Full text of DOI:10.1007/BF02900620

NOTES
from Au nanoparticles[2]. Braun made a 100 nm silver
wire with DNA as a template. Particularly, many shapes of
pure DNA have been synthesized by Seeman, such as
Assemble four-arm DNA
junctions into nanoweb
2-dimensional crystals[12], actuators[13] and cubes[14]
.
In this work, 2-dimensional nanowebs were synthe-
sized from four-arm DNA junctions. If oligonucleotides
with particular sequences are designed, the static four-arm
DNA junction can form with the branch joint immobile.
The static four-arm DNA junction is of a planar cruciform
structure in condition of no multivalent cations in buffer,
so it is an ideal nanopart. We used the four-arm DNA
junction as the core and the double helical DNA with
sticky ends as the linker to synthesize 2-dimensional
nanolattices.
ZHOU Chunqing1, TAN Zukun1, WANG Chen1,
BAI Chunli1 & CAO Enhua2
1. STM Lab, Institute of Chemistry, Chinese Academy of Sciences,
Beijing 100080, China;
2. Institute of Biophysics, Chinese Academy of Sciences, Beijing 100101,
China
Correspondence should be addressed to Bai Chunli (e-mail: zcq@
softhome.net)
Abstract
DNA is of structural polymorphism, which is
useful in nanoarchitecture; especially, four-arm DNA junc-
tions can be used to assemble nanowebs. The static four-arm
DNA junctions were designed and synthesized. One-arm
DNA and two-arm DNA came out simultaneously with the
four-arm DNA junction’s formation. A new method, termed
the two-step method, was proposed and the productivity of
four-arm DNA junctions was increased. A nanoweb was as-
sembled successfully, but it showed irregularity itself. It was
not the same as we expected. We consider that it is a result
from the flexibility of four-arm DNA junction.
1
Materials and methods
(ν) Materials. Tris buffer, polyacrylamide gel, T4
DNA ligase, ethanol, superpure water.
(ξ) Design of DNA’s sequences. The Holliday
DNA junction is a molecular paradigm of homologous
recombination[14]. It is also a kind of four-arm DNA junc-
tion, but its branch joint is mobile. In the Holliday DNA
junctions, the sequences’ homologies play two contradic-
tory roles. On the one hand, only when the sequences are
homologous can this four-arm DNA junction form. On the
other hand, branch migration is a direct consequence of
homology or two-fold sequence symmetry. A four-arm
DNA junction would disassemble when the branch joint
migrates to the end of one double helical DNA strand. So,
this mobile DNA junction is not proper for nanofabrica-
tion. But if we change the sequence symmetry, we can get
an analogue of Holliday DNA junction, a new kind of
four-arm DNA junction, termed static DNA junction, and
the joint could be fixed precisely. A four-arm junction’s
stabilization depends on temperature, the length of arm,
and properties of the buffer, while joint’s fixation profits
from the sequences’ asymmetry.
Keywords: nucleic acid, self-assembly, nanotechnology, atomic force
microscopy.
DNA is well known as the polymeric molecule that
contains the genetic information for life. Moreover, it is
important and cannot be substituted in nanoarchitecture
with unique advantages. (ν) DNA is a kind of nanowire
itself naturally which can be used as template in nanofab-
rication[1–3]. For example, the diameter of canonical
B-double helical nucleic acid is 2 nm. (ξ) DNA is of
structural polymorphism, triplex and quadruplex DNA
have been discovered subsequently[4]. Some special struc-
tures can be synthesized such as four-arm DNA junc-
tions[5,6]. A large number of potentially useful rigid
nanoparts made of polymerized DNA have been built,
including molecular scale rods, rings, cubes[14], tetrahe-
If used as a nanopart, a four-arm DNA junction must
be small and stable. So the construction of static four-arm
DNA junction should accord with the following principles:
(ν) CG base pair is stable than AT. So it is essential to
increase the CG content as highly as possible. (ξ) The
repeated units should be few, so that the mismatch could
be avoided. (ο) The other DNA structures should be kept
away. For example, a sequence with three or more con-
secutive G could form quadruplex DNA, which should be
avoided. (π) Symmetry entails branch migration. The
sequence around the junction site should be of minimum
two-fold symmetry.
drons, hollow tubes, branched joints, etc.[7,8]. (ο) DNA is
available for arbitrary programmed sequences due to con-
venient solid support synthesis. (π) DNA can also be
easily chemically modified, so the application of DNA is
increased for more potential nanotechnological purposes.
(ρ) Particularly, the interactions between DNA and en-
zymes as well as lipids have been studied in detail and
widely. DNA can be recognized and manipulated by a
large battery of enzymes, including ligase, restriction en-
donucleases, kinases and exonucleases. This advantage
ü
11]
can be used in biocomputation and nanofabrication[9
.
Recently, great progress has been made in using
DNA to build nanostructures. Mirkin used DNA as a
linker to direct the formation of macroscopic architecture
The sequences were designed as follows (5Ą-pho-
sphorylation, 3Ą-hydroxylation):
1618
Chinese Science Bulletin Vol. 46 No. 19 October 2001

Fig. 1. Sketch maps showing how to synthesize the 2-dimensional DNA web. (a) Double helical DNA linker with
sticky ends. (b) Static four-arm DNA junction with sticky ends. c) Sketch of a 2-dimensional DNA nanoweb, side
length is 50 bp, 17 nm.
was deposited on a freshly cleaved mica surface and dried
5Ą-GGCTCGCAATCCTGAGCACG-3Ą
(1)
( 2 )
(3)
(4)
for 2 min, then washed twice with superpure water and
dried for use. Samples were imaged with the NANO-
SCOPE ċ type (Digital Instruments) in tapping mode.
The scanning frequency was about 2 Hz. The free ampli-
tude was 1.5 V and the resonant frequency was about 320
kHz.
5Ą-GGCTCGTGCTCACCGAATGC-3Ą
5Ą-GGCTGCATTCGGACTATGGC-3Ą
5Ą-GGCTGCCATAGTGGATTGCG-3Ą
5Ą-AGCCAACCCGTTCTCGGAGCACTGCAGAAC-3Ą(5)
5Ą-AGCCGTTCTGCAGTGCTCCGAGAACGGGTT-3Ą(6)
2
Results and analysis
The structure of the double helical DNA linker with
sticky ends and the static four-arm DNA junction with
sticky ends with the designed sequences are shown in fig.
1.
(ν) Synthesis of the double helical DNA linkers and
the four-arm DNA junctions. It was simple to synthesize
double helical DNA linkers. The two strands (5, 6) were
mixed at stoichiometric ratio (1Ή1, molΉmol), and
(ο) Synthesis of the four-arm DNA junctions and
double helical DNA linkers. The four-arm DNA junc-
tions were synthesized from the former four strands (1ü4)
and the linkers from the latter two strands (5, 6). Samples
were purified from polyacrylamide gel. The bands of
four-arm DNA junctions and double helical DNA linkers
were cut out and eluted in the eluant buffer for 5ü6 h at
room temperature on the shaking table. The products were
centrifugated to eliminate the gel, and precipitated by
adding iced ethanol, then washed twice by 70% ethanol.
(π) Synthesis of the nanoweb. The four-arm DNA
junctions and double helical linkers with sticky ends were
mixed at ratio of 1Ή2 (molΉmol), and annealed at 15ć
(without multivalent cations such as Mg2+). Then T4 DNA
ligases were added to ligate the sticky ends. 12 h later, the
reaction was stopped by adding EDTA to final concentra-
tion of 20 mmol/L. The products were precipitated by
adding iced ethanol and stored in Tris buffer at 0 ü 4ć.
(ρ) Imaging with AFM. The nanoweb samples
were diluted to 1 ng/uL. A 1ü5 uL aliquot of this solution
heated to 90ć, cooled slowly to 55ć and kept constant
for 2 h, then cooled to room temperature. The four-arm
DNA junctions formed with byproducts (one-arm and
two-arm DNA junctions) from the four strands (1ü4).
One-arm DNA junctions were the stable product from any
two of strands (1ü4), and two-arm DNA junctions were
from any three of strands (1ü4). Although the activation
energy was optimum, the formation of four-arm DNA
junction involved four molecules, so the yield was not
high. A new method, termed the two-step method, was
developed to synthesize four-arm DNA junctions. At the
first step, any two of the four strands (1ü4) were mixed
at stoichiometric ratio (1Ή1, molΉmol), and heated to
90ć, cooled slowly to 55ć and kept constant for 2 h,
then cooled to room temperature. So did the other two
strands. Two kinds of one-arm DNA junction were got at
high yield. Then the two intermediates were mixed at 1Ή
1 (molΉmol), incubated at proper temperature for 2 h,
and then cooled slowly to room temperature. The DNA
Chinese Science Bulletin Vol. 46 No. 19 October 2001
1619

NOTES
linkers and junctions were purified from polyacrylamide
gel.
Fig. 2 shows the 12% polyacrylamide gel electro-
phoresis, from which we can state that one-arm, two-arm
and four-arm DNA junctions were stable in proper buffer.
A large number of one-arm DNA junctions formed when
only two strands were involved in reaction (lane 3). While
three strands were involved, one-arm and two-arm DNA
junctions formed simultaneously, and it is worthy of no-
tice that the formation of two-arm DNA junctions was
preponderant (lanes 2, 8). If all the four strands (1ü4)
mixed at the same time and the three junctions including
one-arm, two-arm, four-arm ones formed after incubation
(lanes 1, 9). But the one-arm and two-arm DNA junctions
decreased when we used the two-step method to synthe-
size four-arm DNA junctions (lanes 4ü7). It was found
that the incubation temperature (from 35ć to 55ć) at
the second step did not affect the formation and yield of
four-arm DNA junctions. In conclusion, when four strands
(1ü4) were mixed simultaneously, the two-arm DNA
junctions would form as a stable byproduct. If DNA were
measured accurately, at the first step of the two-step
method, only trace of single strand DNA existed, so at the
second step, two-arm DNA junctions formed less for
shortage of raw materials and four-arm DNA junctions
should be synthesized at very high yield. In this work,
DNA’s concentration was measured by UV spectra and it
was inaccurate for the bases’ bad-distribution. So the ex-
crescent single strand DNA would retain when only two
strands went in reaction and the two-arm junctions were
obvious (lanes 4ü7).
Fig. 2. 12% polyacrylamide gel electrophoresis at 60 V. After 6 h, the
gel was dyed in 0.5 ug/mL EB for 10 min. All the reactants were mixed
at stoichiometric ratio (1Ή1, molΉmol). Lane 1. The four strands (1ü4)
were mixed simultaneously, heated to 90ć, cooled slowly, incubated at
55ć and kept constant for 2 h, then cooled slowly to room temperature.
Lane 2. Reactants were three strands (1ü3), under the same conditions
as of lane 1. Lane 3. Reactants were strands (1, 2) with the same condi-
tions as of land 1. Lane 4. Two-step method was adopted. One-arm DNA
junctions were synthesized from strands (1, 2). So did strands (3, 4),
under the same conditions as of lane 1. Then the two one-arm DNA
junctions were mixed at 1Ή1 (molΉmol), incubated at 55ć which was
kept constant for 2 h, then cooled slowly to room temperature. Lane 5.
The same as lane 4, but the temperature was changed to 45ć. Lane 6.
The same as lane 4, but the temperature was changed to 40ć. Lane 7.
The same as lane 4, but the temperature was changed to 35ć. Lane 8.
The same as lane 2, but reactants were strands (2ü4). Lane 9. The same
as lane 1.
fig. 1(c) can be found in a small area (in the white frame
in fig. 3). The distance between neighboring holes’ center
was about 18 nm, which is near 17 nm (the side length of
designed lattice = 0.34 nm/bph50 bp). Large regular
structure was not found. Although without bivalent and
multivalent cations, the four-arm DNA junction is of a
planar cruciform structure and the angles between
neighboring arms is 90e when the arms were not too
ü
(ξ) Imaging with AFM. Fig. 3 is the AFM image
of the nanowebs assembled from the four-arm DNA junc-
tions and the double helical DNA linkers. It can be stated
that the web was irregular; some holes are big and some
small. The lattice corresponding to what we designed as
long (<3 turns)[14 16]. When the nanowebs formed, the
four-arm DNA junctions were annealed and ligated with
the DNA linkers. The arm’s length increased and the ri-
gidity decreased, so the arm bend and cannot form regular
2-dimensional lattices.
Fig. 3. AFM images. The nanoweb was irregular. The lattice corresponding to what we designed as fig. 1(c) can be found in a small
area (in the white frame in fig. 3). The distance between neighboring holes’ centers was about 18 nm, which is near 17 nm (the side
length of designed lattice = 0.34 nm/bph50 bp).
1620
Chinese Science Bulletin Vol. 46 No. 19 October 2001

7. Zhang, Y., Seeman, N. C., Construction of a DNA-truncated octa-
hedron, J. Amer. Chem. Soc., 1994, 116: 1661.
3
Discussion
8. Seeman, N. C., Molecular craftwork with DNA, The Chemical
Intelligencer, 1995, 1: 38.
9. Ogihara, M., Ray, A., DNA computing on a chip, Nature, 2000,
403: 143.
The four-arm DNA junctions were thermodynami-
cally stable as a nanopart, but their structure is flexible. It
was reported recently that one enzyme, named RuvA[17]
,
10. Liu, Q., Wang, L., Frutos, A. G. et al., DNA computing on surfaces,
Nature, 2000, 403: 175.
11. LaBean, T. H., Yao, H., Kopatsch, J. et al., Construction, analysis,
ligation, and self-assembly of DNA triple crossover complexes, J.
Am. Chem. Soc., 2000, 122: 1848.
12. Liu, F., Sha, R., Seeman, N. C., Modifying the surface features of
two-dimensional DNA crystals, J. Am. Chem. Soc., 1999, 121:
917.
13. Mao, C., Sun, W., Shen, Z. et al., A nanomechanical device based
on the B-Z transition of DNA, Nature, 1999, 397: 144.
14. Seeman, N. C., Kallenbach, N. R., DNA branched junction, Annu.
Rev. Biophys. Biomol. Struct., 1994, 23: 53.
could specially bind with a four-arm DNA junction as
tetramer or octamer[18] even if with bivalent or multivalent
cations. The assembly of RuvA and four-arm DNA junc-
tion was a planar cruciform structure but more rigid than
four-arm DNA junctions alone. A series of enzymes have
been found recently, which can specially recognize and
manipulate four-arm DNA junctions[19,20]. This knowledge
made a solid basis for the use of four-arm DNA junctions
in nanofabrication and biocomputation.
15. Lilley, D. M. J., Clegg, R. M., The structure of the four-arm junc-
tion in DNA, Annu. Rev. Biophys. Biomol. Struct., 1993, 22: 299.
16. Duckett, D. R., Murchie, A., Diekmann, S. et al., The structure of
the Holliday junction and its resolution, Cell, 1988, 55: 79.
17. Shlyakhtenko, L. S., Hsieh, P., Gigoriev, M. et al., A cruciform
structural transition provides a molecular switch for chromosome
structure and dynamics, J. Mol. Biol., 2000, 296: 1169.
18. Roe, S. M., Barlow, T., Brown, T. et al., Crystal structure of an
octameric RuvA-Holliday junction complex, Molecular Cell, 1998,
2: 361.
19. White, M. F., Giraud-Panis, M. E., Pöhler, J. R. G. et al., Recogni-
tion and manipulation of branched DNA structure by junc-
tion-resolving enzymes, J. Mol. Bio., 1997, 269: 647.
20. Gopaul, D. N., Guo, F., Duyne, G. D. V., Structure of the Holliday
junction intermediate in Cre-loxP site-specific recombination, The
EMBO Journal, 1998, 17: 4175.
References
1. Braun, E., Eichen, Y., Sivan, U. et al., DNA-templated assembly
and electrode attachment of a conducting silver wire, Nature, 1998,
391: 775.
2. Mirkin, C. A., Letsinger, R. L., Mucic, R. C. et al., A DNA-based
method for rationally assembling nanoparticles into macroscopic
materials, Nature, 1996, 382: 607.
3. Alivisatos, A. P., Jojnsson, K. P., Peng, X. et al., Organization of
‘nanocrystal molecules’ using DNA, Nature,1996, 382: 609.
4. Shafer, R. H., Stability and structure of model DNA triplexes and
quadruplexes and their interactions with small ligands, Progress in
Nucleic Acid Reseach and Molecular Biology, 1998, 59: 55.
5. Kallenbach, N. R., Ma, R., Seeman, N. C., An immobile nucleic
acid junction constructed from oligonucleotides, Nature, 1983,
305: 829.
6. Wang, Y., Mueller, J. E., Kemper, B. et al., Assembly and charac-
terization of five-arm and six-arm DNA branched junctions, Bio-
chemistry, 1991, 30: 5667.
(Received March 13, 2001; revised April 6, 2001)
Chinese Science Bulletin Vol. 46 No. 19 October 2001
1621