Design, synthesis and biological evaluation of novel inosine 5′-monophosphate dehydrogenase (IMPDH) inhibitors

Article abstract of DOI:10.3109/14756366.2013.793184

This study is based on our attempts to further explore the structure-activity relationship (SAR) of VX-148 (3) in an attempt to identify inosine 5′-mono-phosphate dehydrogenase (IMPDH) inhibitors superior to mycophenolic acid. A five-point pharmacophore developed using structurally diverse, known IMPDH inhibitors guided further design of novel analogs of 3. Several conventional as well as novel medicinal chemistry strategies were tried. The combined structure- and ligand-based approaches culminated in a few analogs with either retained or slightly higher potency. The compounds which retained the potency were also checked for their ability to inhibit human peripheral blood mononuclear cells proliferation. This study illuminates the stringent structural requirements and strict SAR for IMPDH II inhibition.

Full text of DOI:10.3109/14756366.2013.793184

http://informahealthcare.com/enz
ISSN: 1475-6366 (print), 1475-6374 (electronic)
J Enzyme Inhib Med Chem, 2014; 29(3): 408–419
2014 Informa UK Ltd. DOI: 10.3109/14756366.2013.793184
!
ORIGINAL ARTICLE
Design, synthesis and biological evaluation of novel inosine
50-monophosphate dehydrogenase (IMPDH) inhibitors
Torsten Dunkern1*, Sunil Chavan2, Digambar Bankar2, Anuja Patil2, Pritee Kulkarni2, Prashant S. Kharkar2y,
Arati Prabhu2z, Heike Goebel1, Edith Rolser1, Waltraud Burckhard-Boer1, Premkumar Arumugam3, and
Mahindra T. Makhija2
1Global Discovery, Nycomed: a Takeda Company, Nycomed GmbH, Konstanz, Germany, 2Global Discovery, Nycomed: a Takeda Company, Mumbai,
3
India, and GVK Biosciences Private Limited, IDA Nacharam, Hyderabad, India
Abstract
Keywords
This study is based on our attempts to further explore the structure–activity relationship (SAR)
of VX-148 (3) in an attempt to identify inosine 50-mono-phosphate dehydrogenase (IMPDH)
inhibitors superior to mycophenolic acid. A five-point pharmacophore developed using
structurally diverse, known IMPDH inhibitors guided further design of novel analogs of 3.
Several conventional as well as novel medicinal chemistry strategies were tried. The combined
structure- and ligand-based approaches culminated in a few analogs with either retained or
slightly higher potency. The compounds which retained the potency were also checked for
their ability to inhibit human peripheral blood mononuclear cells proliferation. This study
illuminates the stringent structural requirements and strict SAR for IMPDH II inhibition.
3D pharmacophore, advanced chemical
analog, immunosuppressant, IMPDH,
VX-148
History
Received 24 March 2013
Accepted 24 March 2013
Published online 10 May 2013
Introduction
is some ambiguity about the importance of one isoform over the
other in terms of therapeutic utility6. Mycophenolic acid (MPA)
(1, Figure 1), a natural product, is a reversible, potent and
uncompetitive inhibitor of IMPDH. Mycophenolate mofetil
(MMF, CellCeptÕ), a prodrug of MPA, has been approved for
the treatment of acute allograft rejection following kidney
transplant7. MPA and its related forms [MMF or MPA sodium
(MyforticÕ)] cause dose-limiting gastrointestinal (GI) toxicity in
addition to their faster metabolic liability. Thus, high doses
(CellCeptÕ, 1 g b.i.d. and MyforticÕ, 720 mg b.i.d.) are required
to maintain the therapeutic plasma levels8. The competitive
IMPDH inhibitors such as Mizoribine (BredininÕ) and Ribavirin
(VirazoleÕ) (after intracellular activation by phosphorylation) are
nucleoside analogs and have unfavorable tolerability profiles too3.
Thus, there is a need for newer, safer, potent and orally
bioavailable IMPDH inhibitors (selective or nonselective). The
outcome of this search may further help to deduce the importance
of one isoform over the other in a particular disease state.
Nearly a decade ago, a new series of biaryl ureas exemplified
by VX-497 (2, Figure 1) appeared in the literature9. This was
followed by another round of structural modifications of 2 which
led to VX-148 (3, Figure 1)10. Bioisosteric replacement of urea
moiety in 3 led to a new class, N-aryl-2-aminooxazoles [e.g.
BMS-337197 (4, Figure 1)], of IMPDH inhibitors11. There are
several reports of novel and potent IMPDH inhibitors with varied
pharmacokinetic and pharmacodynamic profiles12,13. Despite the
availability of several potent IMPDH inhibitors (Figure 1),
the quest for new inhibitors continues owing to the involvement
of the target in several disease states.
Inosine 50-monophosphate dehydrogenase (IMPDH, EC
1.1.1.205) catalyzes a crucial step in the de novo synthesis of
guanine nucleotides. It oxidizes inosine 50-monophosphate (IMP)
to xanthosine 50-monophosphate (XMP) in a nicotinamide
adenine dinucleotide (NADþ)-dependent manner1,2. Owing to
its critical position in the purine biosynthesis pathway, the
inhibition of IMPDH (leading to reduced cellular levels of
guanine nucleotides) is an important strategy for antiproliferative,
antiviral and antiparasitic effects3. It is also a useful strategy for
producing immunosuppression as IMPDH-dependent de novo
synthesis of guanine nucleotides is responsible for maintaining the
growth and differentiation of human B and T lymphocytes4.
Two isoforms of human IMPDH, I and II (514 amino acids
each, 84% sequence identity) have been identified5. IMPDH I is
constitutively expressed and the majority of this isoform is found
in normal lymphocytes, whereas type II is inducible and is closely
linked to cell differentiation and neoplastic transformation4. There
*Present address: Torsten Dunkern, Global Drug Discovery, Gru¨nenthal
GmbH, 52099 Aachen, Germany.
yPresent address: Prashant S. Kharkar, SPP School of Pharmacy and
Technology Management, SVKM’s NMIMS University, V. L. Mehta
Road, Vile Parle (W), Mumbai-400 056, India.
zPresent address: Arati Prabhu, Piramal Life Sciences, 1 Nirlon Complex,
Goregaon (E), Mumbai-400063, India.
Address for correspondence: Mahindra T. Makhija, Global Discovery,
Nycomed: a Takeda Company, Plot 29-31, Suren Road, Andheri (E),
Mumbai-400093, India. Tel: +91 75077 37444. Fax: +91 22 26185422.
E-mail: makhijatm@yahoo.co.in
This study focuses on our in-house efforts to develop novel
IMPDH inhibitors, which will be devoid of the side effect profiles
of MPA and related drugs. These novel inhibitors can be

DOI: 10.3109/14756366.2013.793184
Novel IMPDH inhibitors 409
Figure 1. Chemical structures of select IMPDH inhibitors used for the pharmacophore model development12
.
potentially useful in several diseases/disorders (e.g. arthritis, pharmacophore model was selected based on: (a) the mutual
inflammatory bowel disease, psoriasis, dermatitis, prevention of agreement between the observed features and the corresponding
transplant rejection, etc.) in which immunosuppression is desired. interactions with IMPDH II residues and (b) pharmacophore
fitness score and related statistics. The model AADHR (A –
Acceptor; D – Donor; H – Hydrophobic; R – Ring features) along
with the interfeature distances is shown in Figure 2.
Compound 3 was selected for further investigations based on
Design strategy
Recently, a chemical feature-based pharmacophore model of
IMPDH inhibitors consisting of six features (one H-bond donor, the in-house early absorption, distribution, metabolism and
one H-bond acceptor, one aromatic ring, one hydrophobic and two excretion (eADME) evaluations (data not shown). In fact, the
excluded volumes) appeared in the literature14. Further, the oxazole-containing compounds were excluded because they were
pharmacophore features were confirmed by molecular docking found to inhibit CYP 3A4 and CYP 2D6 significantly (data not
analyses. In the present study, the molecular docking studies were shown). Compound 3 was divided into three parts – proximal
initially performed using IMPDH II crystal structure (PDB ID (phenyl ring with 4–CN and 3–OMe substitutions), middle (urea)
1JR1)6 to generate the ‘‘binding poses’’, which were subsequently and the distal (remaining part) (Figure 2). Given the importance
used as an input for developing the pharmacophore model. A total of the H-bond acceptor feature (Figure 2), several Part A
of nine potent and structurally diverse IMPDH inhibitors belong- modifications were tried to probe the binding cavity features
ing to classes biarylurea (2, 3 and 5, Charts 1 and 2), (Table 1) with the retention of at least one H-bond acceptor
oxazolamines (4, Figure 1), quinolone (6, Figure 1), indole (7, feature. The distances between the A and R features (Figure 2)
Figure 1), acridone (8, Figure 1) and quinazoline thione (9, spanned a greater range (data not shown) compared to the
˚
Figure 1) in addition to 1 were used in the process. The figures corresponding distance in the pharmacophore model (5.08 A).
depicting the binding modes of few of these inhibitors (2 and 3) Upon obtaining the results from Part A modifications, significant
can be found in the Supplementary material section. A five-point synthetic efforts were invested in finding a suitable urea

410 T. Dunkern et al.
J Enzyme Inhib Med Chem, 2014; 29(3): 408–419
Figure 2. The five-point pharmacophore
model – AADHR – for IMPDH inhibitors
with VX-148 (3, Figure 1) and the design
strategy based on 3 (Parts A, B and C)
utilized for systematic modifications. A:
Acceptor (light pink); D: Donor (cyan); H:
Hydrophobic (green) and R: Ring (orange)
features.
Table 1. In vitro inhibitory and cellular potencies of Part A modifications.
IMPDH II
pIC50
PBMC proliferation
inhibition (pIC50)
Reversal of
PBMC inhibition by GMP (%)
Compound
Part A
1 (MPA)*
2 (VX-497)*
3 (VX-148)*
–
–
6.89
8.46
6.42
6.63
4.71
54.5
5.45
54.5
4.98
54.5
44.5
4.84
54.5
54.5
54.5
54.5
54.5
55
6.9
6.6
7.06
6.25
55
100
100
100
84
–
–
45
–
4-cyano-3-methoxyphenyl
3,4-dimethoxyphenyl
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
3-methoxy-4-methylphenyl
3-methoxy-4-morpholin-4-yl-phenyl
4-(3,6-dihydro-2H-pyran-4-yl-3-methoxyphenyl
3-methoxy-4-(tetrahydropyran4-yl)phenyl
2,3-dihydrobenzofuran-6-yl
2,3-dihydrobenzo[1,4]dioxin-6-yl
3,4-methylenedioxyphenyl
1-methyl-1H-benzoimidazol-5-yl
3-methoxyquinoxalin-6-yl
2,3-dimethoxynaphathalen-6-yl
3-methyl-2,3-dihydrobenzofuran-5-yl
4-methoxyquinazolin-6-yl
2-methyl-4-chromenon-7-yl
4-chromenon-7-yl
55
5.23
NTy
55
–
–
55
5.44
5.45
5.27
5.53
55
5.5
5.38
55
0
0
0
19
–
0
0
–
*Refer Figure 1 for structures.
yNot tested.
replacement (Part B modifications). The combined results of Part Fisher Scientific Inc. (Waltham, MA), VWR International (Radnor,
A and B modifications along with the clues from the pharma- PA), etc., and were used as received unless otherwise indicated. All
cophore model were utilized to design suitable Part C modifica- reactions were performed under an inert atmosphere (argon or N2)
tions. Herein, our attempts to develop novel and potent IMPDH unless otherwise noted. Analytical silica gel 60 F254-coated thin-
inhibitors based on compound 3 with the help of ligand-based and layer chromatography (TLC) plates were purchased from Merck
partly structure-based design strategies are summarized.
Chemicals (Whitehouse Station, NJ), and were visualized with UV
light or by treatment with TLC reagents such as ninhydrin,
Dragandorff’s or phosphomolybdic acid (PMA). Flash-column
chromatography was carried out on Combiflash Rf _silica gel (230–
Materials and methods
General
1
400 mesh). H-NMR spectra were routinely recorded on Bruker
All the chemicals (intermediates and reagents) and solvents 300 MHz FT NMR, with tetramethylsilane (TMS) as an internal
(including dry solvents) were procured from commercial suppliers standard. Reversed-phase high-performance liquid chromatogra-
such as Sigma-Aldrich Chemical Co. (St. Louis, MO), Thermo phy (RP-HPLC) analyses were performed with a Waters Acquity

DOI: 10.3109/14756366.2013.793184
Novel IMPDH inhibitors 411
Scheme 1. Syntheses of compounds 13–27. Reagents and conditions: (a) NaBH4, EtOH, RT; (b) DPPA, DBU, toluene; (c) PPh3, THF:Water; (d) L-
(þ)-tartaric acid, MeOH; (e) NaOH, EtOAc; (f) H2, Pd(OH)2/C, EtOH; (g) KCN, DMSO; (h) 10, CDI, EtOAc, RT; (i) phenyl chloroformate, DCM, RT;
(j) DIPEA, EtOAc, RT; (k) DCM, RT, 6 h.
UPLC instrument (Waters, Milfort, MA) equipped with Waters Chemistry
PDA absorption detector. Analytical ultra-performance liquid
The syntheses of the final compounds are outlined in
chromatography (UPLC) analyses were carried out with a Waters
Schemes 1–6. Synthetic details of the representative compounds
BED C18 column (100 ꢀ 50 mm ꢀ 5 mm) using acetonitrile:water
13, 30 and 45 are shown for reference. Similar details for all other
solvent system (flow rate 1 mL/min, injection volume 10 mL). The
compounds are found in the Supplementary material section.
gradient used was – water 80:acetonitrile 20 (0 min) to water
20:acetonitrile 80 (20 min). The detection was done using UV/Vis
General procedure 1 (Route 1)
detector (ꢀmax 254 nm). For semipreparative analyses, a Waters
Xtera BED C18 column (100 ꢀ 50 mm ꢀ 5 mm) was used. The To a solution of intermediate 12 (190 mg, 0.5 mmol) and the
optical purity of chiral amine was determined using a Thermo GC corresponding aromatic amine (0.5 mmol) in EtOAc (5 mL) was
instrument (Thermo Electron Corporation, San Jose, CA). added N,N-diisopropylethylamine (0.16 mL, 1 mmol) dropwise.
Analytical gas chromatography (GC) analyses were carried out The reaction mixture was refluxed overnight, cooled to room
with a SUPELCO 30 M ꢀ 0.25 mm ꢀ 0.25 mm column (Sigma- temperature (RT) and washed with water (3 ꢀ 10 mL). Combined
Aldrich Chemical Co., St. Louis, MO). Optical rotations were organic layers were separated and dried (anhydrous Na2SO4). The
recorded on a PerkinElmer 241 polarimeter (PerkinElmer, solvent was evaporated and the crude product was purified by
Waltham, MA). Melting points were recorded using a MEL- flash chromatography.
TEMP II capillary melting point apparatus (Laboratory Devices
Inc., Holliston, MA), and are uncorrected.
General procedure 2 (Route 2)
cDNA coding for IMPDH II (accession number BC0155)
(UniProt P12268 IMDH2_HUMAN) was obtained from Open
Biosystems and cloned into pET21 Eschericia coli expression
vector. Escherichia coli DH5a, E. coli DE3 (BL21) and plasmid
pET-21b were purchased from Novagen (Billerica, MA).
Restriction enzymes, Pfu DNA Polymerase, DNA marker were
purchased from NEB. Plasmid DNA purification kit was
purchased from Qiagen (Hilden, Germany). Oligonucleotides,
Protease inhibitor cocktail, IMP, NAD, guanine monophosphate
(GMP), MPA and routine chemicals were obtained from Sigma-
To a solution of intermediate 11 (200 mg, 0.77 mmol) in DCM
(5 mL) at 0 ꢁC was added corresponding isocyanate (0.8 mmol)
and the reaction mixture stirred at RT for 4 h. The reaction was
monitored by TLC. Reaction mixture was extracted with water
(3 ꢀ 10 mL). Organic layers were dried (anhydrous Na2SO4). The
crude product obtained after removing the solvent was purified
using flash chromatography.
General procedure 3 (Route 3)
Aldrich Chemical Co. The 96-well plate for UV assay was Compound 11 (200 mg, 0.77 mmol) and the respective carbamate
purchased from Costar. (0.8 mmol) were dissolved in EtOAc (5 mL). To this solution was

412 T. Dunkern et al.
J Enzyme Inhib Med Chem, 2014; 29(3): 408–419
Scheme 2. Syntheses of compounds 30–35. Reagents and conditions: (a) DCM, RT; (b) (i) CuSO4:SiO2; (ii) methanolic NH3, THF; (c) MeI, K2CO3,
acetone; (d) NaNHCN, 2-propanol, 80 ꢁC, 1 h; (e)1,10-thiocarbonyldi-2(1H)-pyridone, DCM, RT; (f) (i) HCl:1,4-dioxane, RT; (ii) 11b, CDI, EtOAc,
RT; (g) MeI, DMF, RT; (h) EtOH, reflux.
Scheme 3. Syntheses of compounds 36–39. Reagents and conditions: (a) KI (cat.), AcCN, RT; (b) HATU, DIPEA, THF, RT.
added N,N-diisopropylethylamine (0.16 mL, 1 mmol) dropwise. further extracted with EtOAc (2 ꢀ 10 mL). Combined organic
The reaction mixture was refluxed overnight and cooled to RT. phase was dried (anhydrous Na2SO4) and the solvent evaporated
The reaction was terminated by adding water (10 mL). to obtain crude product that was further purified by flash
The organic layer was separated and the aqueous layer was chromatography.

DOI: 10.3109/14756366.2013.793184
Novel IMPDH inhibitors 413
Scheme 4. Syntheses of compounds 40–44.
Reagents and conditions: (a) NaH, THF,
reflux; (b) DIPEA, THF, RT; (c) 1,10-
carbonothionyldipyri-din-2(1H)-one, DCM,
RT; (d) silver trifluoroacetate, TEA, AcCN,
reflux; (e) HATU, DIPEA, THF, RT; (f) (i)
DCM, RT; (ii) HgO, S, EtOH, reflux.
(2R)-1-Cyanobutan-2-yl-[(1S)-1-(3-{[(3,4-dimethoxyphenyl)car-
95:5) to produce 30 as yellow solid. Yield of 30: 50 mg (23%);
bamoyl]amino}phenyl)- ethyl] carbamate (13). General proced- Rf ¼ 0.25 (DCM:MeOH, 95:5); purity (HPLC) ¼ 90%; LC-MS
ure 1 was used. Off-white solid; yield of 13: 100 mg (40%); (ESI) (M þ H)þ: 435.25; HRMS (ESI) (M þ H)þ calcd for
1
Rf ¼ 0.25 (DCM:MeOH 9:1); purity (HPLC) ¼ 92%; LC-MS C23H27N6O3 435.2137, found 435.2139; H NMR (DMSO-d6) ꢁ
(ESI) (M þ H)þ: 441.17; HRMS (ESI) (M þ H)þ calcd for ppm 0.87–0.90 (t, J ¼ 7.4 Hz, 3H), 1.3–1.5 (m, 3H), 1.6–1.8 (m,
C23H29N4O5 441.2129, found 441.2132; 1H-NMR (300 MHz, 2H), 2.73–2.8 (d, J ¼ 5.3 Hz, 2H), 3.82 (s, 4H), 4.586 (m, 2H),
DMSO-d6) ꢁ ppm 0.90 (t, J ¼ 7.4 Hz, 3H), 1.34 (d, J ¼ 6.8 Hz, 5.46 (d, J ¼ 6.0 Hz, 2H), 6.56 (brs, 2H), 6.86–6.88 (d, J ¼ 6.8 Hz,
3H), 1.56–1.70 (m, 2H), 2.84 (d, J ¼ 4.9 Hz, 2H), 3.71 (s,3H), 2H), 7.13–7.16 (t, J ¼ 7.9 Hz, 2H), 7.44–7.47 (d, J ¼ 8.3 Hz, 2H),
3.74 (s, 3H), 4.56–4.74 (m, 2H), 6.86 (s, 2H), 6.92 (d, J ¼ 7.6 Hz, 7.81–7.83 (d, J ¼ 7.9 Hz, 2H).
1H), 7.17–7.24 (m, 2H), 7.26–7.34 (m, 1H), 7.38 (s, 1H), 7.90 (d,
J ¼ 7.9 Hz, 1H), 8.48 (s, 1H), 8.54 (s, 1H).
General procedure 4
(2R)-1-Cyanobutan-2-yl[(1S)-1-{3-[N0-(4-cyano-3-methoxyphe-
nyl)carbamimidamino]-phenyl}ethyl] carbamate (5 mL) amine (1 mmol) was added slowly. Reaction mixture
(30). Intermediate 29a (250 mg, 0.55 mmol)
To a solution of isothiocynate (1 mmol) (Scheme 5) in DCM
(see was stirred at RT for 6 h and extracted with water (3 ꢀ 5 mL).
Supplementary material), NEt3 (0.77 mL, 0.554 mmol), anhyd- Combined organic layers were dried (anhydrous Na2SO4).
rous CuSO4 (1 g), SiO2 (1 g) in THF (5 mL) was stirred at RT for Crude thiourea was purified using flash chromatography. To a
30 min. Methanolic NH3 (5 mL) was added and the reaction solution of thiourea (1 mmol) in acetone (5 mL), MeI
mixture was stirred at RT for another 30 min. It was filtered (2.5 mmol) and K2CO3 (345 mg, 2.5 mmol) were added. The
through celite and filtrate was concentrated in vacuo. Crude contents of the flask were stirred at RT for 4 h and the
product was purified by column chromatography (DCM:MeOH, solvent evaporated in vacuo. The residue was taken up in

414 T. Dunkern et al.
J Enzyme Inhib Med Chem, 2014; 29(3): 408–419
Scheme 5. Syntheses of compounds 45–53. Reagents and conditions: (a) (i) DCM, RT; (ii) MeI, K2CO3, RT; (b) NaNHCN, 2-propanol, reflux;
(c) LiOH, THF:H2O (2:1), RT.
EtOAc (10 mL) and washed with water (3 ꢀ 10 mL). Organic Isopropyl 2-[3-[[(E)-N0-cyano-N-(4-cyano-3-methoxyphenyl)car-
layer was dried (anhydrous Na2SO4) and crude product was bamimidoyl]amino]phen-oxy]acetate (45). Synthesized accord-
purified using flash chromatography to yield a colorless liquid. ing to general procedure 4. Analytical data of 45: white solid;
To a solution of this S-methyl intermediate (1 mmol) in IPA yield of 45: 200 mg (65%); Rf ¼ 0.2 (hexanes:EtOAc 6.5:4.5);
(5 mL), sodium cyanamide (160 mg, 2.5 mmol) was added and purity (HPLC) ¼ 98%; LC-MS (ESI) (M þ H)þ: 408.25; HRMS
the reaction mixture was refluxed for 4 h, cooled to RT and (ESI) (M þ H)þ calcd for C18H16N5O4 408.1661, found 408.1666;
the solvent evaporated in vacuo. The residue was dissolved in 1H NMR (DMSO-d6) ꢁ ppm 1.11–1.40 (m, 6H), 1.99 (s, 1H),
EtOAc (10 mL) and washed with water (3 ꢀ 10 mL). Organic 3.6–3.85 (s, 3H), 4.71 (s, 2H), 4.84–5.12 (m, 1H), 6.62-6.80 (m,
layer was dried (anhydrous Na2SO4) and the solvent 1H), 6.88–6.98 (m, 2H), 7.02 (dd, J ¼ 8.7, 1.9 Hz, 1H), 7.14 (d,
evaporated in vacuo and the crude product was subjected to J ¼ 1.5 Hz, 1H), 7.26 (t, J ¼ 8.3 Hz, 1H), 7.65 (d, J ¼ 8.7 Hz, 1H),
purification by flash chromatography.
9.88 (br. s., 1H), 9.92 (brs, 1H).

DOI: 10.3109/14756366.2013.793184
Novel IMPDH inhibitors 415
Scheme 6. Syntheses of compounds 54–60. Reagents and conditions: (a) 3-Bromoaniline, DCM, RT; (b) MeI, K2CO3, RT; (c) NaNHCN, 2-propanol,
reflux; (d) substituted arylboronic acid, Cs2CO3, SPhos, Pd(OAc)2/EtOAc:toluene (1:1), 100 ꢁC.
Molecular modeling
protocol. Colony 1 resulted in expected band size of 1.6 kb and
further plasmid was amplified from this clone and sequence was
confirmed by automated DNA sequencing. No mutations were
found after the sequence check. The verified clone was trans-
formed into expression host BL21(DE3) PlysS for expression
optimization.
All the molecular modeling studies were performed on a Hewlett-
Packard xw4600 computer workstation with (Hewlett-Packard,
Palo Alto, CA) 2 GB memory and Intel CoreÕ 2 Duo processors
(Intel Corp., Santa Clara, CA) of 2.66 GHz each running under
Red Hat Enterprise Linux (RHEL), version 5 (Red Hat, Inc.,
Cells were grown to an OD600 of 0.3 in lysogeny broth medium
supplemented with 100 mg/mL ampicillin at 37 ꢁC. Protein
expression was induced with the addition of 0.3 mM IPTG
followed by incubation for 5 h at 225 rpm, 25 ꢁC. Cells were
harvested by centrifugation at 4000 rpm, corresponding to 3400g
using HERAEUS Megafuge 1.0R (Kendro Laboratory Products
GmbH, Hanau, Germany) for 30 min and lysed the cells by
sonication at 70% amplitude, with a burst size of 5 min, 30 s on/
off. The clarified cell lysate was subjected for 40% w/v
ammonium sulfate fractionation and pelleted down the protein
by spinning the contents at 10 000 rpm for 30 min at 4 ꢁC. The
pellet was suspended in KH2PO4 buffer and applied the sample on
Sephadex G-25 column to desalt. These fractions were pooled and
the total protein was estimated to be 1.3 mg/mL by Bradford
assay. Bovine serum albumin (BSA) was used to generate
standard curve for the Bradford assay. The activity of the
enzyme with this preparation was found satisfactory with purity of
the enzyme estimated to be 475%.
Raleigh, NC). The molecular modeling operations were per-
¨
15
formed using Schrodinger suite 2011 .
Figure 1 lists the structures of IMPDH inhibitors used for the
development of pharmacophore (IC50 6–30 nM). For developing
the docking protocol, Chinese hamster (Cricetulus griseus)
IMPDH II crystal structure (PDB ID 1JR1)6 was used (details
not shown). Phase, version 3.316,17 was utilized for the develop-
ment of the pharmacophore model with the top ligand poses from
the docking runs as the starting point. The algorithm searches the
common features and scores the hypotheses. The end result of the
scoring process is an overall ranking of all hypotheses from which
user can select best candidate hypothesis using several criteria
such as the alignment of site points and vectors, selectivity,
volume overlap, number of ligands matched, relative conform-
ational energy and activity, etc.16,17. The top-scoring hypothesis
consisting of five features (AADHR) was selected for evaluating
new designs based on the template structure VX-148.
Screening, cloning, expression and purification of
recombinant IMPDH II
Enzyme assay
The assay was performed in a 100 mL final volume in 96-well UV
plates (Costar, 3635) with a reaction buffer composed of 50 mM
Tris–HCl (pH 8.0), 100 mM KCl, 3 mM EDTA and 1 mM
dithiothreitol, 4% v/v DMSO plus or minus inhibitory compound
and 0.75 mg of purified IMPDH II enzyme per well (from 1 mg/
mL stock concentration). The final volume of the enzyme stock
solution per well was 0.75 mL which was insignificant to cause
any change in the final assay buffer composition. The reaction was
initiated by the addition of 300 mM of IMP and 500 mM of NADþ
and the assay was allowed to proceed at 37 ꢁC for 50 min before
stopping by the addition of 15 mM GMP. The NADH generated
was measured by reading the absorbance at 340 nm. At this
wavelength, a background of 50.1 optical density (OD) was
observed with negligible crosstalk between wells. Molecular
Devices Spectramax 384 plus (Molecular Devices, LLC,
Sunnyvale, CA) was used for recording the absorbance, which
allows for selection of up to six wavelengths at a time for
absorbance detection in the UV-vis range (190–1000 nm). Initial
absorbance values (before substrate addition) were subtracted
Cloning strategy and primer designing
The sequences of the clones were verified with the Genebank
sequence (accession number BC015567). Primers were designed
to include restriction enzyme sites HindIII and XhoI to 50- and 30-
ends, respectively. Primers were synthesized from Sigma genosys
(Forw
– GCGTCAAGCTTATGGCCGACTACCTG; Rev –
AGCTGACCTCGAGGAAAAGCCGCTTCTC). PCR was per-
formed using Phusion polymerase enzyme (Finnzyme) for 28
cycles as per the manufacturer’s instructions. Each cycle included
98 ꢁC/10 s, 61 ꢁC/30 s, 72 ꢁC/90 s and final extension at 72 ꢁC for
10 min. Amplified PCR product was gel purified and digested
with HindIII and XhoI restriction enzymes. Digested PCR product
was ligated into pET-21b (Novagen) vector and transformed into
E. coli strain.
Selection and confirmation of clones
Clones from the antibiotic plate were selected and analyzed for
presence of insert by colony PCR using the above mentioned

416 T. Dunkern et al.
J Enzyme Inhib Med Chem, 2014; 29(3): 408–419
from the final absorbance to rule out compound auto fluores- the alicyclic ring might be little bigger in terms of size and shape
cence, if any. Mycophenolic acid (10 mM) used as a positive compared to planar, yet small –CN group. Further, substitution of
control and DMSO as a vehicle control. For IC50 determinations, morpholine in 15 with 3,6-dihydro-2H-pyran (16) gained some
a total of eight concentrations in triplicates with half log dilutions potency, although one log unit less than 3. Our logic that a planar
were used. For majority of the compounds, concentration ranged substituent should behave similar to –CN group was correct to
from 10 mM to 3 nM. Percent inhibition was calculated based on some extent, although a slightly bigger ring size was probably an
the following formula for each concentration and IC50 analyzed impediment to binding. Compound 16 with 4-tetrahydropyran
using the curve fitting program GraphPad Prism software version substitution also proved to be detrimental for IMPDH II inhibition
6 using nonlinear regression analysis.
despite showing good pharmacophore fitness. Having ascertained
the importance of planarity and presence of H-bond acceptor
feature, four new bicyclic analogs 18–20 and 24 (Table 1) were
designed. But to our disappointment, none of these exhibited
pIC50 above five. A second aromatic ring containing at least one
H-bond acceptor feature placed at a nearby position was then
introduced to probe the adjacent areas in the enzyme (exemplified
by 21–23 and 25–27. Compound 21 contained 1-methylbenzimi-
dazole ring system wherein N-3 was postulated to mimic the 4-
CN group, albeit placed at a distance shorter than –CN would
approach. Similarly, 22 had 2-methoxyquinoxaline substructure in
which 2-OMe group was slightly moved away compared to 3. In
case of 23, both the methoxy groups were placed farther due to
insertion of additional phenyl ring. Compound 25 was the
structural isomer of 22 where the positions of the ring N and
–OMe groups were interchanged. The last two designs 26 and 27
had chromone substructure wherein the –C¼O would be a
putative H-bond acceptor. All these modifications did not show
any improvement in potency. Hence, 4-cyano-3-methoxyphenyl
(present in 3) and 3,4-dimethoxyphenyl substructures were fixed
as Part A to be used in the next round of structural alterations.
Several of the compounds with Part A modifications were
further investigated for their ability to inhibit Concanavalin A
(ConA)-induced proliferation of PBMCs. The proliferation of
human PBMCs, which majorly consist of human T- and B-cells, is
highly dependent on the generation of GMP by reduction of IMP
by IMPDH. GMP acts as a precursor for cGMP, which is required
for DNA synthesis, as a prerequisite of proliferation. In accord-
ance with the enzymatic potency, compound 13 turned out to be
the most potent inhibitor of PBMC proliferation in this series of
modifications with a pIC50 of 6.25. This inhibition was almost
completely reversed by the addition of GMP, the end product of
the metabolic pathway IMPDH is part of. This indicates that the
effect of the compound is target specific. As also indicated in
Table 1, examples of compounds were identified (e.g. compounds
%Inhibition ¼ 100 ꢂ ½ðO:D: ꢂ positive controlÞ
ꢂðnegative control ꢂ positive controlÞ ꢃ 100ꢄ
O.D. ꢂ test compound inhibition; positive control ꢂ O.D. with
inhibitor; negative control ꢂ O.D. without inhibitor (100%
activity).
PBMC proliferation assay
Human peripheral blood mononuclear cells (PBMCs) were
isolated from citrate blood by means of a Ficoll–Hypaque
density-gradient centrifugation. After isolation, PBMCs were
seeded into 96-well cell culture plates in 10% v/v fetal calf serum
containing RPMI-10 medium (1.5 ꢀ 105 cell/100 mL/well). Test
compounds were added in a concentration ranging from 10 nM to
10 mM. DMSO to a final concentration of 0.1 % v/v DMSO was
serving as a vehicle control. In order to verify the specificity of a
potential growth inhibition by the IMPDH inhibitors, another
control containing the test compound at the highest concentration
of 10 mM and 100 mM GMP was added. After a period of 30 min
of pre-incubation with the compounds, Concanavalin A was added
to a final concentration of 3 mg/mL. The cells were subsequently
cultured for a period of 48 h. Within the last 12 h of this period,
3H-thymidine (1 mCu/mL) was added to the wells. Finally, cells
were harvested on 96-well filter plates by means of a cell
harvester. After several washing steps with PBS, the filter plates
were dried for 3 h at 60 ꢁC, scintillation fluid was added to each
well, plates were sealed and incorporated. 3H-thymidine was
counted using a TopCount (Perkin Elmer, Waltham, MA). The
IC50 values were calculated using GraphPad Prism 6 (GraphPad
Software, Inc., La Jolla, CA).
Results and discussion
In this study, our attempts to discover potent and novel IMPDH 20, 21, 22, 23, 25, 26), which showed no inhibition of IMPDH but
inhibitors using 3 as a starting point [advanced chemical analog of PBMC proliferation, which was not reversed by 100 mM GMP.
(ACA) approach] are outlined. The division of 3 in three parts Thus, those compounds are supposed to have a further mode of
(Figure 2) allowed us to explore the SAR systematically. For action affecting PBMC proliferation.
designing the analogs of 3, a five-point pharmacophore hypoth-
Part B modifications were aimed at finding a suitable
esis was developed using potent and structurally diverse IMPDH replacement for the linker urea. The donor (D) feature of
inhibitors. The top ranking hypothesis – AADHR (Figure 2) – AADHR five-point pharmacophore, represented by N1 of urea
formed the basis of new molecular entity (NME) design. Three (Figure 2), was found to form a critical H-bond with Asp274 of
out of five features [acceptor (A), hydrophobic (H) and ring (R)] IMPDH II (see Supplementary material section). Most of the Part
resided in Part A of 3, whereas Parts B and C each contained one B designs were based on this observation so as to contain at least
feature, a donor (D) and an acceptor (A), respectively (Figure 2). one H-bond donor. The first two designs were based on the
The first part of our study was to find a suitable replacement for replacement of urea 3 with guanidine (30 and 31, Table 2). The
4-cyano-3-methoxyphenyl (Part A) substructure which would presence of ¼NH instead of ¼O (as in 3) was not tolerated as both
possess corresponding pharmacophoric features A, H and R. these compounds failed to exhibit any inhibition at concentrations
These modifications (13–27) are listed in Table 1 along with their tested. Our next strategy was to replace urea with N-cyanogua-
corresponding pIC50 values for IMPDH II inhibition.
nidine. Compounds 32 and 33 exhibited some retention of
Replacement of 4-CN (acceptor feature) in 3 with 4-OMe IMPDH II inhibitory activity an order of magnitude less than 3.
group (13) led to a slight gain in potency. Substitution of 4-CN Excited by these results, compounds 34 and 35 were designed in
with 4-Me (14) dropped the potency by ꢅ2 log units indicating which ¼NCN was replaced with ¼C-NO2. To our disappoint-
the stringent requirements for H-bond interaction. Another ment, both these compounds were inactive due to their poor
modification wherein –CN was replaced with morpholine ring permeability characteristics (data not shown). Further modifica-
(15) disappointed us in terms of potency despite showing good tions to insert a link in the urea nitrogens and the carbonyl O were
pharmacophore fitness (data not shown). It was speculated that attempted. These designs ‘‘glycinamide (36)’’ and the ‘‘reverse

DOI: 10.3109/14756366.2013.793184
Novel IMPDH inhibitors 417
Table 2. In vitro inhibitory and cellular potencies of Part
modifications.
B
PBMC
proliferation
IMPDH inhibition
Reversal of
PBMC
inhibition
by GMP (%)
Compound R1
Part B
II pIC50
(pIC50)
42
CN
4.61
55
–
43
CN
54.5
54.5
5.87
5.41
0
PBMC
proliferation
IMPDH inhibition
Reversal of
PBMC
inhibition
by GMP (%)
44
CN
0
Compound R1
Part B
II pIC50
(pIC50
)
30
31
32
33
34
35
36
37
38
39
40
CN
OMe
CN
54.5
55
–
NT, not tested.
54.5
5.6
55
–
100
–
glycinamide (37)’’ also failed to show any enzyme inhibition.
Similar results were obtained when essential N1 of urea in 3 was
replaced with O and a methylene linker was introduced between O
and –C¼O (38 and 39, Table 2). Replacing urea with sulfonylurea
also did not help in terms of potency gain. Insertion of piperazine
(41) wherein the donor feature was removed turned out to be
inactive. Other modifications such as conformationally restricting
urea in the form of imidazolidine-2,4-dione (42), insertion of
either monocyclic (43) or bicyclic heteroaromatic ring (44) led to
a similar drop in the potency compared to 3. All these urea
replacements (Part B modifications) turned out to be futile except
N-cyanoguanidine where some retention of inhibitory activity was
observed. Hence, Part B was fixed as –NHC(¼NCN)NH– to be
used in combination with Part A in subsequent modifications in
Part C.
PBMC proliferation inhibition experiments using compounds
with Part B modifications indicated that compound 32 most
potently inhibited proliferation in a manner that is reversible by
adding GMP. However, also in this series, two compounds were
identified, which inhibited PBMC proliferation unspecifically and
which had no activity on IMPDH at all (compounds 43 and 44).
Our attention was then focused on finding a substructure that
would be a replacement for the benzyl carbamate moiety present
on the distal phenyl ring. Our initial efforts were directed at
removing the two chiral centers present in Part C for the ease of
synthesis. The acceptor (A) feature (represented by carbamate
–C¼O, Figure 2) was part of five-point pharmacophore.
Compounds 45 (ester) and 46 (acid) (Table 3) contain this feature
at a similar distance as found in 3. Both these compounds failed to
exhibit any activity. Substitution by N-methyltetrazole group also
led to an inactive compound (47, Table 3). Insertion of a
methylene linker between the tetrazole and the distal Ph ring to
introduce some conformational flexibility did not exhibit any
activity. Further attempts (50–53, Table 3) to vary the distance of
the acceptor feature with or without conformational restriction
(51) also led to a complete loss of inhibitory potency. As a last
resort, an additional Ph ring was introduced with an acceptor
feature placed at different position to tap the features of the
binding cavity. The groups which would behave as acceptor
included –F, –CF3,–COOH and –CN. Thus, any Part C modifi-
cation we tried led to loss of IMPDH II inhibitory activity.
Accordingly, all these compounds also turned out to be inactive
with respect to PBMC proliferation inhibition. In all, Parts A, B
and C modifications demonstrated a steep SAR for IMPDH II
5.38
OMe
CN
5.25
55
54.5
54.5
54.5
54.5
54.5
54.5
54.5
NT
–
OMe
OMe
CN
55
–
55
55
55
55
55
55
–
–
CN
–
OMe
OMe
–
–
41
CN
54.5
–
(continued )

418 T. Dunkern et al.
J Enzyme Inhib Med Chem, 2014; 29(3): 408–419
Table 3. In vitro inhibitory and cellular potencies of Part C modifications.
IMPDH II
pIC50
PBMC proliferation
inhibition (pIC50)
Reversal of PBMC
inhibition by GMP (%)
Compound
R3
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
–OCH2COOCH(CH3)
–OCH2COOH
54.5
54.5
54.5
54.5
54.5
54.5
54.5
54.5
54.5
54.5
54.5
54.5
54.5
54.5
54.5
54.5
55
55
55
55
55
55
55
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
2-methyl-2H-tetrazol-5-yl
1-methyl-1H-tetrazol-5-yl
(2-methyl-2H-tetrazol-5-yl)-methyl
–CH2CH2COOH
–CH¼CHCOOH
–CH2COOH
–COOH
2-fluorophenyl
3-flurophenyl
4-fluorophenyl
NT
NT
55
55
55
55
3-(trifluoromethyl)phenyl
3-carboxyphenyl
4-cyanophenyl
NT
55
5.0
3-cyanophenyl
NT, not tested.
samples and the Analytical Department, Global Discovery, Mumbai, for
their help in recording the required analytical data.
inhibition. It was disappointing to see that most of these
compounds did not turn out to be the potent inhibitors of
IMPDH. Further investigations related to the involvement of other
targets pertaining to the observations of the present investigation
may be helpful to the researchers working in the IMPDH field.
Declaration of interest
There is no conflict of interest.
Conclusions
References
In this study, systematic investigations of the SAR of 3 in order to
discover potent and novel IMPDH II inhibitors were carried out.
A five-point pharmacophore model based on structurally diverse
and potent inhibitors was used to guide the design of new
molecules. The information about the key binding interactions of
these inhibitors obtained from docking analyses was utilized as
well in the design process. Initially, Part A was fixed as 4-cyano-
3-methoxyphenyl or 3,4-dimethoxyphenyl substructures which
were in agreement with the proposed pharmacophore. In our
meticulous attempts to replace the urea (Part B), N-cyanoguani-
dine was identified as a suitable alternative from among the
various substructures tried. Having fixed Parts A and B, further
investigations were centered on doing suitable Part C modifica-
tions. Of the several Part C designs that were tried, none exhibited
promising enzyme inhibitory activity. Hence, it was concluded
that the structural requirements for IMPDH II inhibition were
strict and the binding space was very restrictive in terms of
tolerating any chemical modifications.
1. Crabtree GW, Henderson JF. Rate-limiting steps in the interconver-
sion of purine ribonucleotides in Ehrlich ascites tumor cells in vitro.
Cancer Res 1971;31:985–91.
2. Jackson RC, Weber G, Morris HP. IMP dehydrogenase, an
enzyme linked with proliferation and malignancy. Nature 1975;
256:331–3.
3. Franchetti P, Grifantini M. Nucleoside and non-nucleoside IMP
dehydrogenase inhibitors as antitumor and antiviral agents. Curr
Med Chem 1999;6:599–614.
4. Wu JC. Mycophenolate mofetil: molecular mechanism of action.
Perspec Drug Discov Des 1994;2:185–204.
5. Natsumeda Y, Carr SF. Human type I and II IMP dehydrogenases as
drug targets. Ann NY Acad Sci 1993;696:88–93.
6. Sintchak MD, Fleming MA, Futer O, et al. Structure and
mechanism of inosine monophosphate dehydrogenase in complex
with the immunosuppressant mycophenolic acid. Cell 1996;85:
921–30.
7. Sollinger HW. Mycophenolate mofetil for the prevention of acute
rejection in primary cadaveric renal allograft recipients (US Renal
Transplant Mycophenolate Mofetil Study Group). Transplantation
1995;60:225–32.
8. Sievers TM, Rossi SJ, Ghobrial RM, et al. Mycophenolate mofetil.
Pharmacotherapy 1997;17:1178–97.
Supplementary information
9. Jain J, Almquist SJ, Shlyakhter D, Harding MW. VX-497: a novel,
selective IMPDH inhibitor and immunosuppressive agent. J Pharm
Sci 2000;90:625–37.
The syntheses and the associated spectral data for compounds
14–27, 31–44, 46–60 are given. The dose–response curves for
IMPDH II inhibition and PBMC proliferation inhibition by
reference compounds 1, 2 and 3 are also included. The binding
poses of few reference compounds with IMPDH II (PDB ID 1JR1)
are shown.
10. Jain J, Almquist SJ, Heiser AD, et al. Characterization of
pharmacological efficacy of VX-148, a new, potent immunosup-
pressive inosine 50-monophosphate dehydrogenase inhibitor.
Pharmcol Exper Ther 2002;302:1272–7.
J
11. Dhar TG, Shen Z, Guo J, et al. Discovery of N-[2-[2-[[3-methoxy-4-
(5-oxazolyl)phenyl]amino]-5-oxazolyl]phenyl]-N-methyl-4-morpho-
lineacetamide as a novel and potent inhibitor of inosine monopho-
sphate dehydrogenase with excellent in vivo activity. J Med Chem
2002;45:2127–30.
Acknowledgements
The authors are thankful to Ju¨rgen Volz and colleagues for providing the
HRMS data, Katharina Rossbach and colleagues for providing HTS

DOI: 10.3109/14756366.2013.793184
Novel IMPDH inhibitors 419
¨
¨
12. Dhar TG, Shen Z, Gu HH, et al. 3-Cyanoindole-based inhibitors of 15. Schrodinger Suite 2011 was available from Schrodinger, LLC,
inosine monophosphate dehydrogenase: synthesis and initial struc- New York, NY, 2011.
ture-activity relationships. Bioorg Med Chem Lett 2003;13: 16. Phase, version 3.3, Schrodinger, LLC, New York, NY, 2011.
¨
3557–60.
13. Stamos D, Trudeau M, Bethiel S. Inhibitors of IMPDH enzyme
Patent WO0056331 (Vertex Pharma, USA).
14. Yang N, Wang J, Li J, et al. A three-dimensional pharmacophore
model for IMPDH inhibitors. Chem Biol Drug Des 2011;78:175–82.
17. Dixon SL, Smondyrev AM, Knoll EH, et al. PHASE: a
new engine for pharmacophore perception, 3D QSAR model
development, and 3D database screening. 1. Methodology
and preliminary results. J Comput Aided Mol Des 2006;20:
647–71.