3-Cyanoindole-based inhibitors of inosine monophosphate dehydrogenase: Synthesis and initial structure-activity relationships

Article abstract of DOI:10.1016/S0960-894X(03)00757-1

A series of novel small molecule inhibitors of inosine monophosphate dehydrogenase (IMPDH), based upon a 3-cyanoindole core, were explored. IMPDH catalyzes the rate determining step in guanine nucleotide biosynthesis and is a target for anticancer, immuno

Full text of DOI:10.1016/S0960-894X(03)00757-1

Bioorganic & Medicinal Chemistry Letters 13 (2003) 3557–3560
3-Cyanoindole-Based Inhibitors of Inosine Monophosphate
Dehydrogenase: Synthesis and Initial Structure–Activity
Relationships
T. G. Murali Dhar,* Zhongqi Shen, Henry H. Gu, Ping Chen, Derek Norris,
Scott H. Watterson, Shelley K. Ballentine, Catherine A. Fleener, Katherine A. Rouleau,
Joel C. Barrish, Robert Townsend, Diane L. Hollenbaugh and Edwin J. Iwanowicz*
Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ 08543-4000, USA
Received 30 April 2003; accepted 2 July 2003
Abstract—A series of novel small molecule inhibitors of inosine monophosphate dehydrogenase (IMPDH), based upon a 3-cyano-
indole core, were explored. IMPDH catalyzes the rate determining step in guanine nucleotide biosynthesis and is a target for
anticancer, immunosuppressive and antiviral therapy. The synthesis and the structure–activity relationships (SAR), derived from in
vitro studies, for this new series of inhibitors is given.
# 2003 Elsevier 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
cDNA’s encoding IMPDH have been identified and
isolated. These transcripts labeled type I and type II are
of identical size (514 amino acids).2,3 IMPDH II activity
is markedly upregulated in actively proliferating cell
types including cancers and activated peripheral blood
lymphocytes.4
CellCept1 (mycophenolate mofetil, MMF), a prodrug
of mycophenolic acid (MPA), has clinical utility due to
its inhibition of IMPDH, for the treatment of transplant
rejection. Dose-limiting gastrointestinal (GI) toxicity
exhibited from administration of either MMF or MPA
limits this drug’s potential for treatment of other auto-
immune disorders, such as psoriasis and rheumatoid
arthritis.5
Vertex reported in 2000 a structurally new chemical
series exemplified by VX-497 shown in Figure 1.6 We
recently reported the discovery of BMS-337197 as a
potent inhibitor of IMPDH.7 The N-[3-methoxy-4-(5-
*Corresponding authors. Tel.: +1-609-252-5416; e-mail: murali.dhar@
bms.com (T. G. M. Dhar); eiwanowicz@triadtherapeutics.com
(E. J. Iwanowicz)
Figure 1. The chemical structures of MPA, BMS-337197 VX-497 and
VX-148. IMPDH II inhibitory data for VX-148 as reported by
Vertex.8
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00757-1

3558
T. G. M. Dhar et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3557–3560
oxazolyl)phenyl moiety is a common structural element
that is present in both VX-497 and BMS-337197.
Recently, Vertex disclosed the structure of VX-148, its
second generation inhibitor that is conspicuously devoid
of the terminal oxazole moiety.8 This disclosure
prompted us to reveal our efforts to replace the oxazole
moiety with a nitrile residue for our key chemical series
exemplified by BMS-337197.
The use of the indole core offered the opportunity to
suitably position the terminal nitrile residue and the
flexibility to probe the methyl binding pocket. We rea-
soned that by virtue of being one carbon removed from
the phenyl group, the nitrile residue at the C-3 position
on the indole will be spatially closer to the N–H of Gly
326 and thus will be able to more effectively engage this
crucial interaction between the ligand and the protein
(Fig. 3).
A comparison of inhibitory values for VX-497 and 1
reveals that the carbamate side chain for this chemical
series contributes little in the way of enhancing binding
affinity to IMPDH II (Fig. 2). Similar inhibitory values
are also seen for BMS-337197 and 3. However, the car-
bamate side chain of VX-148 contributes significantly to
the potency of this molecule (compare VX-148 and 2).
Without the side chain to enhance binding affinity, the
two chemical series, exemplified by 2 and 4, are of simi-
lar potency (Fig. 2). Our hypothesis was that the reduc-
tion in the binding affinity may be due to the non-
optimal hydrogen bonding distance between the nitrile
residue of compounds 2 and 4 (Fig. 2) and the N–H of
Gly 326, in comparison to the complementary
measurement for the oxazole moiety of compounds 1
and 3 (vide infra).
The synthetic pathways utilized in the preparation of 3-
cyanoindole based inhibitors of IMPDH are outlined in
Schemes 1–4. The ureas (8–12, Table 1) were prepared
in a straightforward fashion by condensing 6-amino-3-
cyanoindole with the corresponding isocyantes as out-
lined in Scheme 1. The 2-aminooxazole-cyanoindole
series of compounds (13–32, Table 2) were synthesized
via a tandem iminophosphorane/hetrocumelene medi-
ated annulation as outlined in Schemes 2–4.11 It is
important to point out that the BOC and Fmoc groups
that were used to protect the meta and para substituted
benzylic amines prior to cyclization (Schemes 3 and 4)
were not compatible for the benzylic amine that was
ortho substituted. More specifically, polar products were
obtained during the generationof the phenacyl bromide
from the enolether (Scheme 2). This problem was cir-
We recently reported the identification of the 3-(oxazol-
5-yl) indole based IMPDH inhibitors (Fig. 2, 5–7) and
demonstrated that the 3-(oxazol-5-yl) indole group
serves as a excellent replacement of the aniline moiety of
the urea derivative (1) and BMS-337197.9 Based on the
reported protein/inhibitor crystallographic studies,10
our strategy was to investigate an alternative spatial
placement for the nitrile moiety allowing for a more
effective engagement with the N–H of Gly 226 while
maintaining the contribution to binding from the inhib-
itor/protein interactions at the methyl binding pocket
(Fig. 3).
Figure 3. A comparison of binding modes for 3, an analogue of BMS-
337197 and an indole-based inhibitor.
Scheme 1. (a) ClSO2NCO, CH3CN, DMF, 90 ꢀC (64%); (b) NaH,
THF, MeI, 0 ꢀC, (91%); (c) Pd/C, MeOH, 15 psi, 1 h (99%); (d)
dioxane reflux, 15 min (75–85%); (e) 1,10thiocarbonyldi-2(1H)-pyr-
idone, CH2Cl2, rt, 90%.
Figure 2. IC50 values for the oxazole- and nitrile-based IMPDH
inhibitors.

T. G. M. Dhar et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3557–3560
3559
Scheme 4. (a) (BOC)2O, DIPEA, CH2Cl2, rt (100%); (b)
Pd(PPh3)2Cl2, CH2C(OEt)(Bu3Sn), dioxane, 100 ꢀC, 5 h; (c) NBS,
H2O, THF, rt (20% for b, c); (d) NaN3, acetone, H2O, 50 ꢀC (100%);
(e) PPh3, dioxane, 100 ꢀC, 60 min (45%); (f) TFA, CH2Cl2, rt, (100%);
(g) DMF, ClCO-tBu, 0 ꢀC to rt (60%).
Scheme 2. (a) DMF, 80 ꢀC, 18 h, (79%); (b) Pd(PPh3)2Cl2,
CH2C(OEt)(Bu3Sn), dioxane, 80 ꢀC, 5 h (73%); (c) NBS, H2O, THF,
rt; (d) NaN3, 90 min (79% for c, d); (e) PPh3, dioxane, 100 ꢀC, 40 min
(60%); (f) NH2NH2, EtOH, 90 min (60%); (g) CDI, OHCH2CH2CN,
DMF, rt, 90 min, (85%).
Table 1. SAR of cyanoindole ureas
Compd
R1
R2
IMPDH II
IC50, mM
CEM
IC50, mM
1
8
9
10
11
12
NA
Me
Me
Me
Me
H
NA (UREA STD)
H
19
0.63
>10,000
0.67
0.23
1.0
—
—
—
—
—
o-Me
m-Me
p-Me
p-Me
>5000
IMPDH II with an IC50 of 33 nM. The potency of this
compound is approximately 3-fold better than the cor-
responding analogue in the the 3-(oxazol-5-yl) indole
aminooxazole series represented by compound 6. The
fact that the potency of this compound is approximately
10-fold better than the cyanophenyl oxazole compound
4, supports our initial hypothesis about the importance
of the distance between the nitrile and the N–H of Gly
326.
Scheme 3. (a) PPh3, dioxane, 90 ꢀC, 60 min (64%); piperidine, DMF,
rt, (91%); 2-naphthoyl chloride, DMF, Et3N, rt.
cumvented by using the phthalimido protecting group.
A few representative examples are outlined in Schemes
2–4.
Preliminary SAR of the cyanoindole urea series (Table
1) suggested that there was no improvement in either
the enzyme or cell potency when compared to the oxa-
zole urea (1) or the cyanophenyl urea (2) class of com-
pounds. Although these results were disappointing, one
SAR trend is noteworthy. Unlike the 3-(oxazol-5-yl)
indole urea series (5) the corresponding des-methyl
analogue in the cyanoindole series 12 is inactive, sug-
gesting that there are subtle differences in the orienta-
tions of these two classes of compounds with respect to
interaction with N–H of Gly 326 and the methyl binding
pocket.
In order to improve both the enzyme and cell potency, a
series of amines, amides and carbamates off the phenyl
group of compound 13 were synthesized. As Table 2
indicates there were no clear SAR trends both in rela-
tion to the position or the nature of substitution (for
example, compounds 18–20, 23–25 and 28–32). It is also
interesting to note that analogue 17 is significantly less
potent than the corresponding compound in the 3-
(oxazol-5-yl) indole aminooxazole series represented by
compound 7. The carbamates 28 and 29, the side chains
of which are related to VX-148 are also significantly less
potent both in terms of their ability to inhibit IMPDH
II and the T-cell proliferation response. The most
potent compound in this series was compound 26 with
an IC50 of 40 nM against IMPDH II that was compar-
able to the parent compound 13 (IC50=33 nM). How-
ever, both compounds 26 and 13 were significantly less
potent in inhibiting T-cell proliferation in a CEM cell
The SAR trends in the 2-aminooxazole-cyanoindole
series of compounds (13–32) is outlined in Table 2.
Unlike the urea 8, the parent compound in the 2-ami-
nooxazole–cyanoindole series with no substituents on
the phenyl ring (13) is a reasonably potent inhibitor of

3560
T. G. M. Dhar et al. / Bioorg. Med. Chem. Lett. 13 (2003) 3557–3560
ies differ significantly with respect to both BMS-337197
and the 3-(oxazol-5-yl) indole (5–7) series of IMPDH
inhibitors.
Table 2. SAR of 2-aminooxazole-cyanoindole series
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.; Kanno, Y.;
Weber, G; Suzuki, K. J. Biol. Chem. 1990, 265, 5292.
4. Jayaram, H. N.; Grusch, M.; Cooney, D. A.; Krupitza, G.
Curr. Med. Chem. 1999, 6, 561.
5. Sievers, T. M.; Rossi, S. J.; Ghobrial, R. M.; Arriola, E.;
Nishimura, P.; Kawano, M.; Holt, ?. Pharmacotherapy 1997,
17, 1178.
6. Jain, J.; Almquist, S. J.; Shlyakhter, D.; Harding, M. W. J.
Pharm. Sci. 2001, 90, 625.
7. Dhar, T. G. M.; Shen, Z.; Guo, J.; Liu, C.; Watterson,
S. H.; Gu, H. H.; Pitts, W. J.; Fleener, C. A.; Rouleau, K. A.;
Sherbina, N. Z.; McIntyre, K. W.; Witmer, M. R.; Tredup,
J. A.; Chen, B.-C.; Zhao, R.; Bednarz, M. S.; Cheney, D. L.;
MacMaster, J. F.; Miller, L. M.; Berry, K. K.; Harper, T. W.;
Barrish, J. C.; Hollenbaugh, D. L.; Iwanowicz, E. J. J. Med.
Chem. 2002, 45, 2127.
8. Jain, J.; Almquist, S. J.; Heiser, A. D.; Shlyakhter, D.;
Leon, E.; Memmott, C.; Moody, C. H.; Nimmesgern, E.;
Decker, J. Pharm. Exper. Ther. 2002, 302, 1272.
9. Dhar, T. G. M.; Shen, Z.; Fleener, C. A.; Rouleau, K. A.;
Barrish, J. B.; Hollenbaugh, D. A.; Iwanowicz, E. J. Bioorg.
Med. Chem. Lett. 2002, 12, 3305.
Compd
R
IMPDH II
IC50, mM
CEM
IC50, mM
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
H
o-Me
m-Me
p-Me
0.033
0.44
0.49
>0.55
0.266
0.28
0.18
0.26
0.20
0.13
0.10
0.094
0.18
0.040
0.084
0.24
0.420
0.21
0.260
0.420
8.0
—
>10
—
—
1.1
3.5
o-NMeCOCH2OAc
o-CH2NH2
m-CH2NH2
p-CH2NH2
m-NHC(O)Ph
2.8
3.5
5.3
m-CH2NHC(O)Ph
p-CH2NHC(O)Ph
p-CH2NHC(O)Me
p-CH2NHC(O)t-Bu
m-CH2NHC(O)2-napthyl
m-CH2NHC(O)1-napthyl
m-CH2NHC(O)O(CH2)2CN
o-CH2NHC(O)O(CH2)2CN
m-CH2NHC(O)O(CH2)2NMe2
m-CH2NHC(O)O-3-THF
o-CH2NHC(O)O-3-THF
>10
>10
4.9
4.4
2.5
>10
>10
1.4
>10
>10
line when compared to BMS-337197 (BMS-337197
CEM IC50=520 nM vs 4 mM for compound 26).
10. Sintchak, M. D.; Nimmesgern, E. Immunopharmacology
2000, 47, 163.
11. Dhar, T. G. M.; Guo, J.; Shen, Z.; Pitts, W. J.; Gu, H. H.;
Chen, B.-C.; Zhao, R.; Bednarz, M. S.; Iwanowicz, E. J. Org.
Lett. 2002, 4, 2091, and references cited therein.
In summary, a series of cyanoindole based IMPDH inhib-
itors have been designed and synthesized as replacements
for the N-[3-methoxy-4-(5-oxazolyl)phenyl moiety of
VX-497 and BMS-337197. The SAR trends for this ser-