6-Substituted benzopyrans as potassium channel activators: Synthesis, vasodilator properties, and multivariate analysis

Article abstract of DOI:10.1021/jm981047m

During the last 10 years compounds have been discovered which can activate or block KATP channels. In particular, K channel activators (KCA) have been found to be smooth muscle relaxants with their main utility in hypertension and bronchodilation. In this paper we describe the synthesis of new KCA of the benzopyran type with a fixed 4-substituent and a systematic variation in the 6-position. The relaxant potency in rat aorta and trachea was used for biological characterization of the benzopyrans. In both biological test systems, they exhibit potency ranges of more than 3 log units. Structure-activity relationships are investigated by principal component analysis (PCA) and partial least-squares (PLS) analysis. Most striking outliers in an initial PLS analysis of the entire database were the unsubstituted 6-H compound 13 as well as 34 and 35. For the remaining set of 31 compounds, a 3-component PLS model explains the variance in biological activity to 81% in the aortic and to 82% in the tracheal test system. 6- Substituents influence affinity by a direct (presumably dipolar) interaction with the receptor site. According to the 2D-plot of the partial PLS weights, a strong electronegativity as well as high values for the integy moment and for the heat of formation in water dominate the first component; low values for substituent size (as defined by globularity or surface) are in addition favorable for high potency. High lipophilicity and low minimum energies of interaction dominate the second component. Chemical descriptors for the biological potency of the test set in rat aorta and rat trachea are very similar according to the almost identical projection of the Y-variables onto the X-component space.

Full text of DOI:10.1021/jm981047m

J . Med. Chem. 1999, 42, 981-991
981
6-Su bstitu ted Ben zop yr a n s a s P ota ssiu m Ch a n n el Activa tor s: Syn th esis,
Va sod ila tor P r op er ties, a n d Mu ltiva r ia te An a lysis
Raimund Mannhold,*,† Gabriele Cruciani,‡ Horst Weber,§ Horst Lemoine,† Andrea Derix,§ Claus Weichel,§ and
Monica Clementi‡
Institut fu¨r Lasermedizin, Arbeitsgruppe Molekulare Wirkstoff-Forschung, and Institut fu¨r Pharmazeutische Chemie,
Heinrich-Heine-Universita¨t, Universita¨tsstrasse 1, D-40225 Du¨sseldorf, Germany, and Dipartimento di Chimica,
Laboratorio di Chemiometria, Universita` di Perugia, Via Elce di Sotto, 8, I-06123 Perugia, Italy
Received October 5, 1998
During the last 10 years compounds have been discovered which can activate or block KATP
channels. In particular, K channel activators (KCA) have been found to be smooth muscle
relaxants with their main utility in hypertension and bronchodilation. In this paper we describe
the synthesis of new KCA of the benzopyran type with a fixed 4-substituent and a systematic
variation in the 6-position. The relaxant potency in rat aorta and trachea was used for biological
characterization of the benzopyrans. In both biological test systems, they exhibit potency ranges
of more than 3 log units. Structure-activity relationships are investigated by principal
component analysis (PCA) and partial least-squares (PLS) analysis. Most striking outliers in
an initial PLS analysis of the entire database were the unsubstituted 6-H compound 13 as
well as 34 and 35. For the remaining set of 31 compounds, a 3-component PLS model explains
the variance in biological activity to 81% in the aortic and to 82% in the tracheal test system.
6-Substituents influence affinity by a direct (presumably dipolar) interaction with the receptor
site. According to the 2D-plot of the partial PLS weights, a strong electronegativity as well as
high values for the integy moment and for the heat of formation in water dominate the first
component; low values for substituent size (as defined by globularity or surface) are in addition
favorable for high potency. High lipophilicity and low minimum energies of interaction dominate
the second component. Chemical descriptors for the biological potency of the test set in rat
aorta and rat trachea are very similar according to the almost identical projection of the
Y-variables onto the X-component space.
In tr od u ction
cyanoguanidines, and the organic nitrates. Best inves-
tigated subgroup is the benzopyrans. Structure-activity
studies mainly focus on variations in the 4-position of
the benzopyran nucleus.6
Potassium channels comprise the most diverse group
of ion channels. Molecular biology has defined three
families of K channels.1 The first group belongs to the
S4 channel superfamily2 and mainly comprises the
voltage-gated K channels. Another group consists of
channels formed by monomers with a single membrane-
spanning region (minK channels); the third group
represents the inward rectifiers containing two mem-
brane-spanning regions. Members of this group are K
channels gating of which is sensitive to [ATP]. These
KATP channels, first detected by Noma3 in heart muscle,
are dimers with a pore-forming R-subunit and a â-sub-
unit, labeled SUR, due to its binding region for sulfon-
ylureas.
From literature5,6 it is known that 6-substituents have
an outstanding impact on modulating the biological
activity of benzopyran KCA. Small and highly elec-
tronegative 6-substituents were viewed as optimal.
6-Substituents could contribute to receptor affinity
either by a direct interaction with the receptor site or
indirectly by withdrawing electrons from the phenyl
moiety of the benzopyran nucleus and thereby optimiz-
ing charge-transfer interactions of the aromate with the
receptor.
To better understand the molecular meaning of
6-substituents for the biological activity, we synthesized
new benzopyrans with extensive chemical variation
within the 6-position, measured their dilator properties
in rat aorta and trachea, defined structure-activity
relationships on the basis of principal component analy-
sis (PCA) and partial least-squares (PLS) analysis, and
looked for a putative tissue selectivity.
During the last 10 years compounds have been
discovered which can activate or block these KATP
channels. In particular, K channel activators have been
found to be smooth muscle relaxants with their main
utility in hypertension and bronchodilation. The puta-
tive binding region for the KCA is attributed to the
â-subunit.4
There are at least seven classes of KCA,5 of which the
main four are the benzopyrans, the thioformamides, the
Ch em istr y
Synthesis of benzopyran-type potassium channel open-
ers was accomplished according to following proce-
dures: best results were obtained by the method of
Kawase7 (Scheme 1) with the corresponding salicyl
aldehydes 1 and 2 as starting material and cyclization
* Corresponding author: Raimund Mannhold. Phone: (0) 211 81
12759. Fax: (0) 211 81 11374. E-mail: Raimund.Mannhold@uni-
duesseldorf.de.
† Institut fu¨r Lasermedizin, Heinrich-Heine-Universita¨t.
‡ Universita` di Perugia.
§ Institut fu¨r Pharmazeutische Chemie, Heinrich-Heine-Universita¨t.
10.1021/jm981047m CCC: $18.00 © 1999 American Chemical Society
Published on Web 03/02/1999

982 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Mannhold et al.
Sch em e 1a
a
Reagents and conditions: (a) C(CH3)2dCH-COOEt, K2CO3, DMF, 150 °C, 15 h; (b) NBS; (c) NaOH (pellets), Et2O; (d) 2-pyridinol,
EtOH, pyridine, reflux; (e) 2-pyridinol, K2CO3, toluene, 2 h, reflux; (f) NaOH (microlayer; Merck cat. no. 1.01564), dioxane, reflux.
Sch em e 2a
avoid this procedure we boiled the components with
potassium carbonate in toluene, and the trans-configu-
rated pyridones were the only reaction products in al-
most quantitative yield. Dehydration of 9 and 10 leads
to chromenes 13 and 14, which were used as starting
material for the desired compounds.
Acylation of 13 (Scheme 2) under normal Friedel-
Crafts conditions was only successful for the preparation
of 15-20, whereas the substituted aryl ketones 22-27
were not accessible in this way due to a simultaneous
transformation of the benzopyran to an indene ring
system.9 Therefore we used the method of Effenberger
et al.10 with silver triflate and the corresponding acyl
chlorides to produce the mixed carboxylic-sulfuric-
anhydrides in situ. Acceptable yields were obtained for
compounds 22-26 in refluxing 1,2-dichloroethane, and
Sch em e 3a
a
Reagents and conditions: (a) for compounds 15-20, RCOCl,
AlCl3, 1,2-dichloroethane; (b) for compound 21, pyridine-HCl, 170
°C; (c) for compounds 22-27; RCOCl, Ag-triflate, 1,2-dichloro-
ethane, reflux.
with dimethylacrylate in DMF/potassium carbonate at
about 150 °C. The chromenes 3 and 4 were separated
from the reaction mixture on distillation in vacuo. Direct
epoxidation of these chromenes with 3-chloroperbenzoic
acid failed to give the desired oxirans 7 and 8, which
were prepared by alkaline dehydrohalogenation of trans-
bromohydrins 5 and 6; they were in turn obtained from
chromenes 3 and 4 with N-bromosuccinimide (NBS).
Further reactions were similar to the route of Bergmann
and Gericke8 (addition of 2-pyridinol to the epoxides
with catalytic amounts of pyridine in refluxing ethanol).
Unfortunately it was necessary to separate the resulting
mixtures of the pyridones (9, 10) and the (pyridyloxy)-
chromanols (11, 12) by silica gel chromatography. To
a
Reagents and conditions: (a) NH2OH‚HCl, Na2CO3, H2O, 60
°C; (b) NaH2PO2, H2O, AcOH, pyridine, Raney nickel; (c) CH2(CN)2,
AcOH, piperidine, toluene, reflux.

6-Substituted Benzopyrans as K+ Channel Activators
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6 983
Sch em e 4a
methyl ether 14 with pyridine-HCl13 at 170 °C or with
boron tribromide at -10 °C afforded the phenol 34,
which was converted to the acetate 35, the phenacyl
ethers14 36 and 37, and with the corresponding isocy-
anates to a series of urethanes15 (38-42). The fluoro-
sulfonate 43 was prepared from phenol 34 with fluoro-
sulfonyl chloride in absolute pyridine16 at 0 °C. Bi-
makalim (44) and compounds 45-48 (Table 1) were
kindly provided by E. Merck, Darmstadt, Germany, and
are described in the literature;8 44-48 are included in
our test assays to allow a direct comparison of biological
activities of the new benzopyrans with reference com-
pounds.
Resu lts a n d Discu ssion
Structure-activity relationships (SAR) for KCA of the
benzopyran-type, so far available in the literature,5,6
stem from in vitro data (dilation of aortic rings or strips)
or from in vivo experiments in spontaneously hyper-
tensive rats (blood pressure-lowering effect). The great
majority of SAR analyses deal with variations in the
4-position; investigations on 6-substituted benzopyrans
are rather limited with the exception of data on a set of
13 4-carbothioamidobenzopyrans with 6-variation of
intermediate chemical diversity.17,18 According to these
QSAR studies, electronic properties, in particular hy-
drogen bond acceptor properties of the 6-substituent, are
assumed to determine the in vitro potency of the
benzopyrans.
a
Reagents and conditions: (a) NO2BF4, CH2Cl2; (b) iron/H2SO4;
(c) acetonylacetone, reflux.
only the difluoro ketone 27 required higher tempera-
tures (150 °C in nitrobenzene).
Further 6-substituted benzopyrans were prepared as
derivatives of bimakalim8 by common procedures
(Scheme 3). Excellent yields of amidoxime 28 and
aldehyde8 29 were obtained after treatment of the nitrile
with either hydroxylamine or Raney nickel and sodium
hypophosphite,11 respectively. 29 was condensed with
malononitrile to form 30 in high yield.
Preparation of the chromene 33 is shown in Scheme
4. Nitration of 13 with nitronium tetrafluoroborate in
dry dichloromethane afforded 31,8 which was reduced
to the amine 32. Cyclocondensation with acetonylac-
etone by the conventional Paal-Knorr method12 gave
the pyrrole 33.
Here we describe the synthesis (n ) 29) and derive
multivariate SAR analyses (n ) 34) for a database of
6-substituted benzopyrans with extensive chemical
variation in the 6-position. For biological characteriza-
tion of the 34 KCA, their relaxant potency in rat aorta
and trachea was measured according to a protocol,
described in the Experimental Section. Half-maximal
potency was derived from dose-response curves as
Methods for preparation of 6-O-substituted benzopy-
rans are summarized in Scheme 5. Cleavage of the
Sch em e 5a
a
Reagents and conditions: (a) pyridine-HCl/170 °C or BBr3/-10 °C; (b) acetyl chloride, triethylamine, toluene, reflux; (c) R6COCH2Br,
K2CO3, acetone; (d) R7NCO, triethylamine, toluene, reflux; (e) SO2ClF, pyridine, 0 °C.

984 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Mannhold et al.
Ta ble 1. Biological Activity in Rat Aorta and Trachea, Expressed as Reciprocal Half-Maximal Potency (log 1/C ( SD)
log 1/C
compd
R
aorta
trachea
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
H
5.43 ((0.06)
6.55 ((0.05)
7.37 ((0.05)
6.63 ((0.10)
6.61 ((0.07)
6.15 ((0.13)
6.40 ((0.10)
6.49 ((0.05)
6.65 ((0.03)
6.20 ((0.13)
7.08 ((0.13)
6.17 ((0.05)
6.83 ((0.09)
6.76 ((0.14)
6.97 ((0.09)
5.99 ((0.06)
6.91 ((0.05)
6.95 ((0.08)
7.27 ((0.08)
5.44 ((0.05)
5.60 ((0.03)
5.05 ((0.03)
4.63 ((0.04)
5.30 ((0.03)
5.38 ((0.02)
5.30 ((0.02)
5.58 ((0.09)
5.10 ((0.04)
7.95 ((0.05)
7.67 ((0.10)
6.68 ((0.10)
6.21 ((0.12)
7.84 ((0.08)
7.61 ((0.12)
5.05 ((0.08)
5.99 ((0.11)
6.65 ((0.01)
5.85 ((0.07)
6.21 ((0.07)
5.72 ((0.03)
5.84 ((0.08)
5.92 ((0.04)
5.87 ((0.12)
6.09 ((0.11)
6.34 ((0.12)
5.64 ((0.04)
6.25 ((0.08)
5.78 ((0.16)
5.99 ((0.06)
5.51 ((0.06)
6.14 ((0.03)
6.30 ((0.07)
6.78 ((0.06)
5.31 ((0.05)
4.97 ((0.05)
4.69 ((0.07)
4.77 ((0.14)
4.89 ((0.08)
4.22 ((0.16)
4.37 ((0.33)
5.19 ((0.04)
4.54 ((0.04)
7.22 ((0.04)
7.10 ((0.08)
6.32 ((0.12)
5.96 ((0.06)
7.53 ((0.23)
7.37 ((0.05)
methoxy
acetyl
propionyl
benzoyl
4-methoxybenzoyl
2-thienoyl
2-furoyl
4-hydroxybenzoyl
4-nitrobenzoyl
2-fluorobenzoyl
2-nitrobenzoyl
2-methylbenzoyl
2-(trifluoromethyl)benzoyl
2,6-difluorobenzoyl
aminohydroxyiminomethyl
formyl
2,2-dicyanoethenyl
2,5-dimethyl-1-pyrrolyl
hydroxy
acetoxy
phenacyloxy
4-fluorophenacyloxy
phenylcarbamoyloxy
2-fluorophenylcarbamoyloxy
4-fluorophenylcarbamoyloxy
2-(trifluoromethyl)phenylcarbamoyloxy
4-(trifluoromethyl)phenylcarbamoyloxy
fluorosulfonyloxy
cyano
4-pyridyl
thiocarboxamido
bromo
trifluoromethyl
pEC50 value (Table 1). In both test systems the KCA
exhibit potency ranges of more than 3 log units. In rat
aorta the spectrum ranges from a pEC50 value of 7.95
for 43 to 4.63 for 37; in rat trachea the highest potency
(pEC50 ) 7.53) is observed for the 6-bromo derivative
47, and the compound with the weakest potency is 39
with a pEC50 of 4.22.
outlier behavior, we excluded 13, 34, and 35 from the
further PLS analysis.
The variance in biological activity of the remaining
31 benzopyrans is explained by a 3-component model
to 81% in the aortic and to 82% in the tracheal test
systems. Corresponding PLS plots for the 3 components
are comparatively shown in Figure 2. The model tries
to explain the remaining molecular pattern after the
first by the second component and so forth. Character-
istic descriptors dominating the respective components
are shown in a color-coded manner with red coding for
high values and green coding for low values of the
corresponding descriptor. Thus, low molecular volume
is dominating the first, high lipophilicity the second, and
high LUMO values the third component.
The overall quality of the obtained PLS model is
summarized in Figure 3, plotting the experimental
versus the recalculated biological activity data for the
aortic (a) and the tracheal (b) test systems. This figure
elucidates the strong similarity in the SAR for both test
systems and the slight superiority of the model for
tracheal data as compared to the aortic test system.
The 2D-plot of the partial PLS weights in Figure 4
for the second versus the first component displays the
correlation between X-variables and u (Y). The w’s are
the weights that combine the X-variables to the scores
Results of a PLS analysis (t-u score plot) of the 34
KCA are shown in Figure 1. A t-u score plot displays
the observations in the projected X (t) and Y (u) space
and shows how well the Y space (biological activity)
correlates to the X space (chemical descriptors). Most
striking outlier in this plot is the 6-unsubstituted
compound 13, encircled in the figure. For receptor
binding, this finding might indicate an essential role of
6-substitution, lack of which results in the significant
reduction of potency as observed for 13. A first inspec-
tion of the influence of 6-substituents on biological
activity indicates a strong potency increase by prefer-
entially small, electronegative, and rather rigid sub-
stituents (arrow 1), an intermediate increase by CdO-
linked moieties (arrow 2), and a sustained or even
reduced activity by oxygen-bridged groups (arrow 3)
with 13 as reference. In addition to 13, further strong
outliers are 34 and 35; they are primarily responsible
for unexplained variance. Due to their pronounced

6-Substituted Benzopyrans as K+ Channel Activators
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6 985
F igu r e 1. Initial PLS analysis of the entire database: most
striking outlier in this plot is the 6-unsubstituted 13, encircled
in the figure. A first inspection of the influence of 6-substit-
uents on biological activity indicates a strong potency increase
by small, electronegative, and rigid substituents (arrow 1), an
intermediate increase by CdO-linked moieties (arrow 2), and
a sustained or reduced activity by oxygen-bridged groups
(arrow 3) with 13 as the reference.
t, so as to maximize their correlation with Y. X-variables
with high w’s are highly correlated with u (Y). A strong
electronegativity (Ela) and high values for the integy
moment I2 (calculated at -0.5 kcal/mol) and for the heat
of formation in water (HFw) dominate the first compo-
nent; low values (projection onto the negative part of
the X-component) for substituent size (as defined by
globularity G or surface S+) are in addition favorable
for high potency. High lipophilicity (log P) and low
minimum energies of interaction (Emin) likewise domi-
nate the second component.
This pattern of dominating variables might serve as
an explanation for the outlier behavior of the phenol
34, which exhibits strong electron-donating properties
instead of the advantageous electron-withdrawing prop-
erties (see first component) and an extraordinarily low
lipophilicity (see second component). Corresponding
explanations for the other outlier 35 (chemically quite
different from 34) are lacking.
The almost coinciding positioning of the Y-variables
in the PLS weights plot indicates very similar chemical
descriptors for the biological potency of the entire set
of the 6-substituted benzopyrans in both rat aorta and
rat trachea.
Two striking features in our SAR results were treated
in some more detail: (i) the pronounced reduction in
potency when exchanging 6-acetyl in 15 by 6-propionyl
in 16 and (ii) the high activity of 43 exhibiting a
6-fluorosulfonyloxy substituent.
An attempt to explain the potency loss of 16 is given
in Figure 5. Molecules 15 and 16 are superimposed in
order to comparatively show their integy moment vec-
tors. Integy moments measure the imbalance between
the center of mass of a molecule and the position of the
hydrophilic regions around it. Integy moments are
vectors centered in the molecular center of mass. If the
integy moment is high, there is a clear concentration of
hydrated regions in only one part of the molecular
surface. The regions of interaction are calculated with
F igu r e 2. PLS plots for the 3 components of the second PLS
(n ) 31). The model tries to explain remaining “noise” of the
first by the aid of the second component and so forth.
Characteristic descriptors dominating the respective compo-
nents are shown by color coding (red ) high values and green
) low values of the corresponding descriptor). Accordingly, low
volume dominates the first, high log P the second, and high
LUMO values the third component.
GRID using the water probe. The contour level shown
is at -5 kcal/mol. The integy vectors start from the
positive baricenters of the molecules and point toward
the negative baricenters. The two integies for 15 and
16 differ regarding module and direction of the vector;
this comparison indicates a strong sensitivity of the
KCA activity to the directionality of charge distribution
within the 6-varied test compounds.
Comparison of the score and loading plots in Figure
6, obtained from PCA, hints at some chemical interpre-
tation of the high dilator activity of 43. The pattern of
chemical descriptors for 43 exhibits significant differ-
ences from the remaining compounds as far as the
second component is concerned (see score plot); from the
loading plot this difference can be mainly attributed to
a significantly increased electronegativity which in turn
seems to improve receptor affinity. Contribution of

986 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Mannhold et al.
a
F igu r e 4. 2D-plot of the partial weights for the second versus
the first component. A strong electronegativity as well as high
values for the integy moment I2 (calculated at -0.5 kcal/mol)
dominate the first component; low values (projection onto the
negative part of the X-component) for substituent size (as
defined by globularity G or positive surface S+) are in addition
favorable for high potency. High lipophilicity (log P) and low
minimum energies of interaction (Emin) likewise dominate the
second component.
b
F igu r e 3. Plot of experimental versus recalculated biological
activity data for the aortic (a) and tracheal (b) test systems.
The figure elucidates the strong similarity in the SAR for both
test systems and the slight superiority of the model for tracheal
data.
6-substituents to receptor affinity could be either by a
direct (presumably dipolar) interaction with the receptor
site or indirectly by withdrawing electrons from the
attached phenyl moiety and thereby optimizing charge-
transfer interactions of the aromate. The high potency
of 43 with its fluorosulfonyloxy substituent strongly
favors the first alternative.
In summary, our multivariate analysis provides the
following results: (1) 6-Substituents represent part of
the benzopyran pharmacophore; i.e., they directly bind
to the receptor site. KCA activity is strongly sensitive
to the directionality of charge distribution within the
6-substituent; it seems to be particularly favorable for
the 6-fluorosulfonyloxy derivative 43. 6-Substituents
containing a sulfonyl moiety could represent an inter-
esting lead for further optimizing new benzopyrans. (2)
6-Substituents should exhibit a strong electronegativity
and high values for the integy moment, low volume,
high lipophilicity, and low minimum energies of interac-
tion for optimized dilator potency. (3) Tissue selectivity
is not achievable with variation of substitution pattern
in the 6-position.
F igu r e 5. 15 and 16 are superimposed to compare their integy
moment vectors. The regions of interaction are calculated with
GRID using the water probe (contour level -5 kcal/mol). Integy
vectors start from the positive baricenters of the molecules and
point toward the negative baricenters. The two integies for
15 (light gray) and 16 (dark gray) differ regarding module and
direction of the vector.
The present model can be used as a test case to guide
further synthesis. For the simple exchange of F in 43
by Cl, a log 1/C of 7.3 for the aorta and 6.95 for trachea
are predicted. As a more complex alternative -O-CH2-
SO2-Ph was selected yielding predictions of 7.2 and 6.5,
respectively. This could be an interesting candidate
because a limited loss in activity is coupled with a
reduction of lipophilicity, in turn improving its profile
as a putative drug candidate.
Exp er im en ta l Section
A. Ch em istr y. Melting points were taken on a Lindstroem
apparatus and are uncorrected. 1H and H-decoupled 13C NMR

6-Substituted Benzopyrans as K+ Channel Activators
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6 987
F igu r e 6. Principal component analysis (n ) 31). Chemical descriptors for 43 differ from the other compounds in the second
component (see score plot); from the loading plot this difference can be mainly attributed to a significantly increased electronegativity
which in turn seems to improve receptor affinity.
spectra were recorded on a Bruker AC-200 spectrometer. IR
and MS spectra were measured on a Perkin-Elmer 1600 FT-
IR spectrometer and a Finnigan 4000 spectrometer. Elemental
analyses (C, H, N) were carried out in the microanalytical
center (Heinrich-Heine-Universita¨t, Du¨sseldorf) with a Perkin-
Elmer PE2400 CHN elemental analyzer and were within
(0.3% of the calculated values. Solvents were all of anhydrous
grade and were obtained from commercial suppliers; 1,2-
dichloroethane was distilled from P2O5 just prior to use.
Syn th esis of th e Ben zop yr a n s 3 a n d 4. A mixture of the
corresponding aldehyde (12.2 g, 0.1 mol salicyl aldehyde; 15.3
g, 0.1 mol 2-hydroxy-5-methoxybenzaldehyde; Aldrich Chemi-
cals), anhydrous potassium carbonate (40 g), and 80 mL DMF
was heated at 150 °C for 15 h. The resulting dark-brown
mixture was allowed to cool and evaporated to remove DMF.
The residue was partitioned between ether and water, and the
aqueous phase was discarded. The organic layer was succes-
sively washed with 1 N HCl, 1 N NaOH, and water, dried
(MgSO4), evaporated, and distilled in vacuo.
(2), 189 (7), 175 (4), 152 (100), 137 (43), 125 (29). Anal. (C12H15
Br O3) C, H.
-
Syn th esis of Ep oxid es 7 a n d 8. The appropriate bromo-
chromanol (0.02 mol, 5.12 g of 5 or 5.74 g of 6, respectively)
was stirred with a suspension of powdered KOH (5.7 g, 0.1
mol) in 100 mL dry ether for 2 days at room temperature.
Filtration and evaporation gave the crude epoxides, which
were used without further purification.
3,4-Ep oxy-3,4-d ih yd r o-2,2-d im eth yl-2H-1-ben zop yr a n
(7): colorless oil (3.28 g, 93%); 1H NMR (CDCl3) δ 1.25 (s,
3H), 1.58 (s, 3H), 3.47 (d, J ) 4.6 Hz, 1H), 3.90 (d, J ) 4.6 Hz,
1H), 6.81 (‘d’, J ) 8 Hz, 1H), 6.92 (‘t’, J ) 7.5 Hz, 1H), 7.24
(td, J ) 8.0 and 1.6 Hz, 1H), 7.34 (dd, J ) 7.5 and 1.6 Hz,
1H); IR (CHCl3) 2973, 1646, 1385, 1369, 1255 cm-1; MS 176
(5, M), 161 (20), 139 (15), 121 (25), 43 (100).
3,4-E p oxy-3,4-d ih yd r o-6-m et h oxy-2,2-d im et h yl-2H -1-
ben zop yr a n (8): colorless crystals (3.71 g, 90%); mp 82-83
1
°C; H NMR (CDCl3) δ 1.22 (s, 3H), 1.57 (s, 3H), 3.47 (d, J )
4.3 Hz, 1H), 3.78 (s, 3H), 3.86 (d, J ) 4.3 Hz, 1H), 6.74 (d, J
) 8.7 Hz, 1H), 6.80 (dd, J ) 8.7 and 2.7 Hz, 1H), 6.89 (d, J )
2.7 Hz, 1H); IR (KBr) 2977, 2833, 1616, 1499, 1368, 1260, 871
cm-1; MS 206 (61, M), 191 (5), 163 (19), 150 (100), 137 (32),
122 (12). Anal. (C12H14O3) C, H.
2,2-Dim eth yl-2H-1-ben zop yr a n (3): colorless oil7 (3.8 g,
24%); bp19 110 °C; MS 161 (7, M + 1), 160 (16, M), 145 (100).
6-Meth oxy-2,2-d im eth yl-2H-1-ben zop yr a n (4): yellowish
oil7 (2.63 g, 46%); bp0.05 58-60 °C; MS 190 (35, M), 175 (100),
160 (6), 135 (86).
tr a n s-3,4-Dih yd r o-4-(1,2-d ih yd r o-2-oxo-1-p yr id yl)-3-
h yd r oxy-2,2-d im eth yl-2H-1-ben zop yr a n (9). The epoxide
7 (3.52 g, 20 mmol) and 2-pyridinol (2.85 g, 30 mmol) were
taken in 150 mL toluene. Potassium carbonate (14 g, 0.1 mol)
was added, and the mixture was refluxed for 2 h. Inorganic
material was filtered off, and the toluene was washed with
water to remove unreacted 2-pyridinol, then dried (MgSO4),
and evaporated. The residue was recrystallized from ethyl
Syn th esis of th e Br om oh yd r in s 5 a n d 6. The appropriate
chromene (0.05 mol, 8.01 g of 3 or 9.51 g of 4, respectively)
was dissolved in a mixture of 40 mL DMSO and water (0.9 g,
0.05 mol). This solution was treated successively with NBS
(9.8 g, 0.055 mol). The mixture was stirred for 0.5 h and then
poured into 500 mL water and extracted with ethyl acetate.
The organic layer was washed with water to remove DMSO
and finally dried (MgSO4). Evaporation followed by trituration
with hexane gave the bromohydrins.
1
acetate to give 9 (5.15 g, 95%): mp 186 °C; H NMR (CDCl3)
δ 1.34 (s, 3H), 1.52 (s, 3H), 3.86 (dd, J ) 9.5 and 4.1 Hz, after
add. of D2O d, J ) 9.5 Hz, 1H), 4.29 (d, OH), 6.21 (‘t’, J ) 6.8
Hz, 1H), 6.31 (d, J ) 9.5 Hz, 1H), 6.68 (‘d’, J ) 9.0 Hz, 1H),
6.77 (‘d’, J ) 7.5 Hz, 1H), 6.90 (m, 2H), 6.98 (dd, J ) 6.9 and
0.9 Hz, 1H), 7.22 (t, J ) 7.5 Hz, 1H), 7.38 (m, 1H); IR (KBr)
3164, 2975, 1654, 1568, 1542, 1489, 1255, 767 cm-1; MS 253
(21, M-H2O), 238 (75), 159 (37), 43 (100). Anal. (C16H17NO3)
C, H, N.
3-Br om o-3,4-d ih yd r o-2,2-d im eth yl-2H-1-ben zop yr a n -4-
ol (5): colorless crystals (10.4 g, 81%); mp 82 °C; 1H NMR
(CDCl3) δ 1.41 (s, 3H), 1.61 (s, 3H), 2.49 (d, J ) 4.8 Hz, OH),
4.16 (d, J ) 9.4 Hz, 1H), 4.94 (dd, J ) 9.4 and 4.8 Hz, after
add. of D2O d, J ) 9.4 Hz, 1H), 6.8 (‘d’, J ) 7.8 Hz, 1H), 7.00
(‘t’, J ) 7.8 Hz, 1H), 7.20 (‘t’, J ) 7.8 Hz, 1H) 7.39 (‘d’, J ) 7.8
Hz, 1H); IR (KBr) 3407, 2985, 1608, 1586, 1483, 1454, 754
cm-1; MS 258 (11), 256 (11), 161 (10), 122 (100). Anal. (C11H13
BrO) C, H.
-
tr a n s-3,4-Dih yd r o-4-(1,2-d ih yd r o-2-oxo-1-p yr id yl)-3-
h yd r oxy-6-m eth oxy-2,2-d im eth yl-2H-1-ben zop yr a n (10).
According to the procedure described for 9 from the epoxide 8
(4.12 g, 20 mmol), 2-pyridinol (2.85 g, 30 mmol), and potassium
carbonate (14 g, 0.1 mol) in 150 mL toluene and crystallization
of the residue from chloroform/hexane gave 10 (5.54 g, 92%):
3-Br om o-3,4-dih ydr o-6-m eth oxy-2,2-dim eth yl-2H-1-ben -
zop yr a n -4-ol (6): colorless crystals (12.9 g, 90%); mp 123 °C;
1H NMR (CDCl3) δ 1.38 (s, 3H), 1.58 (s, 3H), 2.52 (d, J ) 4.8
Hz, OH), 3.78 (s, 3H), 4.41 (d, J ) 9.5 Hz, 1H), 4.89 (dd, J )
9.5 and 5.0 Hz, after add. of D2O d, J ) 9.5 Hz, 1H), 6.72 (d,
J ) 8.90 Hz, 1H), 6.80 (dd, J ) 8.9 and 2.9 Hz, 1H), 7.00 (d,
J ) 2.9 Hz, 1H); IR (KBr) 3181, 2978, 1496, 1368, 1263, 1222,
904 cm-1; MS 288 (20), 286 (21), 271 (2), 269 (2), 255 (2), 253
1
mp 209 °C; H NMR (CDCl3) δ 1.31 (s, 3H), 1.50 (s, 3H), 3.66
(s, 3H), 3.85 (dd, J ) 9.5 and 4.0 Hz, after add. of D2O d, J )
9.5 Hz, 1H), 4.27 (d, OH), 6.23 (m, 3H), 6.68 (‘d’, J ) 9.9 Hz,
1H), 6.85 (m, 2H), 7.01 (‘dd’, J ) 8.7 Hz, and 1.9 Hz, 1H), 7.38

988 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Ta ble 2. Chemical Data of Ketones 15-27
Mannhold et al.
further purification. The acid chloride was taken in 100 mL
absolute 1,2-dichloroethane, and this solution was dropped into
a stirred suspension of anhydrous AlCl3 (1.32 g, 0.01 mol) in
100 mL 1,2-dichloroethane at 0 °C. Finally, the solution of
benzopyran 13 (2.53 g, 0.01 mol) in 50 mL 1,2-dichloroethane
was added slowly at 0 °C. After stirring at room temperature
overnight the reaction mixture was poured on crashed ice. The
organic layer was washed with 2 N HCl, 0.1 N NaOH, and
water, dried (MgSO4), and evaporated. The residue was
recrystallized as described.
Meth od b for th e P r ep a r a tion of 21. Compound 18 (0.39
g, 1 mmol) was heated with pyridine HCl (1.1 g, 0.01 mol) at
170 °C for 6 h and allowed to cool. The mixture was diluted
with 2 N HCl and extracted with ether. The phenol was
separated from the organic phase with 0.5 N NaOH and
reextracted with ether after acidification of the aqueous layer.
After drying (MgSO4) and evaporation of the ether extract
remained a white solid of 21.
Meth od c for th e P r ep a r a tion of Keton es 22-26. The
acid chlorides (2 mmol) were prepared from the corresponding
acids and thionyl chloride as described in method a. Each was
taken in 50 mL 1,2-dichloroethane, and silver triflate (0.9 g,
3.5 mmol) was added in several portions. After addition of the
benzopyran 13 (0.51 g, 2 mmol) in 50 mL 1,2-dichloroethane,
the mixture was refluxed for 3 h. After cooling and dilution
with 150 mL ether, 10 mL 2 N KOH was added and the
mixture vigorously stirred for 2 h, at room temperature. The
organic phase was washed with water, dried (MgSO4), and
evaporated. The residue was purified by silica gel chromatog-
raphy (if necessary) and crystallized as described (Table 2).
Meth od d for th e P r ep a r a tion of 27. Nitrobenzene was
used as solvent instead of 1,2-dichloroethane, as described in
method c, and the reaction temperature was raised to 160 °C.
The filtrate was distilled in vacuo and the residue purified by
silica gel chromatography with toluene/CH2Cl2/2-propanol (35/
60/5).
yield, prep mp,
recryst
solvent
compd
%
method °C
formula (Mr)
15
16
17
18
19
20
21
22
13
14
15
16
27
93
91
84
92
66
33
25
55
88
10
59
25
19
a
a
a
a
a
a
b
c
c
c
c
c
d
141 ether
131 CH2Cl2/hexane C19H19NO3 (309.4)
C
18H17NO3 (295.3)
107 EtOAc/hexane
138 ether
168 EtOAc/hexane
145 EtOAc/hexane
305 ether
75 EtOAc/hexane
67 ether
79 EtOAc/hexane
146 EtOAc/hexane
132 ether
C
C
C
C
C
C
C
C
C
C
C
23H19NO3 (357.4)
24H21NO4 (387.4)
21H17NO3S (363.4)
21H17NO4 (347.4)
23H19NO4 (373.4)
23H18N2O5 (402.4)
23H18FNO3 (375.4)
23H18N2O5 (402.4)
24H21NO3 (371.4)
24H18F3NO3 (425.4)
23H17F2NO3 (393.4)
68 EtOAc/hexane
(m, 1H); IR (KBr) 3322, 2975, 1654, 1579, 1498, 1378, 1151,
864 cm-1; MS 301 (1, M), 283 (29), 185 (36), 137 (11), 121 (9).
Anal. (C17H19NO4) C, H, N.
tr a n s-3,4-Dih yd r o-3-h yd r oxy-2,2-d im et h yl-4-(2-p yr i-
d yloxy)-2H-1-ben zop yr a n (11). From epoxide 7 (3.52 g,
20 mmol) and 2-pyridinol (2.85 g, 30 mmol) according to the
procedure of Bergmann and Gericke8 (ethanol, pyridine) and
silica gel chromatography (toluene, 35/CH2Cl2, 60/2-propanol,
5) gave 11 (0.89 g, 16%) as a white solid: mp 178 °C (from
1
ethyl acetate); H NMR (CDCl3) δ 1.34 (s, 3H), 1.55 (s, 3H),
3.99 (d, J ) 8 Hz, 1H), 5.80 (d, J ) 8 Hz, 1H), 6.68 (s, OH),
6.90 (m, 4H), 7.22 (m, 1H), 7.35 (‘d’, J ) 7.8 Hz, 1H), 7.70 (m,
1H), 8.16 (m, 1H); IR (KBr) 3279, 2979, 1600, 1583, 1572, 1470,
1433, 1306, 754 cm-1; MS 253 (3, M - H2O), 238 (24), 43 (100).
Anal. (C16H17NO3) C, H, N.
tr a n s-3,4-Dih yd r o-3-h yd r oxy-6-m eth oxy-2,2-d im eth yl-
4-(2-p yr id yloxy)-2H-1-ben zop yr a n (12). According to the
procedure described for 11 from the epoxide 8 (4.12 g, 20 mmol)
and 2-pyridinol (2.85 g, 30 mmol) and recrystallization from
ethyl acetate gave 12 (0.36 g, 6%) as a white solid: mp 85 °C;
1H NMR (CDCl3) δ 1.32 (s, 3H), 1.53 (s, 3H), 3.76 (s, 3H), 3.98
(d, J ) 7.9 Hz, 1H), 5.76 (d, J ) 7.9 Hz, 1H), 6.63 (s, OH),
6.84 (m, 3H), 7.00 (m, 2H), 7.71 (m, 1H), 8.14 (‘dd’, J ) 5.2
and 1.2 Hz, 1H); IR (KBr) 3196, 2835, 1604, 1569, 1493, 1281,
1147 cm-1; MS 301 (2, M), 283 (9), 268 (42), 168 (8), 150 (11),
43 (100). Anal. (C17H19NO4) C, H, N.
6-(Am in oh yd r oxyim in om eth yl)-4-(1,2-d ih yd r o-2-oxo-1-
p yr id yl)-2,2-d im eth yl-2H-1-ben zop yr a n (28). Bimakalim
(44)8 (280 mg, 1 mmol) was dissolved in ethanol and mixed
up with a solution of hydroxylamine hydrochloride (0.42 g, 6
mmol) and sodium carbonate (0.32 g, 3 mmol) in 25 mL water.
The mixture was kept on a water bath at 60 °C for 2 h, cooled,
acidified with 2 N HCl, and extracted with ether. The aqueous
layer was neutralized with NH4OH and saturated with (NH4)2-
SO4. This solution was extracted with ethyl acetate, which was
dried (MgSO4) and evaporated to yield a white solid of 28 (0.27
4-(1,2-Dih yd r o-2-oxo-1-p yr id yl)-2,2-d im eth yl-2H-1-ben -
zop yr a n (13). The benzopyranol 9 (5.43 g, 20 mmol) was
dissolved in 300 mL dioxane, and sodium hydroxide on
microlayer (10 g; Merck 1.01564) was added. The suspension
was refluxed for 2 h and filtered, and the filtrate was
evaporated. The residual white solid was recrystallized from
1
g, 87%): mp 205-207 °C (cyclohexane); H NMR (DMSO-d6)
δ 1.46 (s, 3H), 1.51 (s, 3H), 5.72 (s, NH2), 5.90 (s, 1H), 6.33 (m,
1H), 6.47 (dd, J ) 8.9 and 0.9 Hz, 1H), 6.87 (d, J ) 8.4 Hz,
1H), 6.94 (d, J ) 2.0 Hz, 1H), 7.47 (dd, J ) 8.4 and 2.0 Hz,
1H), 7.57 (m, 2H), 9.52 (s, OH); IR (KBr) 3566, 3338, 1658.
1273 cm-1; MS 311 (11, M), 296 (32), 263 (12), 217 (19). Anal.
(C17H17N3O3‚0.5H2O) C, H, N.
1
ether to yield 13 (3.59 g, 71%): mp 114 °C; H NMR (CDCl3)
δ 1.53 (s, 1H), 1.58 (s, 3H), 5.70 (s, 1H), 6.21 (td, J ) 6.9 and
1.3 Hz, 1H), 6.61 (m, 2H), 6.80 (m, 2H), 7.20 (m, 2H), 7.41 (m,
1H); IR (KBr) 3058, 2976, 1677, 1590, 1532, 760 cm-1; MS 253
(21, M), 238 (100), 159 (38). Anal. (C16H15NO2) C, H, N.
4-(1,2-Dih ydr o-2-oxo-1-pyr idyl)-6-m eth oxy-2,2-dim eth yl-
2H-1-ben zop yr a n (14). According to the procedure described
for 13, the benzopyranol 10 (5.66 g, 20 mmol) was dissolved
in 300 mL dioxane, and sodium hydroxide on microlayer (10
g; E. Merck 1.01564) was added. The residual white solid was
recrystallized from ethyl acetate to yield 14 (3.68 g, 65%): mp
6-F or m yl-4-(1,2-d ih yd r o-2-oxo-1-p yr id yl)-2,2-d im eth yl-
2H-1-ben zop yr a n (29). Bimakalim (44)8 (1.39 g, 5 mmol) and
2 g sodium hypophosphite were dissolved in a mixture of 40
mL water/acetic acid/pyridine (1/1/2); 0.5 g Raney nickel was
added, and the whole mixture was stirred for 2 h at 45 °C.
The catalyst was filtered off, and the filtrate was diluted with
100 mL water and extracted with ether. The organic phase
was washed with 2 N HCl, sodium bicarbonate, and water,
dried, and evaporated. The residue was crystallized from ether
to give a white solid (1.20 g, 85%): mp 172 °C; 1H NMR
(DMSO-d6) δ 1.52 (s, 3H), 1.55 (s, 3H), 6.13 (s, 1H), 6.38 (m,
1H), 6.51 (dd, J ) 9.7 and 1.0 Hz, 1H), 7.06 (m, 2H), 7.60 (m,
2H), 7.78 (dd, J ) 8.4 and 2.0 Hz, 1H), 9.77 (s, 1H); 13C NMR
(DMSO-d6) δ 190.36, 161.78, 158.59, 140.44, 137.64, 134.30,
132.55, 130.01, 129.11, 124.64, 121.97, 118.38, 117.55, 106.30,
78.43, 28.32, 28.14; IR (KBr) 1682, 1666, 1596 cm-1; MS 281
1
119 °C; H NMR (CDCl3) δ 1.51 (s, 3H), 1.56 (s, 3H), 3.67 (s,
3H), 5.74 (s, 1H), 6.21 (m, 2H), 6.64 (‘d’, J ) 9.2 Hz, 1H), 6.72
(dd, J ) 8.8 and 2.8 Hz, 1H), 6.81 (d, J ) 8.8 Hz, 1H), 7.17
(‘dd’, J ) 7.0 and 2.0 Hz, 1H), 7.41 (m, 1H); IR (KBr) 2978,
1673, 1593, 1577, 1267, 862 cm-1; MS 283 (33, M), 268 (100),
225 (6), 189 (43), 173 (17), 145 (19). Anal. (C17H17NO3) C, H,
N.
Gen er a l P r oced u r es for th e Syn th esis of Ben zop yr a n
Keton es (Ta ble 2). Meth od a for th e P r ep a r a tion of
Keton es 15-20. The appropriate acid chloride was freshly
prepared by refluxing (3 h) the corresponding acid (0.01 mol)
in thionyl chloride (11.9 g, 0.1 mol). The excess SOCl2 was
distilled off, and the residue was dried in an exsiccator over
KOH pellets. This material was used in the next step without
(14, M), 266 (53), 187 (52), 115 (19), 78 (100). Anal. (C17H15
NO3) C, H, N.
-
6-(2,2-Dicya n oeth en yl)-4-(1,2-d ih yd r o-2-oxo-1-p yr id yl)-
2,2-d im eth yl-2H-1-ben zop yr a n (30). The aldehyde 29 (0.84
g, 3 mmol) and malononitrile (0.2 g, 3 mmol) were dissolved

6-Substituted Benzopyrans as K+ Channel Activators
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6 989
Ta ble 3. Chemical Data of Carbamates 38-42
in 50 mL toluene. After addition of each 1 drop of acetic acid
and piperidine, the mixture was boiled on a water separator
for 2 h. The toluene phase was washed with 2 N HCl and then
with sodium bicarbonate and water, dried, and evaporated.
The residue was crystallized from ethyl acetate/hexane to give
yield,
%
mp,
°C
IR (KBr)
(cm-1
compd
)
formula (Mr)
38
39
40
41
42
87
85
86
59
68
195
171
185
145
160
1734, 1665
1736, 1666
1734, 1665
1744, 1663
1734, 1662
C
C
C
C
C
23H20N2O4 (388.4)
23H19FN2O4 (406.4)
23H19FN2O4 (406.4)
24H19F3N2O4 (456.4)
24H19F3N2O4 (456.4)
1
a yellow solid of 30 (0.87 g, 88%): mp 202-204 °C; H NMR
(DMSO-d6) δ 1.53 (s, 3H), 1.56 (s, 3H), 6.16 (s, 1H), 6.35 (m,
1H), 6.46 (dd, J ) 9.8 and 1.0 Hz, 1H), 7.11 (d, J ) 8.6 Hz,
1H), 7.36 (d, J ) 2.2 Hz, 1H), 7.56 (m, 2H), 7.84 (dd, J ) 8.6
and 2.2 Hz, 1H), 8.36 (s, 1H); IR (KBr) 2224, 1666, 1582, 1226
cm-1; MS 329 (3, M), 314 (15), 235 (19), 78 (100). Anal.
(C20H15N3O2) C, H, N.
1
crystals; mp 238 °C (from methanol); H NMR (DMSO-d6) δ
1.41 (s, 3H), 1.46 (s, 3H), 5.91 (s, 1H), 5.99 (d, J ) 2.7 Hz,
1H), 6.32 (‘dt’, J ) 6.5 and 1.0 Hz, 1H), 6.47 (‘d’, J ) 9.0 Hz,
1H), 6.57 (dd J ) 8.6 and 2.7 Hz, 1H), 6.69 (d, J ) 8.6 Hz,
1H), 7.55 (m, 2H), 8.97 (s, OH); IR (KBr) 3126, 1658, 1652,
1570, 1537, 1150 cm-1; MS 269 (23, M), 254 (100), 175 (33),
147 (6), 131 (10), 115 (10). Anal. (C16H15NO3) C, H, N.
6-Acetoxy-(1,2-d ih yd r o-2-oxo-1-p yr id yl)-2,2-d im eth yl-
2H-1-ben zop yr a n (35). Phenol 34 (2.69 g, 10 mmol), acetyl
chloride (940 mg, 12 mmol), 1-2 drops of trimethylamine, and
100 mL dry toluene was stirred under reflux for 1 h. After
evaporation the residue was dissolved in methylene chloride,
and the organic layer was washed with 2% sulfuric acid and
2% sodium hydroxide. The organic phase was dried (MgSO4)
and evaporated. The residue was crystallized from methylene
chloride/hexane to give the ester derivative 35 (2.49 g, 80%):
colorless crystals; mp 173 °C from hexane; 1H NMR (CDCl3) δ
1.53 (s, 3H), 1.58 (s, 3H), 2.21 (s, 3H), 5.74 (s, 1H), 6.22 (‘dt’,
J ) 6.8 and 1.2 Hz, 1H), 6.41 (d, J ) 2.3 Hz, 1H), 6.63 (‘d’, J
) 8.4 Hz, 1H), 6.84 (d J ) 8.3 Hz, 1H), 6.90 (dd, J ) 8.3 and
2.3 Hz, 1H), 7.18 (‘dd’, J ) 6.9 and 1.9 Hz, 1H), 7.42 (m, 1H);
IR (KBr) 2974, 1754, 1674, 1593, 1531, 1218, 764 cm-1; MS
311 (6, M), 296 (9), 269 (5), 254 (41), 175 (19), 117 (6), 96 (22),
43 (100). Anal. (C18H17NO4) C, H, N.
4-(1,2-Dih yd r o-2-oxo-1-p yr id yl)-2,2-d im et h yl-6-n it r o-
2H-1-ben zop yr a n (31). This compound is already described
by Bergmann and Gericke,8 but we used another method for
preparation. The benzopyran 13 (0.5 g, 2 mmol) was dissolved
in dry dichloromethane (50 mL), and nitronium tetrafluoro-
borate (0.3 g, 2.2 mmol) was added. After stirring under
anhydrous conditions for 24 h at room temperature, the solvent
was washed with sodium bicarbonate and water, dried (Mg-
SO4), and evaporated. The residue was recrystallized from
ethyl acetate to yield a nearly colorless solid of 31 (0.36 g,
1
60%): mp 164 °C (lit.8 mp 156-158 °C); H NMR (DMSO-d6)
δ 1.54 (s, 3H), 1.57 (s, 3H), 6.25 (s, 1H), 6.41 (m, 1H), 6.54 (d,
J ) 9.3 Hz, 1H), 7.09 (d, J ) 9.0 Hz, 1H), 7.32 (d, J ) 2.6 Hz,
1H), 7.65 (m, 2H), 8.11 (dd, J ) 9.0 and 2.6 Hz, 1H); IR (KBr)
1666, 1594, 1534, 1517, 1481 cm-1; MS 298 (1, M), 283 (21),
238 (73), 159 (38). Anal. (C16H14N2O4) C, H, N.
6-Am in o-4-(1,2-d ih yd r o-2-oxo-1-p yr id yl)-2,2-d im eth yl-
2H-1-ben zop yr a n (32). The nitro compound 31 (0.9 g, 3
mmol) was reduced with an excess of elementary iron powder
(0.3 g) in 100 mL 5 N H2SO4 at room temperature for 3 h. The
remaining solution was filtered and neutralized with NaOH
to pH 7, and tartaric acid was added. This solution was
extracted several times with ethyl acetate, which after drying
and evaporation gave the amine 32 as a light-brown solid (0.52
Gen er a l P r oced u r e for th e Syn th esis of Eth er Der iva -
tives 36-37. Phenol 34 (2.69 g, 10 mmol) was stirred under
reflux with the appropriate phenacyl bromide (11 mmol) and
potassium carbonate in acetone for 1 h. The resultant suspen-
sion was filtrated and evaporated. The residue was purified
by silica gel chromatography (10% acetone/toluene) to give the
ethers as colorless crystals.
4-(1,2-Dih yd r o-2-oxo-1-p yr id yl)-2,2-d im eth yl-6-(p h en -
a cyloxy)-2H-1-ben zop yr a n (36): colorless crystals (1.82 g,
47%); mp 137 °C (from hexane); 1H NMR (CDCl3) δ 1.51 (s,
3H), 1.56 (s, 3H), 5.13 (s, 2H), 5.74 (s, 1H), 6.18 (‘dt’, J ) 6.8
and 1.2 Hz, 1H), 6.32 (‘d’, J ) 1.9 Hz, 1H), 6.59 (‘d’ J ) 9.2
Hz, 1H), 6.77 (m, 2H), 7.14 (‘dd’, J ) 6.9 and 1.9 Hz, 1H), 7.46
(m, 4H), 7.93 (m, 2H); IR (KBr) 2978, 1700, 1672, 1662, 1269,
764 cm-1; MS 387 (14, M), 372 (32), 268 (30), 173 (10), 105
(100). Anal. (C24H21NO4) C, H, N.
4-(1,2-Dih ydr o-2-oxo-1-pyr idyl)-6-(4-flu or oph en acyloxy)-
2,2-d im eth yl-2H-1-ben zop yr a n (37): colorless crystals (2.35
g, 58%); mp 159 °C (from hexane); 1H NMR (CDCl3) δ 1.51 (s,
3H), 1.56 (s, 3H), 5.07 (s, 2H), 5.74 (s, 1H), 6.19 (‘dt’, J ) 6.9
and 1.5 Hz, 1H), 6.31 (d, J ) 2.5 Hz, 1H), 6.60 (‘d’, J ) 9.0 Hz,
1H), 6.77 (m, 2H), 7.15 (m, 3H), 7.39 (m, 1H), 7.98 (m, 2H); IR
(KBr) 3074, 1704, 1666, 1595, 1138 cm-1; MS 405 (8, M), 390
(21), 268 (16), 145 (12), 123 (100). Anal. (C24H20FNO4) C, H,
N.
Gen er a l P r oced u r e for th e Syn th esis of Ca r ba m a tes
38-42. Phenol 34 (2.69 g, 10 mmol) was dissolved in 50 mL
absolute toluene; then 2 drops of triethylamine and 10 mmol
of the appropriate isocyanate were added. The mixture was
stirred for 1 h under reflux. After the solvent was removed,
the residue was crystallized from hexane to give the carbam-
ates 38-42 as colorless crystals (Table 3).
1
g, 65%): mp 157 °C; H NMR (DMSO-d6) δ 1.38 (s, 3H), 1.43
(s, 3H), 4.67 (s, NH2), 5.83 (s, 1H), 5.84 (s, 1H), 6.30 (m, 1H),
6.40 (dd, J ) 8.5 and 2.6 Hz, 1H), 6.45 (d, J ) 8.5 Hz, 1H),
6.57 (d, J ) 8.5 Hz, 1H), 7.52 (m, 2H); IR (KBr) 3423, 3336,
1663, 1590, 1274 cm-1; MS 268 (63, M), 253 (100), 173 (78).
Anal. (C16H16N2O2) C, H, N.
4-(1,2-Dih yd r o-2-oxo-1-p yr id yl)-6-(2,5-d im eth yl-1-p yr -
r olyl)-2,2-d im eth yl-2H-1-ben zop yr a n (33). The amine 32
(0.54 g, 2 mmol) was mixed with 10 mL acetonylacetone and
kept at 160 °C for 2 h. The excess of diketone was distilled in
vacuo and the residue purified by silica gel chromatography
with toluene/acetone (60/40); yellowish solid from methanol/
water (0.4 g, 58%); mp 146 °C; 1H NMR (DMSO-d6) δ 1.51 (s,
3H), 1.55 (s, 3H), 1.88 (s, 6H), 5.71 (s, 2H), 6.07 (s, 1H), 6.24
(d, J ) 2.3 Hz, 1H), 6.31 (m, 1H), 6.46 (d, J ) 9.2 Hz, 1H),
6.96 (d, J ) 8.5 Hz, 1H), 7.09 (dd J ) 8.5 and 2.3 Hz, 1H),
7.55 (m, 2H); IR (KBr) 1662, 1593, 1533 cm-1; MS 346 (14,
M), 331 (13), 252 (17), 236 (11), 208 (53). Anal. (C22H22N2O2)
C, H, N.
Syn th esis of 4-(1,2-Dih ydr o-2-oxo-1-pyr idyl)-6-h ydr oxy-
2,2-d im eth yl-2H-1-ben zop yr a n (34). Meth od a : The ben-
zopyran 14 (5.67 g, 20 mmol) was melted with pyridine
hydrochloride (6.93, 60 mmol) at 170 °C for 2 h. The hot
solution was poured in 150 mL sulfuric acid (2%), and the
product was extracted with methylene chloride. The organic
layer was dried (MgSO4), and evaporation followed by crystal-
lization from methanol gave the phenol 34 (2.21 g, 41%).
Meth od b: 5.67 g (20 mmol) of compound 14 was dissolved
in 40 mL dry dichloromethane and cooled to -10 °C. A mixture
of (25 g, 0.1 mol) boron tribromide and 30 mL dichloromethane
was slowly added, and the solution was stirred for 1 h.
Subsequently the solution was poured into water, and the
product was extracted three times with ethyl acetate. The
combined ethyl acetate phases were dried (MgSO4) and
evaporated. The residue was washed with ether, and the
crystalline solid was purified by recrystallization from metha-
nol to give the phenol derivative 34 (3.12 g, 58%): colorless
4-(1,2-Dih yd r o-2-oxo-1-p yr id yl)-2,2-d im eth yl-2H-1-ben -
zop yr a n -6-yl F lu or osu lfon a te (43). A solution of (2.69 g,
10 mmol) phenol 34 in 30 mL pyridine was stirred in a flask
fitted with a condenser cooled by solid carbon dioxide. The
mixture was chilled to 0 °C, and 1.77 g (15 mmol) of sulfuryl
chloride fluoride was introduced. The mixture was slowly
warmed and held at 30 °C for 1 h. It was then poured into
ice-cold hydrochloric acid, and the product was extracted with

990 J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6
Mannhold et al.
ether. The ethereal extract was washed with water, then with
dilute aqueous sodium chloride, and again with water. After
evaporation the residue was purified by silica gel chromatog-
raphy (ethyl acetate) to give nearly colorless crystals (562 mg,
volumes were computed at 0.25 kcal/mol, negative attractive
volumes in
a range from -0.5 up to -6 kcal/mol. This
procedure yields a minimum of 10 descriptor variables. The
same procedure was applied to the MEP computed by Spartan,
yielding 10 more molecular descriptors. Some of these can be
higly correlated. However, their simultaneous use is to be
preferred since spourious errors can be better detected and
compensated.
1
16%): mp 121 °C (from hexane); H NMR (CDCl3) δ 1.55 (s,
3H), 1.60 (s, 3H), 5.81 (s, 1H), 6.27 (‘dt’, J ) 6.8 and 1.6 Hz,
1H), 6.65 (m, 2H), 6.90 (d, J ) 8.9 Hz, 1H), 7.16 (m, 2H), 7.46
(m, 1H); IR (KBr) 3066, 2980, 1666, 1593, 1534, 1484, 1446,
1232, 1210 cm-1; MS 351 (7, M), 336 (24), 268 (8), 257 (18),
149 (15), 127 (14), 117 (12), 78 (100). Anal. (C16H14FNO5S) C,
H, N.
D. Sta tistica l Ap p r oa ch es. Principal component and PLS
analyses25 were performed with the GOLPE26 software, version
3.1, on a Silicon Graphics workstation.
B. Biologica l Meth od s. Rela xa n t P oten cy in Ra t Aor ta
a n d Tr a ch ea . For biological characterization of the dilator
properties of KCA, male Wistar rats of 250-350 g of body
weight were used. After dissection of aorta and trachea, tissues
were placed in Krebs solution containing (mmol/L) 120 Na+,
5 K+, 2.25 Ca2+, 0.5 Mg2+, 98.5 Cl-, 34 HCO3-, 1HPO42-, and
0.5 SO42-, equilibrated with carbogen (95% O2, 5% CO2).
Ack n ow led gm en t. Part of this work was done while
Raimund Mannhold was a guest of the Laboratory of
Chemometrics, University of Perugia, Italy; he wants
to express his gratitude to Sergio Clementi and Gabriele
Cruciani for their cooperation and hospitality.
Aortic rings of 5-mm width were prepared and denuded from
endothelium, success of which was tested by lacking dilator
response to acetylcholine (1 mmol/L). Tracheal strips were
prepared by splitting tracheal rings on the opposite side of the
ring muscle layer as described by Lemoine and Overlack.19
Su p p or tin g In for m a tion Ava ila ble: Spectroscopic data
(IR, 1H and 13C NMR, MS) on ketones 15-27 and carbamates
38-42. This material is available free of charge via the
Internet at http://pubs.acs.org.
Tissues were mounted in
a
modified Blinks apparatus20
equipped with 7-mL baths and isometric strain gauge trans-
ducers (Swema SG4-45). Pretension of smooth muscle strips
was adjusted to 0.5 g. Experiments were performed at 32.5
°C in Krebs solution, supplemented with (mmol/L) fumarate
(5.0), pyruvate (5.0), glutarate (5.0), bicarbonate (5.0), and
glucose (10.0).
Refer en ces
(1) Hoshi, T.; Zagotta, W. N. Recent advances in the understanding
of potassium channel function. Curr. Opin. Neurobiol. 1993, 3,
283-290.
(2) J an, L. Y.; J an, Y. N. A superfamily of ion channels. Nature 1990,
345, 672.
Contractures were induced with 25 mmol/L K+ and 0.6
µmol/L carbachol in aortic rings and tracheal strips, respec-
tively. After developed force had reached a stable level of
contraction, cumulative concentration-effect curves were
performed for each test compound.
(3) Noma, A. ATP-regulated
KATP-channels in cardiac muscle.
Nature 1983, 305, 147-148.
(4) A¨ mma¨la¨, C.; Moorhouse, A.; Ashcroft, F. M. The sulfonylurea
receptor confers diazoxide sensitivity on the inwardly rectifying
K+-channel KIR 6.1 expressed in human embryonic kidney cells.
J . Physiol. 1996, 494, 709-714.
(5) Edwards, G.; Weston, A. H. Structure-activity relationships of
K+ channel openers. Trends Pharmacol. Sci. 1990, 11, 417-422.
(6) Evans, J . M., Hamilton, T. C., Longman, S. D., Stemp, G., Eds.
Potassium Channels and Their Modulators. From Synthesis to
Clinical Experience; Taylor & Francis: London, 1996.
(7) Kawase Y.; Yamaguchi, S.; Horita, H.; Takeno, J .; Kameyama,
H. A new preparation method of 2,2-dimethyl-2H-chromenes.
Bull. Chem. Soc. J pn. 1982, 55, 1153-1155.
Compounds were dissolved in stock solutions containing
10-2 M DMSO; dilutions were performed in distilled water.
Final concentration of DMSO in the organ bath did not exceed
2% to avoid dilator effects of the solvent. Dose-response curves
were fitted with the software SAS 604; half-maximal potency
of the KCA was derived from the curves as pEC50 value.
C. Ch em ica l Descr ip tor s. Log P data were calculated with
the KOWWIN software.21 The matrix of the other chemical
descriptors was built using different modeling programs.
Geometry optimization was obtained using Spartan22 at the
AM1-acq level. Molecular descriptors such as heat of forma-
tion, ionization potential, dipole moment, HOMO, LUMO,
electronegativity, hardness, and polarity were computed by
Spartan using the QSAR option. The remaining descriptors
were computed from the 3D-interaction energy maps produced
by GRID23 and transformed by VolSurf.24 GRID force field
detects favorable binding sites for chemical probes on the
surface of target molecules. For a drug molecule surrounded
by the water probe, the trapped water molecules are used to
define hydrophilic regions; the hydrophobic probe DRY is
applied for defining hydrophobic regions. Area and volume of
this envelope vary with the level of interaction energies
between probe and target molecule. However, at a fixed
isoenergetic contour level, the hydrophilic area and volume
are well-defined for each molecule. VolSurf is a program to
visualize, quantify, and transform the 3D-energy maps into
useful chemical descriptors. From the 3D-interaction energy
maps VolSurf is able to compute molecular volume, surface,
and globularity, interaction energy moments, and capacity
factors. The globularity is defined as S/Se, in which Se is the
surface area of a sphere of volume V, which is the equivalent
sphere. Globularity equals 1 for spherical molecules. Moreover
VolSurf calculates the integy moment; it is defined similarly
to the dipole moment but calculated from the interaction
energies. Finally capacity factors are calculated, which are
defined as the ratio between the attractive molecular volume
and the molecular surface, which is the amount of negative
interaction per surface unit. Positive repulsive surfaces and
(8) Bergmann, R.; Gericke, R. Synthesis and antihypertensive
activity of 4-(1,2-Dihydro-2-oxo-1-pyridyl)-2H-1-benzopyrans and
related compounds, new potassium channel activators. J . Med.
Chem. 1990, 33, 492-504.
(9) Derix, A. 6-Substituierte Benzopyrane als Modulatoren des KATP
-
Kanals. Synthese, Pharmakologie und Struktur-Wirkungs-
beziehungen. Ph.D. Thesis, Faculty of Science, Heinrich-Heine-
Universita¨t, Du¨sseldorf, Germany, 1998.
(10) Effenberger, F.; Ko¨nig, G.; Klenk, H. Acylierung von Aromaten
mit Carbonsa¨ure-dihalogen-phosphorsa¨ure-anhydriden. Chem
Ber. 1981, 114, 926-936. (b) Effenberger, F.; Epple, G.; Eber-
hard, J . K.; Bu¨hler, U.; Sohn, E. Carbonsa¨ure-trifluormethansul-
fonsa¨ure- und -methansulfonsa¨ure-anhydride als hochwirksame
Acylierungsmittel. - Perfluoralkansulfonsa¨ure-katalysierte Aro-
matenacylierungen. Chem. Ber. 1983, 116, 1195-1208.
(11) Backeberg, O. G.; Staksun, B. A. A novel reduction of nitriles to
aldehydes. J . Org. Chem. 1962, 3961-3963.
(12) Bishop, W. S. New 2,5-dimethylpyrrole derivates. J . Am. Chem.
Soc. 1945, 67, 2261-2262.
(13) Prey, V. Die Spaltung von Phenola¨thern mit Pyridinhydrochlo-
rid. Chem. Ber. 1941, 74, 1219-1225.
(14) Claisen, L.; Eisleb, O. U¨ ber die Umlagerung von Phenol-
allyla¨thern in die isomeren Allylphenole. Liebigs Ann. Chem.
1913, 401, 21-119.
(15) Baker, J . W.; Holdsworth, J . B. The Mechanism of Aromatic Side-
Chain Reactions with Special Reference to the Polar Effects of
Substituents. Part XIII. Kinetic Examination of the Reaction of
Arylisocyanates with Methyl Alcohol. J . Chem. Soc. 1947, 713-
726.
(16) Cramer, R.; Coffman, D. D. New Syntheses of aryl-fluorides and
aryl fluorosulfonates from oxyfluorides of sulfur. J . Org. Chem.
1961, 4164-4165.
(17) Ishizawa, T.; Koga, H.; Ohta, M.; Sato, H.; Makino, T.; Kuro-
maru, K.; Taka, N.; Takahashi, T.; Sato, T.; Nabata, H. Struc-
ture-activity relationships of 6-substituted benzopyran-4-
carbothioamide potassium channel openers. Bioorg. Med. Chem.
Lett. 1993, 3, 1659-1662.

6-Substituted Benzopyrans as K+ Channel Activators
J ournal of Medicinal Chemistry, 1999, Vol. 42, No. 6 991
(18) Ohta, M.; Koga, H.; Sato, H.; Ishizawa, T. Comparative molecular
field analysis of benzopyran-4-carbothioamide potassium chan-
nel openers. Bioorg. Med. Chem. Lett. 1994, 4, 2903-2906.
(19) Lemoine, H.; Overlack, C. Highly potent â2-sympathomimetics
convert to less potent partial agonists as relaxants of guinea-
pig tracheae maximally contracted by carbachol. Comparison of
relaxation with receptor binding. J . Pharmacol. Exp. Ther. 1992,
261, 258-270.
(23) VolSurf program is avalaible at the site: http://chemiome.chm.
unipg.it.
(24) Goodford, P. A Computational Procedure for Determining En-
ergetically Favorable Binding Sites on Important Macromol-
ecules. J . Med. Chem. 1985, 28, 849-857.
(25) Wold, S.; Esbensen, K.; Geladi, P. Principal component analysis.
Chem. Intell. Lab. System 1987, 2, 37-52.
(26) Baroni, M.; Costantino, G.; Cruciani, G.; Riganelli, D.; Valigi,
R.; Clementi, S. Generating Optimal Linear PLS Estimations
(GOLPE): An Advanced Chemometric Tool for Handling 3D-
QSAR Problems. Quant. Struct.-Act. Relat. 1993, 12, 9-20.
(20) Blinks, J . Convenient apparatus for recording contractions of
isolated muscle. J . Appl. Physiol. 1965, 20, 755-757.
(21) Meylan, W. M.; Howard, P. H. Atom/Fragment contribution
method for estimating octanol-water partition coefficients. J .
Pharm. Sci. 1995, 84, 83-92.
(22) Spartan, version 4.1, Wavefunction, Irvine, CA 92612.
J M981047M