Discovery of disubstituted piperidines and homopiperidines as potent dual NK1 receptor antagonists-serotonin reuptake transporter inhibitors for the treatment of depression

Article abstract of DOI:10.1016/j.bmc.2013.02.010

This report describes the synthesis, structure-activity relationships and activity of piperidine, homopiperidine, and azocane derivatives combining NK₁ receptor (NK₁R) antagonism and serotonin reuptake transporter (SERT) inhibition. Our studies culminated in the discovery of piperidine 2 and homopiperidine 8 as potent dual NK₁R antagonists-SERT inhibitors. Compound 2 demonstrated significant activity in the gerbil forced swimming test, suggesting that dual NK₁R antagonists-SERT inhibitors may be useful in treating depression disorders.

Full text of DOI:10.1016/j.bmc.2013.02.010

Bioorganic & Medicinal Chemistry 21 (2013) 2217–2228
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Discovery of disubstituted piperidines and homopiperidines
as potent dual NK1 receptor antagonists–serotonin reuptake
transporter inhibitors for the treatment of depression
a
b
b
b
b
Yong-Jin Wu a, , Huan He , Robert Bertekap , Ryan Westphal , Snjezana Lelas , Amy Newton ,
⇑
Tanya Wallace b, Matthew Taber b, Carl Davis c, John E. Macor a, Joanne Bronson a
a Department of Neuroscience Chemistry, Research and Development, Bristol-Myers Squibb Company, 5 Research Parkway, Wallingford, CT 06492, USA
b Department of Neuroscience Biology, Research and Development, Bristol-Myers Squibb Company, 5 Research Parkway, Wallingford, CT 06492, USA
c Department of Preclinical Candidate Optimization, Research and Development, Bristol-Myers Squibb Company, 5 Research Parkway, Wallingford, CT 06492, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
This report describes the synthesis, structure–activity relationships and activity of piperidine, homopi-
peridine, and azocane derivatives combining NK1 receptor (NK1R) antagonism and serotonin reuptake
transporter (SERT) inhibition. Our studies culminated in the discovery of piperidine 2 and homopiperi-
dine 8 as potent dual NK1R antagonists-SERT inhibitors. Compound 2 demonstrated significant activity
in the gerbil forced swimming test, suggesting that dual NK1R antagonists-SERT inhibitors may be useful
in treating depression disorders.
Received 7 January 2013
Revised 1 February 2013
Accepted 11 February 2013
Available online 19 February 2013
Keywords:
Ó 2013 Elsevier Ltd. All rights reserved.
NK1 receptor (NK1R) antagonist
Serotonin reuptake transporter (SERT)
inhibitor
Dual NK1 receptor antagonists–serotonin
reuptake transporter inhibitor
1. Introduction
antidepressant effect. Indeed, a beneficial effect on the onset of ac-
tion of SSRIs has been observed in the clinical studies with co-
Since their introduction in the 1980s, selective serotonin reup-
take inhibitors (SSRIs) such as paroxetine,1 fluoxetine and citalo-
pram (Chart 1) have enjoyed tremendous clinical and commercial
success due to their improved safety profile over the first-genera-
tion tricyclic antidepressants like imipramine (Chart 1).2 Neverthe-
less, they still display several side effects including gastrointestinal
distress, anxiety, insomnia, weight gain and sexual dysfunction.
Like other currently approved antidepressants, SSRIs also suffer
from slow onset of action, and as a result, a significant number of
depressed patients do not show signs of mood improvement until
4–3 weeks after the initial treatment.3–5 The delay of therapeutic
effects of SSRIs is believed to result from activation of the inhibi-
tory role of serotonin 1A (5-HT1A) autoreceptors. Upon administra-
tion of an SSRI, some of the increased synaptic serotonin (5-HT)
acting at 5-HT1A receptors reduces the firing of serotonergic neu-
rons, resulting in a muted increase of synaptic 5-HT. After initial
desensitization of the 5-HT1A autoreceptors with repeated SSRI
treatment, the serotonergic neurons resume normal firing, allow-
ing an increase of synaptic 5-HT, and thus generating a therapeutic
administration of 5-HT1A antagonists.6,7 Moreover, this combina-
tion therapy has resulted in significant improvements among
SSRI-resistant patients.6,7
Another potential combination strategy might involve neuroki-
nin 1 receptor (NK1R) antagonists8–12 and SSRIs. NK1R antagonists
alone may not be sufficient in treating depression in humans. For
example, aprepitant, an NK1R antagonist currently used in combi-
nation with other antiemetics to help prevent the acute and de-
layed nausea and vomiting associated with emetogenic cancer
chemotherapies, showed some antidepressant effect in early clini-
cal studies,13 but its development for the treatment of depression
was later discontinued after the compound’s effects failed to sepa-
rate from placebo effects in phase III trials.14 Since NK1R antago-
nists indirectly modulate 5-HT function15,16 and attenuate
presynaptic 5-HT1A receptor function, NK1 receptor antagonism
may augment the antidepressant activity of SSRIs. Consistent with
this hypothesis, Bourin et al. demonstrated that GR205171
(Chart 1), a selective and brain penetrant NK1R antagonist, selec-
tively potentiated the antidepressant activity of subactive doses
of two SSRIs, citalopram and paroxetine, in the mouse forced
swimming test (FST).17 Thus, combination therapy of SSRIs with
NK1R antagonists has potential as a new treatment for depression
⇑
Corresponding author. Tel.: +1 203 677 7485; fax: +1 203 677 7702.
E-mail address: yong-jin.wu@bms.com (Y.-J. Wu).
0968-0896/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmc.2013.02.010

2218
Y.-J. Wu et al. / Bioorg. Med. Chem. 21 (2013) 2217–2228
F
Ph
CO2Me
Me
i
ii
Ph
v
CO2Me
Ph
iii
CO2Me
N
F
O
Me
O
CF3
22
23
O
O
O
OH
Me
N
N
N
H
iv
Ph
Ph
Ph
H
O
Ar
O
Ar
Ar
Citalopram (CelexaTM
)
Paroxetine (PaxilTM
)
Fluoxetine (ProzacTM
Lilly
)
Forest
SmithKline Beecham
HO
25
24
27
26
OMe
H
O
CF3
Ph
Ph
N
O
Ar
O
Ph
O
vi
O
Ar
vii
N
N
+
N
N
N
HN
N
H
O
N
H
28
5
O
Me
N
viii
GR-205171
Glaxo Wellcome
.
Me
Imipramine
O
Ph
O
Ar
CF3
Br
O
HN
N
6
N
H
Br
CF3
N
H
Scheme 1. Reagents and conditions: (i) allyl bromide, NaH, DMF, rt, 100%; (ii)
Grubbs I catalyst, CH2Cl2, reflux, 91%; (iii) LiAlH4, ether, 0 °C, 100%; (iv) NaH,
ArCH2Br, DMF, rt, 75%; (v) BH3ꢀTHF, 0 °C; H2O2, NaOH, rt, 50%; (vi) PCC, CH2Cl2, rt,
100%; (vii) sodium azide, trifluoroacetic acid, benzene, 65 °C, 29% (28) and 26% (5);
(viii) BH3ꢀTHF, THF, 65 °C; 1 N HCl, 65 °C, 67%. Ar: 3,5-bis(trifluoromethyl)phenyl.
1
2
Chart 1.
with more rapid onset of action and improved efficacy. However,
combination therapy with two compounds can have potential dis-
advantages, including challenges with differing PK profiles of the
individual components, enhanced risk of drug–drug interactions,
increased number or intensity of side effects, potentiation of
dose-related or idiosyncratic side effects and pharmacokinetic
interactions. An attractive alternative to combination therapy of
two different molecules would be a single compound that has both
serotonin reuptake transporter (SERT) inhibition and NK1R antago-
nism. In 2002, Ryckmans et al. described a series of benzyloxy-
phenethyl piperazine derivatives as dual NK1R antagonists-SERT
inhibitors.18,19 One of the best compounds in the series, (S)-1-(2-
(3,5-dibromobenzyloxy)-1-phenylethyl)piperazine (1, Chart 1),
demonstrated oral activity in an animal model of depression sensi-
tive to both mechanisms.
Ph
Ph
O
CN
CN
Ph
O
CHO
Ph
O
OH
iv
i
ii
iii
O
O
O
O
31
32
30
Ph
29
Ph
O
Ph
Ar
Ar
Ar
O
O
O
vi
v
O
O
ix
OH
35
33
Ph
34
Ph
Ar
Ar
O
O
vii, viii
We also undertook a program to identify dual NK1R antago-
nists-SERT inhibitors. We selected a known potent NK1R antago-
36
7
nist,
4-((3,5-bis(trifluoromethyl)-benzyloxy)-methyl)-4-
phenylpiperidine (2, Chart 1),14 as one of the lead compounds for
our medicinal chemistry program because of its structural similar-
ity to known SSRIs such as paroxetine. We sought to add SERT
inhibitory activity by judiciously incorporating key features of
SSRIs into the molecule, while maintaining NK1R antagonism. It
was in this context that we discovered that 2 was not only a potent
NK1R antagonist but also a potent SERT inhibitor. This finding led
to a series of 4,4-disubstituted piperidines and homopiperidines
designed as potent dual NK1R antagonists-SERT inhibitors. This re-
port details the synthesis, SAR and preliminary behavioral studies
of these compounds.
Scheme 2. Reagents and conditions: (i) ethylene glycol, benzene, Dean–Stark trap,
reflux, 98%; (ii) DIBAL-H, toluene, ꢁ78 °C, 100%; (iii) NaBH4, MeOH, rt, 94%; (iv)
NaH, ArCH2Br, DMF, rt, 38%; (v) HCl, acetone, reflux, 100%; (vi) NaBH4, MeOH, rt,
100%; (vii) MsCl, Et3N, rt; (viii) DBU, benzene, 80 °C, 25% for two steps; (ix) H2, Pd/C,
MeOH, 50 psi, rt, 80%. Ar: 3,5-bis(trifluoromethyl)phenyl.
Scheme 1 describes the synthesis of lactam 5 and piperidine 6
(Table 1). Double
a-allylation of methyl 2-phenylacetate gave
methyl 2-allyl-2-phenylpent-4-enoate 22. Ring closing metathesis
of 22 proceeded smoothly with Grubbs I catalyst to furnish the
cyclopentene derivative 23 in good yield. The ester of 23 was re-
duced with lithium aluminum hydride to give primary alcohol
24. O-Benzylation of 24 under standard conditions led to benzyl
ether 25, hydroboration of the olefin gave alcohol 26, and oxidation
generated cyclopentanone 27. Exposure of this ketone to sodium
azide and trifluoroacetic acid resulted in smooth Schmidt reac-
tion21 to give an approximately 1:1 mixture of two lactams 5
and 28. Reduction of lactam 28 with borane in THF generated
piperidine 6. Both lactam 5 and piperidine 6 shown in Table 1
are racemates.
2. Chemistry
The compounds in Table 1 were prepared as follows. Piperidines
2 and 3 were synthesized following the procedures described by
Stevenson et al.20 Compound 4 was made from 2 via standard
Boc protection. Analogues 5–7 were prepared as outlined in
Schemes 1 and 2.

Y.-J. Wu et al. / Bioorg. Med. Chem. 21 (2013) 2217–2228
2219
Ph
Ph
OEt
Ph
Ph
O
Ar
O
Ar
Ph
v
CN
ii
i
iii
i
ii
34
N
H
N
H
8 9
/
O
O
O
O
37
47
Ph
Ph
CN
Ph
CHO
iv
OH
Scheme 4. Reagents and conditions: (i) sodium azide, trifluoroacetic acid, benzene,
65 °C, 48%; (ii) BH3ꢀTHF, THF, 65 °C; 1 N HCl, 65 °C, 92%. Ar: 3,5-
bis(trifluoromethyl)phenyl.
O
O
O
O
O
O
39
38
40
Ar
Ph
CN
Ph
Ph
CN
Ph
CHO
O
Ar
O
vi
vii
viii
iii
i
ii
Ph
iv
CN
O
O
O
42
Ph
41
49
50
48
Ph
Ph
Ph
OH
O
Ar
O
Ar
O
Ar
O
Ar
Ph
Ph
Ph
vi
v
O
Ar
O
Ar
Ar
HN
HN
O
O
N
H
N
H
O
O
45
46
44
43
ix
HO
Ph
52
Ph
51
54
53
ix
ix
ix
O
Ph
O
vii
O
Ar
O
Ar
viii
Ph
Ph
Ph
Ph
Ph
Ph
O
O
Ar
Ar
O
O
Ar
Ar
O
O
Ar
O
Ar
Ar
HN
N
HN
HN
O
H
N
H
N
H
O
55
56
16
17
9
8
ix
ix
x
x
x
x
Ar
Ph
O
O
Ar
Ph
Ph
N
Ph
N
Ar
O
HN
N
N
N
H
R4
R4
20
21
19
18
10-15
Scheme 5. Reagents and conditions: (i) 4-bromobut-1-ene, NaH, DMF; 4-bromo-
but-1-ene, rt, 38%; (ii) Grubbs I catalyst, CH2Cl2, reflux, 100%; (iii) DIBAL-H, toluene,
ꢁ78 °C, 73%; (iv) NaBH4, MeOH, rt, 100%; (v) NaH, ArCH2Br, DMF, rt, 61%; (vi)
BH3ꢀTHF, 0 °C; H2O2, NaOH, rt, 47%; (vii) PCC, CH2Cl2, rt, 81% (viii) sodium azide,
trifluoroacetic acid, benzene, 65 °C, 26% (71) and 31% (72); (ix) BH3ꢀTHF, THF, 65 °C;
1 N HCl, 65 °C, 89%. Ar: 3,5-bis(trifluoromethyl)phenyl.
Scheme 3. Reagents and conditions: (i) PhMgBr, THF; HCl, rt 100%; (ii) KCN,
trimethylamine hydrochloride, DMF, 93 °C, 92%; (iii) ethylene glycol, benzene,
Dean–Stark trap, reflux, 48%; (iv) DIBAL-H, toluene, ꢁ78 °C, 85%; (v) NaBH4, MeOH,
rt, 87%; (vi) NaH, ArCH2Br, DMF, rt, 71%; (vii) HCl, acetone, reflux, 92%; (viii) sodium
azide, trifluoroacetic acid, benzene, 65 °C, 71%; chiral HPLC separation; (ix)
BH3ꢀTHF, THF, 65 °C; 1 N HCl, 65 °C, 100%; (x) HCHO, HCO2H, CHCl3 or Ti(Oi–Pr)4,
aldehyde, toluene; NaBH4. Ar: 3,5-bis(trifluoromethyl)phenyl.
stereochemistry to the four lactams in Tables 2 and 3 for the pur-
pose of differentiating them in our discussion. The four lactams
were reduced individually to amines 8, 9, 16 and 17. Reductive
alkylation led to tertiary amines 10–15 and 18–19. The stereo-
chemistry of homopiperidines 8–19 shown in Tables 3 and 4 fol-
lows their corresponding lactams.
For a more direct synthesis of 4,4-homopiperidines, an alterna-
tive approach was employed as shown in Scheme 4. This approach
utilized the symmetrical cyclohexanone derivative 34 (Scheme 2)
as the substrate for Schmidt reaction. Exposure of ketone 34 to typ-
ical Schmidt conditions resulted in the formation of a single lactam
47 in good yield. Reduction of 47 gave racemic 8/9.
Scheme 5 details the synthesis of azocane analogs 20 and 21
(Table 4), which were prepared from 2-phenylacetonitrile follow-
ing the same reaction sequence shown in Scheme 1. Azocane 20
was obtained as a racemate, and no attempts were made to sepa-
rate the respective enantiomers due to its moderate SERT inhibi-
tory activity (Table 4).
Scheme 2 describes the synthesis of the cyclohexane derivative
7 (Table 1). The commercially available 4-cyano-4-phenylcyclo-
hexanone (29) was converted to the ketal 30, which was reduced
to the primary alcohol 32 via aldehyde 31. Alcohol 32 underwent
O-benzylation to give benzyl ether ketal 33. The ketal was depro-
tected to give ketone 34, which was reduced to afford alcohol 35.
Cyclohexene derivative 36 was obtained from 35 via mesylation
followed by elimination. Finally, hydrogenation of 36 generated
cyclohexane analog 7.
Scheme 3 describes a divergent synthetic route leading to both
4,4-disubstituted-homopiperidine analogs 8–15 (Table 2) and 3,3-
disubstituted-homopiperidine analogs 16–19 (Table 3). Reaction of
3-ethoxycyclohex-2-enone with phenylmagnesium bromide gave
3-phenylcyclohex-2-enone, which underwent Michael reaction
with potassium cyanide to provide 3-cyano-3-phenylcyclohexa-
none (37). Reaction of 37 with ethylene glycol was followed by
conversion of the resulting ketal 38 to the benzyl ether ketone
42 via a four step sequence. This ketone underwent Schmidt reac-
tion to give a mixture of four lactams (two pairs of enantiomers):
43–46, which were separated by chiral HPLC. The isomeric identity
of the lactams was easily assigned according to spectroscopic data,
but we have not yet determined the absolute stereochemistry of
each enantiomer. Therefore, we arbitrarily assigned the absolute
3. Binding assays
The hNK1R binding affinity of the test compounds in this study
was determined using U373 cells that express the human NK1
receptor. Non-specific binding was measured in the presence of

2220
Y.-J. Wu et al. / Bioorg. Med. Chem. 21 (2013) 2217–2228
Table 2
certain concentrations of L-733,060, a non-peptide NK1 antago-
nist.12 The amount of radioligand bound in the presence and ab-
sence of L-733,060 was analyzed by plotting (ꢁ)log drug
concentration versus the amount of radioligand specifically bound.
The midpoint of the displacement curve (IC50, nM) signifies the po-
tency of the test compounds shown in Tables 1–4. In this assay,
compound 2 exhibited an hNK1R IC50 of 1 nM, which is the same
as that reported in the literature.14
The affinity of the test compounds for the human serotonin
transporter (hSERT) was evaluated by a [125I]RTI-55 binding assay
using recombinant HEK-293 cells that stably express human sero-
tonin transporters. The binding affinities, expressed as IC50s (nM),
were obtained using the same method as described for hNK1R. Un-
der these assay conditions, paroxetine, a well known SSRI, dis-
played an IC50 of 0.15 nM, which is comparable to that described
in the literature.22
Activity of 4,4-disubstituted homopiperidinesa
Compd
Structure
hNK1R (NM)
2.5
hSERT (nM)
2.9
Ph
O
O
O
Ar
Ar
Ar
8
N
H
Ph
9
2.6
2.6
26
26
N
H
Ph
10
N
Me
Ph
The detailed radioligand binding assay protocols are described
in the Experimental section.
O
O
Ar
Ar
11
12
13
14
15
7.2
150
170
N
4. Results and discussion
Me
Ph
Optimization of dual-acting molecules presents unique chal-
lenges in following SAR trends for two different targets. Based on
our preliminary studies (data not shown), the 3,5-bis(trifluoro-
methyl)benzyl ether side chain appeared to be optimal for NK1
activity. Therefore, this side chain was fixed throughout this study.
As summarized in Table 1, N-methylation of the piperidine nitro-
gen in compound 2 (cf. 3) had little impact on the NK1R activity,
but the SERT activity was significantly diminished. The basicity
of the piperidine nitrogen appears to be beneficial to NK1 receptor
binding and essential to SERT activity as the N-Boc analog 4 and
lactam 5 showed much reduced dual NK1R/SERT activities. The
3,3-disubstituted piperidine analog 6 was less potent than the
4,4-disubstituted counterpart 2 by ꢂ20-fold for both targets. The
32
N
Et
Ph
O
O
O
Ar
Ar
Ar
11
>1000
N
Et
Ph
26
91
N
Bn
Ph
74
570
Table 1
N
Bn
IC50 values of 4,4-disubstituted piperidinesa
Compd
Structure
hNK1R (nM)
1.0
hSERT (nM)
1.0
a
See footnote of Table 1.
Ph
O
O
Ar
Ar
2
Table 3
Activity of 3,3-disubstituted homopiperidinesa
N
H
Ph
Compd
Structure
hNK1R (NM)
16
hSERT (nM)
140
Ph
3
4
0.7
14
O
O
Ar
Ar
N
Me
16
HN
Ph
Ph
O
Ar
17
18
410
110
490
860
170
>1000
HN
Ph
N
Boc
Ph
O
O
Ar
Ar
O
Ar
N
5b
57
19
970
25
Me
Ph
O
N
H
19
660
660
Ph
N
O
O
Ar
Ar
6b
Me
HN
Ph
a
See footnote of Table 1.
7
1700
900
>1000
0.2
cyclohexane analog 7 showed virtually no dual activity, again indi-
cating the essentiality of the piperidine basic nitrogen.
paroxetine
a
All values are the mean of at least two separate assay determinations.
Racemate. Ar: 3,5-bis(trifluoromethyl)phenyl.
Table 2 shows the activity of the 4,4-substituted homopiperi-
dine series. Enantiomers 8 and 9 showed the same NK1R binding
b

Y.-J. Wu et al. / Bioorg. Med. Chem. 21 (2013) 2217–2228
2221
Table 4
receptors 5-HT1A, 5-HT1B, 5-HT2A, 5-HT2B, 5-HT2C were all >1
In addition, compound 2 showed excellent selectivity versus all
lM.
Activity of azocane analogsa
other NK receptors (IC50’s for NK2 and NK3 receptors >1 lM).
Compd
Structure
hNK1R (NM)
2.0
hSERT (nM)
51
Ph
O
O
Ar
Ar
6. Behavioral studies on 2
20
Species differences between the human NK1 receptor and NK1
receptors in rats and mice have resulted in many classes of NK1R
antagonists having reduced affinity for rat and mouse NK1 recep-
tors compared to human NK1 receptors. These species differences
make it difficult to test for efficacy in traditional rat and mouse
behavioral models. However, NK1R antagonists generally have sim-
ilar affinity at gerbil and human NK1 receptors. In the case of com-
pound 2, we found comparable potency for binding to both human
and gerbil NK1 receptors (IC50:2 nM for gerbil NK1R versus 1 nM for
human NK1R). Thus, we were able to perform behavioral studies in
gerbils. To investigate the antidepressant effect of piperidine 2, we
utilized the gerbil forced swimming test (FST). This behavioral test
is one of the most widely used preclinical paradigms for predicting
antidepressant activity of drug candidates. The gerbil FST was car-
ried out using the procedures described by Wallace et al.23 In this
test, immobility (i.e., no swimming or struggling) of the animals
was a measure of despair. The compound was administered intra-
peritoneally (IP) 30 min prior to test. The results were expressed as
the immobility time during the 6-min test period (mean S.E.M.)
for the 10 gerbils tested in each group. Figure 1 shows the effect
of various IP doses of compound 2 and fluoxetine at 10 mg/kg
(PO). The lowest IP dose of 2 (0.1 mg/kg) did not modify the immo-
bility times in gerbils when compared to the vehicle-treated con-
trol group. However, 2 decreased immobility time at 0.3, 1.0, and
3.0 mg/kg (IP). When administered at higher doses (10–30 mg/kg,
IP), this compound did not alter immobility presumably due to
untoward sedative effects. As expected, the positive control fluox-
etine significantly decreased immobility in the gerbil FST.
We also evaluated 2 in a locomotor activity assay to determine
whether the results obtained in the gerbil FST were influenced by a
possible increase in spontaneous locomotor activity. The test com-
pound was administered IP at doses shown to be efficacious in the
FST 30 min prior to test, and the spontaneous activity of naïve ani-
mals was measured by determining the total distance travelled. As
shown in Figure 2, compound 2 at the doses of 0.1, 0.3 and 1.0 mg/
kg (IP) produced no significant changes in locomotor activity as
compared with the vehicle-treated control group. Amphetamine
(Amph), a psychostimulant drug, was used as our positive control,
and it significantly increased the locomotor activity at 5 mg/kg (IP).
Thus, we concluded that the reduction of the immobility time in
the gerbil FST observed with piperidine 2 is consistent with an
N
H
Ph
21b
6.7
53
HN
a
See footnote of Table 1.
Racemate.
b
affinity, but they were nearly ten-fold different in SERT inhibition.
As stated previously, we have not yet determined the preferred
absolute stereochemistry. A stereochemical bias was previously
observed in the benzyloxyphenethyl piperazine derivatives re-
ported by Ryckmans et al.19 As found in the piperidine series, N-
methylation of 8 (cf. 10) did not impact the NK1R inhibitory po-
tency but reduced the SERT activity by nearly ten-fold. In contrast,
N-methylation of the enantiomer 9 (cf. 11) resulted in moderate
decrease in both NK1R antagonism (3ꢃ) and SERT inhibition
(6ꢃ). Among the N-alkyl homopiperidine compounds, the N-
methyl analogs 10 and 11 exhibited excellent NK1R inhibitory
activity (IC50 3–7 nM), albeit with diminished SERT activity. There
was little to no enantiomeric preference for NK1 receptor antago-
nism. The rank order of effect of the alkyl group was not identical
for the NK1R and SERT targets: however, a consistent finding was
that introduction of any alkyl group onto the ring nitrogen signif-
icantly decreased SERT inhibitory potency. Similar to the piperi-
dine series, the basic nitrogen is required for homopiperidines as
seen by the reduced activity of the two lactams 43 and 44 (data
not shown).
The data for the 3,3-disubstituted homopiperidines are summa-
rized in Table 3. As compared with their 4,4-disubstituted counter-
parts, the 3,3-disubstituted homopiperidines exhibited much
reduced potency towards both targets. The best compound from
this series (16) was sixfold less potent for NK1 receptor binding
and 48-fold less potent for SERT inhibition relative to that of its
4,4-disubstituted analog (i.e., 8). As expected, the two intermediate
lactams 45 and 46 maintained moderate NK1R binding activity but
lost SERT activity (data not shown).
To further probe the pharmacophore of the dual NK1/SERT bind-
ing interactions, we evaluated the 4,4- and 5,5-disubstituted azo-
cane analogs shown in Table 4. Note that compound 20 is
achiral, and 21 was a racemate. The two azocanes 20 and 21 re-
tained potent NK1 receptor binding affinities, but in terms of SERT
binding, they were less potent than the lead compound 2. Thus the
optimal spatial arrangement between the quaternary carbon and
the basic nitrogen appears to be a two-atom link for dual target
binding interactions. As in the piperidine and homopiperidine ser-
ies, the two intermediate azocanone lactams 55 and 56 possessed
much reduced activity (data not shown).
200
150
*
*
*
*
100
5. Further in vitro profile of compound 2
50
0
In parallel with our efforts to expand the SAR around compound
2, we elected to further profile this analog. Compound 2 was tested
in a commercial screening package (PanLabs) and did not exhibit
0
0.1
0.3
1
3
10
30 Fluox@10
significant (<50% inhibition at 10 lM concentration) interactions
Dose (mg/kg, IP, T-30)
with any of the over 70 other receptors except the following:
NET, DAT, H1, H2, 5-HT2B, Na+ and Ca+2 channels). Follow-up in-
house assays showed the IC50’s for the dopamine and norepineph-
Figure 1. Effects of various intraperitoneal doses of 2 (0.1, 0.3, 1, 3 and 10 mg/kg)
and fluoxetine (10 mg/kg, PO) in gerbil FST. Results are expressed as mean
immobility time S.E.M. ⁄P <0.01. T-30: 30 min pretreatment time.
rine transporters to be > 1 lM. Likewise, the IC50’s for serotonin

2222
Y.-J. Wu et al. / Bioorg. Med. Chem. 21 (2013) 2217–2228
16000
14000
12000
10000
8000
6000
4000
2000
0
from the flasks and the cells rinsed with HBSS without Ca and
without Mg. The cells are then incubated for 5–10 min in 10 mM
Tris-Cl, pH 7.5, 5 mM EDTA before the cells are lifted with a com-
bination of pipetting and scraping, as needed. To prepare mem-
branes, the cell suspension is collected into centrifuge bottles
and homogenized for 30 s with a Polytron homogenizer. The sus-
pension is centrifuged for 30 min @ 32,000ꢃg, 4 °C, then the super-
natant is decanted and the pellet resuspended and homogenized in
50 mM Tris-Cl, pH 7.5, 1 mM EDTA for 10 s. The suspension is then
centrifuged again for 30 min @ 32,000ꢃg, 4 °C. The supernatant is
decanted and the pellet resuspended in 50 mM Tris-Cl, pH 7.5,
1 mM EDTA and briefly homogenized. A Bradford assay (Bio-rad)
is performed and the membrane preparation diluted to 2 mg/ml
with 50 mM Tris-Cl, pH 7.5, 1 mM EDTA. Aliquots are prepared,
and then frozen and stored at ꢁ80 °C.
*
0
0.1
0.3
1
Amph@5
Dose (mg/kg, IP, T-30)
Figure 2. Effect of various intraperitoneal doses of 2 on the spontaneous locomotor
activity in gerbil. Results are expressed as the total distance travelled S.E.M.
⁄P <0.01. T-30: 30 min pretreatment time.
8.2. NK1 radioligand binding assay
Compounds are dissolved in 100% DMSO at a concentration
Table 5
Brain, serum concn and brain/serum ratio of 2a
100ꢃ the desired highest assay concentration, serially diluted 1:3
in 100% DMSO, and 0.6 lL/well of each solution is dispensed to a
Dose (mg/
kg)
Serum concn (nM) Brain concn (nM) Brain-to-serum
ratio
Nunc polypropylene, round bottom, 384 well plate. 100% inhibition
is defined with 0.6 L/well of 1 mM L-733,060 (Sigma L-137) dis-
solved in DMSO. 30
L/well of a 2ꢃ U373 membrane preparation
(267 g/ml in 100 mM Tris-Cl, pH 7.5, 6 mM MgCl2, 0.2% (v/v) Sig-
ma mammalian protease inhibitor cocktail (Sigma P-8340), and
g/ml chymostatin, Sigma C-7268) and 30
l
l
0.1
0.3
1
3
10
30
3.6 (0.5)
12 (11)
8.8 (4.6)
44 (52)
2.3 (0.9)
3.2 (1.1)
3.2 (1.7)
4.2 (1.4)
5.2 (1.6)
2.3 (0.9)
35 (35)
150 (200)
880 (700)
4300 (3600)
7300 (9600)
l
210 (200)
730 (560)
2300 (3200)
4
l
l
L/well of a 2ꢃ radio-
ligand solution (400 pM [125I]Substance P (Perkin Elmer NEX-190)
in 1% (w/v) BSA (Sigma A-2153), 0.1 mg/ml bacitracin, Sigma B-
0125) are added to the well and the reaction incubated for 1 h at
room temperature. The contents of the assay plate are then trans-
ferred to a Millipore MultiscreenHTS GF/B filter plate which has
been pretreated with 0.5% PEI for at least 1 h. The plate is vacuum
filtered and washed with 7 washes of 100 ul/well 20 mM Tris-Cl,
pH 7.5, 0.5% (w/v) BSA chilled to 4 °C. The filtration and washing
is completed in less than 90 s. The plates are air-dried overnight,
a
Compound dosed IP to gerbils as a solution in 100% PEG-400 with brain and
serum samples taken at 36 min post injection. Data are mean values (n = 3) with SD
in parentheses.
antidepressant-like effect and is not confounded by locomotor
effects.
To understand the exposure-effect relationship, we obtained
the brain and serum samples from gerbils following IP administra-
tion of 2. These samples were taken 36 min post-dose, which ac-
counts for the time elapsed between dosing and the end of
evaluation in the behavioral model. Table 5 shows the exposure
data from the doses tested in the gerbil FST. The minimal effective
dose of 2 was 0.3 mg/kg, indicating that a brain concentration of
44 nM may be required to induce a significant decrease on the
immobility test. Compound 2 also showed good brain penetration,
and its brain and serum concentrations increased dose-depen-
dently upon IP administration (Table 5).
12 lL/well of MicroScint scintillation fluid added, and the plates
counted in a Trilux.
8.3. SERT radioligand binding assay
Compounds are dissolved in 100% DMSO at a concentration
100ꢃ the desired highest assay concentration, serially diluted 1:3
in 100% DMSO, and 0.4 lL/well of each solution is dispensed to a
Nunc polypropylene, round bottom, 384 well plate. 100% inhibition
is defined with 0.4 L/well of 1 mM fluoxetine (Sigma F-132) dis-
solved in DMSO. 20
L/well of a 2ꢃ HEK-hSERT membrane prepa-
ration (15 g/ml in 50 mM Tris-Cl, pH 7.5, 120 mM NaCl, 5 mM
KCl) and 20 L/well of
2ꢃ radioligand solution (520 pM
125I]RTI-55 (Perkin–Elmer NEX-272) in 50 mM Tris-Cl, pH 7.5,
l
l
l
7. Conclusion
l
a
[
Piperidine 2 and homopiperidine 8 were identified as potent
dual NK1R antagonists and SERT inhibitors. Behavioral studies of
2 in the gerbil forced swimming test (FST) suggest that dual
NK1R antagonists/SERT inhibitors may have potential for the treat-
ment of depression disorders. Further SAR studies of this class of
dual NK1R antagonists/SERT inhibitors will be reported in due
course.
120 mM NaCl, 5 mM KCl) are added to each well and the reaction
incubated for 1 h at room temperature. The contents of the assay
plate are then transferred to a Millipore MultiscreenHTS GF/B filter
plate which has been pretreated with 0.5% PEI for at least one hour.
The plate is vacuum filtered and washed with 7 washes of 100 lL/
well 50 mM Tris-Cl, pH 7.5, 120 mM NaCl, 5 mM KCl chilled to 4 °C.
The filtration and washing is completed in less than 90 s. The
plates are air-dried overnight, 12 lL/well of MicroScint scintilla-
8. Experimental section
tion fluid added, and the plates counted in a Trilux.
8.1. Preparation of membranes
8.4. Data analysis
Crude membrane suspensions were prepared for the NK1 and
SERT radioligand binding assays from U373 cells or recombinant
HEK-293 cells expressing hNK1 and hSERT, respectively. Cells were
harvested from T-175 flasks as follows. The medium is removed
The raw data are normalized to percent inhibition using control
wells defining 0% (DMSO only) and 100% (selective inhibitor) inhi-
bition which are run on each plate. Each plate is run in triplicate,
and the concentration response curve thus generated is fit

Y.-J. Wu et al. / Bioorg. Med. Chem. 21 (2013) 2217–2228
2223
using the four-parameter dose response equation, Y = Bottom +
(Top–Bottom)/(1+10^((LogIC50-X)*HillSlope)) in order to deter-
mine the IC50 value for each compound. The radioligand concentra-
tion chosen for each assay corresponds to the Kd concentration
determined through saturation binding analysis for each assay.
1H NMR (400 MHz, CDCl3) d 2.70 (4H, s), 3.54 (2H, d, J = 5.6 Hz),
5.75 (1H, s), 7.22–7.35 (5H, m).
8.9. 1-(((1-Phenylcyclopent-3-enyl)methoxy)-methyl)-3,5-
bis(trifluoromethyl)benzene (25)
8.5. General chemistry
To a solution of 24 (257 mg, 1.3 mmol) and 3,5-(bis-trifluoro-
methyl)benzyl bromide (0.41 mL, 2.2 mmol) in DMF (2.5 mL) at
0 °C was added sodium hydride (95% oil dispersion, 75 mg,
3 mmol) and the resulting suspension was stirred at rt for
30 min. Water (2 mL) was added, and the reaction was extracted
with ethyl acetate. The organic layer was separated and concen-
trated in vacuo and the crude product was purified by preparative
TLC eluting with 15% ethyl acetate/85% hexanes to give 25 as an
oily material (382 mg, 75% yield). 1H NMR (400 MHz, CDCl3) d
2.73 (4H, s), 3.51 (2H, s), 4.44 (2H, s), 5.74 (2H, s), 7.21–7.32 (5H,
m), 7.57 (2H, s), and 7.72 (1H, s).
Nuclear magnetic resonance (NMR) spectra were recorded on a
Bruker AM FT instrument operating at 400 or 500 MHz for proton
(1H) and 100 or 125 MHz for carbon (13C). All spectra were re-
corded using tetramethylsilane (TMS) as an internal standard,
and signal multiplicity was designated according to the following
abbreviations: s = singlet, d = doublet, t = triplet, q = quartet,
m = multiplet, brd s or d = broad singlet or doublet. The coupling
constant (J) is in hertz. High-resolution mass spectrometry (HRMS)
data was obtained using a standard flow injection technique on a
Finnigan MAT 900 mass spectrometer in electrospray ionization
(ESI) mode. All tested compounds possessed a purity of not less
than 95% by LC/MS analysis using two methods. LC/MS was per-
formed on a Shimadzu LC-10AS liquid chromatography using a
SPD-10AV UV–Vis detector with mass spectrometry (MS) data
determined using a Micromass LC platform in positive electrospray
ionization mode (ESI+). LC/MS method A: column YMC ODS-A C18
S7 (3.0 ꢃ 50 mm), gradient system 10/90 to 90/10 methanol/water
with a buffer consisting of 0.1% TFA over 2 min, flow rate 5 mL/min,
wavelength 220 nm. Method B: the same as method A except col-
umn Phenomenex Luna C18 S10 (3.0 ꢃ 50 mm) was used.
8.10. 3-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-3-
phenylcyclopentanol (26)
To a solution of 25 (200 mg, 0.5 mmol) in THF (0.55 mL) at 0 °C
was added borane–tetrahydrofuran complex (1.0 M solution,
0.55 mL, 0.6 mmol) dropwise, and the resulting solution was
warmed to rt and stirred at rt for 12 h. The reaction mixture was
cooled to 0 °C, and water (1.0 mL) was added slowly followed by
30% hydrogen peroxide (0.19 mL) and 1 N sodium hydroxide
(0.55 mL). The resulting solution was stirred at rt for 5 min and
then worked up with ethyl acetate. The residue was purified by
preparative TLC eluting with 40% ethyl acetate/60% hexanes to give
26 as an oily material (100 mg, 50% yield). 1H NMR (400 MHz,
CDCl3) d 2.06 (m), 1.9–2.6 (m), 3.42 (s), 3.57 (q, J = 8.8 Hz), 4.45
(s), 4.3–4.6 (m), 7.1-7.5 (m), 7.47 (s), 7.64 (s), 7.74 (s), and 7.76 (s).
8.6. Methyl 2-allyl-2-phenylpent-4-enoate (22)
To
a solution of methyl 2-phenylacetate (74) (500 mg,
3,3 mmol) in DMF (1.0 mL) at 0 °C was added sodium hydride
(95% oil dispersion, 210 mg, 8.3 mmol) and the resulting suspen-
sion was stirred at 0 °C for 10 min. Allyl bromide (0.72 mL,
8.3 mmol) was added, and the reaction mixture was stirred at rt
for 30 min. Saturated sodium chloride was added and the reaction
was extracted with ethyl acetate. The organic layer was separated
and concentrated in vacuo to give 22 as an oily material (770 mg,
100%). This crude product was used directly for the next reaction.
1H NMR (400 MHz, CDCl3) d 2.75 (4H, m), 3.61 (3H, s), 5.01 (4H,
m), 5.45 (2H, m), 7.18–7.30 (5H, m).
8.11. 3-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-3-
phenylcyclopentanone (27)
To a solution of 26 (142 mg, 0.3 mmol) in dichloromethane
(10 mL) at rt was added pyridinium chlorochromate (146 mg,
0.7 mmol and powered 4A° molecular sieves (146 mg), and the
resulting mixture was stirred at rt for 1.5 h and then filtered
through a small pad of silica gel. The filtrate was evaporated in va-
cuo, and the residue was purified by preparative TLC eluting with
30% ethyl acetate/70% hexanes to give 27 as an oily material
(150 mg, 100%). This crude product was used directly for the next
reaction. 1H NMR (400 MHz, CDCl3) d 2.3–2.6 (4H, m), 2.68 (1H, d,
J = 17.6 Hz), 2.76 (1H, d, J = 17.6 Hz), 3.52 (1H, q, J = 9.2 Hz), 3.58
(1H, d, J = 9.2 Hz), 4.52 (2H, s), 7.2-7.4 (5H, m), 7.59 (2H, s), 7.76
(1H, s).
8.7. Methyl 1-phenylcyclopent-3-enecarboxylate (23)
To a solution of 22 (378 mg, 1.6 mmol) in dichloromethane
(41 mL) was added benzylidene-bis(tricyclohexylphosphine)-di-
chloro-ruthenium (68 mg, 0.08 mmol) and the resulting suspen-
sion was heated under reflux for 1 h. The solvent was removed in
vacuo, and the crude product was purified by preparative TLC elut-
ing with 10% ethyl acetate/90% hexanes to give 23 as an oil
(300 mg, 91% yield). 1H NMR (400 MHz, CDCl3) d 2.74 (2H, d,
J = 14.8 Hz), 3.40 (2H, d, J = 14.8 Hz), 3.63 (3H, s), 5.76 (2H, s),
7.22-7.31 (5H, m). 13C NMR (100 MHz, CDCl3) d 42.8, 52.5, 58.4,
126.6, 126.8, 128.7, 129.2, 143.7, and 176.6.
8.12. 5-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-5-
phenylpiperidin-2-one (28) and 4-((3,5-bis(trifluoro-
methyl)benzyloxy)methyl)-4-phenylpiperidin-2-one (5)
A mixture of 27 (34 mg, 0.08 mmol) and sodium azide solution
(3.0 M in H2O, 55 lL, 0.6 mmol) in trifluoroacetic acid (100 lL) was
8.8. (1-Phenylcyclopent-3-enyl)methanol (24)
stirred at 65 °C for 1 h. After allowing to cool to rt, the mixture was
concentrated under vacuum, and saturated sodium bicarbonate
was added. The reaction mixture was extracted with dichloro-
methane, and the organic layer was separated and concentrated
in vacuo, and the crude product was purified by preparative TLC
eluting with 80% ethyl acetate/hexanes to give 28 (10 mg, 29%
yield) and 5 (9 mg, 26% yield). Lactam 5: 1H NMR (400 MHz, CDCl3)
d 7.77 (1H, s), 7.62 (2H, s), 7.2–7.4 (5H, m), 5.55 (1H, br s), 4.53 (2H,
apparent s), 3.62 (1H, d, J = 9.2 Hz), 3.53 (1H, d, J = 9.2 Hz), 3.25
(1H, m), 2.99 (1H, m), 2.91 (1 J, d, J = 17.2 Hz), 2.74 (1H, d,
To a solution of 23 (298 mg, 1.5 mmol) in ether (10 mL) at 0 °C
was added lithium aluminum hydride (1.0 M solution in ether,
1.48 mL, 1.5 mmol), and the resulting suspension was stirred at
0 °C for 30 min. The reaction was quenched with saturated sodium
sulfate (0.5 mL) and then diluted with ether (100 mL). Anhydrous
sodium sulfate was added, and the ether solution was filtered.
The filtrate was evaporated in vacuo to give 24 as an oil (257 mg,
100%). The crude product was used directly for the next reaction.

2224
Y.-J. Wu et al. / Bioorg. Med. Chem. 21 (2013) 2217–2228
J = 17.2 Hz). Lactam 28: 1H NMR (400 MHz, CDCl3) d 7.75 (1H, s),
7.52 (2H, s), 7.2–7.4 (5H, m), 5.91 (1H, br s), 4.52 (1H, d,
J = 12.8 Hz), 4.43 (1H, d, J = 12.8 Hz), 3.88 (1H, m), 3.65 (3H, m),
2.36 (1H, m), and 2.21 (3H, m).
Water was added, and the reaction mixture was extracted with
ether. The ether layer was separated and concentrated in vacuo.
The crude material was purified by silica gel flash chromatography
eluting with from 15–20% ethyl acetate/hexanes to give 33 as a col-
orless oil (1.2 g, 38% yield). 1H NMR (400 MHz, CDCl3) d 7.71 (1H,
s), 7.54 (2H, s), 7.2–7.4 (5H, m), 4.41 (2H, s), 3.95 (4H, m), 3.40
(2H, s), 1.6–2.4 (8H, m).
8.13. 3-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-3-
phenylpiperidine (6)
A solution of 28 (6 mg, 0.014 mmol) in BH3ꢀTHF complex solu-
8.18. 4-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-4-
phenylcyclohexanone (34)
tion (1.5 M in diethyl ether, 46 lL, 0.069 mmol) was heated at
65 °C for 3 h. After the reaction mixture was allowed to cool to
rt, methanol (0.1 mL) and 1 N HCl solution (0.1 mL) were added
and the mixture was stirred at 65 °C for 1 h. The reaction mixture
was concentrated under vacuum and neutralized with 1 N sodium
hydroxide solution (0.1 mL) and extracted with dichloromethane,
and the organic layer was separated and concentrated in vacuo
to give 6 as a colorless oil (4 mg, 67% yield). 1H NMR (CDCl3,
400 MHz) d 7.72 (1H, s), 7.55 (2H, s), 7.2–7.4 (5H, s), 4.47 (1H, d,
J = 13.2 Hz), 4.42 (1H, d, J = 13.2 Hz), 3.57 (1H, d, J = 9.2 Hz), 3.52
(1H, d, J = 9.2 Hz), 3.37 (1H, d, J = 12.8 Hz), 3.10 (1H, d,
J = 12.8 Hz), 2.82 (2H, m), 1.96 (1H, m), 1.68 (1H, m). MS: 418.33
(M+H+).
To a solution of 33 (1.2 g, 2.5 mmol) in acetone (40 mL) at rt was
added 1 N HCl solution (10 mL) and the resulting mixture was
heated at reflux for 1 h. Acetone was removed in vacuo, and the
reaction mixture was extracted with ether. The ether layer was
separated and concentrated in vacuo to give 34 as a colorless oil
(1.1 g, 100%). This crude product was used directly for the next
reaction. 1H NMR (400 MHz, CDCl3) d 7.74 (1H, s), 7.55 (2H, s),
7.2–7.4 (5H, m), 4.45 (2H, s), 3.46 (2H, s), 2.65 (2H, m), 2.34 (4H,
m), and 2.03 (2H, m).
8.19. 4-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-4-
phenylcyclohexanol (35)
8.14. 8-Phenyl-1,4-dioxaspiro[4.5]decane-8-carbonitrile (30)
To a solution of 34 (400 mg, 0.9 mmol) in methanol (5 mL) was
added sodium borohydride (53 mg, 1.4 mmol), and the reaction
mixture was stirred at rt for 10 min. Methanol was removed in va-
cuo, water was added, and usual work-up with ethyl acetate pro-
vided 35 as a colorless oil (416 mg, 100%). This crude product
was used directly for the next step without purification
A mixture of 29 (5 g, 25 mmol), ethylene glycol (7 mL) and
pyridinium p-toluene sulfonate (100 mg) in benzene (100 mL)
was stirred at 111 °C in a Dean-Stark apparatus overnight. Satu-
rated sodium bicarbonate was added, and the reaction was ex-
tracted with ether and the ether layer was concentrated in vacuo
to give 30 (6 g, 98% yield). This crude product was used for the next
step without purification. 1H NMR (400 MHz, CDCl3) d 7.3–7.6 (5H,
s), 3.97 (4H, m), 2.16 (6H, m), and 1.84 (2H, m).
8.20. 1-(((1-Phenylcyclohex-3-enyl)methoxy)methyl)-3,5-
bis(trifluoromethyl)benzene (36)
8.15. 8-Phenyl-1,4-dioxaspiro[4.5]decane-8-carbaldehyde (31)
To a solution of 35 (416 mg, 0.9 mmol) in pyridine (3 mL) at 0 °C
was added methanesulfonyl chloride (0.11 mL, 1.5 mmol), and the
reaction mixture was stirred at 0 °C for 1 h and then rt for 2 h.
Water was added, and usual work-up with ethyl acetate gave the
To a solution of 30 (1.8 g, 7.3 mmol) in toluene (40 mL) at
ꢁ78 °C was added DIBAL-H (1.0 M solution in toluene, 10.9 mL,
10.9 mmol) slowly, and the reaction mixture was stirred at
ꢁ78 °C for 2 h. The reaction mixture was quenched with methanol,
saturated ammonium chloride was added, and the reaction was ex-
tracted with ether and the organic layer was separated and con-
centrated in vacuo to give 31 as a colorless liquid (1.8 g, 100%
yield). This crude material was used directly for the next step. 1H
NMR (400 MHz, CDCl3) d 9.28 (1H, s), 7.2–7.4 (5H, m), 3.92 (4H,
m), 1.3–2.4 (8H, m).
crude
4-((3,5-bis(trifluoromethyl)benzyloxy)methyl)-4-phen-
ylcyclohexyl methanesulfonate. To this crude mesylate were added
DBU (0.97 mL) and DMAP (20 mg), and the resulting mixture was
heated at 80 °C for 12 h. Water and ethyl acetate were added,
and the reaction was worked up as usual. The crude product was
purified by preparative TLC eluting with 15% ethyl acetate/hexanes
to give olefin 36 as a colorless oil (100 mg, 25% yield). 1H NMR
(400 MHz, CDCl3) d 7.73 (1H, s), 7.58 (2H, s), 7.2–7.4 (5H, m),
5.76 (1H, m), 5.74 (1H, m), 4.47 (2H, apparent d), 3.61 (1H, d,
J = 8.8 Hz), 3.52 (1H, d, J = 8.8 Hz), 2.9 (1H, br. d), 2.4 (1H, br. d),
1.7-2.1 (4H, m).
8.16. (8-Phenyl-1,4-dioxaspiro[4.5]decan-8-yl)methanol (32)
To a solution of 31 (1.8 g, 7.3 mmol) in methanol (30 mL) and THF
(10 mL) at 0 °C was added sodium borohydride (276 mg, 7.3 mmol),
in portions, and the resulting suspension was stirred at room tem-
perature for 1 h. The solvent was removed, and saturated sodium
bicarbonate was added. The reaction mixture was extracted with
ethyl acetate and the organic layer was separated and concentrated
in vacuo to give 32 (1.7 g, 94% yield), which was used directly for the
next step. 1H NMR (400 MHz, CDCl3) d 7.2–7.4 (5H, m), 3.90 (4H, m),
3.50 (2H, d, J = 6.0 Hz), 1.1–2.3 (7H, m).1.48–1.81 (7H, m).
8.21. 1-(((1-Phenylcyclohexyl)methoxy)methyl)-3,5-
bis(trifluoromethyl)benzene (7)
A mixture of 36 and 10% Pd/C (5 mg) in ethanol (0.4 mL) was
hydrogenated under hydrogen balloon for 12 h. The reaction mix-
ture was passed through a pad of Celite, and the filtrate was evap-
orated in vacuo to give 7 as a colorless oil (4 mg, 80% yield). 1H
NMR (400 MHz, CDCl3) d 7.72 (1H, s), 7.56 (2H, s), 7.2–7.4 (5H,
m), 4.41 (2H, s), 3.40 (2H, s), 2.20 (2H, m), 1.75 (2H, m), 1.2–1.7
(6H, m).
8.17. 8-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-8-phenyl-
1,4-dioxaspiro[4.5]decane (33)
8.22. 3-Phenylcyclohex-2-enone
To a solution of 32 (1.7 g, 6.9 mmol) and 3,5-(bis-trifluoro-
methyl)benzyl bromide (1.3 mL, 7.2 mmol) in DMF (15 mL) at
0 °C was added sodium hydride (95% oil dispersion, 260 mg,
10.3 mmol) and the resulting suspension was stirred at rt for 2 h.
To a solution of 3-ethoxycyclo-hex-2-enone (2.0 g, 14.3 mmol)
in THF (2 mL) at 0 °C was added phenylmagnesium bromide
(1.0 M solution in THF, 15 mL, 15 mmol) slowly. After the addition,

Y.-J. Wu et al. / Bioorg. Med. Chem. 21 (2013) 2217–2228
2225
the reaction mixture was allowed to warm to rt and stirred at rt for
8.27. 1-(((3,3-Ethylenedioxy-1-phenylcyclohexyl)-
1 h. The reaction was quenched with 1 N HCl solution (15.16 mL)
and then extracted with dichloromethane. The organic layer was
separated and concentrated in vacuo to give 3-phenylcyclohex-2-
enone as yellow solid (2.5 g, 100%). 1H NMR (400 MHz, CDCl3) d
2.16 (2H, dt, J = 4, 12 Hz), 2.48 (2H, t, J = 8 Hz), 2.77 (2H, dt, J = 2,
6 Hz), 6.41 (1H, t, J = 2 Hz), 7.39–7.41 (3H, m), 7.52–7.54 (2H, m).
methoxy)methyl)-3,5-bis(trifluoromethyl)-benzene (41)
To a solution of 40 (726 mg, 2.9 mmol) and 3,5-(bis-trifluoro-
methyl)benzyl bromide (1.1 mL, 5.8 mmol) in DMF (6 mL) at 0 °C
was added sodium hydride (95% oil dispersion, 148 mg, 5.9 mmol),
and the resulting suspension was stirred at rt for 2 h. Saturated so-
dium bicarbonate was added, and the reaction mixture was ex-
tracted with dichloromethane. The dichloromethane layer was
separated and concentrated in vacuo The crude material was puri-
fied by silica gel flash chromatography eluting with a gradient from
10% acetone/90% hexanes to 15% acetone/85% hexanes to give 41
as clear pale yellow sticky oil (990 mg, 71% yield). 1H NMR
(400 MHz, CDCl3) d 1.25 (1H, m), 1.74 (4H, t, J = 8 Hz), 1.90 (1H,
d, J = 12 Hz), 2.13 (1H, m), 2.24 (1H, d, J = 16 Hz), 3.70 (2H, dd,
J = 8, 36 Hz), 3.91 (4H, m), 4.45 (2H, s), 7.21 (2H, t, J = 8 Hz),
7.29–7.38 (4H, m), 7.53 (2H, s), 7.71 (1H, s). MS: 475.14 (M+H+).
8.23. 3-Oxo-1-phenylcyclohexanecarbonitrile (37)
A mixture of 3-phenylcyclohex-2-enone (2.5 g, 14.3 mmol),
potassium cyanide (1.9 g, 29 mmol) and trimethylamine hydrochlo-
ride (2.1 g, 21.5 mmol) in H2O (10 mL) and DMF (57 mL) was heated
at 93 °C for 6 h. After the reaction mixture was allowed to cool to rt,
saturated sodium bicarbonate was added, and the reaction mixture
was extracted with ether. The ether layer was separated and con-
centrated in vacuo to give 2.6 g (92% yield) of the crude 37 which
was directly taken to the next step without further purification.
8.28. 3-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-3-
phenylcyclohexanone (42)
8.24. 3,3-Ethylenedioxy-1-phenylcyclohexanecarbo-nitrile (38)
A mixture of 37 (2.6 g, 13.1 mmol), ethylene glycol (3.7 mL,
65.5 mmol) and pyridinium p-toluene sulfonate (100 mg) in ben-
zene (40 mL) was stirred at 111 °C in Dean–Stark apparatus over-
night. After the reaction mixture was allowed to cool to rt,
saturated sodium bicarbonate was added, and the reaction mixture
was extracted with ether. The organic layer was separated and
concentrated in vacuo to give crude material that was purified by
silica gel flash chromatography eluting with a gradient of 25% ethyl
acetate/75% hexane to 30% ethyl acetate/70% hexane to give 38 as a
white solid (1.7 g, 48% yield from 3-phenylcyclohex-2-enone). 1H
NMR (400 MHz, CDCl3) d 1.56 (1H, dt, J = 4, 16 Hz), 1.70–1.79
(1H, m), 1.89–1.96 (3H, m), 2.06–2.28 (3H, m), 3.89–3.97 (2H,
m), 4.03–4.12 (2H, m), 7.28–7.32 (1H, m), 7.36–7.39 (2H, m),
7.48–7.51 (2H, m). MS: 266.37 (M+Na+).
To a solution of 41 (990 mg, 2.1 mmol) in acetone (4 mL) at rt
was added 1 N HCl solution (3.1 mL), and the resulting mixture
was heated under reflux for 2 h. Saturated sodium bicarbonate
was added, and the reaction mixture was extracted with dichloro-
methane. The organic layer was separated and concentrated in va-
cuo to give 41 as clear pale yellow sticky oil (830 mg, 92% yield).
The crude product was used directly for the next reaction. 1H
NMR (400 MHz, CDCl3) d 1.60 (1H, m), 1.88 (1H, m), 2.19 – 2.24
(2H, m), 2.29–2.32 (2H, m), 2.85 (2H, dd, J = 12, 68 Hz), 3.51 (2H,
dd, J = 12, 56 Hz), 4.52 (2H, dd, J = 16, 18 Hz), 7.22–7.25 (1H, m),
7.32 (4H, d, J = 4 Hz), 7.63 (2H, s), 7.77 (1H, s). MS: 453.37 (M+Na+).
Isomer
A (43) and isomer B (44) of 4-((3,5-bis(trifluoro-
methyl)benzyloxy)methyl)-4-phenyl-azepan-2-one and Isomer A
(45) and isomer B (46) of 6-((3,5-bis(trifluoromethyl)-benzyl-
oxy)methyl)-6-phenylazepan-2-one. A mixture of 42 (730 mg,
1.7 mmol) and sodium azide solution (3.0 M in water, 1.1 mL,
3.4 mmol) in trifluoroacetic acid (5.6 mL) was stirred at 65 °C for
1 h. After cooling down, the mixture was concentrated under vac-
uum, and saturated sodium bicarbonate. The reaction mixture was
extracted with dichloromethane. The organic layer was separated
and concentrated in vacuo and the crude mixture was purified
by silica gel flash chromatography eluting with from 30% ace-
tone/70% hexanes to 40% acetone/60% hexanes to give a mixture
of 43–46 as a colorless sticky oil (540 mg, 71% yield).
8.25. 3,3-Ethylenedioxy-1-phenylcyclohexanecarba-ldehyde
(39)
To a solution of 38 (1.7 g, 6.8 mmol) in toluene (20 mL) at
ꢁ78 °C was added DIBAL-H (1.0 M solution in toluene, 8.9 mL)
slowly. The reaction mixture was stirred for 1 h. The reaction
mixture was allowed to warm to rt, 1 N HCl solution (2.6 mL)
was added, and the reaction mixture was stirred at rt for 1 h.
Saturated sodium bicarbonate was added, and the reaction was
extracted with dichloromethane. The organic layer was separated
and concentrated in vacuo and the crude material was purified
by filtering through a short silica gel pad with dichloromethane
to give 39 as a colorless sticky oil (1.4 g, 85% yield). 1H NMR
The above mixture of 4 isomers was separated by chiral HPLC
(Chiralpak AS semi-prep column, 20 ꢃ 250 mm, 10
lm, solvents
5% EtOH/95% heptane for 83 min., 15% EtOH/ 85% heptane for
17 min., 5% EtOH/95% heptane for 10 min, flow rate 10 mL/min,
UV 205 nm) to give 43 (110 mg), 44 (110 mg), 45 (80 mg) and 46
(90 mg) as colorless oils.
(400 MHz, CDCl3)
d 1.32–1.40 (1H, m), 1.46–1.54 (1H, m),
1.72–1.86 (3H, m), 2.08 (1H, d, J = 16 Hz), 2.59–2.67 (2H, m),
3.94–4.07 (4H, m), 7.15–7.36 (5H, m), 9.36 (1H, d, J = 4 Hz).
Compound 43: Retention time is 44.02 min. with >99.9% ee. 1H
NMR (CDCl3, 400 MHz) d 1.72–1.80 (2H, m), 2.04 (1H, m), 2.51
(1H, d, J = 16 Hz), 2.94 (1H, d, J = 16 Hz), 3.10 (1H, d, J = 16 Hz),
3.17 (1H, m), 3.24 – 3.32 (1H, m), 3.45 (2H, s), 4.46 (2H, s),
5.81 (1H, s), 7.24 (1H, t, J = 8 Hz), 7.34 (2H, t, J = 8 Hz), 7.47 (2H,
d, J = 8 Hz), 7.61 (2H, s), 7.75 (1H, s). 13C NMR (CDCl3, 100 MHz)
d 25.3, 36.7, 42.9, 43.2, 44.1, 71.8, 81.0, 121.8 (quintet, J = 3 Hz),
123.7 (q, J = 271 Hz), 127.2, 127.5 (d, J = 4 Hz), 127.6, 128.8,
8.26. (3,3-Ethylenedioxy-1-phenylcyclohexyl)-methanol (40)
To a solution of 39 (1.4 g, 5.8 mmol) in methanol (25 mL) at 0 °C
was added sodium borohydride (221 mg, 5.8 mmol) in portions,
and the resulting suspension was stirred at rt overnight. Saturated
sodium bicarbonate was added, methanol was removed under vac-
uum, and the reaction mixture was extracted with dichlorometh-
ane. The organic layer was separated and concentrated in vacuo
to give 40 as a clear colorless sticky oil (1.3 g, 87% yield). This crude
product was used directly for the next reaction. 1H NMR (400 MHz,
CDCl3) d 1.48–1.81 (7H, m), 1.88 (1H, d, J = 16 Hz), 2.06 (1H, d,
J = 12 Hz), 2.22 (1H, d, J = 12 Hz), 3.73–4.06 (5H, m), 7.15–7.25
(2H, m), 7.31–7.38 (3H, m). MS: 271.40 (M+Na+).
132.0 (q, J = 33 Hz), 141.4, 142.0, 175.9. Optical rotation, ½a D24
ꢄ
ꢁ9.03 (c 1.77, CH2Cl2). HRMS calcd for C22H22NO2F6 446.1555
(M+H+), found 446.1571.
Compound 44: Retention time is 53.38 min. with >99.8% ee. 1H
and 13C NMR data are the same as 43. Optical rotation, ½a 2D4
ꢄ
+7.10 (c
1.69, CH2Cl2). HRMS calcd for C22H22NO2F6 446.1555 (MH+), found
446.1573.

2226
Y.-J. Wu et al. / Bioorg. Med. Chem. 21 (2013) 2217–2228
Compound 45: Retention time is 73.61 min. with >99.9% ee. 1H
methanol (0.2 mL) was added followed by 1 N HCl solution
(0.2 mL), and the reaction mixture was heated at 65 °C for 2 h.
The mixture was concentrated under vacuum, neutralized with
1 N NaOH solution and extracted with dichloromethane/The or-
ganic layer was separated and concentrated in vacuo to give ( )8/
9 as a colorless sticky oil (35 mg, 92% yield). The spectroscopic data
are identical to those obtained following Scheme 3.
NMR (CDCl3, 400 MHz) d 1.82 (2H, m), 1.97 (1H, m), 2.28 (1H, m),
2.51 (2H, m), 3.50 (1H, d, J = 8 Hz), 3.59 (1H, d, J = 8 Hz), 3.61 (1H,
dd, J = 8, 12 Hz), 3.77 (1H, dd, J = 8, 12 Hz), 4.43 (1H, d, J = 12 Hz),
4.53 (1H, d, J = 12 Hz), 5.72 (1H, br s), 7.23-7.38 (5H, m), 7.57
(2H, s), and 7.76 (1H, s). 13C NMR (CDCl3, 100 MHz) d 19.1, 35.9,
37.0, 45.1, 47.7, 71.9, 76.9, 121.9 (quintet, J = 4 Hz), 123.6 (q,
J = 271 Hz), 127.1, 127.3, 127.5 (d, J = 2 Hz), 129.1, 132.1 (q,
8.33. Isomer A 4-((3,5-bis(trifluoromethyl)benzyloxy)-methyl)-
1-methyl-4-phenylazepane (10)
J = 33 Hz), 141.1, 142.0, 177.9. HRMS calcd for
C22H22NO2F6
446.1555 (M+H+), found 446.1556.
Compound 46: Retention time is 95.79 min. with >99.9% ee. 1H
and 13C NMR data are the same as 45. HRMS calcd for C22H22NO2F6
446.1555 (M+H+), found 446.1549.
A solution of 8 (3 mg), paraformaldehyde (3 mg) and formic
acid (2.2 mg) in chloroform (0.1 mL) was stirred at 65 °C for 12 h.
The mixture was concentrated under vacuum and passed through
an anion exchange cartridge to remove formic acid. The crude
product was purified by preparative TLC eluting with 10% metha-
nol/89% dichloromethane/1% NH4OH to give 10 as a clear sticky
oil (2 mg, 67% yield). 1H NMR (CDCl3, 400 MHz) d 1.67–1.74 (1H,
m), 1.83–1.91 (1H, m), 2.05–2.23 (3H, m), 2.38 (3H, s), 2.42–2.61
(3H, m), 2.79 (2H, m), 3.45 (2H, s), 4.44 (2H, s), 7.22 (1H, t,
J = 4 Hz), 7.32 (4H, d, J = 4 Hz), 7.55 (2H, s), 7.73 (1H, s). 13C NMR
(CDCl3, 125 MHz) d 23.4, 29.8, 34.9, 45.5, 46.7, 53.3, 60.2, 71.8,
81.1, 121.3 (quintet, J = 3.75 Hz), 123.4 (q, J = 271 Hz), 126.3,
126.8, 127.1, 128.4, 131.6 (q, J = 32.5 Hz), 141.4, 145.7. HRMS calcd
for C23H26NOF6 446.1919 (M+H+), found 446.1916.
8.29. Isomer A of 4-((3,5-bis(trifluoromethyl)benzyloxy)-
methyl)-4-phenylazepane (8)
A solution of 43 (90 mg, 0.2 mmol) in boraneꢀTHF complex solu-
tion (1.5 M in diethyl ether, 1.1 mL, 1.7 mmol) was stirred at rt over-
night and then heated at 65 °C for 1 h. After being allowed to cool to
rt, the reaction mixture was concentrated under vacuum, methanol
(0.4 mL) was added followed by 1 N HCl solution (0.4 mL), and the
reaction mixture was heated at 65 °C for 2 h. The mixture was con-
centrated under vacuum, neutralized with 1 N NaOH solution and
extracted with dichloromethane. The organic layer was separated
and concentrated in vacuo to give 8 as a colorless sticky oil
(90 mg, 100%). 1H NMR (CDCl3, 400 MHz) d 1.65–1.73 (1H, m),
1.80–1.87 (1H, m), 1.94–2.01 (1H, m), 2.08 (1H, ddd, J = 4, 8,
16 Hz), 2.31–2.37 (1H, m), 2.48 (1H, dd, J = 4, 16 Hz), 2.86–3.01
(3H, m), 3.16 (1H, dd, J = 4, 16 Hz), 3.42 (2H, s), 4.43 (2H, s), 7.21–
7.23 (1H, m), 7.30–7.36 (4H, m), 7.55 (2H, s), 7.73 (1H, s). 13C NMR
(CDCl3, 100 MHz) d 24.2, 34.0, 36.4, 43.5, 45.6, 48.2, 71.7, 81.5,
121.3 (quintet, J = 3 Hz), 123.3 (q, J = 271 Hz), 126.4, 126.9, 127.0
(d, J = 3 Hz), 128.5, 131.5 (q, J = 33 Hz), 141.2, 143.8. HRMS calcd
for C22H24NOF6 432.1762 (M + H+), found 432.1767.
8.34. Isomer B of 4-((3,5-bis(trifluoromethyl)-benzyloxy)-
methyl)-1-methyl-4-phenylazepane (11)
This compound was made from 9 in the same fashion as 10 to
give 11 as a clear sticky oil (2 mg, 67% yield). 1H and 13C NMR data
are the same as 10. MS: 446.04 (MH+). HRMS calcd for C23H26NOF6
446.1919 (M+H+), found 446.1929.
8.35. Isomer A of 4-((3,5-bis(trifluoromethyl)-benzyloxy)-
methyl)-1-ethyl-4-phenylazepane (12)
8.30. Isomer B of 4-((3,5-bis(trifluoromethyl)-benzyl-
oxy)methyl)-4-phenylazepane (9)
To a solution of 8 (3 mg) in glacial acetic acid (0.05 mL) was
added sodium borohydride (2 mg) in two portions, and the reaction
mixture was stirred at rt for 12 h. The mixture was concentrated un-
der vacuum, saturated sodium bicarbonate was added, and the reac-
tion mixture was extracted with dichloromethane. The organic layer
was separated and concentrated in vacuo to give 12 as a colorless
sticky oil (1.8 mg, 56% yield). 1H NMR (CDCl3, 500 MHz) d 1.57
(3H, br. s), 1.68 (1H, m), 1.79 (1H, m), 2.02–2.10 (2H, m), 2.18 (1H,
dd, J = 10, 15 Hz), 2.39 (1H, m), 2.49 (4H, m), 2.73 (2H, br s), 3.46
(2H, dd, J = 5, 15 Hz), 4.44 (2H, s), 7.21 (1H, t, J = 5 Hz), 7.30–7.35
(4H, m), 7.56 (2H, s), 7.73 (1H, s). 13C NMR (CDCl3, 125 MHz) d
12.3, 23.7, 29.7, 34.9, 45.6, 50.1, 52.2, 57.3, 71.7, 81.2, 121.2 (quintet,
J = 2.5 Hz), 123.4 (q, J = 271 Hz), 126.1, 126.9, 127.1 (d, J = 5 Hz),
128.3, 131.6 (q, J = 32.5 Hz), 141.5, 146.3. HRMS calcd for
This compound was made from 44 in the same fashion as com-
pound 8 to give the product as a clear colorless sticky oil (quanti-
tative yield). 1H and 13C NMR data are the same as 8. HRMS calcd
for C22H24NOF6 432.1762 (M + H+), found 432.1774.
8.31. ( )-4-((3,5-Bis(trifluoromethyl)-benzyl-oxy)methyl)-4-
phenylazepane (47)
A mixture of 34 (100 mg, 0.07 mmol) and NaN3 solution (3.0 M
in H2O, 0.15 mL) in trifluoroacetic acid (0.3 mL) was stirred at 65 °C
for 1 h. Trifluoroacetic acid was removed in vacuo, saturated so-
dium bicarbonate was added, and the reaction mixture was ex-
tracted with dichloromethane. The organic layer was separated
and concentrated in vacuo to give the crude product, which was
purified by preparative TLC eluting with 80% ethyl acetate/hexanes
to give ( )-47 (48 mg, 48% yield). 1H NMR (CDCl3, 400 MHz) d 7.74
(1H, s), 7.53 (2H, s), 7.2–7.4 (5H, m), 5.96 (1H, br. s), 4.45 (2H,
apparent s), 3.36 (2H, s), 3.3 (1H, m), 3.1 (1H, m), 2.4 (4H, m), 1.9
(2H, m). MS: 446.50 (M+H+).
C
24H28NOF6 460.2075 (M+H+), found 460.2080.
8.36. Isomer B of 4-((3,5-bis(trifluoromethyl)-benzyloxy)ethyl)-
1-ethyl-4-phenylazepane (13)
This compound was made from 9 in the same fashion as 12
from 8. 1H and 13C NMR data are the same as 12. HRMS calcd for
C
24H28NOF6 460.2075 (M+H+), found 460.2075.
8.32. ( )-4-((3,5-Bis(trifluoro-methyl)-benzyl-oxy)methyl)-4-
phenylazepane (( )8/9)
8.37. Isomer A of 4-((3,5-bis(trifluoromethyl)-benzyloxy)-
methyl)-1-benzyl-4-phenylazepane (14)
A solution of 47 (38 mg, 0.09 mmol) in boraneꢀTHF complex
solution (1.5 M in diethyl ether, 0.23 mL) was stirred at rt over-
night and then heated at 65 °C for 1 h. After being allowed to cool
down, the reaction mixture was concentrated under vacuum,
A mixture of 8 (3 mg) and benzaldehyde (3 mg) in methanol
(0.15 mL) was stirred at room temperature for 0.5 hr, then sodium
cyanoborohydride (1.0 M solution in THF, 0.02 mL) was added

Y.-J. Wu et al. / Bioorg. Med. Chem. 21 (2013) 2217–2228
2227
followed by 1 drop of trifluoroacetic acid. The mixture was stirred
at rt overnight and concentrated under vacuum. Saturated sodium
bicarbonate was added, and the reaction mixture was extracted
with dichloromethane. The organic layer was separated and con-
centrated in vacuo to give 14 as a colorless oil (1.9 mg, 52% yield).
1H NMR (CDCl3, 400 MHz) d 1.62–1.72 (1H, m), 1.72–1.82 (1H, m),
2.00–2.08 (2H, m), 2.18–2.24 (1H, m), 2.30–2.36 (1H, m), 2.48–2.51
(2H, m), 2.65 (2H, m), 3.45 (2H, brd s), 3.54 (2H, brd s), 4.43 (2H, s),
7.21–7.33 (10H, m), 7.56 (2H, s), 7.73 (1H, s). HRMS calcd for
and the reaction mixture was stirred at rt for 2 h. Saturated sodium
chloride was added, and the reaction mixture was worked up with
ethyl acetate. The residue was purified by silica gel flash chroma-
tography eluting with 10% ethyl acetate/90% hexanes to give 48
as an oil (773 mg, 38%). 1H NMR (400 MHz, CDCl3) d 1.8–2.3 (8H,
m), 4.95 (4H, m), 5.73 (2H, m), and 7.31 (5H, m).
8.44. (Z)-1-Phenylcyclohept-4-enecarbonitrile (49)
C
29H30NOF6 522.2232 (M+H+), found 522.2241.
To a solution of 48 (1.5 g, 6.7 mmol) in dichloromethane
(140 mL) was added benzylidene-bis(tricyclohexylphosphine)-
dichlororuthenium (270 mg, 0.3 mmol) and the resulting suspen-
sion was heated under reflux for 1 h. The solvent was removed in
vacuo, and the crude product was purified by silica gel chromatog-
raphy eluting with 10% ethyl acetate/hexanes to give 49 as an oil
(1.3 g, 100%). 1H NMR (400 MHz, CDCl3) d 1.93 (2H, t, J = 12.4 Hz),
2.11 (2H, m), 2.30 (2H, m), 2.57 (2H, t, J = 12.0 Hz), 5.91 (2H, m),
7.38 (5H, m).
8.38. Isomer B of 4-((3,5-bis(trifluoromethyl)-benzyloxy)-
methyl)-1-benzyl-4-phenylazepane (15)
This compound was made from 9 in the same fashion as 14
from 8. 1H NMR data are the same as 20. HRMS calcd for
C29H30NOF6 522.2232 (M+H+), found 522.2219.
8.39. Isomer A of 3-((3,5-
bis(trifluoromethyl)benzyloxy)methyl)-3-phenyl-azepane (16)
8.45. (Z)-1-Phenylcyclohept-4-enecarbaldehyde (50)
This compound was made from 45 in the same fashion as 8
from 43. 1H NMR (CDCl3, 400 MHz) d 1.4-2.0 (5H, m), 2.23 (1H,
dd, J = 6, 13.6 Hz), 2.74 (1H, m), 2.98 (1H, m), 2.22 (1H, d,
J = 14.4 Hz), 3.27 (1H, d, 14.4 Hz), 3.51 (1H, d, 8.8 Hz), 3.56 (1H,
d, J = 8.8 Hz), 4.58 (2H, s), 7.1–7.4 (5H, m), 7.60 (2H, s), and 7.74
(1H, s). 13C NMR (CDCl3, 125 MHz) d 22.7, 32.0, 35.5, 47.8, 51.2,
56.1, 71.9, 79.7, 121.3 (quintet, J = 4 Hz), 123.4 (q, J = 271 Hz),
126.3, 126.9, 127.1 (d, J = 3 Hz), 128.4, 131.7 (q, J = 25 Hz), 141.5,
To a solution of 49 (1.1 g, 5.6 mmol) in toluene (40 mL) at
ꢁ78 °C was added DIBAL-H (1.0 M solution in toluene, 11.2 mL,
11.2 mmol), and the resulting solution was stirred at ꢁ78 °C for
1 h. The reaction was quenched with saturated ammonium and
then extracted with ethyl acetate. The organic layer was separated
and concentrated in vacuo and the residue was purified by silica
gel chromatography eluting with 10% ethyl acetate/hexanes to give
50 (800 mg, 73%). 1H NMR (400 MHz, CDCl3) d 2.21 (6H, m), 2.37
(2H, m), 5.69 (2H, s), 7.30 (5H, m), and 9.38 (1H, s).
145.0. HRMS calcd for
432.1763.
C
22H24NOF6 432.1762 (M+H+), found
8.46. (Z)-(1-Phenylcyclohept-4-enyl)methanol (51)
8.40. Isomer B of 3-((3,5-
bis(trifluoromethyl)benzyloxy)methyl)-3-phenylazepane (17)
To a solution of 50 (800 mg, 4 mmol) in methanol (30 mL) was
added sodium borohydride (304 mg, 8 mmol), and the resulting
mixture was stirred at room temperature for 1 h. Methanol was re-
moved in vacuo, water was added, and the reaction mixture was
extracted with ethyl acetate The organic layer was separated and
concentrated in vacuo to give 51 as an oil (830 mg, 100%). 1H
NMR (400 MHz, CDCl3) d 1.91 (2H, m), 2.20 (6H, m), 3.55 (2H, s),
5.68 (2H, m), 7.35 (5H, m). 13C NMR (CDCl3, 100 MHz) d 24.0,
32.9, 47.3, 71.7, 126.3, 127.4, 128.6, 131.2, 144.3.
This compound was made from 46 in the same fashion as 8
from 43. 1H and 13C NMR data are the same as 16. HRMS calcd
for C22H24NOF6 432.1762 (M+H+), found 432.1762.
8.41. Isomer A of 3-((3,5-bis(trifluoromethyl)benzyloxy)-
methyl)-1-methyl-3-phenylazepane (18)
This compound was made from 16 in the same fashion as 10
from 8. 1H NMR (CDCl3, 400 MHz) d 1.6 (4H, m), 1.8 (1H, m), 2.1
(1H, m), 2.39 (3H, s), 2.45 (1H, m), 2.65 (1H, m), 2.79 (1H, AB quar-
tet, J = 16 Hz), 2.94 (1H, AB quartet, J = 16 Hz), 3.57 (1H, AB quartet,
J = 12 Hz), 3.71 (1H, d, J = 12 Hz0, 4.48 (1H, d, J = 12 Hz), 4.51 (1H,
AB quartet, J = 12 Hz), 7.2–7.4 (5H, m), 7.60 (2H, s), and 7.73 (1H, s).
13C NMR (CDCl3, 125 MHz) d 22.4, 29.7, 36.6, 46.5, 49.1, 60.5, 64.8,
71.8, 78.6, 121.3 (quintet, J = 3 Hz), 123.4 (q, J = 271 Hz), 126.1,
126.7, 127.1, 128.2, 131.6 (q, J = 32.5 Hz), 141.6, 146.2. HRMS calcd
for C23H26NOF6 446.1919 (MH+), found 446.1924.
8.47. (Z)-5-((3,5-Bis(trifluoromethyl)benzyloxy)-methyl)-5-
phenylcyclohept-1-ene (52)
To a solution of 51 (830 mg, 4.11 mmol) and 3,5-(bis-trifluoro-
methyl)benzyl bromide (0.79 mL, 4.31 mmol) in DMF (8 mL) at
0 °C was added sodium hydride (95% oil dispersion, 156 mg,
6.17 mmol) and the resulting suspension was stirred at rt for 1 h.
The reaction was quenched with water and then extracted with
ethyl acetate. The organic layer was separated and concentrated
in vacuo and the residue was purified by silica gel chromatography
eluting with 10% ethyl acetate/90% hexanes to give 52 as an oily
material (1.1 g, 61%). 1H NMR (400 MHz, CDCl3) d 1.98 (2H, m),
2.15 (4H, m), 2.28 (2H, m), 3.47 (2H, s), 4.42 (2H, s), 5.67 (2H,
m), 7.22 (1H, m), 7.34 (2H, m), 7.39 (2H, m), 7.54 (2H, s), and
7.72 (1H, s). HRMS m/z calcd for C23H21F6O (MꢁH)ꢁ 427.1497,
found 427.1481.
8.42. Isomer B of 3-((3,5-bis(trifluoromethyl)benzyloxy)-
methyl)-1-methyl-3-phenylazepane (19)
This compound was made from 17 in the same fashion as 10
from 8. 1H and 13C NMR data are the same as 24. HRMS calcd for
C
23H26NOF6 446.1919 (M+H+), found 446.1914.
8.43. 2-Butyl-2-phenylhex-5-enenitrile (48)
8.48. 4-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-4-
phenylcycloheptanol (53)
To a solution of 2-phenylacetonitrile (1 mL, 8.7 mmol) in DMF
(1.0 mL) at 0 °C was added sodium hydride (95% oil dispersion,
547 mg, 21.7 mmol) and the resulting suspension was stirred at
0 °C for 10 min. 4-Bromobut-1-ene (2.2 mL, 21.7 mmol) was added,
To a solution of 52 (56 mg, 0.13 mmol) in THF (0.55 mL) at 0 °C
was added borane-tetrahydrofuran complex (1.5 M solution,
0.17 mL, 0.26 mmol) dropwise, and the resulting solution was

2228
Y.-J. Wu et al. / Bioorg. Med. Chem. 21 (2013) 2217–2228
allowed to warm to rt and stirred at rt for 12 h. The reaction mix-
ture was cooled to 0 °C, and water (10.50 mL) was added slowly
followed by 30% hydrogen peroxide (0.19 mL) and 1 N sodium
hydroxide (0.30 mL). The resulting solution was stirred at rt for
5 min and then extracted with ethyl acetate. The organic layer
was separated and concentrated in vacuo and the residue was puri-
fied by preparative TLC eluting with 40% ethyl acetate/60% hexanes
to give 52 as an oily material (27 mg, 47%). 1H NMR (400 MHz,
CDCl3) d 1.3-2.4 (m), 3.40 (s), 3.44 (s), 4.43 (s), 7.32 (m), 7.56 (s),
and 7.73 (s). HRMS m/z calcd for C23H23F6O2 (MꢁH)ꢁ 445.1602,
found 445.1607.
sealed vial for 3 h. The solution was allowed to cool to room tem-
perature, and methanol (0.10 mL) was added slowly followed by
1 N hydrochloric acid (0.10 mL), and the reaction mixture was
heated at 65 °C for 2 h. The solvents were removed in vacuo, 1 N
sodium hydroxide was added, and the reaction mixture was ex-
tracted with dichloromethane. The organic layer was separated
and concentrated in vacuo to give 20 as an oil (16 mg, 89%). 1H
NMR (400 MHz, CDCl3) d 1.58 (5H, m), 2.09 (3H, m), 2.30 (1H,
m), 2.94 (4H, m), 3.44 (1H, d, J = 8.8 Hz), 3.47 (1H, d, J = 8.8 Hz),
4.42 (2H, s), 7.21 (1H, m), 7.37 (4H, m), 7.55 (2H, s), and 7.73
(1H, s). MS: 446.03 (M+H+).
8.49. 4-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-4-
phenylcycloheptanone (54)
8.51. 5-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-5-
phenylazocane (21)
To a solution of 53 (26 mg, 0.058 mmol) in dichloromethane
(0.5 mL) at rt were added pyridinium chlorochromate (25 mg)
and powered 4A° molecular sieves (26 mg), and the resulting mix-
ture was stirred at rt for 1.5 h and then filtered through a small pad
of silica gel eluting with dichloromethane. The filtrate was evapo-
rated in vacuo, and the residue was purified by preparative TLC
eluting with 30% ethyl acetate/70% hexanes to give 54 as an oily
material (21 mg, 81%). 1H NMR (400 MHz, CDCl3) d 1.74 (3H, m),
1.98 (1H, dd, J = 11.6, 1.2 Hz), 2.48 (6H, m), 3.37 (1H, d,
J = 8.8 Hz), 3.39 (1H, d, J = 8.8 Hz), 4.42 (2H, s), 7.28 (1H, m), 7.35
(4H, m), 7.56 (2H, m), and 7.74 (1H, s). HRMS m/z calcd for
This compound was obtained from lactam 56 in the same fash-
ion as 20 from 55. 1H NMR (400 MHz, CDCl3) d 1.57 (4H, m), 1.87
(1H, br s), 2.09 (4H, m), 2.72 (2H, m), 2.88 (2H, m), 3.43 (2H, s),
4.41 (2H, s), 7.21 (1H, m), 7.39 (4H, m), 7.54 (2H, s), and 7.21
(1H, s). MS: 446.03 (M+H+).
References and notes
1. Corby, C. L.; Gabriel, D. J. Serotonin Res. 1997, 4, 47.
2. Isaacson, E. I., Central Nervous System Stimulants. In Wilson and Gisvold’s
Textbook of Organic Medicinal and Pharmaceutical Chemistry, 10th ed., Delgado, J.
N.; Remers, W. A., Eds.; Lippincott-Raven: New York, 1998; Chapter 15, pp 471–
474.
3. Hu, X. H.; Bull, S. A.; Hunkeler, E. M.; Ming, E.; Lee, J. Y.; Fireman, B.; Markson, L.
E. J. Clin. Psychiatry 2004, 65, 959.
C
23H21F6O2 (MꢁH)ꢁ 443.1446, found 443.1457.
6-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-6-phenylazocan-
2-one (55) and 5-((3,5-bis(trifluoro-methyl)benzyloxy)methyl)-5-
phenylazocan-2-one (56). To a solution of 54 (135 mg, 0.3 mmol)
in trifluoroacetic acid (0.6 mL) was added sodium azide (3.0 M
solution in water, 0.2 mL, 0.6 mmol) and the resulting solution
was heated at 65 °C for 1 h. Trifluoroacetic acid was removed in va-
cuo, saturated sodium bicarbonate solution was added, and the
reaction mixture was extracted with ethyl acetate. The organic
layer was separated and concentrated in vacuo and the residue
was purified by preparative TLC eluting with 60% ethyl acetate/
40% hexanes to give 55 (37 mg, 26%) and 56 (44 mg, 31%).
Compound 55: 1H NMR (400 MHz, CDCl3) d 1.5-2.2 (7H, m), 2.63
(1H, dd, J = 5.5, 115.0 Hz), 3.31 (1H, d, J = 8.5 Hz), 3.37 (1H, d,
J = 8.5 Hz), 3.44 (1H, m), 3.77 (1H, m), 4.43 (1H, d, J = 8.0 Hz),
4.46 (1H, d, J = 8.0 Hz), 5.97 (1H, s), 7.26 (3H, m), 7.36 (2H, m),
7.49 (2H, s), and 7.73 (1H, s). 13C NMR (CDCl3, 125 MHz) d 22.9,
30.2, 31.8, 38.8, 40.4, 45.5, 71.8, 84.1, 121.5, 123.4 (q,
J = 262.5 Hz), 126.7, 127.1, 127.5, 128.5, 131.6 (q, J = 37.5 Hz),
141.1 (d, J = 25.0 Hz), 176.8.
4. Gelenberg, A. J.; Chesen, C. L. J. Clin. Psychiatry 2000, 61, 712.
5. Katz, M. M.; Tekell, J. L.; Bowden, C. L.; Brannan, S.; Houston, J. P.; Berman, N.;
Frazer, A. Neuropsychopharmacology 2004, 29, 566.
6. Arborelius, L. I. Drugs 1999, 2, 121.
7. Artigas, F.; Romero, L.; de Montigny, C.; Blier, P. Trends Neurosci. 1996, 19, 378.
8. For a review on NK1R antagonists, see Swain, C. J. In Neurokinin receptor
Antagonists in Progress in Medicinal Chemistry; Ellis, G. P., Luscombe, D. K.,
Oxford, A. W., Eds.; Elsevier Science B. V., 1998; pp 57–81. Chapter 2.
9. Aprepitant: Hale, J. F.; Mills, S. G.; MacCoss, M.; Finke, P. E.; Cascieri, M. A.;
Sadowski, S.; Ber, E.; Chicchi, G. G.; Kurtz, M.; Metzger, J.; Eiermann, G.; Tsou, N.
N.; Tattersall, F. D.; Rupniak, N. M. J.; Williams, A. R.; Rycroft, W.; Hargreaves,
R.; MacIntyre, D. E. J. Med. Chem. 1998, 41, 4607.
10. Vestipitant: Fabio, R. D.; Griffante, C.; Alvaro, G.; Pentassuglia, G.; Pizzi, D. A.;
Donati, D.; Rossi, T.; Guercio, G.; Mattioli, M.; Cimarosti, Z.; Marchioro, C.;
Provera, S.; Zonzini, L.; Montanari, D.; Melotto, S.; Gerrard, P. A.; Trist, D. G.;
Ratti, E.; Corsi, M. J. Med. Chem. 2009, 52, 3238.
11. Casopitant: DiFabio, R.; Alvaro, G.; Griffante, C.; Pizzi, D. A.; Donati, D.; Mattioli,
M.; Cimarosti, Z.; Guercio, G.; Marchioro, C.; Provera, S.; Zonzini, L.; Montanari,
D.; Melotto, S.; Gerrard, P. A.; Trist, D. G.; Ratti, E.; Corsi, M. J. Med. Chem. 2011,
54, 1071–1079.
12. L-733060: Harrison, T.; Swain, B. J. W.; Ball, R. G. Bioorg. Med. Chem. Lett. 1994,
4, 2545.
13. IDdb Meeting Report 2000, July 9–13.
14. Merck & Co Inc. Press Release, November 12, 2003.
15. Santarelli, L.; Gobbi, G.; Debs, P. C.; Sibille, E. T.; Blier, P.; Hen, R.; Heath, M. J.
Proc. Natl. Acad. Sci. U.S.A. 1912, 2001, 98.
16. Haddjeri, N.; Blier, P. NeuroReport 2000, 11, 1323.
17. Chenu, F.; Guiard, B. P.; Bourin, M.; Gardier, A. M. Behav. Brain Res. 2006, 172,
256.
18. Ryckmans, T.; Balancon, L.; Berton, O.; Genicot, C.; Lamberty, Y.; Lallemand, B.;
Pasau, P.; Pirlot, N.; Quere, L.; Talaga, P. Bioorg. Med. Chem. Lett. 2002, 12, 261.
19. Ryckmans, T.; Berton, O.; Crimee, R.; Kogej, T.; Lamberty, Y.; Pasau, P.; Talaga,
P.; Genicot, C. Bioorg. Med. Chem. Lett. 2002, 12, 3195.
20. Stevenson, G. I.; Huscroft, I.; MacLeod, A. M.; Swain, C. J.; Cascieri, M. A.;
Chicchi, G. G.; Graham, M. I.; Harrison, T.; Kelleher, F. J.; Kurtz, M.;
Ladduwahetty, T.; Merchant, K. J.; Metzger, J. M.; MacLntyre, D. E.; Sadowski;
Sohal, B.; Owens, A. P. J. Med. Chem. 1998, 41, 4623.
21. Name Reactions, Y.-J. Wu. In Schmidt Reactions; Li, J., Ed.; John Wiley and Sons:
Hoboken New Jersey, 2009; pp 353–372. Chapter 1.2.6.
Compound 56: 1H NMR (500 MHz, CDCl3) d 1.67 (1H, m), 1.85
(2H, m), 2.06 (2H, m), 2.49 (1H, m), 2.77 (4H, m), 3.33 (1H, d,
J = 8.5 Hz), 3.37 (1H, d, J = 8.5 Hz), 4.34 (1H, d, J = 4.34 (1H, d,
J = 13.0 Hz), 4.44 (1H, d, J = 13.0 Hz), 5.61 (1H, s), 7.28 (3H, m),
7.38 (t, J = 8.0 Hz), 7.50 (2H, s), and 7.73 (1H, s). 13C NMR (CDCl3,
125 MHz) d 27.6, 29.9, 30.1, 40.9, 45.4, 71.8, 83.7, 121.4, 123.4 (q,
J = 262.5 Hz), 126.6, 127.1, 127.5, 127.6, 128.4, 131.7 (q,
J = 37.5 Hz), 141.1 (d, J = 25.0 Hz), 178.0.
8.50. 4-((3,5-Bis(trifluoromethyl)benzyloxy)methyl)-4-
phenylazocane (20)
22. Keverline-Frantz, K. I.; Boja, J. W.; Kuhar, M. J.; Abraham, P.; Burgess, J. P.;
Lewin, A. H.; Carroll, F. I. J. Med. Chem. 1998, 41, 247.
23. Wallace-Boone, T.; Newton, A. E.; Wright, R. N.; Lodge, N. J.; McElroy, J. E.
Neuropsychopharmacology 2008, 33, 1919–1928.
To a solution of 55 (18 mg, 0.04 mmol) in THF (0.10 mL) at rt
was added borane-THF complex (1.50 M solution in THF, 0.10 mL,
0.16 mmol) and the resulting solution was heated at 65 °C in a