Small Molecule Drug Discovery at the Glucagon-Like Peptide-1 Receptor

ArticleinExperimental Diabetes Research 2012(7329):709893 · February 2012with55 Reads
Impact Factor: 4.33 · DOI: 10.1155/2012/709893 · Source: PubMed

The therapeutic success of peptide glucagon-like peptide-1 (GLP-1) receptor agonists for the treatment of type 2 diabetes mellitus has inspired discovery efforts aimed at developing orally available small molecule GLP-1 receptor agonists. Although the GLP-1 receptor is a member of the structurally complex class B1 family of GPCRs, in recent years, a diverse array of orthosteric and allosteric nonpeptide ligands has been reported. These compounds include antagonists, agonists, and positive allosteric modulators with intrinsic efficacy. In this paper, a comprehensive review of currently disclosed small molecule GLP-1 receptor ligands is presented. In addition, examples of "ligand bias" and "probe dependency" for the GLP-1 receptor are discussed; these emerging concepts may influence further optimization of known molecules or persuade designs of expanded screening strategies to identify novel chemical starting points for GLP-1 receptor drug discovery.


Available from: Ana Belén Bueno, Nov 11, 2014
Hindawi Publishing Corporation
Experimental Diabetes Research
Volume 2012, Article ID 709893, 9 pages
Review A rticle
Small Molecule Drug Discovery at the Glucagon-Like
Peptide-1 Receptor
Francis S. Willard,
Ana B. Bueno,
and Kyle W. Sloop
Translational Science and Technologies, Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, IN 46285, USA
Centro de Investigaci
on Lilly, Eli Lilly and Company, 28108 Alcobendas, Madrid, Spain
Endocrine Discover y, Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, IN 46285, USA
Correspondence should be addressed to Kyle W. Sloop, sloop
Received 31 December 2011; Accepted 24 January 2012
Academic Editor: Matteo Monami
Copyright © 2012 Francis S. Willard et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
The therapeutic success of peptide glucagon-like peptide-1 (GLP-1) receptor agonists for the treatment of type 2 diabetes mellitus
has inspired discovery eorts aimed at developing orally available small molecule GLP-1 receptor agonists. Although the GLP-
1 receptor is a member of the structurally complex class B1 family of GPCRs, in recent years, a diverse array of orthosteric and
allosteric nonpeptide ligands has b een reported. These compounds include antagonists, agonists, and positive allosteric modulators
with intrinsic ecacy. In this paper, a comprehensive review of currently disclosed small molecule GLP-1 receptor ligands is
presented. In addition, examples of ligand bias” and “probe dependency” for the GLP-1 receptor are discussed; these emerging
concepts may influence further optimization of known molecules or persuade designs of expanded screening strategies to identify
novel chemical starting points for GLP-1 receptor drug discovery.
1. Introduction
The glucagon-like peptide-1 (GLP-1) receptor is a member
of the peptide hormone binding class B1 (secretin-like re-
ceptors) family of seven transmembrane spanning, h eter-
otrimeric G-protein coupled receptors (GPCRs). The best
characterized physiologic role of the GLP-1 receptor is to
help regulate insulin secretion from pancreatic β cells [1].
GLP-1 binding to the receptor activates Gα
, stimulating
membrane-associated adenylyl cyclases and cyclic 3
(cAMP) production which enhances glucose dependent in-
sulin secretion. The GLP-1 receptor peptide agonists, exen-
atide (exendin-4) and liraglutide, are widely approved med-
icines for the treatment of type 2 diabetes mellitus (T2DM)
Identifying and developing small molecular weight or-
ganic compounds that mimic the orthoster ic binding and
receptor activation properties of GLP-1 peptide agonists is
dicult. Class A GPCRs, for which many therapeutic small
molecules have been developed [3], are structurally distinct
from class B1 GPCRs. Class B1 receptors contain a larger
independently folded globular ectodomain (ECD) at their N-
termini. Peptide ligand binding to the ECD to initiate sig-
naling of class B1 GPCRs is mechanistically dierent com-
pared to class A receptors whose ligands primarily make
contact with residues located w ithin the membra ne spanning
α-helical regions [4]. For class B1 receptors, peptide ligands
make numerous contacts with the ECD and extracellular
loops of the transmembrane bundle [4]. For class A recep-
tors, medicinal chemistry eorts have successfully exploited
the endogenous ligand binding sites within transmembrane
domains [3]. Recent reports solving X-ray crystal structures
of class A GPCRs demonstrate the molecular interactions
used for ligand binding [510].
While basic research eorts to better understand the
intricate mechanisms regulating GLP-1 receptor function are
being aided by advancements in GPCR molecular and struc-
tural biology (see review by Willard and Sloop in this issue
of Experimental Diabetes Research), the field awaits determi-
nation of a high resolution crystal structure of a class B1
GPCR for use as a template to facilitate rational drug design
for these dicult targets. In recent years, though, there have
Page 1
2 Experimental Diabetes Research
been an increasing number of reports showing discovery of
structurally diverse small molecule ligands for the GLP-1
receptor. While the molecular details of compound-receptor
binding are largely not determined, evidence supporting
interaction of ligands with the GLP-1 receptor is provided in
many cases. While several of these molecules only have utility
as research tools, some may represent pharmacophores to
be further optimized for clinical evaluation. Importantly,
although not thought to be utilized endogenously to regulate
GLP-1 receptor signaling, there does appear to be evidence
generated for some scaolds indicating the presence of an
allosteric pocket(s) in the GLP-1 receptor [ 11 13], possibly
located within the transmembrane domains. As a ther a peutic
strateg y, small molecules targeting this pocket may be
optimized to enhance binding and signaling of endogenous
GLP-1 receptor peptides.
The development of orally available modulators of the
GLP-1 receptor for therapeutic evaluation not only requires
identification of specific nonpeptide ligands but also neces-
sitates optimizing molecules to possess appropriate physico-
chemical properties. This is the medicinal chemistry concept
of “drug-like compounds, that is, molecules possessing
functional groups and/or having properties consistent with
the majority of known drugs [14]. Careful analyses of orally
active, marketed drug s have resulted in several proposed
rules for guiding optimization of key physical properties of
compounds. Examples of these include the pioneering “rule
of five” from Lipinski [15](Ruleoffive:MW < 500 Da;
log P<5; number of hydrogen bond donors (HBD) <
5; number of hydrogen bond acceptors (HBA) < 10.
Compounds that violate two or more of these rules have
very low probability of being developed as an oral drug.)
and properties identified by Veber et al. [16](Polarsurface
area (PSA)
140; number of rotatable bonds 10. Com-
pounds that meet these two criteria have high probability
of achie ving good oral bioavailability.). These guidelines
are commonly used in medicinal chemistry strategies, and
therefore, the drug-like profile of several GLP-1 receptor
ligands is evaluated herein.
Below are descriptions of the best characterized small
molecule GLP-1 receptor ligands. Although clinical devel-
opment of any of these compounds is uncertain, the data
suggest small molecules can be identified that target the GLP-
1 receptor. In addition to descriptions of published GLP-1
receptor agonists, other chemotypes, including antagonists
and molecules only reported in the patent literature that
lack thorough biological characterization, are presented.
The comprehensive dissemination of knowledge for small
molecule ligands of this receptor may inspire advances in
chemical and biological approaches for the GLP-1 receptor.
2. Low Molecular Weight GLP-1
Receptor Antagonists
2.1. PNU-126814. A small molecule GLP-1 receptor antag-
onist, PNU-126814, was disclosed in an abstract [17]. This
compound is described to have submicromolar binding
anity for the GLP-1 receptor and inhibit GLP-1 induced
cAMP modulation and insulin secretion in RINmF5 insuli-
noma cells. Unfortunately, the chemical structure of this
compound has not been disclosed.
2.2. T-0632. The first small molecule modulator of the GLP-
1 receptor to be described in the public domain with
both annotated biology and chemistry is the antagonist T-
0632. Beinborn et al. discovered that the cholecystokinin re-
ceptor 1 antagonist, T-0632 (Figure 1 (1)sodium(S)-3-[1-
amino]-6-methoxy-2-oxo-1H-indole]propanoate), is a non-
competitive antagonist of the human GLP-1 receptor with
low micromolar potency [18]. The binding site of this
molecule is hypothesized to be in the ECD of the receptor
as Trp
, within the ECD, was identified as a critical de-
terminant for compound action. Thus, the Trp
Ser mu-
tation in the human receptor results in a
100-fold decrease
in the binding anity of the antagonist. Trp
is within 10
of the peptide binding cleft in the crystal structures of GLP-1
and exendin-4 complexed with the GLP-1 receptor ECD
[19, 20]. Trp
does not make direct contact with the peptide
ligands exendin-4 or GLP-1, but the data suggest that a small
molecule binding event in this region of the protein could
account for inhibitory activity. T-0632 can be considered
an allosteric modulator of the GLP-1 receptor. Additionally,
there is some evidence indicating the compound behaves as
an inverse agonist in a constitutively active GLP-1 receptor
system [21]. Althoug h this compound could be a good tool
for in vivo studies considering its physicochemical properties
(it passes all of the Lipinski and Veber rules), the weak af-
finity of T-0632 for the GLP-1 receptor combined with its
subnanomolar CCK1 antagonist activity renders it largely
inadequate as a research tool to study the GLP-1 receptor.
2.3. 9-Benzylpyr ido[3,4-b]Indoles. Agouron, Inc. described a
family of molecules, exemplified by 2 (Figure 1 (2) 6-(2,5-
: 5,6]pyrido[3,4-b]indol-3(2H)-one),
as GLP-1 receptor antagonists [22]. Unfortunately, the pat-
ent describing these molecules does not include specific
functional data, so the pharmacology of this chemotype of
putative GLP-1 receptor antagonists remains to be eluci-
dated. However, it is apparent that selective class B1 GPCR
antagonists with well-defined pharmacology have utility in
understanding biological systems and their sensitivity to
therapeutic intervention [23].
2.4. Nonselective Glucagon Receptor Antagonists. Avarietyof
weak GLP-1 receptor antagonists have been identified during
investigations of glucagon receptor antagonists as potential
therapeutic agents for T2DM. This finding suggests these
two receptors may share a similar binding pocket. Fig-
ure 1 depicts representative molecules that bind both the
GLP-1 and glucagon receptors. Molecules 3 to 6 (Figure 1
(3) trans-3-[[4-[[(4-tert-butylcyclohexyl)-(ptolylcarbamoyl)
amino]methyl]benzoyl]amino]propanoic acid [24]; (4) N-
zol-5-yl]-4-methyl-2-thienyl]-2-ethyl-butan amide [25]; (5)
trans-4-[[9-tert-butyl-2-oxo-3-(p-tolyl)-1,3-diazaspiro [5.5]
Page 2
Experimental Diabetes Research 3
7 (catechin)
1 (T-0632) 234
567 (
Cl Cl
Figure 1: Chemical structures of GLP-1 receptor antagonists. Representative depictions of (1) T-0632, (2) 9-benzylpyrido[3,4-b]indole,
(3-6) nonselective glucagon receptor antagonists, and (7) catechin.
undecan-1-yl]methyl]- N-(2H-tetrazol-5-yl)benzamide [26];
(6) 4-[[(2Z)-3,6-dimethyl-4-propoxy-2-(p-tolylimino)ben-
zimidazol1yl]methyl]-N-(1H-tetrazol-5-yl)benzamide [27])
bind the GLP-1 receptor with anities in the micromolar
A common characteristic of these compounds is high li-
pophilicity (calculated to be >5 for all four compounds) and
molecular weight (around 500 Da). Unfortunately, it is not
known whether such a high number of hydrophobes is re-
quired for GLP-1 receptor binding; focused optimization of
these series against the GLP-1 receptor has not been reported.
Importantly, the lipophilic nature of these molecules does
not preclude achieving oral exposure as demonstrated by
compound 6 in two animal species [27]. While all of these
compounds are more potent glucagon receptor antagonists,
and little is known about their respective structure activity
relationships (SAR) against the GLP-1 receptor, some of the
molecules could be attractive starting points for identifying
potent and selective GLP-1 receptor ligands.
2.5. Catechin. The polyphenolic natural product, catechin
(7)(Figure 1 (7) trans-2-(3,4-dihydroxyphenyl)chromane-
3,5,7-triol), has been shown to be a functionally selective,
negative allosteric modulator of the GLP-1 receptor [28].
This compound is further discussed in Section 4.
3. Low Molecular Weight GLP-1
Receptor Agonists
3.1. Quinoxalines. Teng, Knudsen et al. at Novo Nor-
disk disclosed a series of quinoxalines exemplified by 8
(Figure 2,(8) 2-[6,7-dichloro-3-(trifluoromethyl)quinoxal-
in-2-yl]sulfanyl-5-methyl-1,3,4-thiadiazole) (usually referr-
ed to in literature as “Compound 1”) and 9 (Figure 2,
2-amine) (usually referred to in literature as “Compound 2”)
[12, 29]. Initial screening using a competitive binding assay
did not provide useful hits from
500,000 compounds. A
change in strategy to perform screening using a functional
assay led to the identification of the quinoxaline scaold
250,000 compounds. Compound 9 is a full agonist in
GLP-1 receptor dependent cAMP accumulation experiments
and shows specificity for the GLP-1 receptor versus other
class B1 GPCRs. Compound 9 is characterized as an ago-
allosteric modulator of the GLP-1 receptor; it displays
intrinsic activit y and also enhances binding of GLP-1 to
theGLP-1receptor[12]. Moreover, compound 9 action is
not blocked by exendin-4
, further supporting an allo-
steric mechanism of action of this molecule. Additional
studies clearly demonstrate that compound 9 increases the
binding anity of both GLP-1 and oxyntomodulin for the
GLP-1 receptor [11]. Together, these data are important
because the results indicate the compound interacts with a
site independent of the orthosteric binding pocket, suggest-
ing the existence of an exploitable allosteric site for small
The definitive experiment to show compound 9 is a
bona fide GLP-1 receptor ligand is the demonstration that
it significantly potentiates glucose dependent insulin secre-
tion in wild type mouse islets but not in islets from
GLP-1 receptor knockout mice [12]. A subsequent report
shows intraperitoneal administration of 9 is insulinotropic
Page 3
4 Experimental Diabetes Research
8 ("compound 1")
13 (BETP)
9 ("compound 2")
14 R = Boc (Boc-5)
16 R = CO-cyclopentyl (S4P)
15 (Boc-5 monomer)
23 R
= CH
24 R = Cyclohexyl
22 R
20 21 (quercetin)
Figure 2: Chemical structures of GLP-1 receptor agonists. Representative depictions of (8–11) quinoxalines, (12) thiophene, (13) pyrim-
idine, (14–16) Boc-5 and derivatives, (17) phenylalanine, (18)azoanthracene,(19) pyrazole, (20) pyrazole-carboxamide, (21) flavonoid, and
(22–24) imidazopyridines.
Page 4
Experimental Diabetes Research 5
and enhances glucose disposal in a glucose tolerance test
Following the disclosure of compounds 8 and 9, further
SAR around the quinoxalines was conducted [31]. A sulfone,
sulfoxide, or thioether in position 2 of the bicycle is essen-
tial for activity, and quinoxalines are superior to other heter-
ocycles as GLP-1 receptor agonists. The investigation also
shows the need for electron-withdrawing groups on the
quinoxaline with 6,7-dichloroquinoxaline being the best
core. It is noted in the SAR describing the optimization of
compound 9 that the quinoxaline analogs are chemically
unstable in the presence of nucleophiles and have high mic-
rosomal metabolism. It can be speculated that the labile
sulfur-containing side chain is responsible for the instability
of these compounds to nucleophiles. The poor chemical sta-
bility precludes longer-term in vivo studies with this com-
There h ave been other eorts aimed at exploring this
scaold. Scientists from the New England Medical Center
claimed a series of 2-thiosubstituted quinoxalines, repre-
sent-ed by compound 10 (Figure 2 (10) 2-(3-methylquin-
oxalin-2-yl)sulfanyl-N-phenyl-acetamide), as weak GLP-1
receptor agonists [21]. Zydus also studied analogs of com-
pound 8 but did not show improved activity over this com-
pound in glucose dependent insulin secretion assays us-
ing RINmF5 insulinoma cells [32]. Recent reports from
the Dong-A Pharmaceutical Company [33, 34]describe
identification of a new series of 2-thioquinoxaline analogs of
8.Compound11 (Figure 2 (11) (5-[6,7-dichloro-3-[1-[1-(1-
2-yl]thiazol-2-ol), disclosed as a racemic mixture, has
100 nM potency in a cAMP response element (CRE)-lu-
ciferase reporter assay. This molecule stimulates insulin se-
cretion in INS-1E insulinoma cells and is selective against
other class B1 GPCRs. Importantly, oral dosing of this com-
pound enhances insulin secretion in a mouse intravenous
glucose tolerance test model [33, 34 ]. It appears the Dong-A
compound may overcome some of the instability issues
observed with the initial quinoxalines to achieve eective
oral exposure.
3.2. Thiophenes. Besides the quinoxaline compounds, Novo
Nordisk also disclosed a second family of GLP-1 receptor
agonists, a series of sulfonyl-thiophenes represented by com-
pound 12 (Figure 2)[35]. No biological data are provided,
and these compounds share a common feature with the
quinoxalines: a sulfonyl group attached to an aromatic ring.
While in this case, the thiophene is an electron-rich ring and
thus less prone to nucleophilic attack, one or two strong
electron-withdrawing carbonyl g roups occur in all of the
examples, leading to speculation that the sulfonyl group in
these systems is also labile.
3.3. Pyrimidines. Our group identified a series of pyrim-
idine based ago-allosteric modulators of the GLP-1 recep-
tor exemplified by r acemic compound 13 (Figure 2,(13)
pyrimidine; BETP) [13].Theparentmoleculeofthisseries
was found by screening a small library, generated from
three-dimensional pharmacophore models, using HEK293
cells stably cotransfected with the human GLP-1 receptor
and a CRE-luciferase reporter. Compound 13 shows GLP-1
receptor dependent activity in both CRE-luciferase and
cAMP accumulation assays in HEK293 cells, and the mole-
cule stimulates glucose dependent insulin secretion in ex
vivo assays of both rodent and human islet preparations.
In combination experiments with GLP-1, compound 13 is
not competitive with
I-GLP-1 but can act in an additive
manner to enhance GLP-1 induced cAMP signaling and
insulin secretion. Consistent with these findings, and similar
to the quinoxalines, compound 13 action is not blocked by
. Importantly, our studies show compound
13 induces insulin secretion in vivo. The molecule is active in
animals undergoing either the intravenous glucose tolerance
test or the hyperglycemic clamp assay. While these in vivo
results are encouraging, significant improvement in various
metabolic liabilities of this compound are necessary before
longer term studies can be explored. Although compound
13 is stable upon incubation in plasma, it shows chemical
instability in the presence of nucleophiles.
3.4. Boc-5. A sig nificant advance in the development of
small molecule GLP-1 receptor agonists is the discovery of
substituted cyclobutanes, exemplified by compound 14
(known as Boc-5) (Figure 2 (14) 1,3-bis [[4-(tert-butoxy-
carbonylamino)benzoyl]amino]-2,4-bis [3-methoxy-4-(thi-
lic acid). B oc-5 was discovere d using a high throughput CRE-
luciferase screen for activators of the rat GLP-1 receptor
[36]. Serendipity played a role in this discovery as the
original compound selected for testing was an olefin that
is half of the size of Boc-5. It was soon realized that the
olefins (depicted as compound 15, monomer of Boc-5)
dimerize in the DMSO solution to become cyclobutanes,
represented by Boc-5 and S4P (compound 16), that are the
real actives. While lacking structural characteristics of drug-
like molecules (Boc-5 violates all of the Lipinsky and Veber
rules), Boc-5 provides a useful proof of concept molecule
for nonpeptide GLP-1 receptor agonists. Pharmacologically,
Boc-5 is a full agonist in the CRE-luciferase assay, while the
closely related molecule S4P is a partial agonist. In a cAMP
accumulation assay, however, both Boc-5 and S4P are partial
agonists. Importantly, both S4P and Boc-5 are functionally
antagonized by exendin-4
and displace
I-GLP-1 in
receptor binding assays. Inhibition of
I-GLP-1 binding
does not appear saturable [36], possibly suggesting an al-
losteric binding mechanism; however, this has yet to be
Boc-5 dose dependently stimulates glucose dependent
insulin secretion in isolated rat islets [36]. Paradoxically,
10 μMofBoc-5ismoreecacious than a saturating con-
centration of GLP-1 in inducing insulin secretion. It is
worth noting that these studies were performed at an ab-
normally high glucose concentration of 25 mM. Boc-5 is
reported to have a plasma half life of approximately 8
hours following intraperitoneal dosing of the compound.
Acute administration of Boc-5 shows an anorectic eect in
C57/B6 mice that lasts over 12 hours in the intraperitoneally
Page 5
6 Experimental Diabetes Research
dosed group and 90 minutes in orally administered animals
[36]. Although no exposure is reported for the oral study,
the large dierence in food intake observed in oral versus
intraperitoneal administration probably indicates poor oral
bioavailability of the compound. Chronic intraperitoneal
administration to db/db mice lowers HbA1c, reduces food
intake, lowers body weight, and enhances insulin secretion
and glucose excursion (using an intrap eritoneal glucose
tolerance test) [36]. Impor t antly, acute eects of Boc-5 are
entirely abrogated by coadministration of exendin-4
suggesting a GLP-1 receptor dependent action [36]. Follow-
up studies confirmed many of these findings using a diet-
induced obesity mouse model [37]. A limited SAR around
Boc-5 has been reported despite synthetic challenges of these
molecules [38]. While some improvement in potency is
shown, the new molecules share the poor physicochemical
properties with S4P and Boc-5, which presumably preclude
these compounds from being oral drugs.
3.5. P henylalanines. Argusina, Inc. disclosed phenylalanine
derivatives as GLP-1 receptor modulators, represented by
compound 17,(Figure 2 (17) 3-[4-[5-[4-(tertbutoxycarbon-
[[5-(p-tolyl)furan-2-carbonyl]amino]propanoic acid) [39].
The molecules are disclosed as racemic mixtures. One can
envision the Argusina compounds as an optimization of the
monomer of Boc-5. Thorough pharmacological evaluation
of these compounds has not been disclosed, although the
initial report suggests that compounds can modulate cAMP
mobilization and insulin secretion in cell culture systems
[39]. From a pharmacokinetic perspective, these molecules
violate the Lipinski and Veber rules (large molecular weight
and high lipophilicity; high PSA), so an improvement of the
physicochemical properties of the scaold would likely be
required to achieve an orally active agent.
3.6. Azoanthracenes. Several patent disclosures from Tran-
stech Pharmaceuticals (TTP) report identification of a num-
ber of azoanthracene and oxadiazoanthracene derivatives as
GLP-1 receptor agonists [4043], exemplified by compound
18,(Figure 2 (18)(S)-3-(4
}propionic acid) [43].
Molecules described in these disclosures have nanomolar
potencies for the GLP-1 receptor in recombinant cell assays
of cAMP, and there is some indication from the data that the
compounds may be partial agonists [42]. A recent publica-
tion disclosed that TTP molecules are eective antidiabetic
agents in preclinical rodent models of T2DM, stimulating
glucose dependent insulin secretion in rodent islets and
improving glucose excursion in an oral glucose tolerance test
[44]. The leading molecule in this class, TTP054 (structure
not disclosed) is reported to currently be under evaluation
in Phase II clinical trials as an oral drug [ 44]. All of the
molecules disclosed by TTP in these patents are of large
molecular weight a nd high lipophilicity (18:MW
880 Da,
calculated log P
= 10, 13 rotatable bonds) compared to
typical orally administered medications, suggesting high
doses of the compound would likely be required to achieve
ecacious exposure.
3.7. Pyrazoles. A series of pyrazoles represented by com-
pound 19 (Figure 2 (19) 4-chloro-2-[[(E)-2-(2,5-dimethyl-
4-nitro-pyrazol-3-yl)vinyl]amino]phenol) were identified by
Kopin and Beinborn as weak agonists of the GLP-1 receptor
[21]. The compound has micromolar activity for the GLP-1
receptor in recombinant cell assays of cAMP [21]. No other
studies have been disclosed with this kind of molecule.
3.8. Pyrazole-Carboxamides. A recent publication reported
discovery of small molecule GLP-1 receptor potentiators as
an ancillary outcome of eorts to identify glucagon receptor
antagonists [45]. The authors used virtual screening of a
library of commercially available drug-like compounds to
search for compounds with physicochemical similarities to
known glucagon receptor antagonists. This was followed by
a homology model based docking approach. Compound 20,
(Figure 2 (20) 3-(4-bromophenyl)-N-(4-methoxyphenyl)-1-
phenyl-pyrazole-4-carboxamide), identified as a potential
candidate for glucagon receptor antagonism, does not show
functional activity at this receptor, but it is observed to
potentiate an EC
concentration of GLP-1 induced cAMP
production in TC6 cells [45]. The authors claim to have
discovered a novel small molecule chemoty pe that poten-
tiates the GLP-1 receptor. If these data are confirmed in
recombinant cell systems where GLP-1 receptor dependence
can be more definitively ascribed, this would be a break-
through discovery w ith respect to GLP-1 receptor allosteric
modulators. The ability to significantly potentiate either
the anity or ecacy (the data in this repor t do not dis-
tinguish between these possibilities) of GLP-1 would be
highly desirable characteristics of a GLP-1 receptor targeted
small molecule. Moreover, this would represent an important
computational and operational approach to discovering class
B1 GPCR positive al losteric modulators.
3.9. Flavonoids. Sexton et al. characterized a series of quer-
cetin-like flavonoids, represented by quercetin (Figure 2
(21) 2-(3,4-dihydroxyphenyl)-3,5,7-trihydroxy-chromen-4-
one), with positive allosteric modulator activity on the GLP-1
receptor [28, 46]. These compounds were originally identi-
fied by Domain Therapeutics as GLP-1 receptor modulators
[46]. A thorough analysis of this chemical scaold demon-
strates this class of molecules positively modulates the anity
and ecacy of GLP-1 receptor peptide ligands. The quercetin
series appears to show ligand bias in that it is specific
for GLP-1 receptor mediated Ca
mobilization but not
cAMP accumulation. Typical of polyphenolic compounds,
a lack of robust functional eects, polypharmacology, and
flat SAR limits the usefulness of this chemotype to in
vitro studies and precludes optimization for pharmacological
purposes. However, the discovery and analysis of this series of
molecules represent a further proof of concept for allosteric
modulators of the GLP-1 receptor.
3.10. Imidazopyridines. There is one published report de-
scribing a series of imidazopyridine based molecules. T he
Page 6
Experimental Diabetes Research 7
initial hit, compound 22 (Figure 2 (22) [3-(8-chloro-6-
methyl-imidazo [1,2-a]py ridin-2-yl)phenyl] 3-methylbut-2-
enoate), was identified from a library of 10,000 compounds
using a GLP-1 receptor functional assay. The strategy used
for optimization was to design a pharmacophore model
from three known agonists (compounds 9 and 12 and the
monomer of S4P, a close analog of monomer 15), although
there is no evidence of structural similarity between these
three series. Nevertheless, analogs of compound 22 that fit
additional features of the pharmacophore were designed.
Both compounds 23 and 24,(Figure 2 (23) 3-[8-chloro-
6-(trifluoromethyl)imidazo [1,2-a]pyridin-2-yl]phenyl] ace-
tate and (24) [3-[8-chloro-6-(trifluoromethyl)imidazo [1,2-
a]pyridin-2-yl]phenyl] cyclohexanecarboxylate), show in-
duc-tion of GLP-1 receptor signaling in GLP-1 receptor
expressing CHO or HEK293 cells [47]. These molecules are
shown to have some agonist activity in heterologous cell lines
expressing the GLP-1 receptor, but further investigation in
other systems is not reported. While these are very small
molecules, and pass the Lipinski and Veber rules, further
optimization of the activity of the compounds against the
GLP-1 receptor, and SAR work to move away from the labile
ester of the phenol, is likely necessary for these compounds
to become useful for oral studies.
4. Ligand Biased Signaling
Studies to more fully characterize several of the small mol-
ecule ligands of the GLP-1 receptor are needed to advance
the field. An emerging concept in GPCR pharmacology is
that of functional selectivity (also known as ligand bias or
stimulus bias) [48, 49]. It is now appreciated that GPCR
ligands can stabilize distinct receptor conformations, which
in some instances, lead to dierential modulation of signal
transduction pathways [48]. To date, there is a single study
reporting that the weak GLP-1 receptor peptide agonists,
oxyntomodulin and glucagon, are biased toward Gα
β-arrestin for coupling to the GLP-1 receptor [50]. These
peptides show low potency and part ial agonism (>50% of
GLP-1 ecacy) for β-arrestin recruitment at the GLP-1
receptor yet are full agonists for cAMP accumulation. In
support of this, another report describes oxyntomodulin
as having stimulus bias for the ERK1/2 activation pathway
relative to cAMP or Ca
mobilization [11]. Further work
is necessary to understand whether the observed in vitro
functional selectivity of the various GLP-1 receptor peptide
ligands is physiologically relevant in vivo. Such studies may
be complicated by the polypharmacology of oxyntomodulin
as it is a dual agonist of both glucagon and GLP-1 receptors
and by receptor reserve phenomena that often lead to the
classification of partial agonists as biased ligands [51].
Sexton et al. demonstrated two GLP-1 receptor ligands
are functionally selective. Quercetin (21)andasubsetofre-
lated naturally occurring flavonoids exhibit functional selec-
tivity as these molecules potentiate GLP-1 peptide signaling
in Ca
mobilization but not cAMP production [11, 28].
Similarly, the flavonoids also display probe dependence in
that the molecules modulate both GLP-1 and exendin-4
action but not oxyntomodulin signal transduction. While
a variety of flavonoid cores were tested in these studies,
only the 3-hydroxyflavone core displayed functional activ-
ity. Unfortunately, flavonoids exert multiple pharmacologic
eects at the concentrations required for the activation of the
GLP-1 receptor. Thus, these are not good starting points for
the identification of more potent allosteric ligands.
An SAR analysis of related compounds identified the pol-
yphenolic natural product, catechin (7), as a probe de-
pendent, functionally selective, and negative allosteric mod-
ulator of the GLP-1 receptor [28]. Both trans enantiomers
of this compound are known as catechin, but it is not clear
from this publication if the racemic compound or one trans
enantiomer was used in this study. Catechin decreases the
ecacy of GLP-1 signaling via cAMP but does not mod-
ulate non-cAMP signaling by the GLP-1 receptor peptide
agonists nor does it significantly alter the pharmacology
of exendin-4 or oxyntomodulin signaling. Analogously, the
small molecule quinoxaline 8 [12] displays a complex profile
of activity consistent with functional selectivity and probe
dependence. Compound 8 shows anity driven positive
allosteric modulator activity for oxyntomodulin, and to a
lesser extent, GLP-1 but not toward exendin-4. This activity
is only observed for the cAMP pathway and not for the Ca
or ERK pathways, demonstrating probe dependence whereby
allosteric modulation of the GLP-1 receptor is dependent on
the species of the bound orthosteric agonist.
These studies are seminal as they provide proof of con-
cept that allosteric modulation of the GLP-1 receptor can
engender pathway specific modulation of signal transduction
outcomes. It would be informative to identify small molecule
ligands of the GLP-1 receptor with biased signaling to
delineate the in vivo consequences of selective modulation of
specific molecular signal transduction mechanisms. Impor-
tantly, this would provide a better understanding of the
molecular mechanisms of the antidiabetic eects of GLP-
1 receptor agonism. At this time, these in vitro studies
demonstrate that probe dependence of allosteric modula-
tors can occur at the GLP-1 receptor and identify critical
aspects to be considered when optimizing small molecule
GLP-1 receptor agonists or modulators. Future eorts also
should be directed at mapping the structural determinants
of al losteric ligand binding and at evaluating interaction
between allosteric and orthosteric sites.
5. Conclusions
Despite 20 years of research following the molecular iden-
tification and cloning of the GLP-1 receptor [52, 53], no
orally available small molecule GLP-1 receptor activator has
been developed for therapeutic use. Encouragingly, however,
the pace of identifying small molecule GLP-1 receptor
ligands is increasing. In addition to the molecules discussed
herein, the field anxiously awaits disclosures from both
Addex Pharmaceuticals S.A. and Vivia Biotech S.L. that have
recently presented data from their smal l molecule GLP-1
receptor modulator programs (see Cambridge Healthtech
Institute, Discovery On Target 2011, Cambridge, MA, USA,
Nov 2–4, 2011). Although the mechanisms of peptide
binding and receptor act ivation of the GLP-1 receptor are
Page 7
8 Experimental Diabetes Research
complex and likely dicult to mimic with low molecular
weight compounds, examples of small molecules working
via an allosteric mode provide compelling evidence that
medicinal chemistry strategies for this target should be
considered. Further, the application of advanced structural
biology methodologies and more sophisticated assay systems
and testing schemes, including work to understand biased
signaling for GLP-1 receptor ligands, will likely be needed to
advance drug-like molecules.
While this paper was under editorial review, a patent ap-
plication from Receptos Inc. (WO2011/156655 A2) was pub-
lished disclosing novel compounds with allosteric modulator
activity at the GLP-1 receptor.
Author’s Contribution
F. S. Willard, A. B. Bueno, and K. W. Sloop contributed
equally to this paper.
Conflict of Interests
All authors are employees of Eli Lilly and Company and may
own company stock or possess stock options.
[1] J. J. Holst, “Glucagon-like peptide-1: from extract to agent.
The claude bernard lecture, 2005, Diabetologia, vol. 49, no.
2, pp. 253–260, 2006.
[2] J. A. Lovshin and D. J. Drucker, “Incretin-based therapies for
type 2 diabetes mellitus, Nature Revie ws Endocrinology, vol. 5,
no. 5, pp. 262–269, 2009.
[3] D.E.Grigoriadis,S.R.J.Hoare,S.M.Lechner,D.H.Slee,and
J. A. Williams, “Drugability of extra cellular targets: discover y
of small molecule drugs targeting allosteric, functional, and
subunit-selective sites on GPCRs and ion channels, Neuropsy-
chopharmacology, vol. 34, no. 1, pp. 106–125, 2009.
[4] S. R. Hoare, “Mechanisms of peptide and nonpeptide ligand
binding to Class B G-protein-coupled receptors, Drug Discov-
ery Today, vol. 10, no. 6, pp. 417–427, 2005.
angstrom crystal structure of a human A2A adenosine recep-
tor bound to an antagonist, Science, vol. 322, no. 5905, pp.
1211–1217, 2008.
[6] T. Warne, M. J. Serrano-Vega, J. G. Baker et al., “Structure of a
β1-adrenergic G-protein-coupled receptor, Nature, vol. 454,
no. 7203, pp. 486–491, 2008.
[7] T. Warne, R. Moukhametzianov, J. G. Baker et al., The struc-
tural basis for agonist and partial agonist action on a beta(1)-
adrenergic receptor, Nature, vol. 469, no. 7329, pp. 241–244,
[8] V. Cherezov, D. M. Rosenbaum, M. A. Hanson et al., “High-
resolution crystal structure of an engineered human β2-
adrenergic G protein-coupled receptor, Science, vol. 318, no.
5854, pp. 1258–1265, 2007.
[9] E. Y. Chien, W. Liu, Q. Zhao et al., “Structure of the human
dopamine D3 receptor in complex w ith a D2/D3 selective
antagonist, Science, vol. 330, no. 6007, pp. 1091–1095, 2010.
[10] T. Shimamura, M. Shiroishi, S. Weyand et al., “Structure of
the human histamine H 1 receptor complex with doxepin,
Nature, vol. 475, no. 7354, pp. 65–72, 2011.
[11] C. Koole, D. Wootten, J. Simms et al., Allosteric ligands of
the glucagon-like peptide 1 receptor (GLP-1R) dierentially
modulate endogenous and exogenous peptide responses in a
pathway-selective manner: implications for drug screening,
Molecular Pharmacology, vol. 78, no. 3, pp. 456–465, 2010.
[12] L. B. Knudsen, D. Kiel, M. Teng et al., “Small-molecule ago-
nists for the glucagon-like peptide 1 receptor, Proceedings of
the National Academy of Sciences of the United States of Amer-
ica, vol. 104, no. 3, pp. 937–942, 2007.
[13] K. W. Sloop, F. S. Willard, M. B. Brenner et al., “Novel small
molecule glucagon-like peptide-1 receptor agonist stimulates
insulin secretion in rodents and from human islets, Diabetes,
vol. 59, no. 12, pp. 3099–3107, 2010.
[14] W. P. Walters and M. A. Murcko, “Prediction of drug-like-
ness, Advanced Dr ug Delivery Reviews, vol. 54, no. 3, pp. 255–
271, 2002.
[15] C. A. Lipinski, “Dr ug-like properties and the causes of poor
solubility and poor permeability, Journal of Pharmacological
and Toxicological Methods, vol. 44, no. 1, pp. 235–249, 2000.
Ward, and K. D. Kopple, “Molecular properties that influ-
ence the oral bioavailability of drug candidates, Journal of
Medicinal Chemistry, vol. 45, no. 12, pp. 2615–2623, 2002.
[17] A. Gronberg, E.-R. Sol, A. Danielsson et al., A low molecular
weight non-peptide antagonist of glucagon-like peptide-1,
Diabetes, vol. 49, supplement 1, p. A251, 2000.
[18] E. C. Tibaduiza, C. Chen, and M. Beinborn, A small mole-
cule ligand of the glucagon-like peptide 1 receptor targets
its amino-terminal hormone binding domain, Journal of
Biological Chemistry, vol. 276, no. 41, pp. 37787–37793, 2001.
[19] S. Runge, H. Thøgersen, K. Madsen, J. Lau, and R. Rudolph,
“Crystal structure of the ligand-bound glucagon-like peptide-
1 receptor extracellular domain, Journal of Biological Chem-
istry, vol. 283, no. 17, pp. 11340–11347, 2008.
[20] C. R. Underwood, P. Garibay, L. B. Knudsen et al., “Crystal
structure of glucagon-like peptide-1 in complex with the
extracellular domain of the glucagon-like peptide-1 receptor,
Journal of Biological Chemistry, vol. 285, no. 1, pp. 723–730,
[21] A. S. Kopin and M. Beingorn, WO2004/203310 Methods and
Compositions for the Treatment of Metabolic Disorders, 2004.
[22] L. K. Truesdale et al., “US 6,469,021 B1 Non-Peptide Antago-
nists Of GLP-1 Receptor And Methods Of Use, U.S. Patent,
Editor, Agouron Pharmaceuticals, San Diego, Calif, USA,
[23] D. R. Gehlert, A. Cippitelli, A. Thorsell et al., “3-(4-Chloro-
2-morpholin-4-yl-thiazol-5-yl)-8-(1-ethylpropyl)-2, 6-
dimethyl-imidazo[1,2-b]pyridazine: a novel brain-penetrant,
orally available corticotropin-releasing factor receptor 1
antagonist with ecacy in animal models of alcoholism,
Journal of Neuroscience, vol. 27, no. 10, pp. 2718–2726, 2007.
[24] J. Lau, G. Behrens, U. G. Sidelmann et al., “New β-alanine
derivatives are orally available glucagon receptor antagonists,
Journal of Medicinal Chemistry, vol. 50, no. 1, pp. 113–128,
[25] J. L. Duy, B. A. Kirk, Z. Konteatis et al., “Discovery and
investigation of a novel class of thiophene-derived antagonists
of the human glucagon receptor, Bioorganic and Medicinal
Chemistry Letters, vol. 15, no. 5, pp. 1401–1405, 2005.
[26] D. M. Shen, F. Zhang, E. J. Brady et al., “Discover y of novel,
potent, and orally active spiro-urea human glucagon receptor
Page 8
Experimental Diabetes Research 9
antagonists, Bioorganic and Medicinal Chemistry Le tters, vol.
15, no. 20, pp. 4564–4569, 2005.
[27] C. Sinz, J. Chang, A. R. Lins et al., “Discovery of cyclic guan-
idines as potent, orally active, human glucagon receptor an-
tagonists, Bioorganic and Medicinal Chemistry Lette rs, vol. 21,
no. 23, pp. 7131–7136, 2011.
[28] D. Wootten, J. Simms, C. Koole et al., “Modulation of the
glucagon-like peptide-1 receptor signaling by naturally occur-
ring and synthetic flavonoids, Journal of Pharmacology and
Exper imental Therapeutics, vol. 336, no. 2, pp. 540–550, 2011.
[29] M. Teng, L. K. Truesdale, and D. Bhumralkar, WO2000/42026
Non-Peptide GLP-1 Agonists, 2000.
[30] N. Irwin, P. R. Flatt, S. Patterson, and B. D. Green, “Insulin-
releasing and metabolic eects of small molecule GLP-1 re-
ceptor agonist 6,7-dichloro-2-methylsulfonyl-3-N-tertbutyl-
aminoquinoxaline, European Journal of Pharmacology, vol.
628, no. 1–3, pp. 268–273, 2010.
cule ago-allosteric modulators of the human glucagon-
like peptide-1 (hGLP-1) receptor, Bioorganic and Medicinal
Chemistry Letters, vol. 17, no. 19, pp. 5472–5478, 2007.
[32] R. H. Bahekar, M. R. Jain, A. A. Gupta et al., “Synthesis and
antidiabetic activity of 3,6,7-trisubstituted-2-(1H-imidazol-
2-ylsulfanyl)quinoxalines and quinoxalin-2-yl isothioureas,
Archiv der Pharmazie, vol. 340, no. 7, pp. 359–366, 2007.
[33] H. S. Moon et al., WO2011/122815 A2 Novel Quinoxilane
Derivatives, 2011, Edited by WIPO.
[34] M. K. Kim et al., “590-P Orally active small molecule GLP-
1R agonist, DA-15864 increases insulin secretion in mice, in
Proceedings of the 70th Scientific Sessions of Annual Meeting of
American Diabetes Association, Orlando, Fla, USA, 2010.
[35] S. M. Knudsen et al., WO2006/003096 Condensed Thiophene
Dernatives And Their Use As Cyclic GLP-1 Agonists, 2006.
[36] D. Chen, J. Liao, N. Li et al., A nonpeptidic agonist of gluc-
agon-like peptide 1 receptors with ecacy in diabetic db/db
mice, Proceedings of the National Acade my of Sciences of the
United States of America, vol. 104, no. 3, pp. 943–948, 2007.
[37] M. He, H. Su, W. Gao et al., “Reversal of obesity and insulin
resistance by a non- peptidic glucagon-like peptide-1 receptor
agonist in diet-induced obese mice, Plos ONE, vol. 5, no. 12,
Article ID e14205, 2010.
[38] Q. Liu, N. Li, Y. Yuan et al., “Cyclobutane derivatives as
novel non-peptidic small molecule agonists of glucagon-like
peptide-1 receptor, Journal of Medicinal Chemistry, vol. 55,
no. 1, pp. 250–267, 2012.
[39] J. Liao et al., “US2011/023482 Phenylalanine Derivatives And
Their Use As Non-Peptide GLP-1 Receptor Modulators,
U.S.P. Application, Editor, Argusina Inc, San Diego, Claif,
USA, 2011.
[40] M. Rao, PCT/US2009/039905 Ligands For The Glp-1 Receptor
And Methods For D iscovery Thereof, I.P. Application, Editor,
TransTech Pharma, 2009.
[41] D. R. Polisetti et al., “PCT/US2010/047661 Solid Composi-
tions Comprising An Oxadiazoanthracene Compound And
Methods Of Making And Using The Same, I.P. Application,
Editor, TransTech Pharma, 2009”.
[42] A. M. M. Mjallli et al., PCT/US2009/036333 Oxadiazoan-
thracene Compounds For The Treatment Of Diabetes, I.P.
Application, Editor, TransTech Pharm, 2009.
[43] D. R. Polisett, E. Benjamin, J. C. Quada et al., WO2011/031620
Solid Composition Comprising an Oxadiazoanthracene Com-
pound and Methods of Making and Using the Same, 2011.
[44] C. Valcarce et al., “Oral glucagon-like peptide-1 (GLP-1)
receptor agonists for the treatment of type 2 diabetes, in
Proceedings of the 4th International Congress on Prediabetes and
the Metabolic Syndrome, Madrid, Spain, April 2011.
[45] C. de Graaf, C. Rein, D. Piwnica et al., “Structure-based dis-
covery of allosteric modulators of two related class B G-pro-
tein-coupled receptors, ChemMedChem, vol. 6, no. 12, pp.
2159–2169, 2011.
[46] S. Schann, S. Mayer, M. Frauli et al., “Sounds of silence: inno-
vative approach for identification of novel GPCR-modulator
chemical entities, in Proceedings of the 238th American Chem-
ical Society National Meeting
, Washington, DC, USA, 2009.
[47] Y. Gong, H. G. Cheon, T. Lee, and N. S. Kang, A Novel 3-
phenyl acetate skeleton and phar macophore model as gluca-
gon-like peptide 1 receptor agonists, Bullet in of the Korean
Chemical Society, vol. 31, no. 12, pp. 3760–3764, 2010.
[48] T. Kenakin, “Functional selectivity and biased receptor signal-
ing, Journal of Pharmacology and Experimental Therapeutics,
vol. 336, no. 2, pp. 296–302, 2011.
[49] P. Keov, P. M. Sexton, and A. Christopoulos, Allosteric mod-
ulation of G protein-coupled receptors: a pharmacological
perspective, Neuropharmacology, vol. 60, no. 1, pp. 24–35,
[50] R. Jorgensen, V. Kubale, M. Vrecl, T. W. Schwartz, and C.
E. Elling, “Oxyntomodulin dierentially aects glucagon-
like peptide-1 receptor β-arrestin recruitment and signaling
through Gαs, Journal of Pharmacology and Experimental
Therapeutics, vol. 322, no. 1, pp. 148–154, 2007.
[51] S. Rajagopal, S. Ahn, D. H. Rominger et al., “Quantifying lig-
and bias at seven-transmembrane receptors, Molecular Phar-
macology, vol. 80, no. 3, pp. 367–377, 2011.
[52] B. Thorens, “Expression cloning of the pancreatic β cell recep-
tor for the gluco- incretin hormone glucagon-like peptide 1,
Proceedings of the National Academy of Sciences of the United
States of America, vol. 89, no. 18, pp. 8641–8645, 1992.
[53] M. P. Graziano, P. J. Hey, D. Borkowski, G. G. Chicchi, and
C. D. Str ader, “Cloning and functional expression of a human
glucagon-like peptide-1 receptor, Biochemical and Biophysical
Research Communications, vol. 196, no. 1, pp. 141–146, 1993.
Page 9