inhibits the growth of breast cancer cells Running title ......Aug 18, 2017 · kinase PI4KIIα has...

1

PI-273, a substrate-competitive, specific small molecule inhibitor of PI4KIIα,

inhibits the growth of breast cancer cells Running title: The First PI4KIIα

Substrate-Competitive Specific Inhibitor

Jiangmei Li1, 7 #, Zhen Gao1, 6 #, Dan Zhao2, 6#, Lunfeng Zhang1,6, Xinhua Qiao1, 6, Yingying Zhao1,

7, Hong Ding2, Panpan Zhang 5，6，7, Junyan Lu2, Jia Liu7, Hualiang Jiang2, Cheng Luo2，4*, Chang

Chen1, 3, 6*

1National Laboratory of Biomacromolecules, CAS Center for Excellence in Biomacromolecules,

Institute of Biophysics, Chinese Academy of Sciences, 15 Datun Road, Chaoyang District,

Beijing, 100101, China

2Drug Discovery and Design Center, State Key Laboratory of Drug Research, Shanghai Institute

of Materia Medica, Chinese Academy of Sciences, Shanghai, China

3Beijing Institute for Brain Disorders, No. 10 Xitoutiao, You An Men, Beijing, 100069, China

4The State Key Laboratory of Toxicology and Medical Countermeasures, Academy of Military

Medical Science, Beijing, China

5 Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences, Shanghai, China

6University of Chinese Academy of Sciences, 19 Yuquan Road, Shijingshan District, Beijing,

100049, China

7Shanghai Institute for Advanced Immunochemical Studies, ShanghaiTech University, Shanghai,

China

#These authors contributed equally to this work.

Running title: The First PI4KIIα Substrate-competitive Specific Inhibitor

Key words: Phosphatidylinositol 4-kinase IIα; Substrate-competitive small-molecule inhibitor;

Virtual screening; Breast cancer

Grant Support: This research was supported by National Key R&D Program of China

(2017YFA0504000, 2016YFC0903100 to C. Chen), Personalized Medicines-Molecular

Signature-based Drug Discovery and Development, the Strategic Priority Research Program of

Research. on August 20, 2021. © 2017 American Association for Cancercancerres.aacrjournals.org Downloaded from

Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited. Author Manuscript Published OnlineFirst on August 21, 2017; DOI: 10.1158/0008-5472.CAN-17-0484

http://cancerres.aacrjournals.org/

2

the Chinese Academy of Sciences (XDA12020316 to C. Chen), the National Natural Sciences

Foundation of China (31500693 to C. Chen, 31101021 and 81472839 to J. Li., and 81430084,

81625022 and 21472208 to C. Luo), the ‘863’ National High-Technology Development

Program of China (0A200202D03 to C. Chen), National Key Scientific Instrument &

Equipment Development Program of China (2012YQ03026010 to C. Luo）, NNCAS-2012-2 to

C. Chen, the Beijing Natural Science Foundation (7132156 to J. Li) and the Science and

Technology Commission of Shanghai Municipality (15431903100 to J. Li)

*Correspondence should be addressed to: Chang Chen, National Laboratory of

Biomacromolecules, Institute of Biophysics, Chinese Academy of Sciences, 15 Datun Road,

Chaoyang District, Beijing 100101, PR China. E-mail: [email protected]. Tel：

86-10-64888406. or Cheng Luo, Drug Discovery and Design Center, State Key Laboratory of

Drug Research, Shanghai Institute of Materia Medica, Chinese Academy of Sciences, 555

Zuchongzhi Road, Pudong District, Shanghai, 201203, PR China. E-mail: [email protected].

Tel：86-21-50271399

Competing financial interests: The authors declare no competing financial interests.



mailto:[email protected]



3

Abstract

While phosphatidylinositol 4-kinase PI4KIIα has been identified as a potential target for

antitumor therapy, the clinical applications of PI4KIIα are limited by a lack of specific

inhibitors. Here we report the first small-molecule inhibitor (SMI) of human PI4KIIα.

Docking-based and ligand-based virtual screening strategies were first employed to

identify promising hits, followed by two rounds of kinase activity inhibition validation.

2-(3-(4-Chlorobenzoyl)thioureido)-4-ethyl-5-methylthiophene-3-carboxamide (PI-273)

exhibited the greatest inhibitory effect on PI4KIIα kinase activity (IC50 = 0.47 μM) and

suppressed cell proliferation. Surface plasmon resonance (SPR) and thermal shift assays

indicated that PI-273 interacted directly with PI4KIIα. Kinetic analysis identified PI-273

as a reversible competitive inhibitor with respect to the substrate phosphatidylinositol

(PI), which contrasted with most other PI kinase inhibitors that bind the ATP binding

site. PI-273 reduced PI4P content, cell viability, and AKT signaling in wild-type MCF-7

cells but not in PI4KIIα knockout MCF-7 cells, indicating that PI-273 is highly selective

for PI4KIIα. Mutant analysis revealed a role of palmitoylation insertion in the

selectivity of PI-273 for PI4KIIα. Additionally, PI-273 treatment retarded cell

proliferation by blocking cells in G2/M, inducing cell apoptosis and suppressing

colony-forming ability. Importantly, PI-273 significantly inhibited MCF-7 cell-induced

breast tumor growth without toxicity. PI-273 is the first substrate-competitive,

subtype-specific inhibitor of PI4KIIα, the use of which will facilitate evaluations of

PI4KIIα as a cancer therapeutic target.




4

Introduction

Phosphatidylinositol (PI) is an essential phospholipid that serves as a metabolic

precursor for both phosphoinositides and inositol (Ins) phosphates (1), and its related

phosphatidylinositol kinases (PIKs) play important roles in a number of human diseases,

including cancer (2-4), malaria (5), Alzheimer's disease (6-8) and diabetes (9). Among

the phosphatidylinositol kinases, PI4KIIα is targeted to the trans-Golgi network (TGN)

and endosomes. In mammalian cells, PI4KIIα is the dominant lipid kinase generating

PI4-phosphate (PI4P) (10), which is not only the precursor of the important regulatory

phosphoinositides PI(4,5)P2 and PI(3,4,5)P3 but also is an emerging regulatory molecule

in trans-Golgi (11) and endosomal trafficking (12,13), as well as phagocytosis (14).

PI4KIIα knockout mice develop late-onset neurological features that resemble human

hereditary spastic paraplegia (15). Recently, an increasing number of reports have

identified PI4KIIα as a potential target for breast cancer therapy (3,4,16-19). We

previously reported that increased PI4KIIα expression promotes tumor growth by

altering the HER2/PI3-kinase (PI3K)/ERK signaling cascades (4) and that dual

inhibition of EGFR and PI4KIIα represents a novel strategy to combat EGFR-dependent

tumors (17). PI4KIIα is also important for the Wnt (20,21) and Akt (3) signaling

pathways. However, in contrast to the extensive study on PI3Ks, pharmacological

manipulation and basic scientific research on PI4KIIα has been limited, largely owing to

a lack of specific inhibitors.

Numerous small-molecule inhibitors (SMIs) of PIKs have been identified via

screening in the past two decades, but most are ATP-competitive inhibitors of PI3Ks




5

(22). Only

2-methyl-5-nitro-2-[(6-bromoimidazo[1,2-α]pyridin-3-yl)methylene]-1-methylhydrazid

e-benzenesulfonic acid (PIK-75)(23,24) and NCGC00012848-02 (NIH-12848) (25)

have been identified as substrate-competitive inhibitors. Isoform selectivity is

complicated by high conservation of the nucleotide-binding pocket amino acid

sequences of the different PIK isoforms, the site at which most SMIs bind. Because

isoform-selective inhibitors may reduce toxicity by decreasing off-target effects,

substrate-competitive inhibitors are urgently needed.

By combining crystal structure data, molecular dynamics (MD) simulations, and

biochemical and mutagenesis studies, we recently identified the putative PI-binding

pocket for PI4KIIα (26). The PI-binding pocket is lined by residues E157, Y159, R181,

L184, T307, D308 and L349 on the palmitoylation insertion and activation loop. In

addition to contributing to the PI-binding pocket, the palmitoylation insertion also

interacts with the membrane via an amphipathic helix, α4, and any perturbation that

weakens its membrane interaction significantly impairs its kinase activity. This insertion

does not occur in other PIKs, making it an ideal site for screening specific inhibitors. In

addition, a high-throughput and non-isotopic assay for PI4-kinase activity has been

developed (27), thereby enabling screens for PI4KIIα-specific inhibitors among large

compound libraries. Structure-based virtual screening (SBVS) has become a powerful

tool in the tool kit of medicinal chemists for rapidly enriching hits from large pools of

virtual compound libraries (28). In this study, we identified a PI substrate-competitive

SMI for PI4KIIα that notably inhibits tumor growth in a breast cancer model without




6

toxicity.

Materials and Methods

Virtual screening.

The SPECS database (http://www.specs.net) is a compound library of 199,970

single-synthesized, well-characterized small molecules. This database has been widely

used in the study of compound screening processes and drug discovery and was used as

the ligand database in this study. To refine the database to include compounds with good

drug-like properties [(molecular weight ≤ 500, log P ≤ 5, number of hydrogen bond

donors ≤ 7, number of hydrogen bond acceptors ≤ 14), molecular solubility -7

(expressed as log S, where S is the solubility in mol/L), ADMET solubility -5 (base

10 logarithm of molar solubility as predicted by regression), ADMET solubility level

2 (assigns the molecule to one of seven solubility classes based on the ADMET

solubility value)], we filtered the database using Pipeline Pilot 7.5 (Scitegic, Inc., San

Diego, CA, USA) prior to screening. The filtered database was used for a subsequent

SHAFTS-3D ligand similarity search. The SHAFTS algorithm was implemented in the

ChemMapper web server, which is a hybrid 3D molecular similarity calculation

approach designed to combine the strength of pharmacophore matching and volumetric

overlay that exhibits satisfactory “scaffold hopping” capability against several

representative kinases and has been utilized for ligand-based virtual screening (29). As

PI stearoyl-arachidonyl acyl chains are long and flexible, a substituted compound with a

butyryl group attached to the oxygen atom of both PI side chains was used. The

structure of the substituted compound was drawn with ChemBio Office 2015, and MM2



http://www.specs.net/


7

minimization was used to obtain the starting structure for the similarity search. Using

the substituted compound as the SHAFTS input molecule, the top 1000 molecules with

similarity scores > 1.0 were reserved, resulting in 822 candidates. The crystal structure

of human PI4KIIα (PDB entry: 4HNE (26)) was used as the receptor for molecular

docking. The docking accuracies of the GLIDE 5.5 (Glide, version 5.5, Schrödinger,

LLC, New York), Gold 5.0 (30-32) and Autodock 4.2 (33,34) programs were evaluated.

The substituent PI molecule was mixed with 50 decoy molecules and docked into the

PI-binding site using the three software programs. Because the GLIDE 5.5 software

provided the best PI substituent ranking, this program was used for subsequent

structural-based virtual screening. Docking studies in Gold 5.0 were performed using

the default genetic algorithm (GA) parameters and automatic settings with 100% search

efficiency. The kinase scoring function implemented in Gold was applied in the docking

calculations, and GA docking was set to be terminated if the best three solutions were

all within 1.5 Å root mean square deviations (RMSD) of each other. In Autodock 4.2,

the 50 decoy molecules and the PI substituent were docked to PI4KIIα using the

Lamarckian GA, and a 0.375 Å grid spacing was set for the energetic map calculations.

The number of GA runs was set to 10, and the standard precision (SP) GLIDE 5.5 mode

was used. The protein structure was prepared using the Protein Preparation Wizard

Workflow with default settings. The PAL loop and activation loop were defined as the

PI-binding site in which the docking grids were created (26). The 822 compounds were

docked into the defined binding site and ranked based on their GLIDE score (G-score).

Considering both the docking scores and binding modes, 522 compounds were selected




8

for further evaluation, which were then structurally clustered into 60 clusters based on

their 2D molecular fingerprints using the Cluster Molecules module in Pipeline Pilot 7.5.

Two or three candidate molecules with relatively simple chemical structures and higher

G-scores within each structural cluster were retained. Ultimately, a total of 142

compounds were selected and purchased for the first-round biochemical assay. Based on

the structures of the 7 active compounds identified by the first-round biochemical assay,

the SHAFTS method was again utilized for scaffold hopping (29). A total of 53

candidates were purchased for the second-round biochemical assay.

Expression and purification for PI4KIIα, PI4KIIβ and PI4KIIIβ.

The constructs PI4KIIα，PI4KIIβ and PI4KIIIβ were expressed as GST fusion proteins

in Escherichia coli BL21-CodonPlus (DE3) competent cells. The cells were grown at

37°C until an OD600 of 1.0 was reached, at which point they were induced by 0.3 mM

isopropyl b-D-thiogalactoside at 16°C for 18 hours. The cells were then harvested and

resuspended in buffer containing 50 mM HEPES (pH 7.5), 1 M NaCl, 2 mM DTT, 20

mM MgCl2, 1 mM phenylmethylsulfonyl fluoride and 1 mg/ml lysozyme. Cell

suspensions were homogenized with a high-pressure cell disruptor (JN BIO) at 30,000

psi, and cell debris was removed by centrifugation at 120,000 g for 40 minutes. Proteins

were purified using a GST-affinity column (GE Healthcare), and the GST tag was

cleaved overnight at 4°C with PreScission protease (GE Healthcare) in buffer

containing 50 mM HEPES (pH 7.5), 300 mM NaCl, 0.1% Triton X-100 and 2 mM

MgCl2.

Kinases activity assay.




9

The proteins used in this study were purchased from the following vendors: PI4KIIIα

from Creative BioMart (catalog no. PI4KA-161H), PI3Kα from Invitrogen (catalog no.

PV4788), PIK3Cδ from Millipore (catalog no. 14-604-K), PIK3Cβ from Millipore

(catalog no. 14-603-K), PIK3Cγ from Invitrogen (catalog no. PR8641C), AKT1 from

BPS (catalog no. 40003), AKT2 from Carna (catalog no. 01-102), and AKT3 from

Carna (catalog no. 01-103, Lot. No. 05CBS-3274). Additionally, PI4KIIα, PI4KIIβ and

PI4KIIIβ were expressed and purified as described above. The ADP-Glo kinase assay

(Promega) was used to evaluate the kinase activities of PI4KIIα, PI4KIIβ, PI4KIIIα,

PI4KIIIβ, PI3Kα, PI3Kδ, PI3Kβ and PI3Kγ according to previous reports with some

modifications(27). Briefly, 5 μl of kinase buffer containing different concentrations of

compounds (25 nM PI4KIIα, 268 nM PI4KIIβ, 83 nM PI4KIIIα, 109 nM PI4KIIIβ, 1.65

nM PI3Kα, 4.8 nM PI3Kβ, 7.6 nM PI3Kγ or 17 nM PI3Kδ) was added to each well of

the assay plate. For the PI4K assay, the buffer contained 50 mM HEPES (pH 7.5), 150

mM NaCl, 0.1% anzergent 3–14, 1 mM EDTA and 20 mM MgCl2, while the buffer for

the PI3K assay contained 50 mM HEPES (pH 7.5), 3 mM MgCl2, 1 mM EGTA, 100

mM NaCl, 0.03% CHAPS and 2 mM DTT. The kinase reaction was initiated by adding

5 μl of the substrate solution (50 μM PI(4,5)P2 and 25 μM ATP for the PI3Ks, 500 μM

PI and 25 μM ATP for the PI4Ks). The assay plates were then covered and incubated at

room temperature for 1 hour, and the reactions were stopped by adding 10 μl of

ADP-Glo reagent (Promega). After equilibration for 40 minutes, 20 μl of kinase

detection reagent (Promega) was added, and the mixture (40 μl total) was incubated for

an additional 60 minutes before reading the luminescence on a Varioskan Flash plate




10

reader (Thermo Scientific).

The kinase activities of AKT1, AKT2 and AKT3 were measured by a mobility shift

assay. Briefly, 15 μl of kinase buffer (50 mM HEPES (pH 7.5), 0.0015% Brij-35)

containing different concentrations of the compounds (2.5 nM AKT1, 0.5 nM AKT2 or

0.7 nM AKT3) was added to each well of the assay plate. Then, a mixture consisting of

10 μl of FAM-labeled peptide (GL Biochem, 3 µM) and ATP (60 μM, 188 μM or 67 μM)

was added to the assay plate, which was then incubated at room temperature for 30

minutes. Finally, the reaction was stopped by adding 25 μl of stop buffer (100 mM

HEPES (pH 7.5), 0.015% Brij-35, 0.2% coating reagent, 50 mM EDTA), and the data

were collected with the Caliper instrument (Perkin Elmer).

Thermal shift assay and cellular thermal shift assay.

The thermal shift assay was performed on a 7500 Fast Real-Time PCR system (ABI).

Each reaction solution contained 5 M PI4KIIα, 5×SYPRO orange (Sigma-Aldrich) and

the test compounds in 20 µl of buffer (50 mM HEPES (pH 7.5), 150 mM NaCl, 2 mM

MgCl2), which was heated from 25 to 95°C at a 1% ramp rate. The melting temperature

(Tm) was calculated by the Boltzmann fitting method using Protein Thermal Shift

Software Version 1.1 (ABI). Each reaction was repeated three times.

The cellular thermal shift assay (CETSA) was carried out based on a previously

described protocol with minor modifications (35,36). For the in vivo assay, MCF-7 cells

were harvested and washed with PBS after the indicated treatments. All buffers were

supplemented with complete protease inhibitor cocktail, and the cell suspensions were

freeze-thawed three times using liquid nitrogen. The soluble fraction (lysate) was




11

separated from the cell debris by centrifugation at 20,000 x g for 30 minutes at 4°C. For

the in vitro assay, the cell lysates were divided into two aliquots; one aliquot was treated

with PI-273, and the other was treated with the diluent DMSO (control) for 30 minutes.

The lysates were then divided into smaller (50 μL) aliquots and heated individually at

different temperatures for 3 minutes (S1000TM thermal cycler, BIO-RAD), followed by

cooling for 3 minutes at room temperature. The appropriate temperatures were

determined in preliminary CETSA experiments. The heated lysates were centrifuged at

20,000 x g for 30 minutes at 4°C to separate the soluble fractions from the precipitates.

The supernatants were transferred to new microtubes and analyzed by sodium dodecyl

sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) followed by western blot.

Detection and quantification of cellular PI(4)P, LPA, PI3P, PI(4,5)P2, PI(3,4,5)P3.

All the cellular lipid contents were measured using the appropriate Mass ELISA kits

(Echelon Biosciences Inc., USA) according to the manufacturer's instructions. The kits

used included the PI(4)P Mass ELISA Kit (catalog no. K4000E), the LPA Mass ELISA

Kit (catalog no. K-2800S), the PI3P Mass ELISA Kit (catalog no. K-3300), the PI(4,5)P2

Mass ELISA Kit (catalog no. K-4500) and the PI(3,4,5)P3 Mass ELISA Kit (catalog no.

K-2500S). Briefly, cellular lipid extracts were analyzed for PI(4)P, LPA, PI3P, PI(4,5)P2

or PI(3,4,5)P3 using the appropriate Mass ELISA Kit described above and normalized

by the protein concentration in the cell lysate. For measurement of PI4P in tumor tissues,

lipids extracted from equal weights (approximately 10 mg) of xenograft tumors were

analyzed for PI4P using the PI4P Mass ELISA Kit described above and normalized by

the tissue weights. Statistical analysis was performed with 6 paired cases (37).




12

Surface plasmon resonance (SPR)-based binding assay.

The SPR binding assays were performed on a Biacore T200 instrument (GE Healthcare)

at 25°C. The PI4KIIα protein was covalently immobilized on a CM5 chip using a

standard amine-coupling procedure in 10 mM sodium acetate (pH 5.0). The chip was

first equilibrated with HBS-EP buffer (10 mM HEPES (pH 7.4), 150 mM NaCl, 3 mM

EDTA, 0.05% (v/v) surfactant P20, 0.1%(v/v) DMSO) overnight. The compounds were

serially diluted in HBS-EP buffer and injected for 120 s (contact phase), followed by

120 s for the dissociation phase. The KD values of the tested compounds were

determined using the Biacore T200 evaluation software (GE Healthcare).

Mouse xenograft experiments.

Eight-week-old male BALB/c nude mice (Weitonglihua, Beijing, China) were allowed

to acclimate for 1 week under specific pathogen-free conditions in the animal facility of

the Institute of Biophysics at the Chinese Academy of Sciences. The mice were

subcutaneously injected with an MCF-7 cell suspension in a mixture of Matrigel and

DMEM, and treatment was initiated after 2 or 4 days. PI-273 was reconstituted in

DMSO and administered by intraperitoneal injection. Tumor volume was monitored via

digital calipering three times each week and calculated using the formula (smallest

diameter2×largest diameter)/2. All procedures involving animals and their care were

approved by the animal ethics committee of the Institute of Biophysics at the Chinese

Academy of Sciences.

Pharmacokinetics.

PI-273 was formulated in a 0.9% saline solution containing 10% DMSO and 0.9%




13

Tween-80. Three male Sprague-Dawley (SD) rats were intravenously treated with 3.33

ml/kg formulated PI-273, and another three were intragastrically treated with 10 ml/kg

formulated PI-273. Blood samples were collected at 0.08, 0.16, 0.33, 0.67, 1, 1.5, 2, 3, 5

hours after the intravenous and intragastrical administrations. One 0.2-ml blood sample

was collected from each rat by retro-orbital bleeding. Blood was collected into

heparin-containing tubes, and plasma was obtained by centrifugation at 4,000 rpm for

10 minutes. All protocols were based on standard operating procedures approved by the

Institutional Animal Care and Use Committee (IACUC) of the Shanghai Institute of

Materia Medica at the Chinese Academy of Sciences. The plasma concentrations of

PI-273 were quantitated by the LC-MS/MS method, and noncompartmental analysis

using Phoenix 1.3 (Pharsight, USA) was performed for all analytical measurements. The

trapezoidal method was used to calculate the area under the concentration-time curve

(AUC), where AUC0-∞ = AUC0-t + Ct/ke (ke is the elimination rate constant). The

mean residence time [(MRT) = AUMC/AUC] and the elimination half-life [(t1/2) =

0.693/ke] were also calculated.

Statistical analysis.

Statistical analyses were performed using two-tailed paired Student’s t-test.

Differences were considered statistically significant when the P-value was <0.01, as

indicated in the legends. All data are expressed as the means ± S.D.

Results

Identification of PI-273 as a PI4KIIα inhibitor based on virtual screening. In this

study, virtual screening was employed to identify promising hits (Fig. 1A). The SPECS




14

database (https://www.specs.net), containing 199,970 compounds, was used as the

ligand database, and compounds with unfavorable physicochemical properties were

filtered using Pipeline Pilot 7.5. The remaining 89,467 compounds were used in a

SHAFTS-3D ligand similarity search with a PI-substituted compound as the input

molecule, resulting in 822 candidates. The 822 compounds were docked into the

PI4KIIα protein pocket near the palmitoylation insertion using the GLIDE 5.5 program.

Based on the G-score and structural clustering, 142 structurally diverse compounds

were selected and purchased for biochemical assays. We then tested the effects of these

compounds on PI4KIIα activity at 100 μM and identified 7 that exhibited greater than

60% kinase activity inhibition (Suppl. Fig. 1A). Measuring the half-maximal inhibitory

concentration (IC50) values of these compounds for kinase activity in vitro (Suppl. Fig.

1B) and cell viability in vivo (Suppl. Fig. 1C) revealed that PI-93 (Suppl. Fig. 1D)

exhibited the strongest inhibitory effect. We then further tested the effect of PI-93 on the

intracellular lipid content. As shown in Suppl. Fig. 1E, the MCF-7 cells incubated with

20 μM PI-93 for 24 hours exhibited lower PI4P levels than DMSO-treated cells, while

the LPA, PI3P, PI(4,5)P2 and PI(3,4,5)P3 levels were not significantly different in either

sample.

Based on the SHAFTS-3D ligand similarity search, another 53 candidates were

selected and purchased for the second round of screening. Six compounds with an

inhibitory effect on PI4KIIα activity stronger than that of PI-93 were selected for further

analysis (Suppl. Fig. 2A, Fig. 1A). The chemical structures of these representative PI-93

analogues are presented in Suppl. Table 1 (in addition to the 6 selected compounds, 5



https://www.specs.net/


15

compounds that shared a common parent nucleus with PI-93 in the first-round screening

and another 3 compounds that had an inhibition effect similar to that of PI-93 in the

second-round screening have also been included) and Fig. 1B. The superimposition of

PI with these six representative active compounds are shown in Suppl. Fig. 2B. The IC50

values of the compounds of interest for PI4KIIα kinase activity in vitro and cell viability

in vivo in MCF-7 cells are presented in Suppl. Table 1.

2-(3-(4-chlorobenzoyl)thioureido)-4-ethyl-5-methylthiophene-3-carboxamide (PI-273)

had the greatest inhibitory effect of MCF-7 cell viability, and its IC50 value in the

biochemical assay was 0.47 μM. In addition, our results (Suppl. Fig. 2C) also indicated

that PI-273 had the best inhibitory effect on the cellular PI4P content among the

analogues (PI-294, PI295, PI-308), which was consistent with the analogues effects on

cell viability and AKT activity. Other PI-273 analogues, such as PI-69/107/114, that had

no effect on PI4KIIα activity also had no effect on MCF-7 cells (Suppl. Table 1). In the

PI4P content assay, PI-273 reduced the intracellular PI4P content in MCF-7 cells in a

dose- and time-dependent manner (Fig. 1C). Furthermore, as shown in Fig. 1D, PI-273

treatment could specifically inhibit PI4P production but not that of other lipids, such as

LPA, PI3P, PI(4,5)P2 and PI(3,4,5)P3. Taken together, these results suggest that PI-273 is

a structurally novel PI4KIIα inhibitor with remarkable potency against the kinase

activity of PI4KIIα.

PI-273 directly binds and reversibly inhibits PI4KIIα. To validate these potential

PI4KIIα inhibitors, direct interactions between PI4KIIα and the candidate compounds

were measured by SPR. As shown in Fig. 2A and Suppl. Fig. 3A, the interactions were




16

dose-dependent. The equilibrium dissociation constant (KD) between PI-273 and

PI4KIIα was approximately 9.49 µM based on the association rate constant (kon = 2.08

×102 M

-1·s

-1) and the dissociation rate constant (koff = 2.13 ×10

-3 s

-1), whereas the KD of

PI-277 was 9.59 µM and the KD of PI-282 was 5.91 µM (these two compounds have the

best solubility among PI-273/274/275/276/282/294/295/308, listed in Suppl. Table 1).

The direct interaction between these compounds and PI4KIIα was further confirmed by

a fluorescence-based thermal shift assay; PI-273 dose-dependently shifted the melting

temperature (Tm) of PI4KIIα by more than 4°C before reaching a plateau. This effect

was also observed in samples treated with PI-93, PI-277 and PI-294, whereas PI-69 did

not affect the PI4KIIα Tm value (Fig. 2B, Suppl. Fig. 3B), consistent with their

individual potencies for PI4KIIα inhibition. To evaluate the in vivo interaction between

PI-273 and PI4KIIα, we conducted a CETSA (35). The CETSA, developed by D.

Molina Martinez and his colleagues, is used for assessing target engagement by drugs in

vivo. The CETSA is based on the biophysical principle of the ligand-induced thermal

stabilization of target proteins, and it will likely become a valuable tool for validating

and optimizing drug target engagement (35,36). As shown in Fig. 2C, intact MCF-7

cells treated with 1 μM PI-273 for 96 hours and cell lysate treated with 20 μM PI-273

for 30 minutes exhibited shifted melting curves for PI4KIIα. Under DMSO treatment,

50% of PI4KIIα was degraded at 60°C, whereas in the PI-273-treated samples, 50%

degradation required temperatures of 70°C (intact cells) and 75°C (cell lysate). Thus,

PI-273 inhibits PI4KIIα activity via a direct binding event. We then assessed the

reversibility of the binding between PI-273 and PI4KIIα. PI4KIIα was preincubated




17

with varying concentrations of PI-273 (0-1 M) for different lengths of time (0–60

minutes) before the kinase reaction was initiated. Only the concentration of PI-273 and

not the incubation time influenced the inhibitory effect of PI-273 on PI4KIIα activity

(Fig. 2D), suggesting that the interaction between PI-273 and PI4KIIα is reversible. As

further confirmation, we evaluated the enzyme kinetics of PI4KIIα, which revealed that

the slope of the velocity-to-enzyme concentration decreased with increasing PI-273

concentration (Fig. 2E), thus indicating that PI-273 is a reversible inhibitor of PI4KIIα.

Taken together, these results strongly indicate that PI-273 directly interacts with

PI4KIIα in a reversible manner.

PI-273 is a PI-competitive inhibitor of PI4KIIα. There are three types of reversible

inhibitors: competitive, uncompetitive and anticompetitive. As described above, PI-273

is a reversible inhibitor of PI4KIIα, and enzyme kinetic analysis was performed to

determine the exact type of this inhibition (Fig. 3A and B). At varying concentrations of

ATP, increasing PI-273 decreased the maximum velocity (Vmax) of the reaction (Fig. 3A

and C), but the apparent affinity (KM, Michaelis-Menten constant) of ATP for the

binding site did not change (Fig. 3A and D), indicating that PI-273 is an

ATP-uncompetitive inhibitor. However, when we varied the concentration of PI, the

Vmax of the reaction remained unchanged (Fig. 3B and C), whereas the KM value for PI

at the binding site increased, indicating that PI-273 is a PI-competitive inhibitor (Fig.

3B and D). We next performed experiments with PI4KIIα mutants to verify this

conclusion (Fig. 3E and Suppl. Table 2). As expected, the Tm of wild-type PI4KIIα was

significantly elevated in the presence of PI-273. However, mutations or deletions in the




18

PI-binding regulation region (for example, the palmitoylation insertion, Δ165-173, or

K165E/K168E/K172E), but not those at the nucleotide-binding site (for example,

K152A, D346A), reduced the ΔTm value. We previously reported that the palmitoylation

insertion modulates kinase activity by tuning the PI-binding pocket (26,38), and these

results confirm that PI-273 is a PI-competitive inhibitor of PI4KIIα.

PI4KIIα is required for the inhibitory effect of PI-273. The selectivity of kinase

inhibitors is crucial to avoid side effects caused by off-target inhibition (39). To validate

the subtype-specific inhibitory effect of PI-273 on PI4KIIα, we tested the inhibitory

effects of PI-273 on 11 related kinases, including four PI4K subtypes (PI4KIIα, PI4KIIβ,

PI4KIIIα and PI4KIIIβ), four PI3K subtypes (PI3Kα, PI3Kβ, PI3Kγ PI3Kδ) and

three AKT subtypes (AKT1, AKT2 and AKT3). As shown in Fig. 4A, the IC50 values of

PI-273 for PI4KIIα were 0.47 µM, consistent with our previous results. The PI-273 IC50

values were much higher for the other kinases than for PI4KIIα, even for the isoform

having the greatest homology to PI4KIIα, PI4KIIβ, which had a PI-273 IC50 value more

than 30-fold higher than that for PI4KIIα. For the other PI4K subtypes (PI4KIIIα and

PI4KIIIβ), whose structures are more similar to those of PI3Ks, the IC50 values were

both more than 100 µM. Although the PI-273 IC50 value for PI3Kβ was only 10-fold

higher than that for PI4KIIα, our results shown in Fig. 1D indicate that PI-273 treatment

cannot reduce PI3P or PI(3,4,5)P3 content, suggesting that PI3Kβ is not a major target of

PI-273 in vivo. All the above results indicate that PI-273 is highly specific for PI4KIIα.

We next tested the specificity of PI-273 for PI4KIIα in vivo. We carried out

genome-scale CRISPR-Cas9 knockout screening for PI-273-resistant genes (Suppl. Fig.




19

4A). Suppl. Fig. 4B shows that PI4KIIα is the most significantly enriched target, whose

function loss can confer strong protection to MCF-7 cells against PI-273. We then

generated PI4KIIα knockout MCF-7 cells using the CRISPR-Cas9 method (Suppl. Fig.

5). As expected, obtaining MCF-7 cells in which PI4KIIα was totally knocked out

proved to be difficult. Among 200 individual monoclones, only one PI4KIIα knockout

cell line was selected, and its sequencing results indicated that PI4KIIα exists as a

multicopy gene in MCF-7 cells (Suppl. Fig. 5A). We then further confirmed the

knockout effects of this clone by western blot analysis (Suppl. Fig. 5B) and

immunofluorescence staining (Suppl. Fig. 5C). In the cell viability assay, the PI4KIIα

knockout greatly decreased the sensitivity to PI-273. Compared to the DMSO-treated

wild-type MCF-7 cells, 10 µM PI-273 could reduce the cell viability by approximately

80%, while this ratio for the PI4KIIα knockout MCF-7 cells was only 25% (Fig. 4B),

indicating that PI4KIIα is essential for the effect of PI-273 on cell viability. Because our

previous works(4,17) and other studies indicated that PI4KIIα can regulate the AKT

signaling pathway(3), we tested the effects of our compounds on AKT signaling. As

shown in Fig. 4C and Suppl. Fig. 6A, PI-273 could suppress the AKT signaling pathway

in a dose- and time-dependent manner. However, other analogues, such as PI-294

(Suppl. Fig. 6B), PI-250 (Suppl. Fig. 6C), PI-295 and PI-308 (Suppl. Fig. 6D), had

much weaker effects on this signaling pathway, consistent with their effects on cell

viability and PI4P content. These results indicate that PI4KIIα is the target for PI-273.

We next addressed whether PI4KIIα is solely essential for the effects of PI-273 in vivo

using PI4KIIα knockout cells, and we found that PI-273 reduced the PI4P content by 40%




20

in wild-type (CTR) cells but had no effect on PI4P in PI4KIIα knockout (KO) cells. By

contrast, PI-273 had no effect on other lipid (LPA, PI3P, PI(4,5)P2 or PI(3,4,5)P3)

content in wild-type cells or PI4KIIα knockout cells (Fig. 4D). Consistent with this

result, PI-273 decreased the p-AKT levels by 50% in wild-type cells but had no effect

on their phosphorylated levels in PI4KIIα knockout cells (Fig. 4E). In addition, PI-273

did not affect the PI4P level (Suppl. Fig. 7A) or the AKT signaling pathway (Suppl. Fig.

7B) in the PI4KIIα knockdown cells generated by siRNA. Taken together, the in vitro

and in vivo experimental results indicate that PI-273 is a PI4KIIα-specific inhibitor.

Effect of PI-273 on breast cancer cells. Compounds PI-273, PI-274, PI-277, PI-294,

PI-295 and PI-308 inhibited PI4KIIα activity both in vitro and in vivo and had varying

effects on the viability of MCF-7 cells. PI-273 had the greatest effect on PI4KIIα

inhibition and was thus selected for further exploration. To examine the biological

effects of PI4KIIα inhibition in greater detail, we tested the inhibitory effect of PI-273

on different breast cancer cell lines: MCF-7, T-47D, SK-BR-3, BT-474, MDA-MB-468,

MDA-MB-231 (KRas mutant), SUM229PE (KRas mutant), SUM159PT (HRas mutant)

and Hs 578T (HRas mutant)(40). The characteristics of the cell lines used in this study,

including the statuses of the primary tumor, origin, estrogen receptor (ER), progesterone

receptor (PR), human epidermal growth factor receptor 2 (HER-2), Ras, mice

tumorigenicity and the PI-273 IC50 value, are summarized in Suppl. Table 3. As shown

in Fig. 5A, Ras-mutant breast cancer cell lines are less sensitive to PI-273, as the PI-273

IC50 values for MCF-7, T-47D, SK-BR-3, MDA-MB-468 and BT-474 cells were 3.5 µM,

3.1 µM, 2.3 µM, 3.9 µM and 2.1 µM, respectively, while the IC50 values for all the




21

Ras-mutant cell lines were higher than 10 µM. The 56-carboxyfluorescein diacetate

succinimidyl ester (CFDA SE) assay was used to evaluate the effects of PI-273 on cell

proliferation. As shown in Suppl. Fig. 8A, 1 µM and 2 µM PI-273 inhibited the cell

proliferation of both MCF-7 and T-47D cells in a time-dependent manner, but these

PI-273 doses did not influence the proliferation of Hs 578T cells. These results are

consistent with the inhibition of these compounds against PI4KIIα in different cell lines

and indicate that MCF-7 and T-47D cells are much more sensitive to PI-273. Cell cycle

regulation by PI-273 was examined by 3,8-diamino-5-[3-(diethylmethylammonio)

propyl]-6-phenylphenanthridinium diiodide staining and flow cytometry. Interestingly,

PI-273 blocked the cell cycle at the G2/M phase (Fig. 5B), similar to the effect of

PIK-75, another substrate-competitive inhibitor of PIKs (PI3Kα)(41). The effect of

PI-273 on cell apoptosis was evaluated by a terminal deoxynucleotidyl transferase dip

nick end labeling (TUNEL) assay. As shown in Fig. 5C, treatment with PI-273 induced

cell apoptosis in all three Ras wild-type breast cancer cells: MCF-7, T-47D and

SK-BR-3. In addition to the cell viability assay, the inhibitory effects of PI-273 on the

tumorigenicity of cells in vitro were evaluated with plate clone-forming tests (Fig. 5D)

and soft agar clone-forming tests (Fig. 5E). PI-273 exhibited prominent drug potency in

MCF-7 and T-47D cells (1 µM PI-273 reduced both plate colony and soft agar colony

growth by more than 50%). However, Hs 578T and MDA-MB-231 cells were less

sensitive to PI-273 than MCF-7 and T-47D cells. To assess whether sensitivity is

dependent on PI4KIIα expression in different cell lines, the PI4KIIα expression levels

(Suppl. Fig. 8B) and PI4P regulation effects (Suppl. Fig. 8C) were measured. The




22

results showed that all the cell lines had high PI4KIIα expression levels, and the PI4P

content could be regulated by PI-273 in all four cell lines tested. Although the PI4KIIα

expression levels were different among these cell lines, such differences could not

distinguish the sensitivity of PI-273. However, the cell sensitivities to PI-273 showed

strong correlation with the Ras status (Fig. 5 A-E and Suppl. Table 3.). Taken together,

these results indicate that PI-273 retards cell proliferation by blocking cells at the G2/M

phase, inducing cell apoptosis and suppressing cell tumorigenicity, and these effects are

dependent on the cell type.

Antitumor effect of PI-273 in a xenograft model. To determine whether PI-273 can

suppress breast cancer in vivo, we engrafted MCF-7 cells, which were confirmed to be

sensitive to PI-273 in the above experiments, into the right flank region of BALB/c

nude mice. Four days after cell injection, the mice were randomized and received either

an intraperitoneal injection of PI-273 25 mg/kg/day or vehicle (5% DMSO). The mice

were sacrificed 24 hours after the 15th

injection. PI-273 profoundly suppressed the

tumor volume (Fig. 6A) and weight (Fig. 6B) in the MCF-7 xenografts compared to the

effects of treatment with the vehicle. Another group of parallel experiments indicated

that an intraperitoneal injection of PI-273 at 50 mg/kg/2 days also inhibited MCF-7

xenograft tumor growth, although with lower efficiency than the 25 mg/kg/day injection

(Suppl. Fig. 9A and B). We also assessed the toxic effects of PI-273 in mice, and no

lethargy, weight loss (Fig. 6C) or other physical indicators of sickness were observed.

Various tissues, including liver, intestine, lung, kidney, spleen and stomach tissue, were

examined by hematoxylin eosin (HE) staining, and no tissue damage (macroscopic or




23

microscopic) was observed (Suppl. Fig. 9C). These studies establish the effectiveness

and safety of PI-273 for antitumor applications. We further assessed the utility of PI-273

in animal experiments. SD rats were treated with 0.5 mg/kg PI-273 intravenously or 1.5

mg/kg PI-273 intragastrically. Blood samples collected retro-orbitally were collected at

0.08, 0.16, 0.33, 0.67, 1, 1.5, 2, 3, 5 hours after the intravenous and intragastrical

administrations. The plasma levels were analyzed by LC-MS/MS, and the summarized

pharmacokinetic results are presented in Table 1. Compound PI-273 is retained in rats at

a half-life of 0.411 hours for intravenous administration and 1.321 hours for

intragastrical administration, and the absolute bioavailability of PI-273 is 5.1%. These

pharmacokinetic results indicate that PI-273 as lead compound showed moderate

pharmacokinetic activities. Histological examination using the TUNEL assay to detect

apoptotic cells revealed a significant increase in the number of apoptotic cells in the

PI-273-treated MCF-7 xenografts compared to those treated with vehicle (Fig. 6D). A

significant decrease in proliferation, measured by Ki-67 staining, was observed in the

PI-273-treated MCF-7 xenografts compared to the vehicle (Fig. 6E). The PI4P content

and p-AKT levels in MCF-7-induced tumors were measured to evaluate the effect of

PI-273 in the xenograft model. Consistent with the data shown above, the PI-273-treated

tumors exhibited reduced PI4P content and p-AKT levels (Fig. 6F). These results

confirm that the PI4KIIα-specific SMI PI-273 suppresses proliferation, survival, and

AKT signaling in MCF-7-induced breast cancer in vivo. The mechanism underlying the

PI-273 antitumor effect is summarized in Fig. 6G; PI-273 inhibits the kinase activity of

PI4KIIα, leading to reduction of the intracellular PI4P content and suppression of the




24

PI3K/AKT signaling pathways, resulting in retarded cell proliferation and increased

apoptosis, and, consequently, tumor growth inhibition. PI-273 shows good potential for

use in clinical trials for human breast cancer.

Discussion

Because PI4KIIα plays important roles in Golgi trafficking(14,42,43) and tumor

progression(3,4,44), SMIs of PI4KIIα have potential as chemical tools for studying the

biological function of PI4KIIα and breast cancer treatment(10,19,42). Here, PI-273 was

screened out and identified as the most potent inhibitor of PI4KIIα activity and breast

cancer cell viability. Biochemistry and kinetic analyses revealed that PI-273 directly

binds and reversibly inhibits PI4KIIα. The fact that PI-273 is a substrate-competitive

inhibitor rather than an ATP-competitive inhibitor is its greatest advantage, as this

results in kinase isoform selectivity. Further study confirmed that PI4KIIα is essential

for the effects of PI-273. As the first substrate-competitive inhibitor in the PI4K family

and the third such inhibitor in the entire PIK family (PIK-75 is a substrate-competitive

inhibitor of PI3K(23) and NIH-12848 is a substrate-competitive inhibitor of

PI5P4Kγ(25)), PI-273 is highly isoform-selective and has significant anti-breast cancer

effects without toxicity, showing its great potential for basic PI4KIIα and

pharmaceutical research.

Numerous small molecules have been previously identified as PIK inhibitors, as

summarized in Supplemental Table 4, and none affect PI4KIIα, except for EGCG(45),

resveratrol(46) and adenosine(47). In fact, we tested the effects of all the potential

candidates on PI4KIIα activity prior to screening, including EGCG, resveratrol,




25

adenosine, PI3K inhibitors (Wortmannin, LY294002, ZSTK474), a PI3Kα inhibitor

(PI103), a PI3Kδ inhibitor (IC87114), a PI3Kγ inhibitor (AS605240), and a PI4KIIIβ

inhibitor (PIK93) (Suppl. Fig. 10). As observed previously, PI4KIIα was insensitive to

PI3K and type III PI4K inhibitors(10,48,49), and the IC50 values of both resveratrol and

EGCG were greater than 100 μM. Moreover, these molecules are all broad-spectrum

inhibitors and thus cannot be used for PI4KIIα subtype-specific functional studies.

Therefore, screening subtype-specific inhibitors for PI4KIIα is of great significance.

As summarized in Supplemental Table 4, hundreds of small molecules were

identified as PIK inhibitors by screening, but most were ATP-competitive, except for

PIK-75(23) and NIH-12848(25), limiting their potential for high isoform selectivity.

The lack of structural information regarding the PIK substrate complex is a barrier to

developing substrate-competitive inhibitors because lipid substrate binding requires the

kinase to adopt a membrane-bound conformation that is likely different from the

structure observed in crystal form. We have recently overcome this limitation by

combining data generated from crystal structure, molecular docking and biochemical

studies to identify the putative PI-binding pocket of PI4KIIα. We determined that a

PI4KIIα-unique insertion, the palmitoylation insertion, affects the conformation of the

substrate-binding pocket and thus regulates its kinase activity(26). In this study, we

found that PI-273 can directly bind the palmitoylation insertion and functions as a

PI-competitive reversible inhibitor (Fig. 2D, 2E, 3A, 3B and Table 1). These results also

confirm our previous MD simulation results that the palmitoylation insertion influences

substrate binding. Our work suggests that MD simulation may be helpful for identifying




26

PIK substrate-binding pockets, thus facilitating substrate-competitive inhibitor

screening. However, with respect to the PI-competitive mechanism of PI-273,

determining whether PI-273 can directly bind the substrate pocket or only induce

allosteric regulation of the substrate-binding site is difficult because of the lack of

inhibitor-soaked crystal structure information. Additional efforts to elucidate the binding

mechanism of PI-273 will be made in the future using NMR, crystal structure, hydrogen

deuterium exchange (HDX) and single-molecule fluorescence resonance energy transfer

(FRET) analysis.

In this study, we demonstrated that PI-273 can inhibit breast cancer cell

proliferation (Suppl. Fig. 8A), block the cell cycle (Fig. 5B) and induce cell apoptosis

(Fig. 5C). We then further confirmed its effects on colony formation (Fig. 5D and 5E)

and on an MCF-7 cell-induced xenograft model (Fig. 6A and 6B), all of which indicated

the suppressive effect of PI-273 on breast cancer growth both in vitro and in vivo.

Because previous studies from our group and others have suggested that in addition to

breast cancer, many other cancers, such as malignant melanoma and thyroid carcinoma,

also feature high PI4KIIα expression(4,16), we suspect that PI-273 may have inhibitory

effects on these types of tumors. In our experiment, we also tested the effects of PI-273

on MDA-MB-435 cells, which have been identified as malignant melanoma

cells(50-52). To our surprise, the MDA-MB-435 cells were even more sensitive to

PI-273 in both the plate clone-forming test and the soft agar clone-forming test

compared to the MCF-7 and T-47D cells (data not shown), suggesting that PI-273 may

suppress melanoma growth and merits further investigation.




27

Because precision medicine is considered the optimal means of completely

conquering cancer, clarifying which patient categories are sensitive to particular drugs is

important. In our study, Hs 578T, MDA-MB-231, SUM229PE and SUM159PT cells

were relatively resistant to PI-273 in terms of cell viability, cell cycle and apoptosis (Fig.

5), which is consistent with previous studies concluding that monotherapies targeting

the PIK signaling pathway are largely disappointing for Ras-mutant cancers(53).

PI3K/AKT and Ras/MEK/ERK are two crucial interlinked growth and survival

signaling pathways in tumors[52], and Torbett et al. indicated that Hs 578T cells with

activated forms of the H-Ras oncogene resulting in the activation of parallel pathways

that are able to compensate for the loss of PI3K signals. Different groups have indicated

that PI3K inhibitor treatment alone does not result in the regression of Ras -mutant lung

tumors(54,55), pancreatic cancer(56) or non-small cell lung cancer(57) because of

Ras/MEK/ERK-dependent signaling activation. However, compared to monotherapies,

combining the two treatments (MEK and PI3K inhibitors) resulted in a significant

survival advantage in these cancers. Taken together, these results suggest that the

pharmacodynamic effects of PI-273 may also be related to Ras activity, which is another

interesting topic to be explored in future work.

Interestingly, because PI-273 treatment and PI4KIIα knockout had negligible

effects on the PI(4,5)P2 and PI(3,4,5)P3 levels (Fig. 1D and Fig. 4D), we hypothesize

that the effect of PI4KIIα inhibitors on AKT activity does not result from the direct

regulation of the PI(4,5)P2 and PI(3,4,5)P3 levels. Instead, numerous studies have

indicated that PI4KIIα is important for receptor activity via trafficking




28

regulation(13,17,58,59). Therefore, it is possible that the regulation of PI4KIIα

inhibitors on AKT activity is dependent on its receptor regulation effect, but this

requires further validation.

In summary, we identified PI-273 as a lead substrate-competitive inhibitor of

PI4KIIα based on a unique insertion, the palmitoylation insertion, of PI4KIIα as a

binding site. Our results also demonstrate the significance, safety, and efficacy of

PI4KIIα as a therapeutic target, and PI-273 will be optimized as a therapeutic agent for

cancer in the future.

Note: Three Supplementary data files accompany this manuscript on the Cancer

Research website. Suppl. data 1 including Suppl. method .Suppl. data 2 incuding 4

tables: Suppl. Table 1. Chemical structures and activity of representative PI-93

analogues; Suppl. Table 2. Activity and △Tm of the PI4KIIα mutants;Suppl. Table 3.

Cell lines used in this study; Suppl. Table 4. Reported PIK inhibitors. Suppl. data 3

incuding 10 figures:Suppl. Fig. 1 First-round screening of PI4KIIα inhibitors;Suppl. Fig.

2 Second-round screening of PI4KIIα inhibitors;Suppl. Fig.3. Direct interaction

between PI-273 analogs and PI4KIIα;Suppl. Fig. 4. PI-273 resistance screens using

CRISPR-Cas9;Suppl. Fig.5. Generate and identify PI4KIIα knock out MCF-7

cells;Suppl. Fig. 6. Effects of PI-273 and its analogs on PI4KIIα downstream signaling

pathways;Suppl. Fig. 7. PI4KIIα is required for the regulation of PI-273 on signaling

transduction in breast cancer cells;Suppl. Fig. 8. PI-273 inhibits the proliferation of

breast cancer cells.;Suppl. Fig. 9 Anti-tumor effect of PI-273 in an MCF-7-induced

xenograft model;Suppl. Fig. 10 Inhibitory effect of 100 μM PIK inhibitors on PI4KIIα.




29

Acknowledgments: We thank Pietro De Camilli for providing the PI4KIIα antibody

and Shane Minogue for providing the full-length human PI4KIIα cDNA.

References

1. Di Paolo G, De Camilli P. Phosphoinositides in cell regulation and membrane dynamics. Nature

2006;443:651-7

2. Ree AH, Flatmark K, Saelen MG, Folkvord S, Dueland S, Geisler J, et al. Tumor

phosphatidylinositol 3-kinase signaling in therapy resistance and metastatic dissemination of

rectal cancer: Opportunities for signaling-adapted therapies. Critical reviews in

oncology/hematology 2015

3. Chu KM, Minogue S, Hsuan JJ, Waugh MG. Differential effects of the phosphatidylinositol

4-kinases, PI4KIIalpha and PI4KIIIbeta, on Akt activation and apoptosis. Cell death & disease

2010;1:e106

4. Li J, Lu Y, Zhang J, Kang H, Qin Z, Chen C. PI4KIIalpha is a novel regulator of tumor growth by its

action on angiogenesis and HIF-1alpha regulation. Oncogene 2010;29:2550-9

5. McNamara CW, Lee MC, Lim CS, Lim SH, Roland J, Nagle A, et al. Targeting Plasmodium PI(4)K to

eliminate malaria. Nature 2013;504:248-53

6. Bothmer J, Markerink M, Jolles J. Evidence for a selective decrease in type 1

phosphatidylinositol kinase activity in brains of patients with Alzheimer's disease. Dementia

1994;5:6-11

7. Zubenko GS, Stiffler JS, Hughes HB, Martinez AJ. Reductions in brain phosphatidylinositol kinase

activities in Alzheimer's disease. Biological psychiatry 1999;45:731-6

8. Landman N, Jeong SY, Shin SY, Voronov SV, Serban G, Kang MS, et al. Presenilin mutations linked

to familial Alzheimer's disease cause an imbalance in phosphatidylinositol 4,5-bisphosphate

metabolism. Proceedings of the National Academy of Sciences of the United States of America

2006;103:19524-9

9. Olsen HL, Hoy M, Zhang W, Bertorello AM, Bokvist K, Capito K, et al. Phosphatidylinositol

4-kinase serves as a metabolic sensor and regulates priming of secretory granules in pancreatic

beta cells. Proceedings of the National Academy of Sciences of the United States of America

2003;100:5187-92

10. Minogue S, Waugh MG. The Phosphatidylinositol 4-Kinases: Don't Call it a Comeback.

Sub-cellular biochemistry 2012;58:1-24

11. Wang J, Sun HQ, Macia E, Kirchhausen T, Watson H, Bonifacino JS, et al. PI4P promotes the

recruitment of the GGA adaptor proteins to the trans-Golgi network and regulates their

recognition of the ubiquitin sorting signal. Molecular biology of the cell 2007;18:2646-55

12. Jovic M, Kean MJ, Dubankova A, Boura E, Gingras AC, Brill JA, et al. Endosomal sorting of VAMP3

is regulated by PI4K2A. Journal of cell science 2014;127:3745-56

13. Henmi Y, Morikawa Y, Oe N, Ikeda N, Fujita A, Takei K, et al. PtdIns4KIIalpha generates

endosomal PtdIns(4)P and is required for receptor sorting at early endosomes. Mol Biol Cell

2016;27:990-1001

14. Pizarro-Cerda J, Payrastre B, Wang YJ, Veiga E, Yin HL, Cossart P. Type II phosphatidylinositol




30

4-kinases promote Listeria monocytogenes entry into target cells. Cellular microbiology

2007;9:2381-90

15. Simons JP, Al-Shawi R, Minogue S, Waugh MG, Wiedemann C, Evangelou S, et al. Loss of

phosphatidylinositol 4-kinase 2alpha activity causes late onset degeneration of spinal cord

axons. Proceedings of the National Academy of Sciences of the United States of America

2009;106:11535-9

16. Waugh MG. Phosphatidylinositol 4-kinases, phosphatidylinositol 4-phosphate and cancer.

Cancer letters 2012;325:125-31

17. Li J, Zhang L, Gao Z, Kang H, Rong G, Zhang X, et al. Dual inhibition of EGFR at protein and

activity level via combinatorial blocking of PI4KIIalpha as anti-tumor strategy. Protein & cell

2014;5:457-68

18. Tokuda E, Itoh T, Hasegawa J, Ijuin T, Takeuchi Y, Irino Y, et al. Phosphatidylinositol 4-phosphate

in the Golgi apparatus regulates cell-cell adhesion and invasive cell migration in human breast

cancer. Cancer research 2014;74:3054-66

19. Waugh MG. Amplification of Chromosome 1q Genes Encoding the Phosphoinositide Signalling

Enzymes PI4KB, AKT3, PIP5K1A and PI3KC2B in Breast Cancer. Journal of Cancer 2014;5:790-6

20. Pan W, Choi SC, Wang H, Qin Y, Volpicelli-Daley L, Swan L, et al. Wnt3a-mediated formation of

phosphatidylinositol 4,5-bisphosphate regulates LRP6 phosphorylation. Science

2008;321:1350-3

21. Qin YB, Li L, Pan WJ, Wu DQ. Regulation of Phosphatidylinositol Kinases and Metabolism by

Wnt3a and Dvl. J Biol Chem 2009;284:22544-8

22. Wu P, Hu Y. Small molecules targeting phosphoinositide 3-kinases. MedChemComm

2012;3:1337-55

23. Zheng Z, Amran SI, Thompson PE, Jennings IG. Isoform-selective inhibition of phosphoinositide

3-kinase: identification of a new region of nonconserved amino acids critical for p110alpha

inhibition. Molecular pharmacology 2011;80:657-64

24. Knight ZA, Gonzalez B, Feldman ME, Zunder ER, Goldenberg DD, Williams O, et al. A

Pharmacological Map of the PI3-K Family Defines a Role for p110α in Insulin Signaling. Cell

2006;125:733-47

25. Clarke JH, Giudici ML, Burke JE, Williams RL, Maloney DJ, Marugan J, et al. The function of

phosphatidylinositol 5-phosphate 4-kinase gamma (PI5P4Kgamma) explored using a specific

inhibitor that targets the PI5P-binding site. The Biochemical journal 2015;466:359-67

26. Zhou Q, Li J, Yu H, Zhai Y, Gao Z, Liu Y, et al. Molecular insights into the membrane-associated

phosphatidylinositol 4-kinase IIalpha. Nature communications 2014;5:3552

27. Tai AW, Bojjireddy N, Balla T. A homogeneous and nonisotopic assay for phosphatidylinositol

4-kinases. Anal Biochem 2011;417:97-102

28. Lu Y, Nikolovska-Coleska Z, Fang X, Gao W, Shangary S, Qiu S, et al. Discovery of a nanomolar

inhibitor of the human murine double minute 2 (MDM2)-p53 interaction through an integrated,

virtual database screening strategy. J Med Chem 2006;49:3759-62

29. Lu W, Liu X, Cao X, Xue M, Liu K, Zhao Z, et al. SHAFTS: A Hybrid Approach for 3D Molecular

Similarity Calculation. 2. Prospective Case Study in the Discovery of Diverse p90 Ribosomal S6

Protein Kinase 2 Inhibitors To Suppress Cell Migration. Journal of Medicinal Chemistry

2011;54:3564-74

30. Jones G, Willett P, Glen RC. Molecular recognition of receptor sites using a genetic algorithm




31

with a description of desolvation. Journal of Molecular Biology 1995;245:43-53

31. Jones G, Willett P, Glen RC, Leach AR, Taylor R. Development and validation of a genetic

algorithm for flexible docking. Journal of Molecular Biology 1997;267:727-48

32. Verdonk ML, Cole JC, Hartshorn MJ, Murray CW, Taylor RD. Improved protein–ligand docking

using GOLD. Proteins: Structure, Function, and Bioinformatics 2003;52:609-23

33. Morris GM, Goodsell DS, Halliday RS, Huey R, Hart WE, Belew RK, et al. Automated docking

using a Lamarckian genetic algorithm and an empirical binding free energy function. Journal of

Computational Chemistry 1998;19:1639-62

34. Huey R, Morris GM, Olson AJ, Goodsell DS. A semiempirical free energy force field with

charge-based desolvation. Journal of Computational Chemistry 2007;28:1145-52

35. Martinez Molina D, Jafari R, Ignatushchenko M, Seki T, Larsson EA, Dan C, et al. Monitoring drug

target engagement in cells and tissues using the cellular thermal shift assay. Science

2013;341:84-7

36. Jafari R, Almqvist H, Axelsson H, Ignatushchenko M, Lundback T, Nordlund P, et al. The cellular

thermal shift assay for evaluating drug target interactions in cells. Nature protocols

2014;9:2100-22

37. Choi S, Hedman AC, Sayedyahossein S, Thapa N, Sacks DB, Anderson RA. Agonist-stimulated

phosphatidylinositol-3,4,5-trisphosphate generation by scaffolded phosphoinositide kinases.

Nature cell biology 2016;18:1324-35

38. Sun F, Li J, Chen C. Puzzle out the regulation mechanism of PI4KIIalpha activity. Science China

Life sciences 2014;57:636-8

39. Miduturu CV, Deng X, Kwiatkowski N, Yang W, Brault L, Filippakopoulos P, et al. High-throughput

kinase profiling: a more efficient approach toward the discovery of new kinase inhibitors.

Chemistry & biology 2011;18:868-79

40. Hollestelle A, Elstrodt F, Nagel JH, Kallemeijn WW, Schutte M. Phosphatidylinositol-3-OH kinase

or RAS pathway mutations in human breast cancer cell lines. Molecular cancer research : MCR

2007;5:195-201

41. Torbett NE, Luna-Moran A, Knight ZA, Houk A, Moasser M, Weiss W, et al. A chemical screen in

diverse breast cancer cell lines reveals genetic enhancers and suppressors of sensitivity to PI3K

isoform-selective inhibition. The Biochemical journal 2008;415:97-110

42. Graham TR, Burd CG. Coordination of Golgi functions by phosphatidylinositol 4-kinases. Trends

in cell biology 2011;21:113-21

43. Wang YJ, Wang J, Sun HQ, Martinez M, Sun YX, Macia E, et al. Phosphatidylinositol 4 phosphate

regulates targeting of clathrin adaptor AP-1 complexes to the Golgi. Cell 2003;114:299-310

44. Pitts TM, Newton TP, Bradshaw-Pierce EL, Addison R, Arcaroli JJ, Klauck PJ, et al. Dual

pharmacological targeting of the MAP kinase and PI3K/mTOR pathway in preclinical models of

colorectal cancer. PloS one 2014;9:e113037

45. Sinha RK, Patel RY, Bojjireddy N, Datta A, Subrahmanyam G. Epigallocatechin gallate (EGCG)

inhibits type II phosphatidylinositol 4-kinases: a key component in pathways of

phosphoinositide turnover. Archives of biochemistry and biophysics 2011;516:45-51

46. Srivastava R, Ratheesh A, Gude RK, Rao KV, Panda D, Subrahmanyam G. Resveratrol inhibits type

II phosphatidylinositol 4-kinase: a key component in pathways of phosphoinositide turn over.

Biochemical pharmacology 2005;70:1048-55

47. Balla A, Balla T. Phosphatidylinositol 4-kinases: old enzymes with emerging functions. Trends in




32

cell biology 2006;16:351-61

48. Minogue S, Anderson JS, Waugh MG, dos Santos M, Corless S, Cramer R, et al. Cloning of a

human type II phosphatidylinositol 4-kinase reveals a novel lipid kinase family. The Journal of

biological chemistry 2001;276:16635-40

49. Waugh MG, Minogue S, Blumenkrantz D, Anderson JS, Hsuan JJ. Identification and

characterization of differentially active pools of type IIalpha phosphatidylinositol 4-kinase

activity in unstimulated A431 cells. The Biochemical journal 2003;376:497-503

50. Sellappan S, Grijalva R, Zhou X, Yang W, Eli MB, Mills GB, et al. Lineage infidelity of MDA-MB-435

cells: expression of melanocyte proteins in a breast cancer cell line. Cancer research

2004;64:3479-85

51. Ellison G, Klinowska T, Westwood RF, Docter E, French T, Fox JC. Further evidence to support the

melanocytic origin of MDA-MB-435. Molecular pathology : MP 2002;55:294-9

52. Ross DT, Scherf U, Eisen MB, Perou CM, Rees C, Spellman P, et al. Systematic variation in gene

expression patterns in human cancer cell lines. Nature genetics 2000;24:227-35

53. Zeitouni D, Pylayeva-Gupta Y, Der CJ, Bryant KL. KRAS Mutant Pancreatic Cancer: No Lone Path

to an Effective Treatment. Cancers 2016;8

54. Downward J. Targeting RAS and PI3K in lung cancer. Nature medicine 2008;14:1315-6

55. Engelman JA, Chen L, Tan X, Crosby K, Guimaraes AR, Upadhyay R, et al. Effective use of PI3K

and MEK inhibitors to treat mutant Kras G12D and PIK3CA H1047R murine lung cancers. Nature

medicine 2008;14:1351-6

56. Junttila MR, Devasthali V, Cheng JH, Castillo J, Metcalfe C, Clermont AC, et al. Modeling targeted

inhibition of MEK and PI3 kinase in human pancreatic cancer. Molecular cancer therapeutics

2015;14:40-7

57. Toulany M, Iida M, Keinath S, Iyi FF, Mueck K, Fehrenbacher B, et al. Dual targeting of PI3K and

MEK enhances the radiation response of K-RAS mutated non-small cell lung cancer. Oncotarget

2016

58. Minogue S, Waugh MG, De Matteis MA, Stephens DJ, Berditchevski F, Hsuan JJ.

Phosphatidylinositol 4-kinase is required for endosomal trafficking and degradation of the EGF

receptor. Journal of cell science 2006;119:571-81

59. Ketel K, Krauss M, Nicot AS, Puchkov D, Wieffer M, Muller R, et al. A phosphoinositide

conversion mechanism for exit from endosomes. Nature 2016;529:408-12




33

Table 1. Pharmacokinetical results of the SD rats after intravenous administration

of 0.5 mg/kg PI-273 and intragastrical administration of 1.5 mg/kg PI-273

respectively.

Pharmacokinetical parameters

PI-273

Intravenously(0.5 mg/kg) Intragastrically(1.5 mg/kg)

Mean ± SD Mean ± SD

Cmax (ng∙ml-1

) 2392.7 ± 46 116 ± 23

Tmax (h) 0.083 ± 0.00 0.278 ± 0.096

t1/2 (h) 0.411 ± 0.087 1.321 ± 0.657

MRT0-12 (h) 0.23 ± 0.072 1.25 ± 0.061

MRT0-∞ (h) 0.23 ± 0.072 1.571 ± 0.378

CL/F (L∙h-1

∙kg-1

) 0.649 ± 0.204 11.4 ± 1.32

Vd/F (L∙kg-1

) 0.38 ± 0.118 21.0 ± 7.86

AUC0-12 (ng∙h∙ml-1

) 816 ± 218 125 ± 8.28

AUC0-∞ (ng∙h∙ml-1

) 816 ± 218 132 ± 16.0




34

Figure legends

Figure 1 Identification of PI4KIIα small-molecule inhibitors. (A) Schematic

representation of the virtual screening strategy adopted in this study. (B) Chemical

structures of PI-273 and PI-69. (C) The effect of PI-273 on cellular PI4P content.

MCF-7 cells were treated with DMSO or the indicated concentrations of PI-273 for the

indicated times. The cells were then harvested for PI4P measurement. (D) MCF-7 cells

were treated with DMSO or 1 μM PI-273 for 24 hours before being collected for the

indicated lipids analyses. The values are presented as the means ± S.D. from three

independent experiments, and all the above experiments were performed three times in

triplicate. * indicates p<0.01.

Figure 2 PI-273 directly binds and reversibly inhibits PI4KIIα. (A) Representative

SPR binding curves for PI4KIIα and 2-fold serial dilutions of PI-273 from 1.25 μM to

20 μM. (B) Thermal shift assay indicating the stabilization of PI4KIIα by PI-273.

Melting curves for 5 μM PI4KIIα protein in the presence of varying concentrations of

PI-273 or PI-69. (C) A cellular thermal shift assay was performed to evaluate the

interaction between PI-273 and PI4KIIα in intact cells (in vivo) and cell lysates (in

vitro). (D) PI4KIIα was preincubated with the indicated concentrations of PI-273 (0-1

μM) for different durations (0–60 minutes) before ATP was added. The graphs represent

normalized % inhibition compared to preincubation time. (E) The initial velocity

kinetics compared to varying PI4KIIα enzyme concentrations (2.5-10 ng/μl) at the

indicated concentrations of PI-273 (0-2 μM). All the above experiments were performed

three times with comparable results.




35

Figure 3 PI-273 is a substrate-competitive inhibitor of PI4KIIα. Enzyme kinetics

analysis of PI4KIIα inhibition by PI-273. Activity assays were performed with varying

concentrations of (A) ATP and (B) PI. Double-reciprocal plots of initial velocities

(Lineweaver-Burk plots) showing uncompetitive inhibition by PI-273 toward ATP and

purely competitive inhibition toward PI as a substrate. The experiment was performed

three times with comparable results. (C) Vmax and (D) KM values for ATP and PI under

varying concentrations of PI-273. (E) Thermal shift assay demonstrating the

stabilization of PI4KIIα variants by PI-273. PI4KIIα (5 μM) variants were treated with

PI-273 (50 μM) and then subjected to the thermal shift analysis. The values are

presented as the means ± S.D. from three independent experiments (Student's t-test),

and this experiment was performed three times in triplicate. * indicates p<0.01 and #ns

indicates no difference.

Figure 4 PI-273 inhibits PI4KIIα downstream signaling transduction in breast

cancer cells. (A) PI-273 IC50 values for PI4KIIα, PI4KIIβ, PI4KIIIα, PI4KIIIβ, PI3Kα,

PI3Kδ, PI3Kβ, PI3Kγ, AKT1, AKT2 and AKT3. (B) The anti-proliferation activity of

PI-273 in wild-type MCF-7 cells and PI4KIIα knockout MCF-7 cells. (C) Specific

regulation of AKT signaling by PI-273 in MCF-7 cells. MCF-7 cells were treated with

the indicated doses of PI-273 for 3 days, followed by treatment with 100 ng/ml EGF for

10 minutes. AKT phosphorylation and protein expression levels were measured by

western blot. The effects of PI-273 on (D) LPA, PI4P, PI3P, PI(4,5)P2, PI(3,4,5)P3

content and (E) AKT signaling in wild-type MCF-7 cells (CTR cells) and PI4KIIα

knockout MCF-7 cells (KO cells). Cells were treated with 1 μM PI-273 or DMSO for




36

24 hours (for lipid ratio detection) or 3 days (for AKT signaling detection), respectively.

Lipid contents were detected by the appropriate Mass ELISA Kit. AKT phosphorylation

and protein expression levels were measured by western blot. All the values are

presented as the means ± S.D. from three independent experiments, and all the above

experiments were performed three times in triplicate. * indicates p<0.01 and #ns

indicates no difference.

Figure 5 PI-273 inhibits the proliferation of breast cancer cells. (A) Cell viability

assays (WST-8 assay) were performed in MCF-7, T-47D, SK-BR-3, BT-474,

MDA-MB-468, MDA-MB-231, SUM229PE, SUM159PT and Hs 578T cells after

treatment with the indicated concentrations of PI-273 for 72 hours. (B) Six breast cancer

cells were treated for 48 hours with DMSO or 2 μM PI-273 and then labeled with PI.

Cells were analyzed by flow cytometry. The graph shows the % of cells in each cell

cycle phase. (C) Apoptosis ratio in cells treated with PI-273. Cells were counted using a

TUNEL kit after being treated with DMSO or 2 μM PI-273 for 48 hours. The % of

apoptotic cells are presented on the y-axis. (D) and (E) Effects of PI-273 on colony

formation in MCF-7, T-47D, Hs 578T and MDA-MB-231 cells. Anchorage-dependent

cell growth and anchorage-independent cell growth were measured using the plate

clone-forming assay and the soft agar clone-forming assay, respectively. Colony

numbers were determined after 7 days of incubation for the plate clone-forming assay

and 14 days for the soft agar clone-forming assay. Scale bar: 200 μm. All the values are

presented as the means ± S.D. from three independent experiments (Student's t-test),

and all the above experiments were performed three times in triplicate * indicates




37

p<0.01 and #ns indicates no difference.

Figure 6 PI-273 suppresses breast cancer in vivo. (A) Tumor growth curve, (B) tumor

weight and (C) mouse weight for MCF-7 xenografts in BALB/c nude mice treated with

25 mg/kg/day PI-273 or an equal volume of vehicle for 16 consecutive days. Statistical

analysis was performed using the paired t-test. (D) TUNEL+ staining of the histological

sections of MCF-7 xenografts. The y-axis represents the ratio of TUNEL+

cells per ten

fields for one tumor sample (n = 5 tumors/treatment). Scale bar: 20 μm. (E) Ki-67

immunohistochemistry staining of xenograft MCF-7 tumor sections. The y-axis

represents the ratio of Ki-67+ cells per ten fields for one tumor sample (n = 5

tumors/treatment). Statistical analysis was performed using Student’s t-test. Scale bar:

100 μm. (F) Effect of PI-273 on the phosphorylation levels of AKT and the PI4P content

in MCF-7 xenograft tumors. All the values are presented as the means ±S.D. from the

indicated samples. * indicates p<0.01. (G) Model of the antitumor effect of PI-273.

PI-273 can specifically bind PI4KIIα, inhibiting its activity and, notably, reducing the

cellular PI4P content, thus resulting in suppressed PI3K/AKT signaling transduction.

This activity blocks cells at the G2/M phase, resulting in proliferation arrest and

apoptosis. Together, these effects can retard the growth of breast tumors.




Published OnlineFirst August 21, 2017.Cancer Res Jiangmei Li, Zhen Gao, Dan Zhao, et al.

, inhibits the growth of breast cancer cellsαinhibitor of PI4KIIPI-273, a substrate-competitive, specific small molecule

Updated version

10.1158/0008-5472.CAN-17-0484doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2017/08/18/0008-5472.CAN-17-0484.DC1

Access the most recent supplemental material at:

Manuscript

Authorbeen edited. Author manuscripts have been peer reviewed and accepted for publication but have not yet

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/early/2017/08/18/0008-5472.CAN-17-0484To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-17-0484

http://cancerres.aacrjournals.org/content/suppl/2017/08/18/0008-5472.CAN-17-0484.DC1

http://cancerres.aacrjournals.org/cgi/alerts


http://cancerres.aacrjournals.org/content/early/2017/08/18/0008-5472.CAN-17-0484


inhibits the growth of breast cancer cells Running title ......Aug 18, 2017 · kinase PI4KIIα has...

Documents

Transcript of inhibits the growth of breast cancer cells Running title ......Aug 18, 2017 · kinase PI4KIIα has...