13

Structure-based drug design case study: p38

Arthur M. Doweyko

inhibitors were found to compete for the ATP binding

INTRODUCTION
site in p38␣. Additionally, statistical analysis of the com-
The overproduction of cytokines has been implicated in a binatorial data indicated a significant contribution to
wide variety of inflammatory diseases such as rheumatoid activity in this series correlated to the presence of the
arthritis, inflammatory bowel disease, psoriasis, multiple 4-methyl-3-benzamido aniline moiety. When the x-ray
sclerosis, osteoporosis, Alzheimer’s disease, and congestive crystal structure of the protein-inhibitor complex for a
heart failure. The ability of p38 mitogen-activated protein member of this triaminotriazine aniline amide series was
kinase (p38 MAPK) to regulate the release and activity determined, the structure/activity relationship (SAR) for
of multiple pro-inflammatory cytokines has attracted the the series was quickly rationalized. Specifically, N-methoxy-
interest of numerous pharmaceutical companies and inde- 4-methyl-3-(4-(methyl(neopentyl)amino)-6-(4-methyl-1,4-
pendent researchers during the past decade or so. Since diazepan-1-yl)-1,3,5-triazin-2-ylamino)benzamide (1) was
its initial discovery in 1994 as a potential molecular tar- cocrystallized with unactivated p38␣ protein (Figure 13.1),
get for a novel class of cytokine suppressive inhibitors (SB- confirming that the series binds to the ATP pocket.20 In a
203580),1 more than 150 patent applications from at least manner similar to ATP, 1 binds to the hinge region of p38␣
thirty pharmaceutical companies have been published, all (characterized by residues 106–110), forming an anchoring
claiming novel p38 inhibitors. Four distinct isoforms of p38 H-bond interaction with Met109. Unlike ATP, 1 makes use
MAPK are known: p38␣ and p38␤ are widely expressed in of an intervening water molecule to form the interaction
eukaryotic cells, including endothelial and inflammatory between Met109 backbone NH and the triazine N3. Also
cells; p38␥ is found in skeletal muscle; and p38␦ is pre- characteristic of other kinases is the presence of a deep
dominantly found in the small intestine, kidneys, and lung hydrophobic pocket near the so-called gatekeeper residue
tissue.2,3 Of these four isoforms, p38␣ has been the most (Thr106, not shown) which provides for one of the more
studied and is believed to be the most physiologically rele- interesting features of kinase inhibitor design in that it rep-
vant. Numerous reviews have been published that focus on resents a space not occupied by ATP and, thus, of potential
both the biology4–7 and chemistry of p38 inhibitors.8–17 The value in the search for inhibitor selectivity against off-target
focus of this chapter is an illustration of p38 inhibitor design kinases. In the present case, the 4-methyl-3-benzamido
guided by structural information obtained both from mod- aniline moiety occupies that space. Further interactions in
eling and actual x-ray crystallographic data. Structure refer- the binding site include H-bonds among Lys53, Glu71, and
ences with a “.pdb” suffix refer to those obtained from the amido NH, between Asp168 backbone NH and amido C=O,
Research Collaboratory for Structural Bioinformatics.18 and between the protonated diazepan nitrogen and Asp168
carboxylate. Interestingly, the distance between triazine N1
and Lys53 is suggestive of an intervening water molecule;
TRIAZINES AND PYRIMIDINES
however, none was evident from the crystallography. The
We begin with a collaborative venture by Bristol-Myers potential for H-bonding between Lys53 and an acceptor
Squibb and Pharmacopeia aimed at the development of atom in other ligands was eventually realized with subse-
a novel series of trisubstituted triazines. High-throughput quent inhibitors. Synthetic efforts targeted variation at all
screening applied to a collection of 2.1 million com- three positions (2,4,6) of the triazine core, and the emer-
pounds derived from a combinatorial library based on gent SAR dovetailed nicely with the crystallographically
the template shown in Scheme 13.1 yielded the 1,3,5- determined binding mode. For example, the methylhydrox-
triaminotriazine aniline amide PS200981, having a p38␣ amate ester was found to have superior binding affinity
IC50 of 1 ␮M.19 Further analyses identified PS166276 with compared to amides in general, a consequence of both its
a p38␣ IC50 of 28 nM having 10× less cytotoxity and small size (complementing the relatively narrow pocket
superior inhibition of lipopolysaccharide-(LPS) induced at that position) and its electron-deficient NH proton
TNF␣ production in THP-1 monocytes (170 nM). These H-bond donor. The use of branched alkyl amines at the

197

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015
 198 Arthur M. Doweyko

Scheme 13.1. Progression of synthetic efforts that culminated in the design and synthesis of trisubstituted tri-
azines with significant p38␣ inhibition. Combinatorial libraries of the type shown led to the identification of
triazines PS200981 and PS166276 (Pharmacopeia) having the unique 4-methyl-3-benzamido aniline head group.
Subsequent SAR work (BMS) led to triazines exemplified by 1, with superior p38␣ inhibition and in vitro activity.

4-position of the triazine was found to yield a number of closing down on the binding site near Ala111. For the sake
active congeners, locating an angular lipophilic element of consistency and facile orientation, the coloring scheme
of the inhibitor along a secondary hydrophobic pocket in (yellow P-loop and Ala111) is maintained in subsequent
p38␣. This pocket is created by the lower rim of the P-loop figures.
The through-water H-bond to Met109 observed for 1
represented an intriguing observation that led to a con-
sideration for its replacement by an H-bond acceptor
built onto an inhibitor core. Such replacements have been
reported as successful in using a cyano moiety to pro-
vide the H-bonding acceptor and span the necessary dis-
tance with inhibitors of scytalone dehydratase21 and epider-
mal growth factor receptor kinase.22 In the present case, a
pyrimidine core was substituted for the triazine of 1 and a
cyano group installed at the 5-position leading to the p38␣
active pyrimidine series illustrated in Scheme 13.2.23 The
5-position was suggested by modeling to have an optimal
trajectory, which was confirmed by the x-ray crystal struc-
ture of the p38␣ complex with 2 (Figure 13.2). The posi-
tion of the cyano N was found almost precisely where the
displaced water O had resided in the 1 complex. The two
hydrophobic pockets (deep and hinge) are occupied in a
Figure 13.1. Major interactions observed in the complex between p38␣ similar manner, with the additional benefit of a likely H-
and triazine 1. H-bonds include Met109 backbone NH/H2O/triazine N3, bond between aniline NH and Thr106. In addition to the
Glu71/hydroxamate NH, Asp168/protonated diazapan, and backbone H-bonding array between ligand amide and Lys53, Glu71,
Asp168 NH/Glu71/hydroxamate O. Hydrophobic interactions include deep
and Asp168, a water molecule is clearly visible between
pocket/4-methyl-3-benzamido aniline and neopentylmethylamino/hinge
hydrophobic pocket. Arrow indicates a potential through water H-bonding pyrimidine N1, Lys53 and Asp168. The strong p38␣ bind-
motif between triazine N3 and Lys53 not observed. ing affinity (IC50 0.41 nM) and hPBMC TNF␣ inhibition

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015
 199 Structure-based drug design case study: p38

Scheme 13.2. The x-ray crystal structure of 1 revealed the presence of a key water molecule providing an H-
bonding interaction between backbone NH at Met109 and triazine N3. The pryimidine scaffold (upper right) provided
a means to replace that water molecule with a 5-cyano. Further synthesis identified 2 as a potent p38␣ inhibitor.
The inset illustrates the major hydrophobic and H-bonding interactions observed at the 4-methyl-3-benzamido
aniline head group and points to a possible water-mediated H-bonding interaction between triazine N1 and Lys53
(not observed).

(IC50 8.7 nM) is consistent with the number of strong inter-

actions observed in the p38␣ ATP binding site, despite
the absence of a substituent at the 2-position (6-position
in the triazine). As the overall binding affinity for either
the triazines or the pyrimidines is significantly affected by
the combination and type of substituents it is difficult to
directly assess the binding contribution due to a specific
substituent. However, in this series, the best combinations
included an amine at the 2-position and a branched alkyl-
amine at the 4-position. The ultimate choice of best amine
relies not only on observed p38␣ binding affinity but on
cytotoxicity screens and cellular activity.

FUSED HETEROCYLICS
A novel structural class of p38␣ MAP kinase inhibitors
was developed as a result of the high-throughput screen-
ing (HTS) hit, pyrrolo[2,1-f ][1,2,4]triazine oxindoles, shown
Figure 13.2. The complex between 2 and p38␣ confirms that the 5- in Scheme 13.3, which exhibited p38␣ IC50 values in the
cyano group that makes a key H-bonding interaction between inhibitor 60- to 80-nM range.24 Substituted phenylaminopyrrolo[2,
and p38␣ is located in the same position as the water molecule in 1. A
1-f ][1,2,4]triazines had been used previously as a template
water molecule was observed in an H-bonding position among pyridyl
N1, Asp168, and Lys53, while backbone NH at Asp168 and Glu71 anchor for kinase inhibitor design,25 and it was envisioned that
the pendant amide. the incorporation of a 4-methyl-3-benzamido aniline (as

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015
 200 Arthur M. Doweyko

Scheme 13.3. High-throughput screening efforts led to the identification of compounds containing the triazine
oxindole core pictured above. Incorporation of the 4-methyl-3-benzamido aniline head group led to a series of
analogs with variations at both ends of this chemotype, namely the use of amides, reverse amides, carbamates,
and hydroxamate esters. Two examples for which x-ray structures with p38␣ were obtained were 3 and 4.

employed in 1 and 2) together with this novel core would

provide an alternative class of p38␣ inhibitors. A number
of analogs were generated with variation at the ester to
increase metabolic stability. These analogs included the use
of amides, reverse amides, and carbamates at the ester posi-
tion as well as variations at the 4-methyl-3-benzamido ani-
line (amides, reverse amides, carbamates, and hydroxam-
ate esters). The series exhibited potent p38␣ inhibition (IC50
1–680 nM) and submicromolar cell activity (LPS/TNF␣).
The question of binding mode arose early in the synthetic
effort and was answered with x-ray crystal structures of
the p38␣ complexes of 3 and 4, having p38␣ IC50 values of
3.1 and 2.2 nM, respectively. The crystal structure of 4 is
shown in Figure 13.3. Despite the distinct possibility that
Met109 NH could H-bond with N1 of the pyrrolotriazine
core, the x-ray data confirmed that this key H-bonding
interaction was occurring with the pendant amide carbonyl
O. This observation was consistent with the emerging SAR
Figure 13.3. The x-ray crystal structure of the p38␣ complex with 4
as broader types of substitutions were tolerated at the C6
reveals Met109 NH H-bonding to the carboxamide O and the pres-
ence of a water molecule engaged in H-bonding between pyrrolotri- position, presumably reflecting their trajectories along the
azine N3, Asp168, and Lys53. The deep pocket is occupied by the 4- hinge region and out into solvent. Once again, as found
methyl-3-benzamido aniline head group while the (S)-␣-methylbenzyl for 2, a water molecule was found in an H-bonding com-
group is located in a hydrophobic channel between P-loop and Ala111 plex between the N3 of the pyrrolotriazine core, Lys53,
(2RG6.pdb). [Source: H.M. Berman, J. Westbrook, Z. Feng, G. Gilliland,
and Asp168. The hydroxamate methyl ester was locked in
T.N. Bhat, H. Weissig, I.N. Shindyalov, P.E. Bourne. The Protein Data Bank
Nucleic Acids Research, 28, pp. 235–242 (2000); see also www.pdb. by H-bonding interactions with backbone Asp168 NH and
org] Glu71. The deep hydrophobic pocket was occupied by the

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015
 201 Structure-based drug design case study: p38

Scheme 13.4. The structures of the Boehringer Ingelheim “allosteric” inhibitors (BI urea-pyrazole and BIRB-796)
are shown at top along with an indication of BIRB-796 interactions with p38␣. The DFG-out conformation of p38␣
refers to the displacement of several residues along the activation loop (Asp168-Phe169-Gly170) providing for a
hydrophobic pocket as an extension of the ATP binding site. The BMS pyrrolotriazines were further developed to
include a structure (3-morpholinobenzamide, 5) that was found to occupy the same DFG-out pocket.

4-methyl-3-benzamido aniline while the ethyl of 3 (not reference as an allosteric site may be somewhat mislead-
shown) and the (S)-␣-methylbenzyl of 4 rested in the outer ing, as the inhibitory effect of binding a small molecule to
hinge pocket. Phe169 site is not allosteric in the classical sense. When
a molecule binds to this newly formed site located adja-
cent to the ATP binding pocket, it directly interferes with
ACCESSING THE DFG-OUT BINDING POCKET
ATP binding by providing a steric block. Normally, allosteric
In 2002 researchers at Boehringer Ingelheim reported a lim- interactions are relegated to those requiring indirect com-
ited set of inhibitors that used a novel p38 MAP kinase munication between isolated sites. The remaining interac-
allosteric binding site.26 The urea-pyrazole and the more tions observed in the x-ray structure for BIRB-796 entail
elaborate BIRB-796 are illustrated in Scheme 13.4. The x- those that have been seen before, namely hinge region
ray crystal structure for BIRB-796 (1KV2.pdb) is shown Met109 H-bonding to pendant morpholino O and an H-
in Figure 13.4. Although part of the inhibitor is located bonding matrix among urea, Glu71, and Asp168 back-
in the ATP-binding site and H-bonds with Met109, the bone NH.
opposite end locates itself in a pocket created by the dis- Accessing the DFG-out conformation of p38 was accom-
placement of part of the activation loop, namely Asp168- plished through an extension of the pyrrolotriazines exem-
Phe169-Gly170 (or DFG). This relocation of the DFG loop plified by 3 and 4. It was found that the installation
(sometimes referred to as DFG-out) vacates a hydrophobic of larger amide groups off the 4-methyl-3-benzamido
pocket formerly occupied by Phe169. The DFG-out config- aniline head group led to congeners with potent p38␣
uration of p38 results in an extended ATP-binding site. Its inhibition (Scheme 13.4). For example, the use of a

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015
 202 Arthur M. Doweyko

Figure 13.4. (Top left) The BIRB-796 x-ray structure in p38␣ illustrates the DFG-out motif, wherein a t-butyl group
occupies the Phe169 pocket while the pendant morpholino O H-bonds to backbone NH at Met109 (1KV2.pdb). (Top
right) Pyrrolotriazine 5 occupies the same Phe169 pocket while the displaced activation loop adopts a different
pose with Leu171 backbone NH engaged in an H-bond to pyrrolotriazine N1 (3BV2.pdb). (Center) A comparison
of overall shape between BIRB-796 and 5 illustrating similar occupation of the deep hydrophobic and DFG-out
pockets.

3-morpholinobenzamide in combination with a C6-(S)-␣- core (Scheme 13.5). The presumption was that presenta-
methylbenzylamide (5) exhibited a p38␣ Ki of 0.44 nM and tion of an H-bond acceptor at roughly the same location
a LPS/TNF␣ IC50 of 18 nM.27 The x-ray crystal structure and trajectory as the cyano group in 2 could lead to a
of the p38␣ complex of 5 confirmed the DFG-out config- novel series of inhibitors that retain the Met109 NH inter-
uration (Figure 13.4). The binding mode of 5 is similar to action thought to be a key H-bonding interaction for nearly
that of 4 in that the H-bonding patterns to Met109, Glu71, all kinase inhibitors. This was achieved by conceptually
and backbone NH at Asp168 are conserved. The notable cyclizing the 5-cyano to the 6-aminoalkyl function, yield-
distinction here is that the pendant morpholinobenzamide ing a pyrazolopyrimidine core, which was further elabo-
group is found deep within the hydrophobic Phe169 pocket. rated both at the N1 and the 4-methyl-3-benzamido ani-
In addition, part of the activation loop has relocated itself line head group to more fully develop an SAR.28 The x-ray
along the outer rim of the ATP-binding site so as to form crystal structure of the unphosphorylated p38␣ complex
a seal, as evidenced by the H-bond between pyrrolotri- of 6, shown in Figure 13.5, confirms that the N2 accep-
azine N1 and backbone NH at Leu171. This feature is dis- tor in the pyrazolopyrimidine core forms an H-bond with
tinct from that reported for BIRB-796. A further comparison Met109 and the pendant methyl amide forms the usual H-
between BIRB-796 and 5 is shown in Figure 13.4, wherein bonding complex with Glu71 and Asp168. Additionally, the
the molecular volume overlap between the two inhibitors is N1-phenyl is located along the hinge region hydrophobic
highlighted. Although BIRB-796 does not make use of the pocket. Although 6 exhibited a good inhibition profile (p38␣
hinge hydrophobic pocket, both inhibitors occupy the deep IC50 14 nM, LPS/TNF␣ IC50 513 nM), further SAR explo-
hydrophobic pocket and the Phe169 pocket in similar ways. ration identified the 1,2-oxazolamide (same as in 2) as supe-
It is clear that relatively large inhibitors can be accommo- rior (p38␣ IC50 5 nM, LPS/TNF␣ IC50 6 nM).
dated by the DFG-out version of p38␣.
THIAZOLES
PYRAZOLOPYRIMIDINES
The discovery of an active thiazole central core repre-
A further elaboration of the pyrimidine chemotype exem- sents an unobvious elaboration of the pyrrolotriazine motif.
plified by 2 led to the discovery of the pyrazolopyrimidine Focused deck screening identified a C2-alkylaminothiazole

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015
 203 Structure-based drug design case study: p38

Scheme 13.5. The conceptual “cyclization” of the 5-cyanopyrimidine core led to the synthesis of the pyrazolopy-
rimidine scaffold, capable of a similar H-bonding trajectory to backbone NH at Met109. An x-ray structure of the
p38␣ complex of 6 confirmed the presence of this key H-bonding interaction.

with moderate p38 activity. This observation suggested illustrated in Scheme 13.6. Synthetic efforts led to the dis-
that replacement of the triazinyl-aniline link with a car- covery of 7 which was found to be a potent inhibitor (p38␣
boxanilide may be a way to retain possible “backside” H- IC50 3.5 nM, LPS/TNF␣ IC50 2.9 nM).29 The x-ray crystal
bonding through water or directly to Lys53, substituting the structure of its p38␣ complex is illustrated in Figure 13.6.
carboxanilide O for the triazinyl N3. In addition, the fused Although 7 is slightly shorter than 3, the added flexibility of
5-membered pyrrolo ring could be replaced with a thia- the carboxanilide linker allows the molecule to adapt itself
zole containing a potential H-bond acceptor N. These two
concepts were embodied in the conceptual transformation

Figure 13.6. The x-ray structure of the p38␣ complex of C2-alkyl-

aminothiazole, 7, illustrates the characteristic deep hydrophobic pocket
occupancy by the 4-methyl-3-benzamido aniline head group and H-
Figure 13.5. The x-ray structure of the p38␣ complex of pyrazolopyrimi- bonding interactions between inhibitor and Met109, Thr106, Glu71, and
dine, 6, illustrates the successful replacement of the 5-cyano group in 2 Asp168. Interestingly, both backbone carbonyl O and NH at Met109 are
with the pyrazolyl N2. (3CG2.pdb) involved in a tandem set of H-bonds. (3BX5.pdb)

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015
 204 Arthur M. Doweyko

Scheme 13.6. The thiazole pictured at top was identified through focused deck screening with modest p38␣ activity.
Coupling this observation with the structure of the pyrrolotriazine, 3, led to the synthesis of 7. Conceptually, the
strategy was to replace the pendant amido carbonyl O–Met109 interaction with the thiazolyl N and incorporate a
possible H-bond acceptor replacement (potential for H-bonding to Lys53/water) for the triazinyl N1 with the central
carboxamido O.

sufficiently to the binding site to engage in productive H- (not shown) and the aliphatic chain of Lys53. Curiously, the
bonding not only to Met109, Glu71, and Asp168 but also to oxalamide portion extends upward and out along the hinge
Thr106 and backbone C=O at Met109 as well. This observa- region, making no direct interactions with protein. It would
tion also reflects previous analyses highlighting the flexibil- appear that the occupancy of the deep hydrophobic pocket
ity of p38␣, specifically regarding the variation in the width is critical in this series and that the Scios design, incorporat-
of the ATP pocket as determined by a comparison of avail- ing a conformationally constrained linker, directs the pen-
able x-ray structures.11 dant benzyl group into that pocket.

INDOLES
Scios recently reported the synthesis and SAR of indole-
based heterocyclic inhibitors of p38␣ shown in Scheme
13.7.30 The authors found that rigidifying the piperidine
linker in Scios 1 led to a significant increase in binding affin-
ity (∼fourteenfold, Scios 2). Further modifications eventu-
ally led to Scios’s first p38 clinical compound, Scios-469
(p38␣ IC50 9 nM). These observations suggested that a sim-
ilar conformational restraint might be achieved through
the incorporation of a fused ring system, specifically, a
connection between the benzyl methylene position and
a proximal piperazine or piperidine ring (as illustrated in
Scheme 13.7). A number of such analogs were synthe-
sized and found to exhibit double-digit nanomolar p38␣
inhibition.31 The binding mode of 8 (p38␣ IC50 13 nM)
was determined by x-ray crystallography (Figure 13.7). Sev-
eral key H-bonding interactions are evident and consistent Figure 13.7. The x-ray structure of the complex between p38␣ and an
with previous observations: through-water Lys53/Asp168 analog of Scios-469 using a conformational constraint in the form of
a fused ring system (imidazopyrazine, 8) illustrates several key interac-
H-bond to the exposed imidazo N and Met109 backbone
tions: central amido carbonyl O H-bonding to Met109 and water molecule
NH H-bonding to the central carboxamide O. The major H-bonding among imidazo N, Asp168, and Lys53. The deep hydrophobic
hydrophobic interaction occurs through the pendant diflu- pocket created between Thr106 (not shown) and Lys53 is occupied by
orophenyl occupancy of the deep pocket created by Thr106 the pendant difluorophenyl. (2QD9.pdb)

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015
 205 Structure-based drug design case study: p38

Scheme 13.7. Scios reported enhanced p38␣ inhibition through the incorporation of conformational constraints in
their series (Scios 1, Scios 2, and Scios-469). The installation of a fused ring system (such as shown in parentheses)
was considered as an alternate approach to conformational constraint, culminating in the potent p38␣ inhibitor, 8.

to the oxazole N to consider a possible additional H-bond

THE 5-MEMBERED HETEROCYCLIC CORE
at that location. Unexpectedly, the P-loop of p38␣ is col-
A very popular p38␣ inhibitor design is based on the lapsed on the inhibitor in such a way that Tyr35 makes a
early work of SmithKline Beecham (SKB) that described tight van der Waals contact with the thiazole sulfur. This
the trisubstituted imidazole, SB-203580 (Scheme 13.8).1,32 may represent a unique hydrophobic interaction scheme
Much effort has been expended on the successful replace- not typically used by p38 inhibitors. Finally, there is the
ment of the central imidazole core with other five- subtle, yet important, hydrophobic interaction engaged by
membered ring systems. Recent efforts have expanded the 2-isopropylamine at the hinge entry. Small alkyl sub-
the approach by focusing on fused ring systems as cen- stituents at this position were observed to modulate the
tral cores. Examples in this regard include the Roche binding affinity (ethyl-butyl, 1.6–16 nM).
pyrrolopyridine.33 Replacement of the Met109-targeting In summary, a number of structure-based design strate-
moiety has also been pursued, exemplified by the two gies were highlighted that focused on the use of novel lig-
Pfizer structures shown in Scheme 13.8.34,35 Parallel efforts and head groups to access the deep hydrophobic pocket
at Bristol-Myers Squibb (BMS) yielded the benzothiazole (most notably, the 3-methyl-5-benzamide system), the
series that included the use of oxazoles and imidazoles as incorporation of H-bond acceptor atoms in the core that
central cores.36 An example oxazole, 9 (p38␣ IC50 6.4 nM, target both the hinge region Met109 interaction as well
LPS/TNF␣ IC50 40 nM), was successfully crystallized with as Lys53/Glu71/Asp168 deeper in the binding site, the
p38␣ (Figure 13.8). As expected, the fluorophenyl moiety is exploitation of the Tyr169 pocket available in the DFG-out
located in the deep hydrophobic pocket. Interestingly, the configuration, the subtle effect of alkyl/aryl group occu-
hinge region Met109 is engaged in two H-bonds to the lig- pancy of the hinge region hydrophobic channel, and unique
and 2-aminothiazole system. Lys53 is found close enough interactions such as the Tyr35 van der Waals interaction

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015
 206 Arthur M. Doweyko

Scheme 13.8. The early work of SKB led to the synthesis of SB-203580 whose p38␣ x-ray structure revealed
key interactions with Met109, Lys53, and the deep hydrophobic pocket. Subsequent efforts by Roche and Pfizer
represent just a few of the variations around the SKB theme that led to potent inhibitors. The use of a fused
heterocycle to access the backbone Met109 NH H-bond was successfully realized with 9. In addition to the usual
interactions, 9 exhibited a unique P-loop collapse that included a tight van der Waals contact between thiazolyl S
and the ring face of Tyr35.

with thiazole sulfur. All of these approaches have led to the REFERENCES
development of leads currently being investigated as poten- 1. Lee, J. C.; Laydon, J. T.; McDonnell, P. C.; Gallagher, T. F.;
tial drug candidates, testifying to the value and impact of a Kumar, S.; Green, D.; McNulty, D.; Blumenthal, M. J.; Heyes,
structure-based design approach in modern drug discovery. 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. A protein kinase involved in the regulation
of inflammatory cytokine biosynthesis. Nature 1994, 372, 739–
746
2. Hale, K. K.; Trollinger, D.; Rihanek, M.; Manthey, C. L. Dif-
ferential expression and activation of p38 mitogen-activated
protein kinase ␣, ␤, ␥ and ␦ in inflammatory cell lineages J.
Immunol. 1999, 162, 4246–4252.
3. Lee, J. C.; Kassis, S.; Kumar, S.; Badger, A.; Adams, J. L. p38
mitogen-activated protein kinase inhibitors – mechanisms
and therapeutic potentials. Pharmacol. Ther. 1999, 82, 389–
397.
4. Chen, Z.; Gibson, T. B.; Robinson, F.; Silvestro, L.; Pearson, G.;
Xu, B.-E.; Wright, A.; Vanderbilt, C.; Cobb, M. H. MAP kinases.
Chem. Rev. 2001, 101, 2449–2476.
5. Dambach, D. M. Potential adverse effects associated with inhi-
bition of p38. ␣/␤ MAP kinases Curr. Top. Med. Chem. 2005, 5,
929–939.
6. Pearson, G.; Robinson, F.; Gibson, T. B.; Xu, B.-E.; Karandikar,
M.; Berman, K.; Cobb, M. H. Mitogen-activated protein (MAP)
kinase pathways: Regulation and physiological functions.
Figure 13.8. The x-ray crystal structure of the p38␣ complex with Endocr. Rev. 2001, 22, 153–183.
benzothiazole-oxazole, 9, illustrates the expected H-bonding interac- 7. Schieven, G. L. The biology of p38 kinase: a central role in
tions at Met109 and Lys53. In addition, the collapse of the P-loop unto inflammation. Curr. Top. Med. Chem. 2005, 5, 921–928.
the inhibitor is evident, along with a tight van der Waals contact between 8. Boehm, J. C.; Adams, J. L. New inhibitors of p38 kinase. Expert
thiazolyl S and Tyr35. (3C5U.pdb) Opin. Ther. Pat. 2000, 10, 25–37.

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015
 207 Structure-based drug design case study: p38

9. Chakravarty, S.; Dugar, S. Inhibitors of p38a MAP kinase. Ann. K. 5-Cyanopyrimidine derivatives as a novel class of potent,
Rep. Med. Chem. 2002, 37, 177–186. selective, and orally active inhibitors of p38␣ MAP kinase.
10. Cirillo, P. F.; Pargellis, C.; Regan, J. The non-diaryl heterocy- J. Med. Chem. 2005, 48, 6261–6270.
cle class of p38 MAP kinase inhibitors. Curr. Top. Med. Chem. 24. Hynes, J., Jr.; Dyckman, A. J.; Lin, S.; Wrobleski, S. T.; Wu,
2002, 2, 1021–1035. H.; Gillooly, K. M.; Kanner, S. B.; Lonial, H.; Loo, D.; McIn-
11. Doweyko, A. M.; Wrobleski, S. T. A comparison of p38 tyre, K. W.; Pitt, S.; Shen, D. R.; Shuster, D. J.; Yang, X.;
inhibitor-protein structures. Am. Drug Discov. 2006, 1, 47–52. Zhang, R.; Behnia, K.; Zhang, H.; Marathe, P. H.; Doweyko,
12. Dumas, J.; Sibley, R.; Riedl, B.; Monahan, M. K.; Lee, W.; A. M.; Tokarski, J. S.; Sack, J. S.; Pokross, M.; Kiefer, S. E.;
Lowinger, T. B.; Redman, A. M.; Johnson, J. S.; Kingery-Wood, Newitt, J. A.; Barrish, J. C.; Dodd, J.; Schieven, G. L.; Left-
J.; Scott, W. J.; Smith, R. A.; Bobko, M.; Schoenleber, R.; Ranges, heris, K. Design, synthesis, and anti-inflammatory properties
G. E.; Housley, T. J.; Bhargava, A.; Wilhelm, S. M.; Shrikhande, of orally active 4-(phenylamino)-pyrrolo[2,1-f ][1,2,4]triazine
A. Discovery of a new class of p38 kinase inhibitors. Bioorg. p38␣ mitogen-activated protein kinase inhibitors. J. Med.
Med. Chem. Lett. 2000, 10, 2047–2050. Chem. 2008, 51, 4–16.
13. Dumas, J.; Smith, R. A.; Lowinger, T. B. Recent developments in 25. Hunt, J. T.; Mitt, T.; Borzilleri, R.; Gullo-Brown, J.; Fargnoli, J.;
the discovery of protein kinase inhibitors from the urea class. Fink, B.; Han, W.-C.; Mortillo, S.; Vite, G.; Wautlet, B.; Wong,
Curr. Opin. Drug Discov. Dev. 2004, 7, 600–616. T.; Yu, C.; Zheng, Z.; Bhide, R. Discovery of the pyrrolo[2,1-
14. Hanson, G. Inhibitors of p38 kinase. Expert Opin. Ther. 1997, f ][1,2,4]triazine nucleus as a new kinase inhibitor template.
7, 729–733. J. Med. Chem. 2004, 47, 4054–4059.
15. Jackson, P. F.; Bullington, J. L. Pyridinylimidazole based p38 26. Pargellis, C.; Tong, L.; Churchill, L.; Cirillo, P. F.; Gilmore, T.;
MAP kinase inhibitors. Curr. Top. Med. Chem. 2002, 2, 1011– Graham, A. G.; Grob, P. M.; Hickey, E. R.; Moss, N.; Pav, S.;
1020. Regan, J. Inhibition of p38 MAP kinase by utilizing a novel
16. Salituro, F. G.; Germann, U. A.; Wilson, K. P.; Benis, G. W.; allosteric binding site. Nat. Struct. Biol. 2002, 9, 268–272.
Fox, T.; Su, M. S.-S. Inhibitors of p38 MAP kinase: therapeutic 27. Wrobleski, S. T.; Lin, S.; Hynes, J.; Wu, H.; Pitt, S.; Shen,
intervention in cytokine-mediated diseases. Curr. Med. Chem. D. R.; Zhang, R.; Gillooly, K. M.; Shuster, D. J.; McIntyre,
1999, 6, 807–823. K. W.; Doweyko, A. M.; Kish, K. F.; Tredup, J. A.; Duke, G. J.;
17. Wrobleski, S. T.; Doweyko, A. M. Structural comparison of p38 Sack, J. S.; McKinnon, M.; Dodd, J.; Barrish, J. C.; Schieven,
inhibitor-protein complexes: a review of recent p38 inhibitors G. L.; Lefterheris, K. Synthesis and SAR of new pyrrolo[2,1-
having unique binding interactions. Curr. Top. Med. Chem. f ][1,2,4]triazines as potent p38␣ MAP kinase inhibitors.
2005, 5, 1005–1016. Bioorg. Med. Chem. Lett. 2008, 18, 2739–2744.
18. Berman, H. M.; Westbrook, J.; Feng, Z.; Gilliland, G.; Bhat, T. N.; 28. Das, J.; Moquin, R. V.; Pitt, S.; Zhang, R.; Shen, D. R.; McIntyre,
Weissig, H.; Shindyalov, I. N.; Bourne, P. E. The Protein Data K. W.; Gillooly, K. M.; Doweyko, A. M.; Sack, J. S.; Zhang, H.;
Bank. Nucleic Acids Res. 2000, 28, 235–242. Kiefer, S. E.; Kish, K. F.; McKinnon, M.; Barrish, J. C.; Dodd,
19. Lin, T. H.; Metzger, A.; Diller, D. J.; Desai, M.; Henderson, I.; J.; Schieven, G. L.; Lefterheris, K. Pyrazolo-pyrimidines:
Ahmed, G.; Kimble, E. F.; Quadros, E.; Webb, M. L. Discov- a novel heterocyclic scaffold for potent and selective
ery and characterization of triaminotriazine aniline amides as p38␣ inhibitors. Bioorg. Med. Chem. Lett. 2008, 18, 2652–
highly selective p38 kinase inhibitors. J. Pharmacol. Exp. Ther. 2657.
2006, 318, 495–502. 29. Hynes, J.; Wu, H.; Pitt, S.; Shen, D. R.; Zhang, R.; Schieven,
20. Leftheris, K.; Ahmed, G.; Chan, R.; Dyckman, A. J.; Hussain, G. L.; Gillooly, K. M.; Shuster, D. J.; Taylor, T. L.; Yang,
Z.; Ho, K.; Hynes, J., Jr.; Letourneau, J.; Li, W.; Lin, S.; Metzger, X.; McIntyre, K. W.; McKinnon, M.; Zhang, H.; Marathe,
A.; Moriarty, K. J.; Riviello, C.; Shimshock, Y.; Wen, J.; Wityak, P. H.; Doweyko, A. M.; Kish, K.; Kiefer, S. E.; Sack, J. S.;
J.; Wrobleski, S. T.; Wu, H.; Wu, J.; Desai, M.; Gillooly, K. M.; Newitt, J. A.; Barrish, J. C.; Dodd, J.; Leftheris, K. The
Lin, T. H.; Loo, D.; McIntyre, K. W.; Pitt, S.; Shen, D. R.; Shus- discovery of (R)-2-(sec-butylamino)-N-(2-methyl-5-(methyl-
ter, D. J.; Zhang, R.; Diller, D.; Doweyko, A.; Sack, J.; Baldwin, carbamoyl)phenyl) thiazole-5-carboxamide (BMS-640994) – A
J.; Barrish, J.; Dodd, J.; Henderson, I.; Kanner, S.; Schieven, potent and efficacious p38␣ MAP kinase inhibitor. Bioorg.
G. L.; Webb, M. The discovery of orally active triaminotriazine Med. Chem. Lett. 2008, 18, 1762–1767.
aniline amides as inhibitors of p38 MAP kinase J. Med. Chem. 30. Mavunkel, B. J.; Chakravarty, S.; Perumattam, J. J.; Luedtke,
2004, 47, 6283–6291. G. R.; Liang, X.; Lim, D.; Xu, Y.-J.; Laney, M.; Liu, D. Y.;
21. Chen, J. C.; Xu, S. L.; Wawrzak, Z.; Basarab, G. S.; Jordan, D. B. Schreiner, G. F.; Lewicki, J. A.; Dugar, S. Indole-based hetero-
Structure-based design of potent inhibitors of scytalone dehy- cyclic inhibitors of p38␣ MAP kinase: designing a conforma-
dratase: displacement of a water molecule from the active site. tionally restricted analogue. Bioorg. Med. Chem. Lett. 2003, 13,
Biochemistry 1998, 37, 17735–17744. 3087–3090.
22. Wissner, A.; Berger, D. M.; Boschelli, D. H.; Floyd, B.; Green- 31. Dhar, T. G. M.; Wrobleski, S. T.; Lin, S.; Furch, J. A.; Nirschl,
berger, L. M.; Gruber, B. C.; Johnson, B. D.; Mamuya, N.; D. S.; Fan, Y.; Todderud, G.; Pitt, S.; Doweyko, A. M.; Sack,
Nilakantan, R.; Reich, M. F.; Shen, R.; Tsou, H.-R.; Upes- J. S.; Mathur, A.; McKinnon, M.; Barrish, J. C.; Dodd, J. H.;
lacis, E.; Wang, Y. F.; Wu, B.; Ye, F.; Zhang, N. 4-Anilino- Schieven, G. L.; Leftheris, K. Synthesis and SAR of p38␣ MAP
6,7-dialkoxyquinoline-3-carbonitrile inhibitors of epidermal kinase inhibitors based on heterobicyclic scaffolds. Bioorg.
growth factor receptor kinase and their bioisosteric rela- Med. Chem. Lett. 2007, 17, 5019–5024.
tionship to the 4-anilino-6,7-dialkoxyquinazoline inhibitors. 32. Boehm, J. C.; Smietana, J. M.; Sorenson, M. E.; Garigipati,
J. Med. Chem. 2000, 43, 3244–3256. R. S.; Gallagher, T. F.; Sheldrake, P. L.; Breadbeer, J.; Bad-
23. Liu, C.; Wrobleski, S. T.; Lin, J.; Ahmed, G.; Metzger, A.; Wityak, ger, A. M.; Laydon, J. T.; Lee, J. C.; Hillegass, L. M.; Gris-
J.; Gillooly, K. M.; Shuster, D. J.; McIntyre, K. W.; Pitt, S.; Shen, wold, D. E.; Breton, J. J.; Chabot-Fletcher, M. C.; Adams,
D. R.; Zhang, R. F.; Zhang, H.; Doweyko, A. M.; Diller, D.; Hen- J. L. 1-Substituted 4-aryl-5-pyridinylimidazoles: a new class
derson, I.; Barrish, J. C.; Dodd, J. H.; Schieven, G. L.; Leftheris, of cytokine suppressive drugs with low 5-lipoxygenase and

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015
 208 Arthur M. Doweyko

cyclooxygenase inhibitory potency. J. Med. Chem. 1996, 39, 35. McClure, K. F.; Letavic, M. A.; Kalgutkar, A. S.; Gabel, C. A.;
3929–3937. Audoly, L.; Barberia, J. T.; Braganza, J. F.; Carter, D.; Carty,
33. Trejo, A.; Arzeno, H.; Browner, M.; Chanda, S.; Cheng, S.; T. J.; Cortina, S. R.; Dombroski, M. A.; Donahue, K. M.; Elliott,
Comer, D. D.; Dalrymple, S. A.; Dunten, P.; Lafargue, J.; Love- N. C.; Gibbons, C. P.; Jordan, C. K.; Kuperman, A. V.; Labasi,
joy, B.; Freire-Moar, J.; Lim, J.; McIntosh, J.; Miller, J.; Papp, J. M.; LaLiberte, R. E.; McCoy, J. M.; Naiman, B. M.; Nelson,
E.; Reuter, D.; Roberts, R.; Sanpablo, F.; Saunders, J.; Song, K. L.; Nguyen, H. T.; Peese, K. M.; Sweeney, F. J.; Taylor, T. J.;
K.; Villasenor, A.; Warren, S. D.; Welch, M.; Weller, P.; White- Trebino, C. E.; Abramov, Y. A.; Laird, E. R.; Volberg, W. A.; Zhou,
ley, P. E.; Zeng, L.; Goldstein, D. M. Design and synthesis of J.; Bach, J.; Lombardo, F. Structure-activity relationships of tri-
4-azaindoles as inhibitors of p38 MAP kinase. J. Med. Chem. azolopyridine oxazole p38 inhibitors: identification of candi-
2003, 46, 4702–4713. dates for clinical development. Bioorg. Med. Chem. Lett. 2006,
34. Dombroski, M. A.; Letavic, M. A.; McClure, K. F.; Barberia, J. T.; 16, 4339–4344.
Carty, T. J.; Cortina, S. R.; Csiki, C.; Dipesa, A. J.; Elliott, N. C.; 36. Liu, C.; Lin, J.; Pitt, S.; Zhang, R. F.; Sack, J. S.; Kiefer, S. E.; Kish,
Gabel, C. A.; Jordan, C. K.; Labasi, J. M.; Martin, W. H.; Peese, K.; Doweyko, A. M.; Zhang, H.; Marathe, P. H.; Trzaskos, J.;
K. M.; Stock, I. A.; Svensson, L.; Sweeney, F. J.; Yu, C. H. Benz- McKinnon, M.; Dodd, J. H.; Barrish, J. C.; Schieven, G. L.; Left-
imidazolone p38 inhibitors. Bioorg. Med. Chem. Lett. 2004, 14, heris, K. Benzothiazole based inhibitors of p38␣ MAP kinase.
919–923. Bioorg. Med. Chem. Lett. 2008, 18, 1874–1879.

Downloaded from https://www.cambridge.org/core. Newcastle University, on 22 Mar 2019 at 15:29:12, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms.
https://doi.org/10.1017/CBO9780511730412.015