Synthesis and evaluation of potent and selective β3 adrenergic receptor agonists containing heterobiaryl carboxylic acids

Article abstract of DOI:10.1016/j.bmcl.2007.05.069

The design, synthesis, and SAR of a novel series of heterobiaryl phenethanolamine β₃ adrenergic receptor agonists are described. The furan analogue 49 was shown to elicit a significant dose-dependent lowering of plasma glucose in a rodent model

Full text of DOI:10.1016/j.bmcl.2007.05.069

Bioorganic & Medicinal Chemistry Letters 17 (2007) 4670–4677
Synthesis and evaluation of potent and selective b3 adrenergic
receptor agonists containing heterobiaryl carboxylic acids
Barry G. Shearer,a,* Esther Y. Chao,a David E. Uehling,a David N. Deaton,a
Conrad Cowan,b Bryan W. Sherman,b Tula Milliken,c Walter Faison,c Kathleen Brown,c
Kimberly K. Adkisond and Frank Leed
aDepartment of Medicinal Chemistry, GlaxoSmithKline, Research Triangle Park, NC 27709, USA
bDepartment of Receptor Biology, GlaxoSmithKline, Research Triangle Park, NC 27709, USA
cDepartment of Pharmacology, GlaxoSmithKline, Research Triangle Park, NC 27709, USA
dDepartment of Research Bioanalysis and Drug Metabolism, GlaxoSmithKline, Research Triangle Park, NC 27709, USA
Received 2 April 2007; revised 18 May 2007; accepted 22 May 2007
Available online 25 May 2007
Abstract—The design, synthesis, and SAR of a novel series of heterobiaryl phenethanolamine b3 adrenergic receptor agonists are
described. The furan analogue 49 was shown to elicit a significant dose-dependent lowering of plasma glucose in a rodent model
of type 2 diabetes.
Ó 2007 Elsevier Ltd. All rights reserved.
Activation of b3 adrenergic receptors (AR) located on
the surface of adipocytes induces lipolysis and stimulates
the upregulation and activation of the mitochondrial
uncoupling protein UCP1. This latter event mediates a
proton conductance pathway that uncouples oxidative
phosphorylation ultimately leading to thermogenesis
and a net increase in energy expenditure.1 The recogni-
tion of the role b3 ARs play in the regulation of meta-
bolic rate has led to significant efforts to identify
potent and selective agonists of the b3 AR for the treat-
ment of diverse disease states including obesity and type
2 diabetes.2 Several classes of b3 AR agonists have been
identified and optimized based upon their ability to acti-
vate rodent b3 ARs. Compounds such as 1 (CL
316243),3 2 (BRL 37344),4 and 3 (CGP 12177A)5 were
shown to elicit anti-obesity effects and exhibit potent
anti-diabetic properties in relevant in vivo rodent
models. These compounds were progressed into human
clinical trials and while some indications of efficacy were
observed, they ultimately failed due to a lack of potency
or deleterious side effects such as tachycardia (b1 AR
effect) and muscle tremor (b2 AR effect) arising from a
lack of b adrenergic receptor selectivity.6
The eventual cloning and expression of the human, rat,
and mouse b ARs revealed significant species variances
in receptor homology.7 Comparison of the b3 AR activ-
ity of numerous agonists on the cloned human and
rodent receptors showed significant interspecies differ-
ences.8 This may, in part, explain the poor efficacy
and side effect profiles observed with the early clinical
candidates and emphasizes the importance of using
cloned human receptor assays for the identification of
selective and efficacious b3 AR agonists. Recently, sev-
eral selective b3 agonists that are potent against the
cloned human b3 AR have been reported. Compounds
such as 4 (BMS-194449),9 5 (L-755507),10 and 6 (L-
770644)11 represent second generation b3 AR agonists
that have progressed into preclinical development
(Fig. 1).
A series of phenethanolamines containing an anilino
phenylacetic acid subunit, as exemplified by 7a, has been
disclosed.12 Compound 7a is a potent human b3 AR full
agonist (pEC50 = 7.8) that also exhibits activity against
the human b1 and b2 ARs (pEC50 = 7.1 and
pEC50 = 7.3, respectively). The homologated phenyl-
propionic analogue 7b was also discovered to be a
potent human b3 AR agonist with slightly improved
selectivity compared to 7a. Recently, we disclosed a
series of biarylaniline phenethanolamines as potent
and selective b3 agonists designed from the aniline
Keywords: b3 Adrenergic receptor; Agonist; Selective; Diabetes.
*
Corresponding author. Tel.: +1 919 483 9860; fax: +1 919 315
0430; e-mail: barry.g.shearer@gsk.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.05.069

B. G. Shearer et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4670–4677
4671
OH
OH
H
H
N
N
O
O
CO2Na
CO2Na
O
CO2H
Cl
Cl
1
2
OCHF2
OH
H
N
CH3SO2NH
HO
O
N
H
H
N
OH
O
N
H
OCHF2
4
3
OH
H
OH
H
O
N
N
O
O
S
HO
NH
N
H
N
O
N
O
S
O
O
N
N
N
N
H
N
H
5
6
Figure 1. Representative b3 agonists.
phenethanolamines 7a and 7b.13 In a continuing effort to
develop this series as novel potent and selective human
b3 AR agonists, we targeted a series of heterobiaryl ana-
logues based upon 7a and 7b.
mediates 10a and 10b are described in Scheme 2. The
synthesis of the 2-anilino-4-oxazolecarboxylic esters 11
and 12 entailed N-Cbz protection of 3-amino- and 4-
aminobenzamide, respectively, followed by treatment
with ethyl bromopyruvate in refluxing EtOH15 and
N-Cbz deprotection via hydrogenation. The analogous
2-anilino-4-thiazolecarboxylic esters 13 and 14 were pre-
pared by converting the N-Cbz protected benzamides 51
and 52 to their corresponding thiobenzamides 53 and 54
with Lawesson’s reagent in refluxing benzene and subse-
quent cyclization with ethyl bromopyruvate.15 Removal
of the N-Cbz protecting group under standard palla-
dium catalyzed hydrogenation proved problematic but
was readily achieved using 30% HBr in acetic acid to
provide the desired anilino thiazoles 13 and 14.
Our general target design strategy is illustrated in Figure
2. The presence of a carboxylic acid or other negatively
charged group has been demonstrated to play an impor-
tant role in providing b3 AR selectivity within certain
phenethanolamine analogues.14 Our target design in-
volved replacing the linking group of the phenylacetic
or phenylpropionic acid moieties with various heteroa-
ryl rings to incorporate conformational restriction in
the region of the carboxylic acid residue and investigate
the influence of the spatial orientation of the carboxylic
acid on b adrenergic receptor activity.
Cyclization of 3-nitro and 4-nitrophenacyl bromide with
ethyl thiooxamate in refluxing EtOH15 efficiently pro-
vides the nitroaryl-2-thiazolecarboxylic esters 55 and 56.
Tin (II) chloride mediated reduction of the nitro groups
gave the desired 4-anilino-2-thiazolecarboxylic esters 15
and 16.16
The synthesis of our b3 AR agonist targets is depicted
in Scheme 1. Treatment of the previously reported ami-
no esters 9a and 9b with DIBAL provided aldehydes
10a and 10b, respectively,12 which readily underwent
reductive amination with a variety of synthetically
prepared heterobiaryl anilines to yield the coupled
products 23–36. Deprotection of both the N-Boc carba-
mate and silyl ether protecting groups was achieved in
a single step using 4 N HCl in dioxane and subsequent
saponification of the ester functionality afforded the de-
sired final targets 37–50. The final targets are shown in
Figure 3.
The synthesis of the 5-anilino-4-oxazolecarboxylic esters
17 and 18 was achieved by treating 3-nitro and 4-nitro-
benzoyl chloride with a THF solution of ethyl iso-
cyanoacetate in the presence of Et3N for 72 h17 followed
by reduction of the nitro groups by palladium catalyzed
hydrogenation.
The syntheses of the novel heterobiaryl anilines utilized
for the reductive amination with the key aldehyde inter-
The 2-anilino-3-furancarboxylic esters 19 and 20
were synthesized through the coupling of the

4672
B. G. Shearer et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4670–4677
n
CO2H
OH
OH
CO2H
H
H
N
Het
N
N
N
H
H
Cl
Cl
7a
n = 1
7b n = 2
8
Figure 2. Heterobiaryl target design.
CO2R
Het
TBSO
Boc
N
TBS
O
Boc
N
H2N
CO2Me
CHO
a
11-22 (Scheme 2)
R
R
b
(Ref. 12)
(25%-94%)
Cl
Cl
9a R = Me
9b R = H
10a R = Me
10b R = H
CO2R
CO2H
TBSO
OH
Boc
N
H
Het
Het
N
c,d
(2%-78%)
N
H
N
H
R
R
Cl
Cl
23-34 R = Me
35-36 R = H
37-48 R = Me
49-50 R = H
Scheme 1. Reagents and conditions: (a) DIBAL/toluene, Et2O, À78 °C; (b) 11–22 (Scheme 2), NaBH(OAc)3, AcOH (cat), CH2Cl2; (c) 4 N HCl/
dioxane, rt; (d) LiOH, MeOH/H2O.
nitrophenyldiazonium salt prepared from the appropri-
ately substituted nitroaniline with 3-furancarboxylic
acid18 followed by esterification and reduction of the ni-
tro groups by hydrogenation. The structurally related 2-
anilino-3-thiophenecarboxylic esters 21 and 22 were syn-
thesized in a similar manner.
with increased intrinsic efficacy and 8- to 10-fold selec-
tivity over both the b1 and b2 AR subtypes. Appreciable
partial agonism was observed at the b1 and b2 ARs.
The 2-arylthiazole-4-carboxylic acid analogue 39 exhib-
ited a slight increase in b3 AR activity compared to the
corresponding oxazole derivative 37. A small decrease in
b1 AR potency was observed resulting in a minor
improvement in b1 selectivity. In contrast, transposition
of the thiazole ring to the meta position of the aniline
ring produced the full agonist 40 with subnanomolar
potency at the b3 AR (pEC50 = 9.1, EMax = 97%) and
significant improvement in selectivity. The measured
selectivity of compound 40 for the b3 AR versus the b1
and b2 AR subtypes was 100- and 200-fold, respectively.
The ability of these compounds to stimulate cAMP
accumulation in vitro was measured in Chinese hamster
ovary (CHO) cells expressing the cloned human b3, b2,
or b1 ARs.19 For each analogue, potency (pEC50) and
efficacy (EMax, the fitted maximal response to compound
expressed as a percent of the maximal response to the
nonselective full b AR agonist isoproterenol) were deter-
mined. These results are summarized in Table 1.
Lead compound 7b was characterized in the above as-
says as a reference and profiled as a potent full human
b3 AR agonist (pEC50 = 7.6, EMax = 104%). However,
this compound also produces a response, albeit submax-
imal, at both the b1 AR (pEC50 = 6.6, EMax = 10%) and
b2 AR (pEC50 = 6.4, EMax = 22%). The 2-(4-aminophe-
nyl)oxazole-4-carboxylic acid analogue 37 exhibits
enhanced potency at all three b ARs as compared to
7b. Although potency at the b3 AR improved insignifi-
cantly, overall selectivity decreased due to 4- and 20-fold
elevations in b2 and b1 AR potencies, respectively. The
isomeric analogue 38 in which the oxazole ring is meta
to the aniline nitrogen gave comparable b3 AR potency
Encouraged by the improvement in selectivity observed
for 2-arylthiazole-4-carboxylic acid analogue 40, we pre-
pared and evaluated isomeric 4-arylthiazole-2 carboxylic
acid analogues. Placement of the thiazole ring para to
the aniline nitrogen as in analogue 41 provided a com-
pound with slightly increased potency at the b3 and b1
ARs compared to the isomeric thiazole derivative 39
resulting in an overall enhanced selectivity profile. The
meta positional isomer 42 displayed an improved selec-
tivity profile relative to compound 41, but was much less
selective than thiazole analogue 40. In addition, both of
the isomeric 4-arylthiazole-2-carboxylic acid analogues
possess increased agonism of the b2 AR.

B. G. Shearer et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4670–4677
4673
X
CO2H
N
OH
OH
H
N
H
N
X
N
H
N
H
N
CO2H
Cl
38 X=O
40 X=S
37 X=O
39 X=S
Cl
S
N
CO2H
OH
OH
H
N
H
N
N
H
N
H
S
N
CO2H
Cl
41
Cl
42
O
N
OH
OH
CO2H
N
H
N
H
N
CO2H
N
H
N
H
O
Cl
44
Cl
43
X
OH
OH
CO2H
H
N
H
N
CO2H
N
H
N
H
X
Cl
46 X= O
48 X= S
Cl
45 X=O
47 X=S
OH
OH
CO2H
CO2H
H
N
H
N
N
H
N
H
O
S
49
50
Cl
Cl
Figure 3. Synthesized heterobiaryl carboxylic acid analogues.
All of the aforementioned analogues position the car-
boxylic acid in an orientation that extends away from
the core structure. In an effort to examine the effect of
positioning the carboxylic acid in a spatial orientation
in closer proximity to the aniline ring, we synthesized
and evaluated a series of heterobiaryl analogues wherein
the aniline ring and the carboxylic acid group were
incorporated adjacent to each other on the heterocyclic
ring system.
lated meta substituted compound 44 is also a potent and
selective full b3 AR agonist. Interestingly, the selectivity
profile for compound 44 is reversed compared to 43 with
greater b3 selectivity being observed versus the b1 AR
than the b2 AR. Both isomers 43 and 44 display signif-
icant partial agonist activity at the b2 AR.
Replacement of the oxazole nitrogen present in com-
pounds 43 and 44 with carbon provides the furan deriv-
atives 45 and 46. Both of these targets are potent and
selective b3 AR agonists with subnanomolar activity.
Both furan analogues also exhibit greater b3 selectivity
over the b2 AR than the b1 AR with the meta substituted
analogue 46 displaying the highest levels of selectivity.
Each furan, however, still exhibits significant partial
agonist activity at the b2 AR.
The 5-(4-aminophenyl)oxazole-4-carboxylic acid ana-
logue 43 is a potent and selective full b3 AR agonist
(pEC50 = 8.9, EMax = 90%). Oxazole 43 is highly selec-
tive versus the b2 AR (320-fold) and modestly selective
versus the b1 AR (20-fold). Notably, this agonist is an
order of magnitude more potent against the b3 AR with
diminished potency at the b1 and b2 ARs than the oxa-
zole isomer 37 with the extended carboxylic acid. Fur-
thermore, the selectivity of 43 is markedly higher than
37 showing a 20- and 50-fold increase in b3 selectivity
over the b1 and b2 ARs, respectively. The structurally re-
Thiophene compounds 47 and 48 are direct analogues of
agonists 45 and 46 whereby the furan oxygen has been
replaced with sulfur. Both thiophene compounds are
highly potent agonists with excellent selectivity profiles.

4674
B. G. Shearer et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4670–4677
b,c
a
R
R
R
NH2
NH2
N
CO2Et
O
O
O
R = 3-NH2
R = 4-NH2
51 R = 3-NHCbz (31%)
11 R = 3-NH2 (30%)
12 R = 4-NH2 (50%)
52 R = 4-NHCbz
(97%)
d
b,e
R
R
NH2
N
CO2Et
S
S
53 R = 3-NHCbz (62%)
54 R = 4-NHCbz (99%)
13 R = 3-NH2 (30%)
14 R = 4-NH2 (44%)
f
g
R
R
R
N
N
Br
CO2Et
S
CO2Et
O
S
R = 3-NO2
R = 4-NO2
55 R = 3-NO2 (78%)
56 R = 4-NO2 (88%)
15 R = 3-NH2 (97%)
16 R = 3-NH2 (96%)
h
c
CO2Et
R
CO2Et
R
R
Cl
N
O
N
O
O
R = 3-NO2
R = 4-NO2
57 R = 3-NO2 (73%)
58 R = 3-NO2 (76%)
17 R = 3-NH2 (89%)
18 R = 3-NH2 (97%)
i,j
k,l
CO2H
R
CO2Me
R
R
NH2
X
X
R = 3-NO2
R = 4-NO2
R = 3-NO2 (28%)
R = 4-NO2 (24%)
R = 3-NO2 (7%)
R = 4-NO2 (38%)
19 X = O
20 X = O
21 X = S
22 X = S
R = 3-NH2 (28%)
R = 4-NH2 (61%)
R = 3-NH2 (70%)
R = 4-NH2 (17%)
59 X = O
60 X = O
61 X = S
62 X = S
Scheme 2. Reagents and conditions: (a) PhCH2OCOCl, Na2CO3, THF, H2O, 0 °C to rt; (b) BrCH2COCO2Et, EtOH, reflux; (c) H2, 10% Pd/C,
EtOH; (d) Lawesson’s reagent, benzene, reflux; (e) 30% HBr/HOAc, CH2Cl2, rt; (f) H2NCSCO2Et, EtOH, reflux; (g) SnCl2, EtOH, reflux; (h)
NCCH2CO2Et, Et3N, THF, rt; (i) NaNO2, concd HCl, H2O, 0 °C; (j) 3-furancarboxylic acid or 3-thiophenecarboxylic acid, CuCl2, acetone, H2O; (k)
H2SO4, MeOH, reflux; (l) H2, 10% Pd/C, MeOH.
The para substituted analogue 47 exhibits subnanomo-
lar activity at the b3 AR receptor (pEC50 = 9.4,
EMax = 96%) with high selectivity over both the b1 and
b2 AR subtypes (250- and 1600-fold, respectively). The
meta substituted thiophene analogue 48 is an extremely
potent b3 AR agonist (pEC50 = 10.2, EMax = 85%) with
remarkable selectivity over both the b1 and b2 ARs
(>20,000- and 13,000-fold, respectively). Compound 48
represents the most potent and selective b3 AR agonist
in this series.
the des-methyl furan and thiophene analogues 49 and
50. The removal of the methyl substituent adjacent to
the phenethylamine serves to eliminate one of the stere-
ogenic centers, potentially simplifying the development
of this class of b3 agonists if acceptable potency, selectiv-
ity, and drug properties are retained. While it is docu-
mented that the inclusion of the methyl substituent
adjacent to the nitrogen with the appropriate stereo-
chemical integrity can influence both b3 AR potency
and selectivity, it is not essential for b AR activity.20
For example, Merck has disclosed a series of potent
and selective des-methyl arylethanolamines bearing a
variety of right hand side phenyl substituents.21
Encouraged by the potencies and selectivities observed
for compounds 45–48, we synthesized and evaluated

B. G. Shearer et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4670–4677
4675
Table 1. Stimulation of cAMP accumulation in CHO cells expressing
human b3, b2, and b1 ARsa
oxazoles 37 and 38 and the thiophenes 47 and 48 with
furans 45 and 46. Those analogues in which the linking
heterocyclic ring was substituted with the carboxylic
acid and aniline ring adjacent to each other were the
most potent and selective b3 AR agonists. This indicates
that the spatial orientation of the carboxylic acid group
is important for b3 AR activity and selectivity. It should
be noted that the majority of the analogues in this series
possess modest to significant activity at the human b2
AR (EMax = 19–76%). Finally, removal of the methyl
substituent from the amino chain resulted in an overall
decrease in b AR potency. Potent b3 AR activity, how-
ever, is retained while preserving high b3 selectivity.
Compound
pEC50a, (Emax, %)b
Selectivityc
b3
b2 b1
b2/b3 b1/b3
ISO
7a
7b
37
38
39
40
41
42
43
44
45
46
47
48
49
50
8.5 (112) 9.8 (116) 9.0 (109)
7.8 (117) 7.3 (90) 7.1 (24)
7.6 (104) 6.4 (22) 6.6 (10)
0.1
3
0.3
5
16
10
0.8
10
7.8 (92)
7.0 (53) 7.9 (23)
6
8
7.8 (115) 6.9 (39) 6.8 (30)
8.2d (100) 7.3d (41) 7.7d (21)
8
3
9.1 (97)
8.8 (100) 7.3 (63) 7.92 (19)
6.8d (48) 7.1 (21)
200
32
100
8
8.7 (97) 7.0 (76) 7.0 (20)
8.9 (90) 6.4 (55) 7.6 (24)
8.4 (102) 6.5 (49) 6.4 (14)
50
50
20
320
79
100
63
The pharmacokinetic properties of the more selective
compounds 40, 43–46, and 48–50 were evaluated in dogs
to determine if their progression into in vivo studies was
appropriate. Each compound was administered intrave-
nously at 0.2 mg/kg and plasma drug levels were deter-
mined by LC–MS/MS. Compounds 40, 43–46, and 48
had modest clearances (5.3–16.5 ml/min/kg) with extre-
mely short half-lives (61 h). The des-methyl thiophene
50, however, displayed low clearance (1.5 ml/min/kg)
and an improved half-life (1.6 h). More significantly,
the des-methyl furan 49 exhibited low clearance
(0.65 ml/min/kg) with an extended half-life of 6.1 h. Fur-
ther studies with furan 49 determined good systemic
exposure via oral administration (F = 60%) could be
achieved when dosed as a suspension. It is worth noting
that the compounds lacking the methyl group on the
amino chain exhibit lower clearances and longer half-
lives than their corresponding methylated analogues.
9.5 (85)
9.6 (87)
9.4 (96)
10.2 (85)
8.1 (91)
7.1 (36) 7.7 (18)
6.9 (38) 7.0 (14)
6.2 (19) 7.0 (14)
250
500
1600
400
250
20,000
630
3200
6.1 (9)
5.6e (<5) 5.3 (<5)
8.9 (103) 5.3 (<5) 5.4e (<2)
5.9e (<7) 13,000
320
4000
a Human b1, b2, and b3 receptors expressed in CHO cells. EC50
,
compound concentration which produces a cAMP response equal to
50% of its maximal response. Values had a standard deviation of
610% (n P 3 unless noted).
b Emax is the fitted maximal value of the concentration-response
expressed as a percent of the maximal response to R-(À)-isoprote-
renol (ISO).
c Defined as the ratio of the pEC50 of b1 or pEC50 of b2 to the pEC50 of
b3.
d Value calculated with n = 2 experiments.
e Response was too weak in two of the three trials to fit a curve. This
value is the average of the maximal observed response rather than the
fitted response.
Based on its potent and selective b3 AR activity and
promising pharmacokinetic profile, the des-methyl furan
compound 4922 was selected for in vivo evaluation in a
rodent model of type 2 diabetes. A dose–response study
was performed in 8-week-old male db/db mice.23 Com-
pound 49 was administered twice daily (BID) for
14 days by oral gavage and the effects on plasma glucose
were measured. The results are summarized in Figure 4.
Efficacious lowering of serum glucose occurred in a
dose-dependent manner with effects on glucose observed
at doses as low as 0.03 mg/kg. The maximal effective
dose of 0.3 mg/kg reduced plasma glucose levels from
428 118 mg/dL in vehicle treated mice to
191 53 mg/dL. Thus, chronic dosing of compound
49 produced effective lowering of plasma glucose in a
The des-methyl furan and thiophene analogues 49 and
50 proved to be potent and highly selective full b3 AR
agonists. The absence of the methyl substituent resulted
in an overall decrease in b AR potency, most notably on
the b1 and b2 ARs. Although the observed decrease in b3
AR potency for the des-methyl furan 49 (pEC50 = 8.1)
and the des-methyl thiophene 50 (pEC50 = 8.9) was 20-
to 30-fold, this loss in potency is readily acceptable given
the subnanomolar potencies of the parent methylated
compounds 46 and 48. The des-methyl furan and thio-
phene analogues 49 and 50 both failed to exhibit appre-
ciable activation of the b1 or b2 ARs and represent
promising potent and selective b3 AR agonists.
The in vitro screening data for these compounds offer
some insights into the SAR of this class of b3 AR ago-
nists. For a given heterobiaryl ring system, meta substi-
tution of the right hand side aniline ring generally led to
a decrease in activity at the b1 and b2 ARs while main-
taining or improving b3 AR potency. As a result, the
meta substituted heterobiaryl analogues generally pro-
vided the better selectivity profile compared to their
structurally related para substituted derivatives. Selec-
tivity was also influenced by the structural features of
the incorporated heterocyclic ring. Replacement of a
heterocyclic oxygen atom with sulfur results in an over-
all selectivity profile enhancement as evidenced by com-
paring the structurally related thiazoles 39 and 40 with
600
Dose
Glucose
(mg/dL)
500
400
300
200
100
0
(mg/kg)
Vehicle
0.03
0.1
428 118
376 153
255 36
191 53
183 42
0.3
1
Vehicle
0.03
0.1
0.3
1
Compound 49 Dose (mg/kg)
Figure 4. Glucose lowering effect of compound 49 in db/db Mice after
14 days of treatment. Data are shown as means SD (n = 10/group).

4676
B. G. Shearer et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4670–4677
5. Liggett, S. B. Mol. Pharmacol. 1992, 42, 634.
6. (a) Weyer, C.; Tataranni, P. A.; Snitker, S.; Danforth, E.,
Jr.; Ravussin, E. Diabetes 1998, 47, 1555; (b) Lipworth, B.
J. Br. J. Clin. Pharmacol. 1996, 42, 291.
7. (a) Emorine, L. J.; Marullo, S.; Briend-Sutren, M.-M.;
Patey, G.; Tate, K.; Delavier-Klutchko, C.; Strosberg, A.
D. Science 1989, 245, 1118; (b) Granneman, J. G.;
Lahners, K. N.; Chaudhry, A. Mol. Pharmacol. 1991,
40, 895; (c) Nahamias, C.; Blin, N.; Elalouf, J.-M.; Mattei,
M. G.; Strosberg, A. D.; Emorine, L. J. EMBO J. 1991,
10, 3721.
rodent model of diabetes consistent with the anti-dia-
betic properties of b3 agonists.24
In summary, we have identified a novel series of hetero-
biaryl carboxylic acid analogues that are potent and
selective full agonists of the human b3 adrenergic recep-
tor. These compounds were designed to explore the
effect of the spatial orientation of the carboxylic acid
on b AR activity by replacing the carbon tether between
the aniline ring and carboxylic acid of the previously
disclosed b3 AR agonists 7a and 7b with various five-
membered heterocyclic ring systems as rigidifying
linkers. The in vitro screening results revealed that both
the structural features of the heterocyclic ring and the
spatial orientation of the carboxylic acid influence
potency and selectivity. In addition, removal of the
methyl substituent from the amino chain afforded inter-
esting results. The des-methyl compounds exhibited an
overall decrease in b AR potency while maintaining high
b3 selectivity. Furthermore, the des-methyl compounds
showed improved pharmacokinetic properties. These
results suggest that removal of the methyl substituent
from the amino chain is permissible when working with
very potent and b3 selective agonists. The results from
these studies are consistent with data previously ob-
served with our biarylaniline phenethanolamine series.13
8. Strosberg, A. D.; Pietri-Rouxel, F. Trends Pharm. Sci.
1996, 17, 373.
9. Washburn, W. N.; Sher, P. M.; Poss, K. M.; Girotra, R.
N.; McCann, P. J.; Gavai, A. V.; Mikkilineni, A. B.;
Mathur, A.; Cheng, P.; Dejneka, T. C.; Sun, C. Q.; Wang,
T. C.; Harper, T. W.; Russell, A. D.; Slusarchyk, D. A.;
Skwish, S.; Allen, G. T.; Hillyer, D. E.; Frohlich, B. H.;
Abboa-Offei, B. E.; Cap, M.; Waldron, T. L.; George, R.
J.; Tesfamariam, B.; Ciosek, C. P., Jr.; Ryono, D.; Young,
D. A.; Dickinson, K. E.; Seymour, A. A.; Arbeeny, C. M.;
Gregg, R. E. Bioorg. Med. Chem. Lett. 2001, 11, 3035.
10. Fisher, M. H.; Amend, A. M.; Bach, T. J.; Barker, J. M.;
Brady, E. J.; Candelore, M. R.; Carroll, D.; Cascieri, M.
A.; Chiu, S.-H. L.; Deng, L.; Forrest, M. J.; Hegarty-
Friscino, B.; Guan, X.-M.; Hom, G. J.; Hutchins, J. E.;
Kelly, L. J.; Mathvink, R. J.; Metzger, J. M.; Miller, R.
R.; Ok, H. O.; Parmee, E. R.; Saperstein, R.; Strader, C.
D.; Stearns, R. A.; Thompson, G. M.; Tota, L.; Vicario, P.
P.; Weber, A. E.; Woods, J. W.; Wyvratt, M. J.; Zafian, P.
T.; MacIntyre, D. E. J. Clin. Invest. 1998, 101, 2387.
11. Shih, T. L.; Candelore, M. R.; Cascieri, M. A.; Chiu, S.-H.
L.; Colwell, L. F.; Deng, L.; Feeney, W. P.; Forrest, M. J.;
Hom, G. J.; MacIntyre, D. E.; Miller, R. R.; Stearns, R.
A.; Strader, C. D.; Tota, L.; Wyvratt, M. J.; Fisher, M. H.;
Weber, A. E. Bioorg. Med. Chem. Lett. 1999, 9, 1251.
12. (a) Uehling, D. E.; Donaldson, K. H.; Deaton, D. N.;
Hyman, C. E.; Sugg, E. E.; Barrett, D. G.; Hughes, R. G.;
Reitter, B.; Adkison, K. K.; Lancaster, M. E.; Lee, F.;
Hart, R.; Paulik, M. A.; Sherman, B. W.; True, T.;
Cowan, C. J. Med. Chem. 2002, 45, 567; (b) Foxton, M.W.
WO9533724, 1995.
From this work, heterobiaryl analogue 49 was shown to
elicit a potent and efficacious reduction of plasma glu-
cose in a rodent model of diabetes. This result in combi-
nation with the excellent in vitro profiles against cloned
human b adrenergic receptors suggests that this novel
class of b3 selective agonists may have relevant utility
for the treatment of type 2 diabetes mellitus in man.
The further development of these compounds will be
reported in due course.
Acknowledgments
13. Uehling, D. E.; Shearer, B. G.; Donaldson, K. H.; Chao,
E. Y.; Deaton, D. N.; Adkison, K. K.; Brown, K. K.;
Cariello, N. F.; Faison, W. L.; Lancaster, M. E.; Lin, J.;
Hart, R.; Milliken, T. O.; Paulik, M. A.; Sherman, B. W.;
Sugg, E. E.; Cowan, C. J. Med. Chem. 2006, 49, 2758.
14. Sher, P. M.; Mathur, A.; Fisher, L. G.; Wu, G.; Skwish,
S.; Michel, I. M.; Seiler, S. M.; Dickinson, K. E. J. Bioorg.
Med. Chem. Lett. 1997, 7, 1583.
The authors acknowledge Gordon Hodgson, Barbara
Reitter, and Elizabeth Sugg for useful discussions
related to this work. The authors also thank Doug
Minick for high resolution mass spectral data.
15. Moody, C. J.; Swann, E.; Houlbrook, S.; Stephens, M. A.;
Stratford, I. J. J. Med. Chem. 1995, 38, 1039.
References and notes
16. Bellamy, F. D.; Ou, K. Tetrahedron Lett. 1984, 25, 839.
17. Suzuki, M.; Iwasaki, T.; Matsumoto, K.; Okumura, K.
Syn. Commun. 1972, 2, 237.
1. Himms-Hagen, J.; Danforth, E., Jr. Curr. Opin. Endocrin.
Diabetes 1996, 3, 59.
2. (a) Collins, P.; Williams, G. J. Clin. Pharmacol. 2001, 51,
13; (b) Weyer, C.; Gautier, J. F.; Danforth, E., Jr. Diabetes
Metab. 1999, 25, 11; (c) Kordik, C. P.; Reitz, A. B. J. Med.
Chem. 1998, 42, 181; (d) Weber, A. E. Ann. Rep. Med.
Chem. 1998, 33, 193.
18. Jurasek, A.; Berecinova, D.; Cakrt, M.; Dandarova, M.;
Friedl, Z.; Stetinova, J.; Kovac, J. Collect. Czech. Chem.
Commun. 1989, 54, 215.
19. Granneman, J. G.; Lahners, K. N.; Rao, D. D. Mol.
Pharmacol. 1992, 42, 964.
3. (a) Largis, E. E.; Burns, M. G.; Muenkel, H. A.; Dolan, J.
A.; Claus, T. H. Drug Dev. Res. 1994, 32, 69; (b) Bloom, J.
D.; Dutia, M. D.; Johnson, B. D.; Wissner, A.; Burns, M.
G.; Largis, E. E.; Dolan, J. A.; Claus, T. H. J. Med. Chem.
1992, 35, 3081.
4. (a) Hollenga, C.; Brouwer, C.; Zaagsma, J. Eur. J.
Pharmacol. 1991, 200, 325; (b) Wilson, C.; Wilson,
S.; Piercy, V.; Sennitt, M. V.; Arch, J. R. S. Eur. J.
Pharmacol. 1984, 100, 309.
20. Oriowo, M. A.; Chapman, H.; Kirkham, D. M.; Sennitt,
M. V.; Ruffolo, R. R., Jr.; Cawthorne, M. A. J. Pharma-
col. Exp. Ther. 1996, 277, 22.
21. Weber, A. E.; Mathvink, R. J.; Perkins, L.; Hutchins, J.
E.; Candelore, M. R.; Tota, L.; Strader, C. D.; Wyvratt,
M. J.; Fisher, M. H. Bioorg. Med. Chem. Lett. 1998, 8,
1101.
22. Synthesis of compound 49: Reductive amination
of aldehyde 10b12b (1.17 g, 2.73 mmol) and methyl

B. G. Shearer et al. / Bioorg. Med. Chem. Lett. 17 (2007) 4670–4677
4677
2-(3-aminophenyl)furan-3-carboxylate
19
(742 mg,
(1H, d, J = 7.5 Hz), 7.09 (1H, t, J = 7.9), 7.26–7.32
(4H, m), 7.39 (1H, s), 7.67 (1H, s). Anal. (C21H21N2O4Cl)
C, H, N.
3.42 mmol) with Na(OAc)3BH (1.20 g, 5.66 mmol) and
HOAc (2 drops) in CH2Cl2 (10 mL) afforded, after work
up and silica gel chromatography eluting with 1:8 EtOAc
in hexane followed by 1:5 EtOAc in hexane, 1.15 g (67%)
of amine 35 as a colorless oil. MS (ES) m/z 629 (MH+).
Intermediate 35 was treated with 4 N HCl/dioxane
(10 mL) and stirred at ambient temperature for 2 h during
which time a white precipitate formed. This mixture was
diluted with diethyl ether (40 mL), stirred for 10 min, and
then the precipitate was collected by suction filtration to
afford, after drying in vacuo, 613 mg (92%) of methyl 2-{3-
[(2-{[(2R)-2-(3-chlorophenyl)-2-hydroxyethyl]amino}ethyl)-
amin]phenyl}-3-furoate dihydrochloride. Saponification
of the isolated ester (613 mg, 1.35 mmol) with LiOHÆH2O
(304 mg, 7.25 mmol) in 3:1 MeOH/H2O (12 mL) followed
by silica gel chromatography eluting with 60:30:1 CHCl3/
MeOH/NH4OH afforded 317 mg (59%) of acid 49 as a
white solid. 400 MHz 1H NMR (DMSO-d6) d 2.60–2.90
(4H, m), 3.16 (2H, br s), 4.74 (1H, d, J = 9.5 Hz), 5.72
(1H, br s), 6.58 (1H, d, J = 7.2 Hz), 6.72 (1H, s), 7.04
23. Compound 49 was administered twice daily (BID) by oral
gavage to male db/db mice (10 mice/group; 60 days of age at
onset of dosing) for 14 days. The reconstitution vehicle for
the in vivo studies was 0.025 M methanesulfonic acid.
Dosing solutions were prepared fresh daily. Prior to the
start of dosing, 10 mice were anesthetized and exsangui-
nated by cardiac puncture for baseline measurements (Day
0 predose values) of postprandial glucose, glycosylated
hemoglobin, and insulin. Subsequently, on days 7 and 14 of
dosing, mice from each dose group were sacrificed and
blood samples for glucose measurements were obtained.
Body weights were measured throughout the study and
there were no significant effects of any of the compounds on
body weight gain in these studies. All serum biochemical
measurements were made using an ILAB600 automated
chemistry Analyzer (Instrumentation Laboratories).
24. Arbeeny, C. M.; Myers, D. S.; Hillyer, D. E.; Bergquist,
K. E. Am. J. Physiol. 1995, 268, E678.