Studies toward the discovery of the next generation of antidepressants. Part 5: 3,4-Dihydro-2H-benzo[1,4]oxazine derivatives with dual 5-HT1A receptor and serotonin transporter affinity

Article abstract of DOI:10.1016/j.bmcl.2005.11.054

The design, synthesis, and structure-activity relationship of two novel classes of benzoxazine derivatives with dual selective serotonin reuptake inhibitors and 5-HT_1A receptor activities are described.

Full text of DOI:10.1016/j.bmcl.2005.11.054

Bioorganic & Medicinal Chemistry Letters 16 (2006) 1338–1341
Studies toward the discovery of the next generation
of antidepressants. Part 5: 3,4-Dihydro-2H-benzo[1,4]oxazine
derivatives with dual 5-HT1A receptor and serotonin
transporter affinity
Dahui Zhou,a,* Boyd L. Harrison,a Uresh Shah,a Terrance H. Andree,b
Geoffrey A. Hornby,b Rosemary Scerni,b Lee E. Schechter,b Deborah L. Smith,b
Kelly M. Sullivanb and Richard E. Mewshawa,*
aChemical and Screening Sciences, Wyeth Research, CN 8000, Princeton, NJ 08543-8000, USA
bNeuroscience, Discovery Research, Wyeth Research, CN 8000, Princeton, NJ 08543-8000, USA
Received 21 October 2005; revised 11 November 2005; accepted 14 November 2005
Available online 5 December 2005
Abstract—The design, synthesis, and structure–activity relationship of two novel classes of benzoxazine derivatives with dual selec-
tive serotonin reuptake inhibitors and 5-HT1A receptor activities are described.
Ó 2005 Elsevier Ltd. All rights reserved.
Selective serotonin reuptake inhibitors (SSRIs) have
achieved great success in treating depression and related
illnesses. SSRIs work by blocking the neuronal reuptake
of serotonin (5-HT), increasing the concentration of
synaptic 5-HT and thus, increasing the activation of
postsynaptic 5-HT receptors. Unfortunately, long-term
treatment is required before clinical efficacy is achieved.
This is speculated to be the result of acute stimulation of
the somatodendritic 5-HT1A autoreceptors that decrease
neuron firing to release serotonin in the forebrain. Only
after desensitization of these autoreceptors, the seroto-
nergic neurons resume normal firing and the eventual in-
crease in 5-HT levels in the major forebrain areas.1–3
more, co-administration of fluoxetine and WAY-
100635 produces an immediate increase in 5-HT levels
in rat frontal cortex using in vivo microdialysis, whereas
this effect cannot be seen with fluoxetine treatment
alone.6 In support of this hypothesis, clinical trials
performed by Artigas et al.7 and Blier and Bergeron8
demonstrated that combination of ( )-pindolol, a non-
selective 5-HT1A antagonist, with the SSRI paroxetine
shortened the onset of antidepressant action to a period
of 3–7 days in contrast to the 2–3 weeks required with
the SSRI alone. Therefore, incorporating both 5-HT1A
antagonism and 5-HT reuptake inhibition within a sin-
gle molecule should provide an immediate increase in
5-HT in frontal cortex, resulting in a rapid onset antide-
pressant. This dual-action feature would thus form the
basis of the next generation of antidepressant
therapy.9–13
It has been proposed that addition of a 5-HT1A antago-
nist component to the action of an SSRI can limit the
negative feedback through blockage of the autoreceptor,
allowing an immediate increase in synaptic 5-HT. Pre-
clinical evidence has shown that acute SSRI-induced
increases in forebrain 5-HT can be significantly aug-
mented by co-treatment with either pindolol or the more
selective 5-HT1A antagonist WAY-100635.4,5 Further-
Previous reports from our laboratories have revealed
two novel series capable of modulating both SSRI and
5-HT1A receptor activities by linking an aryloxyethyl-
amine with a 3-indoletetrahydropyridine (1) or a 3-indo-
lealkylamine (2) through a common basic nitrogen.12,13
In both series, aryloxyethylamines have been successful-
ly utilized to produce potent 5-HT1A ligands, while the
indoletetrahydropyridine or 3-indolealkylamine moie-
ties have served as robust 5-HT uptake inhibitor
Keywords: Antidepressants; Serotonin transporter; 5-HT1A receptor.
*
Corresponding authors. Tel.: +1 732 274 4423; fax: +1 732 274
4505; e-mail: zhoud@wyeth.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.11.054

D. Zhou et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1338–1341
1339
R2
N
H
H
N
moieties. However, shortcomings were observed in both
N
series, that is, lack of selectivity over the a1 receptor. In
order to expand upon the type of structures that can
potentially maintain both desired activities, we decided
to constrain the 5-HT1A pharmacophore within a cyclic
ring, benzoxazine ring, in a manner similar to that pre-
viously exemplified with benzodioxan moiety.14 We are
particularly interested in the benzoxazine system, since
from a synthetic point of view, the benzoxazine analog
offered the advantage of constraining the aryloxyethyl-
amine moiety, as well as allowing us to probe another
region in space via functionalization of the nitrogen of
the oxazine ring. We now report the syntheses and the
structure–activity relationships (SARs) of two new clas-
ses of benzoxazine 3-indolealkylamines 3a–q and ben-
zoxazine 3-indoletetrahydropiridine analogs 4a–e.
b
a
or c
OEt
OH
OH
O
O
O
O
35 R2 = Boc, benzyl
34
31
d-f
or d-e,g
R2
R2
N
3l: h-i, g
3m: h, g
N
NH
H
N
NH2
O
3k: h, g
O
n
36 R2= Boc, benzyl, Me
3l, 3k, 3m
n = 1, 2
3c and 3o: j
3e: j, i
R2
N
NH
H
N
O
3c, 3e, 3o
F
Scheme 2. Reagent and conditions: (a) LiBH4/THF, rt; (b) (t-Boc)2O/
THF, reflux; (c) BnBr/Et3N/CH3CN, rt; (d) TsCl/Py, rt; (e) NaN3/
DMF, 60 °C; (f) Ph3P/THF–H2O, rt; (g) LiAlH4/THF, reflux; (h)
3-indolealkanoic acid/DMAPC/CH2Cl2, rt; (i) TFA/CH2Cl2, rt; (j)
5-fluoro-3-indolepropylbromide/Et3N/DMSO, 90 °C.
Optimized 5-HT1A component
Optimized 5-HT1A component
H
N
R
N
H
N
O
(CH2)n
O
Ar
Ar
R
2
1
NH
5-HT reuptake component
5-HT reuptake component
R2
N
H
N
H
N
O
O
R2
R2
NH
a
b
R3
MeO
MeO
MeO
N
N
OEt
OH
or c
OH
H
O
O
R1
R1
O
(CH2)n
N
N
O
37
O
39
38
R3
R2 = CH3, C2H5, n-pr, Boc
3
N
4
R2
N
H
H
N
d
NH
H
e-f
MeO
MeO
N
OTs
O
O
40
3a-b
F
g or g,f
Schemes 1–4 show the syntheses of target molecules 3a–
q and 4a–e. The condensation of commercially available
2-amino-phenol and 2,3-dibromopropionic acid ethyl
ester afforded ethyl 3,4-dihydro-2H-benzo[1,4]-oxazine-
2-carboxylate 31 in moderate yield (33%). The nitrogen
of the oxazine ring was alkylated under standard condi-
tions and the ester reduced to generate benzoxazine
derivative 32. Alcohols 32 were converted to the tosy-
lates 33 using p-toluenesulfonyl chloride in pyridine,
and coupling of tosylates 33 with different 3-indolepro-
R2
N
NH
N
O
F
OCH3
4a-c, 4e
Scheme 3. Reagent and conditions: (a) LiBH4/THF, rt; (b) (t-Boc)2O/
THF, reflux; (c) R2Br/Et3N/CH3CN, rt; (d) TsCl/Py, rt; (e) 5-fluoro-3-
indolepropylamine/Et3N/DMSO, 90 °C; (f) TFA/CH2Cl2, rt; (g) 5-
fluoro-3-indolepyridine/Et3N/DMSO, 90 °C.
NH
Boc
N
Boc
N
H
N
H
b, c
a
NH2
OH
N
OH
OTs
N
b, e (R2 = Me, n-Pr, i-Pr)
a
O
O
O
or c, e (R2=Ph)
35
OEt
41
4d
R1
R1
O
or d, f (R2 = Et, i-Bu)
31
R1 = H, Cl
O
Scheme 4. Reagent and conditions: (a) TsCl/Py, rt; (b) 3-indolepyri-
dine/Et3N/DMSO, 90 °C; (c) TFA/CH2Cl2, rt.
R2
R2
N
N
g
h
pylamines13 afforded the desired products 3d,f–j,n,p,
and q.
OH
OTs
R1
O
R1
O
32
33
R2
N
NH
H
N
In Scheme 2, reduction of ester 31 with LiBH4 led to
alcohols 34, which were then treated either with di-
tert-butyldicarbonate or benzylbromide to give alcohols
35. Alcohols 35 were converted to amines 36 using a
standard protocol.13 Coupling of 36 to the appropriate
3-indolealkanoic acids, followed by reduction and
deprotection, afforded target compounds 3l,k, and m.
Treatment of amines 36 with 5-fluoro-3-indolepropyl-
bromide13 resulted in target compounds 3c,e, and o.
Where R2 = CH3 (when R1 = Cl),
C2H5, n-pr, i-pr, i-butyl, Ph.
R1
O
n
R3 = H, F; n = 1, 2
3d, 3f-3j, 3n, 3p, 3q
R3
Scheme 1. Reagents and conditions: (a) BrCH2CHBrCO2Et/K2CO3/
acetone, reflux; (b) R2Br or (R2I)/Et3N/CH3CN, reflux; (c) 1,4-cyclo-
hexanedione/p-toluenesulfonic acid/toluene, reflux; (d) (R2CO)2O/
THF, reflux; (e) LiBH4; (f) LiAlH4/THF, rt; (g) TsCl/py, rt; (h) 3-
indoealkylamines/Et3N/DMSO, 90 °C.

1340
D. Zhou et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1338–1341
In Scheme 3, reduction of 3-oxo-benzoxazine deriva-
tive 37 with LiBH4 gave alcohol 38, which was then
treated either with di-tert-butyldicarbonate or different
halides to afford 39. The target compounds 3a and b
were prepared by heating the corresponding tosylates
40 and 5-fluoro-3-indolepropylamine, followed by
deprotection in TFA. Coupling the tosylates 40 with
5-fluoro-3-indolepyridine12 gave target compounds
4a–c and e.
The preparation of the benzoxazine analog 4d is shown in
Scheme 4. Alkylation of tosylate 41 with 3-indole tetrahy-
dropyridine12 gave the desired product 4d in good yield.
All compounds listed in Tables 1 and 2 were the race-
mates.19 Serotonin transporter (r-5-HT-T) and 5-HT1A
receptor (h-5-HT1A) and a1 affinities, as well as 5-
HT1A intrinsic activity (GTPcS Emax) are tabulated in
Tables 1 and 2.
Table 1. Dual SSRI and 5-HT1A receptor activities for compounds 3a–qa
R2
R3
N
H
R1
(CH2)n
N
O
3
N
H
Compound
R1
R2
R3
n
r-5-HT-T Ki (nM)b
h-5-HT1A Ki (nM)c
GTPS Emax (EC50 nM)d
a1 Ki (nM)e
3a
3b
3c
3d
3e
3f
8-OCH3
7-OCH3
H
H
F
F
F
F
F
F
H
H
F
F
H
H
H
F
F
H
H
1
1
1
1
1
1
1
2
1
1
1
1
2
1
1
1
2
3.13 1.73
11.70 4.67
12.43 9.51
15.28 0.39
15.98 0.04
21.13 5.48
22.73 4.63
23.78 5.34
42.58 2.02
51.48 47.41
88.50 33.23
89.75 15.20
108.75 39.24
43% at 100 nM
23% at 100 nM
23% at 100 nM
7% at 100 nM
9.31 1.25
103.00 15.34
8.46 2.27
15.31 2.20
10.98 3.14
2.81 0.10
12.76 2.84
90.42 29.12
75.18 20.45
8.15 0.70
5.86 1.12
6.91 0.52
66 9.2 (52.2)
56 17.7 (181.5)
80 0.7 (167.5)
100 0 (90.0)
100 0 (155.0)
90 7.1 (190.5)
100 0 (196.5)
88 12.7 (329.0)
64 13.4 (336)
88 12.7 (329)
95 7.1 (33.6)
96 6.4 (93.5)
64 7.8 (1385)
—
7.6 3.6
44.5 16.3
H
Me
Et
26% at 100 nM
—
H
H
H
70.0 1.4
—
H
i-Pr
n-Pr
i-Pr
Me
i-Bu
Me
H
3g
3h
3i
H
481 229.8
—
H
7-Cl
H
607 272.9
—
3j
3k
3l
H
28% at 100 nM
26% at 100 nM
H
3m
3n
3o
3p
3q
H
Bn
Ph
Bn
Ph
Ph
76.14 20.62
7.14 0.22
7.25 1.52
30.21 10.96
75.77 8.90
—
65% at 100 nM
H
H
—
—
—
—
—
H
H
—
a Ki, Emax, and EC50 values are means of at least two experiments SEM (performed in triplicate, determined from nine concentrations).
b Binding affinity at rat cortical 5-HT reuptake sites labeled with [3H]-paroxetine.15
c Binding affinity at human 5-HT1A receptors in CHO cells labeled with [3H]-8-OH-DPAT.16
d Stimulation of [35S]-GTPcS binding in CHO cells expressing 5-HT1A receptor;17 Emax refers to maximal agonist effect relative to 5-HT.
e Binding affinity at rat cortical a1 adrenergic receptors labeled with [3H]-prazosin.18
Table 2. Dual SSRI and 5-HT1A receptor activities for compounds 4a–ea
R2
NH
N
R1
N
O
R3
4
Compound
R1
R2
R3
r-5-HT-T Ki (nM)b
h-5-HT1A Ki (nM)c
GTPcS Emax (EC50 nM)d
a1 Ki (nM)e
4a
4b
4c
4d
4e
8-OCH3
8-OCH3
8-OCH3
H
H
F
F
F
H
F
6.94 4.61
12.58 5.76
13.15 1.63
8.37 10.79
22.58 3.64
154.85 3.15
139.9 8.56
35 15.6 (386)
93 9.9 (190)
54 26.9 (418)
62 0.7 (1131)
92 0.7 (135)
14.5 3.6
96.5 29.0
229.0 33.9
16.6 18.1
505.5 222.7
Me
Et
136.45 18.00
804.80 123.33
362.80 46.95
H
n-Pr
8-OCH3
a Ki, Emax, and EC50 values are means of at least two experiments SEM (performed in triplicate, determined from nine concentrations).
b Binding affinity at rat cortical 5-HT reuptake sites labeled with [3H]-paroxetine.15
c Binding affinity at human 5-HT1A receptors in CHO cells labeled with [3H]-8-OH-DPAT.16
d Stimulation of [35S]-GTPcS binding in CHO cells expressing the 5-HT1A receptor;17 Emax refers to maximal agonist effect relative to 5-HT.
e Binding affinity at rat cortical a1 adrenergic receptors labeled with [3H]-prazosin.18

D. Zhou et al. / Bioorg. Med. Chem. Lett. 16 (2006) 1338–1341
1341
Shown in Table 1 are the biological results of target mol-
ecules 3a–q. Consistent with our previous paper (Part
3),13 substitution of the 5-fluoro moieties to the indole
ring resulted in a significant improvement in 5-HT trans-
porter affinity but little effect on the affinity for 5-HT1A
receptor (3c vs k). Interestingly, incorporation of sub-
stitutents at the nitrogen of the oxazine ring had a minor
effect on affinity for the 5-HT1A receptor, indicating a
large tolerance of the 5-HT1A receptor toward structural
variation in this region of the molecule. However, 5-HT
transporter affinity dramatically decreased as the sub-
stitutent on the nitrogen of the oxazine ring increased
in size (3c vs d,f,j,n, and o). A 5-fold increase in 5-HT
transporter affinity was manifested when the methoxy
group was attached at the 8-position (3a vs e). Surpris-
ingly, when the methoxy group was moved to the 7-po-
sition, a 9-fold loss in affinity for 5-HT1A receptor was
found as well as ꢀ4-fold loss in affinity for 5-HT trans-
porter (3a vs b). In contrast to our previous studies
(Parts 2 and 3), three compounds (3c,k, and l) in this
series had very low a1 affinity. Unfortunately, all
compounds from this study were found to be either
5-HT1A agonists or partial agonists as measured by
the GTPcS assay.
pharmacophoric criteria within a single molecule and
to identify novel agents that can represent the next gen-
eration of rapid onset antidepressants.
References and notes
1. Blier, P.; Bergerpm, R. J. Clin. Psychiat. 1998, 16.
2. Kreiss, D. S.; Lucki, I. J. Pharmacol. Exp. Ther. 1995, 274,
219.
3. Hjorth, S.; Auerbach, S. B. Behav. Brain Res. 1996, 73,
281.
4. Romero, L.; Bel, N.; Casanovas, J. M.; Artigas, F. Int.
Clin. Psychopharmacol. 1996, 11, 1.
5. Hjorth, S.; Westlin, D.; Bengtsson, H. J. Neuropharma-
cology 1997, 36, 461.
6. Dawson, L. A.; Nguyen, D. L.; Schechter, L. E. J.
Psychopharacol. 2002, 16, 145.
7. Artigas, F.; Perez, V.; Alvarez, E. Arch. Gen. Psychiat.
1994, 51, 248.
8. Blier, P.; Bergeron, R. J. Clin. Psychopharmacol. 1995, 15,
217.
9. Martinez-Esparza, J.; Oficialdegui, A. M.; Perez-Silanes,
S.; Heras, B.; Orus, L.; Palop, J. A.; Lasheras, B.; Roca, J.;
Mourelle, M.; Bosch, A.; Del Castillo, J. C.; Tordera, R.;
Del Rio, J.; Monge, A. J. Med. Chem. 2001, 44, 418.
10. Takeuchi, K.; kohn, T. J.; Honigschmidt, N. A.; Rocco, V.
P.; Spinazze, P. G.; Koch, D. J.; Atkinsion, S. T.; Hertel,
L. W.; Nelson, D. L.; Wainscott, D. B.; Ahmad, L. J.;
Shaw, J.; Threlkeld, P. G.; Wong, D. T. Bioorg. Med.
Chem. Lett. 2003, 13, 2393.
11. Rocco, V. P.; Spinazze, P. G.; Kohn, T. J.; Honigschmidt,
N. A.; Nelson, D. L.; Wainscott, D. B.; Ahmad, L. J.;
Shaw, J.; Threlkeld, P. G.; Wong, D. T.; Takeuchi, K.
Bioorg. Med. Chem. 2004, 14, 2653.
12. Mewshaw, R. E.; Meagher, K. L.; Zhou, P.; Zhou, D.;
Shi, X.; Scerni, R.; Smith, D. L.; Schechter, L. E.; Andree,
T. H. Bioorg. Med. Chem. Lett. 2002, 12, 307.
13. Mewshaw, R. E.; Zhou, D.; Zhou, P.; Shi, X.; Hornby, G.;
Spangler, T.; Scerni, R.; Smith, D.; Schechter, L. E.;
Andree, T. H. J. Med. Chem. 2004, 47, 3823.
14. Kang, Y.; Stack, G. P. U.S. Patent 5,750,724, 1998;
Husbands, G. M.; Stack, G, P.; Mewshaw, R. E. U.S.
Patent 6,559,169, 2003.
15. Cheetham, S. C.; Viggers, J. A.; Slater, N. A.; Heal, D. J.;
Buckett, W. R. Neuropharmacology 1993, 32, 737.
16. Dunlop, J.; Zhang, Y.; Smith, D. L.; Schechter, L. E.
J. Pharmacol. Toxicol. Methods 1998, 40, 47.
17. Larenzo, S.; Birdsall, N. J. Br. J. Pharmacol. 1993, 109,
1120.
Table 2 depicts the investigation of the indole tetrahydro-
pyridinyl group utilized as a transporter moiety in the
benzoxazine series. Comparing compound 3a to 4a, a
16-fold loss in 5-HT1A affinity was observed. A slight
trend toward lower 5-HT transporter affinity was ob-
served when the nitrogen of the oxazine ring was alkylated
(4a–c and e). A 2-fold loss in affinity 5-HT1A receptor was
found when R2 was an isopropyl group (4a vs e). Surpris-
ingly, though we previously showed that compounds
employing the tetrahydropyridinyl moiety had 5-HT1A
antagonism,12 all the compounds listed in Table 2 were
found to be either full or partial 5-HT1A agonists in the
GTPcS assay in this report. In contrast to most of the
compounds in Table 1, the benzoxazine indole tetrahyd-
ropyridinyl derivatives in Table 2 had higher affinity for
a1 receptor (i.e., 4a,b, and d were more potent for a1
receptor than the 5-HT1A receptor).
In conclusion, we have discovered that the benzoxazine
moiety can be utilized to embrace both the 5-HT1A
pharmacophore along with the 5-HT transporter moiety
to access dual SSRI and 5-HT1A receptor activities. The
selectivity over a1 receptor20 was improved in several
compounds, though most of the compounds in these
two classes were found to function as 5-HT1A receptor
agonists. Studies in our laboratories are continuing to
further understand the 5-HT transporter and 5-HT1A
18. Morrow, A.; Creese, I. Mol. Pharmacol. 1986, 29, 321.
19. For experimental procedures, see: US patents: Mewshaw,
R. E.; Zhou, D. U.S. Patent 6,221,863, 2001; Mewshaw,
R. E.; Shaw, U. U.S. Patent 6,313,114, 2001.
20. Peroutka, S. J.; UÕPrichard, D. C.; Greenberg, D. A.;
Snyder, S. H. Neuropharmacology 1977, 16, 549.