miR-335 negatively regulates osteosarcoma stem cell-like properties by targeting POU5F1
© The Author(s) 2017
Received: 3 December 2016
Accepted: 8 February 2017
Published: 17 February 2017
Evidence is accumulating to link cancer stem cells to the pathogenesis and progression of osteosarcoma. The aim of this study is to investigate the role of miR-335 in osteosarcoma stem cells.
Tumor spheroid culture and flow cytometry were applied to screen out osteosarcoma stem cells. Real-time quantitative PCR was used to detect the expression level of miR-335 in MG63, U2OS and 143B osteosarcoma stem cells. The relationship of miR-335 expression with osteosarcoma stem cells was then analyzed. Transwell assay and transplantation assay were performed to elucidate biological effects of miR-335 on cell invasion and vivo tumor formation. Western Blot and luciferase assays were executed to investigate the regulation of POU5F1 by miR-335.
The expression of miR-335 in osteosarcoma stem cells was lower than their differentiated counterparts. Cells expressing miR-335 possessed decreased stem cell-like properties. Gain or loss of function assays were applied to find that miR-335 antagonist promoted stem cell-like properties as well as invasion. Luciferase report and transfection assay showed that POU5F1 was downregulated by miR-335. Pre-miR-335 resulted in tumor enhanced sensitivity to traditional chemotherapy, whereas anti-miR-335 promoted chemoresistance. Finally, the inhibitory effect of miR-335 on in vivo tumor formation showed that combination of pre-miR-335 with cisplatin further reduced the tumor size, and miR-335 brought down the sphere formation capacity induced by cisplatin.
The current study demonstrates that miR-335 negatively regulates osteosarcoma stem cell-like properties by targeting POU5F1, and miR-335 could target CSCs to synergize with traditional chemotherapeutic agents to overcome osteosarcoma.
KeywordsOsteosarcoma Cancer stem-like cells microRNA POU5F1 Targeted therapy
Osteosarcoma cells are highly aggressive and are widely recognized as the most common primary malignant bone tumor in adolescents . Traditional chemotherapy is associated with systematic toxicity and is less selective, and approximately 10–25% of patients respond poorly to the current chemotherapy . Therefore, effective targeted therapy of osteosarcoma is an urgent problem.
Cancer stem cells (CSCs) are defined as cells with a self-renewal ability that can generate heterogeneous cancer cells, which play a key role in recurrence, metastasis, and drug resistance . Several studies have also confirmed the existence of osteosarcoma stem cells. Gibbs first reported that stem-like cells exist in osteosarcoma, which are capable of forming self-renewable tumor spheres. Honoki reported that ALDH1 in MG63 sarcospheres was significantly higher and that these cells were involved in multidrug resistance . Adhikari et al. have identified mouse and human osteosarcoma stem cells using mesenchymal stem cell markers CD-117 and Stro-1. These markers were preferentially expressed in spheres and doxorubicin-resistant cells . The concept of CSCs provides a theoretical basis for specific targeted therapy. Therefore, studies of the regulation of osteosarcoma stem cells are especially important.
miRNAs are short non-coding RNAs that regulate different aspects of post-transcriptional gene expression, miRNAs usually bind to the 3′-untranslated region (3′-UTR) of targeted mRNAs, to cause translational degradation or target repression and gene silencing . Recently, increasing evidence has suggested that miR-335 plays a role in the regulation of cancer progression. For instance, miR-335 was reported to inhibit small cell lung cancer bone metastases via the IGF-IR and RANKL pathway . It has also been reported that miR-335 targets Bcl-w as an invasion suppressor gene in ovarian cancer . Heyn  reported that miR-335 is crucial for the BRCA1 regulatory cascade in breast cancer development. In osteosarcoma, it has been suggested that miR-335 suppresses tumors by regulating the ROCK1 gene . However, the relationship between miR-335 and osteosarcoma stem cells remains unclear.
In the present study, we showed that the levels of miR-335 were downregulated in osteosarcoma stem cells. Moreover, cells expressing miR-335 possessed decreased stem cell-like properties. These effects were partially caused by regulation of miR-335 targeting POU5F1. Furthermore, we demonstrated that miR-335 increased chemosensitivity and inhibited in vivo tumor formation by targeting osteosarcoma stem cells.
Cell lines and cell culture
MG63, U2OS and 143B cells were purchased from China Center for Type Culture Collection (CCTCC). HEK293 cells were obtained from Shanghai Institute for Biological Sciences of the Chinese Academy of Sciences. All the cells were cultured in RPMI-1640 containing 10% FBS and 1% penicillin/streptomycin. Cells were propagated at 37 °C with 5% CO2 and 100% humidity. Cell viability was determined using trypan blue staining.
Tumor spheroid culture
Cells were plated in six-well ultralow attachment plates (Corning) at a density of 5000 cells/well in RPMI-1640 supplemented with B27 Supplement (Invitrogen), 10 ng/mL human EGF (Sigma-Aldrich), and 10 ng/mL human bFGF (Sigma-Aldrich). Cells were incubated at 37 °C in a humidified atmosphere of 95% air and 5% CO2. Fresh aliquots of EGF and bFGF were added every other day. After culture for 2 weeks, colonies larger than 50 μm in size were regarded as sarcospheres and quantitated by inverted phase contrast microscopy.
For analysis of cell surface markers, osteosarcoma cells were harvested and resuspended in PBS/0.5% normal rabbit serum (Sigma-Aldrich), and blocked on ice for 15 min. Cells were subsequently labelled with Alexa Fluor® 647 anti-human Stro-1 antibody (BioLegend) and PE anti-human CD117 (c-kit) antibody (BioLegend) for 60 min and maintained on ice until analysis. The expression was assessed by flow cytometry and data were analyzed. The double positive (DP) and double negative (DN) cells were then sorted and collected using a Becton–Dickinson FACSort (San Jose).
In order to divide the cells into different subpopulations according to their miR-335 expression status, we incubated the cells with miR-335-5p Hu-Cy5 SmartFlare™ RNA Detection Probe (Millipore), overnight for 16 h. The fluorescence was then detected and cells were sorted into miR-335-high and miR-335-low subpopulation using a Becton–Dickinson FACSort (San Jose). The collected cells were then send for subsequent experiments.
Basing upon the pMIR-REPORT vector the luciferase reporter was constructed (Ambion). Strands of the oligonucleotides of the POU5F1-3′-UTR containing the miR-335 binding site were subcloned into the pMIR-REPORT vector after being synthesized and annealed. Scrambled sequences were also subcloned into the same vector acting as negative control. The POU5F1-3′-UTR-miR-335 or control plasmid was transfected into the HEK293 cells using Lipofectamine 2000 (Invitrogen). Luminescence were assayed using the luminometer.
Anti-miR-335 and pre-miR-335 used in this study were synthesized and confirmed by Shanghai Sangon Biotech Co. Ltd. Scrambled RNA (100 nmol) was act as negative control. We transfected the pre-miR-335, anti-miR-335 and their negative controls into all three cell lines in 6-well plates (1 × 105 cell per well) with 5 μL siPORT NeoFX transfection agent (Ambion) according to the manufacturer’s instructions. The cells were harvested for further experiments at the indicated time.
To test the cell invasion capability, we precoated 8 micron transwells with 10 μg/cm2 matrigel (BD Bioscience), then added 3 × 106 cells into the upper chamber with serum-free medium. The medium containing 10% FBS was put into the lower chamber as chemo-attractant. Cells were allowed to invade for 24 h incubation. Remaining cells which did not invade through the pores were carefully wiped out. Matrigel membranes were fixed in 90% methanol and then stained with crystal violet solution. Five random fields were counted by an inverted microscope (Olympus) examination. Each experiment was repeated at least three times.
Total RNAs including mRNAs and small RNA fraction were isolated from cells using TRIZOL reagent (Invitrogen), and reverse transcriptions were performed by Takara RNA PCR kit (Takara) according to the manufacturer’s instructions. Real-time PCR was performed using a SYBR Green detection system which the U6 small nuclear RNA and β-actin mRNA were used as internal controls. All the reactions were repeated in three times. Forward and reverse primers for miR-335 and U6 were 5′-TCAAGAGCAATAACGAAAAATGT-3′, 5′-GCTGTCAACGATACGCTACGT-3′; and 5′-CGCTTCGGCAGCACATATAC-3′, 5′-TTCACGAATTTGCGTGTCAT-3′; respectively. The primers for POU5F1 and β-actin mRNA were 5′-GAGTGAGAGGCAACCTGGAGAAT-3′, 5′-ACCGAGGAGTACAGTGCAGTGAA-3′; and 5′-GTCCACCGCAAATGCTTCTA-3′, 5′-TGCTGTCACCTTCACCG TTC-3′, respectively.
Western Blot analysis
Total proteins from transfected cells were obtained according to the manufacturer’s instructions (Sigma). We resolved the proteins by 10% SDS-PAGE gel and transferred them onto the PVDF membrane (Millipore). After being blocked in 10% nonfat dried milk for 2 h, the blots were incubated with anti-POU5F1 (1:1000; TA324014, OriGene) or anti-Sox2 (1:700; TA321559, OriGene) at 4 °C overnight. After washing three times with TBST, the blots were incubated with secondary antibody at room temperature for 1 h. After washing three times with TBST, membranes were developed by using enhanced chemiluminescence. β-actin served as an endogenous control.
Cell cytotoxicity assay
The cells in each well containing 100 μL medium were incubated with 10 μL cell counting kit-8 (CCK-8) at 37 °C for 2 h. The optical density (OD) of each well was then measured at 450 nm using a microplate reader.
Animals and transplantation assay
To determine the in vivo tumorigenicity, BALB/C nude mice about 6 weeks old were purchased and maintained at the Center for Animal Experiment at Renmin Hospital of Wuhan University. Osteosarcoma cells (1 × 106) transfected with the pre-miR-335 or vehicle construct were counted by trypan blue staining, and suspended in 10 μL of 50% Matrigel/PBS. The mice were randomly divided into four groups (Vehicle, Cisplatin, pre-miR-335 and cisplatin + pre-miR-335 group), with six mice in each group. For cisplatin treatment, the mice received subcutaneous injection of 5 mg/kg cisplatin per week for 6 weeks. Due to the transient effect of pre-miR-335 transfection, the mice received intratumoral injection with 4 μg of lipofectamine 2000-encapsulated pre-miR-335 every week for the last 4 weeks. Tumors were monitored as long as 6 weeks. Data were accumulated from at least three independent experiments.
All data were expressed as the mean ± standard deviation (SD). Statistical analysis were made between two groups with the t test. One-way ANOVA followed by multiple comparisons among the means was used where more than three means were compared. P < 0.05 was considered as statistically significant.
Expression of miR-335 is downregulated in osteosarcoma stem cells
Cells expressing miR-335 showed decreased stem cell-like properties
Based on the aforementioned results, we separated three osteosarcoma cell lines into miR-335-high and miR-335-low in order to compare their stem cell-like properties (Fig. 1d). CD117 and Stro-1 have been linked to cancer stem cells in osteosarcoma, so we determined whether miR-335-low cells were enriched with cells expressing these markers. The CD117 and Stro-1 double positive rate for MG63, U2OS and 143B miR-335-low cells were significantly higher than the miR-335-high cells (Fig. 1e, f). CSCs possess a transcriptional profile similar to that of embryonic stem cells . The expression of the genes POU5F1 and Sox2, were significantly downregulated in miR-335-high cells compared with miR-335-low cells (Fig. 1g). Spheroid formation has been widely used to assess the in vitro self-renewal potential of stem cell-like cells, so we compared the ability of miR-335-low and miR-335-high cells to form sarcospheres. The miR-335-low cells formed more sarcospheres than miR-335-high cells in all three osteosarcoma cell lines (Fig. 1h). These results suggested that miR-335 negatively correlated with osteosarcoma stem cell-like properties.
Downregulation of miR-335 enhanced osteosarcoma stem cell-like characteristics
miR-335 negatively regulated POU5F1 gene expression
miR-335 decreased the in vivo tumor formation by targeting CSCs
Accumulating evidence has indicated that miRNAs act as key regulators of cell proliferation, differentiation and apoptosis . MicroRNAs have been reported to play roles in the maintenance and regulation of normal and malignant stem cells . Our results showed that miR-335 negatively regulated osteosarcoma stem cell-like properties. First, we showed that the expression of miR-335 in osteosarcoma stem cells was lower than their differentiated counterparts. Second, cells expressing miR-335 possessed decreased stem cell-like properties compared with miR-335 low cells. Third, the miR-335 antagonist promoted stem cell-like properties, whereas miR-335 expression showed the opposite effect. Fourth, we showed that pre-miR-335 treatment reduced tumor cell invasion and resistance, whereas anti-miR-335 promoted tumor cell invasion and resistance. Fifth, we showed that miR-335 decreased the stem cell-like properties induced by cisplatin as assessed by the sphere formation efficiency.
Osteosarcoma is thought to derive from mesenchymal stem cells (MSCs) . miR-335 has been reported to induce cell proliferation, migration and differentiation in human MSCs . Previous studies have also reported that miR-335 is downregulated in osteosarcomas [10, 17]. However, the status of miR-335 in osteosarcoma stem cells remains to be elucidated. miR-335 was shown to be downregulated in breast cancer stem cells and to inhibit CSC growth . In the present study, three different methods were used to isolate osteosarcoma stem cells, and the expression of miR-335 was obviously down-regulated in osteosarcoma stem cells. Additionally, functional experiments showed that downregulation of miR-335 promoted osteosarcoma cells stem cell-like properties. These results were consistent with those of Schoeftner’s study, which showed that miR-335 controlled the Oct4-pRb axis and integrated stem cell self-renewal and cell cycle control .
Previous studies have reported that POU5F1 plays an extremely important role in maintaining the cancer stem cell fate of osteosarcoma . Characterizing mechanism of POU5F1 regulation is therefore of vital interest. In the present study, we showed that POU5F1 gene expression was negatively regulated by miR-335. The Target Scan software  predicted that the POU5F1 3′-UTR was incompletely bound to miR-335, which means that it could repress POU5F1 expression by post-translational regulation. Consistent with this possibility, when we fused the miRNAs regulatory elements (MREs) from the POU5F1 3′-UTR to luciferase, the luciferase reporter analysis showed that exogenous pre-miR-335 and anti-miR-335 regulated the activity, and that suppression occurred in a dose-dependent manner. We further observed that POU5F1 was negatively regulated by miR-335 merely at the protein level through gain- or loss- of function studies. These results further confirmed that miR-335 repressed POU5F1 expression via imperfect complementation to the POU5F1 mRNA 3′-UTR.
CSCs possess intrinsic chemoresistance. The underlying mechanisms include higher levels of the ABC transporter and anti-apoptosis genes, efficient DNA repair, and detoxification by aldehyde dehydrogenase. In addition, CSCs express EMT markers suggesting that they are able to migrate [22, 23]. Consistent with these results, our studies showed that miR-335 significantly decreased drug resistance compared with the anti-miR-335 group and with the control group. Accordingly, the present study showed that miR-335 significantly inhibited the invasive potential of metastasizing cells.
CSC-targeted therapy is a promising procedure to treat tumor cells. We therefore investigated whether miR-335 could selectively inhibit CSCs. Cisplatin alone increased the capacity to form tumor spheres in a osteosarcoma xenograft model, which was consistent with our previous results that conventional chemotherapy enriched for CSCs . However, miR-335 worked together with cisplatin in shrinking the tumor size, as well as decreasing the sphere formation capacity induced by cisplatin. These results suggested that miR335 could be an effective agent to eradicate CSCs.
Nonetheless, there remain some limitations in the current study. Firstly, we did not provide clinical relevance in our study due to the scarcity of osteosarcoma stem cell in human samples. Testing the inhibitory effect of pre-miR-335 on patient-derived xenograft could thus provide more clinical importance. Secondly, it is of great interest to investigate the role of miR-335 in other malignant bone tumor such as Ewing sarcoma and chondrosarcoma. Thirdly, there are plenty of genes targeted by miR-335, it is unknown whether genes except POU5F1 also participate in regulating osteosarcoma stem cells. Finally, we used oligoribonucleotides transfection to regulate miR-335 expression, however, the effect diminished gradually, pre- or anti-miR-335 inducible stable cell lines could provide stronger evidence and further support to our findings.
Taken together, our results for the first time showed the relationship between miR-335 and POU5F1 in osteosarcoma stem cells. Aside from the above limitations, these findings will characterize the underlying mechanisms of osteosarcoma and may facilitate the development of novel therapeutic strategies for clinical application.
cancer stem cells
POU domain, class 5, transcription factor
quantitative real-time polymerase chain reaction
Conceived and designed the experiments: WCG, XDG, LY and ZPZ. Performed the experiments: XDG, LY, ZPZ and GD. Analyzed the data: XDG, LY, ZPZ and TG. Contributed reagents/materials/analysis tools: XDG, LY and ZPZ. Reviewed all data and wrote the paper: WCG, XDG and LY. All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Availability of data and materials
The datasets supporting the conclusions of this article are included within the article and its additional files.
Ethics approval and consent to participate
The care and use of animals has been reviewed and approved by the Institutional Animal Care and Use Committee (IACUC).
The study was supported in part by Grants from the Natural Science Foundation of China (81502575), and Fundamental Research Funds for the Central Universities (2042015kf0069).
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Isakoff MS, Bielack SS, Meltzer P, Gorlick R. Osteosarcoma: current treatment and a collaborative pathway to success. J Clin Oncol. 2015;33(27):3029–35. doi:10.1200/JCO.2014.59.4895.View ArticlePubMedPubMed CentralGoogle Scholar
- Allison DC, Carney SC, Ahlmann ER, Hendifar A, Chawla S, Fedenko A, Angeles C, Menendez LR. A meta-analysis of osteosarcoma outcomes in the modern medical era. Sarcoma. 2012;2012:704872. doi:10.1155/2012/704872.View ArticlePubMedPubMed CentralGoogle Scholar
- Kreso A, Dick JE. Evolution of the cancer stem cell model. Cell Stem Cell. 2014;14(3):275–91. doi:10.1016/j.stem.2014.02.006.View ArticlePubMedGoogle Scholar
- Gibbs CP, Kukekov VG, Reith JD, Tchigrinova O, Suslov ON, Scott EW, Ghivizzani SC, Ignatova TN, Steindler DA. Stem-like cells in bone sarcomas: implications for tumorigenesis. Neoplasia. 2005;7(11):967–76.View ArticlePubMedPubMed CentralGoogle Scholar
- Adhikari AS, Agarwal N, Wood BM, Porretta C, Ruiz B, Pochampally RR, Iwakuma T. CD117 and Stro-1 identify osteosarcoma tumor-initiating cells associated with metastasis and drug resistance. Cancer Res. 2010;70(11):4602–12. doi:10.1158/0008-5472.CAN-09-3463.View ArticlePubMedPubMed CentralGoogle Scholar
- Dong H, Lei J, Ding L, Wen Y, Ju H, Zhang X. MicroRNA: function, detection, and bioanalysis. Chem Rev. 2013;113(8):6207–33. doi:10.1021/cr300362f.View ArticlePubMedGoogle Scholar
- Gong M, Ma J, Guillemette R, Zhou M, Yang Y, Hock JM, Yu X. miR-335 inhibits small cell lung cancer bone metastases via IGF-IR and RANKL pathways. Mol Cancer Res. 2014;12(1):101–10. doi:10.1158/1541-7786.MCR-13-0136.View ArticlePubMedGoogle Scholar
- Cao J, Cai J, Huang D, Han Q, Yang Q, Li T, Ding H, Wang Z. miR-335 represents an invasion suppressor gene in ovarian cancer by targeting Bcl-w. Oncol Rep. 2013;30(2):701–6. doi:10.3892/or.2013.2482.PubMedGoogle Scholar
- Heyn H, Engelmann M, Schreek S, Ahrens P, Lehmann U, Kreipe H, Schlegelberger B, Beger C. MicroRNA miR-335 is crucial for the BRCA1 regulatory cascade in breast cancer development. Int J Cancer. 2011;129(12):2797–806. doi:10.1002/ijc.25962.View ArticlePubMedGoogle Scholar
- Wang Y, Zhao W, Fu Q. miR-335 suppresses migration and invasion by targeting ROCK1 in osteosarcoma cells. Mol Cell Biochem. 2013;384(1–2):105–11. doi:10.1007/s11010-013-1786-4.View ArticlePubMedGoogle Scholar
- Yu L, Fan Z, Fang S, Yang J, Gao T, Simoes BM, Eyre R, Guo W, Clarke RB. Cisplatin selects for stem-like cells in osteosarcoma by activating Notch signaling. Oncotarget. 2016;7(22):33055–68. doi:10.18632/oncotarget.8849.PubMedPubMed CentralGoogle Scholar
- Yu L, Liu S, Zhang C, Zhang B, Simoes BM, Eyre R, Liang Y, Yan H, Wu Z, Guo W, Clarke RB. Enrichment of human osteosarcoma stem cells based on hTERT transcriptional activity. Oncotarget. 2013;4(12):2326–38. doi:10.18632/oncotarget.1554.View ArticlePubMedPubMed CentralGoogle Scholar
- Gammell P. MicroRNAs: recently discovered key regulators of proliferation and apoptosis in animal cells: identification of miRNAs regulating growth and survival. Cytotechnology. 2007;53(1–3):55–63. doi:10.1007/s10616-007-9049-4.View ArticlePubMedPubMed CentralGoogle Scholar
- Roden C, Lu J. MicroRNAs in control of stem cells in normal and malignant hematopoiesis. Curr Stem Cell Rep. 2016;2(3):183–96. doi:10.1007/s40778-016-0057-1.View ArticlePubMedGoogle Scholar
- Walkley CR, Qudsi R, Sankaran VG, Perry JA, Gostissa M, Roth SI, Rodda SJ, Snay E, Dunning P, Fahey FH, Alt FW, McMahon AP, Orkin SH. Conditional mouse osteosarcoma, dependent on p53 loss and potentiated by loss of Rb, mimics the human disease. Genes Dev. 2008;22(12):1662–76. doi:10.1101/gad.1656808.View ArticlePubMedPubMed CentralGoogle Scholar
- Tornero-Esteban P, Rodriguez-Rodriguez L, Abasolo L, Tome M, Lopez-Romero P, Herranz E, Gonzalez MA, Marco F, Moro E, Fernandez-Gutierrez B, Lamas JR. Signature of microRNA expression during osteogenic differentiation of bone marrow MSCs reveals a putative role of miR-335-5p in osteoarthritis. BMC Musculoskelet Disord. 2015;16:182. doi:10.1186/s12891-015-0652-9.View ArticlePubMedPubMed CentralGoogle Scholar
- Liu ZF, Liang ZQ, Li L, Zhou YB, Wang ZB, Gu WF, Tu LY, Zhao J. MiR-335 functions as a tumor suppressor and regulates survivin expression in osteosarcoma. Eur Rev Med Pharmacol Sci. 2016;20(7):1251–7.PubMedGoogle Scholar
- Polytarchou C, Iliopoulos D, Struhl K. An integrated transcriptional regulatory circuit that reinforces the breast cancer stem cell state. Proc Natl Acad Sci USA. 2012;109(36):14470–5. doi:10.1073/pnas.1212811109.View ArticlePubMedPubMed CentralGoogle Scholar
- Schoeftner S, Scarola M, Comisso E, Schneider C, Benetti R. An Oct4-pRb axis, controlled by MiR-335, integrates stem cell self-renewal and cell cycle control. Stem Cells. 2013;31(4):717–28. doi:10.1002/stem.1315.View ArticlePubMedGoogle Scholar
- Siclari VA, Qin L. Targeting the osteosarcoma cancer stem cell. J Orthop Surg Res. 2010;5:78. doi:10.1186/1749-799X-5-78.View ArticlePubMedPubMed CentralGoogle Scholar
- Fromm B, Billipp T, Peck LE, Johansen M, Tarver JE, King BL, Newcomb JM, Sempere LF, Flatmark K, Hovig E, Peterson KJ. A uniform system for the annotation of vertebrate microRNA genes and the evolution of the human microRNAome. Annu Rev Genet. 2015;49:213–42. doi:10.1146/annurev-genet-120213-092023.View ArticlePubMedPubMed CentralGoogle Scholar
- Zhao J. Cancer stem cells and chemoresistance: the smartest survives the raid. Pharmacol Ther. 2016;160:145–58. doi:10.1016/j.pharmthera.2016.02.008.View ArticlePubMedGoogle Scholar
- Chang JC. Cancer stem cells: role in tumor growth, recurrence, metastasis, and treatment resistance. Medicine. 2016;95(1 Suppl 1):S20–5. doi:10.1097/MD.0000000000004766.View ArticlePubMedGoogle Scholar