Discovery and design of benzimidazolone based inhibitors of p38 MAP kinase

Article abstract of DOI:10.1016/j.bmcl.2006.09.014

A new class of benzimidazolone p38 MAP kinase inhibitors was discovered through high-throughput screening. X-ray crystallographic data of the lead molecule with p38 were used to design analogues with improved binding affinity and potency in a cell assay o

Full text of DOI:10.1016/j.bmcl.2006.09.014

Bioorganic & Medicinal Chemistry Letters 16 (2006) 6316–6320
Discovery and design of benzimidazolone based inhibitors
of p38 MAP kinase
Abdelhakim Hammach, Antonio Barbosa, Faith Corbo Gaenzler, Tazmeen Fadra,
Daniel Goldberg, Ming-Hong Hao, Rachel R. Kroe, Pingrong Liu, Kevin C. Qian,
Mark Ralph, Christopher Sarko, Fariba Soleymanzadeh and Neil Moss*
Departments of Medicinal Chemistry and Biologics and Biomolecular Sciences, Boehringer Ingelheim Pharmaceuticals,
Inc., 900 Ridgebury Road, Ridgefield, CT 06877, USA
Received 24 July 2006; revised 31 August 2006; accepted 6 September 2006
Available online 28 September 2006
Abstract—A new class of benzimidazolone p38 MAP kinase inhibitors was discovered through high-throughput screening. X-ray
crystallographic data of the lead molecule with p38 were used to design analogues with improved binding affinity and potency in
a cell assay of LPS-induced TNFa production. Herein, we report the SAR of this new class of p38 inhibitors.
Ó 2006 Elsevier Ltd. All rights reserved.
Tumor necrosis factor alpha (TNFa) and interleukin 1
(IL-1b) are pro-inflammatory cytokines implicated in
many inflammatory diseases such as rheumatoid arthri-
tis (RA), inflammatory bowel disease, and psoriasis. To
date, several anti-TNFa biologics including EnbrelÒ,
RemicadeÒ, and HumiraÒ have been approved by the
FDA for use as anti-inflammatory therapies.1 p38a is
a mitogen activated protein (MAP) kinase that plays
an essential role in the signal transduction pathways
leading to the synthesis of TNFa and IL-1b.2 Efforts
from many pharmaceutical companies have led to the
development of several p38 inhibitor clinical candidates,
including our own BIRB 796 (1).3
We were fortunate to obtain X-ray crystallographic data
of compound 2 bound to p38a (Figs. 2 and 3). Unlike
the diaryl urea based inhibitor 1, which binds to p38a
in its Phe-out conformation,3 compound 2 binds to
p38a in its Phe-in conformation like many conventional
p38a inhibitors.2d The NH and carbonyl of the benzim-
idazolone engage in H-bonding interactions with the
carbonyl of His 107 and the NH of Met 109, respective-
ly. These interactions are analogous to the H-bonding
interactions typically seen between the adenine of ATP
and a kinase. The unsubstituted phenyl group of the
benzophenone portion of 2 occupies the so-called kinase
specificity pocket (KSP) of p38a. The combination of
the molecular potency of 2 together with the structural
information from the X-ray structure encouraged us to
initiate an SAR investigation of this class of p38
inhibitors.
As part of our ongoing efforts toward discovering alter-
native classes of p38a inhibitors that could be developed
as backups to BIRB 796 (1), we conducted a high-
throughput screening campaign against p38a and identi-
fied the benzimidazolone 2 as potentially interesting
starting point (Fig. 1). Compound 2 displayed an IC50
in our high-throughput kinase assay of 130 nM and a
thermal denaturation temperature (Tm)4 with p38a of
50 °C. However, compound 2 did not inhibit TNFa pro-
duction from LPS-stimulated THP-1 cells (Table 1).
The benzimidazolones 16–26 shown in Table 1 were syn-
thesized from one of the 5-substituted-2-halo-nitrophen-
yl derivatives 10, 11, or 12 (representative examples
shown in Scheme 1). Compounds 11 and 12 could be
obtained in good yields via a copper mediated coupling
of phenyl boronic acid 7 with 3-nitro-4-fluoroaniline 8
or with 3-nitro-4-chlorophenol 9, respectively.5 Thus
as shown, reaction of 10–12 with cyclohexylmethyl-
amine provided the 4-amino-3-nitro derivatives 13–15.
Subsequent reduction of the nitro group and condensa-
tion of the resultant diamine with carbonyldiimidazole
Keywords: p38 MAP kinase; Inhibitor; Benzimidazolone.
*
Corresponding author. Tel.: +1 203 798 5101; fax: +1 203 791
6072; e-mail: nmoss@rdg.boehringer-ingelheim.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.09.014

A. Hammach et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6316–6320
6317
O
N
NH2
HN
O
N
O
N
O
O
N
N
N
N
H
H
H
1 (BIRB 796)
2
Figure 1. Structures of BIRB 796 (1) and the lead molecule 2.
Table 1. Effect of the linker (X) and N1-substituents (R) on the
binding affinity to p38a as measured by thermal denaturation
temperature (Tm) and on the inhibition of TNFa from LPS stimulated
THP-1 cells
R
N
O
N
X
H
a
Entry
X
R
Tm (°C)
Inhibition
of TNFa
from THP-1
b
Figure 2. X-ray structure of 2 (green cylinder) complexed with p38a
(Thr 106 to Met 109 as cyan cylinders) highlighting H-bond interac-
tions between 2 and His 107 C@O and Met 109 NH.
cells EC50
(nM)
2
–CO– 4-Guanylbenzyl–
–CO– Ph-CH2–
50.3 0.3 >10,000
3
52.9 0.4
51.8
49.1
1000
880
4
–CO– Cyclohexyl-CH2–
–CO– CH3–
–CO– H–
Triphosphate
binding
5
>5000
>5000
270
6
48.0
region
16
17
21
22
23
24
25
26
–NH– Cyclohexyl-CH2–
Cyclohexyl-CH2–
54.0 0.3
53.2 0.4
55.3
O
NH2
–O–
–NH– t-Bu–
250
74
HN
Adenine
binding
region
HN
Asp168
–NH– NH2CH2(CH3)2C– 48.6 0.1
3400
150
–NH– Cyclopentyl–
–NH– 3-Pyrrolidinyl–
–NH– 2,6-di-F-Benzyl–
–NH– Ph-CH2–
55.3
O
OH
49.6 0.1
55.2
54.2 0.2
1100
220
1
O
6
N
320
O
O
a Values with errors are from two experiments (range/2).
b Values are from one or two experiments (reproducibility in this assay
observed with many compounds typically within 50%).
N
H
HN
O
Met 109
Kinase specificity
pocket (KSP)
NH
yielded the benzimidazolones 4, 16, and 17. The synthe-
sis of compounds 22 and 24 required an additional Boc
deprotection step to liberate the free amine.
His 107
Figure 3. Representation of binding interactions observed from the
complex of compound 2 with p38a.
Scheme 2 depicts the general approach used to function-
alize the 6-position of the benzimidazolone core (synthe-
sis of compound 36 used as an example). Copper
mediated coupling of 2,4-difluoro-5-nitroaniline 18 with
phenylboronic acid 7 afforded the difluoroaniline deriv-
ative 19. Sequential and selective nucleophilic aromatic
substitutions of the two fluorine atoms were achieved
by displacement of the 4-fluoro with cyclohexylmethyl-
amine at room temperature, followed by displacement
of the 2-fluoro with N-methylpiperazine at 100 °C, to

6318
A. Hammach et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6316–6320
TNFa production from LPS-stimulated THP-1 cells.
Z
While the saturated analogue 4 proved equipotent to
compound 3, truncation to the methyl or hydrogen ana-
logues 5 and 6 resulted in a substantial decrease of
potency and a complete loss of binding, respectively
(compound 6 has a Tm equal to apo p38a4).
z
B(OH)2
+
i
L
NO2
NO2
Y
8; Y = NH2, Z = F
9; Y = OH, Z= Cl
10; L = CO,Z =Cl
11; L = NH,Z =F
12; L = O, Z = Cl
7
X-ray co-crystallographic data of 2 also showed that the
oxygen atom of the benzophenone did not partake in
any binding interactions with the protein. However, this
carbonyl group does help link two key pharmacophores,
the hydrogen bond donor/acceptor motif of the benzim-
idazolone, and the phenyl group which binds in the
KSP. A comparison of the calculated relative energies
of the bound conformation and the ground state confor-
mation around the carbonyl region suggested the bound
state conformation to be disfavored by about 1.5 kcal.
We hypothesized that inhibitors with smaller linkers
such as O or NH linker would decrease this energy dif-
ference and thus be more potent. Indeed, replacement of
this carbonyl in 4 by an amino or an ether linker (com-
pounds 16 and 17) resulted in an increase of both molec-
ular and cellular potency.
NH
ii
iii, iv
N
O
L
NO2
N
L
H
13; L = CO
14; L = NH
15; L = O
4; L = CO
16; L = NH
17; L = O
Scheme 1. Reagents and conditions: (i) for 11 and 12, Cu(OAc)2,
NEt3, CH2Cl2, rt, 48 h, 61 and 54%, respectively; (ii) Hunig’s Base,
DMF, cyclohexylmethylamine, 100 °C, 24 h, 59–73%; (iii) 10% Pd/C,
H2, EtOAc, rt, 16 h, 63–82%; (iv) 1,10-carbonyldiimidazole, THF, 12 h,
65–78%.
Further explorations off N1, now using compound 16 as
a frame of reference, indicated a tolerance to various
lipophilic groups (21, 23, 25, and 26). However, modifi-
cation of the N1 group with an amine function resulted
in a substantial decrease in potency (cf. 21 and 22; 23
and 24), highlighting the incompatibility of such polar
substituents in the hydrophobic cleft leading up to the
solvent front.
F
F
B(OH)2
F
F
i
+
HN
NO2
H2N
NO2
19
18
7
Small substitutions on the phenyl ring of 16 (group that
interacts with the KSP) were found to modestly improve
molecular and cellular potency (Table 2). The 2,4-diflu-
oro analogue 30 offered the best overall improvement
increasing Tm by 2.8 °C and decreasing the EC50 in the
TNFa cell assay by around 3- to 4-fold. In contrast,
the 3-position appeared less tolerant to substitution
(31–33).
N
N
N
NH
N
iv, v
N
ii, iii
O
HN
NO2
N
HN
H
20
36
With the aim of improving both the binding and the
physicochemical properties of this class of p38a inhibi-
tors, we investigated whether we could modify our com-
pounds to take advantage of a potential interaction with
Asp 168, located in the phosphate binding domain. The
introduction of appropriately positioned basic substitu-
ents to gain interactions within the phosphate binding
region has been reported to improve inhibitor potency
for certain kinases.6 Based on the observed binding con-
formation, we felt the 6-position of the benzimidazolone
core offered the best directionality to position substitu-
tions within binding distance to Asp 168 (Fig. 3). Unfor-
tunately, introduction of either a piperazinyl (35 and 37)
or the more flexible N-methyl-N0,N0-dimethylaminoeth-
yl (38) led to a slight decrease in Tm (Table 2). Neverthe-
less, incorporation of a basic amine did significantly
improve inhibitor aqueous solubility (5.7 lM for 36 vs
0.2 lM for 16 at pH 7.4).
Scheme 2. Reagents and conditions: (i) Cu(OAc)2, NEt3, CH2Cl2, rt,
48 h, 91%; (ii) cyclohexylmethylamine, DMF, rt, 14 h, 90%; (iii)
N-methylpiperazine, Hunig’s Base, DMF, 100 °C, 24 h, 64%; (iv) 10%
Pd/C, H2, EtOAc, rt, 16 h, 75%; (v) 1,10-carbonyldiimidazole, THF,
16 h, 65%.
provide the tetrasubstituted nitrophenyl derivative 20.
Reduction of the nitro group, followed by condensation
with carbonyldiimidazole, provided the 5,6-diaminosub-
stituted benzimidazolone 36. Compounds 35–38 were
made in a similar manner.
X-ray crystallographic data of compound 2 bound to
p38a showed that the benzyl guanidine substituent lay
in a hydrophobic cleft leading up to the solvent front.
However, the guanidine group itself was not involved
in any particular interactions with the protein. Removal
of this group resulted in an increase of 2.5 °C in the ther-
mal denaturation assay (compound 3, Table 1). More-
over, compound 3 proved efficacious in inhibiting
Interestingly, X-ray crystallographic data of compound
38 with p38a show the dimethylamino group situated

A. Hammach et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6316–6320
6319
Table 2. Effect of the substituents (X)n and R on the binding affinity to p38a as measured by thermal denaturation temperature (Tm) and on the
inhibition of TNFa from LPS stimulated THP-1 cells
R
N
O
N
H
HN
(X)n
a
Entry
(X)n
R
Tm (°C)
Inhibition of TNFa from
THP-1 cells EC50 (nM)
b
16
27
28
29
30
31
32
33
34
35
36
37
38
H
2-F
H
54.0 0.3
55.1 0.1
52.1 0.3
54.6 0.2
56.8 0.7
51.5 0.3
49.6 0.1
48.5 1.2
55
180
66
H
3-F
4-F
H
190
74
H
2,4-di-F
3,4-di-F
3-CF3
3,5-di-Cl
2-CH3
H
H
46
420
H
H
2300
>10,000
64
H
H
Piperazinyl
52.8 0.2
53.4 0.2
53.2
150
150
H
4-CH3-piperazinyl
4-CH3-piperazinyl
N-Methyl-N0,N0-di-CH3-aminoethyl
2-CH3
2-CH3
180
150
52.8
a Values with errors are from two experiments (range/2).
b Values are from one or two experiments (reproducibility in this assay observed with many compounds typically within 50%).
within hydrogen-bonding distance from Asp 168
(Fig. 4). A comparison of the X-ray co-crystal structures
of compounds 38 and 30 provides a potential rationale
for the lack of observed improvement in molecular
potency for 38 (Fig. 4). The co-crystal structure of com-
pound 30 reveals an edge to face interaction between the
aromatic side chain of Tyr 379 and the aromatic core of
the benzimidazolone. In the co-crystal structure of com-
pound 38, neither Tyr 379 nor the entire glycine-rich
loop to which this residue is attached is observed.
Conceivably, in order to accommodate the N-methyl-
N0,N0-dimethylaminoethyl group of compound 38, Tyr
379 and the glycine-rich loop must be displaced, thereby
losing the beneficial edge to face interaction between Tyr
379 and the aromatic core of the benzimidazolone.
Compound 30 possesses a very good overall kinase
selectivity profile showing negligible inhibition of the
following kinases when tested at 10 lM: Btk, Eck,
EGFR, FGFR3, Hek, HGFR, IGF1R, IR, Itk, JAK3,
Lyn, Syk, Tie-2, TXK, VEGFR1, MKK1, ERK2,
JNK1, p38c, p38d, MAPKAP-K1a, MAPKAP-K2,
MSK1, PRAK, PKA, PKCa, PDK1, PKBoPH, SGK,
S6K1, GSK3b, ROCK-II, AMPK, CHK1, CK2, PHOS.
KINASE, LCK, CSK, CDK2/cyclin A, CK1, DYRK
1A, PP2A, NEK6. Compound 30 shows similar inhibi-
tion of p38b at 10 lM as seen for p38a (dose down
not done). It is noteworthy that compound 30 shows
selectivity even against Jnk2a2, a kinase frequently
inhibited by other classes of p38a inhibitors. Compound
30 inhibits Jnk2a2 with an IC50 of 5.5 lM, whereas
BIRB 796 (1) inhibits this enzyme with an IC50 of
0.006 lM.
Tyr 379
Asp 168
Met 109
In conclusion, X-ray crystallographic data of 2 with
p38a were exploited to improve the binding of the benz-
imidazolone class of p38a inhibitors culminating in
compound 30. Most notable were the improvements in
binding and cellular potency resulting from replacing
the carbonyl of the benzophenone with an amino-linker
and optimizing the lipophilic interactions within the
kinase specificity pocket.
His 107
Figure 4. Overlay of p38a co-complexes of compounds 38 (green
cylinder with residues Asp 168, His 107, and Met 109) and 30 (yellow
cylinder with residues Asp 168, His 107, Met 109, and Tyr 379). The
˚
salt bridge (2.9 A) between Asp 168 and the dimethylamino group of
38 highlighted.

6320
A. Hammach et al. / Bioorg. Med. Chem. Lett. 16 (2006) 6316–6320
The IC50s for RO3201195, SB203580, VX-745, and
References and notes
L779,450 have been disclosed in the following references,
respectively. Goldstein, D. M.; Gabriel, T. Curr. Top. Med.
Chem. 2005, 5, 1017; (b) Lee, J. C.; Laydon, J. T.;
McDonnell, P. C.; Gallageher, T. F.; Kumar, S.; Green,
D.; Blumenthal, M. J.; Heys, J. R.; Landvatter, S. W.;
Strickler, J. E.; McLaughlin, M. M.; Siemens, I. R.; Fisher,
S. M.; Livi, G. P.; White, J. R.; Adams, J. L.; Young, P. R.
Nature 1994, 372, 739; (c) Salutoro, F. G. In 27th National
Medicinal Chemistry Symposium, Kansa City, MO, June
13–17, 2000; (d) Liverton, N. J.; Butcher, J. W.; Claiborne,
C. F.; Claremon, D. A.; Libby, B. E.; Nguyen, K. T.;
Pitzenberger, S. M.; Selnick, H. G.; Smith, G. R.; Tebben,
A.; Vacca, J. P.; Varga, S. L.; Agarwal, L.; Dancheck, K.;
Forsyth, A. J.; Fletcher, D. S.; Frantz, B.; Hanlon, W. A.;
Harper, C. F.; Hofsess, S. J.; Kostura, M.; Lin, J.; Luell, S.;
O’Neill, E. A.; Orevillo, C. J.; Pang, M.; Parsons, J.;
Rolando, A.; Sahly, Y.; Visco, D. M.; O’Keefe, S. J. J.
Med. Chem. 1999, 42, 2180.
1. (a) Klinkhoff, A. Drugs 2004, 64, 1267; (b) Barry, J.; Kirby,
B. Expert Opin. Biol. Ther. 2004, 4, 975.
2. (a) Saklatvala, J. Curr. Opin. Pharm. 2004, 4, 372; (b)
Kumar, S.; Boehm, J.; Lee, J. C. Nat. Rev. Drug Discov.
2003, 2, 717; (c) Ono, K.; Han, J. Cell. Signalling 2000, 12,
1; (d) Jakckson, P. F.; Bullington, J. L. Curr. Top. Med.
Chem. 2002, 2, 1009; (e) Cirillo, P.; Pargellis, C.; Regan, J.
Curr. Top. Med. Chem. 2002, 2, 1021.
3. (a) Regan, J.; Breitfelder, S.; Cirillo, P.; Gilmore, T.;
Graham, A. G.; Hickey, E. R.; Klaus, B.; Madwed, J.;
Moriak, M.; Moss, N.; Pargellis, C. A.; Pav, S.; Proto, A.;
Swinamer, A.; Tong, L.; Torcellini, C. J. Med. Chem. 2002,
45, 2994; (b) Regan, J.; Capolino, A.; Cirillo, P. F.;
Gilmore, T.; Graham, A. G.; Hickey, E.; Kroe, R. R.;
Madwed, J.; Moriak, M.; Nelson, R.; Pargellis, C. A.;
Swinamer, A.; Torcellini, C.; Tsang, M.; Moss, N. J. Med.
Chem. 2003, 46, 4676.
5. (a) Chan, D. M. T.; Monaco, K. L.; Wang, R. P.; Winters,
M. P. Tetrahedron Lett. 1998, 39, 2933; (b) Lam, P. Y. S.;
Clark, C. G.; Saubern, S.; Adams, J.; Winters, M. P.; Chan,
D. M. T.; Combs, A. Tetrahedron Lett. 1998, 39, 2941; (c)
Evans, D. A.; Katz, J. L.; West, T. R. Tetrahedron Lett.
1998, 39, 2937; (d) Lam, P. Y. S.; Clark, C. G.; Saubern, S.;
Adams, J.; Averill, K.; Chan, D. M. T.; Combs, A. P.
Synlett 2000, 674.
6. (a) Revesz, L.; Blum, E.; Di Padova, F. E.; Buhl,
T.; Feifel, R.; Gram, H.; Hiestand, P.; Manning,
U.; Rucklin, G. Bioorg. Med. Chem. Lett. 2004, 14,
3601; (b) Revesz, L.; Di Padova, F. E.; Buhl, T.;
Feifel, R.; Gram, H.; Hiestand, P.; Manning, U.;
Wolf, R.; Zimmerlin, A. G. Bioorg. Med. Chem.
Lett. 2002, 12, 2109; (c) Moffat, D.; Davis, P.;
Hutchings, M.; Davis, J.; Berg, D.; Batchelor, M.;
Johnson, J.; O’Connell, J.; Martin, R.; Carabbe, T.;
Delgado, J.; Perry, M. Bioorg. Med. Chem. Lett.
1999, 9, 3351.
4. Kroe, R. R.; Regan, J.; Proto, A.; Peet, G. W.; Roy, T.;
Dickert, L.; Fuschetto, N.; Pargellis, C. A.; Ingraham, R.
H. J. Med. Chem. 2003, 46, 4669, The formula used in this
reference to convert thermal denaturation temperatures
(Tm) to binding affinities (Kd) applies to a Phe-out class of
p38 inhibitors and consequently cannot be accurately
applied to the Phe-in class of inhibitors in this paper. Thus
to allow the reader to approximate conversions of Tm
values to kinase assay IC50s, we provide the Tm values
determined for a range of known Phe-in p38 inhibitors.
Reported Kinase
assay IC50 (nM)
Determined Tm
Compound
(oC)
Roche
RO 3201195
SmithKline
SB203580
Vertex
VX-745
Merck
L 779,450
700
40
50.7
54.9
55.9
59.5
20
0.1