Small molecule ago-allosteric modulators of the human glucagon-like peptide-1 (hGLP-1) receptor

Article abstract of DOI:10.1016/j.bmcl.2007.06.086

Following our previous publication describing the biological profiles, we herein describe the structure-activity relationships of a core set of quinoxalines as the hGLP-1 receptor agonists. The most potent and efficacious compounds are 6,7-dichloroquinoxalines bearing an alkyl sulfonyl group at the C-2 position and a secondary alkyl amino group at the C-3 position. These findings serve as a valuable starting point for the discovery of more drug-like small molecule agonists for the hGLP-1 receptor.

Full text of DOI:10.1016/j.bmcl.2007.06.086

Bioorganic & Medicinal Chemistry Letters 17 (2007) 5472–5478
Small molecule ago-allosteric modulators of the human
glucagon-like peptide-1 (hGLP-1) receptor
Min Teng,a,* Michael D. Johnson,a Christine Thomas,a Dan Kiel,b James N. Lakis,b
Tim Kercher,c Shelley Aytes,c Jarek Kostrowicki,d Dilip Bhumralkar,c Larry Truesdale,c
John May,b Ulla Sidelman,e Janos T. Kodra,f Anker Steen Jørgensen,g
Preben Houlberg Olesen,g Johannes Cornelis de Jong,g Peter Madsen,e Carsten Behrens,f
Ingrid Pettersson,f Lotte Bjerre Knudsen,h Jens J. Holsti and Jesper Lauf
aDepartment of Medicinal Chemistry, Pfizer Global Research and Development,
La Jolla. 10770 Science Center Dr., San Diego, CA 92121-1194, USA
bDepartment of Pharmacology, Pfizer Global Research and Development,
La Jolla. 10770 Science Center Dr., San Diego, CA 92121-1194, USA
cDepartment of Combinatorial Chemistry, Pfizer Global Research and Development,
La Jolla. 10770 Science Center Dr., San Diego, CA 92121-1194, USA
dDepartment of Computational Chemistry, Pfizer Global Research and Development,
La Jolla. 10770 Science Center Dr., San Diego, CA 92121-1194, USA
eProtein Engineering, Diabetes Research Unit, Novo Nordisk A/S, Novo Nordisk Park, 2760 Maaloev, Denmark
fDiabetes Protein Engineering, Diabetes Research Unit, Novo Nordisk A/S, Novo Nordisk Park, 2760 Maaloev, Denmark
gMedChem, Diabetes Research Unit, Novo Nordisk A/S, Novo Nordisk Park, 2760 Maaloev, Denmark
hDepartment of Biology, Diabetes Research Unit, Novo Nordisk A/S, Novo Nordisk Park, 2760 Maaloev, Denmark
iDepartment of Medical Physiology, The Panum Institute, University of Copenhagen, DK-2200 Copenhagen, Denmark
Received 24 May 2007; revised 23 June 2007; accepted 28 June 2007
Available online 4 July 2007
Abstract—Following our previous publication describing the biological profiles, we herein describe the structure–activity relation-
ships of a core set of quinoxalines as the hGLP-1 receptor agonists. The most potent and efficacious compounds are 6,7-dichloro-
quinoxalines bearing an alkyl sulfonyl group at the C-2 position and a secondary alkyl amino group at the C-3 position. These
findings serve as a valuable starting point for the discovery of more drug-like small molecule agonists for the hGLP-1 receptor.
Ó 2007 Elsevier Ltd. All rights reserved.
The b-cell insulin response to an oral glucose load is
greater than the insulin response from glucose adminis-
tered intravenously. This effect, known as the incretin
effect, is due to the secretion of gut hormones during
nutrient absorption, which potentiates glucose-depen-
dent insulin secretion. One of the hormones believed
to make the greatest contribution to normal glucose
homeostasis is glucagon-like peptide 1 (GLP-1).1 GLP-1
is a 30-amino acid peptide derived from the proglucagon
gene.2 It is secreted from the L-cells of the small intestine
in response to nutrient intake. GLP-1 exerts its effect by
binding to the G protein-coupled receptor (GPCR) on
pancreatic b-cells. The resultant increase in intracellular
cAMP initiates cell depolarization and raises the cyto-
solic Ca2+ concentration, ultimately resulting in an aug-
mentation of insulin secretion.3
GLP-1 has demonstrated a number of pharmacological
effects beneficial in the treatment of type 2 diabetes.4 For
example, GLP-1 has been shown to inhibit glucagon
secretion, gastric acid secretion, and gastric emptying,
resulting in reduced food intake and loss of body weight.
Furthermore, it has been demonstrated that GLP-1
Abbreviations: cAMP, cyclic adenosine 30,50-monophosphate; GPCR,
G protein-coupled receptor; hGLP-1, human glucagon-like peptide-1;
hGIP, human gastric inhibitory polypeptide; DTT, 1,4-dithiothreitol.
Keywords: GLP-1; Quinoxaline.
*
Corresponding author. Tel.: +1 858 526 4669; e-mail: min.teng@
pfizer.com
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.06.086

M. Teng et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5472–5478
5473
receptor ligands increase pancreatic b-cell mass by
inducing neogenesis, proliferation, and anti-apoptosis
of the b-cells.5 GLP-1 is also known to stimulate insulin
secretion in a glucose-dependent manner resulting in lit-
tle risk of hypoglycemia. Because type 2 diabetes is char-
acterized by a progressive decline in insulin secretion,
reduced b-cell mass and/or function, increased glucagon
secretion, and is often accompanied by obesity, GLP-1
has attracted enormous attention as a therapeutic target
for the potential treatment of this disease.6
able to stimulate cAMP in a dose-dependent manner
in membranes expressing the hGLP-1 receptor but not
other related GPCRs, such as the glucagon and GIP
receptors. Interestingly, these compounds were unable
to displace 125I-GLP-1 in a competitive binding assay,
but were able to potentiate binding of GLP-1. We have
proposed they be called ago-allosteric modulators. Com-
pound 16f caused the release of insulin from normal
mouse islets and from a perfused rat pancreas. The pre-
liminary studies of the quinoxaline series demonstrated
the feasibility of discovering small molecule agonists
for the hGLP-1 receptor that may lead to a new therapy
for diabetes. We herein describe the synthesis and struc-
As a large peptide, GLP-1 is rapidly degraded in plasma
by dipeptidyl peptidase-IV (DPP-IV).7 Efforts to iden-
tify GLP-1 analogs with longer plasma half-lives have
resulted in the clinical development of liraglutide for
once daily subcutaneous administration.8 Exenatide, an-
other GLP-1 analog originally isolated from the saliva
of the Gila monster, was approved by the FDA in
2005 as a twice daily subcutaneous administration for
the treatment of type 2 diabetes.9 However, as a peptide,
GLP-1 and analogs thereof have to be administered
through intravenous or subcutaneous route. Therefore,
orally active, small molecule agonists of the hGLP-1
receptor are highly desirable.
ture–activity relationships of
sulfonylquinoxalines.
a
core set of 2-
The synthesis of all quinoxalines is outlined in Schemes
1–8. Condensation of disubstituted (X = Cl, OCH3 or
CH3) 1,2-phenylenediamines with an a-ketoester in
DMF/AcOH yielded the corresponding 2-hydroxyquin-
oxalines 1 (a–f, Scheme 1), which were then chlorinated
with POCl3 under refluxing conditions to afford 2-chlo-
roquinoxalines 2(a–f). Displacement of the chloride in
2a by 2-mercapto-5-methyl-1,3,4-thiadiazole in the pres-
ence of 40% KF on alumina in DMF led to 3. Con-
trolled oxidation of 3 by mCPBA afforded 4 and 5.
Compounds 6(a–f) were synthesized by treating 2-chlo-
roquinoxalines 2(a–f) with sodium methanesulfinic acid
in DMF at 100 °C for 12 h. Synthesis of 8(a–d) and 9
is shown in Scheme 2. The synthesis of 3-isopropyl-
6,7-dichloroquinoxaline-2-thiol was carried out by treat-
ing 2,6,7-trichloro-3-isopropylquinoxaline (2a) with
NaSH2H2O in DMF. Subsequent treatment of the thiol
with a variety of alkyl halides in DMF yielded the
Chen et al. recently described nonpeptidic-specific
hGLP-1 receptor agonists.10 At the same time, we re-
ported the discovery of small molecule agonists of the
hGLP-1 receptor.11 Screening of 250,000 wells from
our small molecule libraries against the hGLP-1 recep-
tor, followed by structural modifications, led to the dis-
covery of 6,7-dichloro-2-sulfonylquinoxalines, which
acted as the hGLP-1 receptor agonists in several bio-
chemical and cellular assays. These quinoxalines were
(O)n
Cl
Cl
N
N
S
Cl
Cl
N
N
S
S
S
iv
N
N
N
N
iii
3
4 n=1
5 n=2
X
X
N
N
Cl
Y
X
X
NH2
NH2
X
X
N
N
OH
Y
ii
i
v
X
X
N
SO2CH3
Y
1a: Y = isopropyl X = Cl
1b: Y = CF3 X = Cl
1c: Y = -propyl X = Cl
1d: Y = CO2C2H5 X = Cl
2 (a~f)
N
6 (a~f)
n
1e: Y =
1f: Y =
n
n
-propyl X = OCH3
-propyl X = CH3
Scheme 1. Preparation of quinoxaline derivatives. Reagents and conditions: (i) a-ketoester, DMF, HOAc; (ii) POCl3, reflux; (iii) 2-mercapto-5-
methyl-1,3,4-thiadiazole, KF 40% wt on alumina, DMF, rt, overnight; (iv) mCPBA/DCM, DCM; (v) NaSO2CH3, DMF, 100 °C.
O
O
O
Cl
Cl
N
S
Cl
Cl
N
N
Cl
Cl
Cl
N
S
Cl
Cl
N
N
S
R
R
iii
i
ii
+
N
N
7 (a~e)
9
8 (a~d)
2a
Scheme 2. Preparation of quinoxalines 8(a–d), 9. Reagents and conditions: (i) NaSHÆ2H2O, DMF, rt, 2 h; (ii) Na2CO3, alkyl halide, DMF; (iii)
mCPBA, DCM.

5474
M. Teng et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5472–5478
Cl
Cl
NH2
NH2
COCH(CH3)2
Cl
Cl
N
N
CON(C2H5)2
Cl
Cl
N
N
b
a
11
10
c
N
CO2C2H5
CF3
Cl
Cl
N
12
Scheme 3. Reagents and conditions: (a) i—4-Methyl-2,3-dioxo-pentanoic acid ethyl ester, HOAc, DMF; ii—KOH, EtOH; iii—EDCI, DMAP,
diethylamine, DCM, 48% (three steps); (b) 2,6-Dimethyl-heptane-3,4,5-trione, HOAc, DMF, 27%; (c) 2-Chloro-4,4,4-trifluoro-3-oxo-butyric acid
ethyl ester, EtOH, reflux.
Cl
Cl
N
N
PO(OC2H5)2
CF3
P(OC2H5)3
toluene reflux
13
82%
Cl
Cl
N
N
Cl
Cl
Cl
N
N
NH2
NH3, DMF, 45%
CF3
CF3
CN
14
15
2b
NaCN, DMF
50 oC, 31%
Cl
N
Cl
N
CF3
Scheme 4.
which was subsequently treated with 4,5-dichloropheny-
lenediamine in HOAc/DMF to give 2-ethoxycarbonyl-
quinoxaline. Saponification followed by coupling with
diethylamine led to 10 (Scheme 3). Similarly, 2,6-di-
methyl-3,5-heptadione was oxidized by selenium dioxide
to yield 2,6-dimethyl-heptane-3,4,5-trione. Without iso-
lation, this trione was reacted with 4,5-dichloropheny-
lenediamine to afford 11. Compound 12 was obtained
by reacting 4,5-dichlorophenylenediamine with ethyl
2-chloro-3-keto-4,4,4-trifluorobutyrate in refluxing ethyl
alcohol. Treatment of 2,6,7-trichloro-3-trifluoroquinox-
aline (2b) with triethylphosphite in refluxing toluene
for 12 h afforded 13 (Scheme 4). Compounds 14 and
15 were synthesized by treating 2b with ammonia and
NaCN, respectively. Compounds 16(a–s) were synthe-
sized by treating the 2,3-dichloroquinoxalines with
either 2-propanethiol or a variety of alkylamines in the
presence of cesium carbonate in DMF, followed by
treatment with sodium methanesulfinic acid in DMF
R1
R1
R2
R3
N
N
SO2CH3
R2
R3
N
N
Cl
Cl
i, ii
X
R4
16 (a~s)
R4
Scheme 5. Preparation of quinoxalines 16(a–s). Reagents and condi-
tions: (i) nucleophile, Cs2CO3, DMF, 0 °C; (ii) NaSO2CH3, DMF,
100 °C.
corresponding thioethers 7(a–e). Without further purifi-
cation, the thioethers were oxidized with mCPBA in
DCM to give sulfones 8(a–d) and sulfoxide 9. The syn-
thesis of quinoxalines 10–15 is described in Schemes 3
and 4. 4-Methyl-3-oxo-pentanoic acid ethyl ester was
oxidized by selenium dioxide in refluxing dioxane to
afford 4-methyl-2,3-dioxo-pentanoic acid ethyl ester,
Cl
Cl
Cl
NH2
NH2
Cl
Cl
N
N
Cl
S
Cl
Cl
N
iii
i, ii
CO2C2H5
+
N
O
Cl
18
17
O
O
Cl
Cl
N
N
S
iv
19
Scheme 6. Preparation of Compound 19. Reagents and conditions: (i) HOAc/DMF, rt, 3 days; (ii) POCl3, 0.5 h, reflux, 21% (two steps); (iii) sodium
hydrosulfide, K2CO3, DMF, rt, 4 days, quant; (iv) mCPBA, DCM, 31%.

M. Teng et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5472–5478
5475
Cl
Cl
O
S O
Cl
Cl
O
S O
N
N
Cl
Cl
N
ii
N
i
S
N
H
N
H
20
21
Scheme 7. Reagents: (i) mCPBA, DCM; (ii) isopropyl bromide, K2CO3, DMF, 22% (two steps).
O
Cl
Cl
Cl
Cl
Cl
Cl
CO2CH3
OH
Cl
Cl
CHO
NH2
i
CO2H
NH2
iii
ii
iv
NH
N
O
23
22
24
Cl
Cl
CO2CH3
Cl
Cl
Cl
CO2CH3
SO2CH3
vi
v
N
N
25
26
Scheme 8. Reagents and conditions: (i) NaOBr, NaOH aq 0 °C, 5 min, then 80 °C, 2 min, 72%; (ii) LAH, THF; (iii) MnO2, DCM, 46%; (iv) dimethyl
malonate, NaOMe, MeOH, then 1 N HCl, 87%; (v) POCl3, reflux, 73%; (vi) sodium methanesulfinate, DMF, 100 °C, 1 h, 36%.
Table 1. The effect of the C-2 sulfonyl group on the agonistic activity
(Scheme 5). The synthesis of 19 started with the conden-
at hGLP-1
1
N
8
5
sation of 5-chloro-2-oxo-pentanoic acid ethyl ester with
4,5-dichlorophenylenediamine to yield 2-hydroxyqui-
noxaline, which was then converted into 17 by refluxing
in POCl3. The subsequent sluggish cyclization to 18 was
achieved by using sodium hydrosulfide and K2CO3 in
DMF. Finally, oxidation of 18 by mCPBA produced
the desired sulfone 19 (Scheme 6). Compounds 2112
and 2613 were synthesized according to literature proce-
dures as outlined in Schemes 7 and 8.
7
2
Cl
Cl
X
6
N
3
4
Compound
X
EC50 Emax
(lM) (%)
O
O
S
S
5
6a
8a
8b
1.0
5
47
35
71
63
N
N
To systematically explore the structure–activity relation-
ships of 2-sulfonylquinoxalines as agonists for the hGLP-
1 receptor, our initial efforts included mapping out the
space requirement for the C-2 sulfonyl group. This was
approached through both singleton executions and
parallel synthesis, using intermediate thiols with a well-
established chemistry route shown in Scheme 2. Data
for representative compounds are shown in Table 1. Test
results from the membrane cAMP assay indicated that the
C-2 region tolerated a variety of substituted sulfones,
some of which were relatively bulky. All sulfones were
partial agonists with similar efficacy and potency at the
hGLP-1 receptor. Interestingly, compound 19 had no
measurable activity, which appeared due to the locked
conformation of the sulfone moiety to the quinoxaline
ring (Fig. 1).
SO2CH3
O
O
O
S
S
O
3
O
SO2CH3
O
1.2
O
O
CO2CH3
S
N
H
8c
8d
4.9
3.3
65
64
O
O
O
O
S
S
Data is from the membrane cAMP assay. Efficacy for the test com-
pounds is expressed as percentage activity normalized to that of GLP-1
activity. EC50 is the concentration at which a test compound produces
50% of cAMP response produced by GLP-1. Extrapolation of the
curve is applied when Emax is below 50%. Data is the mean of tripli-
cates with a SD of 10%.
Keeping the C-3 position as an isopropyl or CF3, we ex-
plored the C-2 position using non-sulfonic substituents
possessing either electron-donating or electron-with-
drawing features (Table 2). With the exception of sulfox-
ides 4 and 9, all the non-sulfonic compounds (10–15)
failed to produce measurable cAMP when tested at con-
centrations of up to 10 lM in the membrane assay. These
results indicated that compounds with a sulfone or sulfox-
ide group at the C-2 position have a unique way of bind-
ing. We speculated that unlike cyano, ester, amide, keto or
phosphonate, sulfone and sulfoxide moieties possess the
proper polarization effect on the quinoxaline ring to elicit
the observed GLP-1 agonistic activity.
To explore the C-3 position, we kept the C-2 position as
a methylsulfone (Table 3). Both EC50 and Emax data

5476
M. Teng et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5472–5478
O
O
O
O
O
Cl
Cl
N
N
S
N
Cl
Cl
N
N
S
Cl
Cl
N
N
S
O
H
6c
16f
19
EC50 = 1.2 uM
Emax = 68%
EC50= 0.155 uM
Emax= 85%
NA
Figure 1. Agonistic activities vary with the conformation of the C-2 sulfonyl group.
Table 2. Comparison of potency and efficacy of sulfonic and non-
sulfonic quinoxalines
Table 3. Substitution effects at the C-3 position
Cl
Cl
N
N
SO2CH3
Cl
Cl
N
N
X
Y
X
Compound
X
EC50 (lM)
Emax (%)
Compound
X
Y
EC50
(lM)
Emax
(%)
6b
6c
CF3
2.6
1.2
40
68
24
46
24
46
82
94
85
(CH2)2CH3
CO2C2H5
SCH(CH3)2
NH2
6a
1a
SO2CH3
OH
i-Pr
i-Pr
5
—
35
—
6d
>10
16a
16b
16c
16d
16e
16f
0.661
6.3
O
S
S
NHC2H5
NHCH(CH3)2
0.725
0.214
0.175
0.155
4
i-Pr
5.1
19
N
N
NHCH(CH3)C2H5
NHC(CH3)3
9
10
11
6b
12
13
14
15
SOCH3
i-Pr
i-Pr
i-Pr
CF3
CF3
CF3
CF3
CF3
30
—
—
2.6
—
—
—
—
29
—
—
40
—
—
—
—
CON(C2H5)2
COCH(CH3)2
SO2CH3
CO2C2H5
PO(OC2H5)2
NH2
NH
16g
0.433
69
16h
16i
16j
N(CH3)2
N(CH3)iPr
N(CH3)OCH3
2.2
1.8
2.6
47
75
52
CN
— indicates Emax (%) <2%.
N
16k
2.2
72
suggested that, unlike the C-2 position, the C-3 position
tolerated a variety of functional groups with secondary
amino groups being optimal. Quinoxalines 16(d–f) are
the most potent and efficacious compounds identified
in this study. Potency (EC50) declined with primary
and tertiary amines as seen in 16(b, h–k). Furthermore,
the C-3 region had a poor tolerance for polar functional
groups, as evidenced by the dramatic drop in potency
and efficacy in 16(l–o). We postulated that the enhanced
activity displayed by quinoxalines with a secondary ami-
no group at the C-3 position is the result of an intramo-
lecular H-bond between the NH of the amine and the
S@O of the sulfone (Fig. 1). It seems plausible that this
conformation is necessary for a favorable interaction,
perhaps through crucial H-bond with the hGLP-1 pro-
tein. By contrast, compound 19 has a sulfone moiety
locked within a ring in an unfavorable conformation.
Not surprisingly, this compound showed no activity.
16l
16m
NHNHCOCH3
NH(CH2)2NHCOCH3
14
7.1
32
22
H
N
O
16n
16o
14
38
43
N
O
NH
S
1.4
O
ity compared to 16d. A few asymmetric substituted
quinoxalines were examined. Compounds 16(r, s)
showed weaker activity compared to the corresponding
6,7-dichloroquinoxalines. These data suggested that
chloride substituents at the C-6,7 are required for the
best activity. Generally speaking, electron-donating sub-
stituents at the C-6,7 positions appeared detrimental to
the agonistic activity. A nitro group on the quinoxaline
ring, though strongly electron withdrawing, decreased
the agonistic activity as well.
Next, we examined the quinoxaline ring. We deliberately
chose substituents which would have differing electronic
effects on the benzo ring (Table 4). When the C-3 posi-
tion was an isopropyl group, 6c was the only compound
that showed measurable activity. In compounds with the
C-3 position fixed as a sec-butyl group, replacement of
C-7 chloride with a nitro group (16p) significantly de-
creased both potency and efficacy. Introducing a nitro
group at the C-5 position (16q) also decreased the activ-
Examination of other heterocyclic systems revealed that
agonistic activity at the hGLP-1 receptor appeared more
pronounced for quinoxalines. As shown in Figure 2,
quinoxalines 6(a, d) exhibited agonistic activity while
the corresponding benzimidazole 21 and quinoline 26
showed no measurable activity when tested at concen-
trations of up to 10 lM.

M. Teng et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5472–5478
Table 4. Substitution effects at the quinoxaline ring
5477
R1
R2
R3
N
N
SO2CH3
X
R4
Compound
R1
R2
R3
Cl
R4
X
EC50 (lM)
Emax (%)
6c
6e
H
Cl
H
(CH2)2CH3
(CH2)2CH3
(CH2)2CH3
1.2
—
68
—
—
94
44
82
38
31
85
42
H
CH3O
CH3
Cl
CH3O
CH3
Cl
H
6f
H
H
—
0.175
4.3
16e
16p
16d
16q
16r
16f
16s
H
H
NHCH(CH3)C2H5
NHCH(CH3)C2H5
NHCH(CH3)2
NHCH(CH3)2
NHCH(CH3)2
NHC(CH3)3
H
NO2
Cl
Cl
Cl
H
H
H
0.214
1.5
5.0
H
Cl
Cl
Cl
NO2
H
CF3
H
NO2
Cl
Cl
H
H
Cl
0.155
2.2
H
CF3
NHC(CH3)3
— indicates Emax (%) <2%.
Cl
Cl
N
SO2CH3
CO2CH3
Cl
Cl
N
N
SO2CH3
Cl
Cl
SO2CH3
Cl
Cl
N
N
N
N
SO2CH3
CO2C2H5
6a
26
6d
Emax 24%
21
NA
Emax 35%
NA
EC50 >10 uM
EC50 = 5 uM
Figure 2. Quinoxalines appeared more potent GLP-1 agonists compared to the benzimidazole and quinoline analogs.
Table 5. Correlation between the potency in binding augmentation
(IC50) and the EC50/Emax
good correlation between the potency in binding aug-
mentation (IC50) and the EC50/Emax
.
Compound
EC50 (lM)
Emax (%)
IC50 (lM)
The active quinoxalines showed good stability in simu-
lated gastric fluid media at ambient temperature for at
least 1 h, but they appeared chemically unstable when
treated with strong nucleophiles or bases. For example,
6b was hydrolyzed to 2-hydroxyquinoxaline 1b when
treated with potassium hydroxide in methanol. When
treated with 1,4-dithiothreitol (DTT), the methylsulfone
moiety was replaced by the nucleophile. The active quin-
oxalines had a high microsomal turnover rate when
incubated with human liver microsomes without addi-
tion of co-factors, indicating non-enzymatic breakdown
of the compounds. The stability issues hindered further
in vivo characterization of these compounds, although
some ex vivo experiments were conducted.11
6a
6b
5
35
40
68
24
71
63
29
46
82
85
47
75
—
—
0.253
0.091
0.404
3.1
2.6
1.2
6c
6d
>10
8a
8b
3
1.2
0.551
0.459
1.1
9
30
16c
16d
16f
16h
16i
10
0.725
0.214
0.155
2.2
0.109
0.107
0.032
0.53
1.8
0.46
—
—
>10
>10
11
— indicates Emax (%) <2%.
In conclusion, we have identified a series of 2-sulfonyl-
quinoxalines as small molecule agonists, so called ago-
allosteric modulators, for the hGLP-1 receptor. The
most potent and efficacious compounds are 6,7-dichlo-
roquinoxalines bearing an alkyl sulfonyl group at the
C-2 position and a secondary alkyl amino group at the
C-3 position. It appeared that proper polarization of
the quinoxaline ring and a suitable conformation of
the C-2 sulfonyl group are required for the observed
agonistic activities. The active quinoxalines appeared
stable to gastric fluid conditions, however, improvement
of chemical stability and pharmacokinetic properties is
As mentioned previously, active quinoxalines were able
to activate hGLP-1 receptor in the competitive binding
assay without competing with GLP-1 for the GLP-1
binding site. This suggested that the quinoxalines bind
at an allosteric site on the hGLP-1 receptor. Moreover,
the affinity of GLP-1 for the receptor increased in a
dose-dependent manner upon incubation with the quin-
oxalines. An IC50 was measured and defined as the con-
centration of the quinoxalines at which 50% of the
augmentation of GLP-1 to its receptor was reached by
the quinoxaline. As Table 5 shows, there is generally a

5478
M. Teng et al. / Bioorg. Med. Chem. Lett. 17 (2007) 5472–5478
1685; (e) Holz, C. G. T.; Kuhtreiber, W. M.; Habener, J.
F. Nature 1993, 361, 362.
4. (a) Nauck, M. A.; Kleine, N.; Orskow, C.; Holst, J. J.;
Willms, B.; Creutzfeldt, W. Diabetologia 1993, 36, 741; (b)
Zander, M.; Madsbad, S.; Madsen, J. L.; Holst, J. J.
Lancet 2002, 359, 824.
clearly necessary. The binding mode of these small mol-
ecules is a subject for future investigation and will be re-
ported in due course.
Acknowledgments
5. Vahl, T. P.; D’Alessio, D. A. Exp. Opin. Invest. Drugs
2004, 13, 177.
We thank Dr. Alex Polinski and Dr. Atsuo Kuki for
their enthusiastic support.
6. Murphy, K. G.; Bloom, S. R. Proc. Natl. Acad. Sci.
U.S.A. 2007, 104, 689.
7. Nauck, M. A.; Wollschlager, D.; Werner, J.; Holst, J. J.;
Orskov, C.; Creutzfeldt, W., et al. Diabetologia 1996, 39, 1546.
8. (a) Juhl, C. B.; Hollingdal, M.; Sturis, J.; Jakobsen, G.;
Agerso, H.; Veldhuis, J., et al. Diabetes 2002, 51, 424; (b)
Elbrønd, B.; Jakobsen, G.; Larsen, S.; Agersø, H.; Jensen,
L. B.; Rolam, P.; Sturis, J.; Hatorp, V.; Zdravkovic, M.
Diabetes Care 2002, 25, 1398.
9. Kolterman, O. G.; Buse, J. B.; Fineman, M. S.; Gaines, E.;
Heintz, S.; Bicsak, T. A., et al. J. Clin. Endocrinol. Metab.
2003, 88, 3082.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2007.06.086.
References and notes
10. Chen, D.; Liao, J., et al. Proc. Natl. Acad. Sci. U.S.A.
2007, 104, 943.
1. Kreymann, B.; Ghatai, M. A.; Williams, G.; Bloom, S. R.
Lancet 1987, 2, 1300.
11. Knudsen, L. B.; Kiel, D.; Teng, M., et al. Proc. Natl.
Acad. Sci. U.S.A. 2007, 104, 937.
2. Kieffer, T. J.; Habener, J. F. Endocr. Rev. 1999, 20, 876.
3. (a) Drucker, D.; Phillippe, J.; Mosjov, S.; Chick, W.;
Habener, J. Proc. Natl. Acad. Sci. U.S.A. 1987, 84, 3434;
(b) Wheeler, M. B.; Lu, M.; Dillon, J. S., et al. Endocri-
nology 1993, 133, 57; (c) Lu, M.; Wheeler, M. B.; Leng, X.
H.; Boyed, A. E. D. Endocrinology 1993, 132, 94; (d)
Yada, T.; Itoh, K.; Nakata, M. Endocrinology 1993, 133,
12. Iddon, B.; Kutschy, P.; Robinson, A. G.; Suschitzky, H.;
Kramer, W.; Neugebauer, F. A. J. Chem. Soc., Perkin
Trans. 1: Org. Bioorg. Chem. 1992, 22, 3129.
13. (a) Cordi, A. A.; Desos, P.; Randle, J. C. R.; Lepagnol, J.
Bioorg. Med. Chem. 1995, 3, 129; (b) Desos, P.; Lepagnol,
J.; Morain, P.; Lestage, P.; Cordi, A. A. J. Med. Chem.
1996, 39, 197.