New calcium antagonists: Synthesis, X-ray analysis, and smooth muscle relaxing effect of 3-[O-(benzyl-substituted)-oximino-ethers]-hexahydroazepin-2,3-diones

Article abstract of DOI:10.1016/S0968-0896(99)00090-5

A series of new Z and E 3-[O-(benzyl-substituted)-oximino-ether]-hexahydroazepin-2,3-diones was prepared from the corresponding hexahydroazepin-2,3-diones and examined as smooth muscle relaxants. E and Z structures were assigned by NMR analysis and confirmed for 16 (E and Z) by an X-ray diffraction using synchrotron radiations. The nitrobenzyl derivative 16 was the most potent in vitro as relaxant of rat trachea precontracted with acetylcholine. The E isomer 16b was more potent than the Z isomer 16a. E isomer 16b is more potent than aminophylline to relax both rat trachea and human bronchus.This derivative acts mainly by inhibiting cellular infux of extracellular calcium since it inhibits potently and dose-dependently the contractions of rat trachea to high concentrations of KCl and to CaCl2 in a depolarizing medium. It appears to act weakly by inducing cGMP and cAMP synthesis. Moreover, its relaxing activity is not related to an inhibition of phosphodiesterases, to opening of potassium channels or to induction of prostaglandin synthesis. Therefore, 16b appears to work mainly as a potent calcium antagonist. (C) 1999 Elsevier Science Ltd.

Full text of DOI:10.1016/S0968-0896(99)00090-5

Bioorganic & Medicinal Chemistry 7 (1999) 1655±1663
New Calcium Antagonists: Synthesis, X-ray Analysis, and Smooth
Muscle Relaxing E€ect of 3-[O-(Benzyl-substituted)-oximino-
ethers]-hexahydroazepin-2,3-diones
Hayat El From, a Marie-Helene Pera, a,* Gerard Leclerc, a Duc Tranqui, b
Emmanuelle Corompt, c Germain Bessard c and Philippe Devillier c
a
Â
Groupe de Pharmacochimie Moleculaire, EP 811 CNRS, UFR de Pharmacie, Universite Joseph Fourier. BP 138, F-38243 Meylan Cedex,
Â
France
 Á
Laboratoire de Cristallographie associe a l'UJF, CNRS, BP166, F-38042 Grenoble Cedex, France
b
c
Â
Â
Laboratoire de Pharmacologie UFR de Medecine, Universite Joseph Fourier F-38706 La Tronche Cedex, France
Received 28 October 1998; accepted 12 March 1999
AbstractÐA series of new Z and E 3-[O-(benzyl-substituted)-oximino-ether]-hexahydroazepin-2,3-diones was prepared from the
corresponding hexahydroazepin-2,3-diones and examined as smooth muscle relaxants. E and Z structures were assigned by NMR
analysis and con®rmed for 16 (E and Z) by an X-ray di€raction using synchrotron radiations. The nitrobenzyl derivative 16 was the
most potent in vitro as relaxant of rat trachea precontracted with acetylcholine. The E isomer 16b was more potent than the Z
isomer 16a. E isomer 16b is more potent than aminophylline to relax both rat trachea and human bronchus.This derivative acts
mainly by inhibiting cellular in¯ux of extracellular calcium since it inhibits potently and dose-dependently the contractions of rat
trachea to high concentrations of KCl and to CaCl2 in a depolarizing medium. It appears to act weakly by inducing cGMP and
cAMP synthesis. Moreover, its relaxing activity is not related to an inhibition of phosphodiesterases, to opening of potassium
channels or to induction of prostaglandin synthesis. Therefore, 16b appears to work mainly as a potent calcium antagonist. # 1999
Elsevier Science Ltd. All rights reserved.
Introduction
bronchus and examination of their mechanism of
action.
It is well documented that various amide or lactame
derivatives are endowed with actions on the central
nervous system via a stimulation of various ion chan-
nels.1 In a previous work we have shown that oximino-
ether derivatives exhibited marked antinociceptive and
anticonvulsivant activities;2 therefore, we became inter-
ested in the preparation of new compounds combining
within a single structure an oximino-ether function and
a lactame ring and more precisely the scarcely examined
caprolactame ring.The biological activity of these com-
pounds on airway smooth muscles, which are richly
endowed with K+ and Ca++ channels were eval-
uated. This report describes the preparation and X-ray
analysis of these new oximino-caprolactames with attri-
bution of their Z and E con®gurations, evaluation of
their spasmolytic activity on rat trachea and human
Chemistry
Our strategy for the synthesis of target compounds 11±
18 (Scheme 1) was based on the construction of the
hexahydroazepin-2,3-dione skeleton 4, followed by
attachment of the substituted O-hydroxylamines 5±10.
Duong et al.3 obtained compound 4 by coupling hydra-
zoic acid and 2-[2-(1-cyclohexanone)]-1,3-dioxolane
with polyphosphoric acid. The resulting instable inter-
mediate, gave 4 after a Schmidt transposition, but with
a very poor yield (2%) and the hydrolysis of enamine
was more conveniently accomplished on a silica gel col-
umn to provide pure oxolactam in high yield, but only
with hexa or pentacyclic oxolactam.4 We have prepared
this compound by hydrolysis of the ene-morpholine 3
in acidic medium. This route a€orded hexahydro-
azepin-2,3-dione 4 with 30% yield. Various methods
are available for the preparation of a,a-dichloro-
caprolactam 2.5,6 The use of phosphorus pentachloride
Key words: Hexahydroazepin-2,3-diones; synthesis; X-ray; calcium
antagonist; muscle relaxing.
* Corresponding author. Tel.: +33-4-7604-1006; fax: +33-4-7604-
1007.
0968-0896/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved.
PII: S0968-0896(99)00090-5

1656
H. El From et al. / Bioorg. Med. Chem. 7 (1999) 1655±1663
Scheme 1. Synthesis of oximino caprolactame derivatives. (a) PCl5; (b) morpholine; (c) H2O/H+; (d) NH2OH; (e) KOH/EtOH, R-Br; (f) H2O/H+
(g) pyridine; (h) H2/Pd; (i) CH3-SO2-Cl.
;
without solvent allowed us to notably increase the
reported yields. Treatment of compound 2 with mor-
pholine led to the ene-morpholine 3.7,8 Compounds
5±10 were then prepared according to Garoufalias et
al.9 (Scheme 1), whereas compounds 11±16 (Table 1)
were obtained from 5±10 by reaction with 4 in re¯uxing
ethanol containing 5±10% of pyridine. Reduction of the
nitro derivative 16 using H2/Pd gave the amino deriva-
tive 17, which was ®nally treated by CH3SO2Cl to
obtain the sulfonamide 18. Oximino-ethers 11±18 were
isolated as mixtures of Z and E isomers in which the E
isomer predominated. Isomerisation of 16a to 16b
occurred when 16a (Z) was left aside for some days into
ethylacetate, at room temperature. Ratio of isomers
were assessed by NMR spectroscopy on the basis of
the chemical shifts observed for protons on carbon
adjacent to the oxime double bond.10 In accordance
with assignments reported in the literature for several
oximes and alkylated oximes,11 Z isomers are more
down®eld than their E counterparts. In order to verify
these assignments, an unambiguous structure determi-
nation of the two Z and E isomers, 16a and 16b of 3-[O-
(4-nitrobenzyl)-oximino-ether]-hexahydroazepin-2,3-
diones was ®nally accomplished through X-ray di€rac-
tion. The results obtained are in full agreement with
NMR data.
Table 1. Physiochemical data
Compd
R
mp (ꢀC)
Yield%
Formula (CNH)c
a
5
6
7
CH3
C6H5CH2
4-CNC6H4CH2
229±231b
197±199
134±136
143±145
216±218b
137±139
135±137
69±71
55
25
30
30
35
22
78
10
90
20
80
35
75
30
70
40
60
10
90
20
80
C7H10NOCl
C8H9N2OCl
C7H9N2O3Cl
C7H9N2O3Cl
C7H9N2O3Cl
C7H12N2O2
C7H12N2O2
C13H16N2O2
C13H16N2O2
C14H15N3O2
C14H15N3O2
C13H15N3O4
C13H15N3O4
C13H15N3O4
C13H15N3O4
C13H15N3O4
C13H15N3O4
C13H17N3O2
C13H17N3O2
C14H19N3O4S
C14H19N3O4S
8
9
10
2-NO2C6H4CH2
3-NO2C6H4CH2
4-NO2C6H4CH2
CH3(Z)
CH3(E)
C6H5CH2(Z)
C6H5CH2(E)
4-CNC6H4CH2(Z)
4-CNC6H4CH2(E)
2-NO2C6H4CH2(Z)
2-NO2C6H4CH2(E)
3-NO2C6H4CH2(Z)
3-NO2C6H4CH2(E)
4-NO2C6H4CH2(Z)
4-NO2C6H4CH2(E)
4-NH2C6H4CH2(Z)
4-NH2C6H4CH2(E)
4-CH3SO2NHC6H4CH2(Z)
4-CH3SO2NHC6H4CH2(E)
11a
11b
12a
12b
13a
13b
14a
14b
15a
15b
16a
16b
17a
17b
18a
18b
74±76
141±143
139±141
124±126
119±121
126±128
107±109
120±122
103±105
163±165
154±156
Oil
Oil
a
Aldrich compd.
Literature values: 6, 230ꢀC;19 10, 217ꢀC.19
Symbols in parentheses refer to elements analyzed.
b
c

H. El From et al. / Bioorg. Med. Chem. 7 (1999) 1655±1663
1657
Results and Discussion
X-ray crystallographic study of 16a and 16b
The two isomers, 16a (Z) and 16b (E), were selected for
X-ray crystallographic study. Their molecular structures
are presented showing the atoms labelling (Figs 1 and
2). The con®guration of the C8±N2 double bond (1.275
(6) A) is syn (Z) for 16a and anti (E) for 16b, with dis-
tances O3±C9=3.614 (9) A for the E isomer and 2.569
A for the Z isomer. The caprolactame ring has a chair
or pseudo-chair conformation as determined on the one
hand by the distances of atoms C9, C11 and C12 to the
plane P1 de®ned by C8C13C10N3, on the other hand by
the distances of atoms C12, C9 and N3 to the plane
de®ned by C8H13AC11C10 (H13A is above H13B).
The molecules are held together by hydrogen bonds
between H3...O4 (d=2.119 (9)A) for E isomer and by
H3...N2 (d=2.180 (14)A) for Z isomer.
Figure 2. 16a (Z) crystallographic molecular structure with atoms
labelling.
Pharmacology
Relaxant e€ects of the oximino-caprolactame deriva-
tives on the rat trachea were evaluated. The most potent
oximino-caprolactame derivatives were found to be
compounds 16 (16a+16b), 16b being more potent than
16 and 16a (Fig. 3). The concentration of 16b that
caused a relaxation equal to 50% of the maximal
relaxation induced by aminophylline (IC50) was
36.1Â10 6 M and the maximal relaxation at the highest
caprolactame compound is about 16-fold less potent
than isoprenaline but has a much higher ecacy in rat
trachea (Fig. 4). In human bronchus, by the same way
of comparison, 16b appears fourfold more potent than
aminophylline but ®vefold and 14-fold less potent than
verapamil and cromakalim and also 110- and 750-fold
less potent than salbutamol and isoprenaline, respec-
tively (Fig. 4, Table 3). Compound 16b did not act by
opening ATP-potassium channels, like cromakalim,
since its relaxant e€ect is not inhibited by glibenclamide
4
concentration (3Â10 M) was 88‹7.5%. This com-
pound is about twofold more potent than 16
M) and more than 10-fold more
6
(IC50=68.1Â10
4
potent than 16a (IC50>3Â10 M). The activities of
5
oximino-caprolactame derivatives are shown in Table 2.
(10 M, n=5), a potent and selective blocker of this
type of potassium channels. It did not appear to act by
opening other types of potassium channels that may be
involved in the relaxation of airway smooth muscle since
The comparison with classical relaxing drugs of the
e€ect of 16b has been performed on the rat trachea and
the human bronchus (Fig. 4). On the rat trachea, com-
parison of the IC50 indicates that 16b appears as potent
as verapamil, is about 14-fold less potent than croma-
kalim and 19-fold more potent than aminophylline. In
addition, the comparison with isoprenaline, due to its
maximal e€ect inferior to 50% of the maximal relaxa-
tion with aminophylline,12 can be done by taking into
account the IC25 of isoprenaline (5.35Â10 7 M) and the
6
IC25 of 16b (8.71Â10 M). It shows that the oximino-
Figure 3. Concentration±relaxation curves to 16 (n=23), 16a (n=7)
and 16b (n=17) in rat trachea precontracted with acetylcholine.
Values are mean‹S.E.M. calculated on the number of experiments
4
indicated in parentheses excepted for the data at 3Â10 M con-
centration, for which the mean‹S.E.M. have been calculated on ®ve
to seven experiments. Signi®cant di€erences from 16 are *p<0.05;
**p<00.1 ***p<000.1
Figure 1. 16b (E) crystallographic molecular structure with atoms
labelling.

1658
H. El From et al. / Bioorg. Med. Chem. 7 (1999) 1655±1663
6
Table 2. Relaxant e€ects of the oximino-caprolactame derivatives on
the rat trachea
than verapamil. Indomethacin (10 M, n=6) did not
inhibit the relaxation to 16b indicating that 16b did not
act through the induction of prostaglandin synthesis.
The relaxation induced by 16b is however potentiated by
the selective inhibitor of the cGMP-phosphodiesterase
(type V), zaprinast, at 10 5 M (n=7) but not at 10 6 M
(n=6) (Fig. 7). The potentiation by zaprinast of the
relaxation to 16b (2.6-fold shift to the left of the con-
centration response curve to 16b) is, by far, weaker than
the potentiation by zaprinast of the relaxation to SNP, a
classical activator of the guanylyl cyclase (data not
shown). These data suggest that 16b may weakly pro-
mote the synthesis of cGMP. However, the concentra-
tion±relaxation curve to 16b is not signi®cantly inhibited
by three inhibitors of the soluble guanylyl cyclase,
Compds
Rat trachea
IC50a (95% CI)b
n
I.A.
R.A.
100
16b
17
36.1
(14.2±92.3)
>300
±
0.70‹0.02
16a
11b
12b
13b
14b
15b
18b
7
5
5
4
4
4
4
0.35‹0.04
0
50
0
3.5
43
0
>300
>300
±
0.025‹0.04
0.30‹0.16
0
>300
±
0.033‹0.05
0
5
0
a
6
=valuesÂ10 M.
b
IC50 (concentration inducing 50% of the maximal relaxation
induced by 3 mM aminophylline) with 95% con®dence limits and I.A.
(intrinsic activity) de®ned as the e€ect at 10 4 M, relative to a maximal
relaxation to aminophylline 3 mM. The relative activity (R.A.) is the
ratio between the I.A. of each oximino-caprolactame tested and that
of 1±6b de®ned as 100.
methylene blue (10 4 M, n=12), dipyridamole (10 6 M,
4
n=5) and N-methylhydroxylamine (10
indicating further that the synthesis of cGMP exerts a
M, n=6),
minor role in 16b relaxation pathway. The relaxation to
16b is also weakly but signi®cantly potentiated by roli-
6
pram (10
M, n=16), a selective inhibitor of the
cAMP-phosphodiesterase (type IV) as shown by the 1.5-
fold shift to the left of the concentration-response curve
to 16b (Fig. 7). This borderline e€ect is not due to acti-
vation of b-adrenergic receptors with subsequent synth-
esis of cAMP since the response to 16b was not
the relaxation to 16b was not altered in KCl-enriched
medium (n=4) or in the presence of a non-selective
3
blocker of potassium channels (TEA, 10 M (n=4)
2
6
and 10 M (n=3)) (data not shown). The e€ect of 16b
inhibited by propranolol (10 M, n=7), a potent b1-
on the KCl- and CaCl2-induced contractions on rat
trachea at baseline tone (not contracted with acetyl-
choline) is shown in Figure 5. This compound inhibited
e€ectively and dose-dependently these contractions of
the airway smooth muscle that are directly caused by
the cellular in¯ux of extracellular calcium. 16b, like cal-
cium antagonists,13 displaced concentration±response
curves to the right with depression of the maximal
responses. The comparison of 16b with verapamil (Fig.
6), a classical calcium antagonist acting by blocking the
voltage-sensitive calcium channels, shows that the oxi-
mino-caprolactame compound is 10- to 30-fold less potent
and b2-receptor antagonist. The relaxation to 16b is also
6
5
not potentiated by siguazodan (10 M, n=4; 10 M,
n=6), a selective inhibitor of the type III phosphodies-
terase, which is also a phosphodiesterase important for
the regulation of cAMP breakdown in airway smooth
muscle, notably in murine airways.14 These results,
indicate that, as shown above for cGMP, AMPc do not
clearly participate in the mechanism of the relaxation
induced by 16b. Collectively, these results suggest that
16b may weakly stimulate cGMP and, to a lesser extent,
cAMP syntheses but the increase in cGMP and in cAMP
do not explain the relaxant e€ect of this compound.
Figure 4. Concentration±relaxation curves to 16b (&) and reference relaxant compounds:isoprenaline (!), cromakalim (*), salbutamol (!),
(verapamil (&) and aminophylline (*) in rat trachea (left panel) and human bronchus (right panel) precontracted with acetylcholine. Values are
mean‹S.E.M. of the number of experiments shown in Table 1.

H. El From et al. / Bioorg. Med. Chem. 7 (1999) 1655±1663
1659
Table 3. E€ect of 16b and reference relaxant compounds in rat trachea and human bronchus precontracted with acetylcholine
Compds
Rat trachea
Human bronchus
a
a
n
IC50 (95% CI)b
I.A.
n
IC50 (95% CI)b
I.A.
16b
17
36.1
(14.2±92.3)
ND
0.70‹0.02
7
52.3
(11.5±220)
0.07
(0.03±0.16)
0.47
(0.16±1.10)
0.61‹0.03
0.93‹0.02
0.89‹0.03
Isoprenaline
Salbutamol
Aminophylline
Cromakalim
Verapamil
9
NT
8
0.31‹0.03
Ð
6
10
5
Ð
676
(323±1412)
2.5
(1.5±4.1)
40.7
(13.2±128)
0.15‹0.04
0.67‹0.03
0.68‹0.05
11
9
5
3.8
0.75‹0.06
0.72‹0.08
(0.8±18.2)
10.3
(2.4±42.2)
5
a
6
=valuesÂ10 M; NT=not tested; ND=not determined.
b
IC50 (concentration inducing 50% of the maximal relaxation induced by 3 mM aminophylline) with 95% con®dence limits and I.A. (intrinsic
activity) de®ned as the e€ect at 10 M, or the maximal e€ect, if this occurred at a lower concentration, relative to a maximal relaxation to ami-
nophylline 3 mM.
4
Figure 5. E€ect of 16b in the KCI- and CaCl2-induced contractions in rat trachea at basal tone. Concentration±response curves to KCI or CaCl2
6
5
4
were performed after 20 min incubation with 16b (10 M: *; 10 &; 10 M M, ~) or the solvent (control; *). Values are mean‹of ®ve to 12
experiments. Signi®cant di€erences from control are: *p<0.005; **p<0.01.
Finally, 16b did not act, like theophylline, as a phos-
phodiesterase inhibitor because this compound did not
potentiate the concentration±relaxation curves to acti-
vators of the adenylyl-cyclase, isoprenaline (n=5) and
forskolin (n=6), as well as to the activator of the gua-
nylyl-cyclase, sodium nitroprusside (n=6) (data not
shown).
Experimental
Chemistry
Melting points were determined with a Thomas±
Hoover capillary melting point apparatus and are
uncorrected. Infra-red spectra were obtained on a Pye-
Unicam SP3-100 (Philips) spectrometer. Nuclear mag-
netic resonance spectra were recorded on an AC 200
1
Taken as a whole, the pharmacological data suggest
that 16b acts mainly as a calcium antagonist on airway
smooth muscle. In conclusion, we have observed that 3-
[O-(benzyl-substituted)-oximino-ether]-hexahydroazepin-
2,3-diones may represent a new and valuable class of
calcium antagonists.
FT (Bruker) spectrometer H NMR at 200 MHz and
13C NMR at 50 MHz. Chemical shifts are reported in
parts per million (d) down®eld relative to tetra-
methylsilane. Electron ionization (EI) mass spectra
were obtained on a Nermag R 10-10C spectrometer.
Elemental analyses (C, H, N) were determined by

1660
H. El From et al. / Bioorg. Med. Chem. 7 (1999) 1655±1663
evaporated in vacuo to give a yellow solid of benzhy-
droxamic acid. 80% yield; mp 130±132ꢀC; lit.15=131ꢀC.
To a solution of this acid (25.07 g, 0.183 mol) in ethanol
(28 mL), was added a solution of KOH (10.3 g,
0.184 mol) in water (40 mL) followed by a solution of
alkyl bromide in ethanol. The mixture was heated under
re¯ux for 0.75 h. The resulting solution was cooled
and the precipitate formed was ®ltered. The crystals
were dissolved in EtOH (125 mL). Concentrated HCl
(150 mL) was added to the latter solution, and the mix-
ture was heated under re¯ux for 0.25 h. Evaporation in
vacuo gave a crude residue which was partitioned
between chloroform (200 mL) and water (150 mL). The
aqueous phase was washed several times with CHCl3
and evaporated to give a solid which was recrystallized
from 2 N HCl. Overall yields are low (ca. 30%) because
hydroxamic acid decomposes during the alkylation step
via the Lossen rearrangement to give aniline. Physical
constants are given as an example for compound 7.
Figure 6. Inhibition by verapamil (squares) and 16b (circles) of the
contractile e€ects of KCl at 30 mM (open symbols) and at 60 mM
(closed symbols) in rat trachea at basal tones. Experiments were per-
formed on groups of six to 12 preparations. Points represent mean
with S.E.M. shown by vertical lines.
O-(4-Cyanobenzyl)-hydroxylamine, HCl (7). IR (KBr
1
2%) 3300, 910 cm 1; H NMR (200 MHz, DMSO) d
5.22 (2H, s), 7.45 (2H, m,), 7.64 (2H, m,); 11.2 (2H, s);
13C NMR (50 MHz, DMSO) 75, 118.6, 128.2, 132.2,
142.0. Anal. (C8H9N2OCl) C, H, N.
ꢀ,ꢀ-Dichlorocaprolactam (2). In a 1-L, three-necked
¯ask ®tted with a mechanical stirrer and a re¯ux con-
denser, caprolactame (11.3 g, 0.1 mol) and PCl5 (62.54 g,
0.3 mol) were placed. The mixture was stirred at 90ꢀC
during 0.25 h, cooled to 0 C and hydrolyzed with a con-
centrated aqueous solution of Na2CO3. The solid
formed was ®ltered, and washed successively with H2O,
EtOH, and Et2O. 70% yield; mp 126±128ꢀC [lit.16=
126.5ꢀC]
Morpholino-3-ene caprolactame (3). A mixture of anhy-
drous morpholine (0.23 mol, 20 g) and powdered Drier-
ite was heated under re¯ux in a ¯ask ®tted with a CaCl2
tube. After complete deoxygenation, a,a-dichloro-
caprolactam 2 (2 g, 0.01 mol) was added and the mixture
was heated at 100ꢀC for 6 h. The excess of morpholine
was evaporated in vacuo and the crude residue was
partitioned between CHCl3 and H2O (100 mL each).
The organic layer was dried (MgSO4), ®ltered, and
concentrated in vacuo. Trituration with a minimum of
EtOAc a€orded 3 as a solid. 45% yield; mp 125±127ꢀC
[lit.7=126ꢀC]
Figure 7. In¯uence of rolipram (squares) or zaprinast (circles) on the
concentration±response curves to 16b in rat trachea precontracted
with acetylcholine. Concentration±response curves to 16b were per-
formed after 20 min incubation with saline (control; &, *), rolipram
(n=16, &) or zaprinast 10 M (n=7, *). Values are mean‹S.E.M.
Signi®cant di€erences from control are: *p<0.05; **p<0.01 (student's
test for paired data).
5
CNRS-Vernaison, France, and were within ‹0.4% of
the calculated values.
Hexahydroazepin-2,3-dione (4). To an ice-cooled solu-
tion of enamine 3 (9.8 g, 0.05 mol) in EtOH (50 mL)
was added concentrated HCl (15 mL). After stirring for
0.6 h the ethanol was evaporated in vacuo and CH2
Cl2 (100 mL) was added. The organic phase was
washed with aqueous NaHCO3, two times with water,
dried over MgSO4, and concentrated under reduced
pressure to give 4 as a white solid. 35% yield; mp 67±
69ꢀC.
General procedure for the preparation of compounds
5±10
To a suspension of KOH powder (19.6 g, 0.35 mol) in
anhydrous pyridine (120 mL) at 0ꢀC, hydroxylamine
hydrochloride 13.9 g, 0.2 mol) dissolved in anhydrous
pyridine (100 mL) were added. The ethylbenzoate
(15.0 g, 0.1 mol) was added keeping the temperature
between 0 and 5ꢀC. The resulting mixture was allowed
to stir for 6 h at room temperature. The solid formed
was ®ltered, washed several times with water, with
0.01 N HCl and the residue was dissolved in CH2Cl2.
The organic phase was dried (MgSO4), ®ltered, and
General procedure for the preparation of 3-[O-(aryl)-
oximinoether]-hexahydroazepin-2,3-diones 11±16. The
preparation of 16 is illustrative of the method. To a
solution of hexahydroazepine-2,3-dione
4
(1.01 g,

H. El From et al. / Bioorg. Med. Chem. 7 (1999) 1655±1663
1661
0.008 mol) in absolute EtOH (10 mL) was added 4-
nitrobenzylhydroxamine, HCl 10 (2.5 g, 0.008 mol) and
pyridine (1 mL). The mixture was heated under re¯ux
for 4 h. The solvent was evaporated in vacuo and water
(25 mL) was added. The aqueous solution was extracted
by CH2Cl2 (3Â10 mL). The organic phase was dried
(MgSO4), ®ltered, and evaporated. The mixture of iso-
mers was separated by silica gel column chromatography
(ethylacetate/cyclohexane, 8/2) to a€ord the corre-
sponding Z and E oximino-ethers. Physical constants
are given as examples for compounds 11, 12, and 16.
aminobenzyl)-oximino-ether]-hexahydroazepin-2,3-diones
17 (1 g, 0.004 mol) in CHCl3 (30 mL) were added
CH3SO2Cl (0.92 g, 0.008 mol) and triethylamine (0.19 g,
0.0032 mol). The mixture was heated under re¯ux, dur-
ing 30 min, in a ¯ask ®tted with a CaCl2 tube, then
poured in 100 mL of ice-water. The organic phase was
dried (MgSO4), ®ltered, and evaporated in vacuo. The
residue was puri®ed by silical gel column chromato-
graphy and eluted with ethylacetate to give compound
1
18 as an oil. IR: (KBr 2%) 3250, 1307 cm 1; H NMR
(200 MHz, CDCl3) 18a d 1.72 (4H, m), 2.40 (2H, dd),
2.80 (3H, s), 3.16 (2H, td), 5.20 (2H, s), 6.76 (1H, t),
7.52 (2H, m), 8.00 (2H, m), 8.62 (1H, s); 18b d 1.74 (4H,
m), 2.62 (2H), 2.81 (3H, s), 3.20 (2H, td), 5.25 (2H, s),
6.90 (lH, t), 7.60 (2H, m), 8.20 (2H, m), 8.61 (1H, s); 13C
NMR (50 MHz, CDCl3) d 18a 28.5, 29.4, 29.4, 42.0,
44.0, 75.0, 118.6, 122.5, 128.5, 139.4, 156.2, 168.6; 18b
21.5, 24.7 27.1, 40.1 44.4, 75.6, 118.5, 122.3, 128.5,
139.3, 156.4, 168.3; Anal. (C14H19N3O4S) C, H, N.
3-[O-Methyl oximino-ether]-hexahydroazepin-2,3-diones
1
(11). IR (KBr 2%) 3200, 1700, 1610 cm 1; H NMR
(200 MHz, CDCl3) 11a d 1.74 (4H, m), 2.40 (2H, dd),
3.17 (2H, td), 3.87 (3H, s), 7.26 (1H, t); 11b d 1.77 (4H,
m), 2.51 (2H, dd), 3.24 (2H, td), 3.97 (3H, s), 7.49 (1H,
t); 13C NMR (50 MHz, CDCl3) d 11a 28.6, 29.40, 42.0,
62.2, 156.40 168.9. 11b 21.4, 24.4, 27.1, 40.0, 62.6, 156.6,
169.0. Anal. (C7H12N2O2) C, H, N; MS: (EI,70 eV, I%),
m/e: M+ 156 (79.6%); 87.0 (100%).
X-ray analysis for 16a and 16b
3-[O-Benzyl-oximino-ether]-hexahydroazepin-2,3-diones
1
A syn form isomer, 16a, and an anti form one 16b, were
selected for X-ray crystallographic studies. Crystal data
are listed in Table 4.
(12). IR: (KBr 2%) 3210, 1710, 1600 cm 1; H NMR
(200 MHz, CDCl3) 12a d 1.67 (4H, m), 2.41 (2H, dd),
3.10 (2H, td), 5.15 (2H, s), 7.30 (lH, t), 7.40 (5H, m); 12b
d 1.72 (4H, m), 2.58 (2H, dd), 3.23 (2H, td), 7.35 (2H, s),
7.38 (lH, t); 13C NMR (50 MHz, CDCl3) d 12a 28.0,
29.5, 42.8, 76.0, 127.7, 128.2, 128.3, 138.2, 156.9, 168.8;
12b 21.5, 24.7, 27.1, 24.7, 40.1, 76.9, 127.9, 128.1, 137.0,
157.0, 168.8. Anal. (C13H16N2O2) C, H, N.; MS: (EI,70
eV, I%), m/e: M+ 232 (68,15%); 215 (100%).
16b: The single crystals for X-ray analysis recrystallized
from ethylacetate. A computer-controlled Enraf±Non-
ius CAD-4 di€ractometer with CuK radiation and an
incident beam graphite monochromator were used for
X-ray data collection. Structure determination was
obtained by direct methods using SHELX8617 and
SHELX93.18 All the non-hydrogen, 20 atoms, were
observed in an E map. 16a: The single crystals for
X-ray analysis recrystallized from ethylacetate. Due
to the very small size of a single crystal of Z isomer,
0.14Â0.02Â0.01 mm, the crystal structure solution of
this isomer was less straightforward. The data collection
of this sample was performed by using the X-ray syn-
chrotron radiation; the wavelength is 0.619 A. In addi-
tion to its small size and the instability of the crystal
under X-ray radiation only 739 re¯ections could be
measured. The structure of Z isomer was solved by
direct method as above and di€erence Fourier synth-
eses. Atoms distances to plane P1 and P2:
3-[O-(4-Nitrobenzyl-oximino-ether]-hexahydroazepin-2,3-
1
diones (16). IR: (KBr 2%) 3210, 1690, 1620, 920 cm
;
1H NMR (200 MHz, CDCl3) 16a d 1.72 (4H, m), 2.40
(2H, dd), 3.18 (2H, td), 5.22 (2H, s), 6.78 (lH, t), 7.50
(2H, m), 8.19 (2H, m); 16b d 1.75 (4H, m), 2.60 (2H,
dd), 3.22 (2H, td), 5.31 (2H, s), 7.39 (lH, t), 7.50 (2H,
m), 8.18 (2H, m); 13C NMR (50 MHz, CDCl3) d 16a
28.6, 29.4, 42.1, 74.5, 123.5, 127.7, 145.5, 147.5, 156.2,
168.7; 16b 21.6, 24.8, 27.1, 40.1, 75.5, 123.6, 144.7,
147.5, 158.1, 168.3. Anal. (C13H15N3O4) C, H, N.
3-[O-(4-Aminobenzyl)-oximino-ether]-hexahydroazepin-
2,3-diones (17). To a solution of 16 (1 g, 0.0036 mol) in
absolute EtOH (20 mL) was added 0.1 g of 5% Pd on
charcoal. Reduction of the nitro-derivative, using Pd/H2
during 12 h under atmospheric pressure, gave after ®l-
tration on Celite and evaporation of EtOH in vacuo,
E
Z
to P1
to P2
C9
C11
C12
C12
C9
0.214 (16) A
0.300 (25) A
0.160 (23) A
1.268 (11) A
1.153 (12) A
1.351 (11) A
0.231 (18) A
0.369 (19) A
0.055 (18) A
1.202 (09) A
1.144 (12) A
1.362 (16) A
Ê
Ê
Ê
Ê
compound 17 as a white solid. IR: (KBr 2%) 3300,
1
3225, 1700 1620, 910 cm
;
1H NMR (200 MHz,
CDCl3) 17a d 1,70 (4H, m), 2.12 (2H, s), 2.37 (2H, dd),
3.20 (2H, td), 5.01 (2H, s), 6.50 (1H, t), 6.78 (2H, m),
7.15 (2H, m); 17b d 1.73 (4H, m), 2.10 (2H, s), 2.55 (2H,
dd), 3.23 (2H, td), 5.11 (2H, s), 6.50 (1H, t), 6.78 (2H,
m), 7.19 (2H, m); 13C NMR (50 MHz, CDCl3) d 17a
28.4, 29.3, 29.3, 41.8, 74.9, 115.5, 119.5, 129.0, 156.2,
168.3; 17b 21.9, 24.5, 27.5, 40.0, 75.3, 115.2, 119.5,
129.1, 156.2, 168.0. Anal. (C13H17N3O2) C, H, N.
N3
Pharmacology
Rat isolated trachea. Male rats (250±350 g) were killed
by intraperitoneal injection of 1 mL pentobarbital (6%).
The tracheas were excised from the animals and placed
in Krebs solution (composition in mM: NaCl 118, KCl
5.4, CaCl2 2.5, KH2PO4 0.6, NaHCO3 25 and glucose
3-[O-(4-Methylsulfonylaminobenzyl)-oximino-ether]-hexa-
hydro-azepin-2,3-diones (18). To a solution of 3-[O-(4-

1662
H. El From et al. / Bioorg. Med. Chem. 7 (1999) 1655±1663
Table 4. Crystal data for 16b (E) and 16a (Z): 16 is the 3-[O-(4-nitrobenzyl-oximino-ether]-hexahydroazepin-2,3-diones
16b (E)
16a (Z)
Formula
C
13H15N3O4
277.28
293(2)
C13H15N3O4
277.28
293(2)
0.619
Monoclinic
P21/c
Formula weight
Temperature (K)
Wave length (A)
Crystal system
Space group
a (A)
1.5418
Monoclinic
C2/c
21.141(5)
9.610(3)
13.980(4)
107.33(8)
2711.3(13)
8
6.225(5)
17.885(5)
12.157(5)
92.35(5)
1352.4(13)
4
b (A)
c (A)
b (ꢀ)
V (A3)
Z
Density, calcd (g cm
3
3
3
)
1.359 g.cm
Needle (colourless)
0.860
1168
0.18Â0.08Â0.06
1.362 g.cm
Crystal habit
1
Needle (colourless)
0.103
584
(m mm
F(000)
)
Crystal dimensions (mm)
q range for data collection (ꢀ)
Index range
0.14Â0.02Â0.01
17<q<14.1
3ꢁhꢁ4
10 to 24
24ꢁhꢁ22
0ꢁkꢁ11
9ꢁkꢁ10
0ꢁ1ꢁ16
5ꢁ1ꢁ6
Re¯ections collected
4443
962(0.04)
739
343(0.10)
Independant re¯ections (Rint
Re®nement method
Data/parameters
Goodness-of-®t
R-factor [F2>2s(F2)]
)
Full matrix 1-s on F2
962/182
0.44
0.046 (603 re¯ections)
0.056 (all re¯ections)
Constrained 1-s on F2
343/184
0.63
0.035 (343 re¯ections)
0.0492 (all re¯ections)
wR2 [1<2s(1)] (all intensities)
x,y,w=1/[s2(F2o)+(xP)2+yP]
where p=(F2o)+2Fc2)/3
Extinction coecient
Largest di€. peak and hole (eA
x=0.1706, y=1.9204
None
x=0.0625, y=1.2693
None
0.109 and 0.091
3
)
11.7). Following removal of adhering fat and connective
tissues, the tracheas were cut into four rings. The rings
were suspended in 5 mL organ baths containing Krebs
solution at 37ꢀC, gassed with 95% O2 and 5% CO2 and
equilibrated under an initial tension of 1.5 g. After
equilibration for 1 h, the resting tension was between 0.6
and 1.4 g. Tension was measured isometrically with
strain gauges (UF-1 Pioden) and ampli®ers (EMKA,
France) and displayed on recorders (Sefram, France).
maximal contraction and relaxation was performed to
stabilize the preparations. During the following hour,
the tissues were washed every 15 min. Concentration±
relaxation curves to oximino-caprolactames and refer-
ences compounds (aminophylline, isoprenaline, croma-
kalim, salbutamol, verapamil) were obtained in
preparations precontracted with Ach. After a stable
level of precontraction, in response to a concentration
of Ach giving 50% of the maximal response, the com-
pounds were added to the bath in cumulative con-
centrations. Afterwards, aminophylline 3 mM was
added to the bath to obtain the maximal relaxation.
Only one concentration±response curve of each relaxant
agent was obtained from each ring. Experiments were
conducted on parallel groups of four to eight rings, with
at least one ring serving as control.
Human isolated bronchus. Human bronchial tissues were
removed from patients undergoing surgery for lung
cancer. Just after resection, segments of bronchi with an
inner diameter of 4±6 mm were taken from an area as
far as possible from the malignancy. They were carefully
dissected free of the parenchyma from distal lobar to
segmental airways and transported to the laboratory in
Krebs solution previously aerated with the gas mixture
of 95% O2 and 5% CO2. The tissue was stored at 4ꢀC
until the experiment was carried out (2 to 24 h max-
imum), and two to eight rings of the same bronchus
were prepared. Each bronchial ring was suspended in
Krebs solution under an initial tension of 2 g and trea-
ted in every respect as the rat tracheal rings.
The mechanism of oximino-caprolactame (16b)-induced
relaxation was investigated on the rat trachea prepara-
tion by studying the e€ects of selective modulators of
the di€erent airway smooth muscle pathways of relaxa-
tion. The modulators were added to the Krebs solution
20 min before contraction with acetylcholine, i.e. at
least 30 min before concentration response curve to
16b. In order to test the potential action of 16b as a
calcium antagonist, we have also studied the inhibitory
e€ect of this compound on the contraction induced by
KCl, a depolarizing agent known to contract the airway
smooth muscle by promoting the in¯ux of extracellular
Protocols. Each experiment was initiated by inducing a
maximal contraction of the tracheal or bronchial rings
with acetylcholine (Ach 3 mM) and then a maximal
relaxation with aminophylline 3 mM. This initial

H. El From et al. / Bioorg. Med. Chem. 7 (1999) 1655±1663
1663
calcium, as well as on the contraction directly caused by
extracellular calcium (CaCl2) added to the bath in a
depolarizing medium, according to Advenier et al.13
Acknowledgements
The authors thank F. Thomasson for recording NMR
spectra, and F. Caron for helpful technical assistance in
pharmacology.
Expression of results and statistics. The relaxing e€ects
of compounds are expressed as a percentage of the
maximal relaxation produced by aminophylline (3 mM).
IC50 was de®ned as the concentration of the agent that
caused 50% of maximal relaxation to aminophylline 3
mM. For the studies on the inhibitory activity of 16b on
either KCl- or calcium-induced contraction, results are
expressed as percentages of the maximal contraction
induced either by KCl or by Ach (3 mM). Results are
given as mean‹SEM. Statistical analysis was per-
formed using the Student's t test. p<0.05 values were
considered as signi®cant.
References
1. Khelili, S.; Leclerc, G.; Faury, G.; Verdetti Bioorg. Med.
Chem. 1995, 3, 495.
2. Gilly, C.; Taillandier, G.; Pera, M. H.; Luu-Duc, C.; Rous-
seau, A.; Narcisse, G. Eur. J. Med. Chem. 1993, 28, 905.
3. Duong, T.; Prager, R. H.; Ward, A. D.; Kerr, D. I. B. Aust.
J. Chem. 1976, 29, 2651 .
4. Brouillette, W. J., Einspahr, H. M. J. Org. Chem. 1984, 5113.
5. Wineman, R. J.; Hsu, E. T.; Anagnostopoulos, C. E. J. Am.
Chem. Soc. 1958, 80, 6233.
Drugs. The drugs used and their sources were: iso-
prenaline, aminophylline (PCH, Paris, France), salbu-
tamol sulfate (Glaxo, Paris, France), acetylcholine,
cromakalim, tetraethylammonium chloride (TEA),
sodium nitroprusside (SNP), glibenclamide, methylene
blue, dipyridamole, indomethacin, N-methylhydroxyl-
amine, forskolin (Sigma, St. Louis, USA). The selective
phosphodiesterase (PDE) inhibitors: siguazodan (PDE
III inhibitor), rolipram (PDE IV inhibitor) and zapri-
nast (PDE V inhibitor) were generous gifts from Pro-
fessor Advenier (Paris, France). All stock solutions of
the chemicals were made in distilled water, excepting
glibenclamide in methanol, indomethacin and rolipram
in ethanol, siguazodan and forskolin in dimethyl sulf-
oxide (Sigma, St. Louis, USA) and zaprinast in ammo-
nia (1.6% v/v). All the stock solutions were kept frozen
until use and then diluted in distilled water. The oxi-
mino-caprolactame derivatives were all prepared in
dimethyl sulfoxide (DMSO) (Sigma). At no time did the
maximal concentrations of solvents exceed 0.3% by
volume in the organ bath and produced by themselves
any signi®cant direct e€ect on the basal tone of the rat
trachea or on responses to reference compounds.
6. Watthey, J. W. H.; Stanton, J. L.; Desai, M.; Babiarz, J. E.;
Finn, B. M. J. Med. Chem. 1985, 28, 1511.
7. Kerfanto, M.; Soyer, N. Bull. Soc. Chim. Fr. 1966, 9, 2966.
8. Glushkov, R. G.; Zososova, I. M.; Ovcharova, I. M. Khim.
Geterotsikl. 1979, 7, 954.
9. Garoufalias, S.; Vyzas, A.; Fytas, G.; Foscolos, G. B.;
Chytiroglou, A. Ann. Pharm. Fr. 1988, 46, 97.
10. Karabatsos, G. J.; Taller, R. A. Tetrahedron 1968, 24, 3347.
11. Cozzi, P.; Giordani, A.; Menichincheri, M. J. Med. Chem.
1994, 37, 3588.
12. Anderson, G. P. In Airways Smooth Muscle; Raeburn, D.,
Giembycz, M. A., Eds.; Birkauser Verlag: Basel, 1995; Chap-
ter 1, pp 1±79.
13. Advenier, C.; Cerrina, J.; Duroux, P.; Floch, A.; Renier,
A. Br. J. Pharmacol. 1984, 82, 727.
14. Held, H. D.; Wendel, A.; Uhlig, S. Biochem. Biophys. Res.
Commun. 1997, 231, 22.
15. Bachman, G. B.; Goldmacher, J. E. J. Org. Chem. 1964,
29, 2576.
16. Brenner, M.; Rickendacher, H. R. Helv. Chim. Act. 1958,
21, 181.
17. Sheldrick, G. M. Acta Cryst. 1990, A46, 467.
18. Sheldrick, G. M. Program for the Re®nements of Crystal
Structures; University of Gottingen: Germany, 1993.
19. Brady, O. L.; Klein, L. J. Chem. Soc. 1927, 874.