Development of a Library of Thiophene-Based Drug-Like Lego Molecules: Evaluation of Their Anion Binding, Transport Properties, and Cytotoxicity

Article abstract of DOI:10.1002/chem.201904255

The anion-binding and transport properties of an extensive library of thiophene-based molecules are reported. Seventeen bis-urea positional isomers, with different binding conformations and lipophilicities, have been synthesized by appending α- or β-thiophene or α-, β-, or γ-benzo[b]thiophene moieties to an ortho-phenylenediamine central core, yielding six subsets of positional isomers. Through ¹H NMR, X-ray crystallography, molecular modelling, and anion efflux studies, it is demonstrated that the most active transporters adopt a pre-organized binding conformation capable of promoting the recognition of chloride, using urea and C?H binding groups in a cooperative fashion. Additional large unilamellar vesicle-based assays, carried out under electroneutral and electrogenic conditions, together with N-methyl-d-glucamine chloride assays, have indicated that anion efflux occurs mainly through an H⁺/Cl^? symport mechanism. On the other hand, the most efficient anion transporter displays cytotoxicity against tumor cell lines, while having no effects on a cystic fibrosis cell line.

Full text of DOI:10.1002/chem.201904255

A Journal of
Accepted Article
Title: Development of a library of thiophene-based drug-like LEGO
molecules: evaluation of their anion binding, transport properties
and cytotoxicity
Authors: Paulo Vieira, Margarida Q. Miranda, Igor Marques, Sílvia
Carvalho, Li-Jun Chen, Ethan N. W. Howe, Carl Zhen,
Claudia Y. Leung, Michael J. Spooner, Bárbara Morgado,
Odete A. B. Cruz e Silva, Cristina Moiteiro, Philip A. Gale, and
Vítor Félix
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: Chem. Eur. J. 10.1002/chem.201904255
Link to VoR: http://dx.doi.org/10.1002/chem.201904255
Supported by

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
Development of a library of thiophene-based drug-like LEGO
molecules: evaluation of their anion binding, transport properties
and cytotoxicity
Paulo Vieira,[a] Margarida Q. Miranda,[b] Igor Marques,[b] Sílvia Carvalho,[a] Li-Jun Chen,[c] Ethan N. W.
Howe,[c] Carl Zhen,[c] Claudia Y. Leung,[c] Michael J. Spooner,[d] Bárbara Morgado,[e] Odete A. B. da
Cruz e Silva,[e] Cristina Moiteiro,*[a] Philip A. Gale,*[c] Vítor Félix*[b]
Abstract: The anion binding and transport properties of an extensive library of
thiophene-based molecules are reported. Seventeen bis-urea positional
symptoms, while the cure remains a challenge. Therefore, the
isomers, with different binding conformations and lipophilicities, were
synthetized by the appendage of α-, β-thiophene or α-, β-, γ-benzo[b]thiophene
motifs to the ortho-phenylenediamine central core, yielding six subsets of
positional isomers. Through 1H NMR, X-ray crystallography, molecular
modelling and anion efflux studies, it was demonstrated that the most active
used in the latter approach are restricted to the pamamycins,[4]
transporters display a pre-organised binding conformation able to promote the
recognition of chloride using urea and C-H binding groups in a cooperative
fashion. Additional LUV based assays, carried out under electroneutral and
electrogenic conditions, together with NMDG-Cl assays, indicate that anion
efflux occurs mainly through H+/Cl– symport mechanism. On the other hand, the
the airway surface liquid and reducing its viscosity.[7] Still, the
most efficient anion transporter displays cytotoxicity against tumour cell lines
Most of the current treatments for CF aim to manage the disease
search for an effective treatment is of paramount importance, with
CFTR modulation,[3] and channel replacement therapies being
possible pathways.[2] Potential natural products that could be
duramycins,[5] prodigiosins,[6] and, more recently, amphotericin
B.[7] This unselective channel was shown to restore lung cell’s
ability to secrete HCO3−, thus enhancing the antibiotic activity of
development of a myriad of small drug-like molecules capable of
while having no effects on a cystic fibrosis cell line.
mediating the selective transmembrane anion transport is still
imperative, being a field of intensive research.[8] Most of these
molecules have been designed translating the know–how
Introduction
acquired with the development of anion receptors,[9] coupling
natural[10] or synthetic platforms with suitable binding units, that
operate via conventional or non-conventional hydrogen bonds,[11]
Ion transport across phospholipid cell membranes is crucial to
several biological processes, such as nerve conduction and
via
halogen,[12]
chalcogen,[13]
or
pnictogen
bonding
homeostasis maintenance, being mediated by a combination of
embedded protein ion channels.[1] The dysfunction of these
channels causes serious pathologies, commonly designated as
channelopathies, including types of male infertility and the
genetically inherited disease cystic fibrosis (CF), caused by a
defective transmembrane transport of the bicarbonate and
chloride anions.[1-2]
interactions,[13c] or even through anion-π interactions.[14] Beyond
anion binding strength, the anion transport ability of a synthetic
receptor is also largely dependent of its ability to partition into the
lipid bilayer, i.e. its partition coefficient, and can be tuned by the
addition of structural motifs with different lipophilicities, such as
alkyl chains, aromatic groups and fluorinated units.[15] Moreover,
there is growing evidence that synthetic anionophores are also
able to trigger apoptosis (or disrupt autophagy) in cancer cell lines,
opening
a
new application perspective for these anion
16]
transporters.[10b,
In spite of all these achievements, the
[a]
P. Vieira, Dr. S. Carvalho, Prof. C. Moiteiro
Centro de Química e Bioquímica, Centro de Química Estrutural,
Departamento de Química e Bioquímica
Faculdade de Ciências, Universidade de Lisboa, Campo Grande
1749-016 Lisboa, Portugal.
development of molecules that can make the bench-to-bedside
translation process is still in a relatively early stage.[17]
The main goal of our anion transport endeavour is the
development of anion carriers containing heteroaromatic moieties
widely used in drug design and with the eventual goal of using
these systems as Channel Replacement Therapies. From a
variety of heteroaromatic fragments, the thiophene and
benzo[b]thiophene motifs, present in different classes of drugs
(e.g. antihistamines and analgesics),[18] were selected. These
entities have octanol-water partition coefficients (logP)
comparable to their benzene and naphthalene isosteres (1.81 –
thiophene vs. 2.13 – benzene; and 3.12 – benzo[b]thiophene vs.
3.34 – naphthalene),[19] but possess distinct electronic properties,
as can be seen from the distribution of the electrostatic potential
on the molecular surface (Vs) presented in Figure 1. The presence
of the sulfur atom in thiophene and benzo[b]thiophene results in
these entities being electron deficient, with the VS in front of to the
Cα–H bonds being higher than for the Cβ–H ones. Moreover, the
anisotropic distribution of electron density around the sulfur atoms
yields two σ–holes nearly aligned with the C-S bonds. The
E-mail: cmmoiteiro@fc.ul.pt
[b]
[c]
M. Q. Miranda, Dr. I. Marques, Prof. V. Félix
Department of Chemistry, CICECO – Aveiro Institute of Materials,
University of Aveiro
3810-193, Aveiro, Portugal.
E-mail: vitor.felix@ua.pt
Dr. L.-J. Chen, Dr. E. N. W. Howe, C. Zhen, C. Y. Leung, Prof. P. A.
Gale
School of Chemistry (F11), The University of Sydney
NSW 2006, Australia.
E-mail: philip.gale@sydney.edu.au
Dr. M. J. Spooner
Chemistry, University of Southampton
Southampton SO17 1BJ, UK
B. Morgado, Prof. O. A. B. C. Silva
Department of Medical Sciences, Institute of Biomedicine (iBiMED)
University of Aveiro, 3810-193 Aveiro, Portugal
[d]
[e]
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.

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
chalcogen bonds derived from activated sulfur’s σ–holes have
flanked with two thiophene or benzo[b]thiophene motifs, as
depicted in Scheme 1.
been recently employed in the recognition and transmembrane
transport of chloride by dithienothiophene derivatives.[12a, 12b, 13a,
This scaffold was selected due to its chemical versatility allied to
its successful use in the development of anion transporters.[21]
The thiophene or benzo[b]thiophene moieties were directly linked
to the central ortho-phenylene core via the Cα or Cβ carbon atoms,
leading to the substitution isomers α (3, 4, 8, 9) and β (6, 7, 14,
15), akin to assembling LEGO bricks. Moreover, the
benzo[b]thiophene motif was also bonded to the bis-urea central
core through the phenyl fused ring (Cγ) rather than through the
thiophene, affording 16 and 17. Furthermore, the lipophilicity and
binding affinity of these molecules was mainly tuned with the
fluorination of the central aromatic ring with two fluorine
substituents (2, 4 and remaining odd numbered molecules) or a
single trifluoromethyl group (5),while the fluorination of the phenyl
fused rings distal from the bis-urea binding pocket with two
electron withdrawing CF3 groups (10–13) should mainly affect the
molecule’s lipophilicity. Whereas in all the previous molecules the
heteroaromatic rings are able to communicate via the central
aromatic ring, in α-thiophene derivatives 1 and 2 the methylene
bridges allow to assess how the interruption of the electron
delocalisation influences the anion recognition and transport
properties. On the other hand, playing with the linkage positions
of thiophene and benzo[b]thiophene substituents (Cα, Cβ, and Cγ
carbons) pre-organised conformations, stabilised by putative
intramolecular chalcogen (C–S⋯O=C) or hydrogen (C–H⋯O)
bonds, with different binding geometries and properties were
designed, as demonstrated below.
13b]
Figure 1. Comparison between the distribution of electrostatic potential mapped
on the molecular surface (0.001 e Bohr–1 contour) of benzene (top left),
naphthalene (top right) and their respective isosteres thiophene (bottom left)
and benzo[b]thiophene (bottom right). The colour ranges from purple (below –
15.5 kcal mol–1 ) to red (larger than 21.9 kcal mol–1). The values in of the
decreasing surface maxima (21.9 kcal mol–1) are given along with their location
in front of the C–H bonds (red, pink and salmon spheres), or at the sulfur σ–
holes (yellow spheres).
With this library of seventeen compounds, we undertook a
comprehensive investigation, comprising the molecular design
assisted by computational modelling, synthesis and structural
characterisation followed by experimental binding and
transmembrane transport studies. Molecular dynamics
simulations were also performed to obtain insights into the
interaction of the thiophene-based molecules with the POPC
membrane model. The cytotoxicity of the most promising anion
transporters was also evaluated against cancer and CF cell lines.
Alongside the transport activity and the delivery to the cell
membrane, facile synthesis is also of paramount importance in
the development of drug-like anion carriers,[20] particularly when
we are interested in building and rationalising the transport
properties of a large library of structurally related and chemically
accessible compounds. These principles and the properties of
thiophene and benzo[b]thiophene motifs inspired us to design a
library of ortho-phenylene based bis-urea anion transporters
Scheme 1
This article is protected by copyright. All rights reserved.

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
complexes, the binding arrangement with conformation A is
favoured by ca. 2.1 kcal mol–1.
On the other hand, the energy difference between the A and B
conformations of the chloride complexes of β-benzo[b]thiophene
derivatives 14 and 15 nearly doubles when compared with their
β-thiophene-substituted complexes 6 and 7 (ca. 3.9 kcal mol–1).
Indeed, binding conformation A (Figure 2) presents a synergetic
recognition of chloride via four classical N–H⋯Cl– hydrogen
bonds assisted by C9–H⋯Cl– bonding interactions (see Table S3).
Moreover, in this conformation, the bonded receptors 14 and 15
are nearly planar, while in B they are twisted avoiding the steric
clashes between the C9 benzo[b]thiophenes’ protons and their
adjacent carbonyl groups (see Figure S5).
Results and Discussion
Molecular design and quantum descriptors
The first insights into the binding and conformational preferences
of our library of molecules towards chloride were obtained through
Density Functional Theory (DFT) calculations on 1:1 complexes,
at the M06-2X/6-311++G(d,p) level of theory in a polarizable
continuum model of DMSO,[22] using Gaussian 09.[23] The strength
of the non-covalent interactions was further assessed via the E2
stabilization energies through Natural Bond Orbitals (NBO)
analysis.[24]
In the α-thiophene and α-benzo[b]thiophene-based molecules (3–
5 and 8–13), two conformations A and B were considered (see
Scheme 2). In A, the spatial disposition of the thiophene rings is
locked by two putative chalcogen⋯chalcogen bonds between the
sulfur and oxygen atoms of the carbonyl groups, while in B the
putative C–S⋯O=C non-covalent bonding interactions should be
replaced by C–H⋯O ones. The optimised structures of the
chloride complexes, with the receptors in alternative
conformations, are shown in Figure 2 for 4 and in Figures S2 and
S4 for the remaining α-thiophene and α-benzo[b]thiophene
derivatives. In all complexes, the conformational binding
arrangement A displays two non-covalent intramolecular bonding
contacts, with S⋯O distances and C–S⋯O angles (see Table S1)
consistent with the existence of chalcogen⋯chalcogen bonds,
with an average E2 stabilisation energy of ca. 2.6 kcal mol–1,
derived from the delocalization of the oxygen’ lone pair to the C-
S antibonding orbital (푛푂 → 휎 ∗푆−퐶). Accordingly, in the 3–5 and
8–13 chloride complexes, the conformation A is systematically
preferred over B, with energy differences (ΔEconf = EA – EB)
ranging from 1.9 to 2.7 kcal mol–1 (see Table S2).
The chloride complexes of the β-thiophene-based molecules (6
and 7) were also optimised in two alternative conformations
regarding the relative position of the C7–H and C9–H thiophene
protons to the adjacent carbonyl groups (see Scheme 2 and
Figure S3), due their different acidic character, measured by VS:
in conformation A, the C7–H more acidic protons are close of the
oxygen atoms while in B they are replaced by lesser C9–H ones.
Although the NBO analysis revealed the inexistence of C–H⋯O
bonding interactions in the optimised structures of these
Figure 2. DFT optimised structures of the 1:1 chloride complexes of 4, 15 and
17, with the fluorinated receptors in the A or B conformations. The N–H⋯Cl–
and C–H⋯Cl– hydrogen bonds are drawn in red and blue dashes, respectively,
while the C–S⋯O=C chalcogen bonds are shown as yellow dashes.
Scheme 2
This article is protected by copyright. All rights reserved.

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
Table 1. Quantum descriptors and logP for bis-urea thiophene-based receptors, together with experimental binding and transport data for chloride.
a
E2 (kcal mol–1
푛퐶푙 → 휎 ∗푁−퐻
)
VS,max
(kcal mol–1
94.0
Ka
EC50
k ini
Molecule
logPc
푛퐶푙 → 휎 ∗퐶−퐻
(M–1
)
(mol %)e
(% s-1)f
0.0345
0.0383
b
d
−
−
)
h
1
2
28.2
29.3
2.5
2.3
3.50
3.79
31.84
44.77
-
h
102.5
-
α-thiophenes
h
3
4
5
36.6
37.2
38.0
-
-
-
103.5
112.2
112.9
4.16
4.44
5.03
31.70
44.06
44.56
-
0.0609
0.7457
0.7750
0.3875
0.3604
β-thiophenes
g
h
6
7
34.9
35.6
-
100.3
109.1
3.82
4.11
35.16
67.61
-
0.0493
0.7090
g
-
0.3131
α-benzo[b]thiophenes
g
h
8
38.0
-
107.5
115.9
115.7
123.9
110.6
119.0
6.35
6.63
8.10
8.39
8.10
8.39
41.02
47.53
57.41
70.79
66.07
74.13
-
0.0855
0.2170
0.0347
0.1938
0.0489
0.1792
g
9
38.5
38.7
39.2
39.5
39.9
-
0.6827
g
h
10
11
12
13
-
-
-
0.4379
g
h
-
-
g
h
-
-
β-benzo[b]thiophenes
14
15
35.7
36.7
3.6
3.5
102.6
111.3
6.02
6.30
114.29
163.68
0.1722
0.0089
0.7354
4.3685
γ-benzo[b]thiophenes
h
16
17
34.5
35.4
5.8
5.5
101.1
109.7
6.02
6.30
139.96
207.49
-
0.0917
1.7803
0.0194
a) Stabilization energies estimated by the 2nd-order perturbation theory; b) Most positive value of the electrostatic potential mapped on the molecular surface; c)
ChemAxon Consensus logP;[25] d) Cl– added as TBA salt and the binding constants were determined by 1H NMR titrations in DMSO-d6/0.5% H2O at 293 K using
e)
f)
HypNMR;[26] EC50 defined as the effective concentration needed for 50% activity at t = 270 s; The initial rates of chloride transport (kini) obtained using
standard Cl−/NO3− antiport assay for each transporter (1 mol%); g) E2 < 0.2 kcal mol–1 h) EC50 not determined due to the low transport activity.
;
To explore the role of the C–H⋯Cl– interactions in chloride binding
by the benzo[b]thiophene motifs, the substitution isomers 16 and
17 were designed (see Scheme 1), with the decorating moieties
attached to the bis-urea central core by the Cγ (C9) phenyl carbon.
Likewise 14 and 15, the DFT optimised complexes of 16 and 17
(see Figure 2) in conformation A are energetically favoured (see
Table S2) and show the anion hydrogen bonded to four urea N–
H binding units and two C8–H binding units from the thiophene
moieties. These C–H⋯Cl– bonding contacts are ca. 0.1 Å shorter
than the ones in 14 and 15 (see Table S3), affording higher E2
stabilisation energies (ca. 2.1 kcal mol–1) for 16 and 17 (see Table
1). These findings are also in line with the more acidic character
of the C8 vs. C9 protons of the benzo[b]thiophene fragment (see
Scheme 1 and Figure 1).
3.1 kcal mol–1, while the attachment at C5 position (10 and 11)
affords a greater gain of ca. 8.1 kcal mol–1. Overall, the α-
thiophene and α-benzo[b]thiophene receptors with the sulfur
atoms vicinal to the urea binding units, have higher VS,max values
than in their β-thiophene and β-benzo[b]thiophene substituted
isomers. Likewise, the proximity of the sulfur atom to the urea
binding units in the β-benzo[b]thiophene derivatives 14 and 15
leads to VS,max values 1.5 kcal mol–1 slightly higher than in the γ-
benzo[b]thiophene isomers 16 and 17. Moreover, the red region
of positive electrostatic potential extends from the urea groups to
the C–H binding units.
In the sequence of this energetic and structural analysis, only the
computed binding arrangements with thiophene- and
benzo[b]thiophene-derivatives adopting conformation A will be
used to rationalise their binding and transport properties.
Following our previous works with thiourea-[27] or squaramide-
28]
based[16b,
transporters, the maximum of the electrostatic
Figure 3. Distribution of electrostatic potential mapped on the molecular surface
of 4, 15 and 17 (0.001 e Bohr–1 contour). The colour ranges from -34.0 (purple)
to 106.4 kcal mol–1 (red).
potential surface (VS,max) was also evaluated as a binding
descriptor. The VS,max of 1-17 is located in the red binding region
confined by the urea binding units, as illustrated in Figure 3 for 4,
15 and 17 and in Figures S6 and S7 for the remaining thiophene
and benzo[b]thiophene derivatives. The VS,max values computed
for this library of molecules (see Table 1) indicate that the
fluorination of the central phenyl spacer, with the electron-
withdrawing two fluorine atoms or trifluoromethyl substituents,
leads to an increase of ca. 8.6 kcal mol–1. On the other hand, the
peripheric fluorination of the α-benzo[b]thiophene derivatives 8
and 9, with the insertion of a trifluoromethyl group in the phenyl
fused ring at position C4 (12 and 13) increased the VS,max in ca.
Overall, allying the chemical versatility of thiophene and
benzo[b]thiophene motifs with the selective fluorination of the
receptor skeleta (vide supra), we have designed a series of
seventeen-synthetic drug-like molecules with different putative
binding properties as well as a broad range of lipophilicities, with
logP values ranging from 3.50 to 8.39 (see Table 1), with the
isomeric ones inherently having similar lipophilic character.
This article is protected by copyright. All rights reserved.

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
Synthesis
The library of seventeen symmetrical bis-urea anion transporters
(Scheme 1) was prepared by the addition of isomeric thiophene
and benzo[b]thiophene isocyanates to ortho-phenylenediamine
linkers with different fluorination degrees. The 1-7 molecules were
obtained with high yields through one-pot synthesis using α- and
β-thiophene isocyanates as provided by the chemical suppliers.
The benzo[b]thiophene isocyanates used to produce the target
molecules (8-17) were prepared from the corresponding
carboxylic acids (see Scheme S1) which were previously
converted into acyl azide intermediates, and further underwent a
thermal Curtius rearrangement to yield the required isocyanates.
Full synthetic details and structural characterisation data of all
compounds are given in the Supporting Information.
Figure 4. Perspective views of the asymmetric units of 3 (a) and 5 (b) with
thermal ellipsoids drawn at the 40% probability level. The hydrogen bonds are
drawn as red dashed lines.
Solid-state structures
The single-crystal structures of the free receptors 3 and 5 and
halide complexes of 6, 10, and 12 were determined by X-ray
diffraction. Suitable single-crystals of 14 for X-ray diffraction
The crystal structures of halide complexes of 6, 12 and 14 are
built from asymmetric units containing one receptor, one chloride
and one TBA counter-ion. The asymmetric unit of the chloride
complex of 10 is composed of two independent molecules of
receptor, two chloride and two TBA counter-ions. The geometric
parameters of the hydrogen bonds between the halides and 6, 10,
12, and 14 are gathered in Table S6.
analysis were grown only from
a
solution containing
tetrabutylammonium bromide. The structure of 14 associated with
this halide is also reported.
In the crystal structure of the free α-thiophene 3, the two urea
binding units adopt a divergent spatial disposition (see Figure 4a),
as result of concomitant hydrogen bonds established between
them and carbonyl oxygen atoms of adjacent molecules. These
hydrogen bonds, with N⋯O distances of 2.946(2) and 2.818(2) Å
and N–H⋯O angles of 150 and 149°, are extended throughout
the crystal lattice leading to the formation of 1D-helical chains
(see Figure S53a). The crystal structure of α-thiophene 5 is
generated from an asymmetric unit having two molecules of
receptor (A and B) and two DMSO solvent molecules. The
perspective view of the asymmetric unit of 5, presented in (Figure
4b), shows an urea binding group of A and B assembled by N-
H⋯O=S bridges (N⋯O = 2.9205(3) and 2.8075(3) Å and N–
H⋯O = 150 and 157°) and N-H⋯O=C hydrogen bridges (N⋯O =
2.8308(3) and 2.8109(3) Å and N–H⋯O = 135 and 157°).The
second binding site of B is occupied by the other DMSO
crystallization molecule (N⋯O = 3.0086(3) and 2.8476(3) Å and
N–H⋯O = 162 and 171°), while the second binding site of A is
involved in a 1D network of hydrogen bonds with B (N⋯O =
3.1507(3) and 2.7340(3) Å and N–H⋯O = 137 and 152°) from
neighbouring supramolecular entities, as depicted in Figure S53b.
Overall, both free receptors display a divergent spatial disposition
of two urea binding groups dictated by the hydrogen bonds found
in solid-sate, but with sulfur atoms of α-thiophene in close contact
with adjacent carbonyl groups, as anticipated by the theoretical
calculations (vide supra).
The structure presented in Figure 5 shows the β-thiophene
derivative 6 binding the chloride in 1:1 stoichiometry through four
convergent hydrogen bonds with N⋯Cl– distances and N-H⋯Cl–
angles ranging from 2.7340(3) to 2.7340(3) Å and 137 to 162°,
respectively. Furthermore, the receptor is almost planar with a
dihedral angles (θ), between the planes of the ortho-phenyl
central core and the urea binding units, of 19.1(1) and 22.4(1)°.
The C7–H protons of the -thiophene motif are close of the
neighbouring carbonyl oxygen atom (H⋯O = 2.38 and 2.46 Å),
which is consistent with preferential binging conformation (A)
suggested by DFT calculations and NOESY data (vide infra).
Figure 5. Molecular structure of β-thiophene 6·Cl– with thermal ellipsoids drawn
at the 40% probability level. The four convergent hydrogen bonds are drawn as
red dashed lines.
The halide complexes of benzo[b]thiophene derivatives 14, 10
and 12 display a 2:2 binding stoichiometry in the crystalline state
with each halide hydrogen bonded by two urea binding units from
two receptors as illustrated in Figure 6 for [122·Cl2]2– and
[142.Br2]2– and in Figure S54 for [102·Cl2]2–.These three
complexes display equivalent centrosymmetric binding
arrangements with two urea-binding units each single receptor
adopting a divergent spatial, which contrasts with the convergent
one observed in the crystal structure of β-thiophene complex 6·Cl–.
This article is protected by copyright. All rights reserved.

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
Moreover, in [102·Cl2]2– and [122·Cl2]2– a urea binding unit of each
receptor is twisted relatively to the ortho-phenyl central core with
a θ angle of ca. 50°, while the other is roughly coplanar with a θ
angle of ca. 30° (see Table S7). The extended α-
benzo[b]thiophene decorating motifs are nearly coplanar with the
corresponding urea binding units and the sulfur atoms are close
to the carbonyl oxygen atoms at average S⋯O distances of
2.768(3) and 2.746(2) Å for 10 and 12, respectively. The
dimensions of the N–H⋯Cl– hydrogen bonds, which are in good
agreement with those determined for [6·Cl]–, as evident from the
comparison presented in Table S6.
Chloride binding studies
The chloride binding properties of our library of bis-urea
derivatives were investigated by 1H NMR experiments performed
in DMSO-d6/0.5% H2O at 293 K. The binding constants were
determined with HypNMR[26] using the chemical shift variations of
both internal (N4–H) and external protons (N6–H) of the urea
binding units observed throughout the titrations with
tetrabutylammonium chloride (TBACl). The 1H NMR titration
spectra are presented in Figures S56-S72. Non-linear regression
analyses (Figures S73-S78) indicated a 1:1 stoichiometry for the
anion complexes of 1-17. The binding constants gathered in Table
1
show that the 4,5-difluoro-ortho-phenlyenediamine based
derivatives have systematically higher values than their non-
fluorinated analogous. This finding is entirely consistent with the
increased acidity of the central binding core measured by the
VS,max quantum descriptor (see Table 1). The impact of the
fluorination degree is also evident in the complex of 5, with a
single trifluoromethyl substituent, which has a binding constant
similar to difluorinated analogous 4. The anion recognition by α-
thiophene derivatives 1 and 2 occurs mainly using the N4–H
protons of the bis-urea binding units, as shown by the variations
of N–H resonances upon addition of increasing amounts of
chloride. Indeed, as discernible in the 1H NMR spectra, the
internal N4–H protons present a downfield shift three and five
times larger than the N6–H ones in 1 and 2, respectively. In
contrast, the 1H NMR spectra of α-thiophene derivatives 3-5
display comparable downfield shifts for the four urea proton
resonances (ca. 0.5 ppm) while the chemical shifts of the
remaining protons are almost unmoved during titrations, which
indicates that the anion recognition occurs through the four
convergent N–H⋯Cl– hydrogen bonds with equivalent strength.
An identical binding behaviour is apparent from the 1H NMR
titration spectra of the α-benzo[b]thiophene based receptors,
which is expected, given the lengthened conformation adopted by
9-13 in the DFT optimised structures of their chloride complexes,
with the anion away from the C8–H protons (C8⋯Cl– distances are
ca. 4.44 Å). In addition, in the 2D-NOESY NMR spectra of the α-
benzo[b]thiophene substituted receptors (see Figures S83 and
S84) the correlation observed between the thiophene C8–H and
urea N6–H protons corroborates the existence of the
intramolecular chalcogen⋯chalcogen bonding interactions found
in the DFT optimised structures. In stark contrast, in β-thiophene
derivatives 6 and 7, as shown in Figures S82, both C7–H and C9–
H are correlated with the external urea protons, showing the
fluxional behaviour of these heteroaromatic rings.
Figure 6. Molecular structures of [122·Cl2]2– (a) and [142·Br2]2– (b) with thermal
ellipsoids drawn at the 40% probability level. The hydrogen bonds are drawn as
red dashed lines. A -CF3 substituent of [122·Cl2]2– is disordered over two
alternative tetrahedral positions, but only the position with higher occupation
factor is shown.
In centrosymmetric [142·Br2]2–- the urea binding units are also
twisted relatively to the ortho-phenyl central core by θ angles of
58.3(1) and 31.3(1)°, while being almost coplanar with the β-
benzo[b]thiophene decorating motifs. Moreover, the C7–H
protons of thiophene are close to the carbonyl oxygen atoms at
distances of 2.33 and 2.28 Å. The four independent N–H⋯Br–
hydrogen bonds have equivalent N⋯Br– distances, ranging from
3.4075(3) to 3.4725(3) Å, and N–H⋯Br– angles between 141 and
162°.
The chloride binding by β-benzo[b]thiophene 14 and 15 and γ-
benzo[b]thiophene 16 and 17 compounds occurs with chemical
shifts of the internal and external urea protons moving downfield
by ca. 0.7 and 0.4 ppm, respectively. In contrast to the α-
benzo[b]thiophene substituted receptors, significant downfield
field shifts (ca. 0.4 ppm) are observed for the C9–H phenyl ring
resonances in 14 and 15 and for the C8–H thiophene ring
resonances in 16 and 17 during the titrations. This binding
behaviour is illustrated in Figure 7 with titration spectra of 15 with
chloride and in Figures S69, S71 and S72 for the remaining three
receptors. These structural data provide evidence that the
chloride is synergistically recognised by four convergent N–
This article is protected by copyright. All rights reserved.

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
H⋯Cl– hydrogen bonds and two C–H⋯Cl– bonding interactions.
Further insights into the binding geometries adopted by these
receptors in DMSO-d6 aqueous solution were given by the
NOESY spectra of the β- and γ-benzo[b]thiophene chloride
complexes. The NOESY spectra of 15, presented in Figure 8,
reveals that the external protons of the urea binding units (N6–H)
correlate with the C9–H phenyl protons, meaning that these
protons are close to each other, as depicted in their DFT
optimised structures (see Figure 2 and Figure S5), where the
distances between protons N6–H and C9–H are 2.10 Å,
respectively, and the H⋯Cl– distances of 2.77 and 2.78 Å, in this
order, are consistent with existence of non-conventional hydrogen
bonding contacts. Likewise, in the NOESY spectra of the chloride
complexes of 17 (Figure S88), the thiophene protons C8–H
correlate through the space with the external urea protons, with
distance between the N6–H and C8–H protons of 2.18 and 2.17 Å
in their respective DFT optimised geometries, which also exhibit
the C–H⋯Cl– interactions with H⋯Cl– distances of 2.68 and 2.69
Å. Along the subsets of benzo[b]thiophene isomers, the binding
association constants follow the order 16 > 14 > 8 and 17 > 15 >
9, showing the paramount importance of C–H⋯Cl– for the chloride
binding strength.
Figure 8. Close up of the 2-D NOESY spectrum of 15 (10.3 mM) with Bu4N+Cl–
(8.1 eq) in DMSO-d6 at 293 K, showing key intramolecular NOE cross-peaks
(marked in magenta). NOE cross-peaks of the urea-NHs (H4 and H6) and H6
and H9 were observed and marked with magenta arrows in the structure (see
Figure S86 for the full spectrum).
Anion efflux studies
The anion transport properties of the new library of molecules
towards chloride were determined using the chloride/nitrate
exchange assay (Figure 9a) monitored by a chloride ion selective
electrode (ISE). Typically, unilamellar vesicles prepared from
POPC, were loaded with a buffered NaCl solution (487 mM, with
5 mM phosphate buffer at pH 7.2), suspended in a buffered
NaNO3 solution (487 mM, with 5 mM phosphate buffer at pH 7.2)
and all molecules were added as a DMSO solution and the rate
of chloride efflux was monitored by the chloride ISE. The initial
rates of chloride transport (kini) obtained using this assay are
shown in Table 1 and Figure S90. In the library of thiophene-
based molecules, all non-fluorinated derivatives displayed poor
chloride transport activities except β-benzo[b]thiophene 14. The
fluorinated molecules displayed better activities compared with
their non-fluorinated analogous, indicating the lipophilicity of
these molecules is an important aspect which could influence the
activity of a molecule. The Cl–/NO3– exchange assay showed the
activity sequence of 15 > 17 > 5 ≈ 4 ≈ 14 ≈ 7 > 9 ≈ 11. The same
trend was also confirmed by concentration-dependent Hill
analysis (Table 1 and Figure S91-S98). The β-benzo[b]thiophene
derivative 15 (EC50 = 0.0089) is the best performing molecule,
followed by γ-benzo[b]thiophene derivative 17 (EC50 = 0.0194),
benefiting from the additional C–H⋯Cl– interactions in chloride
binding by the benzo[b]thiophene motif.
In order to understand the behaviour and transport mechanism of
these thiophene-based molecules in more detail, cationophore
coupling assays (Figure 9a) were used to determine whether the
receptor-mediated ion transport occurs in an electrogenic or
electroneutral fashion or in a nonspecific manner. KCl–loaded
unilamellar vesicles were suspended in an inert external K-
Gluconate solution. Ionophore-induced Cl– efflux was measured
by ISE coupling with either valinomycin (an electrogenic K+
transporter) or monensin (an electroneutral M+/H+ transporter). As
shown in Figure 9b, all tested molecules were more efficient in
mediating electroneutral H+/Cl−(OH−/Cl−) than electrogenic Cl–
Figure 7. Selected partial (400 MHz) 1H NMR spectra for the titration of
Bu4N+Cl– into 15 in DMSO-d6/0.5% H2O at 293 K.
The total E2 of the hydrogen bonding interactions (N–H⋯Cl– and
C–H⋯Cl–) are plotted against the log(Ka) in Figure S89. Apart of
1, 2, 6 and 7 (see Scheme 1), the molecules were grouped in two
subsets: a) α-thiophene and α-benzo[b]thiophene derivatives;
and b) β-benzo[b]thiophene and γ-benzo[b]thiophene derivatives.
In the former subset, the E2 data follow a linear relation, with an
R2 of 0.90. Noteworthy, in subset b, characterised by the ability of
14–17 to recognise chloride through cooperativity between N–
H⋯Cl– and C–H⋯Cl– interactions, a linear relationship is also
observed (R2 = 0.87).
This article is protected by copyright. All rights reserved.

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
transport, and both β-benzo[b]thiophene derivative 15 and γ-
benzo[b]thiophene derivative 17 appeared to be only capable of
transporting chloride ions through electroneutral transport but not
electrogenic transport, while the β-benzo[b]thiophene derivative
14 is capable of both electrogenic and electroneutral transport
(Table S8 and Figure S99). Therefore, as the most two active
anion transporters in this series, 15 and 17 function as
electroneutral transports which promote transport via a pure ion
pair H+/Cl− symport event, or via ion exchange.
acids present in the vesicles giving a fatty acid free lipid bilayer.
The results of these thiophene-based molecules are presented in
Figure S100-S123 and summarized in Table 2. As indicated by
the factor of the chloride transport selectivity (FGra) quantified by
dividing EC50(BSA) by EC50 (Gra), these thiophene-based
molecules only show a slight amount of chloride selective
transport. Dividing EC50(BSA) by EC50(OA) give FOA values > 1 for
all receptors, indicating that all these receptors can couple fatty
acid flip-flop and enhance the overall rate of H+/Cl– symport.
Enhancement factors are greater or equal in the presence of OA
compared to Gra, indicating the loss of the chloride transport
selectivity in the presence of OA. The results demonstrate that
these thiophene-based molecules preferentially facilitate H+/Cl–
symport compared to uniport, reinforcing the electroneutral
transport mechanism shown in the cationophore coupled
assay.[30]
Table 2. EC50 values shown for the N-methyl-ᴅ-glucamine chloride (NMDG-
Cl) assay with bis-urea thiophene-based receptors in bovine serum albumin
(BSA) treated vesicles, in the presence of gramicidin (Gra) and in the
presence of oleic acid (OA).
EC50 (mol %)a
Gra
Selectivity
Molecule
b
c
BSA
OA
FGra
FOA
4
5
7
0.0326
0.0237
0.0169
0.0039
0.0005
0.0334
0.0009
0.0021
0.0272
0.0197
0.0094
0.0027
0.0005
0.0103
0.0004
0.0007
0.0258
0.0184
0.0135
0.0028
0.0004
0.0072
0.0003
0.0007
1.2
1.2
1.8
1.4
1.0
3.2
2.7
3.1
1.3
1.3
1.3
1.4
1.4
4.6
3.4
3.2
9
11
14
15
17
a)
b)
Values reported in transporter to lipid molar ratio (mol%);
Factor of
enhancement in the Cl– uniport in the presence of Gra, FGra is calculated by
dividing the EC50 (BSA) by the EC50 (Gra). FGra > 1 indicates Cl– transport
enhancement;
Factor of enhancement in the overall rate of Cl–/H+
c)
cotransport in the presence of OA, FOA is calculated by dividing the EC50 by
the EC50(OA). FOA > 1 indicates the receptor can assist the flip-flop of OA,
increasing pH dissipation.
Molecular dynamics simulations in a bilayer model
Further insights into the transport ability of our library of bis-urea
derivatives were obtained through Molecular Dynamics (MD)
simulations carried out with the chloride complexes of 3/4, 8/9,
14/15 and 16/17, using AMBER18.[31] These receptors were
selected to ascertain how the molecular size, the binding
geometry, and the fluorination of the central phenyl spacer affect
their interaction and diffusion in the POPC bilayer system
composed of 128 lipids and 6500 TIP3P water molecules. The
receptors were described with the General AMBER Force
Field,[32] while lipid14[33] was used for the POPC molecules. The
complexes were placed either in the bilayer core or in the aqueous
phase of the membrane model, affording the respective starting
scenarios 퓦 and 퓛, and simulated for 200 ns in two (3, 4, 8, and
9) or three (14-17) independent MD runs. Remaining MD
simulation details are given in ESI. The diffusion of the complexes
was monitored tracking the distances between the closest bilayer
interface (Pint, defined by the average z position of its 64
phosphorus atoms) and the centres of mass of the phenyl spacer
(PhCOM, defined by the six carbon atoms) and of the N-H binding
Figure 9. Overview of the anion transport ability of the thiophene-based
compounds in large unilamellar vesicles (LUV). (a) Schematic lay-out of the
experiments, where T = transporter, Mon = monensin, and Vln = valinomycin.
(b) Initial rate of chloride transport calculated by exponential or linear fit for each
transporter (1 mol%) under various assay conditions. Detail experimental
conditions are presented in supporting information. (c) Schematic of the N-
methyl-ᴅ-glucamine chloride (NMDG-Cl) assay, monitored by HPTS
fluorescence: i) in bovine serum albumin (BSA) (1 mol%) treated vesicles, ii) in
the presence of gramicidin (Gra) (0.1 mol%) and iii) in the presence of oleic acid
(OA) (2 mol%).
Additionally, we quantified the Cl– > H+/OH– selectivity of these
transporters using a modified N-methyl-ᴅ-glucamine chloride
(NMDG-Cl) assay in three different conditions (Figure 9c), which
involves the use of the proton channel gramicidin D (Gra), and
bovine serum albumin (BSA) (1 mol% with respect to lipid) treated
vesicles.[29] The addition of fatty acid (oleic acid, OA) to the
NMDG-Cl assay allowed an evaluation of the receptors’ ability to
assist the flip-flop of the negatively charged fatty acid head group
facilitating proton transport. BSA was used to sequester any fatty
This article is protected by copyright. All rights reserved.

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
units (N-HCOM, defined by the four nitrogen atoms). The evolution
of the PhCOM⋯Pint and N-HCOM⋯Pint distances is plotted in Figures
S124-S129, together with the counting of the N-H⋯Cl– hydrogen
bonds throughout the MD runs.
Regardless of the starting scenario, 퓦 or 퓛, the bis-urea
molecules diffuse towards the water/lipid interface. Indeed, even
the ones coming from the water phase (퓦) permeate the interface
and achieve equivalent positions as when they diffuse from the
centre of the bilayer (퓛). Overall, the fluorinated molecules show
a well-defined orientation along the MD simulations with the
phenyl spacer and the N-H binding units closer to the bilayer core
exposed to the water phase, respectively. In contrast, in the non-
fluorinated analogous this orientation is less common, being
swapped with ones where the receptor is tilted, with the N-H
binding units away from the water phase. These dispositions are
illustrated with two snapshots taken from the MD runs of 14 and
15, in scenario 퓛 (see Figure 10).
Figure 11. Average number of hydrogen bonds vs the relative position of the
centre of mass of 15 in the second MD run in scenario 퓛. The following colour
scheme was used for the interactions between 15 and chloride ions (green),
water molecules (cyan), POPC head groups (orange), and ester groups
(magenta for the sn-1 chains and purple for the sn-2 chains). The water/lipid
interface is represented as a black line at z = 0 Å.
The counting of N-H⋯Cl– hydrogen bonds throughout the MD
runs carried out in scenario 퓛 indicates that 3, 4, 14, 8, 9, 16 and
17 typically release the anion within the first 50 ns of simulation.
In contrast, in one of the initial MD runs of 15 the complex
dissociation occurs almost at the end of the MD simulation (ca.
180 ns), which lead us to carry out two additional MD runs. In both
simulations, the anion association was maintained throughout the
200 ns, suggesting that 15 has a higher affinity for chloride at the
water/lipid interface of the POPC bilayer. Indeed, this insight is
remarkable considering that the phenyl spacer is oriented to the
bilayer core and that the complexed chloride is exposed to the
competing water molecules amongst the phospholipid head
groups. This result seems to be consistent with the high
association constant of 15 for chloride in DMSO/H2O (see Table
1), as well as its high anion efflux ability. On the other hand, 17
has the largest association constant for the halide, and lower
transport properties. This intriguing result lead us to determine the
association constants of the β- and γ-benzo[b]thiophenes (14-17)
Figure 10. Snapshots of MD runs with 14 and 15 in scenario 퓛, illustrating the
opposite orientations acquired by the fluorinated and non-fluorinated receptors
at the water/lipid interface. The water molecules and aliphatic protons were
hidden for clarity.
–
for H2PO4 , as a prototype of the lipid head group. The association
constants of these four molecules are extremely high, with the
values for 14 and 15 being 11749 and 25119 M-1, respectively,
while for 16 and 17 the values are larger than 105 M-1.[34] Thus,
given that the two pairs of isomeric compounds have the same
lipophilicities (see Table 1), the transport ability of 16 vs 14 and
17 vs 15 seems to be dictated by a delicate balance between their
affinities for chloride and phosphate head groups. Therefore, 15,
with intermediate association constants for both anions, naturally
emerges as the best transporter in this library of small molecules.
Moreover, in the MD simulations in scenario 퓦, the chloride is
promptly hydrated, with the anion release preceding the bis-urea
derivatives permeation of the bilayer. In contrast, in the alternative
scenario 퓛, as the complexes approach the water/lipid interface,
the nearby water molecules promote the chloride release to the
aqueous phase. Regardless of the starting scenario, as illustrated
in Figure 11, with the counting of hydrogen bonds between 15 and
Cl–, phosphate head groups, sn-1 and sn-2 ester groups, and
water molecules as a function of the position of the transporter
along the bilayer normal (z axis), the N-H⋯Cl– interactions are
mainly replaced by hydrogen bonding interactions with the
phosphate head groups (see Figures S130-S135 for the
remaining MD runs). The high affinity of our small molecules for
these hydrogen bonding acceptors (vide infra) might play an
important role in their ability to promote the chloride
transmembrane transport given that the anion uptake necessarily
occurs at the water/lipid interface level, as we have previously
demonstrated.[27-28]
Cytotoxic screening
The cytotoxicity of the non-fluorinated/fluorinated analogous 3/4,
8/9, 14/15, 16/17, on HeLa, MCF-7 and A549 cancer cell lines
was ascertained by the resazurin assay. These studies were
further extended to CFBE cells, a human CF bronchial epithelial
cell model, which has been used on functional studies of mutant
CFTR in native context.[35] These cells were exposed to different
concentrations of compounds for 24 h yielding the IC50 values
presented in Table 3.
This article is protected by copyright. All rights reserved.

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
Compounds 8, 14, 16 and 17 were found to be non-toxic towards
all the cell lines tested. The remaining compounds were cytotoxic
to some extent, with the fluorinated ones being the most toxic, as
determined by the lower IC50 values. In addition, the HeLa cells
were the most sensitive line, with IC50 values of 12.54 and 27.61
μM for 15, and 4, respectively, while 9 and 4 were particularly
cytotoxic for the A549 cell line, with IC50 values of 17.66 and 48.74
μM, respectively. Concerning the MCF-7 cells, only the α-
the N-H binding units oriented towards the water phase prompt to
uptake a chloride, as requested by the anion carrier mechanism.
In contrast, the non-fluorinated analogous swap between multiple
orientations and preferentially interact with phospholipid head
groups, which necessarily limits their transport abilities as
indicated by lower initial rates of chloride transport obtained for
these molecules. Noteworthy, the chloride complex of 15 has
longer lifetimes throughout the MD runs in the membrane system,
which is consistent with its superior anion transport.
thiophene derivatives show
a moderate activity, with the
fluorinated 4 displaying an IC50 value (37.33 μM) that is almost
half of its non-fluorinated analogous 3 (68.32 μM). In respect to
the CFBE cell line, 3, 4 and 9 display IC50 values below 100 μM,
while the most efficient transporter 15 has an IC50 of 124.80 μM.
This low cytotoxic level is significative considering the potential
use of 15 as a Channel Replacement Therapy.
The transport properties achieved with this library of molecules,
inspired us to decorate other supramolecular scaffolds with sulfur
heteroaromaric motifs. This project is ongoing in our laboratories.
Acknowledgements
Table 3. IC50 values (μM) for selected thiophene-based compounds on
The CFBE cell line was kindly provided by Prof. Margarida D.
Amaral (Coordinator of BioISI – Biosystems & Integrative
Sciences Institute, Department of Chemistry and Biochemistry,
Faculty of Sciences, University of Lisbon). This work was
supported by the projects PTDC/QEQ-SUP/4283/2014,
UID/MULTI/00612/2019 and CICECO – Aveiro Institute of
Materials (UID/CTM/50011/2019), financed by National Funds
through the FCT/MEC and co-financed by QREN-FEDER through
COMPETE under the PT2020 Partnership Agreement. M. M. is
thankful for PhD grant PD/BD/142933/2018 from FCT. I.M. is
grateful for a postdoctoral grant (BPD/UI98/6065/2018) under
project “pAGE” (Centro-01-0145-FEDER-000003), co-funded by
Centro 2020 programme, Portugal 2020, European Union,
through the European Regional Development Fund. PAG thanks
the ARC (DP180100612) and the University of Sydney for funding.
HeLA, MCF-7, A549 and CFBE cell lines.a
Compound
HeLa
59.33
27.61
48.73
12.54
A549
74.59
48.74
17.66
83.23
MCF-7
CFBE
86.20
51.41
27.86
124.80
68.32
3
4
37.33
-
-
9
15
a Compounds 8, 14, 16 and 17 have IC50 values superior to 125 μM, being
not reported.
Conclusions
Overall, the experimental binding data reported for the six-
subsets (see Table 1) of thiophene based receptors show that the
fluorination of the receptor skeleton increases the affinity for
chloride. This finding mirrors the acidic character of the two urea-
binding units ascertained by the VS,max quantum descriptor as well
as by the strength of the hydrogen bonding interactions estimated
by E2 energies. It is worth noting that the affinity of β-, and γ-
benzo[b]thiophene derivatives for chloride significantly increase
as result of a synergetic recognition by N-H⋯Cl– and C-H⋯Cl–
hydrogen bonding interactions, as thoroughly documented by 1H
NMR data and theoretical calculations. Moreover, in these
subsets the impact of the fluorination of the central phenyl spacer
on the association constants is noticeable with the fluorinated
analogous showing the higher values (163.7 M-1 for 15 and 207.5
M-1 for 17) of our extensive library of small molecules. On the
other hand, regardless of the fluorination of the aromatic spacer,
the control of the receptors’ conformation, through the C-H⋯O
interactions, has a relevant role in the superior binding affinity of
the β-, and γ-benzo[b]thiophene derivatives (14-17).
Keywords: Anion transport • thiophene-based molecules • efflux
studies • molecular modelling • cytotoxicity
[1]
F. M. Ashcroft, Ion Channels and Disease, Vol. 1, 1st ed., Academic
Press, London, 2000
.
[2]
[3]
B. P. O'Sullivan, S. D. Freedman, The Lancet 2009, 373, 1891-1904.
M. A. Ashlock, R. J. Beall, N. M. Hamblett, M. W. Konstan, C. M.
Penland, B. W. Ramsey, J. M. Van Dalfsen, D. R. Wetmore, P. W.
Campbell, 3rd, Semin. Respir. Crit. Care Med. 2009, 30, 611-626.
E. J. Jeong, E. J. Kang, L. T. Sung, S. K. Hong, E. Lee, J. Am. Chem.
Soc. 2002, 124, 14655-14662.
[4]
[5]
[6]
[7]
T. R. Sheth, R. M. Henderson, S. B. Hladky, A. W. Cuthbert, Biochim.
Biophys. Acta 1992, 1107, 179-185.
S. Ohkuma, T. Sato, M. Okamoto, H. Matsuya, K. Arai, T. Kataoka,
K. Nagai, H. H. Wasserman, Biochem. J 1998, 334, 731-741.
K. A. Muraglia, R. S. Chorghade, B. R. Kim, X. X. Tang, V. S. Shah,
A. S. Grillo, P. N. Daniels, A. G. Cioffi, P. H. Karp, L. Zhu, M. J. Welsh,
M. D. Burke, Nature 2019, 567, 405-408.
[8]
[9]
(a) P. A. Gale, E. N. W. Howe, X. Wu, M. J. Spooner, Coord. Chem.
Rev. 2018, 375, 333-372; (b) X. H. Yu, X. Q. Hong, Q. C. Mao, W. H.
Chen, Eur. J. Med. Chem. 2019, 184, 111782.
P. A. Gale, Acc. Chem. Res. 2011, 44, 216-226.
[10] (a) J. L. Seganish, J. T. Davis, Chem. Commun. 2005, 5781-5783;
The most effective chloride transmembrane transporters are 15
and 17 followed by 14, while 16, the non-fluorinated analogous of
17, is surprisingly almost inactive. Indeed, the anion efflux data
obtained for fluorinated and non-fluorinated analogous of our
library of small molecules indicates that the fluorination of the
central ring potentiates the chloride transport across POPC
bilayer. Accordingly, molecular dynamics simulations show that
the fluorinated molecules reside below the bilayer interface with
(b) J. L. Sessler, L. R. Eller, W. S. Cho, S. Nicolaou, A. Aguilar, J. T.
Lee, V. M. Lynch, D. J. Magda, Angew. Chem. Int. Ed. Engl. 2005
,
44, 5989-5992; (c) P. Iglesias Hernandez, D. Moreno, A. A. Javier, T.
Torroba, R. Perez-Tomas, R. Quesada, Chem. Commun. 2012, 48,
1556-1558; (d) V. Saggiomo, S. Otto, I. Marques, V. Félix, T. Torroba,
R. Quesada, Chem. Commun. 2012, 48, 5274-5276.
[11] (a) O. Jurček, H. Valkenier, R. Puttreddy, M. Novák, H. A. Sparkes,
R. Marek, K. Rissanen, A. P. Davis, Chem. Eur. J. 2018, 24, 8178-
8185; (b) D. H. Sohn, E. Han, J. I. Park, S. J. Cho, J. Kang, Supramol.
Chem. 2018, 30, 169-178.
This article is protected by copyright. All rights reserved.

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
[12] (a) A. V. Jentzsch, D. Emery, J. Mareda, S. K. Nayak, P. Metrangolo,
G. Resnati, N. Sakai, S. Matile, Nat. Commun. 2012, 3, 905; (b) A.
Vargas Jentzsch, A. Hennig, J. Mareda, S. Matile, Acc. Chem. Res.
2013, 46, 2791-2800; (c) A. Vargas Jentzsch, S. Matile, J. Am. Chem.
Soc. 2013, 135, 5302-5303; (d) A. V. Jentzsch, S. Matile, Top. Curr.
Chem. 2015, 358, 205-239.
[13] (a) S. Benz, M. Macchione, Q. Verolet, J. Mareda, N. Sakai, S. Matile,
J. Am. Chem. Soc. 2016, 138, 9093-9096; (b) M. Macchione, M.
Tsemperouli, A. Goujon, A. R. Mallia, N. Sakai, K. Sugihara, S. Matile,
Helv. Chim. Acta 2018, 101, e1800014; (c) L. M. Lee, M. Tsemperouli,
A. I. Poblador-Bahamonde, S. Benz, N. Sakai, K. Sugihara, S. Matile,
J. Am. Chem. Soc. 2019, 141, 810-814.
[14] (a) L. Martinez-Crespo, J. L. Sun-Wang, P. Ferreira, C. F. M.
Mirabella, G. Aragay, P. Ballester, Chem. Eur. J. 2019, 25, 4775-
4781; (b) W. L. Huang, X. D. Wang, S. Li, R. Zhang, Y. F. Ao, J. Tang,
Q. Q. Wang, D. X. Wang, J. Org. Chem. 2019, 84, 8859-8869.
[15] H. Li, H. Valkenier, A. G. Thorne, C. M. Dias, J. A. Cooper, M. Kieffer,
N. Busschaert, P. A. Gale, D. N. Sheppard, A. P. Davis, Chem. Sci.
2019, 10, 9663-9672.
[16] (a) A. M. Rodilla, L. Korrodi-Gregorio, E. Hernando, P. Manuel-
Manresa, R. Quesada, R. Perez-Tomas, V. Soto-Cerrato, Biochem.
Pharmacol. 2017, 126, 23-33; (b) N. Busschaert, S. H. Park, K. H.
Baek, Y. P. Choi, J. Park, E. N. W. Howe, J. R. Hiscock, L. E.
Karagiannidis, I. Marques, V. Félix, W. Namkung, J. L. Sessler, P. A.
Gale, I. Shin, Nat. Chem. 2017, 9, 667-675; (c) N. Akhtar, A. Saha, V.
Kumar, N. Pradhan, S. Panda, S. Morla, S. Kumar, D. Manna, ACS
Appl Mater Interfaces 2018, 10, 33803-33813; (d) R. Quesada, Chem
2019, 5, 1924-1926.
Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V.
Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O.
Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L.
Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J.
Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman,
J. V. Ortiz, J. Cioslowski, and D. J. Fox, Gaussian, Inc., Wallingford
CT, 2009.
[24] (a) F. Weinhold, C. R. Landis, Chem. Educ. Res. Pract. 2001, 2, 91-
104; (b) E. D. Glendening, C. R. Landis, F. Weinhold, J. Comput.
Chem. 2013, 34, 1429-1437.
[25] Calculator Plugins were used for structure property prediction and
calculation, Marvin Sketch 19.2.0-9405, 2019, ChemAxon
(http://www.chemaxon.com).
[26] C. Frassineti, S. Ghelli, P. Gans, A. Sabatini, M. S. Moruzzi, A. Vacca,
Anal. Biochem. 1995, 231, 374-382.
[27] M. J. Spooner, H. Li, I. Marques, P. M. R. Costa, X. Wu, E. N. W.
Howe, N. Busschaert, S. J. Moore, M. E. Light, D. N. Sheppard, V.
Félix, P. A. Gale, Chem. Sci. 2019, 10, 1976-1985.
[28] I. Marques, P. M. R. Costa, Q. M. M, N. Busschaert, E. N. W. Howe,
H. J. Clarke, C. J. E. Haynes, I. L. Kirby, A. M. Rodilla, R. Perez-
Tomas, P. A. Gale, V. Félix, Phys. Chem. Chem. Phys. 2018, 20,
20796-20811.
[29] X. Wu, L. W. Judd, E. N. W. Howe, A. M. Withecombe, V. Soto-
Cerrato, H. Li, N. Busschaert, H. Valkenier, R. Perez-Tomas, D. N.
Sheppard, Y. B. Jiang, A. P. Davis, P. A. Gale, Chem 2016, 1, 127-
146.
[30] X. Wu, J. R. Small, A. Cataldo, A. M. Withecombe, P. Turner, P. A.
Gale, Angew. Chem. Int. Ed. Engl. 2019, 58, 15142-15147.
[31] D. A. Case, I. Y. Ben-Shalom, S. R. Brozell, D. S. Cerutti, T. E.
Cheatham, III, V. W. D. Cruzeiro, T. A. Darden, R. E. Duke, D.
Ghoreishi, G. Giambasu, T. Giese, M. K. Gilson, H. Gohlke, A. W.
Goetz, D. Greene, R Harris, N. Homeyer, Y. Huang, S. Izadi, A.
Kovalenko, R. Krasny, T. Kurtzman, T. S. Lee, S. LeGrand, P. Li, C.
Lin, J. Liu, T. Luchko, R. Luo, V. Man, D. J. Mermelstein, K. M. Merz,
Y. Miao, G. Monard, C. Nguyen, H. Nguyen, A. Onufriev, F. Pan, R.
Qi, D. R. Roe, A. Roitberg, C. Sagui, S. Schott-Verdugo, J. Shen, C.
L. Simmerling, J. Smith, J. Swails, R. C. Walker, J. Wang, H. Wei, L.
Wilson, R. M. Wolf, X. Wu, L. Xiao, Y. Xiong, D. M. York and P. A.
Kollman (2018), AMBER 2018, University of California, San
Francisco.
[17] D. N. Sheppard, A. P. Davis, Nature 2019, 567, 315-317.
[18] (a) D. Gramec, L. Peterlin Masic, M. Sollner Dolenc, Chem. Res.
Toxicol. 2014
, 27, 1344-1358; (b) R. S. Keri, K. Chand, S.
Budagumpi, S. Balappa Somappa, S. A. Patil, B. M. Nagaraja, Eur. J.
Med. Chem. 2017, 138, 1002-1033.
[19] W. M. Haynes, CRC Handbook of Chemistry and Physics, 97th ed.,
CRC Press LLC Taylor & Francis Group, Boca Raton; Florence,
2016
.
[20] C. M. Dias, H. Valkenier, A. P. Davis, Chem. Eur. J. 2018, 24, 6262-
6268.
[21] (a) S. J. Moore, C. J. E. Haynes, J. Gonzalez, J. L. Sutton, S. J.
Brooks, M. E. Light, J. Herniman, G. J. Langley, V. Soto-Cerrato, R.
Perez-Tomas, I. Marques, P. J. Costa, V. Félix, P. A. Gale, Chem.
Sci. 2013, 4, 103-117; (b) C. M. Dias, H. Li, H. Valkenier, L. E.
Karagiannidis, P. A. Gale, D. N. Sheppard, A. P. Davis, Org Biomol
Chem 2018, 16, 1083-1087.
[32] (a) J. Wang, R. M. Wolf, J. W. Caldwell, P. A. Kollman, D. A. Case, J.
Comput. Chem. 2004, 25, 1157-1174; (b) J. Wang, R. M. Wolf, J. W.
Caldwell, P. A. Kollman, D. A. Case, J. Comput. Chem. 2005, 26,
114-114.
[22] J. Tomasi, B. Mennucci, R. Cammi, Chem. Rev. 2005, 105, 2999-
3093.
[33] C. J. Dickson, B. D. Madej, A. A. Skjevik, R. M. Betz, K. Teigen, I. R.
Gould, R. C. Walker, J. Chem. Theory Comput. 2014, 10, 865-879.
–
[23] Gaussian 09, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P.
Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M.
Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery,
Jr., J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K.
N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K.
Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M.
[34] H2PO4 added as TBA salt and the binding constants were
determined by 1H NMR titrations in DMSO-d6/0.5% H2O at 293 K
using HypNMR. The remaining molecules deprotonated in presence
–
of H2PO4
.
[35] L. B. Gottschalk, B. Vecchio-Pagan, N. Sharma, S. T. Han, A. Franca,
E. S. Wohler, D. A. Batista, L. A. Goff, G. R. Cutting, J Cyst Fibros
2016, 15, 285-294.
This article is protected by copyright. All rights reserved.

10.1002/chem.201904255
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents
FULL PAPER
A library of drug-like molecules,
containing thiophene and
Paulo Vieira, Margarida Q. Miranda, Igor
Marques, Sílvia Carvalho, Li-Jun Chen,
Ethan N. W. Howe, Carl Zhen, Claudia
Y. Leung, Michael J. Spooner, Bárbara
Morgado, Odete A. B. da Cruz e Silva,
Cristina Moiteiro,* Philip A. Gale,* Vítor
Félix*
benzo[b]thiophene motifs, was
designed for the transmembrane
transport of chloride. The theoretical
and experimental investigation
showed, that the most active ones
operate the anion transport via a
symport mechanism and taking
advantage of the synergy between N-
H⋯Cl– and C-H⋯Cl– hydrogen bonds.
Page No. – Page No.
Development of a library of
thiophene-based drug-like LEGO
molecules: evaluation of their anion
binding, transport properties and
cytotoxicity
This article is protected by copyright. All rights reserved.