A novel Hsp70 inhibitor specifically targeting the cancer-related Hsp70-Bim protein-protein interaction

Article abstract of DOI:10.1016/j.ejmech.2021.113452

Targeting cancer-related Hsp70-Bim protein-protein interactions (PPIs) offers a new strategy for the design of Hsp70 inhibitors. Herein, we discovered a novel Hsp70 inhibitor, S1g-6, based on the established BH3 mimetics. S1g-6 exhibited sub-μM binding af

Full text of DOI:10.1016/j.ejmech.2021.113452

European Journal of Medicinal Chemistry 220 (2021) 113452
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
A novel Hsp70 inhibitor specifically targeting the cancer-related
Hsp70-Bim protein-protein interaction
Ziqian Wang 1, Ting Song a
a
,
,
**
,
1
, Zongwei Guo , Laura B. Uwituze , Yafei Guo ,
b
a
b
b
b
a
a
a
a
Hong Zhang , Hang Wang , Xiaodong Zhang , Hao Pan , Tong Ji , Fangkui Yin ,
a
a
a, *
Sheng Zhou , Jian Dai , Zhichao Zhang
a
State Key Laboratory of Fine Chemicals, Zhang Dayu School of Chemistry, Dalian University of Technology, Dalian, Liaoning, 116024, China
School of Bioengineering, Dalian University of Technology, Dalian, Liaoning, 116024, China
b
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 7 December 2020
Received in revised form
Targeting cancer-related Hsp70-Bim protein-protein interactions (PPIs) offers a new strategy for the
design of Hsp70 inhibitors. Herein, we discovered a novel Hsp70 inhibitor, S1g-6, based on the estab-
lished BH3 mimetics. S1g-6 exhibited sub-mM binding affinity toward Hsp70 and selectively disrupted
Hsp70-Bim PPI. The target specificity of S1g-6 in situ was validated by affinity-based protein profiling, co-
immunoprecipitation, and cell-based shRNA assays. S1g-6 specifically antagonized the ATPase activity of
Hsp70 upon recruiting Bim and showed selective apoptosis induction in some cancer cell lines over
normal ones through suppression of some oncogenic clients of Hsp70, representing a new class of
antitumor candidates. Hsp70-Bim PPI exhibited cancer-dependent role as a potential anti-cancer target.
29 March 2021
Accepted 2 April 2021
Available online 14 April 2021
Keywords:
Hsp70 inhibitor
Hsp70-bim PPI
©
2021 Elsevier Masson SAS. All rights reserved.
Cancer-related target
Antitumor candidate
1. Introduction
Drug Association.
The critical roles of Hsp70 and its homologs in cancer-related
molecular mechanisms are thought to be mediated by protein-
protein interactions (PPIs) between Hsp70 and a great amount of
proteins, such as co-chaperones DNAJ and Bag-3, apoptosis-related
proteins Bax, p53, caspase 3 and APAF-1, multiple co-chaperones
Hsp90 and HOP, and so on [1e4]. Several Hsp70 inhibitors that
disrupt these PPIs (Fig. 1a), including JG-98/JG-231 (for Bag-3)
[12,13], MAL3-101 (for DNAJ class proteins) [14], YK-5 (for Hsp90
and Hop) [15], BT-44 (for caspase 3) [16], apoptozole (for APAF-1)
[17] and AEAC (for oxidized GAPDH) [18], were developed to
investigate their cancer-related molecular mechanisms. In this re-
gard, specifically targeting cancer-related Hsp70-involved PPIs is an
attractive strategy for Hsp70 inhibitor development, which is
thought to reduce toxicity towards normal cells. For example, JG-98
and JG-231, exhibited obvious selectivity for killing breast cancer
cells over normal cells by specifically targeting the cancer-related
Hsp70-Bag3 PPI [12,13,19].
Heat shock protein 70 (Hsp70) is an important ATP-dependent
molecular chaperone that is abundantly expressed in many ma-
lignant tumors of various origins [1e4]. Hsp70 assists in the correct
folding of many oncogenic proteins, playing critical roles in the
growth, proliferation and apoptosis evasion of cancer cells and
protects cancers from stress resulting from intrinsically misfolded
proteins and extrinsic stimuli, including chemotherapy [1e4].
However, Hsp70 has also been reported as an important facilitator
of many other cellular functions, both cancer-dependent and
otherwise, and the expression level of Hsp70 exhibited no signifi-
cant differences in cancer and normal cells, expressing doubts
about the effectiveness of Hsp70 as a specific target for cancer
therapy [5e8]. Therefore, the clinical application of Hsp70 in-
hibitors is often limited by the toxicity towards normal cells [9e11].
None of them have been clinically approved by the US Food and
Most recently, our work revealed that Bim, a BH3-only member
of the Bcl-2 family proteins, serves as a co-chaperone for Hsp70,
which modulates the folding and stabilization of many Hsp70
oncogenic substrates. In some chronic myeloid leukemia (CML) cell
lines, Bim forms PPI with Hsp70, which facilitates the refolding and
*
Corresponding author.
** Corresponding author.
E-mail addresses: songting@dlut.edu.cn (T. Song), zczhang@dlut.edu.cn
(
Z. Zhang).
1
These authors contributed equally to this work.
https://doi.org/10.1016/j.ejmech.2021.113452
223-5234/© 2021 Elsevier Masson SAS. All rights reserved.
0

Z. Wang, T. Song, Z. Guo et al.
European Journal of Medicinal Chemistry 220 (2021) 113452
Fig. 1. (a) The chemical structure of BH3-mimetic S1. (b) The chemical structure of S1~S1i and S1g-1~S1g-6. (c) The chemical structure of ABT-199, VER-155008 and MKT-077.
stabilization of many oncogenic proteins in downstream pathways
of the fusion tyrosine kinase BCR-ABL, such as AKT and Raf-1, and
subsequently prevent these cells from apoptosis [20].
Herein, through derivation, screening, and optimization of S1, a
Bcl-2 inhibitor developed by our group [21,22], we discovered a
compound that selectively fits the binding surface of Hsp70-Bim
PPI over that of Bcl-2-Bim PPI. These compounds were synthe-
sized via the route published by our group from acenaph-
thoquinone [21,22] (Scheme 1).
novel Hsp70 inhibitor, S1g-6, which exhibited sub-mM binding af-
finity toward Hsp70 and selectively disrupted Hsp70-Bim PPI both
in vitro and in situ. Compared with two well-known Hsp70 in-
hibitors, VER-155008 [23] and MKT-077 [24], that inhibit the
ATPase activity of Hsp70 via competitive binding and allosteric
effect, respectively, S1g-6 exhibited stronger apoptosis induction in
some CML cell lines and showed much lower cytotoxicity to normal
cells, making it a completely new class of Hsp70 inhibitor and a
promising antitumor candidate. With the help of S1g-6, Hsp70-Bim
PPI could be verified as a cancer-related target.
2.2. Structure-activity relationship analysis
The capability of S1~S1i and S1g-1~S1g-6 to disrupt Hsp70-Bim
and Bcl-2-Bim PPIs was detected via fluorescence polarization as-
says (FPAs) using Fam-labeled BimBH3 peptide and recombinant
Hsp70 NBD (residues 1e383) and Bcl-2 (residues 2e206) protein.
The well-known Hsp70 inhibitors, VER-155008 [23] and MKT-077
[
24] (Fig. 1c), which binds distal from and adjacent to Bim-binding
site, respectively, and the Bcl-2 inhibitor ABT-199 [25], were tested
in parallel (Fig. 1c). As shown in Table 1, ABT-199 and S1 only have
the capability to disrupt Bcl-2-Bim but not the Hsp70-Bim complex
2
. Results and discussion
(IC50 > 20
mM). S1 derivatives with a thiophenol or cycloalkyl thiol
group as R
1
exhibited a gain of function to disrupt the Hsp70-Bim
2
.1. Rational design and synthesis
complex (S1a-S1d or S1f-S1g), while compounds bearing amino
substituents as R showed weak potency in disturbing both com-
plexes (S1e, S1h and S1i). Among S1a-S1g, compound S1g with a
cyclohexanethiol group as R exhibited a M range of potency for
disrupting Hsp70-Bim (IC50 ¼ 1.7 M) and a 5-fold selectivity to
Hsp70-Bim over Bcl-2/Bim. Displacing the R group of S1g with
hydrophilic substituents, which led to S1g-4, S1g-5 and S1g-6,
compared with hydrophobic substituents S1g-1, S1g-2 and S1g-3,
further promoted the potency and selectivity to disrupt Hsp70-Bim
over Bcl-2/Bim complex. Among them, S1g-6 showed the best
1
In our previous work, we revealed that Hsp70-Bim PPI is
mediated by the nucleotide-binding domain (NBD) of Hsp70 and
the Bcl-2 homology 3 (BH3) domain of Bim [20]. TROSY-HSQC data
demonstrated that BimBH3 domain inserts into a hydrophobic
groove on the surface of the NBD domain of Hsp70. Moreover,
residues L62 and D67 on BimBH3 domain, two key residues for Bcl-
1
m
m
2
2
-Bim PPI, were also found to play critical roles on the Hsp70-Bim
binding interface, indicating that the binding interface of Hsp70-
Bim showed high similarity to that of Bcl-2/Bim. Therefore, we
chose our previously reported BH3 mimetic (Bcl-2 inhibitor) S1
potency (IC50 ¼ 0.45
mM) and selectivity (ratio > 40) to target
Hsp70-Bim over Bcl-2/Bim. VER-155008 shows no detetable
competition to Hsp70-Bim complex, while MKT-077 exhibited a 5-
fold weaker disruption than S1g-6. The direct binding of S1g-6 to
Hsp70 was further verified by isothermal titration calorimetry
[
21,22], in which the 6-thiamorpholine and 1-carbonyl/2-cyano
groups mimic the binding features of L62 and D67 on Bim,
respectively, as the starting point for developing new Hsp70 in-
hibitors that disrupt the Hsp70-Bim PPI (Fig. 1a).
(ITC), showing dissociation constant K
d
value of 0.96 mM (Fig. 2a).
Herein, we established a library of S1-analogs by replacing the
-thiamorpholine group and the 3-cyano group with a variety of
Compared with the K M of MKT-077, a consistent binding
potency with that determined by FPA illustrated the stronger direct
d
¼ 1.78
m
6
different substituent groups (R and R , Fig. 1b), in order to screen a
1
2
binding and disruption of Hsp70-Bim PPIs of S1g-6 than MKT-077.
2

Z. Wang, T. Song, Z. Guo et al.
European Journal of Medicinal Chemistry 220 (2021) 113452
ꢀ ꢀ ꢀ
3 2 3 3 1 3 2 3
Scheme 1. Synthetic route of S1-analogies. Reagents and conditions: (a) malononitrile, CH CN, 65 C, 3 h; (b) K CO , CH CN, 65 C, 3 h; (c) R H, CH CN, 60 C, 3 h; (d) R H, CH CN,
rt, 2 h.
Table 1
The IC50 value of S1~S1i and S1g-1~S1g-6 for disrupting Hsp70-Bim PPI and Bcl-2-Bim PPI determined by FPA.
Compounds
Hsp70-Bim IC50
(m
M)
Bcl-2/Bim IC50
(
m
M)
Compounds
Hsp70-Bim IC50
(
mM)
Bcl-2/Bim IC50 (mM)
S1
>20
0.8 ± 0.2
1.8 ± 0.6
0.7 ± 0.1
2.8 ± 0.4
6.1 ± 0.3
>20
8.4 ± 1.2
8.6 ± 2.4
14.8 ± 3.1
>20
S1g-1
S1g-2
S1g-3
S1g-4
S1g-5
S1g-6
ABT-199
VER-155008
MKT-077
10.5 ± 0.1
8.4 ± 0.5
13.2 ± 1.8
0.81 ± 0.27
1.0 ± 0.3
0.45 ± 0.10
>20
5.2 ± 0.6
4.8 ± 0.4
16.0 ± 2.7
9.6 ± 1.4
13.5 ± 1.5
18.7 ± 2.3
<0.1
S1a
S1b
S1c
S1d
S1e
S1f
S1g
S1h
S1i
7.7 ± 1.4
2.5 ± 0.5
4.4 ± 1.2
4.5 ± 0.5
15.2 ± 2.7
3.2 ± 0.3
1.7 ± 0.5
18.5 ± 1.9
>20
>20
2.7 ± 0.6
>20
>20
Fig. 2. (a) Binding affinity of S1g-6 and MKT-077 to Hsp70 determined by ITC. Data are from three independent experiments. (b) ATPase activity of Hsp70 treated with BimBH3,
S1g-6, VER-155008 and MKT-077. Error bars show the mean ± s.e.m. (n ¼ 3).
Consequently, the single-turnover ATPase rates of Hsp70 upon
the addition of S1g-6 were detected, using VER-155008 [23] and
MKT-077 [24] as control (Fig. 2b). The rate of ATPase activity of
Hsp70 was significantly increased (4.8 fold) by synthetical BimBH3
peptide, which is consistent with the results in our published work.
Notably, S1g-6 significantly suppressed the BimBH3-stimulated
ATPase activity, with no obvious influence on basal ATPase activ-
ity. On the contrary, VER-155008 and MKT-077 indiscriminately
suppressed the ATPase activity of Hsp70 in the presence and
absence of BimBH3 peptide. These data verified that S1g-6 selec-
tively inhibits the co-chaperone function of Bim to activate the
ATPase activity of Hsp70, rather than directly suppressing the
overall ATPase activity of Hsp70 like the ATP-competitive inhibitor
VER-155008 or allosteric ATPase activity inhibitor MKT-077. .
2.3. Binding mode evaluation of S1g-6
A docking simulation was carried out to predict the binding
mode of S1g-6 in complex of Hsp70. As shown in the top-ranked
binding mode (Fig. 3), S1g-6 could be well-positioned into the
hydrophobic Bim-binding groove on the surface of the nucleotide-
binding domain (NBD) of Hsp70. The cyclohexylthiol group (R ) of
1
S1g-6 occupied the deepest hydrophobic pocket at the bottom of
the groove, which is consistent with the fact that compounds with
1
hydrophobic thiophenol or cycloalkyl thiol group as R showed
better binding affinity than others (Table 1). The 2-carboxyl group
located adjacent to residues T226 and H227 on Hsp70 NBD domain,
the key residues for Bim binding [20], may form H-bonds with
2
them (distance < 3.5 Å). The morpholine group (R ) of S1g-6
3

Z. Wang, T. Song, Z. Guo et al.
European Journal of Medicinal Chemistry 220 (2021) 113452
Fig. 3. Predicted binding models of S1g-6 in complex with Hsp70 (PDB:5AZQ) by docking simulation. Residues with significant chemical shift or peak intensity changes upon Bim
binding in TROSY-HSQC assays [20] were shown in pink color.
pointed to the side chain -COOH group of D206 in Hsp70,
specific target of S1g-6 in living cells and S1g-6 exhibits similar
binding capacities toward Hsp70, Hsc70, and Mortalin.
explaining why compounds with hydrophilic R
enhanced binding capability.
2
groups exhibited
Next, the single-turnover ATPase rates assay showed S1g-6
selectively inhibits the co-chaperone function of Bim to activate the
ATPase activity of Hsc70, rather than directly suppress the overall
ATPase activity of Hsc70, just like it did to Hsp70 (Fig. 4c). Further,
the capability and selectivity of S1g-6 to target Hsp70-Bim PPI in
living cells were confirmed by co-immunoprecipitation (co-IP)
experiments (Fig. 4d). We tested the perturbation of Hsp70-Bim in
BV173 cell line upon treatment with S1g-6, VER-155008, and MKT-
2.4. Validation of the target specificity of S1g-6 in situ
Firstly, we carried out a large-scale chemoproteomics experi-
ment to assess the potential targets of S1g-6 in whole human cell
proteome by affinity-based protein profiling. S1g-6 was incorpo-
rated with a well-established “minimalist” linker to generate the
affinity-based probe, S1g-6-probe, which was synthesized ac-
cording to published methods (Scheme 2). S1g-6-probe was added
to CML cell line BV173 cells lysate in the presence or absence of the
corresponding competitor S1g-6, and then the S1g-6-probe-bind-
ing targets were pulled down from lysate together with this probe
by streptavidin-modified agarose beads. LC-MS/MS analysis
revealed 170 proteins (Supplementary Data S1), and these candi-
dates were further analyzed by corresponding volcano plots as
statistical significance log10 (p-value) of S1g-6-probe enriched
proteins against (log ) of competition ratio (S1g-6-probe with
2
excess competitor S1g-6) via iTRAQ-based quantitative intensity
test. In this assay, only Hsp70 and its homologs, Hsc70 and Mor-
talin, showed enrichment log10 (p-value) ratios and log (S1g-6/
2
vehicle) higher than the significance criterion (>2, Fig. 4a), illus-
trating that S1g-6 selectively targets Hsp70 in the proteome. This
result was further validated by pull-down/western blotting (WB)
experiments with S1g-6-probe and the Hsp70/Hsc70/Mortalin
antibody, which showed that these Hsp70 family proteins were
077, at a gradient of concentrations (0, 1, 3, and 5 mM). The Hsp70-
Bag3 (a co-chaperone exhibiting different binding site with Bim)
complex [19], the target of MKT-077 and its analogies [12,13], was
tested in parallel. As expected, VER-155008 failed to disrupt both
Hsp70-Bim and Hsp70-Bag3 interactions, while MKT-077 inter-
fered with Hsp70-Bag3 PPI. Notably, S1g-6 dose-dependently
inhibited the Hsp70-Bim complex and showed no obvious impact
on the Hsp70-Bag3 complex even at the highest concentration,
which validated that S1g-6 specifically targets the Hsp70-Bim PPI.
MKT-077 exhibited a much weaker Hsp70-Bim PPI disruption.
It is reasonable for MKT-077 to disturb Hsp70-Bim PPI because
its binding site is near the Bim binding site, while VER-155008
binds far from it. The Hsp70-Bim PPI disruption in cells by the three
molecules are consistent with their IC50 values to disrupt Hsp70-
Bim PPI in vitro as shown in Table 1.
We can’t exclude that the other reported Hsp70 inhibitors could
also interfere with Hsp70-Bim PPI even though they bind apart
from the Bim binding site, because allosteric effects may be
involved. However, S1g-6 could exhibit the most potent and reli-
able targeting toward Hsp70-Bim both in vitro and in situ so far.
efficiently enriched by S1g-6-probe and competed in
a
concentration-dependent manner with S1g-6 (Fig. 4b). Taken
together, it can be concluded that Hsp70 family proteins are the
Scheme 2. Chemical structure and synthetic route of S1g-6-probe. Reagents and conditions: (a) 2-(piperazin-1-yl) ethan-1-amine, CH
3
CN, rt, 2 h. (b) Minimalist linker, K
2
3
CO , DMF,
ꢀ
60
C, 4 h.
4

Z. Wang, T. Song, Z. Guo et al.
European Journal of Medicinal Chemistry 220 (2021) 113452
Fig. 4. (a) Volcano plots as statistical significance (log10p-value) of S1g-6-probe enriched proteins against (log
2
) of competition ratio (S1g-6-probe with excess competitor S1g-6)
via iTRAQ-based quantitative intensity test in BV173 cell lysates. (b) Assays of the competitive binding of S1g-6 to Hsp70/Hsc70/Mortalin-S1g-6-probe complexes in cell lysates
determined by Western blot. (c) ATPase activity of Hsp70 treated with BimBH3, S1g-6 and VER-155008. Error bars show the mean ± s.e.m. (n ¼ 3). (d) Western blot and Co-IP
analysis of the levels of Hsp70, Bim, Bag3 and their PPIs in BV173 cells upon treatment with a gradient of concentrations (0e5
mM) of S1g-6, VER-155008 and MKT-077,
respectively, using
b-actin as a loading control. All figures represent the results from n ¼ 3 independent samples.
2
.5. Identification of the anti-cancer molecular biological
cells than VER-155008 and MKT-077. This result also validated the
discovery in our previous work that co-chaperone Bim helps Hsp70
recognize some cancer-related clients, making some cancer cells
addict to Hsp70-Bim function [20]. Moreover, S1g-6 showed higher
killing activity in BV173 and KCL22 than the ATP-competitive in-
hibitor VER-155008. Despite the explanation that nanomolar con-
centrations of Bim is much easier to be competed from Hsp70 than
millimolar concentrations of ATP in the cytosol, the higher anti-
tumor efficiency of S1g-6 than VER-155008 further suggested a
cancer-related role of Hsp70-Bim.
Next, we investigated the transduction signal pathway that
Hsp70-Bim PPI inhibitor S1g-6 induced in cancer cells. In our pre-
vious study, we identified that the Hsp70-Bim PPI regulates the fold
of oncogenic clients AKT and Raf-1 in cancer cells to suppress
apoptosis [20]. Herein, treatment of BV173 and KCL22 cells with
S1g-6 significantly decreased the expression and phosphorylation
mechanism of S1g-6
To investigate whether the specific Hsp70-Bim PPI inhibitor
S1g-6 could induce the selective killing effect for cancer cells over
normal tissue cells, we analyzed the apoptosis induction of S1g-6 in
CML cell lines BV173 and KCL22, and normal tissue cell lines
HEK293 and BaF3. VER-155008 and MKT-077 was used as control.
These cell lines were chosen because Hsp70-Bim PPI was found to
stabilize oncogenic clients and suppress apoptosis in BV173 and
KCL22 cell lines. As shown in Table 2, VER-155008 and MKT-077
exhibited broad-spectrum killing activity over all these cells
(
EC50 ¼ 5.5e8.7
potently induced apoptosis in BV173 and KCL22 (EC50 ¼ 1.42 and
.60 M), and exhibited no obvious cytotoxic effect in HEK293 and
mM and 2.7e6.6 mM). On the contrary, S1g-6
2
m
BaF3 (EC50 > 15 mM), showing much lower toxicity towards normal
Table 2
The EC50 value of S1g-6 and VER-155008 for apoptosis induction in cancer and normal tissue cells determined by Annexin V staining.
Compounds
BV173 EC50
(m
M)
KCL22 EC50
(
m
M)
HEK293 EC50
(m
M)
BaF3 EC50 (mM)
S1g-6
VER-155008
MKT-077
1.4 ± 0.2
5.5 ± 1.9
2.7 ± 0.3
2.6 ± 0.4
6.8 ± 1.1
3.2 ± 0.4
>15
8.7 ± 1.2
6.6 ± 0.6
>15
6.1 ± 1.7
5.4 ± 0.4
5

Z. Wang, T. Song, Z. Guo et al.
European Journal of Medicinal Chemistry 220 (2021) 113452
Fig. 5. (a) Western blot analysis of the levels of AKT, Raf-1, pAKT and pRaf-1 in BV173 and KCL22 cell lines upon treatment with S1g-6 (3
S1g-6 (3 M) induced apoptosis in stable clones from NS/Hsp70 shRNA-transfected BV173 cells determined by Annexin V staining and Western blot showing PARP cleavage. (c) S1g-
(3 M) induced apoptosis in stable clones from NS/Bim shRNA-transfected BV173 cells determined by Annexin V staining and Western blot showing PARP cleavage.
Figures represent the results from n ¼ 3 independent samples, and the data are expressed as the mean ± s.e.m. (n ¼ 3).
mM), using b-actin as a loading control. (b)
m
6
m
levels of AKT and Raf-1, illustrating the capacity of S1g-6 to perturb
the Hsp70-Bim function in living cells (Fig. 5a).
discriminate the cancer-related Hsp70 functions and otherwise.
Further, the dependency of Hsp70-Bim complex in S1g-6-
induced apotosis was validated through cell-based shRNA assays.
As shown in Fig. 5b and c, stable clones from both Hsp70 shRNA-
transfected and Bim shRNA-transfected cells showed obvious
resistance to S1g-6-induced apoptosis which is indicated by
Annexin V staining and PARP cleavage.
4
. Experimental section
4.1. Chemistry
All commercial reagents were purchased and used without
1
further purification unless otherwise stated. H NMR spectra were
obtained with a Bruker AV-500 spectrometer with chemical shifts
reported as ppm (in DMSO, TMS as an internal standard). Mass
spectra were obtained on a HPLC-Q-Tof MS (Micro) spectrometer.
Column chromatography was performed on silica gel 200e300
mesh. The purity of all final products was determined by analytical
HPLC to be ꢁ 95%. HPLC purity of compounds was measured with a
3
. Conclusion
Taken together, through derivation, screening, and optimization
of S1, a small molecular BimBH3 mimetic developed by our group,
we discovered a novel Hsp70 inhibitor, S1g-6, which exhibited sub-
normal-phase HPLC (XBridge C18, 4.6 mm ꢂ 150 mm, 5
mm) with
m
M binding affinity toward Hsp70 and selectively disrupted Hsp70-
two diverse wavelength detection systems.
Bim PPI. The target specificity of S1g-6 in situ was validated by
affinity-based protein profiling, co-immunoprecipitation, and cell-
based shRNA assays, which demonstrated that S1g-6-induced
apoptosis completely depends on disrupting Hsp70-Bim PPI. S1g-6
selectively induced apoptosis in cancer cell lines over normal cell
lines, making it a completely new class of Hsp70 inhibitor and a
promising antitumor candidate. Additionlly, validation of a new
cancer therapeutic target, Hsp70-Bim PPI, is emergeing to
4.2. General procedure for the synthesis of compounds S1-S1i
In a round-bottom flask, a mixture of acenaphthequinone (1.0
ꢀ
equiv) and malononitrile (1.1 equiv) were stirred in CH
for 3 h. The resulting mixture was cooled to rt, and then filtered to
give the desired product 1. The crude product was used for next
3
CN at 65 C
6

Z. Wang, T. Song, Z. Guo et al.
European Journal of Medicinal Chemistry 220 (2021) 113452
step without further purification.
1H), 7.92 (t, J ¼ 8.2 Hz, 1H), 7.07 (d, J ¼ 8.4 Hz, 1H), 4.47 (t, J ¼ 4.4 Hz,
13
In a round-bottom flask, a mixture of 1 (1.0 equiv) and K
2
CO
3
2H), 3.60 (t, J ¼ 4.6 Hz, 2H). C NMR (125 MHz, CDCl
3
): d 175.0,
ꢀ
(
0.1 equiv) were stirred in CH
3
CN at 65 C for 3 h. The resulting
145.6, 144.8, 140.1, 139.1, 134.7, 131.9, 129.9, 127.8, 126.9, 124.7, 124.1,
þ
mixture was cooled to rt, filtered, and washed for 3 times using
CH CN and H O, respectively. The expected product compound 2
was isolated as an orange powder, with yield ¼ 45% over two steps.
To a suspension of compound 2 (1.0 equiv) in CH CN, corre-
sponding R H (4.0 equiv) was added at rt, and the mixture was
120.5, 118.6, 114.9, 113.1, 111.2. ESI-MS: m/z for C17
calc 290.09, found 290.08.
H
12
N
3
O
2
[MþH] ,
3
2
3
4.2.7. 6-(cyclopentylthio)-1-oxo-1H-phenalene-2,3-dicarbonitrile
1
(S1f)
ꢀ
1
further stirred at 60 C for 3 h. Then, the solvent was removed
Red powder, yield ¼ 50%. H NMR (500 MHz, CDCl
3
): d 8.76 (d,
under reduced pressure, and the crude was purified on silica gel
J ¼ 8.0 Hz, 1H), 8.74 (t, J ¼ 7.8 Hz, 2H), 7.78 (t, J ¼ 8.0 Hz, 1H), 7.54 (d,
13
(
Hexane:CH
2
Cl
2
:EA ¼ 30:30:1) to give the expected product S1-S1i,
J ¼ 8.1 Hz, 1H), 3.54 (q, J ¼ 4.4 Hz, 1H), 1.70-1.42
d (m, 8H). C NMR
with yield ¼ 36e55%, respectively.
(125 MHz, CDCl ): 175.7, 140.0, 139.2, 134.7, 131.9, 131.2, 128.4,
3
d
1
27.7, 124.7, 121.9, 120.6, 117.9, 115.6, 45.2, 33.8, 26.8. ESI-MS: m/z
þ
4.2.1. 1-Oxo-6-thiomorpholino-1H-phenalene-2,3-dicarbonitrile
for C20
H
15
N
2
OS [MþH] , calc 331.09, found 331.06.
(
S1)
1
Purple powder, yield ¼ 41%. H NMR (500 MHz, CDCl
3
):
d
8.77 (d,
4.2.8. 6-(cyclohexylthio)-1-oxo-1H-phenalene-2,3-dicarbonitrile
J ¼ 7.7 Hz,1H), 8.48 (d, J ¼ 8.3 Hz,1H), 8.24 (d, J ¼ 8.0 Hz,1H), 7.87 (t,
J ¼ 8.4 Hz, 1H), 7.20 (d, J ¼ 8.1 Hz, 1H), 3.78 (t, J ¼ 4.8 Hz, 4H), 2.99 (t,
(S1g)
1
Red powder, yield ¼ 55%. H NMR (500 MHz, CDCl
3
): d 8.52 (d,
13
J ¼ 4.5 Hz, 4H). C NMR (125 MHz, CDCl
3
):
d
175.2, 155.1, 139.7,
J ¼ 8.0 Hz,1H), 8.29 (d, J ¼ 8.1 Hz,1H), 8.23 (d, J ¼ 8.0 Hz,1H), 7.80 (t,
1
2
3
31.7, 131.4, 129.3, 127.8, 126.5, 122.6, 120.5, 118.6, 115.8, 114.9, 55.6,
4.8. ESI-MS: m/z for C19
32.07.
J ¼ 8.1 Hz, 1H), 7.52 (d, J ¼ 7.8 Hz, 1H), 3.55 (q, J ¼ 4.6 Hz, 1H), 2.14 (t,
þ
13
H
14
N
3
OS [MþH] , calc 332.09, found
J ¼ 4.3 Hz, 2H), 1.78 (t, J ¼ 4.4 Hz, 2H), 1.61-1.46
d
(m, 6H). C NMR
3
(125 MHz, CDCl ): d 175.1, 139.8, 138.6, 134.5, 131.8, 131.1, 129.6,
1
C
28.6, 127.4, 124.8, 121.6, 120.2, 118.3, 115.8, ESI-MS: m/z for
þ
4
.2.2. 6-Phenylthio-1-oxo-1H-phenalene-2,3-dicarbonitrile (S1a)
21
H
17
N
2
OS [MþH] , calc 345.10, found 345.09.
1
Red powder, yield ¼ 48%. H NMR (500 MHz, CDCl
3
): d 8.81 (d,
J ¼ 8.1 Hz, 1H), 8.76 (d, J ¼ 8.0 Hz, 1H), 7.83 (t, J ¼ 8.4 Hz, 1H), 7.77 (t,
4.2.9. 6-(benzylamino)-1-oxo-1H-phenalene-2,3-dicarbonitrile
(S1h)
13
J ¼ 8.3 Hz, 1H), 7.51-7.36
d
(m, 5H), 7.20 (d, J ¼ 8.2 Hz, 1H). C NMR
1
(
125 MHz, CDCl ): 175.5, 139.5, 138.8, 135.7, 134.1, 131.6, 131.2,
3
d
Dark pink powder, yield ¼ 54%. H NMR (500 MHz, DMSO‑d
6
):
1
29.3,128.7,128.2,127.4, 126.1,125.0,120.4,119.1,115.5. ESI-MS: m/z
d
9.08 (d, J ¼ 8.1 Hz, 1H), 8.72 (d, J ¼ 7.9 Hz, 1H), 8.03
d (m, 2H), 7.43
þ
for C21
H
11
N
2
OS [MþH] , calc 339.05, found 339.05.
(d, J ¼ 8.4 Hz, 2H), 7.38 (t, J ¼ 8.3 Hz, 2H), 7.30 (t, J ¼ 8.6 Hz, 1H), 7.02
13
(
d, J ¼ 8.0 Hz, 1H), 4.88 (s, 2H). C NMR (125 MHz, CDCl
3
): d 175.3,
4.2.3. 6-(4-bromophenylthio)-1-oxo-1H-phenalene-2,3-
139.8, 137.4, 131.8, 131.2, 128.5, 127.5, 126.8, 126.2, 125.1, 123.0,
dicarbonitrile (S1b)
120.2, 119.3, 118.2, 115.6, 108.5, 47.9. ESI-MS: m/z for C22
[MþH] , calc 336.10, found 336.11.
H
14
N
3
O
1
þ
Red powder, yield ¼ 45%. H NMR (500 MHz, DMSO‑d
6
):
d
8.66
(
(
t, J ¼ 8.2 Hz, 2H), 8.48 (t, J ¼ 8.1 Hz, 1H), 7.88 (t, J ¼ 8.0 Hz, 1H), 7.70
1
3
d, J ¼ 8.8 Hz, 2H), 7.56 (t, J ¼ 8.1 Hz, 1H), 7.38 (d, J ¼ 8.6 Hz, 2H).
C
4.2.10. 6-(cyclohexylamino)-1-oxo-1H-phenalene-2,3-
NMR (125 MHz, CDCl
3
): 175.1, 140.2, 138.5, 134.7, 131.4, 131.0,
d
dicarbonitrile (S1i)
1
132.7, 131.4, 130.5, 128.8, 128.1, 126.2, 124.7, 121.8, 119.8, 117.6, 114.5.
Dark pink powder, yield ¼ 52%. H NMR (500 MHz, DMSO‑d
6
):
þ
ESI-MS: m/z for C21
H
10
N
2
OSBr [MþH] , calc 416.96, found 416.94.
d
9.23 (d, J ¼ 8.0 Hz, 1H), 8.71 (d, J ¼ 7.8 Hz, 1H), 8.01 (d, J ¼ 8.1 Hz,
H), 7.97 (t, J ¼ 7.8 Hz, 1H), 7.18 (d, J ¼ 8.0 Hz, 1H), 3.93 (m,
1
4
.2.4. 6-(4-Isopropylthiophenol)-1-oxo-1H-phenalene-2,3-
dicarbonitrile (S1c)
Red powder, yield ¼ 52%. H NMR (500 MHz, CDCl
J ¼ 4.0 Hz, 1H), 2.03 (d, J ¼ 4.2 Hz, 2H), 1.83 (d, J ¼ 4.1 Hz, 2H), 1.70
(d, J ¼ 4.4 Hz,1H), 1.57 (q, J ¼ 4.1 Hz, 2H), 1.45 (q, J ¼ 4.0 Hz, 2H),1.21
1
13
3
):
d
8.87 (d,
(q, J ¼ 4.1 Hz, 2H). C NMR (125 MHz, CDCl
3
): d 175.2, 151.1, 139.8,
J ¼ 8.0 Hz, 1H), 8.82 (d, J ¼ 8.2 Hz, 1H), 8.21 (d, J ¼ 8.3 Hz, 1H), 8.06
131.8, 131.4, 131.0, 127.7, 128.8, 125.2, 122.8, 120.1, 118.8, 118.0, 115.4,
þ
(
(
(
t, J ¼ 8.1 Hz,1H), 7.56 (d, J ¼ 8.4 Hz, 2H), 7.43 (d, J ¼ 8.4 Hz, 2H), 7.06
108.8, 61.5, 32.9, 26.1, 25.2. ESI-MS: m/z for C21
328.14, found 328.11.
H
18
N
3
O [MþH] , calc
13
d, J ¼ 8.4 Hz, 1H), 3.03 (m, 1H), 1.33 (d, J ¼ 4.9 Hz, 6H). C NMR
125 MHz, CDCl ): 176.1, 148.2, 139.6, 138.8, 134.4, 132.8, 131.9,
3
d
1
2
3
31.3, 130.7, 128.4, 127.8, 126.1, 125.3, 124.5, 124.1, 118.6, 115.3, 33.6,
4.3. General procedure for the synthesis of compounds S1g-1~S1g-6
þ
4.2. ESI-MS: m/z for C24
H
17
N
2
OS [MþH] , calc 381.10, found
81.09.
To a suspension of compound S1g (1.0 equiv) in CH
3
CN, corre-
sponding R H (30 equiv) was added at rt, and the mixture was
2
4.2.5. 6-(4-methoxyphenylthio)-1-oxo-1H-phenalene-2,3-
further stirred at rt for 6 h. Then, the solvent was removed under
dicarbonitrile (S1d)
reduced pressure, and the crude was purified on silica gel
1
Orange powder, yield ¼ 36%. H NMR (500 MHz, CDCl
3
):
d
8.72
(CH
1~S1g-6, with yield ¼ 12e21%, respectively. The major by-product
is the 6-R substitution product, for thiol group is a better leaving
2
Cl
2
:CH
3
OH ¼ 100:1-20:1) to give the expected product S1g-
(
(
d, J ¼ 8.1 Hz,1H), 8.63 (d, J ¼ 8.0 Hz,1H), 7.77 (t, J ¼ 8.0 Hz, 2H), 7.62
d, J ¼ 8.4 Hz, 2H), 7.34 (d, J ¼ 8.4 Hz, 2H), 7.24 (d, J ¼ 8.1 Hz, 1H),
2
3
.76 (s, 3H). 13C NMR (125 MHz, CDCl
3
):
d
175.4, 159.2, 139.9, 138.6,
group than cyan group.
1
34.5, 132.2, 131.7, 128.7, 128.1, 127.6, 126.1, 124.8, 120.1, 118.7, 115.2,
þ
114.6, 60.0. ESI-MS: m/z for C22
H
13
N
2
O
2
S [MþH] , calc 369.06,
4.3.1. 3-(benzylamino)-6-(cyclohexylthio)-1-oxo-1H-phenalene-2-
found 369.10.
carbonitrile (S1g-1)
1
Yellow powder, yield ¼ 14%. H NMR (500 MHz, CDCl
3
): d 8.64
4.2.6. 6-((2-hydroxyethyl)amino)-1-oxo-1H-phenalene-2,3-
d
(m, 3H), 8.18 (t, J ¼ 8.1 Hz, 1H), 8.08 (t, J ¼ 8.3 Hz, 1H), 7.78 (d,
J ¼ 8.4 Hz, 1H), 7.71-7.64 (m, 3H), 5.14 (d, J ¼ 7.8 Hz, 2H), 3.34 (m,
1H), 2.01 (d, J ¼ 4.3 Hz, 2H), 1.76 (t, J ¼ 4.5 Hz, 4H), 1.65 (t, J ¼ 4.2 Hz,
dicarbonitrile (S1e)
d
d
Orange powder, yield ¼ 41%. 1H NMR (500 MHz, DMSO‑d
):
6
13
d
8.97 (d, J ¼ 8.1 Hz, 1H), 8.64 (d, J ¼ 8.3 Hz, 1H), 8.03 (d, J ¼ 8.4 Hz,
3
4H). C NMR (125 MHz, CDCl ): d 177.6, 175.7, 138.4, 137.0, 134.1,
7

Z. Wang, T. Song, Z. Guo et al.
European Journal of Medicinal Chemistry 220 (2021) 113452
1
32.5, 131.9, 131.2, 128.8, 128.3, 127.8, 126.2, 125.7, 124.8, 124.3,
(piperazin-1-yl)ethan-1-amine (30 equiv) was added at rt, and the
mixture was further stirred at rt for 6 h. Then, the solvent was
removed under reduced pressure, and the crude was purified on
124.0, 123.7, 115.8, 78.3, 53.3, 48.4, 34.2, 25.6, 25.6. ESI-MS: m/z for
þ
C
H
27 25
N
2
OS [MþH] , calc 425.17, found 425.18.
silica gel (CH
with yield ¼ 14%.
To a suspension of compound 3 (1.0 equiv) in DMF, the mini-
malist linker (1.1 equiv) and K CO (3.0 equiv) were added at rt.
2
Cl
2
:CH
3
OH ¼ 20:1) to give the expected product 3,
4.3.2. 3-((cyclohexylmethyl)amino)-6-(cyclohexylthio)-1-oxo-1H-
phenalene-2-carbonitrile (S1g-2)
1
Yellow powder, yield ¼ 15%. H NMR (500 MHz, CDCl
3
):
d
8.56 (t,
2
3
ꢀ
J ¼ 8.4 Hz, 2H), 8.19 (d, J ¼ 8.1 Hz, 1H), 7.77 (t, J ¼ 8.0 Hz, 1H), 7.71
Then, the mixture was stirred at 60 C for 4 h, and the solvent was
(
(
d
t, ¼ 8.3 Hz, 1H), 3.86 (d, J ¼ 4.6 Hz, 2H), 3.38 (q, J ¼ 4.1 Hz, 1H), 2.09
d, J ¼ 4.4 Hz, 2H), 1.76-1.69 (m, 10H), 1.68-1.55 (m, 4H), 1.55-1.47
(m, 5H). 13C NMR (125 MHz, CDCl
): 177.4, 175.1, 137.6, 134.4,
removed under reduced pressure. The crude was purified on silica
d
d
gel (CH
probe, with yield ¼ 77%.
2
Cl
2
:CH
3
OH ¼ 30: 1) to give the expected product S1g-6-
3
d
1
3
4
32.5, 131.8, 130.8, 128.1, 127.7, 124.7, 124.2, 123.9, 115.7, 78.3, 53.3,
8.7, 34.1, 26.1, 25.3. ESI-MS: m/z for C27
31.22, found 431.25.
þ
H
31
N
2
OS [MþH] , calc
4.4.1. 3-((2-(4-(2-(3-(but-3-yn-1-yl)-3H-diazirin-3-yl)ethyl)
piperazin-1-yl)ethyl)amino)-6-(cyclohexylthio)-1-oxo-1H-
phenalene-2-carbonitrile. (S1g-6-probe)
1
4
.3.3. 3-(butylamino)-6-(cyclohexylthio)-1-oxo-1H-phenalene-2-
H NMR (500 MHz, DMSO‑d
6
):
d
8.59 (d, J ¼ 8.2 Hz, 1H), 8.46 (d,
carbonitrile (S1g-3)
J ¼ 7.8 Hz, 1H), 8.41 (d, J ¼ 7.9 Hz, 1H), 7.89 (t, J ¼ 8.1 Hz, 1H), 7.54 (d,
J ¼ 8.2 Hz, 1H), 4.13 (q, J ¼ 6.7 Hz, 2H), 3.46 (m, 8H), 2.92 (t,
J ¼ 6.4 Hz, 2H), 2.76 (s, 1H), 2.63 (t, J ¼ 6.2 Hz, 2H), 2.26 (t, J ¼ 5.4 Hz,
2H), 2.05 (t, J ¼ 4.6 Hz, 2H), 1.81 (m, 5H), 1.74 (t, J ¼ 4.4 Hz, 3H),
1.72-1.53 (m, 5H), 1.25 (m, 4H). C NMR (500 MHz, DMSO‑d
175.2, 173.3, 171.1, 138.7, 134.9, 131.4, 131.3, 128.2, 127.7, 124.8,
124.4, 123.1, 115.8, 98.0, 83.7, 70.0, 68.7, 58.0, 53.4, 47.4, 45.0, 34.2,
Yellow powder, yield ¼ 21%. 1H NMR (500 MHz, DMSO‑d
):
d
6
d
8.62 (d, J ¼ 7.8 Hz, 1H), 8.54 (t, J ¼ 8.0 Hz, 1H), 8.48 (d, J ¼ 8.1 Hz,
1
5
6
H), 7.79
d
(m, 2H), 4.27 (t, J ¼ 4.4 Hz, 2H), 3.68
d
(m, 1H), 2.08
d
d
(m,
(m,
d
13
H), 1.76 (t, J ¼ 4.4 Hz, 2H), 1.63 (t, J ¼ 4.2 Hz, 1H), 1.51-1.30
d
d
6
):
13
H), 0.96 (t, J ¼ 4.1 Hz, 3H). C NMR (125 MHz, DMSO‑d
6
):
d
177.2,
d
175.6, 137.8, 134.0, 131.9, 131.7, 131.1, 128.4, 127.3, 124.2, 123.8, 123.4,
1
15.9, 78.1, 53.4, 42.8, 34.4, 33.0, 25.6, 25.3, 20.4, 13.8. ESI-MS: m/z
33.0, 30.1, 28.0, 25.9, 25.3,11.2. ESI-MS: m/z for C33
39
H N
4
OS [M þ H-
þ
þ
for C24
27
H N
2
OS [MþH] , calc 391.18, found 391.21.
N
2
] , calc 539.29, found 539.27.
4.3.4. 6-(cyclohexylthio)-3-((2-hydroxyethyl)amino)-1-oxo-1H-
4.5. Protein expression and purification
phenalene-2-carbonitrile (S1g-4)
Yellow powder, yield ¼ 15%. 1H NMR (500 MHz, DMSO‑d
):
Constructs of hHsp70 NBD (1e383), hHsc70 NBD (1e383) and
hBcl-2 (2e206) cDNA were subcloned into the pHis vector, and the
resulting proteins with an N-terminal 6 ꢂ His tag were produced in
Escherichia coli strain BL21 (DE3) containing the corresponding
plasmid. The samples were produced in LB medium. The proteins
were further purified by Sephadex G-75 size exclusion chroma-
tography (GE Healthcare), before they were pooled and concen-
trated for biochemical or structural assays.
6
d
8.45 (d, J ¼ 8.7 Hz, 1H), 8.24 (d, J ¼ 8.4 Hz, 1H), 8.14 (d, J ¼ 8.5 Hz,
1
2
1
1
1
5
[
H), 7.74 (d, J ¼ 8.3 Hz, 1H), 7.69 (t, J ¼ 8.1 Hz, 1H), 4.30 (t, J ¼ 5.7 Hz,
H), 4.14 (t, J ¼ 5.4 Hz, 2H), 3.46 (m, 1H), 2.00 (d, J ¼ 4.4 Hz, 2H),
(m, 4H),
177.2, 175.5, 138.1,
34.3, 132.2, 131.7, 131.0, 128.2, 127.5, 124.6, 124.1, 123.6, 115.8, 78.3,
d
.74 (d, J ¼ 4.3 Hz, 2H), 1.58 (d, d, J ¼ 3.9 Hz, 1H), 1.46-1.33
d
.25 ):
d
(m, 2H). 13C NMR (125 MHz, DMSO‑d
6
d
6.4, 53.3, 46.7, 34.3, 25.8, 25.4. ESI-MS: m/z for C22
MþH] , calc 379.15, found 379.13.
23
H N
2
O
2
S
þ
4.6. Fluorescence polarization assays (FPAs)
4.3.5. 3-((2-cyano-6-(cyclohexylthio)-1-oxo-1H-phenalen-3-yl)
amino)propanamide (S1g-5)
A mixture of 10 nM FAM-labeled BimBH3 peptide and 300 nM
Yellow powder, yield ¼ 12%. 1H NMR (500 MHz, DMSO‑d
Bcl-2 (2e206) or Hsp70 NBD (1e383) protein was preincubated in
the assay buffer (100 mM potassium phosphate, pH 7.5; 100 g/mL
6
):
d
8.58 (d, J ¼ 7.8 Hz, 1H), 8.49 (d, J ¼ 7.6 Hz, 1H), 8.46 (t, J ¼ 7.7 Hz,
m
1
3
d
H), 7.81
.65
(m, 1H), 2.02 (t, J ¼ 4.3 Hz, 2H), 1.75 (t, J ¼ 4.5 Hz, 2H), 1.62
(m, 1H), 1.43
): 177.6, 175.1, 174.4, 138.4, 134.1, 132.1, 131.5, 131.0, 128.4,
d
(m, 2H), 4.11 (t, J ¼ 6.1 Hz, 2H), 3.80 (t, J ¼ 5.8 Hz, 2H),
bovine gamma globulin; 0.02% sodium azide). Next, serial dilutions
of compounds were added. After a 30 min incubation, the polari-
zation values were measured using the Spectra Max M5 Detection
System in a black 96-well plate. The polarization value was plotted
as a function of the concentration of the compound, and the IC50
value was determined from the competitive inhibition curve.
d
13
d
(m, 4H), 1.26 d (m, 3H). C NMR (125 MHz,
DMSO‑d
6
d
127.6, 124.8, 124.4, 124.0, 115.7, 78.2, 53.4, 40.4, 38.2, 34.1, 25.7, 25.3.
þ
TOF ESI-MS: m/z for C23
H
24
N
3
O
2
S [MþH] , calc 406.52, found
4
06.51.
4.7. Isothermal titration calorimetry (ITC)
4
1
.3.6. 6-(cyclohexylthio)-3-((2-morpholinoethyl)amino)-1-oxo-
H-phenalene-2-carbonitrile (S1g-6)
The Hsp70 NBD (1e383)-compound (S1g-6/MKT-077) and
Yellow powder, yield ¼ 14%. 1H NMR (500 MHz, DMSO‑d
Hsc70 NBD (1e383)-compound (S1g-6/MKT-077) interactions
6
):
ꢀ
d
8.56 (d, J ¼ 8.4 Hz, 1H), 8.42 (d, J ¼ 8.2 Hz, 1H), 8.29 (d, J ¼ 8.1 Hz,
were characterized at 25 C using ITC200 (GE Healthcare/MicroCal).
1
H), 7.84
d
(m, 2H), 4.02 (t, J ¼ 4.4 Hz, 2H), 3.73
d
(m, 1H), 3.57 (t,
(m, 2H),
A typical experiment included the injection of 13 aliquots con-
taining approximately 0.2 mM ligand solution into a protein solu-
J ¼ 6.4 Hz, 4H), 3.03
.75 (t, J ¼ 4.6 Hz, 2H), 1.61
125 MHz, DMSO‑d ): 178.1, 175.4, 138.8, 134.8, 132.1, 131.3, 128.4,
d
(m, 4H), 2.75 (t, J ¼ 4.7 Hz, 2H), 2.04
d
13
1
(
d
(m, 1H), 1.29-1.23
d
(m, 5H). C NMR
tion of approximately 20
mM in the ITC cell (in a total volume of
6
d
approximately 200 L). An additional set of injections was run in a
m
1
2
4
27.4, 124.8,124.1, 123.5, 115.9, 78.6, 66.6, 56.1, 54.5, 53.2, 40.6, 34.3,
separate experiment with buffer instead of the protein solution as a
control. Before data analysis, the control values were subtracted
from the main experimental data.
þ
5.9, 25.1. ESI-MS: m/z for C26
H
30
N
3
O
2
S [MþH] , calc 448.21, found
48.18.
The Origin 7.0-based software provided by GE/MicroCal was
used for data analysis. The binding isotherms were integrated to
4.4. Procedure for the synthesis of compounds S1g-6-probe
give the enthalpy change ( H), plotted as a function of the molar
D
To a suspension of compound S1g (1.0 equiv) in CH
3
CN, 2-
ratio of the ligand. The DH/molar ratio plot was a sigmoidal,
8

Z. Wang, T. Song, Z. Guo et al.
European Journal of Medicinal Chemistry 220 (2021) 113452
representing the fractional saturation of the binding sites by the
ligand, and therefore the one set of sites model was used as the
centrifugation, the supernatant was removed. The beads were
washed with 0.1% SDS once and PBS four times and then boiled in
SDS loading buffer (2 ꢂ ) [200 mM Tris, 400 mM dithiothreitol
(DTT), and 8% SDS, pH 6.8] for 15 min.
basic option. The dissociation constant K
d
was determined from the
slope of the central linear part of the fractional saturation curve.
4.8. Molecular docking
4.11. iTRAQ-based quantitative proteomic profiling of the
interactome
The 3D structures of the human Hsp70 NBD (Hsp70; PDB 5AZQ)
were obtained from the RCSB Protein Data Bank. The 3D structures
of the inhibitors were generated using Chembio3D Ultra 11.0 fol-
lowed by energy minimization. The AutoDock 4.0 program equip-
ped with ADT was used to perform the automated molecular
docking. Grid maps covering residues that were perturbed by more
than the threshold value of 0.1 ppm on the Hsc70 protein surface
were defined for all inhibitors in the AutoDock calculations using a
grid spacing of 0.6 Å. The GA-LS algorithm was adopted using
default settings, and 100 hybrid GA-LS runs were carried out. A total
of 100 possible binding conformations were generated and
grouped into clusters based on a 1.0 Å cluster tolerance. The
docking models were analyzed and represented using ADT.
For LC-MS/MS protein identification, above pull-down fractions
were separated by SDS-10% polyacrylamide gels, followed by Coo-
massie blue staining. Each lane was cut into multiple gel slices.
Trypsin digestion was performed on each gel slice with an in-gel
trypsin digestion kit (Pierce). The resulting peptide mixtures were
desalted and loaded onto an Acclaim PePmap C18-reversed phase
column (75
rated with reversed phase C18 column (75
m
m ꢂ 2 cm, 3
m
m, 100 Å thermo scientific) and sepa-
m
m ꢂ 10 cm, 5 m,
m
300 Å, Agela Technologies) mounted on a Dionex ultimate 3000
nano LC system. Peptides were eluted using a gradient of 5e80% (v/
v) acetonitrile in 0.1% formic acid over 45 min at a flow rate of
300 nL min-1 combined with a Q Exactive mass spectrometer
(Thermo Fisher Scientific, MA, USA). The eluates were directly
4
.9. Cell lines, reagents and chemical inhibitors
entered Q-Exactive MS (Thermo Fisher Scientific, Waltham, MA,
USA), setting in positive ion mode and data-dependent manner
with full MS scan from 350 to 2000 m/z, full scan resolution at
70,000, MS/MS scan resoltion at 17,500. MS/MS scan with mini-
mum signal threshold 1Eþ5, isolation width at 2 Da. To evaluate the
performance of this mass spectrometry on the iTRAQ labeled
samples, two MS/MS acquisition modes, higher collision energy
dissociation (HCD) was employed. And to optimize the MS/MS
acquisition efficiency of HCD, normalized collision energy (NCE)
was systemically examined, stepped 20%.
Cell lines were purchased from American Type Culture Collec-
tion and used within 6 months from resuscitation. KCL22 and BV-
73 were cultured in RPMI 1640 medium (Thermo Scientific
1
HyClone, Beijing, China). HEK-293T cells were cultured in DMEM
medium (Thermo Scientific HyClone, Beijing, China). BaF3 cells
were cultured in RPMI 1640 medium with 10 ng/ml mouse IL-3
(
R&D Systems, Minneapolis, MN, USA). All of medium were sup-
plemented with 10% heat inactivated fetal bovine serum (FBS;
Gibco BRL, Grand Island, NY, USA) and 100 U/mL penicillin-
ꢀ
streptomycin and all cells were cultured at 37 C and 5% CO
2
.
4.12. Annexin V-FITC staining assay for apoptosis
VER-155008 was purchased from Selleck Chemicals (Shanghai,
China), and MKT-077 was purchased from MedChem Express
Phosphatidylserine exposure was quantified by surface Annexin
(
Monmouth Junction, NJ, USA), all the chemicals were dissolved in
V-FITC staining. Cells were treated with inhibitor in 48 h and then
transferred from a culture well (1.2 ꢂ 10 /well) to a tube and
6
dimethylsulfoxide (DMSO) to a concentration of 10 mM. To obtain
the final concentration, stock solutions were diluted in culture
medium.
washed with PBS containing 1% (v/v) bovine calf serum (Hyclone,
Logan, UT, USA). According to the manufacturer’s instructions, the
cells were incubated with a 1:40 solution of FITC-conjugated
Annexin V (Roche Diagnostics, Germany) in the dark for 10 min
at room temperature and the Annexin V-FITC positive cells were
analyzed by flow cytometry on a BD FACSCalibur (BD Biosciences,
Becton Drive, Franklin Lakes, NJ, USA). Cell Questc software (BD
Biosciences) was used to determine the percentage of apoptosis in
the samples. (The inhibition concentration 50 (IC50) was defined as
the concentration of the drug required to induce apoptosis by 50%.)
4
.10. Affinity-based protein profiling (AfBPP)
Cells were grown to 80e90% confluency in 24-well plates under
conditions described above. The medium was removed, and cells
were washed twice with cold PBS and then treated with 0.5 mL of
the DMEM-containing probe (diluted from DMSO stocks whereby
DMSO never exceeded 1% in the final solution). After 5 h of incu-
bation (37 C, 5% CO
ꢀ
2
), the medium was aspirated, and cells were
washed gently with PBS (2 ꢂ ) to remove excessive probe, followed
by UV irradiation for 20 min on ice. The cells were trypsinized and
pelleted by centrifugation. The cell pellets were resuspended in PBS
4.13. Immunoblotting
Cells were either treated with inhibitor or DMSO (vehicle) and
lysed in RIPA buffer (Solarbio, Beijing, China) supplementing Halt
protease/phosphatase inhibitor cocktail (Pierce Biotechnology,
Rockford, USA) for 30 min on ice and centrifuged at 12,000 g for
(
50
PBS.
One microliter of a freshly premixed click chemistry reaction
cocktail in PBS (2.5 mM Biotin-N from 10 mM stock solution in
mL), homogenized by sonication, and diluted to 1 mg/mL with
ꢀ
3
15 min at 4 C. Protein concentrations were determined using BCA
DMSO, 2.5 mM TBTA from 10 mM freshly prepared stock solution in
deionized water, 25 mM TCEP from 100 mM freshly prepared stock
assay (Beyotime, Shanghai, China) according to the manufacturer’s
instructions. Lysates (150e250
mg) were electrophoretically
solution in deionized water, and 25 mM CuSO
4
from 100 mM
resolved by SDS-PAGE, transferred to PVDF membrane and probed
with the following primary antibodies against: Hsp70 (1:1000,
10995-1-AP; Proteintech), Bag-3 (1:1000, 10599-1-P; Proteintech),
Bim (1:1000, 2933; CST), PARP (1:500, sc-8007; Santa Cruz
Biotechnology), AKT (1:800, PAB1310; Abnova), Raf-1 (1:800,
PAB26752; Abnova), pAKT (1:5000, ab81283; Abcam), pRaf-1
(1:1000, ab173539; Abcam), Actin (1:10000, A01010; Abbkine).
Immune complexes were visualized by the use of Super Signal West
freshly prepared stock solution in deionized water) was added. The
reaction was further incubated for 2 h with gentle mixing, before
being terminated by addition of prechilled acetone (0.5 mL; 30 min
ꢀ
incubation at ꢃ20 C). Precipitated proteins were subsequently
ꢀ
collected by centrifugation (13000 rpm ꢂ 10 min at 4 C), resolu-
bilized in 1% SDS in 50
m
L of PBS and then incubated with avidin-
ꢀ
agarose beads (100
mL/mg of protein) at 4 C overnight. After
9

Z. Wang, T. Song, Z. Guo et al.
European Journal of Medicinal Chemistry 220 (2021) 113452
hypertension, Br. J. Pharmacol. 176 (2019) 1829e1838.
Pico Chemiluminescent Substrate (Pierce Biotechnology) and
detected on a Kodak Image Station 4000 MM Pro (New Haven, CT,
USA).
[
9] L. Shrestha, H.J. Patel, G. Chiosis, Chemical tools to investigate mechanisms
associated with HSP90 and HSP70 in Disease, Cell Chem. Biol. 23 (2016)
158e172.
[
10] M. Ferraro, I. D’Annessa, E. Moroni, G. Morra, A. Paladino, S. Rinaldi,
F. Compostella, G. Colombo, Allosteric Modulators of HSP90 and HSP70: dy-
namics meets function through structure-based drug design, J. Med. Chem. 62
4.14. Co-immunoprecipitation (co-IP)
(2019) 60e87.
Cells were washed twice with ice cold PBS and lysed on ice in IP-
buffer (10 mM HEPES, pH 7.4, 50 mM NaCl, 5 mM MgCl , 1 mM
2
EGTA, 5% glycerol, 0.5% Triton X-100) supplemented with 1% pro-
[11] J.E. Gestwicki, H. Shao, Inhibitors and chemical probes for molecular chap-
erone networks, J. Biol. Chem. 294 (2019) 2151e2161.
[
12] X. Li, S.R. Srinivasan, J. Connarn, A. Ahmad, Z.T. Young, A.M. Kabza,
E.R.P. Zuiderweg, D. Sun, J.E. Gestwicki, Analogues of the allosteric heat shock
protein 70 (Hsp70) inhibitor, MKT-077, as anti-cancer agents, ACS Med. Chem.
Lett. 4 (2013) 1042e1047.
tease/phosphatase inhibitor for 30 min. After centrifugation
ꢀ
(
15 min, 12,000 g, 4 C), supernatants were collected and normal-
[
13] H. Shao, X. Li, M.A. Moses, L.A. Gilbert, C. Kalyanaraman, Z.T. Young,
M. Chernova, S.N. Journey, J.S. Weissman, B. Hann, M.P. Jacobson, L. Neckers,
J.E. Gestwicki, Exploration of benzothiazole rhodacyanines as allosteric in-
hibitors of protein-protein interactions with heat shock protein 70 (Hsp70),
J. Med. Chem. 61 (2018) 6163e6177.
ized to the protein content. 2
mg of antibody were added to an input
volume of 200 l with 5e10 mg/ml protein. As control IPs were
m
done with same amounts of purified rabbit or mouse IgG (Thermo
ꢀ
Fisher, Scientific). After shaking overnight at 4 C, 20
ml protein G
[
14] D.M. Huryna, J.L. Brodskyc, K.M. Brummonda, P.G. Chambersa, B. Eyera,
A.W. Irelandc, M. Kawasumid, M.G. LaPortea, K. Lloyda, B. Manteaua,
P. Nghiemd, B. Quadea, S.P. Seguinc, P. Wipf, Chemical methodology as a
source of small-molecule checkpoint inhibitors and heat shock protein 70
sepharose beads (GE Healthcare, Madison, WI, USA) were added
ꢀ
and the samples incubated for 2 h at 4 C with constant rotation.
Immunocomplexes were washed three times with IP buffer and
(Hsp70) modulators, Proc. Natl. Acad. Sci. Unit. States Am. 108 (2011)
ꢀ
solved by heating for 5 min at 95 C by immunoblotting analysis.
6757e6762.
[
15] A. Rodina, P.D. Patel, Y. Kang, Y. Patel, I. Baaklini, M.J.H. Wong, T. Taldone,
P. Yan, C. Yang, R. Maharaj, A. Gozman, M.R. Patel, H.J. Patel, W. Chirico,
H. Erdjument-Bromage, T.T. Talele, J.C. Young, G. Chiosis, Identification of an
allosteric pocket on human Hsp70 reveals a mode of inhibition of this ther-
apeutically important protein, Chem. Biol. 20 (2013) 1469e1480.
Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
[
16] D.V. Sverchinsky, A.D. Nikotina, E.Y. Komarova, E.R. Mikhaylova, N.D. Aksenov,
V.F. Lazarev, V.A. Mitkevich, R. Suezov, D.S. Druzhilovskiy, V.V. Poroikov,
B.A. Margulis, I.V. Guzhova, Etoposide-induced apoptosis in cancer cells can Be
reinforced by an uncoupled link between Hsp70 and caspase-3, Int. J. Mol. Sci.
1
9 (2018) 2519e2534.
Acknowledgments
[
[
17] S.-K. Ko, J. Kim, D.C. Na, S. Park, S.-H. Park, J.Y. Hyun, K.-H. Baek, N.D. Kim, N.-
K. Kim, Y.N. Park, K. Song, I. Shin, A small molecule inhibitor of ATPase activity
of HSP70 induces apoptosis and has antitumor activities, Chem. Biol. 22
This research was supported by the National Natural Science
Foundation of China (81903462 and 82073703), the China Post-
doctoral Science Foundation (2018M641694) and the Fundamental
Research Funds for the Central Universities (DUT20LK28 and
DUT20YG133).
(2015) 391e403.
18] M.A. Mikeladze, E.A. Dutysheva, V.G. Kartsev, B.A. Margulis, I.V. Guzhova,
V.F. Lazarev, Disruption of the complex between GAPDH and Hsp70 sensitizes
C6 glioblastoma cells to hypoxic stress, Int. J. Mol. Sci. 22 (2021) 1520e1535.
[19] T.A. Colvin, V.L. Gabai, J.L. Gong, S.K. Calderwood, H. Li, S. Gummuluru,
O.N. Matchuk, S.G. Smirnova, N.V. Orlova, I.A. Zamulaeva, M. Garcia-Marcos,
X. Li, Z. Young, J.N. Rauch, J.E. Gestwicki, S. Takayama, M.Y. Sherman, Hsp70-
Bag3 interactions regulate cancer-related signaling networks, Canc. Res. 74
Appendix A. Supplementary data
(2014) 4731e4740.
[
20] Z. Guo, T. Song, Z. Wang, D. Lin, K. Cao, P. Liu, Y. Feng, X. Zhang, P. Wang, F. Yin,
J. Dai, S. Zhou, Z. Zhang, The chaperone Hsp70 is a BH3 receptor activated by
the pro-apoptotic Bim to stabilize anti-apoptotic clients, J. Biol. Chem. 295
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.ejmech.2021.113452.
(2020) 12900e12909.
[
21] Z. Zhang, G. Wu, F. Xie, T. Song, X. Chang, 3-Thiomorpholin-8-oxo-8H-ace-
naphtho[1,2-b]pyrrole-9-carbonitrile (S1) based molecules as potent, dual
inhibitors of B-cell lymphoma 2 (Bcl-2) and myeloid cell leukemia sequence 1
(Mcl-1): structure-based design and structure-activity relationship studies,
J. Med. Chem. 54 (2011) 1101e1105.
Data and materials availability
All data needed to evaluate the conclusions in the paper are
present in the paper and/or the Supplementary Materials. Addi-
tional data related to this paper may be requested from the authors.
[
[
22] T. Song, Q. Chen, X. Li, G. Chai, Z. Zhang, Correction to 3-Thiomorpholin-8-
oxo-8H-acenaphtho[1,2-b]pyrrole-9-carbonitrile (S1) based molecules as
potent, dual inhibitors of B-cell lymphoma 2 (Bcl-2) and myeloid cell leuke-
mia sequence 1 (Mcl-1): structure-based design and structure-activity rela-
tionship studies, J. Med. Chem. 56 (2013) 9366e9367.
23] D.S. Williamson, J. Borgognoni, A. Clay, Z. Daniels, P. Dokurno, M.J. Drysdale,
N. Foloppe, G.L. Francis, C.J. Graham, R. Howes, A.T. Macias, J.B. Murray,
R. Parsons, T. Shaw, A.E. Surgenor, L. Terry, Y.K. Wang, M. Wood, A.J. Massey,
Novel adenosine-derived inhibitors of 70 kDa heat shock protein, discovered
through structure-based design, J. Med. Chem. 52 (2009) 1510e1513.
24] A. Rousaki, Y. Miyata, U.K. Jinwal, C.A. Dickey, J.E. Gestwicki, E.R.P. Zuiderweg,
Allosteric drugs: the interaction of antitumor compound MKT-077 with hu-
man HSP70 chaperones, J. Mol. Biol. 411 (2011) 614e632.
References
[
[
[
1] M.E. Murphy, The HSP70 family and cancer, Carcinogenesis 34 (2013)
181e1188.
2] M.Y. Sherman, V.L. Gabai, Hsp70 in cancer: back to the future, Oncogene 34
2015) 4153e4161.
3] G.D. Lianos, G.A. Alexiou, A. Mangano, A. Mangano, S. Rausei, L. Boni,
G. Dionigi, D.H. Roukos, The role of heat shock proteins in cancer, Canc. Lett.
1
(
[
[
360 (2015) 114e118.
[
[
[
4] J. Wu, T. Liu, Z. Rios, Q. Mei, X. Lin, S. Cao, Heat shock proteins and cancer,
Trends Pharmacol. Sci. 38 (2017) 226e256.
5] F.U. Hartl, A. Bracher, M. Hayer-Hartl, Molecular chaperones in protein folding
and proteostasis, Nature 475 (2011) 324e332.
6] W.B. Pratt, J.E. Gestwicki, Y. Osawa, A.P. Lieberman, Targeting Hsp90/Hsp70-
based protein quality control for treatment of adult onset neurodegenera-
tive diseases, Annu. Rev. Pharmacol. Toxicol. 55 (2015) 353e371.
7] C.L. Klaips, G.G. Jayaraj, F.U. Hartl, Pathways of cellular proteostasis in aging
and disease, J. Cell Biol. 217 (2018) 51e63.
25] A.J. Souers, J.D. Leverson, E.R. Boghaert, S.L. Ackler, N.D. Catron, J. Chen,
B.D. Dayton, H. Ding, S.H. Enschede, W.J. Fairbrother, D.C.S. Huang,
S.G. Hymowitz, S. Jin, S.L. Khaw, P.J. Kovar, L.T. Lam, J. Lee, H.L. Maecker,
K.C. Marsh, K.D. Mason, M.J. Mitten, P.M. Nimmer, A. Oleksijew, C.H. Park,
C.M. Park, D.C. Phillips, A.W. Roberts, D. Sampath, J.F. Seymour, M.L. Smith,
G.M. Sullivan, S.K. Tahir, C. Tse, M.D. Wendt, Y. Xiao, J.C. Xue, H.C. Zhang,
R.A. Humerickhouse, S.H. Rosenberg, S.W. Elmore, ABT-199, a potent and
selective BCL-2 inhibitor, achieves antitumor activity while sparing platelets,
Nat. Med. 19 (2013) 202e208.
[
[
8] B. Rodriguez-Iturbe, M.A. Lanaspa, R.J. Johnson, The role of autoimmune
reactivity induced by heat shock protein 70 in the pathogenesis of essential
10