Discovery of 1 H -pyrazol-3(2 H)-ones as potent and selective inhibitors of protein kinase R-like endoplasmic reticulum kinase (PERK)

Article abstract of DOI:10.1021/jm5017494

The structure-based design and optimization of a novel series of selective PERK inhibitors are described resulting in the identification of 44 as a potent, highly selective, and orally active tool compound suitable for PERK pathway biology exploration both in vitro and in vivo.

Full text of DOI:10.1021/jm5017494

Article
pubs.acs.org/jmc
Discovery of 1H‑Pyrazol-3(2H)‑ones as Potent and Selective
Inhibitors of Protein Kinase R‑like Endoplasmic Reticulum Kinase
(PERK)
Adrian L. Smith,*,† Kristin L. Andrews,‡ Holger Beckmann,¶ Steven F. Bellon,⊥ Pedro J. Beltran,§
Shon Booker,† Hao Chen,# Young-Ah Chung,§ Noel D. D’Angelo,† Jennifer Dao,‡
Kenneth R. Dellamaggiore,§ Peter Jaeckel,¶ Richard Kendall,§ Katja Labitzke,¶ Alexander M. Long,⊥
Silvia Materna-Reichelt,¶ Petia Mitchell,§ Mark H. Norman,† David Powers,‡ Mark Rose,∥
Paul L. Shaffer,⊥ Michelle M. Wu,§ and J. Russell Lipford*,§
†
‡
§
∥
Departments of Medicinal Chemistry, Molecular Structure and Characterization, Oncology Research, and Pharmacokinetics and
Drug Metabolism, Amgen Inc., One Amgen Center Drive, Thousand Oaks, California 91320-1799, United States
⊥
#
Departments of Molecular Structure and Characterization and Protein Technologies, Amgen Inc., 360 Binney Street, Cambridge,
Massachusetts 02142, United States
¶Amgen Research GmbH, Josef-Engert-Straße 11, D-93053 Regensburg, Germany
S
* Supporting Information
ABSTRACT: The structure-based design and optimization of
a novel series of selective PERK inhibitors are described
resulting in the identification of 44 as a potent, highly selective,
and orally active tool compound suitable for PERK pathway
biology exploration both in vitro and in vivo.
may be useful for the treatment of cancer9−11 either as a sole
agent or in combination with other anticancer treatments, thus
spurring recent interest in the identification of UPR pathway
INTRODUCTION
■
The unfolded protein response (UPR) is an evolutionarily
conserved mechanism in mammals by which cells respond to
endoplasmic reticulum (ER) stress.1 Cellular ER stress is
inhibitors.
PERK is an ER transmembrane protein with a stress-sensing
characterized by an accumulation of misfolded proteins in the
domain inside the ER lumen and a cytosolic kinase domain.10
ER lumen resulting from an inability of the cell to adequately
Upon sensing misfolded proteins, PERK is activated by
process protein production in the ER, for example, under
autophosphorylation and oligomerization through release of
nutrient deprivation conditions (hypoxia, glucose starvation) or
BiP/Grp78 from the stress-sensing domain. Activated PERK
where a high secretory load exists. There are three sensors of
phosphorylates and activates its downstream substrate,
misfolded proteins that are known to mediate the UPR through
complementary pathways: protein kinase R-like endoplasmic
eukaryotic initiation factor 2α (eIF2α), which inhibits the
ribosome translation initiation complex in order to attenuate
reticulum kinase (PERK), inositol-requiring enzyme 1 (IRE1),
and activating transcription factor 6 (ATF6).2 PERK, IRE1, and
protein synthesis. This serves to prevent exacerbation of ER
stress by preventing the accumulation of additional misfolded
ATF6 are all located in the ER membrane and sense misfolded
proteins. Although it inhibits general protein synthesis,
proteins in the ER lumen through the molecular chaperone
BiP/Grp78 and/or direct binding of unfolded proteins to the
activated eIF2α causes the translation of specific mRNAs
sensor domains.3 The UPR serves as a mechanism for cellular
involved in restoring ER homeostasis including activating
transcription factor 4 (ATF4). ATF4 mediates the transcription
survival whereby cells are able to adapt to cope with ER stress,
of certain UPR target genes including those for the
endoplasmic-reticulum-associated protein degradation
(ERAD) pathway proteins which target misfolded proteins
but under extreme stress the UPR switches the cellular
machinery toward apoptosis.4−6 The UPR is known to be
active in highly secretory tissues such as the pancreas,7 but
tumors are also thought to utilize the UPR for survival under
stressed conditions such as nutrient deprivation or chemo-
therapeutic insult.8 This suggests that modulators of the UPR
Received: November 11, 2014
© XXXX American Chemical Society
A
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
for ubiquitination and degradation by the proteasome. ATF4
also causes the expression of the transcription factor C/EBP
homologous protein (CHoP), which sensitizes cells to ER
stress-mediated apoptosis, providing a pathway for regulated
removal of severely stressed cells by the organism. There were
no known selective PERK inhibitors at the outset of our work
exploring the UPR, and this prompted us to initiate a search for
tool compounds to validate PERK as an oncology target. The
first potent and selective PERK inhibitors have since been
disclosed by Axten12,13 including GSK2606414, its fluorinated
analog 1, and the analog GSK2656157 which was optimized for
preclinical evaluation (Figure 1). Our research leading to the
by selective C-acylation of the 1H-pyrazol-3-(2H)-one 1218 and
subsequent N-methylation. Reduction of the nitro intermedi-
ates 13a,c to the corresponding amino intermediates 14a,c was
conveniently accomplished with iron in acetic acid. 6-
Quinazolinylboronate ester intermediates 19a−c were prepared
as shown in Scheme 4; intermediates 614 and 17a19 have been
previously described, and 18c was acquired from commercial
sources. The 2-methylaminoquinazoline 18b was most readily
accessed via a three-step process from fluoroaldehyde 15
involving initial formation of the 2-aminoquinazoline 16,
chloro-deamination using a modification of the Sandmeyer
reaction (to give 17b), followed by amination with methyl-
amine. More direct routes from 15 to 18b, such as reaction of
15 with methylguanidine or methylation of 16, were problem-
atic because of regiochemistry issues. The 6-bromoquinazolines
18a−c were readily converted to the corresponding boronate
esters 19a−c.20 The corresponding 2-methyl-6-quinolylboro-
nate ester 20 is commercially available. Benzothiazolylboronate
ester intermediates were accessed as outlined in Scheme 5. The
benzothiazole derivatives 21−24 and 28c were obtained from
commercial sources, and the fluorinated derivatives 28a,b were
accessed using variations on a procedure involving N-acylation
of the 2-aminothiophenols 25 and 26 followed by facile
cyclization−dehydration under reductive conditions to form
the benzothiazoles 28a,b. The bromides 28a−c were then
converted to the corresponding boronate esters 29a−c.
The final stages of the syntheses of the PERK inhibitors 30−
44 are shown Scheme 6. Compounds 30−33, 37−40, and 42−
44 were directly synthesized in one step from the appropriate
benzoyl-1H-pyrazol-3(2H)-one intermediates 13b, 13d, 14a,
and 14c using a Suzuki−Miyaura coupling with the
corresponding boronate esters shown in Schemes 4 and 5.
Compounds 34−36 and 41 were prepared in a generally less
efficient overall two-step process by similarly coupling the 4-(2-
nitrobenzoyl)-1H-pyrazol-3(2H)-one 13a with the correspond-
ing boronate ester, followed by reduction of the nitro group.
The PERK inhibitor 52 was synthesized in a convergent
manner using a Suzuki−Miyaura coupling to join the fragments
48 and 51 (Scheme 7).
Figure 1. Structures of previously disclosed PERK inhibitors.
identification of a novel series of potent, orally active PERK
inhibitors with high selectivity over other kinases is described in
this article. The use of these inhibitors for exploration of PERK
pathway biology as well as PERK target validation in vivo will
be described elsewhere.
RESULTS AND DISCUSSION
■
Chemistry. The screening hit 7, originally prepared as part
of an earlier exploration of B-Raf inhibitor structure−activity
relationships (SARs),14 was readily prepared in three steps (an
amide coupling, a nitro reduction, and a Suzuki−Miyaura
coupling15) as shown in Scheme 1. The screening hits 10a,b
originated from a prior c-MET inhibitor program and were
synthesized from previously reported intermediates16,17 using
more forcing amide coupling conditions (Scheme 2).
Lead Optimization and SAR. No suitable medicinal
chemistry leads for deriving a selective PERK inhibitor tool
compound were known at the outset of this work, and so we
performed a high-throughput screen of the Amgen sample
collection looking for inhibitors of PERK kinase activity
(phosphorylation of the downstream substrate eIF2α). Hits
were assessed for cellular activity in both a mechanism-based
Target molecules were synthesized from the key inter-
mediates shown in Schemes 3−5. 4-Benzoyl-1H-pyrazol-
3(2H)-one intermediates 13a−d (Scheme 3) were prepared
a
Scheme 1. Synthesis of Screening Hit 7
a
Reagents and conditions: (a) TBTU, i-Pr2NEt, DMF, rt, 2 h; (b) Fe, AcOH, EtOH, Δ, 3 h; (c) PdCl2(PPh3)2 (cat.), Na2CO3·H2O, DMF, H2O,
150 °C (microwave), 15 min.
B
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of Screening Hits 10a,b
a
Reagents and conditions: (a) HATU, Et3N, DMF, 90 °C, 16 h.
a
Scheme 3. Synthesis of 4-Benzoyl-1H-pyrazol-3(2H)-one Intermediates 13a−d, 14a, and 14c
a
Reagents and conditions: (a) Ca(OH)2, 1,4-dioxane, 90 °C, 16 h; (b) methyl p-toluenesulfonate, 170 °C, 3 h; (c) Fe, AcOH, EtOH, Δ, 1.5 h.
a
Scheme 4. Synthesis of 6-Quinazolinylboronate Ester Intermediates 19a−c
a
Reagents and conditions: (a) guanidine carbonate, i-Pr2NEt, NMP, 140 °C, 2 h; (b) n-Bu4NCl, (CH3)3SiCl, t-BuONO, 0−50 °C, 1 h; (c) CH3−
NH2, EtOH, 80 °C, 1 h; (d) bis(pinacolato)diboron, Pd(dppf)2Cl2·CH2Cl2, KOAc, 1,4-dioxane, Δ, 2.5 h.
a
Scheme 5. Synthesis of Benzothiazolylboronate Ester Intermediates 29a−c
a
Reagents and conditions: (a) difluoroacetyl chloride, pyridine, CH2Cl2, rt, 2 h. (b) For 28a: fluoroacetic acid, 25, PPh3, Et3N, CCl4, 95 °C, 1.5 h.
For 28b: 27, dithiothreitol, DMF, rt, 16 h. (c) For 29a−c: bis(pinacolato)diboron, Pd(dppf)2Cl2·CH2Cl2, KOAc, 1,4-dioxane, 120 °C (microwave),
30 min.
C
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 6. Synthesis of PERK Inhibitors 30−44
a
Reagents and conditions: (a) boronate ester (Schemes 4 and 5). Method A: Pd(PPh3)4, K3PO4, toluene, MeOH, water, 100 °C, 8 h. Method B:
bis(di-tert-butyl(4-dimethylaminophenyl)phosphine)dichloropalladium(II), Na2CO3, DMF, water, 110 °C, 30 min. (b) For 13a → 34−36, 41: Fe,
AcOH, EtOH, 80 °C, 2.5 h.
a
Scheme 7. Synthesis of PERK Inhibitor 52
a
Reagents and conditions: (a) 45, CDI, THF, 60 °C, 2 h, then 46, rt, 20 h; (b) bis(pinacolato)diboron, Pd(dppf)2Cl2·CH2Cl2, KOAc, 1,4-dioxane,
90 °C, 16 h; (c) NaH, CH3I, DMF, 0 °C → rt, 16 h; (d) NH3, 1,4-dioxane, water, 135 °C (microwave), 2 h; (e) Pd(PPh3)4, K3PO4, toluene, MeOH,
water, 110 °C (microwave), 20 min.
ELISA assay looking at cellular inhibition of PERK
autophosphorylation (using a phospho-PERK T980 (pPERK)
specific antibody in HT1080 cells stably transfected with
doxycycline-inducible T-REx-PERK-FLAG)21 and a functional
imaging assay looking at the rescue of bulk protein synthesis in
stressed U2OS cells (incorporation of the methionine analog L-
azidohomoalanine into protein in methionine-starved media
with thapsigargin-induced ER stress; imaging via click
chemistry22 with an an AlexaFluor 488 alkyne dye). Hits
were also counterscreened against the other known eIF2α
kinases (general control nonderepressible 2 (GCN2), protein
kinase RNA-activated (PKR), and heme-regulated eIF2α kinase
(HRI)), and IRE-1 (the mediator of an alternative ER-stress
pathway). Two series of compounds stood out from the screen,
represented by compounds 7 (PERK IC50 = 41 nM) and 10a/
10b (PERK IC50 = 4/42 nM). Both series demonstrated good
cellular activity but had significant inhibitory activity against
GCN2 (Table 1). GCN2 shares close homology with PERK in
the kinase domain, and since it mediates a different stress-
response pathway in response to amino acid starvation or UV
damage (also signaling through eIF2α), the GCN2 activity
posed a potential problem for evaluating on-target effects of
PERK inhibition. Neither series inhibited IRE-1 (IC50 >50
μM), and selectivity over PKR and HRI was not an issue (7:
PKR IC50 > 10 μM; HRI IC50 > 10 μM. 10a: PKR IC50 = 480
nM; HRI IC50 > 10 μM).
Compound 7 originated from a prior B-Raf inhibitor
program and, as might be expected, had significant activity in
B-Raf mediated assays14 (B-Raf IC50 = 8 nM; A375 pERK IC50
= 11 nM). Apart from the B-Raf activity, 7 was moderately
selective over other kinases in a kinase inhibition panel screen
(Table 1). Given our prior experience with B-Raf inhibitors
enhancing the proliferation of certain cell lines,23 our primary
objective with this series was to completely remove the B-Raf
D
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Table 1. In Vitro Data for PERK Inhibitors 1, 7, 10a, 30−44, and 52
enzyme assay
GCN2
cellular assay
kinase selectivity
protein synthesis
rescue
PERK
CHoP
GCN2
CHoP
b
selectivity ratio
(%)
number of kinases
tested
compd
PERK
B-Raf
pPERK
pGCN2
IC50 (nM) IC50 (nM) IC50 (nM) IC50 (nM)
EC50 (nM)
IC50 (nM) IC50 (nM) IC50 (nM)
1
1
3900
420
290
42
>1000
8
6
13
20
nd
>50000
640
6
387
100
7
41
4
65
310
520
130
310
640
280
130
200
180
5600
190
200
210
170
290
1300
400
52
1200
nd
nd
27
nd
16
24
17
20
9
10a
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
52
>1000
300
nd
12
nd
nd
3
480
480
320
1000
400
650
2300
nd
17
550
100
100
100
100
381
100
100
1
5
8
110
45
nd
140
3
12
66
nd
240
5
10
13
140
30
nd
120
4
36
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
nd
520
6
42
59
nd
1100
6000
nd
5
5
240
>10000
1400
97
35
nd
4
240
4
>10000
20
nd
nd
2
320
760
1100
2100
4800
3000
770
520
50
>50000
1300
1200
3700
1900
>50000
>50000
>50000
100
387
100
100
387
100
387
387
9
82
2
11
6
100
310
610
>50 000
7300
>50000
41
44
4
65
53
0
11
53
6
86
79
0.3
1
88
>50000
>50000
nd
84
0
2
12
89
0.5
a
b
Data represent an average of at least two determinations. nd: not determined. KINOMEscan: percentage of tested kinases with >65% inhibition at
1 μM test concentration.
inhibitory activity while maintaining or improving the activity
against PERK. The inhibitors 10a,b came from a prior c-MET
inhibitor program16 (10a/10b: c-MET IC50 = 6/1 nM). This
series had moderate selectivity over other kinases (10b: >65%
inhibition of 10 out of 47 kinases tested at 1 μM), and the main
objective for optimization was to maintain the excellent activity
against PERK while improving upon the kinase selectivity.
Among the optimization strategies available, combining
structural features of 7 and 10a,b into a hybrid molecule
proved to be particularly successful (Figure 2). Cocrystal
structures of 7 and 10b were obtained with PERK, and the
inhibitors were both observed to bind to the inactive DFG-
out24 conformation of the protein. Overlay of the cocrystal
structures of 7 and 10b with PERK indicated almost identical
alignment of the amide carbonyls of the two inhibitors with
PERK, and the hybrid structure 30 was therefore suggested
from the 10-fold greater potency of 10a compared with 10b
against PERK (4 nM vs 42 nM). Compound 30 maintained
good activity against PERK (Table 1), and it had good overall
kinase selectivity and microsomal stability (rat CLint = 29 μL
min−1 mg−1). Compound 30 retained some residual B-Raf
activity at a level that was found to be sufficient to still cause
activation of MAPK signaling in sensitive cell lines (EC50 = 190
nM in MIA PaCa-2 cells),23 and activity against GCN2 was also
substantially increased resulting in only a 12-fold selectivity
window for PERK over GCN2 at the enzyme level.
The problematic inclusion of GCN2 inhibitory activity, with
its attendant consequences for complicating interpretation of
the functional activity of dual PERK/GCN2 inhibitors, led to
the exploration of assays to give a cellular readout of the GCN2
inhibition being observed (Table 1). A MSD assay was
developed for phospho-GCN2 (pGCN2) giving a direct
mechanistic readout of endogenous GCN2 autophosphoryla-
tion in HT1080 cells stimulated by amino acid starvation. A
CHoP bDNA assay (HT1080 cell line) was also developed for
measuring the relative downstream pathway effects of PERK
and GCN2 inhibition for dual PERK/GCN2 inhibitors such as
30. The PERK pathway was activated with thapsigargin and the
GCN2 pathway was separately activated by amino acid
starvation, both resulting in CHoP mRNA expression by
E
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 2. Design of a hybrid series of PERK inhibitors. Overlay of cocrystal structures of 7 and 10b with PERK suggested hybrid structure 30.
signaling through ATF4. The inhibitory effect of compounds
on CHoP mRNA expression resulting from either PERK
inhibition or GCN2 inhibition could then be measured. As can
be seen from Table 1, downstream signaling through ATF4
appears to be more sensitive to GCN2 inhibition than PERK
inhibition (cf. PERK CHoP vs GCN2 CHoP). Thus,
compound 30 is equipotent at inhibiting PERK and GCN2
autophosphorylation and is similarly equipotent at inhibiting
CHoP mRNA expression through the PERK and GCN2
pathways, despite having some selectivity at the PERK/GCN2
enzyme level.
A brief exploration of substituent SAR around the central
benzoyl ring was performed first. Deletion of the amino group
(31) had a modest effect on PERK activity but increased off-
target GCN2 and B-Raf activity and led to poor microsomal
stability (rat CLint = 497 μL min−1 mg−1). Swapping the methyl
group for an amino group (32) or a chloro group (33) again
had a minimal effect on PERK activity, but activity against
GCN2 and B-Raf was substantial for both compounds and 33
again had relatively poor microsomal stability (rat CLint = 167
μL min−1 mg−1).
Figure 3. Cocrystal structure of 34 solved in complex with PERK at
1.8 Å resolution.
methionine gatekeeper residue (Met888) that accommodates
small substituents from the quinazoline 5-position; the side
chain of the corresponding threonine gatekeeper residue of B-
Raf does not.
The upper aminoquinazoline fragment of 30 proved to be
the most fruitful area for SAR exploration. The 5-
fluoroquinazoline 34 and 5-chloroquinazoline 35 had sub-
stantially reduced activity against B-Raf while maintaining
PERK activity. Compounds 34 and 35 also demonstrated
improved selectivity against a broad panel of kinases (Table 1).
A cocrystal structure of 34 with PERK (Figure 3) illustrates the
binding mode of this series in the ATP binding site of the
PERK kinase domain. The protein adopts a conformation in
which the activation loop is folded out (DFG-out),24 with the
5-phenyl substituent of the pyrazolone occupying the DFG-out
pocket. The 2-phenyl substituent of the pyrazolone occupies a
pocket adjacent to the αC-helix of the kinase. The pyrazolone
carbonyl forms a hydrogen bond with the catalytic lysine
(Lys622). The benzoyl carbonyl forms a hydrogen bond with
the backbone N−H of the DFG Asp955, and the amino-
quinazoline moiety forms hydrogen bonds to the kinase hinge
region Cys891. The source of the improved PERK/B-Raf
selectivity of 34 and 35 relative to 30 arises from the difference
in the gatekeeper residue between the two kinases. PERK has a
Compounds 36−39 represent a key progression in SAR
around the quinazoline hinge-binder. First, replacing the 2-
methylamino substituent of the quinazoline 30 with a 2-methyl
substituent (36) retained PERK activity but was detrimental to
GCN2 activity. Second, our cocrystal structural data suggested
that the quinazoline moiety of compounds such as 34 had the
potential to adopt two alternative binding modes with either
quinazoline ring N atom capable of hydrogen-bonding to the
Cys891 backbone N−H. This was explored by changing the
hinge-binding heterocycle to the benzothiazole 37 (designed to
place the benzothiazole N in proximity to the Cys891 N−H by
analogy with 34, Figure 3). Compound 37 had poor PERK
activity, but switching the point of attachment of the
benzothiazole to the benzoyl core in order to give access to
the alternative binding mode was well-accommodated by
PERK. Thus, benzothiazole 38 maintained excellent activity
against PERK and demonstrated good selectivity against
GCN2, B-Raf, and a broad panel of other kinases. Reinstalling
the 2-methylamino substituent on the benzothiazole (39)
restored activity against GCN2 but retained excellent selectivity
F
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
against a broad panel of kinases. Compound 40 was identified
as the primary circulating metabolite of 39 in rodent studies,
and it had a very similar in vitro profile to 39.
A cocrystal structure of 39 with PERK (Figure 4) revealed
that while the general binding mode to PERK was similar to
trifluoromethyl analog 43 was identified as a selective PERK
inhibitor devoid of measurable GCN2 activity. Selectivity over
other kinases and rat pharmacokinetic properties were both
excellent for 43.
The final piece of SAR around the kinase hinge-binding
moiety involved changing the 2-methylbenzothiazole of 38 for
the analogous 2-methylquinoline of 44. Compound 44 was an
equally potent PERK inhibitor in vitro, with 1000-fold
selectivity over GCN2 and with no measurable effects due to
GCN2 inhibition in the cellular assays. It had exquisite
selectivity over a large panel of kinases, and while the rat
pharmacokinetic properties were only moderate, the mouse
pharmacokinetic properties were significantly better (Table 2).
Oral bioavailability of 44 in rat (and mouse) was significantly
improved compared with that of 38, making it a superior in
vivo tool compound. A cocrystal structure of 44 with PERK
(Figure 5a) revealed that the quinoline methyl group projects
toward the hinge Cys891 carbonyl and the distance (3.4 Å) is
suggestive of a hydrogen bond between a methyl C−H and the
carbonyl oxygen.25 The source of the selectivity imparted by
the quinoline methyl group of 44 over GCN2 is suggested by
comparing the apo structures26 of mouse PERK27 and yeast
GCN228 (Figure 5b), where it is evident that the hinge Cys
residues of the two proteins adopt different conformations in
the apo state. If this difference translates to the protein
structure with inhibitor bound, it would place the hinge Cys
carbonyl of GCN2 closer to the quinoline methyl group of 44,
leading to an unfavorable interaction.
Figure 4. Cocrystal structure of 39 solved in complex with PERK at
1.9 Å resolution.
that of 34, the benzo-fused ring of the benzothiazole moiety
adopted the previously noted alternative binding mode and was
pushed much further back toward the gatekeeper residue
(Met888) compared with 34, thus accounting for the very high
selectivity of 39 over other kinases.
Disclosure of the first potent, selective PERK inhibitors by
Axten12,13 (Figure 1) was contemporaneous with our
identification of 44 as a potent, selective PERK inhibitor. We
prepared the analog 1 as a benchmark and confirmed its activity
(Table 1). The cocrystal structures of this series12,13 with PERK
indicated a conformation of the kinase region different from
that observed with our inhibitors, with the activation loop
folded in (DFG-in). However, these structures clearly indicated
that the addition of a methyl group to the 2-position of the 7H-
pyrrolo[2,3-d]pyrimidine would be accommodated by the
protein in a manner similar to the 2-methylquinoline
substituent of 44. We wondered if this substitution would
further improve upon the already good kinase selectivity of 1
(Table 1) in an analogous manner to the improvement
observed with 30 → 36. Compound 52 was therefore prepared
and profiled. While compound 52 was marginally less potent in
vitro when compared with 1, the kinase selectivity profile was
significantly improved and it had an outstanding rat
pharmacokinetic profile (Table 2). The cocrystal structure of
52 with PERK (Figure 6) confirmed the proposed role of the
additional methyl group, with a distance of 3.2 Å between the
methyl group and Cys891 strongly suggestive of a hydrogen
bond similar to the observation with 44. It should be noted that
both 1 and 52 are significantly more potent in the PERK
cellular assays compared with the compounds described in the
pyrazolone series, despite having apparently similar PERK
enzyme inhibitory activities. This may be a reflection of these
compounds reaching the lower readout limit in the PERK
enzyme assay due to the 1 nM concentration of PERK protein
used in the enzyme assay.
The fluorinated analogs 41−43 were prepared as part of the
SAR exploration of 38. Both the monofluoro- and difluor-
omethyl analogs 41 and 42 were found to be functionally
balanced dual PERK/GCN2 inhibitors. The inhibitor 42 had
superb selectivity over other kinases and excellent rat
pharmacokinetic properties (Table 2). By contrast, the
Table 2. In Vivo Pharmacokinetic Properties of PERK
a
Inhibitors
b
iv (1 mg/kg)
CL
po
MRT
(h)
Vss
AUC
F
compd
species
(L/(h·kg))
(L/kg)
(μM·h)
(%)
c
c
1
rat
3.7
10.4
0.5
3.9
5.5
1.4
2.2
3.1
2.2
5.3
10.8
2.3
5.0
11.9
0.29
0.07
1.0
1.0
0.73
0.56
2.2
2.7
2.6
3.1
2.0
6.5
5.0
8.1
3.6
1.2
1.2
8.2
44
d
d
10a
32
34
35
36
38
39
40
42
43
44
44
52
rat
20.2
e
50
e
rat
0.8
18
c
c
rat
0.57
0.50
2.2
3.8
48
c
c
rat
3.7
49
c
c
rat
1.2
69
c
c
rat
1.4
0.1
3
c
c
rat
0.64
3.0
2.5
35
c
c
rat
1.0
67
c
c
rat
0.94
0.75
1.6
3.4
76
c
c
rat
3.2
83
c
c
rat
1.4
55
c
c
mouse
rat
0.25
0.10
8.8
46
c
c
23.6
98
a
Compounds dosed in male Sprague−Dawley rats and male CD-1
b
c
mice. Compound dosed iv as a DMSO solution. Compound dosed
as a solution at 2.5 mg/kg po in 1% Pluronic F68/1% HPMC/15%
Captisol/83% H2O, pH 2.0, with MsOH. Compound dosed as a
Pharmacology. The pharmacokinetic data presented in
Table 2 for key compounds demonstrate that the PERK
inhibitors were generally well behaved in vivo with good
bioavailability, duration, and exposure. The good pharmacoki-
netic properties of the molecules facilitated in vivo pharmaco-
d
e
solution at 2.0 mg/kg po in 1% Tween 80/99% OraPlus. Compound
dosed as a solution at 2.5 mg/kg po in 1% Tween 80/2% HPMC/97%
H2O, pH 2.0, with MsOH.
G
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 5. (a) Cocrystal structure of 44 solved in complex with PERK at 2.4 Å resolution. (b) Overlay of apo mouse PERK and apo yeast GCN2
crystal structures demonstrating different conformations of the hinge Cys residue of the two proteins in the apo state (circled in red).
Figure 6. Cocrystal structure of 52 solved in complex with PERK at
2.7 Å resolution.
logical evaluation of the compounds. Compounds were
generally well tolerated upon repeat dosing.
A pharmacodynamic (PD) model for measuring in vivo
PERK inhibition was established in athymic nude mice
Figure 7. Inhibition of PERK autophosphorylation 4 h postdose in
doxycycline-inducible PERK HT1080 PD assay with inhibitors dosed
implanted with a doxycycline-inducible HT1080−T-REx−
PERK−FLAG Matrigel plug. The mice were dosed with
doxycycline in their drinking water overnight prior to
implantation, causing overexpression of PERK in the HT1080
plug. The mice were dosed orally with inhibitors, and the tumor
plug was removed at predetermined time points postdose, and
the inhibition of PERK autophosphorylation was measured
(comparing pPERK/total PERK ratios). Dose−response PD
data at the 4 h time point are shown in Figure 7 for the
selective PERK inhibitors 44 and 52. Compound 44 robustly
inhibited PERK autophosphorylation in this assay (ED50 ≈ 3
mg/kg; ED90 ≈ 60 mg/kg at the 4 h time point), and >50%
target coverage was maintained for 24 h in a time course PD
assay when dosed at 100 mg/kg po. Compound 52 was more
potent than 44 in the PD assay (ED50 ≈ 2 mg/kg; ED90 ≈ 30
mg/kg at the 4 h time point), with >50% target coverage being
maintained for 24 h in the time course PD assay when dosed at
3 mg/kg po (reflecting the low clearance of 52).
orally: (a) 44 at 3, 10, 30, 100 mg/kg; (b) 52 at 1, 3, 10, 30 mg/kg. n =
3 mice per group. Controls are Matrigel plugs (vehicle dosing) in the
absence and presence of doxycycline.
1H-pyrazol-3(2H)-one PERK inhibitors. Structural information
was used to optimize potency and selectivity resulting in the
identification of 44 as a potent, exquisitely selective PERK
inhibitor. Following from the information learned in deriving
44, the PERK inhibitor 52 was designed to improve upon the
selectivity of the recently disclosed PERK inhibitor 1.
Compounds 44 and 52 are structurally dissimilar selective
PERK inhibitors, binding to different conformations of the
PERK kinase domain. Both compounds demonstrated excellent
pharmacokinetic properties and were able to give good PERK
target coverage in mouse PD studies. The use of these
compounds for exploration of PERK pathway biology and
PERK target validation, both in vitro and in vivo, will be
described elsewhere.
CONCLUSION
In summary, we have described the design, synthesis,
biochemical, and pharmacological properties of a series of
EXPERIMENTAL SECTION
Chemistry. Reactions were run under inert atmosphere (nitrogen
or argon) and were stirred using a PTFE-coated magnetic stir bar
■
■
H
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
unless noted otherwise. Reactions run at elevated temperature were
performed using a magnetic hot plate stirrer at the temperature
indicated utilizing an oil bath, aluminum heating block, or aluminum
beads for heat transfer. Microwave reactions were run either in a
Biotage microwave reactor set to normal power at the indicated
temperature or in a CEM Discover microwave reactor equipped with
the PowerMAX feature and were performed in sealed microwave
reaction vessels. Solvents and reagents obtained from commercial
sources were used without further purification. Solutions were
concentrated using a rotary evaporator, and solids were collected by
reaction mixture was allowed to cool to rt and was then diluted with 2
N aqueous HCl (50 mL) and stirred for 30 min. The mixture was
diluted with water (100 mL) and extracted with EtOAc (3 × 100 mL).
The combined EtOAc extracts were washed with saturated aqueous
NaHCO3 (200 mL), re-extracting the aqueous layer with EtOAc (100
mL). The combined EtOAc extracts were concentrated, and the
resulting solid was suspended in DCM (5 mL) and sonicated. The
solid was collected by filtration, washing with DCM (1 mL), to give 5
(1.20 g, 78% yield) as a pale-pink solid. 1H NMR (400 MHz, DMSO-
d6): δ 10.22 (s, 1H), 8.04 (d, J = 2.5 Hz, 1H), 7.93 (dd, J = 2.5, 9.1 Hz,
1H), 7.62 (d, J = 7.8 Hz, 1H), 7.27 (d, J = 9.0 Hz, 1H), 6.70 (d, J = 8.2
Hz, 1H), 6.46 (s, 2H), 3.88 (s, 3H), 2.40 (s, 3H). HPLC−MS (ESI)
m/z calculated for C16H14F3IN2O2 + H+ [M + H+]: 451.0. Found:
450.8.
filtration using either a Buchner funnel or sintered funnel. Residual
̈
solvent was removed from products using a vacuum manifold
maintained at approximately 1 Torr. All yields reported are isolated
yields after removal of residual solvents.
Silica gel chromatography refers either to column chromatography
as described by Still29 using EMD silica gel 60 (230−400 mesh) as the
stationary phase or to the use of an automated medium pressure liquid
chromatography system (Teledyne ISCO or Shoko Scientific Purif-
Espoir2) using prepacked silica gel cartridges (RediSep, Interchim
PuriFlash, or SiliCycle). The eluent solvent systems and gradients are
as noted.
2-Amino-N-(4-methoxy-3-(trifluoromethyl)phenyl)-4-meth-
yl-3-(2-(methylamino)quinazolin-6-yl)benzamide (7). A micro-
wave reactor vessel was charged with 5 (0.196 g, 0.435 mmol), N-
methyl-6-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)quinazolin-2-
amine (6)14 (0.161 g, 0.566 mmol), Na2CO3·H2O (0.162 g, 1.31
mmol), PdCl2(PPh3)2 (0.0153 g, 0.0218 mmol), and 10:1 DMF/H2O
(1.8 mL). The vessel was sealed and heated at 150 °C for 15 min in a
microwave reactor. The reaction mixture was diluted with water,
neutralized with TFA, and extracted with EtOAc. The organic extracts
were washed with water and brine and then dried over Na2SO4. The
solution was concentrated in vacuo, and the resulting residue was
purified by silica gel chromatography, eluting with a gradient of 20−
80% EtOAc/hexanes. The product fractions were combined,
concentrated, and slurried in 10:1 Et2O/hexanes. The solid was
collected by filtration to give 7 (0.142 g, 68% yield) as an off-white
solid. 1H NMR (400 MHz, DMSO-d6): δ 10.15 (s, 1H), 9.13 (s, 1H),
8.08 (d, J = 2.5 Hz, 1H), 7.96 (dd, J = 2.5, 9.0 Hz, 1H), 7.58−7.71 (m,
3H), 7.48 (dd, J = 1.9, 8.5 Hz, 1H), 7.37 (d, J = 4.5 Hz, 1H), 7.28 (d, J
= 9.0 Hz, 1H), 6.67 (d, J = 8.2 Hz, 1H), 5.77 (s, 2H), 3.89 (s, 3H),
2.94 (d, J = 4.9 Hz, 3H), 1.98 (s, 3H). 19F NMR (377 MHz, DMSO-
d6): δ −60.80 (s, 3F). 13C NMR (100 MHz, DMSO-d6): δ 167.9,
161.9, 160.3, 152.9, 147.2, 140.3, 136.67 (q, J = 288 Hz), 136.2, 132.1,
130.3, 129.0, 127.8, 127.0, 126.0, 125.0, 120.1, 119.0 (q, J = 6 Hz),
116.9, 116.47 (q, J = 30 Hz), 113.2, 113.1, 56.3, 40.2, 28.1, 20.7.
HRMS (ESI) m/z calculated for C25H22F3N5O2 + H+ [M + H+]:
482.1804. Found: 482.1800.
N-(5-((6,7-Dimethoxyquinolin-4-yl)oxy)pyridin-2-yl)-1-meth-
yl-3-oxo-2,5-diphenyl-2,3-dihydro-1H-pyrazole-4-carboxa-
mide (10a). A mixture of 5-(6,7-dimethoxyquinolin-4-yloxy)pyridin-
2-amine (8)16 (0.175 g, 0.589 mmol) and 1-methyl-3-oxo-2,5-
diphenyl-2,3-dihydro-1H-pyrazole-4-carboxylic acid (9a)17 (0.208 g,
0.706 mmol) in DMF (5 mL) was treated with triethylamine (0.21
mL, 1.47 mmol) followed by HATU (0.336 g, 0.883 mmol). The
mixture was stirred at 90 °C for 16 h. The mixture was allowed to cool
to rt and was then diluted with water (60 mL) and stirred for 10 min.
The precipitate was collected by filtration. The crude material was
purified by silica gel chromatography, eluting with a gradient of 0−
10% MeOH/DCM to give 10a (0.105 g, 31% yield) as a tan solid. 1H
NMR (400 MHz, DMSO-d6): δ 11.41 (s, 1H), 8.50 (d, J = 5.1 Hz,
1H), 8.34 (d, J = 2.7 Hz, 1H), 8.24 (d, J = 9.0 Hz, 1H), 7.53−7.75 (m,
12H), 7.42 (s, 1H), 6.53 (d, J = 5.3 Hz, 1H), 3.96 (s, 3H), 3.95 (s,
3H), 3.18 (s, 3H). 13C NMR (100 MHz, DMSO-d6): δ 162.4, 159.7,
159.7, 155.0, 152.6, 149.4, 149.2, 148.8, 146.5, 146.4, 141.2, 132.8,
131.1, 130.4, 130.2, 129.5, 129.0, 128.2, 127.6, 127.1, 115.0, 114.3,
107.9, 103.1, 99.1, 97.9, 55.7, 55.7, 35.6. HRMS (ESI) m/z calculated
for C33H27N5O5 + H+ [M + H+]: 574.2090. Found: 574.2087.
N-(5-((6,7-Dimethoxyquinolin-4-yl)oxy)pyridin-2-yl)-1-meth-
yl-3-oxo-2-phenyl-5-(pyridin-4-yl)-2,3-dihydro-1H-pyrazole-4-
carboxamide (10b). 10b was prepared as a yellow solid (165 mg,
16% yield) in an analogous manner to 10a from 8 and 1-methyl-3-oxo-
2-phenyl-5-(pyridin-4-yl)-2,3-dihydro-1H-pyrazole-4-carboxylic acid
(9b).16 1H NMR (400 MHz, DMSO-d6): δ 11.31 (s, 1H), 8.78−
8.88 (m, 2H), 8.50 (d, J = 5.3 Hz, 1H), 8.34 (t, J = 8.3 Hz, 1H), 8.25
(s, 1H), 8.22−8.23 (m, 1H), 7.48−7.76 (m, 8H), 7.38−7.47 (m, 1H),
6.54 (t, J = 5.6 Hz, 1H), 3.96 (s, 3H), 3.95 (s, 3H), 3.21 (s, 3H). 13C
NMR (100 MHz, DMSO-d6): δ 161.9, 159.8, 159.5, 152.7, 151.5,
1H, 19F, and 13C NMR spectra were obtained on a Bruker BioSpin
GmbH magnetic resonance spectrometer. NMR spectra are reported
as chemical shifts in parts per million (ppm) relative to an internal
solvent reference. High resolution mass spectra were acquired on a
Synapt G2 HDMS instrument (Waters Corporation, Manchester,
U.K.) operated in positive electrospray ionization mode.
Analytical HPLC−MS was conducted using an Agilent 1100 series
HPLC−mass spectrometry instrument and a Zorbax SB-C18 column
(50 mm × 3.0 mm, 3.5 μm) at 40 °C with a gradient of 5−95%
MeCN/water (modified with 0.1% TFA) over 3 min and a 1.5 mL/
min flow rate. Flow from the UV detector was split (50:50) to the MS
detector, which was configured with an API-ES source. Purities for
final compounds were measured using UV detection at 254 and 215
nm and were ≥95%.
All synthetic starting materials and reagents either were from
commercial sources or were synthesized according to the referenced
literature procedures.
1-(5-(4-Amino-7-methyl-7H-pyrrolo[2,3-d]pyrimidin-5-yl)-
indolin-1-yl)-2-(3-fluoro-5-(trifluoromethyl)phenyl)ethanone
(1). The title compound was prepared as described by Axten.12 1H
NMR (400 MHz, DMSO-d6): δ 8.16 (s, 1H), 8.12 (d, J = 8.2 Hz, 1H),
7.56−7.62 (m, 2H), 7.52 (d, J = 9.8 Hz, 1H), 7.34 (s, 1H), 7.21−7.29
(m, 2H), 6.07 (br s, 2H), 4.27 (t, J = 8.4 Hz, 2H), 4.07 (s, 2H), 3.74
(s, 3H), 3.27 (t, J = 8.3 Hz, 2H). 19F NMR (377 MHz, DMSO-d6): δ
−61.09 (s, 3F), −111.75 (s, 1F). 13C NMR (100 MHz, DMSO-d6): δ
168.0, 162.9, 160.5, 157.1, 151.6, 150.4, 141.7, 140.1, 132.8, 129.9,
127.1, 124.9, 124.0, 123.0, 121.4, 121.1, 116.2, 114.9, 110.9, 99.9, 47.8,
40.8, 30.7, 27.5. HRMS (ESI) m/z calculated for C24H19F4N5O + H+
[M + H+]: 470.1604. Found: 470.1601.
2-Amino-3-iodo-N-(4-methoxy-3-(trifluoromethyl)phenyl)-
4-methylbenzamide (5). A solution of 3-iodo-4-methyl-2-nitro-
benzoic acid (2) (1.052 g, 3.43 mmol) and 4-methoxy-3-
(trifluoromethyl)aniline (3) (0.720 g, 3.77 mmol) in DMF (20 mL)
was treated with i-Pr2NEt (0.72 mL, 4.11 mmol) followed by TBTU
(1.32 g, 4.11 mmol), and the mixture was stirred at rt for 2 h. The
reaction mixture was added dropwise to stirred saturated aqueous
NaHCO3 (100 mL). A tan precipitate formed. The mixture was stirred
for 30 min, and the solid was collected by filtration, washing with
water. The solid was suspended in DMF (40 mL), heated to dissolve,
and added dropwise to stirred 0.5 M aqueous HCl (250 mL). The
resulting precipitate was collected by filtration, washed with water, and
dried to give 3-iodo-N-(4-methoxy-3-(trifluoromethyl)phenyl)-4-
methyl-2-nitrobenzamide (4) as a pale pink solid. HPLC−MS (ESI)
m/z calculated for C16H12F3IN2O4 + H+ [M + H+]: 481.0. Found:
480.8.
The crude 4 (1.65 g, 3.43 mmol) was suspended in EtOH (25 mL)
and AcOH (2.0 mL, 34 mmol). The mixture was heated at gentle
reflux until dissolved and then treated with iron powder (0.765 g, 13.7
mmol). The resulting mixture was heated at gentle reflux for 3 h. The
I
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
149.6, 149.6, 149.5, 149.1, 149.0, 148.7, 146.5, 146.2, 141.2, 135.7,
132.4, 131.2, 129.6, 129.4, 127.5, 127.5, 124.6, 124.5, 115.0, 114.4,
107.6, 103.1, 99.1, 98.3, 55.7, 55.7, 35.3. HPLC−MS (ESI) m/z
calculated for C32H26N6O5 + H+ [M + H+]: 575.2. Found: 575.2.
General Procedure for the Synthesis of Intermediates 13a−
d. A suspension of 2,5-diphenyl-1H-pyrazol-3(2H)-one (12)18 (3.00 g,
12.7 mmol) and calcium hydroxide (2.20 equiv, 2.07 g, 27.9 mmol) in
1,4-dioxane (50 mL) was heated at 50 °C for 20 min. Acid chloride
selected from 11a−d (1.20 equiv, 15.2 mmol) was added portionwise
over 5 min, and the mixture was then heated at 90 °C for 16 h. The
mixture was allowed to cool to rt and was then extracted into EtOAc
(3 × 100 mL) from saturated aqueous NaHCO3 (100 mL). The
combined organic extracts were dried over Na2SO4, filtered, and
concentrated in vacuo. The resulting solid was dried under vacuum
overnight and used without further purification.
6-Bromo-5-chloroquinazolin-2-amine (16). A suspension of 3-
bromo-2-chloro-6-fluorobenzaldehyde (15) (12.88 g, 54.2 mmol) and
guanidine carbonate (12.16 g, 67.5 mmol) in NMP (200 mL) and i-
Pr2NEt (23.5 mL, 135 mmol) was heated at 140 °C for 2 h. The
reaction mixture was allowed to cool to rt, poured into water (1 L),
and treated with brine (100 mL). The precipitate was collected by
filtration and washed with water to give 16 (13.32 g, 95% yield) as a
tan solid. 1H NMR (400 MHz, DMSO-d6): δ 9.30 (s, 1H), 7.93 (d, J =
9.0 Hz, 1H), 7.34 (d, J = 9.0 Hz, 1H), 7.26 (s, 2H). HPLC−MS (ESI)
m/z calculated for C8H5BrClN3 + H+ [M + H+]: 257.9/259.9. Found:
258.0/260.0.
6-Bromo-2,5-dichloroquinazoline (17b). A suspension of 16
(14.45 g, 55.9 mmol) in DCM (350 mL) and DMF (20 mL) was
treated successively with n-Bu4NCl (18.94 g, 68.1 mmol), (CH3)3SiCl
(28.0 mL, 221 mmol), and t-BuONO (20.0 mL, 168 mmol) (added
dropwise over 25 min). The reaction mixture was stirred at room
temperature for an additional 15 min and then heated at 50 °C for 1 h.
The mixture was allowed to cool to rt and extracted into DCM (3 ×
100 mL) from saturated aqueous NaHCO3 (125 mL). The combined
organic extracts were dried (Na2SO4) and concentrated to give 17b
(15.54 g, 100%) as a tan solid. 1H NMR (400 MHz, CDCl3): δ 9.68 (s,
1H), 8.13 (d, J = 9.2 Hz, 1H), 7.81 (d, J = 9.0 Hz, 1H). HPLC−MS
(ESI) m/z calculated for C8H3BrCl2N2 + H+ [M + H+]: 276.9/278.9.
Found: 276.7/278.7.
The solid was mixed with methyl p-toluenesulfonate (20 equiv, 38.4
mL, 254 mmol) and heated at 170 °C for 3 h. The mixture was
allowed to cool to rt, diluted with 1 N aqueous NaOH (250 mL), and
extracted into EtOAc (3 × 150 mL). The combined organic extracts
were dried over Na2SO4, filtered, and concentrated in vacuo. The
crude material was purified by silica gel chromatography eluting with a
gradient of 0−12% MeOH/DCM to give 13a−d.
4-(3-Iodo-4-methyl-2-nitrobenzoyl)-1-methyl-2,5-diphenyl-
1H-pyrazol-3(2H)-one (13a). Tan solid; yield 47% (3.24 g, 6.01
1
General Procedure for the Synthesis of Intermediates 18a,b.
A suspension of 17a (Aurum PharmaTech) or 17b (56 mmol) in
EtOH (125 mL) and methylamine (33% solution in EtOH, 37.5 mL,
300 mmol) was heated at 80 °C for 1 h in a sealed tube. The mixture
was allowed to cool to rt and diluted with water (300 mL). The solid
was collected by filtration and purified by silica gel chromatography,
eluting with a gradient of 0−2.5% MeOH/DCM to give 18a,b.
6-Bromo-5-fluoro-N-methylquinazolin-2-amine (18a). White
solid; yield 100% (2.06 g, 8.04 mmol). 1H NMR (400 MHz, CDCl3):
δ 9.22 (br s, 1H), 7.73 (dd, J = 9.0, 7.6 Hz, 1H), 7.32 (d, J = 9.2 Hz,
1H), 5.47 (br s, 1H), 3.14 (s, 3H). HPLC−MS (ESI) m/z calculated
for C9H7BrFN3 + H+ [M + H+]: 256.0/258.0. Found: 255.9/257.9.
6-Bromo-5-chloro-N-methylquinazolin-2-amine (18b). Off-
mmol). H NMR (400 MHz, CDCl3): δ 7.46−7.72 (m, 7H), 7.41 (s,
3H), 7.33−7.40 (m, 2H), 3.23 (s, 3H), 2.52 (s, 3H). HPLC−MS
(ESI) m/z calculated for C24H18IN3O4 + H+ [M + H+]: 540.0. Found:
539.7.
4-(3-Bromo-4-methylbenzoyl)-1-methyl-2,5-diphenyl-1H-
pyrazol-3(2H)-one (13b). Tan solid; yield 70% (0.80 g, 1.79 mmol).
1H NMR (400 MHz, CDCl3): δ 7.97−8.06 (m, 1H), 7.78−7.83 (m,
1H), 7.30−7.59 (m, 9H), 7.01−7.30 (m, 2H), 3.19 (s, 3H), 2.39 (s,
3H). HPLC−MS (ESI) m/z calculated for C24H19BrN2O2 + H+ [M +
H+]: 447.1/449.1. Found: 446.7/448.7.
4-(3-Bromo-4-nitrobenzoyl)-1-methyl-2,5-diphenyl-1H-pyr-
azol-3(2H)-one (13c). Off-white solid; yield 89% (12.38 g, 25.9
mmol). 1H NMR (400 MHz, CDCl3): δ 8.16 (s, 3H), 7.97 (d, J = 8.2
Hz, 1H), 7.81 (d, J = 8.4 Hz, 1H), 7.60−7.41 (m, 10H), 3.28 (s, 3H).
HPLC−MS (ESI) m/z calculated for C23H16BrN3O4 + H+ [M + H+]:
478.0/480.0. Found: 477.9/479.9.
4-(3-Bromo-4-chlorobenzoyl)-1-methyl-2,5-diphenyl-1H-
pyrazol-3(2H)-one (13d). Tan solid; yield 74% (0.52 g, 1.11 mmol).
1H NMR (400 MHz, CDCl3): δ 8.08−8.13 (m, 1H), 7.79−7.85 (m,
1
white solid; yield 57% (8.74 g, 32.1 mmol). H NMR (400 MHz,
CDCl3): δ 9.34 (s, 1H), 7.80 (d, J = 9.0 Hz, 1H), 7.41 (d, J = 8.8 Hz,
1H), 5.42 (br s, 1H), 3.13 (d, J = 5.1 Hz, 3H). HPLC−MS (ESI) m/z
calculated for C9H7BrClN3 + H+ [M + H+]: 272.0/274.0. Found:
271.9/274.0.
6-Bromo-2-methylquinazoline (18c). 18c was acquired from
commercial sources (Akos Building Blocks).
1H), 7.31−7.56 (m, 11H), 3.21 (s, 3H). HPLC−MS (ESI) m/z
calculated for C23H16BrClN2O2 + H+ [M + H+]: 467.0/469.0. Found:
467.1/469.1.
General Procedure for the Synthesis of Intermediates 19a−
c. A suspension of an aryl bromide selected from 18a−c (8.04 mmol),
bis(pinacolato)diboron (1.13 equiv, 2.31 g, 9.11 mmol), Pd(dppf)2Cl2·
CH2Cl2 (0.10 equiv, 646 mg, 0.79 mmol), and KOAc (3.9 equiv, 3.10
g, 31.6 mmol) in 1,4-dioxane (80 mL) was heated at 110 °C for 2.5 h
and then allowed to cool to rt. The reaction mixture was filtered
through a pad of diatomaceous earth, washing with DCM and MeOH.
The filtrate was concentrated and purified by silica gel chromatog-
raphy, eluting with a gradient of 20−100% EtOAc/hexanes to give
19a−c.
General Procedure for the Synthesis of Intermediates 14a
and 14c. A solution of 13a or 13c (7.42 mmol) in AcOH (8.5 mL,
148 mmol) and EtOH (64 mL) was treated with iron powder (9.8
equiv, 4.06 g, 72.8 mmol) and heated at 80 °C for 1.5 h. The mixture
was allowed to cool to rt and concentrated in vacuo. The residue was
diluted with DCM (40 mL) and 2 M NH3/MeOH (40 mL) and then
filtered through diatomaceous earth washing with 1:1 DCM and 2 M
NH3/MeOH. The filtrate was concentrated in vacuo and purified by
silica gel chromatography, eluting with a gradient of 0−6% (2 M NH3/
MeOH)/DCM to give 14a,c.
5-Fluoro-N-methyl-6-(4,4,5,5-tetramethyl-1,3,2-dioxaboro-
lan-2-yl)quinazolin-2-amine (19a). Off-white solid; yield 99%
1
(2.41 g, 7.95 mmol). H NMR (400 MHz, CDCl3): δ 9.25 (br s,
4-(2-Amino-3-iodo-4-methylbenzoyl)-1-methyl-2,5-diphen-
1H), 7.94 (t, J = 7.5 Hz, 1H), 7.36 (d, J = 8.2 Hz, 1H), 5.47 (br s, 1H),
3.14 (d, J = 5.1 Hz, 3H) 1.39 (s, 12H). HPLC−MS (ESI) m/z
calculated for C15H19BFN3O2 + H+ [M + H+]: 304.2. Found: 222.1
(consistent with hydrolysis of boronate ester).
5-Chloro-N-methyl-6-(4,4,5,5-tetramethyl-1,3,2-dioxaboro-
lan-2-yl)quinazolin-2-amine (19b). Yellow solid; yield 77% (1.33 g,
4.16 mmol). 1H NMR (400 MHz, CDCl3): δ 9.44 (br s, 1H), 7.93 (d,
J = 8.6 Hz, 1H), 7.48 (d, J = 8.6 Hz, 1H), 5.45 (br s, 1H), 3.15 (d, J =
5.1 Hz, 3H), 1.40 (s, 12H). HPLC−MS (ESI) m/z calculated for
C15H19BClN3O2 + H+ [M + H+]: 320.1. Found: 320.2.
yl-1H-pyrazol-3(2H)-one (14a). Tan solid; yield 76% (2.88 g, 5.65
1
mmol). H NMR (300 MHz, CDCl3): δ 7.77 (d, J = 8.2 Hz, 1H),
7.41−7.54 (m, 9H), 6.97 (s, 1H), 6.55−6.60 (m, 1H), 3.13 (s, 3H),
2.43 (s, 3H). HPLC−MS (ESI) m/z calculated for C24H20IN3O2 + H+
[M + H+]: 510.1. Found: 510.0.
4-(4-Amino-3-bromobenzoyl)-1-methyl-2,5-diphenyl-1H-
pyrazol-3(2H)-one (14c). Brown solid; yield 90% (10.42 g, 23.2
1
mmol). H NMR (400 MHz, DMSO-d6): δ 7.86 (d, J = 1.6 Hz, 1H),
7.61−7.54 (m, 5H), 7.52−7.46 (m, 5H), 7.41 (dt, J = 5.7, 3.0 Hz, 1H),
6.72 (d, J = 8.6 Hz, 1H), 6.23 (s, 2H), 3.10 (s, 3H). HPLC−MS (ESI)
m/z calculated for C23H18BrN3O2 + H+ [M + H+]: 448.1/450.1.
Found: 448.0/449.9.
2-Methyl-6-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-
quinazoline (19c). Off-white solid; yield 100% (461 mg, 1.71 mmol).
1H NMR (400 MHz, CDCl3): δ 9.35 (s, 1H), 8.41 (s, 1H), 8.25 (d, J =
J
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
8.6 Hz, 1H), 7.93 (d, J = 8.6 Hz, 1H), 2.92 (s, 3H), 1.40 (s, 12H).
HPLC−MS (ESI) m/z calculated for C15H19BN2O2 + H+ [M + H+]:
271.2. Found: 206.9 (consistent with hydrolysis of boronate ester).
6-Bromo-2-(fluoromethyl)benzo[d]thiazole (28a). Et3N (4.4
mL, 31.6 mmol) was added dropwise to a suspension of 2-amino-5-
bromobenzenethiol (25) (1.27 g, 6.24 mmol), PPh3 (6.10 g, 23.3
mmol), and fluoroacetic acid (0.46 mL, 8.07 mmol) in CCl4 (24 mL).
The reaction mixture was heated at 95 °C for 1.5 h and was then
allowed to cool to rt. The mixture was filtered through diatomaceous
earth, washing with DCM. The filtrate was suspended in Et2O, filtered,
and the filtrate was concentrated and purified by silica gel
chromatography, eluting with a gradient of 50−60% DCM/hexanes
(ESI) m/z calculated for C14H15BF3NO2S + H+ [M + H+]: 330.1.
Found: 329.9.
General Procedure for the Synthesis of PERK Inhibitors 30−
33, 37−40, and 42−44. Method A. A mixture of a 4-benzoyl-1H-
pyrazol-3(2H)-one selected from 13b, 13d, 14a, and 14c (27.5 mmol),
a boronate ester selected from 6, 20−24, 29a, and 29c (2.0 equiv, 55.0
mmol), and K3PO4 (3.0 equiv, 17.5 g, 82 mmol) in toluene (65 mL),
MeOH (35 mL), and water (20 mL) was deoxygenated by bubbling
argon through the mixture for 5 min. Pd(PPh3)4 (0.15 equiv, 4.76 g,
4.12 mmol) was added, and the reaction vessel was then sealed under
an argon atmosphere and heated at 100 °C for 8 h. The mixture was
allowed to cool to rt and was then diluted with 1:1 MeOH/DCM (200
mL) and filtered through a pad of diatomaceous earth. The filtrate was
concentrated and purified by silica gel chromatography, eluting with a
gradient of 0−6% 2 M NH3 in MeOH/DCM to give the title
compound.
Method B. A suspension of a 4-benzoyl-1H-pyrazol-3(2H)-one
selected from 13b, 13d, 14a, and 14c (0.384 mmol), a boronate ester
selected from 6, 20−24, 29a, and 29c (1.3 equiv, 0.507 mmol), bis(di-
tert-butyl(4-dimethylaminophenyl)phosphine)dichloropalladium(II)
(0.10 equiv, 27 mg, 0.038 mmol), and Na2CO3·H2O (3.0 equiv,0.122
g, 1.15 mmol) in DMF (4 mL) and water (0.5 mL) was heated at 110
°C for 30 min. The reaction mixture was allowed to cool to rt and was
then poured into water (30 mL). The resulting precipitate was
collected by filtration, washed with water, and purified by silica gel
chromatography, eluting with a gradient of 0−6% 2 M NH3 in
MeOH/DCM to give the title compound.
1
to give 28a (318 mg, 21% yield) as a rust-colored powder. H NMR
(400 MHz, CDCl3): δ 8.09 (d, J = 1.6 Hz, 1H), 7.90 (d, J = 8.8 Hz,
1H), 7.63 (dd, J = 8.6, 1.8 Hz, 1H), 5.74 (d, J = 48 Hz, 2H). HPLC−
MS (ESI) m/z calculated for C8H5BrFNS + H+ [M + H+]: 245.9/
247.9. Found: 245.9/248.0.
6-Bromo-2-(difluoromethyl)benzo[d]thiazole (28b). Difluor-
oacetyl chloride (4.51 g, 39.4 mmol) was added dropwise to a solution
of 2,2′-disulfanediylbis(4-bromoaniline) (26) (2.00 g, 4.92 mmol) and
pyridine (1.6 mL, 20 mmol) in DCM (20 mL). The mixture was
stirred at rt for 2 h and was then concentrated and extracted into 3:1
EtOAc/MeOH (3 × 20 mL) from water (25 mL). The combined
organic extracts were dried (Na2SO4), filtered, and concentrated to
give crude N,N′-(disulfanediylbis(4-bromo-2,1-phenylene))bis(2,2-di-
fluoroacetamide) (27) (2.77 g, 4.93 mmol), which was used without
further purification. HPLC−MS (ESI) m/z calculated for
C16H10Br2F4N2O2S2 + Na+ [M + Na+]: 582.8/584.8/586.8. Found:
582.5/584.5/586.5.
A mixture of the crude 27 (2.77 g, 4.93 mmol) and dithiothreitol
(3.80 g, 24.6 mmol) in DMF (8 mL) was stirred under inert
atmosphere at rt for 16 h. The reaction mixture was diluted with water
(120 mL) and the precipitate was collected by filtration, washing with
water, to give 28b (2.29 g, 88% yield) as a pale-yellow solid. 1H NMR
(400 MHz, CDCl3): δ 8.13 (dd, J = 1.9, 0.4 Hz, 1H), 7.98 (d, J = 8.7
Hz, 1H), 7.68 (dd, J = 8.7, 2.0 Hz, 1H), 6.92 (t, J = 54 Hz, 1H). 19F
NMR (377 MHz, CDCl3): δ −110.59 (d, J = 55 Hz, 2F). 13C NMR
(100 MHz, CDCl3): δ 162.38 (t, J = 31 Hz), 151.27, 136.50 (t, J = 1
Hz), 130.61, 125.53, 124.75, 120.80 (t, J = 1 Hz), 110.75 (t, J = 240
Hz). HPLC−MS (ESI) m/z calculated for C8H4BrF2NS + H+ [M +
H+]: 263.9/265.9. Found: 263.9/265.9.
For those products isolated as HCl salts, the purified free base was
dissolved in a 2:1 mixture of MeOH/DCM, and 1 N aqueous HCl
(1.05 equiv) was added. The mixture was stirred for 45 min, filtered,
concentrated, and dried. The residue was diluted with hexanes and
sonicated to break up lumps. The solid was collected by filtration,
washing with hexanes, to give the corresponding HCl salt.
General Procedure for the Synthesis of PERK Inhibitors 34−
36 and 41. The 2-nitrobenzoyl-1H-pyrazol-3(2H)-one 13a (3.04 g,
5.64 mmol), a boronate ester selected from 19a−c and 29b (2.5 equiv,
14.3 mmol), bis(di-tert-butyl(4-dimethylaminophenyl)phosphine)-
dichloropalladium(II) (0.10 equiv, 424 mg, 0.599 mmol), and
Na2CO3·H2O (4.8 equiv, 2.86 g, 27.0 mmol) in DMF (40 mL) and
water (4 mL) was heated at 110 °C for 30 min. The reaction mixture
was allowed to cool to rt and was then poured into water (300 mL).
The resulting precipitate was collected by filtration, washing with
water. The solid was purified by silica gel chromatography, eluting with
a gradient of 0−5% MeOH/DCM. The resulting nitro product was
suspended in EtOH (60 mL) and AcOH (6 mL). Iron powder (11
equiv, 3.47 g, 62.1 mmol) was added, and the mixture was heated at 80
°C for 2.5 h. The reaction mixture was allowed to cool to rt and was
filtered through a pad of diatomaceous earth, washing with MeOH,
DCM, and 10:1 MeOH/AcOH. The filtrate was concentrated and
treated with aqueous 5 N NaOH followed by saturated aqueous
NaHCO3 to raise the pH to 7. The suspension was filtered, and the
solid was washed with water. The solid was purified by silica gel
chromatography, eluting with a gradient of 0−6% 2 M NH3 in
MeOH/DCM to give 34−36 and 41.
6-Bromo-2-(trifluoromethyl)benzo[d]thiazole (28c). 28c was
acquired from commercial sources (Oxchem product list).
General Procedure for the Synthesis of Intermediates 29a−
c. A suspension of an aryl bromide selected from 28a−c (7.73 mmol),
bis(pinacolato)diboron (1.50 equiv, 2.95 g, 11.6 mmol), Pd(dppf)2Cl2·
CH2Cl2 (0.15 equiv, 0.95 g, 0.79 mmol), and KOAc (3.0 equiv, 2.28 g,
23.2 mmol) in 1,4-dioxane (25 mL) was heated in a microwave reactor
at 120 °C for 30 min and then allowed to cool to rt. The reaction
mixture was filtered through a pad of diatomaceous earth, washing
with DCM and MeOH. The filtrate was concentrated and purified by
silica gel chromatography, eluting with a gradient of 20−100% EtOAc/
hexanes to give 29a−c.
2-(Fluoromethyl)-6-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-
2-yl)benzo[d]thiazole (29a). Tan solid, yield 100% (417 mg, 1.42
4-(2-Amino-4-methyl-3-(2-(methylamino)quinazolin-6-yl)-
benzoyl)-1-methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one (30). 30
was prepared by method B from 6 and 14a as a rust-colored solid,
1
mmol). H NMR (400 MHz, CDCl3): δ 8.43 (s, 1H), 8.05−8.01 (m,
1
1H), 7.93 (d, J = 8.2 Hz, 1H), 5.77 (d, J = 44 Hz, 2H), 1.39 (s, 12H).
HPLC−MS (ESI) m/z calculated for C14H17BFNO2S + H+ [M + H+]:
294.1. Found: 294.0.
yield 58% (121 mg, 0.22 mmol). H NMR (400 MHz, DMSO-d6): δ
9.11 (s, 1H), 7.78 (d, J = 8.2 Hz, 1H), 7.53−7.67 (m, 11H), 7.35−7.50
(m, 3H), 6.59 (t, J = 9.3 Hz, 1H), 6.52 (br s, 2H), 3.12 (s, 3H), 2.94
(d, J = 4.7 Hz, 3H), 1.94−2.01 (m, 3H). 13C NMR (100 MHz,
DMSO-d6): δ 190.5, 161.9, 161.4, 160.3, 157.8, 151.1, 149.0, 142.9,
136.2, 134.6, 133.8, 130.5, 129.7, 129.2, 129.2, 129.2, 128.9, 128.1,
127.0, 126.5, 125.9, 124.6, 120.1, 116.8, 116.4, 111.2, 37.8, 28.1, 21.0.
HRMS (ESI) m/z calculated for C33H28N6O2 + H+ [M + H+]:
541.2352. Found: 541.2350.
2-(Difluoromethyl)-6-(4,4,5,5-tetramethyl-1,3,2-dioxaboro-
lan-2-yl)benzo[d]thiazole (29b). Pale-yellow solid, yield 90% (2.16
g, 6.95 mmol). 1H NMR (400 MHz, CDCl3): δ 8.47 (s, 1H), 8.12 (d, J
= 8.2 Hz, 1H), 7.98 (d, J = 8.2 Hz, 1H), 6.95 (t, J = 54 Hz, 1H) 1.36
(s, 12H). HPLC−MS (ESI) m/z calculated for C14H16BF2NO2S + H+
[M + H+]: 312.1. Found: 311.9.
6-(4,4,5,5-Tetramethyl-1,3,2-dioxaborolan-2-yl)-2-
1-Methyl-4-(4-methyl-3-(2-(methylamino)quinazolin-6-yl)-
benzoyl)-2,5-diphenyl-1H-pyrazol-3(2H)-one (31). 31 was pre-
pared by method B from 6 and 13b as an off-white solid, yield 43%
(trifluoromethyl)benzo[d]thiazole (29c). Tan solid, yield 89%
1
(0.461 g, 1.40 mmol). H NMR (400 MHz, CDCl3): δ 8.47 (s, 1H),
1
8.18 (d, J = 8.2 Hz, 1H), 8.00 (s, 1H), 1.38 (s, 12H). HPLC−MS
(100 mg, 0.19 mmol). H NMR (400 MHz, DMSO-d6): δ 9.15 (br s,
K
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
1H), 7.62−7.80 (m, 4H), 7.49−7.60 (m, 10H), 7.35−7.48 (m, 3H),
3.15 (s, 3H), 2.94 (d, J = 4.9 Hz, 3H), 2.32 (s, 3H). 13C NMR (100
MHz, DMSO-d6): δ 187.8, 162.1, 161.6, 160.3, 158.2, 150.8, 140.2,
139.9, 135.9, 135.2, 134.2, 133.8, 130.8, 130.5, 130.3, 129.6, 129.2,
128.7, 128.3, 128.0, 127.8, 127.6, 125.4, 124.7, 119.3, 107.7, 37.1, 28.0,
20.4. HRMS (ESI) m/z calculated for C33H27N5O2 + H+ [M + H+]:
526.2243. Found: 526.2244.
1H), 6.45 (br s, 2H), 3.12 (s, 3H), 2.84 (s, 3H), 1.95 (s, 3H). 13C
NMR (100 MHz, DMSO-d6): δ 190.4, 167.5, 161.4, 157.8, 153.9,
148.9, 142.8, 134.5, 134.5, 133.9, 130.5, 129.2, 129.2, 129.2, 128.9,
128.1, 127.1, 126.8, 126.6, 124.6, 123.2, 123.0, 116.8, 116.4, 111.2,
37.8, 21.0, 19.8. HRMS (ESI) m/z calculated for C32H26N4O2S + H+
[M + H+]: 531.1855. Found: 531.1856.
4-(2-Amino-4-methyl-3-(2-methylbenzo[d]thiazol-6-yl)-
benzoyl)-1-methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one (38). 38
was prepared by method A from 14a and 22 as a tan solid, yield 61%
4-(4-Amino-3-(2-(methylamino)quinazolin-6-yl)benzoyl)-1-
methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one (32). 32 was pre-
pared by method B from 6 and 14c as a yellow solid, yield 37% (83
1
(95 mg, 0.18 mmol). H NMR (400 MHz, DMSO-d6): δ 8.04 (d, J =
1
mg, 0.16 mmol). H NMR (400 MHz, DMSO-d6): δ 9.14 (br s, 1H),
8.2 Hz, 1H), 7.90 (d, J = 1.4 Hz, 1H), 7.78 (d, J = 8.2 Hz, 1H), 7.51−
7.65 (m, 9H), 7.38−7.48 (m, 1H), 7.28 (dd, J = 1.6, 8.2 Hz, 1H), 6.58
(d, J = 8.2 Hz, 1H), 6.46 (br s, 2H), 3.12 (s, 3H), 2.84 (s, 3H), 1.94 (s,
3H). 13C NMR (100 MHz, DMSO-d6): δ 190.4, 167.3, 161.4, 157.8,
152.4, 148.9, 142.8, 136.2, 134.6, 133.9, 133.1, 130.5, 129.2, 129.2,
128.9, 128.1, 128.0, 127.0, 126.8, 124.6, 123.3, 122.7, 116.8, 116.4,
111.2, 37.8, 21.0, 19.8. HRMS (ESI) m/z calculated for C32H26N4O2S
+ H+ [M + H+]: 531.1855. Found: 531.1852.
7.73−7.83 (m, 1H), 7.50−7.68 (m, 13H), 7.33−7.48 (m, 2H), 6.77 (d,
J = 8.4 Hz, 1H), 5.85 (br s, 2H), 3.09 (s, 3H), 2.94 (d, J = 4.7 Hz, 3H).
13C NMR (100 MHz, DMSO-d6): δ 186.1, 162.2, 161.7, 160.1, 158.2,
150.8, 150.5, 135.1, 134.6, 133.1, 132.0, 130.9, 130.5, 129.4, 129.1,
128.8, 128.3, 127.5, 127.0, 126.0, 125.3, 124.7, 123.3, 119.6, 113.9,
109.8, 37.7, 28.1. HRMS (ESI) m/z calculated for C32H26N6O2 + H+
[M + H+]: 527.2196. Found: 527.2196.
4-(4-Chloro-3-(2-(methylamino)quinazolin-6-yl)benzoyl)-1-
methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one (33). 33 was pre-
pared by method A from 6 and 13d as a pale-yellow solid, yield 46%
4-(2-Amino-4-methyl-3-(2-(methylamino)benzo[d]thiazol-6-
yl)benzoyl)-1-methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one (39).
39 was prepared by method A from 14a and 23 as a pale-yellow solid,
1
1
(135 mg, 0.25 mmol). H NMR (400 MHz, DMSO-d6): δ 9.17 (br s,
yield 56% (1.12 g, 2.05 mmol). H NMR (400 MHz, DMSO-d6): δ
1H), 7.85 (s, 2H), 7.73−7.82 (m, 2H), 7.44−7.66 (m, 13H), 3.17 (s,
3H), 2.95 (d, J = 4.7 Hz, 3H). 13C NMR (100 MHz, DMSO-d6): δ
186.8, 162.2, 161.5, 160.4, 157.9, 151.2, 138.4, 137.3, 135.2, 135.1,
134.0, 132.3, 131.4, 130.5, 129.7, 129.7, 129.6, 129.2, 128.7, 128.4,
127.9, 127.8, 125.9, 124.6, 119.1, 106.6, 36.8, 28.0. HRMS (ESI) m/z
calculated for C32H24ClN5O2 + H+ [M + H+]: 546.1697. Found:
546.1696.
10.13 (br s, 1H), 7.77 (d, J = 8.2 Hz, 1H), 7.71 (s, 1H), 7.67−7.71 (m,
1H), 7.53−7.64 (m, 9H), 7.43 (t, J = 6.8 Hz, 1H), 7.22 (dd, J = 1.6, 8.2
Hz, 1H), 6.58 (d, J = 8.4 Hz, 1H), 4.43 (br s, 2H), 3.15 (s, 3H), 3.12
(s, 3H), 1.94 (s, 3H). 13C NMR (100 MHz, DMSO-d6): δ 190.4,
167.7, 161.3, 157.8, 148.8, 142.8, 134.5, 133.9, 131.6, 130.5, 129.2,
129.2, 129.2, 129.2, 128.9, 128.8, 128.1, 127.1, 126.6, 124.6, 123.9,
116.8, 116.5, 116.1, 111.1, 37.8, 31.7, 20.9. HRMS (ESI) m/z
calculated for C32H27N5O2S + H+ [M + H+]: 546.1964. Found:
546.1964.
4-(2-Amino-3-(2-aminobenzo[d]thiazol-6-yl)-4-methylben-
zoyl)-1-methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one Hydro-
chloride (40). 40 was prepared by method A from 14a and 24 as a
yellow solid, yield 28% (90 mg, 0.17 mmol). Note: Deacetylation of
the N-acetyl-2-aminobenzothiazole occurred under the reaction
conditions. 1H NMR (400 MHz, DMSO-d6): δ 9.95 (br s, 2H),
7.76 (s, 1H), 7.77 (d, J = 10.0 Hz, 2H), 7.52−7.69 (m, 9H), 7.40−7.47
(m, 1H), 7.26 (dd, J = 1.6, 8.4 Hz, 1H), 6.58 (d, J = 8.4 Hz, 1H), 4.72
(br s, 3H), 3.12 (s, 3H), 1.93 (s, 3H). 13C NMR (100 MHz, DMSO-
d6): δ 190.4, 169.0, 161.3, 157.8, 148.7, 142.8, 138.8, 134.5, 134.0,
132.4, 130.5, 129.3, 129.2, 129.2, 128.9, 128.1, 127.1, 126.4, 125.3,
124.6, 124.5, 116.8, 116.5, 115.2, 111.0, 37.7, 20.9. HRMS (ESI) m/z
calculated for C31H25N5O2S + H+ [M + H+]: 532.1807. Found:
532.1805.
4-(2-Amino-3-(2-(fluoromethyl)benzo[d]thiazol-6-yl)-4-
methylbenzoyl)-1-methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one
(41). 41 was prepared in two steps from 13a and 29a as a yellow solid,
overall yield 14% (42 mg, 0.077 mmol). 1H NMR (400 MHz, DMSO-
d6): δ 8.19 (d, J = 8.4 Hz, 1H), 8.06 (s, 1H), 7.80 (d, J = 8.2 Hz, 1H),
7.52−7.65 (m, 9H), 7.43 (t, J = 6.9 Hz, 1H), 7.38 (d, J = 8.6 Hz, 1H),
6.60 (d, J = 8.4 Hz, 1H), 6.49 (br s, 2H), 5.92 (d, J = 45 Hz, 2H), 3.12
(s, 3H), 1.95 (s, 3H). 19F NMR (377 MHz, DMSO-d6): δ −73.40 (s,
1F). 13C NMR (100 MHz, DMSO-d6): δ 190.4, 166.1 (d, J = 23 Hz),
161.4, 157.8, 151.9, 148.9, 142.8, 135.7, 134.5, 134.3, 134.0, 130.5,
129.2, 129.2, 128.9, 128.7, 128.1, 127.1, 126.6, 124.6, 124.0, 123.9,
116.8, 116.5, 111.2, 81.2 (d, J = 165 Hz), 37.8, 21.0. HRMS (ESI) m/z
calculated for C32H25FN4O2S + H+ [M + H+]: 549.1761. Found:
549.1760.
4-(2-Amino-3-(5-fluoro-2-(methylamino)quinazolin-6-yl)-4-
methylbenzoyl)-1-methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one
(34). 34 was prepared in two steps from 13a and 19a as a yellow solid,
1
overall yield 35% (1.84 g, 3.29 mmol). H NMR (400 MHz, DMSO-
d6): δ 9.28 (br s, 1H), 7.82 (d, J = 8.2 Hz, 1H), 7.41−7.65 (m, 13H),
6.68 (br s, 2H), 6.60 (d, J = 8.4 Hz, 1H), 3.13 (s, 3H), 2.94 (d, J = 4.9
Hz, 3H), 1.98 (s, 3H). 19F NMR (377 MHz, DMSO-d6): δ −124.58
(s, 1F). 13C NMR (100 MHz, DMSO-d6): δ 190.4, 161.3, 160.4, 157.9,
157.1, 155.7, 154.6, 152.9, 149.4, 143.8, 137.1, 134.5, 130.5, 129.2,
129.2, 128.9, 128.1, 127.1, 124.6, 122.1, 120.0, 116.6, 116.4, 114.1,
111.0, 110.2, 110.1, 37.7, 28.0, 20.6. HRMS (ESI) m/z calculated for
C33H27FN6O2 + H+ [M + H+]: 559.2258. Found: 559.2257.
4-(2-Amino-3-(5-chloro-2-(methylamino)quinazolin-6-yl)-4-
methylbenzoyl)-1-methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one
(35). 35 was prepared in two steps from 13a and 19b as a yellow solid,
1
overall yield 9% (54 mg, 0.094 mmol). H NMR (400 MHz, DMSO-
d6): δ 9.35 (br s, 1H), 7.70−7.89 (m, 1H), 7.52−7.68 (m, 10H), 7.46
(d, J = 8.4 Hz, 3H), 6.59 (br s, 3H), 3.12 (br s, 3H), 2.96 (br s, 3H),
1.91 (br s, 3H). 13C NMR (100 MHz, DMSO-d6): δ 190.4, 161.4,
160.4, 158.5, 157.8, 148.9, 143.2, 136.8, 134.5, 134.4, 131.7, 130.5,
129.2, 129.2, 129.2, 128.9, 128.5, 128.2, 127.1, 125.6, 124.6, 124.1,
117.5, 116.5, 116.4, 111.0, 37.7, 28.0, 20.3. HRMS (ESI) m/z
calculated for C33H27ClN6O2 + H+ [M + H+]: 575.1962. Found:
575.1959.
4-(2-Amino-4-methyl-3-(2-methylquinazolin-6-yl)benzoyl)-
1-methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one (36). 36 was pre-
pared in two steps from 13a and 19c as a yellow solid, overall yield
23% (57 mg, 0.11 mmol). 1H NMR (400 MHz, DMSO-d6): δ 9.53 (s,
1H), 7.95−8.08 (m, 2H), 7.82 (d, J = 8.2 Hz, 1H), 7.78 (d, J = 8.4 Hz,
1H), 7.52−7.65 (m, 10H), 6.49−6.69 (m, 3H), 3.13 (s, 3H), 2.82 (s,
3H), 1.96 (s, 3H). 13C NMR (100 MHz, DMSO-d6): δ 190.5, 164.0,
161.4, 160.7, 157.9, 149.1, 148.9, 142.7, 136.8, 135.5, 134.5, 134.2,
130.5, 129.2, 129.2, 129.0, 128.9, 128.2, 128.1, 127.1, 126.0, 124.6,
123.3, 116.8, 116.5, 111.1, 37.8, 26.7, 20.9. HRMS (ESI) m/z
calculated for C33H27N5O2 + H+ [M + H+]: 526.2243. Found:
526.2242.
4-(2-Amino-3-(2-(difluoromethyl)benzo[d]thiazol-6-yl)-4-
methylbenzoyl)-1-methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one
(42). 42 was prepared by method A from 14a and 29b as a tan solid,
1
yield 57% (1.55 g, 2.73 mmol). H NMR (400 MHz, DMSO-d6): δ
8.34 (d, J = 8.4 Hz, 1H), 8.20 (d, J = 1.2 Hz, 1H), 7.85 (d, J = 8.2 Hz,
1H), 7.57−7.73 (m, 10H), 7.44−7.52 (m, 2H), 6.65 (d, J = 8.4 Hz,
1H), 3.87−5.89 (br s, 2H), 3.17 (s, 3H), 1.99 (s, 3H). 19F NMR (377
MHz, DMSO-d6): δ −110.82 (d, J = 41 Hz, 2F). 13C NMR (100 MHz,
DMSO-d6): δ 190.4, 162.1 (t, J = 30 Hz), 161.3, 157.8, 151.4, 148.8,
142.7, 135.6, 135.2, 134.5, 134.1, 130.5, 129.5, 129.2, 129.2, 129.2,
129.2, 128.9, 128.1, 127.1, 126.4, 124.8, 124.6, 116.8, 116.6, 109.9 (t, J
4-(2-Amino-4-methyl-3-(2-methylbenzo[d]thiazol-5-yl)-
benzoyl)-1-methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one (37). 37
was prepared by method A from 14a and 21 as a pale-yellow solid,
1
yield 29% (45 mg, 0.085 mmol). H NMR (400 MHz, DMSO-d6): δ
8.17 (d, J = 8.2 Hz, 1H), 7.70−7.82 (m, 2H), 7.53−7.65 (m, 9H), 7.43
(t, J = 6.8 Hz, 1H), 7.21 (dd, J = 1.6, 8.2 Hz, 1H), 6.59 (d, J = 8.4 Hz,
L
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
= 239 Hz), 37.8, 20.9. HRMS (ESI) m/z calculated for
C32H24F2N4O2S + H+ [M + H+]: 567.1666. Found: 567.1667.
4-(2-Amino-4-methyl-3-(2-(trifluoromethyl)benzo[d]thiazol-
6-yl)benzoyl)-1-methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one
(43). 43 was prepared by method A from 14a and 29c as a pale-yellow
(140 mL) was cooled in an ice−water bath, and NaH (60% in mineral
oil; 1.53 g, 38.2 mmol) was added portionwise over 20 min. The
mixture was stirred for an additional 2 h, and then iodomethane (2.0
mL, 32 mmol) was slowly added. The ice−water bath was removed,
and the mixture was stirred for 16 h. The mixture was poured onto
ice−water (600 mL) and stirred for 20 min. The precipitate was
collected by filtration, washing with water, to give 50 (7.04 g, 78%
yield) as a tan solid. 1H NMR (400 MHz, DMSO-d6): δ 7.88 (s, 1H),
3.79 (s, 3H), 2.66 (s, 3H). HPLC−MS (ESI) m/z calculated for
C8H7ClIN3 + H+ [M + H+]: 307.9. Found: 307.9.
1
solid, yield 46% (95 mg, 0.16 mmol). H NMR (400 MHz, DMSO-
d6): δ 8.40 (d, J = 8.4 Hz, 1H), 8.24 (d, J = 1.2 Hz, 1H), 7.82 (d, J =
8.2 Hz, 1H), 7.52−7.64 (m, 10H), 7.43 (t, J = 6.8 Hz, 1H), 6.61 (d, J =
8.2 Hz, 1H), 6.54 (br s, 2H), 3.13 (s, 3H), 1.95 (s, 3H). 19F NMR
(377 MHz, DMSO-d6): δ −60.72 (s, 1F). 13C NMR (100 MHz,
DMSO-d6): δ 190.4, 161.3, 157.9, 155.1 (q, J = 40 Hz), 151.0, 148.8,
142.6, 136.7, 135.7, 134.5, 134.3, 130.5, 130.2, 129.2, 129.2, 128.9,
128.1, 127.1, 126.1, 125.4, 124.9, 124.6, 119.8 (q, J = 273 Hz), 116.8,
116.5, 111.1, 37.7, 20.9. HRMS (ESI) m/z calculated for
C32H23F3N4O2S + H+ [M + H+]: 585.1572. Found: 585.1576.
4-(2-Amino-4-methyl-3-(2-methylquinolin-6-yl)benzoyl)-1-
methyl-2,5-diphenyl-1H-pyrazol-3(2H)-one Hydrochloride
(44). 44 was prepared by method A from 14a and 20 (Aldrich) as a
pale-yellow solid, yield 54% (8.38 g, 14.9 mmol). 1H NMR (400 MHz,
DMSO-d6): δ 9.03 (d, J = 8.6 Hz, 1H), 8.60 (d, J = 8.8 Hz, 1H), 8.21
(d, J = 1.4 Hz, 1H), 7.99 (d, J = 8.8 Hz, 1H), 7.95 (d, J = 8.8 Hz, 1H),
7.85 (d, J = 8.2 Hz, 1H), 7.53−7.65 (m, 9H), 7.43 (t, J = 6.8 Hz, 1H),
6.64 (d, J = 8.4 Hz, 1H), 3.73 (br s, 3H), 3.14 (s, 3H), 3.05 (s, 3H),
1.95 (s, 3H). 13C NMR (100 MHz, DMSO-d6): δ 190.5, 161.3, 157.9,
157.7, 148.6, 144.9, 142.6, 137.4, 136.5, 134.5, 134.5, 130.6, 130.4,
129.2, 129.2, 129.2, 128.9, 128.1, 127.6, 127.1, 125.4, 124.7, 123.8,
121.5, 117.0, 116.7, 110.9, 37.7, 20.9, 20.4. HRMS (ESI) m/z
calculated for C34H28N4O2 + H+ [M + H+]: 525.2291. Found:
525.2290.
5-Iodo-2,7-dimethyl-7H-pyrrolo[2,3-d]pyrimidin-4-amine
(51). A mixture of 50 (1.76 g, 5.72 mmol), concentrated aqueous NH3
(20 mL, 515 mmol), and 1,4-dioxane (10 mL) was heated in a
microwave reactor at 135 °C for 2 h. The mixture was allowed to cool
to rt and the precipitate was collected by filtration, washing with water,
1
to give 51 (1.28 g, 78% yield) as a tan solid. H NMR (400 MHz,
DMSO-d6): δ 7.33 (s, 1H), 3.65 (s, 3H), 2.39 (s, 3H). HPLC−MS
(ESI) m/z calculated for C8H9IN4 + H+ [M + H+]: 289.0. Found:
288.8.
1-(5-(4-Amino-2,7-dimethyl-7H-pyrrolo[2,3-d]pyrimidin-5-
yl)indolin-1-yl)-2-(3-fluoro-5-(trifluoromethyl)phenyl)-
ethanone (52). A mixture of 48 (1.50 g, 3.34 mmol), 51 (0.874 g,
3.03 mmol), and K3PO4 (1.93 g, 9.10 mmol) in toluene (20 mL),
MeOH (10 mL), and water (5 mL) was treated with Pd(PPh3)4 (350
mg, 0.30 mmol) and sealed under argon. The mixture was heated in a
microwave reactor at 110 °C for 20 min. The reaction mixture was
allowed to cool to rt and was diluted with 50% MeOH/DCM and
filtered through a pad of diatomaceous earth. The filtrate was
concentrated and purified by silica gel chromatography, eluting with a
gradient of 0−100% EtOAc/DCM followed by 0−12% 2 M NH3 in
MeOH/DCM. The resulting product was slurried in EtOAc (50 mL)
for 1 h, and the solid was collected by filtration to give 52 (0.592 g,
1-(5-Bromoindolin-1-yl)-2-(3-fluoro-5-(trifluoromethyl)-
phenyl)ethanone (47).
A suspension of 2-(3-fluoro-5-
(trifluoromethyl)phenyl)acetic acid (45) (25.45 g, 115 mmol) in
THF (200 mL) was treated with 1,1′-carbonyldiimidazole (17.77 g,
110 mmol), and the mixture was stirred at rt for 10 min (until a clear
solution was obtained and the initial effervescence subsided). The
solution was then heated at 60 °C for 2 h, after which time
effervescence had ceased. The solution was allowed to cool to rt, and
5-bromoindoline (46) (19.73 g, 100 mmol) was added. The resulting
solution was stirred for 20 h. The solution was concentrated, dissolved
in EtOAc (400 mL), and washed with 2 N aqueous HCl (2 × 200 mL)
followed by saturated aqueous NaHCO3 (2 × 200 mL). The EtOAc
extract was dried (MgSO4), filtered, concentrated, and recrystallized by
dissolving in EtOAc (100 mL), diluting with hexanes (400 mL), and
allowing it to stand at −20 °C for 16 h. The resulting white solid was
collected by filtration, washing with hexanes, to give 47 (33.65 g, 84%
1
40% yield) as a white solid. H NMR (400 MHz, DMSO-d6): δ 8.11
(d, J = 8.2 Hz, 1H), 7.56−7.62 (m, 2H), 7.52 (d, J = 9.6 Hz, 1H), 7.32
(s, 1H), 7.23 (d, J = 8.2 Hz, 1H), 7.16 (s, 1H), 5.98 (br s, 2H), 4.27 (t,
J = 8.4 Hz, 2H), 4.07 (s, 2H), 3.70 (s, 3H), 3.26 (t, J = 8.2 Hz, 2H),
2.43 (s, 3H). 19F NMR (377 MHz, DMSO-d6) δ −61.06 (s, 3F),
−111.73 (s, 1F). 13C NMR (100 MHz, DMSO-d6): δ 168.0, 161.7 (d,
J = 246 Hz), 159.9, 157.0, 151.4, 141.6, 140.0, 132.8, 130.5 (dq, J = 8,
33 Hz), 130.2, 127.0, 124.8 (q, J = 274 Hz), 124.7, 123.3, 123.0, 121.2,
116.2, 114.8, 110.8, 97.8, 47.8, 40.8, 30.6, 27.5, 25.4. HRMS (ESI) m/z
calculated for C25H21F4N5O + H+ [M + H+]: 484.1761. Found:
484.1763.
Purification of PERK Protein for Structural Studies. Human
PERK (575-1094 Δ670-874) harboring the inactivating D937N
mutation was expressed in E. coli BL21(DE3) cells as a His6-GST-
fusion. Purification of the unphosphorylated protein was performed
using glutathione Sepharose, TEV protease cleavage of the His6-GST
tag, and SEC purification using a Superdex 75 column. Final
purification buffer was 25 mM Tris, pH 7.5, 300 mM NaCl, 10 mM
2-mercaptoethanol, and 10% glycerol. Aliquots of PERK protein were
buffer exchanged into 10 mM HEPES, pH 7.5, 300 mM NaCl, and 5
mM dithiothreitol, and the solution was then concentrated to ∼8−10
mg/mL. Protein was mixed with 3 mol equiv of inhibitors from
DMSO stock solutions and incubated on ice for 1 h prior to setting up
crystallization trials.
Cocrystallization of 10b with PERK. Crystals of PERK with
compound 10b were obtained at 4 °C in hanging drops with 100 mM
BTP, pH 6.5, 300 mM Na2SO4, 14% PEG-3350. Mother liquor
supplemented with ethylene glycol was used as cryoprotectant.
Diffraction data for 10b with PERK were collected on beamline
5.0.2 at the Advanced Light Source (ALS) and processed and scaled
with HKL 2000. The cocrystal structure was solved by molecular
replacement using the structure of GCN2 (PDB code 1ZY4) as the
template. Model building was carried out with Coot, and refinement
was done using Refmac5.
1
yield) as a white crystalline solid. H NMR (400 MHz, DMSO-d6): δ
7.95 (d, J = 8.0 Hz, 1H), 7.43−7.59 (m, 4H), 7.32 (d, J = 8.2 Hz, 1H),
4.23 (t, J = 7.5 Hz, 2H), 4.03 (br s, 2H), 3.15−3.27 (m, 2H). HPLC−
MS (ESI) m/z calculated for C17H12BrF4NO + H+ [M + H+]: 402.0/
404.0. Found: 402.0/404.0.
2-(3-Fluoro-5-(trifluoromethyl)phenyl)-1-(5-(4,4,5,5-tetra-
methyl-1,3,2-dioxaborolan-2-yl)indolin-1-yl)ethanone (48). A
mixture of 47 (33.49 g, 83 mmol), bis(pinacolato)diboron (44.4 g, 175
mmol), KOAc (24.52 g, 250 mmol), and Pd(dppf)Cl2·DCM (3.01 g,
3.69 mmol) in 1,4-dioxane (400 mL) was degassed by bubbling argon
through the mixture for 30 min, and the mixture was then heated at 90
°C for 16 h. The mixture was allowed to cool, concentrated, dissolved
in DCM (400 mL), and purified by silica gel chromatography, eluting
with DCM to give a white solid. The solid was suspended in Et2O (50
mL), stirred for 30 min, diluted with hexanes (300 mL), and
concentrated to remove the Et2O. The suspension was refrigerated in a
−20 °C freezer for 1 h. The resulting white solid was collected by
filtration, washing with hexanes, to give 48 (32.91 g, 88% yield) as a
1
fluffy white solid. H NMR (400 MHz, DMSO-d6): δ 7.88−8.13 (m,
1H), 7.39−7.64 (m, 5H), 4.22 (br s, 2H), 4.05 (br s, 2H), 3.19 (br s,
2H), 1.28 (s, 12H). HPLC−MS (ESI) m/z calculated for
C23H24BF4NO3 + H+ [M + H+]: 450.2. Found: 450.2.
Cocrystallization of 7 with PERK. Crystals of PERK with
compound 7 were obtained at 4 °C in hanging drops with 0.6 M Na/K
tartrate. Mother liquor supplemented with glycerol was used as
cryoprotectant. Diffraction data for 7 with PERK were collected on
4-Chloro-5-iodo-2,7-dimethyl-7H-pyrrolo[2,3-d]pyrimidine
(50). A solution of 4-chloro-5-iodo-2-methyl-7H-pyrrolo[2,3-d]-
pyrimidine (49; Aurum PharmaTech) (8.63 g, 29.4 mmol) in DMF
M
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
beamline 22-ID at the Advanced Photon Source (APS) and processed
and scaled with HKL 2000. The cocrystal structure was solved using
the structure of compound 10b bound to PERK as the initial template.
Model building was carried out with Coot, and refinement was done
using Refmac5.
h followed by the addition of 100 nM thapsigargin (Sigma) for 30 min
to activate the PERK pathway. Newly synthesized protein was then
quantified over a 30 min window using the Click-iT AHA AlexaFluor
488 protein synthesis HCS assay kit (Life Technologies).
PERK CHoP Assay. HT1080 cells were seeded at 2 × 104 cells/mL
in 96-well flat-bottomed plates (Corning) and incubated overnight.
Cells were dosed with PERK inhibitor dose responses for 1 h followed
by the addition of 100 nM thapsigargin for 2 h to induce CHoP. Cells
were lysed in bDNA lysis buffer (Affymetrix) containing 0.17 μg/mL
final concentration of proteinase K (Affymetrix). CHoP and GAPDH
mRNA levels were quantified with the Quantigene 2.0 bDNA assay
kit31 (Affymetrix). CHoP mRNA was normalized to GAPDH mRNA.
pGCN2 MSD Assay. U2OS cells were seeded at 4 × 104 cells/mL
in 96-well flat-bottomed plates (Corning) and incubated for 24 h.
Cells were dosed with PERK inhibitor dose responses for 1 h. Cells
were then washed with DMEM lacking arginine, leucine, and lysine
(Sigma) followed by a 2 h incubation with PERK inhibitor dose
responses in DMEM lacking arginine, leucine, and lysine to activate
GCN2. Cells were then lysed in mPER. High binding plates (MSD)
were then coated with 3 μL of lysate and incubated overnight at 4 °C.
Plates were blocked with 5% blocker A (MSD) for 2.5 h at rt with
shaking. Antibodies to either phospho-GCN2 (EpiTomics) or total-
GCN2 (Cell Signaling) were added for 1 h at rt with shaking. Sulfo-
Tag Goat anti-rabbit (MSD) was then added for 1 h at room
temperature with shaking. Read buffer (MSD) was then added, and
phospho and total GCN2 protein levels were quantified on a
SECTOR Imager 6000 plate reader (MSD).
Cocrystallization of 34 with PERK. Crystals of PERK with
compound 34 were obtained at 4 °C in hanging drops with 100 mM
HEPES, pH 7.0, 180 mM Na/K tartrate, 5% PEG-3350. Mother liquor
supplemented with glycerol was used as cryoprotectant. Diffraction
data for 34 with PERK were collected on beamline 5.0.1 at the
Advanced Light Source (ALS) and processed and scaled with HKL
2000. The cocrystal structure was solved using the structure of
compound 7 bound to PERK as the initial template. Model building
was carried out with Coot, and refinement was done using Refmac5.
Cocrystallization of 39 with PERK. Crystals of PERK with
compound 39 were obtained at 4 °C in hanging drops with 100 mM
MES, pH 6.5, 180 mM Na/K tartrate. Mother liquor supplemented
with glycerol was used as cryoprotectant. Diffraction data for 39 with
PERK were collected on beamline 5.0.1 at the Advanced Light Source
(ALS) and processed and scaled with HKL 2000. The cocrystal
structure was solved using the structure of compound 7 bound to
PERK as the initial template. Model building was carried out with
Coot, and refinement was done using Refmac5.
Cocrystallization of 44 with PERK. Crystals of PERK with
compound 44 were obtained at 4 °C in hanging drops with 100 mM
HEPES, pH 7.0, 400 mM Na/K tartrate. Mother liquor supplemented
with glycerol was used as cryoprotectant. Diffraction data for 44 with
PERK were collected on a Rigaku FR-E home source X-ray generator
equipped with an R-Axis4 detector and processed and scaled with
HKL 2000. The cocrystal structure was solved using the structure of
compound 7 bound to PERK as the initial template. Model building
was carried out with Coot, and refinement was done using Refmac5.
Cocrystallization of 52 with PERK. Crystals of PERK with
compound 52 were obtained at 4 °C in hanging drops with 100 mM
MES, pH 6.5, 150 mM NaCl, 12% PEG-3350. Mother liquor
supplemented with ethylene glycol was used as cryoprotectant.
Diffraction data for 52 with PERK were collected on beamline 5.0.2
at the Advanced Light Source (ALS) and processed and scaled with
HKL 2000. The cocrystal structure was solved using the structure of
compound 7 bound to PERK as the initial template. Model building
was carried out with Coot, and refinement was done using Refmac5.
PERK and GCN2 Enzyme Assays. Human PERK and GCN2
enzyme assays were performed using TR-FRET assays (Cisbio Inc.) in
an analogous manner to previously described LanthaScreen TR-FRET
assays.30 N-Terminal His-tagged PERK and C-terminal His-tagged
GCN2 proteins were expressed in E. coli. His-tagged PERK was used
at 1 nM concentration in the presence of 1 μM ATP (Km
concentration), and His-tagged GCN2 was used at 10 nM
concentration in the presence of 120 μM ATP (0.66 × Km
concentration). The assay substrate was eIF2α (AviTag C-terminal,
His N-terminal; eIF2α(3-315)) at 37.5 nM concentration.
GCN2 CHoP Assay. The same conditions were used as for the
PERK CHoP assay except cells were dosed with PERK inhibitor dose
responses for 1 h followed by a 2 h incubation with PERK inhibitor
dose responses in DMEM lacking arginine, leucine, and lysine to
activate GCN2. Thapsigargin was omitted.
In Vivo PD Assay. Four- to six-week old naive athymic nude mice
(Harlan Laboratories) were treated on day zero with 0.1 mg/mL
doxycycline (Dox) H2O supplied in drinking water ad libitum
overnight. On day 1 of study, HT1080−T-REx−PERK−FLAG cells
(cultured in DMEM + 10% FBS) were trypsinized and counted. All
cells were suspended in 100% growth factor reduced Matrigel. Animals
(n = 3/group) were first treated orally with a single dose of control or
compound(s), and then Matrigel plugs containing 5 × 106 HT1080−
T-REx−PERK−FLAG cells in 500 μL of Matrigel were injected
subcutaneously into the animals’ right flanks. Mice were euthanized
with CO2 asphyxiation, and plug samples were harvested 4 h after
implantation and snap frozen in liquid nitrogen for analysis of PERK
phosphorylation. Plasma samples were also taken for pharmacokinetic
analysis. Tumor cell lysates were prepared by homogenization of the
plugs in ice-cold radioimmunoprecipitation assay buffer [50 mmol/L
Tris (pH 7.5), 150 mmol/L NaCl, 1% NP40, 0.5% sodium
deoxycholate, and 0.1% SDS] containing protease and phosphatase
inhibitors (Sigma) and assayed by pPERK sandwich ELISA using
FLAG antibody (Sigma) for total PERK capture and internally
generated pPERK antibody for pPERK detection.
Cell Culture. HT1080−T-REx−PERK−FLAG cells (Amgen) were
grown in DMEM (Gibco) supplemented with 10% tetracycline-free
FBS (Hyclone), 1× NEAA (Gibco), 5 μg/mL blasticidin (Invitrogen),
and 400 μg/mL Zeocin (Invitrogen). HT1080 cells (ATCC) were
grown in DMEM supplemented with 10% FBS (Gibco) and 1×
NEAA. U2OS cells (ATCC) were grown in McCoy’s 5A (Gibco)
supplemented with 10% FBS, 1× PSG (Gibco), and 1× NEAA. All
cells were grown at 37 °C in the presence of 5% CO2.
pPERK ELISA Assay. HT1080−T-REx−PERK−FLAG cells were
seeded at 3 × 104 cells/mL in 96-well flat-bottomed plates (Corning)
and incubated overnight. PERK was overexpressed by addition of 25
ng/mL doxycycline (Sigma) for 4 h. Then PERK inhibitor dose
responses were added for 1 h. Cells were lysed in mPER (Pierce).
pPERK sandwich ELISA was performed using FLAG antibody
(Sigma) for total PERK capture and internally generated pPERK
(T980) antibody for pPERK detection.
Animal Use Policies. All animals used in the generation of data
included in this manuscript were housed according to all Association
for Assessment and Accreditation of Laboratory Animal Care
specifications. Experimental procedures were done in accordance
with Institutional Animal Care and Use Committee and United States
Department of Agriculture regulations.
ASSOCIATED CONTENT
■
S
* Supporting Information
KINOMEscan kinase selectivity data for compounds 1, 7, 10b,
30−36, 38−44, and 52; X-ray crystallographic data for
compounds 7, 10b, 34, 39, 44, and 52 bound to PERK. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Protein Synthesis Rescue Assay. U2OS cells were seeded at 1 ×
104 cells/mL in 96-well view-plates (PerkinElmer) and incubated
overnight. Cells were dosed with PERK inhibitor dose responses for 1
N
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
regulated translation increases tolerance to extreme hypoxia and
promotes tumor growth. EMBO J. 2005, 24, 3470−3481.
(9) Kim, I.; Xu, W.; Reed, J. C. Cell death and endoplasmic reticulum
stress: disease relevance and therapeutic opportunities. Nat. Rev. Drug
Discovery 2008, 7, 1013−1030.
Accession Codes
PDB codes for PERK with bound inhibitors: 7, 4X7J; 10b,
4X7H; 34, 4X7K; 39, 4X7L; 44, 4X7N; 52, 4X7O.
AUTHOR INFORMATION
Corresponding Authors
*A.L.S.: e-mail, adriansmith@roadrunner.com; phone, (805)
530-2367.
■
(10) Fels, D. R.; Koumenis, C. The PERK/eIF2α/ATF4 module of
the UPR in hypoxia resistance and tumor growth. Cancer Biol. Ther.
2006, 5, 723−728.
(11) Wang, M.; Kaufman, R. J. The impact of the endoplasmic
reticulum protein-folding environment on cancer development. Nat.
Rev. Cancer 2014, 14, 581−597.
*J.R.L.: e-mail, jlipford@amgen.com; phone, (805) 447-2140.
Notes
(12) Axten, J. M.; Medina, J. R.; Feng, Y.; Shu, A.; Romeril, S. P.;
Grant, S. W.; Li, W. H. H.; Heerding, D. A.; Minthorn, E.; Mencken,
T.; Atkins, C.; Liu, Q.; Rabindran, S.; Kumar, R.; Hong, X.; Goetz, A.;
Stanley, T.; Taylor, J. D.; Sigethy, S. D.; Tomberlin, G. H.; Hassell, A.
M.; Kahler, K. M.; Shewchuk, L. M.; Gampe, R. T. Discovery of 7-
methyl-5-(1-{[3-(trifluoromethyl)phenyl]acetyl}-2,3-dihydro-1H-
indol-5-yl)-7H-pyrrolo[2,3-d]pyrimidin-4-amine (GSK2606414), a
potent and selective first-in-class inhibitor of protein kinase R
(PKR)-like endoplasmic reticulum kinase (PERK). J. Med. Chem.
2012, 55, 7193−7207.
(13) Axten, J. M.; Romeril, S. P.; Shu, A.; Ralph, J.; Medina, J. R.;
Feng, Y.; Li, W. H. H.; Grant, S. W.; Heerding, D. A.; Minthorn, E.;
Mencken, T.; Gaul, N.; Goetz, A.; Stanley, T.; Hassell, A. M.; Gampe,
R. T.; Atkins, C.; Kumar, R. Discovery of GSK2656157: an optimized
PERK inhibitor selected for preclinical development. ACS Med. Chem.
Lett. 2013, 4, 964−968.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The contributions of Nick Paras and Jeffrey Petkus are
acknowledged for the original synthesis of compound 7. Josette
Carnahan, Carol Babij, and Manory Fernando are thanked for
obtaining B-Raf assay data on compounds. Andrew Tasker is
thanked for useful discussions leading to the design of 34 and
35.
ABBREVIATIONS USED
■
ATF4, activating transcription factor 4; ATF6, activating
transcription factor 6; BTP, Bis-Tris propane; CDI, 1,1′-
carbonyldiimidazole; CHoP, C/EBP homologous protein;
eIF2α, eukaryotic initiation factor 2α; ELISA, enzyme-linked
immunosorbent assay; ER, endoplasmic reticulum; GCN2,
general control nonderepressible 2; GST, glutathione S-
transferase; HATU, [dimethylamino(triazolo[4,5-b]pyridin-3-
yloxy)methylidene]dimethylazanium hexafluorophosphate;
HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid;
HRI, heme-regulated eIF2α kinase; IRE-1, inositol-requiring
enzyme 1; MES, 2-(N-morpholino)ethanesulfonic acid; MSD,
Meso Scale Discovery; PD, pharmacodynamic; PERK, protein
kinase R-like endoplasmic reticulum kinase; PKR, protein
kinase RNA-activated; pGCN2, phospho-GCN2; pPERK,
phospho-PERK; SAR, structure−activity relationship; TBTU,
O-(benzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium tetra-
fluoroborate; UPR, unfolded protein response
(14) Smith, A. L.; DeMorin, F. F.; Paras, N. A.; Huang, Q.; Petkus, J.
K.; Doherty, E. M.; Nixey, T.; Kim, J. L.; Whittington, D. A.; Epstein,
L. F.; Lee, M. R.; Rose, M. J.; Babij, C.; Fernando, M.; Hess, K.; Le, Q.;
Beltran, P.; Carnahan, J. Selective inhibitors of the mutant B-Raf
pathway: discovery of a potent and orally bioavailable aminoisoquino-
line. J. Med. Chem. 2009, 52, 6189−6192.
(15) Suzuki, A. Carbon−carbon bonding made easy. Chem. Commun.
2005, 4759−4763.
(16) Liu, L.; Norman, M. H.; Lee, M.; Xi, N.; Siegmund, A.; Boezio,
A. A.; Booker, S.; Choquette, D.; D’Angelo, N. D.; Germain, J.; Yang,
K.; Yang, Y.; Zhang, Y.; Bellon, S. F.; Whittington, D. A.; Harmange,
J.-C.; Dominguez, C.; Kim, T.-S.; Dussault, I. Structure-based design of
novel class II c-Met inhibitors: 2. SAR and kinase selectivity profiles of
the pyrazolone series. J. Med. Chem. 2012, 55, 1868−1897.
(17) Bodendorf, K.; Mildner, J.; Lehman, T. Antipyrinealdehyde and
antipyrinic acid. Justus Liebigs Ann. Chem. 1949, 563, 1−11.
(18) Shaw, G.; Sugowdz, G. Isoxazolinones. VI. The hydrogenation
of 5-amino-isoxazoles. A new synthesis of pyrimidines. J. Chem. Soc.
1954, 665−668.
REFERENCES
■
(1) Ron, D.; Walter, P. Signal integration in the endoplasmic
reticulum unfolded protein response. Nat. Rev. Mol. Cell Biol. 2007, 8,
519−529.
(19) Engelhardt, H.; Boehmelt, G.; Kofink, C.; Kuhn, D.; McConnell,
D.; Stadtmueller, H. Preparation of pyrimidinonecarboxylic acid
amides with antiproliferative activity. WO2010007114A2, 2010.
(20) Ishiyama, T.; Murata, M.; Miyaura, N. Palladium(0)-catalyzed
cross-coupling reaction of alkoxydiboron with haloarenes: a direct
procedure for arylboronic esters. J. Org. Chem. 1995, 60, 7508−7510.
(21) Doxycycline-induced overexpression of PERK in HT1080 cells
transfected with the T-REx-PERK-FLAG construct results in
significant PERK autophosphorylation. Therefore, no external UPR
activator (such as thapsigargin) is required in this cell line for
activation of PERK in the assays described.
(2) Shore, G. C.; Papa, F. R.; Oakes, S. A. Signaling cell death from
the endoplasmic reticulum stress response. Curr. Opin. Cell Biol. 2011,
23, 143−149.
(3) Carrara, M.; Prischi, F.; Ali, M. M. U. UPR signal activation by
luminal sensor domains. Int. J. Mol. Sci. 2013, 14, 6454−6466.
(4) Ma, Y.; Hendershot, L. M. The role of the unfolded protein
response in tumor development: friend or foe? Nat. Rev. Cancer 2004,
4, 966−977.
(5) Walter, P.; Ron, D. The unfolded protein response: from stress
pathway to homeostatic regulation. Science 2011, 334, 1081−1086.
(6) Vandewynckel, Y.-P.; Laukens, D.; Geerts, A.; Bogaerts, E.;
Paridaens, A.; Verhelst, X.; Janssens, S.; Heindryckx, F.; van
Vlierberghe, H. The paradox of the unfolded protein response in
cancer. Anticancer Res. 2013, 33, 4683−4694.
(7) Gao, Y.; Sartori, D. J.; Li, C.; Yu, Q.-C.; Kushner, J. A.; Simon, M.
C.; Diehl, J. A. PERK is required in the adult pancreas and is essential
for maintenance of glucose homeostasis. Mol. Cell. Biol. 2012, 32,
5129−5139.
(8) Bi, M.; Naczki, C.; Koritzinsky, M.; Fels, D.; Blais, J.; Hu, N.;
Harding, H.; Novoa, I.; Varia, M.; Raleigh, J.; Scheuner, D.; Kaufman,
R. J.; Bell, J.; Ron, D.; Wouters, B. G.; Koumenis, C. ER stress-
(22) Kolb, H. C.; Sharpless, K. B. The growing impact of click
chemistry on drug discovery. Drug Discovery Today 2003, 8, 1128−
1137.
(23) Carnahan, J.; Beltran, P. J.; Babij, C.; Le, Q.; Rose, M. J.;
Vonderfecht, S.; Kim, J. L.; Smith, A. L.; Nagapudi, K.; Broome, M. A.;
Fernando, M.; Kha, H.; Belmontes, B.; Radinsky, R.; Kendall, R.;
Burgess, T. L. Selective and potent Raf inhibitors paradoxically
stimulate normal cell proliferation and tumor growth. Mol. Cancer
Ther. 2010, 9, 2399−2410.
(24) Nagar, B.; Bornmann, W. G.; Pellicena, P.; Schindler, T.; Veach,
D. R.; Miller, W. T.; Clarkson, B.; Kuriyan, J. Crystal structures of the
O
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
kinase domain of c-Abl in complex with the small molecule inhibitors
PD173955 and imatinib (STI-571). Cancer Res. 2002, 62, 4236−4243.
(25) Steiner, T. Water molecules which apparently accept no
hydrogen bonds are systematically involved in C−H···O interactions.
Acta Crystallogr., Sect. D: Biol. Crystallogr. 1995, 51, 93−97.
(26) We were unable to obtain cocrystal structures of our
compounds with GCN2, but the structure of apo yeast GCN2 is
known. This made comparison of apo GCN2 (PDB accession code
1ZYC) with apo PERK (PDB accession code 3QD2) the most
appropriate.
(27) Cui, W.; Li, J.; Ron, D.; Sha, B. The structure of the PERK
kinase domain suggests the mechanism for its activation. Acta
Crystallogr., Sect. D: Biol. Crystallogr. 2011, 67, 423−428.
(28) Padyana, A. K.; Qiu, H.; Roll-Mecak, A.; Hinnebusch, A. G.;
Burley, S. K. Structural basis for autoinhibition and mutational
activation of eukaryotic initiation factor 2α protein kinase GCN2. J.
Biol. Chem. 2005, 280, 29289−29299.
(29) Still, W. C.; Kahn, M.; Mitra, A. Rapid chromatographic
technique for preparative separations with moderate resolution. J. Org.
Chem. 1978, 43, 2923−2925.
(30) D’Angelo, N. D.; Kim, T.-S.; Andrews, K.; Booker, S. K.;
Caenepeel, S.; Chen, K.; Freeman, D.; Jiang, J.; McCarter, J. D.; San
Miguel, T.; Mullady, E. L.; Schrag, M.; Subramanian, R.; Tang, J.;
Wahl, R. C.; Wang, L.; Whittington, D. A.; Wu, T.; Xi, N.; Xu, Y.;
Yakowec, P.; Zalameda, L. P.; Zhang, N.; Hughes, P.; Norman, M. H.
Discovery and optimization of a series of benzothiazole phosphoinosi-
tide 3-kinase (PI3K)/mammalian target of rapamycin (mTOR) dual
inhibitors. J. Med. Chem. 2011, 54, 1789−1811.
(31) QuantiGene 2.0 Reagent System. User Manual; Panomics, Inc.:
Santa Clara, CA, 2007; http://www.panomics.com/downloads/
UM13074_QG2Manual_RevB_080102.pdf.
P
DOI: 10.1021/jm5017494
J. Med. Chem. XXXX, XXX, XXX−XXX