A ligand-based virtual screening approach to identify small molecules as hERG channel activators

Article abstract of DOI:10.2174/1386207318666150305121841

The hERG potassium channel is currently emerging as a potential target for the treatment of some forms of arrhythmias or to contrast an unintentional channel block caused by drugs. Despite its therapeutic relevance, so far only few compounds are described

Full text of DOI:10.2174/1386207318666150305121841

Send Orders for Reprints to reprints@benthamscience.ae
Combinatorial Chemistry & High Throughput Screening, 2015, 18, 269-280
269
A Ligand-Based Virtual Screening Approach to Identify Small Molecules
as hERG Channel Activators
Elisa Giacomini1,2,§, Rosa Buonfiglio1,§,‡, Matteo Masetti1, Yuhong Wang3, Gea-Ny Tseng3,
Marinella Roberti*,1 and Maurizio Recanatini1
1Department of Pharmacy and Biotechnology, Alma Mater Studiorum - Università di Bologna, via
Belmeloro 6, 40126 Bologna, Italy
2D3 Computation, Istituto Italiano di Tecnologia, via Morego 30, 16163 Genova, Italy
3Department of Physiology & Biophysics, Virginia Commonwealth University, Richmond, VA 23298,
USA
Abstract: The hERG potassium channel is currently emerging as a potential target for the treatment of
some forms of arrhythmias or to contrast an unintentional channel block caused by drugs. Despite its
therapeutic relevance, so far only few compounds are described as able to enhance channel function by
potentiating hERG currents. This gap is also related to the lack of hERG crystal structure which strongly limits the
possibility to employ structure-based techniques in the search and design of novel activators. To overcome this limitation,
in the present work, a ligand-based virtual screening was performed using as separate search queries two conformations of
NS1643, the most deeply investigated and better characterized hERG activator. The library of compounds resulting from
the virtual screening was then clustered based on recurring chemical features, and 5 hits were selected to be evaluated for
their ability to enhance hERG current in vitro. Compound 3 showed a good activating effect, also displaying a mechanism
of action similar to that of NS1643. Moreover, the most interesting compounds were further investigated by synthesizing
in a parallel fashion some analogs, with the aim to get insights about structure-activity relationships.
Keywords: Activators, electrophysiology, hERG potassium channel, ligand-based virtual screening, NS1643, parallel
synthesis.
The hERG-K+ channel belongs to the voltage-gated
family of potassium channels (Kv), and therefore it shows
INTRODUCTION
Ion channels are transmembrane proteins responsible for
the passive and selective transport of ions through biological
membranes, and thereby they contribute to regulate the
excitability and osmotic equilibrium of living cells [1, 2].
These proteins are widely distributed all over the human
body, and they are involved in several processes, such as
transmission of nerve impulses, muscle contraction and
relaxation, cognition, as well as in many other aspects of
physiology [1]. The key role played by ion channels in the
organism is mirrored by the large number of known diseases
associated to their malfunctioning, which makes them
attractive targets in the drug discovery field [3]. Among the
others, the hERG potassium channel (Kv11.1) has aroused
considerable scientific interest due to its relevant
physiological function in the heart and its implication in
arrhythmias.
typical structural features shared by all members of this
family, i.e. the tetrameric assembly of six transmembrane
helices around the pore axis, the voltage sensing domain
(helices S1 - S4), and the pore forming domain (helices S5 -
S6) [4-6]. Along with these similarities, hERG shows several
peculiarities of its own that appear to be responsible for the
promiscuous drugs interactions [7, 8] or correlated to
trafficking defects [9]. However, the most distinctive feature
of the channel relies on its unique gating properties,
characterized by slow activation and deactivation kinetics in
opposition to a very fast voltage-dependent inactivation.
These unusual kinetics find an explanation in the main role
played by hERG in the ventricular cardiac action potential
(AP) and refractoriness of cardiac cells. In cardiomyocytes,
the channel slowly opens when the membrane starts to
depolarize from the holding potential of -80 mV to more
positive values, but once a value of -40 mV is reached, it
rapidly enters in the inactivated state. It is only during late
phase 2 of the AP that hERG quickly recovers from
inactivation causing the typical large tail currents that
significantly contribute to the membrane repolarization.
Since these currents reach their maximum amplitude during
phase 3 of the AP, they are properly referred to as rapidly
activating delayed rectifier currents (IKr) [10].
*Address correspondence to this author at the Department of Pharmacy and
Biotechnology, Alma Mater Studiorum - Università di Bologna, via
Belmeloro 6, 40126 Bologna, Italy; Tel: +(39) 051 2099738;
E-mail: marinella.roberti@unibo.it
§These authors contributed equally to this work.
‡Present address: Computational Chemistry, Chemistry Innovation Centre,
Discovery Sciences. AstraZeneca R&D Mölndal, SE-43183 Mölndal,
Sweden.
Given the description above, it is clear that alterations of
hERG channel function, either due to genetic mutations or
1875-5402/15 $58.00+.00
© 2015 Bentham Science Publishers

270 Combinatorial Chemistry & High Throughput Screening, 2015, Vol. 18, No. 3
Giacomini et al.
drug modulation, could lead to serious health conditions
[11]. Notably, the most important and investigated
correlation between the channel dysfunction and a cardiac
pathology, regards the Long QT Syndrome (LQTS) [12].
LQTS is caused by an abnormal repolarization phase of the
ventricular AP that is diagnosed by a marked prolongation of
the QT interval of the electrocardiogram. People affected by
this condition are predisposed to a rare but potentially fatal
form of ventricular tachyarrhythmia named torsades de
pointes (TdP) [13]. The unintentional hERG blockade by
drugs is the major cause for the acquired LQTS, and over the
past decades many drugs have been withdrawn from the
market because of this unwanted effect [14]. In this scenario,
hERG has always been traditionally considered as an
structure, do severely hamper a reliable VS strategy based on
the knowledge of the three-dimensional structure of the
target. To overcome this limitation, we performed a ligand-
based VS (LBVS) of the Enamine vendor catalog [24] using
two conformations of the NS1643 activator as search queries
(Fig. 1A). Then, the most interesting hits among the final
library were purchased, and some of them were identified as
moderate activators of the hERG-K+ channel. Structure-
activity relationships (SAR) were also undertaken with the
aim to rationalize the effects of substituents on the enhancing
effect of IKr currents.
MATERIALS AND METHODS
Ligand-Based Virtual Screening
antitarget, and many efforts have been undertaken in helping
to identify potential blockers at the early stages of a drug
design process with both in vitro and in silico methods
[15,16]. The concept to consider hERG as a therapeutic
target started to take place only in the past decade [17].
Indeed, compounds able to enhance channel function by
potentiating hERG activity would accelerate the
repolarization and shorten the duration of the AP [5], and
such activators might become useful therapeutics to treat
some forms of arrhythmias or simply to contrast the effect of
the channel block. Despite their unquestionable therapeutic
potential, only few compounds able to enhance IKr currents
have been described so far, and for this reason there is a
compelling need to search for more potent and/or diverse
activators. By screening a large compound library of
previously collected electrophysiology data, Zhang and
coworkers have recently discovered a remarkably potent and
effective hERG activator [18]. Without disregarding the
impressive result, the drawback of this approach is either the
requirement of prior knowledge, or the cost needed to build a
new library from scratch. A well-established alternative
comes from virtual screening (VS) approaches, which have
long been used as valuable tools to screen libraries of up to
millions of compounds in an effective and relatively
inexpensive way. Here, we investigate the possibility to
discover novel hERG modulators by means of such
computational strategies.
The LBVS was performed using the workflow outlined
in Fig. (1B). All the operations hereafter detailed were
performed using programs belonging to the Openeye
Scientific Software toolkit [25] unless otherwise stated.
NS1643, an effective type 2 hERG activator showing a
maximal increase in currents amplitude of more than 300%
at the concentration of 30 μM, and an apparent Kd value of
20 μM [7], was used as a query for the LBVS. Since the
screening was performed so as to match the 3D chemical
features of a reference compound (see below), it is of pivotal
importance to represent the query molecule in its bioactive
conformation [26]. However, in the absence of such an
information, the syn and anti conformations of NS1643 were
considered for the screening (Fig. 1A). Both conformations
were built and optimized with the Gaussian03 program
package [27]. The B3LYP functional was used together with
the split-valence 6-31G(d) basis set. The resulting structures
were used as separate search queries in the step 2 and 3 of
the LBVS.
Step 1. Drug-likeness and scaffold filtering. The Enamine
vendor catalog [24], consisting of more than 106 compounds,
was downloaded from the ZINC database [28] and used as
compound library. The library was then filtered with the
FILTER program [29], in order to discard non-drug like
compounds and at the same time to focus the database
towards desired knowledge-based derived chemical
properties.
Based on the current knowledge, hERG activators can
enhance IKr currents with different mechanisms of action,
and they have been retrospectively classified as type 1 and
type 2 [4]. Both of them attenuate inactivation, but while
type 1 activators severely slow deactivation, supposedly by
binding on the pore domain of the channel, type 2 activators
do not strongly affect deactivation properties, as they
primarily act by shifting the voltage dependence of
inactivation to more positive potentials. 1,3-Bis-(2-hydroxy-
5-trifluoromethyl-phenyl)-urea (NS1643, see Table 1) [19,
20] is considered as one of the most important and
thoroughly examined type 2 activators [21]. It affects hERG
in a concentration-dependent manner working from the
extracellular side of the cell membrane, and its enhancing
effect also involves a slowing of hERG inactivation, as well
as an acceleration of channel activation and slowed
deactivation. The location of the NS1643 binding site in the
channel is still under debate [22, 23], and the possibility of
multiple binding sites triggering diverse pharmacological
effects has also been reported [20, 22, 23]. These
uncertainties, together with the lack of the hERG crystal
The molecular weight (MW), the number of rotatable
bonds (NRB), the octanol-water partition coefficient (logP,
calculated through the XlogP algorithm) [30], and the 2D
polar surface area (2D-PSA) were chosen as basic physico-
chemical descriptors to be used in the filtering. These
quantities were calculated with the Molprop tool for a set of
six known hERG activators (including NS1643, see Table 1),
and the upper and lower intervals of the filter were tuned so
as to include approximately one standard deviation from the
mean for each descriptor. Thus, the filter parameters were set
to: MW = 300 - 500 Da, NRB = 2 - 7, logP = 2.4 - 5.4, and
2D-PSA = 37 - 110 Å2. Besides these properties, the library
was also filtered in order to exclude specific chemical
groups. Reactive or labile groups, toxic groups, and
protecting groups were limited or removed from the
database, in order to minimize the chance to incur in false
positives and false negatives. Thus, we discarded entries
containing groups such as quinone, triflate, phosphoramide,

hERG Channel Activators
Combinatorial Chemistry & High Throughput Screening, 2015, Vol. 18, No. 3 271
Table 1. Known hERG channel activators with the corresponding molecular descriptors used in step 1 of the LBVS protocol.
Activators
MW (Da)
NRB
logP
PSA (A2)
380.06
6
2.718
81.59
NS1643
509.17
384.06
7
6
4.398
5.104
79.73
49.33
RPR260243
PD118057
NS3623
426.01
399.09
351.13
6
6
5
3.784
2.919
3.771
95.59
144.52
65.98
Mallotoxin
A935142
isonitriles, sulfonylnitriles, aldehydes, azides, azocompounds,
peroxides. Default values as implemented in FILTER
program were used. Moreover, dyes were not considered
eligible because of non-specific binding to proteins. Finally,
some functional groups were allowed in moderate quantity,
including hydroxyls (no more than 6), olefins (no more than
4), and amides (no more than 4). After applying this filter,
the total number of structures decreased to about 600.000
entries.
features, a conformational analysis was performed with the
OMEGA software [31]. To minimize the risk of discarding
potentially interesting conformations, we used the default
OMEGA settings of an energy window of 10 kcal mol-1
above the global minimum [32-34]. As a tradeoff between
this requirement and computational efficiency, 50
conformers were generated for each compound, with a
resolution of the conformational search of 1 Å measured on
the heavy atoms RMSD.
Since the following steps of the LBVS procedure relied
on similarity searches based on comparison of 3D chemical
Step 2. Shape-based similarity search. The shape
similarity was evaluated with the Rapid Overlay of Chemical

272 Combinatorial Chemistry & High Throughput Screening, 2015, Vol. 18, No. 3
Giacomini et al.
Fig. (1). (A) Graphical representation of the syn and anti conformations of NS1643 employed as search queries for the LBVS. On the left,
the conformers are shown in sticks and the atoms are colored according to their atomic number, whereas on the right, the molecules are
represented as van der Waals spheres colored according to the ab initio electrostatic potential (red to blue for more negative and more
positive values, respectively). (B) The LBVS workflow.
Structures (ROCS) software [35, 36], which aligns
molecules by optimizing the overlap between their volumes.
The Tanimoto Combo was used as a metric to score and rank
the overlaps. After a preliminary visual inspection, a cutoff
of 0.8 was found as the optimal threshold to filter out
database molecules with a low shape overlap, and was
employed to prune the library. All the other parameters were
set to default values. The conformations were processed with
ROCS as unique entries and stored in the output hits file
based on their individual shape overlay. The 3D structures of
the syn and anti conformations of NS1643 were separately
used as reference structures for two parallel screenings.
syn and anti conformers were used independently to screen
the database.
Step 4. Scaffold clustering. Among the two ensembles of
screened molecules, only entries satisfying simultaneously
the 3D features of the syn and anti conformations were
further considered (consensus scoring). This allowed us to
reduce the database to a number of 158 molecules. Since all
molecules consisted of a pair of diversely substituted
aromatic rings separated by recurring spacers, we decided to
cluster the library based on the chemical features of this
central portion. A hierarchical clustering algorithm was used,
as implemented in the ICM software [41]. In particular, the
tree is built using the UPGMA (Unweighted Pair Group
Method Using Arithmetic Average) algorithm, based on
sequential clustering of the smallest pairwise distance in the
Tanimoto distance matrix of the molecules [42]. In Table 2,
the chemical structures of the 9 clusters are reported. Five
molecules, that is the 3% of this library, were finally selected
to be tested for their ability to enhance hERG currents.
Step 3. Electrostatic-based similarity search. The
electrostatic similarity was evaluated with EON [37] by
comparing the electrostatic potential map of the molecules in
their relative orientation as obtained by ROCS in the
previous step. The MMFF94 [38] force field was used to
calculate partial charges for the compounds of the library,
whereas RESP charges [39, 40] calculated at the B3LYP/6-
31G(d)//B3LYP/6-31G(d) level of theory with the
Gaussian03 package [27] were used for the query molecule
in both conformations. The electrostatic potential of the syn
and anti conformations of NS1643 is shown in Fig. (1A). For
the calculation of partial charges, all the molecules were
considered in the most populated protonation state at neutral
pH. The Electrostatic Tanimoto Combo was used to assess
the electrostatic similarity, and the 1000 top ranked
compounds were stored applying a cutoff value of 1.22.
Similarly to the previous step, the electrostatic maps of the
Chemistry
General Chemical Methods. Reaction progress was
monitored by TLC on pre-coated silica gel plates (Kieselgel
60 F254, Merck) and visualized by UV254 light. Compounds
1-5 (Table 3) were purchased by Enamine, while compounds
6-16 (Table 3) were synthesized as shown in Scheme 1. All
reagents were purchased from Aldrich and were of analytical
grade, so they were used without further purification. All

hERG Channel Activators
Combinatorial Chemistry & High Throughput Screening, 2015, Vol. 18, No. 3 273
Table 2. The nine scaffold clusters identified for the 158 molecules. R and R1 are substituted aromatic rings such as benzene,
pyridine, thiophene, furane, or fused rings variously substituted.
Scaffold 2
Scaffold 5
Scaffold 8
Scaffold 3
Scaffold 1
Scaffold 4
Scaffold 7
Scaffold 6
Scaffold 9
solvents were distilled prior to use. All reactions were
carried out under an inert atmosphere. Compounds were
named relying on the naming algorithm developed by
CambridgeSoft Corporation and used in Chem-BioDraw
Ultra 13.0. 1H-NMR and 13C-NMR spectra were recorded on
Varian Gemini at 400 MHz and 100 MHz respectively.
Chemical shifts (δH) are reported relative to TMS as internal
standard. IR-FT spectra were performed in Nujol and
obtained on a Nicolet Avatar 320 E.S.P. instrument; ν value
max is expressed in cm-1. Mass spectrum was recorded on a
V.G. 7070E spectrometer or on a Waters ZQ 4000 apparatus
operating in electrospray (ES) mode.
N-(2,5-Dimethoxyphenyl)-2-(Phenylthio)Acetamide 1
2-(Phenylthio)acetic acid 17 (0.30 g, 1.78 mmol), 2,5-
dimethoxyaniline 22 (0.32 g, 2.14 mmol) were allowed to
react according to the described general procedure and the
crude product was purified on silica eluting with petroleum
ether and ethyl acetate 9.3:0.7. 1: 0.34 g (yield 63 %); brown
1
solid. H-NMR (400 MHz, CDCl3) δ 3.75 (s, 3H), 3.76 (s,
2H), 3.77 (s, 3H), 6.56 (dd J= 8.8, 2.8 Hz, 1H), 6.76 (d, J=
8.8 Hz, 1H), 7.18-7.22 (m, 1H), 7.26-7.30 (m 2H), 7.37-7.39
(m, 2H), 8.04 (d, J= 3.2 Hz, 1H), 9.22 (br, 1H) ppm. 13C-
NMR (100 MHz, CDCl3) δ 38.6, 55.1, 55.7, 105.2, 108.3,
110.4, 126.4, 127.2, 128.5, 128.7, 133.6, 141.9, 153.2, 165.3
ppm. MS (ES): m/z 326 (M + Na+).
General Parallel Procedure for the Synthesis of Amides
1, 4, 8-12
N-(5-Chloro-2,4-Dimethoxyphenyl)-3a,7a-Dihydro-1H-Indazole-
3-Carboxamide 4
In distinct reactors, carboxylic acids (1 equiv) 17-21, 1-
hydroxybenzotriazole (1.2 equiv), and suitable amines 22-24
(1.2 equiv), were dissolved in dimethylformamide at room
temperature. DCC (1.2 equiv) was added, and the mixtures
were stirred at room temperature overnight. The precipitate
of dicyclohexylurea was filtered off and the filtrate was
added in cold stirring water and extracted with ethyl acetate
x 3. The combined organic layers were dried over Na2SO4
and evaporated under reduced pressure. Each crude product
was purified by flash chromatography on a short pad of silica
gel.
1H-indazole-3-carboxylic acid 18 (0.30 g, 1.85 mmol), 5-
chloro-2,4-dimethoxyaniline 23 (0.42 g, 2.22 mmol) were
allowed to react according to the described general
procedure and the crude product was washed with methanol,
1
and filtered. 4: 0.35 g (yield 43 %); white solid. H-NMR
(400 MHz, (CD3)2CO) δ 3.97 (s, 3H), 4.10 (s, 3H), 7.00(s,
1H), 7.36-7.40 (m, 1H), 7.50-7.55 (m, 1H), 7.74 (tt, J= 0.8,
8.4 Hz, 1H), 8.40-8.42 (m, 1H), 8.73 (s, 1H), 9.43 (br, 1H)
ppm. 13C-NMR (100 MHz, (CD3)2CO) δ 56.9, 57.0, 98.5,
111.6, 113.4, 121.0, 121.1, 122.5, 122.8, 122.8, 123.6, 127.9,
149.0, 149.1, 152.1, 160.8 ppm.
General Procedure to Obtain Hydroxy Amides 6, 7, 13-16
N-(5-Tert-Butyl-2-Methoxyphenyl)-2-(Phenylthio)Acetamide 8
In distinct round bottom flasks, to a stirring solution of
amides (1 equiv) 1, 4, 8-12 in dry dichloromethane, BBr3 1
M was added dropwise (1 equiv for each methoxy group)
under an inert atmosphere at -72 °C. The resulting mixtures
were stirred at room temperature for 20 h. Afterwards, the
reaction was quenched with water, and the aqueous phase
extracted with dichloromethane x 2. The combined organic
layers were dried over Na2SO4 and evaporated under reduced
pressure. Each crude product was purified by flash
chromatography on a short pad of silica gel.
2-(Phenylthio)acetic acid 17 (0.15 g, 0.89 mmol), 5-tert-
butyl-2-methoxyaniline 24 (0.19 g, 1.07 mmol) were allowed
to react according to the described general procedure and the
crude product was placed on a short pad of silica gel and
eluted with petroleum ether and ethyl acetate 9:1. 8: 0.13 g
1
(yield 45 %); white powder; H-NMR (400 MHz, CDCl3) δ
1.28 (s, 9H), 3.76 (s, 2H), 3.79 (s, 3H), 6.77 (d, J= 8 Hz,
1H), 7.03 (m, 1H), 7.21 (s, 1H), 7.20-7.30 (m 2H), 7.39 (dd,
J= 8.4, 1.2 Hz, 2H), 8.41 (d, J= 2 Hz, 1H), 9.19 (br, 1H)
ppm. 13C-NMR (100 MHz, CDCl3) δ 31.4, 34.4, 39.2, 55.8,
109.6, 117.2, 120.6, 126.7, 127.0, 129.2, 129.3, 134.4, 144.0,
146.2, 165.7 ppm.

274 Combinatorial Chemistry & High Throughput Screening, 2015, Vol. 18, No. 3
Giacomini et al.
Table 3. Percentage of activation on hERG current amplitude of synthesized compounds.
Compounds
Id/Ic%
Compounds
Id/Ic%
140%a
134% c
4
1
9
167%a
195%a
160% c
109% c
10
11
3
2
148%a
123%a
67%b
127% c
125% c
116% c
111% c
119% c
12
13
5
6
14
38%a
7
15
16
116%c
8
aTested at 30µM; btested at 100µM; ctested at 20µM.
N-(5-Tert-Butyl-2-Methoxyphenyl)-1H-Indazole-3-Carboxamide 9
2-Bromo-N-(5-Tert-Butyl-2-Methoxyphenyl)Benzamide 10
1H-indazole-3-carboxylic acid 18 (0.20 g, 1.23 mmol), 5-
tert-butyl-2-methoxyaniline 24 (0.27 g, 1.48 mmol) were
allowed to react according to the described general
procedure and the crude product was placed on a short pad of
silica gel and eluted with petroleum ether and ethyl acetate
2-Bromobenzoic acid 19 (0.15 g, 0.75 mmol), 5-tert-
butyl-2-methoxyaniline 24 (0.16 g, 0.89 mmol) were allowed
to react according to the described general procedure and the
crude product was placed on a short pad of silica gel and
eluted with petroleum ether and ethyl acetate 9:1. 10: 0.14 g
1
1
9:1. 9: 0.30 g (yield 74 %); white powder; H-NMR (400
(yield 45 %); red solid; H-NMR (400 MHz, CDCl3) δ 1.35
MHz, CDCl3) δ 1.35 (s, 9H), 3.92 (s, 3H), 6.85 (d, J= 8.4
Hz, 1H), 7.08 (m, 1H), 7.29-7.33 (m, 1H), 7.41-7.45 (m,
1H), 7.51-7.53 (m, 1H), 8.48-8.51 (m, 1H), 9.53 (br, 1H),
10.36 (br, 1H) ppm. 13C-NMR (100 MHz, CDCl3) δ 31.7,
34.6, 56.1, 109.6, 109.9, 117.4, 120.2, 122.2, 122.9, 123.1,
127.4, 127.6, 140.3, 141.6, 144.3, 146.3, 160.5 ppm.
(s, 9H), 3.84 (s, 3H), 6.83 (d, J= 8.8 Hz, 1H), 7.10 (dd, J=
2.4, 8.4 Hz, 1H), 7.31 (dd, J= 8, 2 Hz, 1H), 7.37-7.41 (m,
1H), 7.62-7.66 (m, 2H), 8.29 (br, 1H), 8.64 (d, J= 2.4 1H)
ppm. 13C-NMR (100 MHz, CDCl3)δ 31.5, 34.5, 55.8, 109.6,
117.4, 119.4, 120.7, 127.0, 127.6, 129.8, 131.4, 133.6, 138.3,
144.2, 146.0, 165.2 ppm.

hERG Channel Activators
Combinatorial Chemistry & High Throughput Screening, 2015, Vol. 18, No. 3 275
Scheme 1. Reagents and conditions: a) DCC, OHBt, DMF, 12h, room temp.; BBr3 1M, CH2Cl2, 24h, -72° to room temp.
N-(5-Tert-Butyl-2-Methoxyphenyl)Benzothiophene-2-Carboxamide
11
described above. The crude product was purified on silica
eluting with dichloromethane and methanol 9.9:0.1.6: 0.06 g
(yield 66 %); white powder; 1H-NMR (400 MHz, (CD3)2CO)
δ 3.91 (s, 2H), 6.41-6.44 (m, 1H), 6.70 (d, J= 8.8 Hz, 1H),
7.19-7.24 (m, 1H), 7.29-7.34 (m, 2H), 7.44-7.47 (m, 2H),
7.52 (d, J= 2.8 Hz, 1H), 7.78 (br, 1H), 8.33 (br, 1H), 9.29
(br, 1H) ppm. 13C-NMR (100 MHz, (CD3)2CO) δ 38.9,
108.1, 111.5, 116.4, 127.3, 127.5, 129.8, 130, 135.7, 140.2,
150.9, 167.6 ppm. IR νmax (Nujol) cm-1 3321, 3117, 1649,
1604, 1555, 1460, 1375, 1258, 735, 689. MS (ES): m/z 298
(M + Na+).
Benzothiophene-2-carboxylic acid 20 (0.20 g, 1.12
mmol), 5-tert-butyl-2-methoxyaniline 24 (0.24 g, 1.35
mmol) were allowed to react according to the described
general procedure and the crude product was placed on a
short pad of silica gel and eluted with petroleum ether and
ethyl acetate 9.9:0.1, then washed with petroleum ether. 11:
1
0.25 g (yield 66 %); white powder; H-NMR (400 MHz,
CDCl3) δ 1.34 (s, 9H), 3.93 (s, 3H), 6.85 (d, J= 8.8 Hz, 1H),
7.08-7.11 (m, 1H), 7.40-7.44 (m, 2H), 7.84 (1H), 7.86-7.88
(m, 2H), 8.54 (1H), 8.61 (d, J=2 1H) ppm. 13C-NMR (100
MHz, CDCl3) δ 31.5, 34.5, 55.9, 109.4, 117.4, 120.6, 122.7,
124.9, 125.0, 125.1, 126.4, 126.9, 139.2, 139.6, 141.2, 144.3,
145.8, 160 ppm.
N-(5-Chloro-2,4-Dihydroxyphenyl)-3a,7a-Dihydro-1H-Indazole-
3-Carboxamide 7
To a stirring solution of amide (0.10 g, 0.30 mmol) 4 in 5
mL of dichloromethane dry was added dropwise BBr3 1M
(0.60 mL, 0.60 mmol, 2 equiv) according to the general
procedure described above. The crude product was purified
on silica eluting with dichloromethane and methanol 9.9:0.1.
N-(5-Tert-Butyl-2-Methoxyphenyl)-5-Chlorothiophene-2-Carbox-
amide 12
1
5-Chlorothiophene-2-carboxylic acid 21 (0.20 g, 1.23
mmol), 5-tert-butyl-2-methoxyaniline 24 (0.26 g, 1.48
mmol) were allowed to react according to the described
general procedure and the crude product was placed on a
short pad of silica gel and eluted with petroleum ether and
ethyl acetate 9.9:0.1, then washed with petroleum ether. 12:
7: 0.08 g (yield 66 %); white powder; H-NMR (400 MHz,
(CD3)2CO) δ 6.70 (s, 1H), 7.31-7.32 (m, 1H), 7.45-7.49 (m,
1H), 7.67-7.70 (m, 1H), 8.32 (d, J= 2.8, 1H), 8.34-8.37 (m,
1H), 9.51 (br, 1H) ppm. 13C-NMR (100 MHz, (CD3)2CO) δ
104.9, 110.9, 111.6, 121.1, 121.7, 122.7, 122.7, 123.4, 127.8,
139.4, 142.8, 147.6, 150.3, 161.3 ppm. IR νmax (Nujol) cm-1
3469, 3364, 3222, 1651, 1546, 1514, 1462, 1376. MS (ES):
m/z 326 (M + Na+).
1
0.23 g (yield 58 %); white powder; H-NMR (400 MHz,
CDCl3) δ 1.31 (s, 9H), 3.89 (s, 3H), 6.82 (d, J= 8 Hz, 1H),
6.93 (d, J= 4.4 Hz, 1H), 7.07 (dd, J= 8.4, 2.4 Hz, 1H), 7.36
(d, J= 4.4 Hz, 1H), 8.28 (br, 1H), 8.51 (d, J= 2.4 Hz, 1H)
ppm. 13C-NMR (100 MHz, CDCl3) δ 31.5, 34.4, 55.9, 109.4,
117.4, 120.5, 126.71, 127.1, 135.8, 138.6, 144.3, 145.7,
158.6 ppm.
N-(5-Tert-Butyl-2-Hydroxyphenyl)-2-(Phenylthio)Acetamide 13
To a stirring solution of amide (0.17 g, 0.53 mmol) 8 in 4
mL of dry dichloromethane BBr3 1M (0.53 mL, 0.53 mmol,
1 equiv) was added dropwise according to the general
procedure described above. The crude product was placed on
a short pad of silica gel and eluted with petroleum ether and
N-(2,5-Dihydroxyphenyl)-2-(Phenylthio)Acetamide 6
1
To a stirring solution of amide (0.10 g, 0.33 mmol) 1 in
dichloromethane dry was added dropwise BBr3 1M (0.66
mL, 0.66 mmol, 2 equiv) according to the general procedure
ethyl acetate 9:1.13: 0.15 g (yield 89 %); white powder; H-
NMR (400 MHz, CDCl3) δ 1.23 (s, 9H), 3.81 (s, 2H), 6.81
(d, J= 2.4 Hz, 1H), 6.92 (d, J= 8.4 Hz, 1H), 7.13 (dd, J= 8.4

276 Combinatorial Chemistry & High Throughput Screening, 2015, Vol. 18, No. 3
Giacomini et al.
2.4 Hz, 1H), 7.26-7.28 (m, 1H), 7.31-7.38 (m, 4H), 8.22 (br,
1H), 8.72 (br, 1H) ppm. 13C-NMR (100 MHz, CDCl3) δ 31.3,
33.9, 38.1, 118.9, 119.4, 124.1, 124.7, 127.5, 128.9, 129.6,
133.4, 143.7, 146.4, 168.1 ppm. IR νmax (Nujol) cm-1 3327,
2730, 1645, 1592, 1548, 1509, 1462, 1377, 1310, 1286,
1247, 1200, 1129, 822, 745, 721.
horse serum, penicillin/streptomycin) for 2 - 4 days before
experiments. During voltage clamp experiments, the oocyte
was superfused with a low [Cl-] ND96 solution (to minimize
interference from endogenous Cl currents) at room
temperature with whole oocyte currents recorded using the
double-microelectrode recording configuration and OC-
725C (Warner Instruments, MA) or GeneClamp 500
(Molecular Devices) amplifier. Voltage clamp protocol
generation and data acquisition were controlled by pClamp
10 via DigiData 1440A (Molecular Devices). Current data
were low-pass filtered at 1 kHz (Frequency Devices, MA)
and stored on disks for off-line analysis. The following
software was used for data analysis: pClamp 10, EXCEL
(Microsoft), SigmaPlot, SigmaStat, and PeakFit (SPSS).
N-(5-Tert-Butyl-2-Hydroxyphenyl)-1H-Indazole-3-Carboxamide
14
To a stirring solution of amide (0.25 g, 0.77 mmol) 9 in 5
mL of dry dichloromethane BBr3 1M (0.77 mL, 0.77 mmol,
1 equiv) was added dropwise according to the general
procedure described above, and the crude product was placed
on a short pad of silica gel and eluted with petroleum ether
and ethyl acetate 8:2. 14: 0.17 g (yield 71 %); white powder;
1H-NMR (400 MHz, CD3OD) δ 1.32 (s, 9H), 6.81 (d, J= 8.8
Hz, 1H), 6.99-7.01 (m, 1H), 7.27-7.30 (m, 1H), 7.43 (m,
1H), 7.59 (d, J= 8.4 Hz, 1H), 8.28-8.30 (m, 1H), 8.34 (d, J=
2.4 Hz, 1H) ppm. 13C-NMR (100 MHz, CD3OD) δ 32.0,
35.2, 110.6, 115.5, 118.8, 122.2, 122.7, 123.1, 123.7, 127.0,
128.0, 139.7, 143.2, 143.8, 145.8, 162.9 ppm.
RESULTS AND DISCUSSION
The Enamine database, consisting initially of more than
106 compounds, was virtually screened in search for
potential hERG activators, leading to a library of 158
molecules. Since the crystal structure of the hERG channel is
at present not available, a structure-based VS could not have
been safely performed for two entwined reasons. First, even
though several homology models of the channel have been
reported over the years [23, 43, 44], the percentage of
identity between hERG and the currently available templates
is in general too low to allow a reasonable chance of success
in a VS campaign. Indeed, it is well known that the
performance of structure-based VS is strongly dependent on
the quality of the geometry of the target, and in this case the
picture is worsened by the fact that some activators have
been hypothesized to bind to the so-called “turret” of the
channel. This portion is a distinctive feature of hERG [44],
and in order to model it, one would face the challenging
issue of building a 40 aminoacids segment from scratch,
leading to the possibility to obtain an arguably reliable
geometry. The second motivation that prompted us to
employ a ligand-based VS, was the current uncertainty
concerning the location of the NS1643 binding site. While
some of us reported the pore entrance at the outer mouth of
the selectivity filter as a possible binding site [7], the S5-S6
segment of two adjacent subunits of the channel in proximity
of the pore helix has also been advanced to be implicated in
binding [21]. Recently, through an ingenious topographic
mapping of a homology modeling derived structure, Durdagi
and coworkers have shown that the NS1643 binding to the
channel appears to be extremely complex, and that the
presence of multiple binding sites cannot be excluded [23].
In light of these considerations, we decided to embark on a
LBVS based on the 3D chemical features of NS1643, which
was one of the most potent and effective activators at the
time this study was started. Two conformations of NS1643
were therefore considered during the screening, the syn and
the anti conformations.
N-(5-Tert-Butyl-2-Hydroxyphenyl)Benzo[b]Thiophene-2-Carbo-
xamide 15
To a stirring solution of amide (0.30 g, 0.88 mmol) 11 in
5 mL of dry dichloromethane BBr3 1M (0.88 mL, 0.88
mmol, 1 equiv) was added dropwise according to the general
procedure described above. The crude product was placed on
a short pad of silica gel and eluted with petroleum ether and
ethyl acetate 9:1. 15: 0.13 g (yield 45 %); white powder; 1H-
NMR (400 MHz, CDCl3) δ 1.30 (s, 9H), 6.99 (d, J= 8.4 Hz,
1H), 7.21 (ddd, J= 7.2, 2.4 Hz, 2H), 7.42-7.50 (m, 2H), 7.88-
7.89 (m, 1H), 7.90 (d, J=2.4 Hz, 1H), 7.97 (d, J= 8 Hz, 1H),
8.03 (s, 1H), 8.05 (s, 1H) ppm. 13C-NMR (100 MHz, CDCl3)
δ 31.6, 34.3, 119.4, 119.5, 122.9, 124.6, 124.7, 125.4, 125.5,
127.1, 127.2, 136.9, 139.1, 141.5, 144.2, 146. 4, 161.8 ppm.
N-(5-Tert-Butyl-2-Hydroxyphenyl)-5-Chlorothiophene-2-Carbo-
xamide 16
To a stirring solution of amide (0.30 g, 0.93 mmol) 12 in
5 mL of dry dichloromethane BBr3 1M (0.93 mL, 0.93
mmol, 1 equiv) was added dropwise according to the general
procedure described above. The crude product was placed on
a short pad of silica gel and eluted with petroleum ether and
ethyl acetate 9.5:0.5. 16: 0.23 g (yield 80 %); white powder;
1H-NMR (400 MHz, CDCl3) δ 1.23 (s, 9H), 6.95-6.97 (m,
2H), 7.15-7.17 (m, 1H), 7.20 (d, J=2.4 Hz, 1H), 7.45 (d, J= 4
Hz, 1H), 7.84 (br, 2H) ppm. 13C-NMR (100 MHz, CDCl3) δ
31.4, 34.1, 119.0, 119.1, 124.3, 124.4, 127.3, 128.5, 136.0,
137.1, 144.0, 145.9, 160.0 ppm. IR νmax (Nujol) cm-1 3378,
1633, 1592, 1566, 1549, 1462, 1433, 1376.
Electrophysiology
The 158 molecules library generated through the LBVS
procedure, was carefully examined by visual inspection, and,
since the repeating feature was the presence of two diversely
decorated aromatic rings separated by recurring spacers, we
decided to cluster it according to the functional groups of the
linker itself. In Table 2, the 9 obtained clusters are reported.
Molecules belonging to Scaffold 1 were discarded because
of their high similarity with known hERG activators.
Oocyte Expression, Two-Electrode Voltage Clamp and
Data Analysis
Stage V oocytes were selected for expression. Each
oocyte was injected with 40 nl solution containing 10 ng of
hERG cRNA. After cRNA injection, oocytes were incubated
in an ND96-based medium (ND96 supplemented with 4%

hERG Channel Activators
Combinatorial Chemistry & High Throughput Screening, 2015, Vol. 18, No. 3 277
Similarly, we decided not to take into consideration Scaffold
2 as the electrostatic potential map of the corresponding
molecules did not satisfactorily match the one displayed by
NS1643 in both the syn or anti conformations. Among the
matters the most. Among the tested compounds, 3 showed
the best result, increasing the current amplitude of 195%.
Notably, as shown in Fig. (2), 3 is effective in increasing
outward currents through the hERG/IKr channel during the
plateau phase of cardiac action potential (Fig. 2C), and this
is likely due to a decrease in the degree of hERG/IKr
inactivation (Fig. 2E). However, 3 does not affect the
voltage-dependence of hERG/IKr activation (Fig. 2D) or the
pore conductance of the hERG/IKr channel (Ifully-activated at -
120 mV). This behavior suggested that its mechanism could
be similar to that of NS1643.
remaining clusters, only scaffolds 4-6, and
considered, as they could allow SAR studies to be conducted
taking advantage of combinatorial approach. The
8
were
a
molecules to be tested in vitro were therefore selected on the
basis of their LBVS ranking depending on their commercial
availability, and are shown in Table 3. Specifically, a
phenylthio acetamide derivative (1) was selected from
Scaffold 4, a phenoxy acetamide (2) from Scaffold 5;
Scaffold 6 represented the most populated cluster, therefore
amides 3 and 4, bearing diverse aromatic groups, were
chosen from this ensemble. Finally, an acetamide derivative
(5) was selected from Scaffold 8.
Unfortunately, the poor solubility of the other selected
compounds made their evaluation quite difficult, especially
for 2. However, we determined that 1 increased current
amplitude of 167%, 4 had a modest effect (140%), while 5
did not significantly enhance the currents.
Preliminary assays on molecules 1-5 (Table 3) involved
the evaluation of the increment of the current amplitude
measured before and after drugs application (30 μM) on wt
hERG expressed in oocytes. Channel activation was
performed starting from a holding potential of -80 mV, and
depolarizing up to 60 mV with voltage steps of 20 mV. The
currents were measured at the peak of the bell-shaped test
pulse current-voltage relation, most often at 0 mV. We chose
this physiologically relevant parameter because this is almost
the plateau voltage of cardiac action potential, where IKr
From the above reported results, compound 3 appeared to
be an interesting scaffold worth of further investigation,
whereas the need of more soluble compounds came up to be
a serious issue. With the aim to rationalize our findings with
structure-activity relationships, and at the same time to
obtain more soluble analogs, we decided then to pursue two
roads. On one side, we explored the chemical space around 3
by keeping the 2-hydroxy-5-(tert-butyl)-aniline fragment
resembling the 2-hydroxy-5-trifluoromethyl-phenyl of
NS1643, and coupling it with different carboxylic acids
Fig. (2). Illustration of how hERG current amplitude and gating kinetics are quantified and the effects of compound 3. In all 5 panels,
current traces or data points shown in black are control and those shown in red represent steady-state effects of 3 at a concentration of 30 μM.
(A) Current traces elicited by the diagrammed voltage clamp protocol. The current amplitudes measured at the end of the test pulses (Itest-pulse
)
and the peak of the tail currents (Itail) are noted. The values of Itest-pulse are plotted against test pulse voltage (Vt) in (B); this parameter mimics
the hERG/IKr current amplitudes during the cardiac action potential plateau phase. The values of Itail are plotted against Vt in (C); the
relationship between Itail and Vt reflects the voltage-dependence of hERG/IKr activation. (D) Current trances elicited by the diagrammed
voltage clamp protocol. The peak or plateau phase of the tail current is marked as Ifully-activated. The values of Ifully-activated are plotted against
repolarization voltage (Vr) in (E); this I-V relationship is a measure of the degree and voltage-dependence of hERG/IKr inactivation, seen
here as a voltage-dependent decrease in current amplitude > -30 mV. The amplitudes of inward currents seen at hyperpolarized voltages (e.g.
-120 mV) indicate whether the pore conductance is altered. Statistical analysis: Itest-pulse in the presence of compound 3 is significantly higher
than that of control in the Vt range -20 to +50 mV (t-test, p < 0.05).

278 Combinatorial Chemistry & High Throughput Screening, 2015, Vol. 18, No. 3
Giacomini et al.
Fig. (3). HERG current traces recorded under the control conditions (black), during the exposure to the specified molecules and after
washout of the molecules (blue). In all cases, the following voltage clamp protocol was used: holding voltage -80 mV, currents elicited by 1-
s depolarizing pulses to +20 mV once every 60 s, and tail currents were recorded at -60 mV. The current amplitude increased slightly after
washing out of the molecules 1 and 4: this could be due to mixed activator/inhibitor effects, and the latter was reversed more readily than the
former. On the other hand, compounds 6 and 7 were clearly hERG inhibitors, with varying degree of reversibility after washing out of the
molecules.
bearing heterocycles or groups present in other top-ranked
molecules from the LBVS outcome. On the other side,
despite their poor solubility, 1 and 4 showed a moderate
enhancing effect, therefore we investigated the possibility to
improve their activity by synthesizing the dihydroxy-analogs
6 and 7, respectively. On the same line, we decided to study
the role of methoxy vs hydroxy groups also in the SAR on 3.
The synthetic route to obtain the desired compounds allowed
the reactions to be conducted in a parallel fashion. Amides 1,
4, 8-12 were synthesized through a DCC-HOBt mediated
coupling employing carboxylic acids 17-21 and amines 22-
24, then treating with boron tribromide to afford the
hydroxyl derivatives 6, 7, 13-16 (Scheme 1). All the newly
synthesized compounds (6-17) were then tested for their
ability to modulate the hERG activation (Table 3).
scenario has also been supported by the theoretical and
experimental results reported by Durdagi et al., which
rationalized the dual pharmacological behavior of NS1643
by the presence of multiple spatially separated binding sites
mediating different modulating effects [23]. Interestingly,
going from 7 to 9, thus maintaining the indazole fragment
but coupling it with the 2-methoxy-5-(tert-butyl)-aniline,
once again
a reversed pharmacological activity was
observed, since 9 showed a slight enhancing effect on
current amplitude (134%, see Table 3 and Fig. 4), while a
drop of activity was observed with its hydroxy analog 14,
confirming the complexity of the protein-ligand interaction
in hERG.
Among the methoxy amide 8-12 derivatives, also
compound 10 (Table 3, Fig. 4), which is the methoxy analog
of 3, showed a modest activating effect (160%) on hERG,
suggesting that the 2 Br-phenyl portion could play an
important role. However, compared to parent compound 3, a
loss of activity was detected. Concerning the hydroxyl
derivatives 13-16, surprisingly none was able to show an
enhancing activity.
Even though the di-hydroxy amides 6 and 7 were found
to be more soluble than the parent compounds,
a
controversial result was obtained in the electrophysiology
assays shown in Fig. (3). Indeed, rather than acting as hERG
activators, they turned out to be currents suppressors. Albeit
disappointing, this finding was not completely unexpected,
as even the reference compound NS1643 shows a blocking
behavior at high concentrations [20]. Since the blockers
binding site is known to be highly unspecific in terms of
molecular determinants, it is possible that many hERG
enhancers can bind to multiple sites, and that the net effect
on the channel kinetics is the result of a subtle concentration-
depending balance of the affinities to each site. Such a
Taken together, these results revealed the hydroxyl
function in the ortho position not to be an essential
requirement to obtain an effective modulation of the hERG
channel. Moreover, the amide linkage proved to be a valid
and synthetically accessible bioisosteric replacement of the
urea in the search of novel hERG activators.

hERG Channel Activators
Combinatorial Chemistry & High Throughput Screening, 2015, Vol. 18, No. 3 279
Fig. (4). HERG current traces recorded under the control conditions (black), during the exposure to the specified molecules and
concentrations (red) and after washout of the molecules (blue). The voltage clamp protocol was the same as described for Fig. (3).
[5]
[6]
[7]
Zhou, P.; Babcock, J.; Liu, L.; Li, M.; Gao, Z. Activation of
Human Ether-a-Go-Go Related Gene (hERG) Potassium Channels
by Small Molecules. Acta Pharmacol. Sin., 2011, 32, 781-788.
He, F.-Z.; McLeod, H.L.; Zhang, W. Current Pharmacogenomic
Studies on hERG Potassium Channels. Trends Mol. Med., 2013,
19, 227-238.
Xu, X.; Recanatini, M.; Roberti, M.; Tseng, G. Probing the Binding
Sites and Mechanisms of Action of Two Human Ether-a-Go-Go-
Related Gene Channel Activators, 1,3-Bis-(2-Hydroxy-5-
Trifluoromethyl-Phenyl)-Urea (NS1643) and 2-[2-(3,4-Dichloro-
CONCLUSION
The hERG potassium channel has long been known as an
antitarget in drug discovery, because its undesired block
promotes the prolongation of the plateau phase of the action
potential in myocytes leading to potentially serious cardiac
conditions. Since the discovery of molecules able to
modulate positively the IKr currents, however, the paradigm
of the hERG pharmacology has changed. Indeed, the binding
of small molecules to the channel is no longer necessarily
perceived as an unwanted event, as drugs able to activate the
channel could in principle be exploited for their potential
therapeutic effects. Unfortunately, the advancements in this
field of drug discovery is strongly slowed down by
difficulties due both to the lack of the hERG crystal structure
and to the uncertainties related to the location of the binding
site of activators. Rather than pursuing a problematic
structure-based strategy, in this study, we took advantage of
a ligand-based approach to identify new compounds as
hERG activators by means of a virtual screening campaign.
One of the selected hits among the final library proved to
enhance hERG current of 195%, confirming that our
protocol was successful in providing a new and accessible
scaffold worth to be explored by combinatorial synthetic
strategies.
Phenyl)-2,3-Dihydro-1H-Isoindol-5-Ylamino]-Nicotinic
Acid
(PD3). Mol. Pharmacol., 2008, 73, 1709-1721.
[8]
Di Martino, G.P.; Masetti, M.; Ceccarini, L.; Cavalli, A.;
Recanatini, M. An automated docking protocol for herg channel
blockers. J. Chem. Inf. Model., 2013, 53, 159-175.
[9]
Perrin, M.J.; Subbiah, R.N.; Vandenberg, J.I.; Hill, A.P. Human
Ether-a-Go-Go Related Gene (hERG) K+ channels: function and
dysfunction. Prog. Biophys. Mol. Biol., 2010, 98, 137-148.
Trudeau, M.; Warmke, J.; Ganetzky, B.; Robertson, G. HERG, a
human inward rectifier in the voltage-gated potassium channel
family. Science., 1995, 269, 92-95.
Babcock, J.J.; Li, M. hERG Channel Function: Beyond Long QT.
Acta Pharmacol. Sin., 2013, 34, 329-335.
Recanatini, M.; Poluzzi, E.; Masetti, M.; Cavalli, A.; De Ponti, F.
QT Prolongation through hERG K(+) channel blockade: current
knowledge and strategies for the early prediction during drug
development. Med. Res. Rev., 2005, 25, 133-166.
Kim, P.Y.; Ewer, M.S. Chemotherapy and QT prolongation:
overview with clinical perspective. Curr. Treat. Options
Cardiovasc. Med., 2014, 16, 303.
Sanguinetti, M.C.; Tristani-Firouzi, M. hERG potassium channels
and cardiac arrhythmia. Nature, 2006, 440, 463-469.
Cavalli, A.; Buonfiglio, R.; Ianni, C.; Masetti, M.; Ceccarini, L.;
Caves, R.; Chang, M.W.Y.; Mitcheson, J.S.; Roberti, M.;
Recanatini, M. Computational Design and Discovery of
“Minimally Structured” hERG Blockers. J. Med. Chem., 2012, 55,
4010-4014.
Wang, S.; Li, Y.; Xu, L.; Li, D.; Hou, T. Recent developments in
computational prediction of hERG blockage. Curr. Top. Med.
Chem., 2013, 13, 1317-1326.
Raschi, E.; Vasina, V.; Poluzzi, E.; De Ponti, F. The hERG K+
channel: target and antitarget strategies in drug development.
Pharmacol. Res., 2008, 57, 181-195.
Zhang, H.; Zou, B.; Yu, H.; Moretti, A.; Wang, X.; Yan, W.;
Babcock, J.J.; Bellin, M.; McManus, O.B.; Tomaselli, G.; Nan, F.;
Laugwitz, K.-L.; Li, M. Modulation of hERG potassium channel
gating normalizes action potential duration prolonged by
dysfunctional KCNQ1 potassium channel. Proc. Natl. Acad. Sci.
USA, 2012, 109, 11866-11871.
Hansen, R.S.; Diness, T.G.; Christ, T.; Demnitz, J.; Ravens, U.;
Olesen, S.; Grunnet, M. Activation of Human Ether-a-Go-Go-
Related Gene Potassium Channels by the Diphenylurea 1,3-Bis-(2-
[10]
[11]
[12]
[13]
[14]
[15]
CONFLICT OF INTEREST
The authors confirm that this article content has no
conflict of interest.
[16]
[17]
[18]
ACKNOWLEDGEMENTS
The authors thank Andrea Sartini for technical assistance,
and also acknowledge funding from Istituto Italiano di
Tecnologia, Genova, Italy (Seed project: HERGSYM).
REFERENCES
[1]
Bagal, S.K.; Brown, A.D.; Cox, P.J.; Omoto, K.; Owen, R.M.;
Pryde, D.C.; Sidders, B.; Skerratt, S.E.; Stevens, E.B.; Storer, R.I.;
Swain, N. a. Ion channels as therapeutic targets: a drug discovery
perspective. J. Med. Chem., 2013, 56, 593-624.
[19]
[20]
[2]
[3]
[4]
Gamper, N. Osmosensitivity through the PIP2 Availability: Just
Add Water. J. Physiol., 2010, 588, 3631-3632.
Hydroxy-5-Trifluoromethyl-Phenyl)-Urea
(NS1643).
Mol.
Pharmacol., 2006, 69, 266-277.
Hubner, C.A. Ion Channel Diseases. Hum. Mol. Genet., 2002, 11,
2435-2445.
Perry, M.; Sanguinetti, M.; Mitcheson, J. Revealing the structural
basis of action of hERG potassium channel activators and blockers.
J. Physiol., 2010, 588, 3157-3167.
Casis, O.; Olesen, S.; Sanguinetti, M.C. Mechanism of Action of a
novel human ether-a-go-go-related gene channel activator. Mol.
Pharmacol., 2006, 69, 658-665.

280 Combinatorial Chemistry & High Throughput Screening, 2015, Vol. 18, No. 3
Giacomini et al.
[21]
Szabó, G.; Farkas, V.; Grunnet, M.; Mohácsi, a; Nánási, P.P.
Enhanced repolarization capacity: new potential antiarrhythmic
strategy based on HERG channel activation. Curr. Med. Chem.,
2011, 18, 3607-3621.
Hansen, R.S.; Diness, T.G.; Christ, T.; Wettwer, E.; Ravens, U.;
Olesen, S.; Grunnet, M. Biophysical characterization of the new
human ether-a-go-go-related gene channel opener NS3623 [N-(4-
Bromo-2-(1H-Tetrazol-5-Yl)-Phenyl)-N’-(3'-
Trifluoromethylphenyl)urea]. Mol. Pharmacol., 2006, 70, 1319-
1329.
Durdagi, S.; Guo, J.; Lees-Miller, J.P.; Noskov, S.Y.; Duff, H.J.
Structure-guided topographic mapping and mutagenesis to
elucidate binding sites for the human ether-a-go-go-related gene 1
potassium channel (KCNH2) activator NS1643. J. Pharmacol. Exp.
Ther., 2012, 342, 441-452.
http://www.enamine.net/.
OEChem, version 1.7.7, OpenEye Scientific Software, Inc., Santa
Fe, NM, U. www.eyesopen.com.
[31]
Kirchmair, J.; Wolber, G.; Laggner, C.; Langer, T. Comparative
performance assessment of the conformational model generators
omega and catalyst: a large-scale survey on the retrieval of protein-
bound ligand conformations. J. Chem. Inf. Model., 2006, 46, 1848-
1861.
Quevedo, M.A.; Ribone, S.R.; Briñón, M.C.; Dehaen, W.
Development of a receptor model for efficient in silico screening of
HIV-1 integrase inhibitors. J. Mol. Graph. Model., 2014, 52, 82-90.
Avram, S.I.; Pacureanu, L.M.; Bora, A.; Crisan, L.; Avram, S.;
[22]
[32]
[33]
Kurunczi, L. ColBioS-FlavRC:
A Collection of Bioselective
[23]
Flavonoids and Related Compounds Filtered from High-
Throughput Screening Outcomes. J. Chem. Inf. Model., 2014, 54,
2360-2370.
[34]
Hamza, A.; Wei, N.-N.; Zhan, C.-G. Ligand-based virtual
screening approach using a new scoring function. J. Chem. Inf.
Model., 2012, 52, 963-974.
[24]
[25]
[35]
[36]
ROCS 3.2.0.4: OpenEye Scientific Software, Santa Fe, N.
http://www.eyesopen.com.
[26]
Kirchmair, J.; Distinto, S.; Markt, P.; Schuster, D.; Spitzer, G.M.;
Liedl, K.R.; Wolber, G. How to optimize shape-based virtual
screening: choosing the right query and including chemical
information. J. Chem. Inf. Model., 2009, 49, 678-692.
Hawkins, P.C.D.; Skillman, a G.; Nicholls, A. Comparison of
shape-matching and docking as virtual screening tools. J. Med.
Chem., 2007, 50, 74-82.
[37]
[38]
EON 2.2.0.5: OpenEye Scientific Software, Santa Fe, N.
http://www.eyesopen.com.
[27]
Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A. Jr.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi,
J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.;
Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H.
P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts,
R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth,
G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.;
Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D.
K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.;
Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B.
B.; Liu, G.; Liashenko, A.; Piskorz, P. Komaromi, I.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara,
A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.;
Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, Revision
C.02. (2004) Wallingford CT, Gaussian, I.
Halgren, T.A. Merck Molecular Force Field. I. Basis, Form, Scope,
Parameterization, and Performance of MMFF94. J. Comput.
Chem., 1996, 17, 490-519.
[39]
[40]
[41]
Bayly, C.I.; Cieplak, P.; Cornell, W.; Kollman, P.A. A well-
behaved electrostatic potential based method using charge
restraints for deriving atomic charges: the RESP model. J. Phys.
Chem., 1993, 97, 10269-10280.
Cornell, W.D.; Cieplak, P.; Bayly, C.I.; Kollmann, P.A.
Application of RESP charges to calculate conformational energies,
hydrogen bond energies, and free energies of solvation. J. Am.
Chem. Soc., 1993, 115, 9620-9631.
Abagyan, R.; Totrov, M.; Kuznetsov, D. ICM?A New method for
protein modeling and design: applications to docking and structure
prediction from the distorted native conformation. J. Comput.
Chem., 1994, 15, 488-506.
Michener, C.D.; Sokal, R.R. A Quantitative Approach to a Problem
in Classification. Evolution (N. Y)., 1957, 11, 130.
Masetti, M.; Cavalli, A.; Recanatini, M. Modeling the hERG
potassium channel in a phospholipid bilayer: molecular dynamics
and drug docking studies. J. Comput. Chem., 2008, 29, 795-808.
Tseng, G.-N.; Sonawane, K.D.; Korolkova, Y. V; Zhang, M.; Liu,
J.; Grishin, E. V; Guy, H.R. Probing the outer mouth structure of
the HERG channel with peptide toxin footprinting and molecular
modeling. Biophys. J., 2007, 92, 3524-3540.
[42]
[43]
[28]
Irwin, J.J.; Sterling, T.; Mysinger, M.M.; Bolstad, E.S.; Coleman,
R.G. ZINC: a free tool to discover chemistry for biology. J. Chem.
Inf. Model., 2012, 52, 1757-1768.
[44]
[29]
[30]
FILTER, version 2.02, OpenEye Scientific Santa Fe, NM, USA.
www.eyesopen.com.
Wang, R.; Fu, Y.; Lai, L. A new atom-additive method for
calculating partition coefficients. J. Chem. Inf. Model., 1997, 37,
615-621.
Received: July 23, 2014
Revised: October 8, 2014
Accepted: November 10, 2014