(S)-N-[1-(3-morpholin-4-ylphenyl)ethyl]-3-phenylacrylamide: An orally bioavailable KCNQ2 opener with significant activity in a cortical spreading depression model of migraine

Article abstract of DOI:10.1021/jm034073f

(S)-N-[1-(3-Morpholin-4-ylphenyl)ethyl]-3-phenyl-acrylamide (2) was synthesized as an orally bioavailable KCNQ2 potassium channel opener. In a rat model of migraine, 2 demonstrated significant oral activity in reducing the total number of cortical spreading depressions induced by potassium chloride.

Full text of DOI:10.1021/jm034073f

J . Med. Chem. 2003, 46, 3197-3200
3197
Ch a r t 1
(S)-N-[1-(3-Mor p h olin -4-ylp h en yl)eth yl]-
3-p h en yla cr yla m id e: An Or a lly
Bioa va ila ble KCNQ2 Op en er w ith
Sign ifica n t Activity in a Cor tica l
Sp r ea d in g Dep r ession Mod el of Migr a in e
Yong-J in Wu,*,† Christopher G. Boissard,‡
Corinne Greco,‡ Valentin K. Gribkoff,‡
David G. Harden,# Huan He,† Alexandre L’Heureux,|
Shing Hong Kang,§ Gene G. Kinney,‡
Ronald J . Knox,# J oanne Natale,‡ Amy E. Newton,‡
Sanna Lehtinen-Oboma,§ Michael W. Sinz,§
Digavalli V. Sivarao,‡ J ohn E. Starrett, J r.,†
Li-Qiang Sun,† Svetlana Tertyshnikova,‡
Mark W. Thompson,‡ David Weaver,#
Henry S. Wong, Lei Zhang,§ and
i.e., blockers for cognition enhancement2 and openers
for epilepsy.3 Retigabine (Chart 1) has been shown to
activate human KCNQ2, KCNQ2/3, and KCNQ5 medi-
ated currents. The activation was induced by a shift of
the opening threshold of both the homomeric and
heteromeric potassium channels to more negative
potentials.4-6 Flupirtine (Chart 1), a close analogue of
retigabine, also shifts the activation curves toward neg-
ative voltages, but it is less effective than retigabine.7
Both compounds demonstrated significant analgesic
activity on tactile allodynia in Chung lesioned rats,
indicating the therapeutic potential of KCNQ openers
for the treatment of neuropathic pain.7
Steven I. Dworetzky‡
Bristol-Myers Squibb Pharmaceutical Research Institute,
5 Research Parkway, Wallingford, Connecticut 06492, and
Department of Medicinal Chemistry, Bristol-Myers Squibb
Pharmaceutical Research Institute, 100 de l’Industrie
Boulevard, Candiac, Quebec, J 5R 1J 1, Canada
High-throughput screening of the Bristol-Myers Squibb
compound collection using thallium(I) influx assay8
identified (S)-3-phenyl-N-(1-phenylethyl)acrylamide (1)
as a lead, and subsequent lead optimization led to 2
(Chart 1) as an orally bioavailable KCNQ2 opener.
Acrylamide 2 was evaluated in a model of cortical
spreading depression (CSD). CSD is defined as a wave
of brief neuronal excitation followed by a long-lasting
inhibition that originates from a cerebral focal point and
spreads out at a rate of 2-3 mm/min.9,10 Visual aura
and other prodromal symptoms in patients with classic
migraine appear to proceed in a temporal fashion
similar to experimentally induced spreading depression
in animals.11,12 Therefore, it has been suggested that
CSD may underlie prodromal events that precede classic
migraine, particularly visual aura.13 Recently, spreading
depression has been visualized in migraine patients
using functional magnetic resonance imaging.14 Experi-
mentally, CSD has been shown to produce a transient,
localized hyperemia and plasma protein extravasation,
mimicking an inflammatory process.15 This in turn is
believed to sensitize trigeminal primary afferent neu-
rons, leading to enhanced pain perception.15,16 Anti-
migraine drugs are believed to interfere with these
processes.13,14 This lends support to the hypothesis that
CSD may mediate the visual aura, and possibly other
prodomal symptoms,9 that precede classic migraine
attacks and the ensuing pain.15 Thus, compounds that
interrupt cortical spreading depression may be useful
for the treatment of some types of migraine headache.
This report describes the synthesis of acrylamide 2, its
effects on KCNQ2-mediated currents, and its effect on
cortical spreading depression in rats.
Received April 2, 2003
Abstr a ct: (S)-N-[1-(3-Morpholin-4-ylphenyl)ethyl]-3-phenyl-
acrylamide (2) was synthesized as an orally bioavailable
KCNQ2 potassium channel opener. In a rat model of migraine,
2 demonstrated significant oral activity in reducing the total
number of cortical spreading depressions induced by potassium
chloride.
In tr od u ction . Potassium (K+) channels are consid-
ered to be the most diverse class of ion channels with
greater than 60 genes identified. One or more types of
K+ channels are usually present on excitable cell
membranes, where they have critical roles in cell
function. This has been demonstrated in neurons where
they are responsible, in part, for determining cell
excitability by contributing to membrane repolarization
following depolarization, resting membrane potential,
and regulation of neurotransmitter release. The bio-
physical and pharmacological characteristics of different
K+ channels vary widely, contributing to the complex
physiology of excitable cells. Within the past several
years, several members of the KCNQ gene family have
been identified with some high degree of CNS specificity.
These channels are the molecular substrate of the
cholinergically inhibited “M channel”,1 which regulates,
in part, the subthreshold electrical activity of many
different neurons. The M channel has been reported to
consist of heterotetramers of KCNQ2 and KCNQ3
proteins and possibly KCNQ3 and KCNQ5. Modulators
of the M channel have been under clinical investigation,
Ch em istr y. The synthesis of 2 is shown in Scheme
1. Coupling of 3-phenylacrylic acid (3) with (S)-1-(3-
methoxyphenyl)ethylamine provided amide 4. The meth-
oxy group of 4 was cleaved with boron tribromide, and
the resulting phenol was converted to triflate 5. Pal-
ladium-catalyzed coupling of triflate 5 with morpholine
furnished acrylamide 2.
* To whom correspondence should be addressed. Phone: +1-203-
677-7485. Fax: +1-203-677-7702. E-mail: yong-jin.wu@bms.com.
† Department of Neuroscience Chemistry, Wallingford, CT.
‡ Department of Neuroscience Biology, Wallingford, CT.
#
Department of New Leads Biology, Wallingford, CT.
| Department of Medicinal Chemistry, Candiac, Canada.
§ Department of Preclinical Candidate Optimization, Wallingford,
CT.
Department of Chemical Synthesis, Wallingford, CT.
10.1021/jm034073f CCC: $25.00 © 2003 American Chemical Society
Published on Web 06/12/2003

3198 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 15
Letters
Sch em e 1a
a
Reagents: (i) (S)-1-(3-methoxyphenyl)ethylamine, EDAC‚HCl,
DMAP, Et3N, CH2Cl2, 70%; (ii) BBr3, CH2Cl2, -78 °C, 95%; (iii)
Tf2O, pyridine, CH2Cl2, -78 °C, 80%; (iv) morpholine, Pd2(dba)3
(5 mol %), 2-(di-tert-butylphosphino)biphenyl (10 mol %), K3PO4,
DME, 80 °C, 67%.
F igu r e 3. Effect of 2 (10 µM) on mouse KCNQ2 conductance.
Note the “crossover” in the G-V relationship produced by
compound 2, where there is a hyperpolarizing shift in the
voltage dependence of activation until approximately -20 mV
but a pronounced inhibition of KCNQ2 conductance at poten-
tials above -15 mV. Data shown are the mean ( SEM for n
) 5 cells.
increase in outward current from approximately -70
to +10 mV; at higher voltages the compound resulted
in a decrease in the apparent amplitude of the currents.
Observations similar to this were reported in oocytes
expressing hKCNQ2 currents following application of
retigabine.5
F igu r e 1. Effects of 2 on mKCNQ2 currents at a single
membrane voltage (-40 mV) in oocytes (n ) 4-7).
The experiments described above were carried out in
Xenopus oocytes expressing mKCNQ2. However, oocytes
have been shown to express endogenous KCNQ1 chan-
nels,17 and it is unknown if they likewise express other
endogenous family members. Thus, we examined the
effects of 2 in an alternative expression system to rule
out any oocyte-specific actions of 2. Figure 3 shows the
changes in outward conductance of mKCNQ2 stably
expressed in HEK 293 cells produced by local super-
fusion of 2 (10 µM). Compound 2 appears to linearize
the voltage dependence of activation of KCNQ2 in the
membrane potential range (-110 to about -30 mV),
suggesting that 2 may remove the voltage dependence
of activation of KCNQ2 channels. The voltage for half-
activation (V0.5) was -18.3 mV in control conditions and
-52.1 mV in the presence of 10 µM compound 2, a 33.8
mV shift in the hyperpolarizing direction, thus greatly
increasing the open probability (Po) of KCNQ2 channels
at membrane potentials where the channel open prob-
ability is normally close to zero. Under in vivo physi-
ological conditions, one might expect 2 to produce pro-
found hyperpolarization of neurons expressing KCNQ
channels, and indeed, we have shown 2 to hyperpolarize
HEK 293 cells expressing mKCNQ2 (data not shown).
Similar to the oocyte results, compound 2 also inhibited
mKCNQ2 conductance at more positive test potentials,
in this case, at potentials more positive than ap-
proximately -15 mV. It will be interesting to determine
if such a “crossover” in the conductance-voltage rela-
tionship occurs in KCNQ2/heteromultimer combinations
such as KCNQ2/Q3 and KCNQ3/Q5, which are thought
to constitute some native M channels. In general, the
mechanism and physiological significance of facilitated
conductance at negative potentials and inhibited con-
ductance at more positive potentials are unknown but
would depend on the “crossover” voltage of a compound.
Electrophysiological measurements from neurons would
F igu r e 2. Effects of 2 (20 µM) on the mKCNQ2 current-
voltage relationship in oocytes (n)7).
Resu lts a n d Discu ssion . The effects of 2 on outward
potassium current were determined by using two elec-
trode voltage clamp recordings from Xenopus laevis
oocytes expressing cloned mKCNQ2 channels. The
evaluation in Figure 1 was conducted at a single
membrane voltage (-40 mV), and the effects of this
compound on mKCNQ2 current were expressed as the
percent of compound free control current. Figure 1
demonstrates the concentration-response relationship
obtained from the application of compound 2 to oocytes
expressing the mKCNQ2 construct. The estimated EC50
for the compound is 6.0 µM.
Figure 2 describes the current-voltage (I-V) rela-
tionship (-100 to +40 mV) for compound free control
mKCNQ2 currents (b) and for the effects of 2 (20 µM;
O) on the same oocytes. Application of 2 had a marked
effect on the mKCNQ2 I-V relationship. First, applica-
tion of 2 generated a hyperpolarizing shift in the
threshold for current activation from approximately -70
to -85 mV. Second, compound 2 produced a significant

Letters
J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 15 3199
F igu r e 5. Effects of 2 on resting membrane potential in SH-
SY5Y human neuroblastoma cells (n ) 4).
Ta ble 1. Pharmacokinetic Parameters of 2
PK parameters
iv
rata
dogb
dose, mg/kg
1/2, h
Cl, mL/(min‚kg)
dss, L/kg
5
1
t
2.1 ( 0.5
3 ( 0.3
12.7 ( 0.2
3 ( 0.8
41 ( 4
V
2.5 ( 0.6
po (solution)
dose, mg/kg
11
5 ( 1
93
3
F igu r e 4. Dose-dependent effects of 2 on mKCNQ2 currents
expressed in an HEK 293 stable cell line: (A) concentration-
dependent augmentation of KCNQ2 current in a single cell
voltage-clamped at -40 mV; (B) logistic fit to the mean ( SEM
concentration-response data obtained from three cells.
t
max, h
0.5 ( 0.25
100
F, %
a
Compound dosed in Sprague-Dawley rats as a solution in 90%
PEG-400/10% Tween 80 (n ) 3), iv and in 90% PEG-400/10%
b
Tween 80 (n ) 3), po. Compound dosed in Beagle dogs (n ) 3) as
a solution in 50% PEG-400/50% H2O, iv and po.
help to reveal the physiological effects of this voltage-
dependent alteration of KCNQ2 channels.
Ta ble 2. Brain, Plasma, and Brain/Plasma Data of 2
Figure 4 shows the concentration-dependent facilita-
tion of KCNQ2 current from single cells voltage-clamped
at -40 mV. The threshold concentration of compound
2 for augmenting KCNQ2 current was approximately
300 nM. The onset of compound 2 action was fast (1-2
s) and reached a plateau in approximately 15 s. As
shown in Figure 4A, the effect of compound 2 was
readily reversible after application of compound free
control saline (wash). Measurements from three such
experiments were used to construct the concentration-
response curve shown in Figure 4B. The EC50 for this
compound in this system is about 3.3 ( 0.6 µM and
again is in general agreement with the effects in the
oocyte expression system.
The effects of 2 on membrane potential were also
investigated using a fluorescent membrane potential
assay employing voltage-sensitive dyes in SH-SY5Y
human neuroblastoma cells18 expressing native KCNQ
channels. The relative fluorescence intensity change was
used as an indication of the change in membrane
potential.19 As shown in Figure 5, application of 2 led
to a concentration-dependent hyperpolarization of the
cells. A mean EC50 of 0.69 ( 0.08 µM was calculated.
This greater potency, compared to electrophysiology
measurements, likely reflects differences in the cell
types and assays used in the experimentation and
possibly other KCNQ channels found in these cells.
Compound 2 was characterized in several pharmaco-
kinetic studies (Table 1). Oral bioavailability was excel-
lent in both rat and dog, 93% and 100%, respectively.
The favorable oral bioavailability is presumably due to
its good absorption, consistent with the excellent Caco-2
permeability (Pc 260 nm/s) and moderate clearance. The
oral absorption was rapid in the dog, with peak con-
centration occurring within 0.5 h, whereas the com-
rata
dogb
brain concn, ng/g
plasma concn, ng/mL
B/P ratio
631 ( 136
804 ( 279
0.78
538
640
0.84
a
Compound dosed in Sprague-Dawley rats as a solution in
100% PEG-400 at 20 mg/kg, po (n ) 3), and samples were taken
b
at 6 h. Compound dosed in a Beagle dog as a solution in 90%
PEG-400/10% Tween 80 at 3 mg/kg, po (n ) 1), and samples were
taken at 30 min.
pound was slowly absorbed in rat. Acrylamide 2 was
also shown to have moderate brain penetration based
on the samples taken at the approximate tmax in both
species (Table 2).
Compound 2 was examined for its ability to reduce
the total number of CSDs following intragastric (IG)
administration. CSD was produced by a 10 min applica-
tion of crystalline potassium chloride to a small region
in the rat parietal cortex. The total number of depolar-
izations in the distal lead of each animal was used to
determine the mean number of depolarizations per
animal. These were pooled within each treatment group
and analyzed. Since the data were distributed in a non-
Gaussian fashion, intergroup comparison of depolariza-
tions using the nonparametric Kruskal-Wallis test was
performed. As shown in Figure 6, 2 produced a dose-
related reduction in the total number of depolarizations.
Both 10 and 30 mg/kg doses were effective in signifi-
cantly reducing the total number of depolarization
events. Comparison of the proximal depolarizations
between treatment groups also yielded an identical
pattern [vehicle (PEG 400), 36 ( 1; 2, 3 mg/kg, 36 ( 4;
2, 10 mg/kg, 25 ( 4 (p < 0.05); 2, 30 mg/kg, 23 ( 2 (p <
0.001)].
The ratio of proximal to distal depolarizations (P/D
ratio) was 1.07 ( 0.03 in the vehicle-treated control

3200 J ournal of Medicinal Chemistry, 2003, Vol. 46, No. 15
Letters
disorders characterized by hyperexcitability, such as
migraine headache with aura.
Su p p or tin g In for m a tion Ava ila ble: Experimental de-
tails for the synthesis of 2 and the biological assays. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Refer en ces
(1) Wang, H. S.; Pan, Z.; Shi, W.; Brown, B. S.; Wymore, R. S.;
Cohen, I. S.; Dixon, J . E.; McKinnon, D. KCNQ2 and KCNQ3
Potassium Channel Subunits: Molecular Correlates of the
M-Channel. Science (Washington, D.C.) 1998, 282, 1890-1893.
(2) Costa, A. M.; Brown, B. S. Inhibition of M-Current in Cultured
Rat Superior Cervical Ganglia by Linopirdine: Mechanism of
Action Studies. Neuropharmacology 1997, 36, 1747-1753.
(3) Tatulian, L.; Delmas, P.; Abogadie, F. C.; Brown, D. A. Activation
of Expressed KCNQ Potassium Currents and Native Neuronal
M-Type Potassium Currents by the Anticonvulsant Drug Reti-
gabine. J . Neurosci. 2001, 21, 5535-5545.
F igu r e 6. Effect of 2 (IG) on KCl-induced cortical depolar-
izations. Mean depolarizations in the distal lead from indi-
vidual rats organized as columns based on treatment. Hori-
zontal bar represents the group mean ( SEM.
(4) Rundfeldt, C.; Netzer, R. The Novel Anticonvulsant Retigabine
Activates M-Currents in Chinese Hamster Ovary-Cells Trans-
fected with Human KCNQ2/3 Subunits. Neurosci. Lett. 2000,
282, 73-76.
(5) Main, M. J .; Cryan, J .; Dupere, J . R. B.; Cox, B.; Clare, J . J .;
Burbidge, S. A. Modulation of KCNQ2/3 Potassium Channels
by the Novel Anticonvulsant Retigabine. Mol. Pharmacol. 2000,
58, 253-262.
(6) Wickenden, A. D.; Yu, W.; Zou, A.; J egla, T.; Wagoner, P. K.
Retigabine, a Novel Anti-convulsant, Enhances Activation of
KCNQ2/Q3 Potassium Channels. Mol. Pharmacol. 2000, 58,
591-600.
(7) Ilyin, V. I.; Carlin, K. P.; Hodges, D. D.; Robledo, S.; Woodward,
R. M. Flupirtine: A Positive Modulator of Heteromeric KCNQ2/3
Channels. Presented at the 32nd Annual Meeting of the Society
for Neuroscience, Orlando, FL, November 2-7, 2002; Poster
758.10.
group. This ratio showed a dose-related increase across
treatment groups (3 mg/kg, 1.10 ( 0.03; 10 mg/kg, 1.15
( 0.06; 30 mg/kg, 1.22 ( 0.06). The increase in P/D ratio
for the 30 mg/kg treated group was significant in
comparison with the vehicle-treated group (p < 0.05,
Kruskal-Wallis test with Dunn’s posttest).
We have hypothesized that a KCNQ opener would be
efficacious in a cortical model of hyperexcitability by
virtue of increasing K+ efflux from the cells, stabilizing
the cell membrane, and thus making it harder for
depolarizing stimuli to spread. Consistent with this
hypothesis, 2 reduced the incidence of depolarizations
in a dose-related fashion. The highest dose tested
showed a 42% reduction in the incidence of depolariza-
tions. This pattern of dose-related reduction was also
observed with the depolarizing events recorded by the
proximal lead. Moreover, the ratio of the proximal to
distal depolarizations was significantly higher in the 30
mg/kg treated dose. Thus, in addition to reducing the
incidence of cortical depolarizations per se, 2 also
appears to significantly reduce their spread.
(8) Weaver, C. D. Methods for Detecting Modulators of Ion Channels
Using Thallium(I) Sensitive Assays. PCT Int. Appl. WO
0231508A1, 2002.
(9) Gorji, A. Spreading Depression: A Review of the Clinical
Relevance. Brain Res. Rev. 2001, 38, 33-60.
(10) Somjen, G. G. Mechanisms of Spreading Depression and Hypoxic
Spreading Depression-like Depolarization. Physiol. Rev. 2001,
81, 1065-1096.
(11) Lashley, K. S. Pattern of Cerebral Integration Indicated by
Scotomas of Migraine. Arch. Neurol. Psychiatry 1941, 46, 331-
339.
(12) Leao, A. A. P. Spreading Depression of Activity in the Cerebral
Cortex. J . Neurophysiol. 1944, 7, 359-390.
By use of a fluorescent dye in a FLIPR-based screen,8
preliminary evidence showed 2 to activate other mem-
bers of the KCNQ family. Although 2 activated these
channels, further electrophysiological characterization
is required. In this report we focused on the effects of 2
on KCNQ2. Given that 2 is most likely nonselective, the
activity on homomeric KCNQ2 currents still may be
sufficient for in vivo CSD efficacy, since many neuronal
populations contain only the KCNQ2 channel as deter-
mined by in situ hybridization (data not shown). Finally,
we can estimate the effective brain concentration as
close to or less than the in vitro EC50. At a 10 mg/kg IG
dose of 2, the brain concentration is approximately 1
µM, which effectively reduced the total number of CSDs.
The exact effective brain concentration remains to be
established.
(13) Lauritzen, M. Cortical Spreading Depression in Migraine. Ceph-
alalgia 2001, 21, 757-760.
(14) Cao, Y.; Welch, K. M.; Aurora, S.; Vikingstad, E. M. Functional
MRI-BOLD of Visually Triggered Headache in Patients with
Migraine. Arch. Neurol. 1999, 56, 548-554.
(15) Bolay, H.; Reuter, U.; Dunn, A. K.; Huang, Z.; Boas, D. A.;
Moskowitz, M. A. Intrinsic Brain Activity Triggers Trigeminal
Meningeal Afferents in a Migraine Model. Nat. Med. 2002, 8,
136-142.
(16) Moskowitz, M. A. Neurogenic Inflammation in the Patho-
physiology and Treatment of Migraine. Neurology 1993, 43,
S16-S20.
(17) Barhanin, J .; Lesage, F.; Guillemare, E.; Fink, M.; Lazdunski,
M.; Romey, G. K(V)LQT1 and IsK (minK) Proteins Associate To
Form the I(Ks) Cardiac Potassium Current. Nature (London)
1996, 384, 78-80.
(18) Tosetti, P.; Taglietti, V.; Toselli, M. Functional Changes in
Potassium Conductances of the Human Neuroblastoma Cell Line
SH-SY5Y during in Vitro Differentiation. J . Neurophysiol. 1998,
79, 648-658.
(19) Whiteaker, K. L.; Gopalakrishnan, S. M.; Groebe, D.; Shieh, C.
C.; Warrior, U.; Burns, D. J .; Coghlan, M. J .; Scott, V. E.;
Gopalakrishnan, M. Validation of FLIPR Membrane Potential
Dye for High Throughput Screening of Potassium Channel
Modulator. J . Biomol. Screening 2001, 6, 305-312.
Con clu sion . 2 was identified as a KCNQ2 opener
with excellent oral bioavailability in dogs and rats. In
vivo testing of 2 in the CSD model suggests that KCNQ2
openers may have potential for the treatment of CNS
J M034073F