use

of DNA assemblies in combination with other bionanotechnology

methods and components hold the greatest promise for bottom-up

creation of functional constructs. In one such example, Dai et al.111

reported the fabrication of nanorings of inorganic NPs using a circle of

DNA bound by engineered DNA binding proteins (helicase) fused with

peptides selected for nucleation and binding of Cu2O NPs

(Fig. 10a). The resulting NPs have a dense Cu2O core surrounded

by a low density adsorbed protein shell, which retains biological

functionality (DNA binding) and provides a further level of organized

self-assembly. Since the helicase has 15 permissive sites where peptide

insertion can be carried out, more functional units can be inserted to

create novel multivalent molecular linkers that can organize multiple

building blocks along the DNA template.

Fig. 9 DNA-directed self-assembly of protein molecules. (a) DNA-STV network and conformational change caused by ionic switching. (b) Programmable assembly

of protein arrays on 4 x 4 cross-tile nanogrid. c) Programmable assembly of STV on two-dimensional DNA nanogrids composed of A, B cross-tile building blocks. (d)

TX tile lattice templated protein display. From left to right, STV, thrombin, and selected aptamer binding scFvs. [Part (a) reprinted with permission from107. © 2001

Wiley-VCH. Part (b) reprinted with permission from34. © 2003 American Association for the Advancement of Science. Part (c) and (d) reprinted with permission

from108-110. © 2004 American Chemical Society, © 2005 Wiley-VCH, © 2006 The Royal Society of Chemistry.]

(b)(a) (c)

(d)

NT202p26_35.indd 33NT202p26_35.indd 33 23/02/2007 08:38:4723/02/2007 08:38:47



Keren et al.32,112 recently reported another combination strategy

involving DNA, a DNA binding protein, and inorganic nanomaterials.

RecA, a homologous DNA sequence binding protein, was polymerized

onto ssDNA and used to localize a SWNT at a desired position along

the dsDNA template. The RecA also serves to protect the covered

DNA segment against metallization thereby creating an insulating

gap where the SWNT could sit with its ends contacted by the

conductive metal nanowire, thus creating a field-effect transistor

(Fig. 10b). Mixed biomolecular structures comprised of DNA and

other materials including lipids have also been used for templating

inorganics. For example, a multilamellar structure composed of anionic

DNA and cationic lipid membranes has been used to achieve Cd2+ ion

condensation and growth of CdS nanorods113,114 (Fig. 10c). Lieberman

and coworkers115 recently reported the deposition of monolayer DNA

rafts onto (3-amino-propyl) triethoxysilane stripes on a specially

treated Si surface, which shows the possibility of attaching complex

DNA nanostructures at specific sites and demonstrates a promising

approach for combining bottom-up DNA self-assembly with the top-

down nanolithography technique.

Summary and outlookIn summary, DNA has many unique properties that make it a

promising material for development of self-assembly systems in

nanotechnology. The wide range of DNA tile building blocks

already available makes possible the construction of complex

nanostructures. Combined with the wide variety of other available

nanomaterials, a powerful method for nanofabrication by precise

spatial positioning of functional units is quickly developing. Multiple

functions of DNA can be utilized simultaneously and numerous

techniques can be employed. As the understanding of DNA and

the set of tools available in the molecular toolbox continue to

expand, our ability to engineer novel nanomaterials will continue to

advance.

The outlook for DNA assembly in nanotechnology is very promising

with application areas reaching both down toward the atomic level

and up toward the micron level. The studies summarized here have

helped to flesh out many of the detailed design and construction rules

necessary for successful application of artificial biomolecular structures

to future nanotechnologies.

Fig. 10 Combination strategies for DNA-nanomaterial constructs. (a) TEM image of Cu2O-NPs assembled onto circular DNA. (b) SWNT localization in the middle

of a scaffold λ-DNA molecule mediated by RecA. From top to bottom: RecA nucleoprotein filament (black arrow) bound to template DNA; SWNT bound to RecA

filament; scanning conductance image showing the contrast between the conductive nanotube and the DNA template. Scale bars are 200 nm, 300 nm, and 300 nm,

respectively. (c) DNA-membrane templates for organizing the growth of CdS NPs. TEM images of CdS grown in free solution compared with that templated by

DNA-membrane complexes. [Part (a) reprinted with permission from111. © 2005 American Chemical Society. Part (b) reprinted with permission from112. © 2003

American Association for the Advancement of Science. Part (c) reprinted with permission from113,114. © 2003, 2004 American Chemical Society.]

(b)(a)

(c)

REFERENCES

1. Watson, J. D., and Crick, F. H., Nature (1953) 171, 737

2. Watson, J. D., and Crick, F. H., Cold Spring Harb. Symp. Quant. Biol. (1953) 18,

123

3. Bunka, D. H., and Stockley, P. G., Nat. Rev. Microbiol. (2006) 4, 588

4. Nimjee, S. M., et al., Annu. Rev. Med. (2005) 56, 555

5. Beyer, S., and Simmel, F. C., Nucleic Acids Res. (2006) 34, 1581

6. Mao, C., et al., Nature (1999) 397, 144

7. Yurke, B., et al., Nature (2000) 406, 605

8. Turberfield, A. J., et al., Phys. Rev. Lett. (2003) 90, 118102

9. Liu, D. R., PLoS Biol. (2004) 2, E223

10. Gothelf, K. V., and LaBean, T. H., Org. Biomol. Chem. (2005) 3, 4023

11. Amos, M., Theoretical and Experimental DNA Computation, Springer-Verlag, Berlin

(2005)

12. Stojanovic, M. N., et al., Computing with Nucleic Acids. In Bioelectronics: From

Theory to Applications, Willner, I., and Katz, E. (eds.), Wiley-VCH, Weinheim,

Germany, (2005)

NT202p26_35.indd 34NT202p26_35.indd 34 23/02/2007 08:38:4923/02/2007 08:38:49



13. Xu, J., et al., DNA Structures and their Applications in Nanotechnology, In

Supramolecular Polymers, 2nd edition, Ciferri, A., (ed.) CRC Press, Boca Raton,

Florida, (2005)

14. Murphy, C. J., et al., Science (1993) 262, 1025

15. Arkin, M. R., et al., Science (1996) 273, 475

16. Beratan, D. N., et al., Chem. Biol. (1997) 4, 3

17. Taubes, G., Science (1997) 275, 1420

18. Fink, H.-W., and Schonenberger, C., Nature (1999) 398, 407

19. Porath, D., et al., Nature (2000) 403, 635

20. Braun, E., et al., Nature (1998) 391, 775

21. Kasumov, A. Y., et al., Science (2001) 291, 280

22. Berlin, Y. A., et al., J. Am. Chem. Soc. (2001) 123, 260

23. Berlin, Y., et al., Top. Curr. Chem. (2004) 237, 1

24. Lee, J., et al., Biochem. Cell Biol. (1993) 71, 162

25. Aich, P., et al., J. Mol. Biol. (1999) 294, 477

26. Rakitin, A., et al., Phys. Rev. Lett. (2001) 86, 3670

27. Wettig, S. D., et al., Anal. Sci. (2003) 19, 23

28. Richter, J., et al., Adv. Mater. (2000) 12, 507

29. Patolsky, F., et al., Angew. Chem., Int. Ed. (2002) 41, 2323

30. Ford, W. E., et al., Adv. Mater. (2001) 13, 1793

31. Monson, C. F., and Woolley, A. T., Nano Lett. (2003) 3, 359

32. Keren, K., et al., Science (2002) 297, 72

33. Park, S. H., et al., Appl. Phys. Lett. (2006) 89, 033901

34. Yan, H., et al., Science (2003) 301, 1882

35. Park, S. H., et al., Nano Lett. (2005) 5, 693

36. Park, S. H., et al., Nanotechnology (2004) 15, S525

37. Mirkin, C. A., et al., Nature (1996) 382, 607

38. Alivisatos, A. P., et al., Nature (1996) 382, 609

39. Mucic, R. C., et al., J. Am. Chem. Soc. (1998) 120, 12674

40. Storhoff, J. J., et al., J. Am. Chem. Soc. (2000) 122, 4640

41. Lazarides, A. A., and Schatz, G. C., J. Phys. Chem. B (2000) 104, 460

42. Rosi, N. L., and Mirkin, C. A., Chem. Rev. (2005) 105, 1547

43. Loweth, C. J., et al., Angew. Chem., Int. Ed. (1999) 38, 1808

44. Maeda, Y., et al., Appl. Phys. Lett. (2001) 79, 1181

45. Shchepinov, M. S., et al., Nucleic Acids Res. (1997) 25, 4447

46. Shchepinov, M. S., et al., Nucleic Acids Res. (1999) 27, 3035

47. Watson, K. J., et al., J. Am. Chem. Soc. (2001) 123, 5592

48. Li, Z., et al., Nano Lett. (2004) 4, 1055

49. Li, Y., et al., Nat. Mater. (2004) 3, 38

50. Um, S. H., et al., Nat. Mater. (2006) 5, 797

51. Seeman, N. C., J. Theor. Biol. (1982) 99, 237

52. Kallenbach, N. R., et al., Nature (1983) 305, 829

53. Seeman, N. C., and Kallenbach, N. R., Biophys. J. (1983) 44, 201

54. Seeman, N. C., et al., Nanotechnology (1998) 9, 257

55. Chen, J., and Seeman, N. C., Nature (1991) 350, 631

56. Zhang, Y., and Seeman, N. C., Biophys. J. (1994) 66, A24

57. Zhang, Y., and Seeman, N. C., J. Am. Chem. Soc. (1994) 116, 1661

58. Mao, C., et al., J. Am. Chem. Soc. (1999) 121, 5437

59. Malo, J., et al., Angew. Chem., Int. Ed. (2005) 44, 3057

60. Li, X., et al., J. Am. Chem. Soc. (1996) 118, 6131

61. Winfree, E., et al., Nature (1998) 394, 539

62. Sha, R., et al., Chem. Biol. (2000) 7, 743

63. Liu, F., et al., J. Am. Chem. Soc. (1999) 121, 917

64. Seeman, N. C., Nano. Lett. (2001) 1, 22

65. Zhang, X., et al., J. Am. Chem. Soc. (2002) 124, 12940

66. Yan, H., et al., Nature (2002) 415, 62

67. Liao, S., and Seeman, N. C., Science (2004) 306, 2072

68. LaBean, T. H., et al., J. Am. Chem. Soc. (2000) 122, 1848

69. Yang, X., et al., J. Am. Chem. Soc. (1998) 120, 9779

70. Ding, B., et al., J. Am. Chem. Soc. (2004) 126, 10230

71. Liu, D., et al., J. Am. Chem. Soc. (2004) 126, 2324

72. Reishus, D., et al., J. Am. Chem. Soc. (2005) 127, 17590

73. Ke, Y., et al., J. Am. Chem. Soc. (2006) 128, 4414

74. Wei, B., and Mi, Y., Biomacromolecules (2005) 6, 2528

75. Mathieu, F., et al., Nano Lett. (2005) 5, 661

76. Park, S. H., et al., Nano Lett. (2005) 5, 729

77. Liu, Y., et al., J. Am. Chem. Soc. (2005) 127, 17140

78. Lund, K., et al., J. Am. Chem. Soc. (2005) 127, 17606

79. Park, S. H., et al., Angew. Chem., Int. Ed. (2006) 45, 735

80. He, Y., et al., Angew. Chem. Int., Ed. (2005) 44, 6694

81. Liu, H., et al., Biomacromolecules (2005) 6, 2943

82. LaBean, T. H., et al., Experimental Progress in Computation by Self-Assembly

of DNA Tilings. In DIMACS Series in Discrete Mathematics and Theoretical

Computer Science, Winfree, E., and Gifford, D. K., (eds.) Proceedings of the 5th

DIMACS Workshop on DNA Based Computers, MIT, Cambridge, (1999) 54, 123

83. Yan, H., et al., Proc. Natl. Acad. Sci. USA (2003) 100, 8103

84. Shih, W. M., et al., Nature (2004) 427, 618

85. Rothemund, P. W., Nature (2006) 440, 297

86. Richter, J., et al., Appl. Phys. Lett. (2001) 78, 536

87. Coffer, J. L., et al., Appl. Phys. Lett. (1996) 69, 3851

88. Torimoto, T., J. Phys. Chem. B (1999) 103, 8799

89. Le, J., et al., Nano Lett. (2004) 4, 2343

90. Zhang, J., et al., Nano Lett. (2006) 6, 248

91. Sharma, J., et al., Angew. Chem., Int. Ed. (2006) 45, 730

92. Zheng J., et al., Nano Lett. (2006) 6, 1502

93. Xin, H. J., and Woolley, A. T., J. Am. Chem. Soc. (2003) 125, 8710

94. Dwyer C., et al., Nanotechnology (2002) 13, 601

95. Li, S., et al., J. Am. Chem. Soc. (2005) 127, 14

96. Hazani, M., et al., Chem. Phys. Lett. (2004) 391, 389

97. Williams, K. A., et al., Nature (2002) 420, 761

98. Dujardin, E., et al., Chem. Commun. (2001), 1264

99. Soto, C. M., et al., J. Am. Chem. Soc. (2002) 124, 8508

100. Milam, V. T., et al., Langmuir (2003) 19, 10317

101. Choi, Y. S., et al., Nano Lett. (2004) 4, 391

102. Cassell, A. M., et al., Angew. Chem., Int. Ed. (1998) 37, 1528

103. Niemeyer, C. M., et al., Angew. Chem., Int. Ed. (1998) 37, 2265

104. Niemeyer, C. M., Trends Biotechnol. (2002) 20, 395

105. Niemeyer, C. M., Angew. Chem., Int. Ed. (2001) 40, 4128

106. Niemeyer, C. M., et al., Angew. Chem., Int. Ed. (2000) 39, 3055

107. Niemeyer, C. M., et al., ChemBioChem (2001) 2, 260

108. Li, H., et al., J. Am. Chem. Soc. (2004) 126, 418

109. Liu, Y., et al., Angew. Chem., Int. Ed. (2005) 44, 4333

110. Li, H., et al., Org. Biomol. Chem. (2006) 4, 3420

111. Dai, H., et al., J. Am. Chem. Soc. (2005) 127, 15637

112. Keren, K, et al., Science (2003) 302, 1380

113. Liang, H., et al., J. Am. Chem. Soc. (2003) 125, 11786

114. Liang, H., et al., J. Am. Chem. Soc. (2004) 126, 14157

115. Sarveswaran, K., et al., Langmuir (2006) 22, 11279

116. Mirkin, C. A., Inorg. Chem. (2000) 39, 2258

117. LaBean, T. H., Nat. Mater. (2006) 5, 767

NT202p26_35.indd 35NT202p26_35.indd 35 23/02/2007 08:38:5123/02/2007 08:38:51

Constructing novel materials with DNA

Documents