Ortho -Fluoroazobenzene derivatives as DNA intercalators for photocontrol of DNA and nucleosome binding by visible light

Article abstract of DOI:10.1039/c8ob02343c

We report a high-affinity photoswitchable DNA binder, which displays different nucleosome-binding capacities upon visible-light irradiation. Both photochemical and DNA-recognition properties were examined by UV-Vis, HPLC, CD spectroscopy, NMR, FID assays,

Full text of DOI:10.1039/c8ob02343c

Organic &
Biomolecular Chemistry
View Article Online
View Journal
COMMUNICATION
ortho-Fluoroazobenzene derivatives as DNA
intercalators for photocontrol of DNA and
Cite this: DOI: 10.1039/c8ob02343c
nucleosome binding by visible light†
Received 20th September 2018,
Accepted 19th December 2018
Benedikt Heinrich, a Karim Bouazoune, b Matthias Wojcik,
c
a
Udo Bakowsky c and Olalla Vázquez
*
DOI: 10.1039/c8ob02343c
rsc.li/obc
We report a high-affinity photoswitchable DNA binder, which dis- and strengthened our understanding of genome regulation.
plays different nucleosome-binding capacities upon visible-light However, this approach has mainly been limited to enzyme
irradiation. Both photochemical and DNA-recognition properties inhibitors6–9 and altering nucleosome accessibility has largely
were examined by UV-Vis, HPLC, CD spectroscopy, NMR, FID been disregarded. Indeed, while there is just a handful of
assays, EMSA and DLS. Our probe sets the basis for developing new examples of compounds capable of targeting nucleosomal
optoepigenetic tools for conditional modulation of nucleosomal DNA, to our knowledge, none of these molecules allow spatio-
DNA accessibility.
temporal control of nucleosome binding.10–15
Reversible photoresponsive molecules, which have demon-
strated their potential in diverse areas (material science,16,17
molecular motors18–20 photopharmacology,21–24 molecular
Over the last decades, a large number of studies have estab-
lished that the organization of eukaryotic DNA, by proteins
and RNAs, into a nucleoprotein complex referred to as chroma-
tin is key to regulate genome functions.1 The most abundant
chromatin proteins are the histones. These proteins tightly
wrap DNA into a “beads-on-a-string”-like structure. Each ‘bead’
is referred to as a nucleosome and comprises a histone
octamer around which about 150 base pairs (bp) of DNA are
wrapped. On average, nucleosomes are separated by about
50 bp of free (also called “linker”) DNA.2 This entails that
about 75% of our DNA is wrapped around histones. This
nucleosomal organization restricts access to DNA considerably.
As a result, nuclear processes such as DNA repair, replication
and transcription largely depend on enzymes that can change
nucleosomal DNA accessibility.3–5
Hence, the ability to control chromatin compaction and, in
turn, DNA accessibility using small molecules may provide us
with means to study and, conceivably, control (at least some)
genome functions, in physiological as well as disease contexts.
So far, the use of small-molecule probes as complementary
tools to classic chromatin biochemical approaches has
strongly contributed to deciphering epigenetic mechanisms
containers,25 etc.) represent
a very promising chemical
alternative to optogenetics.26,27 However, the implementation
of photoswitchable genome regulators has been very scarce
and again only focused on histone-modifying enzymes, so
far.28–30 Therefore, we designed and synthesized a novel
photoswitch, which allows light-driven DNA and nucleosome
binding and we studied the consequences of this interaction
(Fig. 1a).
The ability of externally controlling DNA-associated pro-
cesses by switching the state of a photochromic DNA binder
has previously been demonstrated.31,32 However, such
examples are mainly restricted to the use of azo-modified DNA
binders, which involve undesirable UV irradiation.33–38 There
are several strategies to synthesize azobenzene photoswitches
that undergo isomerization upon visible-light irradiation.39–42
For example, the Hecht group has recently optimized the pro-
perties of the classical azobenzenes by introducing σ-electron-
withdrawing fluorine atoms in ortho position to the azo-
benzene unit. Such substitution leads to visible-light switches
with high photoconversions and very long-lived cis-
isomers.43,44 Furthermore, to our knowledge, photoswitchable
DNA binders have never been used in the context of the
nucleosome. Consequently, inspired by the natural DNA minor
groove binder netropsin (1), we synthesized and studied two
water-soluble photosensitive pyrrole hybrids that have identi-
cal number of N-methyl pyrrole rings as netropsin, but incor-
porate an ortho-fluoroazobenzene scaffold between the pyrrole
backbone and also vary in terms of their net charge at physio-
aFachbereich Chemie, Philipps-Universität Marburg, Hans-Meerwein-Straße 4, 35043
Marburg, Germany. E-mail: olalla.vazquez@staff.uni-maburg.de
bInstitut für Molekularbiologie und Tumorforschung, Philipps-Universität Marburg,
Hans-Meerwein-Straße 2, 35043 Marburg, Germany
cInstitut für Pharmazeutische Technologie & Biopharmazie, Philipps-Universität
Marburg, Robert-Koch-Straße 4, 35037 Marburg, Germany
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c8ob02343c
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
Fig. 2 (a) UV-Vis spectra of a 20 μM solution of F4Azo-(PyDp)2 (2) in
10 mM Tris buffer pH 7.6, 50 mM KCl and DMSO (98 : 2), initially at the
thermodynamic state (blue) and after successive irradiation intervals at
520 nm (green). (b) Reversible photochromism upon alternating
irradiation at 520 nm (green) and 405 nm (blue) measured at 335 nm.
Represented data are calculated from three independent experiments.
Fig. 1 (a) Outline of the photocontrollable nucleosome targeting
approach based on visible-light photoswitchable DNA binders. (b)
Structure of the studied molecules, highlighting the key synthetic build-
ing blocks.
logical conditions: F4Azo-(PyDp)2 (2) and F4Azo-(PyDp)(PyOMe)
(3) (Fig. 1b).
The synthesis of F4Azo-(PyDp)2 (2) and F4Azo-(PyDp)
(PyOMe) (3) was straightforward through one-pot condensation
between the previously reported building blocks: 4,45 5,43,44
and 6.46 (Fig. 1 & Schemes S1–S4†). Once synthesized, their
photochemical behaviour was investigated in detail. Thus, the
ortho-fluoroazobenzene building block 5 alone and also in
presence of the pyrrole moieties showed the expected absorp-
tion bands: intense band at λmax = 319 nm assigned to the π →
π* transition together with a weaker one due to a n → π* and
Fig. 3 Evaluation of the cis-isomer stability of F4Azo-(PyDp)2 (2) in
10 mM Tris buffer pH 7.4, 10 mM NaCl and DMSO (95 : 5) by HPLC after
initial irradiation at 520 nm for 2 min (time 0 h). Samples were stored in
total darkness. HPLC chromatograms were recorded after the listed
times. Represented data are
a set, which is representative of the
measurements from three independent experiments.
1
significant intensity decrease of the π → π* band with a slight (Fig. S20 & S23b†). This forced us to use H-NMR analysis for
increase of the n → π* one under 405 nm and 520 nm further investigations (Fig. 4). Thus, the trans/cis ratios were
irradiation, respectively (Fig. S21b†). Nevertheless, the direct calculated by the integration of the aromatic proton signals
insertion of the azophenyl derivative into the pyrrole backbone (6.8 ppm–8.1 ppm). Gratifyingly, in the case of F4Azo-(PyDp)2
affected the parental spectra, as observed for related com- (2) we corroborated the same ratio as the one obtained by
pounds.47 Irradiation of F4Azo-(PyDp)2 (2) entails fast spectro- HPLC (trans/cis ratio 93 : 7 and 30 : 70 for the trans- and cis-
scopic changes.48 Thus, upon irradiation with 520 nm there is state, respectively). In addition, H-NMR studies revealed that,
1
a clear intensity decrease between 287 nm and 500 nm, the indeed, the photochemical behaviour of F4Azo-(PyDp)(PyOMe)
λmax is slightly shifted to 278 nm and a new partly overlapping (3) is analogous to F4Azo-(PyDp)2 (2) (trans/cis ratio 97 : 3 and
band at λmax = 420 nm can be detected (Fig. 2a). The photosta- 27 : 73 for the trans- and the cis-state, respectively).
tionary state is reached after just 2 minutes of irradiation. We
also demonstrated the reversibility of the photoisomerization for explored whether the two compounds are able to interact with
up to 16 cycles without any significant photobleaching (Fig. 2b).
DNA and whether their isomers displayed any differences in
Once the photoisomerization was fully characterized, we
We next determined the isomer ratios at the photostation- DNA-binding affinity. Thermal melting experiments are the
ary state as well as the lifetime of the isomers by integrating most common methodology used for the evaluation of DNA
the peak area of the HPLC chromatograms at the isosbestic interactions; however, we observed decomposition at high
point (287 nm). As expected, the trans-isomer was the thermo- temperatures (Fig. S36†). Consequently, we performed a fluo-
dynamically stable species (trans/cis ratio 92 : 8; Fig. S23a†) rescent intercalator displacement (FID) assay49 compatible
Importantly, the thermal relaxation of the cis-isomer was slow with the isomerization. We also compared our results with the
(cis/trans ratio 77 : 23, after 9 hours stored in the dark, Fig. 3), minor groove binder netropsin (1). Thiazole orange (TO) was
which enabled both DNA and nucleosome binding experiments.
chosen as intercalator since its excitation and emission hardly
Unfortunately, F4Azo-(PyDp)(PyOMe) (3) did not show affect the trans/cis isomerization (Fig. S34†). This experiment
any major change neither by UV-Vis spectroscopy nor HPLC relies on a decrease of fluorescence due to the displacement of
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Communication
Fig. 4 Selected aromatic region of the 1H-NMR spectra (300 MHz) of a 4 mM solution of (a) F4Azo-(PyDp)2 (2) and (b) F4Azo-(PyDp)(PyOMe) (3) in
DMSO-d6 after irradiation to the trans- (top) and cis-isomer (bottom). Aromatic protons are marked in colours in the spectra and molecules.
DNA-bound TO by the competitive binder. Therefore, we first (PyOMe) (3) were able to efficiently replace the bound TO from
determined the apparent binding constant of TO in presence the DNA. The obtained data suggested binding modes beyond
of a hairpin oligonucleotide, which contains the sequence the 1 : 1 stoichiometry. For the mathematical analysis of the
ATTA (dsDNAhAT) with the HypSpec software. Subsequently we FID assays, the experimental data were fitted globally (all wave-
simulated the species distribution for the obtained binding lengths simultaneously) with HypSpec. These experimental
constant (KD = 59.8 13.1 nM) by using the HySS software. data fitted the model adequately (Fig. S33†).
The affinity of TO to dsDNAhAT enabled FID experiments at low
The analysis of the apparent dissociation constants indi-
μM range with more than 95% of TO-DNA complex in our cated that both photoswitchable compounds have high DNA-
experimental conditions (Table S4†), which assures the binding affinity in the nM range (Table 1). Importantly, for
reliability of the assay. Typical fluorescence quenchers such as both azobenzene derivatives (2 and 3), the binding capacity of
DABCYL or BHQ-1 are provided with the diazenyl functional their isomers is different. This demonstrates a conformation-
group (R–NvN–R′), therefore we initially performed control dependent binding mode. In each pair, the trans-isomer is
competition assays in presence of increasing amounts of 5 to always the best binder, which is in agreement with other pre-
rule out artefacts due to quenching effects (Fig. S38†). As vious examples of azobenzene derivatives.33–38 Furthermore,
showed in Fig. 5, both F4Azo-(PyDp)2 (2) and F4Azo-(PyDp) upon addition of the netropsin (1) control, the emission inten-
sity of DNA-bound TO also decreased significantly. The faster
saturation of the netropsin (1) in comparison with our photo-
switchable binders, together with the fact that its binding
curve does not appear to follow a standard isotherm may
be indicative of different binding modes. It was expected that
netropsin (1), as a minor groove binder, could not fully dis-
place the intercalator TO and reach saturation.50 The binding
curve of netropsin (1) with a hairpin oligonucleotide contain-
ing the sequence GGCCC (dsDNAhGC) displayed the standard
Table 1 Binding affinities derived from FID experiments with selected
Fig. 5 (a) Competitive displacement analysis of a 6 μM TO solution in
20 mM NaH2PO4 pH 7.4, 100 mM NaCl and 1 μM of dsDNAhAT with:
netropsin (1, orange stars); F4Azo-(PyDp)2 isomers (trans-2 dark blue
squares; cis-2 dark green squares) and F4Azo-(PyDp)(PyOMe) (trans-3
light blue triangles; cis-3 light green triangles); (b) sequence-selectivity
analysis of netropsin (1, stars) trans-F4Azo-(PyDp)2 (trans-2, squares) in
presence 6 μM TO solution in 20 mM NaH2PO4 pH 7.4, 100 mM NaCl
and 1 μM of dsDNAhAT (blue) or dsDNAhGC (brown). Represented data
and standard deviations are mean values calculated from three indepen-
dent experiments. Data points were fitted with HypSpec, using multi-
variant factor analysis to obtain globally optimized parameters. Lines are
“eye-guides” for assistance in visualizing the binding curves.
hairpin oligonucleotide (dsDNAh)
Compound
dsDNAh site
KD (nM)
Netropsin (1)
Netropsin (1)
cis-F4Azo-(PyDp)(PyOMe) (cis-3)
trans-F4Azo-(PyDp)(PyOMe) (trans-3)
cis-F4Azo-(PyDp)2 (cis-2)
trans-F4Azo-(PyDp)2 (trans-2)
trans-F4Azo-(PyDp)2 (trans-2)
ATTA
GGCCC
ATTA
ATTA
ATTA
ATTA
GGCCC
10
1
n.c.
82
54
6
3
108 14
60
13
1
1
n.c. represents not calculated.
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
isotherm for non-specific interaction i.e. proportional to the
Intriguingly, the addition of F4Azo-(PyDp)2 (2) induced a
concentration of added ligand (Fig. 5b).
remarkable dose-dependent increase in magnitude of the
To get additional insight into its binding mode, we per- signal in the region of 240–480 nm with a clear negative peak
formed exemplarily selectivity experiments with F4Azo-(PyDp)2 at 310 nm. Importantly, the two isomers cis-2 and trans-2
(2) and the additional hairpin dsDNAhGC (Fig. 5b) as well as behaved differently in the CD experiments under the same
circular dichroisim (CD) experiments (Fig. 6).
conditions. In particular, the ICD signals are more pro-
Regarding selectivity, F4Azo-(PyDp)2 (2) showed a slight nounced with trans-2 than with cis-2, which is concurrent with
preference for dsDNAhGC. Of note, all the obtained binding our previous FID experiments (as described above).
constants were derived from indirect assessments, based on
To understand the observed CD signature, we performed
the TO affinity to the corresponding dsDNA and its loss of UV-Vis titrations (Fig. 7). We observed that the addition of
fluorescence when it is displaced by a new binder. Therefore, F4Azo-(PyDp)2 (2) promoted a red shift of the absorbance
the technique is limited to compounds with binding affinities maximum of the dsDNACT and a new partly overlapped band
in the same rage. In addition, ligand-induced condensation or at 420 nm, as in the CD spectra. However, the CD band at
interference with the fluorescence could not be discriminated 310 nm was not clearly detected in our UV-Vis measurements.
from the actual fluorescence loss due to displacement, which Other azo-modified polyamines displayed CD signals at
may explain non-standard isotherms. Consequently, these ∼310 nm in presence of DNA but, to our knowledge, always
values are, strictly speaking, estimations. Despite these with positive ellipticity.34,53 However, it has been reported that
inherent drawbacks, the FID assay is a straightforward method strong insertion of intercalators into the DNA causes a red-
for the comparative characterization of DNA-recognition by the shifted band with high negative CD signals between 240 nm
azobenzene derivatives on the one hand, and on the other and 340 nm,54 which is consistent with our observations.
hand, by netropsin (and compounds with similar binding
modes).
All together our results have shown that F4Azo-(PyDp)2 (2) is
a photoswitchable molecule capable of interacting with the
The CD experiments of F4Azo-(PyDp)2 (2) with the double- DNA through a conformation-dependent binding mode.
stranded calf thymus DNA (dsDNACT) confirmed a non-covalent Finally, we explored the possibility of using F4Azo-(PyDp)2
interaction beyond the external non-specific electrostatic (2) as a controllable nucleosome binder using visible light. For
association with the DNA phosphates. Therefore, the observed this endeavour, we first reconstituted nucleosome core par-
interaction differs from the cationic polyamines such as sper- ticles (NCP) using chicken erythrocyte histones55 and the
mine and spermidine.51 The CD spectra of the dsDNACT alone Widom 601 DNA, which forms a stably positioned nucleo-
showed the characteristic bands of the canonical B-DNA con- some.56 We next used native gel ectrophoretic mobility shift
formation (Fig. 6, grey line): a positive band at 275 nm and a assays (EMSA) to analyse nucleosome reconstitution efficiency.
negative one at 245 nm of similar intensity.52 F4Azo-(PyDp)2 (2) We also used EMSA to assess nucleosome integrity after incu-
is achiral and, hence, optically inactive (Fig. 6, black line). bating nucleosomes with increasing amounts of either netrop-
However, when increasing amounts of F4Azo-(PyDp)2 (2) were sin (1) or F4Azo-(PyDp)2 (2) for three hours. As shown in Fig. 8,
added to the dsDNACT, induced circular dichroic (ICD) signals low concentrations of the minor groove binder netropsin (up
were detected, as expected from DNA binders.
to 4 eq.) do not affect the nucleosome integrity. In the pres-
ence of high netropsin concentrations, the bands became
very diffuse and consequently barely detectable. This phenom-
enon may be due to the formation of many different species
through unspecific electrostatic interactions57,58 and/or aggre-
Fig. 6 CD spectra of a 50 μM solution of dsDNACT in 20 mM NaH2PO4
pH 7.4, 100 mM NaCl with increasing amounts of the F4Azo-(PyDp)2
isomers: (a) trans-2, blue; (b) cis-2, green. F4Azo-(PyDp)2 amounts: 0 eq.
(grey circles); 0.16 eq. (light squares); 0.5 eq. (diamonds); 1.1 eq. (tri-
angles) and 2.8 eq. (dark squares). Black line represents the CD spectra
of a 142 μM (equivalent to 2.8 eq. in the titration) solution of F4Azo-
(PyDp)2 isomers under the same conditions: (a) trans-2 (diamonds); (b)
cis-2 (triangles). Represented data are mean values calculated from two
independent experiments; buffer subtracted; eq. means equivalents:
Fig. 7 UV-Vis spectra of a 12.5 μM solution of dsDNACT in 20 mM
NaH2PO4 pH 7.4, 100 mM NaCl with increasing amounts of the F4Azo-
(PyDp)2 isomers: (a) trans-2, blue; (b) cis-2, green. F4Azo-(PyDp)2
amounts: 0 eq. (grey circles); 0.16 eq. (light squares); 0.5 eq. (diamonds);
1.1 eq. (triangles) and 2.8 eq. (dark squares). Represented data are calcu-
lated from three independent experiments; buffer was subtracted; eq.
mol of compound 2 per mol of dsDNACT
.
means equivalents: mol of compound 2 per mol of dsDNACT.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Communication
lysis, we used NCP at 2 μM concentration, employing plasmids
containing multiple repeats of the Widom 601 sequence for
large-scale DNA preparation.60 As shown in Fig. 9, the particle
size of the NCP is consistent with previous reports.61,62
Increasing amounts of F4Azo-(PyDp)2 (2) clearly change
the size distribution resulting in larger particles than the
canonical NCP (Fig. 9a). This effect was reproducible and,
interestingly, changes were already detectable from the first
F4Azo-(PyDp)2 addition (0.06 eq.), although no alterations in
the EMSA bands were observable under the same conditions at
this concentration (Fig. S46a†). This was also contrary to the
case of netropsin, where low concentrations did not affect
the size of the particle (Fig. S42 & S46d†). Therefore, as
expected, the insertion of F4Azo-(PyDp)2 (2) caused higher
Fig. 8 Evaluation of the interaction of DNA binders with Widom 601
nucleosome core particle (NCP). All experiments were performed with structural distortion than for the minor groove binder 1. To
300 nM NCP and 55 nM dsDNA as control in 10 mM Tris buffer pH 8,
study the possibility that such distortion leads to aggregation
1 mM EDTA, 50 mM NaCl and DMSO (90 : 10) and analyzed by ethidium
bromide staining, after electrophoresis: (a) netropsin interaction: lane 1:
dsDNA; lanes 2–9: NCP + [netropsin] = 0, 32, 16, 8, 4, 2, 1, 0.5 eq.; (b)
F4Azo-(PyDp)2 (2) interaction: lane 7: NCP; lane 6: dsDNA; lane 1–5:
NCP + [cis-2] = 0.5, 0.25, 0.13, 0.06, 0.03 eq.; lane 8–12: NCP + [trans-
2] = 0.5, 0.25, 0.13, 0.06, 0.03 eq. DMSO stocks of F4Azo-(PyDp)2 (2)
were irradiated before EMSA; (c) F4Azo-(PyDp)2 (2) interaction with
in situ irradiation: lane 1: dsDNA; lane 2: NCP; lane 3–4: control with
previous irradiation of NCP + [cis-2] = 0.5, 0.25 eq.; lane 5–6: control
with previous irradiation of NCP + [trans-2] = 0.5, 0.25; lane 7–9 in situ
irradiation at 405 nm for 10 s of NCP + [cis-2] = 0.5, 0.25, 0.13 eq.; (d)
effect of irradiation at 405 nm on NCP: lane 1: dsDNA; lane 2: NCP; lane
3–7: NCP after 1 s, 5 s, 10 s, 30 s, 60 s irradiation. Eq. means equivalents:
mol of compound per mol of base pair; NCP means nucleosome core
particle.
gates, which may be too big to enter the gel. Interestingly, the
binding of our molecule F4Azo-(PyDp)2 (2) apparently triggered
this effect at lower concentrations than netropsin, despite
having the same net charge (Fig. 8a & b). Importantly, there
was a noticeable difference between isomers. In particular, the
intensity of the nucleosome band abruptly ceased to be detect-
able in the presence of 0.5 eq. of the trans-2, while at the same
concentration, the cis-2 failed to fully alter the nucleosome
band. Furthermore, the fact that we circumvented the use of
strong UV light, for the isomerization, facilitated the in situ
application. Thus, the same procedure was repeated but now
performing the irradiation in situ: after incubating the cis-2 for
1 h with the nucleosome, the samples were irradiated at
405 nm for 10 s, in a single isomerization cycle. As a control,
we also included both isomers irradiated prior to the nucleo-
Fig. 9 Effect of F4Azo-(PyDp)2 on: (a) nucleosome (NCP); (b) dsDNA601
at 2 μM concentration in 10 mM Tris 1 mM EDTA pH 8.0 and 50 mM
NaCl monitored by dynamic light scattering (DLS). F4Azo-(PyDp)2 (2)
amounts: 0 eq. (grey); 0.06 eq. (green); 0.13 eq. (blue); 0.25 eq. (black)
some incubation. Gratifyingly, we observed that the in situ
formed trans-2 showed a higher impact than the parental cis-2.
Of note, no irradiation effect on the nucleosome was detect-
able (Fig. 8d).
and 0.5 eq. (red); (c) effect of MgCl2 on NCP at 2 μM concentration in
To complement the EMSA studies and gain preliminary 3.5 mM Tris 0.35 mM EDTA pH 8.0 monitored by DLS. MgCl2 concen-
trations: 0 mM (grey); 2.5 mM (green); 3.75 mM (blue); 5 mM (black);
10 mM (red). Intensity statistics of 10 measurements. Represented data
insights into the effect of F4Azo-(PyDp)2 (2) on the NCPs, we
performed dynamic light scattering (DLS) measurements.
are
a representative one of two measurements from independent
Nucleosomes are colloidal particles of 11 nm wide,59 and
alterations in their hydrodynamic size can assist in the charac-
experiments; eq. means equivalents (mol of compound per mol of base
pair) Represented data are a set, which is representative of the measure-
terization of the interaction with small molecules. In this ana- ments from two independent experiments.
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.

View Article Online
Communication
Organic & Biomolecular Chemistry
via nucleosome disassembly, we performed control experi- Berlin) for the discussions and suggestions regarding the
ments with the analogue set-up but using the dsDNA initial isomerization irradiation experiments; Prof. A. Brehm
employed for the nucleosome assembly (dsDNA601) instead of (IMT Marburg) for critical advice and support; Prof. J-J. Song
the whole NCP, as well as, the precipitated NCP in presence of (Korea Advanced Institute of Science and Technology) for sup-
magnesium chloride; and histone octamer alone. However, the plying the plasmid pUC-19/16X601; M. Alawak and J. Schäfer
latter protein complex did not display any DLS signal under (Phillips-Universität Marburg) for initial discussions regarding
these conditions. 10 mM of magnesium chloride promoted DLS; N. Frommknecht (Phillips-Universität Marburg) for
the precipitation of the NCP (Fig. S43 & S46c†) and the appear- design assistance and construction of LED lamps; COST action
ance of single peak of higher size than the NCP alone but CM1406 (Epigenetic Chemical Biology EPICHEMBIO) for
smaller than the one obtained with highest concentrations support in networking and mutual cooperation. O. V. thanks
of F4Azo-(PyDp)2 (2) (342 nm versus 3580 nm, respectively). the SPP1926 for the Young Investigator Award and Fulbright
Interestingly, the incubation of dsDNA601 in presence of Commission for the Fulbright-Cottrell Award 2016. Finally,
F4Azo-(PyDp)2 (2) (Fig. 9B) displayed DNA-higher order struc- this work was financially supported by Fulbright Commission
tures, as reported with other cationic molecules63–65 such as and the DFG programs: SPP1926 ‘Next Generation
polyamines, surfactants, etc. Qualitatively, the distribution of Optogenetics’, Young Investigator program (grant # GO1011/
the peaks obtained from this titration were similar to the one 11-1) and TRR81: ‘Chromatin Changes in Differentiation and
observed in the presence of NCP (Fig. 9a). Therefore, these Malignancies’ (TRR81/3, Z04).
DLS data suggest that, first, the F4Azo-(PyDp)2 (2) intercalates
into the nucleosomal DNA inducing distortions, which lead to
the formation of high size aggregates. These observations are
consistent with the hypothesis that nucleosome distortion
subsequently leads to disruption. Future experiments will
Notes and references
be required to fully characterize these alterations and their
kinetics.
1 G. Felsenfeld, Cold Spring Harbor Perspect. Biol., 2014, 6(1),
a018200.
In summary, we have designed, synthesized and studied a
novel photoswitchable ortho-fluoroazobenzene DNA binder:
F4Azo-(PyDp)2. We also demonstrated that its DNA interaction
depends on the conformation, which, importantly, can be con-
trolled by visible-light. Furthermore, our study establishes the
possibility of performing photocontrollable nucleosome
binding. To our knowledge this is the first time that nucleo-
some targeting is externally modulated by visible-light photo-
switches. We believe that this approach uncovers the possi-
bility of using photosensitive chemical tools to alter nucleo-
some-based processes. At this point, it is impossible to predict
whether our molecule can preferentially impact specific cell
functions. It is more plausible that, if appropriately taken up
by cells, the molecule would alter cell functions randomly,
2 D. E. Schones, K. Cui, S. Cuddapah, T. Y. Roh, A. Barski,
Z. Wang, G. Wei and K. Zhao, Cell, 2008, 132(5), 887–898.
3 K. Luger, Chromosome Res., 2006, 14(1), 5–16.
4 C. R. Clapier and B. R. Cairns, Annu. Rev. Biochem., 2009,
78, 273–304.
5 C. D. Allis and T. Jenuwein, Nat. Rev. Genet., 2016, 17(8),
487–500.
6 C. D. Allis and T. W. Muir, ChemBioChem, 2011, 12(2), 264–
279.
7 A. Huston, C. H. Arrowsmith, S. Knapp and M. Schapira,
Nat. Chem. Biol., 2015, 11(8), 542–545.
8 J. Shortt, C. J. Ott, R. W. Johnstone and J. E. Bradner, Nat.
Rev. Cancer, 2017, 17(4), 160–183.
9 P. A. Cole, Nat. Chem. Biol., 2008, 4(10), 590–597.
since F4Azo-(PyDp)2 (2) only shows a slight sequence binding 10 J. D. Trzupek, J. M. Gottesfeld and D. L. Boger, Nat. Chem.
preference. Thus, future directions may include the develop- Biol., 2006, 2(2), 79–82.
ment of sequence-specific nucleosome binders to potentially 11 E. Y. Chua, G. E. Davey, C. F. Chin, P. Droge, W. H. Ang and
control cell functions specifically via modulation of DNA tran-
scription and/or repair processes.
C. A. Davey, Nucleic Acids Res., 2015, 43(11), 5284–5296.
12 R. K. Suto, R. S. Edayathumangalam, C. L. White,
C. Melander, J. M. Gottesfeld, P. B. Dervan and K. Luger,
J. Mol. Biol., 2003, 326(2), 371–380.
13 R. S. Edayathumangalam, P. Weyermann, J. M. Gottesfeld,
P. B. Dervan and K. Luger, Proc. Natl. Acad. Sci. U. S. A.,
2004, 101(18), 6864–6869.
Conflicts of interest
There are no conflicts to declare.
14 C. T. McMurray and K. E. van Holde, Proc. Natl. Acad.
Sci. U. S. A., 1986, 83(22), 8472–8476.
15 M. E. Nunez, K. T. Noyes and J. K. Barton, Chem. Biol.,
2002, 9(4), 403–415.
Acknowledgements
The authors gratefully acknowledge: Prof. E. Meggers (Phillips- 16 M. M. Russew and S. Hecht, Adv. Mater., 2010, 22(31),
Universität Marburg) for the accessibility to his S1 laboratory; 3348–3360.
Prof. P. Ballester (ICIQ) for the insight into the use of HypSpec 17 J. Karcher and Z. L. Pianowski, Chem. – Eur. J., 2018,
and HySS software; Prof. S. Hecht (Humboldt-Universität zu
24(45), 11605–11610.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Communication
18 L. A. Huber, K. Hoffmann, S. Thumser, N. Bocher, P. Mayer 41 M. X. Dong, A. Babalhavaeji, S. Samanta, A. A. Beharry and
and H. Dube, Angew. Chem., Int. Ed., 2017, 56(46), 14536–
14539.
19 A. S. Lubbe, Q. Liu, S. J. Smith, J. W. de Vries,
G. A. Woolley, Acc. Chem. Res., 2015, 48(10), 2662–2670.
42 D. Bleger and S. Hecht, Angew. Chem., Int. Ed., 2015, 54(39),
11338–11349.
J. C. M. Kistemaker, A. H. de Vries, I. Faustino, Z. Meng, 43 D. Bleger, J. Schwarz, A. M. Brouwer and S. Hecht, J. Am.
W. Szymanski, A. Herrmann and B. L. Feringa, J. Am. Chem.
Soc., 2018, 140(15), 5069–5076.
20 T. van Leeuwen, A. S. Lubbe, P. Stacko, S. J. Wezenberg and
B. L. Feringa, Nat. Rev. Chem., 2017, 1(12), 96.
Chem. Soc., 2012, 134(51), 20597–20600.
44 C. Knie, M. Utecht, F. Zhao, H. Kulla, S. Kovalenko,
A. M. Brouwer, P. Saalfrank, S. Hecht and D. Bleger,
Chemistry, 2014, 20(50), 16492–16501.
21 J. Broichhagen, J. A. Frank and D. Trauner, Acc. Chem. Res., 45 C. Hotzel, A. Marotto and U. Pindur, Eur. J. Med. Chem.,
2015, 48(7), 1947–1960. 2002, 37(5), 367–378.
22 K. Hull, J. Morstein and D. Trauner, Chem. Rev., 2018, 46 N. R. Wurtz, J. M. Turner, E. E. Baird and P. B. Dervan, Org.
118(21), 10710–10747. Lett., 2001, 3(8), 1201–1203.
23 M. M. Lerch, M. J. Hansen, G. M. van Dam, W. Szymanski 47 R. Sekiya, A. Diaz-Moscoso and P. Ballester, Chemistry,
and B. L. Feringa, Angew. Chem., Int. Ed., 2016, 55(37),
10978–10999.
24 W. A. Velema, W. Szymanski and B. L. Feringa, J. Am.
Chem. Soc., 2014, 136(6), 2178–2191.
25 A. Diaz-Moscoso and P. Ballester, Chem. Commun., 2017,
53(34), 4635–4652.
26 M. Banghart, K. Borges, E. Isacoff, D. Trauner and
R. H. Kramer, Nat. Neurosci., 2004, 7(12), 1381–1386.
27 L. Fenno, O. Yizhar and K. Deisseroth, Annu. Rev. Neurosci.,
2011, 34, 389–412.
2018, 24(9), 2182–2191.
48 C. W. Ong, Y. T. Yang, M. C. Liu, K. R. Fox, P. H. Liu and
H. W. Tung, Org. Biomol. Chem., 2012, 10(5), 1040–1046.
49 D. L. Boger, B. E. Fink, S. R. Brunette, W. C. Tse and
M. P. Hedrick, J. Am. Chem. Soc., 2001, 123(25), 5878–5891.
50 Y. W. Ham, W. C. Tse and D. L. Boger, Bioorg. Med. Chem.
Lett., 2003, 13(21), 3805–3807.
51 H. Deng, V. A. Bloomfield, J. M. Benevides and G. J. Thomas
Jr., Nucleic Acids Res., 2000, 28(17), 3379–3385.
52 J. Kypr, I. Kejnovska, D. Renciuk and M. Vorlickova, Nucleic
Acids Res., 2009, 37(6), 1713–1725.
28 L. Albert, J. Xu, R. W. Wan, V. Srinivasan, Y. L. Dou and
O. Vazquez, Chem. Sci., 2017, 8(6), 4612–4618.
29 S. A. Reis, B. Ghosh, J. A. Hendricks, D. M. Szantai-Kis,
L. Tork, K. N. Ross, J. Lamb, W. Read-Button, B. Zheng,
53 M. Deiana, Z. Pokladek, K. Matczyszyn, P. Mlynarz,
M. Buckle and M. Samoc, J. Mater. Chem. B, 2017, 5(5),
1028–1038.
H. Wang, C. Salthouse, S. J. Haggarty and R. Mazitschek, 54 F. Livolant and M. F. Maestre, Biochemistry, 1988, 27(8),
Nat. Chem. Biol., 2016, 12(5), 317–323. 3056–3068.
30 W. Szymanski, M. E. Ourailidou, W. A. Velema, F. J. Dekker 55 C. L. Peterson and J. C. Hansen, CSH Protoc, 2008, 3(12),
and B. L. Feringa, Chemistry, 2015, 21(46), 16517–16524. 1–8.
31 A. S. Lubbe, W. Szymanski and B. L. Feringa, Chem. Soc. 56 R. D. Makde, J. R. England, H. P. Yennawar and S. Tan,
Rev., 2017, 46(4), 1052–1079. Nature, 2010, 467(7315), 562–566.
32 E. Pazos, J. Mosquera, M. E. Vazquez and J. L. Mascarenas, 57 E. K. Liebler and U. Diederichsen, Org. Lett., 2004, 6(17),
ChemBioChem, 2011, 12(13), 1958–1973. 2893–2896.
33 A. Bergen, S. Rudiuk, M. Morel, T. Le Saux, H. Ihmels and 58 O. Vazquez, J. B. Blanco-Canosa, M. E. Vazquez,
D. Baigl, Nano Lett., 2016, 16(1), 773–780.
J. Martinez-Costas, L. Castedo and J. L. Mascarenas,
34 M. Deiana, Z. Pokladek, J. Olesiak-Banska, P. Mlynarz,
ChemBioChem, 2008, 9(17), 2822–28229.
M. Samoc and K. Matczyszyn, Sci. Rep., 2016, 6, 28605– 59 A. Bertin, S. Mangenot, M. Renouard, D. Durand and
28611. F. Livolant, Biophys. J., 2007, 93(10), 3652–3663.
35 S. Ghosh, D. Usharani, A. Paul, S. De, E. D. Jemmis and 60 P. N. Dyer, R. S. Edayathumangalam, C. L. White, Y. Bao,
S. Bhattacharya, Bioconjugate Chem., 2008, 19(12), 2332–
2345.
S. Chakravarthy, U. M. Muthurajan and K. Luger, Methods
Enzymol., 2004, 375, 23–44.
36 C. Dohno, S. N. Uno and K. Nakatani, J. Am. Chem. Soc., 61 A. Banerjee, P. Majumder, S. Sanyal, J. Singh, K. Jana,
2007, 129(39), 11898–11899. C. Das and D. Dasgupta, FEBS Open Bio, 2014, 4, 251–259.
37 T. Stafforst and D. Hilvert, Angew. Chem., Int. Ed., 2010, 62 Y. Arimura, H. Kimura, T. Oda, K. Sato, A. Osakabe,
49(51), 9998–10001.
H. Tachiwana, Y. Sato, Y. Kinugasa, T. Ikura, M. Sugiyama,
38 T. Goldau, K. Murayama, C. Brieke, S. Steinwand,
M. Sato and H. Kurumizaka, Sci. Rep., 2013, 3, 3510.
P. Mondal, M. Biswas, I. Burghardt, J. Wachtveitl, 63 D. Baigl and K. Yoshikawa, Biophys. J., 2005, 88(5), 3486–
H. Asanuma and A. Heckel, Chem. – Eur. J., 2015, 21(7),
2845–2854.
39 J. Garcia-Amoros and D. Velasco, Beilstein J. Org. Chem.,
2012, 8, 1003–1017.
3493.
64 S. Rudiuk, S. Franceschi-Messant, N. Chouini-Lalanne,
E. Perez and I. Rico-Lattes, Langmuir, 2008, 24(16), 8452–
8457.
40 H. M. Bandara and S. C. Burdette, Chem. Soc. Rev., 2012, 65 A. Venancio-Marques, A. Bergen, C. Rossi-Gendron,
41(5), 1809–1825.
S. Rudiuk and D. Baigl, ACS Nano, 2014, 8(4), 3654–3663.
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem.