Side Population In Epithelial Ovarian Cancer And Its Role In Tumorigenesis

Side population in epithelial ovarian cancer and its role in tumorigenesisObjective: After FACs analysis and method modification, we sorted ovarian cancer cells-derived side population (SP) to identify its stem cell feature, including surface markers expression, capability of colony formation and its role in tumorigenesis. GFP-positive ovarian cancer cell lines were generated by lentiviral transduction. Tumorigenicity of GFP-SP cells were compared to that of non-transducing SP cells and fluorescence image of GFP-positive tumor in vivo was taken to testify whether GFP is able to be a marker to trace cancer cell mobility in vivo.Materials and Methods:1. Human ovarian cancer cell lines, ES-2, HO-8910PM, HO-8910, SK-OV-3 and CAOV-3, were used to perform SP analysis by Hoechst 33342 staining. Cells which were growing well were trypsinized to form single-cell suspension. After stained with Hoechst 33342, Hoechst 33342-negative cells were gated as SP cells. Cells pretreated with Verapamil were used as a control.2. SP cells and non-side population (NSP) cells were sorted respectively. Expression of cell surface markers on SP cells or NSP cells, including CD24, CD44, CD45, CD54, CD62L, CD117 and CD133, were determined by FACs analysis.3. RNA from SP cells and NSP cells was isolated by TRIzol homogenization. Expression of Nanog, Oct-4 and Breast Cancer Resistant Protein (BCRP) were analyzed by RT-PCR. The capability of colony formation and tumorigenesis of SP cells were compared to those of NSP cells by agarose colony formation and tumorigenicity in nude mice.4. In terms of capability of transmigration and invasion of SP cells and NSP cells, the sorted-cells were applied onto Transwell upper chamber in the presence or in the absence of Matrigel. After 12 hours,24 hours or 48 hours, the cells having been migrated to lower chamber were fixed and counted. Microarray gene profile assay was used to determine the gene expression related to cell attachment and migration.5. ES-2 cells and HO-8910PM cells were transduced by lentivirus at various MOI. Expression of transgene was clarified by FACs analysis. The best transducing MOI to different cell line under the control of Ubiquitin promoter was determined referring to GFP expression.6. GFP-ES2 and GFP-HO8910PM were collected by FACsorting and passaged under regular culture condition. GFP-SP cells were sorted after stained with Hoechst33342 and injected subcutaneously into nude mice. The local tumorigenicity of GFP-SP cells at injection site was compared to that of non-transduced cells. The fluorescence image of tumor-bearing mouse was taken under a small animal fluorescence imaging system.Results:1.There was a minor population in ovarian cancer cells which can pump out Hoechst33342 to result in Hoechst33342 staining-negative cells. Taking the difference in Hoechst 33342 staining between this population and major population, it was named as side population.2. The percentage of SP cells was related to gene expression of BCRP. SP cells expressed Nanog, Oct-4 and BCRP by analyses of RT-PCR, and had higher potential of colonigenicity in semisolid agarose gel.3. Sorted-SP cells grew in round-shape, cobblestone-like aggregates in regular cell culture plastic vessels. SP-derived aggregates were more compacted and formed less outgrowth than NSP-derived ones, which were scatter-distributed and spindle-shaped in plastic vessels.4. The differences between SP cells and NSP cells in attachment, transmigration and invasion were determined via Transwell chamber in the presence or in the absence of Matrigel. Cells migrated to the lower chamber were counted after fixation and stained by hematoxylin. The numbers of ES2-SP cells or HO8910PM-SP cells transmigrated through micropore membrane were less than those of NSP cells, either with or without Matrigel.5. Microarray of gene profile showed that expression of adhesion receptors/ligands, including integrin-β, ICAM-5, CAM-1, collagen, laminin and fibronectin, were down-regulated in HO8910PM-SP cells compared to HO8910PM-NSP cells, whereas protocadherin 7 and CXCR4 were up-regulated.6. SP cells can change to NSP phenotype under regular culture condition. However, during two weeks of cultivation, the percentage of Hoechst33342-negative cells in cultured SP population remained higher than that in cultured NSP one.7. Freshly sorted HO8910PM-SP cells were easier to be engrafted at subcutaneous injection site than NSP cells. However, ES2-SP cells showed higher local tumorigenicity only after cultured for a certain period of time when some of SP cells changed to NSP phenotype.8. Lentiviral transduction under the control of Ubiquitin promoter didn't effect on cells capability to pump Hoechst33342 out. GFP-SP cells from HO-8910PM could form GFP-positive tumor in nude mice. There was no difference in tumorigenecity between GFP-SP cells and non-transduced SP cells. Fluorescence in tumor cells at subcutaneous injection site can be detected under a small animal fluorescence imaging system.Conclusions: 1.There is a minor population of Hoechst33342-rejecting SP cells in ovarian cancer cells. SP cells had stem cell features with expression of multi-potential genes.2. Being compared to NSP cells, SP cells had higher tumorigenicity and less capability to attachment to and invasion into matrix. It indicated that, during the process of tumor formation, stem-like SP cells undergo self-renewal to accumulate at primary site. Meanwhile, some SP cells changed to NSP phenotype which represents high potential to infiltrate into surrounding tissues. Both fortunate tumorigenesis.3. Lentiviral transduction does not effect on capability of ovarian cancer SP cells-derived tumor formation. Lentiviral transduction can be a useful tool in the study of tumorigenesis and development of ovarian cancer. GFP protein can be a marker to trace cancer cells in vivo.