C-8 Modifications of 3-alkyl-1,8-dibenzylxanthines as inhibitors of human cytosolic phosphoenolpyruvate carboxykinase

Article abstract of DOI:10.1016/j.bmcl.2007.05.013

New modifications on the C-8 4-aminobenzyl unit of the previously reported 3-alkyl-1,8-dibenzylxanthine inhibitors of cPEPCK are presented. The most active compound reported here is the 5-chloro-1,3-dimethyl-1H-pyrazole-4-sulfonic acid amide derivative 2

Full text of DOI:10.1016/j.bmcl.2007.05.013

Bioorganic & Medicinal Chemistry Letters 17 (2007) 3835–3839
C-8 Modifications of 3-alkyl-1,8-dibenzylxanthines as inhibitors
of human cytosolic phosphoenolpyruvate carboxykinase
Sherrie L. Pietranico,a Louise H. Foley,a,* Nicholas Huby,a Weiya Yun,a Pete Dunten,a,
John Vermeulen,a Ping Wang,a Katherine Toth,a Gwendolyn Ramsey,b
Mary-Lou Gublerb and Stanley J. Wertheimerb
aDepartment of Discovery Chemistry, Roche Research Center, Hoffmann-La Roche Inc., Nutley, NJ 07110, USA
bDepartment of Metabolic Diseases, Roche Research Center, Hoffmann-La Roche Inc., Nutley, NJ 07110, USA
Received 9 March 2007; revised 4 May 2007; accepted 7 May 2007
Available online 22 May 2007
Abstract—New modifications on the C-8 4-aminobenzyl unit of the previously reported 3-alkyl-1,8-dibenzylxanthine inhibitors of
cPEPCK are presented. The most active compound reported here is the 5-chloro-1,3-dimethyl-1H-pyrazole-4-sulfonic acid amide
derivative 2 with an IC50 of 0.29 0.08 lM. An X-ray analysis of a heteroaromatic sulfonamide is presented showing a new p–p
interaction.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Previously we reported1,2 the first GTP-competitive
inhibitors of cytosolic phosphoenolpyruvate carboxyki-
nase (cPEPCK). The PEPCK enzyme catalyzes the
rate-limiting step in gluconeogenesis3 and a number of
papers show that its overexpression in transgenic ani-
mals leads to fasting hyperglycemia and impaired glu-
cose tolerance.4 Reduction of cPEPCK protein levels
using a RNAi has recently been shown to lower blood
glucose levels in diabetic mice5 providing further sup-
port for cPEPCK as a therapeutic target in diabetes.
toma cells, are reported for some of the more potent
compounds. The new modifications at C-8 improved
the in vitro activity by >2-fold over that of xanthine 1
and provided the most active inhibitor reported here,
5-chloro-1,3-dimethyl-1H-pyrazole-4-sulfonic acid {4-
[3-cyclopropylmethyl-1-(2-fluorobenzyl)-2,6-dioxo-2,3,
6,7-tetrahydro-1H-purin-8-ylmethyl]phenyl}amide (2),
with an IC50 in the enzyme assay of 0.29 0.08 lM7
and EC50 in the cellular assay of 2.17 0.57 lM.
The required C-8 3- or 4-aminobenzyl xanthines 3
(R3 = butyl or cyclopropylmethyl) were prepared from
N-trifluoro-3- or 4-aminophenyl acetic acid using the
methods described in earlier papers.1,2 A second route
to 3 employed the N-Boc protected uracil 4 as shown
in Scheme 1. The N-Boc protected uracils 4 were pre-
pared from N-Boc-4-amino phenyl acetic acid8 using
the methods described in Refs. 1 and 2. The 3- or 4-
aminobenzyl xanthines 3 were converted into the sulfon-
amides 2, 10–24, 27 or the trifluoroacetamides 5a,b as
shown in Scheme 1.
Our earlier reports1,2 covered the discovery, X-ray stud-
ies, and modifications at N-1, N-3 as well as limited
modifications at C-8 of a class of xanthines that led to
xanthine 1 as the first submicromolar inhibitor of the
cPEPCK enzyme. We now describe use of the previously
described enzyme assay1 for the study of new modifica-
tions at C-8 of 3-alkyl-1,8-dibenzylxanthines. Results of
a cellular assay,6 evaluating a compound’s ability to in-
hibit PEPCK-dependent gluconeogenesis in rat hepa-
Keywords: cPEPCK inhibitors; Xanthine class; X-ray; Enzyme and
cellular assays.
Initial SAR development at xanthine’s C-8 position had
suggested that the benzyl unit was critical and that an
acetamido substituent on the 4-position of the phenyl
group improved activity.1 However the X-ray studies
of early xanthines2 failed to provide an explanation for
the improved activity when the C-8 benzyl group
contained the 4-acetamido unit. The lack of improved
*
Corresponding author at present address: 1757 Augusta Drive, Fort
Myers, FL 33907, USA. Tel.: +1 315 369 3769; e-mail address:
louf1203@aol.com
Corresponding author concerning the X-ray data: Stanford Linear
Accelerator Center, 2575 Sand Hill Rd., Mailstop 99, Menlo Park,
CA 94025, USA; e-mail address: pete@slac.standford.edu
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.05.013

3836
S. L. Pietranico et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3835–3839
ities observed with the sulfonamides 10 and 11 suggested
further investigation of this compound class might be
warranted.
H
N
F
O
O
H
N
O
C(CH3)3
References 1,2
N
O
N
O
NH2
R3
The benzyl sulfonamide 12 and phenyl sulfonamides
whether unsubstituted 13 (Table 2) or substituted with
electron rich or electron poor groups (data not shown),
however, provided no improvement over the sulfona-
mides 10 and 11. However, changing the aromatic group
on the sulfonamide to a 5-membered nitrogen contain-
ing aromatic heterocycle led to improved activity in
the enzyme assay; compare the thiophenes 14, 15 with
imidazole 16 and the pyrazole 17 and also note the activ-
ities of the thiazole 18 and oxazole 19 in the N-3 cyclo-
propylmethyl class. Although the imidazoles 16 and 20
showed good activity against the enzyme (relative to 6
and 1, respectively) they were inactive in cells. The pyr-
azoles 17, 2, 21, 22 all showed improved activity in the
enzyme assay and compound 2 showed improved activ-
ity in the cellular assay when compared to the corre-
sponding N-acetyl derivative 1. Changing the C-8
aminobenzyl unit in 2 to a 5-amino-2-pyridyl unit re-
sulted in a compound with slightly decreased activity
against the enzyme and a 2-fold decrease in activity in
cells, data not shown. Interestingly, replacing the pyra-
zole unit of 2 with a pyridine, compound 23, gave an
analog less active in the enzyme assay than even the cor-
responding N-acetyl derivative 1. The 8-quinolyl 24
showed improved activity against the enzyme relative
to the pyridyl analog 23 but showed no activity in the
cellular assay.
4
F
O
R3 a = butyl,
H
N
N
b = cyclopropylmethyl
N
N
O
R3
NH2
3
b.
c.
2, 10-24, 27
5a,b
Scheme 1. Routes to 3-alkyl-1,8-dibenzylxanthine sulfonamides and
N-trifluoroacetamides. Reagents and conditions: (a) (1) 4 M HCl,
dioxane, rt; (2) 10% NaOH, MeOH/H2O, 50 ꢁC oil bath; (b)
00
R8 –SðO2Þ–Cl in pyridine, rt; (c) N-(trifluoroacetoxy)succinimide.9
activity of the trifluoroacetyl analogs 5a and b (Table 1),
with their highly electronegative trifluoromethyl group,
over the corresponding acetamide analogs (compounds
6 and 1) made a H-bonding explanation unlikely.
A sampling of other C-8 modifications of compounds in
the N-3 butyl and N-3 cyclopropylmethyl classes is given
in Table 1. The decreased activity shown by the 2,4-
disubstitution on the C-8 benzyl group, compounds 7
and 8, when compared to the 4-substituted analogs 6
and 3a suggested that the presence of an additional 2-
substituent decreased activity relative to compounds
containing only a 4-substituent. Extending the acetyl
unit on C-8 by attaching a dimethylamino group, com-
pound 9, increased solubility but lowered the activity
relative to the corresponding amide analog 1. The activ-
Explanation of some of the increase in activity due to an
aromatic substituent’s attachment to the sulfur of the
sulfonamide was found in the X-ray study of sulfon-
amide 16,10 (Fig. 1). This X-ray showed that the sulfon-
amide linkage allowed a turn that placed the imidazole
ring on top of Phe530; thus Phe530 was now sandwiched
Table 1. IC50 values from the enzyme assay for representative C-8 benzyl modifications of xanthines 6 and 1
F
O
N
F
O
N
H
N
H
N
2'
2'
N
R
N
R
N
N
O
O
4'
4'
R
R
1, 3b, 5b, 9-11
3a, 5a, 6-8
0
0
a,b
IC50 (lM)
Compound
R4
R2
3a
5a
6
Amino
H
H
H
2.56 0.05 (5)
2.30
N-Trifluoroacetylamino
N-Acetylamino
N-Acetylamino
Amino
2.11 0.60(32)
8.70 0.85 (2)
5.65 0.07 (2)
0.69 0.29 (6)
1.56 0.53 (6)
1.80
7
N-Acetylamino
Amino
8
1
N-Acetylamino
Amino
H
H
H
H
H
H
3b
5b
9
N-Trifluoroacetylamino
NH–CO–CH2–N(CH3)2
NH–SO2–CH3
2.07 0.31 (6)
1.25 0.30 (6)
0.96 0.12 (3)
10
11
NH–SO2–N(CH3)2
a The assay protocol is given in Ref. 1.
b IC50 inhibitory values from the enzyme assay. Results showing standard deviation (SD) values were assayed more than once. If multiple assays were
preformed, the number of repetitions is shown in parentheses. Assays were conducted with each compound run in duplicate. A single IC50, without
SD, indicates a single assay with the average of the duplicate run indicated.

S. L. Pietranico et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3835–3839
3837
Table 2. IC50 and EC50 values from the enzyme and cellular assays for representative sulfonamide modifications of the C-8 benzylamino unit of
xanthines 6 and 1
F
F
H
N
O
N
H
N
O
N
N
N
N
N
O
O
R8''
R8''
N
H
N
H
6 & 12-17
1, 2 & 18-24
00
c
e,f,g
Compound
R8
IC50 (lM)a,b [ ]
EC50 (lM)
[ ]c
6
12
13
–C(O)CH3
–S(O)2–Bn
–S(O)2–Ph
2.11 0.60 (32)
31% at 1d
22% at 1d
8.6
ND
ND
SO
2 CO2CH3
14
15
19% at 1d
30% at 1d
ND
ND
S
O2
S
S
O2
S
N
16
17
0.63 0.14 (16)
0.54 0.07 (9)
IA
N
CH3
O2
S
CH3
N
N
3.97 1.5 (5)
Cl
CH3
1
2
–C(O)CH3
0.69 0.29 (6)
0.29 0.08 (38)
4.70 1.6 (7)
2.17 0.57 (9)
O2
S
CH3
N
N
Cl
CH3
O2
S
S
NH2
18
19
0.23 [0.23, 2 = 0.25]
8.20 [8.20, 2 = 1.90, 21 = 4.70]
3.40 [3.40, 21 = 4.00]
N
H3C
O2
S
CH3
0.79 0.26 (6)
O
N
H3C
O2
S
N
CH3
20
21
0.34 0.04 (3) [0.30, 2 = 0.25]
IA
N
CH3
O2
S
CH3
N
N
0.24 0.05 (44)
5.57 0.97 (8)
CH3
O2
S
N
22
23
24
0.39 0.05 (4) [0.43, 2 = 0.32, 21 = 0.23]
1.95 0.07 (2) [1.90, 2 = 0.35]
0.45 0.02 (2)
8.75 0.25 (2) [9.00, 2 = 2.40, 21 = 6.40]
N
CH3
O2
S
5.50 [5.50, 21 = 6.50]
N
N
O2
S
IA
a Legends same as Table 1.
b Legends same as Table 1.
c The numbers in brackets are the average of a duplicate assay on the compound indicated and, for comparison, compounds 2 and/or 21 assayed at
the same time.
d Percent inhibition at the lM concentration indicated.
e EC50 values from the cellular assay6 represent the results of duplicate assays.
f ND, not determined.
g IA, inactive up to 10 lM.

3838
S. L. Pietranico et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3835–3839
chloro-di-methyl substituted pyrazole with a trimethyl
pyrazole, in other xanthine classes these two pyrazole
modifications were found to have nearly identical activ-
ities against the enzyme. Comparison of the N-methyl
analog of sulfonamide 2, the N-methyl-N-chloro-
dimethylpyrazole sulfonamide 26, with 2 showed a 6-
fold decrease in activity in the enzyme assay. The loss
of activity observed with the reverse sulfonamide 25
may be due to steric interactions between the sulfone
oxygens and the methyl groups on the pyrazole ring
and in the N-methyl analog 26 the N-methyl of the N-
methyl sulfonamide with the methyl and halogen on
the pyrazole ring. These interactions, not seen in the
parent compounds, would increase the energy of any
conformational changes required for binding to the pro-
tein. This analysis assumes that the preferred conforma-
tion for sulfonamides is where nitrogen’s lone pair
bisects the sulfone oxygens13 as observed in the X-ray
presented here.
The N-benzyl sulfonamides where the amine attach-
ment was moved to the 3-position of the phenyl unit,
for example, compound 27, showed a decreased activ-
ity in both the enzyme (>3-fold) and cellular (>2-fold)
assays relative to the corresponding 4-substituted ana-
log 2. In the parent C-8 benzyl class the 3-amino and
3-acetamido analogs showed a similar loss in activity
when compared to the corresponding 4-substituted
analogs (data not shown). These results suggested that
while there may be space for groups at the 3-position,
neither the N-benzyl 3-amine, -acetamide nor -sulfon-
amide groups were making any new interactions with
the protein.
Figure 1. Xanthine 16 bound in the purine-binding site of cPEPCK.
Compound 16 is in light blue and the amino acids in the purine-
binding site of cPEPCK are shown in green. In both the xanthine and
cPEPCK the atoms of oxygen, nitrogen, and sulfur are shown in red,
dark blue, and yellow, respectively.
between the imidazole ring and the xanthine core of the
ligand. The lower activities observed for the methylsulf-
onamide 10 and its dimethylamino derivative 11 are in
agreement with the X-ray results. The fact that the
upper face of the xanthine-binding cleft, the area above
the imidazole and sulfonamide moieties, is exposed to
F
F
O
H
F
O
N
O
N
H
N
H
N
N
N
N
H
N
N
N
N
SO2
N
O
N
O
O
CH3
H
N
N
S
SO2
N
CH3
CH3
N
Cl
O2
N
H3C
N
27 IC50 0.94 μM
EC50 3.5 μM
H3C
N
H3C
Cl
25 IC50 6.0 μM
N
H3C
26
IC50 1.50 μM
H3C
solvent may explain the need for more hydrophilic aro-
matic groups. The other interactions of the protein with
the xanthine inhibitor are identical to those already de-
scribed.2 However, the sulfonamide 16 has trapped a
slightly different conformation of the enzyme in the crys-
tal than reported earlier.2 The C-terminal domain (resi-
dues 435–622) is rotated 4.8ꢁ with respect to the
N-terminal domain, in a slightly more open conforma-
tion than seen in the X-ray structure of GTP bound to
PEPCK11 and the previously reported xanthines.
In summary, we present modifications at xanthine’s C-8
aminobenzyl moiety that show 5-membered heteroaro-
matic N-benzylsulfonamides provide a new p–p interac-
tion and further increase the inhibitory activity of the
xanthine class of cPEPCK inhibitors. In the sulfonamide
class the 5-chloro-1,3-dimethylpyrazole analog 2 pro-
vides the most active inhibitor in both the enzyme and
cellular assays.
Acknowledgments
Several N-heteroaromatic benzylsulfonamides (hereafter
called reverse sulfonamides) were prepared12 and all
were less active than the N-benzyl sulfonamides. For
example the direct comparison of the reverse sulfon-
amide 25 with 17 showed at least a 9-fold decrease in
activity in the enzyme assay. While this compares a
We thank Dr. David Fry (Roche) for making us aware
of an analysis by Prof. Franc¸ios. Diederich (ETH-Zur-
¨
ich) and Dr. Klaus Mueller (Roche) of small molecule
X-ray structures showing the conformational preference
for sulfonamides noted here.13

S. L. Pietranico et al. / Bioorg. Med. Chem. Lett. 17 (2007) 3835–3839
3839
were prepared in DMSO and then added to the exper-
imental medium such that the final DMSO content was
0.5%. Experimental medium containing the test com-
pound (500 lL) was added to each well and assays were
preformed in duplicate. After incubating the cells for 6 h
at 37 ꢁC, the medium was collected and stored frozen at
ꢁ20 ꢁC. The amount of glucose in the medium was
determined using a fluorometric assay (AMPLEX RED/
Molecular Probes) and measured using a plate reader.
7. As noted earlier [2] the IC50s reported here might be 9-fold
higher than the true binding constants.
References and notes
1. Foley, L. H.; Wang, P.; Dunten, P.; Ramsey, G.; Gubler,
M.-L.; Wertheimer, S. J. Bioorg. Med. Chem. Lett. 2003,
13, 3607.
2. Foley, L. H.; Wang, P.; Dunten, P.; Ramsey, G.; Gubler,
M.-L.; Wertheimer, S. J. Bioorg. Med. Chem. Lett. 2003,
13, 3871.
3. Cimbala, A. N.; Lamers, W. H.; Nelson, K.; Monahan, J.
E.; Yoo-Warren, H.; Hanson, R. W. J. Biol. Chem. 1982,
257, 7629.
8. Yu, C.; Taylor, J. W. Bioorg. Med. Chem. 1999, 7, 161.
9. Bergeron, R. J.; McManis, J. S. J. Org. Chem. 1988, 53,
3108.
4. (a) Rosella, G.; Zajac, J. D.; Baker, L.; Kaczmarczyk, S.
J.; Andrikopoulos, S.; Adams, T. E.; Proietto, J. Mol.
Endocrinol. 1995, 9, 1396; (b) Valera, A.; Pujol, A.;
Pelegrin, M.; Bosch, F. Proc. Natl. Acad. Sci. U.S.A. 1994,
91, 9151; (c) Sun, Y.; Liu, S.; Ferguson, S.; Wang, L.;
Klepcyk, P.; Yun, J. S.; Freidman, J. E. J. Biol. Chem.
2002, 277, 23301.
˚
10. PDB code: 2 GMV, Resolution 2.30 A, R factor 21.7%.
11. Dunten, P.; Belunis, C.; Crowther, R.; Hollfelder, K.;
Kammlott, U.; Levin, W.; Michel, H.; Ramsey, G. B.;
Swain, A.; Weber, D.; Wertheimer, S. J. J. Mol. Biol.
2002, 316, 215.
´
´
´
5. Gomez-Valades, A. G.; Vidal-Alabro, A.; Molas, M.;
Boada, J.; Bermu´dez, J.; Bartrons, R.; Perales, J. C. Mol.
Ther. 2006, 13, 401.
12. Prepared from ethyl 4-(chlorosulfonyl)phenyl acetate
Kawashima, Y.; Sato, M.; Yamamoto, S.; Shimazaki,
Y.; Chiba, Y.; Satake, M.; Iwata, C.; Hatayama, K. Chem.
Pharm. Bull. 1995, 43, 1132, The sulfonyl chloride was
reacted with the desired heterocyclic amine and then the
ester hydrolyzed to give the needed phenyl acetic acid
derivative for reaction with the uracil according to Ref. 1.
13. Other X-rays showing this phenomenon may be found in
the following references: Yamada, K.; Fujihara, H.;
Yamamoto, Y.; Miwa, Y.; Taga, T.; Timioka, K. Org.
Lett. 2002, 4, 3509; (a) Aggarwal, V. K.; Charmant, J. P.
H.; Ciampi, C.; Hornby, J. M.; O’Brien, C. J.; Hynd, G.;
Parsons, R. J. Chem. Soc. Perkin Trans. 1 2001, 3159; (b)
Hendi, M. S.; Davis, R. E.; Wheeler, K. A. Cryst. Eng.
2000, 3, 209.
6. The PEPCK cellular assay measured the test compounds
inhibition of glucose production in H4IIE rat hepatoma
cells. To maximally activate PEPCK, cells were starved
overnight to deplete their endogenous stores of glycogen.
The H4IIE cells were plated in a 12-well dish at a density
of ꢀ2–3 · 105 cells/well. Cells were incubated overnight at
37 ꢁC in RPMI medium with 10% fetal calf serum (FCS).
The following day, medium was removed and cells were
washed twice in 1 mL phosphate-buffered saline +calcium
and magnesium (PBS +Ca/Mg). To each well was then
added 1 mL RPMI medium (ꢁglucose, ꢁFCS, +1 mg/mL
BSA, +20 mM Na Pyruvate) and incubated overnight at
37 ꢁC. Prior to addition, dilutions of the test compounds