Full text of DOI:10.1038/sj.bjp.0703903

British Journal of Pharmacology (2001) 132, 1154 ± 1160
ã
2001 Nature Publishing Group All rights reserved 0007 ± 1188/01 $15.00
www.nature.com/bjp
E€ects of KRN2391 on ionic currents in rabbit femoral
arterial myocytes
1
1
1Katsuhiko Muraki, Akiko Sasaoka, Minoru Watanabe & *,1Yuji Imaizumi
1Department of Molecular and Cellular Pharmacology, Faculty of Pharmaceutical Sciences, Nagoya City University, Nagoya
467-8603, Japan
1
The e€ects of KRN2391, an ATP-sensitive K+ channel opener (KCO) which also acts as a
nitrate, on ionic membrane currents in rabbit femoral arterial myocytes were examined.
Under whole-cell clamp conditions where cells were superfused with physiological salts solution
2
containing 5.9 mM K+, KRN2391 elicited an outward current at a holding potential of 730 mV.
KRN2391-induced current had a reversal potential of 778 mV and was abolished by glibenclamide
(glib). KRN2391 was approximately 25 times more potent than nicorandil to activate an ATP-
sensitive K+ current (IKATP). On the other hand, 10 mM KRN2391 did not a€ect either voltage-
dependent Ca2+ or delayed recti®er K+ channel currents.
3
In the inside-out patch con®guration, KRN2391 activated 47 pS K+ channels in the presence of
nucleotide diphosphates (NDPs) under the symmetrical 140 mM K+ conditions. Glib and
intracellular ATP reversibly inhibited the activity of the 47 pS K+ channels.
4
The 47 pS K+ channels activated by KRN2391 are similar in their conductance and other
properties to NDP-sensitive K+ channels (KNDP channels) described in other smooth muscles and
the cloned channels. KRN2391 is a potent activator of the 47 pS K+ channels and the activation can
contribute to the KRN2391-induced vasodilation in arterial muscles.
British Journal of Pharmacology (2001) 132, 1154 ± 1160
Keywords: K+ channel opener; KRN2391; ATP-sensitive K+ channels; rabbit femoral artery; nucleotide diphosphates
Abbreviations: ADP, adenosine diphosphate; 4-AP, 4-aminopyridine; ATP, adenosine triphosphate; BK channels, large
conductance Ca2+-dependent K+ channels; gK, single K+ channel conductance; glib, glibenclamide; IBa, inward
Ba2+ current through VDCCs; IK-ATP, ATP-sensitive K+ current; IKds, delayed recti®er K+ currents; KATP
channels, ATP-sensitive K+ channels; KCO, ATP-sensitive K+ channel opener; KNDP channels, NDP-sensitive
K+ channels; KRN2391, N-cyano-N'-(2-nitroxyethyl)-3-pyridinecarboximidamide monomethansulphonate;
NDPs, nucleotide diphosphates; NPo, open-state probability of channels; RFAMs, rabbit femoral arterial
myocytes; TEA, tetraethylammonium; UDP, uridine diphosphate; VDCCs, voltage-dependent Ca2+ channels
Introduction
Opening of K+ channels in vascular smooth muscle cells
causes membrane hyperpolarization toward the equilibrium
potential of K+, resulting in closure of voltage-dependent
Ca2+ channels (VDCCs) and hence vasodilation (Nelson &
Quayle, 1995; Lawson, 1996; Quayle et al., 1997; Kuriyama
et al., 1998). A number of synthetic openers of ATP-sensitive
K+ channels (KATP channels) KCOs, have been developed
for a clinical use in the treatment of systemic hypertension,
coronary angina pectoris and congestive heart diseases
(Weston, 1989; Ligtenberg et al., 1995; Lawson, 1996).
1993; Ogawa, 1994). Since KRN2391 reduces a vascular tone
more potently than nicorandil, it is expected that KRN2391
has more advantages than nicorandil for clinical uses
(Ogawa, 1994). However, despite a large number of studies,
a target channel of KRN2391 has not yet been elucidated.
Based upon biophysical and pharmacological investiga-
tions, it has been proposed that two types of KATP channels
which are sensitive to KCOs are present in smooth muscle
cells (Quayle et al., 1997): small and intermediate conduc-
tance KATP channels (30 ± 50 pS) and large conductance
KATP channels (100 ± 150 pS). NDPs such as ADP, UDP and
GDP e€ectively activate the small and intermediate KATP
channels, and have been described as KNDP channels (Beech
et al., 1993a; Fujita & Kurachi, 2000). Moreover, it has been
found that a channel subunit, Kir 6.1, could form a smooth
muscle KNDP channel with a member of sulphonylurea
receptor, SUR2B (Satoh et al., 1998; Yamada et al., 1997;
Fujita & Kurachi, 2000). It is of importance to identify and
characterize the properties of the channels which are sensitive
to KRN2391 and to compare them with those of KATP
channels in various smooth muscle clones. In the present
study, e€ects of KRN2391 on ionic membrane currents in
rabbit femoral arterial myocytes (RFAMs) were examined
Nicorandil,
a pyridine derivative type KCO, is more
bene®cial than other KCOs because nicorandil is a balanced
arterial- and veno-dilator with a hemodynamic pro®le which
should make it e€ective for the treatment of coronary angina
(Knight et al., 1995). KRN2391, N-cyano-N'-(2-nitroxyethyl)-
3-pyridinecarboximidamide monomethansulphonate, is
a
novel vasodilator which has both a K+ channel opening
action as a KCO and a stimulatory e€ect on guanylate
cyclase as a nitrate (Kashiwabara et al., 1991; Miwa et al.,
*Author for correspondence at: 3-1 Tanabedori, Mizuhoku, Nagoya
467-8603 Japan; E-mail: yimaizum@phar.nagoya-cu.ac.jp

K. Muraki et al
Effects of KRN2391 in femoral arterial myocytes
1155
and the potency to activate ATP-sensitive K+ current (IK-
ATP) was compared with that of nicorandil. The KATP
channel which was a target of KRN2391 was also identi®ed
for the ®rst time and the properties were compared with
those of KATP channels in smooth muscles and the cloned
channels.
developed by Dr John Dempster, University of Strathclyde,
respectively. Membrane current signals were occasionally
printed out using a thermal array recorder (RTA-1200;
Nihon-Koden, Tokyo, Japan). In some experiments, ramp
waveform pulses were applied as a voltage-clamp command
using a multi-pulse generator (FS-1915; NF Electronics,
Tokyo, Japan). All current records were ®ltered at 1 kHz
(4-pole Bessel ®lter, NF Electronics).
Methods
The relative open-state probability of channels (NPo) was
calculated using the following equation:
Cell isolation
N
X
NPo ˆ
ꢀiÃti†=T
All experiments were carried out in accordance with guiding
principles for the care and use of laboratory animals (the
Science and International A€airs Bureau of the Japanese
Ministry of Education, Science, Sports and Culture) and also
with the approval of the ethics committee in Nagoya City
University. Single smooth muscle cells were enzymatically
isolated from the femoral artery of the rabbit. The procedures
for cell-isolation were similar to those previously described
(Imaizumi et al., 1989). Brie¯y, male New Zealand White
rabbits, weighing 1.5 ± 2.0 kg, were anaesthetized with sodium
thiopental (70 mg kg71 i.v.). The isolated femoral artery
(diameter: approximately 100 mm) was cleaned of fat and
connective tissues, and cut open longitudinally. Two pieces of
arterial segments were incubated in Ca2+, Mg2+-free Krebs'
solution (no EGTA) for 20 min at 378C and then placed in
iˆ0
where i is the number of channels open, ti is the time spent by
i channels in the open state, N is the maximum number of
open channels observed in the patch and T is the sampling
time. Since we did not de®ne the total number of channels
present in each patch membrane, we assumed the maximum
number of unitary current levels observed in a patch to be
equal to the number of active channels in the patch. The
single channel data were sampled into a computer by using
the PAT program. Single channel events were detected using
a half amplitude criterion and the all-point amplitude
histogram was ®tted with the Gaussian distribution function.
Solutions
,
the Ca2+ Mg2+-free Krebs' solution containing 0.4%
Ca2+, Ma2+-free Krebs' solution for cell-isolation (mM):
NaCl 112, KCl 4.7, NaHCO3 25, KH2PO4 1.2, glucose 14.
collagenase, 0.1% papain, 0.1% dithiothreitol and 0.1%
trypsin inhibitor for 15 ± 20 min. After the solution was
replaced with Ca2+
The Ca2+
,
Mg2+-free Krebs' solution was gassed with
,
Mg2+-free and enzyme-free Krebs'
mixture of 95% of O2 and 5% CO2. External solution for
whole-cell recording (mM): NaCl 137, KCl 5.9, MgCl2 1.2,
CaCl2 2.2, glucose 14, HEPES 10. To record inward Ba2+
current through VDCCs (IBa), 45 mM NaCl in the external
solution was replaced with 30 mM BaCl2. In some experi-
ments, high K+ external solution (140 mM) was used to
obtain a large IK-ATP. External solution for inside-out patch
con®guration (mM): NaCl 2.9, KCl 140, MgCl2 1.2, glucose
14, HEPES 10, EGTA 1. The pH in each external solution
was adjusted to 7.4 with 10 N NaOH. Pipette solution for
whole-cell recording (mM): KCl 140, MgCl2 4, HEPES 10,
ATP-2Na 1, EGTA 10. When delayed recti®er K+ currents
(Ikds) were recorded, concentration of ATP in the pipette
solution was increased to 5 mM to reduce IK-ATP. To record
IBa, 140 mM KCl was replaced with equimolar CsCl and
5 mM ATP was added to the pipette solution. Pipette solution
for inside-out patch con®guration (mM): NaCl 2.9, KCl 140,
MgCl2 1.2, glucose 14, HEPES 10, CaCl2 2.2. In some
experiments, to reduce the activity of large conductance
Ca2+-dependent K+ channels (BK channels) or VDCCs,
100 ± 300 nM iberiotoxin or 0.1 mM CdCl2, respectively, was
added to the pipette solution. The pH of pipette solutions
was adjusted to 7.2 using 10 N KOH.
solution, the segments were gently agitated with a glass
pipette having a ®re-polished tip.
Electrical recordings
Whole-cell or single channel current recordings were made
from
a single arterial myocyte or a membrane patch,
respectively, using an EPC-7 (List, Germany) or CEZ-2300
ampli®er (Nihon-Koden, Tokyo, Japan). The resistance of
microelectrodes ®lled with the pipette solution was approxi-
mately 2 ± 5 MO for whole-cell and inside-out patch con®g-
urations. The series resistance was partially compensated
electrically under the whole-cell clamp conditions. The cell
capacitance value was determined from the capacity current
elicited by a square hyperpolarizing voltage step from 760 to
770 mV. All experiments were carried out at room
temperature (25+18C).
Data storage and analysis
Membrane currents and voltage signals were monitored on a
storage oscilloscope (VC-6041, Hitachi Tokyo, Japan) and
stored on videotapes after digitized by PCM-recording system
(modi®ed to get DC signal, PCM 501ES; SONY, Tokyo,
Japan). The data on the tape were replayed later and loaded
into a computer (IBM-AT compatible machine) through an
A-D converter (Data translation, DT2801A). Data-acquisi-
tion and analysis for whole-cell and single channel currents
were carried out using AQ/Cell-soft, developed in the
laboratory of Dr Wayne Giles, University of Calgary, and
Single channel current analysis program V7.0C (PAT)
Drugs
The following drugs were used in this study: KRN2391
(supplied from Kirin Brewery Corp., Tokyo, Japan), nicorandil
(Kirin Brewery Corp., Tokyo, Japan), glibenclamide (glib,
Sigma, St Louis, U.S.A.), tetraethylammonium (TEA, Tokyo
Kasei, Tokyo, Japan), 4-aminopyridine (4-AP, Tokyo Kasei,
British Journal of Pharmacology vol 132 (5)

1156
K. Muraki et al
Effects of KRN2391 in femoral arterial myocytes
Tokyo, Japan), adenosine diphosphate (ADP, Oriental, Tokyo,
Japan), adenosine triphosphate (ATP, Oriental, Tokyo, Japan),
uridine diphosphate (UDP, Yamasa, Tokyo, Japan), iberiotox-
in (Peptide Inc., Osaka, Japan).
IKRN2391-out under the physiological K+ gradient was not
large enough at 760 mV for the quantitative analysis, K+
concentration in the external solution was elevated to
140 mM. Cumulative application of nicorandil or KRN2391
at the concentration ranges of 10 ± 1000 mM or 0.01 ± 30 mM,
respectively, elicited inward currents in a concentration-
dependent manner (Figure 2Aa, Ab). These inward currents
were abolished by the addition of 10 mM glib. Figure 2Ba
shows concentration-response relationships of nicorandil and
KRN2391. The current density was obtained by dividing the
amplitude of the inward current with that of each cell-
capacitance and plotted against concentrations of drugs. A
maximum current density induced by nicorandil was not
signi®cantly di€erent from that by KRN2391 (9.2+
2.6 pA pF71, n=6 vs 7.0+1.7 pA pF71, n=7, P40.05).
However, apparent EC50s of nicorandil and KRN2391 to
activate the inward current were 50 and 2 mM, respectively,
indicating that KRN2391 was approximately 25 times more
potent than nicorandil (Figure 2Bb).
E€ects of KRN2391 on IKds and IBa were also examined in
RFAMs. In the presence of 0.1 mM Cd2+, a voltage jump from
a holding potential of 760 mV to potentials between 770 and
+40 mV (for 300 ms) elicited outward currents which were
assumed to be Ikds. The treatment with 10 mM KRN2391 had
little e€ect on IKds (710 mV: 104.4+26.4 pA vs 107.1+
22.0 pA, +10 mV: 207.3+54.2 pA vs 219.0+47.8 pA,
+30 mV: 327.4+87.3 pA vs 347.1+74.9 pA in the presence
and absence of 10 mM KRN2391, respectively, P40.05),
whereas TEA and 4-AP, non-selective K+ channel blockers,
signi®cantly inhibited the Ikds (not shown). Moreover, the
voltage-dependent Ca2+ channel current which was measured
as IBa was not a€ected by 10 mM KRN2391 (IBa at +10 mV,
1.02+0.07 of the control, n=5, P40.05).
Statistics
Pooled data are expressed as mean+s.e.mean in text and ®gures,
and n indicates number of cells used. Statistical signi®cance was
examined using Student's t- or Sche€e's test for two or multiple
groups, respectively. In ®gures, * and ** indicate statistical
signi®cance at P values of 0.05 and 0.01, respectively.
Results
KRN2391-induced membrane current in RFAMs
E€ects of KRN2391 on a holding current across the cell
membrane were examined in RFAMs under a physiological
gradient of K+ concentration (5.9 : 140 mM) in Figure 1.
When
a RFAM was voltage-clamped at a holding of
730 mV, application of 10 mM KRN2391 elicited a slowly
developing outward current (IKRN2391-out, left panel in Figure
1, 44.4+8.1 pA at 730 mV, n=11). Addition of 10 mM glib,
a speci®c inhibitor of KATP channels, to the external solution
abolished the IKRN2391-out. To obtain a current and voltage
(I ± V) relationship of IKRN2391-out, a triangular ramp wave-
form was applied as voltage command pulses before and
during the activation of the current (see the inset in Figure 1).
After subtracting the current obtained at the time
indicated by `1' from that by `2', IKRN2391-out was plotted
against ramp potentials as shown in the right panel in Figure
1. IKRN2391-out reversed at 778.3+3.9 mV (n=5), suggesting
that IKRN2391-out is caused by activation of a K+ current. In
addition, e€ective inhibition of IKRN2391-out by glib indicates
Single channel current activated by KRN2391 in RFAMs
that IKRN2391-out is an IK-ATP
In Figure 2, the potency of KRN2391 to activate IK-ATP
was compared with that of nicorandil. Since the amplitude of
.
Characteristics of the single channel currents activated by
KRN2391 are shown in Figure 3. After the inside-out patch
con®guration was established under the symmetrical high K+
Figure 1 KRN2391-induced outward current in RFAMs. At a holding potential of 730 mV, 10 mM KRN2391 was applied to a
cell in the physiological salt solution containing 5.9 mM K+ followed by 10 mM glib. The pipette was ®lled with solution which
mainly contained 140 mM KCl and 1 mM ATP. The left panel shows a typical current trace. Before and during the activation of the
outward current, triangular ramp waveform pulses (see the inset in right panel) were applied at the time indicated by 1 and 2. A
KRN2391-sensitive current was obtained by subtracting the current recorded at 1 from that at 2 and plotted against the ramp
potentials (the right panel).
British Journal of Pharmacology vol 132 (5)

K. Muraki et al
Effects of KRN2391 in femoral arterial myocytes
1157
Figure 2 Activation of IK-ATP by nicorandil and KRN2391 in high K+ solution. Pipette and external solution contained mainly
140 mM K+. Nicorandil (Aa) or KRN2391 (Ab) was cumulatively applied and thereafter, 10 mM glib was added to the external
solution. The holding potential was 760 mV. (B) Concentration-response relationships of nicorandil (open circles) and KRN2391
(closed circles). Current densities obtained by dividing the current amplitude with each cell-capacitance were plotted against drug-
concentrations (a). The relative current densities normalized to the maximum were expressed in (b).
conditions (140 : 140 mM K+), the simultaneous application
of 10 mM KRN2391 and 1 mM ADP activated a single
channel current whose unitary amplitude was approximately
2 pA at 740 mV (a in Figure 3A). In the lower panel in
Figure 3A, the channel activity was expressed as NPo against
time. The recordings shown in Figure 3A (upper panel) were
obtained at the time shown correspondingly by a, b, c and d
in the lower panel. Exposure to 10 mM KRN2391 without
ADP reduced the channel activity to low levels (b in Figure
3A) and reapplication of both KRN2391 and ADP restored
the activity (c in Figure 3A). The withdrawal of KRN2391 in
the presence of 1 mM ADP abolished the channel activity (d
in Figure 3A).
1 mM NDP (ADP or UDP) were added to the external high
K+ solution. The amplitude of the unitary current was
plotted against each holding potential and the single channel
conductance (gK) was determined by ®tting the data points
to a regression line. Averaged gKs were 47.0+1.9 pS (n=5);
47.4+2.6 pS (n=3; P40.05), when cells were superfused
with solution containing KRN2391 plus either 1 mM ADP or
UDP, respectively. Additionally, both I ± V relationships were
reversed at around 0 mV.
E€ects of glib and ATP on the activity of the KRN2391-
sensitive K+ channel were examined in the inside-out patch
con®guration (Figure 5). The same pipette and external
solutions which mainly contained 140 mM K+ as those in
Figures 3 and 4 were used. As shown in Figure 5Aa, the
presence of 10 mM KRN2391 and 1 mM UDP induced the
channel activity whose unitary current was approximately
1.3 pA at a holding potential of +20 mV. Occasionally,
In Figure 3B, the channel activity under various experi-
mental conditions is summarized. NPo over 30 s was
calculated and the pooled data averaged. Under control
conditions in the absence of either KRN2391 or a NDP, the
channel
activity
was
consistently
low
(NPo:
transient openings of BK channels which had a large
0.00032+0.00032, n=5). Moreover, the application of
10 mM KRN2391 alone had little e€ect on the channel
activity (NPo: 0.00039+0.00039, n=7). Addition of 1 mM
ADP in the presence of KRN2391 markedly increased the
NPo (0.14+0.05, n=6; P50.05 vs control). The addition of
1 mM UDP in the presence of KRN2391 was also e€ective in
a manner similar to that of ADP (0.24+0.06, n=3, P50.01
vs control). In contrast, 1 mM ADP itself did not a€ect
channel activity in the absence of KRN2391 (NPo:
0.0023+0.0023, n=2).
unitary current (45 pA) were detected (large transient lines
in the upper panel in Figure 5A). In a lower panel in
Figure 5A, the channel activity expressed as NPo was
plotted against time. Upon the addition of 10 mM glib to
the external solution, the channel activity almost disap-
peared (Figure 5Ab). However, the withdrawal of glib
followed by the reapplication of 10 mM KRN2391 and
1 mM UDP reactivated the channel (Figure 5Ac). On the
other hand, as illustrated in Figure 5Bb, the addition of
intracellular 1 mM ATP reduced the channel activity, which
was stimulated by the presence of 10 mM KRN2391 and
1 mM UDP. The removal of ATP restored the channel
activity (Figure 5Bc).
Figure 4 shows single channel current recordings at
di€erent potentials (A) and the I ± V relationships of the
KRN2391-sensitive channel (B). Both 10 mM KRN2391 and
British Journal of Pharmacology vol 132 (5)

1158
K. Muraki et al
Effects of KRN2391 in femoral arterial myocytes
Discussion
The present study clearly shows that KRN2391 activates glib-
sensitive currents approximately 25 times more potently than
nicorandil in RFAMs. KRN2391, however, had neither an
e€ect on IBa through VDCCs nor on IKds. A target K+
channel activated by KRN2391 is susceptible to glib, NDPs
and ATP, strongly suggest that KRN2391-sensitive KATP
channels have similar biophysical and pharmacological
properties to those of KNDP channels in other smooth
muscles.
KRN2391-induced outward currents in RFAMs
It had been expected that KRN2391 would open KATP
channels as a KCO and dilate vascular smooth muscles
(Kashiwabara et al., 1991; Miwa et al., 1993; Ogawa, 1994).
The KRN2391-induced vasodilation was e€ectively blocked
by glib (Kashiwabara et al., 1991; Ogawa, 1994). It has been
reported that KRN2391-induced relaxation was more ex-
tensive when vessels were pre-contracted with hormonal
agonists than when pre-contracted with high K+ solution
(Kashiwabara et al., 1991). Additionally, an increase in
86Rb+-e‚ux from vessels and the hyperpolarization of cells
have been observed in the presence of KRN2391 (Kashiwa-
bara et al., 1991; Okada et al., 1993). In the present study,
under the whole-cell clamp conditions where cells were
superfused with the solution containing 5.9 mM K+,
application of KRN2391 elicited an outward current at the
holding potential of 730 mV. The outward current had a
reversal potential of 778 mV which was close to an
equilibrium potential of K+ under the physiological condi-
tions and 0 mV in symmetrical high K+ solution, and was
abolished by an exposure to glib, directly indicating that
KRN2391 activates IK-ATP and acts as a KCO in rabbit
femoral artery.
Figure 3 (A) KRN2391-sensitive 47 pS K+ channel current in
RFAMs. After the establishment of inside-out patch con®guration,
10 mM KRN2391 and 1 mM ADP were applied to a membrane patch
voltage-clamped at 740 mV. The upper panel shows the single
channel current traces recorded at the corresponding time indicated
by a, b, c and d in the lower panel. (B) Summarized data describing
the activity of KRN2391-sensitive 47 pS K+ channels under various
experimental conditions. Number in parentheses indicates the number
of cells used.
Figure 4 I ± V relationships of KRN2391-sensitive 47 pS K+ channels in the inside-out patch con®guration. In the presence of
10 mM KRN2391 and NDPs (1 mM ADP: circles, 1 mM UDP: triangles), single channel currents were recorded at various holding
potentials (A) and these amplitudes plotted against each potential (B).
British Journal of Pharmacology vol 132 (5)

K. Muraki et al
Effects of KRN2391 in femoral arterial myocytes
1159
the presence of KRN2391 may mask the e€ect on Ikds under
the present experimental conditions even though 5 mM ATP
was added to the pipette solution.
Characteristics of KRN2391-sensitive single channel
current
KATP channels which were originally found in cardiac
myocytes are widely distributed in tissues including pancrea-
tic b-cells, skeletal muscle, smooth muscle and neurons.
Despite the similarity in pharmacology, KATP channels in
vascular smooth muscle exhibits di€erent single channel
characteristics and
a distinct mechanism of nucleotide
regulation (Kajioka et al., 1991; Beech et al., 1993a,b; Zhang
& Bolton, 1996; Fujita & Kurachi, 2000). Vascular KATP
channels have less than one-half of the single channel
conductance of the classical KATP channels observed in
cardiac and skeletal muscle cells or pancreatic b-cells.
Moreover, the classical KATP channels open spontaneously
when intracellular ATP is removed, while the vascular KATP
channels require NDPs to open. Consistently, KATP channels
in RFAMs had a 47 pS conductance under the symmetrical
140 mM K+ conditions and their activity was NDP-
dependent, indicating that KATP channels in RFAMS are a
certain type of KNDP channels. Although such di€erences
between the vascular and the classical KATP channels causes
some confusion about the identity of the vascular KATP
channels, it has been found that a channel subunit, Kir 6.1,
may form a smooth muscle KNDP channel with a member of
sulphonylurea receptor, SUR2B (Satoh et al., 1998; Yamada
et al., 1997; Fujita & Kurachi, 2000). Co-expression of
SUR2B and Kir6.1 formed a functional K+ channel with the
features of a KNDP channel: a single channel conductance of
33 pS in symmetric 145 mM K+ solution, NDP-dependent
activation and no spontaneous activation in the absence of
internal ATP (Yamada et al., 1997; Satoh et al., 1998).
Pharmacological properties of KATP channels that respond
to KCOs may be determined by SURs coupled to a Kir
channel subunit. When the Kir6.2 subunit was co-expressed
with SUR2A or SUR2B, the EC50 of nicorandil in activating
the KATP channels was 500 and 10 mM, respectively (Shindo
et al., 1998). On the other hand, the EC50 of nicorandil to
activate IK-ATP in RFAMs was 50 mM. Hence, it is expected
that mainly SUR2B is expressed in RFAMs, while expression
of di€erent SUR isoform cannot be ruled out. Further
molecular biological studies are necessary to elucidate the
type of Kir channel as well as SUR subunits in RFAMs.
In conclusion, KRN2391 potently activates IK-ATP in
rabbit femoral artery, whereas KRN2391 a€ects neither
voltage-dependent Ca2+ nor delayed recti®er K+ channel
currents. The potency of KRN2391 as a KCO is signi®cantly
higher (approximately 25 times) than that of nicorandil. The
47 pS K+ channels which are activated by KRN2391 are
susceptible to glib, ATP and NDPs, and therefore underlie
the IK-ATP observed under whole-cell clamp conditions. The
potent opening of KATP channels by KRN2391 most likely
contributes to the vasodilation by KRN2391.
Figure 5 E€ects of glib (A) and ATP (B) on the activity of
KRN2391-sensitive 47 pS K+ channels. The channel activity was
induced by application of 10 mM KRN2391 and 1 mM UDP in the
inside-out-patch con®guration. The lower panels in (A) and (B)
indicate NPo over 30 s, which was plotted against time. The traces
illustrated in the upper panels were obtained at the corresponding
times indicated by a, b and c in the lower panels, respectively.
In porcine coronary artery, the relaxing e€ect of KRN2391
against agonist-induced contractions was 10 ± 20 times greater
than that of nicorandil (Fukata et al., 1993; Kasai et al.,
1993; Jinno et al., 1992; Miwa et al., 1993). In the present
study, KRN2391 and nicorandil elicited a glib-sensitive
current in a concentration-dependent manner. The maximum
amplitude of the current obtained in the presence of
KRN2391 was similar to that of nicorandil (7 ± 9 pA pF71),
whereas EC50s of KRN2391 and nicorandil were approxi-
mately 2 and 50 mM, respectively, clearly indicating that
KRN2391 is 25 times more potent than nicorandil to activate
IK-ATP as a KCO.
No inhibitory effects of KRN2391 on IBa and IKds
Cromakalim, the ®rst KCO, has inhibitory e€ects on VDCCs
and Ikds in vascular smooth muscle cells (Okabe et al., 1990).
It is, therefore, possible that multiple actions of KRN2391 on
ion channels contribute to its various pharmacological e€ects
on the vasculature. On the other hand, 10 mM KRN2391 did
not a€ect IBa under the conditions where 30 mM Ba2+ was
used as the charge carrier through VDCCs. Moreover, IKd
s
were not a€ected by treatment with 10 mM KRN2391.
However, we cannot completely rule out the possibility that
KRN2391 acts on Ikds because a part of IK-ATP activated in
The authors thank Dr W. Giles and Dr J. Dempster for supplying
data-acquisition and analysis software. KRN2391 and nicorandil
were kindly supplied from Kirin Brewery Co. Ltd (Tokyo, Japan).
British Journal of Pharmacology vol 132 (5)

1160
K. Muraki et al
Effects of KRN2391 in femoral arterial myocytes
References
BEECH, D.J., ZHANG, H., NAKAO, K. & BOLTON, T.B. (1993a). K
channel activation by nucleotide diphosphates and its inhibition
by glibenclamide in vascular smooth muscle cells. Br. J.
Pharmacol., 110, 573 ± 582.
BEECH, D.J., ZHANG, H., NAKAO, K. & BOLTON, T.B. (1993b). Single
channel and whole-cell K-currents evoked by levcromakalim in
smooth muscle cells from the rabbit portal vein. Br. J.
Pharmacol., 110, 583 ± 590.
LIGTENBERG, J.J., VAN HAEFTEN, T.W., LINKS, T.P., SMIT, A.J. &
REITSMA, W.D. (1995). Clinical relevance of ATP-dependent
potassium channels. Neth. J. Med., 47, 241 ± 251.
MIWA, A., KANETA, S., MOTOKI, K., JINNO, Y., KASAI, H., OKADA,
Y., FUKUSHIMA, H. & OGAWA, N. (1993). Vasorelaxant
mechanism of KRN2391 and nicorandil in porcine coronary
arteries of di€erent sizes. Br. J. Pharmacol., 109, 632 ± 636.
NELSON, M.T.
& QUAYLE, J.M. (1995). Physiological roles and
FUJITA, A. & KURACHI, Y. (2000). Molecular aspects of ATP-
sensitive K+ channels in the cardiovascular system and K+
channel openers. Pharmacol. Ther., 85, 39 ± 53.
FUKATA, Y., KANETA, S., OKADA, Y., YOKOYAMA, T., JINNO, Y.,
FUKUSHIMA, H. & OGAWA, N. (1993). Mechanism of action of
KRN2391 in canine coronary vascular bed. Jpn. J. Pharmacol.,
63, 305 ± 311.
IMAIZUMI, Y., MURAKI, K. & WATANABE, M. (1989). Ionic currents
in single smooth muscle cells from the ureter of the guinea-pig. J.
Physiol. (Lond.), 411, 131 ± 159.
JINNO, Y., KASAI, H., OHTA, H., NISHIKORI, K., FUKUSHIMA, H. &
OGAWA, N. (1992). Contribution of cyclic GMP formation to
KRN2391-induced relaxation in coronary artery of the pig. Br. J.
Pharmacol., 106, 906 ± 909.
properties of potassium channels in arterial smooth muscle. Am.
J. Physiol., 268, C799 ± C822.
OGAWA, N. (1994). Pharmacological properties of KRN2391, a
novel vasodilator of the nitrate-potassium channel opener hybrid
type. Gen. Pharmacol., 25, 609 ± 616.
OKABE, K., KAJIOKA, S., NAKAO, K., KITAMURA, K., KURIYAMA,
H. & WESTON, A.H. (1990). Actions of cromakalim on ionic
currents recorded from single smooth muscle cells of the rat
portal vein. J. Pharmacol. Exp. Ther., 252, 832 ± 839.
OKADA, Y., YANAGISAWA, T. & TAIRA, N. (1993). BRL 38227
(levcromakalim)-induced hyperpolarization reduces the sensitiv-
ity to Ca2+ of contractile elements in canine coronary artery.
Naunyn Schmiedebergs Arch. Pharmacol., 347, 438 ± 444.
QUAYLE, J.M., NELSON, M.T.
& STANDEN, N.B. (1997). ATP-
KAJIOKA, S., KITAMURA, K. & KURIYAMA, H. (1991). Guanosine
diphosphate activates an adenosine 5'-triphosphate-sensitive K+
channel in the rabbit portal vein. J. Physiol. (Lond.), 444, 397 ±
418.
sensitive and inwardly rectifying potassium channels in smooth
muscle. Physiol. Rev., 77, 1165 ± 1232.
SATOH, E., YAMADA, M., KONDO, C., REPUNTE, V.P., HORIO, Y.,
IIJIMA, T.
& KURACHI, Y. (1998). Intracellular nucleotide-
KASAI, H., JINNO, Y., KANETA, S., FUKATA, Y., FUKUSHIMA, H. &
OGAWA, N. (1993). Comparison of the e€ects of KRN2391 and
other coronary dilators on porcine isolated coronary arteries of
di€erent sizes. J. Pharm. Pharmacol., 45, 573 ± 575.
KASHIWABARA, T., NAKAJIMA, S., IZAWA, T., FUKUSHIMA, H. &
NISHIKORI, K. (1991). Characteristics of KRN2391, a novel
vasodilator, compared with those of cromakalim, pinacidil and
nifedipine in rat aorta. Eur. J. Pharmacol., 196, 1 ± 7.
KNIGHT, C., PURCELL, H. & FOX, K. (1995). Potassium channel
openers: clinical applications in ischemic heart disease ± overview
of clinical ecacy of nicorandil. Cardiovasc. Drugs Ther., 9
(Suppl 2), 229 ± 236.
KURIYAMA, H., KITAMURA, K., ITOH, T. & INOUE, R. (1998).
Physiological features of visceral smooth muscle cells, with
special reference to receptors and ion channels. Physiol. Rev., 78,
811 ± 920.
LAWSON, K. (1996). Is there a therapeutic future for `potassium
channel openers'? [editorial]. Clin. Sci. (Colch.), 91, 651 ± 663.
mediated gating of SUR/Kir6.0 complex potassium channels
expressed in a mammalian cell line and its modi®cation by
pinacidil. J. Physiol. (Lond.), 511, (Pt 3), 663 ± 674.
SHINDO, T., YAMADA, M., ISOMOTO, S., HORIO, Y. & KURACHI, Y.
(1998). SUR2 subtype (A and B)-dependent di€erential activa-
tion of the cloned ATP-sensitive K+ channels by pinacidil and
nicorandil. Br. J. Pharmacol., 124, 985 ± 991.
WESTON, A.H. (1989). Smooth muscle K+ channel openers; their
pharmacology and clinical potential. P¯ugers Arch., 414 (Suppl
1), S99 ± S105.
È
YAMADA, M., ISOMOTO, S., MATSUMOTO, S., KONDO, C., SHINDO,
T., HORIO, Y. & KURACHI, Y. (1997). Sulphonylurea receptor 2B
and Kir6.1 form a sulphonylurea-sensitive but ATP-insensitive
K+ channel. J. Physiol. (Lond.), 499, 715 ± 720.
ZHANG, H.L. & BOLTON, T.B. (1996). Two types of ATP-sensitive
potassium channels in rat portal vein smooth muscle cells. Br. J.
Pharmacol., 118, 105 ± 114.
(Received September 12, 2000
Revised November 22, 2000
Accepted December 13, 2000)
British Journal of Pharmacology vol 132 (5)