Aminopyrazole-phenylalanine based GPR142 agonists: Discovery of tool compound and in vivo efficacy studies

Article abstract of DOI:10.1021/ml4000854

Herein, we report the lead optimization of amrinone-phenylalanine based GPR142 agonists. Structure-activity relationship studies led to the discovery of aminopyrazole-phenylalanine carboxylic acid 22, which exhibited good agonistic activity, high target selectivity, desirable pharmacokinetic properties, and no cytochrome P450 or hERG liability. Compound 22, together with its orally bioavailable ethyl ester prodrug 23, were found to be suitable for in vivo proof-of-concept studies. Compound 23 displayed good efficacy in a mouse oral glucose tolerance test (OGTT). Compound 22 showed GPR142 dependent stimulation of insulin secretion in isolated mouse islets and demonstrated a statistically significant glucose lowering effect in a mouse model bearing transplanted human islets.

Full text of DOI:10.1021/ml4000854

Letter
pubs.acs.org/acsmedchemlett
Aminopyrazole−Phenylalanine Based GPR142 Agonists: Discovery of
Tool Compound and in Vivo Efficacy Studies
Ming Yu,*,† Mike Lizarzaburu,† Alykhan Motani,† Zice Fu,† Xiaohui Du,† Jiwen (Jim) Liu,† Xianyun Jiao,†
SuJen Lai,† Peter Fan,† Angela Fu,† Qingxiang Liu,† Michiko Murakoshi,‡ Futoshi Nara,‡ Kozo Oda,‡
Ryo Okuyama,‡ Jeff D. Reagan,† Nobuaki Watanabe,‡ Mami Yamazaki,‡ Yumei Xiong,† Ying Zhang,†
Run Zhuang,† Daniel C.-H. Lin,† Jonathan B. Houze,† Julio C. Medina,† and Leping Li†,§
†Amgen Inc., 1120 Veterans Boulevard, South San Francisco, California 94080, United States
‡Daiichi-Sankyo, 1-2-58 Hiromachi, Shinagawa-Ku, Tokyo 140-8710; 1-16-13 Kitakasai, Edogawa-Ku, Tokyo 134-8650, Japan
S
* Supporting Information
ABSTRACT: Herein, we report the lead optimization of
amrinone−phenylalanine based GPR142 agonists. Structure−
activity relationship studies led to the discovery of amino-
pyrazole−phenylalanine carboxylic acid 22, which exhibited
good agonistic activity, high target selectivity, desirable
pharmacokinetic properties, and no cytochrome P450 or
hERG liability. Compound 22, together with its orally
bioavailable ethyl ester prodrug 23, were found to be suitable
for in vivo proof-of-concept studies. Compound 23 displayed good efficacy in a mouse oral glucose tolerance test (OGTT).
Compound 22 showed GPR142 dependent stimulation of insulin secretion in isolated mouse islets and demonstrated a
statistically significant glucose lowering effect in a mouse model bearing transplanted human islets.
KEYWORDS: GPR142 agonist, type 2 diabetes, aminopyrazole−phenylalanine, insulin secretagogue, prodrug,
oral glucose tolerance test, human islet transplant
PR142, a new member in the super family of seven-
Gtransmembrane G-protein-coupled receptors (GPCRs),
was first reported in 2003 by Schioth and colleagues,1 and
further characterized by Schaller in 2006.2 GPR142 is a Gq-
coupled receptor that was expressed predominantly in
pancreatic β-cells. Activation of this receptor triggers an
intracellular signal transduction pathway, which ultimately
leads to β-cell insulin secretion. Tryptophan was recently
identified as an endogenous ligand for GPR142.3 Activation of
the receptor by tryptophan was found to stimulate insulin
secretion in isolated mouse islets in both a dose- and a glucose-
dependent manner and was associated with improved glucose
tolerance.3 These findings suggested that GPR142 plays an
important role in regulating insulin secretion and glucose
homeostasis. This target was therefore deemed suitable for the
discovery of agents for the treatment of type II diabetes with a
low associated risk of hypoglycemia.
We recently reported the discovery of amrinone−phenyl-
alanine GPR142 agonist 14 and the subsequent optimization
campaign, which provided 2 for preliminary proof-of-concept
studies in rodents.5 While these compounds were both highly
potent in a human inositol phosphate (IP) accumulation assay6
(EC50 = 0.090 μM for 1 and EC50 = 0.058 μM for 2), they both
suffered from high in vivo clearance, with rat IV clearance of 3.4
and 5.1 L/h/kg, respectively (Figure 1). Compound 2 was also
found to be a strong CYP inhibitor. While replacing the
pyridone moiety of 1 with a thiadiazole ring was later
Figure 1. Structure of GPR142 agonists.
discovered to reduce rat clearance,5 this heterocyclic compound
was found to be chemically unstable at physiological pH
conditions. Further investigation of this chemical series was
therefore warranted to address the above issues and to provide
a better tool compound for in vivo proof-of-concept studies.
We first investigated the role of the central aromatic ring on
potency as summarized in Table 1. The strategy employed was
to incorporate different heterocyclic rings at the center of the
molecule in order to modulate the overall shape of the
molecule and optimize binding to the receptor. Isooxazole
compound 3 showed a 2-fold loss in intrinsic potency (0.263
Received: March 5, 2013
Accepted: May 1, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ml4000854 | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
protein). In in vivo pharmacokinetic studies, compound 4
displayed lower rat clearance (3.1 L/h/kg) than 7 (higher than
rat liver blood flow rate). Compound 4 was therefore selected
as the template for further SAR exploration.
Table 1. Middle Heterocyclic Ring Replacements
To optimize the middle pyrazole core, various pyrazole
nitrogen substitutions were investigated as illustrated in Table
2. The N-methyl substituent (12) slightly improved the
Table 2. Pyrazole Ring Modification
h-GPR142, IP
ab
,
c
compd
R1
R2
EC50 (μM)
rat LM % remain
4
−H
−CH3
−CH2CF3
−CH2CH2−
−CH2CO2H
H
H
H
0.095
0.037
0.180
2.78
33
8
12
13
14
15
9
1
H
0.516
>90
a
Values were the means of three determinations; standard deviation
b
c
was 30%. Assay in buffer; see ref 6 for assay protocol. Percentage
of compound remaining intact after incubation in rat liver microsomes
(RLM) at 1.0 μM for 30 min at 22 °C.
potency (0.037 μM) versus the unsubstituted pyrazole (4),
but none of the other N-substituents on the pyrazole ring was
as potent.7 Alkyl substitution of the pyrazole nitrogen was later
found to be detrimental to the metabolic stability in rat LM.
The percentage of compound remaining intact dropped from
33% for 4 to 8% for 12. Similarly, low stability in rat LM was
also observed with the hydrophobic triflouroethyl analogue 13
(9%). Conformational rigidification was attempted as exempli-
fied by 14, but locking the 5-amino-pyrazole into coplanar
conformation led to significant loss of potency (2.78 μM)
without any stability benefit (1%). Interestingly the carboxy-
containing pyrazole 15 markedly improved the stability in rat
LM (>90%). Since good stability was also observed previously
on another carboxylic acid compound 11 (63%) (Table 1), we
hypothesized that the physicochemical nature of a carboxylic
acid protects the molecule from microsomal metabolism.
Concurrent SAR campaign on the aminopyridone lead 1
revealed that α-methylamino substitution of the naked pyridine
ring was able to improve the metabolic stability of the
compound.4,5 This modification was incorporated into the
aminopyrazole lead series (Table 3) and improved the stability
from 8% of 12 to 38% of 16 in rat LM. The CYP inhibitory
activity was also notably reduced (86% of 16 vs 16% of 12)
with this optimization. It should also be noted that para-
fluorination of the phenylalanine ring (17) provided a marginal
improvement of stability in rat LM (49%). These modifications
were therefore integrated into the template for our next
optimization.
a
Values were the means of three determinations; standard deviation
b
c
was 30%. See ref 6 for assay protocol. % Fraction of maximal
tryptophan response at 10 μM. Not determined.
d
μM) compared with pyridone compound 1 (0.090 μM). This
compound, however, displayed lower shift of the EC50 value
(0.306 μM) in the presence of 100% of human serum.
Encouraged by this result, we explored additional analogous
heterocycles. Pyrazole 4 was found to restore the EC50 (0.095
μM) in buffer, and more interestingly, it also had low EC50 shift
in the presence of serum (0.164 μM). The 1,2,4-triazole (5)
and 1,3,4-oxadiazole (6) both had significantly lower intrinsic
potency. Thiazole compound 7 was as equally potent as 4
(0.090 μM), but the activity of similar analogues was very
sensitive to the ring orientation, and none of regioisomers 8−
10 showed comparable potency. Finally, thiophene compound
11 displayed a reasonable EC50 in buffer (0.121 μM) but was
10-fold less potent in serum.
The most potent compounds (4 and 7) were profiled in liver
microsomes (LM). Compound 4 had marginally higher
percentage of compound remaining intact (8% and 33%)
than 7 (3% and 5%) after incubation at 1.0 μM for 30 min in
human and rat LM, respectively. It also exhibited better LM
intrinsic clearances (154 and 345 μL/min/mg protein in
human and rat, respectively) than 7 (345 and 459 μL/min/mg
A breakthrough in the metabolic stability of the series was
made when we manipulated the physicochemical properties of
the molecule through substitution of the phenylalanine
nitrogen. As highlighted in Table 4, introduction of acidic
amino substituents as exemplified by 1,2,3-triazole 18, tetrazole
19, and acetic acid 20 resulted in an increase in compound
stability in rat LM. These improvements were also reflected in
B
dx.doi.org/10.1021/ml4000854 | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
these fragments, was also found to be metabolically stable and
free from inhibitory activity against CYP450 isoforms 3A4 and
2D6 (IC50 > 30 μM for both).
Table 3. Elimination of CYP Inhibitory Activity
Compound 22 was further evaluated in both in vitro and in
vivo experiments. In addition to good activity on human, 22
exhibited a GPR142-EC50 of 0.031 and 0.079 μM in mouse and
cynomolgus monkey, respectively, albeit in slightly lower
maximum efficacy (73% and 79%, respectively). It was equally
stable in human LM (>90%) as in rat. Low in vivo clearance
(0.15 L/h/kg) and reasonable half-life (4.1 h) were observed in
rats. In plasma, compound 22 had an unbounded fraction of
16% in human, 10% in cynomolgus monkey, and 1% in rat.
This compound did not show hPXR activation and was also
devoid of hERG liability (IC50 > 30 μM, PatchClamp assay).
Compound 22 was tested in an ex vivo experiment for its
ability to stimulate insulin secretion directly from isolated
mouse islets. As illustrated in Figure 2, compound 22
stimulated insulin secretion from wild-type but not GPR142
deficient islets. These results indicated that the activity of
compound 22 was specific to GPR142.
h-GPR142, IP
RLM
CYP3A4
d
ab
,
c
compd
−R
−H
−NHCH3 −H
−NHCH3 −F
−X
−H
EC50 (μM)
% remain
% activity
12
16
17
0.037
0.037
0.049
8
16
86
71
38
49
a
Values were the means of three determinations; standard deviation
was 30%. See ref 6 for assay protocol. Percentage of compound
remaining intact after incubation in rat liver microsomes (RLM) at 1.0
μM for 30 min at 22 °C. Percentage of CYP activity remaining after
b
c
d
incubation with compound at 3.0 μM.
Table 4. Improvement of Metabolic Stability
Figure 2. Insulin secretion stimulation test of 22 on isolated mice
islets; islets isolated from wild-type and GPR142 knockout mice were
incubated with 3.0 and 10.0 μM concentrations of 22 in the presence
of 16.7 mM glucose; insulin secreted into the media was quantified by
ELISA (ALPCO); mean + SEM, n = 4; statistics by t test vs wild-type;
ns = not significant.
Further profiling revealed that compound 22 had low
permeability across a Caco-2 monolayer (0.9 × 10−6 cm/s).
The poor permeability, presumably due to the zwitterionic
nature of the molecule, resulted in low oral bioavailability (3%
in rat at 2.0 mg/kg and 6% in cynomolgus monkey at 10.0 mg/
kg). Subcutaneous (SC) administration of the compound at 2.0
mg/kg overcame this hurdle with 81% bioavailability in rat and
a maximal drug concentration (Cmax) of 7.5 μM. Similar level of
drug exposure was also observed in cynomolgus monkey at the
same SC dosage (F = 100%, Cmax = 6.3 μM).
a
Values were the means of three determinations; standard deviation
b
c
was 30%. See ref 6 for assay protocol. Percentage of compound
remaining intact after incubation in rat liver microsomes (RLM) at 1.0
μM for 30 min at 22 °C. The compounds were IV dosed at 0.5 mg/
kg. Not determined.
d
e
In order to develop an orally bioavailable tool compound, we
turned to a prodrug strategy and studied the ethyl ester
analogue 23 (Table 5). After po administration of 23 at 2.0
mg/kg in rats, a 40% bioavailability of the parent acid 22 was
achieved with Cmax of 6.8 μM. The compound exposure
correlated well with escalating prodrug dosage and gave a Cmax
of 150 μM in rats at 30.0 mg/kg. Similarly, in cynomolgus
monkeys, prodrug 23 at 2.0 mg/kg delivered free acid 22 in
58% bioavailability with Cmax reaching 2.5 μM. Dosing
cynomolgus monkeys at 10 mg/kg gave a Cmax as high as
12.0 μM and provided unbound drug concentration 15-fold
over the cyno EC50 (0.079 μM) in the IP assay.
the gradual decrease of rat IV clearances. Equally as important
was that compounds 18−20 also maintained potency on
GPR142, with EC50 ranging from 0.026 to 0.071 μM. The
glycine derivative 20 represented a desirable confluence of good
potency (0.042 μM), high RLM stability (>90%), and low in
vivo rat clearance (CL = 1.0 L/h/kg). Amide 21 was also
prepared, but it was found to be metabolically unstable.
Furthermore, with the new acetic acid feature, it was found that
the amino substitution on the pyridine and the 4-fluoro of the
phenylalanine as in compound 20 were no longer necessary for
desirable metabolic stability. Compound 22, after removal of
C
dx.doi.org/10.1021/ml4000854 | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
Table 5. Pharmacokinetic Profile of Tool Compound 22 and Prodrug 23
pharmacokinetic (PK) profile
species
rat
route (compd)
dosage (mg/kg)
Cl (L/h/kg)
0.15
t1/2 (h)
Vdss (L/kg)
0.29
Cmax (μM)
tmax (h)
F (%)
i.v. (22)
p.o. (22)
s.c. (22)
0.5
2.0
2.0
2.0
30.0
4.1
0.17
7.5
0.83
0.38
0.67
0.50
3
81
40
80
a
p.o. (23)
6.8
a
p.o. (23)
150
a
mouse
cyno
p.o. (23)
10.0
6.7
0.41
i.v. (22)
p.o. (22)
s.c. (22)
0.5
0.44
2.3
0.33
10.0
2.0
0.40
6.3
4.0
6
0.50
1.50
0.83
100
58
35
a
p.o. (23)
2.0
2.5
a
p.o. (23)
10.0
12.0
a
Prodrug 23 was dosed; analytical calculations were based on parent-free acid 22.
Compounds 22 and 23 were both evaluated in mice for in
vivo proof-of-concept studies. The first experiment was
conducted with prodrug 23 in B6 mice that were fasted
overnight.8 Sitagliptin (DDP-IV inhibitor) was chosen as the
positive control at an oral dose of 10.0 mg/kg.9,10 Total change
in plasma glucose AUC (−30 to 120 min relative to glucose
challenge) as shown in Figure 3 indicated a strong, dose-
Figure 3. Oral glucose tolerance test of 23 in mice; n = 8 for control
group, and n = 4 for each dose group; plasma glucose excursion was
monitored by blood sample collection and analysis (AccuChek
glucometer) at −30, 0, 20, 40, 60, 90, and 120 min after the glucose
challenge; delta AUC calculated with plasma glucose baseline level at a
time point of 30 min before glucose challenge; mean + SEM; *p <
0.05 vs control by 1-way ANOVA.
Figure 4. Oral glucose tolerance test of 22 in intact vs human islet
transplanted nude mice; crossover study design (n = 12 for KcHIT
mice and n = 8 for intact nude mice, both groups were fasted for 4.0
h); mean + SEM, *p < 0.05 vs vehicle by 1-way ANOVA.
dependent glucose lowering effect. Direct comparison of the
positive control and the 10.0 mg/kg group suggested equivalent
efficiency of 23 and sitagliptin in lowering blood glucose level.
Compound 22 was also used as a tool compound (SC) for
oral glucose tolerance tests11 that were conducted in an intact
nude mouse and a mouse model bearing human islets,
transplanted under the kidney capsule (KcHIT mice).12,13 As
illustrated in Figure 4, direct comparison of plasma glucose
AUC in the time window of 0−90 min revealed a statistically
significant glucose lowering effect in response to 22 in both the
intact and KcHIT mice, which was similar in magnitude to a
near maximal dose of sitagliptin. These data demonstrated that
the efficacy of 22 was indistinguishable between human islet
transplanted mice and mice that bear normal intact mouse
islets. This effect was islet dependent since mice treated with
streptozotocin, but without the reconstitution of healthy islets,
did not exhibit glucose lowering or insulin secretion (data not
shown).
D
dx.doi.org/10.1021/ml4000854 | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
a
Scheme 1. Synthesis of Aminopyrazole GPR142 Agonist 22 and 23
a
Reagents and conditions: (a) CH3NHNH2, MeOH, conc. HCl, 60 °C, 71%; (b) Boc-L-phenylalanine, EDCI, py, DMF; (c) 20% TFA/CH2Cl2, 68%
(two steps); (d) glyoxylic acid, NaB(OAc)3H, HOAc, 1,2-dichloroethane/dioxane, 70 °C, 42% for 22; (e) ethyl bromoacetate, N,N-
diisopropylethylamine, 77% for 23.
Notes
The synthesis of the aminopyrazole−phenylalanine GPR142
agonists is exemplified by 22 as shown in Scheme 1. Treating
commercially available 3-oxo-3-(pyridin-4-yl)propanenitrile 24
with methylhydrazine in the presence of concentrated HCl
provided aminopyrazole 25 as the major regioisomeric
product.14 Amide coupling of 25 with boc-L-phenylalanine
following precedent procedures4,5 offered intermediate 26.
Removal of the N-Boc group from 26 provided the amino
intermediate 27, which after reductive amination with glyoxylic
acid provided the final compound 22 in total yield of 20% after
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Dr. Paul Schnier for the high resonance mass
spectroscopy analysis of the compounds discussed in this letter.
REFERENCES
■
(1) Fredriksson, R.; Hoglund, P. J.; Gloriam, D. I.; Lagerstrom, M.
C.; Schioth, H. B. Seven evolutionarily conserved human rhodopsin G
protein-coupled receptors lacking close relatives. FEBS Lett. 2003, 554,
381−388.
four steps. Reaction of 27 with ethyl bromoacetate under hunig
̈
base mediated condition offered the prodrug 23 in 77% yield.
Similar synthetic sequences were used to provide the remainder
of the compounds in this letter.
(2) Suesens, U.; Hermans-Borgmeyer, I.; Urny, J.; Schaller, H. C.
Characterisation and differential expression of two very closely related
G-protein-coupled receptors, GPR139 and GPR142, in mouse tissue
and during mouse development. Neuropharmacology 2006, 50, 512−
520.
(3) Xiong, Y.; Motani, A.; Reagan, J.; Gao, X.; Yang, H.; Ma, J.;
Schwandner, R.; Zhang, Y.; Liu, Q.; Miao, L.; Luo, J.; Tian, H.; Chen,
J.-L.; Murakoshi, M.; Nara, F.; Yeh, W.-C.; Cao, Z. Submitted for
publication.
(4) Lizarzaburu, M.; Turcotte, S.; Du, X.; Duqutte, J.; Fu, A.; Houze,
J.; Li, L.; Liu, J.; Oda, K.; Okuyama, R.; Reagan, J. D.; Yu, M.; Medina,
J. C. Discovery and optimization of a novel series of GPR142 agonists
for the treatment of type 2 diabetes mellitus. Bioorg. Med. Chem. Lett.
2012, 22, 5942−5947.
(5) Du, X.; Kim, Y.-J.; Lai, S.; Chen, X.; Lizarzaburu, M.; Turcotte, S.;
Oda, K.; Okuyama, R.; Nara, F.; Murakoshi, M.; Fu, A.; Reagan, J.; Liu,
Q.; Zhang, Y.; Motani, A.; Fan, P.; Xiong, Y.; Shen, W.; Li, L.; Houze,
J.; Medina, J. C. Phenylalanine derivatives as GPR142 agonists for the
treatment of Type II diabetes. Bioorg. Med. Chem. Lett. 2012, 22,
6218−6223.
(6) Inositol Phosphate Accumulation Assay-HEK293 cells were
dispensed into a poly-D-lysine tissue culture treated 96-well plate at a
density of 25 000 cells per well. The next day, the cells (∼80−90%
confluent) were transfected with 100 ng of receptor plasmid per well
using Lipofectamine2000 according the manufacturer’s instructions.
Six hours after transfection the media was replaced with inositol free
DMEM/10% dialyzed FCS supplemented with 1.0 μCi/mL triturated
inositol. After incubation overnight, the cells were washed once in
HBSS and then treated with the 100 μL HBSS/0.01% BSA containing
various concentrations of test compounds (prepared as above in
DMSO) and 10 mM LiCl, and incubated at 37 °C for 1 h. The media
was aspirated, and the cells were lysed with ice cold 20 mM formic
acid. After incubation at 4 °C for 5 h, the lysate were added to yttrium
silicate SPA beads, allowed to settle overnight, and read on a Beckman
In summary, the lead optimization campaign on our
aminopyrazole−phenylalanine GPR142 agonist lead series has
led to the discovery of a pyrazole ring as the optimal center ring
replacement, which provided compounds with good potency
and low serum EC50 shift. The acetic acid substituent on the
phenylalanine nitrogen significantly improved in vitro and in
vivo metabolic stability. Tool compound 22 derived from these
optimizations, together with its ethyl ester prodrug 23,
demonstrated robust in vivo efficacy through their ability to
increase insulin secretion and lowering plasma glucose level.
These effects were not unique to mouse native islets, but were
also demonstrable in an in vivo model of human islet function.
Further studies on higher species are currently in process, and
the results will be published elsewhere when they become
available.
ASSOCIATED CONTENT
■
S
* Supporting Information
Synthetic procedures and characterization data for all
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*(M.Y.) Tel: 650-244-2050. Fax 650-837-9630. E-mail: yum@
amgen.com.
Present Address
§Presidio Pharmaceuticals, Inc., 1700 Owens Street, Suite 585,
San Francisco, California 94158, United States.
E
dx.doi.org/10.1021/ml4000854 | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX

ACS Medicinal Chemistry Letters
Letter
TopCount scintillation counter. In measuring the EC50 with serum,
HBSS/0.01% BSA was replaced with 100% human serum.
(7) The investigation of pyrazole N-substitution included alkyl
groups in different sizes and physicochemcial properties (e.g., -Et, -nPr,
-iPr, and -CH2CH2OH), but none of these compounds was as potent
as 12. Moving the substitution onto the other nitrogen of the pyrazole
ring led to significant loss of potency.
(8) In this experiment, the animals were challenged in a single oral
dose of glucose (4.0 kg/kg) at the time point of zero (t = 0 min).
(9) Sitagliptin was dosed 30 min (t = −30 min) before oral glucose
challenge.
(10) In the pharmacokinetic study of 23 on fasted B6 mice at 10.0
mg/kg PO dosage, the plasma drug level of parent 22 was observed to
reach a peak level of 6.7 μM 25 min after dosing. Compound 23 was
administrated at escalating doses of 0.3, 1.0, 3.0, 10.0, and 30.0 mg/kg,
respectively, 30 min before an oral glucose challenge. The selection of
this dosing timing (t = −30 min) was to synchronize the drug tmax of
each group close to the time point of glucose challenge.
(11) In this experiment, the animals were challenged in a single oral
dose of glucose (4.0 kg/kg) at the time point of zero (t = 0 min). Both
the vehicle and the positive control sitagliptin (10.0 mg/kg) were
orally administrated at −60 min. The subject compound 22 (10 mg/
kg) was dosed (SC) at −15 min. The selection of dosing route and
timing were to synchronize the drug tmax of each group close to the
time point of glucose challenge. Blood glucose levels were determined
(AccuChek glucometer) at time points of 0, 5, 15, 25, 30, 45, 60, and
90 min.
(12) These immunodeficient mice (to avoid rejection of the
transplanted human islets) were previously treated with streptozotocin
(an islet specific toxin) to eradicate endogenous mouse islets. Thus,
this model allows for the evaluation of compound efficacy in an in vivo
setting using human islets.
(13) Jansson, J.; Eizirik, D. L.; Pipeleers, D. G.; Hakan Borg, L. A.;
Hellerstrom, C.; Andersson, A. Impairment of glucose-induced insulin
secretion in human pancreatic islets transplanted to diabetic nude
mice. J. Clin. Invest. 1996, 96, 721−726.
(14) For the chemical construction of compounds bearing other
heterocyclic B rings as discussed in Table 1, see Supporting
Information.
F
dx.doi.org/10.1021/ml4000854 | ACS Med. Chem. Lett. XXXX, XXX, XXX−XXX