A Light-up Logic Platform for Selective Recognition of Parallel G-Quadruplex Structures via Disaggregation-Induced Emission

Article abstract of DOI:10.1002/anie.201912027

The design of turn-on dyes with optical signals sensitive to the formation of supramolecular structures provides fascinating and underexplored opportunities for G-quadruplex (G4) DNA detection and characterization. Here, we show a new switching mechanism that relies on the recognition-driven disaggregation (on-signal) of an ultrabright coumarin-quinazoline conjugate. The synthesized probe selectively lights-up parallel G4 DNA structures via the disassembly of its supramolecular state, demonstrating outputs that are easily integrable into a label-free molecular logic system. Finally, our molecule preferentially stains the G4-rich nucleoli of cancer cells.

Full text of DOI:10.1002/anie.201912027

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
International Edition Chemie
www.angewandte.org
Accepted Article
Title: A Light-up Logic Platform for Selective Recognition of Parallel G-
Quadruplex Structures via Disaggregation-Induced Emission
Authors: Marco Deiana, Karam Chand, Jan Jamroskovic, Ikenna Obi,
Erik Chorell, and Nasim Sabouri
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201912027
Angew. Chem. 10.1002/ange.201912027
Link to VoR: http://dx.doi.org/10.1002/anie.201912027
http://dx.doi.org/10.1002/ange.201912027

Angewandte Chemie International Edition
10.1002/anie.201912027
RESEARCH ARTICLE
A Light-up Logic Platform for Selective Recognition of Parallel
G-Quadruplex Structures via Disaggregation-Induced Emission
[a]
[b]
[a]
[a]
[b]
Marco Deiana , Karam Chand , Jan Jamroskovic , Ikenna Obi , Erik Chorell* , and Nasim
Sabouri*[a]
Abstract: The design of turn-on dyes with optical signals sensitive to
the formation of supramolecular structures provides fascinating and
underexplored opportunities for G-quadruplex (G4) DNA detection
and characterization. Here, we show a new switching mechanism that
relies on the recognition-driven disaggregation (on-signal) of an
ultrabright coumarin-quinazoline conjugate. The synthesized probe
selectively lights-up parallel G4 DNA structures via the disassembly
of its supramolecular state, demonstrating outputs that are easily
integrable into a label-free molecular logic system. Finally, our
molecule preferentially stains the G4-rich nucleoli of cancer cells.
enable the use of this class of compounds in high-resolution and
ultrasensitive spectroscopy and microscopy techniques. However,
DIE still remains a poorly established and explored molecular
recognition concept.[ Herein, we have extended DIE towards the
detection of four-stranded G4 DNA structures, which have
multiple regulatory biological functions and are recognized as
novel therapeutic targets in e.g. oncology.[7] To achieve this, we
have coupled the coumarin scaffold with a quinazoline moiety to
afford a more extended heterocyclic donor-π-acceptor scaffold
with ability for self-assembly and G4 binding (Figure 1). One of
the coumarin-quinazoline (CQ) conjugates proved to be capable
of specific detection of parallel over non-parallel and non-G4
topologies. This derivative displayed striking optical changes
upon binding the G4 scaffold, which enabled the construction of a
label-free binary INHIBIT gate that may serve as a proof-of-
principle for the future design of multifunctional molecular logic
circuits. Furthermore, fluorescence imaging of fixed HeLa cells
showed the ability of this compound to preferentially light-up the
nucleoli which accommodate the multi-copy G4-rich ribosomal
genes. Finally, mitochondria staining was observed in living cells
which contain ~ 200 predicted G4 structures per mitochondrial
DNA copy.[8] Thus, this compound has light-up recognition
properties for G4 DNA both in vitro and in cells.
6]
Introduction
Inspired by the complexity and hierarchical organization of
natural occurring systems, chemists have constructed a multitude
of artificial functional materials capable of completing
[1]
sophisticated tasks at a molecular scale. As a result, a vast
variety of mono-, bi- or tridimensional macro- and supramolecular
homo- and heterostructures have been developed in recent years
by exploiting the cooperativity between monomeric units.[2] The
tunable photophysical properties of organic dyes with donor-π-
acceptor scaffolds and hydrogen-bonded/π-surface-based
intermolecular interaction sites have made them particularly
attractive building-blocks for the construction of multistate
nanostructures.[ Among those, coumarin-based compounds are Results and Discussion
3]
considered archetypical dyes with a unique combination of both
valuable optical properties and biological compatibility along with
the tendency to form well-defined supramolecular architectures.[4]
However, in many conventional systems, the typical co-facial π-
stacking arrangements of self-assembled luminophores, even at
sub-micromolar concentrations, results in a partial or complete
suppression of the emissive properties via relaxation of the
excited-states through non-radiative channels (aggregation-
caused quenching, ACQ).[4a,5] From the viewpoint of real-life
applications, this process hampers the widespread applicability of
these types of optical probes and sensors, especially in the bio-
sensing arena. In contrast, we hypothesized that these properties
instead could be advantageous when utilized in the reversed way,
i.e. through disaggregation-induced emission (DIE). This should
Design, synthesis and self-assembly behavior
To identify compounds that signal the presence of a certain
G4 DNA topology through DIE, we designed and synthesized a
small set of CQ (1-6) derivatives (Figure 1).
[
a]
b]
Dr. M. Deiana, Dr. J. Jamroskovic, Dr. I. Obi, Dr. N. Sabouri
Department of Medical Biochemistry and Biophysics, Umeå
University, 90187 Umeå, Sweden
E-mail: nasim.sabouri@umu.se
Dr. K. Chand, Dr. E. Chorell
Figure 1. Synthetic route of coumarin-quinazoline (CQ) derivatives. Reagents
and conditions: (i) Ammonium acetate, Ethyl cyanoacetate, absolute ethanol
15min, reflux, (76-81%); (ii) Isatoic anhydrides, N,N-Diisopropylethylamine,
[
o
DMF, 100 C, 12h (71-77%).
Department of Chemistry, Umeå University, 90187 Umeå, Sweden
E-mail: erik.chorell@umu.se
In the first step, Knoevenagel condensation of commercially
available substituted ortho- hydroxyl benzaldehydes (1a-d) with
ethyl cyanoacetate in presence of ammonium acetate led to in-
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201912027
RESEARCH ARTICLE
situ formation of 3-cynocoumarin, which on subsequent reduction
with another equivalent of ammonium acetate led to formation of
the desired 3-amidinocoumarin intermediates (2a-d). In the
second step, the 3-amidinocoumarin derivatives were reacted
with different isatoic anhydrides under basic conditions to give the
desired CQ (1-6) derivatives in 63-71% yields. The electron
donating group in position 7 was varied with the aim to tune the
fluorescence properties e.g. by a methoxy group (CQ3), a
julolidine group (CQ4 and CQ5) or a diethyl-amino group (CQ1
and CQ2), the latter of which is prone to form nonradiative twisted
intramolecular charge transfer (TICT) states. Additionally,
derivatives with electron withdrawing groups were also
synthesized to enhance the dipolar character of the molecules
the presence of SDS CQ4 underwent gradual disassembly and
exhibited intense photoluminescence. Similar changes in
fluorescence were also observed upon decreasing the water
volume fraction χ(water) in a MeOH/water mixture (Figure S6),
suggesting that photoluminescence is enhanced when the CQ4
aggregates are disassembled in MeOH.
(
CQ2 and CQ5) as well as one derivative with an extended
aromatic backbone (CQ6). Photophysical evaluation of all
compounds showed that CQ4 displayed suitable photostability
and solubility properties compared to its analogues and was thus
used for further in-depth DIE analysis.
Solvent dependence studies on the emission of CQ4
displayed a moderate positive solvatochromic behavior (Figures
S1-S3) and further analysis in aqueous buffered solution showed
a very broad absorption band with maximum (λmax) centered at
72 nm and an apparent extinction coefficient εapp = 2.6 х 10 M-1
4
4
cm-1 (Figure 2A, blue line), suggesting the formation of
aggregated CQ4 molecules. Indeed, atomic force microscopy
Figure 2. A) UV/Vis absorption spectral changes of a buffered aqueous CQ4
solution (cCQ4 = 3.8 µM, Tris buffer c = 50 mM, pH = 7.5) before (blue line) and
after the gradual addition of SDS from 0 to 5 mM (red line). B) AFM height image
on mica of a CQ4 aqueous solution (cCQ4 = 10 µM). C) Evolution of the
absorption profile of CQ4 in the presence of SDS. Black lines correspond to a
fitting with a growth sigmoidal Boltzmann function. The dashed black lines aim
to show the critical concentration of 0.75 х 10-3 M. D) CQ4 emission profile in
the presence of different concentrations of SDS.
(
AFM) revealed the presence of small particles with average
diameter of ~ 33 nm and a height of ~ 3.9 nm (Figure 2B). To
probe if the addition of surfactants could disassemble CQ4, we
carried out titrations with sodium dodecyl sulfate (SDS). At low
concentrations, the surfactant affected the molecular aggregates
weakly (Figure 2C). However, above the critical concentration of
0
.75 х 10-3 M, SDS induced a marked breakup of CQ4 aggregates
as determined by UV/Vis spectroscopy (Figure 2C). This process
led to the formation of a new absorption band centered at 490 nm
G4-CQ4 interactive binding models
In a similar fashion to the SDS-induced disaggregation, we
speculated that binding of CQ4 to the planar π-surfaces of G4
DNA structures may also cause the disassembly of the molecular
aggregates into highly emissive monomeric states. We thus
studied the recognition process of CQ4 towards a large panel of
biologically relevant synthetic and natural G4 DNA structures with
various topologies including parallel, hybrid, and antiparallel G4s
(
Figure 2A, red line), which was ascribed to the monomer
because of the overlap with the associated CQ4 excitation
spectrum (Figure S4), and the concomitant appearance of
isosbestic points at 430 and 510 nm (Figure 2A). The strongly
blue-shifted absorption maximum of CQ4 in the absence of SDS
is a clear evidence for an H-type exciton coupling between co-
facial or “side-by-side” aggregated molecules.[9] However, in
addition to the band centered at 490 nm, the presence of SDS
caused the formation of a new transition band centered at 526 nm,
indicating the formation of slipped-stack packing arrangements
with the “head-to-tail” J-type coupling (Figure 2A).[9] It is important
to note that the gradual addition of SDS induced the concomitant
disappearance of the J-aggregates along with the formation of the
monomeric state with εapp = 5.4 х 104 M-1 cm-1 (Figure 2C).
Concentration-dependent UV/Vis experiments provided the
(
see Table S1 and Figures S7-S9, S10-S35). Indeed, G4 DNA
successfully disassembled CQ4 aggregates and this effect was
strongly pronounced for parallel G4 structures (Figures 3, 4 and
Table 1). The proposed interactive binding model between CQ4
and G4 and non-G4 structures is shown in Figure 3.
a
The association constant (K ) values of CQ4 complexed
with parallel G4 DNA structures were up to three-orders of
magnitude (i.e. ≥ 1000-fold) higher compared to those calculated
in the presence of non-parallel G4 structures (Figure 4A and
Table 1). This outcome can be rationalized by the highly
accessible hydrophobic external G-tetrads in parallel G4
architectures.[6b]
5
-1
aggregation constant (Kagg) = 8.1 ᵡ 10 M (Figure S5). To further
examine the molecular aggregates, we performed steady-state
fluorescence measurements before and after the addition of SDS
(
Figure 2D). In the absence of the surfactant, the CQ4 emission
intensity was remarkably quenched supporting the formation of
new competitive radiationless relaxation pathways that are
available in the co-facial aggregated states. On the other hand, in
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201912027
RESEARCH ARTICLE
note, that the linear response of the fluorescence intensity of CQ4
as a function of c-MYC sG4 concentration provided a limit of
detection (LOD) of 83.5 ± 4.2 nM awarding this probe as one of
the most sensitive parallel G4-binders reported so far (Figure 5,
Figure 3. Schematic illustration of the selective detection mechanism: CQ4
forms molecular aggregates in solution with ACQ signatures (middle). In the
presence of parallel G4s, the formation of stable complexes turns the weakly
emissive aggregated states into highly emissive G4-CQ4 monomeric adducts
through a DIE process (right panel). Conversely, antiparallel and non-G4
structures are unable to fully displace the aggregated state of CQ4 resulting in
a poor fluorescence recovery (left panel).
Importantly, within the same experimental conditions, no
binding was observed for CQ4 in the presence of long genomic
double-stranded DNA (ds-DNA), and the light-up response of
CQ4 was hardly affected by the presence of 50 equivalents
excess of competitive ds-DNA (Figure S36). Parallel G4 over
duplex selectivity can be explained by the difference between
their respective binding surfaces where G4 presents a larger and
more accessible surface area for π-stacking compared to duplex
DNA.[7c]
To further evaluate the selectivity of CQ4 for parallel G4
structures, non-denaturing polyacrylamide gel electrophoresis
(
PAGE) were performed. These gels were loaded with the folded
Figure 4. A) Fluorescence titration of CQ4 in the presence of a large panel of
G4 and non-G4 structures shows a clear-cut preference for parallel G4
topologies. The solid black lines correspond to a 1:1 fitting model at λem. B) Non-
denaturing PAGE images of G4 and non-G4 structures stained with CQ4 (left-
panel). The same gel was also co-stained with thiazole orange TO, (right panel)
to show the specificity of binding. cCQ4 = 1 μM, cTO = 5 μM and coligos = 10 μM,
G4 and non-G4 oligonucleotides, post-stained with CQ4, and
imaged. These images confirmed the outstanding selectivity of
the probe for parallel G4 topologies (Figures 4B and S37). It is
worth to keep in mind that according to previous published papers
most of the parallel G4 structures studied here yielded
heterogeneous solutions with multiple G4 forms (e.g. intra and
intermolecular folding).[10]
λexc = 488 nm.
The strongest DIE effect was observed in the presence of c-
MYC sG4. This G4 structure lacks the terminal flanking
sequences at both 5´- and 3´-ends (compare c-MYC sG4 vs. c-
MYC Pu22) and thus provides better π-stacking possibilities for
the planar conformation of the π-extended CQ4 core, which can
explain the more efficient disaggregation process. To study this in
more detail, the UV/Vis spectral changes were analyzed when
CQ4 was complexed with c-MYC sG4 (Figure 5). This resulted in
the narrowing of the absorption band (isosbestic points at 463 and
5
12 nm) along with the formation of a new red-shifted transition
band centered at 494 nm (Figure 5). The fluorescence emission
intensity of CQ4 was partially quenched in its unbound state
(
photoluminescence quantum yield (PLQY), ΦF(CQ4) = 0.42 ± 0.01).
However, the addition of c-MYC sG4 greatly increased the
fluorescence response of CQ4 providing near-unity PLQY
(ΦF(CQ4:c-MYC sG4) =0.95 ± 0.2), one of the highest values reported
[11]
so far for G4-binders (Figure 5 and Table 1). This result clearly
supports the formation of a c-MYC sG4-CQ4 complex via the
direct interaction of the CQ π-surfaces with the hydrophobic G-
tetrad core of the c-MYC sG4 DNA structure. It is important to
Figure 5. UV/Vis absorption (solid lines) and fluorescence emission spectra
dashed lines) of CQ4 upon gradual addition of c-MYC sG4 (cCQ4 = 3.8 or 2.5
µM, Tris buffer c = 50 mM, KCl = 100 mM, pH = 7.5). Superimposed dotted lines
(
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201912027
RESEARCH ARTICLE
correspond to a 1:1 global fitting model. Inset, calculated limit of detection (LOD)
The mechanism of interaction of CQ4 with c-MYC sG4 was
[
12]
for CQ4-c-MYC sG4 system.
investigated by using the following dissociation model:
inset).[6b,6d] Moreover, the optical changes arising in the
absorption and emission profiles resulted in excellent one-photon
ꢊꢋꢃ
ꢀ
ꢁꢂ + ꢄ − ꢅꢆꢇ ꢈꢉ4 ↔ ꢀꢁꢂ
+ ꢀꢁꢂ: ꢄ − ꢅꢆꢇ ꢈꢉ4
ꢃꢍ1
(ꢎꢏ. 1)
ꢃ
ꢊꢋꢌꢌ
) of 3.9 ᵡ 10 M cm-
4
-1
brightness (defined as the product of εmax ᵡ Φ
F
1.
n
where CQ4 , c-MYC sG4 and CQ4:c-MYC sG4 represent
the CQ4 aggregate, the G4 structure and the complex of CQ4 with
c-MYC sG4, respectively. Such a model would predict an increase
in the apparent rate constant of dissociation of the CQ4 aggregate
with an increase in concentration of c-MYC sG4. As shown in
Figure 7 the apparent rate constant increased with increasing c-
Ionic strength-dependence studies showed that c-MYC sG4
was more stable under K conditions compared to Na or Li+
+
+
conditions, which is fully consistent with the ability of CQ4 to
+
strongly light-up the c-MYC sG4 sequence stabilized by K ions
(
Figures S38-S40).
To further examine the steady-state emission data, the
photophysical properties of the CQ4 excited state was
investigated via time-correlated single photon counting (TCSPC)
measurements (Figure S41). The decay traces of the free and
bound CQ4 molecules were fitted with a biexponential function
MYC sG4 concentration indicating that the G4 structure binds and
catalyzes dissociation of CQ4 molecular aggregates.[12]
(
Table S2). In its free form, CQ4 exhibited an average time-
dependent decay of 3.38 ns, which was increased to 4.76 ns in
the presence of c-MYC sG4. In line with previous results, this
suggest that parallel G4 DNA structures are able to disassemble
the aggregated CQ4 and induce additional geometrical
(
vibrational and rotational) and/or structural (molecular
conformation) restrictions of CQ4 in the G4:CQ4 complex. This
was further supported by the inhibition of the nonradiative
8
-1
relaxation pathways (knr(CQ4) = 1.73 ˣ 10 s vs. knr(CQ4:c-MYC sG4)
=
7
-1
1
.11 ˣ 10 s ) and increase of the radiative channels (kr(CQ4) = 1.23
8
-1
8
-1
ˣ 10 s vs. kr(CQ4:c-MYC sG4) = 1.99 ˣ 10 s ) in the G4:CQ4 complex.
Next, the binding stoichiometry of the complex in solution
was assessed using the mole ratio and Job’s plot methods. We
found
a
clear-cut preference for 1:1 CQ4:c-MYC sG4
stoichiometry (Figures 6, S42 and S43).
Figure 7. Effect of increasing concentration of c-MYC sG4 on the dissociation
kinetics of CQ4 aggregates. Apparent rate constant (k) was obtained by fitting
the kinetic data to a monoexponential function.
Table 1. Summary of the spectroscopic properties and binding constants of
CQ4 complexed with G4 and non-G4 DNA structures.
3
[d]
DNA
λ
nm
max
ε
M
max ᵡ 10
cm
λ
nm
em
K
M
a
Φ
F
τ / ns
Str
-
1
-1
-1[a]
[b]
7
7
7
6
6
5
6
6
6
5
c-MYC sG4
Ceb25
494
495
496
495
494
495
496
496
496
492
40.6
38.2
34.5
36.6
34.2
32.7
29.1
33.4
34.7
28.7
521
519
520
521
521
520
519
520
519
516
3.5 ᵡ 10
1.7 ᵡ 10
1.4 ᵡ 10
2.0 ᵡ 10
1.2 ᵡ 10
5.1 ᵡ 10
1.0 ᵡ 10
1.1 ᵡ 10
1.4 ᵡ 10
3.5 ᵡ 10
0.95
0.97
0.91
0.91
0.88
—
4.76
4.55
—
P
P
P
P
P
P
P
c-MYC Pu22
VAV-1
4.57
4.36
4.54
—
VEGF
bcl-2
Figure 6. Mole ratio and Job’s plot methods for assessing the CQ4:c-MYC sG4
stoichiometry. The vertical dotted lines aim to show the stoichiometry of the
system.
c-MYC Pu24T
ckit-87up
ckit-2
—
sb
—
—
P
sb
A quantitative binding analysis using global fitting on both
the absorption and emission data provided the association
—
—
P
Tel-22
—
—
H
7
-1
a
constant (K ) = 3.5 ᵡ 10 M (Table 1).
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201912027
RESEARCH ARTICLE
4
5
5
Bom17
491
492
495
474
21.6
22.1
28.0
18.0
517
519
519
519
2.6 ᵡ 10
2.0 ᵡ 10
2.3 ᵡ 10
—
—
—
—
—
—
—
AP
AP
TBA
ss-DNA
ds-DNA
—
nG4
nG4
[c]
3.53
[
a] Data fitting with 1:1 binding model was obtained with Bindfit by using
[
13]
multiple global fitting methods (Nelder-Mead method).
The association
constants are evaluated by using the average values of both
spectrophotometric and fluorimetric titrations. [b] Fluorescein is used as
reference standard (0.1 M NaOH, Φ = 0.95). [c] ds-DNA was used with an
F
excess of 50 eq. than respect to the CQ4 concentration. [d] Str = topological
sb
structure, P = parallel, P = snap-back parallel, H = hybrid, AP = antiparallel
and nG4 = non-G4 structure.
Next, by performing electronic circular dichroism (ECD)
spectroscopy we assessed the morphological changes occurring
on the c-MYC sG4 structure in the presence of CQ4 (Figures S44
and S45). The ECD spectra showed that the overall secondary
structure of c-MYC sG4 was fully conserved even upon the
addition of an excess of 4 eq. of ligand.[6b,6c] Therefore, in spite of
Figure 8. A) Schematic illustration of the fluorescence displacement assay
performed by using the G4 end-stacker Phen-DC
3
. B-C) Fluorescence emission
changes on the CQ4-c-MYC sG4 system upon increasing concentration of
the high K
a
value determined, CQ4 does not induce any
3 3
Phen-DC (CQ4 = 2μM, c-MYC sG4 = 2μM and Phen-DC ranged from 0 to 10
topological changes or destabilization effect on the G4 scaffold, a
highly important feature in the design of ideal in vivo G4
fluorescent probes.
μM).
This statement was further investigated by using the DNA
polymerase stop assay for G4-interactive small molecules.[14] If
CQ4 stabilizes the G4 structure, and thereby enhances the
pausing/ arrest of the DNA polymerase on the G4 DNA template,
the position and amount of pausing as well as the amount of full-
length product can be detected and quantified on a denaturing
polyacrylamide gel. CQ4 did not increase DNA replication stalling,
suggesting that CQ4 does not stabilize the G4 structures (Figure
S46). The lack of any well-resolved induced circular dichroism
(
ICD) signals corroborates the current hypothesis of a π-stacking
binding mode. Concentration-dependent CQ4-fluorescence
signal displacement by the well-known G4 end-stacker Phen-DC
3
confirmed the end-stacking ability of our probe (Figure 8).[
15]
These results highlighted the strong competitive behavior of the
two molecules for the same binding site.
Cellular imaging
Confocal laser scanning microscopy (CLSM) was used to
evaluate CQ4 cellular emission fingerprint. Fixed HeLa cells
treated with CQ4 revealed intense green fluorescence signals in
the nuclear and cytoplasmic/mitochondrial regions with clear
peaks in the subnuclear compartments whose appearance was
compatible with that of nucleoli (Figure 9A).[16] Following the in
vitro evidence of CQ4’s high selectivity for parallel DNA G4
structures, we treated cells with RNase to confirm the nature of
the main binding target of the compound (Figure 9B, upper panel).
Thioflavin-T (ThT) was used as control since it is also
intracellularly fluorescent and known to bind RNA G4s making it
Figure 9. A) Confocal fluorescence images of fixed HeLa cells stained with CQ4
1μM) and Hoechst 33342. B) Fluorescence images of HeLa cells stained with
(
Hoechst and CQ4 (1μM) before and after treatment with RNase (upper panel).
Fluorescence images of Hoechst and ThT (1 μM) before and after treatment
with RNase (lower panel). C) Fluorescence displacement assay between CQ4
3
(1μM) and Phen-DC (5 μM). Scale bar = 15 μm. In Figures A and C, Diode 405
nm laser was used for Hoechst λexc = 405 nm, λem 415-450 nm; WLL laser was
used for CQ4 visualization λexc = 472 nm, λem = 482-794 nm. Additional washing
during Phen-DC
3
displacement in Figure 9C caused decrease in overall CQ4
signal intensity. In Figure B, Diode 405 nm laser was used for Hoechst λexc
=
=
405 nm, λem 415-430 nm; for ThT visualization Argon laser λexc = 458 nm, λem
68-782 was used; for CQ4 Argon laser λexc 476 nm, λem 486-782 was used.
[
17]
highly sensitive to RNase treatment (Figure 9B, lower panel).
4
While RNase treatment did not modify CQ4 nucleolar staining, the
nucleolar ThT signal was reduced by the presence of the enzyme,
indicating the ability of CQ4 to preferentially target DNA G4
structures. DNase treatment was not performed due to the
apparent inability of this enzyme to reach the nucleoli.[16c] Instead,
DNA G4 binding was confirmed by using the well-known G4-
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201912027
RESEARCH ARTICLE
binder Phen-DC
3
(Figures 9C and S47). Indeed, Phen-DC
3
showed a weak fluorescence signal in the presence of the non-
completely displaced CQ4 from the subnuclear G-rich
compartment even at low concentration. Finally, CLSM imaging
of living HeLa cells showed co-localization of CQ4 with
Mitotracker, suggesting that CQ4 co-localize with the
mitochondrial network (Figure S48). Mitochondrial targeting has
recently gained considerable attention being a potential target for
cancer treatment with hundreds of possible G4 forming
sequences at the DNA level, with many forming parallel G4
structures.[8]
folded single-stranded c-MYC sG4 sequence. However, upon
addition of K that catalyze the folding of c-MYC sG4, CQ4
+
strongly emitted a fluorescence signal. In contrast, the addition of
the complementary strand, sC4, resulted in the formation of ds-
DNA that could not disaggregate the CQ4 templates thus inducing
no or very weak fluorescence changes both in the absence or the
+
presence of K ions. These results thus fully support the proper
execution of an INHIBIT gate (Figure 10C).
Logic-gate
Conclusion
Although DNA have been used to construct various logic
platforms, the inability to generate detectable signals has forced
the development of DNA-based circuits that rely on suitable
optical reporters.[18] In this context, multitasking supramolecular
fluorescence sensors with distinguishable readout responses are
key components in the ongoing quest towards increasingly
sophisticated, sensitive, and versatile information processing
systems. Therefore, as CQ4 showed several assembly behaviors
associated with different binding conditions, we further examined
its applicability as a building-block for the construction of multi-
output logic circuits. As proof-of-concept, we focused our
attention on the INHIBIT gate, one of the most widely used
Boolean gates in biochemical detections (Figure 10A). The
threshold of the output signals was set at 0.3, and any signal
greater than this value was defined as “1” whereas, those lower
than the threshold signal value were defined as “0”.
In conclusion, a CQ-based dye, CQ4, was designed to display a
supramolecular and solvatochromic feature. The possibility to use
the striking optical changes of CQ4 associated with multiple
assembly states for DIE purposes was evaluated using a large
panel of biologically relevant G4 structures. This revealed that the
probe was able to selectively signal the presence of parallel G4
DNA topologies over both antiparallel and hybrid G4 DNA
topologies as well as non-G4 DNA structures. Moreover, we
applied the optical and supramolecular changes of our sensor in
a tunable label-free DNA-based logic nanoplatform through the
design of a binary INHIBIT gate. The presented recognition
mechanism thus provides a promising and underexplored route to
design selective and sensitive turn-on probes and advanced
multi-output logic circuits. Finally, the ability of CQ4 to target DNA
G4 structures in human cells opens new avenues toward the
design of potential sequence-selective in vivo supramolecular G4
fluorescent probes. A strategy that is currently under investigation
in our laboratories.
Experimental Section
Experimental details, including materials and methods, experimental
procedures, optical studies, G-quadruplex characterization, G-quadruplex
binding studies, microscopy, and chemical synthesis of the compounds,
are provided in the Supporting Information.
Acknowledgements
Work in the Sabouri lab was supported by Knut and Alice
Wallenberg Foundation (KAW2015-0189) and the Swedish
Research Council. MD was supported by a fellowship from the
MIMS Excellence by Choice Postdoctoral Programme. Work in
the Chorell lab was supported by the Kempe foundations (SMK-
Figure 10. A) Diagram of the operational design of the binary INHIBIT gate. B)
Fluorescence spectra of CQ4 with different input conditions. C) Schematic truth
table (left-panel) of the binary INHIBIT gate and associated normalized
fluorescence responses plotted as a bar chart (right-panel). The dotted black
line shows the threshold value (0.3). The blue single-stranded DNA sequence
represents the G-rich oligonucleotide c-MYC sG4, the light-blue circle refers to
K+ ions, the red single-stranded DNA sequence indicates the C-rich
oligonucleotide sC4.
1
632) and the Swedish Research Council. We thank Prof. Sjoerd
Wanrooij for the kind gift of HeLa cells and Dr. Mara Doimo for
her expert advice with microscopy. We acknowledge Dr. Igor
Iashchishyn for assistance with AFM imaging. We also
acknowledge Dr. Irene Martinez Carrasco and the Biochemical
Imaging Center at Umeå University and the National Microscopy
Infrastructure (VR-RFI 2016-00968) for providing assistance in
microscopy.
CQ4 together with the non-folded single-stranded c-MYC
+
sG4 sequence were used as the initial state, whereas K and the
complementary C-strand (hereafter referred as sC4) were used
as the input stimuli. As depicted in Figures 10B and 10C, CQ4
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201912027
RESEARCH ARTICLE
Chem. Int. Ed. 2017, 56, 7520-7524; c) V. Grande, C. -A. Shen, M.
Deiana, M. Dudek, J. Olesiak-Banska, K. Matczyszyn, F. Würthner,
Chem. Sci. 2018, 9, 8375-8381; d) M. Zuffo, A. Guédin, E. -D. Leriche,
F. Doria, V. Pirota, V. Gabelica, J. -L. Mergny, M. Freccero, Nucleic Acid
Res. 2018, 46, e115; e) L. Zhang, J. C. Er, K. K. Ghosh, W. J. Chung, J.
Yoo, W. Xu, W. Zhao, A. T. Phan, Y. -T. Chang, Sci. Rep. 2014, 4, 3776.
a) R. Hänsel-Hertsch, M. Di Antonio and S. Balasubramanian, Nat. Rev.
Mol. Cell Biol. 2017, 18, 279-284; b) S. Balasubramanian, L. H. Hurley,
S. Neidle, Nat. Rev. Drug. Discov. 2011, 10, 261-275; c) S. Neidle, Nat.
Rev. Chem. 2017, 1, 0041.
Keywords: aggregation • coumarin • DNA • G-quadruplex • logic
gate
[1]
a) J. M. Abendroth, O. S. Bushuyev, P. S. Weiss, C. J. Barrett, ACS Nano
015, 9, 7746-7768; b) S. Dhiman, A. Jain, M. Kumar, S. J. George, J.
2
Am. Chem. Soc. 2017, 139, 16568-16575; c) V. Grande, B. Soberats, S.
Herbst, V. Stepanenko, F. Würthner, Chem. Sci. 2018, 9, 6904-6911; d)
Y. Kim, H. Li, Y. He, X. Chen, X. Ma, M. Lee, Nat. Nanotechnol. 2017,
[
7]
8]
12, 551-556; e) T. Tian, Y. Song, L. Wei, J. Wang, B. Fu, Z. He, X. -R.
Yang, F. Wu, G. Xu, S. -M. Liu, C. Li, S. Wang, X. Zhou, Nucleic Acid
Res. 2017, 45, 2283-2293; f) F. della Sala, S. Maiti, A. Bonanni, P.
Scrimin, L. J. Prins, Angew. Chem. Int. Ed. 2018, 57, 1611-1615; g) A.
Sarkar, S. Dhiman, A. Chalishazar, S. J. George, Angew. Chem. Int. Ed.
[
A. Bedrat, L. Lacroix, J. -L. Mergny, Nucleic Acid Res. 2016, 44, 1746-
1
759.
F. Würthner, T. E. Kaiser, C. R. Saha-Möller, Angew. Chem. Int. Ed.
011, 50, 3376-3410.
[
[
[
[
[
9]
2
2017, 56, 13767-13771; h) S. Maiti, I. Fortunati, C. Ferrante, P. Scrimin,
10] H. T. Le, M. C. Miller, R. Buscaglia, W. L. Dean, P. A. Holt, J. B. Chaires,
J. O. Trent, Org. Biomol. Chem. 2012,10, 9393-9404.
L. J. Prins, Nat. Chem. 2016, 8, 725-731.
[
2]
3]
a) D. Pijper, B. L. Feringa, Soft Matter, 2008, 4, 1349-1372; b) Y. Kamiya,
H. Asanuma, Acc. Chem. Res. 2014, 47, 1663-1672.
11] Y. V. Suseela, N. Narayanaswamy, S. Pratihar, T. Govindaraju, Chem.
Soc. Rev. 2018,47, 1098-1131.
[
a) T. Aida, E. W. Meijer, S. I. Stupp, Science 2012, 335, 813-817; b) F.
Würthner, Acc. Chem. Res. 2016, 49, 868-876; c) S. Ghosh, V. K.
Praveen, A. Ajayaghosh, Annu. Rev. Mater. Res. 2016, 46, 235-262.
a) F. Bu, R. Duan, Y. Xie, Y. Yi, Q. Peng, R. Hu, A. Qin, Z. Zhao, B. Z.
Tang, Angew. Chem. Int. Ed. 2015, 54, 14492-14497; b) M. Tasior, D.
Kim, S. Singha, M. Krzeszewski, K. H. Ahn, D. T. Gryko, J. Mater. Chem.
C 2015, 3, 1421-1446; c) X. Liu, J. M. Cole, K. S. Low, J. Phys. Chem. C
12] V. G. Panse, P. Vogel, W. E. Trommer, R. Varadarajan, J. Biol. Chem.
2000, 275, 18698-18703.
13] a) P. Thordarson, Chem. Soc. Rev. 2011, 40, 1305–1323; b) L. K. S. von
Krbek, C. A. Schalley, P. Thordarson, Chem. Soc. Rev. 2017, 46, 2622-
[
4]
2637.
[
14] H. Han, L. H. Hurley, M. Salazar, Nucleic Acid Res. 1999, 27, 537-542.
15] W. J. Chung, B. Heddi, F. Hamon, M. -P. Teulade-Fichou, A. T. Phan,
Angew. Chem. Int. Ed. 2014, 53, 999-1002.
[
2013, 117, 14723-14730; d) X. Liu, J. M. Cole, P. C. Y. Chow, L. Zhang,
Y. Tan, T. Zhao, J. Phys. Chem. C 2014, 118, 13042-13051; e) P. Verma,
H. Pal, J. Phys. Chem. A 2012, 116, 4473-4484; f) L. G. Meimetis, J. C.
T. Carlson, R. J. Giedt, R. H. Kohler, R. Weissleder, Angew. Chem. Int.
Ed. 2014, 53, 7531-7534; g) C. Uttamapinant, K. A. White, H. Baruah, S.
Thompson, M. Fernández-Suárez, S. Puthenveetil, A. Y. Ting, Proc. Natl.
Acad. Sci. USA 2010, 107, 10914-10919; h) G. Signore, R. Nifosì, L.
Albertazzi, B. Storti, R. Bizzarri, J. Am. Chem. Soc. 2010, 132, 1276-
[
16] a) J. Li, X. Yin, B. Li, X. Li, Y. Pan, J. Li, Y. Guo, Anal. Chem. 2019, 91,
5354-5361; b) O. Domarco, C. Kieler, C. Pirker, C. Dinhof, B. Englinger,
J. M. Reisecker, G. Timelthaler, M. D. Garcia, C. Peinador, B. K. Keppler,
W. Berger, A. Terenzi, Angew. Chem. Int. Ed. 2019, 58, 8007-8012; c) F.
Doria, M. Nadai, M. Zuffo, R. Perrone, M. Freccero, S. N. Richter, Chem.
Commun. 2017, 53, 2268-2271.
[
17] S. Zhang, H. Sun, H. Chen, Q. Li, A. Guan, L. Wang, Y. Shi, S. Xu, M.
Liu, Y. Tang, Biochim. Biophys. Acta Gen. Subj. 2018, 1862, 1101-1106.
18] a) H. -Z. He, D. S. -H. Chan, C. -H. Leung, D. -L. Ma, Nucleic Acid Res.
1288.
[
5]
6]
J. Mei, N. L. C. Leung, R. T. K. Kwok, J. W. Y. Lam, B. Z. Tang, Chem.
Rev. 2015, 15, 11718-11940.
[
2
013, 41, 4345-4359; b) C. Yang, S. Yang, J. Li, Y. Du, L. Song, D.
Huang, J. Chen, Q. Zhou, Q. Yang, Y. Tang, Anal. Chem. 2018, 90,
0585-10590.
[
a) D. Zhai, W. Xu, L. Zhang, Y. -T. Chang, Chem. Soc. Rev. 2014, 43,
2402-2411; b) V. Grande, F. Doria, M. Freccero, F. Würthner, Angew.
1
This article is protected by copyright. All rights reserved.

Angewandte Chemie International Edition
10.1002/anie.201912027
RESEARCH ARTICLE
RESEARCH ARTICLE
A versatile supramolecular sensing
platform based on the switching
mechanism of a self-aggregated
coumarin-quinazoline derivative was
used to selectively detect parallel G-
quadruplex (G4s) DNA structures both
in vitro and in cells. The programmed
reorganization of tightly packed
chromophores was further used to
build up a programmable label-free
molecular logic system.
Marco Deiana, Karam Chand, Jan
Jamroskovic, Ikenna Obi, Erik Chorell*,
Nasim Sabouri*
Page No. – Page No.
A Light-up Logic Platform for
Selective Recognition of Parallel G-
Quadruplex Structures via
Disaggregation-Induced Emission
This article is protected by copyright. All rights reserved.