Tetrahydro-3-benzazepines with fluorinated side chains as NMDA and σ1 receptor antagonists: Synthesis, receptor affinity, selectivity and antiallodynic activity

Article abstract of DOI:10.1016/j.ejmech.2019.05.034

The class of tetrahydro-1H-3-benzazepines was systematically modified in 1-, 3- and 7-position. In particular, a F-atom was introduced in β- or γ-position of the 4-phenylbutyl side chain in 3-position. Ligands with the F-atom in γ-position possess higher

Full text of DOI:10.1016/j.ejmech.2019.05.034

Accepted Manuscript
Tetrahydro-3-benzazepines with fluorinated side chains as NMDA and σ receptor
1
antagonists: Synthesis, receptor affinity, selectivity and antiallodynic activity
Simone Thum, Dirk Schepmann, Eva Ayet, Marta Pujol, Francisco R. Nieto, Simon M.
Ametamey, Bernhard Wünsch
PII:
S0223-5234(19)30444-1
DOI:
https://doi.org/10.1016/j.ejmech.2019.05.034
EJMECH 11344
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 22 March 2019
Revised Date: 23 April 2019
Accepted Date: 12 May 2019
Please cite this article as: S. Thum, D. Schepmann, E. Ayet, M. Pujol, F.R. Nieto, S.M. Ametamey, B.
Wünsch, Tetrahydro-3-benzazepines with fluorinated side chains as NMDA and σ receptor antagonists:
1
Synthesis, receptor affinity, selectivity and antiallodynic activity, European Journal of Medicinal
Chemistry (2019), doi: https://doi.org/10.1016/j.ejmech.2019.05.034.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

1
ACCEPTED MANUSCRIPT
Tetrahydro-3-benzazepines with fluorinated side chains as NMDA and σ1
receptor antagonists: Synthesis, receptor affinity, selectivity and antiallodynic
activity
a
a
b
b
c
Simone Thum, Dirk Schepmann, Eva Ayet, Marta Pujol, Francisco R. Nieto ,
d
a,e
Simon M. Ametamey, Bernhard Wünsch*
a
Institut für Pharmazeutische und Medizinische Chemie der Universität Münster,
Corrensstraße 48, D-48149 Münster, Germany
Tel.: +49-251-8333311; Fax: +49-251-8332144; E-mail: wuensch@uni-muenster.de
b
Esteve Pharmaceuticals S.A, Baldiri Reixach 4-8, 08028 Barcelona, Spain
c
Department of Pharmacology and Institute of Neuroscience, School of Medicine and
Biomedical Research Center, University of Granada, Avenida de la Investigación 11,
1
8016 Granada, Spain
d
Department of Chemistry and Applied Biosciences, Institute of Pharmaceutical
Sciences, ETH Zurich, Zurich, Switzerland
e
Cells-in-Motion Cluster of Excellence (EXC 1003-CiM), Westfälische Wilhelms-
Universität Münster, Germany
Abstract
The class of tetrahydro-1H-3-benzazepines was systematically modified in 1-, 3- and
7
-position. In particular, a F-atom was introduced in β- or γ-position of the 4-
phenylbutyl side chain in 3-position. Ligands with the F-atom in γ-position possess
higher GluN2B affinity than analogs bearing the F-atom in β-position. This effect was
attributed to the reduced basicity of β-fluoro amines. 3-Benzazepines with a benzylic
OH moiety show moderate GluN2B affinity, but considerable selectivity over the σ2

2
ACCEPTED MANUSCRIPT
receptor. However, removal of the benzylic OH moiety led to increased GluN2B
affinity, but reduced GluN2B/σ selectivity. With respect to GluN2B affinity the phenol
2
1
7b with a γ-fluorophenylbutyl moiety in 3-position represents the most interesting
fluorinated ligand (K(GluN2B) = 16 nM). Most of the synthesized ligands reveal either
i
similar GluN2B and σ affinity or higher σ affinity than GluN2B affinity. The methyl
1
1
ether 16b shows high σ affinity (K(σ ) = 6.6 nM) and high selectivity over a broad
1
i
1
panel of receptors and transporters. The high antiallodynic activity in the mouse
capsaicin assay proved the σ antagonistic activity of 16b.
1
Keywords
NMDA receptor, GluN2B subunit selective NMDA receptor antagonist, fluorinated
side chain, tetrahydro-3-benzazepines, structure activity relationships, receptor
selectivity, PET, σ receptor selectivity, antiallodynic activity.
1
1
. Introduction
N-Methyl-D-aspartate (NMDA) receptors are indispensable as a control unit for the
glutamatergic network in the central nervous system (CNS), keeping the excitatory
1
neurotransmission balanced. Upon voltage-dependent activation of NMDA receptors
2+
by (S)-glutamate release, physiological Mg blockade of the channel pore is
unlocked and ions can pass. In contrast to the other ionotropic glutamate receptors
(2-amino-3-(3-hydroxy-5-methylisoxazol-4-yl)propionic acid (AMPA) and kainate
2+
receptor), the NMDA receptor mediates influx of Ca ions into neurons and therefore
plays an important role in synaptic plasticity and thus in nervous system
2,3
development, memory and learning.
However, in addition to its positive effects on central nervous system development,

3
ACCEPTED MANUSCRIPT
elevated glutamate concentrations in the synaptic cleft induce neurotoxic effects.
2+
Overactivation of NMDA receptors, resulting in an increased Ca ion influx can lead
finally to excitotoxicity being associated with severe cell damage and subsequent
apoptosis of neurons. This process in turn is linked to the progression of
neurodegenerative disorders like Alzheimer’s and Parkinson’s disease but also other
CNS disorders such as depression, stroke/ischemia, neuropathic pain and alcohol
4,5,6
withdrawal symptoms.
Structurally, the heterotetrameric NMDA receptor is composed of two GluN1 and two
GluN2 subunits. Four GluN2 subunits termed GluN2A-D subunits are known, which
show a distinct expression and developmental pattern as well as a particular
7
physiological role. Therapeutically, selective addressing NMDA receptors with
GluN2B subunit at the so called ifenprodil binding site, not present in subunits
8
GluN2A, GluN2C and GluN2D, has gained increasing attention. The therapeutic
potential of GluN2B negative allosteric modulators with decreased cognitive side
effects has promoted the development of ifenprodil-like drugs.
To verify their exclusive binding to the ifenprodil binding site, to exclude off-target
binding in vivo and identify the required dose for receptor occupation, molecular
9
imaging experiments with positron emission tomography (PET) are essential.
Therefore, high affinity GluN2B ligands, which can be radiolabeled, are of major
interest. Despite considerable efforts, a radioligand for labeling the GluN2B subunit in
10
NMDA receptors for human PET imaging studies is not yet available. Very recently,
11
carbon-11-labeled tetrahydro-3-benzazepin-1-ol [ C]Me-NB1 (2), derived from
ifenprodil (1), was successfully developed as GluN2B PET radiotracer without
suffering from the common problems of already investigated known radioligands.

4
ACCEPTED MANUSCRIPT
11
(Figure 1)
11
Figure 1. Lead structure ifenprodil (1), previously reported radiotracer [ C]Me-NB1
11
(
[ C]2) and design of fluorinated 3-benzazepines 3a and 3b. *Ifenprodil represents
the racemic erythro diastereomer.
To overcome the short physical half-life of carbon-11 (20.3 min), GluN2B ligands
3
a,b amenable for labeling with fluorine-18 (109.8 min) should be developed. In
previous studies, introduction of fluorinated phenylbutyl and phenylpropyl side chains
12,13
into different scaffolds was investigated to achieve high GluN2B affinity.
In these
studies, tetrahydro-3-benzazepine emerged as promising scaffold for addressing the
ifenprodil binding site in the GluN2B subunit. Therefore, we report herein on the
systematic variation and structure-affinity relationship of 3-benzazepines bearing a F-
atom in the side chain, which should be amenable for fluorine-18 labeling. To assess
the impact of the position of the F-atom in the phenylbutyl side chain, the tetrahydro-
3
-benzazepine scaffold decorated with various substituents should be combined with
fluorinated phenylbutyl side chains bearing the F-atom in β- or γ-position relative to
the amino moiety. (Figure 1)

5
ACCEPTED MANUSCRIPT
2
. Synthesis
Scheme 1. Synthesis of tertiary amines 8a, 8b and 9b. Reagents and reaction
14
conditions: (a) p-TsCl, Et N, 4-DMAP, Bu SnO, CH CN:THF=1:1, 14 h, rt, 85 %.
3
2
3
®
(
b) XtalFluor E , Et N•3 HF, CH Cl , 17 h, -78 °C - rt, 72 %. (c) K CO , CH CN, 3 d,
3 2 2 2 3 3
reflux, 14 % (8a), 82 % (8b). (d) H (balloon), Pd/C, CH OH, rt, 3 h, 86 %.
2
3
The preparation of desired 3-benzazepines of type 3 with either a F-atom in β- or γ-
position of the phenylbutyl side chain was envisaged via a nucleophilic substitution
reaction. Therefore, fluorinated phenylbutyl tosylates 7a, bearing the F-atom in β-
13
position, and already described 7b, fluorinated in γ-position, were synthesized.
Scheme 1) Diol 4a, obtained from 4-phenylbut-1-ene by a Sharpless dihydroxylation
reaction with AD-mix, was regioselectively tosylated at the primary alcohol using
(
14
Bu SnO to afford 5a. Subsequent deoxyfluorination of the secondary alcohol with
2
®
(
diethylamino)-difluorosulfonium tetrafluoroborate (XtalFluor E ) provided building
block 7a. To connect the fluorinated phenylbutyl side chains of 7a and 7b with
15,16
benzylated 3-benzazepin-1-ol 6
,
nucleophilic substitution reactions were

6
ACCEPTED MANUSCRIPT
performed using K CO as a base. Hence, tertiary amine 8b, fluorinated in γ-position,
2
3
was obtained in 86 % yield, whereas tertiary amine 8a, with a F-atom in β-position
was isolated in only 14 % yield, even after long reaction periods of up to three days.
Thus, only benzyl ether 8b was subjected to hydrogenolytic cleavage to provide the
phenol 9b. The 3-benzazepin-1-ols 8a, 8b and 9b were obtained as 1:1-mixtures of
diastereomers, which could not be separated by flash chromatography.
Scheme 2. Nucleophilic substitution (a,b) and reductive alkylation (c) for synthesis of
tertiary amines 12a-12c. Reagents and reaction conditions: (a) 7a. K CO , CH CN,
2
3
3
4
d, reflux, 7 % (12a). (b) 7b. K CO , CH CN, 20 h, reflux, 70 % (12b). (c)
2 3 3
15
NaBH(OAc) , CH Cl , 15 h, rt, 12 % (12a), 62 % (13), 57 % (12c ).
3
2
2
Two reaction pathways were investigated to obtain the tertiary amine 12a with a F-
atom in β-position. (Scheme 2) As already described for secondary amine 6
13
(Scheme 1), substitution reaction of commercially available 10 with tosylate 7a
provided the tertiary amine 12a in only 7 % yield. This clearly indicates, that the F-

7
ACCEPTED MANUSCRIPT
atom in β-position hinders the nucleophilic substitution reaction. This assumption is
confirmed by the rather high yield (70 %) of fluorinated 3-benzazepinol 12b prepared
by nucleophilic substitution of secondary amine 10 with γ-fluorotosylate 7b. Both 12a
and 12b with F-atom in β- or γ-position exit as 1:1-mixtures of diastereomers.
In order to improve the yield of the β-fluoro derivative 12a reductive alkylation of
13
17
secondary amine 10 with β-fluoroaldehyd 11a and NaBH(OAc)3 was performed.
Scheme 2) Although the β-fluoro derivative 12a was obtained in slightly higher yield
of 12 %, two pairs of diastereomeric N,O-acetals 13 were obtained as main products
62 %). The same reaction with non-fluorinated 4-phenylbutanal (11c) proves the
(
(
participation of the F-atom in the formation of the N/O-acetal 13, since in this case
the desired tertiary amine 12c was isolated as sole product in 57 % yield. The
generation of acetalic compounds 13 is explained by intramolecular nucleophilic
attack of benzylic alcohol of 10 on the intermediate iminium ion, favored by the
electron withdrawing β-F-atom. The resulting N,O-acetals 13 could not be reduced by
NaBH(OAc)3.
In the next modification, the benzylic OH-moiety of 10 should be removed by
reductive deoxygenation. Thus, treatment of alcohol 10 with Et SiH and BF •OEt
2
3
3
provided 3-benzazepine 14 in 54 % yield. (Scheme 3) N-alkylation of secondary
amine 14 by nucleophilic substitution reaction with tosylate 7b yielded in γ-position
fluorinated tertiary amine 16b. 3-Benzazepine 16a with the F-atom in β-position and
the non-fluorinated phenylbutyl derivative 16c were prepared by reductive alkylation
with α-fluorobutanal 11a and 4-phenylbutynal 11c, respectively.

8
ACCEPTED MANUSCRIPT
Scheme 3. Synthesis of 3-benzazepines by deoxygenation and ether cleavage.
18
Reagents and reaction conditions: (a) Et SiH, BF •OEt , CH Cl , rt, 4 h, 54 %. (b)
3
3
2
2
2
18
4
8 % HBr, 5 h, reflux, 87 %. (c) 7b, K CO , CH CN, 2-3 d, reflux, 80 % (16b); 28 %
2 3 3
(
17b), 14 % (18b). (d) NaBH(OAc) , CH Cl , 15-18 h, rt, 69 % (16a), 12 % (16c);
3 2 2
1
4 % (17a), 68 % (17c), (e) (2-Fluoroethyl) tosylate, K CO , DMF, 19 h, 80 °C, 45 %.
2 3
To obtain the corresponding phenols 17a-c, demethylation of 14 was accomplished
18
using 47 % HBr to afford phenol 15 in 78 % yield. (Scheme 2) N-Alkylation of
secondary amine 15 was performed by reductive alkylation with aldehydes 11a and
1
1c in the presence of NaBH(OAc) and nucleophilic substitution with tosylate 7b. In
3
addition to the N-alkylated phenol 17b (28 %) the double alkylated product 18b was
isolated in 14 % yield after reaction of phenol 15 with tosylate 7b. (Scheme 3)
To introduce a further F-containing modification in 7-position, phenol 17b was
alkylated with (2-fluoroethyl) tosylate to afford the fluoroethyl ether 19b. The late

9
ACCEPTED MANUSCRIPT
stage fluoroethylation of phenol 17b offers the possibility for radiofluorination at the
18
very end of the synthesis by using F-labeled (2-fluorethyl) tosylate. (Scheme 3)
Scheme 4. Synthesis of tertiary amines 21 and 22. Reagents and reaction conditions:
13
(
a) K CO , CH CN, 3 d, reflux, 10 %. (b) NaBH(OAc) , CH Cl , 20 h, rt, 33 %. (c)
2 3 3 3 2 2
DMP, CH Cl , 1 h, rt, 25 % (22b), 32 % (22c).
2
2
Next, tetrahydro-3-benzazepines 21 without further substituents at the benzazepine
scaffold were envisaged to investigate the impact of the different substituents in
7
-position. For this purpose, commercially available tetrahydro-3-benazazepine 20
was reacted with fluorinated tosylate 7b to provide 21b. Reductive alkylation of
secondary amine 20 with fluorinated aldehyde 11a and NaBH(OAc) resulted in
3
13
tertiary amine 21a. Low yields of desired compounds 21a and 21b can be assigned
to decomposition of 20 during storage and purification problems of final products.
(Scheme 4)
In addition to 3-benzazepines without substituent in 1-position (16, 17, 19, 21) and
with an OH-moiety in 1-position (8, 9, 12) able to act as H-bond donor and acceptor,

10
ACCEPTED MANUSCRIPT
an oxo group should be introduced in 1-position, which can only react as H-bond
15
acceptor component. Thus benzylic alcohols 12b and 12c were oxidized with Dess-
Martin-Periodinane to provide the ketones 22b and 22c, respectively. (Scheme 4)
Unexpectedly, the ketones showed a very fast decomposition, which inhibited the
inclusion of the ketones into biological testing.
In order to investigate the effect of the configuration of the F-containing C-atom, the
enantiomers of the γ-fluorophenol 17b were separated by chiral HPLC (column
®
Daicel Chiralpak IA). According to this procedure, both enantiomers (-)-17b and (+)-
1
7b were obtained with 98.6 and 99.8 % ee, respectively (for details see Supporting
Information).
3
3
. Receptor affinity
.1. GluN2B affinity
The affinity towards NMDA receptors with GluN2B subunit (GluN2B affinity) of the
synthesized 3-benzazepines and reference compounds was determined in a
3
competitive radioligand binding assay, using tritium-labeled ifenprodil ([ H]1) as
radioligand. Selectivity for the GluN2B subunit was achieved by employing
membrane preparations from a recombinant cell line stably expressing the GluN1a
19
splice variant and the GluN2B subunit. The results are summarized in Table 1.
Table 1. GluN2B, PCP, σ and σ receptor affinity of synthesized 3-benzazepines
1
2
and reference compounds.

11
ACCEPTED MANUSCRIPT

12
ACCEPTED MANUSCRIPT
a
K ± SEM [nM] (n=3)
i
1
2
b
compd.
R
R
X
β-F
γ-F
H
GluN2B
784
PCP [%]
σ1
σ2
1400
8
8
8
9
9
1
1
1
1
1
1
1
1
1
1
a
OBn
OH
OH
OH
OH
OH
OH
OH
OH
0
393
183
b
OBn
813
0
1800
1
6,20
c
OBn
505 ± 170
422 ± 57
84 ± 18
506
>1 µM
293 ± 58
1400
194
1050
b
OH
γ-F
H
8
130 ± 26
18000
7900
1
6,20
c
OH
35
2a
2b
OCH3
β-F
γ-F
H
0
585
OCH3
630
0
210
4500
15,20
2c
OCH3
706 ± 100
22
182 ± 38 554 ± 27
c
c
c
3
OCH3
N/O acetal γ-F
8 %
0
0 %
0 %
6a
6b
6c
7a
7b
7c
OCH3
H
H
H
H
H
H
H
H
H
H
H
H
H
β-F
γ-F
H
602
120 ± 16
55 ± 10
86 ± 3
16 ± 5
9.9 ± 3
65 ± 7
54 ± 11
249
0
213
740
115
OCH3
21
6.6 ± 2
11 ± 5
OCH3
0
56 ± 1
OH
β-F
γ-F
H
0
70 ± 13 153 ± 36
OH
15
17 ± 5
32 ± 8
39 ± 15
12 ± 3
63 ± 8
49 ± 7
643
OH
nd
(
(
-)-17b
+)-17b
OH
OH
γ-F
γ-F
γ-F
γ-F
β-F
γ-F
H
nd
25 ± 11
10 ± 2
nd
1
1
2
2
2
8b
O(CH ) CHFCH Ph
nd
185
2
2
2
9b
OCH CH F
348
nd
0
25 ± 0.9
108 ± 27
8.0 ± 2.2
10 ± 8
524
2
2
1
3
1a
H
239 ± 11
30 ± 9
227 ± 2
10 ± 0.7
13 ± 2.5
nd
333
1b
H
H
0
48 ± 9
35 ± 13
2
1
1c
>10 µM
nd
Ifenprodil (1)
Eliprodil
125 ± 24 98 ± 34
nd
nd
nd
nd
Nd
Nd
Nd
dexoxadrol
32 ± 7.4
3.4 ± 0.8
nd
(
+)-MK-801
nd
haloperidol
di-o-
nd
6.3 ± 1.6 78 ± 2.3
tolylguanidin
e
nd
nd
89 ± 29
57 ± 18
a
K values of potent compounds were recorded three times (n = 3). For low-affinity
i
compounds with K values at GluN2B-NMDA receptors > 150 nM, the competition
i
curves were recorded only once (single value).

13
ACCEPTED MANUSCRIPT
b
Due to very low affinity towards the PCP binding site of the NMDA receptor, the
3
replacement (in %) of the radioligand [ H](+)-MK-801 from the PCP binding site at a
test compound concentration of 1 µM is given.
c
Due to very low affinity the replacement of the radioligands (in %) from the ifenprodil
binding site, the σ and σ receptors is given.
1
2
nd indicates not determined.
Previous studies have shown that 3-benzazepines without phenolic OH moiety still
22
interact with GluN2B containing NMDA receptors but with reduced affinity. This
finding is in good accordance with our results, since the highest GluN2B affinity was
found for the phenolic 3-benzazepine 17c (K = 9.9 nM). In general, various
i
substituents (H, OH, OCH ) in 7-position of the 3-benzazepine scaffold are well
3
tolerated by the ifenprodil binding pocket, but compounds with a 7-OH moiety (9c,
1
7a-c) result in the highest GluN2B affinity. Even compounds with larger substituents
such as 2-fluoroethoxy (19b) and 4-phenylbutoxy groups (18b) show moderate
GluN2B affinity with K values of 348 nM and 249 nM, respectively.
i
Regarding the position of the F-atom in the phenylbutyl side chain, 3-benzazepines
fluorinated in γ-position relative to the N-atom reveal a remarkably increased GluN2B
affinity compared to compounds bearing the F-atom in the adjacent β-position. This
effect can nicely be observed for unsubstituted 3-benzazepines 21. 3-Benzazepine
2
1b (K = 30 nM) bearing the F-atom in γ-position is 8-fold more potent than 3-
i
benzazepine 21a (K = 239 nM) with the F-atom in β-position. Furthermore, shifting
i
the F-atom from β- to γ-position resulted in 5-fold increased GluN2B affinity of methyl
ethers 16a and 16b, as well as phenols 17a and 17b. The reduced GluN2B affinity of
3
-benzazepines bearing the F-atom in β-position may be explained by a strong effect
on the basicity of the cyclic amine. The basicity reducing effect is considerably lower

14
ACCEPTED MANUSCRIPT
23
for γ-fluoro derivatiyes.
It can be concluded that the preferred position for the
introduction of a F-atom in this compound class is the γ-position.
Removal of the benzylic OH moiety (1-OH group) led to increased GluN2B affinity.
This effect is well demonstrated by the more active 3-benzazepines bearing the F-
atom in γ-position of the side chain. As example, methyl ether 12b with benzylic OH
moiety shows a 5-fold lower GluN2B affinity (K = 630 nM) than methyl ether 16b
i
without benzylic OH moiety (K = 120 nM). An even stronger effect was observed for
i
phenolic 3-benzazepines. Compound 17b without benzylic OH moiety reveals a 26-
fold higher GluN2B affinity (K = 16 nM) than in 1-position hydroxylated counterpart
i
9
b (K = 422 nM). Moderate to low GluN2B affinity was also observed for 3-
i
benzazepines with benzylic OH moiety bearing the F-atom in β-position. As
examples, β- and γ-fluorinated methyl ethers 12a and 12b (K = 506 and 630 nM) as
i
well as benzyl ethers 8a and 8b (K = 784 and 813 nM) show almost the same
i
GluN2B affinities.
Exemplarily, the enantiomers of the most promising fluorinated ligand 17b were
separated and tested. With K-values of 65 nM and 54 nM, both enantiomers (-)-17b
i
and (+)-17b showed almost the same GluN2B affinity. This result led to the
conclusion that the center of chirality bearing the small F-atom in γ-position of the 4-
phenylbutyl side chain is not important for the interaction with the ifenprodil binding
site of GluN2B-NMDA receptors.
Finally, this study demonstrated nicely that a F-atom in the phenylbutyl side chain,
preferably in the γ-position, is well tolerated by GluN2B containing NMDA receptors.
Therefore, the γ-CH moiety of the phenylbutyl side chain represents a promising
2

15
ACCEPTED MANUSCRIPT
position for radiofluorination, i.e. introduction of 18-F by nucleophilic substitution of an
appropriate tosylate precursor. Introduction of a F-atom into the γ-position of the most
potent 3-benzazepine 17c (K = 9.9 nM) led to almost the same GluN2B affinity
i
(
K(17b) = 16 nM).
i
3
.2. Receptor selectivity
To exclude the phencyclidine (PCP, 1-(1-phenylcyclohexyl)piperidine) binding site
within the channel pore of the NMDA receptor as a potential competing binding site,
3
PCP affinity was determined in competitive binding studies using the radioligand [ H]-
(+)-MK-801. Table 1 clearly reveals that the tested 3-benzazepines do not interact
significantly with the phencyclidine binding site, indicating high selectivity for the
ifenprodil binding site over the PCP binding site within the NMDA receptor.
Due to known cross reactivity of NMDA receptor ligands with both σ and σ
1
2
24
receptors and vice versa, affinity of synthesized 3-benzazepines towards σ and σ
1
2
3
receptors was investigated. In the competitive binding assays [ H]-(+)-pentazocine
3
25,26,27
(
σ ) and [ H]ditolylguanidine (σ ) were used as radioligands.
The affinity data
1
2
are summarized in Table 1.
With exception of phenol 9b, 3-benzazepines bearing an OH moiety in benzyl
position (1-position) reveal some selectivity against the σ receptor. For the methyl
2
ethers 12a and 12b the GluN2B/σ selectivity factors are 15 and 7, respectively,
2
whereas benzyl ethers 8a-c show a twofold selectivity against σ receptors. An
2
opposed profile was observed for phenol 9b, the only 3-benzazepine of this series
displaying moderate σ2 affinity (K = 130 nM) and selectivity over GluN2B
i
(
K = 422 nM) and σ receptors (K = 1400 nM). Removal of the benzylic OH moiety
i 1 i

16
ACCEPTED MANUSCRIPT
led to increased GluN2B and σ affinity resulting finally in almost the same K values
2
i
for both receptors indicating loss of GluN2B/σ selectivity. Methyl ether 16c
2
(
K(GluN2B) = 55 nM; K(σ ) = 56 nM) and phenol 17c (K(GluN2B) = 9.9 nM; K(σ )
i i 2 i i 2
=
12 nM) are examples for high affinity but low selectivity GluN2B/ σ ligands.
2
A comparable picture results from analysis of σ affinity data. Only phenols 9b, 9c
1
and 17c reveal a 2-3-fold selectivity for GluN2B over σ receptors. Removal of the
1
OH moiety in 1-position increased the σ affinity drastically and introduction of a F-
1
atom in γ-position of the phenylbutyl side chain further increased the σ receptor
1
affinity. Thus, most of the synthesized compounds show either similar GluN2B and σ1
affinities or considerably higher σ affinity than GluN2B affinity. Methyl ether 16b
1
(
K(σ ) = 6.6 nM), 2-fluoroethyl ether 19b (K(σ ) = 5.0 nM) and unsubstituted 3-
i 1 i 1
benzazepine 21b (K(σ ) = 8.0 nM) with a F-atom in γ-position of the phenylbutyl side
i
1
chain and without OH moiety in benzylic position show very high σ affinity (K(σ )
1
i
1
<
10 nM) and selectivity over GluN2B and σ receptors.
2
4
. Pharmacological in vitro characterization of 16b and 21b
Due to their high σ affinity and selectivity, the γ-fluorinated 3-benzazepines 16b and
1
2
1b were selected for further pharmacological characterization in vitro. The σ1
receptor is found predominantly in the endoplasmic reticulum membrane and in
28,29
mitochondria-associated membranes
Very recently, the σ1 receptor crystal
30,31
structure with various ligands has been solved by Kruse and coworkers.
It plays a
central role in various neurological and neurodegenerative disorders including
alcohol and drug (e.g. cocaine) dependence, depression, anxiety, Alzheimer`s,
32,33,34,35,36,37
Parkinson’s and Huntington’s disease.
σ1 Receptor antagonists (e.g.
38,39,40
S1RA) are able to ameliorate neuropathic pain.
The high density of σ1

17
ACCEPTED MANUSCRIPT
receptors in various human tumor cell lines stimulates the development of σ ligands
1
41,42
for tumor therapy and diagnosis.
Before detailed pharmacological evaluation, the chemical stability of 3-benzazepines
6b and 21b was determined. They were stored at pH 2 (aqueous HCl) and pH 7.4
phosphate buffer). After 24 h, HPLC analysis did not reveal any decomposition
1
(
products of 16b and 21b indicating high chemical stability under these conditions.
At first the selectivity of the σ ligands 16b and 21b over a panel of further targets
1
(receptors and transporters) was determined. As shown in Table 2, affinity towards
δ-opioid (DOR) and κ-opioid (KOR) receptors was not found. However, 3-
benzazepine 21b showed a slight interaction with µ-opioid receptors (MOR), but
could not activate MOR in the cAMP assay. At a concentration of 1 µM, interactions
of the 3-benzazepines 16b and 21b with 5-HT2B and α1A receptors, with serotonin
(SERT), norepinephrine (NET) and dopamine transporters (DAT) could not be
detected. However, 16b and 21b show slight affinity towards hERG channel and 5-
HT1A receptor. Altogether, 3-benzazepines 16b and 21b possess a desirable
selectivity profile, especially compared to lead compound ifenprodil.
Table 2. Affinity of 16b and 21b towards various receptors and transporters.
target
MOR
DOR
KOR
16b
>1000
>1000
>1000
>10000
553
21b
406
a
K [nM]
i
>1000
>1000
>10000
1183
b
EC [nM]
MOR
hERG
SERT
50
c
IC [nM]
50
d
%
inhibition [1 µM]
<50
<50

18
ACCEPTED MANUSCRIPT
NET
DAT
<50
<50
70
<50
<50
72
5
-HT1A
-HT2B
α1A
5
<50
<50
<50
<50
<50
<50
H1
a
b
c
d
K values for opioid receptors.
i
agonistic activity concerning the activation of MOR.
+
inhibition of hERG K channel determined by recording dose/response curves.
Replacement of radioligands at a test compound concentration of 1 µM.
Since animal experiments were planned, metabolism of 3-benzazepines 16b and 21b
was investigated. After incubation of 3-benzazepines 16b and 21b with human,
mouse and rat liver microsomes for 1 h, both test compounds disappeared almost
completely. Obviously, both 3-benzazepines underwent fast metabolism. Additionally,
interactions of 16b and 21b with CYP1A2, 2C9, 2C19, 3A4 and 2D6 were evaluated.
Only CYP2D6 was inhibited by both tertiary amines 16b (55%) and 21b (85 %).
Cytotoxic effects of 16b and 21b were detected in the MTT and NR assays.
5
. In vivo activity of 16b in mechanical allodynia assay
Since methyl ether 16b possesses high σ affinity (K = 6.6 nM) as well as a
1
i
promising selectivity profile (Tables 1 and 2), its antiallodynic activity should be
investigated in an in vivo experiment. It has been shown that σ receptor antagonists
1
40
are useful for the treatment of neuropathic pain such as allodynia. In addition to
preclinical evidence, σ receptor knockout mice did not develop mechanical allodynia
1
43
after intraplantar (ipl) capsaicin injection.
Intraplantar administration of capsaicin provokes protective and defense reflexes up
to 3 min after injection. After this period, hypersensitivity follows to thermal as well as

19
ACCEPTED MANUSCRIPT
44
mechanical stimuli. 30 min before capsaicin administration, 3-benzazepine 16b, the
σ1 receptor antagonist BD-1063 (1-[2-(3,4-dichlorophenyl)ethyl]-4-methylpiperazine,
45
2
3) as positive control as well as the solvent hydroxypropylmethylcellulose (HPMC)
as negative control were applied subcutaneously (sc) to mice. Subsequently,
capsaicin-induced mechanical allodynia was evaluated by stimulation with electronic
46
von-Frey device.
As depicted in Figure 2, 3-benzazepine 16b shows an
antiallodynic effect, which is stronger than the positive control BD-1063 (23). A
statistically significant antiallodynic effect was found at a low dose of 8 mg/kg of 16b,
whereas 32 mg/kg were necessary in case of positive control 23.
Figure 2. Effect of fluorinated 3-benzazepine 16b and reference compound BD-1063
(23) on mechanical allodynia induced by ipl administration of capsaicin in mice.
Animals were treated sc with 16b, 23 or the solvent (HPMC) 30 min prior to capsaicin
injection. Each bar and vertical line represents the mean ± SEM of values obtained in
6
-8 animals. One-way analysis of variance followed by the Bonferroni test was used
to determine statistically significant differences between values obtained in mice
treated with 16b or 23, in comparison with control animals (HPMC-treated, dose 0)
##
value ( P < 0.01) and between the values obtained in mice treated with the same

20
ACCEPTED MANUSCRIPT
dose of 16b and 23 (**P < 0.01).
In order to show that the antiallodynic effect of 16b results from inhibition of σ1
receptors, mice were pretreated with the σ receptor agonist PRE-084 (2-(morpholin-
1
47
4
-yl)ethyl 1-phenylcyclohexane-1-carboxylate, 24, 32 mg/kg) and the experiment
was conducted once more as described above. As shown in Figure 3, the
antiallodynic effect of 3-benzazepine 16b up to a dose of 16 mg/kg was antagonized
completely by PRE-084 (24). However, PRE-084 (24) was not able to antagonize
completely the antiallodynic effect of 3-benzazepine 16b at the highest dose of
3
2 mg/kg. The observed shift to the right of the dose-response curve of 16b proves
unequivocally the correlation between σ1 receptor affinity and the identified
antiallodynic effect.
Figure 3. Effect of fluorinated 3-benzazepine 16b alone and in combination with σ1
receptor agonist PRE-084 (24) on mechanical allodynia induced by ipl administration
of capsaicin to mice. Animals were treated sc with 24 5 min before administration of
1
6b, which was administered sc 30 min before capsaicin. Each bar and vertical line
represents the mean ± SEM of values obtained in 6-8 animals. One-way analysis of
variance followed by the Bonferroni test was used to determine statistically significant

21
ACCEPTED MANUSCRIPT
differences between values obtained in mice treated with 16b or its combination with
##
2
4, in comparison with control animals (dose 0) value ( P < 0.01) and between the
values obtained in mice treated with 16b and 16b + 23 (**P < 0.01).
6
. Conclusion
A series of 3-benzazepines with 4-phenylbutyl substituents in 3-position bearing a F-
atom in either β- or γ-position was synthesized. The impact of the F-atom and the
substitution pattern on affinity and selectivity was evaluated in various
pharmacological assays. Phenolic 3-benzazepine 17b bearing the F-atom in
γ-position reveals the most promising GluN2B affinity (K = 16 nM). In general, 3-
i
benzazepines with benzylic OH moiety show reduced GluN2B affinity compared to 3-
benzazepines without OH-moiety in 1-position. Removal of benzylic OH moiety
resulted in increased GluN2B affinity, but loss of selectivity over the σ receptor.
2
Unfortunately, the 3-benzazepines do not exhibit selectivity over σ receptors.
1
On the other side, some of the fluorinated 3-benzazepines prefer σ receptor over
1
GluN2B receptors with a selectivity factor of up to 18-fold for 16b. The
pharmacological properties and selectivity profile of 16b were further evaluated in in
vitro studies. In a mouse model, 3-benzazeine 16b showed high antiallodynic activity,
which was attributed unequivocally to σ receptor antagonism.
1
7
7
. Experimental
.1. Chemistry, General
Unless otherwise noted, moisture sensitive reactions were conducted under dry
nitrogen. Acetonitrile was dried over molecular sieves (3 Å). CH Cl was distilled over
2
2
CaH . THF was distilled over sodium/benzophenone. Thin layer chromatography
2
(tlc): Silica gel 60 F254 plates (Merck). Flash chromatography (fc): Silica gel 60, 40–64

22
ACCEPTED MANUSCRIPT
µm (Merck); parentheses include: diameter of the column (d), fraction size (v), eluent,
R value. Melting point: Melting point apparatus Mettler Toledo MP50 Melting Point
f
System, uncorrected. MS: microTOF-Q II (Bruker Daltonics); APCI, atmospheric
pressure chemical ionization. IR: FT-IR spectrophotometer MIRacle 10 (Shimadzu)
equipped with ATR technique. Nuclear magnetic resonance (NMR) spectra were
1
13
recorded on Agilent 600-MR (600 MHz for H, 151 MHz for C) or Agilent 400-MR
1
13
spectrometer (400 MHz for H, 101 MHz for
tetramethylsilane and measured referring to CHCl (
C);
δ
in ppm related to
1
δ
= 7.26 ppm ( H NMR) and
δ
=
3
13
1
13
7
7.2 ppm ( C NMR)) and CHD OD (δ = 3.31 ppm ( H NMR) and δ = 49.0 ppm ( C
2
19
NMR)); trichlorofluoromethane (CCl F) was used as reference compound in F NMR
3
spectroscopy; coupling constants are given with 0.5 Hz resolution; the assignments
13
1
of C and H NMR signals were supported by 2-D NMR techniques where
necessary. HPLC: Merck Hitachi Equipment; UV detector: L-7400; autosampler: L-
®
7
200; pump: L-7100; degasser: L-7614; column: LiChrospher 60 RP-select B (5
®
µm); LiChroCART 250-4 mm cartridge; flow rate: 1.0 mL/min; injection volume: 5.0
µL; detection at λ = 210 nm; solvents: A: water with 0.05% (v/v) trifluoroacetic acid; B:
acetonitrile with 0.05% (v/v) trifluoroacetic acid: gradient elution: (A%): 0-4 min: 90%,
4
-29 min: 90 → 0%, 29-31 min: 0%, 31-31.5 min: 0 → 90%, 31.5-40 min: 90%. The
purity of all compounds was determined by this method. The purity of all test
compounds is higher than 95%, unless otherwise noted.
7
7
.3. Synthetic procedures
.3.1. (2-Fluoro-4-phenylbutyl) 4-methylbenzenesulfonate (7a)
®
(
Diethylamino)difluorosulfonium tetrafluoroborate (XtalFluor E , 180 mg, 0.79 mmol,
1
.5 eq) and triethylamine trihydrofluoride (427 µL, 2.62 mmol, 5.0 eq) were dissolved
in CH Cl (6 mL). The solution was cooled to -78 °C. Alcohol 5a (168 mg, 0.52 mmol,
2
2

23
ACCEPTED MANUSCRIPT
1
.0 eq) was added and the mixture was stirred for 1 h at -78 °C and overnight at rt. A
saturated solution of NaHCO (25 mL) was added and the mixture was stirred for
3
2
0 min. The aqueous layer was extracted with CH Cl (3 x 10 mL), then the
2 2
combined organic layers were dried (Na SO ), filtered and the solvent was removed
2
4
in vacuo. The crude product was purified by fc (d = 2 cm, l = 20 cm, V = 7 mL,
cyclohexane:ethyl acetate = 9:1). Colorless oil, yield 122 mg (0.38 mmol, 72 %).
1
C H O S (322.4). R = 0.37 (cyclohexane:ethyl acetate = 8:2). H NMR (400 MHz,
17
19
3
f
CDCl ): δ [ppm] = 1.73-2.07 (m, 2H, PhCH CH CHF), 2.45 (s, 3H, CH ), 2.62-2.82
3
2
2
3
(
m, 2H, PhCH CH CHF), 4.04-4.18 (m, 2H, CHFCH OTs), 4.53-4.70 (dm,
2 2 2
J = 48.5 Hz, 1H, CHF), 7.12-7.23 (m, 3H, 2-Hphenyl, 4-Hphenyl, 6-Hphenyl), 7.25-7.31 (m,
2
6
1
1
H, 3-Hphenyl, 5-Hphenyl), 7.33-7.37 (m, 2H, 3-Htosyl, 5-Htosyl), 7.77-7.81 (m, 2H, 2-Htosyl,
13
-Htosyl). C NMR (101 MHz, CDCl ): δ [ppm] = 21.8 (1C, CH ), 30.8 (d, J = 4.4 Hz,
3
3
C, PhCH CH CHF), 32.7 (d, J = 20.7 Hz, 1C, PhCH CH CHF), 70.7 (d, J = 23.6 Hz,
2
2
2
2
C, CHFCH OTs), 89.6 (d, J = 175.5 Hz, 1C, CHF), 126.4 (1C, C-4phenyl), 128.1 (2C,
2
C-2tosyl, C-6tosyl), 128.5 (2C, C-2phenyl, C-6phenyl), 128.7 (2C, C-3phenyl, C-5phenyl), 130.1
2C, C-3tosyl, C-5tosyl), 132.8 (1C, C-1tosyl), 140.4 (1C, C-1phenyl), 145.2 (1C, C-4tosyl).
(
19
F NMR (376 MHz, CD OD): δ [ppm] = -189.0 (m, 1F). HR-MS (APCI): m/z =
3
+
3
23.1114 (calcd. 323.1112 for C H FO S [M+H] ). Purity (HPLC): 98.5 %,
17
20
3
-1
t = 24.3 min. IR: ṽ [cm ] = 2951 (C-Haliph.), 1597, 1493 (C=Carom.), 1173, 1362
R
(
O-SO ), 814 (disubst. arom.), 748, 698 (monosubst. arom.).
2
7
.3.2.
7-(Benzyloxy)-3-(2-fluoro-4-phenylbutyl)-2,3,4,5-tetrahydro-1H-3-
benzazepin-1-ol (8a)
Secondary amine 6 (104 mg, 0.39 mmol, 1.0 eq) was dissolved in CH CN (8 mL).
3
After addition of tosylate 7a (150 mg, 0.47 mmol, 1.2 eq) in CH CN (3 mL) and
3
K CO (427 mg, 3.09 mmol, 8.0 eq), the suspension was heated to reflux for 3 d. At
2
3

24
ACCEPTED MANUSCRIPT
rt, K CO was filtered off, washed with CH CN and the solvent was removed in
2
3
3
vacuo. The residue was purified by fc three times (1. d = 2 cm, l = 20 cm, V = 7 mL,
cyclohexane:ethyl acetate = 8:2 + 1 % N,N-dimethylethanamine; 2. d = 1 cm,
l = 22 cm, V = 4 mL, CH Cl :CH OH = 99:1
+
1 % N,N-dimethylethanamine;
2
2
3
3
. d = 1 cm, l = 15 cm, V = 4 mL, CH Cl :ethyl acetate = 97:3 + 1 % NH ). Colorless
2
2
3
oil, yield 22.0 mg (0.05 mmol, 14 %). C H FNO
(419.5). R = 0.24
27
30
2
f
1
(
cyclohexane:ethyl acetate = 7:3+ 1 % N,N-dimethylethanamine). H NMR (400 MHz,
CDCl ): δ [ppm] = 1.78-2.08 (m, 2H, NCH CHFCH CH Ph), 2.58-3.09 (m, 8H, 2-H,
3
2
2
2
2
1
1
5
x 4-H, 5-H, NCH CHFCH CH Ph), 3.16-3.32 (m, 2H, 2-H, 5-H), 4.57-4.82 (m, 2H,
2 2 2
-H, CHF), 5.03 (s, 2H, OCH Ph), 6.71-6.77 (m, 2H, 6-H, 8-H), 7.11 (d, J = 8.9 Hz,
2
H, 9-H), 7.18-7.24 (m, 3H, 2-Hphenyl, 4-Hphenyl, 6-Hphenyl), 7.28-7.44 (m, 7H, 3-Hphenyl
,
-Hphenyl, 2-Hbenzyl, 3-Hbenzyl, 4-Hbenzyl, 5-Hbenzyl, 6-Hbenzyl). A signal for the OH proton is
13
not seen in the spectrum. C NMR (101 MHz, CDCl ): δ [ppm] = 31.3, 31.4 (d,
3
J = 4.6 Hz, 1C, NCH CHFCH CH Ph), 35.27, 35.32 (d, J = 20.8 Hz, 1C,
2
2
2
NCH CHFCH CH Ph), 36.95, 37.04 (1C, C-5), 56.6, 56.9 (1C, C-4), 61.5, 61.7 (1C,
2
2
2
C-2), 63.77, 63.79 (d, J = 20.8 Hz, 1C, NCH CHFCH CH Ph), 70.1 (1C, OCH Ph),
2
2
2
2
7
1
2.6, 72.7 (1C, C-1), 91.6, 91.9 (d, J = 171.3 Hz, 1C, CHF), 111.4 (1C, C-8), 117.62,
17.63 (1C, C-6), 126.3 (1C, C-4phenyl), 127.6 (2C, C-2benzyl, C-6benzyl), 128.1 (1C,
C-4benzyl), 128.6 (2C, C-2phenyl, C-6phenyl), 128.68 (2C, C-3phenyl, C-5phenyl), 128.70 (2C,
C-3benzyl, C-5benzyl), 129.8, 129.9 (1C, C-9), 135.6, 135.7 (1C, C-9a), 137.2 (1C,
C-1benzyl), 141.14, 141.18 (1C, C-5a), 141.20 (1C, C-1phenyl), 158.346, 158.354 (1C,
(
19
C-7). Ratio of diastereomers is approx. 1:1.
F NMR (376 MHz, CDCl3):
δ [ppm] = -183.5, -182.7 (m, 1F). HR-MS (APCI): m/z = 420.2321 (calcd. 420.2333
+
-1
for C H FNO [M+H] ). Purity (HPLC): 91.6 %, t = 20.8 min. IR: ṽ [cm ] = 3422
27
31
2
R
(
OH), 3028 (C-H ), 2940 (C-Haliph.), 1605, 1578, 1497 (C=Carom.), 1454 (C-Haliph.),
aryl
1
234 (C-O).

25
ACCEPTED MANUSCRIPT
7
.3.3.
7-(Benzyloxy)-3-(3-fluoro-4-phenylbutyl)-2,3,4,5-tetrahydro-1H-3-
benzazepin-1-ol (8b)
Secondary amine 6(125 mg, 0.46 mmol, 1.0 eq) was dissolved in CH CN (5 mL).
3
After addition of tosylate 7b (165 mg, 0.51 mmol, 1.2 eq) in CH CN (3 mL) and
3
K CO (427 mg, 3.09 mmol, 8.0 eq), the suspension was heated to reflux for 3 d. At
2
3
rt, K CO was filtered off, washed with CH CN and the solvent was removed in
2
3
3
vacuo. The residue was purified by fc (d = 2 cm, l = 20 cm, V = 7 mL,
cyclohexane:ethyl acetate = 7:3 + 1.5 % N,N-dimethylethanamine). Pale yellow resin,
yield 160 mg (0.38 mmol, 82 %). C H FNO (419.5). R = 0.22 (cyclohexane:ethyl
27
30
2
f
1
acetate = 6:4 + 1 % N,N-dimethylethanamine). H NMR (400 MHz, CDCl ): δ
3
[
ppm] = 1.73-1.93 (m, 2H, NCH CH CHFCH Ph), 2.39-3.30 (m, 10H, 2 x 2-H,
2 2 2
2
x 4-H, 2 x 5-H, NCH CH CHFCH Ph), 4.58, 4.59 (d, J = 6.7 Hz, 1H, 1-H), 4.70-4.90
2 2 2
(dm, J = 49.1 Hz, 1H, CHF), 5.03 (s, 2H, OCH Ph), 6.71-6.75 (m, 2H, 6-H, 8-H), 7.10
2
(
d, J = 9.0 Hz, 1H, 9-H), 7.21-7.43 (m, 10H, 2-Hphenyl, 3-Hphenyl, 4-Hphenyl, 5-Hphenyl
,
6
-Hphenyl, 2-Hbenzyl, 3-Hbenzyl, 4-Hbenzyl, 5-Hbenzyl, 6-Hbenzyl). A signal for the OH proton is
13
not seen in the spectrum. C NMR (101 MHz, CDCl ): δ [ppm] = 32.40, 32.42 (d,
3
J = 20.6 Hz, 1C, NCH CH CHFCH Ph), 36.8, 36.9 (1C, C-5), 41.90, 41.94 (d,
2
2
2
J = 21.5 Hz, 1C, NCH CH CHFCH Ph), 55.76, 55.78 (d, J = 3.8 Hz, 1C,
2
2
2
NCH CH CHFCH Ph), 56.0, 56.5 (1C, C-4), 60.7, 61.1 (1C, C-2), 70.1 (1C,
2
2
2
OCH Ph), 72.4, 72.5 (1C, C-1), 92.96, 92.98 (d, J = 171.4 Hz, 1C, CHF), 111.37,
2
1
11.38 (1C, C-8), 117.60, 117.62 (1C, C-6), 126.9 (1C, C-4phenyl), 127.6 (2C, C-2benzyl,
C-6benzyl), 128.1 (2C, C-3phenyl, C-5phenyl), 128.66, 128.68 (1C, C-4benzyl), 128.7 (2C, C-
3phenyl, C-5phenyl), 129.5 (2C, C-2phenyl, C-6phenyl), 129.80, 129.82 (1C, C-9), 135.72,
1
1
35.74 (1C, C-9a), 137.00, 137.02 (d, J = 5.4 Hz, 1C, C-1phenyl), 137.2 (1C, C-1benzyl),
41.18, 141.20 (1C, C-5a), 158.33, 158.34 (1C, C-7). Ratio of diastereomers is 1:1.

26
ACCEPTED MANUSCRIPT
19
F NMR (376 MHz, CDCl ): δ [ppm] = -179.9 (m, 0.5F), -179.3 (m, 0.5F). HR-MS
3
+
(
APCI): m/z = 420.2362 (calcd. 420.2333 for C H FNO [M+H] ). Purity (HPLC):
27
31
2
-
1
9
1
8.7 %, t = 21.2 min. IR: ṽ [cm ] = 3395 (OH), 3028 (C-H ), 2940 (C-Haliph.), 1609,
R
aryl
582, 1497 (C=Carom.), 1454 (C-Haliph.), 1246 (C-O).
7
.3.4.
3-(3-Fluoro-4-phenylbutyl)-2,3,4,5-tetrahydro-1H-3-benzazepine-1,7-diol
(9b)
8
b (100 mg, 0.24 mmol, 1.0 eq) was dissolved in CH OH (4 mL) and the solution was
3
added to a suspension of Pd/C (10 %, 20.0 mg) in CH OH (8 mL). The reaction
3
mixture was stirred for 3 h at rt under H atmosphere (1 bar, balloon). The catalyst
2
®
was removed by filtration over Celite and the solvent was removed in vacuo. The
residue was purified by fc (d = 2 cm, l = 18 cm, V = 7 mL, CH Cl :CH OH = 97:3 +
2
2
3
1
.5 % N,N-dimethylethanamine). Yellow oil, yield 67.5 mg (0.21 mmol, 86 %).
C H FNO (329.4). R = 0.30 (CH Cl :CH OH = 96:4 1 % N,N-dimethyl-
+
20
24
2
f
2
2
3
1
ethanamine). H NMR (400 MHz, CDCl ): δ [ppm] = 1.70-1.91 (m, 2H,
3
NCH CH CHFCH Ph), 2.34-2.62 (m, 3H, 2-H, 4-H, 5-H), 2.65-2.81 (m, 2H,
2
2
2
NCH CH CHFCH Ph), 2.83-3.24 (m, 5H, 2-H, 4-H, 5-H, NCH CH CHFCH Ph), 4.54,
2
2
2
2
2
2
4
2
6
.55 (d, J = 6.6 Hz, 1H, 1-H), 4.68-4.89 (dm, J = 48.8 Hz, 1H, CHF), 6.55-6.59 (m,
H, 6-H, 8-H), 7.00 (d, J = 8.6 Hz, 1H, C-9), 7.20-7.25 (m, 3H, 2-Hphenyl, 4-Hphenyl
-Hphenyl), 7.29-7.34 (m, 2H, 3-Hphenyl, 5-Hphenyl). Signals for the OH protons are not
,
13
seen in the spectrum. C NMR (101 MHz, CDCl ): δ [ppm] = 32.45, 32.48 (d,
3
J = 20.6 Hz, 1C, NCH CH CHFCH Ph), 36.8, 36.9 (1C, C-5), 41.91, 41.95 (d,
2
2
2
J = 21.5 Hz, 1C, NCH CH CHFCH Ph), 55.75, 55.78 (d, J = 3.6 Hz, 1C,
2
2
2
NCH CH CHFCH Ph), 56.1, 56.6 (1C, C-4), 60.7, 61.2 (1C, C-2), 72.6, 72.7 (1C,
2
2
2
C-1), 93.00, 93.04 (d, J = 171.4 Hz, 1C, CHF), 112.6 (1C, C-8), 117.71, 117.74 (1C,
C-6), 126.8 (1C, C-4phenyl), 128.66, 128.67 (2C, C-3phenyl, C-5phenyl), 129.5 (2C, C-

27
ACCEPTED MANUSCRIPT
2phenyl, C-6phenyl), 130.02, 130.04 (1C, C-9), 134.94, 134.95 (1C, C-9a), 137.03,
1
37.04 (d, J = 5.4 Hz, 1C, C-1phenyl), 141.35, 141.37 (1C, C-5a), 155.75, 155.77 (1C,
19
C-7). Ratio of diastereomers is approx. 1:1. F NMR (376 MHz, CDCl ): δ [ppm] = -
3
1
79.6 (m, 0.5F), 179.3 (m, 0.5F). HR-MS (APCI): m/z = 330.1865 (calcd. 330.1864
+
-1
for C H FNO [M+H] ). Purity (HPLC): 88.5 %, t = 15.8 min. IR: ṽ [cm ] = 3244
20
25
2
R
(
OH), 3028 (C-H ), 2924 (C-Haliph.), 1609, 1586, 1497 (C=Carom.), 1454 (C-Haliph.),
aryl
1
227 (C-O).
7
1
.3.5. 3-(2-Fluoro-4-phenylbutyl)-7-methoxy-2,3,4,5-tetrahydro-1H-3-benzazepin-
-ol (12a) and 3-(1-Fluoro-3-phenylpropyl)-8-methoxy-5,6-dihydro-1H,3H-1,4-
methanobenzo-1,3-oxazocine (13)
Procedure 1: Secondary amine 10 (39.0 mg, 0.20 mmol, 1.0 eq) was dissolved in
CH CN (5 mL). After addition of tosylate 7a (72.0 mg, 0.22 mmol, 1.1 eq) in CH CN
3
3
(
5 mL) and K CO (223 mg, 1.61 mmol, 8.0 eq), the suspension was heated to reflux
2 3
for 4 d. At rt, K CO was filtered off, washed with CH CN and the solvent was
2
3
3
removed in vacuo. The residue was purified by fc (d = 1 cm, l = 18 cm, V = 4 mL,
cyclohexane:ethyl acetate = 7:3) to give tertiary amine 12a. Yellow oil, yield 4.5 mg
(
0.01 mmol, 7 %). C H FNO (343.4). R = 0.30 (cyclohexane:ethyl acetate = 1:1).
21 26 2 f
1
H NMR (400 MHz, CDCl ): δ [ppm] = 1.78-2.04 (m, 2H, NCH CHFCH CH Ph), 2.57-
3
2
2
2
2
.94 (m, 7H, 2-H, 4-H, 5-H, NCH CHFCH CH Ph), 3.00-3.09 (m, 1H, 4-H), 3.15-3.32
2 2 2
(
m, 2H, 2-H, 5-H), 3.78 (s, 3H, OCH ), 4.58, 4.60 (d, J = 6.4 Hz, 1H, 1-H), 4.61-4.81
3
(dm, J = 49.9 Hz, 1H, CHF), 6.63 (d, J = 2.6 Hz, 1H, 6-H), 6.66 (dd, J = 8.1/2.6 Hz,
1
6
H, 8-H), 7.11 (d, J = 8.1 Hz, 1H, 9-H), 7.18-7.24 (m, 3H, 2-Hphenyl, 4-Hphenyl
,
-Hphenyl), 7.28-7.33 (m, 2H, 3-Hphenyl, 5-Hphenyl). A signal for the OH proton is not seen
13
in the spectrum. C NMR (101 MHz, CDCl ): δ [ppm] = 31.34, 31.36 (d, J = 4.4 Hz,
3
1
C, NCH CHFCH CH Ph), 35.27, 35.33 (d, J = 20.8 Hz, 1C, NCH CHFCH CH Ph),
2 2 2 2 2 2

28
ACCEPTED MANUSCRIPT
3
7.1, 37.2 (1C, C-5), 55.4 (1C, OCH ), 56.7, 56.9 (d, J = 1.4 Hz, 1C, C-4), 61.6, 61.7
3
(
d, J = 1.3 Hz, 1C, C-2), 63.80, 63.81 (d, J = 20.7 Hz, 1C, NCH CHFCH CH Ph),
2 2 2
7
1
2.7, 72.8 (1C, C-1), 91.7, 92.0 (d, J = 171.0 Hz, 1C, CHF), 110.4 (1C, C-8), 116.74,
16.76 (1C, C-6), 126.3 (1C, C-4phenyl), 128.6 (2C, C-2phenyl, C-6phenyl), 128.7 (2C, C-
3phenyl, C-5phenyl), 129.8, 129.9 (1C, C-9), 135.4, 135.5 (1C, C-9a), 141.1, 141.2 (1C,
19
C-5a), 141.2 (d, J = 0.8 Hz, 1C, C-1phenyl), 159.10, 159.11 (1C, C-7). F NMR
(
376 MHz, CDCl ): δ [ppm] = -183.6 (m, 0.5F), -182.7 (m, 0.5F). HR-MS (APCI):
3
+
m/z = 344.2007 (calcd. 344.2020 for C H FNO [M+H] ). Purity (HPLC): 90.3 %,
21
27
2
-1
t = 17.6 min. IR: ṽ [cm ] = 3399 (OH), 3024 (C-H ), 2936 (C-Haliph.), 1609, 1582,
R
aryl
1
497 (C=Carom.), 1454 (C-Haliph.), 1258 (C-O). Procedure 2: Secondary amine 10
300 mg, 1.55 mmol, 1.0 eq) was added to a stirred solution of freshly prepared
aldehyde 11a (excess) in CH Cl (10 mL). After addition of NaBH(OAc) (658 mg,
(
2
2
3
3
.11 mmol, 2.0 eq), the reaction mixture was stirred for 15 h at rt. A saturated
solution of NaHCO (10 mL), water (5 mL) and CH Cl (5 mL) were added, the
3
2
2
organic layer was separated and the aqueous layer was extracted with CH Cl
2
2
(
(
(
2 x 5 mL) and ethyl acetate (2 x 5 mL). The combined organic layers were dried
Na SO ), filtered and concentrated in vacuo. The crude product was purified by fc
2
4
d = 3 cm,
% N,N-dimethylethanamine) to give 12a and two pairs of diastereomers of 13 in
separate fractions. 12a: Yellow oil, yield 62.0 mg (0.18 mmol, 12 %). C H FNO
l = 22 cm,
V = 20 mL,
cyclohexane:ethyl
acetate = 8:2
+
1
21
26
2
(
343.4). R = 0.30 (cyclohexane:ethyl acetate = 1:1). The spectroscopic data are
f
identical with those of 12a from procedure 1. HR-MS (APCI): m/z = 344.1981 (calcd.
+
3
44.2020 for C H FNO [M+H] ). Purity (HPLC): 90.3 %, t = 17.6 min. 13
21
27
2
R
(
diastereomer 1): Yellow oil, yield 182 mg (0.53 mmol, 34 %). C H FNO (341.4).
21 24 2
1
R = 0.62 (cyclohexane:ethyl acetate = 8:2+ 1 % N,N-dimethylethanamine). H NMR
f
(
400 MHz, CDCl ): δ [ppm] = 1.93-2.20 (m, 2H, CHFCH CH Ph), 2.38-2.49 (m, 1H,
3 2 2

29
ACCEPTED MANUSCRIPT
6
3
-H), 2.70-2.96 (m, 2H, CHFCH CH Ph), 2.99-3.63 (m, 5H, 2 x 5-H, 6-H, 2 x 11-H),
2
2
.76 (s, 3H, OCH ), 4.22-4.40 (dm, J = 49.4 Hz, 0.7H, CHF), 4.28-4.46 (dm,
3
J = 49.2 Hz, 0.3H, CHF), 4.76 (d, J = 6.9 Hz, 0.7H, 1-H), 4.89 (d, J = 7.1 Hz, 0.3H,
1
6
2
-H), 4.91-4.98 (m, 1H, 3-H), 6.60 (dd, J = 8.2/2.6 Hz, 1H, 8-H), 6.64-6.67 (m, 1H,
-H), 6.97 (d, J = 8.2 Hz, 0.7H, 9-H), 6.98 (d, J = 8.2, 0.3H, 9-H), 7.17-7.25 (m, 3H,
-Hphenyl, 4-Hphenyl, 6-Hphenyl), 7.27-7.32 (m, 2H, 3-Hphenyl, 5-Hphenyl). Ratio of
13
diastereomers is 7:3. C NMR (101 MHz, CDCl ): δ [ppm] = 31.15 (d, J = 4.0 Hz,
3
0
.3C, CHFCH CH Ph), 31.19 (d, J = 4.0 Hz, 0.7C, CHFCH CH Ph), 33.4 (d,
2 2 2 2
J = 20.5 Hz, 0.3C, CHFCH CH Ph), 33.8 (d, J = 20.6 Hz, 0.7C, CHFCH CH Ph),
2
2
2
2
3
4.9 (0.3C, C-6), 35.2 (0.7C, C-6), 55.35 (0.7C, OCH ), 55.37 (0.3C, OCH ), 55.6
3 3
(0.7C, C-5), 55.8 (0.3C, C-5), 60.1 (d, J = 1.3 Hz, 0.7C, C-11), 61.0 (d, J = 2.7 Hz,
0
9
.3C, C-11), 80.0 (0.7C, C-1), 80.5 (0.3C, C-1), 91.0 (d, J = 175.5 Hz, 0.7C, CHF),
3.6 (d, J = 175.5 Hz, 0.3C, CHF), 96.8 (d, J = 23.0 Hz 0.7C, C-3), 96.9 (d,
J = 21.9 Hz, 0.3C, C-3), 110.1 (0.3C, C-9), 110.2 (0.7C, C-9), 117.53 (0.3C, C-7),
17.54 (0.7C, C-7), 126.12 (0.7C, C-4phenyl), 126.14 (0.3C, C-4phenyl), 128.57 (0.7C,
C-10), 128.60 (0.3C, C-10), 128.60 (2 x 0.7C, C-3phenyl, C-5phenyl), 128.57 (2 x 0.3C,
C-3phenyl, C-5phenyl), 128.67 (2 x 0.7C, C-2phenyl, C-6phenyl), 128.69 (2 x 0.3C, C-2phenyl
1
,
C-6phenyl), 135.1 (0.7C, C-10a), 135.4 (0.3C, C-10a), 140.3 (0.3C, C-6a), 140.4 (0.7C,
C-6a), 141.4 (0.3C, C-1phenyl), 141.5 (0.7C, C-1phenyl), 158.8 (0.3C, C-8), 158.9 (0.7C,
19
C-8). F NMR (376 MHz, CDCl ): δ [ppm] = -196.3 (m, 0.3F), -188.9 (m, 0.7F). HR-
3
+
MS (APCI): m/z = 342.1843 (calcd. 342.1864 for C H FNO [M+H] ). Purity (HPLC):
21
25
2
-1
4
6.1 % t = 9.2 min; 52.1 %, t = 15.1 min. IR: ṽ [cm ] = 3024 (C-H ), 2932 (C-
R R aryl
Haliph.), 1605, 1578, 1497 (C=Carom.), 1454 (C-Haliph.), 1254 (C-O). 13 (diastereomer
): Yellow oil, yield 148 mg (0.43 mmol, 28 %). C H FNO (341.4). R = 0.50
2
21
24
2
f
1
(
cyclohexane:ethyl acetate = 8:2+ 1 % N,N-dimethylethanamine). H NMR (400 MHz,
CDCl ): δ [ppm] = 1.91-2.20 (m, 2H, CHFCH CH Ph), 2.39-2.49 (m, 1H, 6-H), 2.70-
3
2
2