Geckeler K.E., Nishide H. (Eds.) Advanced Nanomaterials

Подождите немного. Документ загружается.

Structural DNA Nanotechnology: Information - Guided

Self - Assembly

Yonggang Ke , Yan Liu , and Hao Yan

869

Advanced Nanomaterials. Edited by Kurt E. Geckeler and Hiroyuki Nishide

ISBN: 978-3-527-31794-3

28.1

Introduction

Although the Watson – Crick double helical model [1] of DNA has imparted major

impacts on modern biology for more than 50 years, the signiﬁ cance of this

simple – yet very elegant – model is not limited to one particular area. In 1982 [2] ,

Ned Seeman proposed the building of nanostructures from DNA, an idea which

led to the origination of the ﬁ eld now known as “ structural DNA nanotechnology. ”

During the past decade, this ﬁ eld has witnessed much signiﬁ cant progress, and

in this chapter we will discuss the basic concepts and major research directions

of structural DNA nanotechnology, the important progress that has been made in

recent years, and some future perspectives of the ﬁ eld.

DNA, which serves as the genetic information carrier in most organisms on

Earth, is also an ideal candidate for structural nanotechnology, which targets at

controlling and organizing matter at the nanometer scale. First, DNA is a nanom-

eter - scale object itself, with a diameter of 2 nm and a helical repeat of 10 – 10.5

nucleotide pairs, or ∼ 3.5 nm for the common B - form DNA. Second, DNA hybridi-

zation is highly predictable because of the well - known Watson – Crick base - pairing

that guanine (G) pairs with cytosine (C), and adenine (A) with thymine (T). Third,

whilst single - stranded DNA is quite ﬂ exible, the DNA duplex is more rigid and

has a persistence length of approximately 50 nm. It is this combined ﬂ exibility and

rigidity that permits the design of DNA structures to form different geometric

shapes. Furthermore, as a results of advances in modern chemistry and molecular

biology, DNA molecules with any designed lengths, sequences, and a variety of

functionalities can now be synthesized conveniently, and also manipulated by

using the wide range of enzymes that are available for the cleavage, ligation and

ampliﬁ cation of DNA. All of these facilities have provided scientists with an

extreme degree of control for building DNA nanostructures and maneuvering

DNA nanomachines.

Topologically speaking, DNA duplex is a one - dimensional ( 1 - D ) molecule. In

order to create a two - dimensional ( 2 - D ) or three - dimensional ( 3 - D ) structure, it is

870 28 Structural DNA Nanotechnology: Information-Guided Self-Assembly

necessary to use branched objects. For example, Seeman proposed techniques that

would allow DNA strands to assemble into branched junctions that could further

self - assemble into periodic, 2 - D arrays [3] (Figure 28.1 a).

The basic building block here is the branched four - arm junction consisting of

four single - stranded DNA oligonucleotides. To enable DNA junction building

blocks to form higher ordered objects and lattices, single ν stranded DNA over-

hangs called “ sticky ends ” are used to bring DNA junction molecules together.

These sticky ends carries base sequences that are complementary to each other;

for example, the red sticky end is complementary to the black end, and the green

end is complementary to the blue end (Figure 28.1 a). As a result, sticky end cohe-

sion will cause the individual four - arm junctions to be “ glued ” together to form

the 2 - D array. Yet, this simple scheme illustrated a powerful method for building

Figure 28.1 Ned Seeman ’ s original proposal of construction

of a periodic DNA array and its application. (a) Four - arm

junction DNA tiles with sticky ends are connected together to

form a 2 - D periodic array through self - assembly process;

(b) A 3 - D DNA lattice templated protein array could be used

for X - ray crystallography.

28.2 Periodic DNA Nanoarrays 871

DNA nanostructures: ﬁ rst, to design the branched DNA building blocks or “ tiles ” ,

and then to assemble them together using sticky end cohesion.

Although other cohesions have also been used for mortaring DNA nanostruc-

tures – for example, paranemic crossover s ( PX ) cohesion [4] and edge - sharing

[5] – sticky end cohesion is by far the most extensively used in DNA nanostructure

design. Studies of crystal structures have revealed that the duplex formed by com-

plementary sticky ends has an exactly identical structure to B - DNA [6] , a feature

which allows designers to predict and control the relative orientations of the DNA

tiles. It is interesting to note that for sticky ends that are N - bases long, the number

of unique sticky ends can be up to 4

, which provides a library of programmable

molecular interactions.

Ideally, the DNA sequences of a DNA nanostructure should be designed to

achieve the highest stability, so that all other less - stable competitive structures

will be less likely to form. On a practical basis, it is necessary to choose a set of

optimized sequences that minimizes sequence symmetry at the branch points [7]

so as to prevent the branch migration of DNA strands.

As originally proposed by Seeman, one potential application of structural

DNA nanotechnology would be to use the highly ordered self - assembling DNA

scaffolds to organize other types of macromolecule into 3 - D crystals. Moreover,

if the macromolecule could be attached to a 3 - D DNA nanostructure (Figure

28.1 b), the self - assembly of DNA would facilitate the organization of macromol-

ecules into a periodic lattice, the periodicity and parameters of which could be

well deﬁ ned. This would in turn facilitate macromolecule structural analysis using

X - ray diffraction.

Besides the above - described potential application, the DNA nanostructure might

also be used as a scaffold to organize different nanomaterials. An example of this

is the organization of nanoelectronic components into an addressable fashion,

leading to the construction of DNA - templated nanoelectronics/nanophotonics

devices. Indeed, recent developments in the use of 1 - D and 2 - D DNA nanostruc-

tures to template nanoparticles into rationally designed patterns have paved the

way towards this goal.

28.2

Periodic DNA Nanoarrays

Seeman ’ s original proposal has inspired many research groups to construct

a variety of DNA tiles with different sizes and geometries, and assemble them

into 2 - D periodic arrays (Figure 28.2 ). For example, Mao et al. designed and con-

structed DNA parallelograms that would grow into micrometer - sized 2 - D arrays

(Figure 28.2 a) [8] .

A series of DNA tile molecules termed double crossover [9] ( DX ) DNA molecules

were originally created by Seeman, and subsequently utilized in many of the later

studies. The DX tile consists of two parallel DNA helices, joined together by two

crossovers through strand exchange. Winfree et al. successfully built DNA 2 - D

872 28 Structural DNA Nanotechnology: Information-Guided Self-Assembly

arrays through the self - assembly of DX tiles (Figure 28.2 b) [10] . A similar design

strategy, learned from the DX tile construction, was then used by different groups

when designing DNA tiles containing multiple parallel DNA helices; these

included triple crossover tiles [11] , and four - , eight - , and 12 - helix tiles [12, 13] .

Yan et al. used four four - arm DNA junctions to design a cross - shaped tile,

named “ 4 × 4 ” tile [14] . In this design, four four - arm junctions are tethered

together by a long central strand with a dT

loop in between (Figure 28.2 c); the

4 × 4 tiles then self - assembled into a 2 - D square lattice. By removing one arm

from, or adding two more arms to the 4 × 4 tile, Mao ’ s group were able to construct

a three - point - star [15] and a six - point star tile [16] that could self - assemble into 2 - D

lattices with hexagonal and triangular cavities (Figure 28.2 d).

Another family of tiles are tube - like tiles, including three - helix [17] , six - helix

[18] (Figure 28.2 e), and eight - helix bundles [19] . With different sticky end

designs, these tubes tile can be assembled to either 1 - D tubes, 2 - D arrays, or even

3 - D lattices.

28.3

Finite - Sized and Addressable DNA Nanoarrays

For the purpose of DNA - directed macromolecule crystallization, the periodic

arrays represent an excellent choice. However, to build a functional DNA nanoelec-

tronic device, it must be possible to control the size of a DNA array and to

attach functional species at particular locations on the array. Such needs

have driven research teams to develop methods for building ﬁ nite - sized and

addressable arrays.

Park et al. [20] reported the construction of a square - shaped, addressable array

consisting of sixteen 4 × 4 tiles (Figure 28.3 a) through an hierarchical assembly

Figure 28.2 DNA tiles and their 2 - D periodic arrays formed

through sticky end cohesion. (a) A DNA parallelogram tile

consisting of four Holliday junction structures; (b) A double

crossover ( DX ) tile; (c) A 4 × 4 cross - shaped tile. Each arm of

the cross is a four - arm DNA junction; (d) A three - point - star

tile; (e) A six - helix - bundle tile.

28.3 Finite-Sized and Addressable DNA Nanoarrays 873

Figure 28.3 DNA ﬁ nite - sized and addressable

arrays. (a) A 4 × 4 tile array consisting of

16 distinctive 4 × 4 tiles. This is a fully

addressable array on the level of individual

tiles. Streptavidin protein was attached onto

certain DNA tiles to display a letter “ D ” ; (b) A

5 × 5 tile array consisting of 25 eight - helix

tiles. The number of distinctive tiles was

reduced from 25 to 13 by taking advantage of

the array ’ s C

symmetry; (c) Schematic of

Rothemund ’ s scaffolded DNA self - assembly

and atomic force microscopy image of a

nanometer - scale “ smiley face. ” The black

strand is the scaffold DNA that is folded by

other short “ staple ” strands into the

designed shape.

process. In order to demonstrate that the 16 - tile array was fully addressable on the

level of individual tiles, Park and colleagues ﬁ rst functionalized a few tiles on the

array with biotin groups, and then attached streptavidin molecules to the array. In

this way, the protein molecules could be organized to display the letters, “ D ” , “ N ”

and “ A. ”

It is costly to make every tile in an array unique; moreover, the more complex

the system, the greater the chances of self - assembly errors occurring. In order to

build ﬁ nite - sized arrays in a cost - efﬁ cient way, Liu et al. [21] demonstrated a strat-

egy to utilize the geometric symmetry of the array to reduce the number of unique

tiles. For a N - tile ﬁ nite - size array with C

symmetry, the number of unique tiles

required is N / m , if N / m is an integral number, or Int( N / m ) + 1, if N / m is a non-

integral number. Consequently, Liu and coworkers demonstrated two 25 - tile array

examples with C

and C

fold symmetry. The 5 × 5 array with C

symmetry required

13 unique tiles instead of 25 (Figure 28.3 b), while the 5 × 5 array with C

symmetry

required only seven unique cross - shaped tiles instead of 25.

When comparing these two ﬁ nite - size addressable arrays, it is possible to under-

stand the dilemma that research groups often face when designing a DNA struc-

ture. On one hand, the addressability of an array can be increased by introducing

more unique sequences/tiles into the system, although a low yield and a high error

rate will be expected. On the other hand, the symmetry can be utilized to reduce

874 28 Structural DNA Nanotechnology: Information-Guided Self-Assembly

complexity of the system so as to achieve a high yield and a low assembly error

rate, but the high addressability would be lost. Thus, depending on the purpose

of a DNA nanostructure, it is important for a designer to identify a good balance

between the complexity and addressability of the system.

Another important approach when building an addressable DNA nanoarray is

“ nucleated DNA self - assembly. ” This method uses a long natural or synthetic DNA

strand, which serves as a scaffold, to direct the DNA strands or tiles self - assembly.

Yan et al. demonstrated the use of this technique by efﬁ ciently assembling DX

tiles together into barcode - patterned lattices [22] . In 2006, Paul Rothemund

reported an exciting breakthrough, in which he used more than 200 short “ staple ”

DNA strands to fold 7249 base long, single - stranded M13 viral DNA into 2 - D arrays

( “ DNA Origami ” ) with a variety of shapes [23] . Every staple has its unique sequence,

and can only hybridize with predeﬁ ned parts of M13 according to a predetermined

folding path (Figure 28.3 c). The array resulted is fully addressable at the position

of each individual staple strand. In theory, if the scaffold strand is long enough,

it is possible to design any arbitrarily shaped 2 - D DNA array and to functionalize

staples at any position on that array.

28.4

DNA Polyhedron Cages

One branch of structural DNA nanotechnology focuses on the construction of 3 - D

DNA “ cages. ” Potentially, DNA cages can be used for applications such as target -

speciﬁ c drug delivery or nanoparticle site - speciﬁ c functionalization. For delivery

purposes, the cages can be made to encapsulate cargos during their self - assembly

and to display target recognition tags outside the cages. Because DNA duplex is

linear, it is not surprising that most of the DNA cages built to date are polyhedral,

using DNA duplex(es) as straight edges and branching point(s) at the vertices.

When Chen and Seeman assembled the ﬁ rst DNA polyhedron, a cube consisting

of ten DNA strands [24] , they used endonucleases to cleave speciﬁ c edges of the

cube to prove the tube topology with polyacrylamide gel electrophoresis. Many

years later, several groups have only recently developed a number of methods

to build DNA polyhedral cages and more direct ways to prove their formation

(Figure 28.4 ).

Goodman et al. built a DNA tetrahedron by hybridizing four single - stranded

DNAs together in a one - step annealing process (Figure 28.4 a) [25, 26] . In a later

study, the same group built hairpin loops into the tetrahedron and demonstrated

that the edge of this cage could be opened/closed by the addition of “ fuel DNA ”

strands [27] . This mechanism would allow cargos to diffuse into the structure at

the open stage, and to be captured in the close stage. Shih et al. used a 1.7 kilobase

and ﬁ ve short, single - stranded DNAs to construct an octahedron (Figure 28.4 b)

[28] where the edges of the octahedron were double crossovers (DX) or paranemic

crossovers (PX) DNA motifs. One noteworthy feature of this octahedron was that

the 1.7 kilobase DNA was formed through polymerase chain reaction ( PCR ), which

28.4 DNA Polyhedron Cages 875

in turn means that the DNA could be easily ampliﬁ ed to produce large amounts

of the DNA octahedron.

Aldaye et al. recently developed a new approach for building different - shaped

DNA cages [29] . They introduced organic molecules into the circular single -

stranded DNA during the DNA synthesis process (Figure 28.4 c). This allowed the

building of a series of DNA polygons with repeated sequences at each edge and

branched organic molecules at the corners. Linker strands were then used to bring

two or three polygons together to form the DNA cages.

Mao ’ s group reported the design and formation of tetrahedron, dodecahedron,

and buckyballs through hierarchical DNA assembly (Figure 28.4 d) [30] . All three

polyhedra shared the same basic building unit, namely the three - point - star tile

Figure 28.4 DNA polyhedron cages. (a) DNA

tetrahedron: schematic drawing and a physical

model; (b) DNA octahedron model and 3 - D

reconstruction image from cryo - electron

microscopy (EM) experiments; (c) A series

DNA polyhedra. Organic molecules are

incorporated into the DNA circular strands to

help the structures form. (d) DNA

tetrahedron, dodecahedron, and buckyball

assembled from three - point - star tiles. A 3 - D

reconstruction of the cryo - EM images of these

polyhedra is shown on the right.

876 28 Structural DNA Nanotechnology: Information-Guided Self-Assembly

[15] , although at low concentrations (50 – 75 n M ) the tile tended to form discrete

3 - D polyhedra rather than a 2 - D array. Another variable in such a design was the

length of the single - stranded loops at the center part of the tile, which provided a

variable ﬂ exibility to the tile and controlled the angle at which the arms could bend

out of the plane. Longer loops seemed to provide a higher ﬂ exibility at the center

of tiles and allowed them to form polyhedra that required larger bending angles

at the vertexes. For example, ﬁ ve - base loops were found to promote the formation

of tetrahedrons, while three - base loops were suitable for the formation of dodeca-

hedrons. By linking two of the three - point - star tiles, a tile with four arms was

created that could self assemble into a buckyball - shaped DNA structure. In a later

study, a ﬁ ve - point - star tile was shown to self - assemble into an icosahedron, using

a similar design/assembly strategy [31] .

28.5

DNA Nanostructure - Directed Nanomaterial Assembly

As noted at the start of this chapter, one central task of DNA nanotechnology is

to control functional materials at the nanometer scale, and the DNA templated

self - assembly of nanomaterials represents an important direction towards this

goal. Until now, several groups have demonstrated the assembly of nanometer -

scale materials such as metallic nanoparticles, quantum dot s ( QD s) and proteins

on DNA nanostructures (Figure 28.5 ). The recognition events used to capture

these functional materials included protein – protein/small molecule/aptamer

( in vitro selected short DNA or RNA that can speciﬁ cally bind to certain protein)

interactions, functionalized metallic nanoparticle – DNA interactions, and DNA –

DNA hybridization.

By using the well - known interaction between streptavidin and biotin, streptavi-

din protein molecules can be organized onto 2 - D DNA nanoarrays with a control-

led periodicity and spacing [32] . The streptavidin – biotin interaction has also been

shown to be useful for organizing nanoparticles. For example, Sharma et al. suc-

cessfully built a patterned QD array by reacting streptavidin - coated QDs with

biotin groups on the 2 - D DNA nanoarray [33] . By attaching single/multiple copies

of DNA onto the Au nanoparticle surface [34 – 36] , Au nanoparticles could be

rationally organized onto self - assembled DNA nanoarrays through DNA – DNA

hybridization. Recently, Yan ’ s group demonstrated that different - sized Au nano-

particles could be used to control the conformation of DNA nanotubes (Figure

28.5 d) [37] .

Liu et al. incorporated thrombin - binding aptamers into a linear, three - helix DNA

tile array and used aptamer/protein recognition to capture thrombin proteins into

periodic arrays [38] . Recently, Rinker et al. [39] were able to take one step further

by displaying two different thrombin aptamers, each target at a distinctive site of

the protein, on a DNA array. The distances between the two aptamers were sys-

tematically tuned such that a ∼ 5.7 nm spacing led to an optimal “ multivalent

binding, ” which has much higher binding afﬁ nity than does any of the single

aptamers (Figure 28.5 c).

28.6 Concluding Remarks 877

28.6

Concluding Remarks

The ﬁ eld of structural DNA nanotechnology has witnessed numerous break-

throughs during the past decade, although only a small fraction of the prominent

studies are outlined here. In that time, the complexity of the DNA structures has

grown dramatically, from less than 100 nucleotides (four - arm DNA junction) to

more than 10 000 nucleotides ( “ DNA Origami ” ), with series of 3 - D objects having

Figure 28.5 DNA nanostructure - directed nanomaterial

assembly. (a) Streptavidin assembly on DNA 4 × 4 periodic

2 - D array; (b) Quantum dot assembly on a double crossover

(DX) array; (c) Thrombin line assembled on M13 viral DNA

scaffolded 2 - D array; (d) Gold nanoparticles were assembled

on a DX 2 - D array and forced the array to form tubes with

different spiral patterns.

878 28 Structural DNA Nanotechnology: Information-Guided Self-Assembly

been built and characterized. Today, the capability is available to attach proteins,

metallic and semi - conducting nanoparticles onto DNA arrays so as to create a

variety of patterns of these materials. An increasing knowledge of the rules

to design DNA structures has led to the building of more - complex DNA self -

assemblies, and with fewer errors. With such progress, there is much optimism

that DNA nanotechnology can potentially be applied to build DNA nanostructure -

based nanocircuits and to control chemical/biochemical reactions in an ordered

fashion that mimics enzyme cascade reactions. Nonetheless, many challenges

remain. Although some functional materials have been successfully patterned by

DNA arrays, the key ability is still lacking to create well - controlled, multicompo-

nent nanoarchitectures. Whilst the current success with DNA 3 - D objects has been

limited to the building of polyhedron cages for purposes such as protein encap-

sulation and drug delivery, the task remains to create a universal strategy for

building highly ordered 3 - D structures. The “ designer DNA ” nanostructures pro-

duced, with their controlled geometry and topology, might ﬁ nd their use in biologi-

cal applications such as interfacing cellular components through DNA scaffolds.

Yet, much remains to be done in studying the biocompatibility, delivery, and

stability of DNA nanostructures inside living systems. With the ﬁ eld of nanotech-

nology having successfully “ borrowed ” DNA from biological systems, it will not

be surprising to see DNA nanotechnology contribute to the in vivo applications of

nanotechnology in the future.

Acknowledgments

These studies were supported by grants from the National Science Foundation

( NSF ), the Army Research Ofﬁ ce ( ARO ), and the Technology and Research Initia-

tive Fund from Arizona State University to Y.L., and by grants from NSF, ARO,

Air Force Ofﬁ ce of Scientiﬁ c Research, Ofﬁ ce of Naval Research, and the National

Institute of Health to H.Y.

References

1 Watson , J.D. and Crick , F.H.C. ( 1953 ) A

structure for deoxyribose nucleic acid .

Nature , 171 , 737 – 738 .

2 Seeman , N.C. ( 1982 ) Nucleic acid

junctions and lattices . J. Theor. Biol. , 99 ,

237 – 247 .

3 Holliday , R. ( 1964 ) A mechanism for

gene conversion in fungi . Genet. Res. , 5 ,

282 – 304 .

4 Zhang , X. , Yan , H. , Shen , Z. and

Seeman , N.C. ( 2002 ) Paranemic cohesion

of topologically - closed DNA molecules .

J. Am. Chem. Soc. , 124 , 12940 – 12941 .

5 Yan , H. and Seeman , N.C. ( 2003 )

Edge - sharing motifs in structural DNA

nanotechnology . J. Supramol. Chem. , 1 ,

229 – 237 .

6 Qiu , H. , Dewan , J.C. and Seeman , N.C.

( 1997 ) A DNA decamer with a sticky

end: the crystal structure of

d - CGACGATCGT . J. Mol. Biol. , 267 ,

881 – 898 .

7 Seeman , N.C. ( 1990 ) De novo design of

sequences for nucleic acid structural

engineering . J. Biomol. Struct. Dyn. , 8 ,

573 – 581 .