Discovery of a novel trans-1,4-dioxycyclohexane GPR119 agonist series

Article abstract of DOI:10.1016/j.bmcl.2015.04.102

The design and optimization of a novel trans-1,4-dioxycyclohexane GPR119 agonist series is described. A lead compound 21 was found to be a potent and efficacious GPR119 agonist across species, and possessed overall favorable pharmaceutical properties. Com

Full text of DOI:10.1016/j.bmcl.2015.04.102

Bioorganic & Medicinal Chemistry Letters 25 (2015) 3034–3038
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of a novel trans-1,4-dioxycyclohexane GPR119 agonist
series
⇑
Sangdon Han , Sanju Narayanan, Sun Hee Kim, Imelda Calderon, Xiuwen Zhu, Andrew Kawasaki,
Dawei Yue, Juerg Lehmann, Amy Wong, Daniel J. Buzard, Graeme Semple, Chris Carroll, Zhi-Liang Chu,
Hussein Al-Sharmma, Hsin-Hui Shu, Shiu-Feng Tung, David J. Unett, Dominic P. Behan, Woo Hyun Yoon,
Michael Morgan, Khawja A. Usmani, Chuan Chen, Abu Sadeque, James N. Leonard, Robert M. Jones
Arena Pharmaceuticals, 6154 Nancy Ridge Drive, San Diego, CA 92121, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The design and optimization of a novel trans-1,4-dioxycyclohexane GPR119 agonist series is described.
A lead compound 21 was found to be a potent and efficacious GPR119 agonist across species, and
possessed overall favorable pharmaceutical properties. Compound 21 demonstrated robust acute and
chronic regulatory effects on glycemic parameters in the diabetic or non-diabetic rodent models.
Ó 2015 Elsevier Ltd. All rights reserved.
Received 31 March 2015
Revised 23 April 2015
Accepted 30 April 2015
Available online 23 May 2015
Keywords:
T2DM
GPR119 agonist
GDIR
Trans-1,4-dioxycyclohexane
Type 2 diabetes mellitus (T2DM) caused by insulin resistance
and loss of b-cell function is a serious metabolic disorder that
has reached pandemic proportions worldwide.1 Due to a lack of
effective and safe long-term treatments for this chronic disease,
there is a strong demand for novel therapeutic approaches.
GPR119, a class A GPCR, is highly expressed in insulin producing
pancreatic b-cells and incretin releasing intestinal endocrine cells.
Activation of GPR119 elevates intracellular cAMP levels, triggering
glucose-dependent insulin secretion from b-cells and incretin (e.g.,
GLP-1 and GIP) release from intestinal cells.2,3 GLP-1 and GIP pro-
mote b-cell viability and also stimulate insulin secretory activity by
interacting with their receptors expressed on b-cells.4,5 Hence,
orally acting GPR119 agonists may constitute a new therapy for
improvement of glucose tolerance in patients with T2DM.6
Importantly, GPR119 mediated glucose control is expected to be
associated with a low risk of hypoglycemia due to its glucose-de-
pendent mechanism of action.
GPR119 agonists have had a central pyrimidine pharmacophore
as a preferred structural feature. With the goal of identifying alter-
natives to this preferred motif, we replaced the pyrimidine with a
cyclohexane ring (e.g., 3, 4, Fig. 1). The 1,3-diether derivative (3)
was devoid of activity at the human GPR119 receptor (EC50
>100 lM), however, the 1,4-diether analogue (4, a mixture
cis/trans) possessed modest agonist activity in a cyclase assay
(EC50 >200 nM). Encouraged by this result, we separated the two
geometric isomers (cis- and trans-4) to determine if the orientation
of the functional groups was important for receptor activity. The
superior potency of the trans-4 isomer verified the importance of
this geometric configuration (Table 1), and led us to the discovery
of a novel trans-1,4-dioxycyclohexane GPR119 agonist series.
SAR studies began with an investigation of the aryloxy group in
an attempt to identify a motif with the best balance of potency,
efficacy, and physicochemical properties (Table 1). The trans-iso-
mers depicted in Table 1 were prepared in a similar manner as
We previously disclosed the discovery of APD597 (1) and its
favorable preclinical and clinical profile.7,8 Following this, we
recently reported a 5-fluoro-4,6-dialkoxypyrimidine series (e.g.,
2) that exhibited improved in vitro agonist efficacy and possessed
reduced CYP2C9 inhibitory potential.9 Thus far, our prototypical
described in Scheme 1. Alternatively, a Mitsunobu reaction
employing cis-4-((1-methylpiperidin-4-yl)oxy)cyclohexanol (cis-
41) and a corresponding aromatic alcohol was also effective.
Several aromatic rings such as phenyl (4, 5, and 7), picoline (10),
and pyrazine (12, 14) were well tolerated in conjunction with a
methyl sulfone or nitrile. Presumably, these substituents provide
a favorable hydrogen bonding acceptor interaction with GPR119
and contribute to the potency observed. Of the more potent
⇑
Corresponding author.
E-mail address: shan@arenapharm.com (S. Han).
http://dx.doi.org/10.1016/j.bmcl.2015.04.102
0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.

S. Han et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3034–3038
3035
O
GPR119 receptors. Ultimately, the 5-(methylsulfonyl)pyrazine
(12) was selected as the preferred aryloxy group for further SAR
investigation.
MeO2S
MeO2S
MeO2S
N
N
N
O
N
N
H
O
We next turned our SAR effort toward exploration of the piperi-
dine substitution (Table 2). Piperidine motifs incorporating carba-
mates (12, 16–25), amides (26–28), fluorinated alkyl groups (29–
33), and heteroaromatic rings (34–37) were prepared. These were
introduced in the final steps of the synthesis as illustrated in
Scheme 1. Trifluoromethyl analogues (21–23), in particular, dis-
played significantly improved potency relative to our clinical com-
pound (1) and were stable in human, rat, and mouse liver
microsomes. Amide 26 attained good potency and had a low shift
in the serum shift assay, but suffered from significantly reduced
microsomal stability in mouse. Tertiary amines (29–33) comprised
of fluorinated alkyls led to a reduction in GPR119 activity and
reduced microsomal stability (e.g., 30). Based on the combination
of in vitro potency and microsomal stability, a handful of com-
pounds were selected for further consideration. Compound 22 pos-
sessed a similar potency and PK profile relative to 21, however this
enantiomer was less effective in increasing glucose excursion in an
oGTT experiment performed in 129SVE mice. The exceedingly
potent hexafluoro analogue (23) was removed from further consid-
eration as it impaired physical mobility of rodents in a dose depen-
dent manner. Additionally, we were concerned that the activated O-
hexafluoroisopropyl (HFIP) carbamate may form a covalent bond
with serine hydrolases.11 Oxadiazole analogues 34–36 showed
reduced GIP release in normoglycemic 129SVE mice relative to 21,
despite their overall good potency, stability, and pharmacokinetic
properties. For the reasons described above, 21 selected for further
analysis.
CH3
OCH3
1
(APD597, partial Agonist)
O
N
N
N
O
N
O
O
CH3
F
2 (Full Agonist)
O
O
N
O
O
3 (trans/cis-1,3-dioxy)
(trans/cis-1,4-dioxy)
4
Figure 1. Design of a novel dioxycyclohexane GPR119 series.
Table 1
Aryloxy SAR
O
R1
O
R
N
O
O
Compd
R
R1
HTRF cAMP (nM)
c logPc
Starting from commercially available 1,4-dioxaspiro[4.5]decan-
8-ol, 21 was prepared by the synthetic sequence illustrated in
Scheme 1. Aromatic nucleophilic substitution of 1,4-dioxas-
piro[4.5]decan-8-ol, followed by an acid catalyzed deprotection
efficiently gave 39. A cis- enriched mixture of 40 (cis/trans = ꢀ3/2)
was obtained from the NaBH4 reduction, which was separated effi-
ciently to afford isomerically pure (>98%) cis and trans material on
>100 g scale. The separation was carried out by an initial crystal-
lization, followed by trituration with acetone. Stepwise reduction
of the pyridinium salt of 40 subsequently offered the key interme-
diate, trans-4-((1-methylpiperidin-4-yl)oxy)cyclohexanol (41). The
final intermediate 44 was straightforwardly prepared by nucle-
ophilic substitution, copper promoted sulfonylation, and de-
methylation. Intermediate 44 was then converted to 21 via a stan-
dard carbamate formation.
More extensive characterization entailed expanded in vitro
screening and pharmacokinetic analysis. Compound 21 was found
to maintain agonist activity at canine, monkey, and mouse GPR119
receptors (6.1 nM, 144%; 3.4 nM, 73%; 20.5 nM, 130%, respec-
tively). It was stable in liver microsomes across species (human,
rat, mouse half-life (t1/2) >60 min), and also demonstrated good
stability in human hepatocytes (t1/2 >120 min). It was highly bound
to both human and rat plasma proteins (human 97.5%, rat 96.0%),
but less so for mouse (93.5%). In a subsequent Sprague-Dawley
rat PK study, 21 exhibited a low systemic clearance and low
steady-state volume of distribution, while possessing excellent
systemic exposure, terminal half-life, and oral bioavailability
(Table 3). In an abbreviated CNS study, 21 demonstrated a brain
to plasma (b/p) ratio of approximately 1 (brain 820 ng/g, plasma
841 ng/mL) at 2 h post-dose, indicating virtually unrestricted
blood-brain barrier (BBB) penetration.
hGPR119
EC50 (Emax
rGPR119
EC50 (Emax
a
a
)
)
1
2
—
—
—
—
46.0 (75)
13.8 (94)
421.0 (89)
169.0 (93)
3.52
2.45
SO2Me
SO2Me
4
CH3 16.1 (115)
94.0 (120)
2.91
2.91
cis-4
CH3 266.0 (120)
1390.0 (121)
61.0 (119)
F
8.5 (113),
CH3
5
6
2.89
3.09
227.0b
SO2Me
F
CH3 52.0 (106)
167.0 (109)
SO2Me
CN
8.6 (116),
CH3
7
8
9
45.0 (106)
110.0 (108)
201.0 (125)
3.88
2.46
2.46
334.0b
N
CH3 47.0 (111)
SO2Me
SO2Me
N
H
H
13.0 (101)
11.0 (108)
N
10
11
12
89.0 (114)
400.0 (111)
64.0 (110)
2.96
2.84
2.07
SO2Me
N
CH3 262.0 (121)
CON(CH3)2
SO2Me
N
N
N
5.9 (110),
CH3
N
N
N
94.0b
SO2i-Pr
13
14
CH3 18.5 (109)
54.0 (128)
2.91
8.3 (110),
CH3
CN
28.4 (100)
2.19
1.21
144.0b
SO2Me
15
CH3 651.0 (97)
1160.0 (86)
N N
a
Emax are the mean of three or more replicates, and relative to our reference
standard, AR231453.
b
Serum shift assay (Ref. 10).
ChemBioDraw Ultra 12.0.
c
In vivo GPR119 activation and corresponding glucose control
was assessed for 21 in normoglycemic 129SVE mice after oral
administration (Figs. 2 and 3a). Since GIP release is expected upon
activation of GPR119, plasma GIP levels were measured 45 min
post compound administration (Fig. 2). Two different doses of 21
compounds prepared, 12 exhibited the smallest potency shift in
the presence of serum,10 correlating well with the order of c logP
(12 < 14 < 5 < 7). Notably, this new series was associated with a
very high intrinsic activity (Emax >100) at both human and rat

3036
S. Han et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3034–3038
N
N
N
N
OH
c
d, e, f
a
b
O
O
O
O
O
O
O
HO
HO
O
O
38
39
40
41
CF3
O
N
O
HN
N
N
O
O
O
g
j
h
i
O
O
N
O
O
O
N
N
N
N
N
N
N
SO2CH3
Br
SO2CH3
SO2CH3
42
43
44
21
Scheme 1. Synthesis of 21. Reagents and conditions: (a) 4-chloropyridine, t-BuONa, 94%, (b) HCl, 92%, (c) NaBH4, 36%, (d) MeI, (e) NaBH4, (f) H2, Pd/C, 80% (d–f), (g) 2,5-
dibromopyrazine, t-BuOK, 74%, (h) NaSO2Me, (CuOTf)2PhH, N,N-dimethylethane-1,2,-diamine, 70%, (i) 1-chloroethyl carbonochloridate, 90%, (j) (R)-1,1,1-trifluoropropan-2-
ol, CDI, 87%.
Table 2
Piperidine SAR
N
N
O
R
N
MeO2S
O
Compd
R
HTRF cAMP (nM)
LM stabilityb (h, r, m)
>60, >60, >60
hGPR119EC50 (Emax
)
rGPR119EC50 (Emax)
12
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
CO2t-Bu
CO2i-Pr
CO2(1-methylcyclopropan-1-yl)
CO2CH(Me)CH2F (R)
CO2CH(Me)CH2F (S)
CO2CH(CH2F)2
CO2CH(Me)CF3 (R)
CO2CH(Me)CF3 (S)
CO2CH(CF3)2
CO2(tetrahydrofuran-3-yl)
CO2Ph
5.9 (110)
64.0 (110)
119.0 (120)
138.0 (123)
317.0 (115)
356.0 (123)
467.0 (128)
28.8 (120)
49.9 (125)
10.9 (121)
1500.0 (109)
71.8 (104)
66.9 (126)
108.0 (124)
1380.0 (140)
1960.0 (105)
163.0 (124)
80.0 (99)
13.7 (110), 92.7a
25.9 (118)
36.8 (108)
31.1 (105)
65.1 (114)
>60, >60, >60
>60, >60, >60
>60, >60, >60
>60, >60, >60
3.0 (106), 38.0a
4.8 (112), 53.0a
1.2 (104), 45.2a
427.0 (105)
9.6 (116)
COCF2Et
5.3 (107), 28.7a
38.4 (95)
107.0 (107)
776.0 (91)
62.6 (116)
54.8 (96)
>60, >60, 3
21, >60, 37
CO(1-CF3)cyclopentan-1-yl
CO(1-CF3)cyclopropan-1-yl
CH2CF3
1-((1-CF3)cyclopropyl)methyl
1-((1-CF3)cyclobutyl)methyl
1-((1-CF3)cyclopentyl)methyl
1-(1-Fluorocyclopentyl)methyl
3-Isopropyl-1,2,4-oxadiazol-5-yl
5-Isopropyl-1,2,4-oxadiazol-3-yl
48.8 (37)
101.0 (88)
922.0 (119)
97.9 (115)
60.0 (108)
64.9 (114)
267.0 (114)
291.0 (105)
14.0 (102), 64.9a
9.1 (100), 56.3a
10.7 (103), 59.5a
30.3 (97), 276.0a
>60, >60, >60
>60, >60, >60
>60, 59, 56
3-(2-Fluoropropan-2-yl)-1,2,4-oxadiazol-5-yl
5-(2-Fluoropropan-2-yl)-1,2,4-oxadiazol-3-yl
>60, >60, >60
a
Serum shift assay (Ref. 10).
Half life (t1/2, min).
b
Table 3
Pharmacokinetics of 21
Compd
Male Sprague-Dawley rat PKa
b/p @ 2hb
0.98
Cl (L/h/kg)
0.40
Vss (L/kg)
t1/2 (h)
Cmax
(
lg/mL)
AUClast (h⁄lg/mL)
%F
21
2.94
6.45
0.77
5.21
100
a
2 mg/kg (IV)/2 mg/kg in 20% hp-b-CD (PO).
Male SD rats, 2 mg/kg in 20% hp-b-CD (PO).
b

S. Han et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3034–3038
3037
were administered and demonstrated a significant increase in GIP
levels relative to vehicle treated controls. The 30 mg/kg dose was
as efficacious as our tool compound (AR231453)12 in this study.
In an acute oral glucose tolerance test (oGTT), 21 was dosed orally
30 min prior to glucose challenge and blood glucose levels were
measured over 120 min time course (Fig. 3a). Notably, all three
doses produced an equivalent response as that observed for a
3 mg/kg dose of a representative DPP-4 inhibitor (Sitagliptin),
and glycemic suppression relative to vehicle was approximately
40% in an area under curve (AUC) analysis.
Acute glucose control was also evaluated in Zucker Diabetic
Fatty (ZDF) rats (Fig. 3b). Compound 21 was administered 60 min
prior to glucose challenge and demonstrated markedly reduced
blood glucose levels in a dose-dependent manner. It showed com-
parable efficacy to a 3 mg/kg dose of Sitagliptin when dosed at a 3
and 10 mg/kg dose (43% and 58% relative to 57% (Sitagliptin) in an
AUC analysis).
With encouraging acute in vivo data, we investigated the effect
of chronic treatment in Zucker Diabetic Fatty (ZDF) rats (Fig. 4).
Compound 21 was dosed once daily over 4 weeks, and fed glucose
and HbA1c levels were measured at the time of dosing. 21 demon-
strated a significant fed glucose lowering effect after 2 weeks treat-
ment (Fig. 4a) and markedly improved glycated hemoglobin
(HbA1c) levels at day 28 in a dose-dependent manner (1.63%,
1.85%, and 2.58% relative to vehicle) (Fig. 4b). Furthermore, no loss
of the acute drug response over the study period was observed,
thus validating chronic use of a selective GPR119 agonist for glu-
cose control. Lastly, we did not observe any significant drug effect
on body weight irrespective of dose.
Further characterization of 21 revealed no interaction at the
hERG channel in either the [3H]-Astemizole binding assay (IC50
>1000
microsomal cytochrome P450s, 21 exhibited no significant inhibi-
tion of any of the five major isoforms (2C9, 27.5 M; 2C19,
23.4 M; 1A2, 3A4, and 2D6, >50 M) and no inhibitory activity
lM) or patch clamp assay (IC50 >30 lM). In human hepatic
l
l
l
in CYP time-dependent inhibition (Kinact/Ki) assay. Glutathione
(GSH) conjugates were not detected in the GSH trapping assay,
indicating no apparent metabolite-mediated hepatotoxicity.
Further, 21 showed no cytotoxic potential in an Essential Cell
Function (ECF) panel measuring sensitive cellular parameters in
live HepG2 cells.
In summary, we have designed and optimized a new series of
trans-1,4-dioxycyclohexane GPR119 agonists. SAR studies led to
the discovery of the preferred molecule 21 that has potent and effi-
cacious GPR119 activity across species. This lead compound exhib-
ited an excellent ADME and safety profile, and demonstrated
robust regulatory effects on glycemic parameters in the acute
and chronic diabetic or non-diabetic rodent models. Subsequent
investigation of 21 and further development of this novel series
will be disclosed elsewhere.
600
400
200
0
3 mpk
Cmpd 21
PET
30 mpk
30 mpk
AR231453
Figure 2. Effects of 21 on GIP release in male 129SVE mice.
Figure 3. (a) Oral glucose tolerance test (oGTT) of 21 in male 129SVE mice. (b) oGTT in Zucker Diabetic Fatty (ZDF) rats.
(a)
(b)
500
400
300
200
100
0
12
10
8
0.5%MC
21 1 mpk QD
21 3 mpk QD
21 10 mpk QD
*
*
***
10
6
4
2
0
1
3
Cmpd 21 (mpk)
Figure 4. (a) Fed glucose control effect of chronic 28-days administration of 21 in Zucker Diabetic Fatty (ZDF) rats. (b) HbA1c levels in ZDF rats after 28 days.

3038
S. Han et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3034–3038
9. Buzard, D. J.; Kim, S.-H.; Lehmann, J.; Han, S.; Calderon, I.; Wong, A.; Kawasaki,
References and notes
A.; Narayanan, S.; Bhat, R.; Gharbaoui, T.; Lopez, L.; Yue, D.; Whelan, K.; Al-
Shamma, H.; Unett, D. J.; Shu, H.-H.; Tung, S.-F.; Chang, S.; Chuang, C.-F.;
Morgan, M.; Sadeque, A.; Chu, Z.-L.; Leonard, J. N.; Jones, R. M. Bioorg. Med.
Chem. Lett. 2014, 24, 4332.
1. (a) http://www.who.int/mediacentre/factsheets/fs312/en/.; (b) Zimmet, P. Z.
Medicographia 2011, 33, 15.
2. Chu, Z.-L.; Jones, R. M.; He, H.; Carroll, C.; Gutierrez, V.; Lucman, A.; Moloney,
M.; Gao, H.; Mondala, H.; Bagnol, D.; Unett, D.; Liang, Y.; Demarest, K.; Semple,
G.; Behan, D. P.; Leonard, J. Endocrinology 2007, 148, 2601.
3. Chu, Z.-L.; Carroll, C.; Alfonso, J.; Gutierrez, V.; He, H.; Lucman, A.; Pedraza, M.;
Mondala, H.; Gao, H.; Bagnol, D.; Chen, R.; Jones, R. M.; Behan, D. P.; Leonard, J.
Endocrinology 2008, 149, 2038.
10. HTRF cAMP assays (Cisbio) were performed following the manufacturer’s
instructions. For serum shift assays, cells were plated overnight in 384-well
plates. On the day of the assay, growth medium was aspirated from the plates
and replaced with fetal bovine serum (10
diluted in DMSO and then diluted in assay buffer (PBS with 200
0.1% BSA) to achieve a 2Â concentration containing 1% DMSO prior to addition
to assay plates (10 L).
l
L). Test compounds were serially
l
M IBMX &
4. Jia, X.; Brown, J. C.; Ma, P.; Pederson, R. A.; McIntosh, C. H. Am. J. Physiol. 1995,
268, E645.
l
11. Chang, J.-W.; Cognetta, A. B., III; Niphakis, M. J.; Cravatt, B. F. ACS Chem. Biol.
2013, 8(7), 1590.
12. Semple, G.; Fioravanti, B.; Pereira, G.; Calderon, I.; Uy, J.; Choi, K.; Xiong, Y.;
Ren, A.; Morgan, M.; Dave, V.; Thomsen, W.; Unett, D. J.; Xing, C.; Bossie, S.;
Carroll, C.; Chu, Z.-L.; Grottick, A. J.; Hauser, E. K.; Leonard, J. N.; Jones, R. M. J.
Med. Chem. 2008, 51, 5172.
5. Seino, Y.; Fukushima, M.; Yabe, D. J. Diabetes Investig. 2010, 1, 8.
6. Buzard, D. J.; Lehmann, J.; Han, S.; Jones, R. M. Pat. Pharm. Anal. 2012, 1, 285.
7. Semple, G.; Lehmann, J.; Wong, A.; Ren, A.; Bruce, M.; Shin, Y.-J.; Sage, C. R.;
Morgan, M.; Chen, W.-C.; Sebring, K.; Chu, Z.-L.; Leonard, J. N.; Al-Shamma, H.;
Grottick, A. J.; Du, F.; Liang, Y.; Demarest, K.; Jones, R. M. Bioorg. Med. Chem. Lett.
2012, 22, 1750.
8. Katz, L. B.; Gambale, J. J.; Rothenberg, P. L.; Vanapalli, S. R.; Vaccaro, N.; Xi, L.;
Polidori, D. C.; Vets, E.; Sarich, T. C.; Stein, P. P. Clin. Pharmacol. Ther. 2011, 90,
685.