Synthesis and 5-hydroxytryptamine (5-HT) activity of 2,3,4,4a-Tetrahydro-1H-pyrazino[1,2-a]quinozalin-5-(6H)ones and 2,3,4,4a,5,6-hexahydro-1H-pyrazino [1,2-,a] quinoxalines

Article abstract of DOI:10.1016/S0960-894X(00)00400-5

A series of 2,3,4,4a-tetrahydro-1H-pyrazino[1,2-a]quinoxalin-5-(6H)ones and 2,3,4,4a,5,6-hexahydro- 1H-pyrazino[1,2-a]-quinoxalines was shown to exhibit 5-HT(₂C) agonist binding and functional activity. Compound 21R inhibited food intake over 2 h in fasted, male Sprague-Dawley rats with ED₅₀ values of 2 mg/kg (ip) and 10 mg/kg (po). (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0960-894X(00)00400-5

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1991±1994
Synthesis and 5-Hydroxytryptamine (5-HT) Activity of
2,3,4,4a-Tetrahydro-1H-pyrazino[1,2-a]quinoxalin-5-(6H)ones
and 2,3,4,4a,5,6-Hexahydro-1H-pyrazino[1,2-a]quinoxalines
Gregory S. Welmaker,a,* James A. Nelson,a Joan E. Sabalski,a Annmarie L. Sabb,a
John R. Potoski,b Denise Graziano,c Michael Kagan,c Joseph Coupet,d
John Dunlop,d Hossein Mazandarani,d Sharon Rosenzweig-Lipson,d
Stacey Suko€d and Yingxin Zhangd
aMedicinal Chemistry, Chemical Sciences, Wyeth-Ayerst Research, CN8000, Princeton, NJ 08543, USA
bChemical Development, Wyeth-Ayerst Research, 401 N. Middletown Rd, Pearl River, NY 10965, USA
cDiscovery Analytical Chemistry, Chemical Sciences, Wyeth-Ayerst Research, CN8000, Princeton, NJ 08543, USA
dWyeth Neuroscience, Wyeth-Ayerst Research, CN8000, Princeton, NJ 08543, USA
Received 24 May 2000; accepted 29 June 2000
AbstractÐA series of 2,3,4,4a-tetrahydro-1H-pyrazino[1,2-a]quinoxalin-5-(6H)ones and 2,3,4,4a,5,6-hexahydro-1H-pyrazino[1,2-a]-
quinoxalines was shown to exhibit 5-HT2C agonist binding and functional activity. Compound 21R inhibited food intake over 2 h in
fasted, male Sprague±Dawley rats with ED50 values of 2 mg/kg (ip) and 10 mg/kg (po). # 2000 Elsevier Science Ltd. All rights reserved.
Multiple 5-hydroxytryptamine (5-HT; serotonin) recep-
tor types are known to exist and have attracted signi®cant
pharmacological attention, including the 5-HT2 receptor
subfamily.1 The 5-HT2 receptor family currently consists
of three receptor subtypesÐ5-HT2A, 5-HT2B, and 5-HT2C
Ðwhich exhibit similar molecular structure and signal
transduction pathways. Characteristic of the 5-HT2
family of receptors is a relatively low anity for 5-HT, a
high anity for the known 5-HT2 receptor agonist, 2,5-
dimethoxy-4-iodoamphetamine (DOI), and a high anity
for known 5-HT2 receptor antagonists, including ritan-
serin and mesulergine. The 5-HT2 family of receptors
modulates a variety of physiological functions, includ-
ing appetite, mood, nociception, and motor behavior.2
death in the country.5 Obesity is a primary risk factor
for the development of diabetes mellitus, hypertension,
hyperlipidemia, and other cardiovascular disorders. Obe-
sity is also associated with sleep apnea, osteoarthritis,
Obesity is a chronic condition for which no current cure
exists.3 A€ecting approximately one-third of the United
States population, obesity is the most common nutri-
tional problem in the US, and its prevalence worldwide
has risen steadily over the past few decades.4 In 1995,
health care costs attributable to obesity in the United
States were estimated to be nearly $100 billion and it has
been identi®ed as the second leading cause of preventable
Scheme 1. Synthesis of 2,3,4,4a-tetrahydro-1H-pyrazino[1,2-a]quin-
oxalin-5-(6H)ones and 2,3,4,4a,5,6-hexahydro-1H-pyrazino[1,2-a]-
quinoxalines:6 (a) Et3N, DMSO, 60 ꢀC; (b) Fe, AcOH, 60 ꢀC; (c) NaH,
CH3I; (d) H2, 5% Pd/C, EtOH or KOH, H2O:CH3OH, 100 ꢀC or
*Corresponding author. Tel.: +1-732-274-4557; fax: +1-732-274-4505;
e-mail: welmakg@war.wyeth.com
.
HBr, AcOH; (e) BH3 THF, THF.
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00400-5

1992
G. S. Welmaker et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1991±1994
gallbladder disease, degenerative joint disease, respira-
tory disorders, and certain forms of cancer.3,5
and body weight in obese humans by activation of the
5-HT2C receptor.9
There is evidence to suggest that the e€ect of 5-HT on
appetite and food intake is mediated in part by activa-
tion of the 5-HT2C receptors.7 In 1995, Tecott and co-
workers demonstrated that transgenic mice lacking the
5-HT2C receptor became obese due to increased food
intake.2c Additionally, compounds that antagonize the
5-HT2C receptor can lead to increased food intake in
animals and weight gain in humans.8 In 1997, Sargent et
al. reported that m-chlorophenylpiperazine (m-CPP), a
non-selective 5-HT2C partial agonist, reduces appetite
A substructure search through our corporate compound
bank for phenylpiperazine-like agents revealed a series of
2,3,4,4a-tetrahydro-1H-pyrazino[1,2-a]quinoxalin-5-(6H)-
ones, all of which possess the phenylpiperazine moiety,
consistent with m-CPP. Originally evaluated as anti-
hypertensive agents,10 these compounds were chosen to
be screened for their 5-HT2C binding anity. As hits were
identi®ed from this series, additional analogues were syn-
thesized. The general synthetic procedure utilized to pre-
pare these compounds is outlined in Scheme 1.
Table 1. 5-HT2C binding anities for 2,3,4,4a-tetrahydro-1H-pyrazino[1,2-a]quinoxalin-5-(6H)ones and 2,3,4,4a,5,6-hexahydro-1H-pyrazino[1,2-a]-
quinoxalines
5-HT2C binding anitiesa
Antagonist
(Ki (nM)ÆSE or
Agonist
(Ki (nM)ÆSE])
Compounds
R1
R2
R3
R4
R5
X
% inhibition @ 1 mM)
1
H
H
H
Cl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
F
H
H
H
H
H
Cl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
25%
47%
10%
8%
36%
67%
n.d.b
n.d.
n.d.
n.d.
n.d.
n.d.
1R
1S
2
3
4
4R
5
6
7
8
H
OCH3
Cl
Cl
CH3
OH
F
H
H
H
H
H
H
H
H
H
H
H
H
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
F
238Æ24 (3)
34Æ1 (3)
0%
24%
n.d.
n.d.
H
H
CO2CH3
Cl
CF3
26%
1%
33%
46%
21%
5%
0%
7%
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
9
10
11
12
13
14
15
16
17
18
19
20
21
21R
21S
22
23
23R
23S
24
25
26
26R
26S
27R
NHCOCF3
CONH2
NH2
N(n-Bu)2
F
COPh
H
7%
n.d.
2150Æ100 (3)
9200 (1)
15%
48%
n.d.
n.d.
H
H
H
Cl
Cl
Cl
9%
n.d.
1700Æ250 (3)
58Æ9 (9)
32Æ6 (9)
285Æ21 (3)
368Æ33 (3)
84Æ3 (3)
160Æ20 (3)
3600Æ400 (3)
1800Æ350 (3)
22%
18Æ1 (3)
88Æ3 (3)
61Æ1 (3)
92Æ3 (3)
1000 (1)
4Æ1 (16)
3Æ1 (11)
66Æ3 (5)
20Æ3 (8)
9Æ0.3 (3)
25Æ2 (3)
516Æ93 (3)
91Æ2 (3)
n.d.
1Æ0.05 (3)
7Æ1 (3)
8Æ1 (3)
17Æ2 (3)
H
CF3
CF3
CF3
F
H
Cl
Cl
Cl
CF3
H
H,H
H,H
H,H
H,H
H,H
Cl
Cl
Cl
Cl
aReceptor binding studies were performed using standard radioligand binding techniques with receptor membranes isolated from a CHO-k cell line
expressing the human serotonin receptor 5-HT2C (h-5HT2C). Determinations of anities were made using [125I]DOI to label the agonist sites and
[3H]mesulergine to label the antagonist sites. IC50 values were calculated from nine-point-concentration curves in which each concentration was
tested in triplicate, and the Ki was derived using the Cheng±Pruso€ equation:11 Ki=(IC50)/(1+([radioligand]/Kd)). The number of independent
experiments is indicated in parentheses.
bNot determined.

G. S. Welmaker et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1991±1994
1993
Treatment of an appropriately substituted ortho-nitro-
halobenzene with 4-carbobenzyloxypiperazine-2-car-
boxylic acid12 a€ords the (ortho-nitrophenyl)-piper-
azine. Reduction of the nitro group and subsequent
cyclization of the aniline to the lactam is accomplished
using iron in acetic acid. At this point, the lactam can be
alkylated, if desired. The Cbz-protecting group is then
removed by hydrogenolysis. In the cases of substrates
with functionality sensitive to hydrogenolysis, the Cbz-
protecting group is removed by treatment with either
KOH in re¯uxing CH3OH±H2O13 or by HBr in acetic
acid. The resulting quinoxalinone could then be reduced
Table 2. 5-HT2C agonist functional activities for selected compounds
Compounds
5-HT2c agonist functional activitya
EC50 (nM) Æ SE % Emax Æ SE
4(R)
21
21(R)
21(S)
23
23(R)
23(S)
26
26(R)
26(S)
27(R)
86 Æ 13 (4)
95 Æ 2 (4)
12 Æ 0.1 (2)
8 Æ 3 (2)
110 Æ 2.5 (2)
112 Æ 0.5 (2)
88 Æ 4 (2)
108 Æ 5 (3)
97 (1)
290 Æ 27 (2)
136 Æ 46 (3)
19 (1)
2600 (1)
85 (1)
7.6 Æ 0.7 (3)
49 Æ 15 (2)
45 Æ 22 (2)
19 Æ 8 (2)
96 Æ 3 (3)
107 Æ 7 (2)
82 Æ 5 (2)
76 Æ 9 (2)
.
under standard conditions using BH3 THF in THF to
provide the 2,3,4,4a,5,6-hexahydro-1H-pyrazino[1,2-a]-
quinoxaline.
aEC50 values were derived from six-point-concentration±response
curves in which each concentration was tested in duplicate. Emax
values indicate relative ecacy compared to a maximally e€ective 5-
HT concentration. The number of independent experiments is indi-
cated in parentheses.
Initially, the racemic compounds were separated into
their enantiomeric pairs by chiral HPLC. Alternatively,
the required enantiomer was prepared by the same route
utilizing the known (R)- or (S)-4-carbobenzyloxypiper-
azine-2-carboxylic acid.14
Because compound 21R exhibited potent functional
activity as a full agonist, it was tested in vivo using a rat
feeding model.16 Compound 21R, when administered to
fasted, male Sprague±Dawley rats, produced a dose-
dependent decrease in food intake with ED50 values of
2 mg/kg (ip) and 10 mg/kg (po).
All of the analogues were tested for their 5-HT2C antago-
nist binding anity and these results are presented in
Table 1. From the antagonist binding data available, sev-
eral generalizations can be surmised. Monosubstitution
at R2, R3, or R4 has a pronounced e€ect on binding
anity, varying the % inhibition observed at 1 mM from
0% (5, R2=CH3) to 67% (4, R2=Cl) for R2-substitu-
tion, 0% (13, R3=NH2) to 46% (10, R3=CF3) for R3-
substitution, and 15% (17, R4=Cl) to 48% (18, R4=F)
for R4-substitution. In contrast, substitution at R1 or R5
exhibits no e€ect, or perhaps a modest decrease, in
the antagonist anity (1 vs 2 and 1 vs 19). Monochloro-
substitution at R2 (similar to m-CPP) gives a higher
binding anity (4, 67%) than monochloro-substitution
at R1 (2, 8%), R3 (9, 33%), or R4 (17, 15%). Interest-
ingly, reduction of the quinoxalinone (X=O) to the
quinoxaline (X=H,H) exhibited almost no e€ect on
the antagonist binding anity (1 vs 25, 21 vs 26 and
23R vs 27R). The most active compounds are those
with disubstitution at R2 and R3 (21, 23, and 26). In
the case of the quinoxalinones (1, 4, 21 and 23), the
(R)-enantiomer was found to be the eutomer; how-
ever, the 5-HT2C receptor failed to discriminate between
the enantiomers of the quinoxaline 26. All of the
compounds tested exhibited a higher selectively for
agonist binding over antagonist binding, suggestive of
functional agonist activity.
In summary, a series of 2,3,4,4a-tetrahydro-1H-pyr-
azino[1,2-a]quinoxalin-5-(6H)ones and 2,3,4,4a,5,6-hexa-
hydro-1H-pyrazino[1,2-a]quinoxalines were synthesized
and shown to exhibit potent 5-HT2C agonist activity in
vitro and in vivo.
References and Notes
1. (a) For a recent review of central 5-HT receptors and their
function, see: Barnes, N. M.; Sharp, T. Neuropharmacology
1999, 38, 1083. (b) Dekeyne, A.; Girardon, S.; Millan, M. J.
Neuropharmacology 1999, 38, 415 and references therein. (c)
Martin, J. R.; Bos, M.; Jenck, F.; Moreau, J.-L.; Mutel, V.;
Sleight, A. J.; Wichmann, J.; Andrews, J. S.; Berendsen, H. H.
G.; Broekkamp, C. L. E.; Ruigt, G. S. F.; Kohler, C.; van
Delft, A. M. L. J. Pharmacol. Exp. Ther. 1998, 286, 913.
2. (a) Glennon, R. A.; Dukat, M. In Psychopharmacology:
The Fourth Generation of Progress; Bloom, F. E., Kupfer, D.
J., Eds.; Raven: New York, 1995; pp 415±429. (b) Molineaux,
S. M.; Jesseli, T. M.; Axel, R.; Julius, D. Proc. Natl. Acad. Sci.
U.S.A. 1989, 86, 6793. (c) Tecott, L. H.; Sun, L. M.; Akana,
S. F.; Strack, A. M.; Lowenstein, D. H.; Dallman, M. F.;
Julius, D. Nature 1995, 374, 542.
The results of the functional studies measuring [3H]ino-
sitol monophosphate ([3H]IP) formation from CHO
cells expressing human 5-HT2C receptors are depicted
in Table 2.15 In the quinoxalinone series, the (R)-enan-
tiomers of 21 and 23 demonstrated the most potent
activity and functioned as full agonists, while the (S)-
enantiomers were less potent and failed to exhibit a
maximal response when compared to serotonin. In
contrast, in the quinoxaline series (26), there was no
signi®cant di€erence in potency; however, the (R)-iso-
mer did exhibit full agonism whereas the maximal e€ect
produced by the (S)-isomer was less than that observed
with 5-HT.
3. Bray, G. A.; Greenway, F. L. Endocr. Rev. 1999, 20, 805.
4. Rosenbaum, M.; Leibel, R. L.; Hirsch, J. N. Engl. J. Med.
1997, 337, 396.
5. Heck, A. M.; Yanovski, J. A.; Calis, K. A. Pharmacother-
apy 2000, 20, 270.
6. All compounds gave satisfactory spectral data. For example,
see: (R)-8,9-Dichloro-2,3,4,4a-tetrahydro-1H-pyrazino[1,2-a]-
quinoxalin-5-(6H)one, hydrochloride (21R): mp >290 ꢀC; H
1
NMR (400 MHz, DMSO-d6) d 11.00 (s, 1H), 9.58 (s, 2H), 7.12
(s, 1H), 7.02 (s, 1H), 4.03 (dd, 1H, J=11.6, 3.6 Hz), 3.87 (d,
1H, J=10.7 Hz), 3.61 (dd, 1H, J=12.9, 2.0 Hz), 3.41 (d, 1H,
J=9.5 Hz), 3.42±2.99 (m, 3H); IR (KBr) 2950, 2700, 1700,
1590, 1500 cm 1; MS (APCI, m/e (%)) 272 (100, [M+H]+)
+
.
and 274 (65, [M+H] ). Anal. calcd for C11H11Cl2N3O HCl:

1994
G. S. Welmaker et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1991±1994
C, 42.81; H, 3.92; N, 13.62; found: C, 42.45; H, 3.78; N, 13.43;
chiral purity 99.9% (HPLC: Chiralcel AD); [a]2D5 +28ꢀ (c 1,
DMSO). (R)-9-Chloro-8-tri¯uoromethyl-2,3,4,4a-tetrahydro-1H-
pyrazino[1,2-a]quinoxalin-5(6H)-one, hydrochloride (23R): 1H
NMR (400 MHz, DMSO-d6) d 11.0 (s, 1H), 9.47 (br s, 2H),
7.22 (s, 1H), 7.18 (s, 1H), 4.19 (dd, 1H, J=11.6, 3.2 Hz), 4.01
(d, 1H, J=11.8 Hz), 3.62 (d, 1H, J=11.7 Hz), 3.41 (d, 1H,
J=10.1 Hz), 3.14±3.01 (m, 3H); IR (KBr) 3460, 3170, 3020,
2970, 2800, 1700, 1620, 1505, 1450, 1400, 1370, 1300, 1230,
1160, 1110 cm 1; MS (APCI, m/e 306 (100, [M+H]+), 308
(33, [M+H]+). Anal. calcd for C12H12ClF3N3O: C, 42.13; H,
3.54; N, 12.28; found: C, 41.83; H, 3.49; N, 12.01; chiral purity
98% (HPLC: Chirapak AD); [a]2D5 +43ꢀ (c 0.9, DMSO).
7. (a) Kennett, G. A.; Curzon, G. Br. J. Pharmacol. 1991, 103,
2016. (b) Kennett, G. A.; Wood, M. D.; Glen, A.; Grewal, S.;
Forbes, I.; Gadre, A.; Blackburn, T. P. Br. J. Pharmacol.
1994, 111, 797. (c) De Vry, J.; Schreiber, R. Neurosci. Bio-
behav. Rev. 2000, 24, 341.
14. (a) Wu, M. T.; Ikeler, T. J.; Ashton, W. T.; Chang, R. S.
L.; Lotti, V. J.; Greenlee, W. J. Bioorg. Med. Chem. Lett.
1993, 3, 2023. (b) Felder, E.; Ma€ei, S.; Pietra, S.; Pitre, D.
Helv. Chim. Acta. 1960, 43, 888.
15. Functional assay conditions: Agonist stimulated [3H]ino-
sitol monophosphate ([3H]IP) formation was examined in
CHO cells in which the human 5-HT2C receptor subtype was
stably expressed. Cells were incubated overnight in the pres-
ence of 2 mCi/well myo-[3H]inositol to label inositol lipids. On
the day of the assay, cells were washed and then preincubated
with medium containing 10 mM LiCl for 30 min before addi-
tion of receptor agonists over the concentration range 10 10 to
10 5 M for a further 30 min. [3H]IP formation in cell fractions
was measured by liquid scintillation counting and data were
calculated as a percentage of the response observed with a
maximally e€ective concentration of 5-HT (10 mM). Agonist
EC50 values were estimated from log-concentration response
curves using a three-parameter logistic function.
8. (a) Fitton, A.; Heel, R. C. Drugs 1990, 40, 722. (b) Dourish,
C. T. Adv. Biosci. 1992, 85, 179.
16. Feeding behavior conditions: Eight male Sprague±Dawley
rats weighing 150±180 g were separated into individual cages
and acclimated to a powdered diet for 2 weeks. A modi®ed
Latin Square design was used such that each animal received
all treatments. Animals were generally fasted on Monday and
Thursday and food intake was assessed on Tuesday and Fri-
day. Following each 24 h fast, an animal was treated with
either vehicle or a dose of the test compound. Powdered chow
was placed back in the animals' cages and 2 h food intake was
recorded. Data were subjected to one-way ANOVA with
posthoc t-tests to assess group di€erences, and where appro-
priate ED50 values were calculated. The ED50 value is the dose
that produces a 50% reduction in food intake.
9. Sargent, P. A.; Sharpley, A. L.; Williams, C.; Goodall, E.
M.; Cowen, P. J. Psychopharmacology 1997, 133, 309.
10. (a) Freed, M. E.; Potoski, J. R. US Patent 4,032,639, 1977;
Chem. Abstr. 1977, 87, 135404. (b) Freed, M. E.; Potoski, J. R.
US Patent 4,089,958, 1978; Chem. Abstr. 1979, 91, 39528.
11. Cheng, Y.; Pruso€, W. H. Biochem. Pharmacol. 1973, 22,
3099.
12. Henczi, M.; Weaver, D. F. Org. Prep. Proc. Int. 1994, 26,
578. It was found that the amino acid±copper complex could be
used directly in the coupling step in the presence of excess EDTA.
13. Angle, S. R.; Arnaiz, D. O. Tetrahedron Lett. 1989, 30, 515.