Dualistic actions of cromakalim and new potent 2H-1,4-benzoxazine derivatives on the native skeletal muscle KATP channel

Article abstract of DOI:10.1038/sj.bjp.0705233

1. New 2H-1,4-benzoxazine derivatives were synthesized and tested for their agonist properties on the ATP-sensitive K⁺ channels (K_ATP) of native rat skeletal muscle fibres by using the patch-clamp technique. The novel modifications i

Full text of DOI:10.1038/sj.bjp.0705233

British Journal of Pharmacology (2003) 139, 255–262
&
2003 Nature Publishing Group All rights reserved 0007–1188/03
$25.00
www.nature.com/bjp
Dualistic actions of cromakalim and new potent 2H-1,4-
benzoxazine derivatives on the native skeletal muscle KATP channel
1
2
2
1Domenico Tricarico, Mariagrazia Barbieri, Laghezza Antonio, Paolo Tortorella,
2Fulvio Loiodice & *,1Diana Conte Camerino
1Department of Pharmacobiology, Faculty of Pharmacy, via Orabona no. 4, University of Bari, I-70126 Bari, Italy and
2Department of Medicinal Chemistry, Faculty of Pharmacy, via Orabona no. 4, University of Bari, I-70126 Bari, Italy
1
New 2H-1,4-benzoxazine derivatives were synthesized and tested for their agonist properties on the
ATP-sensitive K+ channels (KATP) of native rat skeletal muscle fibres by using the patch-clamp
technique. The novel modifications involved the introduction at position 2 of the benzoxazine ring of
alkyl substituents such as methyl ( – CH3), ethyl ( – C2H5) or propyl ( – C3H7) groups, while
maintaining pharmacophore groups critical for conferring agonist properties.
2
The effects of these molecules were compared with those of cromakalim in the presence or absence
of internal ATP (10ꢀ4 M). In the presence of internal ATP, all the compounds increased the
macropatch KATP currents. The order of potency of the molecules as agonists was ꢀC3H7
(DE50 ¼ 1.63 Â 10ꢀ8 M) 4ꢀC2H5 (DE50 ¼ 1.11 Â 10ꢀ7 M)4 – CH3 (DE50 ¼ 2.81 Â 10ꢀ7 M)4cromak-
slim (DE50 ¼ 1.42 Â 10ꢀ5 M). Bell-shaped dose – response curves were observed for these compounds
and cromakalim indicating a downturn in response when a certain dose was exceeded.
3
In contrast, in the absence of internal ATP, all molecules including cromakalim inhibited the KATP
currents. The order of increasing potency as antagonists was cromakalim (IC50 ¼ 1.15 Â 10ꢀ8 M)X
– CH3 (IC50 ¼ 2.6 Â 10ꢀ8 M)4 – C2H5 (IC50 ¼ 4.4 Â 10ꢀ8 M)4 – C3H7 (IC50 ¼ 1.68Â 10ꢀ7 M) derivatives.
4
These results suggest that the newly synthesized molecules and cromakalim act on muscle KATP
channel by binding on two receptor sites that have opposite actions. Alternatively, a more simple
explanation is to consider the existence of a single site for potassium channel openers regulated by
ATP which favours the transduction of the channel opening. The alkyl chains at position 2 of the 2H-
1,4-benzoxazine nucleus is pivotal in determining the potency of benzoxazine derivatives as agonists or
antagonists.
British Journal of Pharmacology (2003) 139, 255 – 262. doi:10.1038/sj.bjp.0705233
Keywords: KCO; 2H-1,4-benzoxazine; cromakalim; KATP channel; skeletal muscle; patch clamp
Abbreviations: – CH3, methyl; – C2H5, ethyl; – C3H7, propyl; DMSO, dimethylsulphoxide; FDB, flexor digitorum brevis; KATP
,
ATP-sensitive K+ channel; KCOs, potassium channel openers; Kir6.2, inward rectifier K+ channel 6.2 subunit;
log D, logarithm of the distribution coefficient at a particular pH; log P, logarithm of the partition coefficient;
pKa, negative logarithm of the dissociation constant; SUR2A, sulphonylureas receptor 2A subunit
Introduction
The K+ channel openers (KCOs) are a structurally diverse
group of drugs showing a broad spectrum of potentially
therapeutic applications including asthma, urinary incontinence,
hypertension, angina, hypoglycaemia, neuromuscular disorders
and some forms of epilepsy (Longman & Hamilton, 1992;
Schwanstecher et al., 1998). First and second generation KCOs
have been synthesised and tested for their capability to open the
ATP-sensitive K+ channel (KATP) of different tissues leading to
hyperpolarisation of the cells and reduction of the influx of Ca2+
ions through voltage-dependent Ca2+ channels (Lawson, 2000).
Pharmacological investigations have shown that cromaka-
lim, a first-generation KCO, can be effective in the treatment
of neuromuscular disorders. This molecule is able to repolarise
the muscle fibers of hypokalemic periodic paralyses (hypoPP)
patients as well as of K+-depleted rats, an animal model of
hypoPP, and in some myotonic patients is also able to suppress
‘in vitro’ the abnormal hyperexcitability of the fibers (Quasth-
off et al., 1990; Tricarico et al., 1998, 1999). However, the use
of first generation KCOs in skeletal muscle disorders is limited
by their lack of tissue selectivity that in turn is related to their
side effects such as hypotension, hypertricosis and headache
(Andersson, 1992). Knowledge of the tissue-selective expres-
sion of various sulphonylurea receptor (SURs) subunits and
pore forming subunits (kirs) of the KATP channel complexes
has made it possible to investigate tissue-selective KCOs.
However, to date molecules targeting the skeletal muscle KATP
channels are not known.
In order to search for molecules more potent than
cromakalim in activating the muscular KATP channel and
to better investigate their mechanism of action, a series of new
2H-1,4-benzoxazine derivatives have been synthesised.
The novel structural modifications that we performed in the
2H-1,4-benzoxazine nucleus involved the replacement of
one hydrogen atom at position 2 on the benzoxazine ring
with different alkyl substituents such as methyl ( – CH3) ethyl
*Author for correspondence; E-mail: conte@farmbiol.uniba.it

256
D. Tricarico et al
Pharmacology of muscle KATP channel
( – C2H5) or propyl ( – C3H7) groups. This gave us the
opportunity to evaluate the effects of the lipophilicity on
muscular KATP channel activity.
5 Â 10ꢀ3 M glucose and 1 Â 10ꢀ2 M 3-(N-morpholino)propane-
sulphonic acid (MOPS), pH ¼ 7.2. The patch pipette solution
contained 150 Â 10ꢀ3 M KCl, 2 Â 10ꢀ3 M CaCl2 and 1 Â 10ꢀ2
MOPS, pH ¼ 7.2. The bath solution contained 150 Â 10ꢀ3
M
M
The 2H-1,4-benzoxazine derivatives are known to be
structural analogues of benzopyrans, the difference being the
KCl, 5 Â 10ꢀ3 M EGTA and 1 Â 10ꢀ2 M MOPS, pH ¼ 7.2.
Stock solution of ATPK2 (5 Â 10ꢀ3 M) was prepared by
dissolving the chemical in the bath solution. Stock solutions
of cromakalim (7.6 Â 10ꢀ1 M) and of the benzoxazine deriva-
tives (7.6 Â 10ꢀ1 M) were prepared by dissolving the drugs in
dimethylsulphoxide (DMSO). Microliter amounts of the stock
solutions were then added to the bath solution as needed to
obtain concentrations of cromakalim ranging between 10ꢀ12
and 8 Â 10ꢀ4 M, and of the benzoxazine derivatives between
10ꢀ12 and 10ꢀ4 M (DMSO 1.3 Â 10ꢀ8 – 0.05%). Higher con-
centrations could not be tested because of the low solubility of
the substances in the bath solution. The drugs were tested in
the presence or absence of an internal 10ꢀ4 M concentration of
ATPK2.
replacement of the carbon atom at position
4 of the
benzopyran ring with a nitrogen atom. Structure – activity
relations performed on isolated organs have shown that the
introduction of the nitro or amino group in positions 6 and 7
of the 2H-1,4-benzoxazine nucleus, respectively, or the
synthesis of tricyclic derivatives by condensing an oxadiazole
ring to the same part of the nucleus give rise to molecules more
potent than cromakalim or levcromakalim in relaxing the
arterial smooth muscle or in reducing the arterial mean blood
pressure (Matsumoto et al., 1996, 1999, 2000; Caliendo et al.,
1998). In most cases, the vasodilating effects of these
compounds were antagonised ‘in vitro’ by glibenclamide, the
well known KATP channel blocker, suggesting that the KATP
channel is the main target of the action of these molecules in
smooth muscle cells. Furthermore, 2H-1,4-benzoxazine deri-
vatives can be considered cyclic analogues of the anilide
tertiary carbinols (Grant et al., 1994; Matsumoto et al., 1996,
1999; Caliendo et al., 1998). These latter open chain KCOs are
known for their relaxant effects on the bladder and urethra
smooth muscle (Grant et al., 1994). However, the mechanism
of action of these KCOs appears to be complex showing
stimulatory and inhibitory responses (Jow & Numann, 1999;
Teramoto et al., 2001). Similarly pinacidil, an open chain
KCO derivative, activates or inhibits the KATP channels in
skeletal muscle depending on the level of the KATP channel
activity (Hehl & Neumcke, 1994). In addition, inhibitory
responses of diverse K+ channels have been observed with
structural analogues of cromakalim, such as the chromanols
that inhibit KCNQ1 channels, reducing the IKs currents in
heart cells by a stereospecific interaction with the pore-forming
subunit (Seebohm et al., 2001).
DMSO applied at 0.05% concentration to the excised
a M concentration of
10ꢀ4
patches in the presence of
ATPK2 did not increase the KATP channel activity (solvent
control).
We prefer to use ATPK2 instead of MgATP to avoid the
effects of second messengers such as PIP2 or related molecules
and pumps such as the 3Na+/2 K+ ATP-ase, which are known
to induce activation of the KATP channel in excised patches in
the presence of MgATP (Kabakov, 1998; Song & Ashcroft,
2001). Furthermore, it is known that cromakalim and other
KCOs are able to stimulate the native skeletal muscle KATP
channels even in the absence of internal Mg2+ ions (Forestier
et al., 1996).
Synthesis of the new benzoxazine derivatives
The starting compounds were (7)-2-alkyl-2-(4-chloro-phe-
noxy) acetic acids, the preparation of which has been
previously reported (Bettoni et al., 1987, 1992). These acids
were treated with 90% HNO3 at 01C for 2 – 3 h to give the 2-
nitro benzene derivatives which were reduced and cycled to the
corresponding benzoxazinones with iron and HCl (6 N) in
refluxing 1,4-dioxane for 2 – 3 h. The condensation of the
benzoxazinones with 3-aminopyridine, in the presence of TiCl4
and anisole, was carried out in refluxing dry toluene for 6 –
12 h and provided the desired compounds in 10 – 40% yields
(Fryer et al., 1969). The physical properties of the final drugs
are as follows:
In the present work, macropatch KATP currents were
recorded using a patch-clamp technique in the presence of
cromakalim or in the presence of newly synthesised 2H-1,4-
benzoxazine derivatives. The potency of the 2H-1,4-benzox-
azine derivatives relative to cromakalim was evaluated by
constructing dose – response curves for the compounds under
investigation in the presence of internal ATP or in the absence
of the nucleotide.
(R/S)-6-chloro-2-methyl-3-(pyridin-3-yl-amino)-2H-1,4-ben-
1
zoxazine maleate salt. m.p. 148 – 1501C. H-NMR (300 MHz,
Methods
Muscle preparations and single fibre isolation
DMSO-d6): d 1.32 (d, 3H, CH3); 4.92 (q, 1H, CH); 6.23 (s, 2H,
CH ¼ CH); 6.87(d, 1H, benzenic proton); 6.96 (dd, 1H,
benzenic proton); 7.10 (d, 1H, benzenic proton); 7.43 (q, 1H,
pyridinic proton); 8.28 (d, 1H, pyridinic proton); 8.36 (d, 1H,
pyridinic proton); 8.96 (s, 1H, pyridinic proton); 9.80 (bb, 2H,
The flexor digitorum brevis (FDB) muscles were dissected
from male Wistar rats under urethane anaesthesia (1.2 g kgꢀ1).
After dissection, the animals were rapidly killed with an
overdose of urethane according to the ‘Guide for Care and Use
of Laboratory Animals’ prepared by the National Academy of
Sciences. Single muscle fibers were prepared by enzymatic
dissociation (Tricarico et al., 1998).
+
NH+COOH, D2O exchanged). GC – MS, m/z: 27 3 (M , 100);
275 (M++2, 32).
(R/S)-6-chloro-2-ethyl-3-(pyridin-3-yl-amino)-2H-1,4-ben-
1
zoxazine maleate salt. m.p. 149 – 1511C. H-NMR (300 MHz,
DMSO-d6): d 0.95 (t, 3H, CH3); 1.63 (m, 2H, CH2); 3.00 – 5.00
(bb, 2H, NH+COOH, D2O exchanged); 4.68 (t, 1H, CH); 6.22
(s, 2H, CH ¼ CH); 6.89 (d, 1H, benzenic proton); 6.95 (dd, 1H,
benzenic proton); 7.09 (d, 1H, benzenic proton); 7.43 (q, 1H,
pyridinic proton); 8.27(d, 1H, pyridinic proton); 8.37(d, 1H,
Drugs and solutions
The normal Ringer solution contained 145 Â 10ꢀ3 M NaCl,
5 Â 10ꢀ3
M
KCl, 1 Â 10ꢀ3
M
MgCl2, 0.5 Â 10ꢀ3
M
CaCl2,
British Journal of Pharmacology vol 139 (2)

D. Tricarico et al
Pharmacology of muscle KATP channel
257
pyridinic proton); 8.97(s, 1H, pyridinic proton). GC – MS,
m/z: 287(M +, 100); 289 (M++2, 33).
Patch-clamp experiments
(R/S)-6-chloro-2-propyl-3-(pyridin-3-yl-amino)-2H-1,4-ben-
zoxazine (free base). m.p. 180 – 2001C. 1H-NMR (300 MHz,
DMSO-d6): d 0.86 (t, 3H, CH3); 1.30 – 1.70 (m, 4H, CH2 –
CH2); 4.66 (dd, 1H, CH); 6.69 (d, 1H, benzenic proton); 6.78
(dd, 1H, benzenic proton); 6.98 (d, 1H, benzenic proton); 7.22
(q, 1H, pyridinic proton); 8.15 (d, 1H, pyridinic proton); 8.36
(d, 1H, pyridinic proton); 8.81 (s, 1H, pyridinic proton); 9.30
(bb, 1H, NH, D2O exchanged). GC – MS, m/z: 301 (M+, 68);
303 (M++2, 24); 259 (100).
Experiments were performed in inside-out configurations
using the standard patch-clamp technique. Recordings of
channel currents were performed during voltage steps of 6 s
going from 0 mV of holding potential to ꢀ60 mV immediately
after excision, at 201C, in the presence of 150 Â 10ꢀ3 M KCl on
both sides of the membrane in the absence (controls) or in the
presence of 10ꢀ4 M ATP in the bath solution. The macropatch
currents were recorded at 1 kHz sampling rates (fil-
ter ¼ 0.2 kHz) using an Axopatch-1D amplifier equipped with
a CV-4 headstage (Axon Instruments, Foster City, CA,
U.S.A.). Pipettes having an average tip opening area of
6.170.8 mm2 (macropatches ¼ 211) were used to measure the
currents sustained by multiple KATP channels and their
pharmacological properties.
Microanalyses of all the final molecules were within 70.4%
of theoretical values. For pharmacological experiments, these
molecules were used as racemate and free bases.
Molecular modelling studies have shown that the minimal
factors required to confer KCO properties to benzopyran
and other bicyclic compounds is the presence of four common
regions; two representing areas of lipophilic interaction such
as the aromatic ring condensed with the pyran ring and
the alkyl substituent at position 2 of the benzopyran ring,
and the others having hydrogen bonding forming capacity
represented by the oxygen atom of the amide group at position
4 and by the electron-withdrawing group at position 6 of
the benzopyran nucleus (Koga et al., 1993). According to
this pharmacophore model, our molecules showed at least
two lipophilic areas represented by the aromatic ring
condensed with the oxazine nucleus and by an alkyl
substituent at position 2 of the benzoxazine ring. The region
with hydrogen bonding forming capacity is represented
by the pyridylamino group at position 3, while at position 6
of the same nucleus an electron-withdrawing chlorine atom
is present. The novel structural modifications in the 2H-1,4-
benzoxazine nucleus involved the replacement of one
hydrogen atom at position 2 of the benzoxazine ring with
different alkyl substituents such as – CH3, – C2H5 or – C3H7
groups (Figure 1).
The currents flowing through the macropatches excised
from different fibres were digitally averaged, and were
calculated by subtracting the base-line level of the currents
from the open channel level. The base-line level for the KATP
current was measured in the presence of ATP (5 Â 10ꢀ3 M).
Macropatches containing voltage-dependent K+ channels or
inward rectifier K+ channels were excluded from the analysis.
Current amplitude was measured using the Clampfit program
(Axon Instruments, Foster City, CA, U.S.A.). No correction
for liquid junction potential was made, estimated to be
o1.9 mV in our experimental conditions.
No more than two different concentrations of the drugs
were applied to the same excised macropatch. Washout
periods followed the first and the second applications of the
drug solutions. Patches showing rundown or that did not fully
recover during washout after the drug solution applications
were excluded from the analysis.
logP, logD and pKa calculation
The relation between the lipophilicity and the biological
activity of the 2H-1,4-benzoxazine derivatives tested was
evaluated by calculating the log P and log D at pH 7, defined
as logarithm of the partition coefficient and logarithm of the
distribution coefficient at a particular pH, respectively. The
degree of ionisation of our molecules was evaluated by
calculating the pKa defined as the negative logarithm of the
dissociation constant. In our molecules the potential basic
centers are represented by the pyridine nitrogen atom and the
endocyclic nitrogen atom of the amidine function. Theoretical
log P, log D and pKa values were calculated by using ACD
software V. 6.0 (Advanced Chemistry Development Inc.,
Toronto, Ontario, Canada M5H 3V9).
Statistics
Figure 1 Chemical structures of cromakalim and of the newly
synthesised benzoxazine derivatives. The 2H-1,4-benzoxazine deri-
vatives are structural analogues of benzopyrans, the difference being
the replacement of the carbon atom at position 4 with a nitrogen
atom. The new molecules contain different alkyl substituents of
variable length such as the methyl, ethyl and propyl at position 2 of
the benzoxazine nucleus. All the synthesized molecules have a
pyridylamine group with hydrogen bond forming capacity at
position 3 and an electron-withdrawing group such as the chlorine
atom at position 6 of the benzoxazine nucleus. The introduction of
different alkyl substituents at position 2 of the benzoxazine ring
gives the opportunity to evaluate the influence of the increase of
lipophilicity and/or of alkyl chain length in this part of the molecule.
The data are expressed as mean7standard error unless
otherwise specified. The concentration – response relation of
the KATP currents constructed in the presence of internal ATP
fits the product of two equations describing the interaction of a
ligand with two sites mediating opposite effects, the stimula-
tory effect or the inhibitory effect (Rovati & Nicosia, 1994),
while the concentration – response relations of the KATP
currents versus drug concentrations constructed in the absence
of ATP are well fitted by one inhibitory term.
British Journal of Pharmacology vol 139 (2)

258
D. Tricarico et al
Pharmacology of muscle KATP channel
The stimulatory component can be described by the term
n
is the slope factor of the curves calculated in the presence
(equation (1)) or absence (equation (2)) of ATP. The
algorithms of the fitting procedures used are based on a
Marquardt least-squares fitting routine.
ðIdrugþATP ꢀ 1Þ Ã 100 ¼ Amax=ð1 þ ðDE50=½DrugꢁÞ Þ
while the inhibitory component can be described by the term
ð1Þ
n
ðIdrug ꢀ 1Þ Ã 100 ¼ Imax=ð1 þ ð½Drugꢁ=IC50Þ Þ
ð2Þ
Results
For equation (1), Idrug+ATP is the KATP currents measured in
the presence of the molecules under study and in the presence
of internal ATP (10ꢀ4 M); Amax, is the per cent maximal
activation of the KATP currents produced by the molecules
under study, and it is calculated in respect to the current levels
measured in the presence of ATP (10ꢀ4 M) alone. For equation
(2), Idrug, is the KATP currents measured in the presence of the
molecules under study but in the absence of ATP (controls);
Imax, is the per cent maximal inhibition of the KATP currents
produced by the molecules under study, and it is calculated in
respect to the maximal current levels measured in the absence
of ATP; DE50 is the concentration of the drug needed to
enhance the current by 50%, calculated in respect to the
maximal activation produced by the drugs in the presence of
internal ATP; IC50 is the concentration of the drug needed to
reduce the current by 50%, calculated in respect to the
maximal inhibition produced by the drugs in the absence of
internal ATP; [Drug] is the concentration of the drug tested; n
Effects of 2H-1,4-benzoxazine derivatives and cromaka-
lim on muscle KATP channels in the presence of internal
ATP
The effects of increasing concentrations of the newly synthe-
sised 2H-1,4-benzoxazine derivatives or cromakalim on muscle
KATP currents of excised macropatches recorded in the
presence of 10ꢀ4 M ATP in the bath were investigated. We
found that all the new compounds increased the macropatch
KATP currents with different degrees of potency depending on
the length of the alkyl chain at position 2 of the 2H-1,4-
benzoxazine ring (Figures 1; 2a, b).
In the range of concentrations from 10ꢀ8 to 10ꢀ6 M, the 2 –
CH3, 2 – C2H5 and 2 – C3H7,-2H-1,4-benzoxazine derivatives
induced a stimulation of the KATP currents. The 2 – C3H7-2H-
1,4-benzoxazine derivative increased the current concentra-
tion-dependently and was the most potent molecule as KATP
Figure 2 Effects of K+ channel openers on KATP channels of rat skeletal muscle fibres in the presence of internal ATP. (a) Digital
average of KATP current recorded in the excised inside-out macropatches, at ꢀ60 mV (V m), with high cytosolic KCl solutions on
both sides of the membrane, in the absence (controls) (n ¼ 31 macropatches), in the presence of 10ꢀ4 M concentration of ATP (n ¼ 31
macropatches), in the presence of 10ꢀ4 M ATP+the 2H-1,4-benzoxazine derivatives having the methyl (n ¼ 6 macropatches), ethyl
ꢀ4
(n ¼ 5 macropatches) or propyl (n ¼ 7macropatches) groups at position 2 of the benzoxazine ring or in the presence of 10
M
ATP+cromakalim (n ¼ 4 macropatches). C and O in the traces indicate closed and open channel levels, respectively, (b)
Concentration – response relation of the 2H-1,4-benzoxazine derivatives having the – CH3, – C2H5 or – C3H7 groups at position 2 of
the benzoxazine ring, and of cromakalim on the KATP currents inhibited by a 10ꢀ4 M concentration of internal ATP. All the
compounds under investigation including cromakalim produced a concentration-dependent increase of the KATP current at the
lower doses, in contrast inhibiting it at the highest doses, (c) Log P plot versus the DE50 of the 2H-1,4-benzoxazine derivatives and
cromakalim. An inverse relation was observed between the log P, as index of lipophilicity, and DE50 of the compounds under
investigation showing a coefficient of correlation of ꢀ0.97.
British Journal of Pharmacology vol 139 (2)

D. Tricarico et al
Pharmacology of muscle KATP channel
259
channel agonist compared to the other structurally related
analogues (Figure 2a, b).
Effects of 2H-1,4-benzoxazine derivatives and cromaka-
lim on muscle KATP channels in the absence of internal
ATP
However,
the
constructed
concentration – response
curves of the compounds showed a dualistic behaviour; in
fact, bell-shaped concentration – response curves were
observed for these compounds exhibiting a downturn in
response when a certain concentration was exceeded (Figure
2a, b). Cromakalim also induced a significant stimulation of
the current in the range of concentrations from 10ꢀ7 to
In the absence of internal ATP, the 2H-1,4-benzoxazine
molecules under investigation and cromakalim all inhibited
the KATP currents (Figure 3a, b).
In the range of concentrations from 10ꢀ10 to 10ꢀ4 M, the 2H-
1,4-benzoxazine derivatives and cromakalim reduced the KATP
3 Â 10ꢀ4
M
showing
a
downturn in response at higher
currents concentration-dependently, reaching
around 10ꢀ5
a
plateau
concentration (Figure 3a, b). A function
concentrations (Figure 2a, b). On the basis of these findings,
a function describing the interaction of a ligand with two sites
mediating opposite effects, one stimulatory and the other
inhibitory, was used to fit the experimental data. The
calculation of the fitted parameters of equation (1) showed
that the order of potency of the molecules as agonists
expressed as DE50 was – C3H74 – C2H54 – CH34cromaka-
lim (Table 1).
M
describing the interaction of a ligand with one inhibitory
component mediating the inhibitory response was used to fit
the experimental data for all the compounds under investiga-
tion. The calculation of the fitted parameters of equation (2)
showed that the order of potency of the molecules as
antagonists in the absence of internal ATP, expressed as
IC50, was cromakalimX – CH34 – C2H54 – C3H7 derivatives
(Table 1).
To evaluate the role of the lipophilicity and the degree
of ionisation of the 2H-1,4-benzoxazine derivatives in
determining the activation of the KATP channel in our system,
the log P, log D at pH 7and p Ka values were calculated.
No difference was observed between the calculated log P
and log D values for the 2H-1,4-benzoxazine derivatives
under study. Furthermore, the calculation of the pKa
gave values of 4.9170.11 for the protonated form of
the pyridine nitrogen and 3.1370.4 for the protonated
form of the endocyclic nitrogen atom of the amidine
function, respectively. These values were identical for all
the 2H-1,4-benzoxazine derivatives. These findings indicated
that our molecules are expected to be poorly protonated
at the pH of 7.2 used in our experiments and in consequence
of the fact that the lipophilicity plays a major role in
determining the pharmacological activity. In fact, an
inverse correlation was observed between the log P, which is
an index of lipophilicity, and the DE50 of the 2H-1,4-
benzoxazine derivatives (Figure 2c; Table 1). Cromakalim
had the lowest value of log P among the benzoxazine series,
and lower than that experimentally measured (Adlar et al.,
1995).
A direct correlation was observed between the IC50 and the
log P values of the 2 – C3H7, 2 – C2H5 and 2 – CH3-2H-1,4-
benzoxazine derivatives (Figure 3c; Table 1).
Differences in the maximal efficacy as antagonists were
observed within the 2H-1,4-benzoxazine derivatives and
cromakalim (Table 1).
Discussion
We provided evidence that molecules belonging to the class of
the 2H-1,4-benzoxazine derivatives, in the presence of internal
ATP, lead to significant stimulation of the muscle native KATP
channels with a different degree of potency depending on the
alkyl chains at position 2 of the 2H-1,4-benzoxazine nucleus.
This phenomenon is consistent with the hypothesis that the
increase in lipophilicity is a critical factor in determining the
potency of these molecules as muscle KATP channel agonists.
This is supported by the inverse correlation existing between
the calculated log P values, as lipophilicity index, and the DE50
of the molecules as KATP channel agonists.
Differences in the maximal efficacy as agonists were observed
within the 2H-1,4-benzoxazine series and cromakalim (Table 1).
However, to evaluate the possible influence of the different
degree of ionisation of the 2H-1,4-benzoxazine derivatives on
Table 1 Fitting parameters of the concentration–response curves of cromakalim and 2H-1,4-benzoxazine derivatives versus the skeletal
muscle KATP currents in the presence of absence of internal ATP
In the presence of ATP (10ꢀ4 M)
In the absence of ATP
Compounds
DE50 (M)
n
% Maximal activation
IC50m (M)
n
% Maximal inhibition
Cromakalim
Benzoxazine derivatives
R=–CH3
R=–C2H5
R=–C3H7
1.42 Â 10ꢀ5
2.81 Â 10ꢀ7
1.11 Â 10ꢀ7
1.63 Â 10ꢀ8
1.1
63
1.15 Â 10ꢀ8
2.6 Â 10ꢀ8
4.4 Â 10ꢀ8
1.68 Â 10ꢀ7
0.89
65
1.2
1.1
1.3
35
45
54
0.9
0.8
0.9
61
48
41
The parameters reported in the table were calculated using the fitting routine as described in the Methods. The 2H-1,4-benzoxazine
derivatives had a methyl (–CH3), ethyl (–C2H5) or propyl (–C3H7) group at position 2 of the nucleus. DE50 is the concentration of the drug
needed to enhance the current by 50%, calculated in respect to the maximal activation produced by the drugs in the presence of internal
ATP; the per cent maximal activation of the KATP currents produced by the molecules under study is calculated in respect to the current
levels measured in the presence of ATP (10ꢀ4 M) alone as reported in equation (1) of the Methods; IC50 is the concentration of the drug
needed to reduce the current by 50%, calculated in respect to the maximal inhibition produced by the drugs in the absence of internal ATP;
the per cent maximal inhibition of the KATP currents produced by the molecules under study is calculated in respect to the maximal current
levels measured in the absence of ATP as reported in equation (2) of the Methods. n is the slope factor calculated in the presence or absence
of ATP as reported in equations (1) and (2) (see Methods), respectively.
British Journal of Pharmacology vol 139 (2)

260
D. Tricarico et al
Pharmacology of muscle KATP channel
Figure 3 Effects of K+ channel openers on KATP channels of rat skeletal muscle fibres in the absence of internal ATP. (a) Digital
average of KATP current recorded in the excised inside-out macropatches, at ꢀ60 mV (V m), with high cytosolic KCl solutions on
both sides of the membrane, in the absence (controls) (n ¼ 26 macropatches), or in the presence of the 2H-1,4-benzoxazine
derivatives having the methyl (n ¼ 5 macropatches), ethyl (n ¼ 4 macropatches) or propyl (n ¼ 6 macropatches) groups at position 2
on the benzoxazine ring or in the presence of cromakalim (n ¼ 7macropatches). C and O in the traces indicate closed and open
channel levels, respectively. (b) Concentration – response relations of the 2H-1,4-benzoxazine derivatives having the – CH3, – C2H5
or – C3H7 groups at position 2 of the benzoxazine ring, and of cromakalim on the KATP currents in the absence of ATP. All the
compounds under investigation produced a concentration-dependent inhibition of the KATP current, (c) Log P plot versus the IC50 of
the 2H-1,4-benzoxazine derivatives and cromakalim. The calculated coefficient of correlation between the log P and IC50 of the
compounds under investigation was 0.8.
the KATP channel activity, the log D values for these molecules
at pH 7.0 were calculated. The fact that the calculated log D
values were identical to those of log P indicates that the 2H-
1,4-benzoxazine does not affect the KATP channel derivatives in
their protonated forms. This is certainly because of their weak
basic property as demonstrated by the findings that all the 2H-
1,4-benzoxazine derivatives showed quite low calculated pKa
values which were also identical for all the 2H-1,4-benzoxazine
derivatives under study. This means that there is no difference
in their ionisation state which can be explained with the little
influence that the different alkyl groups at position 2 of the
benzoxazine ring exert on the basicity of our molecules. The
main basic centre is in fact represented by the nitrogen atom of
the pyridine ring located at a long distance from the alkyl
groups.
characteristic of the TMD II region is consistent with the
lipophilic properties of the benzopyran molecules and also of
the benzoxazine derivatives (D’hahan et al., 1999; Babenko et
al., 2000).
Another point of interest is the fact that the 2H-1,4-
benzoxazine derivatives and cromakalim in the presence of
internal ATP showed bell-shaped concentration – response
curves stimulating the KATP channels at lower concentrations
but inhibiting the channels at higher concentrations. Further-
more, in the absence of internal ATP, all KCOs tested
inhibited KATP channel currents.
At least two mechanisms could explain these phenomena.
Firstly, these molecules could interact with two sites mediating
opposite effects, one site being a stimulatory site available for
drug binding in the presence of ATP, the other being the
inhibitory site which is unmasked in the absence of nucleotide.
This hypothesis is supported by the fact that our molecules in
the presence of ATP caused a full inhibition of channel
currents at the highest concentration tested. In fact, the general
idea is that the full inhibition of channels or receptors by
agonists at high concentrations is evidence for the existence of
two different sites mediating opposite actions (Rovati &
Nicosia, 1994).
These considerations all together suggest that the increase in
the lipophilicity favours the access of the molecule to the
stimulatory site through the membrane or can permit a
stronger interaction of the molecule with a hydrophobic area
of the binding site. The candidate hydrophobic area where our
molecules can bind is the TMDII region of the SUR (SUR2A)
subunit of muscle KATP channel complex, with TMD 17
playing a prominent role in the binding and affinity of KCO to
SUR (D’hahan et al., 1999; Uhde et al., 1999; Ashcroft &
Gribble, 2000; Moreau et al., 2000). The hydrophobic
Secondly, bell-shaped concentration – response curves with
reduced maximal efficacy could be the result of a partial
British Journal of Pharmacology vol 139 (2)

D. Tricarico et al
Pharmacology of muscle KATP channel
261
agonist action of the molecules on a single binding site (Rovati
& Nicosia, 1994). In this case, the dualistic effects of our KCO
and cromakalim could be the result of ATP-regulated KCO
binding, so that ATP would favour the transduction of
channel opening, whereas in the absence of the nucleotide
channel closure is favoured. Competition experiments per-
formed in skeletal muscle using [3H] P1075 have shown that
the binding of KCO including cromakalim to the membrane
fraction is supported by nucleotides. In most cases, a 1 : 1
stoichiometry was found suggesting the existence of a single
ATP-regulated site for KCO (Dickinson et al., 1997). Similar
results have been obtained for the vascular SUR2B and
pancreatic SUR1 subunits expressed in cell lines in which
nucleotides support binding of various KCOs (Schwanstecher
et al., 1998).
subunit of the KATP channel complex, to the inhibitory effects
of the KCO observed in our experiments cannot be excluded
(D’hahan et al., 1999; Ashcroft & Gribble, 2000).
The fact that pinacidil activates or inhibits the KATP channel
in skeletal muscle depending on the level of KATP channel
activity, or that the anilide tertiary carbinol derivative ZD6169
shows dualistic actions on the smooth muscle and ventricular
KATP channels suggests that the phenomenon that we observed
is not a unique property of the cyclic KCOs but can also be
extended to the open chain KCOs (Hehl & Neumcke, 1994;
Teramoto et al., 2001).
The dualistic mode of action of the KCOs resembles
that observed for some physiological modulators of KATP
channels. In fact, ADP at lower concentrations induces
stimulation of the skeletal muscle KATP channel, while at
higher concentrations it inhibits the channels (Allard &
Lazdunski, 1992). This phenomenon has been explained by
the high affinity interaction of ADP with the second nucleotide
binding fold of the SUR subunit, and with the low affinity
interaction with the ATP inhibitory site located on the kir6.2
subunit of the channel.
Although binding studies support the existence of a single
site for KCOs whose binding is ATP-regulated, we believe that
the existence of two different KCO sites mediating opposite
effects on KATP channel still cannot be excluded. In fact, it is
well known that the binding of a molecule to a receptor site is
also affected by the quaternary structure of the channel in the
native membrane which is conserved in patch-clamp experi-
ments but is lost in binding experiments. This is the case of
KCOs binding to the KATP channel which is an octameric
association of SUR and kir subunits, in which the affinity
measured by binding experiments as well as the stoichiometry
of the reaction do not match those measured by patch-clamp
experiments (Schwanstecher et al., 1998). In our experiments,
the weak correlation observed between the log P and the
potency of the 2 – CH3, 2 – C2H5 or 2 – C3H7-2H-1,4-benzox-
azine derivatives as antagonists (IC50) evaluated in the absence
of ATP suggests that the lipophilicity is not pivotal in
determining the inhibitory effects of KCO. This predicts that
sites different from the agonist sites for KCO may be located at
the interface between the hydrophilic/hydrophobic area.
Furthermore, the contribution of kir6.2, the pore-forming
Currently, we are searching for agonists on native skeletal
muscle KATP channels showing stimulatory effects without the
inhibitory component. This will be useful in selecting
molecules able to restore fully the abnormally reduced skeletal
muscle KATP conductance in hypoPP (Tricarico et al., 1998,
1999). An enantioselective synthesis of the optical isomers of
the tested benzoxazine derivatives is also in progress to
evaluate the influence of the absolute configuration on the
biological activity of these drugs.
This work was supported by Italian – Telethon grant no. 1208. We
thank Dr Sophie Talon for support during the experiments and Dr
Giuseppe Carbonara for helpful hints for the structural characterisa-
tion of the benzoxazine derivatives.
References
ADLAR, M., OKAFO, G., MEENAN, E. & CAMILLERI P. (1995). Rapid
estimation of octanol – water partition coefficients using deoxycho-
late micelles in capillary electrophoresis. J. Chem. Soc. Chem.
Commun., No. 21, 2241 – 2243.
D’HAHAN, N., JACQUET, H., MOREAU, C., CATTY P. & VIVAUDOU
M. (1999). A transmembrane domain of the sulfonylurea receptor
mediates activation of ATP-sensitive K+ channels by K+ channel
openers. Mol. Pharmacol., 56, 308 – 315.
ALLARD, B. & LAZDUNSKI, M. (1992). Nucleotides diphosphates
activate the ATP-sensitive potassium channels in mouse skeletal
muscle. Pflugers Arch., 422, 185 – 192.
ANDERSSON, K.E. (1992). Clinical pharmacology of potassium
channel openers. Pharmacol. Toxicol., 70, 244 – 254.
ASHCROFT, F.M. & GRIBBLE, F.M. (2000). New windows on the
mechanism of action of KATP channel openers. TiPS, 21, 439 – 445.
BABENKO, A.P., GONZALES, G. & BRYAN, J. (2000). Pharmaco-
topology of sulfonylurea receptors. J. Biol. Chem., 275, 717 – 720.
BETTONI, G., FERRORELLI, S., LOIODICE, F., TANGARI, N.,
TORTORELLA, V., GASPARRINI, F., MISITI, D. & VILLANI, C.
(1992). Chiral a-substituted a-aryloxy acetic acids: synthesis,
absolute configuration, chemical resolution and direct separation
by HPLC. Chirality, 4, 193 – 203.
DICKINSON, K.E.J., BRYSON, C.C., COHEN, R.B., ROGERS, L.,
GREEN, D.W. & ATWAL, K.S. (1997). Nucleotide regulation and
characteristics of potassium channel opener binding to skeletal
muscle membranes. Mol. Pharmacol., 52, 473 – 481.
FORESTIER, C., PIERRARD, J. & VIVAUDOU, M. (1996). Mechanism
of action of K channel openers on skeletal muscle KATP channels.
Interactions with nucleotides and protons. J. Gen. Physiol., 107,
489 – 502.
FRYER, R.I., EARLEY, J.V., FIELD, G.F., ZALLY, W. & STERNBACH,
L.H. (1969). A synthesis of amidines from cyclic amides. J. Org.
Chem., 34, 1143 – 1145.
GRANT, T.L., OHNMACHT, C.J.
& HOWE, B.B. (1994). Anilide
tertiary carbinols: a novel series of K+ channel openers. TiPS, 15,
402.
BETTONI, G., LOIODICE, F., TORTORELLA, V., CONTE-CAMERINO,
D., MAMBRINI, M., FERRANINI, E. & BRYANT, S.H. (1987 ).
Stereospecificity of the chloride ion channel: the action of chiral
clofibric acid analogues. J. Med. Chem., 30, 1267– 1270.
CALIENDO, G., GRIECO, P., PERISSUTTI, E., SANTAGATA, V.,
SANTINI, A., ALBRIZIO, S., FATTORUSSO, C., PINTO, A. &
SORRENTINO, R. (1998). Synthesis, biological activity and
conformational study of 1,4-benzoxazine derivatives as potassium
channel modulators. Eur. J. Med. Chem., 33, 957– 967.
HEHL, S. & NEUMCKE, B. (1994). Diverse effects of pinacidil on KATP
channels in mouse skeletal muscle in the presence of different
nucleotides. Cardiovasc. Res., 28, 841 – 846.
KABAKOV, A. (1998). Activation of KATP channels by Na/K pump in
isolated cardiac myocytes and giant membrane patches. Biophys. J.,
75, 2858 – 2867.
KOGA, H., OHTA, M., SATO, H., ISHIZAWA, T. & NABATA, H. (1993).
Design of potent K+ channel openers by pharmacophore model.
Bioorg. Med. Chem. Lett., 3, 625 – 631.
British Journal of Pharmacology vol 139 (2)

262
D. Tricarico et al
Pharmacology of muscle KATP channel
JOW, B. & NUMANN, R. (1999). The effects of ZD6169 on the ATP-
dependent K+ current (IKATP) in isolated cat ventricular myocytes.
Eur. J. Pharmacol., 383, 197– 202.
LAWSON, K. (2000). Potassium channel openers as potential ther-
apeutic weapons in ion channel disease. Kidney Int., 57, 838 – 845.
SCHWANSTECHER, M., SIEVERDING, C., DORSCHNER, H., GROSS,
I., AGUILAR-BRYAN, L., SCHWANSTECHER, C. & BRYAN, J.
(1998). Potassium channel openers require ATP to bind to and act
through sulfonylurea receptors. EMBO J., 17, 5529.
SEEBOHM, G., LERCHE, C., PUSCH, M., STEINMEYER, K., BRUG-
GEMANN, A. & BUSCH, A.E. (2001). A kinetic study on the
stereospecific inhibition of KCNQ1 and Iks by the chromanol 293B.
Br. J. Pharmacol., 134, 1647– 1654.
SONG, D.K. & ASHCROFT, F.M. (2001). ATP-modulation of K(sub-
ATP) channel ATP-sensitivity varies with the type of SUR subunit.
J. Biol. Chem., 276, 7143 – 7149.
TERAMOTO, N., YUNOKI, T., TAKANO, M., YONEMITSU, Y.,
MASAKI, I., SUEISHI, K., BRADING, A.F. & ITO, Y. (2001). Dual
action of ZD6169, a novel K(+) channel opener, on ATP-sensitive
K(+) channels in pig urethral myocytes. Br. J. Pharmacol., 133,
154 – 164.
LONGMAN, S.D.
& HAMILTON, T.C. (1992). Potassium channel
activator drugs: mechanism of action, pharmacological properties,
and therapeutic potential. Med. Res., Rev., 12, 73 – 148.
MATSUMOTO, Y., TSUZUKI, R., MATSUHISA, A., MASUDA, N.,
YAMAGIWA, Y., YANAGISAWA, I., SHIBANUMA, T. & NOHIRA,
H. (1999). Novel potassium channel activators. II1). Synthesis and
pharmacological evaluation of 3,4-dihydro-2H-1,4-benzoxazine
derivatives: modification of the aromatic part. Chem. Pharm. Bull.,
47, 971 – 979.
MATSUMOTO, Y., TSUZUKI, R., MATSUHISA, A., TAKANAMA, K.,
YODEN, T., UCHIDA, W., ASANO, M., FUJITA, S., YANAGISA-
WA, I.
&
FUJIKURA, T. (1996). Novel potassium channel
TRICARICO, D., PIERNO, S., MALLAMACI, R., SIRO BRIGIANI, G.,
CAPRIULO, R., SANTORO, G. & CONTE CAMERINO, D. (1998).
The biophysical and pharmacological characteristics of skeletal
muscle KATP channels are modified in K+ depleted rats, an animal
model of hypokalemic periodic paralysis. Mol. Pharmacol., 54,
197– 206.
TRICARICO, D., SERVIDEI, S., TONALI, P., JURKAT-ROTT, K. &
CONTE CAMERINO, D. (1999). Impairment of skeletal muscle
adenosine triphosphate-sensitive K+ channels in patients with
hypokalemic periodic paralysis. J. Clin. Invest., 103, 675 – 682.
activators: synthesis and structure – activity relationship studies of
3,4-dihydro-2H-1, 4-benzoxazine derivatives. Chem. Pharm. Bull.,
44, 103 – 114.
MATSUMOTO, Y., UCHIDA, A., NAKAHARA, H., YANAGISAWA, I.,
SHIBANUMA, T. & NOHIRA, H. (2000). Novel potassium channel
activators. III. Synthesis and pharmacological evaluation of 3,4-
dihydro-2H-1,4-benzoxazine derivatives: modification at the posi-
tion 2. Chem. Pharm. Bull., 48, 428 – 432.
MOREAU, C., JACQUET, H., PROST, AL., D’HAHAN, N. & VIVAU-
DOU, M. (2000). The molecular basis of the specificity of action of
KATP channel openers. EMBO J., 19, 6644 – 6651.
QUASTHOFF, S., SPULER, A., SPITTELMEISTER, W., LEHMANN-
HORN, F. & GRAFE, P. (1990). K+ channel openers suppress
myotonic activity of human skeletal muscle in vitro. Eur. J.
Pharmacol., 186, 125 – 128.
UHDE, I., TOMAN, A., GROSS, I., SCHWANSTECHER, C.
&
SCHWANSTECHER, M. (1999). Identification of the potassium
channel opener site on sulfonylurea receptors. J. Biol. Chem., 274,
28079 – 28082.
ROVATI, E.G.
with an inhibitory receptor or partial agonism? TiPS, 15,
&
NICOSIA, S. (1994). Lower efficacy: interaction
(Received October 17, 2002
Revised January 23, 2003
Accepted February 4, 2003)
141 – 144.
British Journal of Pharmacology vol 139 (2)