Characterization of HJ-PI01 as a novel Pim-2 inhibitor that induces apoptosis and autophagic cell death in triple-negative human breast cancer

Article abstract of DOI:10.1038/aps.2016.60

Aim: Pim-2 is a short-lived serine/threonine kinase, which plays a key role in metastasis of breast cancer through persistent activation of STAT3. Although the crystal structure of Pim-2 has been reported, but thus far no specific Pim-2-targeted compounds have been reported. In this study, we identified a novel Pim-2 inhibitor, HJ-PI01, by in silico analysis and experimental validation. Methods: The protein-protein interaction (PPI) network, chemical synthesis, molecular docking, and molecular dynamics (MD) simulations were used to design and discover the new Pim-2 inhibitor HJ-PI01. The anti-tumor effects of HJ-PI01 were evaluated in human breast MDA-MB-231, MDA-MB-468, MDA-MB-436, MCF-7 cells in vitro and in MDA-MB-231 xenograft mice, which were treated with HJ-PI01 (40 mg·kg^-1 ·d^-1, ig) with or without lienal polypeptide (50 mg·kg^-1 ·d^-1, ip) for 10 d. The apoptosis/autophage-inducing mechanisms of HJ-PI01 were elucidated using Western blots, immunoblots, flow cytometry, transmission electron microscopy and fluorescence microscopy. Results: Based on the PrePPI network, the potential partners interacting with Pim-2 in regulating apoptosis (160 protein pairs) and autophagy (47 protein pairs) were identified. Based on the structural characteristics of Pim-2, a total of 15 compounds (HJ-PI01 to HJ-P015) were synthesized, which showed moderate or remarkable anti-proliferative potency in the human breast cancer cell lines tested. The most effective compound HJ-PI01 exerted a robust inhibition on MDA-MB-231 cells compared with chlorpromazine and the pan-Pim inhibitor PI003. Molecular dynamics (MD) simulation revealed that HJ-PI01 had a good binding score with Pim-2. Moreover, HJ-PI01 (300 nmol/L) induced death receptor-dependent and mitochondrial apoptosis as well as autophagic death in MDA-MB-231 cells. In MDA-MB-231 xenograft mice, administration of HJ-PI01 remarkably inhibited the tumor growth and induced tumor cell apoptosis in vivo. Co-administration of HJ-PI01 with lienal polypeptide could improve the anti-tumor activity of HJ-PI01 and reduce its toxicity. Conclusion: The newly synthesized compound, HJ-PI01, can induce death receptor/mitochondrial apoptosis and autophagic cell death by targeting Pim-2 in human breast cancer cells in vitro and in vivo.

Full text of DOI:10.1038/aps.2016.60

Acta Pharmacologica Sinica 2016: 1–14
© 2016 CPS and SIMM All rights reserved 1671-4083/16
www.nature.com/aps
Original Article
Characterization of HJ-PI01 as a novel Pim-2 inhibitor
that induces apoptosis and autophagic cell death in
triple-negative human breast cancer
1,
1, 3
Yu-qian ZHAO1, #, Yi-qiong YIN2, #, Jie LIU2, #, Gui-hua WANG1, Jian HUANG1, , Ling-juan ZHU , Jin-hui WANG
*
*
1School of Traditional Chinese Materia Medica, Shenyang Pharmaceutical University, Shenyang 110016, China; 2Department of
Gastrointestinal Surgery, State Key Laboratory of Biotherapy, Collaborative Innovation Center of Biotherapy, West China Hospital,
Sichuan University, Chengdu 610041, China; 3School of Pharmacy, Shihezi University, Shihezi 832002, China
Aim: Pim-2 is a short-lived serine/threonine kinase, which plays a key role in metastasis of breast cancer through persistent activation
of STAT3. Although the crystal structure of Pim-2 has been reported, but thus far no specific Pim-2-targeted compounds have been
reported. In this study, we identified a novel Pim-2 inhibitor, HJ-PI01, by in silico analysis and experimental validation.
Methods: The protein-protein interaction (PPI) network, chemical synthesis, molecular docking, and molecular dynamics (MD)
simulations were used to design and discover the new Pim-2 inhibitor HJ-PI01. The anti-tumor effects of HJ-PI01 were evaluated
in human breast MDA-MB-231, MDA-MB-468, MDA-MB-436, MCF-7 cells in vitro and in MDA-MB-231 xenograft mice, which were
treated with HJ-PI01 (40 mg·kg-1·d-1, ig) with or without lienal polypeptide (50 mg·kg-1·d-1, ip) for 10 d. The apoptosis/autophage-inducing
mechanisms of HJ-PI01 were elucidated using Western blots, immunoblots, flow cytometry, transmission electron microscopy and
fluorescence microscopy.
Results: Based on the PrePPI network, the potential partners interacting with Pim-2 in regulating apoptosis (160 protein pairs) and
autophagy (47 protein pairs) were identified. Based on the structural characteristics of Pim-2, a total of 15 compounds (HJ-PI01 to
HJ-P015) were synthesized, which showed moderate or remarkable anti-proliferative potency in the human breast cancer cell lines
tested. The most effective compound HJ-PI01 exerted a robust inhibition on MDA-MB-231 cells compared with chlorpromazine and the
pan-Pim inhibitor PI003. Molecular dynamics (MD) simulation revealed that HJ-PI01 had a good binding score with Pim-2. Moreover,
HJ-PI01 (300 nmol/L) induced death receptor-dependent and mitochondrial apoptosis as well as autophagic death in MDA-MB-231
cells. In MDA-MB-231 xenograft mice, administration of HJ-PI01 remarkably inhibited the tumor growth and induced tumor cell
apoptosis in vivo. Co-administration of HJ-PI01 with lienal polypeptide could improve the anti-tumor activity of HJ-PI01 and reduce its
toxicity.
Conclusion: The newly synthesized compound, HJ-PI01, can induce death receptor/mitochondrial apoptosis and autophagic cell death
by targeting Pim-2 in human breast cancer cells in vitro and in vivo.
Keywords: Pim-2 inhibitor; HJ-PI01; breast cancer; MDA-MB-231 cells; apoptosis; autophagy; chlorpromazine; PI003; lienal polypeptide
Acta Pharmacologica Sinica advance online publication, 11 Jul 2016; doi: 10.1038/aps.2016.60
three members, Pim-1, Pim-2 and Pim-3, which show high
Introduction
The proviral insertion sites in Moloney murine leukemia homology with one another. Pim-1 and Pim-3 are 71% iden-
virus (Pim) proteins are well-known to be a family of short- tical at the amino acid level, while Pim-1 and Pim-2 share
lived serine/threonine kinases that are highly evolutionarily 61% homology[2]. Elevated expression of the Pim kinases has
conserved in multicellular organisms[1]. They belong to the been detected in hematologic malignancies and certain solid
Ca2+/calmodulin-dependent protein kinase group. Pim has tumors[3].
Regarding cancer biology, Pim family members are weak
# These authors contributed equally to this work.
To whom correspondence should be addressed.
E-mail profhj@163.com (Jian HUANG);
zhulingjuanadele@163.com (Ling-juan ZHU)
Received 2016-01-01 Accepted 2016-04-21
oncogenes that can contribute to tumorigenesis by selectively
*
enhancing the cells’ tumorigenic capabilities[4, 5]. Pim kinases
are overexpressed in solid tumors, such as colon cancer[6],
prostate cancer[7] and hematologic malignancies, including

www.nature.com/aps
2
Zhao YQ et al
lymphomas[8, 9], chronic lymphocytic leukemia (CLL) and becco's modified Eagle’s medium (DMEM), primary fibroblast
acute leukemia[10, 11]. Pim kinases are an active target for drug cell culture medium, fetal bovine serum (FBS), trypsin/EDTA
discovery research. Due to the known implications of Pim-1 (TE) and TRIzol, as well as other cell culture plates, were
in tumorigenesis, most compounds being examined focused obtained from Fisher Scientific (Pittsburgh, PA, USA). The DC
on this kinase, but a few are focused on Pim-2 and Pim- protein assay was purchased from Bio-Rad (Bio-Rad Labora-
3. Pim-2 is a key mechanism of metastasis in breast cancer tories, Hercules, CA, USA). The TUNEL assay and Annexin
through the persistent activation of STAT3, which may lead to V-FLUOS Staining Kit were supplied by Roche (Mannheim,
the epithelial–mesenchymal transition (EMT)[12]. Nevertheless, Germany). The Avidin-biotin-HRP complex was purchased
once Pim-1 was inhibited, Pim-2 expression was increased. from Thermo (Fremont, CA, USA). All of the antibodies used
Moreover, the crystal structure of Pim-2 has been reported, in the experiments were purchased from Cell Signaling Tech-
but no specific Pim-2-targeted compounds have been reported nology (Danvers, MA, USA) and Abcam (Cambridge, UK).
to date. Pim-2 has been proposed to mediate the anti-apop- The lienal polypeptides were purchased from the Fengsheng
totic properties of oncogenes, such as BCR/ABL, FLT3 and Pharmaceutical Company (Jilin, China).
Jak2 mutants[13–18]. To compensate for the loss of Pim-1, Pim-2
expression appears to be a later event in MMLV-induced lym- Network construction
phomas[19, 20]. Thus, the discovery of drugs that target both To construct the Pim-2 network, we modified the global
Pim-2 and Pim-1 is very important and necessary.
human PPI network from PrePPI, which is a database of pre-
Previous reports showed that the phenothiazine derivative dicted and experimentally determined human protein-protein
chlorpromazine, a widely used neuroleptic drug for the con- interactions (PPIs). The predicted interactions are assigned
trol of psychosis, inhibits the proliferation of various tumor a likelihood using a Bayesian framework that combines
cells[21]. The anti-proliferative effect of chlorpromazine has structural, functional, evolutionary and expression informa-
been linked to mitotic arrest through the inhibition of the tion. The unified conceptual framework of the PPI network
motor activity of the mitotic kinesin KSP/Eg5 in HCT116 and was integrated with Cytoscape. Subsequently, we built the
A549 cells and the induction of apoptosis-independent autoph- autophagy- and apoptosis-related PPI network of Pim-2 using
agic cell death via the Beclin-1 dependent pathway in PTEN- the Gene Ontology (GO) consortium with the condition that
null U-87MG glioma cells[22, 23]. Chlorpromazine also inhibited many proteins that interacted with Pim-2 are involved in
intracranial xenograft tumor growth in vivo. A previous study autophagy or apoptosis and are differentially expressed in the
reported that a newly synthesized pan-Pim inhibitor, PI003, iTRAQ analysis .
was obtained by modifying the chemical structure of chlor-
promazine[24]. PI003 induced apoptosis via the death receptor Molecular docking
and mitochondrial pathways by targeting Pim-1, Pim-2 and The initial three-dimensional geometric coordinates of the
Pim-3 in HeLa cells. Similarly to most pan-Pim inhibitors, the X-ray crystal structure of Pim-2 was retrieved from the RCSB
efficacy of PI003 against Pim-2 was relatively poor.
Brookhaven Protein Data Bank (PDB) (http://www.rcsb.org/
In this study, we identified a newly synthesized Pim-2 inhib- pdb/home/home.do) (PDB entry: 2IWI). Then, we used the
itor, HJ-PI01, by in silico analysis and experimental validation. Accelrys Discovery Studio version 3.5 with CHARMm force-
We first constructed the human protein-protein interaction field parameters to dock pre-generated conformations of
(PPI) network and modified it as an autophagy- and apop- drugs into Pim-2 to virtually screen modified inhibitors. We
tosis-related PPI of Pim-2 to reveal its mechanisms. Subse- performed flexible-ligand docking to a rigid receptor with
quently, we synthesized and screened a candidate compound, LibDock, in which the drugs were allowed to be flexible and
HJ-PI01, based on the structural characteristics of Pim-2. The structurally rearranged in response to Pim-2. The results were
apoptosis- and autophagy-inducing efficacies of HJ-PI01 re-ranked by CDOCKER. HJ-PI01 was selected based on its
were verified through in vitro experiments. Then, iTRAQ and scores and structure. Using the Discovery Studio program, we
MS/MS analysis were performed to profile the differentially virtually modified chlorpromazine into HJ-PI01, which exhib-
expressed proteins in HJ-PI01-treated MDA-MB-231 cells to its an obvious increase in every score.
explore the mechanisms of the anti-tumor efficacy of HJ-PI01.
Altogether, we discovered and identified HJ-PI01 as a new Chemical synthesis
compound that targets Pim-2, which sheds new light on breast All reactions requiring anhydrous conditions were performed
cancer therapy with regard to Pim-2.
under an Ar or N2 atmosphere. The chemicals and solvents
were either AR grade or purified by standard techniques.
Thin layer chromatography (TLC) used silica gel plates GF254;
the compounds were visualized by irradiation with UV light
Materials and methods
Reagents
The chemicals, solvents, dimethyl sulfoxide (DMSO), and/or by treatment with a solution of phosphomolybdic acid
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (20% wt in ethanol), followed by heating. Column chromatog-
(MTT), HEPES, Triton X-100, sodium orthovanadate, sodium raphy was performed using silica gel with the eluent indicated
fluoride, edetic acid, PMSF, leupeptin and Tween-20 were pur- in parentheses. 1H-NMR and 13C-NMR were performed on a
chased from Sigma–Aldrich (St Louis, MO, USA). The Dul- Bruker Avance III 400 MHz spectrometer (Bruker, Billerica,
Acta Pharmacologica Sinica

www.chinaphar.com
Zhao YQ et al
3
MA, USA) using CDCl3 or DMSO-d6 as a solvent at room tem- 138.19, 133.76, 128.77, 128.69, 123.37, 122.09, 115.86, 115.79,
perature. The chemical shifts are expressed relative to TMS 114.95, 114.49, 108.61, 108.41, 101.03, 100.83.
(=0 ppm), and the coupling constants J are expressed in Hz.
The purity of the compound screened in the biological assays HJ-PI08
was determined to be ≥97% by HPLC (Agilent 1100 HPLC sys- 1H-NMR (400 MHz, DMSO-d6) δ 7.86–7.45 (m, 2H), 7.32–7.07
tem) with a photodiode array detector. An Atlantis C18 (150 (m, 3H), 7.07–6.85 (m, 2H), 2.08 (s, 3H). 13C-NMR (100 MHz,
mm×4.6 mm, id 5 μm) (Waters, Milford, MA, USA) was used DMSO-d6) δ 170.35, 160.13, 157.61, 150.17, 142.92, 126.73,
with a gradient elution of methanol and HPLC-grade water as 126.19, 124.38, 122.04, 119.09, 117.25, 115.37, 110.02, 26.13.
the mobile phase at a flow rate of 1 mL/min. The HRMS data
were obtained using a Bruker micro-TOF-Q instrument or a HJ-PI09
TOF-MS instrument (Bruker, Billerica, MA, USA).
1H-NMR (400 MHz, DMSO-d6) δ 7.95–7.46 (m, 2H), 7.29–6.80
(m, 5H), 1.42 (s, 9H). 13C-NMR (100 MHz, DMSO-d6) δ 158.51,
155.99, 150.23, 146.34, 129.44, 126.58, 125.13, 124.63, 121.59 (s),
113.68, 111.39, 81.20, 28.33.
Chemical data of the synthetic compounds
HJ-PI01
1H-NMR (400 MHz, DMSO-d6) δ 7.59 (s, 2H), 7.23–7.12 (m,
4H), 7.09 (s, 2H), 2.08 (s, 3H). 13C-NMR (100 MHz, DMSO-d6) HJ-PI10
δ 170.35, 150.17, 126.73, 126.19, 124.47, 124.29, 117.25, 26.13.
1H-NMR (400 MHz, DMSO-d6) δ 7.10 (dd, J=28.0, 24.0 Hz,
2H), 6.95 (d, J=20.0 Hz, 1H), 6.79 (d, J=15.6 Hz, 1H). 13C-NMR
(100 MHz, DMSO-d6) δ 163.72, 161.20, 141.93, 141.26, 133.76
HJ-PI02
1H-NMR (400 MHz, DMSO-d6) δ 7.18–7.10 (m, 1H), 7.01 (dd, (s), 127.74, 123.37, 122.09, 116.71, 114.95, 114.49, 109.13, 108.92,
J=6.3, 3.8 Hz, 1H), 6.98 (t, J=2.2 Hz, 2H), 6.94 (d, J=5.1 Hz, 1H). 101.44.
13C-NMR (100 MHz, DMSO-d6) δ 141.93 (s), 133.76, 123.37,
122.09, 114.95, 114.49.
HJ-PI11
1H-NMR (400 MHz, DMSO-d6) δ 7.60 (ddd, J=15.1, 10.3, 7.3
Hz, 2H), 7.42–7.10 (m, 3H), 7.15–6.68 (m, 3H), 2.08 (s, 3H). 13C-
HJ-PI03
1H-NMR (400 MHz, DMSO-d6) δ 7.67–7.54 (m, 2H), 7.20–7.11 NMR (100 MHz, DMSO-d6) δ 170.35, 167.23, 164.70, 150.17,
(m, 4H), 7.09 (s, 2H), 1.42 (s, 9H). 13C-NMR (100 MHz, DMSO- 127.09, 126.19 (s), 124.38, 117.25, 112.43, 26.13.
d6) δ 150.24, 146.34, 129.44, 126.57, 125.14, 121.59, 113.67, 81.20,
28.33.
HJ-PI12
1H-NMR (400 MHz, DMSO-d6) δ 7.77–7.35 (m, 2H), 7.32–7.05
(m, 4H), 7.07–6.76 (m, 1H), 1.42 (s, 9H). 13C-NMR (100 MHz,
HJ-PI04
1H-NMR (400 MHz, DMSO-d6) δ 7.19–7.06 (m, 2H), 7.06–6.96 DMSO-d6) δ 166.19, 163.67, 150.23, 146.34, 142.86, 129.44 (s),
(m, 2H), 6.96–6.88 (m, 2H), 6.85 (d, J=1.6 Hz, 2H), 2.28 (s, 3H). 127.76, 126.58, 125.13, 122.77, 121.59, 114.41, 113.67, 102.80,
13C-NMR (100 MHz, DMSO-d6) δ 142.61, 141.94, 135.52, 133.76, 81.20, 28.33.
131.62, 123.89, 123.37, 122.09, 116.43, 114.95, 114.49, 114.02,
21.23.
HJ-PI13
1H-NMR (400 MHz, DMSO-d6) δ 7.24 (dd, J=16.0, 2.9 Hz, 1H),
7.16–6.91 (m, 2H), 6.97–6.54 (m, 5H), 2.28 (s, 3H). 13C-NMR
HJ-PI05
1H-NMR (400 MHz, DMSO-d6) δ 7.72–7.48 (m, 2H), 7.26–7.08 (100 MHz, DMSO-d6) δ 162.00, 159.48, 142.61, 138.20, 135.52
(m, 2H), 7.15–6.97 (m, 4H), 2.28 (s, 3H), 2.08 (s, 3H). 13C-NMR (s), 131.63, 128.73, 123.89, 116.43, 115.83, 114.02, 108.51, 100.93,
(100 MHz, DMSO-d6) δ 170.34, 150.16, 142.34, 138.91, 126.73, 21.23.
126.20, 125.48, 124.47, 124.29, 124.20, 123.62, 117.25, 117.10,
26.14, 21.23.
HJ-PI14
1H-NMR (400 MHz, DMSO-d6) δ 7.71 (dd, J=16.0, 2.8 Hz, 1H),
7.56 (dt, J=2.7, 1.7 Hz, 1H), 7.24–6.84 (m, 4H), 2.28 (s, 3H),
HJ-PI06
1H-NMR (400 MHz, DMSO-d6) δ 7.70–7.48 (m, 2H), 7.23–7.05 2.08 (s, 3H). 13C-NMR (100 MHz, DMSO-d6) δ 170.35, 160.13,
(m, 3H), 7.04–6.94 (m, 2H), 2.28 (s, 3H), 1.42 (s, 9H). 13C-NMR 157.61, 142.92, 142.35, 138.91, 125.48, 124.20, 123.62, 122.04,
(100 MHz, DMSO-d6) δ 150.24, 146.34, 143.40, 140.80, 129.44, 119.09, 117.10, 115.37, 110.02, 26.13, 21.23.
128.44, 126.78, 126.57, 125.22, 125.14, 121.59, 113.67, 113.10,
81.20, 28.33, 21.23.
HJ-PI15
1H-NMR (400 MHz, DMSO-d6) δ 7.71 (dd, J=8.0, 1.4 Hz, 1H),
7.56 (d, J=7.4 Hz, 1H), 7.26–6.76 (m, 4H), 2.28 (s, 3H), 1.42 (s,
HJ-PI07
1H-NMR (400 MHz, DMSO-d6) δ 7.29–7.09 (m, 2H), 7.09–6.99 9H). 13C-NMR (100 MHz, DMSO-d6) δ 158.51, 155.99, 150.23,
(m, 2H), 6.95 (ddd, J=25.7, 11.6, 8.2 Hz, 3H), 6.87–6.72 (m, 1H). 143.40, 142.66, 140.80, 128.45, 126.78, 125.22, 124.63, 113.66,
13C-NMR (100 MHz, DMSO-d6) δ 162.00, 159.48, 141.93, 138.22, 113.10, 112.64, 112.44, 111.39, 81.20 (s), 28.33, 21.23.
Acta Pharmacologica Sinica

www.nature.com/aps
4
Zhao YQ et al
Cell culture and the MTT assay
the cells were collected and fixed with 500 μL of PBS and 10
MDA-MB-231 cells, MDA-MB-468 cells, MDA-MB-436 cells mL of 70% ethanol overnight at 4°C; then, after washing twice
and MCF-7 cells were cultured in Dulbecco’s modified Eagle’s with PBS, the cells were incubated with 1 mL of PI staining
medium (DMEM), and HUM-CELL-0056 cells were cultured solution (50 mg/L PI and 1 g/L RNase A) for 30 min at 4°C
in primary fibroblast cell culture medium with 10% (v/v) heat- and measured by flow cytometry (Becton Dickinson, Franklin
inactivated fetal bovine serum and incubated in a humidified Lakes, NJ, USA).
incubator with 5% CO2. The cells were transferred into 96-well
plates at a density of approximately 5.0×104 cells/mL. The iTRAQ and LC-MS analysis
cytotoxicity of the test compounds was measured using the The ER/microsomal vesicle-enriched fractions of MDA-
MTT assay as follows: MDA-MB-231 cells, MDA-MB-468 cells, MB-231 cells treated with or without HJ-PI01 (300 nmol/L
MDA-MB-436 cells and MCF-7 cells were cultured in DMEM, HJ-PI01 for 24 h) were prepared using a method similar to
and HUM-CELL-0056 cells were cultured in primary fibro- that in a previous report[27]. Briefly, the proteins were digested
blast cell culture medium in the presence of a test compound with trypsin (10:1) overnight at 37°C. The resulting peptides
for 24 h, and then 10 μL of MTT (5 mg/mL) were added to were subsequently labeled using an iTRAQ 4-plex kit (AB
the cells in each well. After 4 h of culture, the medium was SCIEX, Dorchester, England) according to the manufacturer’s
removed, and the blue formazan crystals that had formed instructions. To ensure reproducibility and exclude label-
were dissolved in dimethyl sulfoxide. The absorbance of the ing bias, each pair of protein samples (control and HJ-PI01-
formazan product generated from the MTT was measured at treated) were labeled with two different labeling strategy.
570 nm (Bio-Rad Model 680, Bio-Rad, Hercules, CA, USA).
The digested samples were labeled with different iTRAQ tags
as follows: control, iTRAQ 114 and HJ-PI01-treated, iTRAQ
115. After labeling, the samples were pooled and dried. For
Molecular dynamics (MD) simulations
MD simulations were performed with the GROMACS (version the sample fraction, the iTRAQ-labeled samples were recon-
4.5.5) software package to monitor the binding interactions stituted using a strong cation exchange (SCX) buffer A (10
between Pim-2 and HJ-PI01[25]. The topological parameters mmol/L monobasic potassium phosphate, pH 2.75, and 25%
of the ligands were constructed with the Dundee PRODRG acetonitrile), and the pH values of the samples were adjusted
server[26]. The topology of Pim-2 was edited with Amber force to 2.5–3 with phosphoric acid. For the LC-MS analysis, the
field 99SB, and the small molecule was edited with Amber samples were reconstituted in 0.1% formic acid. The MS
general force field. The complex was immersed in a cubic analysis was performed on the TripleTOF 5600 system in the
box of simple point charge (SPC) water molecules. Eight and information-dependent acquisition mode. The mass spectrom-
eleven sodium counter-ions were added by replacing the etry data files were processed by Protein Pilot 4.0 (AB SCIEX,
water molecules to ensure the overall charge neutrality of the Dorchester, UK) using the Paragon algorithm. The mass spec-
simulated receptor system. In this MD process, 200 ps simula- trometry data were searched against all Homo sapiens protein
tions with a time step of 1.2 ns were performed, and the result- sequences in the Uniprot_sprot_201105 database. In the pres-
ing trajectory files were viewed and analyzed using the VMD ent study, only proteins whose four peak area ratios between
software.
the HJ-PI01-treated cells and control cells were >1.2 or <0.8
were accepted as differentially expressed proteins.
Autophagy assay
An electron microscopy analysis (Hitachi 7000, Tokyo, Japan) Western blot analysis
was performed to observe the autophagic vacuoles. The cells The MDA-MB-231 cells were treated with 300 nmol/L HJ-PI01
were transfected with the GFP-LC3 plasmid using the Lipo- for 0, 12, 24, 36 and 48 h. Both the adherent and floating
fectamine 2000 reagent (Invitrogen) according to the manu- cells were collected. The cell pellets were resuspended in
facturer’s instructions. The fluorescence of GFP-LC3 was a lysis buffer, consisting of 50 mmol/L HEPES, pH 7.4, 1%
observed under a fluorescence microscope. 3-Methyl adenine Triton X-100, 2 mmol/L sodium orthovanadate, 100 mmol/
(3-MA) is an established, selective inhibitor of the Class III PI3 L sodium fluoride, 1 mmol/L edetic acid, 1 mmol/L PMSF,
kinase Vps34 and is a reversible inhibitor of autophagy that 10 mg/L aprotinin (Sigma) and 10 mg/L leupeptin (Sigma),
blocks the recruitment of lipids to the autophagosome. We and lysed for 1 h at 4°C. After centrifugation at 12000 rounds
investigated the effect of 3-MA to determine whether HJ-PI01- per minute for 15 min, the protein content of the supernatant
induced cell death involves an autophagic mechanism.
was determined using the Bio-Rad DC protein assay. Equal
amounts of total protein were separated on 12% SDS-PAGE
gels and transferred to nitrocellulose membranes; the mem-
Apoptosis assay
The morphology of the cells undergoing apoptosis was branes were soaked in blocking buffer (5% skimmed milk).
observed under a phase-contrast microscope. For the Annexin The proteins were detected using polyclonal antibodies and
V/PI staining assay, the cells were stained with the Annexin visualized using anti-rabbit or anti-mouse IgG conjugated
V-FLUOS Staining Kit according to the manufacturer’s proto- with peroxidase (HRP), and 3,3’-diaminobenzidine tetrahy-
col (Roche, Mannheim, Germany) and observed under a fluo- drochloride (DAB) was used as the HRP substrate.
rescence microscope. To measure the ratio of apoptotic cells,
Acta Pharmacologica Sinica

www.chinaphar.com
Zhao YQ et al
5
Animal model
binding pocket, LYS40, THR45, VAL62, LEU120 and ALA122,
The experimental protocol was approved by the Animal which may influence the effect of the inhibitor on Pim-2 in the
Experimental Ethics Committee of Shenyang Pharmaceutical subsequent docking studies.
University (Permit Number: SYPU-IACUC-C2015-0721-001).
Healthy female nude mice (BALB/c, 6–8 weeks of age, non- Chemical synthesis and anti-proliferative effect of HJ-PI01
fertile, 18–20 g of weight) were subcutaneously injected with 3-Iodophenol could react with 1-fluoro-2-nitrobenzene in the
MDA-MB-231 cells (1×107 cells/mouse). When the tumors presence of K2CO3 to produce the Williamson reaction product
reached 100 mm3 in volume (calculated as V=L×W×W/2), the 1-iodo-2-(2-nitrophenoxy) benzene, and then the product was
mice were treated with hydroxypropyl-cyclodextrin (vehicle reduced by Fe-HCl; therefore, the nitro group was converted
control), HJ-PI01 (40 mg·kg-1·d-1, po), or HJ-PI01 (40 mg·kg-1·d-1, into an amino moiety. The amino moiety was protected with
po) + the lienal polypeptide (LP) (50 mg·kg-1·d-1, ip). The an acetyl group after treatment with acetic anhydride. The
animals were euthanized after 10 d. The tumor tissues were condensation of the intermediate in the presence of DMEDA
isolated and frozen in liquid nitrogen or immediately fixed in and K2CO3 in refluxing dioxane gave the adduct, which was
formalin.
finally purified by silica-gel column chromatography using
hexane and ethyl acetate as an eluent to obtain the final
products HJ-PI01 to HJ-PI15 (Figure 2). The potential anti-
Immunohistochemical analysis of KI67 and the TUNEL assay
The tumor tissues obtained from the in vivo studies were proliferative activity of all compounds was tested in MDA-
rinsed with PBS and fixed in 4% paraformaldehyde. The sam- MB-231 cells, MDA-MB-468 cells, MDA-MB-436 cells and
ples were dehydrated in a gradient ethanol series, embedded MCF-7 cells (Table 1). These compounds generally exhibited
in paraffin, and sectioned (5 μm). The deparaffinized sections moderate or remarkable anti-proliferative potency, and the
were stained with KI67 antibodies. The samples were incu- most effective compound, HJ-PI01, displayed excellent inhibi-
bated overnight with biotinylated secondary antibodies. The tory activity of up to 76.5% at a 1 μmol/L concentration in
antibodies were detected with an avidin-biotin-HRP complex MDA-MB-231 cells. Thus, we used MDA-MB-231 cells for
(Thermo Scientific, Fremont, CA, USA) and DAB as the chro- the subsequent experiments. According to the molecular
mogen. The nuclei were counterstained with hematoxylin. docking results with Pim-2, we selected HJ-PI01 for the in
For the TUNEL assay, the sections were permeabilized with vitro cellular examination. To examine the effect of HJ-PI01
0.1% Triton X-100 plus 0.1% sodium citrate and then incu- on MDA-MB-231 cell proliferation in comparison with chlor-
bated with 50 μL of the TUNEL reaction mixture (Roche) for promazine and the pan-Pim inhibitor PI003, the MTT assay
60 min at 37°C. After rinsing with PBS three times, 50 μL of was performed to measure the inhibition of MDA-MB-231 cell
converter-POD were added, and the tissues were incubated in proliferation following treatment with different concentra-
a humidified chamber for 30 min at 37°C. DAB substrate was tions of HJ-PI01, chlorpromazine and PI003. HJ-PI01 remark-
then added, followed by counterstaining with hematoxylin. ably inhibited MDA-MB-231 cell growth in a dose-dependent
The positive cells were counted in six fields per tumor sample. manner, and treatment with 300 nmol/L of HJ-PI01 for 24 h
The results are expressed as the average±SD of the tumors in resulted in almost 50% inhibition of MDA-MB-231 cell growth.
each group.
In contrast, 750 nmol/L chlorpromazine and 460 nmol/L PI003
were required to inhibit cell growth (Figure 3A). The results
showed that HJ-PI01 exhibited a substantial improvement in
Statistical analysis
All of the presented data and results were confirmed in at least inhibiting MDA-MB-231 cell growth compared with chlor-
three independent experiments. The data are expressed as promazine and PI003. We also determined the concentration-
the mean±SD. Statistical comparisons were made by one-way dependent effects of HJ-PI01 on normal, non-cancer HUM-
ANOVA or Student’s t-test. P<0.05 was considered statisti- CELL-0056 cells (Human Cardiac Fibroblast Cells) (Figure 3B).
cally significant.
The 300 nmol/L HJ-PI01 treatment resulted in less than 10%
inhibition of cell growth at 24 h, indicating HJ-PI01 has very
weak toxicity in normal non-cancer cells.
Results
Network construction
Based on the PrePPI, we selected 602 proteins that interacted Molecular docking and molecular dynamics (MD) simulations of
with Pim-2 (Supplementary Table S1). Subsequently, we iden- the HJ-PI01-Pim-2 complex
tified the autophagy-related proteins (47 protein pairs) and In the molecular docking assay, we found that a hydrogen
apoptosis-related proteins (160 protein pairs) by functional bond formed between HJ-PI01 and the ILE164 residue of Pim-
annotation, which were used to construct the pathways of 2, whereas the other part of the compound formed a hydro-
Pim-2 (Figure 1A). This result showed the potential mecha- phobic interaction with the TYR203 residue of Pim-2. This
nisms of Pim-2 in autophagy and apoptosis. A variety of binding model showed that HJ-PI01 could influence the ATP
sequence alignments were utilized to further characterize the binding pocket to inhibit Pim-2. Thus, HJ-PI01 had a good
sequence information in one cluster that included Pim-1, Pim-2 binding score with Pim-2 (Figure 4A).
and Pim-3 (Figure 1B). The amino acid sequence of Pim-2
The MD simulations of the HJ-PI01-Pim-2 complex were
displayed important differences in the residues in the ATP performed for 1.2 ns. The RMSD values of the Pim-2 backbone
Acta Pharmacologica Sinica

www.nature.com/aps
6
Zhao YQ et al
Figure 1. The PPI network of Pim-2 and sequence alignment of the kinase domains of the Pim family. (A) All of the proteins that interacted with Pim-
2 are listed. The red, filled circles represent apoptosis-related proteins, and the green, filled circles represent autophagy-related proteins. Moreover,
the green circles with the red border represent apoptosis- and autophagy-related proteins. (B) Sequence alignment of the kinase domains of Pim-1,
Pim-2 and Pim-3. The conserved catalytic residues within the binding site were highlighted in the rectangular boxes, and the regions showing the most
frequent variation were colored in red.
atoms in the initial, minimized structure, which are plotted in deviation. The study demonstrated that the trajectories of the
Figure 4B, were calculated through the phase of the simula- MD simulations of the complex were stable after 1 ns. Based
tion to evaluate the relative stability of the MD trajectories and on the results of the MD simulations, the compound HJ-PI01
the difference in the stabilities in the MD simulations. The directly targeted the ATP-binding site of Pim-2 and stably
HJ-PI01-Pim-2 complex reached equilibrium after 1 ns of the interacted with Pim-2. Considering that Pim-2 contributes
simulation phase. As shown in Figure 4B, the RMSD value for to tumorigenesis, it is reasonable to speculate that the Pim-
the HJ-PI01-Pim-2 complex was 0.4 Å, with less than 0.35 Å 2-targeting compound HJ-PI01 is a promising clinical agent for
Acta Pharmacologica Sinica

www.chinaphar.com
Zhao YQ et al
7
Table 1. Screening the candidate small molecule compounds.
Inhibitory rate at 1 umol/L
Entry
R1
R2
R3
Compound
MDA-MB-231
MDA-MB-468
MDA-MB-436
MCF-7
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
H
H
H
3-CH3
3-CH3
3-CH3
H
H
H
3-F
3-F
H
H
H
H
H
Ac
H
Boc
H
Ac
Boc
H
Ac
Boc
H
Ac
Boc
H
HJ-PI01
HJ-PI02
HJ-PI03
HJ-PI04
HJ-PI05
HJ-PI06
HJ-PI07
HJ-PI08
HJ-PI09
HJ-PI10
HJ-PI11
HJ-PI12
HJ-PI13
HJ-PI14
HJ-PI15
76.5
10.1
21.2
24.3
15.6
31.2
27.8
32.1
33.7
10.6
17.7
37.2
33.7
38.1
21.1
44.0
8.1
39.5
14.7
26.4
27.1
16.9
34.6
19.5
28.3
30.5
7.6
11.4
41.8
35.9
53.6
29.5
22.9
8.8
31.3
27.1
24.3
23.0
10.9
25.3
25.2
14.2
8.9
38.3
37.0
36.5
10.9
12.2
11.4
21.7
38.6
12.7
13.8
24.7
9.0
11.3
14.9
26.8
23.2
6.9
H
6-F
6-F
6-F
H
H
H
6-F
6-F
6-F
3-F
3-CH3
3-CH3
3-CH3
Ac
Boc
Figure 2. Chemical synthesis of HJ-PI01. 3-iodophenol could react with 1-fluoro-2-nitrobenzene in the presence of K2CO3 to produce the Williamson
reaction product 1-iodo-2-(2 nitrophenoxy) benzene, and then the product was reduced by Fe-HCl. The amino group was protected with an acetyl moiety
after treatment with acetic anhydride. The condensation of the intermediate in the presence of DMEDA and K2CO3 in refluxing dioxane yielded the
adducts HJ-PI01 to HJ-PI15.
anti-tumor drug discovery.
suggesting that HJ-PI01 induces MDA-MB-231 cell autophagy
(Figure 5B). Additionally, HJ-PI01 induced a time-dependent
increase in the accumulation of the LC3-II form in MDA-
HJ-PI01 induces autophagic death in MDA-MB-231 cells
As previous evidence has highlighted the important roles of MB-231 cells (Figure 5C). Moreover, the results of the Western
autophagy in the Pim-2-regulated network in breast cancer, blots also displayed a time-dependent increase in the Beclin-1
we explored whether autophagy is induced and required for levels, which indicated that autophagy was activated. In addi-
the HJ-PI01-mediated inhibition of cell proliferation. The for- tion, to examine whether HJ-PI01 could induce autophagic
mation of autophagic vacuoles was assessed by transmission degradation, the steady-state levels of p62, which is associated
electron microscopy. In contrast to the normal morphology of with LC3 turnover and is degraded through the autopha-
the control cells, the HJ-PI01-treated cells showed typical char- gic pathway, were analyzed by Western blot analysis. As
acteristics of autophagy, such as the presence of autophagic expected, the HJ-PI01 treatment also induced p62 degradation
vacuoles that contained cellular organelles (Figure 5A). The in a time-dependent manner. To further confirm the role of
HJ-PI01 treatment resulted in extensive GFP-LC3 localization, HJ-PI01-induced autophagy, an autophagy inhibitor, 3-MA,
Acta Pharmacologica Sinica

www.nature.com/aps
8
Zhao YQ et al
Figure 3. Cell proliferation rate of MDA-MB-231 cells and the effect of HJ-PI01 on Human Cardiac Fibroblast Cells, as measured by the MTT assay.
(A) The cells were treated with different concentrations of HJ-PI01 or chlorpromazine for the indicated times, and the data were representative of
three independent experiments. HJ-PI01 exhibited a 1.56-fold greater inhibitory effect on the MDA-MB-231 cells than chlorpromazine. (B) The HUM-
CELL-0056 cells were treated with various concentrations of HJ-PI01, 100, 200, 300, and 400 nmol/L. HJ-PI01 had a slight effect on HUM-CELL-0056
cells.
Figure 4. Molecular modeling study of the HJ-PI01 and Pim-2. (A) Cartoon showing HJ-PI01 docked into the ATP-binding site of Pim-2. (B) Time-
dependence RMSDs of Pim-2 in 1.2 ns MD simulations.
Acta Pharmacologica Sinica

www.chinaphar.com
Zhao YQ et al
9
Figure 5. HJ-PI01 induces autophagy in MDA-MB-231 cells. (A) The cellular ultrastructure was examined by transmission electron microscopy. (B) The
MDA-MB-231 cells were transfected with GFP-LC3 and then treated with 300 nmol/L HJ-PI01 for 24 h; the increase in the number of GFP-LC3 puncta
was observed under a fluorescence microscope. (C) The cells were treated with 300 nmol/L HJ-PI01 for the indicated time periods, followed by Western
blot analysis to detect the LC3, Beclin-1 and p62 levels. β-Actin was used as an equal loading control. (D) The cells were treated with HJ-PI01 (300
nmol/L) for 24 h with or without pre-treatment with the autophagy inhibitor 3-MA (1 mmol/L). The cell viability ratios were measured using the MTT
assay.
was applied in combination with HJ-PI01. We found that to identify the pathway by which HJ-PI01 induced apoptosis.
treatment with 3-MA increased cell viability compared to the The increased levels of these proteins suggested that the death
cells treated with HJ-PI01 alone (Figure 5D). These results receptor pathway was activated in HJ-PI01-induced apoptosis.
suggested that HP-PI01 induces autophagic cell death in Because Bcl-2 family members play vital roles in regulating
MDA-MB-231 cells.
the mitochondrial apoptotic pathway, the levels of Bax and
Bcl-2 were detected by Western blot analysis. After treatment
HJ-PI01 induces death receptor-dependent and mitochondrial with HJ-PI01, the level of Bax was increased, whereas the
apoptosis
level of Bcl-2 was decreased. Simultaneously, caspase-9 and
Next, we examined the morphological changes in the HJ-PI01- caspase-3 were also activated by the HJ-PI01 treatment. These
treated MDA-MB-231 cells by phase-contrast microscopy. As results clearly indicate that HJ-PI01-induced apoptosis in
shown in the Figure 6A, the control cells displayed a normal MDA-MB-231 cells is mediated by the mitochondrial pathway
cell phenotype. In contrast, the HJ-PI01-treated MDA-MB-231 (Figure 7B). In brief, HJ-PI01 induces apoptosis through the
cells showed typical features of apoptosis, including the death receptor and mitochondrial pathways. In addition, we
obvious appearance of apoptotic bodies. Annexin V, a Ca2+- also observed that HP-PI01 decreased the levels of Pim-2 and
dependent phospholipid binding protein, has a strong binding Pim-2 phosphorylation, but had no effect on Pim-1 and Pim-3
affinity for PS. Thus, we used an Annexin V-FITC/PI staining activity. These results suggested that HJ-PI01 directly inhibits
kit to assess HJ-PI01-induced cell apoptosis. When the cells Pim-2 (Figure 7C). To further confirm the efficacy and speci-
were treated with HJ-PI01, a change in the subcellular localiza- ficity of HJ-PI01 in Pim-2 inhibition, PI003, a pan-Pim inhibi-
tion of phosphatidylserine was observed under a fluorescence tor, was used as a positive control. Accordingly, HP-PI01 dis-
microscope, which is a typical feature of apoptosis (Figure 6B). played more potent inhibition of p-Pim-2 than PI003, whereas
Moreover, apoptosis was further evaluated by measuring the PI003 significantly inhibited Pim-1 phosphorylation and Pim-3
number of cells in the Sub-G1 phase by flow cytometry. As expression, but not Pim-2 (Figure 7D). These results demon-
shown in Figure 6C, HJ-PI01 markedly induced an increase strated that HP-PI01 exhibits good efficacy and specificity for
in the proportion of Sub-G1 cells, which were stained with PI, Pim-2 compared with other pan-Pim inhibitors.
suggesting that HJ-PI01 induced DNA damage. Together,
these results indicate that HJ-PI01 induces apoptotic cell death Proteomics analyses of HJ-PI01-induced autophagy and
in MDA-MB-231 cells.
apoptosis
As shown in Figure 7A, we first examined the expression To explore the molecular mechanisms underlying HJ-PI01-
levels of Fas, FADD and caspase-8 by Western blot analysis induced autophagy and apoptosis, iTRAQ and MS/MS
Acta Pharmacologica Sinica

www.nature.com/aps
10
Zhao YQ et al
Figure 6. HJ-PI01 induces apoptosis in MDA-MB-231 cells. (A) The changes in cellular morphology were examined by phase-contrast microscopy. (B)
The MDA-MB-231 cells were treated with 300 nmol/L HJ-PI01 for 24 h, and the changes in the subcellular localization of phosphatidylserine were
observed under a fluorescence microscope by Annexin V/PI staining. (C) The MDA-MB-231 cells were treated with 300 nmol/L HJ-PI01 for the indicated
time periods, and the number of HJ-PI01-treated cells in the Sub-G1 phase was analyzed by flow cytometry. The increased percentage of cells in the
Sub-G1 phase showed that HJ-PI01 induced apoptosis.
analyses were employed to profile the differentially expressed for KI67, a marker of proliferation, was localized to the cell
proteins in HJ-PI01-treated MDA-MB-231 cells. A total of nuclei. The HJ-PI01 treatment significantly reduced the num-
1777 differentially expressed proteins were defined based on ber of KI67-positive cells compared to the control treatment
a high fold-change, of which 14 proteins were identified in the (Figure 8C), suggesting that HJ-PI01 inhibited tumor cell pro-
Pim-2-related PPI network. A16L1 and CISD2 are autophagy- liferation. Moreover, HJ-PI01 administration resulted in a sta-
related proteins, while PAK4, MP2K2, AKT2 and CSK22 are tistically significant increase in the number of apoptotic bodies
apoptosis-related proteins (Table 2). This result explored the in the tumor, as visualized by TUNEL assay (Figure 8C).
mechanisms of the HJ-PI01-induced autophagy and apoptosis
and the anti-cancer efficacy of Pim-2 inhibition.
Discussion
As a new class of cancer therapeutics, Pim kinase inhibitors
Anti-tumor activity of HJ-PI01 in vivo
have been frequently reported recently. Upon Pim-1 inhibi-
In this experiment, we established a xenograft mouse model tion, the prevalence of increased Pim-2 expression across dif-
using MDA-MB-231 cells to evaluate the efficacy of HJ-PI01 in ferent cancer types suggests that Pim-2 inhibitors may be a
vivo. As shown in Figure 8A, HJ-PI01 significantly inhibited treatment modality for a variety of cancers. In this study, we
tumor growth in nude mice compared with the control group constructed the Pim-2-related PPI network for the first time,
(P<0.01), which was accompanied by obvious decreases in the and all of the possible PPIs were shown. This result assessed
body, liver, spleen and kidney weights of the mice, indicat- the role of Pim-2 in the autophagy and apoptosis pathways.
ing the toxicity of HJ-PI01. Interestingly, when a lienal poly- We further focused on the differences among the Pim kinase
peptide was used in combination with HJ-PI01, it not only family and presented a new binding model for Pim-2 inhibi-
improved the anti-tumor activity of HJ-PI01 but also reduced tors. After selective modification of candidate pan-Pim inhibi-
the toxicity of HJ-PI01 (Figure 8A and 8B). Immunoreactivity tors, we obtained a novel candidate compound, HJ-PI01.
Acta Pharmacologica Sinica

www.chinaphar.com
Zhao YQ et al
11
Figure 7. Immunoblot analysis of the anti- and pro-apoptotic proteins and cell death in the HJ-PI01-treated MDA-MB-231 cells. The MDA-MB-231 cells
were treated with 300 nmol/L HJ-PI01 for the indicated time periods and then harvested and lysed. (A) The levels of the Fas, FADD, and caspase-8
proteins were analyzed by immunoblotting. (B) The levels of the Bax, Bcl-2, caspase-9, caspase-3 and PARP proteins were analyzed by immunoblotting. (C)
The levels of the Pim-2, p-Pim-2, Pim-1 and Pim-3 proteins were analyzed by immunoblotting. (D) The MDA-MB-231 cells were treated with 300 nmol/L
HJ-PI01 or 460 nmol/L PI003 for the indicated time periods and then harvested and lysed. The levels of the Pim-1, Pim-2, Pim-3, p-Pim-1, and p-Pim-2
proteins were analyzed by immunoblotting. β-Actin was used as an equal loading control.
Table 2. PIM2-related differential proteins identified by iTRAQ coupled with LC-MS/MS.
Accession Gene
Protein name
Peptide Cover
Biological function
Fold
No
name
count
(%)
change
Q6NXT1
Q676U5 A16L1
P80217 IN35
ANKRD54 Ankyrin repeat domain-containing protein 54
3
3
6
26.7
30.8
38.1
65
Regulation of intracellular signal transduction
Autophagy:Autophagic vacuole assembly
Cytokine-mediated signaling pathway
Negative regulation of TGF-β receptor signaling
pathway
21.4783
3.9811
2.2909
1.6293
Autophagy-related protein 16-1
Interferon-induced 35 kDa protein
E3 ubiquitin-protein ligase CHIP
Q9UNE7 STUB1
11
O96013 PAK4
Serine/threonine-protein kinase PAK 4
5
29.6
41.8
Apoptotic process:Type I programmed cell
death/Cell cycle/Cell migration
Activation of MAPK signaling pathway/VEGFR
signaling pathway
1.3932
1.3677
P36507
MP2K2
Dual specificity mitogen-activated protein
kinase kinase 2
15
P31751
P19784
P61586
9NYL2
AKT2
RAC-beta serine/threonine-protein kinase
Casein kinase II subunit alpha'
Transforming protein RhoA
Mitogen-activated protein kinase kinase
kinase MLT
4
15
20
4
28.5
50.9
82.9
29.4
Apoptotic process:Type I programmed cell death
Apoptotic process:Type I programmed cell death
Apolipoprotein A-I-mediated signaling pathway
Cell death/cell cycle arrest
0.7311
0.631
0.5445
0.413
CSK22
RHOA
MLTK
P62826
RAN
GTP-binding nuclear protein Ran
Ras GTPase-activating protein-binding protein 2
Glycogen synthase kinase-3 alpha
CDGSH iron-sulfur domain-containing protein 2
27
15
6
92.6
46.3
30.2
17
Mitotic nuclear division
Ras protein signal transduction
Cell migration
0.3767
0.3698
0.1169
0.057
Q9UN86 G3BP2
P49840 GSK3A
Q8N5K1 CISD2
3
Regulation of autophagy
Acta Pharmacologica Sinica

www.nature.com/aps
12
Zhao YQ et al
Figure 8. Anti-tumor effects of HJ-PI01 in vivo. (A, a) Body weights of mice in each group during the treatment of the tumor xenografts. The mice were
treated with hydroxypropyl-cyclodextrin (vehicle control), HJ-PI01 (40 mg·kg-1·d-1, po), or HJ-PI01 (40 mg·kg-1·d-1, po)+lienal polypeptide (LP) (50 mg·kg-1·d-1,
ip). (b–d) Liver, spleen and kidney weights of the mice in each group at the end of treatment. (e) Tumor weights of the mice in each group at the end
of treatment. (B) The tumor volume curves of each group during the treatment of the tumor xenografts. (C) Immunohistochemistry of the proliferation
marker KI67 and TUNEL analysis.
HJ-PI01 inhibited Pim-2 and induced apoptosis and autoph- HJ-PI01 compound for the first time. In some circumstances,
agy in MDA-MB-231 cells. Moreover, the in vivo experiment apoptosis and autophagy seem to be positively or negatively
showed that HJ-PI01 had remarkable anti-tumor efficacy.
interconnected, due to the molecular switches shared between
Our results suggest that HJ-PI01 may be a new therapeutic them[29, 30]. Undoubtedly, there are multiple connections
agent for the treatment of breast cancer. Interestingly, a pre- between apoptotic and autophagic processes that can jointly
vious study has found that Pim-2 kinase can confer primary seal the fate of tumor cells[31]. In this study, the pro-death
hematopoietic cells with resistance to rapamycin treatment[28]
.
autophagy-inducing effect enhanced apoptosis. This study
Therefore, Pim-2 inhibitors may generate better efficacy. More may explain the mechanism of candidate novel compounds.
importantly, we also found that HJ-PI01 induced apoptosis via To the best of our knowledge, although some Pim-1 and pan-
the death receptor and mitochondrial pathways and induced Pim inhibitors have been reported, no Pim-2-specific inhibitors
autophagic cell death as well. Some Pim-targeted compounds have been described[32–36]. Compared with other compounds,
induce apoptosis; here, we reported the double effect of the HJ-PI01 has better druggability because it is based on chlor-
Acta Pharmacologica Sinica

www.chinaphar.com
Zhao YQ et al
13
promazine, which is an FDA-approved drug. Therefore, the
Supplementary information
above-mentioned findings may shed new light on the discov- Supplementary information is available on the website of Acta
ery of additional, small-molecule Pim-2 inhibitors for future Pharmacologica Sinica.
breast cancer therapy.
In this study, we also used proteomics analyses to explore
the molecular mechanisms underlying HJ-PI01-induced
autophagy and apoptosis. Of note, we identified 14 proteins
included in the PPI network, such as the autophagy-related
proteins Atg16L (A16L1) and CDGSH Iron-Sulfur Domain
Containing Protein 2 (CISD2). Atg16L plays an essential role
in autophagy by interacting with Atg12-Atg5 to mediate the
References
1
2
3
Narlik-Grassow M, Blanco-Aparicio C, Carnero A. The PIM family of
serine/threonine kinases in cancer. Med Res Rev 2014; 34: 136–59.
Nawijn MC, Alendar A, Berns A. For better or for worse: the role of Pim
oncogenes in tumorigenesis. Nat Rev Cancer 2011; 11: 23–34.
Blanco-Aparicio C, Carnero A. Pim kinases in cancer: diagnostic,
prognostic and treatment opportunities. Biochem. Pharmacology
2013; 85: 629–43.
conjugation of phosphatidylethanolamine (PE) to LC3[37]
.
4
5
Liu JJ, Lin M, Yu JY, Liu B, Bao JK. Targeting apoptotic and autophagic
pathways for cancer therapeutics. Cancer Lett 2011; 300: 105–14.
Allen JD, Verhoeven E, Domen J, van der Valk M, Berns A. Pim-2
transgene induces lymphoid tumors, exhibiting potent synergy with
c-myc. Oncogene 1997; 15: 1133–41.
CISD2 is a negative regulator of autophagy that contributes
to antagonize BECN1-mediated cellular autophagy in the
endoplasmic reticulum[38]. As shown in the results of our pro-
teomics analysis, Atg16L was upregulated after HP-PI01 treat-
ment, which indicates that autophagy was activated. More-
over, CISD2 expression was significantly decreased, indicating
that HP-PI01 may induce BECN1-mediated autophagy, which
has been validated by our in vitro experiment. Based on these
results, explorations of the potential mechanisms of HJ-PI01-
induced autophagy would further confirm the in-depth mech-
anisms of HP-PI01 in breast cancer therapy.
However, there are still some limitations for HJ-PI01, such
as its low efficacy and increased toxicity. In our study, we
found that treatment with a lienal polypeptide in combina-
tion with HJ-PI01 could improve the anti-tumor efficacy and
reduce the toxicity induced by HJ-PI01. This incredible result
is likely due to two possible reasons. On the one hand, the
lienal polypeptide may have a certain anti-tumor activity.
On the other hand, previous studies have suggested that the
lienal polypeptide has immune modulating activity[39]; thus,
some of the HJ-PI01-regulated pathways or key proteins may
be affected by the lienal polypeptide. However, the detailed
mechanisms remain to be further studied. Additionally, these
results would also shed new light on combination drug thera-
pies in our future studies.
6
Popivanova BK, Li YY, Zheng H, Omura K, Fujii C, Tsuneyama K,
Mukaida N. Proto-oncogene, Pim-3 with serine/threonine kinase
activity, is aberrantly expressed in human colon cancer cells and can
prevent Bad-mediated apoptosis. Cancer Sci 2007; 98: 321–8.
Cibull TL, Jones TD, Li L, Eble JN, Ann Baldridge L, Malott SR,
et al. Overexpression of Pim-1 during progression of prostatic
adenocarcinoma. J Clin Pathol 2006; 59: 285–8.
7
8
Cohen AM, Grinblat B, Bessler H, Kristt D, Kremer A, Schwartz A,
et al. Increased expression of the hPim-2 gene in human chronic
lymphocytic leukemia and non-Hodgkin lymphoma. Leuk Lymphoma
2004; 45: 951–5.
9
Wang Z, Bhattacharya N, Weaver M, Petersen K, Meyer M, Gapter
L, et al. Pim-1: a serine/threonine kinase with a role in cell survival,
proliferation, differentiation and tumorigenesis. J Vet Sci 2001; 2:
167–79.
10 Chen LS, Redkar S, Bearss D, Wierda WG, Gandhi V. Pim kinase
inhibitor, SGI-1776, induces apoptosis in chronic lymphocytic leukemia
cells. Blood 2009; 114: 4150–7.
11 Amson R, Sigaux F, Przedborski S, Flandrin G, Givol D, Telerman A.
The human protooncogene product p33pim is expressed during fetal
hematopoiesis and in diverse leukemias. Proc Natl Acad Sci U S A
1989; 86: 8857–61.
12 Uddin N, Kim RK, Yoo KC, Kim YH, Cui YH, Kim IG, et al. Persistent
activation of STAT3 by PIM2-driven positive feedback loop for
epithelial-mesenchymal transition in breast cancer. Cancer Sci 2015;
106: 718–25.
13 Fu LL, Yang Y, Xu HL, Cheng Y, Wen X, Ouyang L, et al. Identification
of novel caspase/autophagy-related gene switch to cell fate decisions
in breast cancers. Cell Prolif 2013; 46: 67–75.
14 Agrawal S, Koschmieder S, Bäumer N, Reddy NG, Berdel WE,
Müller-Tidow C, et al. Pim2 complements Flt3 wild-type receptor in
hematopoietic progenitor cell transformation. Leukemia 2008; 22:
78–86.
15 Wernig G, Gonneville JR, Crowley BJ, Rodrigues MS, Reddy MM, Hudon
HE, et al. The Jak2V617F oncogene associated with myeloproliferative
diseases requires a functional FERM domain for transformation and
for expression of the Myc and Pim proto-oncogenes. Blood 2008;
111: 3751–9.
16 Mizuki M, Ueda S, Matsumura I, Ishiko J, Schwäble J, Serve H, et al.
Oncogenic receptor tyrosine kinase in leukemia. Cell Mol Biol (Noisy-
le-grand) 2003; 49: 907–22.
17 Adam M, Pogacic V, Bendit M, Chappuis R, Nawijn MC, Duyster J,
et al. Targeting PIM kinases impairs survival of hematopoietic cells
Acknowledgements
We are grateful to Prof Can-hua HUANG (Sichuan University)
for his critical language editing. This study was supported by
grants from the National Natural Science Fundation of China
(No 81303270, U1303124, 81202403, and 81573290), Liaoning
Science and Technology Project (No 2013226027-4), Jilin Sci-
ence and Technology Progress Project (No 20130204058YY),
Applied Basic Research Project of Sichuan Science and Tech-
nology Department (No 2013JY0154) and Shenyang Pharma-
ceutical University Scientific Research Fund (No ZCJJ2013407).
Author contribution
Jian HUANG and Ling-juan ZHU conceived and designed the
experiments; Yu-qian ZHAO performed the experiments; Yi-
qiong YIN, Jie LIU and Gui-hua WANG analyzed the data;
Jin-hui WANG contributed reagents/materials/analysis tools;
and Yu-qian ZHAO and Yi-qiong YIN wrote the paper.
Acta Pharmacologica Sinica

www.nature.com/aps
14
Zhao YQ et al
transformed by kinase inhibitor-sensitive and kinase inhibitor-resistant
and survival. Blood 2005; 105: 4477–83.
forms of Fms-like tyrosine kinase 3 and BCR/ABL. Cancer Res 2006; 29 Xue L, Fletcher GC, Tolkovsky AM. Autophagy is activated by apoptotic
66: 3828–35.
signalling in sympathetic neurons: an alternative mechanism of death
execution. Mol Cell Neurosci 1999; 14: 180–98.
18 Chen Y, Fu LL, Wen X, Liu B, Huang J, Wang JH, et al. Oncogenic and
tumor suppressive roles of microRNAs in apoptosis and autophagy. 30 Yu L, Wan F, Dutta S, Welsh S, Liu Z, Freundt E, et al. Autophagic
Apoptosis 2014; 19: 1177–89.
programmed cell death by selective catalase degradation. Proc Natl
19 Breuer ML, Cuypers HT, Berns A. Evidence for the involvement of pim-
Acad Sci U S A 2006; 103: 4952–7.
2, a new common proviral insertion site, in progression of lymphomas. 31 Liu B, Wen X, Cheng Y. Survival or death: disequilibrating the
EMBO J 1989; 8: 743–8.
oncogenic and tumor suppressive autophagy in cancer. Cell Death Dis
20 Bullock AN, Russo S, Amos A, Pagano N, Bregman H, Debreczeni
2013; 4: e892.
JE, et al. Crystal structure of the PIM2 kinase in complex with an 32 Pierre F, Stefan E, Nédellec AS, Chevrel MC, Regan CF, Siddiqui-Jain
organoruthenium inhibitor. PLoS One 2009; 4: e7112.
21 Hanusova V, Skalova L, Kralova V, Matouskova P. Potential anti-
cancer drugs commonly used for other indications. Curr Cancer Drug
Targets 2015; 15: 35–52.
22 Lee MS, Johansen L, Zhang Y, Wilson A, Keegan M, Avery W, et al.
The novel combination of chlorpromazine and pentamidine exerts
synergistic antiproliferative effects through dual mitotic action.
Cancer Res 2007; 67: 11359–67.
23 Shin SY, Lee KS, Choi YK, Lim HJ, Lee HG, Lim Y, et al. The
antipsychotic agent chlorpromazine induces autophagic cell death
by inhibiting the Akt/mTOR pathway in human U-87MG glioma cells.
Carcinogenesis 2013; 34: 2080–9.
24 Liu Z, He W, Gao J, Luo J, Huang X, Gao C. Computational prediction
and experimental validation of a novel synthesized pan-PIM inhibitor
A, et al. 7-(4H-1,2,4-Triazol-3-yl)benzo[c][2,6]naphthyridines: a novel
class of Pim kinase inhibitors with potent cell antiproliferative activity.
Bioorg Med Chem Lett 2011; 21: 6687–92.
33 Dwyer MP, Keertikar K, Paruch K, Alvarez C, Labroli M, Poker C, et
al. Discovery of pyrazolo[1,5-a]pyrimidine-based Pim inhibitors: a
template-based approach. Bioorg Med Chem Lett 2013; 23: 6178–
82.
34 Dakin LA, Block MH, Chen H, Code E, Dowling JE, Feng X, et al.
Discovery of novel benzylidene-1,3-thiazolidine-2,4-diones as potent
and selective inhibitors of the PIM-1, PIM-2, and PIM-3 protein
kinases. Bioorg Med Chem Lett 2012; 22: 4599–604.
35 Haddach M, Michaux J, Schwaebe MK, Pierre F, O'Brien SE, Borsan C,
et al. Discovery of CX-6258. A potent, selective, and orally efficacious
pan-Pim kinases inhibitor. ACS Med Chem Lett 2012; 3: 135–9.
PI003 and its apoptosis-inducing mechanisms in cervical cancer. 36 Garcia PD, Langowski JL, Wang Y, Chen M, Castillo J, Fanton C, et
Oncotarget 2015; 6: 8019–35.
al. Pan-PIM kinase inhibition provides a novel therapy for treating
25 Pronk S, Páll S, Schulz R, Larsson P, Bjelkmar P, Apostolov R, et al.
hematologic cancers. Clin Cancer Res 2014; 20: 1834–45.
GROMACS 4.5: a high-throughput and highly parallel open source 37 Zavodszky E, Vicinanza M, Rubinsztein DC. Biology and trafficking
molecular simulation toolkit. Bioinformatics 2013; 29: 845–54.
26 Schuttelkopf AW, van Aalten DM. PRODRG: a tool for high-throughput
of ATG9 and ATG16L1, two proteins that regulate autophagosome
formation. FEBS Lett 2013; 587: 1988–96.
crystallography of protein-ligand complexes. Acta Crystallogr D Biol 38 Chang NC, Nguyen M, Shore GC. BCL2-CISD2: An ER complex at the
Crystallogr 2004; 60: 1355–63.
27 Pathmanathan N, Balleine RL. Ki67 and proliferation in breast
cancer. J Clin Pathol 2013; 66: 512–6.
28 Hammerman PS, Fox CJ, Birnbaum MJ, Thompson CB. Pim and Akt
oncogenes are independent regulators of hematopoietic cell growth
nexus of autophagy and calcium homeostasis? Autophagy 2012; 8:
856–7.
39 Jurin M, Zarković N, Ilić Z, Borović S, Hartleb M. Porcine splenic
peptides (Polyerga) decrease the number of experimental lung
metastases in mice. Clin Exp Metastasis 1996; 14: 55–60.
Acta Pharmacologica Sinica