Development of pyrazolo[3,4-d]pyrimidine-6-amine-based TRAP1 inhibitors that demonstrate in vivo anticancer activity in mouse xenograft models

Article abstract of DOI:10.1016/j.bioorg.2020.103901

TNF Receptor Associated Protein 1 (TRAP1) is a mitochondrial paralog of Hsp90 related to the promotion of tumorigenesis in various cancers via maintaining mitochondrial integrity, reducing the production of reactive oxygen species, and reprogramming cellular metabolism. Consequently, Hsp90 and TRAP1 have been targeted to develop cancer therapeutics. Herein, we report a series of pyrazolo[3,4-d]pyrimidine derivatives that are mitochondria-permeable TRAP1 inhibitors. Structure-based drug design guided the optimization of potency, leading to the identification of compounds 47 and 48 as potent TRAP1 and Hsp90 inhibitors with good metabolic and plasma stability as well as acceptable CYP and hERG inhibition. X-ray co-crystallization studies confirmed both 47 and 48 interact with the ATP binding pocket in the TRAP1 protein. Compounds 47 and 48 demonstrated excellent anticancer efficiency in various cancer cells, with limited toxicity over normal hepatocyte and prostate cells. Mouse PC3 xenograft studies showed 47 and 48 significantly reduced tumor growth.

Full text of DOI:10.1016/j.bioorg.2020.103901

Journal Pre-proofs
Development of Pyrazolo[3,4-d]pyrimidine-6-amine-Based TRAP1 Inhibitors
That Demonstrate In Vivo Anticancer Activity in Mouse Xenograft Models
Darong Kim, So-Yeon Kim, Dongyoung Kim, Nam Gu Yoon, Jisu Yun, Ki
Bum Hong, Changwook Lee, Ji Hoon Lee, Byoung Heon Kang, Soosung
Kang
PII:
S0045-2068(20)30057-2
DOI:
https://doi.org/10.1016/j.bioorg.2020.103901
YBIOO 103901
Reference:
To appear in:
Bioorganic Chemistry
Received Date:
Revised Date:
Accepted Date:
8 January 2020
10 April 2020
28 April 2020
Please cite this article as: D. Kim, S-Y. Kim, D. Kim, N. Gu Yoon, J. Yun, K. Bum Hong, C. Lee, J. Hoon Lee,
B. Heon Kang, S. Kang, Development of Pyrazolo[3,4-d]pyrimidine-6-amine-Based TRAP1 Inhibitors That
Demonstrate In Vivo Anticancer Activity in Mouse Xenograft Models, Bioorganic Chemistry (2020), doi: https://
doi.org/10.1016/j.bioorg.2020.103901
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover
page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version
will undergo additional copyediting, typesetting and review before it is published in its final form, but we are
providing this version to give early visibility of the article. Please note that, during the production process, errors
may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier Inc.

Development of Pyrazolo[3,4-d]pyrimidine-6-
amine-Based TRAP1 Inhibitors That Demonstrate In
Vivo Anticancer Activity in Mouse Xenograft
Models
Darong Kim 1, †, So-Yeon Kim2,†, Dongyoung Kim2,†, Nam Gu Yoon2, Jisu Yun3, Ki Bum Hong1,
Changwook Lee2,*, Ji Hoon Lee1,* , Byoung Heon Kang2,*, Soosung Kang3,4,
*
1 New Drug Development Center, Daegu-Gyeongbuk Medical Innovation Foundation (DGMIF),
Daegu 41061, South Korea
2 Department of Biological Sciences, Ulsan National Institutes of Science and Technology
(UNIST), Ulsan, 44919, Republic of Korea
3College of Pharmacy and Graduate School of Pharmaceutical Sciences, Ewha Womans
University, Seoul 03760, South Korea
4Western Seoul Center, Korea Basic Science Institute, Seoul 03760, South Korea
† These authors contributed equally.
*Corresponding Authors|
S. Kang: Tel. 82-2-3277-6619, email sskang@ewha.ac.kr
B.H. Kang: Tel. 82-52-217-2521, email kangbh@unist.ac.kr;
1

J.H. Lee: Tel. 82-53-790-5233, email jhlee@dgmif.re.kr;
C. Lee: Tel. 82-52-217-2534, email changwook@unist.ac.kr
KEYWORDS: TRAP1, selectivity, mitochondria, Hsp90, Anticancer, Drug.
GRAPHIC ABSTRACT:
ABSTRACT
TNF Receptor Associated Protein 1 (TRAP1) is a mitochondrial paralog of Hsp90 related to
the promotion of tumorigenesis in various cancers via maintaining mitochondrial integrity,
reducing the production of reactive oxygen species, and reprogramming cellular metabolism.
Consequently, Hsp90 and TRAP1 have been targeted to develop cancer therapeutics. Herein, we
report a series of pyrazolo[3,4-d]pyrimidine derivatives that are mitochondria-permeable TRAP1
inhibitors. Structure-based drug design guided the optimization of potency, leading to the
identification of compounds 47 and 48 as potent TRAP1 and Hsp90 inhibitors with good
metabolic and plasma stability as well as acceptable CYP and hERG inhibition. X-ray co-
crystallization studies confirmed both 47 and 48 interact with the ATP binding pocket in the
TRAP1 protein. Compounds 47 and 48 demonstrated excellent anticancer efficiency in various
2

cancer cells, with limited toxicity over normal hepatocyte and prostate cells. Mouse PC3
xenograft studies showed 47 and 48 significantly reduced tumor growth.
INTRODUCTION
TNF Receptor Associated Protein 1 (TRAP1) is a mitochondrial paralog of 90 kDa Heat Shock
Protein (Hsp90) closely related to the promotion of tumorigenesis in various cancers. 1 ,2
Mechanistically, TRAP1 facilitates a high cell death threshold against cellular stresses by
maintaining mitochondrial integrity, reducing the production of reactive oxygen species (ROS),
and reprogramming cellular metabolism to enable cancer cells to better adapt to harsh tumor
environments.3,4,5 Inactivation of TRAP1 induces massive apoptosis in cancer cells in vitro and in
vivo, and therefore, several inhibitors targeting mitochondrial TRAP1 have been developed6
(Figure 1).
Figure 1. Representative TRAP1 inhibitors 17, 28, 39, and 4.10 Compounds 2, 3, and 4 can
inactivate TRAP1 both in vitro and in vivo. Compound 1 can bind with and inactivate TRAP1 in
vitro, but cannot inhibit TRAP1 in vivo due to the lack of mitochondrial drug permeation.
Mitochondrion is a double membrane-enclosed organelle that orchestrates various cellular
metabolic pathways.11 In particular, the integrity of the mitochondrial inner membrane (IM) is
3

crucial for the metabolic function of mitochondria in generating a proton gradient across the IM
during oxidative phosphorylation.11 The IM is impermeable to various small molecules including
some therapeutic drugs, and therefore, drug permeability to IM should be considered in order to
efficiently inactivate mitochondrial target proteins in vivo.12 Likewise, many Hsp90 inhibitors
have been reported to inactivate purified recombinant TRAP1 in vitro but cannot inhibit TRAP1
in vivo due to insufficient mitochondrial membrane permeation.13 To inactivate mitochondrial
TRAP1
by
increasing
drug
permeability,
a
mitochondria-targeting
moiety12,
triphenylphosphonium-hexyl (TPP-hexyl), was conjugated with Hsp90 inhibitors.8,9,14 However,
the TPP-hexyl moiety itself can pose a challenge in the development of clinical candidates
because of its heavy (MW= 347) and lipophilic characteristics (cLogP = 8.78).
We previously reported the discovery of DN401 (4)10, a mitochondria-permeable TRAP1
inhibitor without a mitochondria delivery vehicle12, which showed potent anticancer activity in
vitro and in vivo. Recently, we found that 4 could simultaneously inactivate other Hsp90 family
proteins, consequently inducing the degradation of their pro-tumorigenic client proteins localized
in the cytoplasm and endoplasmic reticulum in vivo, and show enhanced anticancer activity.10, 15
Even though 4 inhibited the cytoplasmic Hsp90, it never induced adverse effects (pro-survival
heat shock responses) such as the activation of Hsp70, which is inevitably triggered after
treatment with current Hsp90 inhibitors. Thus, we concluded that mitochondria-permeable
TRAP1 inhibitors with or without Hsp90 inhibitory activity can be used for the development of
potent anticancer drugs.
However, we later found that 4 has poor metabolic stability when incubated with liver
microsomes (human, 35.6%; rat, 6.6%; mouse, 4.0% remaining after 30 min). In a continuing
effort to develop clinically applicable TRAP1 inhibitors, we synthesized various analogues of 4.
4

After obtaining X-ray co-crystal structures with TRAP1 and analyzing the structure-activity
relationship (SAR) of the products, we found that compounds 47 and 48 have better metabolic
stabilities as well as good anticancer activities in both cancer cell-based assays and a mouse
tumor xenograft model.
CHEMISTRY
Pyrazolo[3,4-d]pyrimidine derivatives 8-52 were prepared from commercially available 2-
Amino-4,6-dihydroxpyrimidine (5) in three steps (Scheme 1)10. Intermediate 6 was prepared with
the Vilsmeier-Haack reaction16 via the addition of compound 5 to a mixture of POCl3 and
dimethylformamide at 100 °C for 4 h. The pyrazolo[3,4-d]pyrimidine core (7) was prepared via
the condensation of hydrated hydrazine with intermediate 6. The base-mediated N-benzylation of
pyrazolo[3,4-d]pyrimidine (7) using a diverse range of benzyl bromide provided the desired N-
benzyl pyrazolo[3,4-d]pyrimidine derivatives 8-52 in moderate to good yield.
Scheme 1. Experimental Procedure for the Synthesis of 8-52a
aReagents and condition: a) POCl3, DMF, 100 °C, 4 h, 60~80%; b) hydrazine monohydrate,
TEA, water:THF=1:1, rt, 15 h, 60%; c) benzyl bromide, K2CO3, DMF, 60 °C, 4 h, 20~60%.
3,4-Dimethyl
pyrazolopyrimidine
derivative
56
was
prepared
from
3-
(bis(methylthio)methylene)pentane-2,4-dione (53) in three steps (Scheme 2). A pyrimidine
synthesized from 53 and guanidinium carbonate17 was reacted with hydrazine to yield 55. This
5

compound then underwent N-benzylation with 5-bromo-6-(bromomethyl)benzo[d][1,3]dioxole
to give the target product 56 in moderate yield. When 57 was reacted with ((6-
bromobenzo[d][1,3]dioxol-5-yl)methyl)hydrazine in trimethylamine, cyclized intermediate 58
was produced. Further oxidation with meta-chloroperoxybenzoic acid (mCPBA) followed by a
substitution reaction with ammonia produced the target compound 59. N-Benzylation of 60 with
5-bromo-6-(bromomethyl)benzo[d][1,3]dioxole yielded intermediate 61, followed by oxidation
with mCPBA and substitution with ammonia to produce the target compound 62. Additionally,
hydrolysis of 48 with aqueous NaOH solutions gave the target compound 63 (Scheme 3).
Scheme 2. Experimental Procedure for the Synthesis of 56, 59, and 62a
aReagents and condition: a) guanidinium carbonate, DMF, 120 °C, 8 h, 80%; b) hydrazine
monohydrate, EtOH, reflux, 12 h, 53%; c) 5-bromo-6-(bromomethyl)benzo[d][1,3]dioxole,
K2CO3, DMF, 60 °C, 4-6 h, 50-60%; d) ((6-bromobenzo[d][1,3]dioxol-5-yl)methyl)hydrazine,
TEA, DCM, 0 °C, 2 h, 35%; e) mCPBA, DCM, rt, 3 h, 40%; f) NH3, 120 oC, sealed tube, 2 h, 10-
15 %.
6

Scheme 3. Experimental Procedure for the Synthesis of 56, 60, and 64
aReagents and condition: a) NaOH, Dioxane, 100 °C, 5 days, 24%
RESULTS AND DISCUSSION
TRAP1 binding and cytotoxic activities of analogues
The binding affinity of the synthesized molecules to the ATP pocket of TRAP1 and Hsp90 was
analyzed with a fluorescence polarization assay using fluorescein isothiocyanate (FITC)-labeled
PU-H71.9 The half maximal inhibitory concentration (IC50) values of all tested compounds are
summarized in Tables 1-3. Aiming to primarily improve the TRAP1 inhibitory activity and basic
ADME, structural modification was intensively performed on the phenyl ring of 4 to determine
which steric and electronic features are crucial for cellular activity. The inhibitory activity of the
45 synthesized molecules against TRAP1 and Hsp90 was measured approximately to identify the
most potent compounds (Table 1). These potent molecules were re-evaluated to confirm their
cytotoxicity to cancer cells as well as their inhibitory activity toward TRAP1 and Hsp90 (Table
2).
To confirm steric tolerability on the phenyl ring, various functional groups were introduced into
the ortho-, meta-, and para-positions (R on the Scheme 1). Meta- or para- CF3O-substituted
compounds (18, 28, 29, 37, 38) showed limited potency toward TRAP1 (IC50 > 20 μM), whereas
MeO-, CF3-, Me-, Cl-, or Br- on the meta- or para-position of the phenyl ring was effective, as
7

44-48 and 50-52 were the most potent TRAP1 inhibitors. Additionally, ortho- and para- di-
substitution increased potency, with 41, 42, 48, 51, and 52 having excellent TRAP1 binding
affinity (IC50 = 0.3~2.4 μM for TRAP1).
Among the potent TRAP1 binders, compounds 47, and 48 were found to have very strong
cytotoxic activity with IC50 values of less than 10 μM (Table 2). IC50 values of 4, 47, and 48
were 0.56, >10, and 2.93 μM in FP assay, respectively, indicating the cytotoxic activity was not
related to the Grp94 inhibition. In contrast to the FP assay data, PU-H71 did not show significant
cytotoxic activity in this experimental condition due to insufficient mitochondrial accumulation
of the drug, as we previously reported.8,10 Similarly, compounds 41, 46, 49, 50, and 51, which
displayed good in vitro TRAP1 binding activity, did not show acceptable cytotoxic activity
(Table 2), suggesting deficient mitochondrial membrane penetration of the drugs. The membrane
permeability of small molecules via passive diffusion is highly correlated with the lipophilicity
of the molecules. It is possible that lipophilicity (cLogP) is an important physiochemical property
that affects the permeation of pyrazolo[3,4-d]pyrimidine derivatives into the mitochondria
membrane since compounds that are active in vitro but inactive in vivo generally have low
cLogP values as shown in Table 2-3.
Further modification of chlorine in the 4-position of the pyrazolo[3,4-d]pyrimidine to another
stable isostere18 resulted in the syntheses of 3,4-dimethyl-pyrazolo[3,4-d]pyrimidine (56), 4-
methyl-pyrazolo[3,4-d]pyrimidine (59), 4-methyl-pyrazolo[3,4-d]pyrimidinone (62), and 4-
hydroxy-pyrazolo[3,4-d]pyrimidine (63) (Table 3). However, the 4-methyl pyrazolo[3,4-
d]pyrimidine derivatives 62 and 63 lacked TRAP1-binding affinity. The weak TRAP1 binders,
56 and 59, showed no cytotoxicity to PC3 cancer cells, suggesting a lack of mitochondrial
membrane permeability.
8

Table 1. TRAP1 and Hsp90 binding activities of DN401 derivatives.
TRAP1 FP
IC50 (μM)
Hsp90 N FP
IC50 (μM)
TRAP1 FP
IC50 (μM)
Hsp90 N FP
IC50 (μM)
#
4
R
#
R
0.50
>20
3.5
6.0
24
20
4
>20
10
8
9
25
26
19
14
15
20
15
10
7.0
27
3.5
9.0
11
12
15
20
10
28
29
>20
>20
>20
>20
7.0
13
20
15
30
20
>20
14
15
>20
20
>20
20
31
32
>20
15
>20
>20
16
17
>20
>20
>20
15
33
34
20
>20
9.0
4.0
18
19
>20
>20
>20
>20
35
36
10
>20
>20
8.0
20
21
>20
>20
>20
>20
37
38
>20
>20
20
>20
22
23
6.5
5.0
20
39
40
3.0
>20
>20
>20
>20
IC50 values (μM) using fluorescence polarization (FP) assay were calculated from triplicated
experiments.
Table 2. TRAP1/Hsp90 binding and cytotoxic activities of selectedanalogues.
9

TRAP1 FP
IC50 (μM)
Hsp90 N FP
IC50 (μM)
PC3 cell MTT
IC50 (μM)
#
PU-H71
4
R
M.W.
513.4
382.6
338.2
cLogP
4.94
PSA
101.4
84.8
1.224 ± 0.096
0.464 ± 0.010
1.236 ± 0.063
0.560±0.099
1.702±0.007
0.639±0.014
> 20
1.44 ± 0.03
20
3.10
41
2.95
84.8
42
43
44
45
0.444 ± 0.011
2.431 ± 0.026
1.865 ± 0.075
1.696 ± 0.009
0.295±0.003
2.329±0.061
3.445±0.394
2.575±0.238
9.35 ± 0.04
11.96 ± 0.23
18.40 ± 0.50
15.71 ± 0.48
429.6
317.7
338.6
312.1
3.06
2.62
3.07
3.06
84.8
84.8
66.3
66.3
46
0.478 ± 0.004
0.406±0.003
> 20
325.7
2.40
75.6
47
48
1.983 ± 0.031
0.373 ± 0.004
0.368±0.038
0.298±0.004
0.94 ± 0.04
1.35 ± 0.02
357.7
356.6
3.25
3.21
75.6
66.3
49
50
0.887 ± 0.195
0.939 ± 0.010
3.361±0.074
0.995±0.056
> 20
20
313.7
291.7
2.56
2.85
66.3
66.3
51
52
0.789 ± 0.007
0.447 ± 0.007
1.711±0.244
0.710±0.032
20
325.7
291.7
2.60
2.85
75.6
66.3
15.67 ± 2.05
apKa value was estimated using Jaguar module in Schrodinger software. IC50 values (μM) for
fluorescence polarization (FR) assay and MTT assay (PC3 cells) were calculated from triplicated
experiments. The data are mean ± SEM.
Table 3. TRAP1 binding activities of DN401 derivatives.
TRAP1 FP
IC50 (μM)
Hsp90 FP
IC50 (μM)
PC3 cell MTT
IC50 (μM)
#
R
M.W.
cLogP
PSA
56
2.7
11
>20
376.2
3.15
84.8
10

59
62
1.0
0.50
>20
>20
>20
362.2
378.2
2.88
2.86
84.8
>20
101.5
63
>20
>20
>20
364.2
3.01
105.0
IC50 values (μM) for fluorescence polarization (FR) and MTT assay (PC3 cells) were calculated
from triplicated experiments. The data are mean.
X-ray co-crystal structures of the most promising compounds, 47 and 48, with TRAP1 were
determined and compared with our previous results to assess the SAR (Figure 2). As expected,
the N-5, N-7, and 6-amino of 47 formed water-mediated hydrogen bond networks with Asn119,
Asp158, and Thr251. The N-2 of both 47 and 48 were also directed to the amide of Asn171.
These polar interactions were observed only with TRAP1 and may enhance TRAP1-specific
binding, according to our previous study.10 4-trifluoromethyl-6-methoxy-phenyl (47) and 4-
bromo-6-fluoro-phenyl (48) appear to properly occupy the lipophilic binding pocket formed by
Met163, Ile168, Ile172, Phe205, Trp231, and Ile253, with additional π-πstacking with Phe205 as
shown in Figure 2. It is also possible that the para-substituted halogen on the phenyl ring of 47
or 48 interacts with Trp231 via halogen-πinteractions.19,20 As shown in Figure 3, side views of the
co-crystal structures suggest maximum hydrophobic interactions between the ortho- and para-di-
substituted phenyl in the ligands and TRAP1 as well as possible halogen-π interactions.
11

Figure 2. X-ray co-crystal structures of TRAP1 and its inhibitors. (A), (B), and (C) Views of
inhibitors 47 (PDB code: 7C04), 48 (PDB code: 7C05) and 4 (PDB 5Y3N) bound in the TRAP1
ATP binding site. Water (red spheres), TRAP1 (grey), and key residues (yellow). (D) 2D
schematic views of TRAP1-47 interaction.
12

Figure 3. Comparison phenyl rings of 47 (purple) and 48 (cyan) binding at a lipophilic binding
pocket. Halogens are located 3.5 Å away from the indole ring of Trp231.
Analyses of in vitro ADME
The basic in vitro ADME and Tox properties of compounds 47 and 48 were tested and
compared with those of the lead compound, 4. Microsomal (human, rat, and mouse) and plasma
(human and rat) stability, plasma protein binding (human and rat), CYP inhibition (1A2, 2C9,
2D6, and 3A4), and hERG binding were determined and the results are summarized in Table 4.
The lead compound 4 was rapidly metabolized by human, rat, and mouse microsomes; only ~5%
of compound 4 remained after 30 min of incubation with rat and mouse liver microsomes. When
compound 4 was incubated with human microsomes, ~64% of the molecule disappeared within
30 min. On the other hand, compounds 47 and 48 exhibited significantly improved microsomal
stability with human and rat microsomes as shown in Table 4. A plasma stability test showed all
tested compounds were stable with human and rat plasma. Evaluation of 47 and 48 plasma
protein binding in human and rat microsomes showed a high percentage of binding. Compounds
4, 47, and 48 are poor perpetrators of drug-drug interactions based on their minimal inhibition of
the most abundant CYP450 enzymes such as 1A2, 2C9, 2D6, and 3A4 and have low potential for
cardiotoxicity based on the hERG channel blocking assay.
13

Table 4. In vitro ADME properties of selectedcompounds 47 and 48
Microsomal stabilitya
(% remain after 30 min)
Plasma stability
(% remain after 30min)
Plasma Protein Binding
(% bound)
CYP inhibition assay
(% of control activity)
hERG
Entry
Human
Rat
Mouse
Human
Rat
Human
Rat
1A2 2C9 2D6 3A4 IC50 (μM)
Gamitrinib
-TPP
14.3
21.4
20.9
4
35.6
58.3
82.0
6.6
4.0
4.1
80.2
80.8
89.8
70.8
85.9
64.9
98.8
98.9
99.3
97.8
97.0
96.8
57.2 81.5 84.1 96.6
78.3 79.4 83.7 62.7
77.7 93.6 97.4 100
> 10
> 10
> 10
47
48
50.7
88.5
35.7
aMetabolic stability was determined as percentage of test compound remaining after incubation for 30 min
at 37 °C in liver-microsome preparations (CYP450 and other NADP-dependent enzymes).
TRAP1 and Hsp90 inhibition of 47 and 48
To examine the in vivo TRAP1 inhibitory activity, client degradation and concomitant
mitochondrial dysfunction were analyzed after drug treatment. Both 47 and 48 induced the
degradation of a TRAP1 client protein, succinate dehydrogenase subunit (SDHB), in PC3, but
PU-H71 did not (Figure 4A). Likewise, 47 and 48 increased the production of mitochondrial
ROS and reduced mitochondrial membrane potential (Figure 4B and C), indicating dysfunction
of the cancer mitochondria by TRAP1 inhibition.2, 5-6 Compounds 47 and 48 induced the
degradation of an Hsp90 client protein Akt without the induction of Hsp70, similar to 4, whereas
PU-H71 induced Hsp70 expression (Figure 4A). Overall, 47 showed the strongest TRAP1
inhibitory activity and mitochondrial dysfunction in cancer cells. Considering slightly weaker
binding of 47 with recombinant TRAP1 protein compared with 4 and 48 in FP assay (Table 2),
better in vivo anticancer activity of 47 is closely related to the improved mitochondrial
accumulation of the drug.
14

Figure 4. TRAP1 inhibitory activity. (A) TRAP1 client degradation. PC3 was incubated with the
drugs at a concentration of 10 μM for 6 hours and analyzed by western blot. (B) Mitochondrial
ROS production. Mito-Sox-labeled PC3 cells were treated with 10 μM of the drugs for 6 hours
and analyzed by flow cytometry. (C) Mitochondrial membrane potential. Tetramethylrhodamine
methyl ester-loaded PC3 cells were incubated with 20 μM of the drugs for 4 hours and analyzed
by flow cytometry.
Cytotoxic activity of 47 and 48
The cytotoxic activities of 47 and 48 were further examined in various cancer cell lines
including HeLa, MDA-MB-231-, SK-HEP-1, LN229, PC3, and H460, originating from the
human cervix, breast, liver, brain, prostate, and lung, respectively. Compounds 47 and 48
showed potent cytotoxic activities with IC50 values of 1~4 μM and PC3 was the most affected by
the drugs (IC50=0.939 and 1.347 μM, respectively; Table 2) among the cell lines examined
(Figure 5A). In contrast, the drugs were not toxic to normal mouse hepatocytes and normal
human prostate cells (Figure 5B). The data collectively indicated that 47 and 48 have
cytotoxicity specific to cancer cells but not to normal cells. Similar to the results for TRAP1
client degradation and mitochondrial dysfunction (Figure 4), 47 showed better cancer-specific
cytotoxic activity than 48 and 4 (Figure 5A).
15

Figure 5. Cytotoxic activities of 47 and 48. (A) Cytotoxic activities in cancer cell lines. Cancer
cells were incubated with the drugs as indicated for 24 hours and analyzed by MTT assay. (B)
Cytotoxicity to normal cells. Cells were incubated with the drugs as indicated for 24 hours and
analyzed by MTT assay.
In vivo efficacy of 47 and 48
Finally, we examined and compared the in vivo anticancer activities of 4, 47, and 48 using PC3
xenografts in nude mice. Daily intraperitoneal injection of the drugs reduced tumor growth
(Figure 6A-C) without mouse body weight loss (Figure 6D). Among the drugs, 47 showed
stronger tumor growth inhibition activity than 4 (Figure 6A-C), which is consistent with the
cytotoxicity data for the cancer cell lines (Figure 5A). Consistently, degradation of the TRAP1
client protein SDHB, inhibition of cell proliferation, and induction of cell death were strongest in
47-treated samples in vivo (Figure 6E-G; Figure 7A-C).
16

Figure 6. In vivo anticancer activity of 47 and 48. (A) Xenograft experiment. PC3-implanted
(subcutaneous) nude mice were treated with the compounds (B) Collected tumors. (C) Tumor
weight. (D) Mouse body weight. (E) SDHB expression. Collected tumors were analyzed by
western blotting. Measured band intensity was normalized with β-actin. (F) Suppression of cell
proliferation. Ki-67 staining. (G) Induction of cell death. TUNEL staining. ns, non-significant;
*p=0.0294; **p<0.0063; ***p<0.0002.
17

Figure 7. Analysis of collected tumors from xenograft experiment. (A) TRAP1 client degradation.
Collected tumors were analyzed by western blot. (B) Cell proliferation. Tumor specimens were
stained with anti-Ki-67 antibody and analyzed by a light microscope. (C) Cell death. Tumors
were stained by TUNEL method and analyzed by a fluorescence microscope. The data from A,
B, and C were quantified and are shown in Figure 6E, 6F, and 6G, respectively.
CONCLUSIONS
In summary, metabolically unstable compound 4, which was the initial lead mitochondrial
permeable TRAP1 inhibitor, was optimized to obtain metabolically stable drugs and further
18

improve the anticancer activity in vivo via the synthesis of various derivatives. The TRAP1
binding activity and ADME properties of the products were screened. Structure-based drug
design guided the optimization of potency, leading to the identification of 47 and 48 as potent
TRAP1 inhibitors with good metabolic and plasma stability as well as acceptable CYP and
hERG inhibition. X-ray co-crystal studies confirmed both 47 and 48 interact with the ATP
binding pocket in the TRAP1 protein. Compounds 47 and 48 demonstrated excellent anticancer
efficiency in cervix, breast, liver, brain, prostate, and lung cancer cells, with limited toxicity over
normal hepatocyte and prostate cells. Mouse PC3 xenograft studies showed compounds 47 and
48 reduced tumor growth without body weight loss.
EXPERIMENTAL SECTION
BIOLOGICAL EXPERIMENTAL SECTION
Cell lines and culture conditions: Human cells from Liver (SK-HEP-1), brain(LN229),
cervix(Hela), breast(MDA-MB-231), lung(NCI-H-460) and prostate (PC3) cancers were
purchased from the American Type Culture Collection (ATCC) and maintained as recommended
by the supplier. Cells were 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% CO2. Normal prostate cell RWPE-1 was purchased from the ATCC and
maintained as recommended by the supplier.
Primary hepatocyte isolation and culture: Primary hepatocytes were prepared from 8-week-
old C57BL/6 mice. Briefly, harvested livers from collagenase-perfused mice were dissected and
filtered by a 100-μm cell strainer (BD Biosciences). The collected cells were washed several
19

times with M199/EBSS medium (Hycolone), and hepatocytes were separated by gradient
centrifugation with Percoll (Sigma). The isolated hepatocytes were resuspended and maintained
in M199/EBSS medium containing 10% fetal bovine serum at 37°C in a humidified atmosphere
of 5% CO2.
Recombinant Protein Preparation and Fluorescence Polarization Assays: Recombinant
TRAP1, Grp94N and Hsp90αN proteins were prepared as described previously.9 For
fluorescence polarization assay, the fluorescence probe PU-H71-FITC was synthesized as
described previously.10 Purified recombinant TRAP1 (400nM) and Hsp90αN (100 nM) were
mixed with PU-H71-FITC (10 nM for TRAP1 and 100nM for Hsp90αN ) and incubated for 1hr
at room temperature with various concentrations of inhibitors in fluorescence polarization (FP)
buffer containing 1X PBS, 1 mM DTT, 2 mM MgCl2, 0.1 mg/mL BSA (pH 7.4). For FP analysis
using TRAP1 protein, 0.05% NP40 was added. Fluorescence polarization was finally measured
using a SYNERGY NEO microplate reader (BioTek Instruments, Inc.). To calculate the IC50
values, a linear inhibition curve was obtained for each compound.
ATPase activity assay: ATPase activity of TRAP1 is measured by using the PiColorLock Gold
Phosphate Detection Kit (InnovaBiosciences). Purified TRAP1 1uM was preincubated with
various concentration of inhibitors for 0.5 hours and incubated with 0.2mM ATP for 3 hours at
37 °C in ATPase activity assay buffer containing 100mM Tris, 20mM KCl, and 1mM CaCl2 (pH
7.4). Next, add 20μL mixture of Pi ColorLock Gold reagent and the accelerator (100:1) to each
sample of 100μL and incubate 5 min at RT. After incubation, 10μL stabilizer was added for
stopping color development. The absorbance was measured at 620nm using Microplate reader
(Synergy™ NEO, BioTek).
20

Measurement of reactive oxygen species (ROS), and mitochondrial membrane potential:
To measure mitochondrial superoxide and mitochondrial membrane potential, cells were labeled
with 200 nM Mito-SOX (Invitrogen) and 200 nM TMRM (Invitrogen) with Sytox (Invitrogen)
for 20 min. After washing with 1XPBS, cells were immediately analyzed by flow cytometry with
FACS Calibur™ system (BD Biosciences). Only cells with sytox positive were counted and
analyzed. The collected data were analyzed using FlowJo software (TreeStar).
Analysis of Cell Viability: Cell viability was analyzed using 3-(4,5-dimethyl-thyzozl-2-yl)-2,5-
diphenyltetrazolium bromide (MTT) assays. MTT were treated 48hrs after drug treatment. After
5 hour incubation, MTT was dissolved in dmso and the absorbance was measured using a
SYNERGY NEO microplate reader (BioTek Instruments, Inc.). Absorbance values were
normalized with control and the data was converted to percent viability and analyzed. The mean
IC50 is the concentration of an agent that reduces cell growth by 50% under the experimental
conditions and is the average of at least three independent reproducible statistically significant
measurements.
Western Blot Analysis: Cells were treated with inhibitors 10uM for 6hr. Cell lysates were
separated on 10% SDS-PAGE and transferred onto a PVDF membrane. The membranes were
blocked with with 10% skim milk in tris-buffered saline (TBS) containing 0.05% Tween-20
(TBST) for 1 hr at room temperature and incubated with primary antibodies overnight at 4°C.
The following antibodies are used. anti-TRAP1(BD Biosciences. 1/1000), anti-SDBH(Santa
Cruz. 1/200), anti-AKT(Santa Cruz. 1/200), anti-HSP70(BD Biosciences. 1/1000) and anti-
actin(MP. 1/3000). Membranes were washed three times with TBST for 5min and incubated with
the corresponding secondary antibodies diluted in 10% skim milk in TBST for 1 hr at room
21

temperature. Membranes were washed three times with TBST for 5min at room temperature and
visualized using an enhanced chemiluminescence (ECL) detection kit (BioRad).
In vivo xenograft model: All in vivo experiment were approved by UNIST (UNISTIACUC-17-
13). 1X107 (PC3) cells in 100ul of 1X PBS were injected subcutaneously into both flanks of 7-
week-old male nude mice (Hyochang science). When tumor size reached approximately 100nm3,
vehicle (DMSO) and drugs (30 mg/kg) with 25% cremophor EL (Sigma) and 65% 1X PBS were
injected intraperitoneally every day. Tumor volume was measured daily using electronic calipers
and calculated according to width2 x length x 0.5.
Immunohistochemistry: To analyze cell proliferation in xenograft model, tumor were harvested
and fixed in 10% formalin at 4°C. Formalin-fixed paraffin-embedded tumor samples were
sectioned for histological analysis. After paraffin removal, sections were incubated in 3%
hydrogen peroxide to block endogenous peroxidase. Sections were incubated in antigen retrieval
in citrate buffer and following experiment was performed using IHC kit (Thrermo). In brief, after
permeabilization and blocking, sections were incubated overnight at 4°C with anti-Ki67 antibody
(abcam, 1/2000). Staining were visualized by DAB chromogen and analyzed using virtual
microscope (Olympus). The percent of cells positive for Ki67 per fields was counted.
TUNEL assay: TUNEL staining was performed using in situ cell death detection kit (Roche)
following the manufacturer’s protocol. For TUNEL scoring, stained slides were analyzed using
inverted fluorescence microscope (Olympus). The average numbers of TUNEL positive nuclei
were calculated in five fields (X40) selected from a central region of tumor, not including
necrosis region.
Crystallization and Structure Determination: Protein production for the crystallographic
studies of human TRAP1 N terminal domain (residues 60-294, hTRAP1-N) and TRAP1 lacking
22

C terminal domain (residues 60-561, hTRAP1-NM) was performed as described previously (9).
To crystallize TRAP1 complexed with 47, 50 L of hTRAP1-N protein (20 mg/mL) solution
was mixed with 1.6 L of inhibitor dissolved in DMSO (50 mM), and incubated for 1 hr on ice.
The protein-inhibitor mixture was diluted with buffer containing 20 mM Tris pH 7.5, 150 mM
NaCl, and 5mM DTT to a final volume of 90 L to reduce the DMSO concentration. Crystals
were grown using hanging drop diffusion method at room temperature by mixing 1 L of protein
solution and 1 L of reservoir buffer comprising 120 mM citrate (pH 5.6), 31% PEG 4,000.
Crystals were cryoprotected in a crystallization solution plus 30% glycerol, and were flash frozen
in liquid nitrogen. For hTRAP1-NM in complex with 48, the protein-inhibitor mixture was
prepared as described as above. Crystals of TRAP1-48 complex were grown in well solution
containing 100 mM sodium cacodylate pH 6.76, 100 mM calcium acetate, and 14% PEG 20K at
18°C, and were cryoprotected with 30% glycerol. X-ray diffraction data were collected at the 7A
beamline and 11C beamline of Pohang Accelerator Laboratory (PAL), and were processed by
HKL2000.21 The structures of TRAP1-47 and TRAP1-48 were solved by molecular replacement
with Phaser-MR program22 using the apo structures (PDB: 4Z1I and 5Y3N) as a model. Final
models of TRAP1-47 and TRAP1-48 were refined to 1.70 Å and 2.6 Å resolution, respectively,
via iterative rounds of refinement and rebuilding using Phenix and COOT.23 Crystallographic
statistics are summarized in Table S1. The atomic coordinates and crystallographic structure
factors have been deposited in the Protein Data Bank under the following accession codes: 7C04
and 7C05 for Trap1 in complex with 47 and 48, respectively.
CHEMICAL EXPERIMENTAL SECTION
23

Synthetic Methods and Molecular Characterization: Starting materials were purchased
from Sigma-Aldrich, TCI, Alfa-Aesar, and Enamine and were used without further purification.
Anhydrous solvents were obtained by passage through a solvent purification system composed of
activated alumina and a supported copper redox catalyst. All reactions were performed under an
atmosphere of dry nitrogen. A Teledyne ISCO flash purification system with various prepacked
1
13
silica gel cartridges was used for flash column chromatography. H and C NMR spectra were
1
13
recorded 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 downfield from TMS
(δ 0.0) as the internal standard in DMSO-d6 or CDCl3. MS was performed on a system consisting
of an electrospray ionization (ESI) source in
a
Shimadzu LCMS-2020 liquid
chromatography−mass spectrometer (column: Shim-pack GIS, 100 × 3.0 mm, 3 μm ODS). The
purity of the compounds was evaluated on a Shimadzu reverse-phase analytical HPLC system
(column: Ace C18, 150 × 4.6 mm, 3 μm). Purities of all compounds that were subjected to
biological assay were >95%.
2-Amino-4,6-dichloropyrimidine-5-carbaldehyde (6). N,N-dimethylformamide (90 mmol,
6.61 g) was added dropwise to a solution of phosphorus oxychloride (157 mmol, 14.7 ml) over a
20 min at 0 °C. After adding 2-Amino-4,6-dichloropyrimidine-5-carbaldehyde (39.3 mmol, 5 g)
slowly, the mixture was refluxed for 4h and then concentrated in vacuo. The residue was
carefully added to iced water and placed for 12 hours at room temperature. The resulting crystal
was filtered, washed with water and ethyl acetate, and dried in vacuo to give the target product in
60% yield; 1H NMR (400 MHz, DMSO-d6) δ 10.06 (s, 1H), 8.49 (s, 2H); 13C NMR (100 MHz,
DMSO-d6) δ 185.17, 163.62, 162.14, 113.38; ESI-MS m/z: 169.2 [M+H]+.
4-Chloro-1H-pyrazolo[3,4-d]pyrimidin-6-amine (7). To
a
solution of 2-amino-4,6-
dichloropyrimidine-5-carbaldehyde (10.42 mmol, 2 g) in THF (50 ml) and water (50 ml) was
24

added hydrazine monohydrate (11.46 mmol, 0.56 ml) and Triethylamine (41.7 mmol, 5.82 ml).
After stirring for 15 h at room temperature, organic materials were extracted with Ethyl acetate
(50ml X 3). The combined organic layer was washed with brine and dried over Na2SO4. After
solvents removal in vacuo, the residue was crystallized in dichloromethane and hexane co-
solvent system (DCM:hex=1:10) to give yellow product (58%); 1H NMR (400 MHz, DMSO-d6)
δ 13.24 (s, 1H), 7.94 (s, 1H), 7.13 (s, 1H); ESI-MS m/z: 169.2 [M+H]+.
General procedure I for pyrazolopyrimidine derivatives (8-52). Various benzyl halide was
added to a mixture of 4-chloro-1H-pyrazolo[3,4-d]pyrimidin-6-amine (7, 0.59 mmol, 0.1 g),
K2CO3 (1.18 mmol, 0.16 g) in DMF (5 ml). After stirring at 60 oC for appropriate time (generally
4-8 h), the reaction mixture was cooled, quenched with saturated ammonium chloride solution
(50 ml), and extracted with ethyl acetate (50 ml X 3). The combined organic layer was washed
with brine, dried over Na2SO4, and concentrated in vacuo. The crude products were purified by
column chromatography using silica gel to afford the target products.
4-Chloro-1-(3-methoxybenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (8). General procedure
1
I was applied. Yield: 11%, H NMR (400 MHz, CDCl3) δ 7.90 (s, 1H), 7.23 (t, J = 7.8 Hz, 1H),
6.87 (d, J = 7.7 Hz, 1H), 6.85–6.78 (m, 2H), 5.41 (s, 2H), 5.31 (s, 2H), 3.77 (s, 3H); ESI-MS
m/z: 290.0 [M+H]+.
4-Chloro-1-(3-methylbenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (9). General procedure I
1
was applied. Yield: 29%, H NMR (400 MHz, CDCl3) δ 7.90 (s, 1H), 7.23–7.17 (m, 1H), 7.09
(d, J = 7.1 Hz, 3H), 5.41 (s, 2H), 5.40 (s, 2H), 2.31 (s, 3H); ESI-MS m/z: 274.1 [M+H]+.
25

4-Chloro-1-(4-methylbenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (10). General procedure
1
I was applied. Yield: 16%, H NMR (400 MHz, CDCl3) δ 7.89 (s, 1H), 7.21 (d, J = 7.9 Hz, 2H),
7.12 (d, J = 7.7 Hz, 2H), 5.39 (s, 2H), 5.28 (s, 2H), 2.31 (s, 3H); ESI-MS m/z: 274.1 [M+H]+.
1-(2-Bromo-5-methoxybenzyl)-4-chloro-1H-pyrazolo[3,4-d]pyrimidin-6-amine (11). General
procedure I was applied. Yield: 11%, 1H NMR (400 MHz, CDCl3) δ 7.97 (s, 1H), 7.47 (d, J = 8.8
Hz, 1H), 6.70 (dd, J = 8.8, 3.0 Hz, 1H), 6.35 (d, J = 2.9 Hz, 1H), 5.50 (s, 2H), 5.31 (s, 2H), 3.68
(s, 3H); ESI-MS m/z: 369.9 [M+H]+.
1-(2-Bromobenzyl)-4-chloro-1H-pyrazolo[3,4-d]pyrimidin-6-amine (12). General procedure I
1
was applied. Yield: 17%, H NMR (400 MHz, CDCl3) δ 7.97 (s, 1H), 7.59 (dd, J = 7.8, 1.2 Hz,
1H), 7.22 (td, J = 7.5, 1.2 Hz, 1H), 7.15 (td, J = 7.7, 1.7 Hz, 1H), 6.80 (d, J = 7.6 Hz, 1H), 5.55
(s, 2H), 5.29 (s, 2H); ESI-MS m/z: 337.9 [M+H]+.
4-Chloro-1-(2-chlorobenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (13). General procedure I
1
was applied. Yield: 15%, H NMR (400 MHz, CDCl3) δ 7.96 (s, 1H), 7.40 (dd, J = 7.9, 1.2 Hz,
1H), 7.23 (dd, J = 7.8, 1.7 Hz, 1H), 7.17 (td, J = 7.5, 1.3 Hz, 1H), 6.91–6.85 (m, 1H), 5.58 (s,
2H), 5.26 (s, 2H); ESI-MS m/z: 294.0 [M+H]+.
4-Chloro-1-(3,4-difluorobenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine
(14).
General
procedure I was applied. Yield: 10%, 1H NMR (400 MHz, MeOD) δ 7.83 (s, 1H), 7.11 (ddd, J =
16.6, 9.1, 5.3 Hz, 2H), 7.04–6.92 (m, 1H), 5.30 (s, 2H); ESI-MS m/z: 296.0 [M+H]+.
4-Chloro-1-(4-vinylbenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (15). General procedure I
1
was applied. Yield: 12%, H NMR (400 MHz, CDCl3) δ 7.90 (s, 1H), 7.35 (d, J = 8.2 Hz, 2H),
7.30–7.22 (m, 2H), 6.67 (dd, J = 17.6, 10.9 Hz, 1H), 5.75–5.65 (m, 1H), 5.42 (s, 2H), 5.33 (s,
2H), 5.23 (d, J = 10.9 Hz, 1H); ESI-MS m/z: 286.1 [M+H]+.
26

3-((6-Amino-4-chloro-1H-pyrazolo[3,4-d]pyrimidin-1-yl)methyl)benzonitrile (16). General
1
procedure I was applied. Yield: 15%, H NMR (400 MHz, CDCl3) δ 7.92 (s, 1H), 7.56 (d, J =
15.5 Hz, 3H), 7.44 (t, J = 7.8 Hz, 1H), 5.46 (s, 2H), 5.28 (s, 2H); ESI-MS m/z: 285.1 [M+H]+.
4-Chloro-1-(4-(trifluoromethyl)benzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (17). General
procedure I was applied. Yield: 18%, 1H NMR (400 MHz, CDCl3) δ 7.92 (s, 1H), 7.58 (d, J = 8.0
Hz, 2H), 7.40 (d, J = 7.9 Hz, 2H), 5.49 (s, 2H), 5.37 (s, 2H); ESI-MS m/z: 328.0 [M+H]+.
4-Chloro-1-(4-(trifluoromethoxy)benzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine
(18).
1
General procedure I was applied. Yield: 13%, H NMR (400 MHz, CDCl3) δ 7.91 (s, 1H), 7.34
(d, J = 8.0 Hz, 2H), 7.16 (d, J = 8.0 Hz, 2H), 5.43 (s, 2H), 5.24 (s, 2H); ESI-MS m/z: 344.0
[M+H]+.
1-(2-Bromo-3-fluorobenzyl)-4-chloro-1H-pyrazolo[3,4-d]pyrimidin-6-amine (19). General
1
procedure I was applied. Yield: 15%, H NMR (400 MHz, CDCl3) δ 7.96 (s, 1H), 7.19 (dd, J =
13.3, 7.9 Hz, 1H), 7.06 (t, J = 7.7 Hz, 1H), 6.60 (d, J = 7.7 Hz, 1H), 5.57 (s, 2H), 5.24 (s, 2H);
ESI-MS m/z: 355.9 [M+H]+.
4-Chloro-1-(2-chloro-4-fluorobenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (20). General
procedure I was applied. Yield: 18%, 1H NMR (400 MHz, CDCl3) δ 7.94 (s, 1H), 7.16 (d, J = 6.5
Hz, 1H), 7.00–6.86 (m, 2H), 5.52 (s, 2H), 5.29 (s, 2H); ESI-MS m/z: 312.0 [M+H]+.
4-Chloro-1-(5-fluoro-2-methoxybenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (21). General
1
procedure I was applied. Yield: 15%, H NMR (400 MHz, CDCl3) δ 7.94 (s, 1H), 6.92 (td, J =
8.5, 3.0 Hz, 1H), 6.81 (dd, J = 9.0, 4.3 Hz, 1H), 6.55 (dd, J = 8.8, 2.9 Hz, 1H), 5.46 (s, 2H), 5.25
(s, 2H), 3.84 (s, 3H); ESI-MS m/z: 308.0 [M+H]+.
27

4-Chloro-1-(2-fluoro-5-methylbenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (22). General
1
procedure I was applied. Yield: 16%, H NMR (400 MHz, CDCl3) δ 7.92 (s, 1H), 7.09–7.02 (m,
1H), 6.96 (d, J = 9.1 Hz, 1H), 6.94–6.83 (m, 1H), 5.46 (s, 2H), 5.37 (s, 2H), 2.24 (s, 3H); ESI-
MS m/z: 292.1 [M+H]+.
4-Chloro-1-(4-(methylthio)benzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (23). General
procedure I was applied. Yield: 15%, 1H NMR (400 MHz, CDCl3) δ 7.89 (s, 1H), 7.24 (d, J = 8.4
Hz, 2H), 7.19 (d, J = 8.4 Hz, 2H), 5.39 (s, 2H), 5.28 (s, 2H), 2.44 (s, 3H); ESI-MS m/z: 306.0
[M+H]+.
4-Chloro-1-(2,4-difluorobenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine
(24).
General
1
procedure I was applied. Yield: 9%, H NMR (400 MHz, CDCl3) δ 7.90 (s, 1H), 7.13-7.19 (m,
1H), 6.78-6.85 (m, 2H), 5.45 (s, 2H), 5.28 (s, 2H); ESI-MS m/z: 295.7 [M+H]+.
4-Chloro-1-(2,3-difluoro-4-methoxybenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine
(25).
1
General procedure I was applied. Yield: 20%, H NMR (400 MHz, CDCl3) δ 7.90 (s, 1H), 6.94–
6.86 (m, 1H), 6.71–6.64 (m, 1H), 5.45 (s, 2H), 5.28 (d, J = 12.1 Hz, 2H), 3.87 (s, 3H); ESI-MS
m/z: 326.0 [M+H]+.
4-Chloro-1-(2,4-difluoro-3-methoxybenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine
(26).
1
General procedure I was applied. Yield: 10%, H NMR (400 MHz, CDCl3) δ 7.91 (s, 1H), 6.77-
6.85 (m, 2H), 5.45 (s, 2H), 5.30 (s, 2H), 3.99 (s, 3H); ESI-MS m/z: 325.7 [M+H]+.
4-Chloro-1-(2,3,5,6-tetrafluorobenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (27). General
1
procedure I was applied. Yield: 10%, H NMR (400 MHz, CDCl3) δ 7.88 (s, 1H), 7.03-7.26 (m,
1H), 5.54 (s, 2H), 5.39 (s, 2H); ESI-MS m/z: 331.9 [M+H]+.
28

4-Chloro-1-(2-methoxy-5-(trifluoromethoxy)benzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine
1
(28). General procedure I was applied. Yield: 13%, H NMR (400 MHz, CDCl3) δ 7.94 (s, 1H),
7.11 (d, J = 8.0 Hz, 1H), 6.85 (d, J = 12.0 Hz, 1H), 6.76 (s, 1H), 5.46 (s, 2H), 5.33 (s, 2H), 3.85
(s, 3H); ESI-MS m/z: 373.9 [M+H]+.
4-Chloro-1-(2-fluoro-5-(trifluoromethoxy)benzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine
1
(29). General procedure I was applied. Yield: 19%, H NMR (400 MHz, CDCl3) δ 7.93 (s, 1H),
7.12 (d, J = 4.0 Hz, 2H), 7.02 (dd, J = 4.0, 2.4 Hz, 1H), 5.49 (s, 2H), 5.36 (s, 2H); ESI-MS m/z:
361.8 [M+H]+.
4-Chloro-1-(2-methoxy-6-(trifluoromethyl)benzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine
1
(30). General procedure I was applied. Yield: 12%, H NMR (400 MHz, CDCl3) δ 7.80 (s, 1H),
7.44 (t, J = 8.0 Hz, 1H), 7.35 (d, J = 8.0 Hz, 1H), 7.08 (d, J = 8.0 Hz, 1H), 5.55 (s, 2H), 5.44 (s,
2H), 3.69 (s, 3H); ESI-MS m/z: 357.9 [M+H]+.
4-Chloro-1-(2-fluorobenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (31). General procedure I
was applied. Yield: 18%, 1H NMR (400 MHz, CDCl3) δ 7.91 (s, 1H), 7.05-7.14 (m, 4H), 5.51 (s,
2H), 5.25 (s, 2H); ESI-MS m/z: 277.9 [M+H]+.
4-Chloro-1-(2-fluoro-4-(trifluoromethyl)benzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (32).
1
General procedure I was applied. Yield: 12%, H NMR (400 MHz, CDCl3) δ 7.93 (s, 1H), 7.55-
7.58 (m, 1H), 7.47 (dd, J = 4.0, 1.8 Hz, 1H), 7.20 (t, J = 8.0 Hz, 1H), 5.53 (s, 2H), 5.27 (s, 2H);
ESI-MS m/z: 345.9 [M+H]+.
4-Chloro-1-(2,3,4,5-tetrafluorobenzyl)-1H-pyrazolo[3,4-d]pyrimidin-6-amine (33). General
1
procedure I was applied. Yield: 18%, H NMR (400 MHz, CDCl3) δ 7.93 (s, 1H), 6.80 (ddd, J =
13.2, 9.3, 2.5 Hz, 1H), 5.47 (s, 2H), 5.28 (s, 2H); ESI-MS m/z: 331.9 [M+H]+.
29