Identification of novel and potent 2-amino benzamide derivatives as allosteric glucokinase activators

Article abstract of DOI:10.1016/j.bmcl.2009.01.053

The identification and structure-activity-relationships (SARs) of novel 2-amino benzamide glucokinase activators are described. Compounds in this series were developed to be potent GK activators, and their binding mode to the GK protein was determined by crystal structure analysis. In vivo pharmacokinetic and acute in vivo efficacy studies of compound 18 are also described.

Full text of DOI:10.1016/j.bmcl.2009.01.053

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1357–1360
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of novel and potent 2-amino benzamide derivatives
as allosteric glucokinase activators
*
Teruyuki Nishimura , Tomoharu Iino, Morihiro Mitsuya, Makoto Bamba, Hitomi Watanabe,
Daisuke Tsukahara, Kenji Kamata, Kaori Sasaki, Sumika Ohyama, Hideka Hosaka,
Mayumi Futamura, Yasufumi Nagata, Jun-ichi Eiki
Banyu Tsukuba Research Institute, Banyu Pharmaceutical Co. Ltd., Okubo-3, Tsukuba, 300-2611 Ibaraki, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The identification and structure–activity-relationships (SARs) of novel 2-amino benzamide glucokinase
activators are described. Compounds in this series were developed to be potent GK activators, and their
binding mode to the GK protein was determined by crystal structure analysis. In vivo pharmacokinetic
and acute in vivo efficacy studies of compound 18 are also described.
Received 25 November 2008
Revised 13 January 2009
Accepted 15 January 2009
Available online 21 January 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Glucokinase
Glucokinase activator
Diabetes
Glucokinase (GK), a member of the hexokinase family,1 cata-
lyzes the first step in glycolysis involving the phosphorylation of
glucose to glucose 6-phosphate. GK plays an important role as a
glucose sensor in maintaining plasma glucose homeostasis by
enhancing insulin secretion from pancreatic b-cells and glucose
metabolism in the liver.2,3 Therefore activation of GK is expected
to be a novel therapeutic strategy for the treatment of type II dia-
betes. Consistent with this rationale, several groups have reported
that the glucokinase activators (GKAs) demonstrated antidiabetic
efficacy in rodent models.4–7 Recently, AstraZeneca8,9 and OSI Pros-
idion10,11 have also defined the SARs of small molecule allosteric
activators of GK. In addition, a number of GKAs have been disclosed
in the patent literature and reviewed in peer edited journals.12–15
As part of our own effort, we have supplied detailed descriptions
of in vitro features and mechanisms of action of our GK activator
(compound 18), effects on GK and glucokinase regulatory protein
(GKRP) interactions, and the GK activator’s effect on pancreatic islets
and hepatocytes.16 In this paper, we describe the discovery and de-
tailed SAR studies of a novel series of potent allosteric GKAs built on
a 2-amino benzamide scaffold. We also discuss the SARs based on
the binding mode between the GK protein and our own activator
as revealed by crystal structure analysis.17
activation of recombinant human GK enzyme.18 In vitro GK assay
was conducted in two different glucose concentration, 2.5 and
10 mM which were simulated to be fasting and postprandial blood
glucose conditions, respectively.
At first, with regard to the core structure of the GKAs, both the
reduction of the amide linkage to benzylamine and substitution on
the amide NH were not tolerated (data not shown). These SARs
suggest that the benzamide template would account for the signif-
icant GK activation. Preliminary SAR studies on the screening hits
(1a and b) also indicated the importance of the substituent on 5-
position of the benzene ring, along with the presence of an NH2
group in the 2-position, in the activation of the target GK enzyme
(Table 1). The removal of a chlorine atom from the 5-position of the
Table 1
Investigation of preliminary SAR on benzamide lead 1
O
S
R4
R1
R2
N
H
N
R3
Compound
R1
R2
R3
R4
2.5 mM Glc
EC50
M)a
10 mM Glc
(
l
EC50 (l
M)a
A high throughput screen of our compound collection was car-
ried out in order to identify the small molecule GKA and the thia-
zole amide series (1a and b) were identified based on their
1a
1b
2
3
4
Cl
Cl
H
H
Cl
H
H
H
Cl
H
NH2
NH2
NH2
NH2
H
H
Me
H
Me
Me
11
2.6
1.4
>30
>30
6.3
6.5
>30
>30
17
* Corresponding author. Tel.: +81 29 877 2000; fax: +81 29 877 2029.
E-mail address: teruyuki_nishimura@merck.com (T. Nishimura).
a
EC50 values are the mean of at least two independent assays (Ref. 18).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.01.053

1358
T. Nishimura et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1357–1360
benzene ring (2), or the migration of the Cl atom to the 4-position
(3), resulted in loss of potency. Compound 4, containing no NH2
group on the benzene ring, showed reduced GK activation potency.
Initially, to explore the SARs in the series, the aniline moiety was
kept on the benzamide template and various substitutions were
introduced at the 5-position to probe the requirement for potency.
Figure 1 depicts the preparation of compounds using 5-fluoro-
2-nitrobenzoic acid as the starting material. Treatment of 5-flu-
ety of heteroaromatics with feasible physicochemical properties
and metabolic stability, we next examined the thioether deriva-
tives. These studies demonstrated that ethyl-thioether (12) was
more potent than the corresponding ether compound (5); on the
other hand, GK activity of phenylthioether (13) and 2-pyr-
idinylthioether derivatives (14) slightly reduced the potency com-
pared with the phenoxy compound (6). Although thio-imidazole
(15) and 1,2,4-triazole (17), both possessing an NH group in their
ring structure, significantly reduced the potency, the correspond-
ing N-methylated 5-membered ring systems (16 and 18) were
effective at improving the GK activation.
With these data in hand we elected to focus further on the N-
Me-triazolylthio compound 18 to achieve a balance between
intrinsic potency and physical properties such as aqueous solubil-
ity and metabolic stability. Table 3 depicts the SAR associated with
the aromatic ring on the amide portion. Among them, 2-thiazolyl
(18) and [1,2,4]-thiadiazol-5-yl (19 and 20) amides are very effec-
tive for potent GK activation and the 2-pyridine (21) ring system
was tolerated. In contrast, 3- and 4-pyridyl (22 and 23) and phenyl
(24) derivatives showed no activation of GK enzyme. Examination
of the aromatic ring systems on the amide suggested that the nitro-
gen atom at the 2-position on the aromatic ring is necessary for GK
activation.
oro-2-nitrobenzoic acid with oxalyl chloride and
a catalytic
amount of DMF afforded the corresponding benzoyl chloride. The
benzoyl chloride then reacted with the appropriate commercially
available aromatic amines to give benzamides. Reactions of 2-
nitrobenzamides with alkoxide, phenol or thiol nucleophiles
(K2CO3/DMF or TEA/CH3CN) provided the 5-substituted-2-nitro
benzamides. Subsequently the 2-amino benzamides were obtained
by reduction of the nitro compounds catalyzed by iron powder and
saturated aqueous NH4Cl.
During the course of our optimization studies concerning the 5-
position substituent on the benzene ring (Table 2), the presence of
hydrophobic groups such as phenoxy (6) afforded enhanced GK po-
tency, and introduction of the substitution on the ortho (7) posi-
tion was more effective for GK activation than substitutions on
both the meta (8) and para (9) positions. Introduction of a polar
group such as methoxy (10) and methylsulfonyl (11) onto the
ortho position was also tolerated. In order to introduce a wide vari-
Next we examined substitutions on the thiazole ring in the
compound (Table 4). A methyl substitution on both 4- and 5-posi-
tion on the thiazole moiety showed no significant effects on their
intrinsic potencies. Results regarding substituents on the 4-posi-
tion of the thiazole ring in the compound, electron withdrawing
groups such as CO2Et (27), CF3 (29) and hydrophilic group like
CH2OH (28) showed a significant reduction of potency in compar-
ison with compounds 25 and 18.
Table 3
SAR around aromatic ring on the amide moiety
Figure 1. Reagents and conditions: (a) (COCl)2, DMF (cat.), CH2Cl2 followed by Aryl-
NH2, TEA, CH3Cl; (b) EtONa or EtSNa, THF; (c) Phenol or Thiol, K2CO3/DMF or TEA/
CH3CN; (d) Fe powder, sat-NH4Cl, i-PrOH.
O
Ar
S
N
N
N
H
N
NH2
Table 2
Effects of substitution on 5-position of the benzene ring
Compound
Ar
2.5 mM Glc
EC50
M)a
10 mM Glc
EC50
(
l
(l
M)a
O
S
S
X
18
19
20
21
22
23
0.42
0.49
0.64
1.2
0.14
R1
N
N
H
N
NH2
N
N
S
S
0.12
0.18
0.23
>10
>10
>10
Compound R1
X
2.5 mM Glc EC50
M)a
10 mM Glc EC50
(l
M)a
(
l
N
N
1b
5
Cl
Et
–
O
O
O
O
O
O
O
S
S
S
S
S
6.5
6.8
1.4
1.7
6
C6H5
0.70
0.26
0.60
1.7
0.41
0.51
0.78
0.92
1.2
0.13
0.10
0.14
0.40
0.12
0.13
0.18
0.20
0.22
0.38
0.10
7
8
9
10
11
12
13
14
15
16
2-F-C6H4
3-F-C6H4
4-F-C6H4
2-MeO-C6H4
N
2-MeSO2-C6H4
Et
>10
>10
>10
N
N
C6H5
2-Pyridine
1H-Imidazol-2-yl
1.6
0.23
1-Me-1H-Imidazol-2-
yl
4H-[1,2,4]-Triazol-3-yl
4-Me-4H-[1,2,4]-
Triazol-3-yl
17
18
S
S
2.4
0.42
0.54
0.14
24
a
a
EC50 values are the mean of at least two independent assays (Ref. 18).
EC50 values are the mean of at least two independent assays (Ref. 18).

T. Nishimura et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1357–1360
1359
Table 4
the small and large domains approximately 20 Å from and opposite
the glucose binding site. The crystal data suggested that there are
important H-bond and hydrophobic interactions involved in bind-
ing to the GK protein and inducing GK activation. The compound
makes three hydrogen bonds with the enzyme: both the NH and
nitrogen atom in the amino-thiazole moiety pair with the back-
bone of Arg63, and the NH on aniline pairs with Tyr215. In addi-
tion, the aniline moiety on the benzene ring may works as an
important functional group which making an intramolecular H-
bond interaction with the carbonyl group on its amide linkage,
thus fixing the conformation in a beneficial way. The N-Me-triazole
moiety on the benzene ring occupies the hydrophobic space on the
allosteric binding site.
The X-ray data and revealed binding mode correlate well with
our SARs which were developed during the modification of this
structure class. In order to bind to the GK enzyme and show potent
GK activation, an amino-heteroaromatic amide core possessing the
nitrogen atom at the 2-position of the heteroaromatic ring is crit-
ical. Furthermore, because the aniline moiety works as both an in-
tra- and intermolecular hydrogen bonding acceptor, introduction
of a substitution on the aniline portion and removal of aniline moi-
ety resulted in a loss of potency.
Effects of substitution on the thiazole ring
R5
R4
O
S
S
N
N
N
H
N
N
NH2
Compound
R5
R4
2.5 mM Glc
EC50
10 mM Glc
(l
M)a
EC50 (l
M)a
25
26
18
27
28
29
H
Me
H
H
H
H
H
Me
CO2Et
CH2OH
CF3
0.35
0.33
0.42
1.1
1.6
2.7
0.13
0.11
0.14
0.42
0.42
0.93
H
a
EC50 values are the mean of at least two independent assays (Ref. 18).
We finally turned our attention to the SARs of the aniline moi-
ety on the core benzene ring (Table 5). However, replacement of
NH2 (18) by hydrogen (30) or NO2 group (31) resulted in a loss
of potency. The introduction of a methyl group (32) on the aniline
portion reduced GK potency.
In parallel with our SAR studies, the co-crystal structure of GK
protein with allosteric activators was solved and the binding mode
of our own GKA series was revealed (Fig. 2).19,20 Compound 18
binds to an allosteric binding site on the hinge region between
These SAR studies allowed the identification of several potent
activators of human glucokinase. In addition, an activator which
provided an acceptable pharmacokinetic and metabolic profile
was identified and evaluated in animals. We describe here the
pharmacokinetic profile and glucose-lowering efficacy of com-
pound 18 in rats. During an in vivo pharmacokinetic study of
male Wistar rats, compound 18 demonstrated excellent bioavail-
ability (Table 6). In a Wistar rat Oral Glucose Tolerance Test
Table 5
SAR of the aniline moiety of compound 18
Table 6
O
S
Pharmacokinetic parameter of compound 18 in male Wistar rats
S
N
N
N
N
H
N
R3
O
S
S
N
N
N
N
H
Compound
R3
2.5 mM Glc
10 mM Glc
N
EC50
(l
M)a
EC50 (l
M)a
NH2
18
30
31
32
NH2
H
NO2
0.42
7.3
>10
1.1
0.14
1.5
>10
0.33
Distribution volume
(L/kg)
Clearance (mL/
min/kg)
t1/2 (iv)
(h)
Cmax
M)
tmax
(h)
Fpo
(%)
(
l
NHMe
0.4
14.3
0.9
8.5
0.5
ꢀ100
a
EC50 values are the mean of at least two independent assays (Ref. 18).
Doses: 1 mg/kg (iv) and 3 mg/kg (po).
Figure 2. Overall structure of GK protein with compound 18 and binding mode of compound 18; Final 2Fo À Fc electron density (1.2
r level) for compound 18 is shown as
orange mesh. This figure was prepared using the CCP4 package21 and PyMOL22

1360
T. Nishimura et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1357–1360
Figure 3. In vivo efficacy data of compound 18 in male Wistar rat OGTT23. (a) Plasma glucose lowering and (b) AUC reduction. ‘Ã’ Values of p < 0.05 were considered
statistically significant for vehicle group.
M.; Reynet, C.; Schofield, K. L.; Shah, V. K.; Spindler, F.; Taylor, A.; Turton, R.;
Williams, G. M.; Wong-Kai-In, P.; Yasuda, K. J. Med. Chem. 2008, 51, 4340.
12. Sarabu, R.; Grimsby, J. Curr. Opin. Drug Discov. Dev. 2005, 8, 631.
(OGTT), oral administration of compound 18 demonstrated glu-
cose-lowering efficacy in a dose-dependent manner, and showed
significant reduction of plasma glucose levels at 10 and 30 mg/kg
(Fig. 3).23
In conclusion, we have identified a series of 2-amino benzamides
as small molecule allosteric activators of glucokinase. The binding
mode of these molecules was revealed and detailed SAR on GK acti-
vation was developed and analyzed. The triazole compound 18 was
developed and has demonstrated promising in vivo glucose-lower-
ingeffectsinanacuteratmodel. FurthermodificationsandSARstud-
ies based on the crystal structure analyses are subjects for future
investigation and will be reported in due course.
13. Guertin, K. R.; Grimsby, J. Curr. Med. Chem. 2006, 13, 1839.
14. Coghlan, M.; Leighton, B. Expert Opin. Investig. Drugs 2008, 17, 145.
15. Sarabu, R.; Berthel, S. J.; Kester, R. F.; Tilley, J. W. Expert Opin. Ther. Patents 2008,
18, 759.
16. Futamura, M.; Hosaka, H.; Kadotani, A.; Shimazaki, H.; Sasaki, K.; Ohyama, S.;
Nishimura, T.; Eiki, J.; Nagata, Y. J. Biol. Chem. 2006, 281, 37668.
17. Kamata, K.; Mitsuya, M.; Nishimura, T.; Eiki, J.; Nagata, Y. Structure 2004, 12,
429.
18. Glucokinase activities were measured by the glucose-6-phosphate
dehydrogenase coupled continuous spectrophotometric assay. EC50 values
were designated as concentration of compounds which give half of maximal
increased enzymatic activities of glucokinase by the compound at 2.5 and
10 mM glucose. Compound 18 was used an internal control across all assay
plates for data validation. The EC50 values of this compound are 0.42 0.09 and
0.14 0.04 lM at 2.5 and 10 mM glucose, respectively. All compounds in this
References and notes
series investigated exhibited equivalent Emax values to the maximal response
evoked by compound 18 (8.0- to 8.3-fold and 1.6- to 1.8-fold over control
levels at 2.5 and 10 mM glucose). Detailed assay methods and conditions are
described in Ref. 16.
1. Printz, R. L.; Magnuson, M. A.; Granner, D. K. Annu. Rev. Nutr. 1993, 13, 463.
2. Matschinsky, F. M. Diabetes 1996, 45, 223.
19. Protein and crystals were obtained according to established procedures.17
Crystals were soaked in 0.5 mM compound 18 overnight in mother liquor
containing 5% DMSO. Diffraction data were collected on beamline BL32B2 at
the SPring-8, at 100 K (a Rigaku R-axis V image plate). Data processing and data
reduction were carried out using programs from the HKL2000 (HKL Research,
Inc.) and the CCP4 package. Compound 18 was modeled into the electron
density using Afitt (OpenEye Scientific Software). The protein–compound
complex model was refined using CNX (Accelrys), and the final structure has
been deposited in the Protein Data Bank with the deposition code 3FR0
together with structure factors and detailed experimental conditions.
20. Crystallographic statistics for the GK-compound 18 complex are as follows:
space group P6522, unit cell 84.1, 84.1, 323.2 Å, resolution 2.70 Å, 17,901
reflections from 361,813 observations give 99.8% completeness with Rsym of
3. Matschinsky, F. M.; Glaser, B.; Magnuson, M. A. Diabetes 1998, 47, 307.
4. Grimsby, J.; Sarabu, R.; Corbett, W. L.; Haynes, N. E.; Bizzarro, F. T.; Coffey, J. W.;
Guertin, K. R.; Hilliard, D. W.; Kester, R. F.; Mahaney, P. E.; Marcus, L.; Qi, L. D.;
Spence, C. L.; Tengi, J.; Magnuson, M. A.; Chu, C. A.; Dvorozniak, M. T.;
Matschinsky, F. M.; Grippo, J. F. Science 2003, 301, 370.
5. Efanov, A. M.; Barrett, D. G.; Brenner, M. B.; Briggs, S. L.; Delaunois, A.; Durbin, J.
D.; Giese, U.; Guo, H. H.; Radloff, M.; Gil, G. S.; Sewing, S.; Wang, Y.; Weichert,
A.; Zaliani, A.; Gromada, J. Endocrinology 2005, 146, 3696.
6. Coope, G. J.; Atkinson, A. M.; Allott, C.; McKerrecher, D.; Johnstone, C.; Pike, K.
G.; Holme, P. C.; Vertigan, H.; Gill, D.; Coghlan, M. P.; Leighton, B. Br. J.
Pharmacol. 2006, 149, 328.
7. Fyfe, M. C. T.; White, J. R.; Taylor, A.; Chatfield, R.; Wargent, E.; Printz, R. L.;
Sulpice, T.; McCormack, J. G.; Procter, M. J.; Reynet, C.; Widdowson, P. S.;
Wong-Kai-In, P. Diabetologia 2007, 50, 1277.
6.2% and mean I/r(I) of 9.2. The final model containing 3506 protein, 135
8. McKerrecher, D.; Allen, J. V.; Bowker, S. S.; Boyd, S.; Caulkett, P. W. R.; Currie, G.
S.; Davies, C. D.; Fenwick, M. L.; Gaskin, H.; Grange, E.; Hargreaves, R. B.; Hayter,
B. R.; James, R.; Johnson, K. M.; Johnstone, C.; Jones, C. D.; Lackie, S.; Rayner, J.
W.; Walker, R. P. Bioorg. Med. Chem. Lett. 2005, 15, 2103.
water, 1 salt, 12 glucose and 26 compound atoms has an R-factor of 22.1% (Rfree
using 5% of the data 27.5%). Mean temperature factors for the protein and the
ligand are 44.4 and 33.3 Å2, respectively.
21. Collaborative. Acta Crystallogr. D 1994, 50, 760.
9. McKerrecher, D.; Allen, J. V.; Caulkett, P. W. R.; Donald, C. S.; Fenwick, M. L.;
Grange, E.; Johnson, K. M.; Johnstone, C.; Jones, C. D.; Pike, K. G.; Rayner, J. W.;
Walker, R. P. Bioorg. Med. Chem. Lett. 2006, 16, 2705.
10. Castelhano, A. L.; Dong, H. Q.; Fyfe, M. C. T.; Gardner, L. S.; Kamikozawa, Y.;
Kurabayashi, S.; Nawano, M.; Ohashi, R.; Procter, M. J.; Qiu, L.; Rasamison, C.
M.; Schofield, K. L.; Shah, V. K.; Ueta, K.; Williams, G. M.; Witter, D.; Yasuda, K.
Bioorg. Med. Chem. Lett. 2005, 15, 1501.
22. PyMOL molecular graphics system, DeLano Scientific, San Carlos, CA. Available
from: <http://www.pymol.org>.
23. Oral glucose tolerance test (OGTT); Nine-week-old male Wistar rats fasted
overnight before performing the test. The rats were orally administered
compound 18 or vehicle alone (0.5% methylcellulose solution) followed
30 min later by an oral glucose challenge (2 g/kg). Plasma glucose
concentrations were measured just prior to and following the glucose
challenge (30, 60, 90, and 120 min). AUC values were calculated from the
data (from À30 to 120 min).
11. Bertram, L. S.; Black, D.; Briner, P. H.; Chatfield, R.; Cooke, A.; Fyfe, M. C. T.;
Murray, P. J.; Naud, F.; Nawano, M.; Procter, M. J.; Rakipovski, G.; Rasamison, C.