Partial structures of ketoconazole as modulators of the large conductance calcium-activated potassium channel (BKCa)

Article abstract of DOI:10.1016/j.bmcl.2005.11.001

A series of partial structures of ketoconazole has been synthesized and tested for activity on the large conductance calcium-activated potassium channel (BK) in bovine smooth muscle cells. This has provided openers and blockers of the channel. The results suggest that the phenyl and phenoxy moieties are important for interaction with BK, whereas the imidazole group is unimportant. The properties of the phenoxy moiety seem to determine whether the compounds act to open or block the channel.

Full text of DOI:10.1016/j.bmcl.2005.11.001

Bioorganic & Medicinal Chemistry Letters 16 (2006) 887–890
Partial structures of ketoconazole as modulators of the
large conductance calcium-activated potassium channel (BKCa)
Eoin C. Power,a,*, C. Robin Ganellina and David C. H. Bentonb
aDepartment of Chemistry, University College London, 20 Gordon Street, London WC1H 0AJ, UK
bDepartment of Pharmacology, University College London, Gower Street, London WC1E 6BT, UK
Received 23 March 2005; revised 1 November 2005; accepted 2 November 2005
Available online 15 December 2005
Abstract—A series of partial structures of ketoconazole has been synthesized and tested for activity on the large conductance
calcium-activated potassium channel (BK) in bovine smooth muscle cells. This has provided openers and blockers of the channel.
The results suggest that the phenyl and phenoxy moieties are important for interaction with BK, whereas the imidazole group is
unimportant. The properties of the phenoxy moiety seem to determine whether the compounds act to open or block the channel.
ꢀ 2005 Elsevier Ltd. All rights reserved.
The large conductance calcium-activated potassium
channel (BK) is a ubiquitously expressed potassium
channel that has been implicated in a variety of physio-
logical processes.1 The channels are activated both by
depolarization and by elevated calcium levels. In smooth
muscle, they have been shown to maintain myogenic
tone by shortening the action potential.2 Furthermore,
BK channels have been implicated in the regulation of
neurotransmitter release. Given its importance in main-
taining myogenic tone,3 activation of the BK channel
presents a possible mechanism for the treatment of asth-
ma or hypertension. In addition to this, a potassium
channel opener has been shown to confer neuroprotec-
tion in a rat model of permanent stroke.4
Several agents have been reported to activate the BK
channel. Natural products such as soyasaponins9 and
CAF60310 have been shown to open the channel. In
addition to these, synthetic biaryls such as NS-004,11
NS160812 and 3-aryloxindoles4 activate BK. The anti-
fungal drug ketoconazole (Chart 1) has been reported
to activate BK channels in ferret portal vein13 and sub-
sequently shown to inhibit the BK channel in GH3 pyra-
midal neurons in rat.14 Given its structural divergence
from the other known BK channel openers, it was seen
as an interesting target for investigation.
Taking ketoconazole as the lead, various partial struc-
tures (Fig. 1) were synthesized and tested to discover
which parts of the ketoconazole molecule may be impor-
tant for activity.
BK channels are membrane-bound channels that con-
tain a- and b-subunits, examples of which have been
cloned from a variety of tissues.5–7 Several types of
b-subunits exist and this, along with splice variation of
the a-subunit, contributes to the heterogeneity of native
BK channels.8 This may prove to be important in the
design of therapeutic agents that act by activation of
the BK channel.
Compound 1 was synthesized according to the published
route.15 Hydrolysis of the benzoyl ester yielded the alco-
hol 2 that was reacted with p-toluene sulfonyl chloride
to yield 3. Reaction of 3 with the appropriate phenoxide
in DMF or DMSO yielded compounds 4, 5, 6, 7 and 8
(Scheme 1, Table 1). To obtain 6, the aniline was protected
as an imine. The fully protected product after reaction
with 3 was not isolated. After flash chromatography, a
mixture of aniline 6 and imine was obtained. This was
treated with hydrazine to convert any remaining imine
to the desired product 6. Compound 9 was obtained by
treating 8 with KOH in i-PrOH according to a literature
procedure.18 It was also considered worthwhile to test
the benzoyl 1 and tosyl 3 derivatives, which were synthetic
intermediates.
Keywords: Potassium channels; Opener; Blocker; Large conductance;
BKCa.
*
Corresponding author. Tel.: +39 0577 381436; fax: +39 0577
381202; e-mail: epower@sienabiotech.it
Present address: Sienabiotech S.p.A., Via Fiorentina, 53100 Siena,
Italy.
0960-894X/$ - see front matter ꢀ 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.11.001

888
E. C. Power et al. / Bioorg. Med. Chem. Lett. 16 (2006) 887–890
Cl
Cl
N
N
HO
N
O
O
N
O
O
Cl
i,ii,iii
Cl
O
R
CAF603
Soyasaponins
O
O
1
4,5,6,7,8,9
Cl
Cl
O
CF3
CF3
X
N
X
Scheme 1. Synthesis of ketoconazole analogues. Reagents and condi-
tions: (i) 1, NaOH, dioxane reflux; (ii) 2, tosyl chloride, pyridine; (iii) 3,
phenol, NaH, DMSO. 0–80 ꢁC.
OH
N
H
N
OH
O
O
H
O
Aryloxindoles
NS-004
Cl
N
N
O
Table 1. Structures of compounds 1–9
OH
Cl
H
H
CF3
N
N
Substituent
Activity BK
O
1
2
3
4
5
6
7
8
9
COPh
Blocker
Opener
None
O
H
Tosyl
Phenyl
N
Cl
Blocker
Blocker
Opener
Blocker
Ketoconazole
N
NS-1608
O
4-Bromophenyl
4-Aminophenyl
4-(N-Piperidino)phenyl
Chart 1. Known BK channel activators.
4-(4-Acetyl-N-piperazino)phenyl (ketoconazole) Opener
4-(N-Piperazino)phenyl
Opener
Cl
i
O
HO
N
HO
NH2
10
Cl
N
O
N
O
O
ii,iii
O
Br
11
12
HO
N
N
N
N
O
iv
O
N
NH
O
N
Figure 1. Structure of ketoconazole showing the notional removal of
subunits.
O
iv, v
13
The amine group in 6 was introduced via benzylimino-
phenol 1016 which was formed in good yield by stirring
4-aminophenol and benzaldehyde in methanol at room
temperature. To synthesize the piperidino-compound
7, 4-bromoanisole was aminated according to a Buch-
wald–Hartwig procedure17 and demethylated using
BBr3 in dichloromethane to yield the phenol 11. 1-Ace-
tyl-4-(4-methoxyphenyl)piperazine 12 and acetyl-4-(4-
hydroxyphenyl)piperazine 13 were prepared according
to literature procedures19 (Scheme 2). Compound 13
was used to prepare compounds 8 (Table 1), 16 and 17
(Scheme 3, Table 2).
HO
N
N
HO
N
NH
Scheme 2. Synthesis of 4-substituted phenols. Reagents and condi-
tions: (i) Benzaldehyde MeOH, rt; (ii) Pd2(dba)3, P(o-Tol)3, piperidine,
LHMDS, toluene 80 ꢁC; (iii) BBr3, DCM, À78 ꢁC–rt; (iv) Ac2O,
pyridine, DCM; (v) K2CO3, MeOH, rt.
OTs
R1
i,ii
O
iii
R1
O
O
14 &15
R1
O
O
O
The compounds 14 and 15 were synthesized by forma-
tion of the acetal followed by tosylation (Scheme 3,
Table 2). Substitution of the tosylate 14 with phenox-
ide of 13 in DMSO yielded 16. Similarly 17 was syn-
thesized from 15. For these initial investigations 16
was tested as the racemate and 17 as a mixture of
diastereomers.
N
N
O
16 & 17
Scheme 3. Synthesis of tosylates 14 and 15 and replacement by
substitution to give 16 and 17. Reagents and conditions: (i) Glycerol,
PTSA, benzene/butanol, reflux Dean–Stark; (ii) tosyl chloride, pyri-
dine; (iii) 13, NaH, DMSO. 0–80 ꢁC.

E. C. Power et al. / Bioorg. Med. Chem. Lett. 16 (2006) 887–890
Table 2. Structures of compounds 14–17
889
ed compound was applied to at least three different cells
and the effect is reported in Table 3 as the mean change
in current at +80 mV SEM. As an internal control the
effectiveness of ketoconazole (30 lM) was tested on
every cell included in the study. Since ketoconazole
has already been shown to act on BK, it was used as
the reference agent in this study.13
R1
Stereochemistry
14
15
16
17
Methyl
Racemic
2,4-Dichlorophenyl
Methyl
2,4-Dichlorophenyl
Racemic & 1:1 cis and trans
Racemic
Racemic & 1:1 cis and trans
A culture of bovine pulmonary artery (BPASM) derived
by explant culture was kindly supplied by Dr R. Corder,
William Harvey Research Institute, London, UK. The
cells used in the study were from passages 12 to 19. Cul-
tures were maintained in DMEM (Invitrogen) supple-
mented with 10% foetal calf serum, 2 mM L-glutamine
and 100 U mlÀ1 penicillin and 100 lg mlÀ1 streptomy-
cin. For electrophysiology, the cells were plated onto
35 mm tissue culture dishes and used within 1–3 days.
Stock solutions of test compounds were prepared in
DMSO at a concentration of 2 · 10À2 M and diluted
in bath solution to give the desired final concentration.
The concentration of DMSO did not exceed 0.5% v/v
and at this concentration had no visible effect on the
current.
Activity of partial structures (Table 3): Major dissection
of the ketoconazole molecule gave weakly active or inac-
tive fragments; thus, 2 and 12 were notionally obtained
by splitting of the ether linkage. Compound 2, derived
from the removal of the phenylpiperazine moiety,
showed a little activity. Compound 12, having the phe-
nylpiperazine moiety found in ketoconazole 8, was inac-
tive (no effect at 100 lM) and was still not rendered
active when combined with the dioxalane ring as in 16
(no effect at 100 lM). Thus, it appears that either the
2,4-dichlorophenyl group and/or the imidazolyl ring of
ketoconazole are very important for activity.
Whole cell BK currents were recorded using an EPC7
patch-clamp amplifier essentially as described.20 Briefly,
cells were superfused with a solution containing (in
mM): NaCl 136, KCl 6, MgCl2, 1, CaCl2 2, glucose 10
and HEPES 10. pH was adjusted to 7.4 with 1 M
NaOH. The pipette filling solution contained (in mM):
KCl 136, MgCl2 1, CaCl2 1, HEPES 10 and EGTA 2.
The pH was adjusted to 7.2 with 1 M KOH. Free
Ca2+ was calculated to be 150 nM. Patch pipettes were
fabricated from 1.5-mm borosilicate capillaries (Clarke
ElectroMedical), fire polished and coated with Sylgard
Resin and had resistances of 3–5 MX when filled with
internal solution. All experiments were conducted at
room temperature (23–25 ꢁC).
Maintaining the imidazole and dichlorophenyl moieties,
while varying the acetylpiperazine group, yielded some
interesting results. Removing the acetyl group gave the
piperazine 9, which was a more effective channel opener
than ketoconazole. Splitting of the acetylpiperazine ring
to leave the structure having a p-aminophenoxy group
gave compound 6, which was also more potent than
ketoconazole.
Routinely cells were clamped at a holding potential of
À20 mV. Voltage dependent activation of BK currents
could be elicited by applying 100 mS voltage jumps to
more positive potentials. The effect of test compounds
was assessed by measuring the current activated on
stepping from the holding potential to +80 mV and
expressed as the difference between the current in the
presence and absence of test compounds as a percentage
of the current in its absence. Each concentration of test-
On the other hand, removing the acetylpiperazine ring
entirely gave the phenoxy compound 4, which blocked
the channel. Likewise, replacing the acetylpiperazine
ring by other groups such as piperidine (7) or bromine
(5) gave channel blockers. Replacing the phenoxy group
by benzoate gave a weak blocker (1) or by p-toluenesulf-
onate gave compound (3), which had no effect at 30 lM.
Removal of the imidazole ring gave compound 17,
which had potency similar to that of ketoconazole as a
channel opener. This finding is similar to the finding that
clotrimazole can block the BK channel even after
removal of the imidazole moiety.13 Given that this moi-
ety is also crucial for the cytochrome P450 activity of
ketoconazole, the authors would suggest that the action
on BK is not due to a potentiating effect via cytochrome.
Also it has been shown that ketoconazole acts on
excised patches of smooth muscle cells which suggests
a direct action or an action close to the channel.13
Table 3. Quantitative results for the compounds tested
Activity BK
n
% Inhibition (concn) % Opening (concn)
1
2
3
4
5
6
7
8
Blocker
Opener
None
3
3
40 8 (30 lM)
47 3 (100 lM)
2
30 (30 lM)
Blocker
Blocker
Opener
Blocker
Opener
3
4
3
3
5
73 7 (30 lM)
95 35 (30 lM)
566 200 (30 lM)
80 6 (10 lM)
74 14 (10 lM)
218 38 (30 lM)
348 32 (100 lM)
310 83 (30 lM)
31
9
Given that the piperazines 8 and 9, and the aniline 6, act
as channel openers, it would appear that an anilino
nitrogen atom is the common feature conferring this
mode of action. Such a group might act as a hydro-
gen-bond acceptor or increase the electron density in
the phenoxy ring. It is therefore surprising to find that
piperidine 7 (which also possesses an anilino N) is a
9
Opener
3
12 None
16 None
17 Opener
3
9
17 (100 lM)
3
17 6 (100 lM)
4
235 57 (30 lM)
Activity is shown as the % decrease (blockade) or increase (opening) of
baseline current at the given concentration and n = number of
observations.

890
E. C. Power et al. / Bioorg. Med. Chem. Lett. 16 (2006) 887–890
7. Tseng-Crank, J.; Godinot, N.; Johansen, T. E.; Ahring, P.
K.; Strøbæk, D.; Mertz, R.; Foster, C. D.; Olesen, S.-P.;
Reinhart, P. H. Proc. Natl. Acad. Sci. U.S.A. 1996, 93,
9200.
blocker. Perhaps, the differentiation between these com-
pounds is that 6, 8 and 9 have hydrophilic substituents,
whereas for 7, the piperidino group is lipophilic.
8. Dworetzky, S. I.; Boissard, C. G.; Lum-Ragan, J. T.;
McKay, M. C.; Post-Munson, D. J.; Trojnacki, J. T.;
Chang, C.-P.; Gribkoff, V. K. J. Neurosci. 1996, 16,
4543.
9. McManus, O. B.; Harris, G. H.; Giangiacomo, K. M.;
Feigenbaum, P.; Reuben, J. P.; Addy, M. E.; Burka, J. F.;
Kaczorowski, G. J.; Garcia, M. L. Biochemistry 1993, 32,
6128.
Overall these results show that the phenyl groups seem
to be important for binding to the BK channel, whereas
the imidazole group does not. They also point to the
properties of the phenoxy moiety as being crucial for
determining the type of activity.
10. Ondeyka, J. G.; Ball, R. G.; Garcia, M. L.; Dombrowski,
A. W.; Sabnis, G.; Kaczorowski, G. J.; Zink, D. L.; Bills,
G. F.; Goetz, M. A.; Schmalhofer, W. A.; Singh, S. B.
Biorg. Med. Chem. Lett. 1995, 5, 733.
11. Olesen, S.-P.; Munch, E.; Watjen, F. J.; Drejer, R.
J. Neuroreport 1994, 5, 1001.
Acknowledgment
The authors wish to acknowledge and thank Prof Don
Jenkinson for helpful discussions on this work.
12. Strobaek, D.; Christopherson, P.; Holm, N. R.; Moldt, P.;
Ahring, P. K.; Johansen, T. E.; Olesen, S.-P. Neurophar-
macology 1996, 45, 903.
References and notes
13. Rittenhouse, A. R.; Parker, C.; Brugnara, C.; Morgan, K.
G.; Alper, S. L. Am. J. Physiol. 1997, 273, C45.
14. Wu, S.-N.; Li, H.-F.; Jan, C.-R.; Shen, A. Y. Neuropha-
racology 1999, 38, 979.
1. Gribkoff, V. K.; Starrett, J. E.; Dworetsky, S. I. Adv.
Pharmacol. 1997, 319.
2. Shimizu, S.; Yokoshiki, H.; Sperelakis, N.; Paul, R. J.
J. Vasc. Res. 2000, 37, 16.
15. Heeres, J.; Backx, L. J. J.; Mostmans, J. H.; Van Cutsem,
J. J. Med. Chem. 1979, 22, 1003.
16. Isse, A. A.; Abdurahman, A. M.; Vianello, E. J. Chem.
Soc. Perkin. 2 1996, 597.
17. Louie, J.; Hartwig, J. F. Tet. Lett. 1995, 36, 3609.
18. Heeres, J.; Backx, L. J. J.; Van Cutsem, J. J. Med. Chem.
1984, 22, 894.
19. Chapman, D. R.; Bauer, L.; Waller, D. P.; Zanveld, L. J.
D. J. Heterocycl. Chem. 1990, 27, 2063.
3. Assano, M.; Nomura, Y.; Ito, K.; Uyama, Y.; Imaizumi,
Y.; Watanabe, M. J. Pharm. Exp. Ther. 1995, 275, 775.
4. Hewawasam, P.; Erway, M.; Moon, S. L.; Knipe, J.;
Weiner, H.; Boissard, C. G.; Post-Munson, D. J.; Gao, Q.;
Huang, S.; Gribkoff, V. J. J. Med. Chem. 2002, 45, 1487.
5. Tseng-Crank, J.; Foster, C. D.; Krause, J. D.; Mertz, R.;
Godinot, N.; DiChiara, T. J.; Reinhartt, P. H. Neuron
1994, 13, 1315.
6. Vogalis, F.; Vincent, T.; Qureshi, I.; Schmalz, F.; Ward,
M. W.; Sanders, K. M.; Horowitz, B. Am. J. Physiol. 1996,
271, G629.
20. Olesen, S-P.; Munch, E.; Moldt, P.; Drejer, J. Eur. J.
Pharmacol. 1994, 251, 53.