Conformationally restricted homotryptamines 3. Indole tetrahydropyridines and cyclohexenylamines as selective serotonin reuptake inhibitors

Article abstract of DOI:10.1016/j.bmcl.2007.03.040

A series of indole tetrahydropyridine and indole cyclohexenylamines was prepared, and their binding affinities at the human serotonin transporter (SERT) were determined. In particular, a nitrile substituent at the C5 position of the indole ring gave poten

Full text of DOI:10.1016/j.bmcl.2007.03.040

Bioorganic & Medicinal Chemistry Letters 17 (2007) 3099–3104
Conformationally restricted homotryptamines 3. Indole
tetrahydropyridines and cyclohexenylamines as selective
serotonin reuptake inhibitors
Jeffrey A. Deskus,* James R. Epperson, Charles P. Sloan, Joseph A. Cipollina,
Pierre Dextraze, Jingfang Qian-Cutrone, Qi Gao, Baoqing Ma, Brett R. Beno,
Gail K. Mattson, Thaddeus F. Molski, Rudolph G. Krause, Matthew T. Taber,
Nicholas J. Lodge and Ronald J. Mattson
Bristol-Myers Squibb Pharmaceutical Research Institute, 5 Research Parkway, Wallingford, CT 06492-7660, USA
Received 30 January 2007; revised 13 March 2007; accepted 13 March 2007
Available online 16 March 2007
Abstract—A series of indole tetrahydropyridine and indole cyclohexenylamines was prepared, and their binding affinities at the
human serotonin transporter (SERT) were determined. In particular, a nitrile substituent at the C5 position of the indole ring gave
potent SERT activity. The stereochemistry of the N,N-dimethylamine substituent was determined for the most potent indole cyclo-
hexenylamine, 6a. The enantiomers of 6a were energy minimized and compared to other conformationally restricted SSRIs. Com-
pound 6a was found to give a dose–response similar to the SSRI fluoxetine in microdialysis studies in rats.
ꢀ 2007 Elsevier Ltd. All rights reserved.
5-Hydroxytryptamine (5-HT, serotonin) is a neurotrans-
mitter that plays an important role in a variety of
physiological functions in the central nervous system
and peripheral tissues. Consequently, the modulation
of serotonin function via therapeutic agents continues
to be an active and promising area of drug discovery
research.
side chain of 5-methoxy serotonin was replaced with a
tetrahydropyridin-4-yl moiety.3 Similar analogs have
been examined at various 5-HT receptor subtypes.4
We have previously reported SERT binding results with
homotryptamines 1,5 and conformationally restricted
analogs 2,5a 3,6 and 46 (Fig. 1). Based on SAR studies,
we have hypothesized that modifying the RU 24969 side
chain to an N-methyl tetrahydropyrid-4-yl moiety (5)
might yield analogs with high affinity for SERT. Also
of interest was the impact of extending the nitrogen
one more carbon away from the indole as in dimethyl-
amino cyclohexen-4-yl analogs (6). We now report the
The selective serotonin reuptake inhibitors (SSRIs) are
effective antidepressants, and are relatively safe despite
some recognized issues.1 Recent SSRI research has
focused on compounds with added properties that may
result in a more rapid onset of antidepressant action
such as serotonin transporter (SERT) inhibition com-
bined with 5-HT1A, 5-HT1B, or NK1 antagonism.2
R
X
R
RU 24969
X=OMe
The conformational restriction of serotonergic ligands is
a precedented way to improve the binding and selectiv-
ity of these agents. One of the earliest conformationally
restricted homo-serotonin analogs was the selective
5-HT1B agonist RU 24969, in which the amino ethylene
N
H
N
H
4
5
6
X=F
CH2NMe2
1
2
3
-CH2CH2CH2NMe2
X=H,CN
-H2C
N
Me
N
X=H,CN
CH2NMe2
NMe2
X=CN
Keywords: SSRI; Serotonin transporter; Homotryptamine.
*
Corresponding author. Tel.: +1 203 677 6560; fax: +1 203 677
7702; e-mail: jeffrey.deskus@bms.com
Figure 1.
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.03.040

3100
J. A. Deskus et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3099–3104
Table 1. SERT binding affinities11 of tetrahydropyrid-4-yl indoles
SERT activity of N-methyl tetrahydropyrid-4-yl indoles,
5a–h, and N,N-dimethylamino cyclohexen-4-yl indoles,
6a–h (Fig. 1).
N
N-Methyl tetrahydropyrid-4-yl indoles 5a–h were pre-
pared by the reaction of commercially available 5-substi-
tuted indoles with N-methyl piperidone, as is shown in
Scheme 1.
R
N
H
Compound
R
SERT IC50 (nM)11
N,N-Dimethylamino cyclohexen-4-yl indoles 6a–h were
prepared by reductive amination of commercially avail-
able 1,4-cyclohexanedione mono-ethylene ketal with
dimethylamine, then ketal hydrolysis, and coupling with
the analogous indoles,7 as shown in Scheme 2.
5a
5b
5c
5d
5e
5f
CN
NO2
F
19
40
3
6
160 50
210 50
240 50
690 270
730 270
910 90
Br
Cl
H
5g
5h
OMe
Me
Since our previous work on indole SSRIs demonstrated
high eudesmic ratios for the enantiomers of compounds
3 and 46 (similar to the case of (S)- and (R)-citalopram8),
6a was resolved by chiral supercritical fluid chromato-
graphy (SFC).9 One enantiomer, R-6a, was then con-
verted to its N-tosyl derivative (p-TsCl, NaHMDS,
DMF, 0 ꢁC, 30 min.) to determine its absolute stereo-
chemistry by X-ray crystallography (Fig. 2).10
Table 2. SERT binding affinities11 of N,N-dimethylamino cyclohexen-
4-yl indoles, and of the enantiomers of 6a
N
The SERT binding affinities11 of compounds 5a–h are
shown in Table 1. Results for compounds 6a–h and
the enantiomers of 6a are shown in Table 2.
R
N
H
N-Methyl tetrahydropyrid-4-yl compound 5f is less po-
tent than 1 (X = H, SERT IC50 58 6 nM),5a with the
same carbon linker length on the side chain. This loss
of binding affinity may be in part due to the lack of flex-
ibility present in the tetrahydropyrid-4-yl ring system.
By moving the nitrogen outside of the ring system and
incorporating the N,N-dimethylamine moiety, cyclohex-
en-4-yl compound 6f was found to have much improved
Compound
R
SERT IC50 (nM)11
6a
CN
CN
CN
NO2
F
1.6 0.3
1.1 0.2
0.72 0.11
S-6a
R-6a
6b
1
14
0.3
6
6c
6d
Br
Cl
5.3 0.9
19
6e
5
6f
6g
H
3.1 0.5
150 30
770 230
OMe
Me
6h
N
SERT potency. Compound 6f is greater than 200-fold
more potent than 5f. The difference in the two bridging
ring systems is consistent throughout the series, as the
N,N-dimethylamino cyclohexen-4-yl analogs are gener-
ally 10- to 40-fold more active than their N-methyl tetr-
ahydropyrid-4-yl counterparts. The combination of
carbon chain length extension and the preferred amino
functionality contributes to the SERT binding differ-
ences between these two groups of compounds.
R
N
O
R
a
+
N
H
N
H
5 a-h
Scheme 1. Reagents and condition: (a) pyrrolidine in ethanol/reflux.
O
N
In terms of SAR trends for the indole substituent, the
potency in both series is markedly influenced by the
5-substituent on the indole. The nitro (5b and 6b) and
cyano (5a and 6a) analog, were the most potent com-
pounds, suggesting that electron withdrawing substitu-
ents on the indole ring might be favored for SERT
binding. The same was true for the previously reported
5-cyano compounds 1 (X = CN, SERT IC50 2.0
0.4 nM)5a and 2 (X = CN, SERT IC50 2.0 0.3 nM),5a
as they were also the most potent analogs in that series.
a,b
N
O
O
O
+
R
R
c
N
H
N
H
6 a-h
As for the effect of chirality, the R-enantiomer12 of 6a
was found to be nearly equipotent to the corresponding
Scheme 2. Reagents and conditions: (a) Me2NH, NaBH(OAc)3; (b) aq
HCl; (c) pyrrolidine/reflux.

J. A. Deskus et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3099–3104
3101
S-6a and R-6a, were then compared directly in a single,
20-point concentration experiment to determine their
binding affinities at SERT, DAT, and NET.13 The re-
sults of this selectivity study are shown in Table 4.
Both the racemate 6a and S-enantiomer S-6a showed
similar levels of selectivity for SERT over DAT (>140-
fold) and NET (>60-fold). Interestingly, the R-enantio-
mer R-6a gave much higher selectivity ratios for SERT,
>700-fold over DAT, and >900-fold over NET. For
comparison, the SSRI fluoxetine (SERT Ki 0.72 nM,
DAT Ki 1900 nM, and NET Ki 440 nM)6 has SERT
selectivity over DAT (>2600-fold) and NET (>600-
fold). Our previously reported SSRI compound 3 dem-
onstrated very high selectivity for the serotonin trans-
porter over DAT and NET (SERT Ki 0.18 nM, DAT
Ki 2100 nM, and NET Ki 4600 nM).6
In vivo studies with the racemate 6a were performed due
to the limited availability of the single enantiomers. In
microdialysis studies,14 6a robustly increased extracellu-
lar serotonin concentrations in the frontal cortex of
awake, freely moving rats. The maximal response was
achieved at 1 mg/kg (ip) as shown in Figure 3. In oral
studies, 6a increased serotonin levels to 150–200 percent
above baseline. This effect was similar to that produced
by fluoxetine, both given at 10 mg/kg po (Fig. 4).
Figure 2. ORTEP drawing of N-1-tosyl derivative of R-6a with
ellipsoids drawn at 30% probability level for non H-atoms and
arbitrary scaled open circles for H-atoms. Carbon and hydrogen atoms
are not labeled.
S-enantiomer. 6a and its enantiomers are not as potent
as 3 (SERT Ki 0.18 0.02 nM),6 which might suggest
that the constraint of the cyclohexen-4-yl ring system
gives a somewhat less favorable interaction with the
serotonin transporter as compared to the trans cyclopro-
pane moiety of 3.
In order to relate the potency of 6a to other potent
SSRIs, low-energy conformations of the individual
enantiomers were examined.15 Figure 5 shows the pre-
dicted lowest-energy conformation of R-6a and second
lowest-energy conformation of S-6a superimposed with
low-energy conformations of the SSRIs (1S,4S)-sertra-
line and (S)-citalopram (av SERT IC50 0.73 nM (n = 2)
and 1.45 nM (n = 2), respectively,16 Fig. 6). The pre-
dicted energy of the conformation of S-6a shown in
To examine selectivity, the most active compounds at
SERT from this study were evaluated for binding affin-
ity at the human dopamine (DAT) and norepinephrine
(NET) transporters. The results are shown in Table 3,
where percent inhibition was determined at a single con-
centration of 1 lM.
Table 4. SERT, DAT, and NET binding affinities13 of 6a and its
enantiomers
Many of the compounds were weak to moderate inhib-
itors at either DAT, NET or both of these transporters.
Compound 6d was the only strong inhibitor of both
DAT and NET transporters while having single digit
nanomolar affinity at SERT. Compound 6a showed
moderate affinity at DAT and a higher percent inhibi-
tion at NET. This compound and its enantiomers,
Compound
SERT
IC50 (nM)13
DAT
IC50 (nM)13
NET
IC50 (nM)13
6a
0.34
0.56
0.40
97
80
57
37
S-6a
R-6a
290
390
10mg/kg
Compound 6a ip
3.0 mg/kg
1.0 mg/kg
0.3 mg/kg
0.1 mg/kg
Vehicle
Table 3. Inhibition of dopamine and norepinephrine transporters
400
350
300
250
200
150
Compound
SERT IC50
(nM)a
DAT,
% inhibitionb
NET,
% inhibitionb
5a
5b
6a
6b
6c
6d
6e
6f
19
40
3
6
56
73
71
74
23
95
16
67
15
27
90
81
35
99
34
70
1.6 0.3
1
14
0.3
6
100
50
5.3 0.9
19
3.1 0.5
0
5
0
20
40
60
80 100 120 140 160 180 200 220 240 260 280 300
Time (min)
a From Tables 1 and 2.
b Test concentration 1 lM.
Figure 3. Dose–response curves for compound 6a given ip.

3102
J. A. Deskus et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3099–3104
F
Compound 6a (10 mg/kg po)
Fluoxetine (10 mg/kg po)
Vehicle (po)
400
350
300
250
Cl
N
200
150
100
50
Cl
NHMe
O
N
0
(S)-citalopram
(1S,4S)-sertraline
0
20 40 60 80 100 120 140 160 180 200 220 240 260 280 300
Figure 6.
Time (min)
Figure 4. Compound 6a versus fluoxetine and vehicle given po.
compounds should be able to interact with the same
H-bond acceptor in the serotonin transporter.
the figure is 0.1 kcal/mol above that of the putative glo-
bal minimum. The conformations of sertraline and (S)-
citalopram shown in Figure 5 have energies within
1.5 kcal/mol of their respective predicted global minima.
The overlay shows good correspondence between two
key elements of the SSRI pharmacophore in each mole-
cule, namely a substituted aromatic ring and a basic
amine moiety.
Overall, the results of the molecular modeling analysis
suggest that the enantiomers of compound 6a are capa-
ble of adopting similarly shaped low-energy conforma-
tions which match key elements of the SSRI
pharmacophore. Even with such a good fit in terms of
energy minimized conformations, it is the differences in
the conformations of the respective enantiomer pairs
that are reflected in their eudesmic ratios. The SERT
affinity ratio of the enantiomers of 6a is only 1.5, while
that of 3 and its enantiomer is nearly 50-fold.6
Compound 4 and its enantiomer also show a significant
difference, here about 7-fold.6
This overlay also suggests that each enantiomer of com-
pound 6a may share common SERT binding interac-
tions with sertraline and (S)-citalopram. Of particular
note is the overall three-dimensional shape similarity
of the low-energy conformers of both enantiomers of
compound 6a, as shown in Figure 5. This suggests that
they can bind to the serotonin transporter in an analo-
gous manner, forming similar intermolecular interac-
tions with the protein. The high degree of similarity is
consistent with the nearly equal IC50 values determined
for the enantiomers.
In conclusion, extending the basic nitrogen out of the
ring system and increasing the carbon side-chain length,
as with cyclohexen-4-yl compound 6a, gave a 10-fold in-
crease in affinity for the serotonin transporter over that
of tetrahydropyridine 5a. Compound 6a and especially
its R-enantiomer R-6a demonstrate selectivity for the
serotonin transporter over dopamine and norepineph-
rine transporters, but not to the same extent as the ratios
reported for 3. In vivo, 6a demonstrated a robust dose–
response in the microdialysis experiment, increasing
serotonin levels to a similar extent as fluoxetine
(10 mg/kg for both compounds given po). R-6a and
S-6a demonstrated nearly equal affinity for SERT.
Modeling studies suggest that both enantiomers can
adopt similar conformations which overlap well with
our potent and highly selective compounds, 3 and 4. This
In Figure 7, the same predicted low-energy conformers
of both enantiomers of compound 6a described above
are shown overlaid with putative global energy mini-
mum conformers of the potent SSRIs 3 (SERT Ki
0.18 0.02 nM) and 4 (SERT Ki 14 1.9 nM).6 In this
case as well, the correspondence between the aromatic
rings is quite good, and the close grouping of the puta-
tive hydrogen bond acceptor site points suggests that the
Figure 5. Superimposition of R-6a (gray carbon atoms, ball and stick
representation) and S-6a (gray carbon atoms, stick representation)
with the SSRIs sertraline ((1S,4S)-stereochemistry, orange carbon
atoms, stick representation) and (S)-citalopram (yellow carbon atoms,
stick representation). Magenta spheres are putative hydrogen bond
acceptor site points utilized in the RMS fitting procedure. The image
was created with DS Viewer ProTM 6.0 (Accelrys, Inc., San Diego, CA,
2005).
Figure 7. Superimposition of R-6a (gray carbon atoms, ball and stick
representation) and S-6a (gray carbon atoms, stick representation)
with compounds 3 (orange carbon atoms, stick representation) and 4
(yellow carbon atoms, stick representation). Magenta spheres are
putative hydrogen bond acceptor site points utilized in the RMS fitting
procedure. The image was created with DS Viewer ProTM 6.0 (Accelrys,
Inc., San Diego, CA, 2005).

J. A. Deskus et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3099–3104
3103
hexane, then 5% methanol in dichloromethane. Com-
pound 6a (1.3 g, 44% yield) was recovered and converted
to an HCl salt. TOF HRMS m/e 266.1667 (M+H)+; 1H
NMR (400 MHz, acetone-d6): d 8.26 (s, 1H), 7.56 (d,
J = 8.4, 1H), 7.50 (s, 1H), 7.38 (m, 1H), 6.22 (m, 1H), 2.60
(m, 1H), 2.27 (s, 6H), 2.46 (m, 3H), 2.21 (m, 1H), 2.05 (m,
1H), and 1.55 (m, 1H).
work helps to further define the SAR and stereochemical
requirements of the indole alkyl amine SSRI series.
References and notes
1. Ananth, J. Psychother. Psychosom. 1998, 67, 61.
2. (a) Rocco, V. P.; Spinazze, P. G.; Kohn, T. J.; Honigsch-
midt, N. A.; Nelson, D. L.; Wainscott, D. B.; Ahmad, L.
J.; Shaw, J.; Threlkeld, P. G.; Wong, D. T.; Takeuchi, K.
Bioorg. Med. Chem. Lett. 2004, 14, 2653; (b) Oficialdegui,
A. M.; Martinez, J.; Perez, S.; Hears, B.; Irurzun, M.;
Palop, J. A.; Tordera, R.; Lasheras, B.; del Rio, J.;
Monge, A. Farmaco 2000, 55, 345; (c) Evrard, D. A.;
Zhou, P.; Yi, S.; Zhow, D.; Smith, D.; Sullivan, K. M.;
Hornby, G. A.; Scheckter, L. E.; Andree, T. H.; Mew-
shaw, R. E. Bioorg. Med. Chem. Lett. 2005, 15, 911; (d)
Ryckmans, T.; Balancon, L.; Berton, O.; Genicot, C.;
Lamberty, Y.; Lallemand, B.; Pasau, P.; Pirol, N.; Quere,
L.; Talaga, P. Bioorg. Med. Chem. Lett. 2002, 12, 261; (e)
Ryckmans, T.; Berton, O.; Grimee, R.; Kogej, T.; Lam-
berty, Y.; Pasau, P.; Talaga, P.; Genicot, C. Bioorg. Med.
Chem. Lett. 2002, 12, 3195.
3. (a) Euvrard, C. R.; Boissier, J. R. Eur. J. Pharmacol 1980,
63, 65; (b) Guillaume, J.; Dumont, C.; Laurent, J.;
Nedelec, L. Eur. J. Med. Chem. 1987, 22, 33; (c) Macor,
J. E.; Burkhart, C. A.; Heym, J. H.; Ives, J. L.; Lebel, L.
A.; Newman, M. E.; Neilson, J. A.; Ryan, K.; Schulz, D.
W.; Torgersen, L. K.; Koe, B. K. J. Med. Chem. 1990, 33,
2087.
8. Chen, F.; Larsen, M. B.; Sanchez, C.; Wiborg, O. Eur.
Neuropsychopharamacol. 2005, 15, 193.
9. Preparative
SFC
separation:
Chiralpak
AS-H,
30 · 250 mm, 5 lm, 88% CO2/12% ethanol with 0.1%
diethylamine, 50 mL/min @ 150 bar, peak 2 corresponds
to R-6a.
10. Full crystallographic data have been deposited to the
Cambridge Crystallographic Data Center (CCDC refer-
ence number 615920). Copies of the data can be obtained
free of charge via the internet at http://
www.ccdc.cam.ac.uk.
11. The SERT binding affinities were determined using
membrane homogenates from HEK-293 cells that stably
expressed human serotonin transporters (HEK-hSERT
cells). Membrane homogenates were incubated with 2 nM
[3H]citalopram (specific activity = 85 Ci/mmol) and
increasing concentrations of test compounds for 1 h at
25 ꢁC in a total volume of 250 lL. Amount of radioligand
bound in the presence and absence of a competitor was
analyzed by plotting (À)log drug concentration versus the
amount of radioligand specifically bound. Non-specific
binding was defined with 10 lM fluoxetine. The midpoint
of the displacement curve (IC50 nM) signified the potency.
n = 3 for each reported IC50; each n represents a duplicate
measurement taken at five different concentrations and
reported as the average SEM (standard error measure-
ment). n = 2 for compounds with an IC50 > 500. n = 5 for
compound 6a.
4. Cole, D. C.; Ellingboe, J. W.; Lennox, W. J.; Mazanda-
rani, H.; Smith, D. L.; Stock, J. R.; Zhang, G.; Zhou, P.;
Schechter, L. E. Bioorg.. Med. Chem. Lett. 2005, 15, 379.
5. (a) Schmitz, W. D.; Denhart, D. J.; Brenner, A. B.; Ditta,
J. L.; Mattson, R. J.; Mattson, G. K.; Molski, T. F.;
Macor, J. E. Bioorg. Med. Chem. Lett. 2005, 15, 1619; (b)
Denhart, D. J.; Mattson, R. J.; Ditta, J. L.; Macor, J. E.
Tetrahedron Lett. 2004, 45, 3803.
12. Analytical data for (R)-3-(4-(dimethylamino)cyclohex-1-
enyl)-1H-indole-5-carbonitrile, compound R-6a: HPLC
purity: >98%; >99%ee; Optical rotation: +51.9 (methanol,
c = 3.0 mg/mL).
6. Mattson, R. J.; Catt, J. D.; Denhart, D. J.; Deskus, J. A.;
Ditta, J. L.; Higgins, M. A.; Marcin, L. R.; Sloan, C. P.;
Beno, B. R.; Gao, Q.; Cunningham, M. A.; Mattson, G.
K.; Molski, T. F.; Taber, M. T.; Lodge, N. J. J. Med.
Chem. 2005, 48, 6023.
13. SERT, DAT, and NET binding affinities were determined
using membrane homogenates from stably transfected
HEK-293 cell lines expressing the human form of the
transporters. Membrane homogenates were incubated
with 125I labeled ligands. RTI-55 (Perkin-Elmer) was used
for SERT (260 pM) and DAT (125 pM). A custom labeled
125I Nisoxetine (Perkin-Elmer) was used for NET
(300 pM). The reactions were carried out in a 384-well
format and harvested in 384-well filter plates. The IC50
data was determined from a 20-point curve. Non-specific
binding was defined using 10 lM Fluoxetine for SERT,
100 lM Desipramine for NET, and 10 lM GBR-12935 for
DAT.
7. Preparation of 6a: 1,4-cyclohexanedione mono-ethylene
ketal and N,N-dimethylamine were dissolved in dichloro-
methane at room temperature with magnetic stirring.
Sodium triacetoxyborohydride was added and the reaction
mixture was stirred for 16 h. A 2 N sodium hydroxide
solution was added, and the mixture was extracted with
dichloromethane, dried over magnesium sulfate, filtered,
and concentrated under vacuum to an oil. The crude oil
was dissolved in diethyl ether at room temperature with
stirring. A 3 M solution of hydrochloric acid was added
and stirring continued for 2 h. The mixture was neutral-
ized, extracted into ethyl acetate, dried over magnesium
sulfate, filtered, and then concentrated under vacuum to
an oil that was used without further purification. This
ketone (1.9 g, 13.5 mmol, 1.2 equiv) and 5-cyanoindole
(1.75 g, 11.2 mmol) were dissolved in 10 mL of ethanol at
room temperature with stirring. Pyrrolidine (2.8 mL,
33.7 mmol, 3 equiv) was added and the reaction mixture
was heated to reflux with stirring for 24 h. The mixture
was cooled to room temperature, diluted with ethanol and
ether, and extracted with 3 M HCl solution. The aqueous
layer was neutralized, extracted with ethyl acetate, dried
over magnesium sulfate, filtered, and concentrated under
vacuum to an oil. The oil was purified by silica gel flash
column chromatography, eluting with 50% ethyl acetate in
14. Taber, M. T.; Wright, R. N.; Molski, T. F.; Clarke, W. J.;
Brassil, P. J.; Denhart, D. J.; Mattson, R. J.; Lodge, N. J.
Pharmacol. Biochem. Behav. 2005, 80, 521.
15. The conformations of the R- and S-enantiomers of
compound 6a were identified as follows: conformational
searches for the N-protonated states of each enantiomer
¨
were performed using MacroModelꢂ 9.1 (Schrodinger,
LLC, New York, NY, 2005),a the OPLS2001 force-field,b
GB/SA water solvation model,c and PRCG minimization
algorithm.d The geometry of each conformer of enantio-
mers S-6a and R-6a was then optimized using density
functional theory (DFT) at the RB3LYP/6-31+G* levele,f
with solvation effects approximated by a self-consistent
reaction field (SCRF) water solvation model. Jaguar 6.5
(Schro¨dinger, LLC, New York, NY, 2005) was used to
perform the DFT calculations. The above computational
protocol is the same as described in reference 6 except for

3104
J. A. Deskus et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3099–3104
the version of the Jaguar program used. The conforma-
In both figures, all other compounds were fit to sertraline,
even though it is not shown explicitly in Figure 7; (a)
Mohamadi, F.; Richards, N. G. J.; Guida, W. C.;
Liskamp, R.; Lipton, M.; Caufield, C.; Chang, G.;
Hendrickson, T.; Still, W. C. J. Comput. Chem. 1990, 11,
440; (b) Jorgensen, W. L.; Maxwell, D. S.; Tirado-Rives, J.
J. Am. Chem. Soc. 1996, 118, 11225; (c) Still, W. C.;
Tempczyk, A.; Hawley, R. C.; Hendrickson, T. A.
J. Am. Chem. Soc. 1990, 112, 6127; (d) Polak, E.; Ribiere,
G. Rev. Fr. Inf. Rech. Oper. 1969, 16, 35; (e) See: Jaguar
6.5 Operating Manual; Schrodinger, LLC: New York, NY,
2005, and references therein; (f) Hehre, W. J.; Radom, L.;
Schleyer, P. v. R.; Pople, J. A. Ab Initio Molecular Orbital
Theory; John Wiley: New York, 1986, Chapter 4; (g)
Gundertofte, K.; Bøgesø, K. P.; Liljefors, T. A Stereose-
lective Pharmacophoric Model of the Serotonin Re-
Uptake Site. In Computer-Assisted Lead Finding and
Optimization: Current Tools for Medicinal Chemistry;
van de Waterbeemd, H., Testa, B., Folkers, G., Eds.;
Verlag Helvetica Chimica Acta: Basel, Switzerland, 1997;
pp 443–459.
tions of (1S,4S)-sertraline, (S)-citalopram (Fig. 6), 3, and 4
used in this study are those shown in Figure 5a of
reference 6, and were identified as described therein.
Compounds 3 and 4 correspond to (+)-12a and (À)-16e,
respectively, within reference 6. The DFT-optimized
lowest-energy conformer of R-6a and second lowest-
energy conformer of S-6a were found to superimpose well
with the DFT-optimized low-energy conformer of sertra-
line shown in Figure 5 using a three-point RMS fitting
procedure. The points used for the superposition with
sertraline are phenyl ring centroids (the dichlorophenyl
ring in sertraline was used), basic nitrogen atoms, and
˚
putative hydrogen bond acceptor site points located 2.8 A
from each basic nitrogen atom along the N-H vectors.g
The RMS fitting of (S)-citalopram to sertraline included a
fourth pair of points corresponding to the centroids of the
fluorophenyl ring in (S)-citalopram, and the non-chlori-
nated phenyl ring in sertraline. The superposition shown
in Figure 7 was generated by RMS fitting the low-energy
conformers of R-6a, S-6a, 3, and 4 to the sertraline
reference conformer as described above for R-6a and S-6a.
16. In–house results.