Potent and selective, sulfamide-based human β3-adrenergic receptor agonists

Article abstract of DOI:10.1016/j.bmcl.2004.03.089

A series of sulfamide-based analogs related to L-796568 were prepared and evaluated for their biological activity at the human β₃- adrenergic receptor (AR). This modification allows for a significant reduction in molecular weight, while maintai

Full text of DOI:10.1016/j.bmcl.2004.03.089

Bioorganic & Medicinal Chemistry Letters 14 (2004) 3235–3240
Potent and selective, sulfamide-based human b3-adrenergic
receptor agonists
Robert L. Dow,a,* Ernest S. Paight,a Steven R. Schneider,a John R. Hadcock,a
Diane M. Hargrove,a Kelly A. Martin,a Tristan S. Maurer,b Nancy A. Nardone,a
David A. Tessb and Paul DaSilva-Jardinea
aCardiovascular and Metabolic Diseases, Pfizer Global Research and Development, Groton, CT 06340, USA
bPharmacokinetics, Dynamics and Metabolism, Pfizer Global Research and Development, Groton, CT 06340, USA
Received 4 March 2004; revised 26 March 2004; accepted 29 March 2004
Abstract—Aseries of sulfamide-based analogs related to L-796568 were prepared and evaluated for their biological activity at the
human b3-adrenergic receptor (AR). This modification allows for a significant reduction in molecular weight, while maintaining
single-digit nanomolar potencies at the b3-AR and high selectivities versus the b1- or b2-AR.
Ó 2004 Elsevier Ltd. All rights reserved.
It has been fifty-five years since the receptor systems that
are activated by the endogenous catecholamines,
adrenaline, and noradrenaline, were classified into
a- and b-subtypes.1 Twenty years later the b-adreno-
ceptor (b-AR) classes were further subdivided into the
b1- and b2-subtypes.2 In this classification system the
lipolytic response (i.e., hydrolysis of fatty acid stores),
induced by b-AR agonists in adipocytes, was initially
defined as a b1-AR-mediated response.2 With the advent
of selective b-AR antagonists by the early 1980s it
became clear that the lipolytic response was not wholly
mediated via the b1-subtype, leading to the proposal of a
third member of the beta-adrenoceptor family (b3-AR).3
Drug discovery efforts directed toward optimization of
the pharmacological properties associated with the
b3-AR, such as thermogenesis (energy expenditure) in
brown adipose tissue and lipolysis in white adipose tis-
sue, lead to the identification of several potent chemo-
types (see Fig. 1).4 Based on the tissue expression of the
b3-AR there is therapeutic potential for these agents in
the treatment of obesity, type-II diabetes, and urinary
incontinence.
none of these compounds produced significant responses
in the disease endpoints of type-II diabetes or obesity.4
It was during this period that the human b3-AR was
cloned and expressed.5 Subsequent investigation
revealed that the rat b3-AR pharmacology was signifi-
cantly different from the human receptor, with the for-
mer being significantly more responsive to agonists.6
Since optimization of the early agonist structures 1 were
carried out utilizing the rat receptor, research efforts
shifted to identification of agents that were more effec-
tive against the human b3-AR.
One of the series to arise from efforts to make more
potent agonists of the human b3-AR was a series of
compounds in which the acid moiety of 1 is replaced by
a sulfonamide functionality.7 These efforts ultimately led
to the discovery of L-796568 (2), which was advanced to
human clinical trials. After a single-dose of L-796568 in
healthy, overweight to obese male subjects modest
increases in energy expenditure (8%) were observed at a
dose of 1000 mg.8 However, upon chronic treatment
(28 days) there were no statistically significant differ-
ences in energy expenditure or body composition.9
Pharmacokinetic evaluation of L-796568 revealed that it
has low oral bioavailability in preclinical species, with
poor absorption being the major factor.10 This poor
absorption is most likely due to the high molecular
weight (MW ¼ 625) and lipophilicity (cLogP ¼ 5.3) of
L-796568, attributes that have been associated with this
pharmacokinetic issue.11 Unfortunately, the large
Anumber of agents were advanced into human clinical
trials during the late 1980s and early 1990s, however
Keywords: b3-Adrenergic receptor; Agonist; Sulfamide.
* Corresponding author. Tel.: +1-860-4414423; fax: +1-860-7154610;
e-mail: robert_l_dow@groton.pfizer.com
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.03.089

3236
R. L. Dow et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3235–3240
OH
H
R = 3-Cl-Ph, PhOCH2, 3-pyridyl, 5-(2-aminopyridyl)
N
X
R
CO2H
X = CH2 or CH2O
Y = CH2, OCH2, bond
Y
1
OH
H
N
O
O
S
N
N
H
N
CF3
S
2 L-796568
Figure 1.
O
O
O
a,b
N
NO2
Cl
Cl
3
c, d
O
O
OH
H
N
N
e
O
O
NRR'
O
O
NRR'
S
S
N
N
H
H
Cl
Cl
4
5
Scheme 1. Reagents and conditions: (a) 4-nitrophenethylamine, N,O-bis(trimethylsilylamide), DMSO, 75 °C; (b) carbonyldiimidazole, CH2Cl2, 57%
overall; (c) stannous chloride, EtOH, reflux, 100%; (d) RR0NSO2Cl, pyridine, 1,2-dichloroethane )35 to 60 °C, 69% for 4d; (e) KOH, EtOH, 80 °C,
40% for 5d.
sulfonamide functionalities found in this structural class
are necessary for the b3-AR binding affinity and selec-
tivity. Based on these results a series of sulfamide-con-
taining analogs were investigated to determine where
this pharmacophore modification would potentially
enhance b3-AR binding affinity/selectivity, allowing
for reductions in molecular weight of the resulting
agonists.
For the corresponding analogs containing the 3-pyridyl
‘head group’ the ethanolamine core was constructed via
13
alkylation of tosylate 7
with 4-nitrophenethylamine
and protection of the secondary amine as the tert-but-
ylcarbamate (Scheme 2). Conversion of the nitro group
to the sulfamide functionality and deprotection of the
ethanolamine core afforded analogs 10. Preparation of
homologs of 10 in which the phenethylamine portion is
replaced by phenoxyethylamine (13) is outlined in
Scheme 3. Reaction of 7 with ethanolamine and pro-
tection of the amine afforded 11. Incorporation of the
phenoxy-based sulfamide substituent proceeded via
Mitsunobu coupling with 4-nitrophenol and reduction
of the nitro group. Reaction with sulfamoyl chlorides
and deprotection afforded targets 13.
Synthetic routes to these sulfamide-based derivatives are
detailed in Schemes 1–4. For the chlorophenyl analogs
(Scheme 1), reaction of commercially available (R)-3-
chlorophenylstyrene oxide with excess 4-nitropheneth-
ylamine and reaction with carbonyldiimidazole afforded
oxazolidinone 3. Reduction of the nitro group and
reaction with the appropriate sulfamoyl chloride,12 fol-
lowed by deprotection provided 5. The N-methyl
derivative 6 was prepared via Eq. 1.
Scheme 4 details the synthetic route to b,b-dimethyl-
phenethylamine-based derivatives 17. Condensation of
hindered amine 1414 with tosylate 7 could be achieved by
heating at 100 °C for 48 h in DMSO. Attempted pro-
tection of the amino functionality as the tert-butylcar-
bamate failed due to steric hindrance. This result
suggested that selective reaction with the aniline might
be possible, thus obviating the need for protection of the
secondary amine. Reduction of the nitro functionality of
15, followed by reaction with sulfamoyl chlorides pro-
vided moderate to good yields of the monosulfamides
OH
H
a) LiHMDS, MeI
N
4a
O
N
O
NMe2
S
Me
b) KOH, EtOH
c) 4M HCl in dioxane
Cl
6
ð1Þ

R. L. Dow et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3235–3240
3237
TBDMSO
TBDMSO
N
BOC
N
OTs
a,b
Cl
N
Cl
NO2
8
7
c, d
TBDMSO
N
BOC
N
OH
.
H
N
.
2HCl
e,f
O
O
NRR'
O
O
NRR'
S
S
N
N
N
H
H
9
10
Scheme 2. Reagents and conditions: (a) 4-nitrophenethylamine, DMSO, 100 °C; (b) (Boc)2O, THF, 45% overall; (c) 10% Pd/carbon, H2, ethyl
acetate, 77%; (d) RR0NSO2Cl, pyridine, 1,2-dichloroethane, )35 to 60 °C, 55% for 9d; (e) tetrabutylammonium fluoride, THF; (f) 4 M HCl in dioxane
87% overall for 10d.
TBDMSO
TBDMSO
N
BOC
N
OTs
a,b
OH
Cl
N
Cl
11
c, d
7
.
.
2HCl
H
N
NH2
TBDMSO
N
BOC
N
NRR'
OH
H
N
S
O
O
e, f,g
O
O
N
12
13
Scheme 3. Reagents and conditions: (a) ethanolamine, diisopropylethylamine, DMSO, 80 °C; (b) (Boc)2O, THF, 87% overall; (c) 4-hydroxynitro-
phenol, diisopropylazo-dicarboxylate, triphenylphosphine, THF; (d) 10% Pd/C, ammonium formate, MeOH, 50% overall; (e) RR0NSO2Cl, pyridine,
1,2-dichloroethane )35 to 60 °C; (f) tetrabutylammonium fluoride, THF; (g) 4 M HCl in dioxane, 70% overall for 13b.
TBDMSO
TBDMSO
H
N
H2N
a
OTs
+
Me Me
Me Me
Cl
N
NO2
NO2
Cl
N
15
7
14
b, c
OH
TBDMSO
N
.
H
N
H
N
.
2HCl
d, e
O
O
NRR'
O
O
NRR'
Me Me
Me Me
S
S
N
N
N
H
H
17
16
Scheme 4. Reagents and conditions: (a) diisopropylethylamine, DMSO, 100 °C, 48 h, 63%; (b) 10% Pd/C, ammonium formate, MeOH, 93%; (c)
RR0NSO2Cl, pyridine, 1,2-dichloroethane )35 to 60 °C, 40% for 16j; (d) tetrabutylammonium fluoride, THF; (e) HCl, dioxane/MeOH 90% overall
for 17j.

3238
R. L. Dow et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3235–3240
Table 1. b-Adrenergic activities for series 515
NRR0
b3-AR EC50 lM (IA)
b2-AR IA @ 30 lM
b1-AR IA @ 30 lM
2
0.003 (92)
0.299 (75)
0.464 (63)
0.537 (62)
0.617 (84)
9.9 (105)
1
6
7
5
5
1
1
5
5a
5b
5c
5d
6
Dimethylamino
Cyclohexylamino
N-Cyclohexyl-N-methylamino
Piperidinyl
10
6
4
Dimethylamino
2
16. Deprotection of the alcohol moiety then afforded the
desired aminoalcohols 17.
replacement of the two carbon linker between the sec-
ondary amine and phenyl group (attached to the sulf-
amide functionality in this case) with an ethoxy-based
linker provided up to an order of magnitude improve-
ment in potency. This modification was incorporated
into the current series, utilizing both a relatively small
sterically demanding sulfamide substituent (13a) and the
more sterically demanding 4-methylpiperidine derivative
13b. Incorporation of the oxygen atom in the linker
leads to a significant loss in potency relative to the ethyl-
linked homologs (Table 3).
Early b3-AR agonist programs identified the 3-chloro-
phenyl moiety to be a good pharmacophore replacement
for the catechol group of the endogenous hormones.7 As
a baseline, sulfamide targets 5 were initially prepared
and tested for b3-AR agonist activity and selectivity
versus b1- and b2-AR. These analogs would also serve
the role of beginning to define the structure–activity-
relationship (SAR) around the sulfamide functionality.
As mentioned above, the focus of sulfamide substituent
SAR was focused on analogs derived from low molec-
ular weight amines. As can be seen from the examples in
Table 1, both 2°- and 3°-sulfamide-based analogs con-
taining the 3-chlorophenylethanolamine core were 100–
200-fold less potent than 2. Alkylation of the aniline
nitrogen of the sulfamide functionality leads to a further
order of magnitude reduction in potency (6 vs 5a).
Pitha and co-workers showed that incorporation of a
quaternary carbon in the phenethyl linker next to the
secondary amine of b-AR agonists can produce dra-
matic increases in potency.14 This work was carried out
in a rat reticulocyte binding assay (presumably assessing
mostly b1-AR binding) and does not address the
potential for such a modification to impact b3-AR
binding. However, extension of this concept to the sul-
famide series had a positive impact on b3-AR agonist
activity (Table 4). The better analogs, such as the cis-3,5-
dimethylpiperdinyl-17e and anilino-based 17f were sin-
gle-digit nanomolar, full agonists of the human b3-AR.
This structural modification did not enhance b1- or b2-
AR agonism, with selectivities of >5000-fold for the
more potent analogs.
Next, we investigated incorporation of the 3-pyridyl-
ethanolamine core found in 2 (Table 2). While this
modification does not produce enhanced b3-AR func-
tional potency for the dimethylamino (10a) and cyclo-
hexylamino (10b) sulfamide analogs, other 3°-sulfamides
such as 10c and 10d exhibit 4–6-fold improvements in
potency. Based on the in vitro profile of the piperidinyl
analog 10d a series of functionalized piperidine deriva-
tives were investigated. Incorporation of lipophilic
substituents into either the 3-, 4-, or 5-positions (10e–g)
of the piperidine ring result in further improvements in
agonist potency. b3-AR agonist potencies were now in
the 20–30 nM range and selectivities versus b1- and b2-
AR were >1000-fold. Before exploring further
improvements in the in vitro profile of this series,
through modifications of the sulfamide functionality,
the SAR of the phenylethyl linker region of these com-
pounds was evaluated.
Analogs 17a–f did not significantly increase oxygen
consumption (surrogate measure of energy expenditure)
in rats when dosed orally. This is presumably the result
of their high intrinsic clearances. For example, the rat
Table 3. b-Adrenergic activities for series 1315
NRR0
b3-AR
EC50 lM
(IA)
b2-AR
EC50 lM
(IA)
b1-AR
EC50 lM
(IA)
13a Dimethylamino 0.943 (96)
13b 4-Methylpiperidinyl 1.36 (98)
4
7
3
5
Previous efforts (unpublished) in our laboratories had
shown for certain series of b3-AR agonists that
Table 2. b-Adrenergic activities for series 1015
NRR0
b3-AR EC50 lM (IA)
b2-AR IA @ 30 lM
b1-AR IA @ 30 lM
2
0.003 (92)
0.265 (84)
1.46 (71)
1
11
0
1
6
10a
10b
10c
10d
10e
10f
10g
Dimethylamino
Cyclohexylamino
0
N-Cyclohexyl-N-methylamino
Piperidinyl
0.180 (67)
0.099 (77)
0.036 (80)
0.023 (100)
0.024 (95)
16
6
15
7
4-Methylpiperidinyl
4-Phenylpiperidinyl
cis-3,5-Dimethylpiperidinyl
8
6
2
17
2
11

R. L. Dow et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3235–3240
Table 4. b-Adrenergic activities for series 1715
3239
NRR0
b3-AR EC50 lM (IA)
b2-AR IA @ 30 lM
b1-AR IA @ 30 lM
2
0.003 (92)
0.038 (98)
0.010 (91)
0.047 (106)
0.025 (110)
0.004 (89)
0.007 (107)
0.165 (103)
0.041 (79)
0.018 (105)
0.008 (85)
1
7
1
10
12
20
17
7
17a
17b
17c
17d
17e
17f
17g
17h
17i
17j
Dimethylamino
4-Methylpiperidinyl
4-Phenylpiperidinyl
Cyclohexylamino
6
9
19
11
8
cis-3,5-Dimethylpiperidinyl
Anilino
29
28
7
2-Methoxyethyl-amino
N-Methylpiperizinyl
N-Benzylpiperizinyl
cis-3,5-Dimethylmorpholinyl
24
4
13
8
46
10
and human microsomal intrinsic clearances for 17e are
466 and 130 mL/min/kg, respectively. Since the in vitro
clearances tended to track with the molecular weight/
lipophilicity of the sulfamide group, a series of analogs
were prepared in which polar/ionic functionality was
incorporated into the sulfamide. Afew representative
examples are provided in Table 4 (17g–j). Incorporation
of small alkoxy (17g) or tertiary amine (17h) containing
sulfamide functionalities lead to significantly reduced
clearances, though they also suffered significant reduc-
tions in b3-AR agonist activity. Adding additional
lipophilicity (17i) to the tertiary amine of 17h enhances
the functional response, but not surprisingly reacquires
microsomal lability. Substitution of the 4-methylene
substituent of the high potency 3,5-dimethylpiperidinyl
analog 17e with an oxygen atom providing morpholine
17j, maintains single-digit nanomolar potency and
excellent selectivity. This activity profile for 17j is further
confirmed through b-AR binding studies that demon-
strate high b3-AR binding affinity (7 nM), with relatively
weak binding to the b1-AR (9750 nM) and b2-AR
(1400 nM).
References and notes
1. Ahlquist, R. P. Am. J. Physiol. 1948, 153, 586.
2. Lands, A. M.; Arnold, A.; McAuliff, J. P.; Luduena, F. P.;
Brown, T. G. Nature 1967, 214, 597.
3. Tan, S.; Curtis-Prior, P. B. Int. J. Obes. 1982, 7, 409.
4. For reviews of the subject see: (a) Dow, R. L. Exp. Opin.
Invest. Drugs 1997, 6, 1811; (b) Weber, A. E. Ann. Rep.
Med. Chem. 1998, 33, 193.
5. Emorine, L. J.; Marullo, S.; Briend-Sutren, M.-M.; Patey,
G.; Devalier-Klutchko, C.; Strosberg, A. D. Science 1989,
245, 1118.
6. Liggett, S. Mol. Pharmacol. 1992, 42, 634.
7. Mathvink, R. J.; Tolman, J. S.; Chitty, D.; Candelore, M.
R.; Cascieri, M. A.; Colwell, L. F., Jr.; Deng, L.; Feeney,
W. P.; Forrest, M. J.; Hom, G. J.; MacIntyre, D. E.;
Miller, R. R.; Stearns, R. A.; Tota, L.; Wyvratt, M. J.;
Fisher, M. H.; Weber, A. E. J. Med. Chem. 2000, 43, 3832,
and references cited therein.
8. van Baak, M. A.; Hul, G. B. J.; Toubro, S.; Astrup, A.;
Gottesdiener, K. M.; DeSmet, M.; Saris, W. H. M. Clin.
Pharmacol. Ther. 2002, 71, 272.
9. Larsen, T. M.; Toubro, S.; van Baak, M. A.; Gottesdiener,
K. M.; Larson, P.; Saris, W. H. M.; Astrup, A. Am.
J. Clin. Nutr. 2002, 76, 780.
10. Stearns, R. A.; Miller, R. R.; Tang, W.; Kwei, G. Y.;
Tang, F. S.; Mathvink, R. J.; Naylor, E. M.; Chitty, D.;
Colandrea, V. J.; Weber, A. E.; Colletti, A. E.; Strauss, J.
R.; Keohane, C. A.; Feeney, W. P.; Iliff, S. A.; Chiu, S.-H.
L. Drug Metab. Disp. 2002, 30, 771.
More importantly, the rat and human microsomal
intrinsic clearances of this compound are reduced to 165
and 11 mL/min/kg, respectively. With a molecular
weight of 463 and a cLogP ¼ 2.6 there is an expectation
that this compound should also not have the oral
absorption issues associated with 2.
11. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney,
P. J. Adv. Drug Del. Rev. 1997, 23, 3.
12. Sulfamoyl chlorides were prepared by treating the appro-
priate sodium sulfamate (synthesized via reported proce-
dure: Wiley, R. A.; Pearson, D. A.; Schmidt, V.; Wesche,
S. B.; Roxon, J. J. J. Med. Chem. 1983, 26, 1077) with of
phosphorus oxychloride (2 equiv) at 80 °C in 1,2-dichloro-
ethane for 18 h, cooling, adding an equal volume of
hexanes, filtering, and concentrating the filtrate in vacuo.
13. Dow, R. L.; Schneider, S. R. European Patent Application
1138685, 2001; Chem. Abstr. 2001, 135, 288694.
14. Milecki, J.; Baker, S. P.; Standifer, K. M.; Ishizu, T.;
Chida, Y.; Kusiak, J. W.; Pitha, J. J. Med. Chem. 1987, 30,
1563.
Increases in oxygen consumption in the rat has been
utilized as a surrogate endpoint for determination of
increased energy expenditure produced by b3-AR
agonists.16 When dosed intraperitoneally at 30 mg/kg,
17j produces a 38 6% increase in oxygen consumption
relative to the control animals.17 This effect was dose
responsive with 10 mg/kg resulting in a 26 3% increase.
When dosed orally in rats at 30 mg/kg, 17j elicits a
35 3% increase in oxygen consumption. That this oral
response is similar to that seen from intraperitoneal
administration supports the hypothesis that 17j has
good oral absorption.
15. Agonist activities (EC50) were assessed by measuring
cAMP levels in CHO cells expressing cloned human
b-adrenergic receptors. Intrinsic activities (IA) represent
the percentage of the maximal response attained by
isoproterenol. NT ¼ Not tested.
In conclusion, this study has detailed the discovery of a
novel series of highly potent and selective b3-AR agon-
ists. These compounds have physiochemical properties
that are projected to lead to good oral absorption, which
is supported by preliminary in vivo efficacy studies.
16. Depocas, F.; Hart, J. S. J. Appl. Physiol. 1957, 10, 388.
17. Animals are removed from general housing first thing in
the morning (7:00–7:30 am) and are deprived of food and

3240
R. L. Dow et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3235–3240
water for the length of the oxygen consumption measure-
monotosylate salt) or vehicle (n ¼ 4 each). Oxygen
consumption measurements continue for 2 h. Oxygen
consumption values associated with periods of high
ambulatory activity (>100 counts/10 min) are excluded
from all calculations, as well as the first 5 values of
the run and the first value after dosing (settle-down
periods).
ments. Animals are weighed (310–350 g), marked, and
placed into individual activity monitored chambers
(1700 ꢀ 1700 ꢀ 500). The system is calibrated and the run
started (8:00–8:30 am). Oxygen consumption measure-
ments are made every 10 min for 3 h, at which time the
animals are dosed with test compounds (17j dosed as the