Isolation And Identification Of Cancer Stem Cells In Brain Gliomas And Analyses Of Biological Characteristics

[BACKGROUND]Gliomas, especially high grade glioblastomas (WHO grade IV), are characterized with highly malignancy, rapid growth, poor prognosis, resistance to radiotherapy and chemotherapy, and early relapse with a 5-year-survival rate of just 20%. Moreover, the tumor cells are highly heterogeneous and diversified in terms of their genetic backgrounds between individual cases. Therefore, the mechanism underlying glioma initiation, cellular regulation and growth characteristics are still remain poorly understood.The cancer stem cell hypothesis implies that only a relatively small fraction of cells (1%), termed cancer stem cell (CSC), possess the stem cell ability such as indefinite proliferation, extensive self-renewal, and multi-linkage differentiation, and they are the true sources of tumor initiation, metastasis and relapse. Bonnet et al are the first who have isolated the leukemia CSC from acute myeloid leukaemia and the researches on brain CSC are pioneered by Ignatova et al, Singh et al and Galli et al who have identified brain CSC from human medulloblastoma and gliomas of different grades. These studies have provide insights into our understanding of glioma cancer biology and paved the way for the future CSC-targeted therapy.The most widely adopted brain CSC identification methods are floating tumor spheres culture, CD133 magnetic activated cell sorting (MACS) and flow cytometry sorting (FCMS) for side population (SP). However, the differences in the specimens analyzed and specific identification methods applied usually lead to large controversy over the percentage and biological characteristics of the brain CSC isolated. Kondo et al and Li et al have identified a rare cell faction(0.4%, 1%) with CSC properties from long-term in vitro cultured rat C6 glioma cell line. Moreover, Huang et al also have proved the existence of cancer stem-like cells in SHG-44 glioma cell line. However, there are still some researches believe that there is no cancer stem cell in long-term in vitro cultured cell lines and they are dramatically different from their matched primary tumor in terms of genetic backgrounds. Our preliminary studies also showed that rat C6 cells are different from human primary glioblastoma cells in many respects, such as tumor sphere formation and tumorigenesis. Therefore, a systematic comparison was made between the identified CSCs from C6 glioma cell line and human glioblastomas.[OBJECTIVES]To identify the CSCs from primary human glioblastomas and rat C6 glioma cell line and explore their biological characteristics; Three different identification methods are used to identify the CSCs, comparing their composition and characteristics, and the optimal identification method is suggested; To determine the differences between the CSCs from primary human glioma and in vitro cultured glioma cell line and to further illuminate the role CSC plays in the initiation and development of glioma.[METHODS]Part 1: Both serum-containing and serum-free medium were used to culture the 5 human primary glioblastomas and the impact of alteration in culture medium on the growth pattern of glioblastoma cells was observed; FCMS was used to determine the percentage of CD133 positive cells in primary tumor cells and floating tumor spheres; Single-cell clonal culture was used to determine the percentage of cells with self-renewal capacity; Cytoimmunofluorescence and immunohistochemistry staining were used to analyze the multi-linkage differentiation capacity and co-expression of differentiations markers of the human glioblastoma cells. Xenograft tumor formation in nude mice was used to determine the in vivo tumorigenecity of the glioblastomacells.Part 2: Both serum-containing and serum-free medium were used to culture the rat C6 glioma cell line and the impact of alteration in culture medium on the growth pattern of C6 cells was observed; FCMS was used to determine the percentage of CD133 positive cells and differentiated progenies of the C6 cells in different culture medium; Consecutive single-cell clonal culture in serum-free medium was used to determine the percentage of cells with self-renewal capacity; Cytoimmunofluorescence and immunohistochemistry staining were used to analyze the multi-linkage differentiation capacity and co-expression of differentiations markers of the C6 cells. Xenograft tumor formation in nude mice was used to determine the in vivo tumorigenecity of the C6 cells.Part 3: CD133 MACS was used to isolate CD133 positive cell fractions from the CD133 negative cell fractions in both human primary glioblastoma and C6 line; Western Blot and immunofluorscence assay were used to determine the purity of the cell separated. Brdu staining was used to clear the proliferative capacity of the cells separated. Clonal analysis, immunofluorescence assay, and in vivo nude mice transplantation were used to determine the composition and biological characteristics of the CSCs in the separated cell fractions. Immunofluorescence assay was used to determine the toxicity of Hoechst 33342 to tumor cells and the expression of ABCG2.[RESULTS]一,Isolation and identification of cancer stem cells in human glioblastomas.1. About 5.8-13.9% human primary glioblastoma cells were capable of forming tumor spheres and 20-34.5% of the tumor sphere cells were endowed with self-renewal capacity, multi-linkage differentiation and in vivo tumorigenecity.2. Human primary glioblastoma cells could readily transform their growth pattern in serum-containing and serum-free medium and co-expression of CD133 and GFAP/NSE was found in few cells.3. FCMS and immunohistochemistry assay showed CD133 positive rate in primary glioblastoma cells were much lower than in the tumor spheres they formed in serum-free medium.4. Nude mice transplantation showed only human glioblastoma cells which could form tumor-spheres were tumorigenic and the xenograft tumor cells also contained CSCs.二,Isolation and identification of cancer stem cells in C6 glioma cell line5. C6 cells could transform its growth pattern in different culture medium and serum-containing medium did not alter its tumor-sphere formation capacity in serum-free medium.6. >60% C6 cells and their progenies were capable of forming tumor spheres in serum-free medium and >90% C6 cells and their progenies were capable of extensive self-renewal in serum-containing medium. All these cells had characteristics of neuronal and glial linkage differentiation and in vivo tumorigenecity.7. FCMS and immunohistochemistry assay showed a 30% CD133 positive rate in long-term serum-containing medium cultured C6 cells and >85% in serum-free meium.8. C6 cells co-expressed CD133 and NSE/GFAP in serum-containing medium, which is rarely seen in serum-free medium.9. Nude mice transplantation showed all C6 single cells were tumorigenic and the xenograft tumor cells had the similar CSC properties as their mother cells.三,Identification of cancer stem cells with different methods10. Both C6 CD133 positive and negative cell fractions contamed cancer stem-like cells with different proliferation, clone formation and tumorigenecity capacities.11. CD133 negative cell fraction in human glioblastomas contained no CSC and not all CD133 positive cells were capable of forming clones.12. CD133 negative cells in C6 cell line could generate CD133 positive cells in serum-containing medium but not the human glioblastoma cells.13. Hoechst 33342 was toxic to both C6 and human glioblastoma cells.14. C6 cell line and human glioblastoma cells contained 1% cells which were ABCG2 positive. Only this rare cell fraction still had CSC properties when they were incubated with Hoechst33342 for 2 hours.[CONCLUTIONS]1. Both human glioblastoma cells and C6 cells contained cancer stem-like cells. 2. Human glioblastoma cells differed from C6 cells in terms of their CSC percentage, self-renewal capacity, differentiation marker expression, transformation in serum-containing/free medium and in vivo tumorigenecity.3. Both C6 CD133 positive and negative fractions contained cancer stem-like cells while only CD133 positive human glioblastoma cells had self-renewal capacity, multi-linkage differentiation and in vivo tumorigenecity.4. Hochest 33342 was toxic to both human glioblastoma cells and C6 cells and thus exclude those cells that could not pump out Hochest 33342, for example ABCG2 negative cells, from being potential CSCs.5. Identifying CSCs by their biological characteristics yield a much higher CSC percentage than by CD133 MACS or SP FCMS. Therefore finding a more reliable CSC marker should be the aim of future work.