2,3-Disubstituted acrylamides as potent glucokinase activators

Article abstract of DOI:10.1016/j.bmcl.2010.08.029

The phenylacetamide 1 represents the archtypical glucokinase activator (GKA) in which only the R-isomer is active. In order to probe whether the chiral center could be replaced, we prepared a series of olefins 2 and show in the present work that these com

Full text of DOI:10.1016/j.bmcl.2010.08.029

Bioorganic & Medicinal Chemistry Letters 20 (2010) 5673–5676
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
2,3-Disubstituted acrylamides as potent glucokinase activators
b
a
a
b
Achyutharao Sidduri a, , Joseph S. Grimsby , Wendy L. Corbett , Ramakanth Sarabu , Joseph F. Grippo ,
Jianping Lou a, Robert F. Kester a, Mark Dvorozniak b, Linda Marcus b, Cheryl Spence b, Jagdish K. Racha c,
David J. Moore c
*
a Department of Discovery Chemistry, Roche Research Center, 340 Kingsland Street, Nutley, NJ 07110, USA
b Department of Metabolic Diseases, Roche Research Center, 340 Kingsland Street, Nutley, NJ 07110, USA
c Department of eADME Non-Clinical Safety, Roche Research Center, 340 Kingsland Street, Nutley, NJ 07110, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The phenylacetamide 1 represents the archtypical glucokinase activator (GKA) in which only the R-iso-
mer is active. In order to probe whether the chiral center could be replaced, we prepared a series of ole-
fins 2 and show in the present work that these compounds represent a new class of GKAs. Surprisingly,
the SAR of the new series paralleled that of the saturated derivatives with the exception that there was
greater tolerance for larger alkyl and cycloalkyl groups at R2 region in comparison to the phenylaceta-
mides. In normal Wistar rats, the 2,3-disubstituted acrylamide analog 10 was well absorbed and demon-
strated robust glucose lowering effects.
Received 28 May 2010
Revised 4 August 2010
Accepted 5 August 2010
Available online 11 August 2010
Keywords:
Discovery of novel GK activators for the
treatment of diabetes
Ó 2010 Elsevier Ltd. All rights reserved.
Type 2 diabetes affects over 150 million people worldwide.1 It is
associated with decreased function of pancreatic b-cells, decreased
insulin action and excess hepatic glucose production.2 Glucokinase
(GK), a member of the hexokinase family, acts as a b-cell glucose
sensor and couples glucose metabolism to insulin release. GK also
exerts control over hepatic glucose metabolism by influencing glu-
cose uptake and production.3,4 As such, we hypothesized that GK
activators could represent a novel and promising approach for
the treatment of type 2 diabetes.5,6 Roche scientists discovered
the GK activator 1 and showed that it lowers blood glucose levels
in rodents and humans.7,8
R2
H
N
H
R3
N
N
O
R
S
O
S
R1
O
O
1
2
SC1.5 = 0.127 uM
Figure 1.
In order to understand the importance of the chiral center in 1
(Fig. 1), we report here an investigation of the effects of replacing it
with a double bond resulting in the discovery of 2,3-disubstituted
acrylamide GKA series 2. These analogues were proved to be po-
tent GK activators. We also describe a robust 4-step synthesis of
the 2,3-disubstituted acrylamide analogues that allowed for the ra-
pid exploration of the SAR and report the pharmacokinetic proper-
ties and potent glucose lowering activity of 10, a prototypical
member of this series.
The preparation of this series is outlined in Scheme 1. Thus, the
trans-olefin 4 was prepared by taking advantage of a highly regio-
and stereoselective 1,4-addition of an alkyl organocopper species9
to methyl propiolate to obtain, in turn a vinylcopper intermediate,
that upon iodinolysis produced the pure trisubstituted olefin 4
where the R2 and the iodide are cis to each other. The required
organocopper reagent could be generated in situ either from an
organozinc10 or organomagnesium9 intermediate. The intermedi-
ate 4 was treated with activated zinc dust and the resulting vinyl-
zinc intermediate was coupled with a substituted bromo or
iodobenzene in the presence of Pd(dba)2 and triphenylphosphine
to give the desired styrene 5 with retention of the E-configura-
tion.11 Saponification of the methyl ester was achieved with 1.0N
aqueous sodium hydroxide in ethanol. Finally, the R3 moiety was
introduced via an in situ formed acyl bromide intermediate using
N-bromosuccinimide and triphenylphosphine at 0 °C for 1.5 h fol-
lowed by treatment with R3–NH2 at 0 °C to room temperature to
obtain the target compounds 2.
The tetrazole substituted bromobenzene and iodobenzene
starting materials were prepared in two steps following a literature
procedure as shown in Scheme 2.12
The compounds were assayed as previously described.7,8 Briefly,
the assay format couples the glucokinase mediated formation of
* Corresponding author. Tel.: +1 973 235 7143: fax: +1 973 235 7122.
E-mail address: achyutharao.sidduri@roche.com (A. Sidduri).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.08.029

5674
A. Sidduri et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5673–5676
R2
I
cyclopentyl analog 10. A similar modification in the saturated ser-
ies led to a complete loss of activity.8 In our experience, the ready
accommodation of the R2 region of larger and homologated cyclo-
alkyl groups is unique to this olefinic class of GK activators.
In order to investigate the stereochemical preference of these
olefins and the impact of further subsitution, we prepared the Z-
isomer and tetra-substituted derivatives. The Z-regioisomer 15 is
ꢀ20-fold less potent than the corresponding E-regioisomer 10
and the tetra-substituted olefin 16 was inactive (Fig. 2), possibly
due to conformational effects leading to inhibition of essential
hydrogen bonding between N–H and the protein.13
a, b, c, d
e, f
R2
X
O
O
4
3
R2
R2
H
O
g, h
N
R3
O
O
R
Further improvement in potency was achieved by optimizing
the heteroaromatic ring (R3) and the R- and R1-substituents on
the benzene ring while maintaining the cyclopentyl ring at R2
(Table 2). The SAR at the R3 region in the current series follows very
similar trends to the saturated series. Several smaller substituents
such as bromide, chloride, methyl, and cyano were acceptable at
the 5-position of the thiazol-2-yl ring. Of these, the chloro 17
and bromo 18 derivatives gave a >5-fold improvement in potency.
Other heterocycles such as pyridin-2-yl were also tolerated (21 vs
22), but with a small reduction in potency. Replacement of thiazol-
2-yl ring with an open chain N-methylacetamide resulted a nine-
fold less potent compound (25 vs 26) which is in parallel with re-
sults obtained in the saturated series.8
Also consistent with our experience with the saturated ana-
logues,8 the benzene ring accommodates a variety of electron
withdrawing groups at the 3- and 4-positions. Some of the 3,4-
disubstituted analogues were somewhat more potent than the 4-
monosubstituted compound 10. The methanesulfonyl moiety at
the 4-position can be substituted by a 5-methyl tetrazol-1-yl group
(22 vs 27) without loss of potency. On the other hand, substitution
with chloride reduced activity fivefold (22 vs 24).
R
R1
R1
5
2
Scheme 1. Reagents and conditions: (a) X = I, Zn dust, 1,2-dibromoethane, TMSCl,
THF, rt, 2–5 h, or X = Br, Mg turnings, THF, rt, 4–5 h; (b) CuCNÁ2LiCl, THF, À70 °C to
À10 °C, 10 min; (c) methyl propiolate, THF, À70 °C to À30 °C, 3 h; (d) I2, THF, À70 °C
to rt, 1 h; (e) Zn dust, 1,2-dibromoethane, TMSCl, THF, rt, 3–4 h; (f) substituted
bromobenzene or iodobenzene, Pd(dba)2, PPh3, THF, 40–45 °C, 15 h; (g) 1.0 N aq
NaOH, EtOH, 40 °C, 4–5 h; (h) NBS, PPh3, CH2Cl2, 0 °C, 1.5 h, and then 2-amino-R3,
0 °C to rt, 15 h.
X
X
Y
a, b
N
H2N
N
N
R1
R1
N
6
7
R1 = F or Cl
X = Br or I
Y = CH3 or CF3
Scheme 2. Reagents and conditions: (a) Acetic anhydride or trifluoroacetic
anhydride, THF, 0 °C to rt, 15 h; (b) NaN3, acetonitrile, CH2Cl2, trifluoromethane-
sulfonic anhydride, 0 °C to rt, 15 h.
Combination of the optimal substituents at each position R, R1,
R2, and R3 led to the very potent analogues described in Table 3.
Several GK activators incorporating cyclopentylmethyl, cyclohexyl,
and cycloheptyl moieties had SC1.5 values in the single digit (31
and 35) to low double digit nanomolar range (29, 32, 36 and 37).
The positive impact of addition of strongly electron withdrawing
trifluoromethyl- (28–31) and nitro- (32) groups at the R1 position
of the aromatic ring are consistent with parallel observations in
saturated series.14 We also note that the replacement of a 5-methyl
by a 5-trifluoromethyl group on the tetrazol-1-yl ring caused a
fourfold reduction in potency (33 vs 34).
glucose-6-phosphate to NADH which is measured spectrophoto-
metrically. Concentration response curves were constructed and
the concentration of drug molecule causing a 50% increase (SC1.5
)
in GK activity, relative to baseline, was calculated.
As anticipated from our experience with the saturated deriva-
tives,8 a significant improvement in potency was observed through
the replacement of the lower alkyl moieties at R2 with cycloalkyl
groups (Table 1). Interestingly, the cyclopentyl analog was twofold
more potent than that of the corresponding saturated compound
(1 vs 10). In contrast to the SAR of the saturated derivatives, the
cyclohexyl 11 and cycloheptyl 12 were more potent than the
cyclopentyl derivative 10. As a further surprise, the homologated
cyclopentylmethyl analog 14 retained equivalent activity to the
H
N
H
N
N
N
O
S
O
S
S
S
Table 1
O
O
O
O
SAR summary of R2 modifications
1
10
R2
SC 1.5 = 0.127 µM
SC 1.5 = 0.06 µM
H
N
N
O
S
CH3
S
H
N
O
O
H
N
N
N
Compd
R2
SC1.5 (lM)
O
S
8
9
Ethyl
1.2
0.45
0.06
0.029
0.021
0.073
0.053
O
S
S
S
Isopropyl
Cyclopentyl
Cyclohexyl
Cycloheptyl
Cyclooctyl
O
O
O
O
10
11
12
13
14
16
15
SC 1.5 = Inactive
SC 1.5 = 1.18 µM
Cyclopentylmethyl
Figure 2. Comparison of E-, Z- and tetra-substituted olefins.

A. Sidduri et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5673–5676
5675
Table 2
One concern with this class of compounds was the potential for-
mation of a GSH adduct by a 1,4-Michael addition of GSH to the
SAR summary of R, R1, and R3 modifications
a,b-unsaturated acrylamide. To understand whether this possibil-
ity constituted a liability for this series, compound 10 was labeled
with 14C at the carbonyl carbon and the resulting analog was incu-
bated with primary cryopreserved hepatocytes from rat, dog,
monkey, and human for 30 min. During this study, six major
metabolites were identified (Fig. 3). Four of them are the products
of mono-oxidation at the 2- and 3-positions of the cyclopentane
ring (M1 and M2) and their corresponding ketones (M3 and M4).
The other two metabolites were identified as the thiourea deriva-
tives (M5 and M6) resulting from oxidative cleavage of the thiazole
ring of the M1 and M2 metabolites. Basically, the total metabolism
varies widely from species to species as shown in Table 4, although
the major metabolites form in all species tested were M5 and M6.
Interestingly, we found no trace of a GSH adduct formed during
this study.
H
N
R3
O
R
R1
Compd
R
R1
R3
SC1.5 (lM)
17
18
19
20
21
22
23
24
SO2Me
SO2Me
SO2Me
SO2Me
SO2Me
SO2Me
SO2Me
Cl
H
H
H
H
5-Chloro-thiazol-2-yl
5-Bromo-thiazol-2-yl
5-Methyl-thiazol-2-yl
5-Cyano-thiazol-2-yl
Pyridin-2-yl
Thiazol-2-yl
Thiazol-2-yl
Thiazol-2-yl
0.012
0.008
0.049
0.097
0.096
0.045
0.033
0.215
Cl
Cl
Br
Cl
Based on the in vitro and metabolic stability data, compound 10
was selected for pharmacokinetic and in vivo efficacy studies in
rats to evaluate the potential of this class of compounds.
H3C
25
26
27
F
Thiazol-2-yl
0.102
0.904
0.04
N
N
As shown in Table 5 and 10 was completely absorbed and dis-
played a high clearance and moderately high volume of distribu-
tion resulting in an elimination half-life of 1.7 h in rats (Table 5).
The oral bioavailability (F%) in mice (10 mg/kg), dogs (10 mg/kg),
and monkey (100 mg/kg) was 58%, 88%, and 37%, respectively.
In normal Wistar rats, 10 caused a robust dose-dependent
reduction in blood glucose levels after oral doses of 3–30 mg/kg.
At the higher dose, the effects persisted more than 8 h (Fig. 4).
N
N
H3C
H
N
CH3
F
N
O
N
N
N
H3C
Cl
Thiazol-2-yl
N
N
N
N
Table 3
SAR summary of R, R1, R2, and R3 modifications
H
N
N
R2
O
S
H
N
S
10
R3
O
O
O
R
R1
OH
Compd
R
R1
R2
R3
SC1.5 (lM)
28
29
30
31
32
SO2Me
SO2Me
SO2Me
SO2Me
SO2Me
H3C
CF3
CF3
CF3
CF3
cC5H9CH2 Thiazol-2-yl
0.037
0.010
H
cC6H11
cC6H11
cC6H11
Thiazol-2-yl
5-Bromo-pyridin-2-yl 0.036
5-Bromo-thiazol-2-yl 0.004
N
N
O
S
NO2 cC6H11
Thiazol-2-yl
0.013
S
O
O
M1 and M2
OH
O
33
Cl
Cl
cC6H11
Thiazol-2-yl
0.025
N
N
N
N
F3C
H
N
H
N
NH2
N
34
cC6H11
Thiazol-2-yl
0.092
N
N
O
S
O
S
N
S
S
N
O
O
O
O
M5 and M6
35
36
SO2Me
SO2Me
H3C
H
CF3
cC7H13
cC7H13
5-Bromo-thiazol-2-yl 0.004
5-Bromo-thiazol-2-yl 0.011
M3 and M4
Figure 3. Metabolites of compound 10.
37
Cl
cC7H13
5-Bromo-thiazol-2-yl 0.012
N
N
N
N
Table 4
Metabolites profile of 14C radiolabelled compound 10 at 30
l
M concentration
Primary cryopreserved
hepatocytes
M1 and
M2
M3 and
M4
M5 and
M6
Total %
metabolism
Based on these results, we believe that the working model pro-
posed previously for chiral compounds such as 1 would be applica-
ble: donor–acceptor hydrogen bonding from the 2-amino
heteroaromatic ring, a hydrophobic cyclic aliphatic R2 and an aro-
matic ring which tolerates a range of electron withdrawing groups
Rat
Dog
Monkey
Human
8.0
0.6
37.1
13.7
2.9
0.0
12.3
7.3
5.2
0.2
16.5
15.6
16.1
0.8
65.9
36.6
8
in the 3- and 4-positions.

5676
A. Sidduri et al. / Bioorg. Med. Chem. Lett. 20 (2010) 5673–5676
Table 5
ylacetamide derivative 1. This modification resulted in the identi-
fication of a number of potent GK activators with good exposure
in a variety of species and robust efficacy in lowering glucose levels
in vivo.
Pharmacokinetic parameters of compound 10 in rats following intravenous and oral
administration
AUC0-24
(ng h)
T1/2
(h)
CL
Vdss
(mL/kg)
F (%)
(mL/h/kg)
iv (5 mg/kg)
3150
1.7
1543
2643
po (30 mg/kg)
23998
100
Acknowledgment
I would like to thank Dr. Jefferson Tilley for his advice and assis-
tance in writing this Letter.
80
60
40
20
References and notes
*
*
*
*
1. Wild, S.; Roglic, G.; Green, A.; Sicree, R.; King, H. Diabetes Care 2004, 27, 1047.
2. Ramol-Halsted, B. A.; Edelman, S. V. Primary Care 1999, 26, 771.
3. Matschinsky, F. M. Diabetes 1996, 45, 223.
*
*
*
4. Matschinsky, F. M. Nat. Rev. Drug Disc. 2009, 8, 399.
*
5. Sarabu, R.; Grimsby, J. Curr. Opin. Drug Discovery Dev. 2005, 8, 631.
6. Guertin, K. R.; Grimsby, J. Curr. Med. Chem. 2006, 13, 1839.
7. Grimsby, J.; Sarabu, R.; Corbett, W. L.; Haynes, N. E.; Bizzaro, F. T.; Coffey, J. W.;
Guertin, K. R.; Hilliard, D. H.; Kester, R. F. K.; Mahaney, P. E.; Marcus, L.; Qi, L.;
Spence, C. L.; Tengi, J.; Magnuson, M. A.; Chu, C. A.; Dvorozniak, M. T.;
Matschinsky, F. M.; Grippo, J. F. Science 2003, 301, 370.
*
*
*
*
Vehicle
#10 (30 mg/kg)
#10 (10 mg/kg)
#10 (3 mg/kg)
8. Haynes, N. E.; Corbett, W. L.; Bizzaro, F. T.; Coffey, J. W.; Guertin, K. R.; Hilliard,
D. H.; Holland, G.; Kester, R. F. K.; Mahaney, P. E.; Marcus, L.; Qi, L.; Spence, C. L.;
Tengi, J.; Magnuson, M. A.; Chu, C. A.; Dvorozniak, M. T.; Matschinsky, F. M.;
Grippo, J. F.; Grimsby, J.; Sarabu, R. J. Med. Chem. 2010, 53, 3618.
9. Normant, J. F.; Alexakis, A. Synthesis 1981, 841.
0
-60
60
180
300
420
Time (min)
Rats (n=5) were fasted for two hours prior to dose.
*, p < 0.05 compared to Vehicle using a Student’s -test.
10. Knochel, P.; Singer, R. D. Chem. Rev. 1993, 93, 2117.
11. Knochel, P.; Janikaram Rao, C. Tetrahedron 1993, 49, 29.
t
12. Thomas, E. W. Synthesis 1993, 767.
Figure 4. Efficacy with compound 10 in Wistar rats.
13. Dunten, P.; Swain, A.; Kammlott, U.; Crowther, R.; Lukacs, C. M.; Levin, W.;
Reik, L.; Grimsby, J.; Corbett, W. L.; Magnuson, M. A.; Matschinsky, F. M. In
Glucokinase and Glycimic Disease-From Basics to Therapeutics; Matschinsky, F.
M., Magnuson, M. A., Eds.; Karger AG: Basel, Switzerland, 2004; pp 145–154.
14. Sarabu, R.; Tilley, J.; Grimsby, J. Acc. Drug Discovery 2011, 51.
In conclusion, we prepared a novel class of 2,3-disubstituted
acrylamide GK activators as achiral derivatives of the chiral phen-