ARTICLES
Chinese Science Bulletin 2006 Vol. 51 No. 24 2973—2976
DOI: 10.1007/s11434-006-2223-9
Analogic China map
constructed by DNA
QIAN Lulu1, WANG Ying2, ZHANG Zhao1,
ZHAO Jian1, PAN Dun2, ZHANG Yi2, LIU Qiang1,
FAN Chunhai2, HU Jun1,2 & HE Lin1,3
1. Bio-X Center, Shanghai Jiao Tong University, Shanghai 200030,
China;
2. Shanghai Institute of Applied Physics, Chinese Academy of Sciences,
Shanghai 201800, China;
3. Institute for Nutritional Sciences, Shanghai Institutes of Biological
Sciences, Chinese Academy of Sciences, Shanghai 200031, China
Correspondence should be addressed to Liu Qiang (email: [email protected]), Fan Chuanhai (email: [email protected]), Hu Jun (email: jhu@
sjtu.edu.cn), or He Lin (email: [email protected])
Received September 30, 2006; accepted November 19, 2006
Abstract In this research, a nanoscale DNA
structure of analogic China map is created. The
nanostructure of roughly 150 nm in diameter with a
spatial resolution of 6 nm is purely constructed by
folding DNA. The picture observed by atomic force
microscopy (AFM) is almost identical with the designed shape. The DNA origami technology invented
by Rothemund in 2006 is employed in the construction of this shape, which has proved the capability of
constructing almost any complicated shape enabled
by DNA origami, and provides new bottom-up method
for constructing nanostructures.
Keywords: DNA, origami, self-assembly, nanostructure, China map,
AFM.
Compared to conventional top-down routines of
shaping, bottom-up self-assembly for constructing
shapes (especially nanoscale shapes) is undoubtedly an
important method. The accumulated knowledge and
technology on DNA and the coding capacity of DNA
itself have made DNA self-assembly the most promising self-assembly technology.
New methods on constructing nanoscale shapes with
DNA self-assembly emerge frequently. In 1989, Seeman[1] first proposed a DNA branched junction as the
basic self-assembly unit. After that, Yan et al.[2] have
improved these four-arm junctions which self-assemble
into DNA nanogrids, as have been clearly observed by
AFM. In 1998, Winfree et al.[3] constructed a DNA
double crossover molecule, called DX motif. Each DX
motif contains four sticky ends, and they self-assemwww.scichina.com
www.springerlink.com
bled into two-dimensional DNA crystals. DX motif can
be designed to make DNA triangles[4] and nanotubes[5].
In 2000, LaBean et al.[6] constructed a DNA triple
crossover molecule, called TX motif. Each TX motif
contained six sticky ends, and allowed for better twodimensional DNA arrays. The common idea of DNA
self-assembly is to patch the big shape by assembling
small basic units through Watson-Click complementary.
This shared property has induced their common limitation; that is, it will be very difficult, if possible, to construct non-periodical complicated shapes.
In 2006, Rothemund[7] first proposed a completely
new DNA self-assembly method, that is, DNA origami,
which could successfully construct a variety of relatively complicated nanoscale shapes and patterns. This
technology is undoubtedly a breakthrough in the field
of DNA self-assembly. With DNA origami, Rothemund[7] has constructed six desired shapes of ~100 nm
in diameter, including square, rectangle, star, disk with
three holes, etc., and several patterns including the map
of western hemisphere. The method of DNA origami is
folding a long single-stranded DNA molecule by
choosing a number of short complementary oligonucleotides to hold it in place. The structures can be programmed to desired shapes. Meanwhile, each oligonucleotide can also serve as a pixel to form designed patterns on the surfaces of those structures.
In this research, DNA origami technology is used to
construct an analogic China map shape by folding a
7-kb single-stranded DNA by over 200 complementary
oligonucleotides, resulting in a double-stranded DNA
structure of ~150 nm in diameter. Imagine the long single-stranded DNA as a soft rope. Fill DNA into the
raster in horizontal direction with specified offset to
simulate the given shape, and short oligonucleotides
perform like a number of staples, binding the long single-stranded DNA in appropriate sites. This structure is
observed by AFM, and a nanoscale China map shape is
successfully built. This research used DNA origami to
construct an asymmetrical complicated two-dimensional shape, and therefore provided a further proof of
the complicated nanoscale shape constructing capacity
enabled by DNA origami technology.
1
1.1
Material and method
Design of DNA sequence
To fold DNA into China map shape with DNA origami technology, the sequence design strategy comprises: Choose a long single-stranded DNA (called
2973
ARTICLES
scaffold strand), and choose appropriate binding site
and design complementary short oligonucleotides
(called staple strands) to hold the scaffold strand into a
desired shape.
The design is performed in two steps:
(i) Design the scaffold according to the outline of
China map. Treat the scaffold strand as a long line,
filling the raster to form the China map shape by folding the strand in horizontal direction. The two parts of
northwest and northeast of China map has make it a
concave shape, so this shape cannot be achieved only
by folding from north to south. As shown in Fig. 1,
separate the main part of China map into east part and
west part by the longitude on Hainan Island. Also
separate the convex part of northeast and northwest into
east and west parts. The folding starts with the south
end of Hainan Island, goes from south to north to finish
west part first, and after reaching the most north point
of northwest part, goes from north to south. Finish the
folding of northeast part in the same way, and continue
folding on east coast, which will end, again, with
Hainan Island. Taiwan Island is the most difficult part,
as it is not conjoint with the mainland and not in the
central longitude. This issue is roughly solved by linking lines between the island and the mainland. The
linking lines are part of the scaffold but not counted
into the horizontal folding. Islands in South China Sea
are not included in this analogic shape.
The length of each folding line corresponds with the
number of bases on the scaffold strand. One turn of
DNA double-helix is around 10.67 bases. Each offset
has to be an integer of half turns (e.g. 10.67×0.5×3≈16
bases). One turn is around 3.6 nm in length and 2 nm in
width. The gap between horizontal helices depends on
the spacing of crossovers, which is set at 1.5 turn.
Therefore, the inter-helix gap is ~1 nm. This data shall
be considered into the length-width proportions when
(a)
we are designing the scaffold. Meanwhile, the constraint of ~7000-nt in full length shall also be satisfied.
The total number of designed bases is supposed to be
close to, but not exceeds, the full length.
Therefore, the designed scaffold shown in Fig. 1 can
be represented as the following data set.
Each entry is the number of bases for each line on
the scaffold, which is proportional to the offset. This set
of data, which is obtained by measuring the China map,
is also the input for next step programming: (16 16 6 6
32 32 112 112 107 107 112 112 208 208 230 230 256
256 267 267 272 112 112 112 80 80 48 48 32 48 16 16
16 48 48 64 64 256 80 80 64 64 32 32 48 48 32 32 16
32 16 48 48 80 80 64 64 64 64 32 32 16 112 64 64 80
80 96 96 112 112 112 112 96 96 128 16 16 128 48 48 6
6 16 16). Each integer has to be the multiple times of
16, or with residue 6 or 11 when divided by 16. Under
that presupposition, the integer shall be adjusted to be
as proportional to offset as possible to achieve a tight
simulation of the shape. The specification of residue 6
and 11 will minimize the strain. This issue is discussed
in the Supplementary Note S1 of Ruthemund paper[7].
(ii) Generate the sequence of staple strands by program. In order to make the folded rope-like DNA to
be fixed to the shape, it is necessary to staple between
horizontal lines and vertical gaps, which is implemented by Watson-Crick complementary. We can generate the sequence of staple strands according to the
known scaffold sequence and A-T C-G pair matching
principals. Staple strands could be classified into eight
types of a―h, as shown in Fig. 2 (a). The four types of
a―d are used for horizontal fixing, as shown in Fig. 2
(b); their combination can form different types of horizontal fixings. Some staple strands of certain type lack
several bases, but it is still considered as this type. For
instance, at the upper border of the shape, some staple
strands of type a or b will lack 8-nt in 3′; at the left or
(b)
Fig. 1. The map of the main part of China (a) and the corresponding designed scaffold (b).
2974
Chinese Science Bulletin
Vol. 51 No. 24 December 2006
ARTICLES
Fig. 2. Eight types of staple strands (a), staples to fix horizontal lines (b) and staples to fix vertical gaps (c). Thick black lines represent scaffold strand,
thick blue lines represent staple strands (arrows direct from 5′ to 3′) and thin black lines represent the complementary of base pairs.
right border, there will be instances missing more bases.
The four types of e―h are used to fix vertical gaps, as
shown in Fig. 2 (c). Their combination can fix vertical
gaps for different situations. Most staple strands are of
32-nt, except those appearing on the shape border. All
the spacing of crossovers is 1.5 turns (16-bp).
According to the above described designing strategy,
we developed sequence design software in PERL. The
7249-nt circular genomic DNA sequence from the virus
M13mp18 is the input of the program. Identify a 20-bp
long stem-loop structure, which contains a BsrB I
cleavage site. Simulate to cleave 73-nt off by BsrB I
enzyme, resulting in a 7176-nt long single-stranded
DNA sequence. Then generate 219 staple sequences, 8
sequences of 23-nt oligonucleotides complementing the
remaining scaffold sequence, and 2 sequences of 32-nt
oligonucleotides complementing the scaffold sequence
linking Taiwan Island. The program output is shown in
Fig. 3. The PDF of this output is in the supplementary
material, zooming in to see each base on scaffold and
staple, and the strand direction indicated by > and <.
Fig. 3. Sequence design for the main part of China map. Scaffold is
represented in black and staples are represented in different colors.
1.2
Experimental steps
(i) Digestion. Incubate the circular single-stranded
M13mp18 DNA in restriction buffer at 37℃ for 15
min, and then digest it with BsrB I (New England Bio
labs) for 2 h. Purify digested DNA by phenol-chloro-
Fig. 4. AFM image of a well-formed DNA analogic main part of China map (a) and AFM image of a number of DNA analogic main part of China
maps (b).
www.scichina.com
www.springerlink.com
2975
ARTICLES
form extraction and ethanol precipitation.
(ii) Mixing. Mix M13mp18 DNA purified in (i)
with synthesized 229 short DNA oligonucleotides (Invitrogen) in a 100-μL volume (160 nmol/L of each
short strand, 1.6 nmol/L M13mp18 DNA).
(iii) Annealing. Anneal the mixture from 94℃ to
4℃ in a PCR machine at a rate of 0.01℃/s.
(iv) AFM imaging. Use AFM to observe the assembled nanostructure.
1.3
strand with greater length, the precision will be improved.
3
AFM imaging
Samples were prepared by deposition of 5 μL onto
freshly cleaved mica.
Imaging was performed in Tapping Mode under 30
μL 1× TAE/Mg2+ buffer on a Digital Instrument Nanoscope III a Multimode AFM (Veeco) with J scanner,
using an NP-S oxide-sharpened silicon nitride tip
(Veeco). The tip-surface interaction was minimized by
optimizing the scan set-point.
2
Result
AFM images shows that, as designed, the folded
DNA analogic China map shape is of ~150 nm length,
~120 nm width, and 2 nm height (Fig. 4(a)).
We define product ratio as the ratio between
well-formed structures (Fig. 4(b)-a and -b) to recognizable structures (Fig. 4(b)-c), to roughly estimate the
goodness of sequence design. In our design and experiment, the product ratio is 59% with sample size 125.
The two main factors influencing product ratio are
self-assembled shape distortion (Fig. 4(b)-c) and the
accumulation of staple strands on self-assembled
shapes (e.g. the bright pixel d in Fig. 4(b)). Additionally,
some shapes aggregate (Fig. 4 (b)-b and -d), while others do not (Fig. 4(b)-a and -c). But generally speaking,
the stacking of shapes along blunt-ended helices is not
severe. Furthermore, because the designed shape is
asymmetric, some new phenomena occur, which will
never be observed in symmetric design. We observed
that only 9% of the well-formed shapes lied in the designed orientation, the other 91% was reversely oriented. This phenomenon could possibly imply that the
self-assembled shape is not straight planar.
The size and precision of constructed analogic China
map shape is limited by the length of M13mp18
(~7000-nt). Provided with current precision, there is
still discrepancy with a real China map at the shape
border. Particularly, Taiwan Island and Hainan Island
differ more. If the shape is constructed with a scaffold
2976
Discussion
In this paper, a particular nanoscale shape is constructed by DNA origami technology. The ultimate target of DNA origami technology is to construct more
practical useful nanostructures. There are the following
several ingredients to be explored further: (1) The
choice of scaffold strand. The length of scaffold strand
will decide the size and precision of the constructed
nanostructure. So, it is better to consider the size and
precision at first and calculate the appropriate scaffold
length. Since long single-stranded DNA synthesis is
very expensive, we need to find longer naturally-existing long single-stranded DNA beside M13mp18. (2)
Transform DNA nanostructure into other materials. It
will generate many other practical applications if DNA
nanostructure could be converted into other materials,
such as potential applications in integrated circuit industry. Therefore, the DNA metallization technology is
of key importance for this application. (3) Three-dimensional nanostructures can be viewed as a pile of
vertically accumulated two-dimensional sections. If
these two-dimensional sections can be linked by staple
strands in certain way, this conceit might be realized.
Acknowledgements This work was supported by Shanghai
Municipal Science and Technology Commission (Grant No.
03DZ14025).
References
1
2
3
4
5
6
7
Seeman N C. Nanoscale assembly and manipulation of branched
DNA: A biological starting point for nanotechnology. In: Lewis J,
Quel J L, et al, eds. Nanocon Proceedings. Bellevue: NANOCON,
1989. 101―107
Yan H, Reif J H, LaBean T H. DNA-templated self-assembly of
protein arrays and highly conductive nanowires. Science, 2003, 301:
1882―1884[DOI]
Winfree E, Liu F, Seeman N C, et al. Design and self-assembly of
two-dimensional DNA crystals. Nature, 1998, 394: 539―544[DOI]
Rothemund P W. Papadakis N, Winfree E. Algorithmic self-assembly of DNA Sierpinski triangles. PLoS Biol, 2004, 2: 2041―
2053
Rothemund P W, Ekani-Nkodo A, Winfree E, et al. Design and
characterization of programmable DNA nanotubes. JACS, 2004,
126: 16344―16352[DOI]
LaBean T H, Reif J H, Seeman N C, et al. Construction, analysis,
ligation, and self-assembly of DNA triple crossover complexes.
JACS, 2000, 122: 1848―1860[DOI]
Rothemund P W. Folding DNA to create nanoscale shapes and patterns. Nature, 2006, 440: 297―302[DOI]
Chinese Science Bulletin
Vol. 51 No. 24 December 2006