Paralog Specificity Determines Subcellular Distribution, Action Mechanism, and Anticancer Activity of TRAP1 Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.7b00978

Although Hsp90 inhibitors can inhibit multiple tumorigenic pathways in cancer cells, their anticancer activity has been disappointingly modest. However, by forcing Hsp90 inhibitors into the mitochondria with mitochondrial delivery vehicles, they were conv

Full text of DOI:10.1021/acs.jmedchem.7b00978

Subscriber access provided by UNIVERSITY OF ADELAIDE LIBRARIES
Article
Paralog specificity determines subcellular distribution, action
mechanism, and anticancer activity of TRAP1 inhibitors
Hye-Kyung Park, Hanbin Jeong, Eunhwa Ko, Geumwoo Lee, Ji-Eun Lee, Sang Kwang Lee, An-Jung Lee, Jin
Young Im, Sung Hu, seong heon kim, Ji Hoon Lee, Changwook Lee, Soosung Kang, and Byoung Heon Kang
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b00978 • Publication Date (Web): 17 Aug 2017
Downloaded from http://pubs.acs.org on August 17, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Paralog specificity determines subcellular
distribution, action mechanism, and anticancer
activity of TRAP1 inhibitors
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
,§
1,§
2
2
1
Hye-Kyung Park , Hanbin Jeong , Eunhwa Ko , Geumwoo Lee , Ji-Eun Lee , Sang Kwang
2
1
1
1
2
2
Lee , An-Jung Lee , Jin Young Im , Sung Hu , Seong Heon Kim , Ji Hoon Lee , Changwook
1,
2,3,
1,
Lee *, Soosung Kang *, and Byoung Heon Kang *
1
(
2
(
3
Department of Biological Sciences, Ulsan National Institutes of Science and Technology
UNIST), Ulsan, 44919, South Korea
New Drug Development Center, DaeguꢀGyeongbuk Medical Innovation Foundation
DGMIF), Daegu, 41061, South Korea
College of Pharmacy and Graduate School of Pharmaceutical Sciences, Ewha Womans
University, Seoul, 03760, South Korea
§
Authors contributed equally
*
Corresponding authors
KEYWORDS: TRAP1, Hsp90, Mitochondria, Inhibitors, Paralog selectivity, Anticancer
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT: Although Hsp90 inhibitors can inhibit multiple tumorigenic pathways in
cancer cells, their anticancer activity has been disappointingly modest. However, by forcing
Hsp90 inhibitors into the mitochondria with mitochondrial delivery vehicles, they were
converted into potent drugs targeting the mitochondrial Hsp90 paralog TRAP1. Here, to
improve mitochondrial drug accumulation without using the mitochondrial delivery vehicle,
we increased freely available drug concentrations in the cytoplasm by reducing the binding of
the drugs to the abundant cytoplasmic Hsp90. After analyzing Xꢀray cocrystal structures, the
purine ring of the Hsp90 inhibitor 2 (BIIB021) was modified to pyrazolopyrimidine
scaffolds. One pyrazolopyrimidine, 12b (DN401), bound better to TRAP1 than to Hsp90,
inactivated the mitochondrial TRAP1 in vivo, and exhibited potent anticancer activity.
Therefore, the rationale and feasible guidelines for developing 12b can potentially be
exploited to design a potent TRAP1 inhibitor.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
ACS Paragon Plus Environment

Page 3 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
TRAP1 is a mitochondrial Hsp90 paralog that controls cytoprotection, metabolic
1ꢀ3
reprogramming, and drug resistance in cancer cells. TRAP1 possesses a structure similar to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
that of the cytosolic Hsp90, with increased structural similarity to the ATPꢀbinding pockets at
4
ꢀ5
the Nꢀterminal domains (NTDs). Various Hsp90 inhibitors targeting Hsp90 NTDs have
6
ꢀ14
been developed as potential anticancer drugs (Figure 1) , and many of them have exhibited
15ꢀ16
inhibitory activity against TRAP1 in vitro.
However, we have previously reported that
current Hsp90 inhibitors do not inactivate the mitochondrial TRAP1 because of an
4
, 17
insufficient mitochondrial drug accumulation,
which is closely related to their modest
1
6, 18
anticancer activity in vivo.
Because the forced mitochondrial delivery of Hsp90 inhibitors
increases their anticancer activity via a novel mode of action, improving mitochondrial drug
4, 17ꢀ20
accumulation is urgently required for transforming moderate drugs into potent ones.
We hypothesized that an insufficient mitochondrial accumulation of Hsp90 inhibitors
reflects a strong binding to abundant (~2% of total cellular proteins) cytoplasmic Hsp90
21ꢀ23
proteins
before entry into the mitochondria. Indeed, the concentration of freely available
portion of the inhibitor may possibly be too low to reach the mitochondria. Therefore, we
decided to develop inhibitors with strong TRAP1ꢀbinding, but weak Hsp90ꢀbinding, activities
4, 6
based on the structural analyses of Hsp90 inhibitors bound to TRAP1 and Hsp90.
On
comparing the structures of TRAP1 and Hsp90 complexed with inhibitors, we identified that
inhibitorꢀtarget protein interactions have very similar configurations. The pyrimidine of the
6
7
purine ring and bromopiperonyl in 1 (PUꢀH71) or the pyridinyl of 2 (BIIB021) were wellꢀ
occupied by the active sites of both the chaperones, and the imidazole ring of purine was
4
relatively exposed to the solvent (Supporting Information, Figure S1). However, a part of the
active site lid (Leu172–Phe201) structure was disordered (Figure 2A) and its flanking
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
conserved residues, Asn171 and Gly202, had different configurations in TRAP1 compared
with those in Hsp90 (Figure 2B,C). Thus, to generate TRAP1ꢀselective interactions, we
modified the imidazole ring of the purine scaffold, which is relatively close to the flanking
regions of the disordered TRAP1 active site lid in the crystal structures.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
7
8
9
10
11 12 13
14
Figure 1. Representative Hsp90 inhibitors. 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , and 9 .
4
ACS Paragon Plus Environment

Page 5 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Structural comparison between TRAP1 and Hsp90. (A) Ribbon diagram of Hsp90
(cyan; PDB: 2FWZ) and TRAP1 (pink; PDB: 4Z1F) complexed with 1. The dotted line
indicates structurally disordered region. (B) Closeꢀup view. The structural difference between
the interactions of Hsp90 and TRAP1 with 1 is highlighted. (C) Sequence alignment of
TRAP1 and Hsp90. Structurally disordered region in the complex structure of TRAP1ꢀ1 and
its flanking amino acid residues are indicated.
CHEMISTRY
Pyrazolopyrimidin derivatives 12a and 12b were prepared from various benzyl halides in
two steps; a benzylhydrazine formation and a cyclization for pyrazole ring formation. The
benzylhydrazine formation was performed via addition of excess amount of hydrated
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
hydrazine into benzyl bromide solution in MeOH. Pyrazole ring formation was performed via
mixing equivalent amount of 2ꢀaminoꢀ4,6ꢀdichloropyrimidineꢀ5ꢀcarbaldehyde with a
benzylhydrazine under TEA basic condition (Scheme 1).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 1. experimental procedure for the synthesis of 12a-b
a
Reagents and conditions: (a) NH
2
NH
2
·ΗCl, MeOH, 0 to 60
2 2
°C, 3 h; (b) HCl, CH Cl , ꢀ2 °C,
13 h; (c) 2ꢀaminoꢀ4,6ꢀdichloropyrimidineꢀ5ꢀcarbaldehyde, TEA, CH
2
Cl , ꢀ2 C, 13 h
2
°
The synthesis of 16 is illustrated in Scheme 2. Tandem condensation of ethyl
bromopyruvate with dibenzylamine and aminoguanidine bicarbonate provided the known
24
intermediate
Debenzylation via Pd/C catalyzed hydrogenation at hydrogen atmosphere, followed by amide
bond formation with the activated ester 2,5ꢀdioxopyrrolidinꢀ1ꢀyl 2ꢀ(benzo[ ][1,3]dioxolꢀ5ꢀ
yl)acetate gave amide intermediate 15 in moderate yield. Cyclization of 15 under POCl
condition and reflux with extra POCl afforded chlorinated product 16.
3ꢀaminoꢀ6ꢀ((dibenzylamino)methyl)ꢀ1,2,4ꢀtriazinꢀ5(4H)ꢀone
(14).
d
3
3
Scheme 2. experimental procedure for the synthesis of 16
a
Reagents and conditions: (a) dibenzylamine, ethyl acetate, reflux, 2 h; (b) aminoguanidine
bicarbonate, EtOH, reflux, 12 h; (c) Pd/C, H , AcOH, H O, 55 °C, 4 h; (d) 2,5ꢀ
2
2
dioxopyrrolidinꢀ1ꢀyl 2ꢀ(benzo[
CH CN, reflux, 5 h
3 3
d][1,3]dioxolꢀ5ꢀyl)acetate, TEA, CH CN, 50 °C; (e) POCl ,
3
6
ACS Paragon Plus Environment

Page 7 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 20 was prepared from glycine in three steps. Reductive amination of glycine
with benzo[d][1,3]dioxoleꢀ5ꢀcarbaldehyde gave compound 18. Cyclization of 18 with 2ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
aminoꢀ4,6ꢀdichloropyrimidineꢀ5ꢀcarbaldehyde, followed by hydrolysis under aqueous LiOH
condition afforded compound 20 in good yield.
Scheme 3. experimental procedure for the synthesis of 20
a
Reagents and conditions: (a) benzo[
4 2 2
d][1,3]dioxoleꢀ5ꢀcarbaldehyde, NaBH , TEA, CH Cl ,
50 °C, 12 h; (b) 2ꢀaminoꢀ4,6ꢀdichloropyrimidineꢀ5ꢀcarbaldehyde, DBU, DMF, room
2
temperature, 2 h; (c) LiOH, THF/H O, 65 °C, 12 h
RESULTS AND DISCUSSION
Among various purine mimic derivatives, guanine mimics with the piperonyl side chain
derived from 2, such as 12a, 16, and 20, were discovered to have favorable inhibitory activity
toward TRAP1 (Figure 3A). Additional syntheses of pyrazolopyrimidines provided us with
21 (Figure 3B), which exhibited stronger inhibitory activities against TRAP1 (IC50 = 138
nM) but comparable inhibitory activities against Hsp90 (IC50 = 97 nM) compared with those
exhibited by 2 (Figure 3C). Subsequently, the pyridinyl of 21 was replaced with
bromopiperonyl to yield 12b, which further improved the inhibitory activities against TRAP1
but reduced those against Hsp90 (Figure 3C). Compound 12b exhibited highly potent
inhibition of TRAP1 (IC50 = 79 nM) but weak inhibition of Hsp90 (IC50 = 698 nM) compared
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
with that exhibited by the purine scaffold Hsp90 inhibitors 1 and 2 (Table 1). To our
knowledge, 12b is the most selective TRAP1 inhibitor (over Hsp90) identified till date.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Development of the mitochondrial TRAP1 inhibitor 12b. (A) Initially synthesized
2
mimic molecules. (B) Structures of 21 and 12b. (C) Inhibitor binding to TRAP1 and Hsp90.
25
Fluorescence polarization was measured with purified recombinant proteins and 1ꢀFITC3,
as described in the Biological Experiment Section, compared with maximum and minimum
millipolarization (mP) values, and presented as the mean ± SEM of two independent
experiments.
Table 1. Drug IC50 values. Inhibitory activities of the drugs against TRAP1 and Hsp90 in the
fluorescence polarization experiments
Name
IC50 for TRAP1 (ꢁM)
IC50 for Hsp90 (ꢁM)
1
2
0.257± 0.001
0.535 ± 0.001
6.335 ± 0.541
0.079 ± 0.001
0.089 ±0.003
0.052 ±0.001
8.932 ±3.472
0.698 ±0.021
1
2a
1
2b
8
ACS Paragon Plus Environment

Page 9 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
6
0
>10
>10
>10
2
8.434 ± 0.641
0.138 ± 0.001
21
0.097 ±0.006
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
We determined the crystal structures of human TRAP1ꢀNM (residues 60–561) in complex
with 12b and 21. Structural analyses revealed that both pyrazolopyrimidines of 12b and 21
26
established additional πꢀπ stacking interactions with the conserved Phe201 residue in
TRAP1 (Figure 4A) compared with the structures of Hsp90 or TRAP1 complexed with the
4, 6
purine derivatives 1 and 2 (Supporting Information, Figure S2 and S3). Furthermore, 12b
exhibited TRAP1 inhibition that was superior to 1, 2, and 21, reflecting the importance of
additional waterꢀmediated hydrogen bonds between the pyrazole Nꢀ2 and the Asn171 amide
(
Figure 4AꢀC; Supporting Information, Figure S2). Thus, the TRAP1ꢀspecific binding of 12b
was enhanced by establishing additional interactions with Asn171 and Phe201, located on the
flanked residues of the disordered TRAP1 structure.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Comparison of inhibitorꢀbound structures. (A) Determined Xꢀray cocrystal
structures of 21 (left; PDB: 5Y3O) and 12b (right; PDB: 5Y3N) bound to human TRAP1.
Dotted yellow lines and red spheres indicate intermolecular hydrogen bonds and water
molecules, respectively. (B) Structural comparison of TRAP1–1 (left; PDB: 4Z1F) and
TRAP1–12b (right). (C) Structural comparison between the crystal structure of Hsp90–2
(left; PDB: 3O6O) and the computational model structure of Hsp90ꢀ12b (right). Structures
are presented in the same orientation. Hsp90 and TRAP1 are colored cyan and pink,
respectively. Oxygen and nitrogen atoms are colored red and blue, respectively. The
27
computational binding experiment was completed using AutoDock Vina. All structural
figures were prepared using PyMOL (www.pymol.org).
10
ACS Paragon Plus Environment

Page 11 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 12b triggered the signature responses indicative of TRAP1 inactivation
1
8, 28
including the loss of mitochondrial membrane potential,
overproduction of mitochondrial
29ꢀ31
32ꢀ33
ROS,
elevation of CHOP expression,
and subsequent discharge of cytochrome c in
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
8, 28
HeLa cells within 6 hours (Figure 5AꢀD).
In contrast, these responses were not observed
1
1
in the cancer cell (Figure 5AꢀD) after treatment with the Hsp90 inhibitor 2, which most
potently inhibited purified recombinant TRAP1 in vitro (Figure 5E). 12b degraded Hsp90
client proteins less effectively compared with Hsp90 inhibitors and did not induce expression
3
4
of the biomarker (Hsp70) for Hsp90 inhibition during both shortꢀterm (6 hours) and longꢀ
term (36 hours) drug treatment (Figures 5C and 6), likely due to reduced binding affinity
toward Hsp90 (Figure 3C). Hsp70 suppresses cancer cell death, which is closely related to the
moderate anticancer activity of and drugꢀresistance development against the Hsp90
35ꢀ36
inhibitors.
Compound 12b completely avoided Hsp70 induction, the adverse effect of
Hsp90 inhibitors.
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 5. TRAP1 inhibition by 12b. (A) The measurement of mitochondrial membrane
potential. HeLa cells were treated with 20 ꢁM of FCCP, 6, and 12b for 4 hours. The cells
were stained with JCꢀ1 and analyzed using flow cytometry. (B) Mitochondrial ROS
production. HeLa cells were treated with drugs for 6 hours, and the cells were stained with
MitoꢀSox and analyzed using flow cytometry. (C) Expression of Hsp70 and Hsp90 client
proteins. HeLa cells were treated with 6 and 12b at 20 ꢁM for 6 hours. The expression of
CHOP, Chk1, Akt, Hsp70, and βꢀactin was analyzed using Western blotting. (D) Cytochrome
c release assay. HeLa cells were treated with drugs at 20 ꢁM for 4 hours, and the cytoplasmic
fraction was analyzed using western blotting. (E) Inhibition of TRAP1 ATPase activity. Data
are presented as means ± SEM from experiments performed in triplicate.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Analysis of cytoplasmic Hsp90 inhibition. (A) Short exposure to inhibitors.
Cytoplasmic Hsp90 inhibition was assessed by analyzing Hsp70 induction and client protein
degradation. Prostate and cervical human cancer cell lines, PC3 and HeLa, were treated with
drugs (10 ꢁM) for 6 hours as indicated and were analyzed by western blotting. (B) Long
exposure to inhibitors. HeLa cells were incubated with 6 and 12b for 36 hours as indicated
and were analyzed by western blotting. Gamitrinib (22) is a mitochondriaꢀtargeted TRAP1
17
inhibitor. 2, 5, and 6 are representative Hsp90 inhibitors with purine, geldanamycin, and
37
38ꢀ40
resorcinol scaffolds, respectively. Akt, Chk1, and Cdk4 are Hsp90 client proteins.
12
ACS Paragon Plus Environment

Page 13 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 12b exhibited cytotoxic activity that was superior to Hsp90 inhibitors in several
human cancer cells (Figure 7A,B). The cytotoxicity is largely dependent on TRAP1
inhibition of the drug, considering significant reduction in cancer cell viability at low drug
concentrations (~5 ꢁM) (Figure 7A) without signs of cytoplasmic Hsp90 inhibition (Figure
6B). However, normal human corneal cells and mouse primary hepatocytes were only
marginally affected by 12b treatment, whereas Hsp90 inhibitors caused considerably more
cytotoxicity in these normal cells (Figure 7C). Taken together, these data suggested that 12b
has superior cancerꢀspecific cytotoxicity but reduced cytotoxic effects in normal cells
compared with those of other Hsp90 inhibitors. Thus, to examine anticancer activities in vivo,
30 mg/kg 12b was intraperitoneally administered once daily to PC3 xenograft nude mice.
Compound 12b treatment reduced tumor growth (Figure 7D) and increased apoptotic cell
death in tumor tissues (Figure 8A), and did not cause weight loss or histologic abnormalities
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(Figures 7E and 8B). Moreover, in agreement with in vitro results, tumors isolated from 12bꢀ
treated animals showed reduced expression of Hsp90 client proteins and induction of CHOP
expression without elevated Hsp70 expression (Figure 8C).
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. In vitro and in vivo anticancer activities of 12b. (A) Cytotoxic activities of 12b
against several cancer cell lines. Cancer cells were incubated with the indicated drug
concentrations for 24 hours, and cell viability was analyzed using MTT assay. Data are
presented as means ± SEM from two independent experiments performed in triplicate. (B)
Cell death induction by 12b. HeLa and PC3 cells were incubated with 20 ꢁM 6 or 12b for 24
hours, labeled with FITCꢀDEVDꢀfmk and propidium iodide, and analyzed using flow
cytometry. 12b elevated dead cell population (PI positive) with apoptotic phenotype
(AnnexinV positive), whereas the Hsp90 inhibitor 6 did not trigger cell death. (C) Cytotoxic
activity in normal cells. Mouse primary hepatocytes and human corneal cells were incubated
with various drug concentrations for 24 hours, and cell viability was measured using MTT
assay. Data are presented as means ± SEM from two independent experiments performed in
duplicate; *p < 0.05; **p < 0.003; ***p < 0.0001 (12b vs. other drugs). (D) In vivo anticancer
14
ACS Paragon Plus Environment

Page 15 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
7
activity of 12b. PC3 cells (1 × 10 ) were subcutaneously implanted into nude mice. After the
establishment of tumors, mice were intraperitoneally injected with vehicle or 30 mg/kg 12b
daily, and tumor sizes were measured using calipers. (E) Mouse body weights. Vehicle
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
DMSO) and 12bꢀtreated mouse body weights were measured at the end of the experiment
and are presented in the bar graph; n.s., nonꢀsignificant.
Figure 8. In vivo activity of 12b. (A) Cell death induction in vivo. After 12b treatment of
PC3ꢀimplanted nude mice, as in Figure 7D, tumor tissues from vehicle (DMSO)ꢀ and 12bꢀ
treated mice were analyzed by TUNEL staining and examined using a fluorescence
+
microscope (left). Data are presented as means ± SEM; ***p< 0.0001. Dead cell (TUNEL )
population was markedly elevated in the 12b treated group. (B) Histological analyses of
drugꢀtreated mouse organs. Organs harvested from vehicle (DMSO)ꢀ or 12bꢀtreated mice
were fixed with formalin. Fixed tissues were stained with H&E and analyzed by dot slide
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
microscope. No histologic abnormality was observed in 12bꢀtreated mice. (C) In vivo mode
of action of 12b. After 12b treatment of PC3 implanted nude mice, as shown in Figure 7D,
tumor tissues from vehicle (DMSO)ꢀ and 12bꢀtreated mice were analyzed by western blotting
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(left). Quantitation of the band intensity (right). The data are the mean ± SEM; n.s. for p>0.05;
*p<0.05; **p<0.01. 12b treatment elevated CHOP expression, indicative of TRAP1
inhibition, without inducing Hsp70, a biomarker for Hsp90 inhibition.
CONCLUSIONS
In summary, for developing a TRAP1 inhibitor not requiring a mitochondrial delivery
vehicle that occasionally causes nonꢀspecific toxicity and poor ADME properties, we
hypothesized that the binding affinity to the abundant cytoplasmic Hsp90 should be
minimized to increase freely available drug concentrations around the mitochondria.
Collectively, our findings indicated that the development of the pyrazolopyrimidine 12b for
optimized TRAP1 and Hsp90 binding changed the subcellular drug distribution and
subsequently generated a potent TRAP1 inhibitor with a novel mode of action.
EXPERIMENTAL SECTION
Synthetic Methods and Molecular Characterization: Starting materials were purchased
from SigmaꢀAldrich, TCI, and AlfaꢀAesar and were used without further purification.
Solvents were purified by passage through a solvent column composed of activated alumina
and a supported copper redox catalyst. All reactions were performed under an atmosphere of
dry argon. A Biotage Initiator microwave system was used for microwaveꢀassisted reaction.
Thinꢀlayer chromatography was carried out on Merck precoated silica gel 60 F254 plates. A
Teledyne ISCO flash purification system with various prepacked silica gel cartridges was
1
13
used for flash column chromatography. H and C NMR spectra were recorded in the
16
ACS Paragon Plus Environment

Page 17 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
13
indicated solvent on a Bruker AVANCE III HD (400 and 100 MHz for H and C,
respectively) spectrometer. Chemical shifts are reported as δ values in parts per million
6 3
downfield from TMS (δ 0.0) as the internal standard in DMSOꢀd or CDCl . MS was
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
performed on a system consisting of an electrospray ionization (ESI) source in a Shimadzu
LCMSꢀ2020 liquid chromatographyꢀmass spectrometer (column: Shimꢀpack GIS, 100 x 3.0
mm, 3ꢁm ODS). The purity of the compounds was evaluated on a Shimadzu reverseꢀphase
analytical HPLC system (column: Ace C18, 150 x 4.6 mm, 3ꢁm). Purities of all compounds
that were subjected to biological assay were >95%.
Preparation of Hsp90 inhibitors: 1, 2, 3, 4, 5, and 6 were purchased from Tocris, Toronto
Research Chemicals, and SigmaꢀAldrich.
Piperonylhydrazine 11a and 11b: Hydrazine monohydrate (6.33 mL, 130 mmol) was
added to a solution of 3ꢀpiperonylcloride (1aꢀb: 130 mmol) in MeOH (40 mL) at 0 °C,
followed by stirring for 30 min at the same temperature. After stirring for 2.5 h at 60 °C, the
reaction mixture was cooled to room temperature and quenched by addition of water (24 mL).
The solvent was distilled to approximately 55 mL under reduced pressure. To the residue, 2
N NaOH (aq. 160 mL) was added; this was then extracted with CH
was dried over Na SO , filtered, and then concentrated under reduced pressure. The resulting
residue was dissolved in CH Cl (100 mL) and then cooled to 0 °C. Subsequently, 4 N HCl in
2 2
Cl . The organic layer
2
4
2
2
dioxane (38 mL, 150 mmol) was added to the solution, and the mixture was stirred at 0 °C for
3 h. The resulting precipitate was collected via filtration, followed by washing with CH Cl
isopropyl ether, and CH Cl , and it was finally dried to obtain piperonylhydrazine
1
2
2
,
2
2
1
hydrochloride (11a-b) as a white salt (78% for 11a, 82% for 11b). 11a: H NMR (400 MHz,
CDCl ) δ 6.84 (s, 1H), 6.78 (bs, 2H), 5.96 (s, 2H), 3.83 (s, 2H), 2.98 (bs,3H); MS (ESI, m/z)
3
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
+
1
calculated for C
8
H
11
N
2
O
2
[M+1] 167.1, found 167.0. 11b: H NMR (400 MHz, DMSOꢀd
6
) δ
1
3
7
.21 (s, 1H), 7.11 (s, 1H), 6.07 (s, 2H), 5.74 (br, 3H), 4.03 (s, 2H); C NMR (100 MHz,
DMSOꢀd
6
) δ 148.3, 147.6, 128.2, 114.2, 112.9, 110.7, 102.6, 52.9; MS (ESI, m/z) calculated
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
for C
1-(Benzo[
aminepiperonyl hydrazine (12a) and 1-((6-bromobenzo[
chloro-1 -pyrazolo[3,4- ]pyrimidin-6-amine (12b, DN401): A solution of triethylamine
1.2 mL, 8.4 mmol) in CH Cl (3.0 mL) was added dropwise to piperonylhydrazine
hydrochloride (11a or 11b) (2.5 mmol) and 2ꢀaminoꢀ4,6ꢀdichloropyrimidienꢀ5ꢀcarbaldehyde
(0.4 g, 2.1 mmol) in CH Cl (10 mL) at 0°C for 15 min. After stirring for 1.5 h at 0°C, the
reaction mixture was poured into 0.2 N HCl (aq) and then extracted with CH Cl (×3). The
combined organic layer was washed with brine and dried over MgSO . The resulting residue
8
H10BrN
2
O
2
[M+1] 245.0, found 245.5.
d
][1,3]dioxol-5-ylmethyl)-4-chloro-1
H
-pyrazolo[3,4-
d]pyrimidin-6-
d
][1,3]dioxol-5-yl)methyl)-4-
H
d
(
2
2
2
2
2
2
4
was purified by column chromatography (CH
2
Cl
2
to 1:4 EtOAc/CH
2
Cl
2
) to obtain the target
1
compound 12a (0.60 g, 95%) or 12b (0.79 g, 99%) as a white solid. 12a: H NMR (400 MHz,
MeOD) δ 7.97 (s, 1H), 6.93 (s, 2H), 6.48 (s, 1H), 5.97 (bs, 3H), 5.44 (s,2H); MS (ESI, m/z)
+
1
calculated for C13
5 2
H11ClN O [M+1] 304.1, found 304.0.; 12b: H NMR (400 MHz, DMSOꢀ
13
d ) δ 8.04 (s, 1H), 7.39 (s, 2H), 7.25 (s, 1H), 6.48 (s, 1H), 6.04 (s, 2H), 5.32 (s,2H); C NMR
6
(100 MHz, DMSOꢀd
6
) δ 161.4, 155.9, 153.5, 147.7, 147.2, 132.8, 128.6, 112.5, 112.3, 108.9,
+
106.0, 102.0, 49.49; MS (ESI, m/z) calculated for C13
82.0.
-Amino-6-((dibenzylamino)methyl)-1,2,4-triazin-5(4
H
10BrClN
5
O
2
[M+1] 382.0, found
3
3
H
)-one (14):
A
solution of
dibenzylamine (4.05 g, 20.5 mmol) and ethyl bromopyruvate (2.0 g, 10.3 mmol) in ethyl
aetate (10 mL) was heated to reflux for 2 h. The reaction mixture was cooled to room
temperature and filtered to remove precipitated solid. The filtrates were concentrated under
reduced pressure. The residue and aminoguanidine bicarbonate (1.4 g, 10.3 mmol) in EtOH
18
ACS Paragon Plus Environment

Page 19 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(10 mL) were heated to reflux for 12 h. After cooling to ambient temperature, the mixture
was concentrated under reduced pressure. The crude residue was purified by flash column
1
chromatography (EA/Hex) to afford the target compound 14 (487 mg, 14.9 % yield). δ H
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NMR (400 MHz, DMSOꢀd
6
) δ 12.08 (s, 1H), 7.54–7.04 (m, 10H), 6.74 (s, 2H), 3.64 (s, 4H),
+
3.47 (s, 2H); MS (ESI, m/z) calculated for C18
20
H N
5
O [M+1] 224.1, found 224.5.
N
-((3-Amino-5-oxo-4,5-dihydro-1,2,4-triazin-6-yl)methyl)-2-(benzo[
acetamide (15): Under hydrogen atmosphere, a mixture of 14 (343 mg, 1.07 mmol) and 10%
Pd/C (11 mg, 0.01 mmol) in AcOH (0.2 mL) and H O (3 mL) was stirred for 4 h at 55 °C.
d][1,3]dioxol-5-yl)
2
After cooling to room temperature, the mixture was filtered through a plug of celite, and the
filtrate evaporated to dryness. The residue was dissolved in TEA (0.42 ml, 2.61 mmol) and
3
CH CN (2 mL) and heated to 50 °C. 2,5ꢀdioxopyrrolidinꢀ1ꢀyl 2ꢀ(benzo[d][1,3]dioxolꢀ5ꢀ
yl)acetate (296 mg, 1.07 mmol) was added to the solution and the solution was stirred until
no starting material was shown in TLC. After cooling to room temperature, the mixture was
filtered, the residue washed with water/CAN, and the filtrate dried in vacuum to give the
1
crude titled compound (151 mg, 46.6% yield). H NMR (400 MHz, DMSOꢀd
6
) δ 12.06 (s,
1H), 8.11 (s, 1H), 6.81 (m, 3H), 5.97 (s, 2H), 4.06 (d, J = 5.7 Hz, 2H), 3.72 (s, 2H); MS (ESI,
+
m/z) calculated for C H N O [M+1] 304.1, found 304.9.
1
3
14
5
4
7
-(Benzo[
To a suspension of 15 (150 mg, 0.5 mmol) in CH
heated to reflux for 5 h. After cooling to 0 °C, a saturated aqueous K
d
][1,3]dioxol-5-ylmethyl)-4-chloroimidazo[5,1-
CN (3 mL) was added POCl
CO (5 mL) was added
f][1,2,4]triazin-2-amine (16):
3
3
(97 uL) and
2
3
dropwise to the mixture at a rate to keep temperature below 30 °C. Once the quench was
complete (pH= 9), the suspension was stirred at room temperature for 20 min and then
filtered to obtain the crude solid product. The solid was washed with water, and dried in
vacuum. The dried solid (37 mg) was dissolved in POCl
3
(3 mL) and heated to reflux for 10 h.
The mixture was cooled, concentrated under vacuum, and poured into ice water. The pH of
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
the mixture was adjusted to 8 using saturated aqueous NaHCO . The organic material was
extracted with EA, dried over MgSO , and filtered. The filtrated was concentrated under
4
reduced pressure and the residue was purified by column chromatography to give the target
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
compound 16 (10 mg, 11% yield). H NMR (400 MHz, CDCl
3
) δ 7.75 (s, 1H), 7.04 (s, 2H),
6.83 (d, J = 1.5 Hz 1H), 6.80 (d, J = 7.9 Hz 1H), 6.68 (dd, J = 8.0, 1.6 Hz 1H), 5.91 (s, 2H),
+
4
.17 (s, 2H): MS (ESI, m/z) calculated for C13
Methyl (benzo[
ester hydrochloride (3 g, 23.9 mmol), piperonal (4.3 g, 28.7 mmol) in CH
added TEA (3.9 ml, 28.7 mmol). After stirring for 1 h a room temperature, NaBH
H11ClN
5
O
2
[M+1] 304.1, found 304.9.
d
][1,3]dioxol-5-ylmethyl)glycinate (18): To a solution of Glycine methyl
Cl (5 ml) was
(497 mg,
2
2
4
13.14 mmol) was added to the mixture at 0 °C. The reaction mixture was stirred for 12 h at
room temperature, after which water (30 mL) and ethyl acetate (30 mL) were added to the
mixture. The organic layer was partitioned, dried over MgSO filtered, and concentrated
4
,
under reduced pressure. The crude residue was purified by flash column chromatography
1
(
EA/Hex) to give the titled compound 18 as colorless oil (3.5 g, 65.6 %). H NMR (400 MHz,
CDCl
3
) δ 6.83 (s, 1H), 6.74 (s, 2H), 5.93 (s, 2 H), 3.71 (s, 3 H), 3.69 (s, 2H), 3.39 (s, 2H);
+
MS (ESI, m/z) calculated for C11
H
14NO
4
[M+1] 224.1, found 224.0.
Methyl 2-amino-7-(benzo[
d
][1,3]dioxol-5-ylmethyl)-4-chloro-7H-pyrrolo[2,3-
d
]pyrimidine-6-carboxylate (19): To a solution of 18 (590 mg, 2.64 mmol) in DMF was
added TEA (0.57 mL, 3.2 mmol) and 2ꢀaminoꢀ4,6ꢀdichloropyrimidineꢀ5ꢀcarbaldehyde (500
mg, 3.2 mmol). After stirring for 6 h at room temperature, DBU (0.45 mL, 3.0 mmol) was
added. The mixture was stirred for a further 12 h at 60 °C, then diluted with EA and water.
The organic layer was dried over MgSO , filtered, and concentrated under reduced pressure.
4
The crude residue was purified by flash column chromatography (EA/Hex) to give the title
1
compound 19 as yellow solid (200 mg, 70% yield). H NMR (400 MHz, DMSOꢀd
6
) δ 7.23
(bs, 2H), 7.02 ꢀ7.19 (m, 1H), 6.80 (d, J = 8.07 Hz, 1H), 6.74 (s, 1H), 6.56 (d, J = 8.07 Hz,
20
ACS Paragon Plus Environment

Page 21 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4 4
1H), 5.96 (s, 2H), 5.51 (s, 2H), 3.78 (s, 3H); MS (ESI, m/z) calculated for C16H14ClN O
+
[
M+1] 361.1, found 361.5.
2-Amino-7-(benzo[d][1,3]dioxol-5-ylmethyl)-4-chloro-7H-pyrrolo[2,3-d]pyrimidine-6-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
carboxylic acid (20): A mixture of 19 (150 mg, 0.42 mmol) and LiOH (15 mg, 0.64 mmol)
in THF/water (2:1) was stirred for overnight at 65 °C. The reaction mixture was cooled to
room temperature, acidified with 1N HCl, and extracted with EA. The organic phase was
dried over MgSO filtered, and concentrated under reduced pressure. The crude residue was
4
,
1
purified by flash column chromatography to give 20 (126 mg, 84% yield). H NMR (400
MHz, DMSOꢀd ) δ 13.07 (bs, OH), 7.06 (s, 1H), 6.80 (d, J = 8.0 Hz, 1H), 6.73 (d, J = 1.7 Hz,
H), 6.55 (dd, J = 8.0, 1.6 Hz, 1H), 5.95 (s, 2H), 5.53 (s, NH ); MS (ESI, m/z) calculated for
6
1
2
+
C
15
H
12ClN
4
O
4
[M+1] 347.1, found 347.2.
4-Chloro-1-((4-methoxy-3,5-dimethylpyridin-2-yl)methyl)-1H-pyrazolo[3,4-
7
1
d
]pyrimidin-6-amine (21, DN320): 21 was prepared using a published procedure : H NMR
(
400 MHz, DMSOꢀ d ) δ 8.09 (s, 1H), 7.91 (s, 1H), 5.53 (s, 2H), 3.79 (s, 3H), 2.26 (s, 6H);
6
13C NMR (100 MHz, DMSOꢀd
6
) δ 164.7, 161.8, 156.2, 154.7, 153.2, 148.3, 132.6, 126.3,
+
125.5, 106.8, 59.2, 49.2, 11.9, 9.6; MS (ESI, m/z) calcd for C14
H
15ClN
6
O [M+1] 318.8,
found 320.0.
Biological Experimental Section
Chemicals, plasmids, and antibodies. All chemicals were purchased from Sigma. Antiꢀ
CHOP and antiꢀCdk4 antibodies were purchased from Cell Signaling; antiꢀAkt and antiꢀChk1
were purchased from Santa Cruz Biotechnology; antiꢀβꢀactin was obtained from MP
Biomedicals; antiꢀcytochrome c and antiꢀHsp70 were purchased from BD Biosciences.
Cells and culture conditions. Human cells from cervix (HeLa), liver (SKꢀHEPꢀ1), brain
(T98G), lung (H460), and prostate (PC3) cancers were purchased from the American Type
Culture Collection (ATCC) and were maintained as recommended by the supplier. Cells were
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
cultured in DMEM or RPMI medium (GIBCO) containing 10% fetal bovine serum (FBS;
GIBCO) and 1% penicillin/streptomycin (GIBCO) at 37°C in a humidified atmosphere of 5%
2
CO . Human corneal cells were purchased from ATCC and were cultured in ATCC corneal
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
medium (ATCC) according to the manufacturer’s instructions. Primary hepatocytes were
4
isolated from 8ꢀweekꢀold BALB/c mice as previously described and were cultured in
M199/EBSS medium (Hyclone) containing 10% FBS at 37°C in a humidified atmosphere of
5% CO2.
Analyses of cell viability and apoptosis induction. Cell viability was determined using 3ꢀ
(4,5ꢀdimethylꢀthyzoylꢀ2ꢀyl)ꢀ2,5ꢀdiphenyltetrazolium bromide (MTT) assays and was
quantified by measuring absorbance of the tetrazolium dye at 595 nm using an Infinity M200
microplate reader (TECAN). Absorbance values were normalized to the DMSO control and
data were expressed as percent viability. To measure apoptosis induction, DNA contents
(propidium iodide) and externalized phosphatidylserine (Annexin V) using apoptosis kit
(Molecular Probe). Labeled cells were quantified using the FACS Calibur™ system (BD
Biosciences). Data were processed using FlowJo software (TreeStar). Cytochrome c release
was determined by Western blotting after mitochondrial fractionation as described
1
8
previously. Apoptosis in the tumor sections was assessed using in situ apoptosis detection
kit (Millipore) according to the manufacturer’s protocols and analyzed using an FV1000 laser
confocal scanning microscope (Olympus).
Recombinant protein preparation and fluorescence polarization assays. Recombinant
TRAP1 and Hsp90α were prepared (Supporting Information, Figure S4) as described
4
previously. For fluorescence polarization experiments, the fluorescence probe 1ꢀFITC3 was
2
1
synthesized as described previously, and 10 nM 1ꢀFITC3 and 100 nM protein were
incubated for 24 hours at 4°C with various concentrations of inhibitors in FP buffer
2 4 2 4
containing 135 mM NaCl, 2.7 mM KCl, 4.3 mM Na HPO , 1.4 mM KH PO , 1 mM DTT, 2
22
ACS Paragon Plus Environment

Page 23 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mM MgCl
2
, 0.1 mg/mL BSA, and 0.05% NP40 (pH 7.3). Fluorescence polarization was
finally measured using a SYNERGY NEO microplate reader (BioTek Instruments, Inc.).
ATPase Activity Assay. ATPase activity was measured with the PiColorLock Gold
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
Phosphate Detection Kit (Innova Biosciences) as described previously. Briefly, 200 nM
TRAP1 was preincubated with various concentrations of inhibitors in assay buffer (100 mM
2
Tris, 20 mM KCl, and 6 mM MgCl , pH 7.0) for 30 min at 37°C. Then, the mixture was
incubated with 0.2 mM ATP for 3 h and the 20 ꢁL mixture of Pi Color Lock Gold reagent
and the accelerator (100:1) was added to the 80 ꢁL ATP hydrolysate sample. After 5 min
incubation at 25°C, color development was stopped by the addition of 10 ꢁL stop solution,
and absorbance was measured at 620 nm with an Infinity M200 microplate reader (Tecan).
Analysis of mitochondrial membrane potential and ROS production. To examine
mitochondrial membrane potential or superoxide production, HeLa cells were incubated with
5
ꢁg/mL JCꢀ1 (Molecular Probes) or 200 nM MitoꢀSOX Red (Invitrogen) for 20 min at 37°C.
Subsequently, the cells were washed with PBS, trypsinized, resuspended in 500 ꢁL of PBS,
and analyzed using the FACS Calibur system (BD Bioscience).
Crystallization and structure determination. Protein production for human TRAP1
lacking the C terminal domain (residues 60–561, hTRAP1ꢀNM) was performed as described
4
, 41
previously.
purified hTRAP1 NM domain was mixed with 3.7
on ice and adjusted to a final volume of 90 l in buffer containing 25 mM TrisꢀCl (pH 7.5),
To crystallize hTRAP1ꢀNM and inhibitor complexes, 50
ꢀl of 20 mg/mL
ꢀ
l of 10 mM inhibitors in DMSO for 1 h
ꢀ
150 mM NaCl, and 5 mM DTT to reduce the DMSO concentration to 4%. Complexes were
then crystallized using the hanging drop diffusion method at room temperature by mixing 1
ꢀ
l of protein solution with 1 ꢀl of reservoir buffer. Crystals of the TRAP1–21 complex grew
in reservoir buffers comprising 100 mM sodium cacodylate (pH 6.76), 14%–16% 8K
polyethylene glycol (PEG), and 100 mM calcium acetate. The bestꢀdiffracted crystals of the
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
TRAP1–21 complex were obtained in reservoir buffer containing 14%–16% 20K PEG
instead of 8K PEG. Crystals were then cryoprotected in a crystallization solution
supplemented with 30% glycerol and were flash frozen in liquid nitrogen. Xꢀray diffraction
data were collected in the 5C beamline of the Pohang Accelerator laboratory (PAL), and data
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
42
were processed using HKL2000 software. Both crystals had almost the same space group
(P41212) and unit cell (a = b = 69.5 Å, c = 253.0 Å) as previously reported for the TRAP1–1
4
complex crystal. Electron densities of inhibitors were calculated using difference Fourier
4
methods with the unliganded TRAP1 structure from TRAP1–1 (PDB 4Z1F), and were
43
subsequently built into residual electron densities using Coot. Final models for TRAP1–12b
(PDB 5Y3N) and TRAP1–21 (PDB 5Y3O) were refined to 2.4 and 2.7 Å resolution,
44
respectively, using Phenix. Crystallography statistics are summarized in Supporting
Information Table S1. All structural figures were generated using PyMOL
(http://www.pymol.com).
Tumor xenograft experiments. All experiments involving animals were approved by
7
UNIST (IACUCꢀ12ꢀ003ꢀA). Cancer cells (1 × 10 PC3) were suspended in 200 ꢁL of PBS
and were injected subcutaneously into both flanks of 6ꢀweekꢀold BALB/c nu/nu male mice
(Charles River Laboratories). Tumors are then allowed to grow to an average volume of
3
approximately 100 mm and animals were randomly grouped (two tumors/mouse, five
mice/group). Subsequently, DMSO (vehicle) and 12b dissolved in DMSO were each mixed
with propylene glycol at a 1:1 ratio (final propylene glycol concentration, 50%) and was
administered intraperitoneally (30 mg/kg) every day. Tumors were measured using calipers,
2
and tumor volumes were calculated using the following formula: V = 1/2 × (width) × length.
At the end of experiment, animals were euthanized, and organs including brain, heart, kidney,
liver, lung, spleen, stomach, and tumor were collected for histology and western blotting.
Harvested organ specimens were also fixed in 10% formalin and embedded in paraffin for
24
ACS Paragon Plus Environment

Page 25 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
histological analyses. Briefly, sections (5 ꢁm) were placed on highꢀadhesive slides, were
stained with H&E, and were scanned using the Dotslide system (Olympus) with 20×
magnification. For Western blot analyses, tissue samples were homogenized in RIPA buffer
containing 50 mM Tris (pH 8.0), 150 mM NaCl, 1% NPꢀ40, and 0.25% Nꢀdeoxycholate, and
protease inhibitor and phosphatase inhibitor cocktails (Calbiochem), and soluble factions
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
2
were used for analysis as described previously. The band intensity was quantified using
ImageJ software (National Institute of Health, USA).
Statistical analyses. MTT experiments were duplicated and repeated independently at least
three times. Statistical analyses were performed using the software program Prism 7.0
(GraphPad). Differences were identified using unpaired tꢀtests and were considered
significant when p < 0.05.
ASSOCIATED CONTENT
Supporting Information. The Supporting Information is available free of charge on the ACS
Publications website.
Additional structural comparison; TRAP1 SDSꢀPAGE analysis; crystallography parameters
and refinements (PDF)
Molecular formula strings (CSV)
Accession codes. The atomic coordinates and crystallographic structure factors can be
accessed using PDB code 5Y3O (21) and 5Y3N (12b) in the Protein Data Bank
(www.rcsb.org). Authors will release coordinates upon article publication.
AUTHOR INFORMATION
Corresponding Author
Changwook Lee, Ph.D.
Phone: +82ꢀ52ꢀ217ꢀ2534, Eꢀmail: changwook@unist.ac.kr
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Soosung Kang, Ph.D.
Phone: +82ꢀ2ꢀ3277ꢀ6619, Eꢀmail: sskang@ewha.ac.kr
orcid.org/0000ꢀ0001ꢀ7016ꢀ2417
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Byoung Heon Kang, Ph.D.
Phone: +82ꢀ52ꢀ217ꢀ2521, Eꢀmail: kangbh@unist.ac.kr
orcid.org/0000ꢀ0001ꢀ5902ꢀ0549
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank the staff of the 5C beamline of the PAL for assistance with synchrotron facilities
and the staff of the UNISTꢀOlympus Biomedical Imaging Center for technical support. This
work was supported by the 2015 UNIST Research Fund (1.150098.01), the National
Research Foundation of Korea (NRF) grants funded by MEST (2015R1D1A1A01058016;
2010ꢀ0028684; 2016R1A2B2012624), and the Korea Drug Development Fund funded by
MSIP, MOTIE, and MOHW (KDDFꢀ201512ꢀ02).
.
ABBREVIATIONS USED
CypD, cyclophilin D; DMAG, 17ꢀ(dimethylaminoethylamino)ꢀ17ꢀdemethoxygeldanamycin;
Hsp90, 90 kDa heat shock protein; MTT, 3(4,5ꢀdimethylꢀthyzoylꢀ2ꢀyl)2,5
diphenyltetrazolium bromide; NTD, Nꢀterminal domain; SDSꢀPAGE, sodium dodecyl sulfate
26
ACS Paragon Plus Environment

Page 27 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
polyacrylamide gel electrophoresis; TPP, triphenylphosphonium; TRAP1, tumor necrosis
factor receptorꢀassociated protein 1; TRAP1ꢀNM, Cꢀterminal deleted TRAP1.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
REFERENCES
(1) Felts, S. J.; Owen, B. A.; Nguyen, P.; Trepel, J.; Donner, D. B.; Toft, D. O. The hsp90ꢀ
related protein TRAP1 is a mitochondrial protein with distinct functional properties. J. Biol.
Chem. 2000, 275, 3305ꢀ3312.
(
2) Chen, B.; Piel, W. H.; Gui, L.; Bruford, E.; Monteiro, A. The Hsp90 family of genes in
the human genome: insights into their divergence and evolution. Genomics 2005, 86, 627ꢀ637.
3) Matassa, D. S.; Amoroso, M. R.; Maddalena, F.; Landriscina, M.; Esposito, F. New
(
insights into TRAP1 pathway. Am. J. Cancer Res. 2012, 2, 235ꢀ248.
(4) Lee, C.; Park, H. K.; Jeong, H.; Lim, J.; Lee, A. J.; Cheon, K. Y.; Kim, C. S.; Thomas, A.
P.; Bae, B.; Kim, N. D.; Kim, S. H.; Suh, P. G.; Ryu, J. H.; Kang, B. H. Development of a
mitochondriaꢀtargeted Hsp90 inhibitor based on the crystal structures of human TRAP1. J.
Am. Chem. Soc. 2015, 137, 4358ꢀ4367.
(5) Gewirth, D. T. Paralog specific Hsp90 inhibitors ꢀ a brief history and a bright future.
Curr. Top. Med. Chem. 2016, 16, 2779ꢀ2791.
(6) Immormino, R. M.; Kang, Y.; Chiosis, G.; Gewirth, D. T. Structural and quantum
chemical studies of 8ꢀarylꢀsulfanyl adenine class Hsp90 inhibitors. J. Med. Chem. 2006, 49,
4953ꢀ4960.
(7) Biamonte, M. A.; Boehm, M. F.; Hong, K. D.; Hurst, D.; Kasibhatla, S. R.; JeanꢀYves Le,
B.; Shi, J.; Zhang, L. Novel heterocyclic compounds as Hsp90ꢀinhibitors. WO 2005028434
A3, 2004.
(8) Whitesell, L.; Cook, P. Stable and specific binding of heat shock protein 90 by
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 33
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
geldanamycin disrupts glucocorticoid receptor function in intact cells. Mol. Endocrinol. 1996,
10, 705ꢀ712.
(
9) Egorin, M. J.; Rosen, D. M.; Wolff, J. H.; Callery, P. S.; Musser, S. M.; Eiseman, J. L.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Metabolism of 17ꢀ(allylamino)ꢀ17ꢀdemethoxygeldanamycin (NSC 330507) by murine and
human hepatic preparations. Cancer Res. 1998, 58, 2385−2396.
(
10) Jez, J. M.; Chen, J. C.; Rastelli, G.; Stroud, R. M.; Santi, D. V. Crystal structure and
molecular modeling of 17ꢀDMAG in complex with human Hsp90. Chem. Biol. 2003, 10,
61−368.
11) Brough, P. A.; Aherne, W.; Barril, X.; Borgognoni, J.; Boxall, K.; Cansfield, J. E.;
3
(
Cheung, K. M.; Collins, I.; Davies, N. G.; Drysdale, M. J.; Dymock, B.; Eccles, S. A.; Finch,
H.; Fink, A.; Hayes, A.; Howes, R.; Hubbard, R. E.; James, K.; Jordan, A. M.; Lockie, A.;
Martins, V.; Massey, A.; Matthews, T. P.; McDonald, E.; Northfield, C. J.; Pearl, L. H.;
Prodromou, C.; Ray, S.; Raynaud, F. I.; Roughley, S. D.; Sharp, S. Y.; Surgenor, A.;
Walmsley, D. L.; Webb, P.; Wood, M.; Workman, P.; Wright, L. 4,5ꢀdiarylisoxazole Hsp90
chaperone inhibitors: potential therapeutic agents for the treatment of cancer. J. Med. Chem.
2008, 51, 196−218.
(
12) Lin, T. Y.; Bear, M.; Du, Z.; Foley, K. P.; Ying, W.; Barsoum, J.; London, C. The
novel HSP90 inhibitor STAꢀ9090 exhibits activity against Kitꢀdependent and ꢀindependent
malignant mast cell tumors. Exp. Hematol. 2008, 36, 1266−1277.
(13) Woodhead, A. J.; Angove, H.; Carr, M. G.; Chessari, G.; Congreve, M.; Coyle, J. E.;
Cosme, J.; Graham, B.; Day, P. J.; Downham, R.; Fazal, L.; Feltell, R.; Figueroa, E.;
Frederickson, M.; Lewis, J.; McMenamin, R.; Murray, C. W.; O’Brien, M. A.; Parra, L.;
Patel, S.; Phillips, T.; Rees, D. C.; Rich, S.; Smith, D. M.; Trewartha, G.; Vinkovic, M.;
Williams, B.; Woolford, A. J. Discovery of (2,4ꢀdihydroxyꢀ5ꢀisopropylphenyl)ꢀ[5ꢀ(4ꢀ
methylpiperazinꢀ1ꢀylmethyl)ꢀ1,3ꢀdihydrois oindolꢀ2ꢀyl]methanone (AT13387), a novel
28
ACS Paragon Plus Environment

Page 29 of 33
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
inhibitor of the molecular chaperone Hsp90 by fragment based drug design. J. Med. Chem.
2010, 53, 5956−5969.
(14) Kakashima, T.; Ishii, T.; Tagaya, H.; Seike, T.; Nakagawa, H.; Kanda, Y.; Akinaga, S.;
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Soga, S.; Shiotsu, Y. New molecular and biological mechanism of antitumor activities of
KWꢀ2478, a novel nonasnamycin heat shock protein 90 inhibitor in multiple myeloma cells.
Clin. Cancer Res. 2010, 16, 2792−2802.
(
15) Patel, P. D.; Yan, P.; Seidler, P. M.; Patel, H. J.; Sun, W.; Yang, C.; Que, N. S.; Taldone,
T.; Finotti, P.; Stephani, R. A.; Gewirth, D. T.; Chiosis, G. Paralogꢀselective Hsp90 inhibitors
define tumorꢀspecific regulation of HER2. Nat. Chem. Biol. 2013, 9, 677ꢀ684.
(16) Kang, B. H.; Altieri, D. C. Compartmentalized cancer drug discovery targeting
mitochondrial Hsp90 chaperones. Oncogene 2009, 28, 3681ꢀ3688.
(
17) Kang, B. H. TRAP1 regulation of mitochondrial life or death decision in cancer cells
and mitochondriaꢀtargeted TRAP1 inhibitors. BMB Rep. 2012, 45, 1ꢀ6.
(18) Kang, B. H.; Plescia, J.; Song, H. Y.; Meli, M.; Colombo, G.; Beebe, K.; Scroggins, B.;
Neckers, L.; Altieri, D. C. Combinatorial drug design targeting multiple cancer signaling
networks controlled by mitochondrial Hsp90. J. Clin. Invest. 2009, 119, 454ꢀ464.
(19) Butler, L. M.; Ferraldeschi, R.; Armstrong, H. K.; Centenera, M. M.; Workman, P.
Maximizing the therapeutic potential of HSP90 inhibitors. Mol. Cancer Res. 2015, 13, 1445ꢀ
1451.
(
20) Kang, B. H.; Siegelin, M. D.; Plescia, J.; Raskett, C. M.; Garlick, D. S.; Dohi, T.; Lian,
J. B.; Stein, G. S.; Languino, L. R.; Altieri, D. C. Preclinical characterization of
mitochondriaꢀtargeted small molecule hsp90 inhibitors, gamitrinibs, in advanced prostate
cancer. Clin. Cancer Res. 2010, 16, 4779ꢀ4788.
(
21) Taldone, T.; Patel, P. D.; Patel, M.; Patel, H. J.; Evans, C. E.; Rodina, A.; Ochiana, S.;
Shah, S. K.; Uddin, M.; Gewirth, D.; Chiosis, G. Experimental and structural testing module
29
ACS Paragon Plus Environment