Synthesis and SAR of new pyrrolo[2,1-f][1,2,4]triazines as potent p38α MAP kinase inhibitors

Article abstract of DOI:10.1016/j.bmcl.2008.02.067

A novel series of compounds based on the pyrrolo[2,1-f][1,2,4]triazine ring system have been identified as potent p38α MAP kinase inhibitors. The synthesis, structure-activity relationships (SAR), and in vivo activity of selected analogs from this class o

Full text of DOI:10.1016/j.bmcl.2008.02.067

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2739–2744
Synthesis and SAR of new pyrrolo[2,1-f][1,2,4]triazines
as potent p38a MAP kinase inhibitors
Stephen T. Wrobleski,a,* Shuqun Lin,a John Hynes, Jr.,a Hong Wu,a Sidney Pitt,b
Ding Ren Shen,b Rosemary Zhang,b Kathleen M. Gillooly,b David J. Shuster,b
Kim W. McIntyre,b Arthur M. Doweyko,c Kevin F. Kish,c Jeffrey A. Tredup,c
Gerald J. Duke,c John S. Sack,c Murray McKinnon,b John Dodd,a Joel C. Barrish,a
Gary L. Schievenb and Katerina Leftherisa
aDepartment of Immunology Chemistry, Bristol-Myers Squibb, Princeton, NJ 08543-4000, USA
bDepartment of Immunology Drug Discovery, Bristol-Myers Squibb, Princeton, NJ 08543-4000, USA
cDepartment of Structural Biology and Modeling, Bristol-Myers Squibb, Princeton, NJ 08543-4000, USA
Received 12 November 2007; revised 22 February 2008; accepted 27 February 2008
Available online 4 March 2008
Abstract—A novel series of compounds based on the pyrrolo[2,1-f][1,2,4]triazine ring system have been identified as potent p38a
MAP kinase inhibitors. The synthesis, structure–activity relationships (SAR), and in vivo activity of selected analogs from this class
of inhibitors are reported. Additional studies based on X-ray co-crystallography have revealed that one of the potent inhibitors from
this series binds to the DFG-out conformation of the p38a enzyme.
ꢀ 2008 Elsevier Ltd. All rights reserved.
The lack of convenient and effective treatments for
chronic debilitating inflammatory diseases such as rheu-
matoid arthritis, Crohn’s disease, and psoriasis represents
a significant unmet medical need. The recent success of the
anti-cytokine biological agents anakinra (Kineret),1 an
IL-1 receptor antagonist, etanercept (Enbrel),2 a soluble
TNF receptor fusion protein, and infliximab (Remicade)3
and adalimumab (Humira), both TNF-a monoclonal
antibodies, has demonstrated clinical benefit in the treat-
ment of inflammatory diseases.4 However, due to the well
known disadvantages common to these protein-based
therapies such as high cost and subcutaneous or intrave-
nous administration, orally active small molecules that
can effectively act as anti-cytokine agents would clearly
be of added benefit to patients.5
of four isoforms (a, b, c, and d) whereby p38a is be-
lieved to be the predominant isoform involved in the
inflammatory response.7 As a result, the development
of orally active small molecule p38a inhibitors has been
actively pursued by many researchers.8 In collaboration
with others, we previously identified and reported
substituted triaminotriazines9 and 5-cyanopyrimidines10
as novel and potent p38a inhibitors. In an ongoing
effort to identify structurally diverse analogs, we have
recently discovered a series of substituted pyrrolo[2,
1-f][1,2,4]triazines as novel inhibitors of the p38a
MAP kinase.11 This paper describes the synthesis and
preliminary structure–activity relationships (SAR) of
compounds based on the general structure depicted in
Figure 1 where modifications of the R1 and R2 substit-
uents have recently been evaluated. In addition, results
from X-ray co-crystallographic studies of an analog
bound to unphosphorylated p38a and in vivo evalua-
tion of select compounds in a murine model of acute
inflammation will be discussed.
The p38 mitogen-activated protein kinase (MAPK)
pathway has been proven to play a central role in
the regulation of the proinflammatory cytokines
TNF-a and IL-1b.6 The p38 MAPK family consists
Exploration of the C-6 SAR used analog 1 as a starting
point based on published work from our own laborato-
ries and those of others.12,13 Using the route previously
reported,12 the synthesis began with the known pyrrole
Keywords: p38; Kinase; Inflammation; Pyrrolotriazines; TNF-a; IL-1.
Corresponding author. Tel.: +1 609 252 4873; fax: +1 609 252
6601; e-mail: stephen.wrobleski@bms.com
*
0960-894X/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.02.067

2740
S. T. Wrobleski et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2739–2744
in human peripheral blood mononuclear cells
O
(PBMCs).16 With the exception of the carboxylic acid
10, the cellular potency correlated well with the enzyme
inhibition potency. The most active compounds in cells
included the ethyl amide 13 and the (S)-a-methylbenzyl-
amide 17 which were equipotent at TNF-a inhibition
(IC50 = 2 nM). Lack of cellular activity for the acid 10
was likely due to poor cell permeability.
HN
N
H
R2
5
N
R1
6
N
N
7
Figure 1. General structure of new pyrrolo[2,1-f][1,2,4]triazine based
p38a inhibitors.
To evaluate the aryl amide SAR, the C-6 position was
fixed using the (S)-a-methylbenzyl amide (Table 2). Sub-
stitution at the meta or para positions of the aryl ring
generally provided the most potent compound.
214 (Scheme 1). Deprotonation with NaH and reaction
with either O-(2,4-dinitrophenyl)hydroxylamine or
monochloramine15 followed by cyclization with form-
amide provided intermediate 3. Subsequent chlorination
using POCl3 afforded 4 which was coupled with various
functionalized anilines 5 to yield intermediate 6. Finally,
hydrolysis of the C-6 ester and amide formation under
standard conditions provided the desired analogs 7 for
initial evaluation using in vitro assays.16
The optimal substituents were found to be the 4-cyano
(27) and the 3-morpholino (28, 30) groups along with
the previously identified 3-fluoro, 5-morpholino disub-
stitution pattern (17). The unsubstituted analogs 21
and 29 and the 3-methyl substituted compound 22 had
slightly decreased potency whereas the 3,5-disubstituted
CF3 analog 25 and 2,6-dichloro analog 26 did not show
significant activity when tested up to 1 lM. In addition,
replacement of the phenyl ring with a 4-pyridyl group
provided equipotent analogs against the enzyme, albeit,
with a slightly decreased cellular activity relative to their
phenyl ring counterparts (21 vs 29 and 28 vs 30).
O
O
N
HN
N
H
N
R1
N
F
N
1
The C-6 SAR showed that the most active compounds
at inhibiting the enzyme contained either a primary or
secondary amide (11–17, 20), an ester (9), or a carbox-
ylic acid (10) (Table 1). The tertiary amide 19 and the
C-6 unsubstituted pyrrolotriazine 8 had decreased po-
tency relative to these analogs. In addition, the (S)-enan-
tiomer of the a-methylbenzyl amide (17) was nearly
50-fold more potent against p38a than the corresponding
(R)-enantiomer (18). At the time, this finding suggested
to us that the a-methylbenzyl group might be occupying
a lipophilic pocket in the p38a active site that had been
utilized by other reported p38 inhibitors containing a
similar preference for the (S)-a-methylbenzyl group over
its enantiomeric counterpart.17 This hypothesis was later
validated by X-ray co-crystallographic studies of a clo-
sely related analog to 17 (vide infra).
Replacement of the morpholine group with other het-
erocycles was also investigated in the C-6 ethyl amide
series (Table 3). Incorporation of an N-methyl pipera-
zine ring (32) or five-membered heteroaromatic groups
(33, 34) resulted in reduced enzymatic potency relative
to the morpholine substitution (31). Although the imid-
azole analog 34 and the c-lactam 27 did have respect-
able enzymatic potency in the nanomolar range, the
cellular potency for both compounds was >100 nM.
As a result, these compounds were not further inves-
tigated.
Combination of the optimal pyrrolotriazine C-6 substit-
uents and aryl amide side chain substituents provided
the most potent compounds as summarized in Table 4.
In addition, the (S)-1-methoxy-2-propan-2-yl C-6
amides 38–40 were also prepared. Although these deriv-
atives were slightly less potent than the ethyl and
a-methylbenzyl amides in most cases, they did show
Having identified a series of analogs with potent inhibi-
tory activity against the enzyme, the compounds were
subsequently evaluated for inhibition of TNF-a release
Cl
OH
CO2Et
a, b
c
N
N
EtO2C
EtO2C
EtO2C
N
N
NH
N
N
2
5
4
3
O
O
O
R1
H2N
N
R1
HN
N
HN
N
R1
H
H
H
e, f
O
N
d
N
EtO2C
N
N
R2
N
N
N
R3
6
7
Scheme 1. Synthesis of pyrrolo[2,1-f][1,2,4]triazine based p38 inhibitors. Reagents and conditions: (a) NaH, DMF then DnpONH2 or NH2Cl in
Et2O; (b) formamide, 165 ꢁC; (c) POCl3, 110 ꢁC; (d) DMF, 55 ꢁC; (e) 1 N aq NaOH, THF, 50 ꢁC; (f) R2R3NH, EDCI, HOBt, DMF.

S. T. Wrobleski et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2739–2744
2741
Table 3. Heterocycle modifications
Table 1. Pyrrolotriazine C-6 SAR
O
O
O
Het
HN
N
H
N
HN
N
H
O
N
N
6
R1
N
NH
N
N
F
N
Compound
R1
H
p38a
IC50 (nM)
TNFa IC50
(nM)
Compound Het
p38a IC50 (nM) TNFa IC50 (nM)
O
N
8
9
27
2
98
4
>1000
31
0.98a
924
123
7
45
—
N
–CO2Et
–CO2H
–C(O)NH2
10
11
12
13
14
15
16
1
3
19
14
2
32
33
34
–C(O)NHMe
–C(O)NHEt
4
N
0.42a
–C(O)NH-n-Pr
–C(O)NH-i-Pr
–C(O)NH-(S)-
1-methoxy-
5
5
2
4
N
N
6
—
7
2-propan-2-yl
–C(O)NH-(S)-
a-methylbenzyl
–C(O)NH-(R)-
a-methylbenzyl
–C(O)N(Me)2
–C(O)NHCH2
CH2N(Me)2
N
126
N
N
17
18
0.54a
23
2
48
O
35
16
633
19
20
63
10
212
18
a Ki determination.
a Ki determination.
co-complexed with purified, unphosphorylated p38a
was solved.18 The key binding interactions between 30
and the p38a enzyme are illustrated in Figure 2. Four
hydrogen bonds are apparent between the inhibitor
and the protein. These include two hydrogen bonds
from the pendant diaryl amide linker to Glu71 and
Asp168 and a hydrogen bond between the pyrrolotri-
azine N1 and Leu171 residue. In addition, a key hydro-
gen bond between the pyrrolotriazine C6-amide
carbonyl and the p38a hinge region Met109 residue is
also observed. The coplanar orientation of the C-6
amide and pyrrolotriazine ring system in the bound state
explains the decreased activity of the C6-tertiary amide
19 which most likely prefers a nonplanar orientation.
Further analysis of the co-complex reveals that the bind-
ing of 30 requires a large protein conformational change
in the conserved Asp-Phe-Gly (DFG) motif to accom-
modate the morpholino group. This protein conforma-
tion has been commonly referred to as the DFG-out
conformation19 and allows the morpholine group to oc-
cupy a large hydrophobic pocket normally occupied by
the Phe169 side chain. The presence of the morpholine
group in this lipophilic pocket lined by the hydrophobic
residues Leu74, Leu75, Val83, Ile141, Ile146, and Ile166
(not shown) explains why most analogs lacking this
group are significantly less potent. In addition, com-
pounds containing groups larger than a fluorine on
opposite sides of the phenyl ring (25, 26) are significantly
less potent (>1 lM) presumably due to unfavorable
interactions with the Glu71 side chain which is posi-
tioned on the opposite side of the phenyl ring relative
to the lipophilic pocket. While the decreased potency
of the piperazine analog 32 can easily be rationalized
Table 2. Aryl amide side chain SAR
O
R
HN
N
1
H
X
O
NH
2
N
3
N
N
Compound
R
H
X
p38a IC50 TNFa
(nM)
IC50 (nM)
21
22
23
24
25
26
27
28
17
29
30
C
C
C
C
C
C
C
C
C
N
N
19
94
52
259
313
72
3-Methyl
4-Methyl
3-CF3
13
32
3,5-Di-CF3
2,6-Dichloro
4-Cyano
>1000
>1000
3.40a
0.46a
0.54a
14
—
—
11
29
3-Morpholino
3-F, 5-Morpholino
H
3-Morpholino
2
117
43
0.44a
a Ki determination.
slightly improved aqueous solubility relative to the other
derivatives (data not shown). Unfortunately, this im-
proved aqueous solubility did not translate into im-
proved cellular activity.
To gain insight into the binding mode of this novel series
of p38 inhibitors, an X-ray crystal structure of analog 30

2742
S. T. Wrobleski et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2739–2744
Table 4. Optimal pyrrolotriazine C6 substituents with optimal aryl amide substitutions
O
R
HN
N
H
1
X
2
N
R1
3
N
N
Compound
R1
X
R
p38a IC50 (nM)
TNFa IC50 (nM)
31
36
13
37
38
39
40
28
30
17
27
–C(O)NHEt
–C(O)NHEt
–C(O)NHEt
–C(O)NHEt
C
3-Morpholino
3-Morpholino
3-F, 5-Morpholino
4-Cyano
0.98a
2
45
287
2
N
C
C
C
C
C
C
N
C
C
0.42a
23
46
29
7
–C(O)NH-(S)-1-methoxy-2-propan-2-yl
–C(O)NH-(S)-1-methoxy-2-propan-2-yl
–C(O)NH-(S)-1-methoxy-2-propan-2-yl
–C(O)NH-(S)-a-methylbenzyl
–C(O)NH-(S)-a-methylbenzyl
–C(O)NH-(S)-a-methylbenzyl
3-Morpholino
3-F, 5-Morpholino
4-Cyano
12
2
16
65
2
3-Morpholino
3-Morpholino
3-F, 5-Morpholino
4-Cyano
0.46a
0.44a
0.54a
3.4a
18
2
–C(O)NH-(S)-a-methylbenzyl
6
a Ki determination.
ing site, consistent with other reported p38a inhibitors
that contain this moiety.17 In addition, an indirect
hydrogen bond between the pyrrolotriazine N-3 and
the Lys53 may be possible through the intermediacy of
a water molecule (not shown). Finally, it is also appar-
ent from this structure that the DFG-out protein con-
formation nicely accommodates a unique H-bond
between the pyrrolotriazine N1 of 30 and the Leu171
residue. Interestingly, this is in contrast to the reported
binding mode of structurally related quinazoline-based
p38 inhibitors where the quinazoline core orients differ-
ently within the active site to form a direct hydrogen
bond between the quinazoline N1 and the hinge region
Met109 residue.20
All compounds having significant cellular potency
(IC50 < 100 nM) were tested by oral administration in
an acute in vivo murine model where the inhibition of
LPS-stimulated TNF-a production was measured. Com-
pounds 28, 30, and 31 were found to be the most potent
in this model, significantly inhibiting TNF-a production
by 87%, 89%, and 84%, respectively (Fig. 3).
Figure 2. Binding interactions between 30 and unphosphorylated p38a
˚
based on X-ray crystallographic analysis (2.4 A resolution, PDB entry
3BV2). Hydrogen-bond distances are given in angstroms with key
protein residues labeled.
by noting that a protonated amine would not be favored
in the lipophilic pocket, the potency differences between
analogs 33, 34, and 35 could not be easily explained
based on the X-ray structure of 30. In addition, a ratio-
nale for the significant in vitro potency of the 4-cyano
substituted analogs which lacked the morpholino group
(27, 37, 40) could not be gleaned from the structure of
30. These compounds would not be expected to bind
to the DFG-out conformation since they do not contain
a large lipophilic substituent normally required to dis-
place the Phe169 side chain.
In conclusion, we have developed a novel series of pyr-
rolo[2,1-f][1,2,4]triazine p38a MAP kinase inhibitors
having potent activity against the enzyme. Derivatives
28, 30, and 31 show significant inhibition of LPS-stimu-
lated TNF-a production when orally administered in an
in vivo murine model. X-ray crystallographic studies of
compound 30 show that the inhibitor binds to the DFG-
out protein conformation and forms hydrogen-bond
interactions with key residues (Met109, Glu71, and
Asp168) that have been utilized by other reported
p38a inhibitors.21 Despite these similarities, the pyrrolo-
triazine inhibitors such as 30 are unique in that they
bind to p38a in an orientation that is significantly differ-
ent than the structurally related quinazoline-based
inhibitors. In addition, the pyrrolotriazine N1 in the
case of 30 forms a unique hydrogen-bond interaction
Other interesting features from the co-complex with 30
include the placement of the ortho-methyl substituted
phenyl ring within the hydrophobic selectivity pocket
and the orientation of the chiral a-methylbenzyl group
against a hydrophobic surface near the edge of the bind-

S. T. Wrobleski et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2739–2744
2743
8. For reviews of this area see: (a) Hynes, J., Jr.; Leftheris, K.
Curr. Top. Med. Chem. 2005, 5, 967; (b) Goldstein, D. M.;
Gabriel, T. Curr. Top. Med. Chem. 2005, 5, 1017; (c)
Diller, D. J.; Lin, T. H.; Metzger, A. Curr. Top. Med.
Chem. 2005, 5, 953; (d) Dominguez, C.; Powers, D. A.;
Tamayo, N. Curr. Opin. Drug Disc. Dev. 2005, 8, 421.
9. Leftheris, K.; Ahmed, G.; Chan, R.; Dyckman, A. J.;
Hussain, Z.; Ho, K.; Hynes, J., Jr.; Letourneau, J.; Li, W.;
Lin, S.; Metzger, A.; Moriarty, K. J.; Riviello, C.;
Shimshock, Y.; Wen, J.; Wityak, J.; Wrobleski, S. T.;
Wu, H.; Wu, J.; Desai, M.; Gillooly, K. M.; Lin, T. H.;
Loo, D.; McIntyre, K. W.; Pitt, S.; Shen, D. R.; Shuster,
D. J.; Zhang, R.; Diller, D.; Doweyko, A.; Sack, J.;
Baldwin, J.; Barrish, J.; Dodd, J.; Henderson, I.; Kanner,
S.; Schieven, G. L.; Webb, M. J. Med. Chem. 2004, 47,
6283.
4000
3000
2000
1000
0
Vehicle
28
30
31
10. Liu, C.; Wrobleski, S. T.; Lin, J.; Ahmed, G.; Metzger, A.;
Wityak, J.; Gillooly, K. M.; Shuster, D. J.; McIntyre, K.
W.; Pitt, S.; Shen, D. R.; Zhang, R. F.; Zhang, H.;
Doweyko, A. M.; Diller, D.; Henderson, I.; Barrish, J. C.;
Dodd, J. H.; Schieven, G. L.; Leftheris, K. J. Med. Chem.
2005, 48, 6261.
Figure 3. LPS-induced TNF-a inhibition by 28, 30, and 31 in mouse.
BALB/c female mice (Harlan), 6–8 weeks of age, were used.
Compounds were dosed (10 mg/kg) in poly(ethylene glycol)
(MW = 300; PEG 300) to mice (n = 8/treatment) by oral gavage in a
volume of 0.1 mL. Control mice received PEG300 alone (‘Vehicle’).
Thirty minutes later, mice were injected intraperitoneally with 50 lg/kg
lipopolysaccharide (LPS; E. coli O111:B4; Sigma). Blood samples were
collected 90 min after LPS injection. Serum was separated and
analyzed for the level of TNF-a by commercial ELISA assay
(BioSource) according to the manufacturer’s instructions. Data shown
are means SD. *p < .05 versus Vehicle, ANOVA. Positive control
(compd 11b from Ref. 11) gave 95% TNF-a inhibition in this assay.
11. Hynes, J., Jr.; Dyckman, A. J.; Lin, S.; Wrobleski, S. T.;
Wu, H.; Gillooly, K. M.; Kanner, S. B.; Lonial, H.; Loo,
D.; McIntyre, K. W.; Pitt, S.; Shen, D. R.; Shuster, D. J.;
Yang, X.; Zhang, R.; Behnia, K.; Zhang, H.; Marathe, P.
H.; Doweyko, A. M.; Tokarski, J. T.; Sack, J. S.; Pokross,
M.; Kiefer, S. E.; Newitt, J. A.; Barrish, J. C.; Dodd, J.;
Schieven, G. L.; Leftheris, K. J. Med. Chem. 2008, 51, 4.
12. (a) Hunt, J. T.; Mitt, T.; Borzilleri, R.; Gullo-Brown, J.;
Fargnoli, J.; Fink, B.; Han, W.-C.; Mortillo, S.; Vite, G.;
Wautlet, B.; Wong, T.; Yu, C.; Zheng, X.; Bhide, R.
J. Med. Chem. 2004, 47, 4054; (b) Fink, B. E.; Vite, G. D.;
Mastalerz, H.; Kadow, J. F.; Kim, S.-H.; Leavitt, K. J.;
Du, K.; Crews, D.; Mitt, T.; Wong, T. W.; Hunt, J. T.;
Vyas, D. M.; Tokarski, J. S. Bioorg. Med. Chem. Lett.
2005, 15, 4774; (c) Borzilleri, R. M.; Zheng, X.; Qian, L.;
Ellis, C.; Cai, Z.-W.; Wautlet, B. S.; Mortillo, S.;
Jeyaseelan, R.; Kukral, D. W.; Fura, A.; Kamath, A.;
Vyas, V.; Tokarski, J. S.; Barrish, J. C.; Hunt, J. T.;
Lombardo, L. J.; Fargnoli, J.; Bhide, R. S. J. Med. Chem.
2005, 48, 3991.
with the Leu171 residue. Future efforts to further ex-
plore the promising potential of pyrrolo [2,1-f][1,2,4]tri-
azines as p38a MAP kinase inhibitors will be reported in
due course.
Acknowledgments
We, the authors, thank our colleagues Dr. Brian E.
Fink, Dr. Kyoung S. Kim, and Dr. Alaric D. Dyckman
for helpful suggestions in the preparation of this
manuscript.
13. Cumming, J. G.; McKenzie, C. L.; Bowden, S. G.;
Campbell, D.; Masters, D. J.; Breed, J.; Jewsbury, P.
Bioorg. Med. Chem. Lett. 2004, 14, 5389, and references
therein.
References and notes
14. Kamijo, S.; Kanazawa, C.; Yamamoto, Y. J. Am. Chem.
Soc. 2005, 127, 9260.
15. Hynes, J., Jr.; Doubleday, W. W.; Dyckman, A. J.;
Godfrey, J. D., Jr.; Grosso, J. A.; Kiau, S.; Leftheris, K.
J. Org. Chem. 2004, 69, 1368.
1. (a) Kavanaugh, A. Adv. Ther. 2006, 23, 208; (b) Fleisch-
mann, R. Expert Rev. Clin. Immunol. 2006, 2, 331.
2. Grunke, M.; Kalden, J. R. Expert Rev. Clin. Immunol.
2005, 1, 313.
3. (a) Saleem, B.; Mackie, S.; Emery, P. Expert Rev. Clin.
Immunol. 2006, 2, 193; (b) Cottone, M.; Mocciaro, F.;
Modesto, I. Expert Opin. Biol. Ther. 2006, 6, 401.
4. (a) Papadakis, K. A. Expert Rev. Clin. Immunol. 2006, 2,
11; (b) Mease, P. J. Expert Opin. Biol. Ther. 2005, 5, 1491;
(c) Lee, S. J.; Kavanaugh, A. Therapy 2005, 2, 13.
5. Wagner, G.; Laufer, S. Med. Res. Rev. 2006, 26, 1.
6. (a) Adams, J. L.; Badger, A. M.; Kumar, S.; Lee, J. C.
Prog. Med. Chem. 2001, 38, 1; (b) Lee, J. C.; Laydon, J. T.;
McDonnell, P. C.; Gallagher, T. F.; Kumar, S.; Green, D.;
McNulty, D.; Blumenthal, M. J.; Heyes, J. R. Nature
1994, 372, 739.
7. (a) Hale, K. K.; Trollinger, D.; Rihanek, M.; Manthey, C.
L. J. Immunol. 1999, 162, 4246; (b) Allen, M.; Svensson,
L.; Roach, M.; Hambor, J.; McNeish, J.; Gabel, C. A. J.
Exp. Med. 2000, 191, 859; (c) Fearns, C.; Kline, L.; Gram,
H.; Di Padova, F.; Zurini, M.; Han, J.; Ulevitch, R. J.
J. Leukocyte Biol. 2000, 67, 705.
16. Protocols for in vitro assays are as follows: p38 enzyme
assays were performed in V-bottomed 96-well plates. The
final assay volume was 60 lL prepared from three 20 lL
additions of enzyme, substrates (MBP and ATP) and test
compounds in assay buffer (50 mM Tris, pH 7.5, 10 mM
MgCl2, 50 mM NaCl, and 1 mM DTT). Bacterially
expressed, activated p38 was pre-incubated with test
compounds for 10 min prior to initiation of reaction with
substrates. The reaction was incubated at 25 ꢁC for 45 min
and terminated by adding 5 lL of 0.5 M EDTA to each
sample. The reaction mixture was aspirated onto a pre-wet
filtermat using a Skatron Micro96 Cell Harvester (Ska-
tron, Inc.), then washed with PBS. The filtermat was then
dried in a microwave oven for 1 min, treated with
MeltilLex A scintillation wax (Wallac), and counted on
a Microbeta scintillation counter Model 1450 (Wallac).
Inhibition data were analyzed by nonlinear least-squares
regression using Prizm (GraphPadSoftware). The final

2744
S. T. Wrobleski et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2739–2744
concentration of reagents in the assays are ATP, 1 lM;
16 h by centrifugation at 5 ꢁC and stored at À80 ꢁC prior
to purification. All purification steps were done at 4 ꢁC,
and were as follows. The cell paste from the culture
expressing noncleavable pentahistidine-tagged p38a was
lysed by homogenization at 8000–9000 psi (APV Rannie
Mini-Lab 8.30H) in 25 mM Hepes, 500 mM NaCl, 50 mM
imidazole, 5% glycerol (v/v), 2 mM b-mercaptoethanol,
1 lg/mL Leupeptin, and 1 lg/mL Pepstatin, pH 7.5. The
lysate was clarified by ultracentrifugation (45 min at
30,000g) and filtration (1.2 lm syringe filter) before Ni-
affinity chromatography (Pharmacia Biotech, Chelating
Sepharose Fast Flow), eluting with an imidazole gradient
from 50 to 500 mM final concentration in this buffer.
Pooled fractions were dialyzed into 25 mM Hepes, 50 mM
NaCl, 5% glycerol (v/v), 1 mM EDTA, 2 mM DTT, and
1 lg/mL Leupeptin, pH 7.5, and loaded onto a Pharmacia
Resource Q anion exchange column followed by elution
with a NaCl gradient to 0.6 M in the same buffer.
Fractions containing target protein were combined and
dialyzed into 25 mM Hepes, 200 mM NaCl, 5% glycerol
(v/v), 1 mM EDTA, and 2 mM DTT, pH 7.5 and loaded
onto a Pharmacia Superdex 200 (26/60) column and eluted
with an isocratic gradient with the same buffer. The final
fractions were combined and dialyzed into 25 mM Tris,
50 mM NaCl, 5% glycerol (v/v), 1 mM DTT, pH 7.4. Co-
crystals of (His)5-p38a 30 inhibitor complexes grew
spontaneously in hanging drop vapor-diffusion trials
conducted at 4 ꢁC with drops containing 1 lL of concen-
trated protein–inhibitor complex mixed with an equal
volume of reservoir solution (10% PEG 8000, 200 mM
KCl, 100 mM MgOAc, 50 mM sodium cacodylate, pH
6.5). Prior to data collection, the crystals were slowly
transferred to a stabilization solution (8% glycerol, 30%
PEG 8000, 200 mM KCl, 100 mM MgOAc, 50 mM
sodium cacodylate, pH 6.5) and then flash-frozen in liquid
[c-33P]ATP, 3 nM, MBP (Sigma, #M1891), 2lg/well; p38,
10 nM; and DMSO, 0.3%.
TNF-a production by LPS-stimulated PBMCs: heparin-
ized human whole blood was obtained from healthy
volunteers. Peripheral blood mononuclear cells (PBMCs)
were purified from human whole blood by Ficoll-Hypaque
density gradient centrifugation and resuspended at a
concentration of 5 · 106/mL in assay medium (RPMI
medium containing 10% fetal bovine serum). 50 lL of cell
suspension was incubated with 50 lL of test compound
(4· concentration in assay medium containing 0.2%
DMSO) in 96-well tissue culture plates for 5 min at rt.
100 lL of LPS (200 ng/mL stock) was then added to the
cell suspension and the plate was incubated for 6 h at
37 ꢁC. Following incubation, the culture medium was
collected and stored at À20 ꢁC. TNF-a concentration in
the medium was quantified using a standard ELISA kit
(Pharmingen-San Diego, CA). Concentrations of TNF-a
and IC50 values for test compounds (concentration of
compound that inhibited LPS-stimulated TNF-a produc-
tion by 50%) were calculated by linear regression analysis.
17. (a) Liverton, N. J.; Butcher, J. W.; Claiborne, C. F.;
Claremon, D. A.; Libby, B. E.; Nguyen, K. T.; Pitzen-
berger, 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; (b) Fitzgerald, C. E.; Patel, S.
B.; Becker, J. W.; Cameron, P. M.; Zaller, D.; Pikounis, V.
B.; O’Keefe, S. J.; Scapin, G. Nat. Struct. Biol. 2003, 10,
764, Based on Ref. 17b the p38a binding mode from X-ray
studies for a p38 inhibitor showing a preference for a (S)-
a-methylbenzyl group is depicted below:.
˚
nitrogen. Data to 2.4 A resolution were collected at the
IMCA-CAT beamline BM-17 at the Advanced Photon
Source, Argonne, IL. On a MarCCD detector, reduced
with the programs DENZO and SCALEPACK, (Otwi-
nowski, Z.; Minor, W. Processing X-ray data collected in
oscillation mode. Methods Enzymol. 1997, 276, 307–326.)
and refined by program autoBuster (Global Phasing, Ltd,
Cambridge, U.K.). The final crystallographic refinement
R-factor was 21.2%. The coordinates have been deposited
in the Protein Data Bank as entry 3BV2.
CF3
N
NH
Asp168
N
N
Met109
Gly110
19. Pargellis, C.; Tong, L.; Churchill, L.; Cirillo, P. F.;
Gilmore, T.; Graham, A. G.; Grob, P. M.; Hickey, E.
R.; Moss, N. l.; Pav, S.; Regan, J. Nat. Struct. Biol. 2002,
9, 268.
HN
(S)-α-methylbenzyl
20. Based on Ref. 13, the p38a binding mode for a quinazoline
p38 inhibitor based on X-ray crystallography studies is
depicted below:
Thr106
18. The gene encoding human p38a MAP kinase (isoform 2,
residues 2-360) was subcloned into a pET28a vector
(Novagen) between the NcoI and BamHI restriction sites,
conferring expression of a noncleavable N-terminal pen-
tahistidine-tagged enzyme. The resulting plasmid was used
to transform a W3110 (DE3) strain of E. coli (American
Type Culture Collection). Cells were grown at 37 ꢁC into
late log phase (OD600 of 11) in an oxygen-sparged 1-L
fermenter using enriched Don’s M101 medium [46 mM
potassium phosphate, 23 mM ammonium sulfate, 4% (w/
v) yeast extract (Becton Dickenson), 5% (w/v) Hy-Soy
peptone (Quest Scientific), 2 mM magnesium sulfate, 2%
glycerol (v/v), and 50 lg/mL kanamycin sulfate], chilled to
20 ꢁC, and induced with 1 mM isopropyl b-D-thiogalac-
topyranoside (MP Biomedicals). Cells were harvested after
Glu71
N
NH
Met109
N
O
H
N
O
N
Asp168
O
N
O
N
21. Wrobleski, S. T.; Doweyko, A. M. Curr. Top. Med. Chem.
2005, 5, 1005.