MicriRNA Expression Profile Of Gastric Cancer Stem Cells In MKN45Cancer Cell Line

Part1Suspension Culture System as a model of isolation of gastriccancer stem cell in gastric cancer line MKN-45Objective: To isolate cancer stem cell in gastric cancer line MKN-45under suspensionculture system, and establish a cell model that exhibits the characteristics of CSCs for theresearch in ganstric cancer stem cell.Methods: The human gastric cancer cell line MKN-45obtained from the Cell Bank of theChinese Academy of Sciences, Shanghai, China was cultured in RPMI-1640mediumcontaining10%fetal bovine serum （FBS） and plated at the density of1x106live cells per75-cm2flask. When the cells attached, we passaged them upon confluence. Spheroidbodies were derived by placing the parental cells into serum-free RPMI-1640culturemedium containing1%N-2supplement,2%B-27supplement （Invitrogen）,1%antibioticmixture （Gibco）,20ng/ml human FGF-2and100ng/ml EGF （Chemicon）. The parentalcells were plated in96-well ultra-low attachment plate （Corning） at100cells per well.Two weeks later, plates were analyzed for spheroid body formation and were quantifiedusing an inverted microscope （Olympus） at x40and x100magnification. After primaryspheroid body reached the size of approximately200-500cells per spheroid body, thespheroid bodies were dissociated at the density of1,000cells per milliliter and100singlecell suspension （100μl） was seeded in each well of a96-well ultra-low attachment plate（Corning） in serum-free medium described above. Two weeks later, wells were analyzedfor subspheroid body formation. The subspheroid body cells were cultured in vitro andpassaged more than six generations. Results: MKN-45parental cells were cultured in serum-free medium described in themethods section. In this condition, cells grew as non-adherent, three-dimensional spheroidclusters, called spheroid body. The self-renewing capacity of these spheroid body-formingcells was assessed by dissociating them into single cell and growing in serum-free mediumdescribed in the methods section. At day3, the spheroid bodies appeared to be taking shape.At day7, the spheroid bodies were formed substantially. At day14, the spheroid bodieswere completely formed. At day21, the spheroid bodies had become well-roundedstructures composed of numerous, compacted cells. With the increase of passage, thespheroid body cells showed an increased ability to form sub-spheroid body. At passagesixth, these cells derived from spheroid body cells generated sub-spheroid bodies again at29.70±6.21%compared with3.30±1.49%of parental cells （p<0.01）.Conclusion: Spheroid body-forming cells from gastric cancer cell line MKN45undersuspension culture system possess the capablity of self-renewal and proliferation which isan important characteristic of CSCs.Part2Identification of cancer stem cell in human gastric cancer lineMKN-45Objective:To identify the characteristics of spheroid body-forming cells from gastriccancer cell line MKN45under suspension culture system， and establish a practical andcost-effective model that exhibits the characteristics of CSCs for the research in gastriccancer stem cell.Methods:①The expression levels of stemness genes（Oct4, Sox2, Nanog and CD44） weredetermined by Quantitative real-time PCR;②The expression levels of stemness genesrelated protein（Oct4, Sox2, Nanog and CD44） were determined by Western blot analysisand Immunofluorescence staining;③The rates of resistance to drugs were assessed usingMTT assay.2,000healthy spheroid body-forming cells or parental cells per well were plated in96-well plates in100μl RPMI-1640medium （4wells per group） withchemotherapeutic drugs （5-Fu, DDP） or control PBS. At each time point （24and48h）,10ml MTT solution was added to each well and the plate was incubated for4h at37C, thenthe medium was replaced by150ml DMSO. To assess the effect of drug resistance ofspheroid body-forming cells or parental cells, we treated the dissociated cells with5-Fu （50μg/ml,100μg/ml,200μg/ml） alone, DDP （10μg/ml,20μg/ml,40μg/ml） alone,5-Fu plusDDP （50μg/ml+10μg/ml;100μg/m+20μg/ml;200μg/ml+40μg/ml） or control PBSfor24and48h. MTT assay is based on mitochondrial conversion of MTT to yellowishformazan, being indicative of the number of viable cells. The number of viable cells wasevaluated by absorbance OD450nm （Abs） using Model680microplate reader.④In vivotumorigenicity experiments, male athymic nude mice （nu/nu）,6to8weeks old, werehoused under pathogen-free conditions in the barrier animal facility. Equal number（1x104,2x104,2x105,2x106） of freshly dissociated cells was suspended in200μl PBS, thespheroid body-forming cells were injected subcutaneously into the right rear flank of eachmouse （6mice per group） and the parental cells were injected subcutaneously into the leftrear flank of each mouse, we examined the tumorigenic capacity of spheroid body-formingcells and parental cells. The mice were observed for tumor growth every7-10days over8weeks and then sacrificed by cervical dislocation. The grafts were removed, fixed with10%buffered formalin, and stained with hematoxylin and eosin （H&E）.Results:①Quantitative real-time PCR and western blot analysis were performed onspheroid body-forming cells and parental cells. The expression levels of Oct-4, Sox2,Nanog and CD44in spheroid body-forming cells were2.92±0.44、4.49±0.20、2.34±0.22、1.18±0.04. respectively, while The expression levels of Oct-4, Sox2, Nanog and CD44inparental cells were1.02±0,06、1.00±0.06、1.00±0.00、1.00±0.05，respectively. The resultsshowed that the cells expressing Oct-4, Sox2, Nanog and CD44were significantly more inspheroid body-forming cells than in parental cells（P＜0.05）.②To examine the subcellularlocalization of Oct-4, Sox2, Nanog and CD44in spheroid body-forming cells,immunofluorescent staining of Oct-4, Sox2, Nanog and CD44was performed. Oct-4, Sox2 and Nanog proteins were positively stained within the perinuclear and cytoplasm of thespheroid body-forming cells, and CD44was positively stained mainly in the membrane.Dual staining of Oct-4/CD44, Sox2/CD44, and Nanog/CD44indicated that the embryonalproteins （Oct-4, Sox2and Nanog） were colocalized with CD44in the spheroidbody-forming cells.③The MKN-45spheroid body-forming cells exhibited generalresistance to5-Fu and DDP in the treatment of24h. Compared with the parental MKN-45cells, the survival rates of spheroid body-forming cells were higher under the treatment of50μg/ml,100μg/ml and200μg/ml5-Fu （1.3-fold,1.4-fold and1.7-fold, respectively）;10μg/ml,20μg/ml and40μg/ml DDP （1.7-fold,3.2-fold and3.4-fold, respectively）. Whereas,under the treatment for48h, the relative survival rates were not significantly increased,neither5-Fu nor DDP. But in the treatment of5-Fu combined with DDP, the survival ratesof MKN-45spheroid body-forming cells were increased2.6-fold,2.9-fold and2.7-fold,respectively, under the treatment for24h; and5.1-fold,4.8-fold and5.4-fold, respectively,for48h.④The tumorigenicity experiments in vivo showed that as few as2x104cells fromthe MKN-45spheroid body were able to form a tumor when subcutaneously injected intonude mice, while2x106parental cells were needed. This was100times higher than that ofspheroid body-forming cells. Moreover, spheroid body-forming cells generatedsubcutaneous tumors with larger volume compared with those generated from parentalcells. H&E examination of xenografts derived from MKN-45spheroid body-forming cellsshowed that these tumors closely resembled those from the parental cells, mainlycontaining differentiated cells.Conclusions: Spheroid body-forming cells from the gastric cancer cell line MKN-45cultured in stem cell-conditioned medium possessed gastric CSC properties, such aspersistent self-renewal, extensive proliferation, drug resistance, high tumorigenic capacityand overexpression of stemness gene and related proteins （Oct4, Sox2, Nanog and CD44）,compared with the parental cells. Suspension Culture is a practical and cost-effectivemodel for the research in ganstric cancer stem cell. More importantly, Oct4/CD44、Sox2/CD44、 Nanog/CD44maybe a type of gastric CSCs,respectively. Part3MicroRNA expression profile of gastric cancer stem cells in theKN-45cancer cell lineObjective: This study aimed to determine the miRNA profile of gastric cancer stem cellsin the MKN-45cancer cell line and to explore the role of microRNAs in gastric CSCsMethods:①miRNA extraction and reverse transcription: Total RNA, including themiRNAs, was extracted from the spheroid body-forming cells and the MKN-45parentalcells, and purified by using the miRNeasy mini kit （Qiagen, Hilden, Germany） accordingto the manufacturer’s instructions. The RNA concentration was quantified by using aNanoDrop2000spectrophotometer （Thermo Fisher Scientific, Waltham, USA）. Theintegrity and the quality of RNA were evaluated by Agilent2100Bioanalyzer （AgilentTechnologies, Santa Clara, USA）. RNAs with a2100RIN （RNA integrity number）₆.0and28S/18S₀.7were used for the miRNA array analysis and reverse transcription （RT）.An aliquot of1.0mg of total RNA was converted to cDNA by using the miScript II RT kit（Qiagen）.②miRNA array assay: Qualified RNA samples were further detected by themiRNA array at Shanghai Biochip, Co., Ltd.（Shanghai, China）, by using Agilent’s humanmiRNA microarray, version18.0, which contains1887human and89human viral miRNAsequences from the Sanger v18.0miRNA database. Labeling and hybridization wereperformed according to the manufacturer’s protocols. Signals on the slides were scannedby using an Agilent G2565BA microarray scanner.③Quantitative real-time polymerasechain reaction analysis: Quantitative real-time polymerase chain reaction （qRT-PCR） wasperformed by using the miScript SYBR Green PCR kit （Qiagen） according to themanufacturer’s instructions on an LC480II RT-PCR platform （Roche, Indianapolis, USA）.A universal reverse primer was provided by the manufacturer, and each forward primerwas specific to the mature miRNA listed in the Sanger miRNA database. The expression ofthe miRNAs was normalized by using U6small RNA as the endogenous control. Theanalyzed miRNAs included: miR-181a-5p, miR-27a-3p, miR-34a-5p, miR-21-5p, miR-483-5p, miR-22-3p, miR-4270, miR-16-5p, let-7b-3p, and miR-29a-3p. The experiments Three online software programs, miRanda http://microrna. sanger.ac.uk, TargetScanhttp://www. Target scan.org, PicTar http://www.ncrna.org/KnowledgeBase/link-Database/mirna_target_database, and were used for bioinformatics analysis and targetprediction for the10analyzed miRNAs above.Results:①We used an miRNA microarray to evaluate the miRNA expression profiles inthe spheroid body-forming cells and the parental cells. The miRNA expression pattern wasfound to be significantly different. A total of182miRNAs with a more than2-fold changewere differentially expressed between the spheroid body-forming cells and the parentalcells. Of these miRNAs, nine were over-expressed in the spheroid body-forming cells,while the others （173） were all under-expressed.②We performed qRT-PCR for10miRNAs: miR-181a-5p, miR-27a-3p, miR-34a-5p, miR-21-5p, miR-483-5p, miR-22-3p,miR-4270, miR-16-5p, let-7b-3p, and miR-29a-3p. The qRT-PCR results for the10testedmiRNAs were0.49±0.05,0.81±0.14,2.57±0.07,0.74±0.05,2.63±0.31,0.68±0.28,1.61±0.08,1.64±0.14,1.18±0.07, and0.44±0.03（expressed as the relative ratio betweenthe spheroid body-forming cells and the parental cells±standard deviation）. Seven of the10tested miRNAs gave qRT-PCR results that were concordant with the microarray data, withmiR-16-5p, let-7b-3p, and miR-34a-5p being the exceptions, indicating a concordance rateof70.0%.③Target prediction of the miRNAs was carried out by using three onlinesoftware programs, miRanda, PicTar, and TargetScan. To increase the specificity （at thecost of lower sensitivity）, we integrated the results of three target prediction programs andexamined only the intersection. We found that384miRNA: mRNA target pairs werepredicted by all of the three programs. Among the10analyzed miRNAs, miR-181a-5p,miR-27a-3p, miR-34a-5p, miR-21-5p, miR-22-3p, miR-4270, let-7b-3p, and miR-29a-3phave predicted targets, while there were no predicted targets for miR-483-5p andmiR-16-5p. We ranked the miRNA–target gene interactions and generated a network byusing CytoScapewRto visualize the predicted miRNA–mRNA interactions. There weremore than ten to hundreds of predicted target genes for each miRNA. Several genes were identified as the potential targets of two or more of the miRNAs. Most of these targetswere relevant to the regulation of actin cytoskeleton, focal adhesion, extracellularmatrix–receptor interaction, and the pathways in cancer. Especially, several genes wereassociated with some pivotal signaling pathways of the ‘stem cell genes’, such as Notch,Wnt/b-catenin, mitogen-activated protein kinase （MAPK）, mammalian target of rapamycin（mTOR）, and Janus kinase/signal transducer and activator of transcription （JAK-STAT）,etc.Conclusions: We identified a profile of miRNAs expressed differentially in gastric CSCs,which provided a starting point to explore the functions of these miRNAs. Evaluatingcharacteristic miRNAs of gastric CSC may be a new method for studying gastric cancerand developing therapeutic strategies aimed at eradicating the subpopulation of cancerstem cells in gastric cancer.