Synthesis of benzofuran, benzothiophene, and benzothiazole-based thioamides and their evaluation as KATP channel openers

Article abstract of DOI:10.1002/cmdc.201000297

Several series of benzofurans, benzothiophenes, and benzothiazoles, all featuring the thioamide group, were synthesized and tested as novel K _ATP channel openers in artificial cell systems: CHO cells transfected with SUR1/Kir6.2, and HEK 293 cells transfected with SUR2B/Kir6.1; these served as model systems for insulin-secreting pancreatic β cells and smooth muscle cells, respectively. All compounds were investigated with respect to their binding affinity for the SUR2B-type K_ATP channels using [ ³H]P1075 as radioligand. Selected compounds were also tested as agonists in intact cells using DiBAC₄(3) and DyeB (R7260) as membrane potential dyes. Remarkable affinity for SUR2B/Kir6.1 channels in the single-digit micromolar range was observed. In addition, benzothiazole-derived thioamides with sterically demanding, lipophilic substituents showed >100-fold selectivity in favor of SUR2B/Kir6.1. A one-carbon spacer between the heterocyclic skeleton and the thioamide moiety was observed to be crucial for affinity and selectivity. Two of the most potent and selective compounds were studied by crystal structure analyses.

Full text of DOI:10.1002/cmdc.201000297

DOI: 10.1002/cmdc.201000297
Synthesis of Benzofuran, Benzothiophene, and
Benzothiazole-Based Thioamides and their Evaluation as
KATP Channel Openers
Andreas Fischer,[a] Claas Schmidt,[b] Stefan Lachenicht,[a] Dagmar Grittner,[b]
Marcus Winkler,[c] Thomas Wrobel,[b] Achim Rood,[b] Horst Lemoine,[b] Walter Frank,[d] and
Manfred Braun*[a]
Several series of benzofurans, benzothiophenes, and benzo-
thiazoles, all featuring the thioamide group, were synthesized
and tested as novel KATP channel openers in artificial cell sys-
tems: CHO cells transfected with SUR1/Kir6.2, and HEK 293
cells transfected with SUR2B/Kir6.1; these served as model sys-
tems for insulin-secreting pancreatic b cells and smooth
muscle cells, respectively. All compounds were investigated
with respect to their binding affinity for the SUR2B-type KATP
channels using [3H]P1075 as radioligand. Selected compounds
were also tested as agonists in intact cells using DiBAC4(3) and
DyeB (R7260) as membrane potential dyes. Remarkable affinity
for SUR2B/Kir6.1 channels in the single-digit micromolar range
was observed. In addition, benzothiazole-derived thioamides
with sterically demanding, lipophilic substituents showed
>100-fold selectivity in favor of SUR2B/Kir6.1. A one-carbon
spacer between the heterocyclic skeleton and the thioamide
moiety was observed to be crucial for affinity and selectivity.
Two of the most potent and selective compounds were stud-
ied by crystal structure analyses.
Introduction
ATP-sensitive potassium channels, usually called KATP channels,
are ubiquitously distributed and present in most excitable
cells. By tuning the permeability of potassium ions, they link
the metabolic state of the cell to the excitability of the mem-
brane. In pancreatic b cells, they regulate insulin secretion. In
cardiac cells, they decrease the duration of the action poten-
tial. They also control the vessel tone in smooth muscle cells
and are responsible for the transmitter release in neurons.
With regard to their structure, KATP channels consist of octam-
ers of two membrane proteins: the sulfonylurea receptor (SUR)
and the inward rectifying potassium channel (Kir). To form an
effective KATP channel, four Kir (Kir6.x) channels associate with
four SUR subunits. Kir falls into two subtypes: Kir6.1 and Kir6.2,
whereas three different subtypes (SUR1, SUR2A, and SUR2B) of
the sulfonylurea receptor are known. The individual SUR iso-
form is responsible for the tissue-specific properties of the KATP
channels: SUR1/Kir6.2 in pancreatic b cells, SUR2A/Kir6.2 in car-
diac cells, and SUR2B/Kir6.1 in vascular smooth muscle cells. In
view of the multitude of physiological functions, KATP channels
represent promising targets for various drugs. In the past, not
only were antagonists like the anti-diabetic sulfonylureas de-
veloped, but agonists, the so-called potassium channel open-
ers (PCOs or KCOs), were also studied intensively because they
are regarded as tools to interfere with the excitability of vari-
ous cells. Hypertension, asthma, urinary incontinence, and car-
diac ischemia are only some of the potential indications for
PCOs.[1]
Figure 1. Structures of representative KATP channel openers.
(Figure 1): 1. the cyanoguanidines such as pinacidil (1), 2. the
thiadiazine-S,S-dioxides such as diazoxide (2) with its manifold
analogues [BPDZ73 (3) may be considered a typical representa-
[a] A. Fischer,+ Dr. S. Lachenicht, Prof. Dr. M. Braun
Institute of Organic & Macromolecular Chemistry
University of Dꢀsseldorf, Universiꢁtsstr. 1, 40225 Dꢀsseldorf (Germany)
Fax: (+49)211-81-15079
E-mail: braunm@uni-duesseldorf.de
[b] C. Schmidt,+ D. Grittner, T. Wrobel, A. Rood, Prof. Dr. H. Lemoine
Institute of Laser Medicine, Molecular Drug Research
Medical Faculty, University of Dꢀsseldorf (Germany)
[c] Dr. M. Winkler
Institute of Pharmacology & Toxicology, Medical Faculty
University of Tꢀbingen (Germany)
[d] Prof. Dr. W. Frank
Although a large number of compounds with various chemi-
cal structures have proven themselves as potent agonists of
KATP channels, most of the drugs rely on three skeletons
Institute of Inorganic & Structural Chemistry
University of Dꢀsseldorf (Germany)
[+] These authors contributed equally to this work.
ChemMedChem 2010, 5, 1749 – 1759
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1749

M. Braun et al.
MED
tive], and 3. the benzopyran and chromene family including
lev-cromakalim (4), KC-399 (5), and the related benzothiazine 6
as representative drugs.[2]
isothiocyanate 9b to give 3-hydroxybenzofuran-2-carbothioa-
mide 10i. It formed completely as the enol tautomer accord-
ing to its spectroscopic data.
Despite the impressive progress that has been made with
the benzopyrans and benzothiazines in terms of affinity, the
development of tissue-selective PCOs remains a challenge, and
high affinity combined with selectivity toward the different
KATP channels composed of various combinations of SUR and
Kir isoforms are highly desirable. In pursuing this goal, we
were guided by the idea that heterocyclic skeletons, hitherto
unused in KATP-active drugs, might be promising candidates for
the development of new and selective PCOs. For this purpose
we combined, for the first time, the thioamide pharmaco-
phore[3] with the benzofuran, benzothiophene, and benzothia-
zole skeletons.
The position where the thioamide group is attached at the
heterocyclic skeleton was expected to have a substantial influ-
ence on the affinity for KATP channels. To prove this, benzofuran
12 carrying the thioamide residue at position 3 was prepared.
The synthesis of 12 started with 2-methylbenzofuran 11, itself
available from 7a by metallation and treatment with methyl
iodide using a modified published procedure.[5] A Friedel–
Crafts-type acylation[6] performed with methyl isothiocyanate
9a permitted the introduction of the thioamide moiety at posi-
tion 3 to give benzofuran 12 in a completely regioselective
manner (Scheme 2). The substituted coumaranone 11 b[7]
served us as starting material for the preparation of carbamate
13, which was obtained by reduction with sodium borohydride
and subsequent treatment with ethyl isothiocyanate 9b. For
reasons of comparison, the regioisomeric benzothiophene-de-
rived thioamides 14 and 15, prepared according to published
procedures,[8] were also included in the evaluation as potential
PCOs.
Results and Discussion
Synthesis
The heterocyclic skeletons of benzofuran, 3-coumaranone,
benzothiophene, and benzothiazole were linked to the thio-
amide moiety that was introduced in all cases by using alkyl or
aryl isothiocyanates as electrophilic reagents. In a first series
(Scheme 1), we took advantage of the known acidity of benzo-
furan at carbon atom 2.[4] Thus, treatment of benzofuran 7a
with n-butyllithium generated 2-lithiobenzofuran 8a, which
was quenched with isothiocyanates 9a–e to give the respec-
tive thioamides 10a–e. In an analogous manner, thioamide
10h was prepared from trifluoromethyl-substituted benzofuran
7c via the lithiated intermediate 8c. When 5-bromobenzofuran
7b was used as the starting material, the deprotonation was
performed with lithium diisopropylamide instead of n-butyl-
lithium to avoid a bromine–lithium exchange. The relatively
low yields of thioamides 10 f and 10g resulting from lithioben-
zofuran 8b and isothiocyanates 9a and 9b was probably
caused by incomplete deprotonation due to the weaker base
and by the undesired addition of diisopropylamine to the iso-
thiocyanates. 3-Benzofuranone 11 was treated with sodium hy-
dride to generate the enolate prior to the reaction with ethyl
Scheme 2. Synthesis of benzofuran-derived thioamide 12 and carbamate 13.
Regioisomeric benzothiophenes 14 and 15. Reagents and conditions:
a) 1. nBuLi, THF, ꢀ788C, 1 h, 2. MeI, ꢀ788C!RT, 16 h, 68%; b) 9a, AlCl3,
MeNO2, RT, 24 h, 35%; c) NaBH4 (12 equiv), EtOH, 08C!RT, 99%; d) nBuLi,
9b, Et2O, ꢀ788C!RT, 29%.
In two additional series, we determined whether a combina-
tion of the thioamide moiety with the benzothiazole skeleton
would lead to compounds with activity at KATP channels. Thus,
benzothiazoles 16a and 16b were first coupled directly with
isothiocyanates, taking advantage of the known acidity at posi-
tion 2 of the ring skeleton.[9] The trifluoromethyl-substituted,
hitherto unknown benzothiazole 16b was obtained from 2-
amino-4-(trifluoromethyl)thiophenol and formaldehyde in the
presence of scandium trifluoromethanesulfonate in 72%
yield.[10] The lithiation of benzothiazoles 16a and 16b at the 2-
position was performed at ꢀ1008C with n-butyllithium. The
labile intermediates thus generated were quenched at the
same temperature with isothiocyanates 9a, 9b, and 9e to give
the thioamides 17a–e (Scheme 3).
Scheme 1. Synthesis of benzofuran-2-carbothioamides 10. Reagents and con-
ditions: a) nBuLi, THF, ꢀ788C, 1 h; b) R2N=C=S (9), ꢀ788C!RT, 16 h; c) NaH,
THF, then 9b, RT, 18 h.
1750
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1749 – 1759

KATP Channel Openers
which stably express the arterial smooth-muscle-type KATP
channel. By using the agonist radioligand [3H]P1075, the pKD
values listed in Table 1 were determined. All the compounds
investigated exhibited afinity; however, the afinity varied con-
siderably, and the binding constants range from the three-digit
to the single-digit micromolar level.
Table 1. Dissociation constants for compounds 10, 12–15, 17, 19, 20,
and 2 determined by radioligand binding at HEK 293 (SUR2B/Kir6.1) cells.
Scheme 3. Synthesis of benzothiazole-2-carbothioamides 17. Reagents and
conditions: a) nBuLi, THF, ꢀ1008C, 1 h; b) 9, ꢀ758C!ꢀ508C, 20 min.
[a]
[a]
Entry
Compd
pKD ꢁASD[b]
Entry
Compd
pKD ꢁASD[b]
1
3
5
7
10a
10c
10e
10g
10i
3.82ꢁ0.04
4.39ꢁ0.03
3.60ꢁ0.09
4.44ꢁ0.02
4.08ꢁ0.02
4.18ꢁ0.03
4.22ꢁ0.06
<3.30
4.08ꢁ0.05
4.83ꢁ0.03
5.39ꢁ0.03
4.30ꢁ0.03
3.88ꢁ0.04
2
4
6
8
10
12
14
16
18
20
22
24
10b
10d
10 f
10h
12
4.23ꢁ0.03
4.38ꢁ0.03
4.34ꢁ0.03
3.99ꢁ0.05
3.95ꢁ0.04
4.64ꢁ0.03
4.12ꢁ0.03
4.12ꢁ0.04
4.31ꢁ0.03
5.03ꢁ0.03
4.12ꢁ0.02
4.97ꢁ0.02
Finally, we envisaged incorporating a one-carbon spacer be-
tween the benzothiazole and the thioamide moieties. For this
purpose, thiazoles 18, carrying an alkyl side chain at position 2,
were submitted to deprotonation[11] and subsequently
quenched with isothiocyanates. Thus, thioamides 19a–d were
synthesized (Scheme 4). They contain either a methylene
9
11
13
15
17
19
21
23
25
14
15
17a
17c
17e
19b
19d
20b
13
17b
17d
19a
19c
20a
2
[a] pKD values (ꢀlogm) calculated by nonlinear regression according to
Equation (1); see Experimental Section. [b] Asymptotic standard devia-
tions estimated by nonlinear regression.
In a first series of benzofurans 10a–e (Table 1 Entries 1–5),
the influence of the substituent at the nitrogen atom of the 2-
carbothioamide residue was tested. The affinity increases from
methyl to ethyl (Entries 1 and 2) to n- or tert-butyl, the latter
compounds 10c and 10d exhibiting very similar pKD values
(Entries 3 and 4). These results suggest that lipophilic residues
at the nitrogen atom are favorable with respect to affinity. On
the other hand, a phenyl group at this position is deleterious
(Entry 5). An additional bromine substituent at position 5 of
the heterocyclic core leads to a slight enhancement in affinity
(Entries 6 vs. 1, and 7 vs. 2). The electron-withdrawing trifluoro-
methyl group, which has been found to be advantageous in
PCOs based on the benzopyran skeleton, did not lead to stron-
ger binding in the case of benzofuran 10h, which is equipped
with that group at position 5. Compound 10i, which features a
hydroxy group in addition to the thioamide residue, did not
provide higher affinity either.
Scheme 4. Synthesis of “spacered” benzothiazole-derived thioamides 19. Re-
agents and conditions: a) PhLi, Et2O, ꢀ788C, 1 h; b) 9, ꢀ788C!RT, overnight.
spacer (in 19a) or a branched spacer (in 19b–d). The latter
compounds were prepared as racemic mixtures.
To evaluate the role of the carbothioamide in relation to the
carboxylic amide, lithiated benzothiazoles 18a and 18c were
allowed to react with tert-butylisocyanate to give the carboxyl-
ic amides 20a and 20b, respectively (Scheme 5).
In vitro biological studies
In view of the fact that benzofurans, benzothiophenes, and
benzothiazoles have not been recognized as KATP channel
openers, the binding affinity of all the compounds 10, 12–15,
17, 19, and 20 was measured in HEK 293 (SUR2B/Kir6.1) cells,
To determine whether the position of the thioamide group
influences binding at the SUR2B-type KATP channel, furans 10a
and 12 (Table 1, Entry 1 vs. Entry 10) as well as the benzothio-
phenes 14 and 15 (Entry 11 vs. Entry 12) were compared. The
marginal difference between the pKD values of compounds
10a and 12 cannot be simply attributed to the substitution
pattern of the thioamide residue, as benzofuran 12 differs
from 10a by an additional methyl substituent. Clearer insight
comes from the regioisomeric benzothiophenes 14 and 15:
the attachment of the thioamide group at position 2 of the
heterocyclic ring is more favorable than the 3-substitution pat-
tern, and the pKD values differ by ~0.5 log units. On the other
Scheme 5. Synthesis of benzothiazole-derived amides 20. Reagents and con-
ditions: a) nBuLi, Et2O, ꢀ788C, 1 h; b) Me3CNCO, ꢀ788C!RT, overnight. 20a:
36%, 20b: 10%.
ChemMedChem 2010, 5, 1749 – 1759
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1751

M. Braun et al.
MED
hand, comparison of the activity of compounds 10a and 15
(Entry 1 vs. Entry 12) clearly reveals that replacement of the
oxygen atom in the five-membered ring with sulfur leads to an
enhancement in affinity of 0.82 log units. This result suggests
that an “enlarged” five-membered heterocyclic ring containing
sulfur instead of oxygen might more effectively fill the space
available in the receptor. Therefore, we designed and synthe-
sized benzothiazole-derived thioamides, anticipating that the
additional nitrogen atom would lead to an additional hydro-
philic binding site at the receptor.
fects and to determine pEC50 values in CHO (SUR1/Kir6.2) and
HEK 293 (SUR2B/Kir6.1) cells.[13] Membrane hyperpolarization
by increasing the outward K+ current is a necessary prerequi-
site to decrease the influx of Ca2+ ions, which, in turn, decreas-
es the insulin release of the b cell (SUR1/Kir6.2) and the relaxa-
tion of smooth muscle (SUR2B/Kir6.1). Membrane potential can
be measured by fluorescence using fluorescent dyes such as
DiBAC4(3) and DyeB (R7260) which makes it possible to test a
manifold of cells of the same age cultivated in 12-well strips,
thereby increasing the reliability of results with decreasing ex-
perimental effort. The reliability of membrane hyperpolariza-
tion for the detection of agonistic versus antagonistic effects
has been highlighted.[14]
However, the affinity of neither benzothiazoles 17a and 17b
nor derivatives 17d and 17e, with the electron-withdrawing
trifluoromethyl substituent at position 5, surpassed that of
benzothiophene 15 or of bromo-substituted benzofuran 10g
(Entry 12 vs. Entries 13, 14, 16, and 17). Here again, the N-
phenyl unit diminished the affinity substantially (Entry 15).
Thus we turned to benzothiazoles containing a spacer
carbon atom between the heterocyclic ring and the thioamide
moiety, guided by the idea that this pattern might provide
more flexibility in the pharmacophoric units thus permitting a
better fit and enhanced affinity at the SUR2B receptor. Whereas
compound 19a, with a methylene spacer, showed a pKD value
of 4.31 (Entry 18), benzothiazoles 19b and 19c, featuring a
branched spacer, led to a substantial increase in affinity (En-
tries 19 and 20). Finally, the combination of a branched spacer
with a bulky tert-butyl residue at the nitrogen atom provided
the highest affinity in the single-digit micromolar range
(Entry 21). The low affinity of the carboxylic amides 20a and
20b clearly reveals the role of the carbothioamide group as a
true pharmacophore. Thus, the pKD values of compounds 19d
and 20b, with a substitution of sulfur by oxygen and other-
wise identical skeleton, differ by more than one log unit
(Entry 21 vs. 23). For reasons of comparison, the pKD value of
the drug diazoxide (2) is also included (Entry 24).[12] Finally, car-
bamate 13 represents a swap of the lipophilic and hydrophilic
regions relative to the benzofurans 10; this dramatic modifica-
tion is deleterious to the affinity (Entry 25).
Several compounds that showed significant affinity accord-
ing to the data obtained from radioligand binding (Table 1)
were assayed for KATP activity by means of the membrane
polarization measurements, the results of which are listed in
Table 2. For reasons of comparison, the corresponding data for
diazoxide (2) are also included. The representative experiment
with thioamide 19d displayed in Figure 2b demonstrates the
exponential decline of the membrane potential with time,
induced by increasing concentrations of 19d in HEK 293
(SUR2B/Kir6.1) cells. The agonistic effects of compounds were
reversed upon addition of glibenclamide (not shown), a KATP
channel inhibitor, confirming a KATP mechanism. Remarkably,
benzothiazole 19d (at 25 mm) failed to induce membrane hy-
perpolarization in SUR1-type KATP channels (Figure 2a), whereas
it was effective in hyperpolarizing membrane potential in the
SUR2B-type channels.
Most of the compounds were characterized by a two- to
threefold (i.e., 0.3–0.5 log units) higher functional potency
(EC50) than binding affinity measured as pKD (Table 1), reflecting
the fact that in artificial cell lines that overexpress KATP chan-
nels, only part of the functional KATP channels are necessary for
membrane hyperpolarization, whereas all of the binding sites
equally contribute to radioligand binding. In all cases, the
membrane potentials were determined using DiBAC4(3). In
addition, measurements with benzothiazoles 19b–d were also
performed with DyeB, an alternative membrane potential dye,
which turned out to be less sensitive to quenching effects.[15]
In a recent study on modified diazoxides, we had observed
that sterically demanding lipophilic residues at the heterocyclic
ring caused not only enhanced affinity but also remarkable se-
lectivity between the SUR2B/Kir6.1 and the SUR1/Kir6.2 recep-
tors.[12] Therefore, we were also
Table 2. Membrane hyperpolarization (pEC50 values and selectivity) at HEK 293 and CHO cells.
interested to determine the se-
lectivity of the current com-
pounds. However, with no ago-
nist radioligands for SUR1-type
KATP channels yet available which
would permit the determination
of KD values for SUR1-agonists,
the selectivity of compounds for
different channel types must be
measured by functional meth-
ods. The measurement of ago-
nist-induced membrane hyper-
polarization is an adequate
method to discriminate between
agonistic and antagonistic ef-
[a]
Entry
Compd
pEC50
HEK 293(SUR2B/Kir6.1)
Selectivity [D][b]
CHO(SUR1/Kir6.2)
1
2
3
4
5
6
7
2
15
5.14ꢁ0.02
4.72ꢁ0.02
4.62ꢁ0.02
3.82ꢁ0.03
5.4ꢁ0.06%[d]
6.3ꢁ1.1%[e]
2.5ꢁ1.5%[f]
7.8ꢁ1.0%[h]
0.42
0.18
0.90
4.80ꢁ0.03
10d
19b
19c
19d
20b
4.72ꢁ0.04
5.07ꢁ0.02 (4.93ꢁ0.02)[c]
5.08ꢁ0.03 (5.14ꢁ0.03)[c]
5.41ꢁ0.03 (5.45ꢁ0.03)[c]
3.97ꢁ0.02[c,g]
>1.4
>1.3
>2.0
[i]
–
[a] Membrane potentials determined with DiBAC4(3); pEC50 values (ꢀlogm) ꢁASD calculated by nonlinear re-
gression according to Equation (2); see Experimental Section. [b] Selectivity calculated as D=pEC50(SUR2B/
Kir6.1)ꢀpEC50(SUR1/Kir6.2). [c] Membrane potential determined with DyeB. [d] pEC50 =3.45ꢁ0.09 at 25 mm (extrapo-
lated). [e] pEC50 =3.69ꢁ0.03 at 25 mm (extrapolated). [f] pEC50 =3.30ꢁ0.02 at 25 mm (extrapolated). [g] Extrapo-
lated. [h] pEC50 =3.5ꢁ0.1 at 25 mm (extrapolated). [i] D value not given because both pEC50 values are extrapo-
lated.
1752
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1749 – 1759

KATP Channel Openers
with the “spacered”, tert-butyl-substituted compound 19d,
which not only displays the highest affinity in the series but
also the highest selectivity. Its affinity for the SUR2B/Kir6.1
receptor of HEK 293 cells surpasses that for the CHO (SUR1/
Kir6.2) cells by more than two orders of magnitude. In a sense,
this result parallels our recent finding that the introduction of
bulky, lipophilic residues in the side chain of diazoxide leads to
a remarkable enhancement in selectivity in favor of the SUR2B
receptor. These results are illustrated for selected compounds,
benzothiophene 15, benzofuran 10d, and benzothiazole 19d
in comparison with diazoxide (2) in Figure 3, which displays
the measurement of the membrane potential of HEK 293
(SUR2B/Kir6.1) and CHO (SUR1/Kir6.2) cells. In particular, it
shows the remarkable selectivity of thioamide 19d. Here
again, the crucial role played by the carbothioamide pharma-
cophore is illustrated by the marginal membrane polarization
of the carboxylic amide 20b that differs from thioamide 19d
only in the exchange of sulfur for oxygen (Entry 7). Comparison
of the membrane hyperpolarization with HEK 293 and CHO
cells also reveals that replacement of the carbothioamide
group with a carboxylic amide is not only deleterious to affini-
ty but also provides only marginal selectivity (Entry 7 vs. 6).
As the most potent compounds 19b–d are used as racemic
mixtures, the question of differences in biological activity be-
tween individual enantiomers might arise. However, due to
facile enolization, racemization under physiological conditions
is expected to occur. This was proven by keeping compound
19d in a neutral solution of DMSO and D2O at room tempera-
ture for one day: a complete deuterium exchange occurring at
the stereogenic carbon center (in addition to the carbothio-
amide nitrogen atom) under these conditions clearly proves
enolization and, in an indirect manner, the configurational in-
stability at the stereogenic center.
Figure 2. Representative experiments demonstrating the measurement of
membrane hyperpolarization in a) CHO cells transfected with SUR1/Kir6.2
and b) HEK 293 cells transfected with SUR2B/Kir6.1 with thioamide 19d.
Membrane potential was measured by fluorescence using the bis-oxonol
dye DiBAC4(3). Maximum effects were induced with the thienothiadiazine
derivative NNC414 and with the cyanoguanidine P1075 in SUR1- and SUR2B-
type channels, respectively. Changes in DiBAC4(3) fluorescence (arbitrary
units, data sampling at a frequency of 1 Hz) were normalized to maximum
effects. Note that 25 mm 19d in panel a) coincides with its ꢀlog value of 4.6
in panel b).
Crystal structure analyses were obtained for two of the most
potent compounds, 19b and 19d. They are shown in Figure 4.
The hydrogen-bond donor (NH of the thioamide group) and
concomitant hydrogen-bond acceptor (the thiazole nitrogen
atom) lead to a dimeric arrangement in the unit cell. Thus, the
crystal structures, in combination with the structure–activity re-
lationship, reveal three features that might be important for
the affinity and selectivity as KATP channel openers: 1) the hy-
drogen-bonding domain, 2) a region of high electron density
arising from the lone pairs at the sulfur atoms, and 3) the lipo-
philic side chain at the thioamide moiety. A model based
thereupon is illustrated in Figure 5 for thioamide 19d. Work
intended to evaluate and verify the three-binding-sites model
through various heterocyclic skeletons and ring sizes is in
progress.
In comparing the drug diazoxide (2) with benzothiophene
15 regarding affinity and selectivity toward the two receptors,
SUR2B/Kir6.1 and SUR1/Kir6.2, it is evident that 15 is inferior in
both respects to diazoxide (Table 2, Entries 1 and 2). Although
replacement of sulfur with oxygen in the five-membered het-
erocycles does not improve affinity (Table 2, Entry 2 vs. 3), the
membrane potential measurement of benzofuran 10d reveals
a substantial increase in selectivity in favor of the SUR2B-type
KATP channels (Table 2, Entry 3). This effect may be caused by
the sterically demanding, lipophilic tert-butyl residue at the
thioamide moiety of the benzofuran 10d. Another favorable
effect came from the combination of a spacer carbon atom
between the heterocyclic skeleton and the thioamide residue
and a lipophilic group at the thioamide nitrogen atom. These
structural features were realized in benzothiazoles 19b–d
(Table 2, Entries 4–6). Thus, the optimum result was obtained
Conclusion
In summary, this is the first demonstration that the heterocyclic
skeletons of benzofuran, benzothiophene, and benzothiazole
combined with a thioamide moiety provide compounds with
substantial activity as KATP channel openers. With regard to af-
finity, the best results were obtained if the benzothiazole het-
erocycle is linked to the thioamide group through a branched
ChemMedChem 2010, 5, 1749 – 1759
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1753

M. Braun et al.
MED
Figure 5. Three-binding-sites model of the PCO 19d.
selectivity. They clearly prevent benzothiazoles from
binding to the SUR1 receptor, so that the selectivity
for SUR2B-type KATP channels is >100-fold higher.
This selectivity distinguishes the benz-annulated
compounds also developed from the established
drug diazoxide (2) that is in clinical use.
Experimental Section
Syntheses
Figure 3. Comparison of the effects of a) diazoxide (2), b) benzothiophene 15, c) benzo-
furan 10d, and d) benzothiazole 19d on the membrane potential (E) of HEK 293 cells
Melting points (uncorrected) were determined with a
Bꢁchi melting point apparatus 540. IR spectra: Bruker
Vector 22. NMR spectra (d values in ppm): Bruker AM-
200-SY and AM-500. Mass spectra: Finnigan MAT 8200.
TLC: Silica gel 60 F254 (Merck). Column chromatography:
Macherey–Nagel Kieselgel 60 and Merck Kieselgel 60,
mesh size 0.04–0.063 mm. Elemental analyses were car-
ried out with a Perkin–Elmer CHN Analyzer 2400 at the
~
*
transfected with SUR2B/Kir6.1 ( ) and on CHO cells transfected with SUR1/Kir6.2 ( ). For
compound 19d no significant hyperpolarization could be measured in CHO (SUR1/Kir6.2)
cells; the dashed line represents a hypothetical curve according to Eq. (1), indicating an
upper limit for the effectiveness of 19d on SUR1-type KATP channels.
Institute of Pharmaceutical Chemistry (University of Dꢁsseldorf). All
reactions involving organometallic compounds were carried out
under an atmosphere of anhydrous N2. THF and Et2O were pre-
dried with KOH and distilled under N2 from sodium/benzophe-
none. They were taken from the distillation flask, which was closed
by a septum, with syringes or cannulae. Reactions at temperatures
<08C were monitored by a thermocouple connected to a resist-
ance thermometer (Ebro).
General procedure for the preparation of benzofuran-2-carbothio-
amides 10a–e: A 100 mL two-necked flask was equipped with a
magnetic stirrer, a connection to a combined N2/vacuum line and
closed with a septum. The air in the flask was replaced by N2, and
benzofuran 7a or 7c (20.0 mmol), dissolved in 15 mL dry THF, was
injected by syringe. While stirring, the mixture was cooled to
ꢀ788C in a dry-ice/acetone bath. A 1.6m solution of n-butyllithium
in hexane (12.5 mL, 20.0 mmol) was injected slowly. After stirring
for 1 h at the same temperature, the corresponding isothiocyanate
9a–e, dissolved in 2 mL dry THF, was added by syringe. The solu-
tion was allowed to reach room temperature while stirring over-
night. After the addition of pentane (50 mL), the mixture was fil-
tered and the filtrate was concentrated in a rotary evaporator. The
oily residue was purified by column chromatography.
Figure 4. Diagrams of the molecular structures of a) 19b and b) 19d; dis-
placement ellipsoids are drawn at the 50% probability level, radii of H
atoms are chosen arbitrarily, and only H atoms involved in hydrogen bond-
ing are labeled. Selected geometric parameters, 19b [19d] bond lengths {in
ꢂ}: S1ꢀC7 1.721(2) [1.736(3)], S1ꢀC1 1.736(2) [1.731(2)], S2ꢀC10 1.658(2)
[1.658(2)], N1ꢀC1 1.291(3) [1.303(2)], N1ꢀC2 1.395(3) [1.396(3)], N2ꢀC10
1.320(3) [1.317(3)], N2ꢀC11 1.464(3) [1.493(3)], C1ꢀC8 1.511(3) [1.510(3)], C2ꢀ
C3 1.387(3) [1.403(3)], C2ꢀC7 1.395(3) [1.387(3)], C3ꢀC4 1.371(4) [1.373(4)],
C4ꢀC5 1.381(4) [1.383(4)], C5ꢀC6 1.370(4) [1.368(4)], C6ꢀC7 1.400(3)
[1.393(3)], C8ꢀC9 1.524(3) [1.530(3)], C8ꢀC10 1.537(3) [1.549(3)], C11ꢀC12
1.481(4) [1.526(3)], C11ꢀC13 [1.530(3)], C11ꢀC14 [1.519(3)], N2ꢀH1 0.85(2)
[0.89(2)], H1···N1’ 2.16(2) [2.24(3)]; bond angles {in 8}: N2ꢀH1···N1’ 178(2)
[162(2)], S1ꢀC1ꢀC8ꢀC9 ꢀ14.7(3) [ꢀ56.8(2)].
According to this procedure, the following compounds were
obtained:
N-Methylbenzofuran-2-carbothioamide
(10a)
from
7a
(20.0 mmol) and methyl isothiocyanate (9a) (20.0 mmol): Yield:
3.4 g (89%); mp: 123.68C; Rf =0.5 (pentane/EtOAc 1:1); 1H NMR
(500 MHz, CDCl3): d=3.32 (d, J=4.73 Hz, 3H, CH3), 7.22 (d, J=
7.88 Hz, 1H, 6-H), 7.33–7.40 (m, 2H, 4-H and 5-H), 7.59 (d, J=
7.88 Hz, 1H, 7-H), 7.66 (s, 1H, 3-H), 8.27 (brs, 1H, NH); 13C NMR
one-carbon spacer. The concept of introducing sterically de-
manding, lipophilic residues turned out to be fruitful for high
1754
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1749 – 1759

KATP Channel Openers
(125 MHz, CDCl3): d=32.5 (CH3), 112.1 (C4), 113.8 (C3), 123.2 (C7),
124.3 (C6), 127.9 (C3’), 128.5 (C5), 153.0 (C7’), 154.9 (C2), 184.5 (C=
S); IR (KBr): n˜ =3324.7, 2362.2, 1578.6, 1526.3, 1474.2, 1444.3,
1363.4, 1291.7, 1161.5, 1121.5, 1047.2, 977.7, 939.4, 886.0, 831.3,
781.2, 751.3, 691.6, 612.8, 502.4 cmꢀ1; MS (EI, 70 eV): m/z (%)=192
([M+1]+, 13), 191 ([M]+, 100), 162 (24), 161 (32), 158 (18), 150 (35),
131 (18), 121 (12), 118 (35), 90 (11), 89 (24), 63 (9); Anal. calcd for
C10H9NOS (191.25): C 62.80, H 4.74, N 7.32, found: C 62.72, H 4.62,
N 7.11.
N-Ethyl-3-methyl-5-trifluoromethylbenzofuran-2-carbothioamide
(10h) from 7c (1.37 mmol) and ethyl isothiocyanate (9b)
(1.37 mmol): Yield: 0.35 g (88%); mp: 109.68C; H NMR (500 MHz,
1
CDCl3): d=1.33 (t, J=7.3 Hz, 3H, CH2-CH3), 2.73 (s, 3H, 3-CH3), 3.80
(dq, J=5.9 Hz, J=7.1 Hz, 2H, CH2), 7.45 (d, J=8.8 Hz, 1H, 6-H),
7.61 (d, J=8.2 Hz, 1H, 7-H), 7.84 (s, 1H, 4-H), 8.81 (brs, 1H, NH);
13C NMR (125 MHz, CDCl3): d=10.5 (3-CH3), 12.4 (CH2-CH3), 38.6
(CH2), 110.9 (C4), 117.9 (C3), 118.0 (C7), 123.1 (C6), 123.5 (C3’), 129.6
(CF3), 145.8 (C7’), 151.8 (C2), 182.1 (C=S); MS (EI, 70 eV): m/z (%)=
288 ([M+1]+, 16), 287 ([M]+, 100), 286 (16), 268 (6), 255 (7), 254
(43), 244 (14), 243 (30), 242 (13), 231 (9), 229 (5), 227 (20), 226 (73),
225 (9), 224 (6), 200 (7), 199 (19), 170 (5), 169 (6), 159 (7), 151 (15),
44 (7).
N-Ethylbenzofuran-2-carbothioamide (10b) from 7a (10.0 mmol)
and ethyl isothiocyanate (9b) (10.0 mmol): Yield: 1.9 g (93%); mp:
1
272.78C; Rf =0.5 (CHCl3); H NMR (500 MHz, CDCl3): d=1.35 (t, J=
7.4 Hz, 3H, CH3), 3.84 (dq, J=5.9 Hz, J=7.0 Hz, 2H, CH2), 7.22 (t,
J=7.7 Hz, 1H, 6-H), 7.36 (t, J=8.0 Hz, 1H, 5-H), 7.40 (d, J=8.5 Hz,
1H, 4-H), 7.59 (d, J=7.9 Hz, 1H, 7-H), 7.66 (s, 1H, 3-H), 8.13 (s (br),
1H, NH); 13C NMR (125 MHz, CDCl3): d=13.9 (CH3), 40.6 (CH2), 112.1
(C4), 114.0 (C3), 123.2 (C7), 124.3 (C6), 127.9 (C3’), 128.5 (C5), 152.9
(C7’), 154.9 (C2), 183.3 (C=S); MS (EI, 70 eV): m/z (%)=206 ([M+1]+,
17), 205 ([M]+, 100), 204 (27), 172 (12), 161 (37), 144 (42), 118 (33),
89 (22), 44 (14); Anal. calcd for C11H11NOS (205.28): C 64.39, H 5.36,
N 6.82, found: C 64.15, H 5.48, N 6.93.
General procedure for the preparation of 5-bromobenzofuran-2-car-
bothioamides 10 f,g: A 100 mL two-necked flask was equipped
with a magnetic stirrer, a connection to a combined N2/vacuum
line and closed with a septum. The air in the flask was replaced by
N2, and diisopropylamine (0.22 mL, 0.159 g, 1.58 mmol) and dry
THF (10 mL) were injected by syringes. After cooling to ꢀ788C by
means of a dry-ice/acetone bath, a 1.6m solution of n-butyllithium
in hexane (0.99 mL, 1.58 mmol) was injected and stirring was con-
tinued for 30 min at 08C. Thereafter, the mixture was cooled to
ꢀ158C. A solution of 5-bromobenzofuran 7b (0.300 g, 1.52 mmol)
and THF (5 mL) generated under N2, was added slowly by syringe.
Stirring was continued for 2 h at the same temperature. Then, iso-
thiocyanate 9a or 9b (1.2 mmol), dissolved in 2 mL dry THF, was
added, and stirring was continued at ꢀ158C for 2 h. The mixture
was poured into H2O and extracted with CHCl3 (3ꢃ150 mL). The
combined organic layers were dried with Na2SO4 and concentrated
in a rotary evaporator. The oily residue was purified by column
chromatography.
N-n-Butylbenzofuran-2-carbothioamide
(10c)
from
7a
(4.64 mmol) and n-butyl isothiocyanate (9c) (4.64 mmol): Yield:
1.07 g (99%); mp: 49.28C; Rf =0.6 (CHCl3); 1H NMR (500 MHz,
CDCl3): d=0.92 (t, J=7.4 Hz, 3H, CH3), 1.41 (sext, J=7.6 Hz, 2H,
CH2-CH3), 1.70 (quint, J=7.4 Hz, 2H, CH2-CH2-CH3) 3.79 (q, J=
6.5 Hz, 2H, -NH-CH2), 7.20 (t, J=7.4 Hz, 1H, 6-H), 7.34 (t, J=7.4 Hz,
1H, 5-H), 7.39 (d, J=8.5 Hz, 1H, 4-H), 7.57 (d, J=7.9 Hz, 1H, 7-H),
7.64 (s, 1H, 3-H), 8.15 (brs, 1H, NH); 13C NMR (125 MHz, CDCl3): d=
14.2 (CH3), 20.7 (CH2-CH3), 30.7 (CH2-CH2-CH3), 45.5 (NH-CH2), 112.1
(C4), 114.0 (C3), 123.1 (C7), 124.2 (C6), 127.8 (C3’), 128.5 (C5), 153.0
(C7’), 154.9 (C2), 183.4 (C=S); MS (EI, 70 eV): m/z (%)=234 ([M+1]+,
14), 233 ([M]+, 85), 200 (51), 191 (52), 162 (24), 161 (100), 150 (27),
144 (65), 118 (10), 89 (21); Anal. calcd for C13H15NOS (233.33): C
66.92, H 6.48, N 6.00, found: C 66.72, H 6.69, N 6.00.
According to this procedure, the following compounds were
obtained:
5-Bromo-N-methylbenzofuran-2-carbothioamide (10 f) from 7b
(1.58 mmol) and methyl isothiocyanate (9a) (1.58 mmol): Yield:
120 mg (29%); mp: 170.48C; Rf =0.4 (CHCl3); 1H NMR (500 MHz,
CDCl3): d=3.39 (d, J=4.7 Hz, 3H, CH3), 7.34 (d, J=8.8 Hz, 1H, 7-H),
7.52 (dd, J=2.2 Hz, J=8.8 Hz, 1H, 6-H), 7.66 (d, J=0.6 Hz, 1H, 3-H),
7.80 (d, J=1.9 Hz, 1H, 4-H), 8.32 (brs, 1H, NH); 13C NMR (125 MHz,
CDCl3): d=32.6 (CH3), 112.8 (C3), 113.6 (C5), 117.4 (C7), 125.6 (C4),
130.3 (C3’), 130.8 (C6), 153.5 (C7’), 153.8 (C2), 184.0 (C=S); MS (EI,
70 eV): m/z (%)=272 ([M+1]+, 12), 271 ([M]+, 86), 270 (25), 269
(100), 242 (12), 241 (11), 240 (14), 239 (19), 238 (15), 236 (16), 230
(51), 229 (10), 228 (41), 227 (14), 211 (9), 209 (10), 198 (23), 196
(27), 149 (15), 121 (13), 95 (15), 89 (22), 88 (15), 87 (14), 74 (13), 44
(12); Anal. calcd for C10H8BrNOS (270.15): C 44.46, H 2.98, N 5.18,
found: C 44.11, H 2.35, N 4.72.
N-tert-Butylbenzofuran-2-carbothioamide
(10d)
from
7a
(4.64 mmol) and tert-butyl isothiocyanate (9d) (4.64 mmol): Yield:
241 mg (22%); mp: 80.38C; Rf =0.6 (CHCl3); 1H NMR (500 MHz,
CDCl3): d=1.26 (s, 9H, CH3), 7.19 (t, J=7.4 Hz, 1H, 6-H), 7.33 (t, J=
8.0 Hz, 1H, 5-H), 7.38 (d, J=8.5 Hz, 1H, 4-H), 7.56 (d, J=7.6 Hz, 1H,
7-H), 7.62 (s, 1H, 3-H), 8.04 (brs, 1H, NH); 13C NMR (125 MHz,
CDCl3): d=28.4 (-CH3), 56.3 (-C(CH3)3), 112.1 (C4), 113.3 (C3), 123.0
(C7), 124.2 (C6), 127.6 (C3’), 128.7 (C5), 153.9 (C7’), 154.7 (C2), 182.3
(C=S); MS (EI, 70 eV): m/z (%)=234 ([M+1]+, 15), 233 ([M]+, 86),
232 (41), 177 (33), 161 (23), 150 (30), 145 (13), 144 (100), 143 (11),
89 (13); Anal. calcd for C13H15NOS (233.33): C 66.92, H 6.48, N 6.00,
found: C 66.89, H 6.56, N 5.90.
N-Phenylbenzofuran-2-carbothioamide
(10e)
from
7a
(10.1 mmol) and phenyl isothiocyanate (9e) (10.1 mmol): Yield:
2.42 g (95%); mp: 152.68C; Rf =0.8 (CHCl3); 1H NMR (500 MHz,
CDCl3): d=7.21–7.26 (m, 2H, phenyl o-H), 7.36–7.42 (m, 3H, phenyl
m,p-H), 7.45 (d, J=8.5 Hz, 1H, 4-H), 7.61 (d, J=7.9 Hz, 1H, 7-H),
7.75 (s, 1H, 3-H), 7.83 (d, J=8.2 Hz, 2H, 5-H and 6-H), 9.68 (brs,
1H, NH); 13C NMR (125 MHz, CDCl3): d=112.2 (C4), 115.1 (C3), 123.2
(C7), 124.0 (C6), 124.5 (C3’), 127.5 (phenyl C-2), 128.2 (phenyl C-4),
128.7 (C5), 129.4 (phenyl C-3), 138.4 (phenyl C-1), 153.5 (C7’), 154.8
(C2), 181.4 (C=S); MS (EI, 70 eV): m/z (%)=254 ([M+1]+, 20), 253
([M]+, 97), 252 (93), 220 (42), 161 (100), 144 (15), 110 (15), 101 (11),
89 (15); Anal. calcd for C15H11NOS (253.32): C 71.12, H 4.38, N 5.53,
found: C 70.39, H 4.46, N 5.54.
5-Bromo-N-ethylbenzofuran-2-carbothioamide (10g) from 7b
(1.35 mmol) and ethyl isothiocyanate (9b) (1.35 mmol): Yield:
78 mg (20%); Rf =0.5 (CHCl3); H NMR (500 MHz, CDCl3): d=1.34 (t,
1
J=7.4 Hz, 3H, CH3), 3.82 (dq, J=5.5 Hz, J=7.3 Hz, 2H, CH2), 7.28
(d, J=8.8 Hz, 1H, 7-H), 7.44 (dd, J=2.2 Hz, J=8.8 Hz, 1H, 6-H), 7.59
(s, 1H, 3-H), 7.72 (d, J=1.9 Hz, 1H, 4-H), 8.11 (brs, 1H, NH);
13C NMR (125 MHz, CDCl3): d=13.7 (CH3), 31.3 (CH2), 112.8 (C3),
113.6 (C5), 117.4 (C7), 125.6 (C4), 130.3 (C3’), 130.8 (C6), 153.5 (C7’),
153.8 (C2), 184.0 (C=S); MS (EI, 70 eV): m/z (%)=286 ([M+1]+, 14),
285 ([M]+, 99), 284 (34), 283 (100), 282 (23), 252 (13), 250 (15), 242
(35), 241 (24), 240 (35), 239 (23), 225 (12), 224 (36), 223 (17), 222
(35), 198 (29), 196 (30), 143 (11), 89 (11), 44 (27).
ChemMedChem 2010, 5, 1749 – 1759
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1755

M. Braun et al.
MED
3-Hydroxy-N-ethylbenzofuran-2-carbothioamide (10i): A 100 mL
two-necked flask, equipped with a magnetic stirrer and a connec-
tion to a combined N2/vacuum line, was charged with NaH (60%
dispersion paraffin, 1 g, 25 mmol). The flask was closed with a
septum and the air in the flask was replaced by N2. Dry THF
(20 mL) was added. Under stirring in an ice bath, a solution of 11
(1.0 g, 7.46 mmol) in dry THF (15 mL) was added by syringe. After
stirring at room temperature for 30 min, ethyl isothiocyanate 9b
(0.65 g, 7.46 mmol) was injected. After stirring at room tempera-
ture for 18 h, cold brine (5 mL) and 2n HCl (20 mL) were added.
The mixture was extracted with CHCl3 (3ꢃ150 mL). The combined
organic layers were washed with 50 mL portions of dilute NaOH
and 1n HCl and deionized H2O and dried with Na2SO4. The solvent
was removed in a rotary evaporator and the residue was exposed
to oil-pump vacuum for 12 h to give pale-yellow, oily 10i. Yield:
1.6 g (97%); 1H NMR (500 MHz, CDCl3): d=1.31 (t, J=7.3 Hz, 3H,
CH3), 3.76 (dq, J=6.1 Hz, J=7.1 Hz, 2H, CH2), 7.21 (t, J=7.7 Hz, 1H,
5-H), 7.28 (d, J=8.2 Hz, 1H, 6-H), 7.41 (t, J=8.0 Hz, 1H, 4-H), 7.49
(brs, 1H, NH), 11.16 (s, 1H, OH); 13C NMR (125 MHz, CDCl3): d=14.2
(CH3), 38.8 (CH2), 112.2 (C7), 121.4 (C4), 122.1 (C5), 123.6 (C6), 130.0
(C3’), 130.9 (C2), 151.0 (C3), 151.4 (C7’), 180.8 (C=S); MS (EI, 70 eV):
m/z (%)=222 ([M+1]+, 13), 221 ([M]+, 100), 188 (27), 177 (11), 176
(21), 160 (62), 159 (15), 121 (18); Anal. calcd for C10H8BrNOS
(221.28): C 59.71, H 5.01, N 6.33, found: C 59.84, H 4.95, N 6.36.
magnetic stirrer, a connection to a combined N2/vacuum line and
closed with a septum. The air in the flask was replaced by N2, and
benzothiazole 16a or 16b (9.1 mmol), dissolved in 15 mL of dry
THF, was injected by syringe. By means of a liquid N2/pentane
bath, the solution was cooled to ꢀ1008C, and a 1.6m solution of
n-butyllithium in hexane (5.7 mL, 9.1 mmol) was added at such a
rate that the temperature, monitored by an electronic inlet ther-
mometer, did not exceed ꢀ908C. After warming to ꢀ758C under
stirring, isothiocyanates 9a, 9b, or 9e (9.1 mmol) were injected.
Stirring was continued at the same temperature for 20 min. H2O
(20 mL) was then added and the mixture was extracted with CHCl3
(3ꢃ150 mL). The combined organic layers were dried with Na2SO4
and the solvent was removed in a rotary evaporator. The residue
was submitted to column chromatography or recrystallization from
EtOH (17d).
According to this procedure, the following compounds were
obtained:
N-Methyl-1,3-benzothiazole-2-carbothioamide (17a) from 15a
(9.1 mmol) and methyl isothiocyanate (9a) (9.1 mmol): Yield:
366 mg (19%); mp: 120.28C; Rf =0.8 (CHCl3); 1H NMR (500 MHz,
CDCl3): d=3.33 (d, J=5.0 Hz, 3H, CH3), 7.39 (t, J=7.4 Hz, 1H, 6-H),
7.46 (t, J=8.0 Hz, 1H, 5-H), 7.85 (d, J=8.2 Hz, 1H, 7-H), 7.95 (d, J=
8.2 Hz, 1H, 4-H), 9.28 (brs, 1H, NH); 13C NMR (125 MHz, CDCl3): d=
33.0 (CH3), 122.4 (C7), 124.9 (C4), 126.9 (C6), 127.4 (C5), 139.3 (C7’),
153.5 (C3’), 168.8 (C2), 186.1 (C=S); MS (EI, 70 eV): m/z (%)=210
([M+2]+, 6), 209 ([M+1]+, 9), 208 ([M]+, 67), 181 (9), 180 (14), 179
(100), 178 (13), 175 (7), 167 (8), 135 (24), 134 (5), 108 (11), 102 (6),
74 (5); Anal. calcd for C9H8N2S2 (208.3): C 51.89, H 3.87, N 13.45,
found: C 51.92, H 3.70, N 13.23.
(O-2,2-Diallyl-2,3-dihydrobenzofuran-3-yl)ethylcarbamothioate
(13): A 100 mL flask was charged with 11 b (0.9 g, 4.2 mmol), which
was dissolved in dry EtOH (25 mL). Within 1 h, NaBH4 (1.91 g,
50.4 mmol) was added in small portions at 08C. The mixture was
hydrolyzed with 20 mL H2O and extracted with Et2O (3ꢃ75 mL).
The combined organic layers were dried with Na2SO4 and the sol-
vent was removed in a rotary evaporator to obtain the correspond-
ing alcohol. Yield: 0.81 g (89%); 1H NMR (500 MHz, CDCl3): d=
2.30–2.71 (m, 4H, CH2), 4.98–5.20 (m, 4H, =CH2), 5.23 (d, J=
17.3 Hz, 1H, CH-OH), 5.72–6.14 (m, 2H, allyl-CH), 6.80–6.84 (m, 1H,
7-H), 6.89–6.94 (m, 1H, 5-H), 7.22–7.28 (m, 1H, 4-H), 7.34–7.39 (m,
1H, 6-H); MS (EI, 70 eV): m/z (%)=216 ([M]+, 90), 174 (41), 147 (34),
133 (100), 121 (39), 91 (50), 77 (40).
N-Ethyl-1,3-benzothiazole-2-carbothioamide (17b) from 15a
(9.1 mmol) and ethyl isothiocyanate (9b) (9.1 mmol): Yield: 0.83 g
1
(41%); mp: 84.18C; Rf =0.8 (CHCl3); H NMR (500 MHz, CDCl3): d=
1.35 (t, J=7.3 Hz, 3H, CH3), 3.80 (dt, J=7.2 Hz, J=13.0 Hz, 2H,
CH2), 7.37 (t, J=7.7 Hz, 1H, 6-H), 7.44 (t, J=8.0 Hz, 1H, 5-H), 7.83
(d, J=8.2 Hz, 1H, 7-H), 7.94 (d, J=8.2 Hz, 1H, 4-H), 9.17 (brs, 1H,
NH); 13C NMR (125 MHz, CDCl3): d=13.7 (CH3), 41.2 (CH2), 122.4
(C7), 125.0 (C4), 126.9 (C6), 127.3 (C5), 139.4 (C7’), 153.5 (C3’), 169.1
(C2), 184.8 (C=S); MS (EI, 70 eV): m/z (%)=222 ([M]+, 55), 181 (9),
180 (13), 179 (100), 178 (16), 161 (12), 135 (21), 108 (7), 102 (6);
Anal. calcd for C10H10N2S2 (222.33): C 54.02, H 4.53, N 12.60, found:
C 54.12, H 4.56, N 12.42.
A 100 mL two-necked flask was equipped with a magnetic stirrer
and a connection to a combined N2/vacuum line. The flask was
closed with a septum and the air in the flask was replaced by N2.
The crude alcohol (0.4 g, 1.85 mmol) was added and dissolved in
dry Et2O (25 mL). The resulting solution was cooled to ꢀ788C by
means of a dry-ice/acetone bath. Under stirring at this tempera-
ture,
a 1.6m solution of n-butyllithium in hexane (1.4 mL,
2.22 mmol) was added by syringe and stirring was continued for
1 h. Then 9b (0.16 g, 1.85 mmol) was injected. The solution was al-
lowed to reach room temperature under stirring overnight. The
mixture was hydrolyzed with 20 mL H2O and extracted with Et2O
(3ꢃ75 mL). The combined organic layers were dried with Na2SO4
and the solvent was removed in a rotary evaporator. The residue
was purified by column chromatography. Yield: 0.16 g (29%); Rf =
0.56 (CHCl3); 1H NMR (500 MHz, CDCl3): d=1.31 (t, J=7.3 Hz, 3H,
CH3), 3.76 (dq, J=6.1 Hz, J=7.1 Hz, 2H, CH2), 7.21 (t, J=7.7 Hz, 1H,
5-H), 7.28 (d, J=8.2 Hz, 1H, 6-H), 7.41 (t, J=8.0 Hz, 1H, 4-H), 7.49
(brs, 1H, NH), 11.16 (s, 1H, OH); 13C NMR (125 MHz, CDCl3): d=14.2
(CH3), 38.8 (CH2), 112.2 (C7), 121.4 (C4), 122.1 (C5), 123.6 (C6), 130.0
(C3’), 130.9 (C2), 151.0 (C3), 151.4 (C7’), 180.8 (C=S); MS (EI, 70 eV):
m/z (%)=303 ([M]+, 4), 262 (12), 199 (32), 191 (52), 163 (11), 158
(100), 129 (45); Anal. calcd for C17H21NO2S (303.42): C 67.29, H 6.98,
N 4.62, found: C 67.28, H 7.22, N 4.69.
N-Phenyl-1,3-benzothiazole-2-carbothioamide (17c) from 15a
(9.1 mmol) and phenyl isothiocyanate (9e) (9.1 mmol): Yield: 0.76 g
(45%); mp: 156.28C; Rf =0.8 (CHCl3); H NMR (500 MHz, CDCl3): d=
7.23 (tt, J=0.5 Hz, J=7.4 Hz, 1H, phenyl 4-H), 7.37–7.42 (m, 3H,
phenyl 2-H and 3-H), 7.47 (dt, J=1.3 Hz, J=7.7 Hz, 1H, phenyl 3-
H), 7.85 (d, J=7.6 Hz, 1H, 7-H), 7.99 (dd, J=7.9 Hz, J=12.3 Hz, 3H,
4-H, 5-H and 6-H), 10.92 (brs, 1H, NH); MS (EI, 70 eV): m/z (%)=
271 ([M+1]+, 21), 270 ([M]+, 78), 269 (100), 238 (10), 237 (55), 167
(22), 161 (15), 135 (23), 109 (12), 108 (11), 77 (16).
1
N-Ethyl-5-(trifluoromethyl)benzothiazole-2-carbothioamide
(17d) from 15b (0.49 mmol) and ethyl isothiocyanate (9b)
1
(0.49 mmol): Yield: 68 mg (48%); mp: 104.08C; H NMR (500 MHz,
CDCl3): d=1.45 (t, J=7.3 Hz, 3H, CH3), 3.89 (dq, J=5.7 Hz, J=
7.3 Hz, 2H, CH2), 7.67–7.71 (m, 1H, 6-H), 8.02–8.06 (m, 1H, 7-H),
8.29–8.31 (m, 1H, 5-H), 9.21 (brs, 1H, NH); 13C NMR (125 MHz,
CDCl3): d=13.2 (CH3), 40.9 (CH2), 121.8 (CF3), 121.8 (C6), 121.9 (C4),
122.7 (C7), 129.7 (C5), 142.1 (C7’), 152.7 (C3’), 170.8 (C2), 183.7 (C=
S); MS (EI, 70 eV): m/z (%)=290 ([M]+, 64), 271 (3), 247 (100), 229
General procedure for the preparation of benzothiazole-2-carbothio-
amides 17a–e: A 100 mL two-necked flask was equipped with a
1756
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1749 – 1759

KATP Channel Openers
(16), 203 (14), 170 (5), 157 (5), 44 (17); Anal. calcd for C11H9F3N2S2
(290.33): C 45.51, H 3.12, N 9.65, found: C 45.97, H 3.08, N 9.26.
1.27 (t, J=7.3 Hz, 3H, NCH2-CH3), 2.20–2.39 (m, 2H, CH-CH2-CH3),
3.61–3.76 (m, 2H, NCH2-CH3), 4.46 (dd, J=6.0 Hz, J=8.3 Hz, 1H,
CH), 7.41 (t, J=7.6 Hz, 1H, 6-H), 7.50 (t, J=7.7 Hz, 1H, 5-H), 7.88 (d,
J=8.0 Hz, 1H, 7-H), 8.00 (d, J=8.2 Hz, 1H, 4-H), 9.32 (brs, 1H, NH);
13C NMR (125 MHz, CDCl3): d=11.7 (CH-CH2-CH3) 13.3 (NCH2-CH3),
31.7 (CH-CH2-CH3), 41.2 (NCH2-CH3), 61.0 (CH) 121.7 (C4), 122.6 (C7),
125.3 (C6), 126.3 (C5), 134.5 (C7’), 152.7 (C3’), 170.6 (C2), 200.1 (C=
S); MS (EI, 70 eV): m/z (%)=264 ([M]+, 45), 231 (19), 203 (14), 192
(18), 176 (100), 162 (91), 148 (12), 109 (17); Anal. calcd for
C13H16N2S2 (264.41): C 59.05, H 6.10, N 10.59, found: C 59.09, H
6.12, N 10.58.
N-Phenyl-5-(trifluoromethyl)benzothiazole-2-carbothioamide
(17e) from 15b (0.49 mmol) and phenyl isothiocyanate (9e)
(0.49 mmol): Yield: 130 mg (78%); mp: 160.08C; Rf =0.74 (CHCl3);
1H NMR (500 MHz, CDCl3): d=7.32–7.36 (m, 1H, 7.67–7.71 (m, 1H,
phenyl p-H), 7.47–7.52 (m, 2H, phenyl o-H), 7.69–7.73 (m, 1H, 6-H),
8.04–8.08 (m, 1H, 7-H), 8.04–8.08 (m, 2H, phenyl m-H), 8.36–8.38
(m, 1H, 4-H), 10.97 (brs, 1H, NH); 13C NMR (125 MHz, CDCl3): d=
122.0 (-CF3), 122.0 (C6), 122.2 (C4), 122.7 (C7), 127.2 (phenyl o-C),
129.2 (phenyl m-C), 129.6 (C5), 129.8 (phenyl p-C), 137.9 (NH-C),
142.7 (C7’), 152.5 (C3’), 172.2 (C2), 180.5 (C=S); MS (EI, 70 eV): m/z
(%)=338 ([M]+, 66), 337 (100), 305 (38), 235 (15), 77 (25).
2-(Benzothiazol-2-yl)-N-tert-butylpropanethioamide (19d) from
18d (0.67 g, 4.1 mmol) and tert-butylisothiocyanate (9d) (0.52 g,
4.1 mmol): Yield: 0.66 g (58%); mp: 124.78C; Rf =0.64 (EtOAc);
1H NMR (500 MHz, CDCl3): d=1.54 (s, 9H, C(CH3)3), 1.82 (d, J=
7.2 Hz, 3H, CH-CH3), 4.49 (q, J=7.2 Hz, 1H, CH), 7.39–7.43 (m, 1H,
6-H), 7.48–7.52 (m, 1H, 5-H), 7.87–7.90 (m, 1H, 7-H), 7.97–8.00 (m,
1H, 4-H), 9.21 (brs, 1H, NH); 13C NMR (125 MHz, CDCl3): d=23.9
(CH-CH3) 27.3 (C(CH3)3), 55.7 (C(CH3)3), 56.0 (CH), 121.8 (C4) 122.6
(C7), 125.3 (C6), 126.3 (C5), 134.4 (C7’), 152.7 (C3’), 172.1 (C2), 199.8
(C=S); MS (EI, 70 eV): m/z (%)=278 ([M]+, 22), 189 (13), 163 (100),
109 (15), 57 (20); Anal. calcd for C14H18N2S2 (278.44): C 60.39, H
6.52, N 10.06, found: C 60.90, H 6.36, N 9.58.
General procedure for the preparation of benzothiazole-2-carbothio-
amides 19a–d and carboxylic amides 20a,b: A 100 mL two-necked
flask was equipped with a magnetic stirrer, a connection to a com-
bined N2/vacuum line and closed with a septum. Benzothiazole
18a, 18b or 18c (10.0 mmol) was inserted and the air in the flask
was replaced by N2. By means of a dry-ice/acetone bath, the solu-
tion was cooled to ꢀ788C, and a 1.6m solution of n-butyllithium in
hexane (6.6 mL, 10.5 mmol) was added at such a rate that the tem-
perature, monitored by an electronic inlet thermometer, did not
exceed ꢀ758C. After stirring for one hour at ꢀ788C, isothiocyanate
9b or 9d (10.0 mmol) or tert-butylisocyanate was injected. The re-
action mixture was stirred overnight and allowed to warm up to
room temperature. Then, H2O (20 mL) and saturated ammonium
chloride solution (20 mL) was added and the mixture was extracted
with CHCl3 (4ꢃ75 mL). The combined organic layers were dried
with Na2SO4 and the solvent was removed in a rotary evaporator.
The residue was submitted to column chromatography or recrys-
tallization.
2-(Benzothiazol-2-yl)-N-tert-butylacetamide (20a) from 18a
(1.49 g, 10.0 mmol) and tert-butylisocyanate (0.99 g, 10.0 mmol):
Yield: 0.89 g (36%); mp: 154.68C; Rf =0.04 (CHCl3), 1H NMR
(500 MHz, CDCl3): d=1.36 (s, 9H, C(CH3)3), 3.99 (s, 2H, CH2), 6.90
(brs, 1H, NH), 7.41 (t, J=7.6 Hz, 1H, 6-H), 7.50 (t, J=7.6 Hz, 1H, 5-
H), 7.90 (d, J=7.9 Hz, 1H, 7-H), 7.99 (d, J=8.2 Hz, 1H, 4-H);
13C NMR (125 MHz, CDCl3): d=28.6 (C(CH3)), 42.1 (CH2), 51.6
(C(CH3)) 121.6 (C4), 122.5 (C7), 125.3 (C6), 126.3 (C5), 135.1 (C7’),
152.5 (C3’), 165.5 (C2), 165.6 (C=O); MS (EI, 70 eV): m/z (%)=248
([M]+, 5), 233 (2), 176 (3), 149 (100), 108 (5), 57 (15); Anal. calcd for
C13H16N2OS (248.34): C 62.87, H 6.49, N 11.28, found: C 62.72, H
6.74, N 11.08.
2-(Benzothiazol-2-yl)-N-ethylethanethioamide (19a) from 18a
(1.49 g, 10.0 mmol) and ethyl isothiocyanate (9b) (0.88 g,
10.0 mmol): Yield: 0.39 g (16%); mp: 109.88C; Rf =0.22 (n-hexane/
EtOAc, 2:1), 1H NMR (500 MHz, CDCl3): d=1.30 (t, J=7.3 Hz, 3H,
CH3), 3.72 (dq, J=5.4 Hz, J=7.3 Hz, 2H, CH2-CH3), 4.57 (s, 2H,
CH2CS), 7.40–7.45 (m, 1H, 6-H), 7.49–7.54 (m, 1H, 5-H), 7.87–7.91
(m, 1H, 7-H), 7.99–8.03 (m, 1H, 4-H), 9.53 (brs, 1H, NH); 13C NMR
(125 MHz, CDCl3): d=12.9 (CH3), 41.3 (CH2-CH3), 48.8 (CH2-CS) 121.8
(C4), 122.6 (C7), 125.6 (C6), 126.5 (C5), 134.5 (C7’), 152.4 (C3’), 165.6
(C2), 194.7 (C=S); MS (EI, 70 eV): m/z (%)=236 ([M]+, 33), 203 (21),
175 (7), 149 (100), 101 (9); Anal. calcd for C11H12N2S2 (236.36): C
55.90, H 5.12, N 11.85, found: C 55.91, H 5.28, N 11.64.
2-(Benzothiazol-2-yl)-N-tert-butylpropanamide (20b) from 18b
(1.63 g, 10.0 mmol) and tert-butylisocyanate (0.99 g, 10.0 mmol):
Yield: 0.27 g (10%); mp: 135.28C; 1H NMR (500 MHz, CDCl3): d=
1.32 (s, 9H, C(CH3)3), 1.70 (d, J=7.2 Hz, 3H, CH3), 4.00 (q, J=7.2 Hz,
1H, CH), 6.57 (brs, 1H, NH), 7.39 (t, J=7.2 Hz, 1H, 6-H), 7.48 (t, J=
7.2 Hz, 1H, 5-H), 7.88 (d, J=8.0 Hz, 1H, 7-H), 7.98 (d, J=8.1 Hz, 1H,
4-H); 13C NMR (125 MHz, CDCl3): d=18.7 (CH3), 28.5 (C(CH3)), 46.9
(CH), 51.4 (C(CH3)) 121.7 (C4), 122.6 (C7), 125.1 (C6), 126.1 (C5),
134.9 (C7’), 152.7 (C3’), 169.4 (C2), 172.1 (C=O); MS (EI, 70 eV): m/z
(%)=190 ([M]+, 2), 163 (100), 135 (2), 109 (12), 57 (16); Anal. calcd
for C14H18N2OS (262.37): C 64.09, H 6.91, N 10.68, found: C 63.91, H
6.87, N 10.73.
2-(Benzothiazol-2-yl)-N-ethylpropanethioamide (19b) from 18b
(1.63 g, 10.0 mmol) and ethyl isothiocyanate (9b) (0.88 g,
10.0 mmol): Yield: 0.34 g (14%); mp: 96.98C; Rf =0.23 (n-hexane/
EtOAc 1:1); 1H NMR (500 MHz, CDCl3): d=1.27 (t, J=8.1 Hz, 1H,
CH2-CH3), 1.86 (d, J=7.2 Hz, 3H, CH-CH3), 3.61–3.76 (m, 2H, CH2-
CH3), 4.62 (q, J=7.2 Hz, 1H, CH), 7.41 (ddd, J=1.1 Hz, J=7.5 Hz,
J=7.8 Hz, 1H, 6-H), 7.50 (ddd, J=1.2 Hz, J=7.3 Hz, J=8.3 Hz, 1H,
5-H), 7.88 (d, J=7.8 Hz, 1H, 7-H), 8.00 (d, J=8.1 Hz, 1H, 4-H), 9.26
(s, 1H, NH); 13C NMR (125 MHz, CDCl3): d=12.9 (CH2-CH3), 23.6 (CH-
CH3), 41.3 (CH2-CH3), 53.5 (CH) 121.8 (C4), 122.7 (C7), 125.4 (C6),
126.3 (C5), 134.4 (C7’), 152.7 (C3’), 171.7 (C2), 201.1 (C=S); MS (EI,
70 eV): m/z (%)=250 ([M]+, 50), 217 (78), 205 (7), 162 (100), 136
(15), 109 (32); Anal. calcd for C12H14N2S2 (250.38): C 57.56, H 5.64, N
11.19, found: C 57.76, H 5.58, N 11.14.
5-(Trifluoromethyl)benzothiazole (15b): A 100 mL two-necked
flask was equipped with a magnetic stirrer, a connection to a com-
bined N2/vacuum line and closed with a septum. The air in the
flask was replaced by N2, and 2-amino-4-(trifluoromethyl)thiophe-
nol (1.9 g, 2.5 mmol), aqueous formaldehyde (37% solution, 0.42 g,
14 mmol), scandium(III) triflate (0.35 g, 0.70 mmol) and dry THF
(40 mL) were injected. The mixture was stirred for three days at
room temperature under through-flow of compressed air. After
this period EtOAc (300 mL) was added and the organic layer was
washed with an aqueous solution of K2CO3 (5%, 40 mL) and brine
(40 mL). The combined organic extracts were dried with Na2SO4
and the solvent was removed in a rotary evaporator. The residue
was submitted to column chromatography. Yield: 0.36 g (72%);
2-(Benzothiazol-2-yl)-N-ethylbutanethioamide (19c) from 18c
(1.77 g, 10.0 mmol) and ethyl isothiocyanate (9b) (0.88 g,
10.0 mmol): Yield: 0.24 g (9%); mp: 129.88C; Rf =0.43 (CHCl3);
1H NMR (500 MHz, CDCl3): d=0.97 (t, J=7.4 Hz, 3H, CH-CH2-CH3),
ChemMedChem 2010, 5, 1749 – 1759
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1757

M. Braun et al.
MED
Rf =0.47 (CHCl3); MS (EI, 70 eV): m/z (%)=203 ([M]+, 100), 184 (18),
176 (12), 157 (21), 132 (13).
and cultured for 3 days in volumes of 200 mL per well before test-
ing. Chinese hamster ovary (CHO) cells stably expressing human
SUR1 and Kir6.2 were cultured under identical conditions as HEK
293 cells using G418 as selection marker and culturing cells only
for 1.5 days in 12-well strips before testing.
Crystal structure determinations
Membrane potential assays with DiBAC4(3): Confluent monolay-
ers cultivated in 12-well strips (96-well format) were rinsed twice
with 200 mL assay buffer (120 mm NaCl, 2.0 mm KCl, 1.0 mm MgCl2,
2.0 mm CaCl2, 5 mm glucose) supplemented with 20 mm HEPES [N-
(2-hydroxyethyl)piperazine-N’-(2-ethanesulfonic acid)], pH 7.4) at
308C before 200 mL assay buffer supplemented with 5 mm
DiBAC4(3) was added.[19] Labeling of cells cultured in 12-well strips
was continuously observed in 12-channel fluorescence detectors at
308C until the distribution of DiBAC4(3) across the plasma mem-
brane reached equilibrium (60–90 min). Test compounds were
added in a volume comprising 1/10 total buffer volume and gently
mixed resulting in a time-dependent decrease in fluorescence indi-
cating hyperpolarization of the membrane potential. Maximum hy-
perpolarization was induced with 30 mm NNC414[20] in CHO (SUR1/
Kir6.2) cells and with 3 mm KC399[2h,i] or with 10 mm P1075 in HEK
293 (SUR2B/Kir6.1) cells.
Crystals of compounds 19b and 19d suitable for X-ray study were
investigated with a Stoe Imaging Plate Diffraction System using
graphite-monochromatized MoKa radiation at T=291 K. Unit cell
parameters were determined by least-squares refinements on the
positions of 8000 and 2156 reflections in the range 2.88<q<25.88
and 2.08<q<25.28, respectively. In the case of 19b the monoclinic
lattice and systematic extinctions are consistent with space group
types Cc and C2/c, in the case of 19d the anorthic lattice is consis-
¯
¯
tent with space group types P1 and P1. C2/c and P1 proved to be
the correct space groups in the course of structure refinements.
The structures were solved by direct methods,[16] and the positions
of all the H atoms were found via DF syntheses. Refinements[17] by
full-matrix least-squares calculations on F2 converged to the indica-
tors above. Anisotropic displacement parameters were refined for
all atoms heavier than hydrogen. Positional parameters and iso-
tropic displacement parameters were refined for the H atoms in-
volved in hydrogen bonding. Idealized bond lengths and angles
were used for the CH3, CH2, and CH groups; the riding model was
applied for their H atoms. In addition, the H atoms of the CH3
groups were allowed to rotate around the neighboring CꢀC
bonds. Isotropic displacement parameters of the H atoms were
kept equal to 150, 120, 120, and 120% of the equivalent isotropic
displacement parameters of the parent primary, secondary, tertiary,
and “aromatic” carbon atoms, respectively. Crystal data and further
details: (19b) C12H14N2S2, Mr =250.39 gmolꢀ1, monoclinic, C2/c, a=
22.8724(14) ꢂ, b=7.2817(4) ꢂ, c=16.4463(9) ꢂ, b=102.283(7)8, V=
2676.4(3) ꢂ3, Z=8, Dcalcd =1.243 Mgmꢀ3, m=0.374 mmꢀ1, F(000)=
1056, crystal size 0.38ꢃ0.30ꢃ0.15 mm3, MoKa (l=0.71073 ꢂ),
2.79<q<25.008, total of 16958 reflections, 2237 independent re-
flections [Rint =0.034], completeness to q=25.008: 94.7%, Full-
matrix least-squares refinement on F2, 2237 data, 151 parameters,
GOF on F2 =1.106, R1 =0.036, wR2 =0.075 [I>2s(I)], R1 =0.051,
Fluorescence was effected with blue light (l=488 nm) excitation,
which was reflected by a dichroidic beam splitter (505 nm) mount-
ed at 458 to illuminate adherent cells from the bottom of the 12-
well strips. Excitation was performed with short light pulses of
30 ms to minimize photobleaching of the dye. Emitted light again
passed the dichroidic beam splitter and a long-pass emission filter
(OG515, Schott, Mainz, Germany) before it was detected by low
noise silicone photodiodes (OSD15-5T, Centronic, Croydon, UK).
After a >100000-fold amplification of the photocurrent with a
two-stage amplifier (operational amplifiers OPA111 and OPA121,
Texas Instruments, USA) fluorescent signals were digitized and sent
(RS232, 38400 baud) to a personal computer (Pentium, 1 GHz).
Computer software was programmed to allow continuous registry
of fluorescence data, their display on a conventional screen, and to
file data for subsequent analysis. As all electronic components
were chosen to be smaller than 9 mm (i.e. 96-well format); each of
the 12 channels could be equipped with independent excitation
and detection units, so that fluorescence data could be continu-
ously monitored at a high data sampling rate (1 Hz).[21]
wR2 =0.077 (all data), D1max/D1min
:
0.20/ꢀ0.16 eꢂꢀ3
;
(19d)
C14H18N2S2, Mr =278.44 gmolꢀ1, triclinic, P1, a=7.9491(15) ꢂ, b=
9.6764(17) ꢂ, c=9.9800(17) ꢂ, a=95.40(2)8, b=97.55(2)8, g=
¯
92.96(2)8, V=756.0(2) ꢂ3, Z=2,
D
calcd =1.223 Mgmꢀ3
,
m=
0.337 mmꢀ1, F(000)=296, crystal size 0.23ꢃ0.17ꢃ0.12 mm3, MoKa
(l=0.71073 ꢂ), 2.07<q<25.008, total of 10018 reflections, 2237
independent reflections [Rint =0.073], completeness to q=25.008:
97.0%, Full-matrix least-squares refinement on F2, 2237 data, 171
parameters, GOF on F2 =0.996, R1 =0.039, wR2 =0.068 [I>2s(I)],
Membrane potential assays with DyeB: DyeB is the membrane
potential dye R7260 from Molecular Probes manufactured for use
in high-throughput screening devices.[13] Confluent monolayers cul-
tivated in 12-well strips (96-well format) were incubated with
0.125 mgmLꢀ1 DyeB. Fluorescence was effected with green light
(l=505 nm) excitation. Emitted fluorescence light was filtered with
a long-pass filter of 530 nm (OG530, Schott, Mainz, Germany). All
other procedures were as described above for the use of
DiBAC4(3).
R1 =0.080, wR2 =0.073 (all data), D1max/D1min: 0.19/ꢀ0.19 eꢂꢀ3
.
CCDC 768915 (19b) and CCDC 768916 (19d) contain the supple-
mentary crystallographic data (excluding structure factors) for this
paper. These data can be obtained free of charge from The Cam-
bridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_
request/cif.
Data analysis
Competition binding: Experiments were performed in the pres-
ence of the radioligand ([3H]P1075, [L*]) and increasing concentra-
tions of competing ligands [L]. Displacement curves were analyzed
by nonlinear regression as reported[22] according to Equation (1):
Pharmacological methods
Cell culture: HEK 293 (human embryonic kidney) cells stably ex-
pressing murine SUR2B and Kir6.1[18] were cultured at 378C in a hu-
midified atmosphere of 93% air and 7% CO2 in a-MEM (modified
Eagle’s medium, Invitrogen) with glucose supplemented with 10%
fetal calf serum and 0.15 mgmLꢀ1 zeocin and 0.3 mgmLꢀ1 G418 as
selection medium. Cells were seeded out in poly(l-lysine)-coated
12-well strips (96-well format) at a density of 60000 cells per well
*
*
Bs ð½LꢂÞ ¼ B0 ½Lꢂ=f½Lꢂ þ KD ð1 þ ½L ꢂ=KL Þg
ð1Þ
in which B0 and Bs ([L]) represent the specific binding of L* to KATP
channels in the absence and presence of L, KL* denotes the dissoci-
ation constant of the radioligand L*, and KD is the equilibrium dis-
1758
www.chemmedchem.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1749 – 1759

KATP Channel Openers
kahashi, H. Nabata, Bioorg. Med. Chem. Lett. 1993, 3, 2005–2010; i) U.
Uhrig, H.-D. Hçltje, R. Mannhold, H. Weber, H. Lemoine, J. Mol. Graph.
Model. 2002, 21, 37–45; j) V. Cecchetti, V. Calderone, O. Tabarrini, S. Sa-
batini, E. Filipponi, L. Testai, R. Spogli, E. Martinotti, A. Fravolini, J. Med.
Chem. 2003, 46, 3670–3679; k) E. Carosati, H. Lemoine, R. Spogli, D.
Grittner, R. Mannhold, O. Tabarrini, S. Sabatini, V. Cecchetti, Bioorg. Med.
Chem. 2005, 13, 5581–5591.
sociation constant of the test compound L. Experimental data
were analyzed after transformation of data to obtain homoscedas-
ticity[21] which resulted in reliable estimates of parameters (B0, Bns,
and pKD) and asymptotic standard deviations (ASD). Data were
fitted to the hyperbola defined by nonlinear regression using the
SAS software package STAT. Data points in the figures are means
ꢁSEM.
[3] T. J. Brown, R. F. Chapman, D. C. Cook, T. W. Hart, I. M. McLay, R. Jordan,
J. S. Mason, M. N. Palfreyman, R. J. A. Walsh, M. T. Withnall, J.-C. Aloup, I.
Cavero, D. Farge, C. James, S. Mondot, J. Med. Chem. 1992, 35, 3613–
3624. For further potassium channel openers containing a thioamide
moiety, see references [2h,i].
Membrane potential assays: Equilibrium effects of ligand (L)-in-
duced changes in fluorescence induced with KATP channel openers
were analyzed according to Equation (2):
[4] H. Gilman, J. W. Morton, Jr., Org. React. 1954, 8, 258–304.
[5] H. W. Gschwend, H. R. Rodriguez, Org. React. 1979, 26, 1–360.
p
p
p
I ð%Þ ¼ 100 ꢃ ½Lꢂ =f½Lꢂ þ EC50
g
ð2Þ
´
[6] Cf. T. Jagodzinski, Synthesis 1988, 717–720.
Concentration–effect curves were fitted by nonlinear regression
analysis according to Lemoine.[23] Exponents p indicate the Hill co-
efficient correcting for the steepness of curves.
[7] M. Braun, A. Hessamian-Alinejad, B. Fꢅaux de Lacroix, B. Hernandez Al-
varez, G. Fischer, Molecules 2008, 13, 995–1003.
´
[8] T. S. Jagodzinski, Pol. J. Chem. 1992, 66, 653–659.
[9] H. Gilman, J. A. Beel, J. Am. Chem. Soc. 1949, 71, 2328–2331.
[10] Cf. T. Itoh, K. Nagata, M. Miyazaki, A. Ohsawa, Heterocycles 2000, 52,
1037–1040.
[11] a) H. Chikashita, S. Ikegami, T. Okumura, K. Itoh, Synthesis 1986, 375–
377; b) M. V. Costa, A. Brembilla, D. Roizard, P. Lochon, J. Heterocycl.
Chem. 1991, 28, 1541–1544.
[12] S. Lachenicht, A. Fischer, C. Schmidt, M. Winkler, A. Rood, H. Lemoine,
M. Braun, ChemMedChem 2009, 4, 1850–1858.
[13] S. M. Gopalakrishnan, S. A. Buckner, I. Milicic, D. R. Groebe, K. L. Whiteak-
er, D. J. Burns, U. Warrior, M. Gopalakrishnan, J. Pharmacol. Exp. Ther.
2002, 300, 910–917.
Drugs and materials: DiBAC4(3) was from Molecular Probes (Invi-
trogen, Karlsruhe, Germany), DyeB (R7260) was from Molecular De-
vices (1311 Orleans Drive, Sunnyvale, CA, USA), glibenclamide was
from Sigma (Deisenhofen, Germany), and [3H]P1075 was from Phar-
macia Biotech. P1075 was a gift from Leo Pharma (Ballerup, Den-
mark), NNC414 was from Dr. J. B. Hansen (Novo Nordisk Research,
2760, Mꢄløv, Denmark). PCOs were dissolved in DMSO and further
diluted with incubation buffer (final concentration in the mem-
brane potential assays: <0.1%, in radioligand binding: <1%).
[14] H. Lemoine, C. Teschner, L. Kꢆstner, J. Kꢁper, D. Grittner, A. Rood,
Naunyn-Schmiedeberg’s Arch. Pharmacol. 2007, 375 (Suppl. 1), 45.
[15] When N-phenyl-substituted thioamides (such as 10e, 17c, and 17e)
were submitted to the membrane potential assay using DiBAC4(3), a
limitation of the method became relevant, inasmuch as the apparent
affinities in the membrane potential measurement were substantially
higher than those obtained from the radioligand experiments, so that
pEC50 values surpassed the pKD values by ~3 log units. It seems that
these compounds interfere directly with the fluorescence mechanism
of DiBAC4(3). Therefore, the experiments were repeated with wild-type
HEK 293 cells that are expected to be resistant to PCO action. As the
relevant compounds markedly inhibited DiBAC4(3) fluorescence in the
wild-type cells, the apparent high affinity must be interpreted as an ar-
tifact. The problem was overcome by using the fluorescent dye DyeB.
[16] G. M. Sheldrick, SHELXS86: Program for the Solution of Crystal Struc-
tures, University of Gçttingen, Germany, 1985.
Acknowledgements
This work was supported by the Deutsche Forschungsgemein-
schaft (Br 604/18-1). The technical assistance of Bernd Mꢀnster-
mann and his team (Institute of Laser Medicine) in the construc-
tion of the new fluorescence detectors is gratefully acknowledged.
We thank Ms. Eleonore Hammes for technical support during X-
ray data collection.
Keywords: crystal structures
probes · heterocycles · structure–activity relationships
· drug design · fluorescent
[17] G. M. Sheldrick, SHELXL97: Program for the Refinement of Crystal Struc-
tures, University of Gçttingen, Germany, 1997.
[18] D. Stephan, M. Winkler, P. Kꢁhner, U. Russ, U. Quast, Diabetologia 2006,
49, 2039–2048.
[19] T. R. Miller, R. D. Taber, E. J. Molinari, K. L. Whiteaker, L. M. Monteggia,
V. E. S. Scott, J. D. Brioni, J. P. Sullivan, M. Gopalakrishnan, Eur. J. Pharma-
col. 1999, 370, 179–185.
[20] F. E. Nielsen, S. Ebdrup, A. F. Jensen, L. Ynddal, T. B. Bodvarsdottir, C.
Stidsen, A. Worsaae, H. C. M. Boonen, P. O. G. Arkhammar, T. Fremming,
P. Wahl, H. T. Korno, J. B. Hansen, J. Med. Chem. 2006, 49, 4127–4139.
[21] H. Lemoine, H. A. Rood, International Patent Application WO 2006/
056168A1; [Chem. Abstr. 2006, 145, 3768].
[1] a) S. Seino, T. Miki, Prog. Biophys. Mol. Biol. 2003, 81, 133–176 and refer-
ences therein; b) J. Bryan, L. Aguilar-Bryan, Curr. Opin. Cell Biol. 1997, 9,
553–559; c) I. Uhde, A. Toman, I. Gross, C. Schwanstecher, M. J. Schwan-
stecher, J. Biol. Chem. 1999, 274, 28079–28082; d) C. Moreau, A.-L.
Prost, R. Dꢅrand, M. Vivaudou, J. Mol. Cell. Cardiol. 2005, 38, 951–963;
e) U. Quast, D. Stephan, S. Bieger, U. Russ, Diabetes 2004, 53 (Suppl. 3),
156–164.
[2] a) P. Lebrun, V. Devreux, M. Hermann, A. Herchuelz, Eur. J. Pharmacol.
1988, 156, 283–286; b) P. Lebrun, V. Devreux, M. Hermann, A. Her-
chuelz, J. Pharmacol. Exp. Ther. 1989, 250, 1011–1018; c) T. C. Hamilton,
A. H. Weston, Gen. Pharmacol. 1989, 20, 1–9; d) B. Pirotte, J. Fontaine,
P. Lebrun, Curr. Med. Chem. 1995, 2, 573–582; e) S. Sebille, P. de Tullio,
B. Becker, M.-H. Antoine, S. Boverie, B. Pirotte, P. Lebrun, J. Med. Chem.
2005, 48, 614–621; f) G. C. Rovnyak, S. Z. Ahmed, C. Z. Ding, S. Dzwonc-
zyk, F. N. Ferrara, W. G. Humphreys, G. J. Grover, D. Santafianos, K. S.
Atwal, A. J. Baird, L. G. McLaughlin, D. E. Normandin, P. G. Sleph, S. C.
Traeger, J. Med. Chem. 1997, 40, 24–34; g) E. Salamon, R. Mannhold, H.
Weber, H. Lemoine, W. Frank, J. Med. Chem. 2002, 45, 1086–1097; h) H.
Koga, H. Sato, J. Imagawa, T. Ishizawa, S. Yoshida, I. Sugo, N. Taka, T. Ta-
[22] B. Ehle, H. Lemoine, A. J. Kaumann, Naunyn-Schmiedeberg’s Arch. Phar-
macol. 1985, 331, 52–59.
[23] H. Lemoine, Quant. Structure–Activity Relat. 1992, 11, 211–218.
Received: July 20, 2010
Revised: August 11, 2010
Published online on September 15, 2010
ChemMedChem 2010, 5, 1749 – 1759
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1759