Branched DNA that forms a solid at 95°C

Article abstract of DOI:10.1002/anie.201006992

At arm's length: Branched DNA with exceptionally short sticky ends (see picture) assembles into a new material at temperatures where genomic DNA is fully denatured, if branching geometry and linker rigidity favor crystallization. The process is shown by modeling and for synthetic material assembled from dilute aqueous buffer solution. Copyright

Full text of DOI:10.1002/anie.201006992

DOI: 10.1002/anie.201006992
DNA-Based Assembly
Branched DNA That Forms a Solid at 958C**
Arunoday Singh, Mariyan Tolev, Martin Meng, Konstantin Klenin, Oliver Plietzsch,
Christine I. Schilling, Thierry Muller, Martin Nieger, Stefan Brꢀse, Wolfgang Wenzel, and
Clemens Richert*
Control over the structure of materials may be achieved by
using predictable interactions, such as base pairing. Base
pairing between DNA strands is emerging as one of the most
versatile design principles of nanoconstruction.[1] A range of
hybridization[2] and folding motifs[3] of linear and circular
DNA have been reported. The flexibility of the design has
been further expanded by linking oligonucleotides to syn-
thetic branching elements or “cores”.[4,5] The resulting con-
struct can have properties not found in natural DNA. This
includes DNA-coated gold nanoparticles[6] that assemble into
three-dimensional aggregates, the melting transitions of
which are exceptionally sharp.[7] Nanoparticle size and
linker structure affect the association behavior,[8] and crystal-
lization may be induced in favorable cases.[9]
For DNA hybrids with organic cores, the effect of linking
the DNA to a branching element can be more dramatic still.
Four-arm hybrid 1 (Scheme 1) with its tetrahedral core was
recently shown to assemble into a macroscopic material, even
though its oligonucleotide arms are just dimers.[10] The
assembly process is sequence specific, as demonstrated by
mismatch controls, but the UV-melting transitions are broad,
not sharp as in the case of gold nanoparticles. Shortly after the
publication of the unusually stable assemblies of 1, the first
designed DNA crystals were reported.[11] The fact that the
association of the rigid triangle motifs that serve as rigid
“cores” in these crystals is also driven by no more than dimer
“sticky ends” again suggests that the rules for 3D construction
Scheme 1. Structures of the DNA hybrids employed: (CG)4TPM (1),
of periodic assemblies are quite different from those of linear
(CG)6HPX (2), and (CG)4TTPA (3). Abbreviations of cores are derived
from those of the corresponding alcohols: TPM=tetrakis(hydroxyphe-
nyl)methane, HPX=hexakis(hydroxyphenyl)-p-xylene, TTPA=tetrakis-
(triazoylphenyl)adamantane.
DNA.[12]
[*] A. Singh, M. Tolev, Prof. C. Richert
Institut fꢀr Organische Chemie, Universitꢁt Stuttgart
70569 Stuttgart (Germany)
We have modeled the assembly processes of hybrids by
using an effective coarse-grained model to better understand
the effect of rigidity and core geometry on assembly. The
theoretical results motivated the synthesis of new hybrids
with greater propensity to assemble into three-dimensional
structures. Herein, we report two such hybrids, namely 2 with
six DNA arms and pseudo-octahedral core, and 3, which
forms a material from micromolar aqueous solution at 958C.
We performed Brownian dynamics simulations with geo-
metric forms that hybridize via the ends of their arms as
models for DNA hybrids. These uncharged, coarse-grained
models can reflect some key properties, such as size,
coordination, and energy of bond formation as a function of
bond angle and distance. To observe the assembly processes,
the simulations had to be performed at high concentration
(3 mm), and environmental effects (such as salt concentration,
Fax: (+49)711-685-64321
E-mail: lehrstuhl-2@oc.uni-stuttgart.de
M. Meng, O. Plietzsch, C. I. Schilling, Dr. T. Muller, Prof. S. Brꢁse
Institute of Organic Chemistry and Center for Functional Nano-
structures (CFN), Karlsruhe Institute of Technology (KIT)
76131 Karlsruhe (Germany)
Dr. M. Nieger
Laboratory of Inorganic Chemistry, Department of Chemistry
University of Helsinki, 00014 Helsinki (Finland)
Dr. K. Klenin, Priv.-Doz. Dr. W. Wenzel
Institute for Nanotechnology
Karlsruhe Institute of Technology (KIT)
76021 Karlsruhe (Germany)
[**] This work was supported by CFN (project no. C5.01-03). We thank
Harald Henning for contributing to the synthesis of (CC)4TTPA.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201006992.
Angew. Chem. Int. Ed. 2011, 50, 3227 –3231
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3227

Communications
electrostatic screening), required for quantitative prediction
search for a core with pseudo-octahedral structure but no
increase in flexibility over 1.
of the assembly temperature, cannot be adequately described.
Simulations at half the concentration show a weak depend-
ence of the assembly temperature on concentration, with a Tm
depression of 4–88C.
Figure 1a shows typical results of a tempering protocol.
Hybrids with tetrahedral cores and flexible linkers were
found to have a large propensity to form both five- and six-
Hybrid 2 was assembled starting from dimethyl tereph-
thalate (4) via 1,4-phenylenebis[bis(4’-methoxyphenyl)me-
thanol] (5) and HPX (6, hexakis(p-hydroxyphenyl)-p-
xylene;[13] see the Supporting Information). A crystal struc-
ture of 6 showed the expected pseudo-octahedral geometry
À
(Figure 2a). The molecules are linked by intermolecular OH
O hydrogen bonds forming a two-dimensional network of
channels parallel to the crystallographic a and b axes with
Figure 1. Modeling assembly processes. a) Order, but not energy, as a
function of temperature for simulated adiabatic cooling of simplified
models of hybrids with flexible tetrahedral (red), rigid tetrahedral
(blue), or octahedral structure (green). b) Energy per bond (in units of
artificial duplex strength) for tetrahedral hybrids with one active DNA
arm (black, reference system) or four active arms and rigid (blue) or
flexible (red) core–DNA linkage. Models used are without the charge
of actual DNA or its counterions.
Figure 2. a) Molecular structure of 6 viewed along the xylene axis.
b) Crystal packing with solvent benzene as guest. See the Supporting
Information for further details.
membered rings, thus leading to an arrangement without
long-range order (red structure in Figure 1a; see the Sup-
porting Information for definition of order parameter). The
broad distribution of local bonding arrangements may explain
why broad melting transitions are observed for 1.[10] However,
for stiffer hybrids (that is, small angular range for hybrid-
ization) we observed mostly six-membered rings that coalesce
into extended crystalline structures reminiscent of the two
phases of ice in their local arrangement (blue structure in
Figure 1a). This indicates a significantly stronger propensity
for the emergence of long-range order for the rigid system,
even though the bond energy per particle in the final structure
is very similar to that of the more flexible system.
We then modeled the assembly of octahedrally coordi-
nated hybrids, which were expected not to suffer from the
problem of multiphase crystallites. Compared to the tetrahe-
dral cores we found an increased tendency to form ordered
structures with a near perfect coordination throughout the
entire crystallites (Figure 1a, green structure). This led to the
53.1% porosity, filled by five benzene guest molecules per
unit cell and formula unit (Figure 2b).
Based on 6, solid-phase synthesis on controlled-pore glass
with a combination of 5’- and 3’-phosphoramidites and an on-
support phosphitylation step gave 2 in 5% overall yield, after
purification under denaturing conditions. Adamantane hybrid
3 was prepared next, using a solution-phase synthesis that
involved a cycloaddition and phosphoramidite coupling to
elaborate the DNA arms[14] (see Scheme S3 in the Supporting
Information). The strong tendency of this hybrid to self-
assemble required unconventional steps. A phosphoramidite
coupling at 408C was followed by treatment with a mixture of
oxidizing agents and detritylation with aqueous acid. Final
deprotection with a mixture of ammonia and methylamine,
and thorough washing of the basic solution, gave a crude
product that was > 85% pure (see Figure S4 in the Supporting
Information). For HPLC, 3 was dissolved at 808C with 10%
CH3CN and chromatographed with 10 mm buffer in the
presence of organic solvent at 558C.
3228
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 3227 –3231

Figure 3 shows absorbance profiles of 1–3 at 10 mm hybrid
concentration, measured at 260 nm. Assays started under
denaturing conditions (10 mm NaOH, 958C), followed by
Figure 4. Photographs of cuvettes containing: a) samples of hybrids
1–3 (10 mm), 150 mm NaCl, and 100 mm MgCl2 at 208C under ambient
light (1 shows slight turbidity, both 2 and 3 show precipitation), and
b) samples of 2 (38 mm), ethidium bromide (EB, 120 mm), and the
precipitate of 2 soaked in EB (2+EB) under a hand-held UV lamp
(254 nm+366 nm).
Table 1: UV melting points [8C] of assemblies formed by branched DNA
hybrids at 10 mm concentration.
Compound
Buffer[a]
+Na+[b]
+Mg2+[c]
(CG)4TPM (1)
(CG)6HPX (2)
(CG)4TTPA (3)
n.t.[d]
14Æ0.8
25.1Æ1.3
solid[g]
solid[e]
solid[f]
solid[g]
18.7Æ0.9
61.5Æ0.7
[a] 10 mm phosphate, 1.5 mm NaOAc, pH 7. [b] +150 mm NaCl. [c] +
100 mm MgCl2. [d] No cooperative transition. [e] Turbidity at ꢀ108C.
[f] Turbidity at ꢀ258C. [g] Precipitation at ꢀ958C.
Figure 3. UV-monitored thermal association and melting profiles (heat-
ing: open symbols, cooling: filled symbols) of 10 mm 1–3 in a) 10 mm
phosphate buffer, 1.5 mm NaOAc, or b) plus 150 mm NaCl and
100 mm MgCl2. Increases in absorbance in the cold are caused by
scattering on particles, while the sharp decrease in absorbance for 3 in
the heat is a result of precipitation.
Control experiments with hybrids featuring non-self-comple-
mentary DNA arms, namely (TC)6HPX and (CC)4TTPA,
demonstrated that the association is sequence specific, as
predicted by the Watson–Crick base-pairing rules (see
Figures S12–S15 in the Supporting Information). Similar
controls for 1 were reported previously, together with a
series of hybrids with longer DNA arms.[10] Analysis of
precipitates harvested after 15 h at 48C by UV spectroscopy
and MALDI-TOF mass spectrometry confirmed that the
precipitates contain intact hybrids (see Figures S18–S20 in the
Supporting Information). Staining with duplex-specific inter-
calator YOYO-1 indicates the formation of base pairs, and
electron micrographs of particles show the formation of
regularly shaped nanocrystals in precipitates (see Figures S17,
S23, and S24 in the Supporting Information). For 3, sufficient
material was available to perform exploratory elemental
analyses of precipitates. The CHN content is in agreement
with approximately 40% water and 24 equivalents of salt in
the assemblies (see the Supporting Information).
Soaking preformed precipitates of pseudo-octahedral
hybrid 2 with intercalators for 15 h at 48C led to the uptake
of approximately one equivalent of ethidium bromide per
predicted CG:GC duplex and approximately 0.7 equivalents
of methylene blue per predicted duplex (Figure 4b and the
Supporting Information). This suggests that 1) a large pro-
portion of the DNA arms engages in duplex formation, and
2) sufficient porosity exists for the intercalators to enter the
material. Both properties are in agreement with the predicted
neutralization at 958C. At the hybrid concentration chosen
and minimal salt concentration, 1 showed no transition,
whereas both 2 and 3 gave association and melting curves with
little hysteresis. Additional precipitation/melting curves
acquired under other buffer conditions are presented in the
Supporting Information.
After addition of NaCl and MgCl2 and cooling, all hybrids
showed macroscopic signs of assembly (Figure 4). The onset
of particle formation depended strongly on structure
(Table 1). Without magnesium, only 3 assembled visibly.
After addition of divalent cations, six-arm hybrid 2 started to
form particles below 258C and flexible four-arm hybrid 1
below 108C. For rigid hybrid 3 the propensity to assemble is
most dramatic. This hybrid starts precipitating even at 958C,
as evidenced by a sharp decrease in absorbance and settling of
the material. The assemblies melted slowly during subsequent
heating, as expected for solids (hysteresis between heating
and cooling curves). The material formed by 3 did not fully
redissolve upon reheating to 958C.
Linear DNA does not precipitate under our buffer
conditions (see Figure S16 in the Supporting Information).
Angew. Chem. Int. Ed. 2011, 50, 3227 –3231
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3229

Communications
and 100 mm MgCl2 were added at 958C, and melting curves were
measured with cooling or heating rates of 0.58CminÀ1 in a spectrom-
eter with detection at 260 nm.
Crystal structure: CCDC 785245 (6) contains the supplementary
crystallographic data for this paper. These data can be obtained free
of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif. See also the Supporting Infor-
mation.
structure of 2 (green line, Figure 1a). Larger crystallites than
those shown in Figures S23 and S24 in the Supporting
Information are required to confirm the structure of the
solids by X-ray crystallography. So far, we have been unable
to grow such crystals.
To rationalize the ability of hybrids to assemble with just
dimer DNA arms, we performed additional simulations.
When three arms of a tetrahedral model system were
deactivated to simulate duplex formation without multi-
valency, a much lower assembly temperature was found
(black line, Figure 1b). So, the increase in melting temper-
ature when proceeding from conventional DNA dimers to 1
and then 2 found in calculations is in qualitative agreement
with our experimental data. The additional massive increase
in assembly temperature for 3 over 1 appears to be largely the
result of electrostatics, as 3 has only four negative charges
whereas 1 and 2 have eight and twelve, respectively.
Received: August 11, 2010
Revised: December 18, 2010
Published online: March 4, 2011
Keywords: DNA · molecular modeling · nanostructures ·
.
oligonucleotides
Our results show how core structure and salt content
affect the assembly process of branched DNA hybrids. By
using CG “zippers” and organic cores, new materials are
generated in a simple manner. The discovery of compounds
such as 3 shows that many factors, including 1) increased
rigidity, 2) reduced charge, 3) optimal fraction of organic,
non-DNA matter, and 4) appropriate counterions, must come
together to achieve hybridization, and indeed solidification at
temperatures where linear DNA remains fully dissociated.
Growing macroscopic crystals and learning the rules for
embedding functional molecules[15] or substrates in the
cavities of hybrid assemblies are the next steps in the
development of hybrid-based functional materials.
In conclusion, we have shown the dramatic effect of
multivalency and a rigid, uncharged linker on the association
of branched DNA in dilute aqueous solution. Two base pairs
per DNA arm suffice to induce formation of a new material at
958C. To the best of our knowledge, the hybridization
properties of 3 are unprecedented for intermolecular duplex
formation in the field of nucleic acid chemistry.[16]
[1] N. C. Seeman, Nature 2003, 421, 427 – 431.
[2] Selected references: a) J. T. Fu, Y. C. Tse-Dinh, N. C. Seeman, J.
Mol. Biol. 1994, 236, 91 – 105; b) Z. Shen, H. Yan, T. Wang, N. C.
Seeman, J. Am. Chem. Soc. 2004, 126, 1666 – 1674; c) R. P.
Goodman, I. A. T. Schaap, C. F. Tardin, C. M. Erben, R. M.
Berry, C. F. Schmidt, A. J. Turberfield, Science 2005, 310, 1661 –
1665; d) J. Zimmermann, M. P. J. Cebulla, S. Monninghoff, G.
von Kiedrowski, Angew. Chem. 2008, 120, 3682 – 3686; Angew.
Chem. Int. Ed. 2008, 47, 3626 – 3630.
[3] Selected references: a) P. W. K. Rothemund, Nature 2006, 440,
297 – 302; b) H. Gu, J. Chao, S. J. Xiao, N. C. Seeman, Nat.
Nanotechnol. 2009, 4, 245 – 248; c) S. M. Douglas, H. Dietz, T.
Liedl, B. Hꢀgberg, F. Graf, W. M. Shih, Nature 2009, 459, 414 –
418.
[4] a) F. A. Aldaye, A. L. Palmer, H. F. Sleiman, Science 2008, 321,
1795 – 1799; b) U. Feldkamp, C. M. Niemeyer, Angew. Chem.
2006, 118, 1888 – 1910; Angew. Chem. Int. Ed. 2006, 45, 1856 –
1876; c) K. V. Gothelf, T. H. LaBean, Org. Biomol. Chem. 2005,
3, 4023 – 4037.
[5] Selected references: a) L. H. Eckardt, K. Naumann, W. M.
Pankau, M. Rein, M. Schweitzer, N. Windhab, G. von Kiedrow-
ski, Nature 2002, 420, 286; b) K. M. Stewart, J. Rojo, L. W.
McLaughlin, Chem. Commun. 2003, 2934 – 2935; c) K. M. Stew-
art, L. W. McLaughlin, J. Am. Chem. Soc. 2004, 126, 2050 – 2057;
d) K. M. Stewart, J. Rojo, L. W. McLaughlin, Angew. Chem.
2004, 116, 5932 – 5935; Angew. Chem. Int. Ed. 2004, 43, 5808 –
5811; e) F. A. Aldaye, H. F. Sleiman, Angew. Chem. 2006, 118,
2262 – 2267; Angew. Chem. Int. Ed. 2006, 45, 2204 – 2209.
[6] a) C. A. Mirkin, R. L. Letsinger, R. C. Mucic, J. J. Storhoff,
Nature 1996, 382, 607 – 609; b) A. P. Alivisatos, K. P. Johnsson,
X. Peng, T. E. Wilson, C. J. Loweth, M. P. Bruchez, Jr., P. G.
Schultz, Nature 1996, 382, 609 – 611; c) P. Hazarika, B. Ceyhan,
C. M. Niemeyer, Angew. Chem. 2004, 116, 6631 – 6633; Angew.
Chem. Int. Ed. 2004, 43, 6469 – 6471.
Experimental Section
Modeling: A system consisting of 100 particles in the gas phase was
subjected to Brownian dynamics simulated annealing and the final
structures were analyzed. Particles interacted through four or six
“arms”, each of which represented a DNA linker attached to the core
(corresponding to tetrahedral and octahedral symmetry, respec-
tively). Two arms belonging to different particles interacted by an
effective distance- and angle-dependent potential (see the Supporting
Information). An order parameter was introduced by associating a
straight (undirected) line in space with each arm. In a highly ordered
or periodic system the orientations of these lines are strongly
correlated and fall into a small number of large clusters, otherwise
there are many different clusters of lines with different orientations.
We characterized the degree of order at any point in the simulation by
computing the fraction of particles contributing to the largest cluster
of lines in the system (see the Supporting Information).
[7] T. A. Tanton, C. A. Mirkin, R. L. Letsinger, Science 2000, 289,
1757 – 1760.
[8] a) R. Jin, G. Wu, Z. Li, C. A. Mirkin, G. C. Schatz, J. Am. Chem.
Soc. 2003, 125, 1643 – 1654; b) R. J. Macfarlane, M. R. Jone, A. J.
Senesi, K. L. Young, B. Lee, J. Wu, C. A. Mirkin, Angew. Chem.
2010, 122, 4693 – 4696; Angew. Chem. Int. Ed. 2010, 49, 4589 –
4592.
[9] a) D. Nykypanchuk, M. M. Maye, D. van der Lelie, O. Gang,
Nature 2008, 451, 549 – 552; b) S. Y. Park, A. K. R. Lytton-Jean,
B. Lee, S. Weigand, G. C. Schatz, C. A. Mirkin, Nature 2008, 451,
553 – 556.
UV-melting analysis: Lyophilized hybrid samples were taken up
in NaOH solution (100 mL, 10 mm) and heated to 958C for full
denaturation, followed by neutralization with acetic acid (100 mL,
10 mm) at 958C and addition of buffer (10 mm triethylammonium
acetate or 10 mm phosphate buffer) to establish conditions for base
pairing. For samples “with salt”, stock solutions giving 150 mm NaCl
[10] M. Meng, C. Ahlborn, M. Bauer, O. Plietzsch, S. A. Soomro, A.
Singh, T. Muller, W. Wenzel, S. Brꢁse, C. Richert, ChemBio-
Chem 2009, 10, 1335 – 1339.
3230
www.angewandte.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2011, 50, 3227 –3231

[11] J. Zheng, J. J. Birktoft, Y. Chen, T. Wang, R. Sha, P. E.
Constantinou, S. L. Ginnell, C. Mao, N. C. Seeman, Nature
2009, 461, 74 – 77.
[12] C. Richert, M. Meng, A. Singh, Small 2009, 5, 2782 – 2783.
[13] T. Hatano, T. Kato, Tetrahedron 2008, 64, 8368 – 8380.
[14] O. Plietzsch, C. I. Schilling, M. Tolev, M. Nieger, C. Richert, T.
Muller, S. Brꢁse, Org. Biomol. Chem. 2009, 7, 4734 – 4743.
[15] For all-DNA structures with embedded proteins, see for
example: a) C. M. Erben, R. P. Goodman, A. J. Turberfield,
Angew. Chem. 2006, 118, 7574 – 7577; Angew. Chem. Int. Ed.
2006, 45, 7414 – 7417; b) P. J. Paukstelis, J. Am. Chem. Soc. 2006,
128, 6794 – 6795.
[16] For exceptionally stable intramolecular duplexes, see for exam-
ple: K. M. Herrlein, R. L. Letsinger, Angew. Chem. 1997, 109,
622 – 624; Angew. Chem. Int. Ed. Engl. 1997, 36, 599 – 601.
Angew. Chem. Int. Ed. 2011, 50, 3227 –3231
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3231