Ovarian cancer cells with CD133+ phenotype is more resistant against Ngai Bun Boesenbergia pandurata extract than original ovarian cancer cells

Progress in Stem Cell, 7(1):290-295  
Original Research  
Open Access Full Text Article  
Ovarian cancer cells with CD133+ phenotype is more resistant  
against Ngai Bun Boesenbergia pandurata extract than original  
ovarian cancer cells  
Oanh Thi-Kieu Nguyen1,2, Phuc Van Pham1,2,3,4,*  
ABSTRACT  
Introduction: Ovarian cancer is one of the most common cancers in women. Due to the diffi-  
culty in early detection and treatment of ovarian cancer, many research studies and clinical trials  
Use your smartphone to scan this  
QR code and download this article  
have been developed to discover more efficient therapies. Besides Western medicine, traditional  
medicine has gained increased interest as a research field with potential to lead to the production of  
marketable therapeutic products. With the diversity of tropical plants in Asia, traditional medicine  
has been very popular and has served as a traditional therapy for generations. The Ngai bun (Boe-  
senbergia pandurata) root is used not only as a food spice but also in ethnomedicine. This study  
aimed to compare the anti-tumor activity of Boesenbergia pandurata root extract against ovarian  
cancer cells and CD133+ovarian cancer cells that were enriched from the original ovarian cancer  
cells. Methods: Crude extract of Boesenbergia pandurata roots were prepared in two kinds of sol-  
vents (methanol and chloroform). The ovarian cancer cells OVP-10 were used in this study. The  
population of CD133+ ovarian cancer cells (CD133+OVP-10) were sorted from the OVP-10 cancer  
cells. Both OVP-10 cells and CD133+OVP-10 cells were treated with these crude extracts. Adipose-  
derived stem cells (ADSCs) were used as control normal cells for all assays. The anti-tumor activity  
of extracts were evaluated based on the IC50 values. Results: Based on the IC50 index, the chlo-  
roform extract had an anti-tumor activity higher than that of methanol extract, on both OVP-10  
and CD133+OPV-10 cells (IC50 of methanol and chloroform extracts were 330.1 16.9 µg/mL and  
246.5 21.2 µg/mL, respectively, for OVP-10 cells; IC50 of methanol and chloroform extracts were  
411.8 83.7 µg/mL and 307 9.2 µg/mL respectively, for CD133+OVP-10 cells). The results also  
showed that CD133+OVP-10 cells were more resistant to chloroform extract than were OVP-10 cells  
(307 9.2 µg/mL vs. 246.5 21.2 µg/mL, respectively, for CD133+OVP-10 vs. OVP-10 cells, p <  
0.05). Conclusion: The chloroform extract of Boesenbergia pandurata roots displayed strong anti-  
tumor activity against ovarian cancer cells OVP-10 and CD133+OVP-10; the latter cells were found  
to be more resistant than the original ovarian cancer cells.  
1Stem Cell Institute, University of Science  
Ho Chi Minh City, Viet Nam  
2Vietnam National University Ho Chi  
Minh City, Viet Nam  
3Cancer Research Laboratory, University  
of Science Ho Chi Minh City, Viet Nam  
4Laboratory of Stem Cell Research and  
Application, University of Science Ho  
Chi Minh City, Viet Nam  
Correspondence  
Phuc Van Pham, Stem Cell Institute,  
University of Science Ho Chi Minh City,  
Viet Nam  
Key words: CD133+ cancer stem cells, ovarian cancer stem cells, ovarian cancer, Boesenbergia  
pandurata  
Vietnam National University Ho Chi Minh  
City, Viet Nam  
Cancer Research Laboratory, University  
of Science Ho Chi Minh City, Viet Nam  
irinotecan, and camptothecins) and microorganisms  
INTRODUCTION  
In 2019, the American Cancer Society estimated ovar-  
Laboratory of Stem Cell Research and  
Application, University of Science Ho Chi  
Minh City, Viet Nam  
(e.g. doxorubicin, mitomycin, and bleomycin)3.  
Doxorubicin (DOX) is the most popular anti-cancer  
ian cancer as the leading cause of death in gyneco-  
Email: pvphuc@hcmuns.edu.vn;  
drug, which is currently widely used for treatment of  
many kinds of human cancers, both solid and hema-  
tological4. DOX was shown to induce resistance in  
3D spheroids, at a rate higher than that exhibited in  
standard 2D cell culture5. Additionally, tirapazamine  
(TPZ; 3-amino-1,2,4-benzotriazine 1,4 dioxide) is a  
new class of cytotoxic drugs with a focus on treating  
hypoxic mammalian cells6. When cultured in 2D,  
the toxicity of DOX with target cells is more sensi-  
tive than TPZ. Cell viability of cells with the testing  
drug was identified by using the Alamar Blue assay.  
By incorporating a fluorometric/colorimetric growth  
logical disease, ranking 5th of the 10 leading cancer  
types in women1. Ovarian cancer is a rare disease, in  
which early detection is difficult and surgical strate-  
gies are the first step in treatment for this cancer2.  
However, the extent of surgery depends on how far  
the ovarian cancer has spread; chemotherapy must be  
used in the next steps to eradicate any residual cancer  
cells still present in the body aꢀer surgery. e goal  
of chemotherapy is to destroy the cancer by inhibit-  
ing the proliferation of cancer cells. Chemotherapy  
is a potential treatment for prolonging the cancer pa-  
tient’s life. ere are many kinds of anti-cancer drugs  
History  
Received: 22 January 2020  
Accepted: 05 March 2020  
Published: 19 March 2020  
DOI : 10.15419/psc.v7i1.408  
Copyright  
© Biomedpress. This is an open-  
access article distributed under the  
terms of the Creative Commons  
Attribution 4.0 International license.  
from natural sources, such as plants (e.g. vincristine, indicator based on detection of metabolic assay, the  
Citethisarticle: NguyenOT, PhamPV. OvariancancercellswithCD133+ phenotypeismoreresistant  
against Ngai Bun Boesenbergia pandurata extract than original ovarian cancer cells. Progress Stem  
Cell; 7(1):290-295.  
290  
Progress in Stem Cell, 7(1):290-295  
CD133+ cell sorting  
Alamar blue assay can quantitatively measure the pro-  
liferation of various human and animal cells. Many  
new drugs and therapies have been under develop-  
ment and are being tested in ongoing clinical trials,  
with the aim of finding the most effective therapy of  
ovarian cancer.  
OVP-10 cells were labelled with CD133 magnetic  
beads and sorted by magnetic-activated cell sorting  
(MACS), per the instructions of the Human CD133  
MicroBead kit (kit and MACS sorting multi-stand  
were obtained from Miltenyi Biotec Inc., Auburn,  
Recently, besides modern medicine, ethnomedicine CA). e CD133+ cancer cells were expanded to ob-  
tain enough cells for the bioassays.  
has also become a potential direction for the devel-  
opment of new drugs. Ethnomedicine is a study of  
the traditional medicine based on the use of plant  
compounds. Ngai bun (Boesenbergia pandurata) is  
Drugs  
e standard drugs used in this study were doxoru-  
a ginger species belonging to the Zingiberaceae fam- bicin (DOX) and tirapazamine (TPZ); both were pur-  
chased from Sigma-Aldrich. Ngai bun extract was  
isolated from fresh root, following a previously pub-  
lished protocol, and dissolved in the different sol-  
vents7.  
ily; it is cultivated in tropical countries such as South-  
east Asia, India, Sri Lanka and South China. is  
species has many local names, such as “ngai bun” in  
Vietnam7, “temu kunci” in Indonesia8, krachai (or  
krachai-dang) in ailand, and Chinese key (or fin-  
gerroot) in English9. is plant is mainly used as a  
spice and contains several prenylated chalcones and  
flavonoids with many biological activities, including  
antifungal, antibacterial, anti-inflammatory and anti-  
Alamar Blue assay  
e cell viability of ovarian cancer cells was tested by  
Alamar Blue assay. Cells were plated in a 96-well plate  
at a density of 2 x 104 cells/well. Aꢀer plating for  
24 h, cells were treated with the drugs at six differ-  
ing concentrations (2000, 1000, 500, 250, 125, 62, and  
0 µg/mL) for 48 h. e culture medium was then  
removed and the wells were replaced with fresh me-  
dia. As a negative control, fresh media was also added  
to empty wells. All wells were added with 10 µl of  
the Alamar Blue solution and then re-incubated at  
370C, 5% CO2 for 4 h. Data was collected by using  
a DTX880 system (Beckman-Coulter, Brea, CA), and  
fluorescence was monitored at 530-560 nm excitation  
wavelength and 590 nm emission wavelength.  
cancer functions7,8  
.
e cancer stem cell (CSC) theory proposes that in  
the cancer cell population, there exists a small popu-  
lation of stem cells which have the capability of self-  
renewal and tumor-initiation. Recent research studies  
had revealed that CSCs in ovarian cancer are not only  
responsible for primary tumor growth and metastasis  
but also for chemotherapy resistance10. In studies on  
ovarian cancer stem cells, CD133 is one of most pop-  
ular cell surface markers which has been used for the  
isolation of CSCs from various type of cancers, such  
as breast, colon, ovarian and prostate cancer10,11  
.
Statistical analysis  
In this research study, we aimed to test the toxicity  
of Ngai bun (Boesenbergia pandurata) extract in dif-  
ferent solvents on ovarian cancer OVP-10 cells and  
CD133+ sorted OVP-10 cells (CD133+OVP-10 cells)  
to assess the therapeutic potential of this plant for  
ovarian cancer treatment.  
Each experiment was repeated three times. e IC50  
and significant differences between mean values were  
calculated by using GraphPad Prism 7.0 (GraphPad  
Inc., La Jolla, CA), with p-value < 0.05 set as statistical  
significance.  
RESULTS  
MATERIALS AND METHODS  
Isolation of human ovarian cancer  
CD133+OVP-10 cells  
Cell culture  
Human ovarian cancer OVP-10 cells were grown in  
DMEM/F12 with 10% fetal bovine serum (Sigma-  
Aldrich, St. Louis, MO). e cells were maintained  
in cell culture flasks (25-cm2 and 75-cm2) in an in-  
cubator set at 370C and 5% CO2. Cells were cultured  
and passaged two to three times before processing in  
the bioassays.  
Human ovarian cancer OVP-10 cells were cultured  
and expanded in 75-cm2 flasks. When cells reached  
approximately 70% confluency (Figure 1 B), cells  
will be trypsinized and subcultured in DMEM/F12  
medium supplemented with 10% FBS. Aꢀer pass-  
ing two to three times, OVP-10 cells were subjected  
to bioassays, and/or sorted for CD133+ cells using  
291  
Progress in Stem Cell, 7(1):290-295  
MACS (Figure 1A). CD133+OVP-10 cells were la- with CD133+OVP-10 cells, no significant differences  
belled with CD133 magnetic beads, isolated and cul- were observed between the two cell lines; the IC50  
of the corresponding cells (ADSCs vs. CD133+OVP-  
10 cells) with MeOH solvent, respectively, were 497.2  
32.4 µg/mL and 411.8 83.7 µg/mL (Figure 3B)  
(p > 0.05). In Figure 3C, with the same cells in a  
different solvent (CHCl3), the IC50 index of OVP-10  
cells was 246.5 21.2 µg/mL, which was significantly  
different from the IC50 of 474.6 18.8 µg/mL for  
ADSCs (p < 0.05). e results inFigure 3D showed  
that CD133+OVP-10 cells treated with CHCl3 extract  
corresponded to an IC50 index that was significantly  
lower than that for ADSC cells (307 9.2 µg/mL vs.  
474.6 18.8 µg/mL, respectively) (p < 0.05).  
tured to confluency before treatment with the com-  
pounds (Figure 1C). e morphology of OVP-10 cells  
and CD133+OVP-10 cells showed no significant dif-  
ferences; however, in culture, the proliferation time of  
CD133+OVP-10 cells was slightly longer than that of  
OVP-10 cells.  
Testing an ovarian cell model for drug  
screening with standard drugs  
e cell concentration also affected the in vitro bioas-  
say for anti-cancer drug screening12. e growth of  
OVP-10 cells was tested and the model for screening  
the different extracts was optimized. As seen in Fig-  
ure 2A, OVP-10 cells continued to proliferate aꢀer 7  
days. e bioassay was performed in 3 days: on the  
first day, cells were plated in wells and incubated for  
24 h; next, cells were treated with compounds for 48  
h; on the third day, cells were processed for Alamar  
blue assay and the IC50 indexes were calculated. Dur-  
ing these days, OVP-10 cells were still stable and there  
was little increase in cell number (Figure 2A). e  
OVP-10 cells were treated with standard drugs (dox-  
orubicin or tirapazamine) to confirm that the OVP-  
10 cell model could be used for screening. Doxoru-  
bicin is popular standard drug which used as a control  
in many studies of drug screening. Specifically, dox-  
orubicin only affected the monolayer cell model but  
not the three-dimensional (3D) cell culture model.  
In contrast, tirapazamine was only effective in the  
3D cell culture model. As shown in Figure 2B and  
Figure 2C, the IC50 of doxorubicin in OVP-10 cells  
(168.9 2.3 nM) and that of CD133+OVP-10 cells  
(567.7 95.7 nM) were highly different (p < 0.01).  
DISCUSSION  
Ovarian cancer is the most serious gynecologic can-  
cer, typically diagnosed at an advanced stage13,14  
.
e current standard treatment for ovarian cancer is  
surgery and chemotherapy. e chemotherapy strat-  
egy is faced with many obstacles including cancer  
metastasis and resistance of tumor with drugs. is  
has motivated the development of drug discovery to  
help find novel potentially therapeutic compounds for  
anti-cancer treatment. In drug screening, cells must  
be in the proliferation stage and should be stable in  
the testing with drugs. In this study, the OVP-10 cells  
continued to proliferate for 7 days and were suitable  
for our assay. Aꢀer 24 h of plating, OVP-10 cells  
showed a low increase in cell number and aꢀer 3 days  
(i.e. the day of the Alamar blue assay), cells were still  
in log phase of proliferation.  
Despite the disadvantage of the 2D model and the  
development of the 3D model, the 2D model is still  
very popular for drug screening. Evaluation of the 2D  
model using in vitro bioassays, such as MTT or Ala-  
mar Blue assay, is necessary to assess the efficiency of  
the 2D model in anti-cancer drug screening12. e  
cell concentration and drug concentration parame-  
ters have a great effect on the success of the in vitro  
bioassays. As shown in this study, the cell concen-  
tration for plating in 96-well plates was 1000-2500  
cells/well, and 6 parameters of drugs or extract con-  
centrations were at the very least required for calcula-  
Ngai bun extract dissolved in chloroform  
solvent had a greater effect on both OVP-  
10 and CD133+OVP-10 cancer cells than  
methanol solvent  
e IC50 index of Ngai bun extract in chloroform sol-  
vent (CHCl3) had a greater effect on killing ovarian  
cancer OVP-10 cells when compared with the effect  
of this extract on adipose-derived stem cells (ADSCs),  
which were used as the control. e IC50 of methanol  
(MeOH) and chloroform (CHCl3) solvent were 330.1  
16.9 µg/mL and 246.5 21.2 µg/mL, respec-  
tively. As indicated inFigure 3A, the MeOH extract  
had a different effect on ADSCs and OVP-10 cells,  
tion of the IC50 index15  
.
Use of the traditional extract from plants could kill  
cancer cells with fewer effects on normal cells. In  
this study, Ngai bun (Boesenbergia pandurata) extract  
was demonstrated to be toxic for OVP-10 cells, when  
compared with adipose derived stem cells as control.  
When comparing different Boesenbergia species (B.  
as shown by the IC50 index of 497.2 32.4 µg/mL armeniaca, B. rotunda, or B. pulchella var attenuate),  
and 330.1 16.9 µg/mL, respectively. However, Jing et al. showed that Boesenbergia rotunda extract in  
292  
Progress in Stem Cell, 7(1):290-295  
Figure 1: Isolation and expansion of CD133+OVP-10 cells from OVP-10 ovarian cancer cells. (A) Procedure  
of CD133+OVP-10 cells isolation by magnetic activated cell sorting. (B) Human ovarian cancer OVP-10 cells. (C)  
Human ovarian carcinoma CD133+OVP-10 cells. Pictures were taken at 20 X magnification.  
methanol had the strongest inhibitory effects against the IC50 of Ngai bun extract which could kill 50% of  
CD133+OVP-10 cells, that concentration could kill  
more than 50% of OVP-10 cells but less than 50% of  
ADSC cells. is shows that the dose of drug used for  
treatment must be chosen carefully.  
In this study, two kind of solvents were chosen to eval-  
uate which was the best solvent for dissolving Ngai  
bun (Boesenbergia pandurata) extract, and still main-  
tain the functions of the extract. As observed, the  
chloroform-dissolved extract induced better toxicity  
than the methanol-dissolved extract. Besides the ap-  
propriate concentrations, the suitable solvent is also a  
key factor for determining the success of drug discov-  
ery.  
CaOV3 ovarian cancer and different types of can-  
cers, such as breast cancer MDA-MB231 (IC50 66.5 ꢀ  
2.12 µg/mL), MCF7 (IC50 51 µg/mL), cervical can-  
cer HeLa (IC50 66.5 2.12 µg/mL), and colon cancer  
HT-29 (IC50 52 2.12 µg/mL). Boesenbergia genera  
is potentially potent extract for treatment of ovarian  
cancer. OVP-10 cells is one of the targets for investi-  
gation of drug cytotoxicity against ovarian cancer16  
.
Moreover, the synergistic anti-tumor effect would be  
combined to develop new therapies for ovarian can-  
cer treatment17. From different studies, the extract of  
Boesenbergia genera exhibit robust potency that can  
be utilized as an potential candidates for the devel-  
opment of new anti-cancer drugs. Advances in drug  
discovery will require identifying and developing new  
and innovative marketable pharmaceutical products.  
Future studies from this research should focus further  
on the discovery of such compounds.  
Besides the toxicity towards ovarian cancer OVP-10  
cells in a dose-dependent manner, Ngai bun (Boesen-  
bergia pandurata) extract in chloroform was demon-  
strated to inhibit the cell viability of CD133+OVP-  
10 cells, representing ovarian cancer stem cells. At  
CONCLUSION  
Overall, the data obtained from this study shows  
that Ngai bun (Boesenbergia pandurata) chloroform-  
dissolved extract is more toxic on OVP-10 cells  
than on CD133+OVP-10 cells. e cytotoxicity of  
the chloroform extract was also higher that of the  
methanol extract.  
293  
 
Progress in Stem Cell, 7(1):290-295  
Figure 2: Model for screening the efficiency of Ngai bun extract. (A) Proliferation of human ovarian cancer  
OVP-10 cells in 96-well plates over10 days. (B, C) Effect of standard drugs (doxorubicin or tirapazamine) on human  
ovarian cancer OVP-10 cells (B) and CD133+ sorted human ovarian cancer (CD133+OVP-10) cells. ∗∗∗: p < 0.001.  
Abbreviations: DOX: dororubicin, TPZ: tirapazamine  
Figure 3: The IC50 of human ovarian cancer OVP-10 cells and CD133+OVP-10 cells in different solutions.  
(A) methanol (MeOH), (B) chloroform (CCl3). Each experiment was processed three times, and statistical analysis  
was performed by GraphPad Prism 7.0 with *p<0.05 (** p<0.0021, ***p<0.0002). Abbreviations: ADSCs: Adipose  
derived stem cells.  
294  
   
Progress in Stem Cell, 7(1):290-295  
7. Hai N, Phong L, Mai N, Nhan N. Flavanones from the rhizomes  
of Boesenbergia pandurata. Science and Technology Devel-  
opment Journal. 2019;2(4). Available from: https://doi.org/10.  
8. Chahyadi A, Hartati R, Wirasutisna K, Elfahmi. Boesenbergia  
Pandurata Roxb. An Indonesian Medicinal Plant: Phytochem-  
istry, Biological Activity, Plant Biotechnology Procedia Chem-  
istry. 2014;13:13–37. Available from: https://doi.org/10.1016/j.  
9. Eng-Chong T, Yean-Kee L, Chin-Fei C, Choon-Han H, Sher-  
Ming W, Li-Ping C, et al. Boesenbergia rotunda: From Eth-  
nomedicine to Drug Discovery. Evidence-Based Complemen-  
tary and Alternative Medicine. 2012;PMID: 23243448. Avail-  
10. SusterNK, Virant-KlunI. Presenceandroleofstemcellsinovar-  
ian cancer. World J Stem Cells. 2019;11(7):383–397. PMID:  
383.  
ABBREVIATIONS  
CSC: Cancer stem cell  
DOX: Doxorubicin  
IC50: the half-maximal inhibitor concentrations  
TPZ: Tirapazamine  
CONFLICT OF INTEREST  
e authors report no conflicts of interest in this work.  
AUTHORS’ CONTRIBUTION  
All authors equally contributed in this work and ap-  
proved the final version of manuscript for submission.  
ACKNOWLEDGEMENT  
11. Keyvani V, Farshchian M, Esmaeili SA, Yari H, Moghbeli M,  
Nezhad SR, et al. Ovarian cancer stem cells and targeted  
therapy. Journal of Ovarian Research. 2019;12(1):120. PMID:  
e authors thank to Vietnam National University, Ho  
Chi Minh City (VNU-HCM) for funding this project,  
under grant number A2015-18-01/HD-KHCN.  
12. Lieberman M, Patterson G, Moore R. In vitro bioassays for  
anticancer drug screening: effects of cell concentration and  
other assay parameters on growth inhibitory activity. Cancer  
letters. 2001;173(1):21–29. Available from: https://doi.org/10.  
13. Le D, Kubo T, Fujino Y, Sokal D, Vach T, Pham T, et al. Repro-  
ductive factors in relation to ovarian cancer: a case-control  
study in Northern Vietnam. Contraception. 2012;86(5):494–  
499. PMID: 22579106. Available from: https://doi.org/10.1016/  
14. Cortez A, Tudrej P, Kujawa K, Lisowska K. Advances in ovarian  
cancertherapy. CancerChemotherPharmacol. 2018;81(1):17–  
38. PMID: 29249039. Available from: https://doi.org/10.1007/  
REFERENCES  
1. Siegel R, Miller K, Jemal A. Cancer statistics. CA: a cancer  
journal for clinicians 2019. 2019;69(1):7–34. PMID: 30620402.  
2. Jammal M, Lima C, Murta E, Nomelini R. Is Ovarian Cancer  
Prevention Currently Still a recommendation of Our Grand-  
parents? Revista brasileira de ginecologia e obstetricia : re-  
vista da Federacao Brasileira das Sociedades de Ginecologia e  
Obstetricia. 2017;39(12):676–685. PMID: 29179244. Available  
3. Kathiresanb P. In vitro cytotoxicity MTT assay in Vero, HepG2  
and MCF -7 cell lines study of Marine Yeast. J App Pharm Sci.  
2015;5(3):080–084. Available from: https://doi.org/10.7324/  
4. Yang F, Teves S, Kemp C, Henikoff S. Doxorubicin, DNA torsion,  
and chromatin dynamics. Biochimica et Biophysica Acta (BBA)  
- Reviews on Cancer. 2014;1845(1):84–89. PMID: 24361676.  
5. Nunes A, Costa E, Barros A, de Melo-Diogo D, Correia I. Es-  
tablishment of 2D Cell Cultures Derived From 3D MCF-7  
Spheroids Displaying a Doxorubicin Resistant Profile. Biotech-  
nology Journal. 2019;14(4):1800268. PMID: 30242980. Avail-  
6. Reddy S, Williamson S. Tirapazamine: a novel agent targeting  
hypoxic tumor cells. Expert opinion on investigational drugs.  
2009;18(1):77–87. PMID: 19053884. Available from: https://  
15. Hughes J, Rees S, Kalindjian S, Philpott K. Principles of  
early drug discovery.  
British journal of pharmacology.  
2011;162(6):1239–1249. PMID: 21091654. Available from:  
16. Jakubowska-Mucka A, Sienko J, Zapala L, Wolny R, Lasek W.  
Synergistic cytotoxic effect of sulindac and pyrrolidine dithio-  
carbamate against ovarian cancer cells. Oncology reports.  
2012;27(4):1245–1250. PMID: 22266802. Available from:  
17. Anyst K, Janyst M, Siernicka M, Lasek W. Synergistic anti-  
tumor effects of histone deacetylase inhibitor scriptaid and  
bortezomib against ovarian cancer cells. Oncology reports.  
2018;39(4):1999–2005. Available from: https://doi.org/10.  
295  
                                 
pdf 6 trang yennguyen 14/04/2022 2000
Bạn đang xem tài liệu "Ovarian cancer cells with CD133+ phenotype is more resistant against Ngai Bun Boesenbergia pandurata extract than original ovarian cancer cells", để tải tài liệu gốc về máy hãy click vào nút Download ở trên

File đính kèm:

  • pdfovarian_cancer_cells_with_cd133_phenotype_is_more_resistant.pdf