Phenoxypyrimidine inhibitors of p38α kinase: Synthesis and statistical evaluation of the p38 inhibitory potencies of a series of 1-(piperidin-4-yl)-4-(4-fluorophenyl)-5-(2-phenoxypyrimidin-4-yl) imidazoles

Article abstract of DOI:10.1016/S0960-894X(01)00163-9

As a continuation of our work with 1,4,5 substituted imidazole inhibitors of p38α, we report a series of 1-(4-piperidinyl)-4-(4-fluorophenyl)-5-(2-phenoxy-4-pyrimidinyl) imidazoles related to 7. The compounds have IC₅₀'s for inhibition of p38α ranging from 6.0 to 650 nM. Statistical analysis of the p38α inhibitor potencies shows a correlation of IC₅₀'s with the electron donating strength of low molecular weight substituents.

Full text of DOI:10.1016/S0960-894X(01)00163-9

Bioorganic & Medicinal Chemistry Letters 11 (2001) 1123–1126
Phenoxypyrimidine Inhibitors of p38ꢀ Kinase:
Synthesis and Statistical Evaluation of the p38 Inhibitory
Potencies of a Series of 1-(Piperidin-4-yl)-4-(4-fluorophenyl)-5-
(2-phenoxypyrimidin-4-yl) Imidazoles
Jeffrey C. Boehm,a,* Michael J. Bower,b Timothy F. Gallagher,a Shouki Kassis,c
d
StephenR. Johnson and Jerry L. Adamsa
aDepartment of Medicinal Chemistry, GlaxoSmithKline Pharmaceuticals, 709 Swedeland Road, Kingof Prussia, PA 19406, USA
bDepartment of Physical and Structural Chemistry, GlaxoSmithKline Pharmaceuticals, 709 Swedeland Road,
Kingof Prussia, PA 19406, USA
cDepartment of Bone and Cartilage Biology, GlaxoSmithKline Pharmaceuticals, 709 Swedeland Road,
Kingof Prussia, PA 19406, USA
dDepartment of Cheminformatics, GlaxoSmithKline Pharmaceuticals, 709 Swedeland Road, Kingof Prussia, PA 19406, USA
Received 2 January 2001; accepted 20 February 2001
Abstract—As a continuation of our work with 1,4,5 substituted imidazole inhibitors of p38a, we report a series of 1-(4-piperidinyl)-
4-(4-fluorophenyl)-5-(2-phenoxy-4-pyrimidinyl) imidazoles related to 7. The compounds have IC50’s for inhibition of p38a ranging
from 6.0 to 650 nM. Statistical analysis of the p38a inhibitor potencies shows a correlation of IC50’s with the electron donating
strength of low molecular weight substituents. # 2001 Elsevier Science Ltd. All rights reserved.
Previously, we described a series of 4-aryl-5-(pyridin-4-yl)
imidazole p38 inhibitors containing substitution at the
imidazole N-1, and exemplified by 1.1 Later, we found
that replacement of the pyridyl group with a variety of
substituted pyrimidinyl rings afforded more potent
inhibitors (2–4).2 Subsequent work seeking to enhance
the in vitro and in vivo potency of this series led to the
development of analogues such as 5, an d6, containing a
piperidinyl at the imidazole N-1.3
Another variation of this general structure is the phe-
noxypyrimidine 7, a potent inhibitor of p38a. For com-
parison, the prototypical p38a inhibitor SB-203580
inhibits the enzyme with an IC50 of 48 nM while the
corresponding value for 7 was 19 nM. In the present
report we describe the effects of substitutiononthe
phenoxy substituent of 7 onp38 a inhibition.
Solution-phase synthesis of 1-(piperidin-4-yl)-4-(4-
fluorophenyl)-5-(2-aryloxypyrimidin-4-yl)
imidazoles
proceeded through either the methyl or propyl sulfides
9b,c4 by procedures similar to those described pre-
viously (Scheme 1).5 As depicted, slight modificationof
the solution-phase methodology permitted solid-phase
synthesis.6 About half of the compounds in Table 1
were prepared by the resin-based approach.
*Corresponding author. Tel.: +1-610-270-6595; fax: +1-610-270-
4490; e-mail: jeffrey_c_boehm@sbphrd.com
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00163-9

1124
J. C. Boehm et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1123–1126
Scheme 1. (a) Neat, 100 ꢁC; (b) thiourea, NaOMe; (c) alkyl halide (methyl iodide, propyl bromide); (d) Merrifield resin; (e) 3 N HCl 23 ꢁC then
Na2CO3; (f) TFA, ꢀ; (g) 4-amino-1-Boc-piperidine, CH2Cl2; (h) 11b or 11c, K2CO3, DMF; (i) 11d, TBD, CH2Cl2; (j) oxone, THF, H2O; (k) tetra-
butylammonium oxone, CH2Cl2; (l) MCPBA; (m) ArOH, NaH, THF thenTFA thenOH ꢀ; (n) ArOH, TMS2Nꢀ Na+, THF thenTFA.
It has beenreported that p38 a inhibition is improved in
a series of 2-(4-fluorophenyl)-3-(pyridin-4-yl) pyrrolo-
pyridines when electron rich substituents were placed on
the core heterocycle pyridine.7 This is likely the result of
increased electron density on the 3-pyridyl nitrogen. Co-
crystallographic studies of the imidazole p38 inhibitors
have previously shownthat this pyridyl nitrogenforms
a hydrogen bond to the ATP binding region of p38 and
is a key interaction required for potent inhibition by the
pyridyl imidazole class of p38 inhibitors.8 With this in
mind, it seemed that the potency of the analogues in
Table 1 might correlate with the electrondensity of
phenoxy substituents.
Further analysis was restricted to the low molecular
weight compounds in Table 1 (Mr < 460). A pool of 62
structural features was analyzed using a stochastic opti-
mizationmethod. From these, three structural features
were selected which afforded a multiple linear regression
model which is statistically sound, with R=0.92 and
F=36.6 with the removal of one outlier.11 The para-
meters are the sum of the squared partial atomic
charges (SSAC), the autocorrelated topological distance
three lone-pair electronegativity, and the solvent-acces-
sible surface area of atoms with a negative partial charge.
The sum of the squared partial atomic charges (SSAC)
of the structure is calculated over all the atoms inthe
P
As a simple test of this idea, the Hammet spara for the
4-phenoxy substituted analogues were plotted against
log IC50.9 Consistent with the hypothesis, a correlation
was obtained (R=0.64). If a single high molecular
weight analogue (15ao) is omitted, the correlation
improves (R=0.78) (Fig. 1).
compound,
ðq2i Þ. Of the three structural features,
SSAC has the least effect onthe IC 50’s. The auto-
correlated12,13 electronegativity of lone pair electrons
feature is calculated as:
X
AENLPð3Þ ¼
PiPj
This improvement is understandable on the basis of a
10
i;j
comparisonof the MacroModel
structural searches
for 7 and 15ao, in which the low energy binding con-
formations prevalent for 7 differ from those determined
for 15ao (Fig. 2). The latter inhibitor appears to favor
conformations in which the phenoxy p-stacks over the
4-fluorophenyl rather than over the piperidine. It is
likely that the preferred binding conformation is the low
energy conformation for 7 rather thanthat of 15ao. If
this conformational effect is generally true, then it may
be expected that the IC50’s of analogues containing lar-
ger, aromatic substituents will not correlate with elec-
tronic effects. This prediction is supported by the
observationthat the feature which correlates most with
the log(1/IC50) data for all the compounds in Table 1 is
the molecular weight (R=ꢀ0.75).
Figure 1. p38 inhibition correlates with the electron donating strength
of the low Mr 4-phenoxy substituents.

J. C. Boehm et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1123–1126
Table 1. IC50 for p38 inhibition by phenoxypyrimidines (15)
1125
9
Compound
R
IC50 Mr spara
(nM)
15a
15b
15c
15d
15e
2,4-Dimethyl
2-Hydroxy
4-Ethyl
2,5-Dimethyl
2-Fluoro
4-Hydroxy
6.0 443
6.6 431
7.4 443 ꢀ0.15
10.7 443
13.0 433
15f
14.0 431 ꢀ0.37
15g
15h
15i
2-Methyl
15.0 429
15.2 443
2,3-Dimethyl
4-Methoxy
3-Methoxy
Figure 2. Comparisonof low energy (MacroModel) structures for
(A) and 15ao (B).
7
15.7 445 ꢀ0.27
15j
18.7 445
15k
7
15l
15m
15n
15o
15p
15q
15r
15s
15t
15u
15v
15w
15x
15y
15z
4-Methyl
H
3-Hydroxy
19.0 429 ꢀ0.17
19.2 415
20.0 431
21.6 443
23.0 459
25.0 445
34.3 443
0
3,4-Dimethyl
3,4-Methylenedioxy
2-Methoxy
2,6-Dimethyl
4-Isopropyl
3,5-Dimethyl
3-Carboxamidyl
4-Phenyl
electron withdrawing groups are on the least potent
inhibitors.
These features represent a correlative relationship that is
not necessarily causal. However, the statistical analysis
seems to support the generalization that small electron
rich substituents ortho to the phenoxy oxygen have the
greatest effect on boosting potency. Conversely, large
41.0 457 ꢀ0.15
41.0 443
56.0 458
60.0 491 ꢀ0.01
3-N-Methylcarboxamidyl
4-tert-Butyl
4-Carboxy
75.0 472
79.0 471 ꢀ0.20
82.0 459 0.45
85.4 487 0.45
89.7 458 0.36
95.0 433
electron deficient substituents para to the phenoxy oxy-
14
4-Carboxyethyl
4-Carboxamidyl
3-Fluoro
genhave a deleterious effect.
The results canbe par-
tially rationalized as the consequence of changes in the
H bond acceptor properties of the pyrimidine N-4.
15aa
15ab
15ac
15ad
15ae
15af
15ag
15ah
15ai
15aj
15ak
15al
15am
15an
15ao
15ap
15aq
15ar
15as
4-Chloro
4-Carboxymethyl
4-Phenoxy
3-Trifluormethyl
2-Acetamido
2-Propionamido
3,4-Dichloro
4-Carboxypropyl
4-Cyano
114 449 0.23
119 473 0.45
130 507
132 483
134 472
151 486
165 484
174 501
177 440 0.66
189 500
209 451
The best p38a inhibitors in Table 1 have potencies less
than 10 nM. This level of activity is an improvement
over the p38 inhibitors reported by our group up to this
time.15
3-N-Isopropylcarboxamidyl
3,4-Difluoro
References and Notes
4-Trifluoromethyl
4-Fluoro
3-N,N-Dimethylcarboxamidyl
4-Benzyloxy
3-(Piperazin-1-yl)carboxamidyl
215 483 0.54
300 433 0.06
312 486
1. Boehm, J. C.; Smietana, J. M.; Sorenson, M. E.; Garigipati,
R. S.; Gallagher, T. F.; Sheldrake, P. L.; Bradbeer, J.; Badger,
A. M.; Laydon, J. T.; Lee, J. C.; Hillegass, D. E.; Griswold,
D. E.; Breton, J. J.; Chabot-Fletcher, M. C.; Adams, J. L. J.
Med. Chem. 1996, 39, 3929.
2. Adams, J. L.; Boehm, J. C.; Kassis, S.; Gorycki, P. D.;
Webb, E. F.; Hall, R.; Sorenson, M.; Lee, J. C.; Ayrton, A.;
Griswold, D. E.; Gallagher, T. F. Bioorg. Med. Chem. Lett.
1998, 8, 3111.
317 521 ꢀ0.23
406 527
3-(Piperazin-1-yl)carboxamidomethyl 453 541
3-(Piperidin-1-yl)carboxamidyl
4-Methylsulfonyl
599 526
650 493 0.72
3. (a) Jackson, R. J.; Bolognese, B.; Hillegass, L.; Kassis, S.;
Adams, J.; Griswold, D. E.; Winkler, J. D. J. Pharmacol. Exp.
Ther. 1998, 284, 687. (b) Adams, J. L.; Gallagher, T. F.;
Boehm, J. C.; Kassis, S.; Gorycki, P. D.; Gum, R. J.; Webb, E.
F.; Sorenson, M. E.; Smietana, J. M.; Garigapati, R. S.; Hall,
R. F.; Aryton, A.; Badger, A.; Griswold, D. E.; Young, P. R.;
Lee, J. C. Book of Abstracts; XVth International Medicinal
Chemistry Symposium; Cambridge, UK, September 1998.
4. Bredereck, H.; Sell, R.; Effenberger, F. Chem. Ber. 1964,
97, 3407.
where Pi is the atom-based lone-pair electronegativity of
atom i. Atoms i and j are atom pairs separated by three
bonds. This feature encodes how the lone-pair electro-
negativity is distributed across the molecule. For this
data, substitutionat the 2 positionof the phenoxy has a
greater effect onpotency thansubstitutionat the 3 or 4
positions. The presence of the partial negative surface
area is likely encoding the presence of electron-with-
drawing groups that, for this data, have larger exposed
surfaces of heteroatoms. The partial negative surface
area has the highest correlation, with the log(1/IC50)
p38 inhibition data (R=ꢀ0.76). The solvent-exposed
5. Sisko, J.; Kassick, A. J.; Mellinger, M.; Filan, J. J.; Allen,
A.; Olsen, M. A. J. Org. Chem. 2000, 65, 1516.
6. Gallagher, T. F.; Boehm, J. C.; Osifo, I. K.; Kassis, J.; Lee,
J. C.; Aryton, A.; Griswold, D. E.; Adams, J. L.; Wang, Z.;

1126
J. C. Boehm et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1123–1126
Goldsmith, E. J. Book of Abstracts, 215th ACS National
Meeting, Dallas, 29 March–2 April 1998.
7. Henry, J. R.; Rupert, K. C.; Dodd, J. H.; Turchi, I. J.;
Wadsworth, S. A.; Cavender, D. E.; Fahmy, B.; Olini, G. C.;
Davis, J. E.; Pellegrino-Gensey, J. L.; Schafer, P. H.; Sie-
kierka, J. J. J. Med. Chem. 1998, 41, 4196.
8. (a) Tong, L.; Pav, S.; White, D. M.; Rogers, S.; Crane,
K. M.; Cywin, C. L.; Brown, M. L.; Pargellis, C. A. Nat.
Struct. Biol. 1997, 4, 31116. (b) Wilson, K. P.; McCaffrey,
P. G.; Hsiao, K.; Pazhinisamy, S.; Galullo, V.; Bemis, G. W.;
Fitzgibbon, M. J.; Caron, P. R.; Murcko, M. A.; Su, M. S. S.
Chem. Biol. 1997, 4, 423. (c) Wang, Z.; Canagarajah, B. J.;
Boehm, J. C.; Kassis, S.; Cobb, M. H.; Young, P. R.; Abdel-
Meguid, S.; Adams, J. L.; Goldsmith, E. J. Structure (Lon-
don) 1998, 6, 117.
high potency based on the inhibition of other 2-methyl sub-
stituted analogues (15a, 15d, 15g, an d15h); the model predicts
a value in line with this expectation. In actuality, 15p is sub-
stantially less potent. We hypothesize that the 2,6-substitution
partially occludes the 4-pyrimidinyl N thus making H-bond
formationless favorable. As it is the olny 2,6-substituted
example inthis data set, this hypothesis is only weakly sup-
ported.
12. Moreauo, G.; Broto, P. Nouv. J. Chim. 1980, 4, 359.
13. Wagener, M.; Sadowski, J.; Gasteiger, J. J. Am. Chem.
Soc. 1995, 117, 7769.
14. Of interest is a comparison of 15e, 2-fluoro (13 nM), 15z,
3-fluoro (95 nM), and 15am, 4-fluoro (300 nM). This unex-
pected steric effect may illustrate steric differences between
electron donating conjugative effects and electron withdrawing
inductive effects in aryl fluorides.
15. Gallagher, T. F.; Seibel, G. L.; Kassis, S.; Laydon, J. T.;
Blumenthal, M. J.; Lee, J. C.; Lee, D.; Boehm, J. C.; Fier-
Thompson, S. M.; Abt, J. W.; Sorenson, M. E.; Smietana,
J. M.; Hall, R. F.; Garigapati, G. S.; Bender, P. E.; Erhard,
K. F.; Krog, A. J.; Hofmann, G. A.; Sheldrake, P. L.;
McDonnell, P. C.; Kumar, S.; Young, P. R.; Adams, J. L.
Bioorg. Med. Chem. 1997, 5, 49.
9. Hansch, C.; Leo, A.; Hoekman, D. ExploringQSAR.
Hydrophobic, Electronic, and Steric Constants; American
Chemical Society: Washington, DC, 1995.
10. Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Lis-
kamp, R.; Lipton, M.; Caufield, C.; Chang, G.; Hendrickson,
T.; Still, W. C. J. Comput. Chem. 1990, 11, 440.
11. The lone outlier in this model is the 2,6-dimethyl sub-
stituted analogue (15p). This compound is expected to have