Light Regulation of DNA Minicircle Dimerization by Utilizing Azobenzene C-Nucleosides

Article abstract of DOI:10.1002/chem.201706003

Although DNA has the ability to form almost every desired shape, the usability of DNA nanostructures can be limited due to the lack of functionality. To functionalize nanoscale structures, light-responsive moieties like photoswitchable azobenzenes can be introduced into DNA. Upon UV irradiation, the isomerization of the azobenzene moiety causes destabilization of the neighboring base pairs leading to decreased binding ability. The linker strategy of the azobenzene to the DNA alters the performance of the switching behavior significantly. We hereby report the utilization of four different azobenzene C-nucleosides and compare their features in a nanoarchitecture model with the help of gel-electrophoresis and atomic force microscope-imaging.

Full text of DOI:10.1002/chem.201706003

DOI: 10.1002/chem.201706003
Communication
&
Photoswitchable Nanoarchitectures |Hot Paper|
Light Regulation of DNA Minicircle Dimerization by Utilizing
Azobenzene C-Nucleosides
Nikolai Grebenovsky,[a] Thomas Goldau,[a] Michael Bolte,*[b] and Alexander Heckel*[a]
radiation with different wavelengths is desirable. So-called
Abstract: Although DNA has the ability to form almost
photoswitches, for example, azobenzenes,[16] are currently used
every desired shape, the usability of DNA nanostructures
in a variety of DNA nanodevices, like nano-tweezers,[17] capsule
can be limited due to the lack of functionality. To function-
structures,[18–20] light-modulated tetrahedral structures[21] and
alize nanoscale structures, light-responsive moieties like
many more.[22]
photoswitchable azobenzenes can be introduced into
Despite the fast growing variety of photoswitchable DNA
DNA. Upon UV irradiation, the isomerization of the azo-
nanoarchitectures, most designs rely on a linker system be-
benzene moiety causes destabilization of the neighboring
tween an azobenzene moiety and the oligonucleotide, which
base pairs leading to decreased binding ability. The linker
has been established by Asanuma et al. in 2001.[23,24] It consists
strategy of the azobenzene to the DNA alters the per-
of a diol-linker derived from d-threoninol, which is incorporat-
formance of the switching behavior significantly. We
ed as acyclic surrogate for the native nucleoside (Figure 1a,
hereby report the utilization of four different azobenzene
C-nucleosides and compare their features in a nanoarchi-
tecture model with the help of gel-electrophoresis and
atomic force microscope-imaging.
The fact, that the three dimensional structure of oligonucleo-
tides is directly determined by the sequence of the nucleobas-
es and their hybridization partners, led to the discovery of a
huge variety of two- and three-dimensional constructs.
Although these DNA nanoarchitectures[1–3] and DNA origami[4]
are already very complex structures, it is tempting to add func-
tionality to utilize them.[5] This can be achieved in several ways,
for example as DNA–protein hybrids,[6,7] nanoparticle-doped
DNA motifs,[8] organochemical conjugates,[9] DNA nanopores[10]
as well as many others.[11,12] One additional approach of DNA
nanoarchitecture functionalization is the incorporation of light-
responsive elements. In contrast to most chemical stimuli, light
does not pollute the system and is harmless for biological ap-
plications if biocompatible wavelength and doses are used.
Light can be applied with high spatio-temporal control and
dose-precision.[13–15] To harness the energy of light and trans-
late it into a defined motion or reaction, a reversible system,
which can be switched between different states by alternate ir-
Figure 1. a) Overview of the different azobenzene modifications used in this
study, b) by switching the azobenzene-moieties with visible or UV light to
trans or cis state, respectively, the dimerization of the azobenzene-modified
DNA minicircles can be controlled.
tAzo). Different linker strategies were investigated but very few
gained a popularity comparable to the so-called tAzo linker
system.[22] However, its switching performance can still be im-
proved-especially regarding its operation at room temperature:
Recently, azobenzene C-nucleosides have been introduced to
successfully control the hybridization properties of RNA[25] and
DNA.[26]
[a] N. Grebenovsky, T. Goldau, Prof. Dr. A. Heckel
Institute of Organic Chemistry and Chemical Biology
J. W. Goethe University Frankfurt
Max-von-Laue-Straße 7, d-60438 Frankfurt am Main (Germany)
E-mail: heckel@uni-frankfurt.de
Those C-nucleosides mimic natural nucleosides and repre-
sent a promising new approach of photocontrol of oligonu-
cleotide dimerization behavior (Figure 1a, mAzo).[27] Since pre-
vious studies suggested that the orientation of the nucleosidic
linker and the azobenzene in the DNA strand are crucial for its
switching capabilities,[25] in this study we additionally introduce
a methoxy-group on the 2’-position of the azobenzene C-nu-
cleoside (Figure 1a, MeO-mAzo). We expected this to alter ori-
[b] Dr. M. Bolte
Institute for Inorganic Chemistry
J. W. Goethe University Frankfurt
Max-von-Laue-Straße 7, d-60438 Frankfurt am Main (Germany)
E-mail: bolte@chemie.uni-frankfurt.de
Supporting information and the ORCID identification number(s) for the au-
thor(s) of this article can be found under:
https://doi.org/10.1002/chem.201706003.
Chem. Eur. J. 2018, 24, 1 – 5
1
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
entation and flexibility of the nucleoside similar to the differ-
ence of these features in DNA and RNA. Additionally, the inves-
tigation of halogen-doped[28] or heteroarene-modified[29] azo-
benzenes, which allow for example a bathochromically shifted
photoisomerization, might lead to interesting developments of
photoresponsive nanostructures for biological applications,
since penetration depth in biological tissues increases with
red-shifted absorbance.[30] Therefore, recently developed azo-
benzenes including pyrazole heterocycles have been investi-
gated here as well, since their capability of cis to trans photo-
isomerization upon irradiation at 530 nm is highly increased
compared to unmodified azobenzene.[31]
phosphorylated by T4 polynucleotide kinase (T4PNK) and ligat-
ed by T4 ligase, followed by denaturing PAGE and gel extrac-
tion. The double-stranded DNA minicircles consist of the
single-stranded ring (strands 1+2), three strands used as struc-
ture-giving elements (strands 3, 4, 5) and one strand, which is
responsible for the dimerization of the minicircles (strand X).
This modular approach allows changes of the binding site by
simply exchanging one strand, without altering the whole
design. The DNA minicircles were annealed as mentioned
above. For gel experiments, no further purification was
needed, whereas a molecular weight cut-off by spin-filtration
was applied before preparing samples for AFM imaging. DNA
rings and their dimers could clearly be visualized by AFM for
qualitative analysis (Figure 2b). We measured the rings to be
approximately 20 nm in diameter, 0.5 nm in height on the
poly-l-ornithine-coated surface and with a bridge of 10 nm in
length on the average, which is in accordance with previous
findings (Figure 2c, the arrow indicates the bridge between
the rings).[32,36]
To compare the influence of these different photoswitches
on hybridization behavior and its application in DNA nanoarch-
itectures, a nanoarchitecture model system was needed, which
should represent the following key features: it should be
simple in design, so that it can be produced fast and easily.
The design should be modular, so that a screening of different
modifications can be done without altering the whole struc-
ture. Additionally, the structure should be planar, so that the
imaging of the structures by scanning probe microscopy tech-
niques is easy. For this reason we utilized azobenzene C-nu-
cleoside-modified DNA minicircles as a nanoarchitecture model
system. These lariat-shaped DNA minicircles, which bear a 10-
base-long single-stranded overhang as binding region, can di-
merize to form handcuff-shaped structures (Figure 1b). This
model-system has been chosen, since the minicircles are easy
to synthesize and can be visualized both by atomic force mi-
croscopy (AFM) and polyacrylamide gel electrophoresis (PAGE).
Similar systems have already been used to investigate dimeri-
zation features of polyamide tethered DNA minicircles,[32–33]
DNA–rotaxanes and –catenanes[35–37] and upon uncaging of
photocleavable protection groups.[38]
The sequence design for the binding region was a crucial
step in this study. C-nucleosidic linkers have shown to be most
effective if substituted for nucleobases in the binding se-
quence (Figure 3a, replacement motif).[25,39] For this, a screen-
Figure 3. a) Schematic display of the binding regions used in this study,
b) dimerization-ratios [%] which could be achieved upon irradiation of the
samples at 365 nm, 420 nm and 530 nm.
The chemical synthesis of the azobenzene C-nucleoside
phosphoramidites and the subsequent oligonucleotide synthe-
sis is described in detail in the Supporting Information, Sec-
tion 1 and 2. For the preparation of DNA minicircles, the inner
single-stranded ring has to be synthesized first by splint liga-
tion (Figure 2a). For that, the two strands forming the
168 bases long single-stranded ring (Figure 2a, strands 1+2),
were annealed with two splints by heating to 908C and pas-
sively cooling over a time of 30 minutes (Figure 2a, strand 5
and splint). The annealed construct was then subsequently
ing by exchanging one, two, and three nucleobases in the
binding sequence with azobenzene C-nucleosides was carried
out. We found, that introducing one modification still yields a
high dimerization ratio of the DNA minicircles in native PAGE,
but with little to no photocontrol of the dimerization behavior
(for more details see Supporting Information S20, Strand 6+
Figure 2. a) Synthetic approach of creating lariat-type DNA minicircles, b) AFM Image of dimerized DNA minicircles on previously poly-l-ornithine treated
mica-surface (dimers highlighted by rectangles), c) close-up AFM-image of a DNA minicircle dimer. The arrow indicates the linking bridge between the rings.
&
&
Chem. Eur. J. 2018, 24, 1 – 5
www.chemeurj.org
2
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
11-1). Three modifications, on the other hand, introduce such a
sterical restrain to the binding sequence, that dimerization
almost completely ceased in either photostationary state (S20,
Strand 6+11-3). First signs of a photocontrolled dimerization
behavior could be shown for two modifications, although not
in desirable extent (S20, Strand 6+11-2). To reduce the steric
repulsion caused by the azobenzene-modifications, abasic sites
were introduced in the counter strand on the corresponding
positions, so that the bulky azobenzene moieties have more
space to intercalate between the neighboring nucleobases.
This motif was tested for two and three azobenzene modifica-
tions. It could be shown, that good photocontrol could be ach-
ieved for two azobenzene modifications, were three modifica-
tions still lacked sufficient dimerization (S20, 6-5+11-2 and 6-
5+11-3). Utilizing this sequence pattern, the four different azo-
benzene-moieties were incorporated and compared in their
ability to alter the dimerization behavior upon irradiation with
appropriate wavelength.
Importantly, for the mAzo, MeO-mAzo, p-pyrAzo and m-
pyrAzo designs it was irrelevant if the trans to cis isomerization
was performed before or after annealing of the two single
rings. This is a feature that we had investigated in previous
studies[25,26] and is in contrast to the behavior acyclic linker sys-
tems where it is difficult to operate the photoswitch at room
temperature once the duplex has been formed.
In summary, we could show that azobenzene C-nucleosides
can be used to introduce a significant photocontrol into
higher-ordered DNA motifs and objects. Although the dimeri-
zation ratios in the trans state are slightly lower than in the
native sequence, azobenzene C-nucleoside linkers are a power-
ful tool, when it is necessary to switch off dimerization after
trans-to-cis isomerization. This can be used to build DNA nano-
machines and other DNA structures, which rely on a high ratio
in conformational switching with the possibility to switch di-
merization off to a high degree.
Figure 3 gives an overview of the results of the dimerization
and its photoswitching obtained with this new set of sequence
design rules: With a native 10mer complementary region a di-
merization degree of 44% could be achieved in 1ꢁTA-buffered
native PAGE (50mm Tris, 100mm NaOAc, pH 8.0) and an in-gel
temperature of 288C. With two mAzo residues-installed in the
replacement motif and with abasic sites on the counter strand-
a dimerization of 39% was achieved in the trans state. This di-
merization degree is slightly smaller compared to the wild
type and shows, that even in the optimized binding region
with abasic sites some sterical repulsion remains. After irradia-
tion of a 50 pmol sample for 120 s with 365 nm (250 mW) the
dimerization ratio could be decreased to only 6% in the cis
photostationary state. Exchanging the mAzo residues for MeO-
mAzo units resulted in a somewhat reduced dimerization
switching amplitude with only 30% dimerization in the trans
and 11% in the cis photostationary state. Under the same con-
ditions, in the photostationary state the p-pyrAzo residue and
the m-pyrAzo residues both led to a switching performance
almost identical to the one of mAzo within error limits-when
operated with 530 nm (50 pmol, 600 mW, 120 s) for the cis to
trans isomerization to account for the different light absorb-
ance properties. Both the overall decreased binding affinity in
trans state compared to the native binding sequence as well
as the high photocontrol on the dimerization ratio can be ex-
plained by the rigidity of the nucleosidic linker. Visualizing the
azobenzene as a photoswitchable lever, the C-nucleoside linker
system resembles a rigid anchor point, which supports the
prying movement of the lever better than a flexible anchor
point of acyclic linker systems. It keeps the photoswitch in
place, avoiding evasion by alternative placements in the major
or minor groove of the duplex or a reduction of dimerization
switching amplitude by loss of conformational perturbation
due to too many internal degrees of freedom. In our series of
investigated photoswitchable C-nucleoside residues MeO-
mAzo showed a less optimal performance. This shows once
more that subtle changes in the linker can have significant ef-
fects on the performance of the photo switch.
Experimental Section
Organic syntheses including spectroscopic data are described in
detail in Supporting Information Section 1. Synthesis of single
stranded DNA Rings: A mixture of strand 1, 2, splint and strand 5
was heated to 908C and passively cooled down to room tempera-
ture to anneal the constructs, followed by 5’-phosphorylation and
ligation. After ligation, samples were purified by 5% denaturing
PAGE and gel extraction. For synthesis of double stranded DNA
rings, single stranded ring was mixed with strand 3, 4, 5 and an
azobenzene doped strand X and annealed as mentioned before
(for further details, see Supporting information Section 4). For irra-
diation experiments 2–3 pmol of each azobenzene-modified ring
where mixed after irradiation at 365 nm, 420 nm or 530 nm (details
listed in Supporting Information Section 5.3) with the same
amount of binding partner in ultrapure water, before mixing with
loading buffer and pipetting into pockets, to ensure that needed
salt concentration for sufficient dimerization of the 10mer binding
site is present at the same time for every sample. Gels were run
and imaged as described in Supporting Information Section 5. The
quantitative analysis of the narrow gel bands was carried out with
the program GelBandFitter from Mitov et al. according to the de-
scribed procedure.[40]
Acknowledgements
The authors would like to thank the Deutsche Forschungsge-
meinschaft (INST 161–761-1 FUGG and SFB 902) and the “Dr.
Illing Stiftung fꢂr Molekulare Chemie” for funding. Also the au-
thors would like to thank their colleagues Thomas Halbritter,
ˇ
Matiss Reinfelds, Patrick Seyfried and Dr. Tomas Slanina for sci-
entific discussion.
Conflict of interest
The authors declare no conflict of interest.
Keywords: AFM · DNA nanoarchitecture · photoregulation ·
photoswitches
Chem. Eur. J. 2018, 24, 1 – 5
www.chemeurj.org
3
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ

Communication
[1] N. C. Seeman, Nature 2003, 421, 427–431.
[24] H. Asanuma, T. Takarada, T. Yoshida, D. Tamaru, X. Liang, M. Komiyama,
Angew. Chem. Int. Ed. 2001, 40, 2671–2673; Angew. Chem. 2001, 113,
2743–2745.
[25] T. Goldau, K. Murayama, C. Brieke, S. Steinwand, P. Mondal, M. Biswas, I.
Burghardt, J. Wachtveitl, H. Asanuma, A. Heckel, Chem. Eur. J. 2015, 21,
2845–2854.
[26] T. Goldau, K. Murayama, C. Brieke, H. Asanuma, A. Heckel, Chem. Eur. J.
2015, 21, 17870–17876.
[27] A. S. Lubbe, W. Szymanski, B. L. Feringa, Chem. Soc. Rev. 2017, 46, 1052–
1079.
[2] A. Heckel, M. Famulok, Biochimie 2008, 90, 1096–1107.
[3] DNA in Supramolecular Chemistry and Nanotechnology (Eds.: E. Stulz,
G. H. Clever), Wiley, Germany, 2015.
[4] P. Wang, T. A. Meyer, V. Pan, P. K. Dutta, Y. Ke, Chem 2017, 2, 359–382.
[5] Y. Krishnan, F. C. Simmel, Angew. Chem. Int. Ed. 2011, 50, 3124–3156;
Angew. Chem. 2011, 123, 3180–3215.
[6] B. Saccꢃ, C. M. Niemeyer, Chem. Soc. Rev. 2011, 40, 5910–5921.
[7] F. Praetorius, H. Dietz, Science 2017, 355, eaam5488.
[8] F. N. Gꢂr, F. W. Schwarz, J. Ye, S. Diez, T. L. Schmidt, ACS Nano 2016, 10,
5374–5382.
[28] D. Blꢅger, J. Schwarz, A. M. Brouwer, S. Hecht, J. Am. Chem. Soc. 2012,
134, 20597–20600.
[29] a) J. Calbo, C. E. Weston, A. J. P. White, H. S. Rzepa, J. Contreras-Garcꢆa,
[9] N. V. Voigt, T. Tørring, A. Rotaru, M. F. Jacobsen, J. B. Ravnsbæk, R. Subra-
mani, W. Mamdouh, J. Kjems, A. Mokhir, F. Besenbacher, K. Vesterager
Gothelf, Nat. Nanotechnol. 2010, 5, 200–203.
[10] S. Krishnan, D. Ziegler, V. Arnaut, T. G. Martin, K. Kapsner, K. Henneberg,
A. R. Bausch, H. Dietz, F. C. Simmel, Nat. Commun. 2016, 7, 12787.
[11] R. Schreiber, J. Do, E. Roller, T. Zhang, V. J. Schꢂller, P. C. Nickels, J. Feld-
mann, T. Liedl, Nat. Nanotechnol. 2014, 9, 74–78.
[12] O. I. Wilner, I. Willner, Chem. Rev. 2012, 112, 2528–2556.
[13] C. Brieke, F. Rohrbach, A. Gottschalk, G. Mayer, A. Heckel, Angew. Chem.
Int. Ed. 2012, 51, 8446–8476; Angew. Chem. 2012, 124, 8572–8604.
M. J. Fuchter, J. Am. Chem. Soc. 2017, 139, 1261–1274; b) V. Adam, D. K.
ˇ
Prusty, M. Centola, M. Skugor, J. S. Hannam, J. Valero, B. Klçckner, M. Fa-
mulok, Chem. Eur. J. https//doi.org/10.1002/chem.201705500.
[30] J. L. Sandell, T. C. Zhu, J. Biophotonics 2011, 4, 773–787.
[31] C. E. Weston, R. D. Richardson, P. R. Haycock, A. J. P. White, M. J. Fuchter,
J. Am. Chem. Soc. 2014, 136, 11878–11881.
[32] T. L. Schmidt, A. Heckel, Small 2009, 5, 1517.
[33] T. L. Schmidt, C. K. Nandi, G. Rasched, P. P. Parui, B. Brutschy, M. Famu-
lok, A. Heckel, Angew. Chem. Int. Ed. 2007, 46, 4382; Angew. Chem.
2007, 119, 4460.
[34] G. Rasched, D. Ackermann, T. L. Schmidt, P. Broekmann, A. Heckel, M. Fa-
mulok, Angew. Chem. Int. Ed. 2008, 47, 967; Angew. Chem. 2008, 120,
981.
[35] D. Ackermann, T. L. Schmidt, J. S. Hannam, C. S. Purohit, A. Heckel, M.
Famulok, Nat. Nanotechnol. 2010, 5, 436.
´
[14] W. Szymanski, J. M. Beierle, H. A. V. Kistemaker, W. A. Velema, B. L. Ferin-
ga, Chem. Rev. 2013, 113, 6114–6178.
[15] P. Klꢄn, T. Solomek, C. G. Bochet, A. Blanc, R. Givens, M. Rubina, V. Popik,
A. Kostikov, J. Wirz, Chem. Rev. 2013, 113, 119–191.
[16] D. Blꢅger, S. Hecht, Angew. Chem. 2015, 127, 11494–11506; Angew.
Chem. Int. Ed. 2015, 54, 11338–11349.
[17] X. G. Liang, H. Nishioka, N. Takenaka, H. Asanuma, ChemBioChem 2008,
9, 702–705.
[36] D. P. N. GonÅalves, T. L. Schmidt, M. B. Koeppel, A. Heckel, Small 2010, 6,
1347–1352.
[37] T. L. Schmidt, A. Heckel, Nano Lett. 2011, 11, 1739.
[38] T. L. Schmidt, M. B. Koeppel, J. Thevarpadam, D. P. N. GonÅalves, A.
Heckel, Small 2011, 7, 15, 2163–2216.
[18] F. Tanaka, T. Mochizuki, X. G. Liang, H. Asanuma, S. Tanaka, K. Suzuki, S.
Kitamura, A. Nishikawa, K. Ui-Tei, M. Hagiya, Nano Lett. 2010, 10, 3560–
3565.
[19] D. Bhatia, S. Mehtab, R. Krishnan, S. S. Indi, A. Basu, Y. Krishnan, Angew.
Chem. Int. Ed. 2009, 48, 4134–4137; Angew. Chem. 2009, 121, 4198–
4201.
[39] P. Mondal, M. Biswas, T. Goldau, A. Heckel, I. Burghardt, J. Phys. Chem. B
2015, 119, 11275–11286.
[40] M. I. Mitov, M. L. Greaser, K. S. Campbell, Electrophoresis 2009, 30, 848–
851.
[20] D. Bhatia, S. Surana, S. Chakraborty, S. P. Koushika, Y. Krishnan, Nat.
Commun. 2011, 2, 339.
[21] D. Han, J. Huang, Z. Zhu, Q. Yuan, M. X. You, Y. Chen, W. H. Tan, Chem.
Commun. 2011, 47, 4670–4672.
[22] J. Li, X. Wang, X. Liang, Chem. Asian J. 2014, 9, 3344–3358.
[23] H. Asanuma, H. Kashida, Y. Kamiya, Chem. Rec. 2014, 14, 1055–1069.
Manuscript received: December 19, 2017
Accepted manuscript online: January 16, 2018
Version of record online: && &&, 0000
&
&
Chem. Eur. J. 2018, 24, 1 – 5
www.chemeurj.org
4
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Communication
COMMUNICATION
Light-responsive DNA nanoarchitec-
ture can be created by incorporating
light-switchable azobenzene moieties
into the DNA. Upon irradiation, the iso-
merization of the azobenzene-moiety
causes destabilization of the basepairs
in circumfering basepairs leading to de-
creased binding ability. The introduction
of azobenzene C-nucleosides in DNA
nanoarchitecture leads to superior
switching ratios with a high off-rate.
&
Photoswitchable Nanoarchitectures
N. Grebenovsky, T. Goldau, M. Bolte,*
A. Heckel*
&& – &&
Light Regulation of DNA Minicircle
Dimerization by Utilizing Azobenzene
C-Nucleosides
Chem. Eur. J. 2018, 24, 1 – 5
www.chemeurj.org
5
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ