Discovery of Irreversible p97 Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.9b00144

Inhibitors of human p97 (also known as valosin-containing protein) have been actively pursued because of their potential therapeutic applications in cancer and other diseases. However, covalent and irreversible p97 inhibitors have not been well explored.

Full text of DOI:10.1021/acs.jmedchem.9b00144

Article
pubs.acs.org/jmc
Cite This: J. Med. Chem. XXXX, XXX, XXX−XXX
Discovery of Irreversible p97 Inhibitors
Rui Ding,†,§ Ting Zhang,‡,§ Daniel J. Wilson,† Jiashu Xie,† Jessica Williams,† Yue Xu,‡ Yihong Ye,‡
and Liqiang Chen*,†
†Center for Drug Design, College of Pharmacy, University of Minnesota, Minneapolis, Minnesota 55455, United States
‡Laboratory of Molecular Biology, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health,
Bethesda, Maryland 20892, United States
S
* Supporting Information
ABSTRACT: Inhibitors of human p97 (also known as
valosin-containing protein) have been actively pursued
because of their potential therapeutic applications in cancer
and other diseases. However, covalent and irreversible p97
inhibitors have not been well explored. Herein, we report our
design, synthesis, and biological evaluation of covalent and
irreversible inhibitors of p97. Among an amide and a reverse
amide series we synthesized, we have identified a p97 inhibitor
whose functional irreversibility has been established both in
vitro and in cells. Also importantly, mass spectrometry reveals three potential cysteine residues labeled by this compound, and
mutagenesis together with computer modeling suggests Cys522 as a major site, which when modified, could compromise the
function of p97. Taken together, this new inhibitor may provide a template for designing more potent p97 inhibitors with
covalent and irreversible characteristics.
for p97.9 While D1 is relatively inactive in terms of ATP
hydrolysis compared to the D2 domain, the ATP binding in
INTRODUCTION
■
Covalent chemical inhibitors targeting biological enzymes have
D1 promotes hexamer formation and overall structural
been historically avoided in drug discovery because of potential
integrity. The D2 domain on the other hand is believed to
risks of severe immune response and idiosyncratic toxicity as a
provide the major mechanical force.6 The energy generated by
result of on- or off-target covalent modifications;1 however,
D2 ATP hydrolysis allows p97 to extract ubiquitinated proteins
this view has changed recently as several covalent inhibitors
from membranes, cellular structures or protein complexes,
have emerged as attractive drug candidates, especially in cancer
therapy.2−4 Compared to noncovalent inhibitors, covalent
inhibitors offer unique advantages, including superior bio-
chemical efficiency, nonequilibrium binding that allows for
effective competition with endogenous substrates of high
acting as a “segregase” in a wide range of cellular processes,
including endoplasmic reticulum-associated protein degrada-
tion (ERAD),10 mitochondria-associated degradation,11 and
ribosome-associated degradation.12 In general, proteins
released by p97 are delivered to the 26S proteasome for
concentration, lower dosage, and targeting of undruggable
degradation. p97 is also involved in autophagy,13 another form
of protein degradation and recycling. Therefore, p97 is
intimately associated with protein quality control and homeo-
stasis. Furthermore, p97 is involved in cellular functions, such
as DNA repair and cell cycle progression, through its segregase
activity.14
binding sites.3 Although early covalent inhibitors usually
originated from natural products, de novo design of covalent
inhibitors has received increasing attention recently. This
exciting research area has been greatly bolstered by recent
Food and Drug Administration approvals of several covalent
kinase inhibitors that had been designed and developed for
cancer treatment.3
The adenosinetriphosphatase (ATPase) p97, which is also
known as valosin-containing protein, belongs to the AAA +
(ATPase associated with various cellular activities) ATPase
family.5 p97 is a homohexamer, with each subunit comprising
an N-domain, two AAA ATPase domains (D1 and D2), and a
short unstructured C-terminal tail.6 Upon adenosine 5′-
triphosphate (ATP) binding to D1, the N-domain adopts an
axial position, a remarkable conformational change from the
initial coplanar position.7,8 Being flexible, the N-domain can
interact with a wide range of adaptor proteins that render
differential subcellular localization and substrate recognition
Given that drugs targeting the 26S proteasome have been
used in clinics to treat cancer and the intimate connection of
p97 with the proteasome in protein homeostasis regulation,
p97 has also been actively pursued as an anticancer target.
Higher levels of p97 have been reported in patients with
various cancer types, including colorectal cancer, pancreatic
cancer, thyroid cancer, breast cancer, squamous cell carcinoma,
gastric carcinoma, osteosarcoma, and lung cancer.15 In
addition, elevated p97 is associated with poor clinical
Received: January 23, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 1. Selected known p97 inhibitors.
outcomes.16 Therefore, it has been hypothesized that
inhibition of p97 might be a promising anticancer therapy,
especially in cancers where survival is critically dependent on
cellular protein homeostasis.15 This concept is further
bolstered by the fact that a p97 inhibitor has recently entered
into a phase I clinical trial.17 In addition to cancer, p97 has
been heavily investigated because mutations in p97 gene have
been linked to pathogenesis of inclusion body myopathy,
Paget’s disease of bone, and frontotemporal dementia
(IBMPFD), as well as a small subset of inherited amyotrophic
lateral sclerosis (ALS, Lou Gehrig’s disease).18 Recent studies
have pinpointed a gain-of-function activity as a potential
underlying basis for pathogenesis in IBMPFD.8,19,20 As a result,
inhibition of p97 may also represent a potential strategy for
therapeutic intervention in IBMPFD and potentially ALS.
Since p97 is implicated in a variety of human diseases and its
inhibition constitutes a promising therapeutic avenue, there
has been a growing list of p97 inhibitors,16 some of which are
shown in Figure 1. Eeyarestatin I (1a, EerI) was the first
reported p97 inhibitor that blocked ERAD and stabilized a
fluorescence-tagged proteasome reporter protein Ub-v-GFP.21
As a bifunctional compound, EerI’s nitrofuran-containing
(NFC) group probably binds the D1 domain and its aromatic
moiety is capable of interacting with hydrophobic mem-
branes.22 Further efforts have led to EerI analogues with a
truncated aromatic moiety22 and a derivative (1b) with
improved aqueous solubility.23 In addition to EerI, several
D2-targeting inhibitors have been recently discovered. N2,N4-
Dibenzylquinazoline-2,4-diamine (2, DBeQ) features a quina-
zoline core structure and mainly targets the D2 domain.24
Subsequent structure−activity relationship (SAR) studies
performed on DBeQ gave rise to ML240 (3) and ML241,
both of which selectively bind the D2 domain.25 More
recently, inhibitors built on a pyrimidine core structure, such as
4a and 4b, have been reported.17,26 CB-5083 (4b) has been
advanced into a phase I clinical trial in solid tumors, validating
p97 as an anticancer target.17 Compound 5 also features a
pyrimidine scaffold but a different substitution pattern.27
Photoaffinity experiments suggest that compound 5 and its
analogues also target the D2 domain. Alkylsulfanyl-1,2,4-
triazoles as exemplified by NMS-873 (6) were identified as
allosteric inhibitors as they bind into a tunnel between the D1
and D2 domains.28,29 UPCDC30245 (7) is also an allosteric
inhibitor, occupying a distinct region at the interface of D1 and
D2 domains as revealed by a cryo-electron microscopy (cryo-
EM) study.30 Upon binding, compound 7 is believed to
interfere with the cross-talk between the domains and
subsequent conformational changes that are required for
functional p97. In addition to the p97 inhibitors shown in
Figure 1, new compounds also include withaferin A
analogues,31 indole amides,32 trifluoromethyl and pentafluor-
osulfanyl indoles,33 chlorinated analogues of dehydrocurvular-
in,34 and oxaspirols.35
The majority of the reported p97 inhibitors are noncovalent
and reversible. Since p97 is one of the most abundant cellular
proteins, covalent and irreversible inhibition under non-
equilibrium conditions should provide a more effective
means to neutralize p97, avoiding high doses that may cause
nonspecific off-target side effects. Compared to noncovalent
inhibitor counterparts designed to target the ATP-binding D1
or D2 site, irreversible blockage may also allow more effective
competition with ATP.
A few covalent p97 inhibitors have been suggested, including
EerI and NMS-859 (8). EerI is believed to serve as a prodrug,
and its anti-p97 activity requires the NFC group to be
converted into a reactive metabolite in cells even though the
exact nature of the active species still remains unknown.36
Compound 8 features an electrophilic α-chloroacetamide,
which has been shown to selectively modify Cys522, a residue
buried in the D2 active site, but the potency is low.22 In this
work, we report our design, synthesis, and evaluation of
covalent and irreversible p97 inhibitors based on the core
structure of compounds 4a and 4b, which are among the most
potent and specific reversible p97 inhibitors tested to date.17
RESULTS AND DISCUSSION
■
Design of Covalent p97 Inhibitors. A general approach
to designing covalent inhibitor is to attach an electrophile onto
an existing reversible compound. A desirable parent compound
possesses a high reversible binding affinity that is not
compromised by the introduced electrophile. These two
features would allow the resulting inhibitor to efficiently
dock into the binding site, where the electrophile forms a
covalent bond with a nearby amino acid residue. For the
B
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 2. Design of irreversible p97 inhibitors.
a
Table 1. Evaluation of p97 Inhibitors
a
cmpd
core
X
R
Kiapp (nM)
GSH reactivity t1/2 (h)
GSH adduct
EC50 (μM)
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
4a
4b
I
I
I
I
I
I
I
I
II
II
II
II
II
II
II
II
II
III
III
CH2
CH2
CH2
CH2
CH2
CH2
O
CCH
CN
30.6 3.8
28.1 3.6
24.5 3.3
24.1 3.3
37.3 4.3
40.8 4.5
18.6 2.9
42.4 4.6
37.9 4.3
33.2 4.0
43.0 4.7
26.9 3.5
31.4 3.8
20.8 3.1
32.8 3.9
32.3 3.9
26.6 3.5
28.6 3.6
26.1 7.0
>24
>24
no
no
0.470
0.468
1.41
0.517
0.723
0.608
1.08
0.524
0.131
0.436
0.463
1.57
CH2CCH
CH2CN
CCMe
Et
>24
no
CCH
CCMe
CCH
CCMe
CCEt
CCnPr
CHCH2
Et
CCH
CCMe
CHCH2
>24
>24
0.21
16.4
13.5
21.6
5.08
no
no
O
CH2
CH2
CH2
CH2
CH2
CH2
O
O
O
CH2
O
yes
yes
yes
yes
yes
0.892
1.01
0.21
16.2
4.31
yes
yes
yes
0.0787
0.600
0.880
0.128
0.146
a
EC50 values were determined in MIA PaCa-2 cells.
electrophile, the reactivity needs to be fine-tuned to avoid
excessive activity, which can lead to nonspecific modification of
protein residues and consequent off-target toxicity. Assay
conditions have been developed to gauge the intrinsic
reactivity of a wide range of electrophiles and the consequence
of structural modifications.37−40 In short, a desired covalent
inhibitor should contain structural elements that initiate tight
binding to its target and subsequent efficient trapping of a
target residue.
To obtain covalent p97 inhibitors, we decided to use the
reversible inhibitors 4a and 4b as the parent compounds
because of their potency and specificity, both of which are
expected to bind the D2 ATP-binding site.26 A previous
computational docking study performed on compound 4a has
suggested that the phenyl group projects into a narrow cavity
that is normally occupied by the purine ring of ATP, while the
primary amide on the indole ring is more solvent-exposed.26
More importantly, examination of the residues in proximity to
compound 4a revealed three cysteine residues (Cys522, 691,
and 695), which could potentially be targeted by a properly
positioned electrophile. Among these three cysteine residues,
Cys522 is buried in a narrow cavity and it is the target of a
known p97 inhibitor 8, in part due to its small size. In
principle, Cys522 can be trapped with an electrophile
appended on the phenyl ring in compound 4a. Unfortunately,
previous structure−activity relationship (SAR) studies sug-
gested that only minor structural changes can be tolerated on
the phenyl ring;26 therefore, we decided to keep the ring intact.
Instead, we opted to focus on the primary amide, upon which
an electrophile can be introduced by minor alterations of the
functionality of this compound.
We devised two chemotypes derived from compounds 4a
and 4b (Figure 2). In the so termed amide series, as
represented by the general structure 9, the original primary
amide in 4a and 4b was extended to give a secondary amide, to
which an electrophile was appended. In the reverse amide
series as represented by the general structure 10, the primary
amide was replaced with a reverse amide equipped with an
C
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthesis of Compounds 11-27
a
Reagents and Conditions: (a) Pd2(dba)3, XPhos, Cs2CO3, dioxane, 100 °C; (b) LiOH, H2O/MeOH/tetrahydrofuran (THF); (c) amine,
N,N,N’,N′-tetramethyl-O-(1H-benzotriazol-1-yl)uronium hexafluorophosphate (HBTU), Et3N, dimethylformamide (DMF); (d) SnCl2, MeOH, 75
°C; (e) H2, Pd/C, MeOH; (f) carboxylic acid, N,N′-dicyclohexylcarbodiimide (DCC), CH2Cl2; or carboxylic acid, N-ethyl-N′-(3-
dimethylaminopropyl)carbodiimide (EDC), 1-hydroxybenzotriazole (HOBt), DMF; or acryloyl chloride, K2CO3, acetone/H2O.
electrophile. Building on the general structures of 9 and 10, we
prepared and evaluated a series of potential covalent p97
inhibitors using 4a or 4b as the reference compound (Table 1).
Specifically, for the amide series built on core I, we prepared
and tested compounds 11−18. In compound 11, we explored a
terminal alkyne that has been shown to react with active-site
cysteine residues.41,42 In compound 12, we examined a nitrile
group that has been well exploited in the design of covalent
cysteine protease inhibitors.43 Compounds 13 and 14 were
also synthesized to investigate the effect of a longer spacer
between the amide functionality and the terminal alkyne or
nitrile group, respectively. In compound 15, a methyl group
was appended upon the original terminal alkyne to investigate
if a steric hindrance would impede the anticipated nucleophilic
attack by a thiol. We also synthesized compound 16, a
reversible inhibitor that contained a simple propyl secondary
amide of a comparable length but without an electrophilic
capacity. Compounds 11−16 were based on 4a, while
inhibitors 17 and 18 were built on 4b, featuring a terminal
alkyne and a methyl-capped alkyne, respectively.
For the reverse amide series built on core II, we generated
compounds 19−27, in which a Michael acceptor was installed.
While compound 19 contained a propiolamide functionality,
its terminal alkyne was further capped with alkyl groups of
increasing length (namely, methyl, ethyl, and propyl), leading
to compounds 20−22, respectively. These three compounds
were synthesized to probe simple alkyl groups and their steric
effect on the resulting inhibitors, providing a possibility to fine-
tune the reversibility. For a similar purpose, we prepared
compound 23, which had an acrylamide with attenuated
electrophilicity relative to propiolamide. To provide an
example of a reversible inhibitor with similar structural
elements, we synthesized compound 24 devoid of a Michael
acceptor. Like inhibitors 17 and 18 in the amide series, we also
investigated compounds 25−27, which were based on 4b and
featured a propiolamide, a methyl-capped propiolamide, and
an acrylamide, respectively.
Chemical Synthesis. The chemical synthesis of the new
p97 inhibitors is depicted in Scheme 1. To prepare the amide
inhibitors based on compound 4a, known chloride 28a26 and
indole 2926 underwent a palladium-mediated C−N coupling
reaction to give methyl ester 30a, hydrolysis of which afforded
carboxylic acid 31a. Amide-coupling reactions were performed
on acid 31a, and various amines gave inhibitors 11−16. To
obtain the amide inhibitors based on compound 4b, an
identical synthetic sequence was performed with chloride
28b26 and indole 29 as starting materials, leading to
compounds 17−18. To synthesize the reverse amide inhibitors
based on compound 4a, known chloride 28a and indole 3244
were subjected to a palladium-mediated C−N coupling
reaction to afford nitrate 33a. Upon reduction of 33a, the
resulting aniline 34a was coupled with carboxylic acids or
treated with an acyl chloride to give compounds 19−24. The
reverse amide inhibitors based on compound 4b were obtained
in a similar manner, starting with chloride 28b and indole 32.
Biochemical Evaluation of p97 Inhibitors. We estab-
lished an ATPase assay in which inorganic phosphate released
by the p97-catalyzed reaction was continuously monitored
using a method previously developed by Webb.45 The method
relies on purine nucleoside phosphorylase (PNP), which uses
D
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
inorganic phosphate to convert 7-methyl-6-thioguanosine
(MesG) into 2-amino-6-mercapto-7-methylpurine. The latter
product was then detected spectrophotometrically. Since
parent compounds 4a and 4b were reported to possess high
inhibitory activity,26 we chose to measure them and new
analogues under tight-binding conditions, fitting data to the
Morrison equation to generate Kiapp values.46 In our hands,
compounds 4a and 4b both exhibited low nanomolar Kiapp
values (Table 1), which were comparable to those reported
previously.26
assay conditions with the possible exceptions of compounds 19
and 25.
Antiproliferative Activity. To evaluate the anticancer
activity of the newly identified p97 inhibitors, we tested their
antiproliferative activity in MIA PaCa-2 cells, a pancreatic
cancer cell line. Consistent with their biochemical anti-p97
activity, our p97 inhibitors showed high antiproliferative
activity. The majority of the compounds possessed an EC50
value <1 μM regardless of their core structures (Table 1).
Compounds 4a and 4b displayed remarkably high antiprolifer-
ative activity, which is consistent with their reported anticancer
property.26 Compounds 11−18 from the amide series
exhibited reduced anticancer activity, suggesting that elongat-
ing the original primary amide did not gain any cytotoxic
activity, at least in MIA PaCa-2 cancer cell lines. Among
compounds selected from the reverse amide series, 19 and 25
had low EC50 values comparable to those of compounds 4a
and 4b, a property that is attributable to their reactive
propiolamide groups, which probably account for compound
19’s cytotoxicity in Ub-v-GFP cells (see below). In
comparison, a slightly reduced activity was observed in
compounds 20 and 26, both of which had a methyl capping
group. An ethyl capping group did not further decrease
anticancer activity as seen in compound 21 versus 20.
However, a longer propyl capping group led to significantly
diminished activity. Furthermore, the intrinsic electrophilicity
of a Michael acceptor was also an important factor as seen in
23 versus 19 and 27 versus 25. In short, new p97 inhibitors all
showed potent anticancer activity. Although there is no unified
SAR on antiproliferative activity in MIA CaPa-2 cells, the
intrinsic electrophilicity of and the steric hindrance surround-
ing a Michael acceptor are clearly important determinants of
anticancer activity in the reverse amide series.
Reversibility under the Biochemical Assay Condi-
tions. Since compound 20 from the reverse amide series could
form a covalent bond with GSH, we next evaluated its
reversibility in blocking wild-type (WT) p97 ATPase activity.
Compounds 4b and 3, both of which lacked a reactive group
and were therefore expected to be reversible, were used as
references. Compounds were first tested using dialysis because
their tight-binding nature precluded the use of the classic jump
dilution experiment. After being treated with dimethyl
sulfoxide (DMSO) as a control or with compounds 20 (750
nM), 4b (1610 nM), and 3 (3200 nM) at concentrations 10
times of their corresponding IC50, the reaction wells were
subjected to a two-step dialysis to remove inhibitors. The
DMSO wells retained full activity, indicating that there was no
unfolding of p97 during dialysis (Table 2). Surprisingly,
compound 4b acted largely as an irreversible inhibitor with
only ∼20% activity recovered. Because compound 4b lacks a
Michael acceptor, the observed irreversibility might be due to
high affinity to p97. Indeed, extremely tight binders may
appear irreversible under similar biochemical assay condi-
tions47,48 because they can rebind the enzyme before migrating
Examination of Kiapp values in Table 1 revealed that the
newly synthesized compounds all displayed anti-p97 activity
similar to that of 4a or 4b. This finding was not completely
unexpected because the parent compounds 4a and 4b already
possessed high potency with Ki values in the nanomolar range;
therefore, relatively minor structural modifications were
unlikely to drastically alter the inhibitory activity. In short,
we identified amides and reverse amides that possessed
inhibitory activity comparable to that of the known p97
inhibitor 4b. Further evaluation of these new compounds was
therefore justified.
Measurement of Intrinsic Reactivity. We used a
modified liquid chromatography (LC)−tandem mass spec-
trometry (MS/MS) assay to test selected compounds for their
intrinsic reactivity toward glutathione (GSH),38,40 an abundant
cellular cysteine-containing tripeptide. A moderate activity
would suggest cysteine reactivity, and thus a capacity to act as
an irreversible inhibitor, but high activity may indicate the
possibility of nonspecific off-target reactivity in cells. The
reactivity, measured in half-life values, is indicated in Table 1.
Compounds 11−12, 15, and 17−18 showed no reactivity
toward GSH as indicated by t1/2 > 24 h and no detection of
GSH adducts. Lack of reactivity is probably due to the
attenuated nucleophilicity of noncatalytic GSH cysteine
relative to the catalytic cysteines that can be successfully
captured by terminal alkynes or nitriles in properly oriented
cysteine protease inhibitors.
In contrast to the amide series, compounds selected from
the reverse amide series showed appreciable albeit varied
reactivity toward GSH. Compound 19 exhibited high intrinsic
reactivity as judged by t1/2 < 30 min, a result that is in line with
the reactivity reported for analogous propiolamides. This high
reactivity is likely responsible for compound 19’s cytotoxicity
(see below). As expected, steric capping of the terminal alkyne
in compound 19 with a small alkyl group as seen in
compounds 20−22 markedly reduced the intrinsic reactivity
by >60-fold. When the propiolamide in 19 was replaced with
an acrylamide, the resulting compound 23 possessed
significantly attenuated reactivity due to the reduced electro-
philicity of acrylamide. When compounds 25−27 were tested,
their t1/2 values were almost identical to those determined for
compounds 19, 20, and 23, respectively, indicating that an
extra oxygen atom in the former three compounds had
minimal effect on their intrinsic reactivity. Consistent with the
observed GSH reactivity, GSH adducts were detected for
compounds 19−23 and 25−27.
a
Table 2. Enzymatic Activity Recovered (%) after Dialysis
We also determined the reactivity toward dithiothreitol
(DTT), a thiol-containing reducing agent that was present in
our p97 biochemical assay. Since the amide series showed no
reactivity toward GSH, we focused on the reverse amide series
(Table S1). The DTT reactivity generally mirrored those
toward GSH. Therefore, we concluded that DTT had no
significant effect on the compounds under the biochemical
20
4b
3
DMSO
WT p97
C522A
p97 delta N
3
6
0
1
1
19
60 21
11
4
82
100
ND
6
100
100
100
a
a
ND, not determined.
E
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 3. Inhibition of p97 stabilizes Ub-v-GFP. (A) Experimental scheme. (B) Whole-cell extract of Ub-v-GFP cells treated with the inhibitors as
indicated (4 h) was subject to immunoblotting (IB) analysis. Anti-p97 blot serves as a loading control. (C) As in (B), except that the indicated
inhibitors (2 μM) were tested.
out of a dialysis cassette. Rebinding is more likely to occur with
enzymes bearing multiple binding sites for inhibitor. Indeed,
because the six protomers of p97 showed positive cooperation
with regard to ATP hydrolysis,49,50 reversing its activity would
require complete removal of the inhibitor from all six subunits
of a p97 hexamer. Consistent with our model, when 4b was
tested with a p97 mutant (C522A) that bore a mutation near
the inhibitor binding site, its reversibility was significantly
enhanced (Table 2), presumably due to reduced affinity.
Additionally, compound 3, a weaker p97 inhibitor, exhibited
stronger reversibility for WT p97 than 4b with 82% activity
recovered after dialysis. Likewise, this compound appeared
almost completely reversible when tested with the p97 C522A
mutant (Table 2, see below). Collectively, these results
supported the notion that the apparent irreversibility of 4b
in WT p97 was due to its high binding affinity and the intrinsic
positive cooperation among p97 subunits, therefore validating
our dialysis assay conditions. Under these conditions, when
WT p97 was treated with compound 20, merely 3% activity
was recovered after dialysis. This result indicated that 20 is an
irreversible p97 inhibitor under the assay conditions.
Inhibition and Recovery of p97 Activity in Cells. We
next evaluated selected compounds for their p97 inhibitory
activity in cells stably expressing the proteasome substrate Ub-
v-GFP whose degradation also requires p97 (Figure 3A).24
Inhibition of p97 ATPase activity is expected to stabilize this
short-lived protein, leading to its accumulation in cells. Indeed,
in cells treated with MG132 (20 μM), a proteasome inhibitor,
the level of Ub-v-GFP was dramatically increased (Figure 3B,
lane 2 vs 1). Likewise, cells exposed to these newly developed
p97 inhibitors, even at 2 μM, accumulated Ub-v-GFP to a
similar level to cells treated with 20 μM MG132 (Figure
3B,C). Among the chemicals tested, compound 19 did not
increase Ub-v-GFP as much as other compounds, but this was
due to a strong toxicity unrelated to p97 inhibition, which led
to reduced cell number (Figure 3B, lanes 3−5). To rule out
that our compounds blocked Ub-v-GFP degradation through
inhibition of the proteasome, we monitored a signal sequence-
deleted T-cell receptor (TCR) variant (ΔSS TCRα-YFP),
whose degradation is mediated by proteasome but does not
require p97.21,22 Degradation of ΔSS TCRα-YFP was inhibited
by MG132 but not by compound 20 or NMS-873 (Figure S1),
excluding the possibility that our compounds inhibited Ub-v-
GFP degradation by blocking the proteasome. All together,
these results demonstrate that the newly synthesized chemicals
can inhibit p97 function in mammalian cells.
To explore the reversibility of these p97 inhibitors in cells,
we developed a drug washout assay (Figure 4). To this end,
Ub-v-GFP cells were either treated continuously with an
inhibitor (2 μM) for 4 h (treatment A) or treated for 1 h and
subsequently cultured in inhibitor-free medium for three
additional hours (treatment B) (Figure 4A). The total
incubation time was the same for both schemes; therefore,
for an irreversible inhibitor, a short period of inhibition (1 h,
treatment B) should be sufficient to render complete p97
inactivation even after compound withdrawal. Consequently,
the level of Ub-v-GFP accumulation between the two
treatment conditions should be similar. Indeed, when cells
were exposed to the irreversible proteasome inhibitor MG115,
similar levels of Ub-v-GFP accumulation were observed
between the two treatment conditions (Figure 4D). By
contrast, the previously reported reversible p97 inhibitor
NMS-87329 only caused significant Ub-v-GFP accumulation
under treatment A. Thus, the assay could effectively discern
the differential reversibility of inhibitors in cells. When we
tested compound 4b using this assay, as predicted from its lack
F
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 4. Reversibility of newly identified p97 inhibitors in cells. (A) Experimental design. (B) Ub-v-GFP reporter cells were treated with the
indicated inhibitors (2 μM) following either the A or B scheme. Cell lysates were analyzed by immunoblotting. (C) Structure of compound 20
(LC-1028). (D) Compound 20 is an irreversible inhibitor. Cells treated with the indicated inhibitors were lysed and analyzed by immunoblotting.
(E) A373 cells stably expressing US11 and HA-tagged MHC class I heavy chain H2A were treated with the indicated inhibitor (2 μM) for 1 h. The
cells were then incubated in inhibitor-free medium that contains cycloheximide (50 μg/mL). At the indicated time points, a portion of the cells was
harvested. Cell lysates prepared were analyzed by immunoblotting. (F) Compound 20 is a potent p97 inhibitor in cells. The cells treated with the
indicated compounds were analyzed by immunoblotting.
G
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 3. Summary of the p97 Peptides Covalently Modified by Compound 20 As Detected by LC−MS/MS
a
ModScore >20 indicates that the covalent modification by compound 20 was localized at the proposed residue with >99% certainty. C# (in red)
indicates a cysteine residue that was modified by compound 20 through a conjugate addition reaction at the alkyne group of compound 20.
of a Michael acceptor, it only induced strong accumulation of
Ub-v-GFP under treatment A conditions, but not under
treatment B conditions (Figure 4B). This result confirms our
hypothesis that 4b is a reversible inhibitor, but its reversibility
is masked by the tight affinity to p97 in vitro.
exhibited by compounds 20−22 could be attributed to the
possibility that a longer alkyl cap impeded the necessary
conformation change and subsequent covalent bond for-
mation.
In in vitro studies, compounds 4b and 20 both behaved as
irreversible inhibitors; however, in cells, compound 4b was
largely reversible, whereas 20 was irreversible. This discrepancy
may be explained by the different biological settings in which
the compounds were tested. In cells, when a reversible
inhibitor like 4b is released from p97, it can be pumped out of
the cells by ABC transporters, manifesting its reversible nature;
however, in dialysis tubes, it may rebind p97 quickly, so it
behaves like an irreversible inhibitor. Alternatively, compounds
4b upon release from p97 may be subjected to cellular
metabolism, giving rise to inactive species, which would reduce
its effective concentration. By contrast, for covalent inhibitor
compound 20, cellular metabolism should have a minimal
effect on its activity as covalent bonding with p97 may prevent
metabolism. While future efforts are needed to elucidate the
different cellular behavior of compounds 20 and 4b, our
washout experiment corroborated that compound 20 was an
irreversible inhibitor, a conclusion that was further supported
by our LC−MS/MS experiment (see below).
When compound 19 was tested under the same
experimental conditions, severe cytotoxic activity compro-
mised a definite determination of reversibility (Figure 3B). In
comparison, compound 20, which contained a methyl-capped
propiolamide (Figure 4C) showed no cytotoxicity within the 4
h treatment period. More importantly, it was clearly
irreversible as judged by almost the same level of Ub-v-GFP
accumulation regardless of the treatment scheme (Figure 4B).
To further test the irreversibility of compound 20, we tested
the degradation of another p97 substrate, MHC class I heavy
chain, in cells co-expressing the human cytomegalovirus
protein US11. US11 is a type I membrane protein that
induces the degradation of newly synthesized MHC class
heavy chain.51 When these cells were treated with either NMS-
873 or compound 20 at 2 μM for 1 h, the MHC class I heavy
chain accumulated in these cells. After drug washout and
addition of cycloheximide to prevent protein synthesis, cells
treated with NMS-873 degraded MHC class I heavy chain
rapidly with t1/2 less than 1 h. By contrast, cells treated with
compound 20 turned over MHC class I heavy chain at a much
reduced rate compared to NMS-873 (Figure 4E). Collectively,
these results strongly indicate that the effect of compound 20 is
largely irreversible in cells. A dose titration experiment showed
that treating cells with 0.5 μM compound 20 could lead to
similar levels of Ub-v-GFP accumulation to cells treated with
20 μM MG132, demonstrating the potency of this compound
(Figure 4F). Interestingly, compound 21, which featured an
ethyl capping group, also displayed a high degree of
irreversibility (Figure 4B). However, when a propyl group
was attached to the terminal alkyne, the resulting compound
22 completely lost irreversibility. These findings suggest that
the methyl cap in compound 20 might induce a steric effect
that hinders a nucleophilic attack on its Michael acceptor. This
notion is supported by a drastic increase of t1/2 observed for 20
versus 19 in the GSH reactivity assay (Table 1). Gratifyingly,
the methyl cap did not lead to complete loss of irreversibility.
However, a longer cap like an ethyl or propyl group had a
detrimental effect on inhibition irreversibility. Since the
intrinsic GSH reactivity determined for compounds 20−22
were comparable (Table 1), the difference in reversibility could
not be simply explained by increased steric hindrance induced
by a longer alkyl cap. Because our computational study of
compound 20 and its analogues (see below) suggested that a
significant conformation change was required for formation of
a covalent bond between compound 20 and its potential target
residue Cys522 (vide infra), the differential reversibility
In addition to the reverse amide series, we also tested
compounds 11, 15, 17, and 18, which belonged to the amide
series. Consistent with their lack of intrinsic GSH reactivity,
these inhibitors were essentially all reversible under the assay
conditions (Figure 4B). In summary, we have identified
compound 20 as a potent irreversible inhibitor of p97, which
justified further investigation of its mode of action.
Identification of Residues Covalently Modified by
Compound 20. To experimentally identify the residue(s)
modified by compound 20, recombinant p97 was treated with
compound 20 with DMSO as a control. In addition, we also
treated p97 with compound 4b as an additional control. After
sodium dodecyl sulfate-polyacrylamide gel electrophoresis
(SDS-PAGE) separation, the p97-containing bands were
excised and analyzed by LC−MS/MS. The resulting tandem
spectra were searched for covalent modifications that resulted
from a cysteine residue’s conjugate addition to the alkyne
group of compound 20. The probability of the proposed
modification was scored by ModScore, a program that was
based on the algorithm Ascore.52 A ModScore of >20 indicated
covalent modification, a specified site with >99% certainty.52
Among the 12 cysteine residues present in p97, only three
(Cys105, 522, and 535) were found to be modified (Tables 3
and S2, and Figure S2). Among them, Cys105 and Cys522
appeared more prominent as they were each identified several
times (21 and 5, respectively) with ModScore >20. As
expected, when p97 was treated with 4b, no covalent
H
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 5. Proposed binding modes and interactions of selected p97 inhibitors docked into the cryo-EM structure of human p97 (PDB 5FTK). (A)
Binding modes of compounds 4b (magenta) and 19 (green) at the active site of the D2 domain. (B) Binding modes of compounds 20 (magenta)
and 21 (green) at the active site of the D2 domain. (C) Two-dimensional (2D) representation of potential interactions between compound 20 and
p97. (D) Potential binding mode of compound 20 after formation of a covalent bond with Cys522. Upon covalent bond formation, a significant
change in the orientation of compound 20 was observed. The phenyl group, which was originally placed at the interface of subunits D and E,
flipped and now mainly interacted with residues from subunit E.
modification by 4b was detected. We therefore focused our
study on Cys105 and C522.
experiments. Remarkably, the detected peptide sequence
(covalently modified) was identical to the one observed for
compound 20 (Table 3). Thus, compound 20 was able to label
Cys522, a residue susceptible to covalent modification by small
molecules.
We next tested the reversibility of compound 20 using the
p97 C522A mutant. Surprisingly, it remained essentially
irreversible for C522A, an observation that could potentially
exclude Cys522 as a main target. However, as discussed above,
it is also possible that compound 20 appeared irreversible
because it is a tight binder for both WT p97 and p97 C522A
similarly to 4b for WT p97.
Previously reported iodoacetamide and compound 8 contain
chemically reactive α-iodo- and α-chloroacetamide warheads,
respectively, which are functional groups undesired in drug
discovery. By contrast, compound 20 features a relatively inert
methyl-capped amide warhead (as judged by its low intrinsic
reactivity toward GSH or DTT). Thus, it is more suitable for
drug development. Additionally, Cys522 is buried within the
D2 ATPase active site, indicative of a specific interaction with
compound 20. Taken together, as a covalent p97 inhibitor,
compound 20 is chemically inert until it binds to p97 and
subsequently traps its target residue. With these desired
features, compound 20 serves as an excellent template for
further developing irreversible p97 inhibitors.
Cys105 is solvent-exposed on the surface of the N domain,
and it does not contribute to the ATP-binding pockets (Figure
S3). These features render it more accessible to compound 20,
giving rise to the relatively high abundance of the identified
peptide. However, previous studies have demonstrated that the
N domain is dispensable for p97 ATPase activity.53 When a
p97 mutant lacking the N domain was tested, compound 20
still effectively inhibited it in an irreversible manner (Table 2),
further suggesting that Cys105 is unessential for compound
20-mediated p97 inhibition. Thus, although compound 20
might form a covalent interaction with Cys105, this
modification could not solely account for the ATPase
inhibition by compound 20. Nonetheless, because the N
domain is extensively involved in interactions with a number of
co-factors critical for p97’s biological functions, future
investigation on the consequence of compound 20’s covalent
modification of Cys105, especially in cells, is justified.
The relatively abundant Cys522-containing peptides identi-
fied by our LC−MS/MS experiment together with its
proximity to the D2 active site strongly suggested it as a
major modification site relevant to p97 inhibition by
compound 20. Cys522 has been shown to be the target
residue of iodoacetamide54 and compound 828 based on MS
I
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Noncompetitive Inhibition by Compound 20. To
determine if compound 20 was competitive with ATP, we
measured its Kiapp values at increasing ATP concentrations
(0.312−5.00 mM) (Table S3). Under these conditions, 20 was
not conclusively ATP-competitive, a finding that could be
explained by several reasons. First, as a covalent and
irreversible inhibitor, compound 20 has an intrinsic advantage
over a noncovalent inhibitor when competing with endoge-
nous ATP. Second, compound 20 binds p97 with a Kd value in
the low nanomolar range, whereas the affinity of ATP for D1
and D2 domain is ∼2 μM (Kd).55 Therefore, ATP may not
effectively compete with compound 20 even at higher
concentrations. Third, there is a positive corporation with
regard to ATP hydrolysis among the six p97 protomers.49,50
Hence, ATP may need to replace all of the inhibitor molecules
in the p97 hexamers to restore the enzymatic capacity, posing a
high barrier for ATP to compete with compound 20.
that the inhibitor adopted an orientation similar to the one
predicted by our model.
Next, we used the same docking method to investigate the
interactions of a few selected new inhibitors with p97.
Intriguingly, compound 19 (green, Figure 5A) adopted a
conformation and engaged in hydrogen bonding in a manner
similar to that of compound 4b (Figure S6). The terminal
acetylene of compound 19 fitted into a small pocket formed by
residues, including Thr623 and Ala622. By contrast, com-
pounds 20, 21, and 22, which contained an extra methyl, ethyl,
and propyl cap, respectively, exhibited a different mode of
binding. Take compound 20 (magenta, Figure 5B) as an
example, unlike 4b, its indole moiety instead of the phenyl ring
protruded into the narrow cavity that is normally occupied by
the adenine ring of ADP. As a result, the phenyl ring was
sandwiched between adjacent subunits D and E, participating
in mainly lipophilic interactions with the surrounding residues,
including Pro636 from the subunit D. Several structural
features contribute to this new binding mode. First, the small
pocket formed by Thr623 and Ala622 could not accommodate
the extra methyl cap of compound 20 that is absent from
compound 19. Second, the linear reverse amide snugly fitted
into a thin tunnel (mainly formed by lipophilic residues) at the
back of the adenine-binding cavity, engaging in lipophilic
interactions with the tunnel residues (Figure 5C). In addition,
the amide carbonyl formed a hydrogen bond with the
backbone NH of Gly480, which in 5FTK interacted with the
N1 nitrogen of the adenine ring. Compounds 21 (green,
Figure 5B) and 22 (Figure S7) displayed an identical binding
mode to 20 because the thin tunnel was open-ended and
therefore could accommodate the longer-cap groups in these
two compounds.
To examine whether the thin tunnel could also be accessed
by other reverse amides, we also performed docking study on
compounds 23 and 24, which contained a terminal double
bond and an ethyl group, respectively. Intriguingly, neither
compound was able to adopt the same binding mode displayed
by compounds 20−22, indicative of the importance of the
acetylene group and the short alkyl cap. We proposed that a
straight acetylene group allowed the whole reverse amide side
chain to adopt a conformation that perfectly matched the
contour of the thin tunnel in p97. Moreover, the short alkyl
caps were expected to interact with the lipophilic residues
within the thin tunnel, enhancing binding affinity.
Notably, the structural model of compound 20 in complex
with p97 positioned the alkyne functional group close to
Cys522 (5.7 Å between the sulfur atom of Cys522 and the
terminal carbon atom of the acetylene functionality),
suggesting that a nucleophilic attack by Cys522 and
subsequent formation of a covalent bond was possible. This
notion was further supported by our covalent docking57 using
Cys522 as a nucleophile (Table S5, Figures S8 and S9),
demonstrating that a covalent bond between Cys522 and 20
was indeed computationally feasible (Figure 5D).
In addition to the reverse amide series, we also performed
molecular docking on compounds 11 and 15 selected from the
amide series. Compound 11 exhibited a binding mode similar
to compounds 4b and 19 (Figure S10). Interestingly,
compound 15, which contained an extra methyl cap, was
also able to adopt an analogous binding conformation even
though there was a minor difference in the triple-bond
orientation, in contrast to the drastic change in binding mode
observed for 19 versus 20.
Computational Study. Our studies showed that com-
pound 20 modified Cys522, a residue that is positioned near
the p97 D2 active site, where 4b and other analogous
inhibitors bind. Also interestingly, this residue has been shown
to be covalently modified by structurally simple iodoaceta-
mide54 and compound 8.28 Furthermore, mutating Cys522
reduced the irreversibility of 4b and 3 but did not affect the
reversibility of compound 20, suggesting that 20 might utilize a
unique binding mode to engage p97 and subsequently form a
covalent bond with Cys522. To obtain insight into the binding
mode of compound 20 and its analogues, we performed a
computational docking study (Table S4) using a cryo-EM
structure of hexameric human p97 (Protein Data Bank (PDB)
5FTK) that consists of subunits A−F.30 We first focused on
the D2 active site. Because the nucleotide-binding pocket is
shaped by residues from two adjacent subunits, both subunits
might contribute to inhibitor binding. We therefore used the
combined subunits E and D as a template to generate a
docking grid that encompassed the D2 active site (E subunit)
and the surrounding residues, including those from subunit D.
Using this docking grid, we first examined compound 4b
(magenta, Figure 5A), whose phenyl group was assumed to
project into a narrow cavity occupied normally by the adenine
ring of adenosine 5′-diphosphate (ADP), as suggested by a
previous computational modeling study.26 In this model, the
primary amide of 4b was proposed to interact with Ser664 and
Lys663, which were located opposite to the ADP phosphate
binding site. In our study, compound 4b could be docked into
the same D2 ATPase active site, but the primary amide of 4b
engaged a different set of interactions with p97, forming
hydrogen bonds with Lys524, Asn624, and Asp577 within the
ADP phosphate binding site (Figure S4). The discrepancy is
probably due to the fact that we used two adjacent subunits as
the template, whereas the previous study used one isolated
subunit in docking computation. When two adjacent subunits
were included in the model, a relatively crowded environment
created at the opposite side of the ADP-binding pocket by the
intruding residues from the neighboring subunit prevented the
primary amide functionality from forming hydrogen bonds.
Consistent with this interpretation, when we performed
computational docking using a grid generated from the
isolated subunit E, compound 4b was able to adopt a docking
pose (Figure S5) identical to that proposed previously,
validating our computing methodology.26 In further support
of our model, a recent X-ray crystallography study56 of
compound 4b in complex with p97 delta N (3.77 Å) revealed
J
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
was then diluted with water (30 mL) and extracted with dichloro-
methane (30 mL × 3). The combined organic layer was dried over
Na2SO4 and filtered. The filtrate was concentrated in vacuo, and the
residue was purified by flash column chromatography using EtOAc/
hexanes (50%) to afford compound 11 as a white solid (27.2 mg,
100%). 1H NMR (CDCl3, 600 MHz) δ 8.13 (d, J = 8.3 Hz, 1H), 7.46
(d, J = 7.5 Hz, 1H), 7.38−7.29 (m, 5H), 7.08 (t, J = 7.9 Hz, 1H), 6.77
(s, 1H), 6.31 (t, J = 5.1 Hz, 1H), 5.03 (t, J = 5.5 Hz, 1H), 4.74 (d, J =
5.5 Hz, 2H), 4.31 (dd, J1 = 5.1 Hz, J2 = 2.5 Hz, 2H), 2.82−2.78 (m,
2H), 2.62 (s, 3H), 2.41−2.38 (m, 2H), 2.28 (t, J = 2.5 Hz, 1H),1.95−
1.86 (m, 4H). High-resolution mass spectrometry (HRMS) (ESI+)
calcd for C28H28N5O (M + H)+ 450.2288, found 450.2287.
In short, our docking study revealed that compound 20 was
able to adopt a binding mode distinct from their parent
compounds 4a and 4b. This novel binding mode allowed for
covalent bond formation between compound 20 and Cys522.
A covalent bond between compound 20 and p97 might
account for the irreversibility of 20.
CONCLUSIONS
■
We have designed and synthesized two series of p97 inhibitors
based on the core structure of the known compounds 4a and
4b. While both the amide and reverse amide series exhibited
largely comparable anti-p97 and antiproliferative activity,
assessment of the latter group yielded compound 20 as a
potential covalent and irreversible p97 inhibitor, which
featured a methyl-capped acetylene electrophilic group.
Evaluation of compound 20 under the dialysis conditions
established it as an irreversible inhibitor. This conclusion was
further supported by its functional irreversibility in cells. LC−
MS/MS study revealed cysteine 522 as a candidate for covalent
modification by compound 20, and our computational study
yielded a plausible model that is consistent with this notion. In
short, we have identified compound 20 as a potent and
irreversible inhibitor that covalently modifies p97. Whether
compound 20 is suitable for clinical study remains to be tested,
but as an irreversible inhibitor, its pharmacokinetics and
pharmacodynamics are expected to be uncoupled and its
duration of action should be prolonged, leading to a less
frequent dosing schedule. Such desired properties will be
explored for compound 20 and its advanced analogues.
Importantly, its unique binding mode and structural features
provide a starting point for designing improved p97 inhibitors
regardless of their mode of inhibition (covalent or non-
covalent).
1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-N-
(cyanomethyl)-2-methyl-1H-indole-4-carboxamide (12). Com-
pound 12 was prepared from acid 31a in a fashion similar to the one
1
described for compound 11. White solid, 27.1 mg, yield 100%. H
NMR (CDCl3, 600 MHz) δ 8.14 (d, J = 8.2 Hz, 1H), 7.43 (d, J = 7.5
Hz, 1H), 7.38−7.29 (m, 5H), 7.08 (t, J = 7.9 Hz, 1H), 6.75 (s, 1H),
6.50 (t, J = 5.7 Hz, 1H), 5.04 (t, J = 5.5 Hz, 1H), 4.74 (d, J = 5.5 Hz,
2H), 4.40 (t, J = 5.7 Hz, 2H), 2.82−2.78 (m, 2H), 2.62 (s, 3H),
2.42−2.38 (m, 2H), 1.95−1.86 (m, 4H). HRMS (ESI+) calcd for
C27H27N6O (M + H)+ 451.2241, found 451.2241.
1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-N-
(but-3-yn-1-yl)-2-methyl-1H-indole-4-carboxamide (13). Com-
pound 13 was prepared from acid 31a in a fashion similar to the one
1
described for compound 11. White solid, 28.7 mg, yield 100%. H
NMR (CDCl3, 600 MHz) δ 8.11 (d, J = 8.3 Hz, 1H), 7.48 (d, J = 7.4
Hz, 1H), 7.38−7.28 (m, 5H), 7.08 (t, J = 7.9 Hz, 1H), 6.80 (s, 1H),
6.54 (t, J = 6.0 Hz, 1H), 5.04 (t, J = 5.5 Hz, 1H), 4.75 (d, J = 5.5 Hz,
2H), 3.67 (q, J = 6.2 Hz, 2H), 2.82−2.78 (m, 2H), 2.62 (s, 3H), 2.56
(td, J1 = 6.4 Hz, J2 = 2.6 Hz, 2H), 2.42−2.37 (m, 2H), 2.05 (t, J = 2.6
Hz, 1H), 1.94−1.86 (m, 4H). HRMS (ESI+) calcd for C29H30N5O (M
+ H)+ 464.2445, found 464.2450.
1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-N-(2-
cyanoethyl)-2-methyl-1H-indole-4-carboxamide (14). Com-
pound 14 was prepared from acid 31a in a fashion similar to the
one described for compound 11. White solid, 27.0 mg, yield 96%. 1H
NMR (CDCl3, 600 MHz) δ 8.13 (d, J = 8.2 Hz, 1H), 7.45 (d, J = 7.6
Hz, 1H), 7.38−7.28 (m, 5H), 7.08 (t, J = 7.9 Hz, 1H), 6.77 (s, 1H),
6.64 (t, J = 6.0 Hz, 1H), 5.03 (t, J = 5.5 Hz, 1H), 4.74 (d, J = 5.5 Hz,
2H), 3.74 (q, J = 6.3 Hz, 2H), 2.82−2.78 (m, 2H), 2.77 (t, J = 6.3 Hz,
2H), 2.62 (s, 3H), 2.41−2.38 (m, 2H), 1.94−1.86 (m, 4H). HRMS
(ESI+) calcd for C28H29N6O (M + H)+ 465.2397, found 465.2403.
1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-N-
(but-2-yn-1-yl)-2-methyl-1H-indole-4-carboxamide (15). Com-
pound 15 was prepared from acid 31a in a fashion similar to the one
EXPERIMENTAL SECTION
■
Chemical Synthesis. All commercial reagents were used as
provided unless otherwise indicated. An anhydrous solvent-dispensing
system (J.C. Meyer) using two packed columns of neutral alumina
was used for drying THF, Et2O, and CH2Cl2, whereas two packed
columns of molecular sieves were used to dry DMF. Solvents were
dispensed under argon. Flash chromatography was performed with
Ultra Pure silica gel (SiliCycle) or with RediSep Rf silica gel columns
on a Teledyne ISCO CombiFlash Rf system using the solvents as
indicated. Nuclear magnetic resonance spectra were recorded on a
Varian 600 MHz spectrometer with Me4Si or signals from residual
solvent as the internal standard for 1H. Chemical shifts are reported in
ppm, and signals are described as s (singlet), d (doublet), t (triplet), q
(quartet), m (multiplet), br s (broad singlet), and dd (double
doublet). Values given for coupling constants are of first order. High-
resolution mass spectra were recorded on an Agilent TOF II TOF/
MS instrument equipped with either an electrospray ionization (ESI)
or an atmospheric pressure chemical ionization interface. Analysis of
sample purity was performed on an Agilent 1200 Infinity series high-
performance liquid chromatography (HPLC) system with a
Phenomenex Gemini C18 column (5 μm, 4.6 × 250 mm2). HPLC
conditions were as follows: solvent A = water, solvent B = MeCN or
MeOH, and flow rate = 2.0 mL/min. Compounds were eluted with a
gradient of from 10 to 100% MeCN/water or from 10 to 100%
MeOH/water in 15 min. Purity was determined by the absorbance at
254 nm. All tested compounds have a purity of ≥95%.
1
described for compound 11. White solid, 22.2 mg, yield 90%. H
NMR (CDCl3, 600 MHz) δ 8.13 (d, J = 8.3 Hz, 1H), 7.46 (d, J = 7.4
Hz, 1H), 7.38−7.29 (m, 5H), 7.08 (t, J = 7.9 Hz, 1H), 6.78 (s, 1H),
6.24 (br s, 1H), 5.00 (br s, 1H), 4.75 (d, J = 5.3 Hz, 2H), 4.25 (q, J =
2.3 Hz, 2H), 2.83−2.78 (m, 2H), 2.63 (s, 3H), 2.42−2.38 (m, 2H),
1.95−1.87 (m, 4H), 1.84 (t, J = 2.3 Hz, 3H). HRMS (ESI+) calcd for
C29H30N5O (M + H)+ 464.2445, found 464.2439.
1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-2-
methyl-N-propyl-1H-indole-4-carboxamide (16). Compound 16
was prepared from acid 31a in a fashion similar to the one described
for compound 11. White solid, 13.3 mg, yield 59%. 1H NMR (CDCl3,
600 MHz) δ 8.11 (d, J = 8.3 Hz, 1H), 7.44 (d, J = 7.5 Hz, 1H), 7.38−
7.29 (m, 5H), 7.08 (t, J = 7.8 Hz, 1H), 6.72 (s, 1H), 6.16 (t, J = 5.7
Hz, 1H), 5.02 (t, J = 5.5 Hz, 1H), 4.75 (d, J = 5.5 Hz, 2H), 3.48 (q, J
= 6.7 Hz, 2H), 2.82−2.78 (m, 2H), 2.62 (s, 3H), 2.41−2.37 (m, 2H),
1.94−1.86 (m, 4H), 1.71−1.65 (m, 2H), 1.02 (t, J = 7.4 Hz, 3H).
HRMS (ESI+) calcd for C28H32N5O (M + H)+ 454.2601, found
454.2600.
1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-2-
methyl-N-(prop-2-yn-1-yl)-1H-indole-4-carboxamide (11). To
a solution of acid 31a (25.0 mg, 0.0606 mmol) and 2-propynylamine
(3.7 mg, 0.067 mmol) in DMF (1 mL) was added HBTU (25.3 mg,
0.667 mmol) and triethylamine (36.7 mg, 0.364 mmol). The resulting
mixture was allowed to stir at room temperature (rt) overnight and
1-(4-(Benzylamino)-7,8-dihydro-5H-pyrano[4,3-d]-
pyrimidin-2-yl)-2-methyl-N-(prop-2-yn-1-yl)-1H-indole-4-car-
boxamide (17). Compound 17 was prepared from acid 31b in a
fashion similar to the one described for compound 11. White solid,
14.9 mg, yield 68%. H NMR (CDCl3, 600 MHz) δ 8.18 (d, J = 8.1
Hz, 1H), 7.47 (d, J = 7.3 Hz, 1H), 7.40−7.30 (m, 5H), 7.11 (t, J = 7.8
1
K
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Hz, 1H), 6.82 (s, 1H), 6.29 (br s, 1H), 4.76 (d, J = 5.2 Hz, 2H), 4.61
(br s, 1H), 4.58 (s, 2H), 4.35−4.30 (m, 2H), 4.07 (t, J = 5.4 Hz, 2H),
2.92 (t, J = 5.4 Hz, 2H), 2.65 (s, 3H), 2.29 (t, J = 2.4 Hz, 1H). HRMS
(ESI+) calcd for C27H26N5O2 (M + H)+ 452.2081, found 452.2079.
1-(4-(Benzylamino)-7,8-dihydro-5H-pyrano[4,3-d]-
pyrimidin-2-yl)-N-(but-2-yn-1-yl)-2-methyl-1H-indole-4-car-
boxamide (18). Compound 18 was prepared from acid 31b in a
fashion similar to the one described for compound 11. White solid,
reaction mixture was then diluted with water (30 mL) and extracted
with CH2Cl2 (30 mL × 3). The combined organic layer was dried
over Na2SO4 and filtered. The filtrate was concentrated in vacuo and
the residue was purified by flash column chromatography using
EtOAc/hexanes (40%) to afford compound 23 as a yellow solid (28.8
1
mg, 55%). H NMR (CDCl3, 600 MHz) δ 7.81 (br d, J = 5.6 Hz,
2H), 7.44 (br s, 1H), 7.37−7.28 (m, 5H), 7.05 (t, J = 8.1 Hz, 1H),
6.44 (d, J = 16.9 Hz, 1H), 6.32 (dd, J1 = 16.9 Hz, J2 = 10.2 Hz, 1H),
6.28 (s, 1H), 5.75 (d, J = 9.8 Hz, 1H), 5.01 (s, 1H), 4.74 (d, J = 5.5
Hz, 2H), 2.82−2.77 (m, 2H), 2.59 (s, 3H), 2.40−2.35 (m, 2H),
1.93−1.84 (m, 4H). HRMS (ESI+) calcd for C27H28N5O (M + H)+
438.2288, found 438.2296.
1
15.1 mg, yield 67%. H NMR (CDCl3, 600 MHz) δ 8.16 (d, J = 8.3
Hz, 1H), 7.45 (d, J = 7.3 Hz, 1H), 7.38−7.29 (m, 5H), 7.09 (t, J = 7.8
Hz, 1H), 6.81 (s, 1H), 6.25 (br s, 1H), 4.77−4.70 (m, 3H), 4.57 (s,
2H), 4.24 (s, 2H), 4.06 (t, J = 5.4 Hz, 2H), 2.90 (t, J = 5.4 Hz, 2H),
2.63 (s, 3H), 1.84 (s, 3H). HRMS (ESI+) calcd for C28H28N5O2 (M +
H)+ 466.2238, found 466.2234.
N-(1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-2-
methyl-1H-indol-4-yl)propionamide (24). Compound 24 was
prepared from amine 34a in a fashion similar to the one described for
N-(1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-2-
methyl-1H-indol-4-yl)propiolamide (19). To a solution of amine
34a (18 mg, 0.047 mmol) and propiolic acid (9.9 mg, 0.14 mmol) in
CH2Cl2 (2 mL) was added DCC (29.0 mg, 0.141 mmol), and the
mixture was stirred at rt for 30 min. The suspension was then filtered
and the filtrate was concentrated in vacuo. The residue was purified
by flash column chromatography using EtOAc/hexanes (40%) to
1
compound 20. White solid, 19.1 mg, yield 63%. H NMR (CDCl3,
600 MHz) δ 7.80 (d, J = 8.1 Hz, 1H), 7.75 (d, J = 7.6 Hz, 1H), 7.38−
7.28 (m, 5H), 7.25 (br s, 1H), 7.05 (t, J = 8.0 Hz, 1H), 6.25 (s, 1H),
4.97 (s, 1H), 4.75 (d, J = 5.3 Hz, 2H), 2.84−2.76 (m, 2H), 2.60 (s,
3H), 2.46 (q, J = 7.5 Hz, 2H), 2.42−2.35 (m, 2H), 1.95−1.85 (m,
4H), 1.30 (t, J = 7.5 Hz, 3H). HRMS (ESI+) calcd for C27H30N5O (M
+ H)+ 440.2445, found 440.2443.
1
afford compound 19 as a yellow solid (17.6 mg, 86%). H NMR
(CDCl3, 600 MHz) δ 7.83 (d, J = 8.3 Hz, 1H), 7.70 (d, J = 7.8 Hz,
1H), 7.61 (br s, 1H), 7.39−7.29 (m, 5H), 7.05 (t, J = 8.1 Hz, 1H),
6.29 (s, 1H), 4.98 (t, J = 5.6 Hz, 1H), 4.75 (d, J = 5.6 Hz, 2H), 2.93
(s, 1H), 2.83−2.78 (m, 2H), 2.61 (s, 3H), 2.42−2.37 (m, 2H), 1.95−
1.86 (m, 4H). HRMS (ESI+) calcd for C27H26N5O (M + H)+
436.2132, found 436.2137.
N-(1-(4-(Benzylamino)-7,8-dihydro-5H-pyrano[4,3-d]-
pyrimidin-2-yl)-2-methyl-1H-indol-4-yl)propiolamide (25).
Compound 25 was prepared from amine 34b in a fashion similar to
the one described for compound 19. White solid, 13.6 mg, yield 79%.
1H NMR (CDCl3, 600 MHz) δ 7.85 (d, J = 8.3 Hz, 1H), 7.68 (s, 1H),
7.65 (d, J = 7.7 Hz, 1H), 7.38−7.33 (m, 2H), 7.33−7.28 (m, 3H),
7.05 (t, J = 8.0 Hz, 1H), 6.30 (s, 1H), 4.76−4.69 (m, 3H), 4.55 (s,
2H), 4.08−4.03 (m, 2H), 2.93 (s, 1H), 2.92−2.87 (m, 2H), 2.60 (s,
3H). HRMS (ESI+) calcd for C26H24N5O2 (M + H)+ 438.1925, found
438.1918.
N-(1-(4-(Benzylamino)-7,8-dihydro-5H-pyrano[4,3-d]-
pyrimidin-2-yl)-2-methyl-1H-indol-4-yl)but-2-ynamide (26).
Compound 26 was prepared from amine 34b in a fashion similar to
the one described for compound 20. White solid, 15.0 mg, yield 83%.
1H NMR (CDCl3, 600 MHz) δ 7.83 (d, J = 8.4 Hz, 1H), 7.68 (d, J =
7.7 Hz, 1H), 7.53 (s, 1H), 7.38−7.28 (m, 5H), 7.04 (t, J = 8.0 Hz,
1H), 6.30 (s, 1H), 4.76−4.69 (m, 3H), 4.55 (s, 2H), 4.05 (t, J = 5.2
Hz, 2H), 2.92−2.87 (m, 2H), 2.60 (s, 3H), 2.02 (s, 3H). HRMS
(ESI+) calcd for C27H26N5O2 (M + H)+ 452.2081, found 452.2084.
N-(1-(4-(Benzylamino)-7,8-dihydro-5H-pyrano[4,3-d]-
pyrimidin-2-yl)-2-methyl-1H-indol-4-yl)acrylamide (27). Com-
pound 27 was prepared from amine 34b in a fashion similar to the
one described for compound 23. White solid, 17.3 mg, yield 82%. 1H
NMR (CDCl3, 600 MHz) δ 7.84 (d, J = 7.0 Hz, 1H), 7.78 (d, J = 6.5
Hz, 1H), 7.42 (br s, 1H), 7.38−7.28 (m, 5H), 7.06 (t, J = 8.1 Hz,
1H), 6.45 (d, J = 16.9 Hz, 1H), 6.37−6.30 (m, 1H), 6.29 (s, 1H),
5.76 (d, J = 9.7 Hz, 1H), 4.71 (s, 3H), 4.55 (s, 2H), 4.05 (t, J = 5.5
Hz, 2H), 2.94−2.93 (m, 2H), 2.60 (s, 3H). HRMS (ESI+) calcd for
C26H26N5O2 (M + H)+ 440.2081, found 440.2082.
Methyl 1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-
yl)-2-methyl-1H-indole-4-carboxylate (30a). To a solution of
chloride 28a26 (323 mg, 1.18 mmol) and indole 2926 (223 mg, 1.18
mmol) in dioxane (8 mL) were added Pd2(dba)3 (108 mg, 0.118
mmol), XPhos (112 mg, 0.236 mmol), and Cs2CO3 (768 mg, 2.36
mmol) under Ar atmosphere. The resulting mixture was stirred at 100
°C overnight and allowed to cool to rt. The mixture was diluted with
water (30 mL) and extracted with CH2Cl2 (30 mL × 3). The
combined organic layer was dried over Na2SO4 and filtered. The
filtrate was concentrated in vacuo and the residue was purified by flash
column chromatography using EtOAc/hexanes (25%) to afford
compound 30a as an orange solid (466 mg, 93%). 1H NMR
(CDCl3, 600 MHz) δ 8.20 (d, J = 8.2 Hz, 1H), 7.85 (d, J = 7.6 Hz,
1H), 7.37−7.28 (m, 5H), 7.08 (t, J = 7.9 Hz, 1H), 7.01 (s, 1H), 5.00
(t, J = 5.5 Hz, 1H), 4.74 (d, J = 5.5 Hz, 2H), 3.96 (s, 3H), 2.82−2.78
(m, 2H), 2.40−2.36 (m, 2H), 1.94−1.86 (m, 4H). HRMS (ESI+)
calcd for C26H27N4O2 (M + H)+ 427.2129, found 427.2137.
N-(1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-2-
methyl-1H-indol-4-yl)but-2-ynamide (20). To a solution of
amine 34a (27 mg, 0.070 mmol) and 2-butynoic acid (8.9 mg, 0.11
mmol) in DMF (1 mL) were added EDC hydrochloride (27.0 mg,
0.141 mmol) and HOBt (9.5 mg, 0.0704 mmol), and the mixture was
stirred at rt overnight. The reaction mixture was then diluted with
water (30 mL) and extracted with CH2Cl2 (30 mL × 3). The
combined organic layer was dried over Na2SO4, and filtered. The
filtrate was concentrated in vacuo and the residue was purified by flash
column chromatography using EtOAc/hexanes (40%) to afford
1
compound 20 as a white solid (24.5 mg, 77%). H NMR (CDCl3,
600 MHz) δ 7.80 (d, J = 8.3 Hz, 1H), 7.72 (d, J = 7.9 Hz, 1H), 7.52
(br s, 1H), 7.37−7.28 (m, 5H), 7.03 (t, J = 8.1 Hz, 1H), 6.29 (s, 1H),
5.01 (t, J = 5.6 Hz, 1H), 4.74 (d, J = 5.6 Hz, 2H), 2.82−2.77 (m, 2H),
2.60 (s, 3H), 2.40−2.35 (m, 2H), 2.01 (s, 3H), 1.93−1.85 (m, 4H).
HRMS (ESI+) calcd for C28H28N5O (M + H)+ 450.2288, found
450.2283.
N-(1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-2-
methyl-1H-indol-4-yl)pent-2-ynamide (21). Compound 21 was
prepared from amine 34a in a fashion similar to the one described for
1
compound 20. White solid, 15.5 mg, yield 75%. H NMR (CDCl3,
600 MHz) δ 7.79 (d, J = 8.2 Hz, 1H), 7.71 (d, J = 7.6 Hz, 1H), 7.54
(s, 1H), 7.38−7.27 (m, 5H), 7.03 (t, J = 8.0 Hz, 1H), 6.29 (s, 1H),
5.03 (br s, 1H), 4.73 (d, J = 5.1 Hz, 2H), 2.83−2.77 (m, 2H), 2.59 (s,
3H), 2.42−2.34 (m, 4H), 1.94−1.85 (m, 4H), 1.24 (t, J = 7.5 Hz,
3H). HRMS (ESI+) calcd for C29H30N5O (M + H)+ 464.2445, found
464.2436.
N-(1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-2-
methyl-1H-indol-4-yl)hex-2-ynamide (22). Compound 22 was
prepared from amine 34a in a fashion similar to the one described for
1
compound 20. White solid, 17.3 mg, yield 82%. H NMR (CDCl3,
600 MHz) δ 7.79 (d, J = 8.3 Hz, 1H), 7.71 (d, J = 7.6 Hz, 1H), 7.52
(s, 1H), 7.38−7.27 (m, 5H), 7.03 (t, J = 8.0 Hz, 1H), 6.29 (s, 1H),
5.03 (br s, 1H), 4.73 (d, J = 5.2 Hz, 2H), 2.83−2.77 (m, 2H), 2.60 (s,
3H), 2.41−2.36 (m, 2H), 2.34 (t, J = 7.1 Hz, 2H), 1.94−1.84 (m,
4H), 1.69−1.61 (m, 2H), 1.05 (t, J = 7.4 Hz, 3H). HRMS (ESI+)
calcd for C30H32N5O (M + H)+ 478.2601, found 478.2608.
N-(1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-2-
methyl-1H-indol-4-yl)acrylamide (23). To a solution of K2CO3
(33.4 mg, 0.242 mmol) and acryloyl chloride (22 mg, 0.24 mmol) in
acetone (2 mL) and water (0.5 mL) at 0 °C was added amine 34a (46
mg, 0.121 mmol), and the mixture was stirred at 0 °C for 1 h. The
Methyl 1-(4-(Benzylamino)-7,8-dihydro-5H-pyrano[4,3-d]-
pyrimidin-2-yl)-2-methyl-1H-indole-4-carboxylate (30b). Com-
L
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
pound 30b was prepared from intermediates 28b26 and 2926 in a
fashion similar to the one described for compound 30a. White solid,
90 mg, yield 83%. 1H NMR (CDCl3, 600 MHz) δ 8.24 (d, J = 8.2 Hz,
1H), 7.86 (d, J = 7.6 Hz, 1H), 7.37−7.28 (m, 5H), 7.09 (t, J = 7.8 Hz,
1H), 7.03 (s, 1H), 4.73 (d, J = 4.9 Hz, 2H), 4.70 (t, J = 4.9 Hz, 1H),
4.55 (s, 2H), 4.05 (t, J = 5.5 Hz, 2H), 3.95 (s, 3H), 2.89 (t, J = 5.5 Hz,
2H), 2.64 (s, 3H). HRMS (ESI+) calcd for C25H25N4O3 (M + H)+
429.1921, found 429.1923.
1-(4-(Benzylamino)-5,6,7,8-tetrahydroquinazolin-2-yl)-2-
methyl-1H-indole-4-carboxylic acid (31a). To a solution of 30a
(443 mg, 1.04 mmol) in water (7.5 mL), MeOH (7.5 mL), and THF
(22 mL) was added LiOH (249 mg, 10.4 mmol). The resulting
mixture was stirred at 80 °C for 7 h. After cooling to rt, the reaction
mixture was diluted with water (30 mL), acidified with 1 N HCl to
pH = 2, and then extracted with CH2Cl2 (30 mL × 3). The combined
organic layer was dried over Na2SO4 and filtered. The filtrate was
concentrated in vacuo and the residue was purified by flash column
chromatography using MeOH/CH2Cl2 (5%) to afford compound 31a
as a white solid (288 mg, 84%). 1H NMR (CDCl3, 600 MHz) δ 8.24
(d, J = 8.2 Hz, 1H), 7.96 (d, J = 7.6 Hz, 1H), 7.38−7.29 (m, 5H),
7.13−7.09 (m, 2H), 4.99 (t, J = 5.5 Hz, 1H), 4.75 (d, J = 5.5 Hz, 2H),
2.84−2.80 (m, 2H), 2.66 (s, 3H), 2.41−2.37 (m, 2H), 1.95−1.87 (m,
4H). HRMS (ESI+) calcd for C25H25N4O2 (M + H)+ 413.1972, found
413.1979.
2-(4-Amino-2-methyl-1H-indol-1-yl)-N-benzyl-7,8-dihydro-
5H-pyrano[4,3-d]pyrimidin-4-amine (34b). A mixture of 33b
(140 mg, 0.337 mmol) and 10% Pd/C (10 mg) in MeOH (28 mL)
was stirred under H2 atmosphere (balloon) at rt for 2 h. The reaction
mixture was filtered, and the filtrate was concentrated in vacuo. The
residue was purified by flash column chromatography using EtOAc/
hexanes (50%) to afford compound 34b as a white solid (80 mg,
61%). 1H NMR (CDCl3, 600 MHz) δ 7.54 (d, J = 8.4 Hz, 1H), 7.34−
7.31 (m, 2H), 7.30−7.325 (m, 3H), 6.90 (t, J = 7.9 Hz, 1H), 6.40 (d,
J = 7.4 Hz, 1H), 6.23 (s, 1H), 4.70 (d, J = 5.4 Hz, 2H), 4.57 (t, J = 5.4
Hz, 1H), 4.47 (s, 2H), 4.03−3.98 (m, 2H), 3.76 (br s, 2H), 2.88−
2.83 (m, 2H), 2.57 (m, 3H). HRMS (ESI+) calcd for C23H24N5O (M
+ H)+ 386.1975, found 386.1968.
Purification of p97 Protein. His-tagged p97 was purified using
Ni-NTA beads (QIAGEN). Protein eluted from the Ni column was
further purified by size exclusion chromatography on a Superdex 200
HR (10/30) column in 50 mM Tris−HCl, pH 8.0, 150 mM
potassium chloride, 5% glycerol, 2 mM magnesium chloride, and 1
mM DTT. Purified protein was stored at −80 °C in 50 mM Tris/
HCl, pH 8.0, 150 mM potassium chloride, 2 mM magnesium
chloride, and 5% glycerol.
Expression and purification of p97 C522A was performed as
described previously.34
p97 Biochemical Assay. Reactions were carried out in duplicate
in 96-well half-area UV-transparent microplates (Corning) at a final
volume of 100 μL. Assays were initiated by adding 0.028 mg/mL
(approximately 300 nM) p97 enzyme to the reaction buffer (50 mM
Tris−HCl pH 8.0, 150 mM KCl, 1 mM MgCl2, 15% glycerol)
containing 0.2 mM ATP, 1 mM DTT, 2 U/mL PNP (Sigma-Aldrich),
and 0.2 mM MesG (Berry & Associates) that contained inhibitors
(0−11.1 μM). The reactions were monitored continuously at 360 nm
on an i3 multimode plate reader (Molecular Devices), and the initial
velocities were fit to the Morrison equation using Prism 5
(GraphPad).
1-(4-(Benzylamino)-7,8-dihydro-5H-pyrano[4,3-d]-
pyrimidin-2-yl)-2-methyl-1H-indole-4-carboxylic acid (31b).
Compound 31b was prepared from ester 30b in a fashion similar to
the one described for compound 31a. White solid, 87 mg, yield 100%.
1H NMR (CD3OD, 600 MHz) δ 7.86 (d, J = 7.6 Hz, 1H), 7.68 (d, J =
8.1 Hz, 1H), 7.38−7.26 (m, 5H), 7.10 (t, J = 8.0 Hz, 1H), 7.08 (s,
1H), 4.78 (s, 2H), 4.67 (s, 2H), 4.08 (t, J = 5.0 Hz, 2H), 2.89 (t, J =
5.0 Hz, 2H), 2.44 (s, 3H). HRMS (ESI+) calcd for C24H23N4O3 (M +
H)+ 415.1765, found 415.1757.
N-Benzyl-2-(2-methyl-4-nitro-1H-indol-1-yl)-5,6,7,8-tetrahy-
droquinazolin-4-amine (33a). Compound 33a was prepared from
intermediates 28a26 and 3244 in a fashion similar to the one described
for compound 30a. Yellow solid, 265 mg, yield 97%. 1H NMR
(CDCl3, 600 MHz) δ 8.22 (d, J = 8.2 Hz, 1H), 8.07 (d, J = 8.2 Hz,
1H), 7.38−7.35 (m, 2H), 7.34−7.29 (m, 3H), 7.12 (s, 1H), 7.07 (t, J
= 8.2 Hz, 1H), 5.11 (t, J = 5.6 Hz, 1H), 4.74 (d, J = 5.6 Hz, 2H),
2.83−2.78 (m, 2H), 2.65 (s, 3H), 2.43−2.39 (m, 2H), 1.96−1.87 (m,
4H). HRMS (ESI+) calcd for C24H24N5O2 (M + H)+ 414.1925, found
414.1924.
N-Benzyl-2-(2-methyl-4-nitro-1H-indol-1-yl)-7,8-dihydro-
5H-pyrano[4,3-d]pyrimidin-4-amine (33b). Compound 33b was
prepared from intermediates 28b26 and 3244 in a fashion similar to
the one described for compound 30a. White solid, 90 mg, yield 83%.
1H NMR (CDCl3, 600 MHz) δ 8.26 (d, J = 8.2 Hz, 1H), 8.09 (d, J =
GSH Reactivity Assay. To assess reactivity with GSH, the test
compounds (1 μM) were incubated at 37 °C with 5 mM GSH and a
stable internal standard in 0.1 M potassium phosphate buffer (pH 7.4,
10% acetonitrile) in a total volume of 1 mL. The reactions were
initiated upon addition of GSH, and the samples were analyzed by
directly injected the mixture into LC−MS/MS immediately after the
GSH addition and then at predetermined intervals within a period of
24 h. Control reactions were run in the absence of GSH. The
disappearance of parent compounds was monitored as % remaining
relative to time zero, and the data were fitted to first-order kinetics by
plotting the natural log of % remaining as a function of time. The
pseudo-first-order rate constant (k), which is the negative slope of the
linear fitting, was used to calculate the reaction half-life (t1/2 = 0.693/
k). The formation of GSH conjugates was determined at the end of
incubation to confirm the reactions of test compounds with GSH.
Antiproliferation Assay in MIA PaCa-2 Cells. The assay was
performed as described previously.58
Reversibility under the Biochemical Assay Conditions. The
p97 enzyme was diluted to 0.028 μg/μL in assay buffer without
glycerol (50 mM Tris, pH 8.0, 150 mM KCl, 1 mM MgCl2, and 1 mM
DTT). Reactions contained 750 nM compound 20, 1610 nM
compound 4b, 3200 nM 3 (10 × IC50), or DMSO (1% of final
volume). The reactions were incubated at rt for 20 min to allow for
binding of the inhibitors to the enzyme. After the initial incubation,
the reactions were moved to hydrated D-Tube Dialyzer Mini 6−8
kDa MWCO cassettes (Millipore) and placed in separate beakers
each containing 500 mL of the above assay buffer (4000 × dilution).
The beakers were placed on separate stirring plates and stirred
overnight (16 h) at 4 °C. After the initial dialysis, the cassettes were
placed in 500 mL of fresh buffer and dialyzed as above for an
additional 6 h.
8.2 Hz, 1H), 7.41−7.36 (m, 2H), 7.35−7.31 (m, 3H), 7.16 (s, 1H),
7.11 (t, J = 8.2 Hz, 1H), 4.76 (d, J = 5.1 Hz, 2H), 4.70 (t, J = 5.1 Hz,
1H), 4.60 (s, 2H), 4.10−4.07 (m, 2H), 2.94−2.90 (m, 2H), 2.68 (m,
3H). HRMS (ESI+) calcd for C23H22N5O3 (M + H)+ 416.1717, found
416.1715.
2-(4-Amino-2-methyl-1H-indol-1-yl)-N-benzyl-5,6,7,8-tetra-
hydroquinazolin-4-amine (34a). To a solution of 33a (220 mg,
0.533 mmol) in MeOH (10 mL) was added SnCl2 (705 mg, 3.73
mmol). The resulting mixture was stirred at 75 °C overnight. After
cooling to rt, the reaction was quenched with saturated Na2CO3 (30
mL). The resulting suspension was filtered and the filtrate was
extracted with CH2Cl2 (30 mL × 3). The combined organic layer was
dried over Na2SO4 and filtered. The filtrate was concentrated in vacuo
and the residue was purified by flash column chromatography using
MeOH/CH2Cl2 (3%) to afford compound 34a as a yellow solid (109
1
mg, 53%). H NMR (CDCl3, 600 MHz) δ 7.51 (d, J = 8.3 Hz, 1H),
7.36−7.32 (m, 4H), 7.30−7.27 (m, 1H), 6.90 (t, J = 8.0 Hz, 1H),
6.40 (d, J = 7.4 Hz, 1H), 6.24 (s, 1H), 4.90 (t, J = 5.5 Hz, 1H), 4.74
(d, J = 5.5 Hz, 2H), 3.76 (br s, 2H), 2.81−2.76 (m, 2H), 2.59 (s, 3H),
2.36−2.32 (m, 2H), 1.91−1.83 (m, 4H). HRMS (ESI+) calcd for
C24H26N5 (M + H)+ 384.2183, found 384.2182.
The enzymatic activity after dialysis was determined based on the
above p97 biochemical assay. Assays were initiated by adding 50 μL of
the overnight reactions to 50 μL of assay buffer (50 mM Tris, pH 8.0,
150 mM KCl, 2 mM MgCl2, 15% glycerol (w/v), 0.2 mM ATP, 1 mM
DTT, 2 U/mL PNP, and 0.2 mM MesG, all given as final
M
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Kiapp values at increasing ATP concentrations; active site
(within D2 domain) Glide XP docking scores; CovDock
docking scores; degradation of ΔSS TCRα-YFP in the
presence of proteasome or p97 inhibitors; examples of
tandem mass spectra of peptides that were covalently
modified by compound 20 at Cys105, 522, and 535;
three cysteine residues (Cys105, 522, and 535) in
human p97 (PDB 5FTK, subunit E) as potential targets
of compound 20; 2D representation of potential
interactions between compound 4b and p97; proposed
binding mode and interactions of compound 4b docked
into human p97 (PDB 5FTK) using subunit E as a
template; 2D representation of potential interactions
between compound 19 and p97; 2D representation of
potential interactions between compound 22 and p97;
proposed binding modes of S1−3 docked (XP mode)
into human p97 (PDB 5FTK) at the active site of the
D2 domain; proposed binding modes of S1−3
covalently docked into human p97 (PDB 5FTK) at
the active site of the D2 domain; proposed binding
modes and interactions of compounds 11 and 15 docked
into human p97 (PDB 5FTK) at the active site of the
D2 domain (PDF)
concentrations in 100 μL total). Additionally, compounds 20 (750
nM), 4b (1610 nM), and 3 (3200 nM) were added to wells with the
overnight DMSO reactions as negative controls. All wells were mixed
by pipetting and read for 2 h in an i3 multimode microplate reader in
continuous absorbance mode set to 360 nm.
Inhibition and Recovery of p97 Activity Ub-v-GFP Cells. Ub-
v-GFP cells (3 × 105, HEK293T-based) were seeded in 12-well poly-
lysine D-coated plates and allowed to grow overnight. p97 inhibitor (2
μM) was added to each well. Before harvesting, the control group was
incubated at 37 °C for 4 h (scheme A); for the recovery group,
medium containing the drug was removed carefully after 1 h
incubation. The cells were washed once with PBS and incubated in
drug-free medium for another 3 h. Whole-cell extract prepared in the
Laemmli buffer was subject to standard SDS-PAGE and immunoblot-
ting analysis.
Identification of Covalent Modifications by Compound 20.
Recombinant p97 (10 μg; final concentration, 2 μM) and 10 μM
compound 20 or DMSO were incubated in 50 μL of 1× ATPase
buffer (50 mM Tris/HCl, pH 8.0, 150 mM KCl, 1 mM magnesium
chloride, and 15% glycerol) for 3 h at room temperature. The mixture
was subjected to SDS-PAGE on a 4−12% Bis-Tris gel. After staining
the gel with colloidal Coomassie Blue (Invitrogen), the p97 band was
cut off and analyzed at the Taplin Mass Spectrometry Facility at
Harvard Medical School to identify residues that were covalently
modified by compound 20.
The p97 gel band was digested using a modified in-gel trypsin
digestion procedure,59 and the resulting peptides were separated by
nanoscale HPLC. The eluted peptides were detected (electrospray
ionization), isolated, and further fragmented to give tandem mass
spectra on an LTQ-Orbitrap mass spectrometer (Thermo Fisher
Scientific, San Jose, CA). The raw mass MS/MS images were
converted into mzXML format, processed, and then searched using
the SEQUEST program (version 28) with methionine oxidation
(+15.9949) and covalent modification by compound 20 (+449.2216,
conjugate addition the alkyne) being dynamically allowed. Sites of
cysteine modification were assigned using ModScore, a variation of
the Ascore algorithm.52
Molecular Modeling. The docking study was carried out using
the Schrodinger modeling package (version 2015-4).60 The solved
cryo-EM structure of human p97 (PDB 5FTK)30 was taken from the
Protein Data Bank. The protein structure was first processed using
protein preparation wizard in Maestro. During the preparation,
missing hydrogens atoms were added and the missing residues were
added to the cryo-EM structure using Prime. After the het states were
generated, the state of the lowest energy was selected and the energy
minimization was performed using OPLS 2005 force field. The
receptor grid, which encompassed the active site of the D2 domain at
the interface of subunits D and E, was then generated around the
endogenous ligand ADP using the Receptor Grid Generation
application. The p97 inhibitors were drawn in Maestro and processed
using the LigPrep application to generate conformation and ionization
states within pH 7 2 (Epik). The p97 ligands generated by LigPrep
were docked into the receptor grid in the standard Glide XP mode.
Post-docking minimization was performed and up to three best poses
per ligand were written out. All poses were examined manually, and
the pose of the best docking score was retained unless noted
otherwise. Covalent docking was performed using the CovDock57
application with Cys522 as the nucleophilic residue, which underwent
a conjugate addition to an alkyne (carbonyl-activated) as preset by
CovDock. Structural visualization and representation were performed
with PyMOL.61
Molecular formula strings (CSV)
AUTHOR INFORMATION
Corresponding Author
*E-mail: chenx462@umn.edu. Phone: 612-624-2575. Fax:
612-624-8154.
ORCID
■
Liqiang Chen: 0000-0002-4229-863X
Author Contributions
§R.D. and T.Z. contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by the Center for Drug Design in
the Academic Health Center of the University of Minnesota
and by the intramural research program of the NIDDK at NIH.
The authors thank Eli Chapman (Department of Pharmacol-
ogy and Toxicology, College of Pharmacy, The University of
Arizona) for his generous gift of p97-C522A-pET14b
construct. They thank Di Xia (NCI) for critical reading of
the manuscript. The University of Minnesota Supercomputing
Institute provided all of the necessary computational resources.
The authors also thank Ross Tomaino of the Taplin Mass
Spectrometry Facility at Harvard Medical School for his expert
help with the MS experiments.
ABBREVIATIONS
■
ALS, amyotrophic lateral sclerosis; DBeQ, N2,N4-dibenzylqui-
nazoline-2,4-diamine; DCC, N,N′-dicyclohexylcarbodiimide;
D T T , d i t h i o t h r e i t o l ; E D C , N - e t h y l - N ′ - ( 3 -
dimethylaminopropyl)carbodiimide; ERAD, endoplasmic re-
ticulum-associated protein degradation; GSH, glutathione;
HBTU, N,N,N′,N′-tetramethyl-O-(1H-benzotriazol-1-yl)-
uronium hexafluorophosphate; HOBt, 1-hydroxybenzotriazole;
IBMPFD, inclusion body myopathy, Paget’s disease of bone,
and frontotemporal dementia; MesG, 7-methyl-6-thioguano-
sine; NFC, nitrofuran-containing; PNP, purine nucleoside
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jmed-
chem.9b00144.
■
S
DTT reactivity of selected compounds; list of p97
peptides that were covalently modified by compound 20
N
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
membrane proteins with a new small molecule inhibitor, eeyarestatin.
Mol. Biol. Cell 2004, 15, 1635−1646.
(22) Wang, Q.; Shinkre, B. A.; Lee, J. G.; Weniger, M. A.; Liu, Y.;
Chen, W.; Wiestner, A.; Trenkle, W. C.; Ye, Y. The ERAD inhibitor
Eeyarestatin I is a bifunctional compound with a membrane-binding
domain and a p97/VCP inhibitory group. PLoS One 2010, 5,
No. e15479.
(23) Ding, R.; Zhang, T.; Xie, J.; Williams, J.; Ye, Y.; Chen, L.
Eeyarestatin I derivatives with improved aqueous solubility. Bioorg.
Med. Chem. Lett. 2016, 26, 5177−5181.
(24) Chou, T. F.; Brown, S. J.; Minond, D.; Nordin, B. E.; Li, K.;
Jones, A. C.; Chase, P.; Porubsky, P. R.; Stoltz, B. M.; Schoenen, F. J.;
Patricelli, M. P.; Hodder, P.; Rosen, H.; Deshaies, R. J. Reversible
inhibitor of p97, DBeQ, impairs both ubiquitin-dependent and
autophagic protein clearance pathways. Proc. Natl. Acad. Sci. U.S.A.
2011, 108, 4834−4839.
phosphorylase; SAR, structure−activity relationship; TCR, T-
cell receptor
REFERENCES
■
(1) Nakayama, S.; Atsumi, R.; Takakusa, H.; Kobayashi, Y.;
Kurihara, A.; Nagai, Y.; Nakai, D.; Okazaki, O. A zone classification
system for risk assessment of idiosyncratic drug toxicity using daily
dose and covalent binding. Drug Metab. Dispos. 2009, 37, 1970−1977.
(2) Singh, J.; Petter, R. C.; Baillie, T. A.; Whitty, A. The resurgence
of covalent drugs. Nat. Rev. Drug Discovery 2011, 10, 307−317.
(3) Bauer, R. A. Covalent inhibitors in drug discovery: from
accidental discoveries to avoided liabilities and designed therapies.
Drug Discovery Today 2015, 20, 1061−1073.
(4) Jackson, P. A.; Widen, J. C.; Harki, D. A.; Brummond, K. M.
Covalent modifiers: a chemical perspective on the reactivity of
alpha,beta-unsaturated carbonyls with thiols via hetero-Michael
addition reactions. J. Med. Chem. 2017, 60, 839−885.
(5) Ye, Y.; Tang, W. K.; Zhang, T.; Xia, D. A mighty “protein
extractor” of the cell: structure and function of the p97/CDC48
ATPase. Front. Mol. Biosci. 2017, 4, 39.
(25) Chou, T. F.; Li, K.; Frankowski, K. J.; Schoenen, F. J.; Deshaies,
R. J. Structure-activity relationship study reveals ML240 and ML241
as potent and selective inhibitors of p97 ATPase. ChemMedChem
2013, 8, 297−312.
(26) Zhou, H. J.; Wang, J.; Yao, B.; Wong, S.; Djakovic, S.; Kumar,
B.; Rice, J.; Valle, E.; Soriano, F.; Menon, M. K.; Madriaga, A.; Kiss
von Soly, S.; Kumar, A.; Parlati, F.; Yakes, F. M.; Shawver, L.; Le
Moigne, R.; Anderson, D. J.; Rolfe, M.; Wustrow, D. Discovery of a
first-in-class, potent, selective, and orally bioavailable inhibitor of the
p97 AAA ATPase (CB-5083). J. Med. Chem. 2015, 58, 9480−9497.
(27) Cervi, G.; Magnaghi, P.; Asa, D.; Avanzi, N.; Badari, A.; Borghi,
D.; Caruso, M.; Cirla, A.; Cozzi, L.; Felder, E.; Galvani, A.; Gasparri,
F.; Lomolino, A.; Magnuson, S.; Malgesini, B.; Motto, I.; Pasi, M.;
Rizzi, S.; Salom, B.; Sorrentino, G.; Troiani, S.; Valsasina, B.; O’Brien,
T.; Isacchi, A.; Donati, D.; D’Alessio, R. Discovery of 2-
(cyclohexylmethylamino)pyrimidines as a new class of reversible
valosine containing protein inhibitors. J. Med. Chem. 2014, 57,
10443−10454.
(28) Magnaghi, P.; D’Alessio, R.; Valsasina, B.; Avanzi, N.; Rizzi, S.;
Asa, D.; Gasparri, F.; Cozzi, L.; Cucchi, U.; Orrenius, C.; Polucci, P.;
Ballinari, D.; Perrera, C.; Leone, A.; Cervi, G.; Casale, E.; Xiao, Y.;
Wong, C.; Anderson, D. J.; Galvani, A.; Donati, D.; O’Brien, T.;
Jackson, P. K.; Isacchi, A. Covalent and allosteric inhibitors of the
ATPase VCP/p97 induce cancer cell death. Nat. Chem. Biol. 2013, 9,
548−556.
(29) Polucci, P.; Magnaghi, P.; Angiolini, M.; Asa, D.; Avanzi, N.;
Badari, A.; Bertrand, J.; Casale, E.; Cauteruccio, S.; Cirla, A.; Cozzi,
L.; Galvani, A.; Jackson, P. K.; Liu, Y.; Magnuson, S.; Malgesini, B.;
Nuvoloni, S.; Orrenius, C.; Sirtori, F. R.; Riceputi, L.; Rizzi, S.;
Trucchi, B.; O’Brien, T.; Isacchi, A.; Donati, D.; D’Alessio, R.
Alkylsulfanyl-1,2,4-triazoles, a new class of allosteric valosine
containing protein inhibitors. Synthesis and structure-activity relation-
ships. J. Med. Chem. 2013, 56, 437−450.
(30) Banerjee, S.; Bartesaghi, A.; Merk, A.; Rao, P.; Bulfer, S. L.; Yan,
Y.; Green, N.; Mroczkowski, B.; Neitz, R. J.; Wipf, P.; Falconieri, V.;
Deshaies, R. J.; Milne, J. L.; Huryn, D.; Arkin, M.; Subramaniam, S.
2.3 A resolution cryo-EM structure of human p97 and mechanism of
allosteric inhibition. Science 2016, 351, 871−875.
(31) Tao, S.; Tillotson, J.; Wijeratne, E. M.; Xu, Y. M.; Kang, M.;
Wu, T.; Lau, E. C.; Mesa, C.; Mason, D. J.; Brown, R. V.; La Clair, J.
J.; Gunatilaka, A. A.; Zhang, D. D.; Chapman, E. Withaferin A analogs
that target the AAA + chaperone p97. ACS Chem. Biol. 2015, 10,
1916−1924.
(32) Alverez, C.; Bulfer, S. L.; Chakrasali, R.; Chimenti, M. S.;
Deshaies, R. J.; Green, N.; Kelly, M.; LaPorte, M. G.; Lewis, T. S.;
Liang, M.; Moore, W. J.; Neitz, R. J.; Peshkov, V. A.; Walters, M. A.;
Zhang, F.; Arkin, M. R.; Wipf, P.; Huryn, D. M. Allosteric indole
amide inhibitors of p97: identification of a novel probe of the
ubiquitin pathway. ACS Med. Chem. Lett. 2016, 7, 182−187.
(33) Alverez, C.; Arkin, M. R.; Bulfer, S. L.; Colombo, R.; Kovaliov,
M.; LaPorte, M. G.; Lim, C.; Liang, M.; Moore, W. J.; Neitz, R. J.;
Yan, Y.; Yue, Z.; Huryn, D. M.; Wipf, P. Structure-activity study of
(6) Xia, D.; Tang, W. K.; Ye, Y. Structure and function of the AAA +
ATPase p97/Cdc48p. Gene 2016, 583, 64−77.
(7) Tang, W. K.; Li, D.; Li, C. C.; Esser, L.; Dai, R.; Guo, L.; Xia, D.
A novel ATP-dependent conformation in p97 N-D1 fragment
revealed by crystal structures of disease-related mutants. EMBO J.
2010, 29, 2217−2229.
(8) Tang, W. K.; Xia, D. Altered intersubunit communication is the
molecular basis for functional defects of pathogenic p97 mutants. J.
Biol. Chem. 2013, 288, 36624−36635.
(9) Buchberger, A.; Schindelin, H.; Hanzelmann, P. Control of p97
function by cofactor binding. FEBS Lett. 2015, 589, 2578−2589.
(10) Zhang, T.; Ye, Y. The final moments of misfolded proteins en
route to the proteasome. DNA Cell Biol. 2014, 33, 477−483.
(11) Xu, S.; Peng, G.; Wang, Y.; Fang, S.; Karbowski, M. The AAA-
ATPase p97 is essential for outer mitochondrial membrane protein
turnover. Mol. Biol. Cell 2011, 22, 291−300.
(12) Meyer, H.; Bug, M.; Bremer, S. Emerging functions of the
VCP/p97 AAA-ATPase in the ubiquitin system. Nat. Cell Biol. 2012,
14, 117−123.
(13) Bug, M.; Meyer, H. Expanding into new markets–VCP/p97 in
endocytosis and autophagy. J. Struct. Biol. 2012, 179, 78−82.
(14) Dantuma, N. P.; Acs, K.; Luijsterburg, M. S. Should I stay or
should I go: VCP/p97-mediated chromatin extraction in the DNA
damage response. Exp. Cell Res. 2014, 329, 9−17.
(15) Fessart, D.; Marza, E.; Taouji, S.; Delom, F.; Chevet, E. P97/
CDC-48: proteostasis control in tumor cell biology. Cancer Lett. 2013,
337, 26−34.
(16) Chapman, E.; Maksim, N.; de la Cruz, F.; La Clair, J. J.
Inhibitors of the AAA + chaperone p97. Molecules 2015, 20, 3027−
3049.
(17) Anderson, D. J.; Le Moigne, R.; Djakovic, S.; Kumar, B.; Rice,
J.; Wong, S.; Wang, J.; Yao, B.; Valle, E.; Kiss von Soly, S.; Madriaga,
A.; Soriano, F.; Menon, M. K.; Wu, Z. Y.; Kampmann, M.; Chen, Y.;
Weissman, J. S.; Aftab, B. T.; Yakes, F. M.; Shawver, L.; Zhou, H. J.;
Wustrow, D.; Rolfe, M. Targeting the AAA ATPase p97 as an
approach to treat cancer through disruption of protein homeostasis.
Cancer Cell 2015, 28, 653−665.
(18) Tang, W. K.; Xia, D. Mutations in the human AAA + chaperone
p97 and related diseases. Front. Mol. Biosci. 2016, 3, 79.
(19) Zhang, T.; Mishra, P.; Hay, B. A.; Chan, D.; Guo, M. Valosin-
containing protein (VCP/p97) inhibitors relieve Mitofusin-dependent
mitochondrial defects due to VCP disease mutants. Elife 2017, 6,
No. e17834.
(20) Blythe, E. E.; Olson, K. C.; Chau, V.; Deshaies, R. J. Ubiquitin-
and ATP-dependent unfoldase activity of P97/VCP*NPLO-
C4*UFD1L is enhanced by a mutation that causes multisystem
proteinopathy. Proc. Natl. Acad. Sci. U.S.A. 2017, 114, E4380−E4388.
(21) Fiebiger, E.; Hirsch, C.; Vyas, J. M.; Gordon, E.; Ploegh, H. L.;
Tortorella, D. Dissection of the dislocation pathway for type I
O
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
bioisosteric trifluoromethyl and pentafluorosulfanyl indole inhibitors
of the AAA ATPase p97. ACS Med. Chem. Lett. 2015, 6, 1225−1230.
(34) Tillotson, J.; Bashyal, B. P.; Kang, M.; Shi, T.; De La Cruz, F.;
Gunatilaka, A. A.; Chapman, E. Selective inhibition of p97 by
chlorinated analogues of dehydrocurvularin. Org. Biomol. Chem. 2016,
14, 5918−5921.
(35) Wijeratne, E. M.; Gunaherath, G. M.; Chapla, V. M.; Tillotson,
J.; de la Cruz, F.; Kang, M.; U’Ren, J. M.; Araujo, A. R.; Arnold, A. E.;
Chapman, E.; Gunatilaka, A. A. Oxaspirol B with p97 inhibitory
activity and other oxaspirols from Lecythophora sp. FL1375 and
FL1031, endolichenic fungi inhabiting Parmotrema tinctorum and
Cladonia evansii. J. Nat. Prod. 2016, 79, 340−352.
(36) Wang, Q.; Li, L.; Ye, Y. Inhibition of p97-dependent protein
degradation by Eeyarestatin I. J. Biol. Chem. 2008, 283, 7445−7454.
(37) Schwartz, P. A.; Kuzmic, P.; Solowiej, J.; Bergqvist, S.; Bolanos,
B.; Almaden, C.; Nagata, A.; Ryan, K.; Feng, J.; Dalvie, D.; Kath, J. C.;
Xu, M.; Wani, R.; Murray, B. W. Covalent EGFR inhibitor analysis
reveals importance of reversible interactions to potency and
mechanisms of drug resistance. Proc. Natl. Acad. Sci. U.S.A. 2014,
111, 173−178.
(49) Nishikori, S.; Esaki, M.; Yamanaka, K.; Sugimoto, S.; Ogura, T.
Positive cooperativity of the p97 AAA ATPase is critical for essential
functions. J. Biol. Chem. 2011, 286, 15815−15820.
̈
(50) Hanzelmann, P.; Schindelin, H. Structural basis of ATP
hydrolysis and intersubunit signaling in the AAA + ATPase p97.
Structure 2016, 24, 127−139.
(51) Wiertz, E. J.; Jones, T. R.; Sun, L.; Bogyo, M.; Geuze, H. J.;
Ploegh, H. L. The human cytomegalovirus US11 gene product
dislocates MHC class I heavy chains from the endoplasmic reticulum
to the cytosol. Cell 1996, 84, 769−779.
(52) Beausoleil, S. A.; Villen, J.; Gerber, S. A.; Rush, J.; Gygi, S. P. A
probability-based approach for high-throughput protein phosphor-
ylation analysis and site localization. Nat. Biotechnol. 2006, 24, 1285−
1292.
(53) Ye, Y.; Meyer, H. H.; Rapoport, T. A. Function of the p97-
Ufd1-Npl4 complex in retrotranslocation from the ER to the cytosol:
dual recognition of nonubiquitinated polypeptide segments and
polyubiquitin chains. J. Cell Biol. 2003, 162, 71−84.
(54) Noguchi, M.; Takata, T.; Kimura, Y.; Manno, A.; Murakami, K.;
Koike, M.; Ohizumi, H.; Hori, S.; Kakizuka, A. ATPase activity of
p97/valosin-containing protein is regulated by oxidative modification
of the evolutionally conserved cysteine 522 residue in Walker A motif.
J. Biol. Chem. 2005, 280, 41332−41341.
(55) Briggs, L. C.; Baldwin, G. S.; Miyata, N.; Kondo, H.; Zhang, X.;
Freemont, P. S. Analysis of nucleotide binding to P97 reveals the
properties of a tandem AAA hexameric ATPase. J. Biol. Chem. 2008,
283, 13745−13752.
(56) Tang, W. K.; Odzorig, T.; Jin, W.; Xia, D. Structural basis of
p97 inhibition by the site-selective anti-cancer compound CB-5083.
Mol. Pharmacol. 2018, 286−293.
(38) Flanagan, M. E.; Abramite, J. A.; Anderson, D. P.; Aulabaugh,
A.; Dahal, U. P.; Gilbert, A. M.; Li, C.; Montgomery, J.;
Oppenheimer, S. R.; Ryder, T.; Schuff, B. P.; Uccello, D. P.;
Walker, G. S.; Wu, Y.; Brown, M. F.; Chen, J. M.; Hayward, M. M.;
Noe, M. C.; Obach, R. S.; Philippe, L.; Shanmugasundaram, V.;
Shapiro, M. J.; Starr, J.; Stroh, J.; Che, Y. Chemical and computational
methods for the characterization of covalent reactive groups for the
prospective design of irreversible inhibitors. J. Med. Chem. 2014, 57,
10072−10079.
̈
(39) Jost, C.; Nitsche, C.; Scholz, T.; Roux, L.; Klein, C. D.
(57) Zhu, K.; Borrelli, K. W.; Greenwood, J. R.; Day, T.; Abel, R.;
Farid, R. S.; Harder, E. Docking covalent inhibitors: a parameter free
approach to pose prediction and scoring. J. Chem. Inf. Model. 2014,
54, 1932−1940.
Promiscuity and selectivity in covalent enzyme inhibition: a
systematic study of electrophilic fragments. J. Med. Chem. 2014, 57,
7590−7599.
(40) Cee, V. J.; Volak, L. P.; Chen, Y.; Bartberger, M. D.; Tegley, C.;
Arvedson, T.; McCarter, J.; Tasker, A. S.; Fotsch, C. Systematic study
of the glutathione (GSH) reactivity of N-arylacrylamides: 1. effects of
aryl substitution. J. Med. Chem. 2015, 58, 9171−9178.
(58) Ai, T.; Willett, R.; Williams, J.; Ding, R.; Wilson, D. J.; Xie, J.;
Kim, D. H.; Puertollano, R.; Chen, L. N-(1-Benzyl-3,5-dimethyl-1H-
pyrazol-4-yl)benzamides: antiproliferative activity and effects on
mTORC1 and autophagy. ACS Med. Chem. Lett. 2017, 8, 90−95.
(59) Shevchenko, A.; Wilm, M.; Vorm, O.; Mann, M. Mass
spectrometric sequencing of proteins silver-stained polyacrylamide
gels. Anal. Chem. 1996, 68, 850−858.
(41) Sommer, S.; Weikart, N. D.; Linne, U.; Mootz, H. D. Covalent
inhibition of SUMO and ubiquitin-specific cysteine proteases by an in
situ thiol-alkyne addition. Bioorg. Med. Chem. 2013, 21, 2511−2517.
(42) Ekkebus, R.; van Kasteren, S. I.; Kulathu, Y.; Scholten, A.;
Berlin, I.; Geurink, P. P.; de Jong, A.; Goerdayal, S.; Neefjes, J.; Heck,
A. J.; Komander, D.; Ovaa, H. On terminal alkynes that can react with
active-site cysteine nucleophiles in proteases. J. Am. Chem. Soc. 2013,
135, 2867−2870.
̈
̈
(60) Schrodinger Release 2015-4; Schrodinger, LLC: New York, NY,
2015.
(61) The PyMOL Molecular Graphics System, version 1.8;
̈
Schrodinger, LLC.
(43) Oballa, R. M.; Truchon, J. F.; Bayly, C. I.; Chauret, N.; Day, S.;
Crane, S.; Berthelette, C. A generally applicable method for assessing
the electrophilicity and reactivity of diverse nitrile-containing
compounds. Bioorg. Med. Chem. Lett. 2007, 17, 998−1002.
(44) Moskalev, N.; Makosza, M. A novel method of indole ring
system construction: one-pot synthesis of 4-and 6-nitroindole
derivatives via base promoted reaction between 3-nitroaniline and
ketones. Tetrahedron Lett. 1999, 40, 5395−5398.
(45) Webb, M. R. A continuous spectrophotometric assay for
inorganic phosphate and for measuring phosphate release kinetics in
biological systems. Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 4884−4887.
(46) Morrison, J. F. Kinetics of the reversible inhibition of enzyme-
catalysed reactions by tight-binding inhibitors. Biochim. Biophys. Acta,
Enzymol. 1969, 185, 269−286.
(47) Witte, J. F.; Bray, K. E.; Thornburg, C. K.; McClard, R. W.
‘Irreversible’ slow-onset inhibition of orotate phosphoribosyltransfer-
ase by an amidrazone phosphate transition-state mimic. Bioorg. Med.
Chem. Lett. 2006, 16, 6112−6115.
(48) McClerren, A. L.; Endsley, S.; Bowman, J. L.; Andersen, N. H.;
Guan, Z.; Rudolph, J.; Raetz, C. R. A slow, tight-binding inhibitor of
the zinc-dependent deacetylase LpxC of lipid A biosynthesis with
antibiotic activity comparable to ciprofloxacin. Biochemistry 2005, 44,
16574−16583.
P
DOI: 10.1021/acs.jmedchem.9b00144
J. Med. Chem. XXXX, XXX, XXX−XXX