Characterization of a novel high-potency positive modulator of K v7 channels

Article abstract of DOI:10.1016/j.ejphar.2013.03.039

K_v7 channel activators decrease neuronal excitability and might potentially treat neuronal hyperexcitability disorders like epilepsy and mania. Here we introduce NS15370 ((2-(3,5-difluorophenyl)-N-[6-[(4-fluorophenyl) methylamino]-2-morpholino-3-pyridyl]acetamide)hydrochloride, an in vitro high-potency chemical analogue of retigabine, without effects on GABA _A receptors. NS15370 activates recombinant homo- and heteromeric K_v7.2-K_v7.5 channels in HEK293 cells at sub-micromolar concentrations (EC₅₀～100 nM, as quantified by a fluorescence based Tl⁺-influx assay). In voltage clamp experiments NS15370 exhibits a complex, concentration-dependent mode-of-action: At low concentrations it accelerates voltage-dependent activation rates, slows deactivations, and increases steady-state current amplitudes. Quantified by the peak-tail current method, the V₁₂ value of the steady-state activation curve is shifted towards hyperpolarized potentials at concentrations ～100 times lower than retigabine. However, in contrast to retigabine, NS15370 also introduces a distinct time-dependent current decrease, which eventually, at higher concentrations, causes suppression of the current at depolarized potentials, and an apparent cross-over of the voltage-activation curve. In brain slices, NS15370 hyperpolarizes and increases spike frequency adaptation of hippocampal CA1 neurons and the compound reduces the autonomous firing of dopaminergic neurons in the substantia-nigra pars compacta. NS15370 is effective in rodent models of hyperexcitability: (i) it yields full protection against mouse 6 Hz seizures and rat amygdala kindling discharges, two models of partial epilepsia; (ii) it reduces (+)-MK-801 hydrogen maleate (MK-801)-induced hyperactivity as well as chlordiazepoxide (CDP)+d-amphetamine (AMP)-induced hyperactivity, models sensitive to classic anti-psychotic and anti-manic treatments, respectively. Our findings with NS15370 consolidate neuronal K_v7 channels as targets for anti-epileptic and psychiatric drug development.

Full text of DOI:10.1016/j.ejphar.2013.03.039

European Journal of Pharmacology 709 (2013) 52–63
Contents lists available at SciVerse ScienceDirect
European Journal of Pharmacology
journal homepage: www.elsevier.com/locate/ejphar
Neuropharmacology and analgesia
Characterization of a novel high-potency positive modulator of
Kv7 channels
William Dalby-Brown, Carsten Jessen, Charlotte Hougaard, Marianne L. Jensen, Thomas
A. Jacobsen, Karin S. Nielsen, Helle K. Erichsen, Morten Grunnet, Philip K. Ahring,
n
Palle Christophersen , Dorte Strøbæk, Susanne Jørgensen
NeuroSearch A/S, Pederstrupvej 93, DK-2750 Ballerup, Denmark
a r t i c l e i n f o
a b s t r a c t
Article history:
Kv7 channel activators decrease neuronal excitability and might potentially treat neuronal hyperexcit-
ability disorders like epilepsy and mania. Here we introduce NS15370 ((2-(3,5-difluorophenyl)-N-[6-[(4-
fluorophenyl)methylamino]-2-morpholino-3-pyridyl]acetamide)hydrochloride, an in vitro high-potency
chemical analogue of retigabine, without effects on GABAA receptors. NS15370 activates recombinant
homo- and heteromeric Kv7.2–Kv7.5 channels in HEK293 cells at sub-micromolar concentrations
(EC50∼100 nM, as quantified by a fluorescence based Tl+-influx assay). In voltage clamp experiments
Received 15 January 2013
Received in revised form
15 March 2013
Accepted 24 March 2013
Available online 3 April 2013
Keywords:
Retigabine
NS15370
Neuronal Kv7 channel
Epilepsy
NS15370 exhibits
a complex, concentration-dependent mode-of-action: At low concentrations it
accelerates voltage-dependent activation rates, slows deactivations, and increases steady-state current
amplitudes. Quantified by the peak-tail current method, the V value of the steady-state activation curve
½
is shifted towards hyperpolarized potentials at concentrations ∼100 times lower than retigabine.
However, in contrast to retigabine, NS15370 also introduces a distinct time-dependent current decrease,
which eventually, at higher concentrations, causes suppression of the current at depolarized potentials,
and an apparent “cross-over“ of the voltage-activation curve. In brain slices, NS15370 hyperpolarizes and
increases spike frequency adaptation of hippocampal CA1 neurons and the compound reduces the
autonomous firing of dopaminergic neurons in the substantia-nigra pars compacta. NS15370 is effective
in rodent models of hyperexcitability: (i) it yields full protection against mouse 6 Hz seizures and rat
amygdala kindling discharges, two models of partial epilepsia; (ii) it reduces (+)-MK-801 hydrogen
maleate (MK-801)-induced hyperactivity as well as chlordiazepoxide (CDP)+d-amphetamine (AMP)-
induced hyperactivity, models sensitive to classic anti-psychotic and anti-manic treatments, respectively.
Our findings with NS15370 consolidate neuronal Kv7 channels as targets for anti-epileptic and
psychiatric drug development.
Bipolar disorder
CA1 pyramidal neurons
& 2013 Elsevier B.V. All rights reserved.
1. Introduction
channels remain scientifically interesting due to their role in
excitable tissues, combined with the fact that mutations in four
Kv7 channels constitute a family of five members, Kv7.1–Kv 7.5,
corresponding to the genes KCNQ1–KCNQ5. Kv7.1 is a non-
neuronal channel, whereas Kv7.2–Kv7.5 were previously consid-
ered to be mainly expressed in neuronal tissue, where they serve
pivotal physiological functions in the control of excitability (see
Jentsch, 2000). It has recently become clear that Kv7.4 and Kv7.5
channels are also expressed in smooth muscle cells, where they
play crucial roles in regulating contractility (see Greenwood and
Ohya, 2009). Kv7.2 and Kv7.3 do have a predominant neuronal
expression, and mainly exist as functional heteromers. The Kv7
out of five KCNQ genes have been connected to human diseases
(Jentsch, 2000). Benign familial neonatal convulsions are asso-
ciated with loss-of-function mutations in KCNQ2 and KCNQ3
(Biervert et al., 1998; Charlier et al., 1998), whereas loss-of-
function mutations in KCNQ4 can result in autosomal dominant
deafness (Kubisch et al., 1999). Cardiac arrhythmias have been
associated with both loss- and gain-of-function mutations in
KCNQ1 (Jespersen et al., 2005; Chen et al., 2003).
The first report of Kv7 channel activity was the characterization
of the so-called M-current from symphathetic neurons (Brown and
Adams, 1980) which was later demonstrated to be mediated by
Kv7.2+Kv7.3 heteromers (Wang et al., 1998) although Kv7.4 and
Kv7.5 possess the same biophysical characteristics: These are slow
voltage-dependent activation at sub-threshold potentials, com-
paratively fast deactivation, and little inactivation (Jensen et al.,
n
Corresponding author. Tel.: +45 4460 8222; fax: +45 4460 8080.
E-mail addresses: pc@aniona.com,
palle.christophersen@gmail.com (P. Christophersen).
0014-2999/$ - see front matter & 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ejphar.2013.03.039

W. Dalby-Brown et al. / European Journal of Pharmacology 709 (2013) 52–63
53
2007). An important additional characteristic is inhibition upon
muscarinic receptor activation, hence the name M-channel (Jensen
et al., 2007; Brown and Passmore, 2009). Generally, GPCRs signal-
ling via the Gq/11 pathway may decrease neuronal Kv7 channel
activity by degradation of phosphatidylinositol 4,5-bisphosphate
(PIP2; Suh and Hille, 2002) an essential membrane constituent for
Kv7 activity. Being under combined voltage-dependent and meta-
botropic influence render Kv7 channels important in controlling
intrinsic excitability as well as network activity (Brown and
Passmore, 2009). Activators of Kv7 channels have gained attention
for treatment of hyperexcitability disorders like epilepsy, mania,
chronic pain, as well as migraine, anxiety and addiction to
psychostimulants (Korsgaard et al., 2005; Munro et al., 2007;
Hansen et al., 2007; Dencker et al., 2008). Retigabine, the proto-
type activator, recently entered the market as a new mode-of-
action drug for treatment of complex partial seizures (Fattore, and
Perucca, 2011). Retigabine is a low-potency (micromolar) non-
selective activator of Kv7.2–Kv7.5 channels and a high-dose anti-
epileptic in the clinic. Since retigabine also acts as positive
modulator of GABAA receptors (Otto et al., 2002; Rundfeldt and
Netzer, 2000a), it remains a discussion whether the clinical
efficacy is solely due to activation of Kv7 channels.
We identified and characterized a novel Kv7.2–Kv7.5 activator,
NS15370 ((2-(3,5-difluorophenyl)-N-[6-[(4-fluorophenyl)methyla-
mino]-2-morpholino-3-pyridyl]acetamide)hydrochloride,
WO
2010/060955) a retigabine analogue, which possesses ∼100-fold
higher potency, different mode-of-action, and improved selectivity
towards GABAA receptors. We demonstrate dampening of neuro-
nal activity in brain slices and show efficacy of NS15370 in rodent
models of epilepsy, psychosis and mania. We conclude that
activation of Kv7 channels underlie these effects.
2. Materials and methods
2.1. Chemistry
NS15370 was synthesized at NeuroSearch A/S. For further
details please refer to Supplementary Fig. 1. The compound may
be prepared as the free base or as the hydrochloride salt. For the
Fig. 1. Comparison of NS15370 and Retigabine in FLIPR-based Tl+ influx assays. (A) Chemical structure of the Kv7 activator NS15370 (((2-(3,5-difluorophenyl)-N-[6-[(4-
fluorophenyl)methylamino]-2-morpholino-3-pyridyl]acetamide)hydrochloride) in comparison with retigabine. (B) (Left) Fluorescence emission followed over time on a
fluoremetric imaging plate reader (FLIPR) from HEK293 cells stably expressing Kv7.2+7.3 and loaded with BTC-AM. In the first addition, either Cl− free assay buffer (dashed
line) or Cl− free assay buffer containing 0.3 μM NS15370 (full line) is added, as indicated by the arrow. In the second addition, a stimulus buffer containing Tl+ (2 mM) as well
as K2SO4 (5 mM) for depolarization is added, as indicated by the arrow. (Right) Concentration-response curves based on the Tl+-influx assay. The two compounds were tested
in the concentration ranges from 0.1 nM to 10 μM (NS15370, filled circles) or 0.1 nM to 100 μM (retigabine, open circles). The effect at any given concentration was
background corrected and normalized to the positive control response (30 μM retigabine). The data represent at least three independent experiments. (C) Concentration–
response curves with NS15370 and retigabine on concatenated Kv7.3–7.2 wildtypes (filled circles) and concatenated point mutated (Kv7.3−Kv7.2 mt) constructs.

54
W. Dalby-Brown et al. / European Journal of Pharmacology 709 (2013) 52–63
present study, NS15370 was prepared as the hydrochloride salt.
The compound has a Mw of 492.9 Da (Mw of the free base is
456.5 Da), a measured LogP and LogD7.4 of 4.2, a measured pKa of
3.61 and a solubility of 0.18 mg/ml and 0.26 mg/ml in H2O and 15%
methyl-β-cyclodextrin, respectively. For in vitro experiments
NS15370 was dissolved in dimethyl sulfoxide (DMSO) at 10 mM
and, subsequently, diluted in the extracellular solution. For in vivo
experiments NS15370 was handled as described in the individual
methods sections.
day and transferred to peel-off flasks in three different densities
and selection with 150 μg/ml hygromycin was initiated the next
day. Transfected cells were grown in selection media for approxi-
mately 14 days, followed by harvesting and further propagation of
the appearing colonies. The expression of functional hKv7.x chan-
nels was verified using FLIPR and patch clamp measurements for
biophysical and pharmacological characterization. The HEK293
cell lines stably expressing hKv7.2+7.3 or hKv7.4 channels were
generated as described earlier (Schrøder et al., 2001; Søgaard et al.,
2001), respectively.
All tissue culture media were from Life Technologies (Nærum,
Denmark). The cells were grown in Dulbecco's minimum essential
medium (DMEM) supplemented with 10% fetal calf serum (FCS)
and 2 mM Glutamax as well as the relevant selection antibiotics
and maintained at 37 1C in a 5% CO2 atmosphere.
2.2. Biology
2.2.1. Plasmids
cDNA encoding human (h) Kv7.2, previously described in
Schrøder et al. (2001), was subcloned using NheI and XbaI into
the custom made pNS3h vector, containing an IRES element to
ensure translation of hygromycin-resistance and hKv7.2 from the
same transcript. In the same way, cDNA encoding hKv7.5, pre-
viously described in Dupuis et al. (2002), was inserted in pNS3h
using BamHI and EcoRI.
2.2.3. FLIPR-experiments
2.2.3.1. Cell seeding. HEK293 cells stably expressing concatenated
hKv7.3−7.2, hKv7.3−7.2 mt, hKv7.2+7.3, hKv7.2, hKv7.4 or hKv7.5
were seeded on poly- -lysine (10 μg/ml) coated 384-well, clear-
D
cDNA for hKv7.3 (Schrøder et al., 2001) and hKv7.2 were
concatenated as follows: Unique restrictions sites were introduced
in hKv7.2 (BamHI) and hKv7.3 (AgeI) using site-directed mutagen-
esis. Briefly, plasmids were uracilated and used as template in a
mutagenesis reaction, in which mutagenic oligonucleotides and T7
DNA polymerase were used to introduce mutations. Mutated
plasmids were identified by the introduction or elimination of
restriction sites and verified by sequencing. The sequences of the
mutagenic oligonucleotides were: hKv7.2, BamHI: cttcgtg-
gggctggatcccggcgcgcccgacT and hKv7.3 AgeI: TGCCTCTGT-
CGTCCACcGGtGATGGGATTTCTGATTC (Eurofins, MWG Operon,
Ebersberg, Germany). A synthetic cDNA sequence covering the
C-terminal of hKv7.3 downstream of the AgeI site, a linker with the
amino acid sequence GSQ10 and the N-terminal part of hKv7.2
between the ATG and the BamHI site with the following sequence
was purchased from Genscript (Piscataway, NJ, USA): CACcGGtG-
ATGGGATTTCTGATTCAGTATGGACCCCTTCCAATAAGCCCATTggtagt-
caacaacagcaacagcagcaacagcagcaaATGGTGCAGAAGTCGCGCAACGG-
CGGCGTATACCCCGGCCCGAGCGGGGAGAAGAAGCTGAAGGTGGGC-
TTCGTGGGGCTGGAtCCC. The plasmid containing the cDNA for
hKv7.2 with the introduced BamHI site was digested using SpeI
and BamHI and used as vector fragment in a ligation reaction with
the linker fragment digested with AgeI and BamHI and hKv7.3
digested with SpeI and AgeI using T4 DNA ligase. The result-
ing construct was verified by restriction enzyme digestion and
sequencing.
bottomed, black-walled Optiplates (Corning Inc, NY, USA). The cells
were seeded at a density of ꢀ3 ꢁ 106 cells/ml in 20 μl DMEM
containing 10% FCS per well and left overnight at 37 1C in a 5% CO2
incubator.
2.2.3.2. Fluorescence based Tl+ influx assay. Prior to an experiment,
the cells were loaded with the fluorescent dye benzothiazolecou-
marinacetoxymethyl ester (BTC-AM, Life Technologies, Nærum,
Denmark), essentially as described by Weaver et al. (2004). In
brief, the cells were washed thrice in Cl−-free assay buffer (in mM:
140 Na+-gluconate, 2.5 K+-gluconate, 6 Ca2+-gluconate, 1 Mg2
+-gluconate, 5 glucose, 10 HEPES, pH 7.3 with NaOH or H3PO4),
the buffer was aspirated, and 25 μl Cl−-free loading buffer (assay
buffer containing an additional 2 mM BTC-AM, 2 mM amaranth and
1 mM tartrazine (Sigma-Aldrich, Brøndby, Denmark)) was added
to each well. The cells were incubated at 37 1C for 1 h and,
subsequently, transferred to a fluorimetric imaging plate reader
(FLIPR, Molecular Devices, Sunnyvale, CA, USA). For excitation, the
488 nm line of an argon laser was used, whereas a 540730 nm
band-pass filter was inserted on the emission side. A DMSO stock
solution of the compound was diluted in Cl−-free assay buffer
(final DMSO concentration o0.1%). The stimulus buffer consisted
of Cl−-free assay buffer supplemented with Tl2SO4 (final [Tl+]:
2 mM), K2SO4 at a concentration optimized individually for each
cell line to elicit a maximal and robust response in the presence of
a positive modulator of Kv7 channels (Kv7.2+7.3, Kv7.3−7.2, and
Kv7.3−7.2 mt: 5 mM; Kv7.4 and Kv7.5: 10 mM) as well as the
quenchers amaranth (2 mM) and tartrazine (1 mM). Assay plates
Using the mutagenesis methods described above, a mutated,
concatenated construct (hKv7.3–7.2 mt) was prepared using the
following oligonucleotides: hKv7.2, W236L, AfeI: GGAGCTGGTCA-
CaGCgctGTACATCGGCTTCCTTT and hKv7.3, W265L, AfeI: GCAAA-
were tested using
a two-addition protocol (1st add.: test
compound, 2nd add.: stimulus buffer). The cells were stimulated
in quadruplicate, i.e. four individual wells were stimulated per
concentration. In addition, 16 wells/plate served as negative
controls (buffer addition) and another 16 wells/plate as positive
controls. Full efficacies were defined by: retigabine: Kv7.3−7.2
(30 mM), Kv7.2+7.3 (30 mM), Kv7.4 (100 mM); ICA-27243: Kv7.3
−7.2 mt (30 mM); BMS-204352: Kv7.5 (30 mM); BMS-204352 was
used because we saw unspecific effects of retigabine when applied
at concentrations high enough to elicit a maximal response in the
Kv7.5 assay. The data from the individual wells were expressed as
fold increase (maximal signal (after addition of stimulus buffer)
divided by the average baseline value (immediately prior to the
addition)). Next, the data were background corrected by
subtraction of the averaged value of the fold increase for the
negative control wells. The data were subsequently analyzed by
normalizing the corrected fold increase values to the maximal
GAACTCATCACaGCgctGTACATCGGTTTCCTGA
(Eurofins,
MWG
Operon, Ebersberg, Germany) and the concatenated construct
described above as template. The resulting construct was verified
by restriction enzyme digestion and sequencing.
2.2.2. Cell lines
Stable human embryonic kidney (HEK) 293 cell lines expres-
sing hKv7.2, hKv7.4, hKv7.5, concatenated hKv7.3–7.2, concatenated
mutated hKv7.3–7.2 mt or co-transfected with equal molar
amounts of hKv7.2+hKv7.3 were generated as follows: One day
prior to transfection HEK293 cells were plated in a cell culture T25
flask (Thermo Fisher Scientific, Roskilde, Denmark) to 50–80%
confluence. The cells were transfected using the Lipofectamine
Plus kits (Life Technologies, Nærum, Denmark) according to the
manufacturer's instructions. The cells were trypsinized the next

W. Dalby-Brown et al. / European Journal of Pharmacology 709 (2013) 52–63
55
positive control response and fitting the data to the Hill equation,
as described in Hougaard et al. (2007). EC50 values and efficacy
(% of positive control) values are reported.
Individual slices were transferred to a submersion-style record-
ing chamber (Luigs & Neumann, Ratingen, Germany) and perfused
at 2 ml/min with aCSF maintained at 30 1C using a feedback-
controlled heater (Warner Instruments, Hamden, CT, USA). Whole-
cell patch clamp recordings were made from glutamatergic CA1
pyramidal neurons in hippocampus or dopaminergic neurons in
the substantia nigra pars compacta (SN). Cells were visualised at
40 ꢁ magnification using an up-right Olympus microscope
(BX51WI) equipped with oblique illumination. Patch pipettes were
pulled from borosilicate glass tubing (O.D. 1.32 mm) using a
horizontal electrode puller (Zeitz instruments). Pipette solutions
contained (in mM): 135 CH3KSO4, 10 KCl, 10 HEPES, 1 MgCl2,
2 Na2–ATP, 0.4 Na–GTP, pH 7.2 with KOH for hippocampus slices or
135 K+–gluconate, 10 KCl, 10 HEPES, 1 MgCl2, 5 Mg–ATP, 0.5
Na–GTP, pH 7.2 with KOH for midbrain slices. Electrodes
(3–7 MΩ) were advanced under positive pressure using an elec-
tronically controlled micromanipulator (Eppendorf, Radiometer,
Denmark). Following high resistance seal formation the mem-
brane was ruptured by suction and experiments performed using
an EPC-9 amplifier (HEKA, Lambrecht, Germany). Experimental
control, data acquisition, and basic analyses were done with the
Patchmaster (HEKA) software package. Input resistance was esti-
mated by measuring the change in membrane voltage produced
by a hyperpolarizing current pulse. For midbrain slices, some
experiments were conducted in the presence of the D2 antagonist,
Sulpiride (10 mM), which induced a more regular autonomous
firing pattern.
2.2.4. Electrophysiological experiments
2.2.4.1. Cell seeding. The cell cultures were split by trypsinization
1–2 times a week at 80–90% confluence. Cells for patch clamp
experiments were seeded on poly-D-lysine coated cover slips
(3.5 mm diameter, VWR International, Herlev, Denmark) placed
in 35 mm petri dishes on the day of the experiment.
2.2.4.2. Whole-cell voltage clamp. For each experiment a cover slip
was transferred to a 15 μl recording chamber and superfused at a
rate of 1 ml/min with an extracellular saline containing (in mM):
144 NaCl, 4 KCl, 2 CaCl2, 1 MgCl2 and 10 HEPES (pH adjusted to
7.4 with NaOH). Pipettes were pulled (DMZ electrode puller, Zeitz
Instruments, Augsburg, Germany) from borosilicate glass (Vitrex
Medica, Herlev, Denmark) and filled with an intracellular solution
containing (in mM): 150 KCl, 5.4 CaCl2, 1.75 MgCl2, 10 EGTA and 10
HEPES. Di-Na–ATP (4 mM) and Na–GTP (0.4 mM) were added on
the day of experiment and pH subsequently adjusted to 7.2 with
KOH. Experiments were controlled and data were sampled using
Pulse software and an EPC-9 amplifier (HEKA electronics,
Lambrecht, Germany) connected to the computer via the ITC-16
interface. Current traces and measurements of peak tail currents
were exported to IGOR (WaveMetrics, Lake Oswego, OR) for
further analysis. The filled pipettes had resistances of 1.8–3 MΩ
and the electrodes were off-set balanced before attaching a cell.
Capacitative transients were automatically cancelled after
establishment of the GΩ seal and again after establishment of
the whole-cell mode. The series resistances were in the range
3–5 MΩ, updated between each application of the voltage protocol
and compensated by 80%. Leak- and zero-subtractions were not
applied.
2.2.5. In vivo experiments
2.2.5.1. Animals. Sprague Dawley rats (Taconic) were used for
brain slice preparations and male Wistar rats (Taconic) were
used for kindling experiments. Female NMRI mice (Taconic) were
used for the 6 Hz testing. Male c57BL/6j mice (22–25 g, Harlan
laboratories) were used to investigate drug effects on
chlordiazepoxide-hydrochloride (CDP) plus d-amphetamine
sulphate (d-AMP) (both Sigma-Aldrich)-induced hyperactivity. All
animals were kept separately in ventilated closed racks
(Scantainer, Scanbur, Karlslunde, Denmark) at constant
temperature (21 1C) and humidity (60–70%) with a 12-hour light/
12-hour dark schedule (lights on at 0600 h). Female NMRI mice
(20–30 g, Harlan laboratories, Horst, The Netherlands) were used
for studying compound-induced effects on exploratory locomotor
activity, (+)-MK-801-maleate (MK-801, Sigma-Aldrich, Brøndby,
Denmark)-induced hyperactivity, and anticonvulsant activity in
the 6 Hz assay. Group housing and isolation from males were
chosen to prolong estrus above the normal 4–5 days (Ma et al.,
1999) and to make them cycle synchronously (Koyama, 2004).
Food (standard laboratory pellets) and water were available ad
libitum. Experiments were conducted during daytime in the light
phase. Animals were habituated to the experimental room
approximately 24 h before start of experiments and each animal
was only used once. All experimental procedures carried out in
these studies were in compliance with the European Communities
Council Directive of 24. November 1986 (86/609/EEC) and
approved by the Danish Animal Welfare Committee, appointed
by the Danish Ministry of Justice.
2.2.4.3. Analysis. The peak tail currents were measured and
depicted as a function of the activating step potential (Vm) and
fitted to a Boltzmann equation of the following form:
ItailðVmÞ ¼ Ið−∞Þ þ ðI ð∞Þ=ð1 þ expð−ðVm−V Þ=SÞÞÞ
½
tail
Itail(Vm) is the peak tail current recorded at −120 mV, after a
preceding activation to a given membrane potential (Vm). I(−∞) is
the background voltage-independent current and Itail(∞) is the
maximal fitted Itail. V is the potential where the current reaches
½
50% of (Itail(∞)−I(−∞)) and S is the slope factor.
2.2.4.4. Brain slice preparations. Hippocampus: 21–30 day-old
Sprague Dawley rats (Taconic, Lille Skensved, Denmark) were
decapitated and brains rapidly dissected into ice-cold modified
artificial cerebrospinal fluid (aCSF) of the following composition
(in mM): 75 sucrose, 85 NaCl, 2.5 KCl, 4 MgCl2, 0.5 CaCl2, 1.25
NaH2PO4, 26 NaHCO3, 16 glucose, saturated with 95% O2, 5% CO2.
Horizontal slices (300 mm) were cut using a vibrating tissue slicer
(VT1200 Leica, Ballerup, Denmark) and placed in a home-made
holding chamber at room temperature in aCSF of the following
composition (in mM): 127 NaCl, 2.5 KCl, 2 MgCl2, 2 CaCl2, 1.25
NaH2PO4, 26 NaHCO3, 16 glucose, bubbled with 95% O2 plus 5%
CO2. Midbrain: horizontal slices (300 mm) were prepared from
11–13 day-old Sprague Dawley rats (Taconic) in ice-cold aCSF
consisting of (in mM): 124 NaCl, 4 KCl, 1.2 MgSO4, 2.45 CaCl2, 1.25
NaH2PO4, 25.7 NaHCO3, 11 glucose, 0.15 ascorbic acid, bubbled
with 95% O2 plus 5% CO2. Slices were placed in aCSF in a home-
made chamber at room temperature for a minimum of 1 h prior to
experiments.
2.2.5.2. Amygdala kindling. Rats were anesthetized (Hypnorm/
Midazolam mixture sub cutaneous (s.c.) administration) and
stereotaxically implanted with a bipolar stimulating electrode
(Plastics One Inc., Roanoke, VA, USA, E363/2 twisted stainless
steel electrodes isolated with polyamide) in the right amygdala.
Stimulation was initiated after a postoperative recovery period of
at least 7 days and applied once or twice daily using an isolated
stimulator (A365, World Precision Instruments, Sarasota, Florida,

56
W. Dalby-Brown et al. / European Journal of Pharmacology 709 (2013) 52–63
USA). The subjects were stimulated at the after discharge
threshold intensity starting at 100 μA (biphasic square-wave
pulses, 1 ms duration, 50 Hz for 1 s). The electroencephalogram
(EEG) was recorded prior to and after each stimulation via a
1401plus analogue-digital converter and Spike2 (Windows version
6, Cambridge Electronic Design Limited, Cambridge, UK). The
behavioural manifestations of each seizure were scored
according to the classification of Racine (1972); (1) mouth and
facial movements, (2) head nodding, (3) forelimb clonus,
(4) rearing and (5) rearing and falling onto the cage floor.
Repeated stimulations resulted in prolonged afterdischarge
durations and progression of the behavioural manifest. Rats were
defined as fully kindled when stage five seizures appeared on
three consecutive stimulations. NS15370 or vehicle (0.5%
hydroxypropyl-methycellulose (HPMC)) was administered i.p. to
rats 30 min before evaluation. Afterdischarge durations were
expressed as % inhibition of the control level (identified two
days earlier) and data were analysed using one-way ANOVA
followed by Bonferroni t-test for multiple comparison against
vehicle treated rats.
2.2.5.6. MK-801-induced hyperactivity. The mice were placed
individually into test cages and left undisturbed. After 30 min of
habituation to the environment, vehicle or MK-801 was
administered intra peritoneally (i.p.; 0.2 mg/kg, dissolved in 0.9%
NaCl) and recording of animal movements commenced. NS15370
(suspended in 0.5% HPMC) was injected p.o. immediately before
habituation, i.e. 30 min before test start. Locomotor activity was
measured for a total of 60 min. All compounds were administered
in a dose volume of 10 ml/kg. Behavioural output was processed
and analyzed as described in Section 2.2.5.5.
2.2.5.7. CDP+d-AMP-induced hyperactivity. Male c57BL/6j mice
were injected p.o. with vehicle or NS15370 (suspended in 0.5%
HPMC), placed individually into test cages and left undisturbed for
30 min to habituate. After 30 min of habituation to the
environment, vehicle or a mixture of CDP+d-AMP (dissolved in
5% glucose) was injected i.p. (yielding final doses of 2.5 mg/kg and
4 mg/kg, respectively) and recording of animal movements
commenced. The dosing protocol allowed for assessment of drug
effect on spontaneous locomotor activity per se (NS15370 at 3 mg/
kg and 10 mg/kg+vehicle) as well as drug effects on CDP+d-AMP
induced hyperactivity (NS15370 at 3 mg/kg and 10 mg/kg+CDP+d-
AMP mixture). The mixture of CDP+d-AMP and NS15370 was
administered in a dose volume of 10 ml/kg. Locomotor activity was
measured for a total of 60 min. Behavioural output was processed
and analyzed as described in Section 2.2.5.5.
2.2.5.3. 6 Hz seizure test. Limbic partial seizures were induced via
corneal stimulation (6 Hz, 0.2 ms rectangular pulse width, 3 s
duration) using a Grass S48 stimulator, a constant current unit
(CCU1) and an isolation unit (SIU5) all driven by a stimulus
generator (Master-8, AMPI, Jerusalem, Israel). Mice were orally
pre-treated with either NS15370 (1, 3, 10 mg/kg in 0.5% HPMC) or
vehicle (n¼8) 30 min before the individual mice were challenged
with a current intensity of 32 mA (the calculated current intensity
that induces seizures in 97% of the population). The 6 Hz-induced
seizure was characterized by one of the following behavioral
components: stun, forelimb clonus, twitching of the vibrissae or
Straub-tail (Barton et al., 2001). Mice were scored as protected or
not protected in the 6 Hz test by the presence of any of these
behaviours for more than 3 s. The dose of NS15370 protecting 50%
of the animals from seizures (ED50), was calculated by logistic
regression analysis (GraphPad Prism version 5.00 for Windows,
GraphPad Software, San Diego, California, USA).
3. Results
3.1. In vitro characterization of NS15370 on recombinant Kv7
channels
3.1.1. Tl+-fluorescence assays
NS15370, shown in Fig. 1A (left panel), was synthezised as part
of a medicinal chemistry program on Kv7.x positive modulators
(see Supplementary Fig. 1). As an estimate of potency at Kv7
channels, project compounds were evaluated in an optimized
fluorescence-based Tl+ influx assays using HEK293 cells stably
expressing Kv7 channels. Fig. 1B (left panel) shows an example
with Kv7.2+7.3 channels, where the fluorescence signal is followed
over time in the presence (full line) and absence (dashed line) of
NS15370 (0.3 mM, first addition). The second addition containing
Tl+ induced a marked increase in fluorescence only in the presence
of NS15370, indicating increased Tl+ uptake due to activation of
2.2.5.4. Pharmacokinetic evaluation of NS15370. Chemicals and
Reagents: Water purified in a Milli-Q water purification system
(Millipore), HPLC grade solvents and other chemicals of analytical
grade were used. LC-MS analysis: LC-MS analysis was performed
using Cohesive turboflow TLX-1 system coupled to triple
quadrupole mass spectrometer (TSQ, ThermoFinningan). The
mass spectrometer was run in SRM-mode (SRM-transition:
457-348). An Xterra MS C8 IS (3.5 μm, 2.0 mm ꢁ 20 mm)
column was used. Water with formic acid (0.2%) and acetonitril
was used as mobile phases performing gradient elution.
Kv7 channels. The effect of NS15370 was quantified in
a
concentration-response regimen (Fig. 1B, right panel, closed cir-
cles), and the EC50 value was estimated at 0.1370.1 mM with an
efficacy of 100711% (n¼4) compared to retigabine (30 mM)
serving as positive control. The corresponding EC50 value for
retigabine itself (open circles) was estimated at 1.571.1 mM and
an efficacy of 103717% of the control wells (n¼119). NS15370 was
furthermore tested on Kv7.2, Kv7.4, Kv7.5 as well as Kv7.3–7.2
(concatenated heteromers, as control for subunit stoechiometry,
see also Fig. 1C, left panel) channels expressed stably in HEK293
cells (Table 1). As seen, NS15370 potently (EC50∼100 nM) activated
all Kv7 channel sub-type combinations in this assay. In contrast,
retigabine was overall less potent (range 12–80 fold compared to
NS15370) and in particular exhibited lower potency for Kv7.4 and
Kv7.5. Lastly, we also checked the effect of NS15370 by testing its
effects on the mutated Kv7.3–7.2 construct (7.3:W265L; 7.2:
W236L) that disrupt the retigabine binding site (Fig. 1C, left panel,
open circles). As for retigabine (Fig. 1C, right panel) activation of
Kv7 channels by NS15370 depended on the tryptophan in the
S5 helix.
2.2.5.5. Exploratory locomotor activity. The study was conducted as
previously reported (Butini et al., 2009). In short, drug effects on
exploratory locomotor activity were assessed in homecages
containing automated activity frames (20 cm ꢁ 40 cm ꢁ 18 cm,
TSE systems GmbH) and equipped with infrared sensors (6 ꢁ 2)
for acquisition of animal movements (also used in Sections 2.2.5.5
and 2.2.5.6). The mice were placed individually into novel home
cages, whereafter animal movements were recorded immediately
for a period of 30 min. Mice were injected per orally (p.o.) with
vehicle (0.5% hydroxylpropylmethyl cellulose, HPMC) or test
compound (suspended in 0.5% HPMC) in a dose volume of
10 ml/kg 30 min before test start. Behavioural output was
subsequently processed on a computer (ActiMot software, TSE
systems GmbH) and analyzed by two-way ANOVA with time and
drug as variables.

W. Dalby-Brown et al. / European Journal of Pharmacology 709 (2013) 52–63
57
Table 1
Selectivity profiles of NS15370 and retigabine between selected Kv7 sub-types. EC50
values estimated from concentration–response curves (Fig. 1) using a FLIPR-based
Tl+ influx assay. NS15370 and retigabine were tested at concentrations ranging
from 0.00316 nM to 31.6 μM on the Kv7 subtypes: Kv7.2, concatenated Kv7.3−7.2,
non-concatenated Kv7.2+7.3, Kv7.4 and Kv7.5, and the effect at any given concen-
tration was background corrected and normalized to the positive control response
defined by: 30 mM retigabine (Kv7.2, concatenated Kv7.3−7.2 and Kv7.2+7.3), 100 μM
retigabine (Kv7.4) and 30 mM BMS-204352 (Kv7.5). The EC50 values were estimated
as described in Section 2. The data represents at least three independent
experiments.
NS15370
Retigabine
Kv7.2
EC50 (μM)
Max (% ctrl)
0.170.1
127736
2.871.4
100726
Kv7.2+7.3
EC50 (μM)
Max (% ctrl)
0.1370.1
100711
1.571.1
103 717
Kv7.3−7.2
EC50 (μM)
Max (% ctrl)
0.0470.04
102736
0.971
103714
Kv7.4
EC50 (μM)
Max (% ctrl)
0.170.1
115713
8.174
119 724
Kv7.5
EC50 (μM)
Max (% ctrl)
0.1570.1
7272
1274
54711
Fig. 2. Comparison of NS15370 and Retigabine in whole-cell patch clamp experiments
using HEK293 cells expressing Kv7.2+7.3 channels. (A) A HEK293 cell expressing Kv7.2
+7.3 channels was clamped at −90 mV in the whole-cell configuration. Current traces
obtained before (black traces) and during the application of retigabine (3 μM, gray
trace) is shown in the left panel and the effect of NS15370 (0.03 μM) on the same cell is
shown in the right panel (gray trace). The voltage was stepped to −30 mV for 1 s every
5 s in order to activate the channels and subsequently to −60 mV which deactivates
the channels in the absence of activators (see protocol on Fig.). (B) The currents
mediated by Kv7.2+7.3 channels during the voltage steps to −30 mV were measured at
50 ms (circles) and 950 ms (squares) as indicated by the symbols in (A) and plotted as
a function of time. Retigabine and NS15370 were included in the bath solution during
the periods indicated by the bars.
3.1.2. Whole cell voltage-clamp assays.
We validated the effects of NS15370 and characterized its
mode-of-action in whole cell recordings using Kv7.2+7.3 and
Kv7.4 as representatives of hetero- and homomeric channels.
Fig. 2A shows currents elicited from a Kv7.2+7.3 expressing cell,
in the control situation and in the presence of NS15370 (0.03 μM)
or retigabine (3 μM). Both NS15370 and retigabine strongly accel-
erated the activation rate upon a depolarization to −30 mV, and
almost abolished deactivation upon the ensuing hyperpolarizing
step to −60 mV, clearly validating NS15370 as a Kv7 channel
activator. In contrast to retigabine NS15370 did not increase the
current amplitude recorded at the end of the 1 s voltage step to
−30 mV, rather it showed a tendency to decrease (see below).
Fig. 2B, which shows a time-course read-out from this experiment,
highlights the preferential activating effect of NS15370 in the early
phase of the depolarizing step and even a slight inhibition in the
late phase (depolarization for 1 s).
In order to characterize the effects of NS15370 on the voltage-
dependence of Kv7.2+7.3 and Kv7.4 in comparison to retigabine, we
performed a series of tail current analyses at different compound
concentrations (Figs. 3 and 4). Fig. 3A and B exemplify the effects of
NS15370 (0.03 μM and 0.3 μM) on Kv7.2+7.3 currents elicited by
increasing depolarizations. At 0.03 μM the depolarization-evoked
sustained currents at −60 mV and −30 mV were slightly increased
as were the amplitudes and durations of the corresponding tail
−35 mV. Fig. 4 compiles these purely descriptive analyses performed
for a range of concentrations of both NS15370 and retigabine on Kv7.2
+7.3 as well as Kv7.4. Evidently, NS15370 is a much more potent
activator than retigabine (V vs. concentration; Fig. 4A) on both
½
subtypes, having effects at 0.03 μM corresponding to those seen for
retigabine at 3 μM. However, NS15370 also induces
a clear
concentration-dependent current depression (Imax vs. concentration,
Fig. 4B), which is not observed for retigabine. Overall, the electro-
physiological data on relative potency and selectivity are in agreement
with the FLIPR data, which did not, however, reveal the strong
inhibiting effect of NS15370 at depolarized potentials.
NS15370 was without any detectable effect on the GABAA
receptors combinations most abundantly expressed in the brain
(α1β2γ2; α3β2γ2; α3β2γ2) in concentrations up to 30 μM, when tested
on Xenopus laevis oocytes expressing the human receptors (in
contrast to retigabine, see Supplementary Fig. 2). Furthermore, in a
LeadProfilingScreen (Ricerca) comprising 66 ion channel and 7TM
receptor binding assays, NS15370 (10 μM) was found only to
inhibit radioligand binding to Nav and Cav channels (Batrachotoxin
and Nitrendipine assays, respectively), and in subsequent func-
tional in house patch clamp studies, the IC50 values were esti-
mated at 410 μM (data not shown). Furthermore, NS15370
showed no effects on action potential parameters such as rise-
time, height and broadness (Sections 3.1.3 and 3.1.4).
currents elicited by
a deactivating hyperpolarization step to
−120 mV. However, upon depolarization to 0 mV, NS15370 decreased
the amplitude of the depolarization-elicited current and the morphol-
ogy changed from a sustained to an apparantly inactivating current. At
0.3 μM (Fig. 3B) this pattern was accentuated further: The step to
−60 mV revealed a pure activation with increased sustainedand tail
currents, whereas the current induced by steps to −30 mV and 0 mV
were strongly depressed in the late phase, which also suppressed the
peaks of the tail-currents. The peak tail currents vs. the preceeding
voltage steps are plotted in Fig. 3C and D. The 0.03 μM Fig. shows a
relatively pure left-shift (V effect) of the activation curve, whereas the
½
0.3 μM curve exhibits a complex mix of pronounced activation (V
effect, from −38 mV to −87 mV), and current depression (Imax effect;
from −1.1 nA to −0.7 nA) which makes the curves “crossover“ at
3.1.3. Current clamp recordings—Hippocampal CA1 neurons
Kv7 channels are important in the regulation of action potential
firing in the hippocampus (Gu et al., 2005), which prompted us to
½

58
W. Dalby-Brown et al. / European Journal of Pharmacology 709 (2013) 52–63
Fig. 3. Effects of NS15370 on Kv7.2+7.3 activation curves obtained from whole-cell patch clamp experiments. Whole-cell voltage clamp experiments were conducted using
HEK293 cells stably expressing Kv7.2+Kv7.3. The cells were clamped at −90 mV from where the channels were activated by 1 s steps to potentials between −90 mV and
10 mV using 10 mV increments with 5 s or 10 s in between. The potential was stepped to −120 mV after each depolarising step. (A) Current traces upon depolarization to
−60 mV, −30 mV and 0 mV before (black traces) and in the presence of NS15370 applied at a concentration of 0.03 μM (gray traces). (B) Same as (A) but using 0.3 μM
NS15370. The arrow heads indicate where the peak tail currents were measured. (C) Peak tail currents as a function of the preceding step potential. Open circles are before
and filled circles are in the presence of NS15370 (0.03 μM). (D) Same as (C) but obtained from an experiment with 0.3 μM NS15370.
test the effect of NS15370 on evoked firing in CA1 pyramidal neurons
(Fig. 5A). Depolarizing current pulses elicited trains of action poten-
tials characterized by strong spike frequency adaptation (Control).
Application of NS15370 (1 μM) gradually decreased evoked firing
frequency and increased spike frequency adaptation of all CA1
neurons tested. In most experiments only a single action potential
remained in the presence of NS15370, which was accompanied by a
472 mV (n¼4) hyperpolarization of the membrane potential. The
effect of NS15370 was reversible and a subsequent application of the
Kv7 channel inhibitor, XE991 (10 μM), reduced spike frequency
adaptation and facilitated repetitive firing.
3.1.4. Current clamp recordings—Dopaminergic neurons
Kv7 channels are expressed in the midbrain (Koyama and
Appel, 2006; Hansen et al., 2006), and administration of retiga-
bine reduces activity and signalling of dopaminergic neurons in
substantia nigra pars compacta and ventral tegmental area
(Hansen et al., 2006; Sotty et al., 2009). Here we explored the
effect of NS15370 on dopaminergic neurons in rat midbrain
slices (Fig. 5B). Dopaminergic neurons exhibit
a
slow
(1.270.2 Hz, n¼4), spontaneous pacemaker-like firing pattern
with deep after-hyperpolarizations following the individual
spikes (Control). Application of NS15370 (1 μM) rapidly slowed

W. Dalby-Brown et al. / European Journal of Pharmacology 709 (2013) 52–63
59
Fig. 4. Effects of NS15370 and retigabine on parameters derived from activation curves using Kv7.2+7.3 or Kv7.4 expressing HEK293 cells. (A) Shift in V obtained from
½
activation curves plotted as a function of the compound concentration. Filled symbols are data on NS15370 and open symbols are experiments with retigabine. Circles are
from cells expressing Kv7.2+Kv7.3 and squares are from Kv7.4 expressing cells. (B) Change in the fitted maximal current obtained from activation curves and plotted as a
function of the compound concentration. Symbols are as in (A).
Fig. 5. NS15370 reduces the firing frequency of hippocampal CA1 neurons and
substantia nigra compacta dopaminergic neurons. (A) Trains of action potentials
elicited by 600 ms current injections (160 pA) from the resting membrane potential
(−66 mV, no bias current). The number of action potentials elicited by current
injections was decreased in the presence of 1 μM NS15370. This effect was reversed
upon prolonged wash-out. Application of XE991 (10 μM) induced an increase in the
number of action potentials elicited by the current injections. The Fig. is repre-
sentative of five experiments. (B) Pacemaker-like firing recorded under control
condition is inhibited by superfusion with 1 μM NS15370. Application of NS15370
initially reduced the firing frequency, but a complete silencing of spontaneous
activity was observed upon prolonged exposure to the compound. Application of
XE991 (10 μM), in the continuous presence of NS15370, effectively restored pace-
maker activity. The illustrated experiment was conducted in the presence of 10 μM
of the D2 receptor antagonist sulpiride, similar results were achieved in the absence
of sulpiride (n¼3).
Fig. 6. Anticonvulsant effects of NS15370 in amygdala-kindled rats. NS15370
administered i.p. to fully kindled Wistar rats 30 min before electrical stimulation
in the amygdaloid focus. Symbols represent the percentage inhibition of after
discharge duration (mean7SEM, n¼6–10) compared to the control level identified
2 days prior to the test day. Data were analysed by one way ANOVA followed by
Bonferroni t-test for multiple comparison versus the vehicle group, ***Po0.001.
1988). NS15370 was given i.p. to amygdala-kindled rats with a pre-
treatment time of 30 min. Amygdala-kindled rats displayed a
significant increase in after discharge duration (main effect of
treatment: F(4, 40)¼17, 11, Po0.001 as compared to vehicle
treatment) induced by electrical stimulation at the amygdaloid
focus. Upon administration of NS15370, the duration of the after
discharge was inhibited in a dose-dependent manner, as seen from
Fig. 6. Full inhibition was observed at 30 and 45 mg/kg (i.p. both
Po0.001), whereas 10 mg/kg inhibited the after discharge dura-
tion by 70% (Po0.001). The effect of 3 mg/kg (5%) was not
significant, implying an ED50 value between 3 mg/kg and 10 mg/
kg. Vehicle control responses (1%) were not significantly different
from control levels recorded 2 days earlier.
the firing frequency by increasing the length of the after-
hyperpolarizations and ultimately silenced the neuron comple-
tely. Co-application of XE991 (10 μM) reversed the inhibitory
effect of NS15370 on basal spike activity and re-established
regular pacemaker firing.
3.2. In vivo characterization—Epilepsy models
NS15370 was given p.o. (1 mg/kg, 3 mg/kg and 10 mg/kg) to
mice in the 6 Hz model of limbic seizures (corneal stimulation,
6 Hz, 32 mA) and the results are shown in Fig. 7A. NS15370 dose-
dependently protected an increasing fraction of mice against
seizures with full protection (eight out of eight mice) at 10 mg/
kg, implying an ED50 value between 3 mg/kg and 10 mg/kg. In
To evaluate the anticonvulsant potential of NS15370, the
compound was tested in the rat amygdala kindling model of
complex partial seizures and in the mice 6 Hz model of limbic
seizures since these models reflect aspects of clinical partial
seizures in humans (Barton et al., 2001; Löscher and Schmidt,

60
W. Dalby-Brown et al. / European Journal of Pharmacology 709 (2013) 52–63
Fig. 7. Anticonvulsant effects of NS15370 in 6 Hz induced seizures in mice. (A) Dose-response relationship showing the anticonvulsant effect of NS15370 orally administered to
female NMRI mice 30 min before induction of limbic seizures with corneal electrical stimulation (6 Hz, 32 mA). Data are plotted as percentages of protected mice with the actual
numbers of animals (protected/total) aligned with each point on the graph. (B) Time dependent correlation between protective effects and plasma concentrations of NS15370. Upper
panel: Percentage of protected mice vs. time after dosing of 10 mg/kg NS15370. Lower panel: Average plasma concentrations vs. time after dosing (same mice as in upper panel).
Fig. 8. Anti-psychotic and anti-manic effects of NS15370 in mice. Effect of NS15370 on (A) MK-801-induced hyperactivity in female NMRI mice, (B) CDP+d-AMP-induced
hyperactivity and (C) spontaneous exploratory locomotor activity both in male c57BL/6j mice. NS15370 was administered p.o. 30 min before test start. MK-801 (0.2 mg/kg)
and CDP (2.5 mg/kg)+d-AMP (4 mg/kg) mixture were administered i.p. immediately before test start. Results were statistically evaluated by subjecting data to two way
ANOVA followed by Fishers LSD test for multiple planned comparisons (MK-801-induced hyperactivity and exploratory locomotor activity in NMRI mice) and one way
ANOVA followed by Fishers LSD test (CDP+d-AMP hyperactivity and effects on spontaneous activity in c57BL/6j mice). V: vehicle; mix: CDR+d-AMP-mixture; MK: MK-801;
###: Po0.001 vs. vehicle; ***: Po0.001 vs. CDP+d-AMP or MK-801.
order to investigate the time-dependence of the antiepileptic
action as well as its correlation to the systemic exposure of
NS15370, we performed a series of experiments with varying
pre-treatment times before seizure induction. As seen from Fig. 7B,
the time-dependent degree of protection (Upper panel) correlated
with the mean plasma concentration of NS15370 (Lower panel).
effect on spontaneous exploratory locomotor activity was investigated
by administering the compound to naïve female NMRI mice intro-
duced into novel environments, as shown in Fig. 8C. NS15370 had no
effect on exploratory locomotor activity at 3 mg/kg and 10 mg/kg but
caused a significant depression at 30 mg/kg (main effect of treatment:
F(3, 161)¼27.89, Po0.001, Po0.001 for 30 mg/kg as compared to
vehicle treatment).
3.3. In vivo characterization—Models of mania and psychosis
When hyperactivity was induced in male c57BL/6j by a mixture
of CDP+d-AMP, 3 mg/kg and 10 mg/kg of NS15370 significantly
reduced the excessive locomotor activity, as seen in Fig. 8B (main
effects of treatment: F(2, 19)¼5.2, P¼0.017; P¼0.019 and 0.009 for
3- and 10 mg/kg versus CDP+d-AMP, respectively). These doses did
not significantly affect locomotor activity when administered to
naïve c57BL/6j mice.
Accordingly, NS15370 was able to reduce NMDA antagonist—as
well as CDP+d-AMP-induced hyperactivity in doses not affecting
spontaneous locomotor activity per se. Reduction of spontaneous
activity required 10 times higher doses of NS15370 compared to
NS15370 was tested in two mice models sensitive to antipsy-
chotic- and antimanic treatment, respectively, namely, MK-801-
induced hyperactivity and CDP+d-AMP-induced hyperactivity.
As seen in Fig. 8A, NS15370 significantly and dose-dependently
reduced hyperactivity induced by the NMDA antagonist, MK-801, in
female NMRI mice (main effect of treatment: F(4, 419)¼35.79,
Po0.001) with a minimal effective dose of less than or equal to
3 mg/kg (Po0.001, P¼0.0.082 and Po0.001 for 3 mg/kg, 10 mg/kg
and 30 mg/kg, respectively, as compared to MK-801 treatment). Drug

W. Dalby-Brown et al. / European Journal of Pharmacology 709 (2013) 52–63
61
the doses that reduced MK-801-induced hyperactivity in naïve
NMRI mice, and at least three fold higher doses compared to
effective doses in the CDP+d-AMP assay in c57BL/6j mice.
channel activation and other antiepileptic principles, such as Nav
block (Lamotrigine, Phenytoin) or synaptic vesicle 2A binding (Leve-
tiracetam) (Bialer et al., 2009; Barton et al., 2001).
Whereas NS15370 showed very potent and dose-dependent antic-
onvulsant effects in the mouse 6 Hz assay, no such correlation could
be made in the rat amygdala kindling model probably due to poor
exposure of NS15370 in rats. However, by applying i.p. administration,
a sufficient exposure was achieved to obtain full protection (98% and
97% at 30 mg/kg and 45 mg/kg, respectively) against the after dis-
charge duration following a partial seizure, which is believed to be the
test variable exhibiting the highest clinical relevance (Large et al.,
2012). Comparable effects have been reported on retigabine in
amygdala kindled rats (Tober et al., 1996), where 92% and 77%
protection against after discharges were obtained (5 mg/kg i.p. and
15 mg/kg p.o., respectively), compared to the control level determined
2 days earlier. The Kv7 activator ICA-27243 has also been reported to
produce clear anticonvulsant effects in the rat amygdala kindling
model, without exhibition of any after discharge at the highest dose
tested, i.e. displaying full protection (Roeloffs et al., 2008). Accordingly,
NS15370 showed significant anticonvulsant properties in both a
mouse and a rat model of partial seizures. Due to the poor metabolic
properties of NS15370 in rat, but not in mouse, potencies are difficult
to compare between species. However, full protection against seizures
as well as a complete inhibition of after discharge duration, were
observed in both models, indicative of a fully efficacious compound.
It has previously been shown that retigabine potentiates GABA-
activated currents at concentrations ≥10 μM (Rundfeldt and
Netzer, 2000a), and it was speculated that this could contribute
to the anticonvulsive effects of retigabine (van Rijn and Willems-
van Bree, 2003). However, in contrast to retigabine, NS15370 is
devoid of GABAA receptor activation (up to 30 μM), suggesting that
the activation of Kv7 channels in itself is sufficient to elicit
antiepileptic activity in rodents. This is in congruence with what
has been reported for the Kv7 activator ICA-27243, which is also
without effect on GABAA channels (Roeloffs et al., 2008).
Consistent with the pronounced stimulation of Kv7 channels in
hippocampal CA1 neurons, we also find that NS15370 reduces the
frequency of autonomous firing of midbrain dopaminergic neurons in
a XE991-sensitive manner, signifying activation of Kv7 channels, which
is in accordance with effects of retigabine observed both in vitro and
in vivo (Sotty et al., 2009; Hansen et al., 2006). In addition, the present
study clearly demonstrates that NS15370 reduces locomotor hyper-
activity induced by MK-801 as well as by the combination of CDP+d-
AMP. Reductions in hyperactivity were apparent in doses at least 3–10
fold lower than doses affecting spontaneous locomotor activity,
indicating that the effects were not likely to be confounded by non-
specific, sedative/motor imparing effects. Similar effects have been
reported for retigabine in the acute CDP+d-AMP assay (Redrobe and
Nielsen, 2009) and in more disease-related models such as AMP-
induced sensitization of locomotor activity in mice (Dencker and
Husum, 2010) and PCP-induced sensitization of locomotor activity in
rats (Sotty et al., 2009).
4. Discussion
This study describes synthesis and characterization of NS15370,
a novel activator of Kv7 channels. NS15370 is an analogue of
retigabine, which like this molecule interacts with all homomeric
and heteromeric Kv7.2–Kv7.5 channels via the S5 residue Trp-265
(Kv7.3 numbering). NS15370 is 10–100 times more potent than
retigabine in vitro (EC50ꢀ100 nM vs. 2–8 μM in the Tl+-assay) but
has no effect on GABAA receptors (retigabine potentiates α1β2γ2, α3β2γ2,
and α5β2γ2 receptors, see also Rundfeldt and Netzer, 2000a; van Rijn
and Willems-van Bree, 2003). From a selectivity/potency point-of-
view NS15370 is thus a more selective Kv7-tool for use in brain slices
and whole animal experiments. We conclude that the demonstrated
dampening of neuronal excitability in slices as well as the anti-
epileptic and anti-psychotic/manic effects in rodents is due to activa-
tion of Kv7 channels.
NS15370 displays a complex concentration-dependent mode-
of-action as revealed by voltage-clamp analysis. At low concentra-
tions (0.03 μM and lower), activation of both Kv7.2+Kv7.3 and Kv7.4
monomers is a relatively clean 20 mV hyperpolarizing shift of the
activation curve (V effect). A similar shift in V was only observed
with retigabine at a concentration of 10 μM (see also Main et al.,
2000), underpinning the much higher potency of NS15370 (∼100–
½
½
300 times). Hyperpolarizing shifts in V are also the main mode-
½
of-action of other pharmacological activators of Kv7, such as BMS-
204352 (Schrøder et al., 2001), (S)-1 acrylamide (Bentzen et al.,
2006) and ICA-27243 (Wickenden et al., 2008). However, the
magnitude of the maximal shift in V is less pronounced for these
½
compounds than what is seen with NS15370. At higher concentra-
tions of NS15370, current depression becomes prominent and
results in a complex picture, consisting of a strongly activating
hyperpolarizing shift in V combined with an inhibition of the
½
plateau current at sustained depolarizations. In our hands this
effect is not seen with retigabine (in agreement with Wickenden
et al., 2000 and Tatulian et al., 2001, but see Rundfeldt and Netzer,
2000b). A similar mixed mode-of-action has been described for
(S)-2, an acrylamide Kv7 modulator (Blom et al., 2009).
M-channels are thought to mediate early spike frequency adapta-
tion in hippocampal CA1 neurons by contributing to the Ca2+-inde-
pendent component of the medium after-hyperpolarization (Gu
et al., 2005). Consistent with this, we demonstrate that application
of NS15370 profoundly augments spike frequency adaptation upon
injection of depolarizing current pulses, an effect that was amelio-
v
rated by the selective K 7 channel inhibitor XE991. Overall, these
findings support the notion that activation of Kv7 channels in the
hippocampus serve as a molecular brake during conditions of
hippocampal hyperexcitability. In agreement with this NS15370
exhibits pronounced anticonvulsant effect by producing full protec-
tion in two models of electrically induced partial seizures, the mouse
6 Hz seizure- and the rat amygdala kindling model, consolidating Kv7
channels as an effective target for antiepileptic drugs. Importantly,
the anticonvulsant effects were observed at concentrations 3–10 fold
lower than doses affecting spontaneous locomotor activity. This is in
accordance with what has been reported for the Kv7 activators
retigabine (Bialer et al., 2009) and ICA-27243 (Roeloffs et al., 2008)
in the mouse 6 Hz seizure assay (32 mA); however, NS15370 was
slightly more potent in the 6 Hz assay compared to both retigabine
and ICA-27243. Importantly, when raising the stimulation intensity
to 44 mA, in order to recruit more brain structures (Barton et al.,
2001), both retigabine and NS15370 (data not shown) still effectively
blocked seizures, which may point to a clear difference between Kv7
Given the lack of subtype-selectivity of NS15370, we are unable to
state which sub-types carry the main responsibility for the observed
anti-psychotic and anti-manic effects. It could be speculated that
Kv7.4 and Kv7.5 are not the main targets, since BMS-204352, a
compound displaying some selectivity toward Kv7.4 and Kv7.5, shows
reduced (or lack of) effect in identical behavioural models (Redrobe
and Nielsen, 2009 ;Nielsen, personal communication).
In summary, we have identified and characterized NS15370—a
novel Kv7 positive modulator with structural similarity to retiga-
bine, but displaying much higher potency, distinct biophysical
properties and improved selectivity against GABAA receptors. Our
findings consolidate Kv7 channels as an anti-epileptic target and
suggest that Kv7 channels may carry potential as novel targets for
anti-psychotic/anti-manic treatments. However, since NS15370 is

62
W. Dalby-Brown et al. / European Journal of Pharmacology 709 (2013) 52–63
a pan-Kv7.2–7.5 positive modulator, the peripheral Kv7-related
side effects observed for retigabine, such as urinary retention
(Brickel et al., 2012), are most likely also to be found for a
compound like NS15370, and further chemical optimization will
be needed in order to obtain significant advances in that respect.
Chen, Y.H., Xu, S.J., Bendahhou, S., Wang, X.L., Wang, Y., Xu, W.Y., Jin, H.W., Sun, H.,
Su, X.Y., Zhuang, Q.N., Yang, Y.Q., Li, Y.B., Liu, Y., Xu, H.J., Li, X.F., Ma, N., Mou, C.P.,
Chen, Z., Barhanin, J., Huang, W., 2003. KCNQ1 gain-of-function mutation in
familial atrial fibrillation. Science 299, 251–254.
Dencker, D., Dias, R., Pedersen, M.L., Husum, H., 2008. Effect of the new anti-
epileptic drug retigabine in a rodent model of mania. Epilepsy Behav. 12, 49–53.
Dencker, D., Husum, H., 2010. Antimanic efficacy of retigabine in a proposed mouse
model of bipolar disorder. Behav. Brain. Res. 207, 78–83.
Dupuis, D., Schrøder, R.L., Jespersen, T., Christensen, J.K., Christophersen, P.,
Jensen, B.S., Olesen, S.P., 2002. Activation of KCNQ5 channels stably expressed
in HEK293 cells by BMS-204352. Eur. J. Pharmacol. 437, 129–137.
Authorship contributions
Fattore, C., Perucca, E., 2011. Novel medications for epilepsy. Drugs 71, 2151–2178.
Greenwood, I.A., Ohya, S., 2009. New tricks for old dogs: KCNQ expression and role
in smooth muscle. Br. J. Pharmacol. 156, 1196–1203.
Gu, N., Vervaeke, K., Hu, H., Storm, J.F., 2005. Kv7/KCNQ/M and HCN/h, but not KCa2/SK
channels, contribute to the somatic medium after-hyperpolarization and excitability
control in CA1 hippocampal pyramidal cells. J. Physiol. 566, 689–715.
Hansen, H.H., Andreasen, J.T., Weikop, P., Mirza, N., Scheel-Krüger, J., Mikkelsen, J.D.,
2007. The neuronal KCNQ channel opener retigabine inhibits locomotor activity
and reduces forebrain excitatory responses to the psychostimulants cocaine,
methylphenidate and phencyclidine. Eur. J. Pharmacol. 570, 77–88.
Hansen, H.H., Ebbesen, C., Mathiesen, C., Weikop, P., Rønn, L.C., Waroux, O., Scuvée-
Moreau, J, Seutin, Mikkelsen, J.D., 2006. The KCNQ channel opener retigabine
inhibits the activity of mesencephalic dopaminergic systems of the rat.
J. Pharmacol. Exp. Ther. 318, 1006–1019.
Hougaard, C., Eriksen, B.L., Jørgensen, S., Johansen, T.H., Dyhring, T.H., Madsen, L.S.,
Strøbæk, D., Christophersen, P., 2007. Selective positive modulation of the SK3
and SK2 subtypes of small conductance Ca2+-activated K+ channels. Br. J.
Pharmacol. 151, 655–665.
ꢂ
Participated in research design: Hougaard, Jacobsen, Nielsen,
Erichsen, Strøbæk, Jørgensen, Ahring, Christophersen
Conducted experiments: Hougaard, Jacobsen, Nielsen, Erichsen,
Strøbæk, Jørgensen
Contributed with new reagents or analytical tools: Jensen,
Ahring, Jessen, Dalby-Brown
Performed data analysis: Hougaard, Jacobsen, Nielsen, Erichsen,
Strøbæk, Jørgensen
ꢂ
ꢂ
ꢂ
ꢂ
Wrote or contributed to the writing of the manuscript: Dalby-
Brown, Jensen, Hougaard, Jacobsen, Nielsen, Erichsen, Grunnet,
Strøbæk, Jørgensen, Ahring, Christophersen
Jensen, H.S., Grunnet, M., Olesen, S.P., 2007. Inactivation as a new regulatory
mechanism for neuronal Kv7 channels. Biophys. J. 92, 2747–2756.
Jentsch, T.J., 2000. Neuronal KCNQ potassium channels: physiology and role in
disease. Nat. Rev. Neurosci. 1, 21–30.
Acknowledgements
The excellent technical assistance from Anja Dahl Nielsen, Susanne
Kalf Hansen, Vibeke Meyland Schmidt, Jette Sonne, Anne Stryhn
Meincke, Lene Gylle Larsen, Hanne Nord Søndergård, Henrik Marcher,
Sanne Laulund and Nina Nørager (all NeuroSearch) and Ibrahim
Bulduk (Mercachem, The Netherlands) is greatfully acknowledged.
Jespersen, T., Grunnet, M., Olesen, S.P., 2005. The KCNQ1 potassium channel: from
gene to physiological function. Physiology (Bethesda) 20, 408–416.
Korsgaard, M.P.G., Hartz, B.P., Brown, W.D., Ahring, P.K., Strøbaek, D., Mirza, N.R.,
2005. Anxiolytic effects of Maxipost (BMS-204352) and retigabine via activa-
tion of neuronal Kv7 channels. J. Pharmacol. Exp. Ther. 314, 282–292.
Koyama, S., Appel, S.B., 2006. Characterization of M-current in ventral tegmental
area dopamine neurons. J. Neurophysiol. 96, 535–543.
Koyama, S., 2004. Primer effects by conspecific odors in house mice:
a new
perspective in the study of primer effects on reproductive activities. Horm.
Behav. 46, 303–310.
Appendix A. Supporting information
Kubisch, C., Schroeder, B.C., Friedrich, T., Lütjohann, B., El-Amraoui, A., Marlin, S.,
Petit, C., Jentsch, T.J., 1999. KCNQ4, a novel potassium channel expressed in
sensory outer hair cells, is mutated in dominant deafness. Cell 96, 437–446.
Large, C.H., Sokal, D.M., Nehlig, A., Gunthorpe, M.J., Sankar, R., Crean, C.S.,
Vanlandingham, K.E., White, H.S., 2012. The spectrum of anticonvulsant efficacy
of retigabine (ezogabine) in animal models: implications for clinical use.
Epilepsia 53, 425–436.
Supplementary data associated with this article can be found in the
online version at http://dx.doi.org/10.1016/j.ejphar.2013.03.039.
Löscher, W., Schmidt, D., 1988. Which animal models should be used in the search
for new antiepileptic drugs? A proposal based on experimental and clinical
considerations. Epilepsy Res. 2, 145–181.
References
Ma, W., Miao, Z., Novotny, M.V., 1999. Induction of estrus in grouped female mice
(Mus domesticus) by synthetic analogues of preputial gland constituents.
Chem. Senses 24, 289–293.
Main, M.J., Cryan, J.E., Dupere, J.R., Cox, B., Clare, J.J., Burbidge, S.A., 2000.
Modulation of KCNQ2/3 potassium channels by the novel anticonvulsant
retigabine. Mol. Pharmacol. 58, 253–262.
Munro, G., Erichsen, H.K., Mirza, N.R., 2007. Pharmacological comparison of antic-
onvulsant drugs in animal models of persistent pain and anxiety. Neurophar-
macology 53, 609–618.
Otto, J.F., Kimball, M.M., Wilcox, K.S., 2002. Effects of the anticonvulsant retigabine
on cultured cortical neurons: changes in electroresponsive properties and
synaptic transmission. Mol. Pharmacol. 61, 921–927.
Redrobe, J.P., Nielsen, A.N., 2009. Effects of neuronal Kv7 potassium channel
activators on hyperactivity in a rodent model of mania. Behav. Brain Res. 198,
481–485.
Roeloffs, R., Wickenden, A.D., Crean, C., Werness, S., McNaughton-Smith, G., Stables,
J., McNamara, J.O., Ghodadra, N., Rigdon, G.C., 2008. In vivo profile of ICA-27243
Barton, M.E., Klein, B.D., Wolf, H.H., White, H.S., 2001. Pharmacological character-
ization of the 6 Hz psychomotor seizure model of partial epilepsy. Epilepsy Res.
47, 217–227.
Bentzen, B.H., Schmitt, N., Calloe, K., Brown, D.W., Grunnet, M., Olesen, S.P., 2006.
The acrylamide (S)-1 differentially affects Kv7 (KCNQ) potassium channels.
Neuropharmacology 51, 1068–1077.
Bialer, M., Johannessen, S.I., Levy, R.H., Perucca, E., Tomson, T., White, H.S., 2009.
Progress report on new antiepileptic drugs: a summary of the Ninth Eilat
Conference (EILAT IX). Epilepsy Res. 83, 1–43.
Biervert, C., Schroeder, B.C., Kubisch, C., Berkovic, S.F., Propping, P., Jentsch, T.J.,
Steinlein, O.K., 1998.
A potassium channel mutation in neonatal human
epilepsy. Science 279, 403–406.
Blom, S.M., Schmitt, N., Jensen, H.S., 2009. The acrylamide (S)-2 as a positive and
negative modulator of Kv7 channels expressed in Xenopus laevis oocytes. PLoS
One4, 0008251e8251, i:10.1371/journal.pone.
Brickel, N., Gandhi, P., Vanlandingham, K., Hammond, J., Derossett, S., 2012. The
urinary safety profile and secondary renal effects of retigabine (ezogabine): a
first-in-class antiepileptic drug that targets KCNQ (Kv7) potassium channels.
Epilepsia 53, 606–612.
[N-(6-chloro-pyridin-3-yl)-3,4-difluoro-benzamide],
a potent and selective
KCNQ2/Q3 (Kv7.2/Kv7.3) activator in rodent anticonvulsant models. J. Pharma-
col. Exp. Ther. 326, 818–828.
Brown, D.A., Adams, P.R., 1980. Muscarinic suppression of a novel voltage-sensitive
K+ current in a vertebrate neurone. Nature 283, 673–676.
Rundfeldt, C., Netzer, R., 2000a. Investigations into the mechanism of action of the
new anticonvulsant retigabine. Interaction with GABAergic and glutamatergic
neurotransmission and with voltage gated ion channels. Arzneimittelforschung
50, 1063–1070.
Rundfeldt, C., Netzer, R., 2000b. The novel anticonvulsant retigabine activates
M-currents in Chinese hamster ovary-cells tranfected with human KCNQ2/3
subunits. Neurosci. Lett. 282, 73–76.
Schrøder, R.L., Jespersen, T., Christophersen, P., Strøbæk, D., Jensen, B.S., Olesen, S.P.,
2001. KCNQ4 channel activation by BMS-204352 and retigabine. Neurophar-
macology 40, 888–898.
Brown, D.A., Passmore, G.M., 2009. Neural KCNQ (Kv7) channels. Br. J. Pharmacol.
156, 1185–1195.
Butini, S., Gemma, S., Campiani, G., Franceschini, S., Trotta, F., Borriello, M., Ceres, N.,
Ros, S., Coccone, S.S., Bernetti, M., De Angelis, M., Brindisi, M., Nacci, V.,
Fiorini, I., Novellino, E., Cagnotto, A., Mennini, T., Nielsen, K.S., Andreasen, J.T.,
Scheel-Kruger, J., Mikkelsen, J.D., Fattorusso, C., 2009. Discovery of a new class
of potential multifunctional atypical antipsychotic agents targeting dopamine
D3 and serotonin 5-HT1A and 5-HT2A receptors: design, synthesis, and effects
on behavior. J. Med. Chem. 52, 151–169.
Søgaard, R., Ljungstrøm, T., Pedersen, K.A., Olesen, S.P., Jensen, B.S., 2001. KCNQ4
channels expressed in mammalian cells: functional characteristics and phar-
macology. Am. J. Physiol. Cell. Physiol. 280, C859–C866.
Charlier, C., Singh, N.A., Ryan, S.G., Lewis, T.B., Reus, B.E., Leach, R.J., Leppert, M.,
1998. A pore mutation in a novel KQT-like potassium channel gene in an
idiopathic epilepsy family. Nat. Genet. 18, 53–55.

W. Dalby-Brown et al. / European Journal of Pharmacology 709 (2013) 52–63
63
Sotty, F., Damgaard, T., Montezinho, L.P., Mørk, A., Olsen, C.K., Bundgaard, C.,
Husum, H., 2009. Antipsychotic-like effect of retigabine [N-(2-Amino-4-(fluor-
obenzylamino)-phenyl)carbamic acid ester], a KCNQ potassium channel opener,
via modulation of mesolimbic dopaminergic neurotransmission. J. Pharmacol.
Exp. Ther. 328, 951–962.
Suh, B.C., Hille, B., 2002. Recovery from muscarinic modulation of M current
channels requires phosphatidylinositol 4,5-bisphosphate synthesis. Neuron 35,
507–520.
Tatulian, L., Delmas, P., Abogadie, F.C., Brown, D.A., 2001. Activation of expressed
KCNQ potassium currents and native neuronal M-type potassium currents by
the anti-convulsant drug retigabine. J. Neurosci. 21, 5535–5545.
Tober, C., Rostock, A., Rundfeldt, C., Bartsch, R., 1996. D-23129: a potent antic-
onvulsant in the amygdala kindling model of complex partial seizures. Eur. J.
Pharmacol. 303, 163–169.
van Rijn, C.M., Willems-van Bree, E., 2003. Synergy between retigabine and GABA in
modulating the convulsant site of the GABAA receptor complex. Eur. J.
Pharmacol. 464, 95–100.
Wang, H.S., Pan, Z., Shi, W., Brown, B.S., Wymore, R.S., Cohen, I.S., Dixon, J.E.,
McKinnon, D., 1998. KCNQ2 and KCNQ3 potassium channel subunits: molecular
correlates of the M-channel. Science 282, 1890–1893.
Weaver, C.D., Harden, D., Dworetzky, S.I., Robertson, B., Knox, R.J., 2004.
A
thallium-sensitive, fluorescence-based assay for detecting and characterizing
potassium channel modulators in mammalian cells. J. Biomol. Screen. 9,
671–677.
Wickenden, A.D., Krajewski, J.L., London, B., Wagoner, P.K., Wilson, W.A., Clark, S.,
Roeloffs, R., McNaughton-Smith, G., Rigdon, G.C., 2008. N-(6-chloro-pyridin-3-
yl)-3,4-difluoro-benzamide (ICA-27243): a novel, selective KCNQ2/Q3 potas-
sium channel activator. Mol. Pharmacol. 73, 977–986.
Wickenden, A.D., Yu, W., Zou, A., Jegla, T., Wagoner, P.K., 2000. Retigabine, a novel
anti-convulsant, enhances activation of KCNQ2/Q3 potassium channels. Mol.
Pharmacol. 58, 591–600.