Novel diamide-based inhibitors of IMPDH

Article abstract of DOI:10.1016/S0960-894X(02)00132-4

A series of novel amide-based small molecule inhibitors of inosine monophosphate dehydrogenase is described. The synthesis and the structure-activity relationships (SARs) derived from in vitro studies are presented.

Full text of DOI:10.1016/S0960-894X(02)00132-4

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1323–1326
Novel Diamide-Based Inhibitors of IMPDH
Henry H. Gu,* Edwin J. Iwanowicz, Junqing Guo, Scott H. Watterson, Zhongqi Shen,
William J. Pitts, T. G. Murali Dhar, Catherine A. Fleener, Katherine Rouleau,
N. Z. Sherbina, Mark Witmer, Jeffrey Tredup and Diane Hollenbaugh
Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ 08543-4000, USA
Received 24 October 2001; accepted 5 February 2002
Abstract—A series of novel amide-based small molecule inhibitors of inosine monophosphate dehydrogenase is described. The
synthesis and the structure–activity relationships (SARs) derived from in vitro studies are presented. # 2002 Elsevier Science Ltd.
All rights reserved.
Inosine monophosphate dehydrogenase (IMPDH), a
key enzyme in the de novo synthesis of guanosine
nucleotides, catalyzes the irreversible NAD-dependent
oxidation of inosine-50-monophosphate (IMP) to xan-
thosine-50-monophosphate (XMP).1 Two distinct genes
encoding IMPDH have been identified and the gene
products are labeled type I and type II. The proteins are
of identical size (514 amino acids)2ꢀ4 and share 84%
sequence identity. B- and T-lymphocytes depend on the
de novo, rather than the salvage pathway, in the gen-
eration of nucleotides. Due to the unique reliance of
lymphocytes on the de novo pathway, IMPDH is an
attractive target for selectively inhibiting an immune
response, without undue inhibition of other proliferating
cells.
Dose-limiting gastrointestinal (GI) toxicity is observed
from oral administration of either MPA or CellCept1
in a clinical setting. Seminal contributions by Vertex
have shown that biarylureas, exemplified by VX-497,
are highly potent inhibitors of IMPDH catalytic activity
(Fig. 1).9 This compound is currently under evaluation
in the clinic and is expected to have an improved thera-
peutic window with regard to GI toxicity-limiting dosing.
Our focus is the identification and development of
potent, selective inhibitors of IMPDH with improved
physical and pharmacological properties and fewer side
effects. In this report, we outline the synthesis and
Mycophenolic acid (MPA, Fig. 1) and some of its deri-
vatives have been shown to be potent, uncompetitive,
reversible inhibitors of human IMPDH type I and type
II.5,6 MPA has been demonstrated to block the response
of B- and T-cells to mitogen or antigen. IMPDH inhi-
bitors, such as MPA and the prodrug CellCept1, are
useful therapies in the treatment of transplant rejection
and have been found to have effects in autoimmune
disorders, such as psoriasis and rheumatoid arthritis.7
In addition, various cancer cell lines are sensitive to
IMPDH inhibition, and thus clinical trials evaluating
IMPDH inhibitors for their antileukemic effects have
been undertaken.8
*Corresponding author. Fax: +1-609-252-6601; e-mail: henry.gu@
bms.com
Figure 1. Chemical structures of MPA, VX-497, and diamides I and II.
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00132-4

1324
H. H. Gu et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1323–1326
biological evaluation of a new class of inhibitors of
IMPDH, based on the diamide backbones I and II (Fig. 1).
0.052 mM (IMPDH II). With R3 fixed as H, other lipo-
philic residues were examined at the R4 position. Ana-
logues with R4 as benzyl Ie, cyclohexyl If, n-propyl Ig,
and iso-propyl Ih were all inactive when tested at the 10
mM concentration. In contrast, the t-butyl analogue Ii,
displayed an IC50 of 0.010 mM. Assessment of Ib and Ij,
both bearing methyl substitution at R3, demonstrated
the necessity for substitution with hydrogen to achieve
high potency. The R4 phenyl of Ia may be effectively
substituted. Compound Ic, functionalized with a carba-
mate side chain from VX-497, displayed an IC50 value
of 0.030 mM.
The synthesis of diamide analogues 8 and 9 is shown in
Scheme 1. The 3-methoxy-4-(5-oxazolyl)-aniline 5 was
prepared from 2-methoxy-4-nitro-toluene 1 on a multi-
gram scale utilizing a synthetic procedure described by
Vertex.10 Acylation of 5 with ethyl oxalyl chloride gave
the corresponding oxalamic acid ethyl ester, which was
hydrolyzed to generate the oxalamic acid 6. In a similar
synthetic sequence, the malonamic acid 7 was prepared
utilizing ethyl malonyl chloride as the acylating agent.
Amides 8 and 9 were synthesized from 6 and 7, respec-
tively, by treatment with the corresponding amine and
utilizing BOP as the coupling agent.11 The syntheses of
anilines 12 and 16, described in Scheme 2, were pre-
pared according to literature procedures.10,12 Com-
pound Ic was prepared from compound 6 utilizing [(3-
aminophenyl)methyl] carbamic acid [(3S)-tetrahydro-3-
furanyl] ester B under the standard BOP-coupling con-
ditions.13 Compounds Iq and Ir were prepared from
4-(5-oxazolyl)-aniline and 12, respectively, utilizing the
reaction sequence for the conversion of 5 to 8.
In general, analogues where R4 has a quaternary center
at the site of attachment retain potency. Compound Io,
featuring a carboxylic acid moiety, is the exception in
this group. Notably, when the gem-dimethyl residues of
R4 were fashioned into a cyclopentane ring giving Il,
potency was maintained. We found that small structural
changes to the 3-methoxy-4-(5-oxazolyl)-aniline moiety
led to significant changes in potency. Where replace-
ment of the methoxy residue with a bromide (Is) or
chloride (It) moiety was tolerated, removal of the
methoxy residue (Iq) was not. Similarly, utilizing
4-methyl-5-oxazolyl aniline as a replacement for 5 giv-
ing Ir, led to a significant loss in potency.
The structure–activity relationships for the inhibition of
IMPDH type II catalytic activity are summarized in
Table 1. In the ethanediamide series I, several lipophilic
residues were examined in the R3 and R4 positions.
Compound Ia (R4=Ph) displayed an IC50 value of
Propanediamide analogue IIa with phenyl as the R4
residue displayed potent inhibitory activity against
Scheme 1. Reagents and conditions: (a) HOAc, Ac2O, concd H2SO4, CrO3, 0–10 ꢁC (51%); (b) concd HCl, dioxane, reflux (91%); (c) TosMIC,
K2CO3, MeOH, reflux (84%); (d) 40 psi H2, Pd/C, EtOH, (95%); (e) ClCOCOOEt, Et3N, CH2Cl2 (95%); (f) NaOH, EtOH (90%); (g) ClCOCH2-
COOEt, Et3N, CH2Cl2 (100%); (h) NaOH, EtOH (96%); (i) BOP, NMM, DMF, HN-R3R4 (85–99%).
Scheme 2. Reagentsꢁand conditions: (a) Me-TosMIC, K2CO3, MeOH, reflux (79%); (b) 40 psi H2, Pd/C, EtOH, (95%); (c) HOAc, Ac2O, concd
reflux (95%).
.
H2SO4, CrO3, 0–10 C (28%); (d) concd HCl, dioxane, reflux (99%); (e) TosMIC, K2CO3, MeOH, reflux (95%); (f) SnCl2 2H2O, EtOH/EtOH,

H. H. Gu et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1323–1326
Table 1. SAR of diamides I and II for the inhibition of IMPDH II
1325
Compd
R1
R2
R3
R4
IMPDH II IC50, (mM)
Ia
Ib
Ic
Id
Ie
If
Ig
Ih
Ii
Ij
Ik
Il
Im
In
Io
Ip
Iq
Ir
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
OMe
H
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
4-Me-5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
5-Oxazolyl
H
Me
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
Ph
Ph
A
NHPh
CH2Ph
Cyclohexyl
n-Propyl
iso-Propyl
t-Butyl
t-Butyl
0.052
>10
0.030
>10
>10
>10
>10
>10
0.010
>10
0.019
0.018
0.074
0.047
>10
0.030
>10
>10
C(Me)2CH2OH
1-CH2OH-Cyclopentyl
C(Me)2CO2Me
C(Me)2CH2N(Me)2
C(Me)2CO2H
C(Me)2Ph
t-Butyl
H
Br
Cl
OMe
OMe
t-Butyl
t-Butyl
t-Butyl
Ph
m-Me–Ph
Is
It
IIa
IIb
0.050
0.055
0.40
0.44
optimize this series of analogues to achieve oral activity
in a B- and T-cell mediated pharmacodynamic model
will be reported in due course.
References and Notes
1. Jackson, R. C.; Weber, G. Nature 1975, 256, 331.
2. Collart, F. R.; Huberman, E. J. Biol. Chem. 1988, 263,
15769.
3. Natsumeda, Y.; Ohno, S.; Kawasaki, H.; Konno, Y.;
Weber, G.; Suzuki, K. J. Biol. Chem. 1990, 265, 5292.
4. Collart, F. R.; Huberman, E. US Patent 5,665,583, 1997;
Chem. Abstr. 1997, 113, 186070.
5. Anderson, W. K.; Boehm, T. L.; Makara, G. M.; Swann,
R. T. J. Med. Chem. 1996, 39, 46.
6. Nelson, P. H.; Eugui, E.; Wang, C. C.; Allison, A. C. J.
Med. Chem. 1990, 33, 833.
Figure 2. Uncompetitive inhibition of Il versus NAD+ (see ref 14 for
details).
IMPDH II. Substitution of the phenyl moiety at the
meta position, which gives IIb, produced similar results.
MPA is a potent inhibitor of mammalian IMPDHs
(Ki=10 nM for IMPDH II) and displays uncompetitive
kinetics with respect to both IMP and NAD+. Like-
wise, VX-497 displays similar kinetics of inhibition.
Compound Il was found to be an uncompetitive inhi-
bitor of with a Kii=23ꢂ3 nM (see Fig. 2).14 Due to the
structural similarity to VX-497, this was not unex-
pected.
7. Sievers, T. M.; Rossi, S. J.; Ghobrial, R. M.; Arriola, E.;
Nishimura, P.; Kawano, M. Holt Pharmacother. 1997, 17,
1178.
8. Markland, W.; McQuaid, T. J.; Jain, J.; Kwong, A. D.
Antimicrob. Agents Chemother. 2000, 44, 859.
9. Sintchak, M. D.; Nimmesgern, E. Immunopharmacology
2000, 47, 163.
10. Armistead, D. M.; Badia, M. C.; Bemis, G. W.; Bethiel, R.
S.; Frank, C. A.; Novak, P. M.; Ronkin, S. M.; Saunders, J. O.
PCTInt. Appl., WO 9740028, 1997; Chem. Abstr. 1997, 128,
3680.
11. BOP reagent: benzotriazol-1-yloxy-tris(dimethyl-amino)-
phosphonium hexafluoro-phosphate.
12. Possel, O.; van Leusen, A. M. Tetrahedron Lett. 1997, 48,
4229.
13. Gu, H. H.; Dhar, T. G. M.; Iwanowicz, E. J. PCT Int.
Appl., WO 0026197, 2000; Chem. Abstr. 2000, 132, 334449.
We have identified two new series of novel diamide-
based inhibitors of IMPDH II catalytic activity. Com-
pound Il, a member of the ethanediamide structural
class, is an uncompetitive inhibitor of IMPDH II with
respect to NAD+ with a Kii=23ꢂ3 nM. Studies to

1326
H. H. Gu et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1323–1326
variable concentrations of NAD+ (10, 20, 40, and 80 mM;
Scheme for the synthesis of B:
Km=27 mM). The protein concentration used for Ki determi-
nations was 10 nM, and reactions were initiated by adding
enzyme. A solution of inhibitor was prepared in DMSO (20
mM), and diluted serially. Four concentrations of Il were used
to calculate Ki (5, 10, 20, and 50 nM), maintaining a final
volume of 2.5% v/v DMSO. Samples were analyzed using a
Cary Model 3E Uv–vis spectrophotometer equipped with a
Peltier temperature controller maintained at 37 ꢁC. Software
provided by the instrument manufacturer was used to calcu-
late the initial rates of reaction as the linear least-squares fit to
each data set. To determine the Ki value, the initial velocity
data (rate per min) were fitted globally using the competitive,
uncompetitive, and non-competitive models with KinetAsyst
kinetics software (IntelliKinetics, Princeton, NJ, USA). The
best fits were obtained to the uncompetitive model of inhibi-
tion for Il, and shown above in Figure 1. Note the point at the
origin is an artifact of the fitting program.
14. The enzymatic activity of human IMPDH II was quanti-
tated using a procedure similar to reported methods.15,16 The
conversion of NAD+ to NADH was followed spectro-
photometrically at 340 nm [d(e)=6220 Mꢀ1 cmꢀ1]. Quartz
cuvettes (1.0 cm path length) were used, with a total volume of
1.00 mL. The assay buffer consisted of 50 mM Tris, pH 8.0,
100 mM KCl, 2 mM EDTA, 3 mM DTT, 1.00 mM IMP, and
15. Carr, S. F.; Papp, E.; Wu, J. C.; Natsumeda, Y. J. Biol.
Chem. 1993, 268, 27286.
16. Xiang, B.; Yalor, J. C.; Markham, G. D. J. Biol. Chem.
1996, 271, 1435.