Cancer stem cells from glioblastoma multiforme: culturing and phenotype

Transcription

1 Page 1 of 6 Cancer Stem Cells Cancer stem cells from glioblastoma multiforme: culturing and phenotype DS Kyurkchiev* Abstract Introduction Glioblastoma multiforme is the most common and malignant brain tumour. The contemporary concept of its development is based on cancer stem cells which originate from the transformation of the normal neural stem cells present in the brain. Both neural stem cells and cancer stem cells express markers such as Nestin, Sox-2, CD133, CD44 and occasionally glial fibrillary acidic protein. Both cell types have similar signal transduction pathways and a number of properties as self-renewal, differentiation and proliferation capacity. Various approaches to culturing cells isolated from glioblastoma multiforme and the corresponding expression of markers for cancer stem cells are the object of this review. Cells isolated from glioblastoma can be cultured in serum-free media containing epidermal growth factor and fibroblast growth factor or in media with foetal bovine serum. It is possible to culture the cells in media combining these basic approaches. Discussion According to the concept that is predominant in the literature, neurospheres represent the best model of cancer stem cells when cells isolated from glioblastoma are cultured. Cells in neurospheres express Nestin, Sox-2, CD133, CD44; * Corresponding author dsk666@gmail.com Laboratory of clinical immunology, University Hospital St. Ivan Rilski Department of Clinical Laboratory and Clinical Immunology, Medical University of Sofia, 15 Acad. Ivan Geshov Street, 1431 Sofia,Bulgaria they have genetic alterations similar to the tumour and tumourigenicity. In contrast, adherent glioblastoma cells seem to lose the expression of Nestin and CD133 and start to express glial fibrillary acidic protein; they have weaker tumourigenicity and possess genetic changes different from those in the primary tumour. However, there are plenty of data which prove that this concept is not absolutely true. Semi-adherent cellular growth has been described with simultaneous expression of Nestin and glial fibrillary acidic protein, and quite often the adherent cells have a higher expression of markers for cancer stem cells in comparison with the neurospheres. Conclusion It can be stated that the protocol for glioblastoma multiforme cells cultured as neurospheres is not the only one and not necessarily the best model for cancer stem cells. Introduction Glioblastoma multiforme (GBM) is the most malignant and the most common primary brain tumour 1. Currently generally accepted is the hypothesis that cancer stem cells (CSCs) are the basic pathology factors leading to development of GBM and these cells are the results from the transformation of normal neural stem cells (NSCs) 2 6. Neural stem cells NSCs are described at any stage of the development from the embryo to the adult organism, and they are located in their specific niches in the subventricular zones, in the subgranular zone and in the dentate gyrus of the hippocampus In these niches NSCs are situated around the blood vessels. They are in communication with other cells and the extracellular matrix and the basic factors regulating their proliferation and destiny are epidermal growth factor (EGF) and basic fibroblast growth factor (bfgf) 11. The niches are characterised by a relative hypoxia which seems to be very important for the survival and proliferation of NSCs 12. Various cellular types are described in the niches, such as neuroblasts (type A), NSCs (type B) and transitory amplifying progenitors (type C) and all these cells are surrounded by ependymal cells 1,13. NSCs are pluripotent cells capable of self-renewal and differentiation as a result of which they lose their stemness 12. An obligatory condition for this is the presence of blood serum and the absence of EGF and bfgf which are the factors responsible for proliferation and sustaining the NSCs in undifferentiated state 2,7,9,14. The proliferative capacity of the NSCs and their association with blood vessels renders the cells highly mobile and makes possible their movement to zones with hypoxia 10,12. NSCs express specific phenotypic markers and the most important are the following: Nestin is cytoplasmic microfilament protein correlated with stemness and related to the organisation of the cytoskeleton, cellular signalling, organogenesis and cellular metabolism. In the course of differentiation the NSCs lose expression of Nestin and start to express βiii-tubulin and glial fibrillary acidic protein (GFAP) 5,7,8,12. GFAP has been described as a marker of astrocyte differentiation but still it can be co-expressed parallel to Nestin by NSCs 9,11.

2 Page 2 of 6 Sox-2 is transcription factor expressed by NSCs with cytoplasmic localisation which is connected to the process of differentiation 7,9. CD44 is a transmembrane molecule with a role in the adhesion between cells and the extracellular matrix which is expressed on the surface of NSCs 15,16. CD133 is a specific surface marker for NSCs 9,17 and its significance will be discussed in detail further. Cancer stem cells The CSCs hypothesis postulates that GBM formation is due to transformation of NSCs into CSCs 4,12. This might be due to mutations in the NSCs or to genetic changes in partially differentiated progenitor cells originating from the NSCs 7,14. It has been shown that partially differentiated cell can be transformed to CSCs after treatment with EGF and bfgf 18 and these changes can be observed with type B cells (NSCs), with type C cells and even with differentiated cells 9,10. Transformation of NSCs or progenitors in CSCs is considered to be related to normal cellular mechanisms but in abnormal order, at abnormal time and intensity 10. It is accepted that the regulatory Hedgehog systems which are related to genes for stemness and self-renewal are affected, as well as EGF signal pathway affecting the differentiation. Mutations in PTET (tumour suppressor gene) are also described. As a result, a maturation arrest is observed leading to intensive cell divisions and lack of differentiation 10. The resultant CSCs are described as a small subset of slowly dividing cells with cell renewal capacity, tumourigenicity, association with genes specific to NSCs, capability for migration and differentiation in cell populations similar to the primary tumour 2,4,5,19. CSCs are cells at the end of a process of dedifferentiation and they have common signal pathways with NSCs, vigorous proliferative potential regulated by EGF and bfgf and ability to differentiate as neurons and glial cells in the absence of these factors but in the presence of blood serum 2,7. The cellular composition of GBM is described as organised chaos because pleomorphic tumour cells, proliferating blood vessels, infiltrating inflammatory cells and necrosis are observed. The CSCs which are about 1% 3% of the cellular composition are in fact the organiser of the tumour 1,14. These cells seem to be at the top of the hierarchy of the tumour cells and they can by themselves form a GBM 20. Various types of CSCs are described in reference to their tumourigenic potential and genetic anomalies depending on their location in different areas of the tumour. These facts raise the question whether CSCs are transformed cell types or just specific functional state of cells 2,12. As far as the expression of markers is concerned CSCs similar to NSCs express Nestin either alone or together with expression of GFAP 4,7,8. Expression of Sox-2 is observed but quite often as aberrant expression as a sign of disorganisation of the process of differentiation 4,8. In contrast to the cytoplasmic expression of Sox-2 specific to NSCs, Sox-2 in CSCs is expressed in the cell nucleus 7. Another marker used for phenotypic characterisation of CSCs is CD44 which is with a role for tumour invasion and is detected on cells from the invasive part of the tumour 7. CD133 is a marker which is believed to be expressed under hypoxic conditions 7,21 and its significance is quite contradictive in relation to CSCs. Some authors claim that CD133 is the marker which identifies the CSCs 1,8,14,18,20, its expression is a poor prognostic sign 7 and only the CD133+ cells have tumourigenic properties 14,22. A hierarchy is described as the CD133+ cells at the top Table 1 Comparison between NSCs and CSCs in reference to their major features and expression of specific phenotype markers Self-renewal Yes Yes Differentiation Yes Yes Multipotency Yes Yes Signal pathways NSCs Similar Genetic alterations Similar Nestin Yes Yes GFAP Yes or no Yes or no Sox-2 Yes (predominantly cytoplasmic) Yes (predominantly nuclear) CD133 Yes Yes (disputable) CD44 Yes Yes CSCs, cancer stem cells; GFAP, glial fibrillary acidic protein; NSCs, neural stem cells. CSCs

3 Page 3 of 6 The author has referenced some of its own studies in this methodology. These referenced studies have been conducted in accordance with the Declaration of Helsinki (1964) and the protocols of these studies have been approved by the relevant ethics committees related to the institution in which they were performed. All human subjects, in these referenced studies, gave informed consent to participate in these studies. There are two basic approaches to culture cells isolated from GBM after the tissue sample is homogenised mechanically, followed by treatment with collagenase, papain and DNase and the single-cell suspension obtained is cultured in 1. Serum-free medium supplemented with 10 ng EGF, 10 ng bfgf and B27 with the aim to obtain neurospheres (NSs)4,5,19,23,25, Medium supplemented with 10% foetal bovine serum (FBS) with the aim to obtain adherent cells (ACs) in the culture18,19,23,26,27. Additionally, some other protocols for GBM cells cultures are used with the aim to obtain some intermediate forms of cell cultures: 1. Culture medium supplemented with EGF, bfgf and 10% FBS28; 2. Serum-free medium with EGF and bfgf in culture plates coated with laminin to obtain AC9,29; 3. Cultures in medium with 10% FBS in non-adherent plates (coated with poly 2-hydroxyethyl methacrylate (polyhema) to obtain serum spheres 30. Discussion Neurospheres As mentioned above one of the forms of growth of GBM cells is the NSs as serum-free medium with EGF and bfgf are needed for their formation26. This is cell growth type which is common for both NSCs and CSCs2. The NSs are heterogeneous aggregates (Figure 1a) and it is believed that they originate from a single cell and its progeny in the NSs are cells from type B or type C9,12. NSs have self-renewal capacity, clonal origin and after dissociation of the structure they form secondary spheres9. It is assumed that the NSs represent the most malignant zones of GBM2, they are formed from CSCs and have the same genetic and phenotypic characteristics of the primary tumour4,19. The phenotypic changes are manifested by expression of stemmarkers typical for NSCs such as Nestin, Sox-2, CD44, CD133 and lack of GFAP2,29. Some authors presume that the CD133+ cells form the NSs and in the same time they express Nestin and GFAP simultaneously22, while others claim that CD133 cells Figure 1: Different types of growth in vitro of GBM isolated cells: (a) NSs grown in serum-free medium with EGF and bfgf, (b) ACs in medium with 10% BFS without EGF and bfgf and (c) semi-acs grown in medium supplemented with FBS, EGF and bfgf. Magnification 40x. ACs, adherent cells; bfgf, basic fibroblast growth factor; EGF, epidermal growth factor; FBS, foetal bovine serum; NSs, neurospheres; GBM, glioblastoma multiforme. of it and they are the real CSCs14,22. Other authors claim that CD133 cannot be used as a universal marker for CSCs9. The hierarchical connection between CD133+ and CD133 cells is not accepted23 and it is claimed that CD133-cell are tumourigenic with self-renewal capacity and proliferation equal to the CD133+ cells4,7,8,24. CD133 expression is detected in 60% of GBMs tested, and it is believed that this molecule can migrate from the cytoplasm to the cell membrane23. Some GBMs express processed form of the CD133 which is not recognised by many of the monoclonal antibodies used9. Some additional markers, except the abovementioned, are described both in NSCs and CSCs such as Musashi-1, CXCR4 and CD157. It is obvious from the data presented above that CSCs have a lot of similarities with the NSCs (Table 1) and this is the basis for the CSCs hypothesis for development of GBM. The aim of this article was to discuss culturing and the phenotype of CSCs from GBM.

4 Page 4 of 6 can form NSs too 13. Data are reported that CD133 is expressed rarely and its expression is quite weak in NSs 26 and another study demonstrates that from 11 GBM samples growing as NSs only four of them are CD133 positive, whereas in seven samples from this series there is very weak expression or lack of this marker. All samples show tumourigenicity and form NSs when cultured in vitro 23. Adherent cells Cells isolated from GBM grow as ACs when cultured in the presence of FBS but without growth factors 26. They form a monolayer characterised initially by its heterogeneity which later turns into a homogeneous monolayer of fibroblast-like cells (Figure 1b). The ACs are genetically and phenotypic different from the primary tumour and are capable of self-renewal 2,4,9,26. It is assumed that the serum leads to irreversible alterations associated with partial differentiation of the GBM cells and this is reflected on the markers expression 29. A lower expression of Nestin and Sox-2 is observed compared with the NSs, but a higher expression of GFAP and βiii-tubulin is observed 2,8,19,26,29. Still, simultaneous expression of Nestin and GFAP has been reported with AC 27. Cultures with serum lead to loss of the CD133 expression 19,26 and as a whole it is assumed that ACs are characterised with partial stemness 2, loss of GBM characteristics, lower capacity of self-renewal and questionable tumourigenicity 9,29. Comparison between neurospheres and adherent cells and intermediate models Published data predominantly claim that the NSs are a better model of CSCs 29 with higher expression of Nestin, Sox-2 and CD133 and lower expression of differentiation markers such as GFAP 8,26. The NSs have the same genetic alterations as the primary tumour; they are with higher tumourigenicity and higher capacity of self-renewal compared with the ACs. However, it should be stated that the differences between NSs and ACs are not so salient. Some intermediate forms between NSs and ACs have been described and most commonly defined as semi-adherent cultures (Figure 1c) in which both NSs and ACs are observed 5,18,28. Moreover, ACs are described which in the presence of serum form aggregates on the bottom of the wells which strongly resemble NSs and when transferred to serumfree medium grow as NSs 26. Although the NSs are accepted as a better model for CSCs as compared with ACs 29, there are data which cast doubt on the strict distinction between both cell types as far as the fraction of CSCs are concerned. When NSs are transferred to medium with serum they grow as ACs and vice versa ACs grow as NSs in serum-free medium with EGF and bfgf. This raises the question about some crucial changes in the cell behaviour 2 and because of that models for cultures are designed which combine the conditions for NSs and ACs. When GBM cells are cultured in serum-free medium with EGF and bfgf in tissue plate coated with laminin, the NSs adhere and grow as ACs and this way the EGF and bfgf get access to all cells which are not accessible otherwise. Almost all cells were shown to be positive for Nestin, Sox-2, CD133 and CD44 which accepted as proof of this assumption 9. Another intermediate model uses plates coated with polyhema which prevents the adhesion and the GBM cells are cultured with FBS. Serum spheres are obtained and the cells show stronger expression of Nestin, Sox-2 and CD44 compared with ACs and NSs. The serum spheres do not express CD133 and show an increased migration activity and formation of cell colonies 30. The intermediate model that we have developed and used is characterised with culturing of GBM isolated cells in medium supplemented simultaneously with EGF, bfgf and 10% FBS. In all cases we observed growth of ACs and just one case the growth was defined as semi-adherent. Confocal microscopy and flow cytometry studies of the expression of CSCs markers showed that in GBM samples about 97.1% of the cultured cells expressed Nestin, 79.3% Sox-2+, 78.8% CD44+ and 71.3% GFAP+. Expression of CD133 was detected just in two GBM samples 28. These results show that the model used is closer to the other intermediate models than the classical models, such as NSs and ACs. In spite of serum presence the cultured cells expressed in higher percentages CSCs markers with exception of CD133. Although EGF and bfgf were present in the culture medium, the cells grew as АCs and just in a single case they were semi-adherent cells but we never observed NSs under these culture conditions (Table 2). Conclusion It can be stated that the protocol for GBM cells cultured as NSs is not the only one and not necessarily the best model for CSCs. Plenty of data have been reported pointing to the assumption that the NSs cells expressing specific markers cannot explain completely the concept of CSCs. The reasons are that NSs do not always express higher levels of markers for CSCs compared with ACs and intermediate models, however ACs and intermediate models cells have many properties of CSCs. Abbreviations list ACs, adherent cells; bfgf, basic fibroblast growth factor; CSCs, cancer stem cells; EGF, epidermal growth factor; FBS, foetal bovine serum; GBM, glioblastoma multiforme; GFAP, glial fibrillary acidic protein; NSCs, neural stem cells; NSs, neurospheres; polyhema, poly 2-hydroxyethyl methacrylate.

5 Page 5 of 6 Table 2 Type of culture growth and expression of markers for CSCs depending on the culture conditions for GBM isolated cells Culture conditions Growing Nestin Sox2 CD133 CD44 GFAP EGF, FGF, B27 NSs / FBS АCs +/ +/ +/ EGF, FGF, B27 laminin ACs / FBS, polyhema Serum spheres No data EGF, FGF, FBS ACs or semi-acs / ACs, adherent cells; CSCs, cancer stem cells; EGF, epidermal growth factor; FBS, foetal bovine serum; GBM, glioblastoma multiforme polyhema, poly 2-hydroxyethyl methacrylate. References 1. McLendon R, Rich J. Glioblastoma stem cells: a neuropathologist s view. J Oncol Aug;2011: Schiffer D, Mellai M, Annovazzi L, Piazzi A, Monzeglio O, Caldera V. Glioblastoma cancer stem cells: basis for a functional hypothesis. Stem Cell Discov Jun;2: Galli R, Binda E, Orfanelli U, Cipelletti B, Gritti A, De Vitis S, et al. Isolation and characterization of tumorigenic, stemlike neural precursors from human glioblastoma. Cancer Res Oct;64(19): Gürsel D, Shin B, Burkhardt J, Kesavabhotla K, Schlaff C, Boockvar J. Glioblastoma stem-like cells-biology and therapeutic implications. Cancers Jun;3(2): Günther H, Schmidt N, Phillips H, Kemming D, Kharbanda S, Soriano R, et al. Glioblastoma-derived stem cell-enriched cultures form distinct subgroups according to molecular and phenotypic criteria. Oncogene May;27(20): Altaner C. Glioblastoma and stem cells. Neoplasma Mar;55(5): Dimov I, Tasić-Dimov D, Conić I, Stefanovic V. Glioblastoma multiforme stem cells. Sci World J Apr;11: Matsuda Y, Yoshimura H, Suzuki T, Ishiwata T. Nestin: neural stem/progenitor cell marker in brain tumors. In: Lichtor T, editor. Evolution of the molecular biology of brain tumors and the therapeutic implications, 1st ed. InTech; Rijeka, Croatia Chapter 23, pp Gilbert C, Ross A. Cancer stem cells: cell culture, markers, and targets for new therapies. J Cell Biochem Dec;108(5): Sanai N, Alvarez-Buylla A, Berger M. Neural stem cells and origin of gliomas. N Engl J Med Aug;353(8): Villa A, Rubio F, Navarro B, Bueno C, Martínez-Serrano A. Human neural stem cells in vitro. A focus on their isolation and perpetuation. Biomed Pharmacother Mar;55(2): Schiffer D, Mellai M, Annovazzi L, Piazzi A, Monzeglio O, Caldera V. On the origin and growth of gliomas. Anticancer Res Jun;30(6): Facchino S, Abdouh M, Bernier G. Brain cancer stem cells: current status on glioblastoma multiforme. Cancers Mar;3(2): Cho D, Lin S, Yang W, Lee H, Hsu D, Lin H, et al. Targeting cancer stem cells for treatment of glioblastoma multiforme. Cell Transplant Feb;22(4): Liu Y, Han S, Wu Y, Tuohu T, Xue H, Cai J, et al. CD44 expression identifies astrocyte-restricted precursor cells. Dev Biol Dec;276(1): Pluchino S, Gritti A, Blezer E, Amadio S, Brambilla E, Borsellino G, et al. Human neural stem cells ameliorate autoimmune encephalomyelitis in non-human primats. Ann Neurol Sep;66(3): Doetsch F, Petreanu L, Caille I, Garcia-Verdugo J-M, Alvarez-Buylla A. EGF converts transit-amplifying neurogenic precursors in the adults brain into multipotent stem cells. Neuron Dec;36(6): Pavon L, Marti L, Sibov T, Miyaki L, Malheiros S, Mamani J, et al. Isolation, cultivation and characterization of CD133+ stem cells from human glioblastoma. Einstein Apr Jun;10(2): (in English, Portuguese). 19. Yuan X, Curtin J, Xiong Y, Liu G, Waschsmann-Hogiu S, Farkas D, et al. Isolation of cancer stem cells from adult glioblastoma multiforme. Oncogene Dec;23(58): Nakahata A, Suzuki D, Rodini C, Pereira M, Janjoppi L, Okamoto O. Human glioblastoma cells display mesenchymal stem cell features and form intracranial tumors in immunocompetent rats. J Stem Cells. 2010;5(3): Pistollato F, Persano L, Puppa A, Rampazzo E, Basso G. Isolation and expansion of regionally defined human glioblastoma cells in vitro. In: Schlaeger T, editor. Current protocols in stem cell biology, 1st ed. Wiley online library; Chapter 3, Unit 3.4, pp Singh S, Hawkins C, Clarke I, Squire J, Bayani J, Hide T, et al. Identification of human brain tumour initiating cells. Nature Nov;432(7015): Brescia P, Ortensi B, Fornasari L, Levi D, Broggi G, Pelicci G. CD133 is essential for glioblastoma stem cell maintenance. Stem Cells May;31(5): Beier C, Kumar P, Meyer K, Leukel P, Bruttel V, Aschenbrenner I, et al. The cancer stem cell subtype determines immune infiltration of glioblastoma. Stem Cells Dev Oct;21(15): Mao P, Joshi K, Li J, Kim S, Li P, Santana-Santos L, et al. Mesenchymalglioma stem cells are maintained by activated glycolytic metabolism involving aldehyde dehydrogenase 1A3. Proc Natl Acad Sci USA May;110(21): Caldera V, Mellai M, Annovazzi L, Piazzi A, Lanotte M, Cassoni P, Schiffer D. Antigenic and genotypic similarity between

6 Page 6 of 6 primary glioblastomas and their derived neurospheres. J Oncol. 2011; 2011: Veselska R, Kuglik P, Cejpek P, Svachova H, Neradil J, Loja T, et al. Nestin expression in the cell lines derived from glioblastoma multiforme. BMC Cancer Feb;6(32). 28. Kyurkchiev D, Naydenov E, Tumangelova-Yuzeir K, Ivanova-Todorova E, Belemezova K, Bochev I, et al. Cells isolated from human glioblastoma multiforme express progesterone induced blocking factor (PIBF). Cell Mol Neurobiol May;34(4): Lee J, Kotliarova S, Kotliarov Y, Li A, Su Q, Donin N, et al. Tumor stem cells derived from glioblastomas cultured in bfgf and EGF more closely mirror the phenotype and genotype of primary tumors than do serum-cultured cell lines. Cancer Cell May;9(5): Hong X, Chedid K, Kalkanis S. Glioblastoma cell line-derived spheres in serumcontaining medium versus serum-free medium: a comparison of cancer stem cell properties. Int J Oncol Jul;41(5):