Effects of MELK depletion by shRNA or inhibition by OTSSP167 in cell lines were assessed by colony for-mation and MTT proliferation assays, Western blotting apoptosis, andflow cytometry c
Trang 1MELK expression in ovarian cancer correlates with poor outcome and its
inhibition by OTSSP167 abrogates proliferation and viability of ovarian
cancer cells
Reto S Kohlera,1, Henriette Kettelhacka,1, Alexandra M Knipprath-Mészarosb, André Fediera,
Andreas Schoetzaua, Francis Jacoba,c,⁎ , Viola Heinzelmann-Schwarza,b,⁎⁎
a
Ovarian Cancer Research, Department of Biomedicine, University Hospital Basel, University of Basel, Switzerland
b
Hospital for Women, Department of Gynecology and Gynecological Oncology, University Hospital Basel, University of Basel, Switzerland
c
Glyco-Oncology, Ovarian Cancer Research, Department of Biomedicine, University Hospital Basel, University of Basel, Switzerland
H I G H L I G H T S
• Elevated MELK in serous ovarian cancer associates with poor progression-free survival
• Silencing and inhibition of MELK results in apoptosis in serous ovarian cancer cell lines
• Drug-resistant cells retain sensitivity to MELK-inhibitor OTSSP167
a b s t r a c t
a r t i c l e i n f o
Article history:
Received 12 November 2016
Received in revised form 7 February 2017
Accepted 7 February 2017
Available online xxxx
Objective Maternal embryonic leucine-zipper kinase (MELK) shows oncogenic properties in basal-like breast cancer, a cancer subtype sharing common molecular features with high-grade serous ovarian cancer We exam-ined the potential of MELK as a molecular and pharmacological target for treatment of epithelial ovarian cancer (EOC)
Methods/materials Bioinformatic analysis was performed on nine OC transcriptomic data sets totaling 1241 patients Effects of MELK depletion by shRNA or inhibition by OTSSP167 in cell lines were assessed by colony for-mation and MTT (proliferation) assays, Western blotting (apoptosis), andflow cytometry (cell cycle analysis) Results Elevated MELK expression was correlated with histological grading (n = 6 data sets, pb 0.05) and progression-free survival (HR 5.73, pb 0.01) in OC patients and elevated MELK expression in other cancers with disease-free survival (n = 3495, HR 1.071, pb 0.001) Inhibition or depletion of MELK reduced cell prolifer-ation and anchorage-dependent and -independent growth in various OC cell lines through a G2/M cell cycle ar-rest, eventually resulting in apoptosis OTSSP167 retained its cytotoxicity in Cisplatin- and Paclitaxel-resistant IGROV1 and TYK-nu OC cells and sensitized OVCAR8 cells to Carboplatin but not Paclitaxel
Conclusion MELK inhibition by OTSSP167 may thus present a strategy to treat patients with aggressive, pro-gressive, and recurrent ovarian cancer
© 2017 The Authors Published by Elsevier Inc This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/)
Keywords:
TCGA
Apoptosis
Disease outcome
Oncogenic growth
Drug resistance
1 Introduction
Epithelial ovarian cancer (EOC) is the leading cause of death from a
gynecologic cancer and patients with high-grade serous ovarian cancer
(HGSOC), the most common and aggressive type, have a 5-year survival
ofb20% Effective screening or early detection is limited and hence pa-tients commonly present at an advanced stage Overall survival has, de-spite improved surgical techniques and drug regimens, not changed significantly for several decades, in part owing the considerable risk for disease recurrence even after full remission following cytoreductive surgery and platinum- and taxane-based chemotherapy[1–3] Hence novel drugs or additional therapeutic strategies are required to improve patient survival
MELK (maternal embryonic leucine zipper kinase) is a highly con-served serine/threonine kinase initially found in a wide range of early
Gynecologic Oncology xxx (2017) xxx–xxx
⁎ Corresponding author at: Department of Biomedicine, University Hospital Basel,
Hebelstrasse 20, 4031 Basel, Switzerland.
⁎⁎ Correspondence to: V Heinzelmann-Schwarz, University Hospital Basel, Spitalstrasse
21, 4031 Basel, Switzerland.
E-mail addresses: francis.jacob@unibas.ch (F Jacob), viola.heinzelmann@usb.ch
(V Heinzelmann-Schwarz).
1
Contributed equally.
http://dx.doi.org/10.1016/j.ygyno.2017.02.016
0090-8258/© 2017 The Authors Published by Elsevier Inc This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).
Contents lists available atScienceDirect
Gynecologic Oncology
j o u r n a l h o m e p a g e :w w w e l s e v i e r c o m / l o c a t e / y g y n o
Trang 2embryonic cellular stages It is proposed to play prominent roles in cell
cycle control, cell proliferation, apoptosis, drug resistance, cell renewal,
embryogenesis, and oncogenesis[4] MELK is therefore believed to have
an oncogenic driver function This may be attributed to its capacity to
disable critical cell-cycle checkpoints, reduce DNA replication stress,
and stimulating proliferation by its capacity to increase the threshold
for DNA-damage tolerance[5]
Recent studies indicate that MELK is highly expressed in several
human cancers including breast, gastric, and lung cancers[6] However,
there seems to be some subtype specificity within cancer types as MELK
is overexpressed in basal-like (BBC) but not in luminal breast cancer and
normal breast tissue, correlating with poor prognosis and tumor
aggres-siveness[7] MELK silencing by shRNA reduced oncogenic colony
forma-tion and induced apoptosis in BBC cells but not in other types of breast
cancer cells and pharmacological MELK inhibition by OTSSP167 reduced
BBC tumor but not luminal cancer growth in mice[7] OTSSP167 is a
se-lective small molecule inhibitor of MELK developed in 2012 and was
shown to inhibit proliferation in several cancer cells of different origin
[8,9] OTSSP167 has been examined in an Australian clinical phase I
study for its safety in healthy volunteers and is at present being
evaluat-ed in a clinical phase I study on various solid tumors but not including
ovarian tumors[4]
Owing the lack of respective data in ovarian cancer and owing the
fact that BBC and HGSOC share striking similarities regarding molecular
and genetic profiles[10,11], i.e thatN90% of BBC carry TP53 mutations,
which is also a hallmark in HGSOC[1], that frequencies of RB1 and CMYC
pathway mutations are also significantly similar in BBC and HGSOC[11,
12], that transcriptomic profiles are shared by both cancer types[13,14]
and that BRCA1/2 mutations are frequently found in BBC[12] and
HGSOC[15–17], we investigated the putative role of MELK in EOC
Specifically, we evaluated MELK gene and protein expression in
transcriptomic data sets as well as in vitro in ovarian cancer, normal
ovarian surface epithelial, and fallopian tube cell lines, respectively;
compared the transcriptomics data to patient outcome parameters;
and evaluated the effect of shRNA and OTSSP167 on proliferation,
onco-genic growth, cell cycle progression, and apoptosis in various (including
drug-resistant) ovarian cancer cell lines
2 Materials and methods
2.1 Data acquisition and bioinformatical analysis
MELK expression was initially evaluated in our oligonucleotide array
data published in Heinzelmann et al 2004[10] Publicly available
transcriptomic data sets were downloaded from Gene Expression
Om-nibus (http://www.ncbi.nlm.nih.gov/geo/) Information including gene
expression accession numbers of the data sets are listed in
Supplemen-tary Table S1 All publicly available data sets were searched for the
pres-ence of the following criteria: 1) serous ovarian cancer, 2)
histopathological grade, 3) and normal versus ovarian cancer Statistical
analysis andfigures were obtained through the use of the software R
version 3.1.3 (www.R-project.org) Gene expression data for the
“Can-cer Cell Line Encyclopedia were accessed through the cbioportal using
R (www.cbioportal.org) Following R packages were therefore applied:
“cgdsr”, “catspec”, “compareGroups”, “party”, “ggplot2”, and “Rcpp”
2.2 Cell culture and MELK-inhibitor
HOSE6-3, OVCAR3, OVCAR4, OVCAR5, OVCAR8, SKOV3, A2780, BG1,
IGROV1, OAW42, T47D, MCF7, and MDA-MB-468 cells were cultured in
RPMI supplemented with 10% fetal bovine serum (FBS) and penicillin/
streptomycin TOV112D was cultured in DMEM supplemented with
10% FBS and penicillin/streptomycin EFO27 were cultured in RPMI
con-taining 20% FBS including 1 mM sodium pyruvate and
penicillin/strep-tomycin FT33-Tag (hTERT + SV40 large T), FT190 (hTERT + SV40
TAg), FT194 (hTERT + SV40 Tag), FT237 (hTERT + p53 shRNA +
CDK4-R24C), and FT246 (hTERT + p53 shRNA + CDK4-R24C) (kind gifts by Dr Drapkin) were cultured in DMEM F12/50 without HEPES supplemented with 2% Ultroser (USG, Pall Corporation, USA) and peni-cillin/streptomycin[18] TYK-nu and TYK-nu(R) cells were obtained from JCRB cell bank, Japan and were cultured in MEM Eagle supple-mented with 10% FBS and penicillin/streptomycin[19] Cell lines were characterized by STR profiles and routinely tested for mycoplasma in-fection MELK-inhibitor OTSSP167 (hydrochloride) was purchased from MedchemExpress (Monmouth Junction, NJ, USA)
2.3 Western blot analysis Western blot analysis was used to detect expression of MELK and MDR1 in cell lines and assess apoptosis in OTSSP167-treated or shMELK-treated cells Cells were harvested and lysed in cell lysis buffer (9803, Cell Signaling Technology, Bioconcept, Allschwil, Switzerland) supplemented with 0.1% SDS, 0.5% sodium deoxycholate and 1 mM PMSF Protein concentrations were determined using Pierce™ BCA assay kit (Pierce, Life Technologies Europe BV, Zug, Switzerland) A total of 30μg of protein was separated by sodium dodecyl sulfate poly-acrylamide gel electrophoresis (SDS-PAGE), followed by blotting onto PVDF membrane (Bio-Rad, Cressier, Switzerland) according to standard protocols Proteins were detected by specific primary antibodies and the respective secondary horseradish peroxidase-conjugated antibodies The primary antibodies were: anti-MELK (EPR 3981, Source Bioscience, Berlin, Germany), anti-MDR1 (sc-13,131, Santa Cruz Biotechnology, Labforce AG, Muttenz, Switzerland), anti-cleaved poly(ADP-ribose) po-lymerase (PARP-1) (9541; Cell Signaling Technology), anti-Tubulin (2148, Cell Signaling Technology), and anti-actin (A5441, Sigma, Buchs, Switzerland) The secondary antibodies were: anti-rabbit IgG-HRP (7074, Cell Signaling Technology) and anti-mouse IgG-IgG-HRP (7076, Cell Signaling Technology) Complexes were visualized by en-hanced chemiluminescence (SuperSignal West Dura, Pierce) according
to the manufacturer's instructions using a chemiluminescent imaging system (ChemiDoc XR, Bio-Rad)
2.4 shRNA-mediated knockdown of MELK expression and effects on apo-ptosis and colony formation
This included the construction of the plasmids, the production of the lentivirus, and the generation of stable cell lines To obtain pLKO-Tet-shMELK vector that expresses doxycycline-inducible short-hairpin RNA against human MELK, the following oligonucleotide pairs[7]
were inserted into Tet-pLKO-puro (21915, Addgene, Cambridge, MA, USA) using AgeI and EcoRI restriction sites: 5′-ccg ggc ctg aaa gaa act cca att act cga gta att gga gtt tct ttc agg ctt ttt g−3′ and 5′-aat tca aaa agc ctg aaa gaa act cca att act cga gta att gga gtt tct ttc agg c−3′ Targeting sequences are underlined Constructs were verified by Sanger DNA sequencing (Source Biosciences, Berlin, Germany)
HEK293T cells were transfected with pLKO-Tet-shMELK together with packaging vectors pMD2.G (12,259, Addgene) and pCMV8.74 (22,036, Addgene) using JetPEI transfection reagent (Chemie Brunschwig, Basel, Switzerland) Medium was replaced 24 h after trans-fection and after another 48 h supernatant was harvested,filtered through a 0.45μm filter, aliquoted and stored at −80 °C To generate stable inducible cell lines, HOSE6-3 and OVCAR8 cells were infected with viral supernatant supplemented with 8μg/ml polybrene After
24 h medium was changed and 1μg/ml puromycin was added after an-other 24 h Stable cell lines were maintained in RPMI containing 0.5μg/
ml puromycin shMELK expression was induced by doxycycline addi-tion (0.5μg/ml) for 72 h
Effects of MELK-knockdown on apoptosis and colony formation in soft-agar were determined For apoptosis analysis, cells (300′000) were seeded, fresh medium with or without doxycycline (0.5μg/ml) was added for cells expressing pLKO-Tet-shMELK on the next day, lysed after 48 h, and stored for Western blot analysis For colony
Trang 3formation, 5′000 cells were suspended in 0.4% low melt agarose and
plated onto a layer of 0.6% low melt agarose in 12-well plates On the
next day, fresh medium with or without doxycycline (0.5μg/ml) was
added to the wells for cells expressing pLKO-Tet-shMELK After
14 days colonies werefixed (4% formaldehyde) and stained (0.05%
crys-tal violet)
2.5 OTSSP167-mediated colony formation and anchorage-independent
growth inhibition and apoptosis induction
For the colony formation assay, 200 cells were seeded into 12-well
plates and treated with 100 nM OTSSP167 for 14 days Colonies were
fixed (4% formaldehyde) and stained (0.05% crystal violet) for 1 h For
the anchorage-independent growth assay, 5′000 cells suspended in
0.4% low melt agarose were plated onto a layer of 0.6% low melt agarose
in 12-well plates On the next day, fresh medium with or without
100 nM OTSSP167 was added Colonies werefixed and stained as
de-scribed For apoptosis analysis, 300′000 cells seeded into 12-well plates
were treated with 100 nM OTSSP167 for 48 h and harvested for Western
blot analysis (PARP-1 cleavage) as described above
2.6 Cell cycle analysis
Cell cycle analysis was performed byflow cytometry Cells (300′000)
were seeded on 6-well plates and treated with various concentrations of
OTSSP167 for 24 h On the next day, adherent andfloating cells were
harvested and resuspended in PBS,fixed by adding 500 μl −20 °C 70%
ethanol and stored on ice for at least 30 min Cells were then rehydrated
with cold PBS and stained overnight at 4 °C with Hoechst 33367 (BD
Bioscience, Allschwil, Switzerland) The DNA content was analyzed
using a FACS Fortessaflow cytometer (BD Biosciences)
2.7 Confocalfluorescence microscopy
Cells were grown on polylysine glass slides attached to a 8-well
chamber,fixed with 4% para-formaldehyde, permeabilized with 0.3%
Triton X-100 and blocked with 5% (w/v) BSA fraction V (Sigma)
dis-solved in PBS Cells were stained with anti-tubulin (2148, Cell Signaling
Technology), Alexa Fluor 647 Phalloidin (8940, Cell Signaling
Technolo-gy) and counterstained with ProLong Gold Antifade Reagent with DAPI
(8961, Cell Signaling Technology) according to manufacturers' protocol
Fluorescence images were taken with a LSM 780 confocal microscope
(Zeiss, Feldbach, Switzerland)
2.8 MTT cell viability assay
This assay was performed to determine the effect of OTSSP167 alone
or a combination of OTSSP167 with either Carboplatin or Paclitaxel on
proliferation and cell viability Cells were seeded into 96-well plates
and treated on the next day with various concentrations of OTSSP167
or a combination with Carboplatin or Paclitaxel for 72 h Cells were
then incubated with 500μg/ml (final concentration) MTT dye (in PBS)
for 4 h, followed by removal of the medium, dissolution of the violet
crystals with 200μl DMSO, and optical density measurement (OD,
ab-sorbance at 540 nm) (Synergy H1 Hybrid Reader, BioTek, Luzern,
Swit-zerland) Data are presented as relative cell viability (OD as percentage
of untreated control) as a function of drug concentration IC50values
were calculated by linear extrapolation to compare drug sensitivity
Each experiment was performed independently at least three times in
quadruplicates
2.9 Generation of drug-resistant cells
In addition to the Cisplatin-resistant TYK-nu(R) cells and the
respec-tive sensirespec-tive parental TYK-nu cells generated previously[19], IGROV1
and OVCAR8 cells were step-wise exposed to increasing concentrations
of Paclitaxel or Carboplatin (IGROV1) or OTSSP167 (OVCAR8) in order
to generate drug-resistant sublines by a protocol described before
[20] The principle of selection was the clonal growth in the presence
of increasing drug concentrations, assuming that cells acquire new fea-tures in an irreversible fashion by chronic drug exposure Briefly, 100′
000 cells are seeded into 12-well plates and exposed to IC50 concentra-tion of Paclitaxel, Carboplatin or OTSSP167 for 48 h Remaining viable cells were allowed to recover in drug-free culture medium until reaching confluency This cycle was repeated with incrementing drug concentrations until a stable resistance was reached or no viable cells were left Stable resistance was verified by MTT assay performed over several passages of this resistant cell population and arbitrarily defined
by a resistance factor (IC50 value ratio of resistant to parental cells) of 2.0 or above
2.10 Statistical analysis Comparisons between cancer and control or the various cancer sub-types were examined using one-way ANOVA Survival was investigated using Cox proportional hazard regression model, conditional inference trees, and Kaplan-Meier analysis Comparisons of IC50-values in drug-resistant and -sensitive cells were done by the two-tailed Student t-test A p-valueb 0.05 was considered statistically significant All statisti-cal evaluations were done using the statististatisti-cal software R version 3.1.3 (www.R-project.org) including following packages:‘ctree’ in R package party[21]and survival[22]
3 Results 3.1 Increased MELK expression in EOC correlates with histological grading and shorter disease-free survival
Two recent studies reported elevated MELK expression in several human cancers including epithelial ovarian, breast, gastric, and lung cancer[6]and a correlation between elevated MELK expression and poor prognosis and tumor aggressiveness in basal-like breast cancer (BBC)[7] In order to expand on these studies, we reviewed the TCGA data base (PANCAN12) and its bioinformatical analysis revealed that MELK expression was also a predictor of poorer disease-free survival
in cancers in general (n = 3495, HR 1.071, CI [1.031 to 1.112], p-value based on Cox regression 0.000364), i.e regardless of the cancer subtype (Fig 1A) and that elevated MELK expression was present also in cancers (n = 3599) of the head and neck regions, bladder, colon, rectum, and ovary (Fig 1B)
MELK expression in ovarian cancer, our cancer of interest, ranked be-tween cancers with the origin rectum and endometrium Worthwhile mentioning that our previously performed gene expression profiling
on various ovarian tumors and healthy ovarian surface epithelium al-ready in 2004 demonstrated an interesting profile for KIAA0175, later
to be named MELK[10,11]: KIAA0175 was generally low expressed in normal ovarian surface epithelium, slightly elevated in mucinous and endometrioid borderline (low-malignant) tumors, and at higher but varying levels primary and metastatic ovarian cancer samples of differ-ent histotypes (serous, endometrioid, and mucinous) (Fig 2A) In the present study we determined and validated MELK expression in ovarian cancer searching the publicly accessible GEO Gene Expression Omnibus database for the respective transcriptomic data sets and compared MELK expression with clinico-pathological characteristics and outcome data provided along with the transcriptomic data Nine transcriptomic data sets were identified comprising 1241 patients subdivided into se-rous ovarian carcinomas (SOC, n = 1144), sese-rous benign tumors (SBenign, n = 20), serous borderline tumors (SBL, n = 43), and tissue from normal ovarian surface epithelium (HOSE, n = 34)
(Supplementa-ry Table S1) The comparison of HOSE versus SOC revealed a statistically significant (p b 0.01; five data sets) increase in MELK expression in SOC (Supplementary Fig S1A) One data set showed that MELK expression
Trang 4progressively increased from benign to borderline to SOC
(Supplemen-tary Fig S1B) MELK expression positively correlated with the
histolog-ical grade, i.e lesser histopathologhistolog-ical differentiation (grade) the higher
MELK expression (Supplementary Fig S1C) TCGA data set analysis
sep-arated high-grade SOC (HGSOC) into“low” and “high” MELK and
identi-fied high MELK expression as a predictor for poor progression-free
survival (HR 5.73, p = 0.0036) (Fig 2B) These results indicate that
MELK is elevated in SOC compared to normal ovarian tissue and that
MELK expression increases with the histological grading, i.e with
ag-gressiveness of SOC, and correlates with shorter progression-free
survival
3.2 Elevated MELK in HGSOC cell lines
To study MELK expression in ovarian cancer cell lines we accessed
the Cell Line Encyclopedia (CCLE) for MELK expression in a large set of
ovarian cancer cell lines Twenty-eight cell lines with an MELK
expres-sion level between MDA-MB-468 (BBC and positive for
MELK-associat-ed oncogenesis[7]) and T47D (luminal breast cancer and negative for
MELK-associated oncogenesis[7]) were identified Those with high
(e.g SKOV3), intermediate (e.g OVCAR8, IGROV1), and moderate (e.g
OVCAR4) MELK gene expression (Fig 3A) were chosen for Western
blot analysis (Fig 3B) Cell lines representing“normal” ovary surface
ep-ithelium (HOSE6-3) and fallopian tube epep-ithelium (FT33-tag, FT190,
FT194, FT237, FT246) were also included MELK expression was highest
in EOC cells, low or poorly detectable in almost all fallopian tube cells
and HOSE6-3 cells (Fig 3B) Both normal controls have been recently suggested to be the origin of EOC and their subtypes[23,24]
3.3 Depletion of MELK by shRNA induces apoptosis and inhibits colony formation
We determined whether depletion of MELK protein by shRNA-me-diated knockdown induces apoptosis and abrogates oncogenic growth
in EOC cells The results show that doxycycline-induced MELK silencing induced PARP-1 cleavage (measure for ongoing apoptosis) in OVCAR8 cells but not HOSE6-3 cells (Fig 4A: top panel) and inhibited colony for-mation in soft-agar (representing oncogenic growth) in OVCAR8 cells (bottom)
3.4 MELK-inhibitor OTSSP167 induces G2/M cell cycle arrest, inhibits pro-liferation and colony formation, and activates apoptosis preferentially in HGSOC cell lines
The observation that shRNA-mediated depletion of MELK induced apoptosis and inhibited proliferation prompted us to determine the ef-fects of pharmacological inhibition of MELK by OTSSP167 on prolifera-tion and survival OTSSP167 has been reported to inhibit proliferaprolifera-tion
in several cancer cells of different origin in vitro and in vivo[8,9]by in-ducing a cell cycle arrest in G2/M[25] We therefore determined the ef-fects of OTSSP167 on cell cycle transition and cell morphology, on colony formation, apoptosis, and proliferation in a panel of selected cell lines The results demonstrate that OTSSP167 induced a marked
Fig 1 MELK is elevated in various cancers and is a predictor disease-free survival in all TCGA cancer samples regardless of the cancer subtype (A) Kaplan-Meier curve for disease-free survival for low and high MELK expression in all TCGA cancer samples (B) Transcriptomic data set analysis Box-whisker plots (data log transformed) showing MELK gene expression
in the various cancers compared to “normal” kidney (ovarian cancer indicated by arrow).
Fig 2 MELK expression is associated with aggressive ovarian cancer subtypes Bar chart showing MELK expression in the different subtypes as published in Heinzelmann et al 2004 [10,11]
(A) (B) Kaplan-Meier curve for low and high MELK expression (threshold determined by conditional inference trees) and progression-free survival (TCGA data set for high-grade SOC).
Trang 5arrest at the G2/M transition of the cell cycle in BG1 and OVCAR8 cells
(Supplementary Fig S2A) In addition, OTSSP167-treated cells displayed
enlarged nuclei and altered cell morphology and tubulin and F-actin
dis-tribution, typically seen in cells arrested at the G2/M transition
(Supple-mentary Fig S2B)
OTSSP167 inhibited anchorage-independent growth in soft agar
(Fig 4B) and anchorage-dependent colony formation (Fig 4C) Western
blotting demonstrated that 100 nM OTSSP167 for 48 h produced marked PARP-1 cleavage in HGSOC cell lines whereas no or weak PARP-1 cleavage was observed in normal ovarian surface epithelial cells (HOSE6-3) and in the fallopian tubal cells (FT33-tag, FT190, FT237) (Fig 4D) Breast cancer cell lines MDA-MB-468 and T47D were used as positive and negative apoptosis controls, respectively MTT-assay results (Fig 4E) showed that OTSSP167 inhibited proliferation
Fig 3 (A) Differential MELK expression in ovarian cancer cell lines as retrieved from the “Cancer Cell Line Encyclopedia (CCLE)” Broad Institute data base Data are normalized and cell lines are sorted by descending MELK expression MELK expression in MDA-MB-468 basal-like breast cancer cells (indicated in red) and T47D (blue) and MCF-7 (green) luminal breast cancer cells are shown as references (B) Western blot displaying MELK expression in cell lines representing the ovarian surface epithelium, fallopian tube epithelium, and HGSOC Whole cell protein extracts were produced, separated, and analyzed as described in “ Materials and methods ” Tubulin is the sample loading control.
Fig 4 Effects of MELK depletion and pharmacological inhibition of MELK activity on cell cycle transition, morphology, colony formation, apoptosis, and proliferation (A) Apoptosis induction (top panel) and oncogenic growth inhibition in soft-agar (bottom) in shRNA-mediated MELK-depleted cells (B) OTSSP167 inhibits anchorage-independent growth in soft agar (C) OTSSP167 reduces colony growth formation (anchorage-dependent) (D) Western blot for PARP1-cleavage in cell lines representing the ovarian surface epithelium (HOSE), the fallopian tube epithelium (FT), and ovarian cancer cell lines of various histotypes Whole cell protein extracts of cell cultures treated with 100 nM OTSSP167 for 48 h were produced, separated, and analyzed Tubulin, sample loading control; luminal (T47D) and basal (MDA-MB-468) breast cancer cell lines: negative and positive MELK controls, respectively (E) MTT-assay results for the cell lines indicated presented as IC50 concentrations sorted by the histotypes (fallopian tubes, ovary surface epithelium, HGSOC) Cells were exposed to OTSSP1167 for 72 h Mean ± SD of at least four independent experiments performed in quadruples For further details, see “ Material and methods ”.
Trang 63-times more efficiently in EOC (IC50range: 24 to 82 nM) than in FT cells
(IC50range: 134 to 160 nM) BG1 and OVCAR3 cells were to most
sensi-tive among the EOC cells (IC50: 24 nM and 36 nM) HOSE6-3 cells were
as sensitive as the two least sensitive EOC cells (TYK-nu and IGROV1)
3.5 OTSSP167 retains efficiency in drug-resistant HGSOC cells and does not
produce drug resistance acquisition
Drug resistance (both intrinsic and acquired) is one major and
criti-cal obstacle in cancer management and the development of novel
anti-cancer compounds which are highly effective also in recurrent tumors is
of great importance We therefore evaluated whether ovarian cancer
cell lines resistant to platinum compounds and Paclitaxel (both
current-ly used in ovarian cancer treatment) remain sensitive to OTSSP167
MTT-assay results demonstrate that 4-fold (p = 0.011, n = 4)
Cisplat-in-resistant TYK-nu(R) cells (Fig 5A) and 9.3-fold (p = 0.002, n = 3)
Paclitaxel-resistant IGROV1-PXL cells (Fig 5B) were as sensitive to
OTSSP167 as their parental counterparts The IGROV1-PXL cells were
generated in our laboratory by step-wise exposing IGROV1 cells to
in-creasing Paclitaxel concentrations, starting with 1μM Paclitaxel (details
see“Materials and methods”) and, unlike the Cisplatin-resistant
TYK-nu(R) cells, overexpressed multidrug resistance protein 1 (MDR1)
(Fig 5C) In contrast, four cycles of Carboplatin exposure in IGROV1
cells (starting concentration: 20μM) failed to yield a stable and
signifi-cantly Carboplatin-resistant IGROV1 subpopulation (1.22-fold, p =
0.338, n = 3)
We also determined whether OTSSP167 causes resistance
acquisi-tion We repeatedly exposed OVCAR8 cells to OTSSP167 (starting at
75 nM) and were able to produce a subline with reduced OTSSP167
sen-sitivity (1.45-fold (p = 0.123, n = 3;Fig 5D) after four cycles: no viable
cells were found when OTSSP167 concentration was increased to
120 nM or higher in two independent attempts This OVCAR8-OTS
sub-line did not show cross-resistance to Carboplatin, Paclitaxel or
Doxorubicin (Fig 5D) Taken together, these results indicate that OTSSP167 remains effective in drug-resistant ovarian cancer cells and suggest that OTSSP167 is unlikely to cause cross-resistance acquisition
in these cells
3.6 OTSSP167 sensitizes OVCAR8 cells to Carboplatin but not Paclitaxel Addressing the chemo-sensitizing property of OTTSSP167 we deter-mined whether sub-lethal concentrations of OTSSP167 sensitize OVCAR8 cells to Carboplatin and Paclitaxel MTT-assay results showed that the combination of Carboplatin and OTSSP167 was more cytotoxic than each compound itself (Fig 5E), whereas the combination of Pacli-taxel with OTSSP167 was less cytotoxic than each compound itself (Fig
5F) In contrast to a narrow additive effect for Carboplatin a clear-cut an-tagonistic effect was observed with Paclitaxel
4 Discussion Unlike for other cancers MELK and its inhibitor OTSSP167 have not yet been investigated in ovarian cancer Here we (i) compared MELK ex-pression, clinico-pathological outcome parameters in nine independent transcriptomic data sets; (ii) investigated the effects of genetic and pharmacological MELK inhibition on proliferation, oncogenic growth, and apoptosis; and (iii) evaluated the efficiency of MELK-inhibitor OTSSP167 in drug-resistant ovarian cancer cell lines, its possibility to cause resistance acquisition, and its chemosensitizing potential in these cells
Four majorfindings emerge from this study: Firstly, MELK was ele-vated in EOC and in particular increased towards aggressiveness of SOC; secondly, elevated MELK expression correlated with poorer disease outcome (progression-free survival); Thirdly, MELK inhibition/deple-tion abrogated proliferainhibition/deple-tion and oncogenic growth by arresting cells at the G2/M transition of the cell cycle and induced apoptosis in EOC
Fig 5 OTSSP167 and drug resistance Sensitivity to OTSSP167 of Cisplatin-resistant TYK-nu(R) and -sensitive TYK-nu (A) and Paclitaxel-resistant IGROV1-PXL and -sensitive IGROV1 cells (B) (C) MDR1 expression (Western blot) in parental IGROV1 and TYK-nu cells and in their respective resistant counterparts (pxl5 and nu(R)) Tubulin is sample loading control and Pos is positive MDR1 control (D) Sensitivity to OTSSP167 of OVCAR8 and OVCAR8-OTS cells (obtained by four cycles of step-wise exposure to increasing OTSSP167 concentrations: details in
“ Materials and methods ”) and their cross-sensitivity to Carboplatin, Paclitaxel, and Doxorubicin Effect of co-treatment of OVCAR8 cells with OTSSP167 and Carboplatin (E) or Paclitaxel (F) on cell viability: white bars (Carboplatin or Paclitaxel alone), grey bars (Carboplatin or Paclitaxel plus 50 nM OTSSP167), black bars (Carboplatin or Paclitaxel plus 70 nM OTSSP167) MTT-assay results presented as mean ± SD of at least 3 independent experiments performed in quadruples.
Trang 7cells Fourthly, MELK-inhibitor OTSSP167 retained its effectivity in
Pac-litaxel-resistant (IGROV1) and Cisplatin-resistant (TYK-nu) cells and
sensitized OVCAR8 cells to Carboplatin but not to Paclitaxel From this
we may conclude that elevated MELK in EOC is a possible therapeutic
target and that OTSSP167 is a potential novel compound against
(drug-resistant) aggressive ovarian cancers, in particular EOC,
undiffer-entiated cancers, and MMMT
Although elevated MELK expression has been previously reported in
various cancers[6], we were one of thefirst group to identify MELK as
an interesting target in epithelial ovarian cancer[10,11]: its gene
ex-pression profile was outstanding in our results published in 2004,
when MELK was still only known as KIAA0175 Since the development
of the small-molecule MELK-inhibitor OTSSP167 in 2012[8,9],
investi-gations into this gene have continuously grown almost exponentially
In the present study we addressed MELK expression particularly in
HGSOC, the most common and lethal gynecological cancer in women
with a poor overall survival[1]in more detail and clearly demonstrated
elevated MELK in ovarian cancer on the gene expression level, on
pro-tein level in tissues, and in cell lines In contrast, MELK was poorly
expressed in normal surface epithelial, fallopian tube cells, and in
be-nign tumors, indicating a clear-cut and significant preference of MELK
expression for ovarian cancer Intriguingly, MELK expression correlated
with increasing histopathological grades and with a shorter
progres-sion-free survival in the TCGA data set Ourfinding is therefore clinically
relevant as it proposes MELK expression as marker for tumor
aggres-siveness and progression and as prognostic predictor for poor outcome
and expands a recent study showing that MELK expression correlated
with early recurrence and poor prognosis in hepatocellular carcinoma
[26] These characteristics of MELK have been reported for BBC, where
(unlike in luminal breast cancer and normal breast) MELK was highly
expressed and correlated with poor prognosis[7] The comparability
of our results with HGSOC and those reported for BBC may be not
sur-prising and may be attributed to the similarities between these two
can-cers[11,12] Despite its specificity towards aggressive cancer types, it
remains to be elucidated whether MELK can drive oncogenic
transformation
Not less important is ourfinding that OTSSP167 significantly
re-duced proliferation and oncogenic growth at low nanomolar
concentra-tions in EOC cells, associated with a marked arrest at the G2/M
transition of the cell cycle and apoptosis OTSSP167 also inhibited
prolif-eration and induced apoptosis in fallopian tube epithelial cells but at a
considerable lesser extent Accordingly, the anti-proliferative and
pro-apoptotic effect of OTSSP167 has also been reported with other cancer
cells including BBC cells and myeloma cells[7,27–30] MELK
phosphor-ylates CDC25B which is required for the entry in mitosis[31]: this event
is abrogated by MELK inhibition, resulting in the G2/M arrest[25] MELK
seems to disable critical cell-cycle checkpoints, reduce replication stress,
and stimulate proliferation by its capacity to increase the threshold for
DNA-damage tolerance[5] Roles for FOXM1, a transcription factor
reg-ulated by MELK via direct phosphorylation[4,27,32], and for TP53[33,
34]have been suggested, and OTSSP167.has been shown to decrease
MELK and FOXM1[32]
We observed abrogation of proliferation and oncogenic growth and
apoptosis induction also in EOC cells where MELK expression was
si-lenced by doxycycline-inducible shRNA This is consistent with the
BBC cell study[7]and indicates that not only the enzymatic inhibition
of MELK but also its loss results in growth inhibition and apoptosis
This view is supported by a recent study[5]showing that the
structur-ally unrelated MELK inhibitor MELK-T1 provoked the rapid
protea-some-mediated degradation of MELK: this was associated with a rapid
and long-lasting ataxia telangiectasia-mutated (ATM) activation,
phos-phorylation of checkpoint kinase 2 (CHK2), a strong phosphos-phorylation of
TP53, a prolonged up-regulation of p21, and a down-regulation of
FOXM1 target genes The authors conclude that MELK is a key
stimula-tor of proliferation by its ability to increase the threshold for
DNA-dam-age tolerance and propose that targeting MELK by the inhibition of both
its catalytic activity and its protein stability might sensitize tumors to DNA-damaging agents or radiation therapy The latter has indeed been shown recently as genetic and pharmacological knockdown/inhibition
of MELK radio- and chemo-sensitized rectal cancer[35]and breast cancer cells[36]
Another intriguingfinding is related to the issue of OTSSP167 and drug resistance in EOC cells Our data indicate that OTSSP167 retains its effectivity in Paclitaxel- or Cisplatin-resistant cells and that overexpres-sion of MDR1 in Paclitaxel-resistant OVCAR8 cells does not impair OTSSP167 sensitivity The latter suggests that OTSSP167 efficacy is MDR1-independent, but is opposed to a recent study reporting OTSSP167 resistance in ABCB1 transporter-overexpressing myeloma cells [7,27–30] Our data also indicate that repeated exposure to OTSSP167 cause drug resistance acquisition in OVCAR8 cells, at least to some extent which may be considered minor (1.45-fold) when compared
to those obtained for Paclitaxel (9.3-fold) and Cisplatin (4.0-fold) Regard-less of whether or not considered minor the observed resistance to OTSSP167 is not associated with cross-resistance to Paclitaxel, Carboplatin, and Doxorubicin
Interestingly, our results confirm for Carboplatin the reported sensitiz-ing property of OTSSP167 to DNA damagsensitiz-ing agents[35]but also provide
an opposing example with Paclitaxel: as a microtubule-poison Paclitaxel kills cells by inducing mitotic arrest and/or interferes with the interphase
[37], but it is unknown how MELK inhibition encounters the cytotoxic ef-fect of Paclitaxel Thisfinding, though deriving from one cell line only, may be of clinical interest, in particular for the design of combinational clinical trial
These characteristics of OTSSP167 together with its radio and (at least for some drugs) chemosensitizing property are clinically significant and may underline the intriguing advantages of OTSSP167 Interestingly, a re-cent study suggests that OTSSP167 has“off-MELK” effects, i.e that OTSSP167 has other target kinases and may have additional mechanisms
of action for cancer cell killing[38] For instance, in addition to MELK, OTSSP167 inhibits Aurora B and BUB1, both of which compromise mitotic checkpoint regulation and hence contribute to cell killing in a MELK-inde-pendent manner Whether and to which extent these kinases can be accounted for the apoptotic effect of OTSSP167 observed in ovarian cancer cells remains open, especially since shRNA against MELK recapitulated the effects seen with OTSSP167 in our study
Our results collectively propose MELK as an attractive drugable target in patients with (refractory) ovarian cancer This warrants the evaluation of OTSSP167 in clinical trials, either as single compound or in combination with chemotherapeutic drugs (e.g Carboplatin) or radiation OTSSP167 may not only present an inter-esting alternative to the current treatment of the aggressive forms
of ovarian cancer, the survival of which has, despite improved surgi-cal techniques and drug regimens, only changed marginally[1,39,40]
but may also provide new hope to improve the poor survival rate of this disease
Conflict of interest statement All authors declare no conflict of interest.
Funding and acknowledgements This work was supported by Swiss National Science Foundation (310030_156982, 310030_143619 and 32 to VHS); OncoSuisse Grant (KFS_3013-08-2012 to VHS), Krebsliga Beider Basel (06-2013
to VHS) We thank Prof R Drapkin (Penn Ovarian Cancer Research Center, University of Pennsylvania, Philadelphia, PA, USA) for gener-ously providing FTSEC cells We thank Danny Labes and Emmanuel Traunecker (Flow Cytometry Facility) and Michael Abanto and Beat Erne (Microscopy Facility) for all necessary support We also thank Monica Nunez Lopez for her experimental contributions to this manuscript
Trang 8Appendix A Supplementary data
Supplementary data to this article can be found online athttp://dx
doi.org/10.1016/j.ygyno.2017.02.016
References
[1] D.D Bowtell, S Bohm, A.A Ahmed, P.J Aspuria, R.C Bast Jr., V Beral, et al.,
Rethink-ing ovarian cancer II: reducRethink-ing mortality from high-grade serous ovarian cancer,
Nat Rev Cancer 15 (2015) 668–679.
[2] R.E Bristow, R.S Tomacruz, D.K Armstrong, E.L Trimble, F.J Montz, Survival effect of
maximal cytoreductive surgery for advanced ovarian carcinoma during the
plati-num era: a meta-analysis, J Clin Oncol 20 (2002) 1248–1259.
[3] O.W Foley, J.A Rauh-Hain, M.G del Carmen, Recurrent epithelial ovarian cancer: an
update on treatment, Oncology 27 (2013) 288–294 (98).
[4] R Ganguly, A Mohyeldin, J Thiel, H.I Kornblum, M Beullens, I Nakano, MELK-a
conserved kinase: functions, signaling, cancer, and controversy, Clin Transl Med.
4 (2015) 11.
[5] L Beke, C Kig, J.T Linders, S Boens, A Boeckx, E van Heerde, et al., MELK-T1, a
small-molecule inhibitor of protein kinase MELK, decreases DNA-damage tolerance
in proliferating cancer cells, Biosci Rep 35 (2015).
[6] D Gray, A.M Jubb, D Hogue, P Dowd, N Kljavin, S Yi, et al., Maternal embryonic
leucine zipper kinase/murine protein serine-threonine kinase 38 is a promising
therapeutic target for multiple cancers, Cancer Res 65 (2005) 9751–9761.
[7] Y Wang, Y.M Lee, L Baitsch, A Huang, Y Xiang, H Tong, et al., MELK is an oncogenic
kinase essential for mitotic progression in basal-like breast cancer cells, elife 3
(2014), e01763.
[8] S Chung, H Suzuki, T Miyamoto, N Takamatsu, A Tatsuguchi, K Ueda, et al.,
Devel-opment of an orally-administrative MELK-targeting inhibitor that suppresses the
growth of various types of human cancer, Oncotarget 3 (2012) 1629–1640.
[9] S Chung, Y Nakamura, MELK inhibitor, novel molecular targeted therapeutics for
human cancer stem cells, Cell Cycle 12 (2013) 1655–1656.
[10] V.A Heinzelmann-Schwarz, M Gardiner-Garden, S.M Henshall, J Scurry, R.A.
Scolyer, M.J Davies, et al., Overexpression of the cell adhesion molecules DDR1,
Claudin 3, and Ep-CAM in metaplastic ovarian epithelium and ovarian cancer,
Clin Cancer Res 10 (2004) 4427–4436.
[11] Cancer Genome Atlas Research N, Integrated genomic analyses of ovarian
carcino-ma, Nature 474 (2011) 609–615.
[12] Cancer Genome Atlas N, Comprehensive molecular portraits of human breast
tu-mours, Nature 490 (2012) 61–70.
[13] M.D Iglesia, B.G Vincent, J.S Parker, K.A Hoadley, L.A Carey, C.M Perou, et al.,
Prog-nostic B-cell signatures using mRNA-seq in patients with subtype-specific breast
and ovarian cancer, Clin Cancer Res 20 (2014) 3818–3829.
[14] A Prat, B Adamo, C Fan, V Peg, M Vidal, P Galvan, et al., Genomic analyses across
six cancer types identify basal-like breast cancer as a unique molecular entity, Sci.
Rep 3 (2013) 3544.
[15] J Brouwer, J Kluiver, R.C de Almeida, R Modderman, M.M Terpstra, K Kok, et al.,
Small RNA sequencing reveals a comprehensive miRNA signature of
BRCA1-associ-ated high-grade serous ovarian cancer, J Clin Pathol (2016).
[16] K.P Pennington, T Walsh, M.I Harrell, M.K Lee, C.C Pennil, M.H Rendi, et al.,
Germline and somatic mutations in homologous recombination genes predict
plat-inum response and survival in ovarian, fallopian tube, and peritoneal carcinomas,
Clin Cancer Res 20 (2014) 764–775.
[17] A.M Patch, E.L Christie, D Etemadmoghadam, D.W Garsed, J George, S Fereday,
et al., Whole-genome characterization of chemoresistant ovarian cancer, Nature
521 (2015) 489–494.
[18] A.M Karst, R Drapkin, Primary culture and immortalization of human fallopian tube
secretory epithelial cells, Nat Protoc 7 (2012) 1755–1764.
[19] N Yoshiya, S Adachi, Y Misawa, H Yuzawa, T Honda, K Kanazawa, et al., Isolation
of cisplatin-resistant subline from human ovarian cancer cell line and analysis of its
cell-biological characteristics, Nihon Sanka Fujinka Gakkai Zasshi 41 (1989) 7–14.
[20] A Fedier, K.J Dedes, P Imesch, A.O Von Bueren, D Fink, The histone deacetylase in-hibitors suberoylanilide hydroxamic (Vorinostat) and valproic acid induce irrevers-ible and MDR1-independent resistance in human colon cancer cells, Int J Oncol 31 (2007) 633–641.
[21] T Hothorn, K Hornik, A Zeileis, Unbiased recursive partitioning: a conditional infer-ence framework, J Comput Graph Stat 15 (2006) 651–674.
[22] T.M Thernau, P.M Grambsch, Modeling Survival Data: Extending the Cox Model, Springer-Verlag, New York, 2000.
[23] A Flesken-Nikitin, C.I Hwang, C.Y Cheng, T.V Michurina, G Enikolopov, A.Y Nikitin, Ovarian surface epithelium at the junction area contains a cancer-prone stem cell niche, Nature 495 (2013) 241–245.
[24] R Perets, G.A Wyant, K.W Muto, J.G Bijron, B.B Poole, K.T Chin, et al., Transforma-tion of the fallopian tube secretory epithelium leads to high-grade serous ovarian cancer in Brca;Tp53;Pten models, Cancer Cell 24 (2013) 751–765.
[25] S Li, Z Li, T Guo, X.F Xing, X Cheng, H Du, et al., Maternal embryonic leucine zipper kinase serves as a poor prognosis marker and therapeutic target in gastric cancer, Oncotarget 7 (2016) 6266–6280.
[26] H Xia, S.N Kong, J Chen, M Shi, K Sekar, V.P Seshachalam, et al., MELK is an onco-genic kinase essential for early hepatocellular carcinoma recurrence, Cancer Lett.
383 (2016) 85–93.
[27] K Joshi, Y Banasavadi-Siddegowda, X Mo, S.H Kim, P Mao, C Kig, et al., MELK-de-pendent FOXM1 phosphorylation is essential for proliferation of glioma stem cells, Stem Cells 31 (2013) 1051–1063.
[28] R Kuner, M Falth, N.C Pressinotti, J.C Brase, S.B Puig, J Metzger, et al., The maternal embryonic leucine zipper kinase (MELK) is upregulated in high-grade prostate can-cer, J Mol Med 91 (2013) 237–248.
[29] M.L Lin, J.H Park, T Nishidate, Y Nakamura, T Katagiri, Involvement of maternal embryonic leucine zipper kinase (MELK) in mammary carcinogenesis through in-teraction with Bcl-G, a pro-apoptotic member of the Bcl-2 family, Breast Cancer Res 9 (2007) R17.
[30] A.T Stefka, J.H Park, Y Matsuo, S Chung, Y Nakamura, A.J Jakubowiak, et al., Anti-myeloma activity of MELK inhibitor OTS167: effects on drug-resistant Anti-myeloma cells and putative myeloma stem cell replenishment of malignant plasma cells, Blood Cancer J 6 (2016), e460.
[31] N Davezac, V Baldin, J Blot, B Ducommun, J.P Tassan, Human pEg3 kinase associ-ates with and phosphorylassoci-ates CDC25B phosphatase: a potential role for pEg3 in cell cycle regulation, Oncogene 21 (2002) 7630–7641.
[32] T Kato, H Inoue, S Imoto, Y Tamada, T Miyamoto, Y Matsuo, et al., Oncogenic roles
of TOPK and MELK, and effective growth suppression by small molecular inhibitors
in kidney cancer cells, Oncotarget (2016).
[33] C Gu, Y.K Banasavadi-Siddegowda, K Joshi, Y Nakamura, H Kurt, S Gupta, et al., Tumor-specific activation of the C-JUN/MELK pathway regulates glioma stem cell growth in a p53-dependent manner, Stem Cells 31 (2013) 870–881.
[34] M Fischer, M Quaas, L Steiner, K Engeland, The p53-p21-DREAM-CDE/CHR path-way regulates G2/M cell cycle genes, Nucleic Acids Res 44 (2016) 164–174.
[35] S Choi, J.L Ku, Resistance of colorectal cancer cells to radiation and 5-FU is
associat-ed with MELK expression, Biochem Biophys Res Commun 412 (2011) 207–213.
[36] C Speers, S.G Zhao, V Kothari, A Santola, M Liu, K Wilder-Romans, et al., Maternal embryonic leucine zipper kinase (MELK) as a novel mediator and biomarker of radioresistance in human breast cancer, Clin Cancer Res (2016).
[37] B.A Weaver, How Taxol/paclitaxel kills cancer cells, Mol Biol Cell 25 (2014) 2677–2681.
[38] W Ji, C Arnst, A.R Tipton, M.E Bekier 2nd, W.R Taylor, T.J Yen, et al., OTSSP167 ab-rogates mitotic checkpoint through inhibiting multiple mitotic kinases, PLoS One 11 (2016), e0153518.
[39] S Vaughan, J.I Coward, R.C Bast, A Berchuck, J.S Berek, J.D Brenton, et al., Rethink-ing ovarian cancer: recommendations for improvRethink-ing outcomes, Nat Rev Cancer 11 (2011) 719–725.
[40] S Banerjee, S.B Kaye, New strategies in the treatment of ovarian cancer: current clinical perspectives and future potential, Clin Cancer Res 19 (2013) 961–968.