1,4-Diazepanes derived from (S)-serine - Homopiperazines with improved σ1 (sigma) receptor affinity and selectivity

Article abstract of DOI:10.1016/j.ejmech.2008.03.033

Starting from the proteinogenic amino acid (S)-serine chiral non-racemic 1,4-diazepanes 4 with a hydroxymethyl residue in position 2 are synthesized and pharmacologically evaluated. The key step in the synthesis is the formation of the bicyclic system 8 b

Full text of DOI:10.1016/j.ejmech.2008.03.033

Available online at www.sciencedirect.com
European Journal of Medicinal Chemistry 44 (2009) 519e525
http://www.elsevier.com/locate/ejmech
Original article
1,4-Diazepanes derived from (S )-serine e Homopiperazines
with improved s1 (sigma) receptor affinity and selectivity
*
Stephan Bedurftig, Bernhard Wunsch
¨
¨
Institut fu¨r Pharmazeutische und Medizinische Chemie, Westfa¨lische Wilhelms-Iniversita¨t Mu¨nster, Hittorfstraße 58-62, 48149 Mu¨nster, Germany
Received 21 January 2008; received in revised form 27 March 2008; accepted 27 March 2008
Available online 4 April 2008
Abstract
Starting from the proteinogenic amino acid (S )-serine chiral non-racemic 1,4-diazepanes 4 with a hydroxymethyl residue in position 2 are
synthesized and pharmacologically evaluated. The key step in the synthesis is the formation of the bicyclic system 8 by consecutive nucleophilic
substitution of the chloropropionamide 7 with primary amines and intramolecular aminolysis. Both reaction steps require catalysis with the
Lewis acid Ti(O-iPr)4. Homologation of the piperazine to the 1,4-diazepane ring results in a remarkable improvement of s1 receptor affinity
and s1/s2 selectivity. The 1,4-dibenzyl derivative 4a interacts with a Ki value of 7.4 nM with s1 receptors and shows a 53-fold selectivity
for s1 receptors over s2 receptors.
Ó 2008 Elsevier Masson SAS. All rights reserved.
Keywords: s Receptor ligands; 1,4-Diazepanes; Homopiperazines
R
1. Introduction
CH3
The class of s receptors is subdivided into at least two sub-
types, which are termed s1 and s2 receptor. Whereas the s1 re-
ceptor has been cloned from various tissues and species [1e4]
the s2 receptor subtype is not cloned so far. Both receptor sub-
types are found in the central nervous system as well as in tis-
sues of the periphery (e.g. liver, lung, heart). The exact
understanding of the physiological role of both s receptor
subtypes still remains to be elucidated [5,6]. However, it is
well established that s receptors are involved in several physi-
ological and pathophysiological processes. Therefore, ligands
interacting with s receptors are of particular interest for the
development of novel antipsychotics [7,8], antidepressants
[9], anti-cocaine agents [10e12], and antitumor agents [6,13,14].
In literature some s1 receptor ligands based on the pipera-
zine heterocycle are described. These s1 receptor ligands
include the very potent 1,4-disubstituted piperazines 1 [15]
N
N
N
N
Cl
Cl
1
2
R
N
R
N
H
H
N
N
OH
OH
3
4
* Corresponding author. Tel.: þ49 251 8333311; fax: þ49 251 32144.
Fig. 1. Comparison of hydroxymethyl substituted 1,4-diazepanes (homopiper-
azines) 4 with known piperazine based s1 receptor ligands 1e3.
E-mail address: wuensch@uni-muenster.de (B. Wunsch).
¨
0223-5234/$ - see front matter Ó 2008 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2008.03.033

520
S. Bedu¨rftig, B. Wu¨nsch / European Journal of Medicinal Chemistry 44 (2009) 519e525
and 2 [10] (compare Fig. 1). Recently we have shown that pi-
perazines 3 with a hydroxymethyl residue in position 2 also in-
teract with s1 receptors. High s1 receptor affinity is achieved
for derivatives with a second aromatic system within the resi-
due R in position 4. Thus, the dibenzyl derivative 3a
(R ¼ CH2Ph, Ki ¼ 38 nM) represents one of the most potent
compounds of this series [16].
In order to broaden the relationships between the structure
of piperazine derivatives and their s1 receptor affinity homol-
ogous ligands were envisaged. Herein we wish to report on the
synthesis of 2-hydroxymethyl substituted 1,4-diazepanes (ho-
mopiperazines) 4, which represent ring homologs of pipera-
zines 3. The ring enlargement results in an increased s1
receptor affinity and, moreover, a considerable improvement
of selectivity against the s2 receptor.
3. Receptor binding studies
The s receptor affinities of the 1,4-diazepanes 4 were deter-
mined in competition experiments with radioligands. In the s1
assay, homogenates of guinea pig brains served as receptor
material. The s1 selective ligand [3H]-(þ)-pentazocine was
employed as radioligand, and the non-specific binding was de-
termined in the presence of a large excess of haloperidol
[18,19]. Rat liver was the source for s2 receptors in the s2 as-
say. Since a s2 selective radioligand is not commercially avail-
able, the non-selective radioligand [3H]-ditolylguanidine was
used in the presence of an excess of non-radiolabeled (þ)-pen-
tazocine (100 nM) for selective labeling of s1 receptors.
Performing the s2 assay in the presence of an excess of
non-tritiated 1,3-di(o-tolyl)guanidine led to the non-specific
binding of the radioligand [18,19].
In addition to the s1 and s2 receptor affinities the interac-
tions with m-opioid and k-opioid receptors as well as with the
phencyclidine binding site of the NMDA receptor were deter-
mined [20]. At first a screening with two rather high concen-
trations (1 and 10 mM) of the test compounds was performed.
Only when considerable inhibition of the radioligand binding
was observed at a concentration of 10 mM the exact Ki values
were determined.
2. Chemistry
The synthesis of the hydroxymethyl substituted 1,4-diaze-
panes 4 started with the methyl ester 5 of the proteinogenic
amino acid (S )-serine [16]. At first the primary amine of 5
was acylated with chloropropionyl chloride to obtain the chlo-
ropropionamide 6 in 74% yield. Reaction of the hydroxyamide
6 with benzaldehyde dimethyl acetal led to an inseparable mix-
ture of diastereomeric oxazolidines cis-7 and trans-7 (ratio
30:70) (Scheme 1).
4. Results and discussion
In contrast to our experience with the corresponding chlo-
roacetamides [16] the chloropropionamides 7 did not react
with benzylamine upon simple heating to reflux. Only when
the Lewis acid Ti(O-iPr)4 was added the conversion of the
chloropropionamide 7 into the bicyclic system 8a took place
(Scheme 1). Also, the synthesis of the (dimethoxyethyl)
substituted derivatives 8b by reaction of 7 with aminoacetalde-
hyde dimethyl acetal was only possible in the presence of the
Lewis acid Ti(O-iPr)4. We assume that the Lewis acid acti-
vates the chloropropionamide substructure for the nucleophilic
substitution as well as the ester moiety for the intramolecular
aminolysis. The diastereomeric bicyclic compounds cis-8a,b
and trans-8a,b were separated by flash column chromatogra-
phy and, subsequently, the relative configuration of the prod-
ucts was assigned by nuclear Overhauser effect.
It should be noted that the direct intramolecular aminolysis
of linear dipeptides made up of an a and a b-amino acid to
form 1,4-diazepanediones is very difficult [17]. However, the
preformation of the oxazolidine ring in 7 reduces the confor-
mational flexibility and induces a favorable orientation of
the crucial functional groups for the intramolecular
aminolysis.
In Table 1 the s receptor affinities of the 1,4-diazepaes 4a
and 4b are compared with the s receptor affinities of the cor-
responding piperazine derivatives 3a and 3b and reference
compounds. The Ki value of the 1,4-dibenzyl-1,4-diazepane
4a towards the s1 receptor is 7.4 nM indicating a 5-fold
increase of its s1 receptor affinity compared with the 1,4-
dibenzylpiperazine 3a (Ki ¼ 38 nM). Additionally the s2
receptor affinity of the 1,4-diazepane 4a is 2-fold lower
than the s2 receptor affinity of the piperazine 3a. Taking
the increased s1 and decreased s2 receptor affinity of the
1,4-diazepane 4a together results in a considerable improve-
ment of the s1/s2 selectivity (selectivity factor 53 instead
of 5).
This observation was confirmed by synthesis and pharma-
cological evaluation of the homologous 1,4-diazepane 4b of
the (dimethoxyethyl) substituted piperazine 3b, which showed
negligible affinity to both s receptor subtypes (compare Table
1, entry 2). For the homologous 1,4-diazepane 4b a s1 recep-
tor affinity of 320 nM was determined, which demonstrates
again a considerable affinity increase compared with the cor-
responding piperazine derivative 3b (Ki > 10,000 nM). The
s2 receptor affinity of 4b (Ki ¼ 9010 nM) is in the same range
as the s2 receptor affinity of the piperazine 3b
(Ki > 10,000 nM). These data for 4b support the idea that
ring enlargement of the piperazine ring to a 1,4-diazepane
ring leads to increase of the s1 receptor affinity and s1/s2
selectivity.
In the last step the bicyclic systems 8a and 8b were reduced
with LiAlH4. During this reduction both lactam carbonyl moi-
eties as well as the oxazolidine ring were reduced to give
directly the hydroxymethyl substituted 1,4-diazepanes 4a
and 4b, respectively. Since the chiral center in position 10 of
the bicyclic systems 8a and 8b was destroyed in this reaction
step, a mixture of diastereomers cis-8 and trans-8 was
employed for the LiAlH4 reduction, respectively.
In order to get a rough impression of the selectivity of the
1,4-diazepanes 4 the affinity to m-opioid and k-opioid recep-
tors as well as the phencyclidine binding site of the NMDA

S. Bedu¨rftig, B. Wu¨nsch / European Journal of Medicinal Chemistry 44 (2009) 519e525
521
H3CO
N
O
H
H3CO
Cl
H3CO
H2N
O
H
O
H
Cl
(c) or (d)
(b)
(a)
O
HCl
O
N
H
OH
O
HO
.
5 HCl
6
7
R
R
4N
R
N
O
N
O
H
H
(e)
2
H
7
9
N
N
1
8
+
O
O
N
OH
O
O
H
10
H
cis-8a,b
trans-8a,b
4a,b
4,8
a
R
CH2C6H5
CH2CH(OCH3)2
b
Scheme 1. (a) ClꢀCH2CH2COCl, NEt3, CH2Cl2, 1 h, ꢀ5 ꢁC, 1 h, rt, 74%. (b) PhCH(OCH3)2, TosOH, toluene, 2 h, reflux, 74%. (c) BnNH2, NEt3, CH3CN, 24 h,
reflux, then Ti(O-iPr)4, 24 h, reflux, 25% (cis-8a), 67% (trans-8a). (d) H2NCH2CH(OCH3)2, NEt3, CH3CN, 24 h, reflux, then Ti(O-iPr)4, 96 h, reflux, 8% (cis-8b),
34% (trans-8b). (e) LiAlH4, THF, 96 h, reflux, 46% (4a), 40% (4b).
receptor was investigated. In all experiments with concentra-
tions of 10 mM of the test compounds 4a and 4b the inhibition
of the radioligand binding was lower than 50% indicating IC50
values greater than 10 mM. Thus, both 1,4-diazeoanes 4a and
4b interact selectively with s1 receptors.
diazepanes 4 is presented. When compared with the pipera-
zines 3 the homologous 1,4-diazepanes 4 show considerable
increase of s1 receptor affinity and s1/s2 selectivity.
6. Experimental
5. Conclusion
6.1. Chemistry, general
In this communication a method for the preparation of
chiral non-racemic 2-hydroxymethyl substituted 1,4-
Unless otherwise noted, moisture sensitive reactions were
conducted under dry nitrogen. Thin layer chromatography
Table 1
s Receptor affinities of hydroxymethyl substituted 1,4-diazepanes 4 compared with the corresponding piperazines 3 and reference compounds
R
N
R
N
H
H
N
N
OH
OH
3
4
Entry
Compd
R
Ki ꢂ SEM (3) (nM)
s1 ([3H]-(þ)-pentazocine)
s2 ([3H]-ditolylguanidine)
s1/s2 Selectivity
1
2
3
4
5
6
7
3a [16]
3b [16]
CH2Ph
38 ꢂ 2.4
>10,000
7.4 ꢂ 0.9
320 ꢂ 18
2.2 ꢂ 0.3
164 ꢂ 47
265 ꢂ 32
177 ꢂ 16
>10,000
374 ꢂ 59
9010 ꢂ 641
34 ꢂ 2.3
64 ꢂ 11
5
CH2CH(OCH3)2
CH2Ph
CH2CH(OCH3)2
e
53
4a
4b
28
16
Haloperidol
Ditolylguanidine
BMY 14802
0.39
1.5
391 ꢂ 62

522
S. Bedu¨rftig, B. Wu¨nsch / European Journal of Medicinal Chemistry 44 (2009) 519e525
(tlc): Silica gel 60 F254 plates (Merck). Flash chromatography
(fc): Silica gel 60, 0.040e0.063 mm (Merck); parentheses
include: diameter of the column [cm], eluent, fraction size
[mL], Rf e optical rotation: polarimeter 241 (PerkineElmer);
1.0 dm tube; concentration c [g/100 mL]; temperature 20 ꢁC.
Elemental analyses: Vario EL (Elementaranalysesysteme
GmbH). MS: MAT 312, MAT 8200, MAT 44, and TSQ
7000 (Finnigan); EI ¼ electron impact, CI ¼ chemical ioniza-
tion. High resolution MS (HRMS): MAT 8200 (Finnigan).
IR: IR spectrophotometer 1605 FT-IR (PerkineElmer).
1.046 g, 74%) as a pale yellow, viscous oil, ratio cis-7:trans-
7 ¼ 30:70. C14H16ClNO4 (297.7) Calcd. C 56.97, H 5.41, N
4.70, Found C 57.68, H 5.67, N 4.22. MS (EI): m/z
(%) ¼ 299/297 (M, 1/3), 240/238 (M ꢀ CO2CH3, 3/8), 206
(M ꢀ ClC2H4CO, 26), 148 (2-phenyl-1,3-oxazolidine, 100),
105 (PhCHO, 67). MS (CI): m/z (%) ¼ 300/298 (MHþ, 20/
59). IR (film): n (cmꢀ1) ¼ 2953 (m, CeH), 1749 (s, C]O
ester), 1663 (s, C]O amide), 1431 (s, CeH). 1H NMR
(CDCl3): d ¼ 2.31 (dt, J ¼ 16.5/6.6 Hz, 0.7H, ClCH2CH2CO,
trans), 2.45e2.54 (m, 0.3H, ClCH2CH2CO, cis), 2.59 (dt,
16.6/6.8 Hz, 0.7H, ClCH2CH2CO, trans), 2.72e2.82 (m,
0.3H, ClCH2CH2CO, cis), 3.54e3.79 (m, 2H, ClCH2CH2CO),
3.81 (s, 3 ꢃ 0.23H, OCH3, cis, rotamer 1), 3.83 (s, 3 ꢃ 0.70H,
OCH3, trans), 3.86 (s, 3 ꢃ 0.07H, OCH3, cis, rotamer 2),
4.07e4.14 (m, 0.23H, 5-HA, cis, rotamer 1), 4.12 (dd,
J ¼ 14.3/7.1 Hz, 0.17H, 5-HA, trans, rotamer 2), 4.19e4.29
(m, 0.07H, 5-HA, cis, rotamer 2, and 0.53H, 5-HA, trans,
rotamer 1, and 0.3H, 5-HB, cis, rotamer 1 and 2, and 0.53H,
5-HB, trans rotamer1), 4.44e4.47 (m, 0.17H, 5-HB, trans, ro-
tamer 2), 4.68e4.72 (m, 0.3H, 4-H, cis), 4.81e4.89 (m, 0.7H,
4-H, trans), 6.11 (s, 0.53H, 2-H, trans, rotamer 1), 6.27 (s,
0.17H, 2-H, trans, rotamer 2), 6.32 (s, 0.23H, 2-H, cis, rotamer
1), 6.50 (s, 0.07H, 2-H, cis, rotamer 2), 7.35e7.53 (m, 4H,
arom), 7.71e7.74 (m, 1H, arom).
1
(br ¼ broad, m ¼ medium, s ¼ strong). H NMR (300 MHz),
13C NMR (75 MHz): Unity 300 FT NMR spectrometer (Var-
ian), d in ppm related to tetramethylsilane, coupling constants
are given with 0.5 Hz resolution; the assignments of 13C and
1H NMR signals were supported by 2D NMR techniques.
6.2. (þ)-Methyl (2S )-2-(3-chloropropanoylamino)-
3-hydroxypropanoate (6)
A suspension of (S )-serine methyl ester hydrochloride [16]
(5$HCl, 10.0 g, 64.3 mmol) in CH2Cl2 (100 mL) was cooled
(ꢀ5 ꢁC). Triethylamine (17.9 mL, 128.6 mmol) and subse-
quently a solution of 3-chloropropanoyl chloride (5.9 mL,
61.1 mmol) in CH2Cl2 (15 mL) were added dropwise. The
reaction mixture was stirred for 1 h at ꢀ5 ꢁC and for 1 h at
rt. Then Et3N$HCl was filtered off, the solvent was evaporated
in vacuo and the residue was purified by fc (2 ꢃ 8 cm, ethyl
acetate, 100 mL, Rf ¼ 0.52) to yield 6 (9.95 g, 74%) as
6.4. (ꢀ)-(7S,10R)-5-Benzyl-10-phenyl-9-oxa-1,
5-diazabicyclo[5.3.0]decane-2,6-dione (cis-8a) and
(þ)-(7S,10S )-5-benzyl-10-phenyl-9-oxa-1,5-
diazabicyclo[5.3.0]decane-2,6-dione (trans-8a)
a
pale yellow, viscous oil. [a]589 ¼ þ21.7 (c ¼ 0.85,
CH2Cl2). C7H12ClNO4 (209.6) Calcd. C 40.11, H 5.77, N
6.68, Found C 40.04, H 5.97, N 6.45. MS (EI): m/z
(%) ¼ 152/150 (M ꢀ CO2CH3, 12/39), 118 (M ꢀ ClC2H4CO,
2). MS (CI): m/z (%) ¼ 229/227 (M þ NHþ4 , 33/100), 212/
210 (MHþ, 16/47). IR (film): n (cmꢀ1) ¼ 3312 (br, OH),
2956 (m, CeH), 1737 (s, C]O ester), 1649 (s, C]O amide),
1535 (s, CeN). 1H NMR (CDCl3): d ¼ 2.56 (t, J ¼ 6.0 Hz, 1H,
A
solution of
7
(cis-7:trans-7 ¼ 30:70, 0.278 g,
0.93 mmol), benzylamine (0.10 mL, 0.93 mmol) and triethyl-
amine (0.13 mL, 0.93 mmol) in acetonitrile (35 mL) was
heated to reflux for 24 h. Then titanium(IV) tetraisopropylate
(0.41 mL, 1.4 mmol) was added and the mixture was heated
to reflux for additional 24 h. Removal of the solvent in vacuo
followed by fc (3 cm, ethyl acetate, 8 mL) gave cis-8a
(Rf ¼ 0.65) and trans-8a (Rf ¼ 0.39).
OH ), 2.71e2.77 (m, 2H,
ClCH2CH2CO
and/or
ClCH2CH2CO), 3.79e3.85 (m, 2H, ClCH2CH2CO and/or
ClCH2CH2CO), 3.81 (s, 3H, CO2CH3), 3.91e4.05 (m, 2H,
3-H), 4.71 (dt, J ¼ 7.3/3.6 Hz, 1H, 2-H), 6.64 (s broad, 1H,
NH ).
cis-8a. Colorless viscous oil, yield 79 mg (25%).
[a]589 ¼ ꢀ32.6 (c ¼ 0.50, CH2Cl2). C20H20N2O3 (336.4).
HRMS: Calcd. 336.1474, Found 336.1472 (ꢀ0.2 ppm). MS
(EI): m/z (%) ¼ 336 (M, 24), 259 (M ꢀ phenyl, 3), 245
(M ꢀ benzyl, 21), 105 (PhCHO, 30), 91 (benzyl, 100). IR
(film): n (cmꢀ1) ¼ 2927 (m, CeH), 1642 (s, C]O), 1430
6.3. Methyl (2R,4S )- and (2S,4S )-3-
(3-chloropropanoyl)-2-phenyl-1,3-oxazolidine-
4-carboxylate (cis-7 and trans-7)
1
(m, CeH). H NMR (CDCl3): d ¼ 2.51 (ddd, J ¼ 18.1/12.8/
4.3 Hz, 1H, 3-H), 2.65 (dt, J ¼ 18.3/3.2 Hz, 1H, 3-H), 3.29
(dt, J ¼ 15.9/4.0 Hz, 1H, 4-H), 3.92 (ddd, J ¼ 15.7/12.7/
2.9 Hz, 1H, 4-H), 4.32 (dd, J ¼ 10.8/8.5 Hz, 1H, 8-H), 4.53
(d, J ¼ 14.7 Hz, 1H, PhCH2N), 4.78e4.83 (m, 2H, 7-H and
8-H), 4.81 (d, J ¼ 14.6 Hz, 1H, PhCH2N), 6.47 (s, 1H, 10-
A solution of benzaldehyde dimethyl acetal (2.15 mL,
14.31 mmol), 6 (1.0 g, 4.77 mmol) and p-toluenesulfonic
acid monohydrate (21 mg, 0.11 mmol) in toluene (20 mL)
was heated to reflux for 2 h in a DeaneStark trap. The reaction
mixture was concentrated to a volume of about 5 mL. After
addition of Et2O (20 mL) to the cooled solution, the organic
layer was washed with a saturated solution of NaHCO3
(2 ꢃ 10 mL), water (10 mL) and brine (10 mL). The organic
layer was dried (Na2SO4), evaporated in vacuo and the residue
was purified by fc (5 cm, CH2Cl2/ethyl acetate 9:1, 30 mL,
Rf ¼ 0.55) to provide a mixture of cis-7/trans-7 (yield
1
H), 7.27e7.43 (m, 10H, arom). H NOE: after irradiation at
d ¼ 6.47 ppm (10-H) a NOE was found at d ¼ 4.8 ppm (7-H
and 8-H). 13C NMR (CDCl3): d ¼ 35.9 (1C, C-3), 41.6 (1C,
C-4), 50.2 (1C, PhCH2N), 56.6 (1C, C-7), 68.6 (1C, C-8),
91.9 (1C, C-10), 126.7 (2C, C-4 [Bn] and C-4 [Ph]), 128.1
(1C, C-2 or C-3 or C-5 or C-6 arom), 128.3 (2C, C-2 and/or
C-3 and/or C-5 and/or C-6 arom), 128.4 (2C, C-2 and/or

S. Bedu¨rftig, B. Wu¨nsch / European Journal of Medicinal Chemistry 44 (2009) 519e525
523
C-3 and/or C-5 and/or C-6 arom), 128.8 (1C, C-2 or C-3 or C-
5 or C-6 arom), 128.9 (2C, C-2 and/or C-3 and/or C-5 and/or
C-6 arom), 136.2 (1C, C-1 [Bn] or C-1 [Ph]), 137.8 (1C, C-1
[Bn] or C-1 [Ph]), 167.3 (1C, C-2 or C-6), 167.4 (1C, C-6 or
C-2).
82). IR (film): n (cmꢀ1) ¼ 2938 (m, CeH), 2834 (m, CeH),
1638 (s, C]O), 1428 (m, CeH). 1H NMR (CDCl3):
d ¼ 2.78 (ddd, J ¼ 17.7/5.3/3.3 Hz, 1H, 3-H), 2.91 (ddd,
J ¼ 17.7/11.3/4.2 Hz, 1H, 3-H), 3.39 (dd, J ¼ 14.1/5.2 Hz,
1H, NCH2CH(OCH3)2), 3.42 (s, 3H, OCH3), 3.43 (s, 3H,
OCH3), 3.55 (dt, J ¼ 15.5/4.8 Hz, 1H, 4-H), 3.84 (dd,
J ¼ 13.9/5.0 Hz, 1H, NCH2CH(OCH3)2), 4.01 (ddd, J ¼ 15.1/
11.6/3.5 Hz, 1H, 4-H), 4.30e4.36 (m, 1H, CH(OCH3)2),
4.48e4.55 (m, 2H, 8-H), 4.80 (t, J ¼ 7.6 Hz, 1H, 7-H), 6.51
(s, 1H, 10-H), 7.27e7.38 (m, 5H, arom). 1H NOE: after
irradiation at d ¼ 6.51 ppm (10-H) a NOE at d ¼ 4.80 ppm
(7-H) was not detected.
trans-8a. Colorless viscous oil, yield 209 mg (67%).
[a]589 ¼ þ76.2 (c ¼ 0.55, CH2Cl2). C20H20N2O3 (336.4).
HRMS: Calcd. 336.1474, Found 336.1472 (ꢀ0.2 ppm). MS
(EI): m/z (%) ¼ 336 (M, 22), 259 (M ꢀ phenyl, 5), 245
(M ꢀ benzyl, 13), 105 (PhCHO, 26), 91 (benzyl, 100). IR
(film): n (cmꢀ1) ¼ 2928 (m, CeH), 1644 (s, C]O), 1429
1
(m, CeH). H NMR (CDCl3): d ¼ 2.53 (ddd, J ¼ 17.4/11.0/
4.2 Hz, 1H, 3-H), 2.78 (ddd, J ¼ 17.4/6.4/3.0 Hz, 1H, 3-H),
3.43 (ddd, J ¼ 15.5/6.5/4.7 Hz, 1H, 4-H), 3.83 (ddd,
J ¼ 15.3/11.0/3.0 Hz, 1H, 4-H), 4.39 (dd, J ¼ 9.5/7.1 Hz, 1H,
8-H), 4.57 (d, J ¼ 14.7 Hz, 1H, PhCH2N), 4.59 (t,
J ¼ 9.2 Hz, 1H, 8-H), 4.82 (d, J ¼ 14.3 Hz, 1H, PhCH2N),
4.81e4.86 (m, 1H, 7-H), 6.49 (s, 1H, 10-H), 7.27e7.38 (m,
6.6. (ꢀ)-[(2R)-(1,4-Dibenzyl)-1,4-diazepan-
2-yl]methanol (4a)
Under N2 atmosphere LiAlH4 powder (0.271 g,
7.13 mmol) was added to a stirred solution of 8a (mixture
of cis/trans isomers, 0.240 g, 0.713 mmol) in THF (40 mL).
The reaction mixture was heated to reflux for 96 h. Under
N2 atmosphere and cooling (ice bath) water (0.5 mL), 3 M
NaOH (0.5 mL) and water (0.5 mL) were successively added.
The suspension was refluxed for 30 min. After cooling down
to rt, the precipitate was filtered off and the solvent was
removed in vacuo. Purification of the residue by fc (2 cm,
ethyl acetate, 3 mL, Rf ¼ 0.34) furnished 4a as a colorless
oil, yield 0.101 g (46%). [a]589 ¼ ꢀ8.3 (c ¼ 0.585, CH2Cl2).
C20H26N2O (310.4) Calcd. C 77.37, H 8.44, N 9.02, Found
C 76.83, H 7.94, N 8.33. HRMS: Calcd. 310.2045, Found
310.2044 (ꢀ0.1 ppm). MS (EI): m/z (%) ¼ 310 (M, 1), 279
(M ꢀ CH2OH, 45), 219 (M ꢀ benzyl, 4), 91 (benzyl, 100).
1
10H, arom). H NOE: after irradiation at d ¼ 6.49 ppm (10-
H) a NOE at d ¼ 4.85 ppm (7-H) was not detected.
6.5. (ꢀ)-(7S,10R)-5-(2,2-Dimethoxyethyl)-10-phenyl-
9-oxa-1,5-diazabicyclo[5.3.0]decane-2,6-dione (cis-8b)
and (þ)-(7S,10S )-5-(2,2-dimethoxyethyl)-10-phenyl-9-
oxa-1,5-diazabicyclo[5.3.0]decane-2,6-dione (trans-8b)
A
solution of
7
(cis-7:trans-7 ¼ 30:70, 0.218 g,
0.73 mmol), aminoacetaldehyde dimethyl acetal (79 mL,
0.73 mmol) and triethylamine (0.10 mL, 0.73 mmol) in
acetonitrile (25 mL) was heated to reflux for 24 h. Then
titanium(IV) tetraisopropylate (0.32 mL, 1.1 mmol) was
added and heating was continued for additional 96 h. The
solvent was removed in vacuo and the residue was purified
by fc (3 cm, ethyl acetate, 8 mL) to yield cis-8b
(Rf ¼ 0.30) and trans-8b (Rf ¼ 0.18).
MS (CI): m/z (%) ¼ 311 (MHþ, 100). IR (film):
n
(cmꢀ1) ¼ 3416 (br, OH), 2932 (m, CeH), 1452 (m, CeH).
1H NMR (CDCl3): d ¼ 1.72e1.83 (m, 2H, 6-H), 2.54 (ddd,
J ¼ 12.2/7.9/4.3 Hz, 1H, 5-H or 7-H), 2.64e2.81 (m, 4H,
3-H, 5-H, and 7-H), 2.86e2.92 (m, 1H, 2-H), 2.99 (ddd,
J ¼ 14.3/7.4/3.2 Hz, 1H, 7-H or 5-H), 3.46 (d, J ¼ 5.5 Hz,
2H, CH2OH), 3.64 (s, 2H, PhCH2N4), 3.84 (d, J ¼ 13.7 Hz,
1H, PhCH2N1), 3.93 (d, J ¼ 13.7 Hz, 1H, PhCH2N1), 7.27e
7.37 (m, 10H, arom). The signal for the proton of the OH
group could not be detected. 13C NMR (CDCl3): d ¼ 27.2
(1C, C-6), 49.8 (1C, C-5 or C-7), 56.3 (1C, C-3, C-5 or C-
7), 56.4 (1C, C-3, C-5 or C-7), 57.0 (1C, N1CH2Ph), 62.5
(1C, C-2), 62.9 (1C, CH2OH), 63.9 (1C, N4CH2Ph), 127.1
and 127.2 (2C, 2 ꢃ C-4 arom), 128.30 and 128.32 (4C,
2 ꢃ C-3 and 2 ꢃ C-5 arom), 128.78 and 128.82 (4 C,
2 ꢃ C-2 and 2 ꢃ C-6 arom), 139.2 and 139.7 (2C, 2 ꢃ C-1
arom).
cis-8b. Colorless viscous oil, yield 20 mg (8%).
[a]589 ¼ ꢀ28.3 (c ¼ 0.375, CH2Cl2). C17H22N2O5 (334.4).
HRMS: Calcd. 334.1529, Found 334.1530 (þ0.1 ppm). MS
(EI): m/z (%) ¼ 334 (M, 4), 303 (M ꢀ OCH3, 7), 153
(M ꢀ PhCHO ꢀ CH(OCH3)2, 9), 105 (PhCHO, 5), 91 (benzyl,
6), 75 (CH(OCH3)2, 100). IR (film): n (cmꢀ1) ¼ 2938 (m, Ce
1
H), 2835 (m, CeH), 1639 (s, C]O), 1428 (m, CeH). H
NMR (CDCl3): d ¼ 2.71 (dt, J ¼ 18.3/2.9 Hz, 1H, 3-H), 2.88
(ddd, J ¼ 18.3/13.2/4.6 Hz, 1H, 3-H), 3.41e3.47 (m, 1H, 4-
H), 3.43 (s, 6H, 2 ꢃ OCH3), 3.48 (dd, J ¼ 13.8/5.5 Hz, 1H,
NCH2CH(OCH3)2), 3.65 (dd, J ¼ 13.9/4.8 Hz, 1H,
NCH2CH(OCH3)2), 4.05 (ddd, J ¼ 15.9/13.2/2.7 Hz, 1H, 4-
H), 4.22e4.29 (m, 1H, 8-H), 4.47 (t, J ¼ 5.0 Hz, 1H,
CH(OCH3)2), 4,72e4.79 (m, 2H, 7-H and 8-H), 6.48 (s, 1H,
1
10-H), 7.32e7.43 (m, 5H, arom). H NOE: after irradiation
at d ¼ 6.48 ppm (10-H) a NOE was found at d ¼ 4.75 ppm
6.7. (ꢀ)-[(2R)-1-Benzyl-4-(2,2-dimethoxyethyl)-1,4-
diazepan-2-yl]methanol (4b)
(7-H and 8-H).
trans-8b. Colorless viscous oil, yield 82 mg (34%).
[a]589 ¼ þ22.2 (c ¼ 0.58, CH2Cl2). C17H22N2O5 (334.4).
HRMS: Calcd. 334.1529, Found 334.1530 (þ 0.1 ppm). MS
(EI): m/z (%) ¼ 334 (M, 1), 105 (PhCHO, 1), 75 (CH(OCH3)2,
100). MS (CI): m/z (%) ¼ 335 (MHþ, 2), 303 (M ꢀ OCH3,
Under N2 atmosphere LiAlH4 powder (0.233 g, 6.13 mmol)
was added to a stirred solution of 8b (mixture of cis/trans iso-
mers, 0.205 g, 0.613 mmol) in THF (40 mL). The reaction
mixture was heated to reflux for 96 h. Under N2 atmosphere
and cooling (ice bath) water (0.5 mL), 3 M NaOH (0.5 mL)

524
S. Bedu¨rftig, B. Wu¨nsch / European Journal of Medicinal Chemistry 44 (2009) 519e525
and water (0.5 mL) were successively added. The suspension
was refluxed for 30 min. After cooling down to rt, the precip-
itate was filtered off and the solvent was removed in vacuo.
Purification of the residue by fc (2 cm, ethyl acetate/ethanol,
5:1, 3 mL, Rf ¼ 0.35) furnished 4b as pale yellow oil, yield
concentrations of the test compound, 3 nM [3H]-pentazocine,
and buffer (50 mM TriseHCl, pH 7.4) in a total volume of
500 mL for 120 min at 37 ꢁC. The incubation was terminated
by rapid filtration through presoaked Whatman GF/B filters
(0.5% polyethylenimine in water for 2 h at 4 ꢁC) using
a cell harvester. After washing four times with 2 mL of cold
buffer 3 mL of scintillation cocktail were added to the filters.
After at least 8 h bound radioactivity trapped on the filters was
counted in a liquid scintillation analyzer. Non-specific binding
was determined with 10 mM haloperidol.
75 mg (40%.). [a]589 ¼ ꢀ3.9 (c ¼ 0.505,
CH2Cl2).
C17H28N2O3 (308.4). HRMS: Calcd. 308.2099, Found
308.2092 (ꢀ0.7 ppm). MS (EI): m/z (%) ¼ 308 (M, 3), 277
(M ꢀ CH2OH, 75), 233 (M ꢀ CH(OCH3)2, 49), 186
(M ꢀ benzyl ꢀ CH2OH, 28), 91 (benzyl, 100). MS (CI): m/z
(%) ¼ 309 (MHþ, 100), 277 (M ꢀ CH2OH, 31). IR (film): n
(cmꢀ1) ¼ 3306 (br, OeH), 2937 (m, CeH), 2871 (m, CeH),
7.3. s2 Receptor binding assay [18]
1
1446 (m, CeH). H NMR (CDCl3): d ¼ 1.67e1,86 (m, 2H,
6-H), 2.53e2.67 (m, 4H, 3-H, 5-H and 7-H), 2.68e2.82 (m,
3H, CH2CH(OCH3)2, and 3-H or 5-H or 7-H), 2.85e2.93
(m, 1H, 2-H), 2.95 (ddd, J ¼ 13.9/7.0/3.1 Hz, 1H, 3-H, 5-H
or 7-H), 3.37 (s, 3H, OCH3), 3.39 (s, 3H, OCH3), 3.51 (dd,
J ¼ 11.1/6.1 Hz, CH2OH), 3.64 (dd, J ¼ 11.0/6.2 Hz, 1H,
CH2OH), 3.83 (d, J ¼ 15.1 Hz, 1H, PhCH2N), 3.85 (d,
J ¼ 14.7 Hz, 1H, PhCH2N), 4.49 (t, J ¼ 5.4 Hz, 1H,
CH(OCH3)2), 7.24e7.32 (m, 5H, arom). The signal for the
proton of the OH group could not be detected. 13C NMR
(CDCl3): d ¼ 27.6 (1C, C-6), 49.5 (1C, C-3, C-5 or C-7),
52.9 and 53.8 (2C, 2 ꢃ OCH3), 55.8 (1C, NCH2CH(OCH3)2),
57.6 (1C, C-3, C-5 or C-7), 57.9 (1C, NCH2Ph), 60.3 (1C, C-3,
C-5 or C-7), 62.7 (1C, C-2), 63.1 (1C, CH2OH), 103.0 (1C,
CH(OCH3)2), 127.0 (1C, C-40), 128.3 (2C, C-30 and C-50),
128.8 (2C, C-20 and C-60), 139.7 (1C, C-10).
For the s2 assay rat liver membranes were prepared as de-
scribed in Ref. [18]. The membrane preparation (about 60 mg
of protein) was incubated with 3 nM [3H]-ditolylguanidine
(2220 GBq/mmol, American Radiolabeled Chemicals, Inc.)
and different concentrations of test compounds in buffer
(50 mM TriseHCl, pH 8.0) in the presence of 100 nM
(þ)-pentazocine. The total volume was 250 mL. The incuba-
tion (120 min, 25 ꢁC) was stopped by addition of 2 mL of
ice-cold buffer (10 mM TriseHCl, pH 8.0) followed by rapid
filtration through presoaked Whatman GF/B filters (0.5% pol-
yethylenimine in water for 2 h at 4 ꢁC) using a cell harvester.
After the sample was washed three times with 2 mL of cold
buffer, a total volume of 3 mL of scintillation cocktail was
added to the filters. After at least 8 h bound radioactivity trap-
ped on the filters was counted in a liquid scintillation analyzer.
Non-specific binding was determined with 10 mM non-
radiolabeled ditolylguanidine.
7. Receptor binding studies
7.1. General
Acknowledgement
Homogenizer: PotterÒS (B. Braun Biotech International).
Ultraturrax: EuroturraxÒ T20 (Ika Labortechnik). Centrifuge:
High speed cooling centrifuge model J2-HS (Beckman). Fil-
ter: Whatman glass fiber filters GF/B, presoaked in the me-
dium described below before use. Filtration was performed
with a Brandel 24-well cell harvester. Scintillation cocktail:
Rotiscint Eco Plus (Roth). Liquid scintillation analyzer: Tri-
Carb 2100 TR (Canberra Packard), counting efficiency 66%.
All experiments were carried out in triplicates. IC50 values
were determined from competition experiments with at least
6 concentrations of test compounds and were calculated
with the program GraphPad PrismÒ 3.0 (GraphPad Soft-
ware) by non-linear regression analysis. Ki values were cal-
culated according to Cheng and Prusoff [21]. The Ki values
are given as mean value ꢂ SEM from three independent
experiments.
Financial support of this work by the Deutsche Forschungs-
gemeinschaft and the Fonds der Chemischen Industrie is grate-
fully acknowledged. Thanks are also due to the Degussa AG,
Janssen-Cilag GmbH, and Bristol-Myers Squibb for donation
of chemicals and reference compounds.
References
[1] M. Hanner, F.F. Moebius, A. Flandorfer, H.-G. Knaus, J. Striessnig,
E. Kempner, H. Glossmann, Proc. Natl. Acad. Sci. U.S.A. 93 (1996)
8072e8077.
[2] R. Kekuda, P.D. Prasad, Y.-J. Fei, F.H. Leibach, V. Ganapathy, Biochem.
Biophys. Res. Commun. 229 (1996) 553e558.
[3] P. Seth, Y.-J. Fei, H.W. Li, W. Huang, F.H. Leibach, V. Ganapathy, J.
Neurochem 70 (1998) 922e931.
[4] P.D. Prasad, H.W. Li, Y.-J. Fei, M.E. Ganapathy, T. Fujita, L.H. Plumley,
T.L. Yang-Feng, F.H. Leibach, V. Ganapathy, J. Neurochem. 70 (1998)
443e451.
7.2. s1 Receptor binding assay [18]
[5] L. Brasili, Pharm. Acta Helv. 74 (2000) 201e203.
[6] C.S. John, B.B. Lim, B.J. Vilner, B.C. Geyer, W.D. Bowen, J. Med.
Chem. 41 (1998) 2445e2450.
[7] M. Abou-Gharbia, S.Y. Ablordeppey, R. Glennon, in: J.A. Bristol (Ed.),
Annual Reports in Medicinal Chemistry, vol. 28, Academic, San Diego,
1993, pp. 1e10.
For the s1 assay guinea pig brain membranes were pre-
pared as described in Ref. [18]. The test was performed with
the radioligand [3H]-pentazocine (1036 GBq/mmol; NENÔ
Life Science Products). The thawed membrane preparation
(about 150 mg of protein) was incubated with various
[8] F. Eiden, H. Lentzen, Pharm. Unserer Zeit 25 (1996) 250e259.
[9] Monograph Igmesine Hydrochloride, Drugs Fut. 24 (1999) 133e140.

S. Bedu¨rftig, B. Wu¨nsch / European Journal of Medicinal Chemistry 44 (2009) 519e525
525
[10] A. Foster, H. Wu, W. Chen, W. Williams, W.D. Bowen, R.R. Matsumoto,
A. Coop, Bioorg. Med. Chem. Lett. 13 (2003) 749e751.
[11] R.R. Matsumoto, Y. Liu, M. Lerner, E.W. Howard, D.J. Bracket, Eur. J.
Pharmacol. 469 (2003) 1e12.
[15] B.R. de Costa, X.-S. He, J.T.M. Linders, C. Dominguez, Z.Q. Gu,
W. Williams, W.D. Bowen, J. Med. Chem. 36 (1993) 2311e2320.
[16] S. Bedurftig, B. Wunsch, Bioorg. Med. Chem. 12 (2004) 3299e3311.
¨
¨
[17] H. Bieraugel, H.E. Schoemaker, H. Hiemstra, J.H. van Maarseveen, Org.
¨
Biomol. Chem. 1 (2003) 1830e1832.
[12] R.R. Matsumoto, K.A. McCracken, B. Pouw, J. Miller, W.D. Bowen,
W. Williams, B.R. de Costa, Eur. J. Pharmacol. 411 (2001) 261e273.
[13] K.W. Crawford, W.D. Bowen, Cancer Res. 62 (2002) 313e322.
[14] S.-R. Choi, B. Yang, K. Plossl, S. Chumpradit, S.P. Wey, P.D. Acton,
K. Wheeler, R.H. Mach, H.F. Kung, Nucl. Med. Biol. 28 (2001)
657e666.
[18] C.A. Maier, B. Wunsch, J. Med. Chem. 45 (2002) 438e448.
¨
[19] C.A. Maier, B. Wunsch, J. Med. Chem. 45 (2002) 4923e4930.
¨
[20] S. Soukara, C. Maier, U. Predoiu, A. Ehret, R. Jackisch, B. Wunsch, J.
¨
Med. Chem. 44 (2001) 2814e2826.
[21] Y. Cheng, W.H. Prusoff, Biochem. Pharmacol. 22 (1973) 3099e3108.