University of Calgary PRISM: University of Calgary's Digital Repository
Graduate Studies The Vault: Electronic Theses and Dissertations
2016 The Role of ING5 in Maintaining Stemness of Brain Tumor Initiating Cells
Wang, Fangwu Jr
Wang, F. J. (2016). The Role of ING5 in Maintaining Stemness of Brain Tumor Initiating Cells (Unpublished master's thesis). University of Calgary, Calgary, AB. doi:10.11575/PRISM/28327 http://hdl.handle.net/11023/3233 master thesis
University of Calgary graduate students retain copyright ownership and moral rights for their thesis. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission. Downloaded from PRISM: https://prism.ucalgary.ca UNIVERSITY OF CALGARY
The Role of ING5 in Maintaining Stemness of Brain Tumor Initiating Cells
by
Fangwu Wang
A THESIS
SUBMITTED TO THE FACULTY OF GRADUATE STUDIES
IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE
DEGREE OF MASTER OF SCIENCE
GRADUATE PROGRAM IN BIOCHEMISTRY AND MOLECULAR BIOLOGY
CALGARY, ALBERTA
AUGUST, 2016
© Fangwu Wang 2016 Abstract
Brain tumor initiating cells (BTICs) are believed to account for the recurrence of glioblastomas following treatment. Recent studies have shown that the stemness of BTICs and intratumoral differentiation hierarchy are determined largely on the epigenetic level. The ING family of epigenetic regulators function in diverse growth regulatory, metastasis and chemoresistance pathways, through targeting different histone acetyltransferase (HAT) and histone deacetylase
(HDAC) complexes to the H3K4me3 mark to alter histone acetylation. ING5, a stoichiometric unit of three HAT complexes, has been directly implicated in the maintenance of epidermal stem cells.
Here we found that ING5 was highly expressed in BTICs and rapidly downregulated upon in vitro differentiation. Ectopic expression of ING5 promoted self-renewal, prevented lineage differentiation and increased the stem cell pool in the BTIC population, accompanied by an elevated expression of stem cell core transcription factors OCT4, OLIG2 and Nestin. ING5 enhanced the activity of the PI3K/AKT and MEK/ERK pathways in the absence of growth factors to sustain self-renewal of BTICs over serial sphere passage. Transcriptome analysis indicated
ING5 was an inducer of the intracellular calcium signaling and follicle stimulating hormone pathways, which were confirmed to co-operatively enhance the self-renewal of BTICs. This study identifies ING5 as a positive regulator of BTIC stem cell character, whose expression negatively correlates with patient prognosis, especially in the Proneural subtype and tumors with low SOX2 expression, therefore suggesting that altering histone acetylation status and the signaling pathways induced by ING5 may provide useful clinical targets to reduce recurrence in glioblastoma.
ii Acknowledgements
First, I would like to express my gratitude to my supervisor, Dr. Karl Riabowol, for his continuous guidance, encouragement and support during my Master’s study and research. Second, I would like to thank the members of my supervisory committee, Dr. Jennifer Cobb and Dr. Derrick
Rancourt, for their very helpful advice and support. And I would like to thank the members in our lab, Mr. Arash Nabbi, Dr. Subhash Thalappilly, Dr. Yang Yang, Ms. Nancy Adam, and the previous student Dr. Uma Rajarajacholan, for their unconditional help and precious advice throughout the two years. I would also like to thank Dr. Alice Wang, Dr. Artee Luchman and Mr.
Charles Chesnelong for their valuable inputs and technical supports about our research system.
Finally, I would like to thank my dear parents and families for their understanding and kind encouragement during my Master’s study.
iii Table of Contents
Abstract ...... ii Acknowledgements ...... iii Table of Contents ...... iv List of Tables ...... vi List of Figures and Illustrations ...... vii List of Symbols, Abbreviations and Nomenclature ...... ix
CHAPTER I: INTRODUCTION ...... 1 1.1 Brain tumor epidemiology ...... 1 1.2 Molecular subtypes of Glioblastoma ...... 4 1.2.1 Driver mutations in GBM ...... 4 1.2.2 Molecular subtypes of GBM ...... 5 1.2.3 Epigenetic subtypes ...... 5 1.3 Tumor initiating cells in GBM ...... 7 1.3.1 The cancer stem cell model ...... 7 1.3.2 Isolation and characterization of BTICs ...... 14 1.3.3 Cell markers for CNS lineages and BTIC differentiation ...... 16 1.3.4 The cellular origin of BTICs and GBM ...... 18 1.4 Mechanisms of BTIC stemness ...... 20 1.4.1 Pluripotency factors in BTIC regulation ...... 20 1.4.2 Neurodevelopmental factors in BTIC regulation ...... 21 1.4.3 Mitogenic pathways and tumor suppressors in BTIC regulation ...... 22 1.4.4 Epigenetic regulation of BTICs ...... 24 1.5 The INhibitor of Growth (ING) family of epigenetic regulators ...... 28 1.5.1 Major functions of ING1 in growth regulation and cancer biology ...... 28 1.5.2 Structural features of the ING proteins ...... 30 1.5.3 Epigenetic functions of ING proteins ...... 31 1.5.4 Functions of ING5 in cancer ...... 32 1.5.5 ING5 in stem cell maintenance ...... 33 1.6 Objectives ...... 38
CHAPTER TWO: MATERIALS AND METHODS ...... 39 2.1 BTIC cultures and sphere formation assays ...... 39 2.2 Sphere cell differentiation and immunofluorescence ...... 39 2.3 pCI plasmid transfection ...... 40 2.4 PiggyBac stable cell lines ...... 42 2.5 siRNA transfection ...... 43 2.6 lentiviral-based shRNA system ...... 44 2.7 Quantitative Real-Time PCR ...... 45 2.8 Gene expression microarray and data analysis ...... 45 2.9 Western blotting ...... 45 2.10 Flow cytometry ...... 46 2.11 Cell division symmetry analysis ...... 46 2.12 Live cell calcium imaging ...... 47 2.13 Chemicals ...... 47
iv 2.14 Statistics ...... 47
CHAPTER THREE: RESULTS ...... 48 3.1 Determining the expression of ING5 during BTIC differentiation ...... 48 3.2 Determining the function of ING5 in BTIC self-renewal ...... 52 3.3 The effects of ING5 on BTIC differentiation ...... 64 3.4 Examining the function of ING5 in BTIC self-renewal in the absence of growth factors ...... 64 3.5 Transcriptome assay in ING5 knockdown cell lines ...... 69 3.6 Determining the regulatory function of ING5 on intracellular calcium levels ...... 75 3.7 Determining the effects of calcium levels on BTIC self-renewal ...... 81 3.8 Induction of the FSH pathway by ING5 ...... 87 3.9 Determining the role of the ING5-induced FSH pathway in BTIC self-renewal ....92 3.10 Determining the role of the PHD motif in ING5-induced stemness maintenance 97 3.11 The correlation between ING5 expression levels and GBM prognosis ...... 100
CHAPTER FOUR: DISCUSSION ...... 105
CHAPTER FIVE: CONCLUSIONS ...... 117
REFERENCES ...... 118
v List of Tables
Table 1. Top 25 Gene Ontology hits enriched in shR-ctr group using GSEA analysis. (NES = normalized enrichment score.) ...... 76
Table 2. Gene list of Quantity of FSH from the IPA downstream function...... 89
Table 3. Crosstabulation of ING5 and SOX2 showing positive correlation of gene expression across GBM samples (P = 3.0178E-9)...... 103
Table 4. Gene list of Quantity of Steroid from the IPA downstream function...... 112
Table 5. Gene list of Quantity of Thyroid Hormone from the IPA downstream function. ...113
vi List of Figures and Illustrations
Figure 1. ING5 expression decreases during embryonic stem cell differentiation (A and B from Vladislav Alekseev)...... 35
Figure 2. Protocols for transient and stable ING5 overexpression and knockdown experiments in BT 189 cells...... 41
Figure 3. The BTIC in vitro differentiation model...... 49
Figure 4. The expression of ING5 decreases during differentiation in BTIC lines...... 50
Figure 5. ING5 enhances sphere forming ability of BT 189 cells...... 53
Figure 6. BT 189 stable shRNA cell lines have lower sphere forming capability...... 55
Figure 7. The alteration of ING5 protein levels by various manipulations in BT 189 cell line...... 56
Figure 8. ING5 induces the expression of stem cell factors and downregulates differentiation markers...... 58
Figure 9. ING5 overexpression increases the CD133+/CD44+ population in BTICs...... 59
Figure 10. ING5 knockdown decreases the CD133+ population in BTICs...... 60
Figure 11. ING5 increases the frequency of asymmetric division of BTICs...... 62
Figure 12. ING5 overexpression inhibits the acquisition of differentiation morphology in iPB cells...... 65
Figure 13. ING5 overexpression prevents lineage differentiation...... 66
Figure 14. ING5 knockdown promotes lineage differentiation...... 67
Figure 15. ING5 sustains self-renewal over serial passages in the absence of growth factors...... 68
Figure 16. ING5 overexpression induces activation of the PI3K/AKT and MEK/ERK signaling pathways...... 70
Figure 17. Inhibiting the PI3K/AKT and MEK/ERK signaling pathways promotes neuronal differentiation...... 71
Figure 18. MEK/ERK pathway inhibition decreases the CD133 positive population of BTICs...... 73
Figure 19. Transcriptome analysis in shRNA knockdown cell lines...... 74
vii Figure 20. GSEA analysis shows that Calcium channel activity is induced by ING5...... 77
Figure 21. ING5 upregulates genes encoding for calcium channel components...... 78
Figure 22. ING5 increases intracellular calcium levels in live cells...... 79
Figure 23. High ING5 expression level is correlated with high intracellular calcium levels...... 82
Figure 24. ING5 knockdown decreases the proportion of cells with high intracellular calcium levels...... 83
Figure 25. ING5 overexpression increases the proportion of cells with high intracellular calcium levels...... 84
Figure 26. Calcium level elevation increases CD133 positive cells in BTICs...... 85
Figure 27. Calcium level affects BTIC self-renewal ability...... 86
Figure 28. Calcineurin is a downstream effector of calcium signaling in self-renewal regulation...... 88
Figure 29. ING5 induces the FSH pathway in BTICs...... 90
Figure 30. Calcium level and the FSH signaling pathway positively regulate self-renewal ability of BTICs...... 93
Figure 31. The FSH pathway promotes stem cell properties of BTICs...... 95
Figure 32. The PHD motif is required for the stem cell maintenance function of ING5...... 98
Figure 33. Model for how ING5 functions in the maintenance of BTIC self-renewal...... 99
Figure 34. ING5 levels negatively correlate with survival of GBM...... 101
Figure 35. The subtype-specific correlation of ING5 expression level with GBM survival. 102
Figure 36. The potential effect of ING5 expression on GBM survival is affected by SOX2 status...... 104
viii List of Symbols, Abbreviations and Nomenclature
Symbol Definition AKT AKT Serine/threonine Kinase BTIC Brain Tumor Initiating Cell CaN Calcineurin CPA Cyclopiazonic Acid CSC Cancer Stem Cell EGF Epidermal Growth Factor ERK Extracellular Signal-regulated Kinase ESC Embryonic Stem Cell FGF Fibroblast Growth Factor FSH Follicle Stimulating Hormone GBM Glioblastoma Multiforme H3K4me3 Trimethylated Histone H3 Lysine 4 HAT Histone Acetyltransferase HDAC Histone Deacetylase Complex IDH Isocitrate Dehydrogenase ING Inhibitor of Growth iPB Inducible PiggyBac MEK Mitogen-activated Protein Kinase Kinase NSC Neural Stem Cell PDGF Platelet-derived Growth Factor PHD Plant Homeodomain PI3K Phosphoinositide 3-kinase RTK Receptor Tyrosine Kinase TCGA The Cancer Genome Atlas TMZ Temozolomide
ix 1
CHAPTER I: INTRODUCTION
1.1 Brain tumor epidemiology
Tumors of the human central nervous system have been classified and graded by the World Health
Organization (WHO), a system that has been used and accepted globally (Gonzales 2001, Louis et al. 2007). The WHO classification is mainly based on histopathological features, tumor origin/localization and survival data. Under the “tumors of neuroepithelial tissue” category, there are tumors from various cellular origins, including astrocytic tumors, oligodendroglial tumors, neuronal tumors, embryonal tumors and several others. A grading system was introduced by WHO to assess a range of malignancy scale across these tumors. Glioblastoma Multiforme (GBMs) in the astrocytic tumor category and embryonal tumors are the most malignant Grade IV tumors.
They are characterized by high mitotic activity, widespread infiltration, vascular proliferation, necrosis and rapid disease evolution (Louis et al. 2007). The prognosis of these Grade IV patients are highly dependent on the availability of effective treatment. Because of the age at diagnosis
(median age of 64 years) and an inadequacy of current anti-tumor therapies, GBM patients have the poorest prognosis among all the brain tumor types. The five-year survival of GBM patients is only 5.1 % (Ostrom et al. 2015). GBM also has the highest rate of incidence, accounting for 46.1% of primary malignant tumors in the central nervous system, with 17,000 new diagnoses per year in the United States (Ostrom et al. 2015, Dolecek et al. 2012).
The most consistent and significant risk factors for GBM are age and radiation exposure
(Bondy et al. 2008). GBM is thought to arise from multistep genetic alterations with accumulating oncogenic mutations, therefore there is a clear trend of increased risk with age. The cranial X- irradiation for cancer therapy in childhood is also associated with a higher risk for developing
2
GBM in adulthood (Bondy et al. 2008). The cases relating hereditary syndromes and family tumor history to GBM incidence are rare and the effects of environmental and lifestyle factors on the risk of GBM are modest (Gu et al. 2009, Ohgaki 2009). GBM is 1.6-fold more common in men compared to women and 2 times higher in the white ethnicity compared to the black, which suggests the possible effects of genetic polymorphisms across ethnical populations (Ostrom et al.
2015).
The current treatment modality for the Grade IV GBM is maximal neurosurgical resection followed by adjuvant radiotherapy combined with chemotherapy. The DNA alkylating agent temozolomide (TMZ) is widely used, which extends the median survival from 12 months to 15 months, with a doubled 2-year survival rate to 27 % (Omuro and DeAngelis 2013). Patients with silenced expression of a DNA repair enzyme, O6-methylguanine-DNA methyltransferase
(MGMT) are more likely to benefit from radiation-TMZ combined therapy (Hegi et al. 2005).
However, the radiation and TMZ regimens are not able to prevent tumor recurrence and progression after surgery. Multiple factors may account for the failure of treatment, including insufficient delivery of chemotherapy dosages across the blood-brain barrier, local diffusion and invasion of the tumor cells, and most importantly, the remarkable intratumoral heterogeneity and fast tumor evolution which frequently leads to the development of treatment resistance.
The emergence of next-generation sequencing technology has allowed confirmation of traditional diagnosis with molecular insights for better predictions of the malignancy of tumors and responsiveness to treatments. For example, mutations in isocitrate dehydrogenase (IDH) and co-deletion of 1p and 19q chromosome arms were identified as typical markers for lower-grade gliomas and related to better prognosis than for glioblastomas (Brat et al. 2015). Molecular
3 characterization of GBMs from large numbers of patients is becoming an important approach to discover therapeutic and diagnostic target candidates.
4
1.2 Molecular subtypes of Glioblastoma
1.2.1 Driver mutations in GBM
Genomic analysis has been widely applied to cancer studies over the past years to identify genetic alterations and to help determine the molecular basis for the tumor onset. The genome of cancer cells is altered on multiple levels, including DNA sequence changes, copy number alterations and chromatin rearrangements. Through molecular screening in large GBM cohorts, recurrent chromosomal alterations and co-selection of genes in these aberrant chromosome territories were identified (Bredel et al. 2009). The Cancer Genome Atlas (TCGA) research network has published a series of reports on the multidimensional genomic analysis of malignant gliomas (2008, Brennan et al. 2013, Brat et al. 2015, Ceccarelli et al. 2016), which catalogued the most frequent genomic alterations that could be regarded as driver events for glioblastoma pathogenesis. The most prevalent mutations converge on three core signaling pathways: receptor tyrosine kinase (RTK) signaling, the p53 axis and the retinoblastoma (RB) tumor suppressor pathway, with respective frequencies of 88 %, 78% and 87% (TCGA 2008). Activation of the RTK pathways can be caused by abnormalities in the epidermal growth factor receptor (EGFR), erb-b2 receptor tyrosine kinase
2 (ERBB2), mutations of the phosphoinositide 3-kinase (PI3K) complex and alterations of platelet- derived growth factor receptor, alpha polypeptide (PDGFRA). A subsequent TCGA report integrating multi"omics" analyses provided a more comprehensive landscape of GBM from 543 patients (Brennan et al. 2013) and revealed new oncogenic events resulting from somatic mutations and chromosome rearrangements. Telomerase reverse transcriptase (TERT) promoter mutations and mutations in chromatin remodelers have been identified as potential drivers in GBM (TCGA
2008). Recurrent patterns of mutations were also displayed in GBMs, such as concurrence in
5 parallel pathways and mutual exclusivity in the same pathway, reflecting a pattern of intratumoral heterogeneity and selection during the tumor evolution.
1.2.2 Molecular subtypes of GBM
In 2010, Roel Verhaak et al. and the TCGA group conducted hierarchical clustering on the gene expression datasets of GBM and identified four major subtypes, the Proneural, Neural, Classical and Mesenchymal, with 210 signature genes assigned to each subtype (Verhaak et al. 2010). The transcriptome profile of each subtype is closely matched with the gene expression signature of its respective neural cell type. For example, the Proneural group resembles oligodendrocytic cell type while the Classical group is enriched with the astrocytic signature. The Neural group displays differentiated neuronal signatures, while the Mesenchymal group is associated with astroglial and microglia markers. This study also demonstrated a strong link between genetic aberrations and particular subtypes. For example, alterations of PDGFRA/IDH1, EGFR/CDKN2A and neurofibromin 1 (NF1) define Proneural, Classical and Mesenchymal subtypes, respectively. This study dissected the gene expression heterogeneity of GBMs, hinted at the distinct cellular origins of tumors and provided insights into predicting patient outcome based on specific mutations in each subtype.
1.2.3 Epigenetic subtypes
A DNA methylation profiling study conducted by TCGA group further profiled the epigenetic signatures for the four subtypes (Brennan et al. 2013). Six DNA methylation clusters were identified in all GBM samples, which largely overlapped with the epigenetic subtypes defined by
Sturm et al (Sturm et al. 2012). The Proneural subtype consists of two distinct epigenomic classes, the Glioma-CpG-Island-Methylator-Phenotype (G-CIMP) associated with IDH mutations and the non-G-CIMP signature enriched in PDGFRA-mutated samples. In addition, a high rate of
6 mutations in chromatin remodelling genes (CMGs) was observed in three epigenetic subclasses.
The alterations of mixed-lineage leukemia (MLL) and histone deacetylase complex (HDAC) family members were enriched in the M2 subtype (associated with Mesenchymal subtype). This suggests a functional relationship between DNA methylation, chromatin remodeling in different molecular subtypes.
7
1.3 Tumor initiating cells in GBM
1.3.1 The cancer stem cell model
The concept of cancer stem cells (CSCs) is based on the heterogeneity of the cellular composition in cancer. The phenotypical and functional heterogeneity of cancer cells was well documented in many tumors several decades ago (Fidler and Kripke 1977, Heppner 1984). It has been noted that some germ lineage cancers, neuroblastomas and myeloid leukemia contained cells with variable degrees of differentiation, and the postmitotic cells with limited proliferation potential could be derived from the undifferentiated cancer cells (Kleinsmith and Pierce 1964, Ambros et al. 2002,
Fearon et al. 1986).
Hypothesis and supporting evidence
The cancer stem cell model includes two major hypotheses (Dick 2008): 1). the heterogeneity of cancer is derived from the hierarchical organization of cancer stem cells and their differentiated progeny cells, primarily based on the epigenetic hierarchy rather than genetic variabilities; 2). the small proportion of cancer stem cells are tumor-initiating, while their differentiated progeny cells have very limited or do not have tumor initiating ability, therefore the cancer stem cells are also known as tumor initiating cells. This model suggests that cancer progression is mainly driven by the small population of cancer stem cells, therefore they should be the major target for cancer elimination.
Since the identification of cellular markers to efficiently enrich stem-like cells in cancer, the cancer stem cell concept gained increasing credence. The subpopulations of live cancer cells from leukemia and breast cancers were separated by fluorescence activated cell sorting (FACS) using flow cytometry and their tumorigenic potential directly compared (Bonnet and Dick 1997, Al-Hajj et al. 2003). In 1990s, two studies identified the cancer initiating cells in human acute myeloid
8 leukemia (AML) by transplantation into non-obese diabetic mice with severe combined immunodeficiency disease (NOD/SCID) (Lapidot et al. 1994, Bonnet and Dick 1997). The frequency of AML-initiating cells from patient peripheral blood was estimated to be one unit in
250,000 cells. Lapidot et al showed that the AML-initiating cells were within the CD34+ CD38- fraction. When 2 × 105 CD34+ CD38- cells were transplanted, high levels of AML-colony-forming units (AML-CFU) were observed, however, the same number of CD34- or CD34+ CD38+ cells were not engrafted and contained no AML-CFU (Lapidot et al. 1994). Bonnet et al showed that
5× 103 CD34+ CD38- cells were sufficient to initiate AML in NOD/SCID mice, however, 100 times as many CD34- or CD34+ CD38 + cells did not engraft (Bonnet and Dick 1997). The self- renewal capabilities of CD34+ CD38- cells were indicated by high levels of AML-CFU and serial transplantation in vivo (Bonnet and Dick 1997). The differentiation potentiality of CD34+ CD38- cells injected into mice was indicated by the presence of differentiated lineage markers of cells derived from the CD34+ CD38- AML-initiating cells (Bonnet and Dick 1997).
Al-Hajj et al showed in breast cancer as few as 100 CD44(+)CD24(-/low)Lineage(-) cells were able to form tumors in mice, whereas tens of thousands of cells with alternate phenotypes failed to form tumors. The CD44(+)CD24(-/low)Lineage(-) population was capable of generating the phenotypic heterogeneity found in the initial tumor and could be serially transplanted in vivo, indicating the differentiation and self-renewal capabilities reminiscent of normal stem cells (Al-
Hajj et al. 2003).
Other studies followed a similar approach and supported the cancer stem cell model in solid tumors including colorectal cancer (Dalerba et al. 2007, O'Brien et al. 2007), pancreatic cancer (Li et al.
2007) and ovarian cancer (Curley et al. 2009). O'Brien et al found colon cancer-initiating cells
(CC-ICs) were enriched in the CD133+ subpopulation, and injection of 1000 cancer cells with
9 highest levels of CD133 could reproducibly generate tumor in NOD/SCID mice (O'Brien et al.
2007). However, of 47 mice injected with CD133- cells (dose range: 2 × 103 to 2.5 × 105 cells), only one mouse transplanted with the highest cell dose (2.5 × 105) generated a tumour. The
CD133+ CC-ICs could initiate tumor during serially transplantation and the degree of differentiation of CD133+ xenografts resembled the orignial tumors. With a different set of markers, Dalerba et al indicated a (EpCAM)high/CD44+ cancer initiating cell population in solid colorectal cancer tissues (Dalerba et al. 2007). They found 200 to 500 (EpCAM)high/CD44+ cells could frequently generate tumor while 104 (EpCAM)low/CD44- cells consistently failed to form tumor. However, the self-renewal and differentiation properties of this cancer initiating cell population were not characterized in this study.
The hierarchical organization of cancers with CSCs at the apex has been gradually acknowledged, and it has been accepted that the cancers harboring CSCs are primarily propagated from this small cell population with exceptional tumorigenic capability rather than the differentiated bulk tumor cells. The stem cell-like properties of CSCs also provided explanations for the resilience of tumors to therapies and metastatic relapse after years of initial treatment (Aguirre-Ghiso 2007).
Controversies:
Nevertheless, some controversial aspects about the cancer stem cell model arise in many recent studies, suggesting the cancer stem cell model applies to some, but not all cancer types, and the current cellular markers may not be sufficient to fully define and purify this population in some cases.
First, it is known that not all cancer initiating cells are cancer stem cells, but some studies failed to address the stem cell-like properties of the indicated cancer stem cell populations. Moreover,
10 the intratumoral hierarchical organization is not clear in some cancer types. Although the lineage differentiation is obvious in some cancers, such as AML (Bonnet and Dick 1997) and glioblastomas (Galli et al. 2004), the morphology of tumorigenic and nontumorigenic cells is indistinguishable in breast cancer (Al-Hajj et al. 2003). In addition, there has been no direct evidence that the differences between cancer stem cells and non-stem cancer cells are primarily due to the differentiation hierarchy on the epigenetic level rather than the clonal advantage from genetic differences.
Second, although many studies inject freshly-dissected cells from the tumor tissues into
NOD/SCID mice, some studies transplant cells that have undergone a short period of in vitro culture or from cell lines of long-term culture. The culture systems used can possibly change the original properties of cancer cells, and epigenetic modification is especially susceptible to environmental changes. For example, mitogenic stimulation of EGF used in the isolation and culture of normal neural stem cells (Reynolds and Weiss 1992) and glioma initiating cells (Galli et al. 2004, Lee et al. 2006) was also reported to induce dedifferentiation of mature neural lineages in combination with INK4a/ARF inactivation in mice (Bachoo et al. 2002), indicating it might potentially promote a stem-like state of some highly plastic non-stem cancer cells.
Furthermore, the cancer stem cell model suggests that only a small proportion of stem-like cells at the apex of tumor hierarchy are capable of initiating tumor. However, the widely used xenograft system in immunodeficient mice has been implicated to dramatically underestimate the tumorigenic capability of injected cells of several cancer types, due to the rejection by the residual nature killer cells in the NOD/SCID mice. It has been shown that manipulation of the function of the immune systems in these mice could make a big difference to the engraftment rate of human hematopoietic stem cells and leukemia progenitor cells (McKenzie et al. 2005, Feuring-Buske et
11 al. 2003). Some recent studies suggest the cancer initiating cells can make up a much higher proportion of total cancer cells than expected, using more permissive immunodeficient mouse models. Kelly et al showed that at least one in ten of Eµ-myc pre-B/B lymphoma cells from transgenic mice could generate tumor in normal congenic mice (Kelly et al. 2007). Another transgenic study of mouse malignant peripheral nerve sheath tumor showed that the tumorigenic cells could be as high as 18 % from Nf1(+/-); Ink4a/Arf(-/-) mice (Kelly et al. 2007). Using a more immunocompromised IL2Rγnull NOD/SCID mouse xenograft system, Quintana et al showed that approximate one in four of unselected melanoma cells from patients could form tumors (Quintana et al. 2008). Although not all of these cancer initiating cells are necessarily cancer stem cells, these studies suggest the rate of tumorigenic cells are greatly underestimated in many cases and if there is such a high proportion of cancer initiating cells, then the clonal evolution model may apply better to these cancers instead of the cancer stem cell model (Shackleton et al. 2009).
In addition, whether the CSC properties are intrinsically stable or the identity between CSCs and non-CSCs is interconvertible is not clear. Chaffer et al observed the spontaneous conversion of primary human mammary epithelial cells to CD44hi stem-like state in vitro. The de novo generated
CD44hi cells display the mammary progenitor traits, including 3D ductal structure formation and mammosphere formation capabilities. After oncogenic transformation, the non-stem cells are even more likely to convert to CD44hi stem-like cancer cells (Chaffer et al. 2011). Two human breast cancer cell lines derived from primary tumors show stochastic transitions between luminal, basal and stem-like states in vitro, based on the expression of cell surface markers CD44, CD24 and
EpCAM (Gupta et al. 2011). In this study, only the stem-cell fraction could seed tumors (tumors were detected in four of four injected animals) among the sorted subpopulations, however, when the irradiated carrier cells were admixed, all three fractions (luminal, basal and stem-like) were
12 equally capable of forming tumors, indicating the transitions from non-tumorigenic cells to stem- like tumor initiating cells. Based on these in vitro observations, the interconversion between non- stem cancer cells and CSCs is likely to happen in vivo under acute stresses from treatment modalities. Therefore, the cancer stem cell needs to be adapted in the cases where the stem-like state is interconvertible. There are critical questions to be solved in the future: 1) at what percentage and to what extent the non-stem cells and CSCs are interconvertible in vivo; 2) whether the interconvertible cells could be distinguished from cells that could not transition to other states;
3) what the tumorigenic cell population is made of in the interconvertible cell populations.
Last but not the least, the cellular markers used to enrich cancer stem cells are often controversial.
These markers are important indicators for the intrinsic differences between CSCs and non-stem cancer cells which can be subtle and indistinguishable by morphology. CD133 or Prominin-1, is a pentaspan transmembrane protein and antigen for human hematopoietic stem cells (Miraglia et al.
1997a). It has been shown to enrich cells capable of long-term self-renewal and xenografting in immunodeficient mice in a range of cancers, including glioblastoma (Singh et al. 2004), colon cancer (O'Brien et al. 2007), hepatocellular carcinoma (Ma et al. 2010), and childhood ALL (Cox et al. 2009). CD133 enriches the proportion of colon initiating cells from 1/ 5.7 × 104 to 1/ 262,
(O'Brien et al. 2007) and enriches brain tumor initiating cells from less than 1/105 to at least 1/100
(Singh et al. 2004).
By contrast, several studies showed that the CD133 negative cell populations derived from GBM patients were also capable of generating tumors in immunocompromised mice, although the frequency of tumor initiating cells was not as high as those in the CD133 positive cell populations.
Ogden et al showed that injection of 1 × 105 CD133- cells from 2 of 4 GBM samples could xenograft in immunocompromised mice (Ogden et al. 2008). Another study found that when 10
13 spheroids (each 200-300 µm in diameter) of CD133- cells from GBM patients were injected into nude immnodeficient rats, 19 of 28 rats developed tumors (Wang et al. 2008). Beier et al established CD133- cell lines from GBM patients under medium conditions favoring the growth of stem cells, which could form adherent spheres and have three lineage differentiation potentiality
(Beier et al. 2007). 105 or 106 cells from these CD133- cell lines could induce glioblastoma-like lesions in immunodeficient nude mice. However, a same number of CD133- cells derived from
CD133+ cell lines did not form tumor. Since this study characterized and maintained the CD133-
CSC lines in the presence of mitogenic factors, LIF and N27, the properties of these cells could possibly be changed or selected under the stem cell-favoring culture conditions.
The existence of tumor initiating cells in the CD133- population can be caused by several possibilities: 1) there are stem-like cells that do not express CD133 in contrast to the CD133+ CSC populations, in which case CD133 is not a robust marker for CSCs; 2) a small proportion of
CD133- cells may convert to CD133+ cells during tumorigenesis in vivo; 3) the tumor initiating cells in the CD133- population are not cancer stem cells but the cells gained advantages through clonal evolution, so there might not be an obvious hierarchical organization generated by these cells; 4) the higher number of injected cells in the above studies (over 105) might serve as carrier cells and increase the survival time and the probability of xenograft, since host immune rejection and environmental stresses are critical factors for successful xenograft; 5) a small number of cells with very low CD133 expression are included in the CD133- population which still possess cancer stem cell properties.
Since the fluorescence or magnetic activated cell sorting is antibody-based, the specificity and sensitivity of antibodies is critical for the outcomes of relevant experiments. However, there are caveats for the utilization of different antibody clones in CSC studies. The majority of CD133
14 related experiments have been done with two monoclonal antibodies (mAbs), AC133 and AC141, which led to the original identification of the CD133 antigen on hematopoietic progenitors (Yin et al. 1997). Although the two mAbs have been extensively used, the exact epitopes they recognize have not been well characterized. It has been reported that they recognized different glycosylated epitopes on CD133 protein (Miraglia et al. 1997b). Since different CD133 antibody clones may recognize different epitopes, discrepancies of CD133 distribution using different clones have been reported in multiple studies (Barrantes-Freer et al. 2015, Hermansen et al. 2011). Therefore, comparisons of studies based on the immunoreactivity to CD133 using different antibody clones should be performed with caution. In addition, the most widely used clones AC133 and AC141, could not represent the mRNA and protein levels of CD133 in some cases (Bidlingmaier, Zhu, and
Liu 2008).Since AC133 and AC141 have been successfully used to identify and purify CSC populations, it is possible that certain glycosylation status rather than the expression levels of
CD133 is a better marker for CSCs. The molecular information for the exact binding epitopes and the sensitivities of different clones in different experiments need to be better characterized in the future.
1.3.2 Isolation and characterization of BTICs
The evidence of adult neural stem cells (NSCs) and neurogenesis throughout adulthood has altered the perception about the origin of brain neoplasms. In 1992, Reynolds and Weiss isolated and cultured NSCs from adult mouse brain, which showed proliferation ability under EGF treatment and neuronal and astrocytic differentiation potentials (Reynolds and Weiss 1992). It has been revealed that two major regions in the adult brain of mammals account for constitutive neurogenesis where multipotent neural stem cells reside: the subventricular zone (SVZ) lining the
15 lateral walls of the forebrain lateral ventricles and the subgranular zone (SGZ), the inner layer of dentate gyrus within the hippocampus (Peretto et al. 1999, Alvarez-Buylla and Garcia-Verdugo
2002). In both regions neural stem cells (Type B cells) first generate transiently amplifying progenitors (Type A and D), which subsequently differentiate into mature glia and neurons.
With the same method used for NSC isolation and in vitro growth (Reynolds and Weiss
1992), neural stem-like cells were discovered in the tumor tissues of several types of malignant gliomas (Hemmati et al. 2003, Ignatova et al. 2002, Singh et al. 2003). The cells were enzymatically dissociated and plated in serum free medium with supplements favoring stem cell growth (EGF, bFGF). These cells can form neurosphere-like clones, undergo differentiation into multiple lineages, and maintain self-renewal abilities over long-term passage in vitro, which is reminiscent of the properties of NSCs. However, they differ from NSCs in their highly proliferative nature and dysregulated differentiation programs (Hemmati et al. 2003, Ignatova et al. 2002). The following studies took a further step to test their tumor initiating ability in vivo by injecting a small number of CD133+ cells into the brain of non-obese diabetic, severe combined immunodeficient (NOD-SCID) mice (Singh et al. 2004, Kelly et al. 2009). Less than 100 CD133+ cells were able to generate and phenocopy the original tumor, while in contrast, 105 CD133- cells could not cause a tumor. The expression of neuronal and astrocytic lineage markers within the
CD133+ xenograft indicated multilineage differentiation in vivo. Tumors generated from secondary transplantation of 1000 CD133+ cells derived from the primary xenograft indicated self-renewal capability in vivo. These NSC-like cells were defined as brain tumor initiating cells
(BTICs). Kelly et al used a different protocol to isolate unsorted potential BTICs in the absence of exogenous mitogenic factors, and they reduced the number of injected cells to 10 which generated tumors resembling human GBMs (Kelly et al. 2009).
16
The BTIC culture system using serum-free medium supplemented with growth factors well preserves their tumor initiating and self-renewal properties, which mirrors the in vivo biology of the original tumor when cells are transplanted into mice (Lee et al. 2006), suggesting a more reliable model to study GBMs than traditional cancer cell lines. However, a disadvantage of using growth factors for in vitro culture is that they may induce stem cell features in the highly plastic cancer cells rather than reflect their actual phenotypes in vivo. One study has implied that growth factors promoted neurosphere growth of isolated BTICs but were dispensable for their survival and self-renewal (Kelly et al. 2009). In this case, the stemness of BTICs is likely to be an intrinsic property and not from mitogen-induced epigenetic remodelling.
A clinically relevant property of BTICs is their relative insensitivity to radiation and chemotherapies, since these cells are able to circumvent genotoxic insults through being quiescent in the cell cycle and their enhanced capacity for DNA damage repair (Bao et al. 2006). Studies using in vivo lineage tracking and kinetic analysis indicated that BTICs and their progeny became predominant during the re-growth of tumors and were the main source for tumor recurrence after therapies (Chen et al. 2012, Gao et al. 2013).
1.3.3 Cell markers for CNS lineages and BTIC differentiation
During neurogenesis in early developmental stages, a small number of multipotent neuroepithelial cells in the neural tube give rise to proliferating progenitor cells, which subsequently generate a heterogeneous population of neurons and glial cells. Specific cell marker identifying different classes of neuronal cells are crucial to understand the molecular mechanisms of stem cell differentiation in the central nervous system (CNS) during early mammalian neurogenesis.
McKay’s group first immunized mice with human neuronal cell populations from a certain period
17 of development and obtained monoclonal antibodies that stained for cells of different types and different developmental stages (Hockfield and McKay 1985). The neural stem cell marker Nestin, was discovered by an antibody identifying a transient radial glial cell population, which was important in neuronal migration guidance (Hockfield and McKay 1985, Lendahl, Zimmerman, and
McKay 1990). Aside from neuroepithelial stem cells, Nestin is also seen in hair follicle sheath progenitor cells and human central nervous system tumors (Li et al. 2003, Tohyama et al. 1992).
Different stages of axon migration and maturation are marked by distinct cytoskeletal proteins. The stem cell specific class VI intermediate filament Nestin, will be replaced by vimentin and successively NF-L, NF-M and NF-H during axon development (Nixon and Shea 1992). The expression of class III ß-tubulin (Tubb3, Tuj1) appears in relatively late stages of neurogenesis, accompanied by the repression of the early stage intermediated filament protein vimentin (Butler,
Robertson, and Gallo 2000). Microtubule-associated proteins MAP5 and MAP2 were also identified specifically in differentiated neurons. Since they show a substantially higher expression during axon extension in developing neonatal brains compared to adult brains, they may play a role in modulating microtubule function during neurogenesis (Riederer, Cohen, and Matus 1986,
Caceres, Banker, and Binder 1986). Two differentiated types of macroglial cells, astrocyte and oligodendrocytes are specifically stained by anti-GFAP (Glial fibrillary acidic protein) (Bignami et al. 1972) and anti-O4 antibodies (Sommer and Schachner 1981).
The frequency and time-course of each cell type generated during development was well defined by using an in vitro differentiation model of embryonic stem cells (Fraichard et al. 1995).
After induction by retinoic acid, Nestin-positive precursors were first identified. Then, neuron-like cells positive for neuron-specific antigens (MAP2, MAP5, synaptophysin) emerged. A smaller proportion of oligodendrocytes and astrocytes stained for O4 and GFAP appeared shortly after
18 neurons. GABAergic and cholinergic neurons positive for glutamic acid decarboxylase (GAD) or acetylcholinesterase (AchE) activity, respectively, constitutes around 10-20 % of the neuronal-like cell population.
Because BTICs share many similarities with neural stem cells and molecular profiling of
BTICs suggests a stem cell or progenitor origin of tumorigenesis (Cusulin et al. 2015), markers specific for neural stem cells are also used for identification of BTICs, such as Nestin and CD133.
In addition, BTICs can differentiate into multiple lineages corresponding to those marked by
Tubb3, GFAP and O4. It is unclear whether differentiation of BTIC in vivo follows the route of neural stem cell specification during neurogenesis. However, it has been reported that GBM tissues displayed aberrant elevated levels of Nestin and Tubb3, which may be linked to malignancy
(Dahlstrand, Collins, and Lendahl 1992, Martinez-Diaz et al. 2003), suggesting these neuronal antigens may play distinct roles in tumor tissues compared to their functions in development processes.
1.3.4 The cellular origin of BTICs and GBM
Understanding the cellular origin of BTICs is critical to decipher the acquisition of their stemness and the predominant signaling pathways sustaining their self-renewal. The origin of BTICs is associated with the origin of tumorigenesis, which is investigated using mouse models. It has been implicated that gliomas could be induced from both neural stem cells and mature neural lineages, and the latter is achieved through cell dedifferentiation.
Neural stem cells and early progenitors intrinsically possess proliferation and self-renewal capability and higher cellular plasticity than mature neurons and astrocytes, therefore they are more permissive for tumor transformation. Transgenic mouse strains with p53 inactivation and
19
NF1 loss generated tumors exclusively in the SVZ regions, indicating an NSC/progenitor origin of gliomagenesis (Zhu et al. 2005). Introduction of PTEN heterozygosity into the p53/NF1 tumor suppressor model accelerated tumor formation from neural stem/progenitor cells, whereas the mature glial cells could not be transformed in this model (Kwon et al. 2008, Alcantara Llaguno et al. 2009). A recent study suggested that the adult lineage-restricted progenitors such as neural progenitors and oligodendrocyte progenitor cells were also capable of GBM transformation by p53/NF1/PTEN inactivation, and the generated tumors retained the features of original progenitor types (Alcantara Llaguno et al. 2015). This implies the diversity of GBM subtypes may result from the cell of origin rather than the divergent differentiation paths from the neural stem cell origin of tumor.
Tumor initiation from differentiated lineages usually requires a combination of oncogene activation with tumor suppressor inactivation to reverse the terminal differentiation state.
Activation of both RAS and AKT pathways were not able to induce transformation of mature astrocytes (Holland et al. 2000). However, loss of INK4a/ARF combined with KRAS activation allowed for gliomagenesis from astrocytes (Uhrbom et al. 2002). Similarly, EGFR activation could dedifferentiate and transform astrocytes in the INK4a/ARF-/- model (Bachoo et al. 2002). The protooncogene c-MYC and activated RAS and AKT pathways co-operatively cause astrocyte dedifferentiation and tumor induction (Lassman et al. 2004). These studies indicate that in the course of tumorigenesis, mature glial cells first involve a dedifferentiation process, which releases cells from the cell cycle arrest. The oncogenic events in these models were also shown to promote cell proliferation and dedifferentiation in culture.
20
1.4 Mechanisms of BTIC stemness
The stemness feature of BTICs can be hijacked from normal stem cells or induced by oncogenic mutations during malignant transformation, as noted below.
1.4.1 Pluripotency factors in BTIC regulation
The pluripotency state of cells can be induced in somatic cells using a reprogramming process by co-expressing four transcription factors, OCT4, SOX2, c-MYC and KLF4, which establish a particular molecular circuitry in pluripotent cells (Takahashi and Yamanaka 2006, Jaenisch and
Young 2008). These factors are also highly expressed in many types of cancers (Schoenhals et al.
2009, Guo et al. 2011). SOX2 is not only a master regulator in pluripotent cells, but is also a persistent marker for neural stem cells from early embryonic development throughout adulthood
(Ellis et al. 2004), and plays an important role in maintenance of NSCs (Wegner and Stolt 2005).
The SOXB1 group (SOX1, SOX2 and SOX3) proteins prevent premature cell cycle exit and neuronal differentiation, by sequestering proneural proteins and abolishing their DNA binding activity. SOX2 is strongly expressed in the SVZ and SGZ neurogenic regions in normal brains and has an overall high expression in gliomas, particularly in the proliferative stem cell regions
(Vanner et al. 2014). SOX2 maintains the tumor initiating ability of glioma stem cells and make the cells resistant to differentiation by regulating a group of stemness-related genes (Favaro et al.
2014, Berezovsky et al. 2014). Unlike SOX2, OCT4 is not present in normal NSCs but is activated in gliomas through DNA hypomethylation of its promoter (Shi et al. 2013, Wu et al. 2016).
Consistent with its function in embryonic stem cells, OCT4 can induce a stem cell-like state in
GBM cells and its expression is correlated with poor prognosis and tumor recurrence (Lopez-
Bertoni et al. 2015, Rodini et al. 2012). The pluripotency factors, Nanog and c-MYC, have also
21 been shown to induce stem-like features and tumor initiating capabilities in mature astrocytes
(Moon et al. 2011, Lassman et al. 2004).
1.4.2 Neurodevelopmental factors in BTIC regulation
Gliomagenesis is believed to be associated with deregulated neurogenesis programs. Some neurodevelopment regulators have also been indicated to affect the stem cell properties of BTICs.
The basic helix-loop-helix (bHLH) factors in the inhibitor of differentiation (ID) and hairy and enhancer of split (HES) families are required to maintain the proliferative state and self-renewal of neural progenitors (Nam and Benezra 2009, Kageyama and Ohtsuka 1999). During neocortex development, ID and HES proteins are downregulated with a concurrent increase of expression of neuronal lineage-associated bHLH factors (Neurog1, Neurog2, Ascl1)(Ross, Greenberg, and Stiles
2003). ID1 is a transcription repressor, which inhibits the expression of the GTPase activating protein for RAP1 (Rap1GAP) and maintains RAP1-mediated cell adhesion that is critical for the anchorage of neural stem cells to their niche (Lasorella, Benezra, and Iavarone 2014). This ID-
RAP1 cell adhesion axis was shown to maintain the aggressive Mesenchymal subtype of GBMs
(Niola et al. 2013). Notch1 is a well-known factor for neural stem cell maintenance. The Notch1-
HES signaling pathway antagonizes the proneural bHLH factor Ascl1 to inhibit differentiation
(Kageyama and Ohtsuka 1999, Androutsellis-Theotokis et al. 2006). Activation of the AKT pathway by Notch signaling was seen in both normal neural stem cells and glioma cells, which enhanced cell proliferation activity (Zhao et al. 2010). These studies indicate that neural stem cells and BTICs are driven by some common signaling pathways. On the contrary, ectopic expression of neuronal lineage regulators, such as BMP4, Neurog2 and SOX11 (Gomes, Mehler, and Kessler
2003, Hide et al. 2009, Su et al. 2014), promotes BTIC differentiation and inhibits tumor progression
22
1.4.3 Mitogenic pathways and tumor suppressors in BTIC regulation
Epidermal growth factor receptor (EGFR) and fibroblast growth factor receptor (FGFR) signaling pathways are required for maintaining the self-renewal of neural stem cells in vitro (Reynolds and
Weiss 1992, Shimazaki, Shingo, and Weiss 2001). These pathways can even induce a stem cell- like state in non-stem cells, which leads to cell dedifferentiation and malignant transformation as discussed above. The very high frequency of mutations in the RTK signaling pathways in GBM suggests that mitogen-driven stem cell properties may underlie the highly malignant and aggressive nature of GBMs. The effect of growth factors (EGF, FGF, PDGF) on BTIC self-renewal include increased proliferation and inhibition of lineage differentiation. EGFR signaling blocks neuronal differentiation whereas PDGFR activation inhibits differentiation to oligodendrocytes and astrocytes (Ayuso-Sacido et al. 2010, Jiang et al. 2011). However, they all seem to block neurogenesis and promote the expansion of OLIG2 positive transit amplifying progenitors which can give rise to oligodendrocytes, glia cells and certain types of neurons (Gonzalez-Perez et al.
2009, Jackson et al. 2006, Ligon et al. 2006).
The downstream PI3K/AKT and RAS/ERK pathways were also shown to regulate cell fate determination during neurodevelopment. The RAS/ERK signaling controls a Neurog2-ASCL1 genetic switch, that promotes gliogenesis at the expense of neurogenesis. At high levels of ERK activation, ASCL1+ progenitors are biased to glial lineages and initiate astrocytomas, whereas at moderate level of ERK activation, ASCL1 promotes both neuronal and glial differentiation and causes gliomas with intermixed components (Li et al. 2014). ERK promotes expression of the neural progenitor marker OLIG2 and gliogenesis. This effect is antagonized by a negative regulator of the RAS signaling, Neurofibromin 1 (NF1) (Wang et al. 2012). The PI3K/AKT pathway also plays an important role in sustaining stem cell survival. It was shown that AKT could
23 phosphorylate OCT4 to promote a proliferative stem-like feature in U87 GBM cells (Zhao et al.
2015). In contrast, the PI3K specific phosphotase, Phosphatase and tensin homolog (PTEN), has an opposite effect to promote neuronal differentiation and the survival of mature neurons
(Lachyankar et al. 2000). PTEN inactivation enhances PI3K pathway and promotes an undifferentiatied state in neural cells leading to gliomagenesis (Zheng et al. 2008).
The tumor suppressors in the RB signaling axis (p16INK4a, p19ARF, p53) are involved in the regulation of neural stem cell proliferation and responsible for declined neurogenesis activity with age. Loss-of-function of these genes is frequently seen in high grade gliomas (Brennan et al.
2013). Loss of p16INK4a and p53 induces a higher rate of proliferation and rescues the age-related decrease of self-renewal potential in SVZ neural progenitors (Molofsky et al. 2006, Meletis et al.
2006). Some of the tumor suppressors are associated with the terminal differentiation state of brain cells. For example, combined loss of p16INK4a and p19ARF, but not p53, causes dedifferentiation of astrocytes under activation by EGFR signaling (Bachoo et al. 2002). p16INK4a and p19ARF synergize to maintain the terminal differentiation state in neural cells. In contrast, the p53-p21 axis does not seem to facilitate neural lineage differentiation. Loss of p53 induces proliferation as well as rapid differentiation in neural progenitors (Gil-Perotin et al. 2006), and p21 deletion causes a long-term exhaustion of NSCs after a transient propagation (Kippin, Martens, and van der Kooy
2005). This indicates that p53 and p21 may function to maintain the quiescence state of neural progenitors rather than to promote differentiation. OLIG2 is universally expressed in malignant gliomas and promotes a proliferation-competent state of cells, probably through its function to suppress p21 expression (Ligon et al. 2007).
24
1.4.4 Epigenetic regulation of BTICs
Epigenetic mechanisms are an important layer of gene expression regulation that include DNA methylation, post-translational modifications of histones, incorporation of histone variants, nucleosome positioning, and functions of non-coding RNAs. TCGA genome-wide sequencing revealed the interplay of genetic and epigenetic alterations in the pathogenesis of GBM. For instance, the mutation status of IDH and H3F3A (a hotspot in pediatric GBMs) defines DNA methylation subgroups, which are correlated with differential expression levels of transcription factors (OLIG1/2, FOXG1) (Sturm et al. 2012). The G-CIMP DNA hypermethylation pattern induced by IDH mutations disturbs the boundary between topological chromatin domains and causes aberrant activation of PDGFRA expression (Flavahan et al. 2016). Besides the DNA methylation abnormality in GBM, 46 % of GBM samples harbor mutations associated with chromatin modification genes (CMG), although the functional relevance of most of these genes regarding the biology of GBM remains obscure.
From pluripotent cells to somatic stem cells, epigenetic mechanisms play an important role in stemness maintenance and cell fate determination consistent with Waddington's epigenetic landscape model (Ladewig, Koch, and Brustle 2013). Embryonic stem cells (ESCs) and induced pluripotent stem cells (iPSCs) are characterized by hyperdynamic chromatin structures, with prevalence of the bivalent mark (combined H3K4me3 and H3K27me3) (Bernstein et al. 2006) and histone hyperacetylation modification (Melcer et al. 2012, Markowetz et al. 2010). These chromatin features are critical for their plasticity and responsiveness to various developmental stimuli. Similarly, the multipotent NSCs have higher epigenetic plasticity than their differentiated progeny and they also prime the lineage specific genes with bivalent marks. During neurodevelopment, these genes will either be activated or suppressed by loss of bivalency with
25 retained monovalent H3K4me3 or H3K27me3 marks (Burney et al. 2013). A study of the ESC- derived neural differentiation trajectory shows during each lineage commitment event there is a gain of DNA methylation indicating the silencing of stem cell-associated genes, whereas the differentiation-associated genes are "primed" with active H3K27ac or H3K4me marks on the transcription factor binding sites prior to their activation in the next developmental stage (Ziller et al. 2015). It is possible that the mature neural lineages can obtain a stem-cell like phenotype if the sequential epigenetic events during neural development are reversed.
The epigenetic landscape is distinct between BTICs and differentiated GBM cells and is involved in maintaining BTIC identity and preventing interconversion between the two cell types.
It was suggested that BTICs failed to undergo terminal differentiation in response to BMP, due to resistant chromatin loci occupied by SOX proteins and incomplete DNA methylation reconfiguration (Caren et al. 2015). This epigenetic barrier between BTICs and non-stem GBM cells can be traversed through a process resembling iPSC reprogramming by introducing four transcription factors (SOX2, OLIG2, POU3F2, SALL2) into the differentiated GBM cells and epigenetically converting them into BTICs (Suva et al. 2014). However, how the BTIC-specific epigenetic landscape is established and maintained through epigenetic modifiers is largely unknown. Most studies of epigenetic regulators in BTICs show loci-specific roles instead of global effects of specific modification types. The polycomb repressive complex-2 (PRC2), catalyzing the repressive mark H3K27me3, is a well-known regulator for stem cell maintenance. It can target an overlapping group of genes in embryonic stem cells and BTICs, which promote differentiation
(Signaroldi et al. 2016). Particularly, PRC2 silences the factors in the BMP signaling pathway to make BTICs resistant to differentiation stimuli (Signaroldi et al. 2016, Lee et al. 2008). PRC2 affects epigenetic plasticity of BTICs and is required for bidirectional conversion between BTICs
26 and differentiated cancer cells, through targeting both stemness and differentiation-promoting factors (Natsume et al. 2013). Bmi1 is the catalytic component of another polycomb group complex (PRC1) which results in monoubiquitination on histone H2A. Bmi1 maintains self- renewal in both neural stem cells and glioma stem cells by silencing the p16INK4a/p19ARF locus and p21WAF1 gene (Fasano et al. 2007, Bruggeman et al. 2007). The mixed-lineage leukemia family members MLL1 and MLL5 were shown to promote stemness of BTICs through two different mechanisms. MLL1 induced by hypoxia activates the expression of hypoxia-inducible factor 2a (HIF2a) and its target gene vascular endothelial growth factor (VEGF) by catalyzing
H3K4me3 modification, in contrast MLL5 with defective histone methyltransferase activity, induces local chromatin condensing and promotes BTIC stemness properties (Gallo et al. 2015).
So far, how histone acetylation is involved in BTIC maintenance has been poorly understood possibly due to a lack of knowledge about the substrate specificity of histone acetylation modifiers.
It was implicated that the global histone acetylation level was positively regulated by phosphorylated AKT, which increased the nuclear acetyl-coenzyme A to coenzyme A ratio (Lee et al. 2014). It was also noted that GBMs had a higher level of total H3 acetylation than normal tissues (Nagarajan and Costello 2009). Deregulated histone acetylation enzymes were frequently found in gliomas as reviewed in (Spyropoulou et al. 2013). Regarding each specific histone acetylation modifier in GBM, the function largely depends on the targeted genes, which should be taken into consideration when applying anti-tumor strategies like HDAC inhibition. For example, histone acetylation may suppress glioma cell proliferation when p21WAF1 is induced by hyperacetylation and the application of HDAC inhibitor would prove beneficial (Kamitani et al.
2002, Wetzel et al. 2005). However, histone acetylation may also promote self-renewal when
27
EGFR is activated by p300 (Erfani et al. 2015). Overall, there is still a paucity of knowledge about the functions of specific histone acetylation enzymes in BTIC and GBM development.
28
1.5 The INhibitor of Growth (ING) family of epigenetic regulators
The mechanisms behind cell growth is tightly linked with genomic integrity surveillance to prevent oncogenesis. The ING (INhibitor of Growth) family of proteins have been discovered to play a pivotal role in sensing and reacting to cellular stresses and further influencing cell cycle progression and programmed cell death under genotoxic and senescent conditions (Vieyra,
Toyama, et al. 2002) (Garkavtsev et al. 1998). Changes in the levels of INGs are commonly seen in multiple types of cancer, and appear to have a causal relationship with the proliferation and invasion properties of cancer cells. The ING family members were initially recognized as the type-
II tumor suppressors, frequently dysregulated or mislocalized tumor suppressors in cancer.
However, increasing lines of evidence implicate the INGs in playing more complex and diverse roles in cancer depending on cell type and genomic/epigenomic context. INGs localized in the nucleus interact with factors involved in DNA replication, DNA repair and the chromatin remodeling machinery, with their chromatin-binding and protein-protein interacting domains
(Tallen and Riabowol 2014). INGs are also capable of participating in the transduction of various proliferation- and stress-related signals by shuttling between cytoplasm and nucleus.
1.5.1 Major functions of ING1 in growth regulation and cancer biology
The first gene of the ING family was discovered and cloned in a study screening for candidate tumor suppressor genes using subtractive hybridization between several breast cancer cell lines and normal mammary epithelial cells (Garkavtsev et al. 1996). The identified gene encodes the 33 kDa ING1 protein, which prevents tumor transformation and causes G1 phase arrest when overexpressed. Consistent with its tumor suppressive function, the expression of ING1 increases in senescent diploid fibroblasts and inhibition of ING1 extends the population doubling capability of normal diploid fibroblasts (Garkavtsev and Riabowol 1997). ING1 can also sensitize normal
29 and cancer cells to apoptosis and promote DNA damage repair mechanisms under various stress conditions through p53-dependent and independent pathways, suggesting a critical mechanism for how it prevents tumor transformation and protects genomic integrity (Scott, Boisvert, et al. 2001,
Cheung and Li 2002, Bose et al. 2013, Feng, Bonni, and Riabowol 2006). A series of studies were conducted to examine the genetic and expression alterations of ING1 in various cancer types.
Somatic mutations in the ING1 gene is relatively rare in esophageal squamous cell carcinomas
(Gunduz et al. 2000), breast cancers (Toyama et al. 1999) and gastrointestinal carcinoma (Oki et al. 1999), but occur with a high incidence (20 %) in melanoma, consistent with DNA damage repair serving as a major etiological factor in this cancer type (Campos et al. 2004). Compared to genetic alterations, downregulation of ING1 expression is very commonly seen in a wide range of cancers (Nouman et al. 2002, Walzak et al. 2008, Tokunaga et al. 2000, Yu et al. 2004).
Remarkably, ING1 gene knockout leads to vulnerability of mice to radiation insults and higher incidence of lymphomas, indicating a critical function of ING1 in genome stability maintenance especially in the lymphatic system (Kichina et al. 2006).
ING1 is associated with the p53/p21-RB axis in many critical functions, including growth control (Garkavtsev et al. 1998), apoptosis (Shinoura et al. 1999) and senescence (Soliman et al.
2008, Rajarajacholan, Thalappilly, and Riabowol 2013), through direct interaction with p53 or by regulating its target genes. ING1 activates p21 expression through histone acetylation (Kataoka et al. 2003). ING1 can also stabilize p53 at the protein level by promoting acetylation and preventing polyubiquitination-meidated degradation (Soliman et al. 2008, Thalappilly et al. 2011). ING1, specifically the ING1a isoform, also induces the expression of p16INK4a in an age-dependent manner to cause the senescent phenotype (Soliman et al. 2008). It has been shown that ING1a
30 increased the expression of p16INK4a and RB through the endocytosis-related factor Intersectin
2 (ITSN2) (Rajarajacholan, Thalappilly, and Riabowol 2013).
Functions for ING1 in malignant gliomas have been indicated in several reports. The mRNA level of ING1 is reduced in gliomas, which is negatively correlated with tumor grades (Tallen et al. 2004). Moreover, there is a frequent mislocalization of ING1b to the cytoplasm in brain tumor tissues, which may abolish its tumor suppressor functions in this tumor type (Vieyra et al. 2003).
ING1 is involved in the sensitivity of GBM cells to drug-induced apoptosis (Tallen et al. 2008), and synergizes with the effect of the HDAC inhibitor TSA in p53/p16 deficient tumors, by activating the proapoptotic proteins caspase 3 and Fas-associated death domain (FADD)
(Tamannai et al. 2010).
1.5.2 Structural features of the ING proteins
Phylogenetic analysis indicated ING1 to ING5 were all present in vertebrates, and homologs of
INGs were highly conserved across organisms from plants, fungi, invertebrates to vertebrates (He et al. 2005). On the carboxyl terminus, there are plant homeodomains (PHD, a form of zinc finger), which interact with post-translationally modified core histone proteins found mostly in chromatin- regulatory proteins. The PHD motif of INGs has the highest affinity for the trimethylated H3 lysine
4 residue (H3K4me3) (Pena et al. 2006), a histone mark usually present on active promoters. INGs possess a nuclear localization signal (NLS) region adjacent to the PHD. In case of ING1, two nucleolar targeting sequences are embedded within the NLS (Scott, Bonnefin, et al. 2001).
Deletion of the NLS region abolished nuclear localization and the ability to regulate gene transcription of INGs (Russell et al. 2008).
The lamin interaction domain (LID) is a unique feature of the ING family across the human proteome. The interaction between lamin A and ING1 is for the nuclear localization and the
31 function of ING1 in apoptosis, while disrupted lamin-ING1 interaction causes phenotypes reminiscent of Hutchinson-Gilford progeria syndrome (Han et al. 2008). Interestingly, embryonic stem cells (ESCs) are devoid of lamin A and start expressing it when they are differentiating.
Forced expression of lamin A leads to chromatin rigidity and reduced plasticity in ESCs (Melcer et al. 2012). Whether ING proteins participate in lamin A-induced chromatin remodeling during differentiation is an intriguing question.
The leucine zipper-like (LZL) region is present in all INGs except for ING1, and contributes to the capability of ING2 to induce muscle differentiation (Eapen et al. 2012). Some structural features are present only in a subset of ING alternative splicing products, such as the PCNA-
Interacting Protein (PIP) motif in ING1b which is important for UV-induced DNA damage response (Scott, Bonnefin, et al. 2001), and the polybasic region (PBR) in ING1 and ING2 isoforms that binds to stress-inducible phosphatidylinositol monophosphate (PtIn-MP) (Kaadige and Ayer 2006) and also serves as a ubiquitin interaction motif or UIM (Thalappilly et al. 2011).
1.5.3 Epigenetic functions of ING proteins
Chromatin remodeling capability is shared among ING members, since they all possess a plant homeodomain that specifically targets H3K4me3 (Shi et al. 2006) and all INGs are incorporated in to one or more histone acetylation complexes. Yeast homologs of the ING family, Pho23, Yng1, and Yng2 respectively interact with Sin3-HDAC, NuA3 HAT and NuA4 HAT (Howe et al. 2002,
Loewith et al. 2000). In human cells, ING1/2 were identified as stoichiometric components of the mSin3A-HDAC complex, while ING3/4/5 are associated with the MYST family of HATs (Tip60,
HBO1, MOZ/MORF) (Kuzmichev et al. 2002, Doyon et al. 2006). The ING1b isoform also interacts with a group of HATs, such as TRRAP, PCAF and p300, whereas ING1a mainly binds to HDAC1 and inhibits histone acetylation (Vieyra, Loewith, et al. 2002). INGs may play an
32 important role to target the histone modification enzymes to the vicinity of chromatin through interaction with the DNA-binding protein PCNA or targeting the histone mark H3K4me3, and further induce a DNA repair cascade or activate gene expression (Vieyra, Loewith, et al. 2002,
Schafer et al. 2013).
ING5 has been shown to interact with the HBO1, MOZ and MORF HAT complexes (Doyon et al. 2006). Analysis of the molecular architecture of the quartet MOZ/MORF complex indicates
ING5 may not directly bind to the MOZ/MORF histone acetyltranferases, but facilitate the assembly of the complex by enhancing the affinity between BRPF1 and EAF6, two other components in the complex. Moreover, ING5 drastically stimulates the catalytic activity of HATs and elevates acetylation levels on H3K4me3-modified H3, suggesting the functional importance of the histone-targeting function of ING5 through its PHD motif (Ullah et al. 2008). The ING5-
MOZ/MORF-HATs has been shown to modify both H3 and H4, while the ING5-HBO1-HAT switches substrate between H3 and H4, depending on which scaffold protein is incorporated
(Lalonde et al. 2013). ING5 can form a complex with p53 and the Tip60 HAT specifically under
DNA damage condition, which causes p53 K120 acetylation and activation of its target genes BAX and GADD45 (Liu et al. 2013).
1.5.4 Functions of ING5 in cancer
ING5 was first identified from a human genome search for ING homologous genes (Feng, Hara, and Riabowol 2002) (Shiseki et al. 2003). It was indicated to interact with and modulate p53 in the DNA damage response and apoptosis (Shiseki et al. 2003, Liu et al. 2013) and may link p53- dependent stress response with HBO1-mediated DNA replication licensing by bridging the two factors to the replication origin sites (Iizuka et al. 2008). ING5 functions in the DNA damage response and in apoptosis appear to generate opposite effects in different cancer types, promoting
33 therapy sensitivity in bladder cancer cells (Li et al. 2015) while causing chemoresistance in the gastric cancer cells (Gou et al. 2015). This suggests that the function of ING5 may be highly context-dependent. Besides a tumor suppressive role, ING5 was also indicated to enhance proliferation in multiple cancer cell lines (MCF7, HCT116, U2OS, HeLa) (Doyon et al. 2006,
Linzen et al. 2015), probably through the HBO1-mediated DNA replication initiation mechanism.
The genetic alteration of ING5 is not frequent in tumorigenesis, with the highest rate of all abnormalities being seen in pancreatic cancer (15.6 %) and the highest mutation rate of 2.8 % in colorectal cancer, according to TCGA data. The alteration of ING5 levels varies among different cancers and a discrepancy between ING5 mRNA and protein levels exist in several cancer types
(Xing et al. 2011, Zheng et al. 2011). The primary function of ING5 in different cancers is complicated by its subcellular localization shift. The five identified splicing variants of ING5, two of which lack the NLS region, may also account for the variability in its localization (Cengiz et al.
2010). It was suggested that the nuclear level of ING5 protein was negatively correlated with tumor progression and grades, whereas cytoplasimc ING5 correlated in an opposite manner (Zheng et al.
2011, Li et al. 2010). Despite these observations, the genetic and epigenetic mechanisms of ING5 in cancer development remain obscure.
1.5.5 ING5 in stem cell maintenance
ING proteins, as essential growth regulators and epigenetic factors, have been implicated in regulation of cell differentiation, germ cell development and cell reprogramming. ING2 is required for myogenic differentiation mediated by the activity of Sin3A-HDAC1 complex (Eapen et al.
2012). ING2 is also required for normal spermatogenesis by regulating chromatin acetylation in spermatocytes (Saito et al. 2010). ING3 regulates oocyte maturation by facilitating the asymmetric division of oocytes (Suzuki et al. 2013). Moreover, ING3 is highly expressed in metaphase II
34 oocytes and was identified as a potential oocyte reprogramming factor (Awe and Byrne 2013).
ING4 drives prostate epithelial cell differentiation and its disruption stimulates prostate tumorigenesis (Berger et al. 2014).
ING5 has been shown by several studies to potentially regulate stem cell maintenance.
Knockdown of ING5 in epidermal stem cells increased cell differentiation and ING5 was identified as a factor in the core epigenetic network that maintained epidermal stem cell self-renewal (Mulder et al. 2012). ING5 ChIP-sequencing showed genome-wide occupancy of ING5 on promoters of genes marked by H3K4me3, which are subsequently silenced during differentiation. This suggests
ING5 may target the HAT complexes to H3K4me3-modified chromatin and promote gene activation. Moreover, upregulation of ING5 levels was repeatedly seen in pluripotent cells compared to differentiated lineages (Vlismas et al. 2016, Hussein et al. 2014), suggesting it may function as a common factor for cell plasticity in various stem cell systems. Consistently, our preliminary data shows that ING5 is highly expressed in mouse embryonic stem cells (ESCs) and dramatically decreased at the mRNA and protein levels during differentiation into the embryoid body in three individual ESC lines (Figure 1).
The MYST family of HATs are important epigenetic regulators in stem cells and development.
The frequent MOZ translocation in acute myeloid leukemia is associated with oncogenesis, suggesting its critical role in development of the hematopoietic system. MOZ is required for hematopoietic stem cell maintenance through its HAT activity (Thomas et al. 2006, Perez-Campo et al. 2009), whereas the function of MORF, a paralog of MOZ, is specific to maintaining neural stem cells/progenitors in mammals (Merson et al. 2006, Sheikh et al. 2012). HBO1 is critical in embryonic development and responsible for more than 90% of H3K14ac in embryos, which
35
(Figure 1A and 1B: Vladislav Alekseev & Karl Riabowol)
Figure 1. ING5 expression decreases during embryonic stem cell differentiation (A and B from Vladislav Alekseev).
36
Figure 2. ING5 expression decreases during embryonic stem cell differentiation (A and B from Vladislav Alekseev). (A) mRNA level of ING5 before and after embryoid body differentiation in three mouse embryonic stem cell lines, E14, R1 and D3. E14 cells are shown in the photomicrographs. (B) Protein levels of ING5 and pluripotency factor OCT4 before and after embryoid body differentiation. (C) Positive and negative controls for ING5 protein detection. The whole blot incubated with ING5 antibody was shown on the left and the amount of total protein was shown by Amido black staining. The major band between 28k Da and 35 kDa indicates ING5 protein with a molecular weight of 28-29 kDa (240 aa). For the positive control, iPB stable overexpression cells were established from the BT 189 cell line. Cells were grown as suspended spheres and upon induction, dissociated into single cells and treated with 1x (30 µg/mL) cumate for 72 hours before the spheres were directly lysed in Laemmli sample buffer. For the negative control, shRNA stable cells were established from the BT 189 cell line. Cells were grown as suspended spheres and upon induction, dissociated into single cells and treated with 100 ng/mL doxycycline for 7 days (cells were passed once at the fourth day) before the spheres were directly lysed in Laemmli sample buffer.
37 controls expression of embryonic patterning genes (Kueh et al. 2011). MOZ and HBO1 homozygous mutations are lethal whereas MORF mutants exhibit severe defects in the central nervous system. Moreover, the deletion of BRPF1 which, like ING5, binds to MOZ, MORF, and
HBO1 acetyltransferases, causes more severe developmental abnormality including vascular, hematopoietic defects and neural tube closure failure leading to embryonic lethality (You et al.
2015). These studies implicate MYST acetyltransferases and HAT subunits as essential regulators for development and stem cell maintenance, and hint at a potential role for ING5 in these processes.
38
1.6 Objectives
The major aim of this project is to determine the role of ING5 in the stemness features of BTICs, a stem cell-like population in glioblastoma. As a stoichiometric component of three HAT complexes, ING5 has been shown to modulate the activity and substrate targeting of histone acetyltransferases to potentially regulate gene expression. However, how ING5 affects gene expression by histone modification and the genes targeted by ING5 remain unclear. Several studies and our preliminary data indicate a potential function of ING5 in stem cell maintenance, but the mechanisms by which ING5 affects stemness regulation have not been directly investigated so far.
BTICs are supposed to contribute to tumor initiation and recurrence in GBM, the highly aggressive and devastating form of brain tumor. The role of ING5 in cancer stem cells and in GBM has not been reported. Considering the function of ING5 in stem cell regulation, it would be interesting to investigate its role in stemness maintenance of BTICs and the underlying molecular mechanisms.
We hypothesized that ING5 might regulate the stem cell properties of BTICs through epigenetic mechanisms, and intended to reveal the target genes and signaling pathways of ING5 in this process, therefore contributing to our understanding of the stem cell phenotypes of BTIC, so we also aimed to identify therapeutic strategies to target these cells in GBM treatment.
39
CHAPTER TWO: MATERIALS AND METHODS
2.1 BTIC cultures and sphere formation assays
BTIC lines BT 12, BT 134 and BT 189 were isolated and established from adult GBM patients as described previously (Kelly et al. 2009). Cells were cultured as neurospheres in either serum free medium (SFM) alone (BT 12) or SFM supplemented with 20 ng/mL EGF (Peprotech), 20 ng/mL
FGF (R&D Systems) and 2 µg/mL heparin solution (STEMCELL) (BT 134 and BT 189) depending on the growth requirement for each BTIC line. Cells were dissociated with Accutase solution (Innovative Cell Technologies) and plated at a density of 20,000 cells/mL. For sphere formation assays, 200 viable single cells were seeded in each well of 96-well plates and were grown for 12-14 days. 30 µL of fresh medium was added to the existing medium every three days.
The complete field of each well was captured under a 4x microscope objective and the diameter of each sphere measured. At least three wells (600 cells) were counted for each condition in one experiment. Sphere formation assays with treatment of protein kinase inhibitors, calcium modulators, FSHR blocking antibody and other reagents were conducted in the absence of growth factors. BT 12 and BT 134 cells were used in one experiment shown in Figure 3D. Except for
Figure 3D, all experiments, including transient ING5 manipulation and stable cell line construction were done with BT 189 cells. Serial sphere assays were conducted by harvesting all of the cells from the last passage and replating them into new wells at the same dilution.
2.2 Sphere cell differentiation and immunofluorescence
Dissociated BTICs were plated on poly-L-ornithine coated glass coverslips (coated with 20 µg/mL poly-L-ornithine water solution) and induced to differentiate in SFM supplemented with 1 % fetal
40 bovine serum (Invitrogen) for 5-7 days. The medium was replaced once at the fourth day by fresh medium containing 1 % fetal bovine serum. The stem cell and lineage markers were detected by immunofluorescence (IF). For IF experiments, cells were fixed with 4 % formaldehyde for 15 minutes, permeabilized with 0.5 % Triton X-100 for 3 minutes and then blocked in 3 % BSA for
1 hour at room temperature. Coverslips were then incubated with the primary antibodies against
Nestin (1: 800; MAB-1259, R&D Systems), Tubb3 (1:400; MRB-435P, Biolegend), GFAP
(1:200; Z0334, Dako) or FSHR (1:30, MAB65591, R&D Systems) at 4 °C for 16 hours. Coverslips were washed twice with 1x PBS (phosphate buffered saline) gently between each step. After incubation with fluorophore-conjugated secondary antibodies (Invitrogen) for 1 hour at 37 °C, coverslips were stained with DAPI or Hoechst 33258 (Sigma) and examined under an Olympus
IX71 wide-field microscope.
2.3 pCI plasmid transfection
The pCI-ING5 plasmid was constructed by cloning ING5 cDNA into the pCI empty vector
(Promega) using Not I and Kpn I restriction sites. For BTIC transfection, lipofectamine 2000
(Invitrogen) reagent was used according to the manufacturer’s instructions. BTICs were seeded at a density of 12-16 × 104 /mL on poly-L-ornithine-coated plates 16-24 hours prior to transfection.
Transfection was performed in serum free medium in the absence of growth factors. For 24 well transfection, 100 µL transfection complex (containing 750 ng DNA and 1.8 µL transfection reagent) was added to 500 µL existing medium. The medium was replaced with fresh SFM 6-8 hours after transfection and growth factors were added one day after transfection if required for the following experiment. 2 days after transfection, cells were collected or dissociated first into single cells for the following experiments (Figure 2A).
41
Figure 2. Protocols for transient and stable ING5 overexpression and knockdown experiments in BT 189 cells. (A) Time axis showing the time points of each step before and after transfection with pCI plasmids and siRNAs. (B) Schematic of the iPB vector and time axis showing the time points for iPB cell line establishment, selection and induction. (C) Schematic of
42 the pINDUCER10 vector and time axis showing the time points for shRNA cell line establishment, selection and induction.
2.4 PiggyBac stable cell lines
Stable overexpressing cell lines were established using the PiggyBac transposon system (Systems
Biosciences). The whole length ING5 cDNA was cloned directly from pCI-ING5 vector into the inducible PiggyBac vector (PBQM530A-1 cumate switch, Systems Biosciences) using the NotI and NheI restriction sites. The PBQM530A-1 vector has a size of 9.5 kb. The major elements include (from 5’ to 3’ direction) a CMV-CUO switch element, a multiple cloning site (MCS), an internal ribosome entry site (IRES), a copGFP tag, a CymR (repressor) and a puromycin resistance gene (PuroR) (Figure 2B).
The copGFP protein has a shorter emission wavelength (max 502 nm) than the traditional GFP and is almost not detectable under the filter of widefield fluorescence microscope (528 ± 19 nm) and the FITC channel of flow cytometer (max 525 nm). In the experiments with Fluo3-AM staining
(emission max 526 nm), the background intensity of iPB stable cell lines is as low as BT 189 cells.
The inducible expression is regulated by the cumate switch. The constitutively expressed CymR repressor binds the cumate operator sequences (CUO) with high affinity. The repression is alleviated through the addition of cumate which binds to CymR.
The PiggyBac transposon is a mobile genetic element which transposes from the vector to chromosomes through a “cut and paste” mechanism. The transposase recognizes inverted terminal repeat sequences (ITRs) located on both ends of the transposon vector and moves the contents between the two ITRs into TTAA chromosomal sites. The PiggyBac vector and the plasmid encoding a transposase (PB210PA-1, Systems Biosciences) were co-transfected into BT 189 cells
43 by electroporation. Two days after transfection, 350 ng/mL puromycin was added to cells and the medium was changed every three days containing 350 ng/mL puromycin. After 10-14 days, positive clones were pooled to make stable overexpressing cells. Expression of ING5 was induced by cumate (Sigma) at 30 µg/mL (1x) for 2-3 days before further experiments were conducted
(Figure 2B).
For the construction of iPB-ING5-FLAG, a FLAG tag (DYKDDDDK) was added to the c- terminus of full-length ING5 during the PCR amplification of the ING5 coding sequence from pCI-ING5 vector. For the iPB-ING5-ΔPHD construction, the PHD motif on the c-terminus (186-
240 aa) was excluded from ING5 and a FLAG tag was added to the c-terminus of truncated ING5 during the PCR amplification from pCI-ING5 vector. The sequence was confirmed by DNA sequencing. Then the PCR products were double digested with NotI and NheI and inserted into empty PBQM530A-1 vector as described above.
2.5 siRNA transfection
For knockdown experiments, the following siRNA targeting ING5 was designed and synthesized: siR-ING5 (sense, GGAAUACAGUGACGACAAATT; anti-sense,
UUUGUCGUCACUGUAUUCCTT). si-ING5 and scrambled control siRNA were transfected using lipofectamine 2000 according to the manufacturer’s protocol. BTICs were seeded at a density of 10 × 104 /mL on poly-L-ornithine-coated plates 16-24 hours prior to transfection.
Transfection was performed in serum free medium in the absence of growth factors. For 24 well transfection, 100 µL transfection complex containing 1.8 µL transfection reagent was added to 500
µL existing medium. The final concentration of siR-ING5 and siR-ctr added to cells was 200nM.
The medium was replaced with fresh SFM 6-8 hours after transfection and growth factors were
44 added one day after transfection if required. 2 days after transfection, cells were collected or dissociated first into single cells for the following experiments (Figure 2A).
2.6 lentiviral-based shRNA system
The following two shRNA sets targeting ING5 were designed and synthesized: shRNA-1 (targeted sequence ATCAGCTGGAAGTTCCTCTGAA), shRNA-2 (targeted sequence
TTTATCCACCATCTCGTAGGTC). The mir30-styled shRNAs were synthesized as single stranded DNA oligos with common ends corresponding to part of the endogenous mir30 miRNA flanking sequence. These common sequences were used to prime a PCR reaction which included restriction sites for the following double digestion. The PCR products were then digested and ligated into the digested empty pINDUCER-10 lentiviral vector (Meerbrey et al. 2011).
The pINDUCER-10 is a bicistronic system with a constitutively expressed reverse tet- transactivator (rtTA) and puromycin resistance gene driven by the Ubc promoter, and a turboRFP- shRNA cassette driven by the tetracycline response element (TRE2) promoter (Figure 2C). Upon the addition of doxycycline, transcription of the turboRFP-shRNA cassette is activated.
For lentivirus packaging, the pINDUCER vector was co-transfected with four plasmids encoding for gag, pol, rev and vsv proteins. The supernatants from HEK293T cells were collected and virus particles isolated using an ultracentrifuge. BT 189 cells were infected with the virus particles. Two days after infection, 350 ng/mL puromycin was added to cells and the medium was changed every three days containing 350 ng/mL puromycin. After 2-3 weeks, positive clones were pooled to make stable overexpressing cells. Expression of shRNAs was induced by doxycycline at 100 ng/mL for
6-8 days before further experiments were conducted (Figure 2C).
45
2.7 Quantitative Real-Time PCR
Total RNA was isolated using TRIzol (Invitrogen) according to the manufacturer’s instructions.
Reverse transcription was conducted using the Applied Biosystems cDNA Reverse Transcription kit. Real-time qPCR reactions are carried out using Maxima SYBR Green qPCR Mastermix
(Fermentas) on the QuantStudio 6 Flex Real-Time PCR System (Applied Biosystems).
2.8 Gene expression microarray and data analysis shRNA expression was induced for 7 days with doxycycline before total RNA was extracted and genomic DNA removed using an RNeasy Mini Kit (QIAGEN). RNA quality was tested using a
Bioanalyzer (Agilent Technologies). The transcriptome assay was conducted using Affymetrix
Human Gene 2.0 ST arrays, performed in the facility of The Centre for Applied Genomics
(Toronto). The raw data was normalized and fold changes calculated using the Affymetrix software package (Expression Console, Transcriptome Analysis Console v3.0). Downstream function analysis was performed using IPA (Ingenuity Pathway Analysis)
(www.qiagen.com/ingenuity). Gene set enrichment analysis was performed using GSEA software
(www.broadinstitute.org/gsea/index.jsp).
2.9 Western blotting
Dissociated cells or spheres were lysed in Laemmli sample buffer, sonicated for 10 seconds and boiled for 10 min. The total protein was loaded and resolved on SDS-PAGE gels and transferred to nitrocellulose membranes. Membranes were blocked in 5% skim milk in PBS for 1 hour and then incubated with primary antibodies for 16-18 hours at 4 °C. Blots were then washed with PBS and incubated with horseradish peroxidase-conjugated secondary antibodies (Santa Cruz) for 1
46 hour. After the membrane was washed with PBS, proteins were visualized using enhanced chemiluminescence (ECL). Primary antibodies used were rabbit polyclonal anti-ING5 (1:2000,
10665-1-AP, Proteintech), rabbit monoclonal anti-pAKT (1:1000, 4058, Cell Signaling
Technology), rabbit monoclonal anti-AKT (1:1000, 4685, Cell Signaling Technology), rabbit monoclonal anti-p-p42/44 (1:1000, 4370, Cell Signaling Technology) and the same antibodies used for immunofluorescence (see 2.2).
2.10 Flow cytometry
For CD133 and CD44 staining, cells were dissociated, washed with PBS and then incubated with conjugated antibodies CD133-PE (1:100; 12-1338, eBioscience) and CD44-FITC (1:50; ab19622, abcam) at 4 °C for 30 minutes. Antibodies were diluted in cold PBS containing 2 % FBS and 1mM
EDTA. For isotype controls, cells were incubated with PE Mouse IgG1 kapa (400111, Biolegend) and FITC Rat IgG2b (400633, Biolegend) at matched concentrations. After washing with cold
PBS containing 2% FBS and 1mM EDTA, cells were re-suspended in 300 µL PBS and examined on an LSR II flow cytometer (BD Biosciences). Data analysis was performed using FlowJo software (Tree Star, Inc., Ashland, OR, http://www.treestar.com). For intracellular calcium level detection, cells were loaded with 5 µM Fluo3-AM for 45 min at 37 °C in dark, washed with PBS twice, and checked on the flow cytometer.
2.11 Cell division symmetry analysis
The division mode analysis was performed on mitotic pairs seeded at low density on coated coverslips in the absence of growth factors. iPB cells were synchronized using the double thymidine block method. After the second block with 1.5mM thymidine, cells were dissociated
47 and seeded at low density on poly-L-ornithine coated coverslips for 18-20 hours before being fixed and immunostained for Nestin as described in 2.2.
2.12 Live cell calcium imaging
Intracellular calcium levels were detected using the cell membrane permeable calcium dye Fluo3-
AM (Sigma). BTICs were dissociated and seeded on poly-L-ornithine coated plates at a density of
6-8 × 104 /mL on the previous day in the absence of growth factors. Attached cells were then incubated in medium containing 7 µM Fluo3-AM at 37 °C for 30 min. After washing with PBS, cells were de-esterified for 30 min in fresh medium (with no serum and growth factors), and examined under a fluorescence microscope. For co-transfection experiments, the calcium level was quantified by subtracting the initial fluorescence intensity from the fluorescence after excitation, using ImageJ software.
2.13 Chemicals
Chemicals used in this study included Tofacitinib, PX-866 and PD184352 (LC laboratories), ionomycin, Cyclopiazonic Acid, SKF96365, Nifedipine and Cyclosporin A (Alomone Labs), KN-
93 (Sigma) and BAPTA-AM (R & D Systems).
2.14 Statistics
All t tests were performed in GraphPad PrismTM for comparisons between two groups, and results were displayed as the mean ± SEM. Pearson correlation and Kaplan Meier analyses are performed using SPSS Statistics software (IBM Corp).
48
CHAPTER THREE: RESULTS
3.1 Determining the expression of ING5 during BTIC differentiation
According to our previous experiments in mouse embryonic stem cells, the expression of ING5 immediately decreases as ESCs differentiate towards the three germ layers. The decreasing trend is consistent in three independent ESC lines (Figure 1). Next we tested whether the expression of
ING5 was also altered in BTICs during differentiation. The tumor initiating ability and stemness properties of BTICs is best preserved in serum-free medium supplemented with growth factors in vitro (Lee et al. 2006), and upon serum treatment they gradually undergo lineage differentiation.
We used 1% fetal bovine serum (FBS) to induce differentiation in the BT 189 cell line and differentiated cells were identified by the neuronal marker β-tubulin class III (Tubb3) and the astrocyte marker glial fibrillary acidic protein (GFAP) (Figure 3). We observed decreased ING5 mRNA and protein levels within the first two days of serum-induced differentiation (Figure 4A,
4B). Immunofluorescence also showed ING5 reduction and a loss of nuclear localization in differentiated cells (Figure 4C). Since the genetic background varies greatly among BTIC lines, we tested the protein level of ING5 in two other cell lines, BT 12 and BT 134, both of which showed precipitous decline of ING5 during differentiation (Figure 4D). These results indicate that
ING5 is expressed at a higher level in stem cells than differentiated cells and therefore may play a role in maintaining stem cell properties of BTICs.
49
Figure 3. The BTIC in vitro differentiation model. Morphology of self-renewing spheres and immunofluorescence of neuronal (Tubb3) and glial (GFAP) lineage markers in differentiated cells.
Scale bar = 200 µm.
50
Figure 4. The expression of ING5 decreases during differentiation in BTIC lines.
51
Figure 5. The expression of ING5 decreases during differentiation in BTIC lines. (A) mRNA level of ING5 decrease during differentiation in BT 189 cells. N=4. Values displayed as mean ±
SEM. (B) protein levels of ING5 decrease during differentiation in BT 189 cells. N=3. Values displayed as mean ± SEM. BT 189 cells dissociated from the same flask of spheres were either cultured as spheres in the presence of EGF and FGF (sphere group in B) or seeded on poly-L- ornithine coated plates and cultured in the medium containing 1 % FBS and no growth factors for at least five days (differentiation group in B). Spheres or differentiated cells were collected at indicated time points and the mRNA and protein levels in these cells were detected. (C)
Immunofluorescence of ING5 in undifferentiated cells (upper) and in cells differentiated for 5 days
(lower). BT 189 cells were dissociated and seeded on poly-L-ornithine coated coverslips. For the undifferentiated group, cells were cultured in the absence of serum and growth factors for 1 day before being fixed with 4 % formaldehyde. For the differentiated group, cells were cultured in the medium containing 1 % FBS and no growth factors for 5 days before fixation. Representative fields were shown for the IF experiment. Scale bar = 20 µm. (D) ING5 protein levels decrease in
BT 12 and BT 134 cells. Experiments were performed with the same method as in BT 189 cells.
52
3.2 Determining the function of ING5 in BTIC self-renewal
To test the functional relevance of ING5 in BTIC stem cell properties, we performed sphere formation assays, which measures the clonogenic capability of each single cell, in ING5 gain- and loss-of-function systems (Figure 2). Sphere formation from dissociated cells cultured at a very low density was observed over a two-week time course. ING5 overexpression in BT 189 cells by plasmid transfection increased sphere formation rate and the size of spheres (Figure 5A, 5B).
Conversely, ING5 knockdown by si-ING5 reduced the number and size of spheres (Figure 5C,
5D). To target ING5 more homogeneously in cell populations we constructed two doxycycline- inducible shR-ING5 lines (shR1, shR2) using the lentivirus system from BT 189 cells and found that the two knockdown lines independently showed reduced sphere formation capability (Figure
6). For further experiments we mainly used shR2 (hereafter shR-ING5) that had a better (~80%) knockdown efficiency (Figure 7D). The alteration of ING5 protein levels by various manipulations is shown in Figure 7. As the ING5-HBO1 complex was reported to be required for
DNA synthesis in breast cancer cells (Doyon et al. 2006), we examined the cell cycle profile of
BTICs using flow cytometry after ING5 manipulation. We did not observe changes in cell cycle in response to ING5 (data not shown), suggesting that enhanced sphere formation capability resulted from increasing self-renewal rather than directly affecting BTIC proliferation.
53
Figure 6. ING5 enhances sphere forming ability of BT 189 cells.
54
Figure 7. ING5 enhances sphere forming ability of BT 189 cells. (A, B) ING5 transient overexpression by plasmid transfection increases sphere formation rates and sphere sizes in sphere formation assays. N = 3. Values displayed as mean ± SEM. ** P < 0.01 and * P < 0.05 (unpaired t test). (C, D) ING5 transient knockdown by siRNA decreases sphere formation rates and the average volume of BTIC spheres. N = 3. Values displayed as mean ± SEM. * P < 0.05 (unpaired t test). BT 189 cells were seeded on poly-L-ornithine coated plates 16-24 hours before transfection.
Transfection with pCI plasmids or siRNAs was performed in the absence of serum and growth factors. One day after transfection cells were supplemented with growth factors. Transfection efficiency is about 50 % in BTICs. Two days after transfection spheres were dissociated and 200 viable single cells were seeded in each well of 96-well plates and were grown for 12-14 days. 30
µL of fresh medium was added to the existing medium every three days. The complete field of each well was captured under a 4x objective and the diameter of each sphere measured using
ImageJ software. At least three wells (600 cells) were counted for each condition in one experiment. The sphere formation rate was calculated by counting the number of spheres with a diameter over 50 µm or over 100 µm.
55
Figure 8. BT 189 stable shRNA cell lines have lower sphere forming capability. (A) Sphere formation rates for cell lines stably expressing shRNAs against ING5 (shR1 and shR2-ING5) or control non-targeting shRNA (shR-ctr). N = 3. Values displayed as mean ± SEM. ** P < 0.01 and
* P < 0.05 (unpaired t test). shRNA stable cell lines were established from BT 189 cells (Figure
2C). Expression of shRNAs was induced by doxycycline at 100 ng/mL for 6-8 days before cells were dissociated and 200 viable single cells were seeded in each well of 96-well plates and were grown for 12-14 days. 30 µL of fresh medium containing 100 ng/mL dox was added to the existing medium every three days. Sphere formation rates were calculated as described in Figure 5. (B)
Fluorescence of the RFP reporter in stable cell lines superimposed with differential interference contrast (DIC) images. Scale bar = 100 µm.
56
Figure 9. The alteration of ING5 protein levels by various manipulations in BT 189 cell line.
(A) Transient ING5 overexpression by transfection with pCI plasmids. The protocol for plasmid transfection was described in Figure 2A. N=3. Representative blot was shown. (B) Induced ING5 overexpression by cumate at 30 µg/mL for 3 days in iPB stable expression cell lines. The procedure for iPB cell line construction was shown in Figure 2B. N=3. Representative blot was shown. (C)
Transient ING5 knockdown by siRNA. The protocol for siRNA transfection was described in
Figure 2A. N=3. Representative blot was shown. (D) ING5 knockdown in two shRNA stable cell lines constructed from BT 189 cells. Expression of shRNAs was induced by dox (100 ng/mL) for
7 days. The procedure for shRNA cell line construction was shown in Figure 2C. N=3.
Representative blots were shown. α-Tubulin (α-Tub) and β-actin (ACTB) were used as protein loading controls.
57
Assessing the expression levels of stem cell core transcription factors OCT4, OLIG2, and neural stem cell markers CD133 and Nestin showed that all were upregulated 3-5-fold by ING5
(Figure 8A). Induction of Nestin and repression of neuronal lineage marker Tubulin β III
(Tubb3) was confirmed by western blot in ING5 overexpressing cells (Figure 8B) and opposite effects were seen in ING5 knockdown cells (Figure 8C).
BTICs grow as heterogeneous populations with a range of self-renewing capabilities in vitro.
For more homogeneous and titratable expression of ING5, we constructed stable ING5- overexpressing clones using the PiggyBac transposon system (Woltjen et al. 2009). To characterize the stem cell hierarchy in BTICs we utilized two cell surface markers CD44 and
CD133, which are correlated with the stemness properties and tumor initiating abilities of glioma cells (Pietras et al. 2014, Liu et al. 2009, Singh et al. 2004). Induction of ING5 expression by cumate resulted in a higher percentage of CD44 and C133 positive cells in iPB-ING5 cells compared to controls (Figure 9), and an even greater decrease in the CD44 positive population in knockdown cell lines (Figure 10).
Both CD44 and CD133 have been reported to enrich cancer stem cells in multiple cancer types (Singh et al. 2004, O'Brien et al. 2007, Dalerba et al. 2007, Cox et al. 2009), however, it is still not clear which marker better characterize the stem cell hierarchy in GBM. Here we show that
ING5 is positively correlated with the expression of both CD44 and CD133 in iPB cell lines.
However, unfortunately the shRNA cell lines contain a very strong RFP label which interfere with the detection of CD133-PE signal in this experiment. Although we were not able to directly compare the CD133 expression in shRNA cells with iPB cell lines, it was shown in shRNA cells that the ING5 expression was positively correlated with CD44, consistent with the data shown in iPB cells.
58
Figure 10. ING5 induces the expression of stem cell factors and downregulates differentiation markers. (A) RT-qPCR analysis of stem cell markers and stem cell core transcription factors after
ING5 overexpression. BT 189 cells were transfected with pCI plasmids and three days after transfection cells were harvested and RNA was extracted for qPCR analysis. N = 3. Values displayed as mean ± SEM. (B-C) Western blot analysis of neural stem cell marker Nestin and neuronal lineage differentiation marker Tubb3 in response to ING5 overexpression (B) and knockdown (C). For overexpression, BT 189 cells were transfected with pCI plasmids and three days after transfection cells were harvested and total protein was extracted for western blot. For shRNA knockdown, stable shRNA cells were induced by 100 ng/mL dox for 7 days and total protein was extracted. N=3. Representative blots were shown. α-Tubulin (α-Tub) and β-actin
(ACTB) were used as protein loading controls.
59
Figure 11. ING5 overexpression increases the CD133+/CD44+ population in BTICs. (A) Flow cytometry analysis of cells positive for the stem cell markers CD133 and CD44 in iPB-ctr and iPB-
ING5 overexpression groups. iPB cells were induced by cumate at 30 µg/mL for 3 days before dissociation and FACS analysis. Gates were set according to the isotype control. N=3.
Representative bivariate plots were shown. (B) Statistics of the flow cytometry data. MFI = median fluorescence intensity.
60
Figure 12. ING5 knockdown decreases the CD133+ population in BTICs. (A) Flow cytometry analysis of shRNA cells with a RFP tag positive for the stem cell marker CD44. shRNA cells were induced by dox at 100ng/mL for 6-8 days before dissociation and FACS analysis. Gates were set according to the isotype control. N=3. Representative bivariate plots were shown. (B) Statistics of the flow cytometry data. MFI = median fluorescence intensity.
61
Tissue stem cells usually achieve self-renewal through asymmetric cell division in which two daughter cells unequally inherit fate determinants for stemness and lineage commitment
(Konno et al. 2008). In neural stem cells, CD133 is not only a stem cell marker, but also specifically marks the apical membrane (Kosodo et al. 2004), which initiates subsequent asymmetric division.
Therefore, we tested whether increased ING5 in BTICs was associated with a change of cell division symmetry. In our mitotic pair analysis, the distribution of Nestin was used as an indicator for three division modes (Cusulin et al. 2015), symmetric proliferating, symmetric differentiating and asymmetric division (Figure 11A). As expected, symmetric proliferating division comprised the majority of mitotic pairs in both control and ING5 overexpression groups (sym-pro in Figure
11B), perhaps due to the perturbed asymmetric machinery reported in malignant cells (Chang,
Wang, and Wang 2012). Despite the high frequency of symmetric division, ING5 induced a more than two-fold increase in the asymmetric division mode (highlighted by the white arrowhead in
Figure 11A), consistent with a role in promoting asymmetric division and maintaining stemness.
62
Figure 13. ING5 increases the frequency of asymmetric division of BTICs.
63
Figure 14. ING5 increases the frequency of asymmetric division of BTICs. (A) An example of cell division symmetry based on the immunofluorescence of Nestin with the Red arrow indicating symmetric differentiating division and the white arrow asymmetric division.
Asymmetric division was characterized as more than 2 fold difference of Nestin fluorescence intensity between the two daughter cells and one of the daughter cell displays very weak signal of
Nestin, otherwise the mitotic pair is defined as symmetric division. If the two daughter cells both display very weak signal of Nestin and show differentiated morphology, it is defined as symmetric differentiating division, and the rest of mitotic pairs are symmetric proliferating division. (B)
Mitotic pair analysis of symmetric proliferating (sym-pro), symmetric differentiating (sym-diff) and asymmetric (asym) cell division. The division mode analysis was performed on mitotic pairs seeded at low density on coated coverslips in the absence of growth factors. Synchronized iPB cells were dissociated and seeded at low density on poly-L-ornithine coated coverslips for 18-20 hours before being fixed and immunostained for Nestin. N=3. Values displayed as mean ± SEM.
* P < 0.05 (unpaired t test).
64
3.3 The effects of ING5 on BTIC differentiation
Next we asked whether forced expression of ING5 could inhibit serum-induced differentiation. In the inducible PiggyBac (iPB) cell lines, ING5 expression caused resistance to lineage differentiation with a delayed acquisition of morphological changes compared to control cells
(Figure 12). Immunostaining for the neural stem cell marker Nestin and the neuronal lineage marker Tubb3 on these cells indicated a higher level of Nestin and reduced expression of Tubb3 in iPB-ING5 cells (Figure 13). Conversely, in shRNA knockdown lines Tubb3 was increased while Nestin was downregulated (Figure 14). These results suggest that ING5 prevents lineage commitment of BTICs under serum induction.
3.4 Examining the function of ING5 in BTIC self-renewal in the absence of growth factors
EGF and FGF are required for in vitro propagation of neural stem cells and some BTIC lines, suggesting that mitogenic signaling is needed for sustaining self-renewal. To test whether the function of ING5 was associated with these pathways, we deprived cells of growth factors and performed serial sphere formation assays. Sphere formation was greatly impaired in the absence of EGF and FGF, but could be rescued by ING5 overexpression (Figure 15). In fact, ING5 increased sphere formation to a greater extent under growth factor deprivation than in the presence of growth factors, and maintained self-renewal in serial sphere passages when control cells lost nearly all of their self-renewal capability (Figure 15A). To identify pathways through which ING5 might be exerting this effect, we treated cells with the kinase inhibitors Tofacitinib, PX-866 and
PD184352 that target the JAK/STAT, PI3K/Akt and MEK pathways, respectively. The JAK1/3 inhibitor Tofacitinib impaired sphere formation in both control and ING5 overexpressing groups
(Figure 16A).
65
Figure 15. ING5 overexpression inhibits the acquisition of differentiation morphology in iPB cells. Morphological changes of iPB-ctr and iPB-ING5 cells before (Day 0) and after (Day 1-3) differentiation induced by 1 % FBS. iPB cells were induced by cumate at 30 µg/mL for 2 days before the cells were dissociated and seeded onto poly-L-ornithine coated 12 well plates in the presence of 1 % FBS, 30 µg/mL cumate and no growth factors. Cell morphology was recorded every day for 5 days and representative fields were displayed. N=3. Scale bar = 100 µm.
66
Figure 16. ING5 overexpression prevents lineage differentiation. Immunofluorescence of neural stem cell marker Nestin and neuronal lineage marker Tubb3 in differentiated iPB control and iPB-ING5 cell lines. iPB cells were induced by cumate at 30 µg/mL for 3 days before they were dissociated and seeded onto poly-L-ornithine coated 12 well plates in the presence of 1 %
FBS, 30 µg/mL cumate and no growth factors. Differentiation was induced for 5 days before the cells were fixed and immunostained for neural markers. N=3. Representative figures were shown.
Scale bar = 200 µm.
67
Figure 17. ING5 knockdown promotes lineage differentiation. Immunofluorescence of Nestin and Tubb3 in differentiated shRNA cell lines with RFP reporter. shRNA cells were induced by
100 ng/mL dox for 7 days before they were dissociated and seeded onto poly-L-ornithine coated
12 well plates in the presence of 1 % FBS, 100 ng/mL dox and no growth factors. Differentiation was induced for 5 days before the cells were fixed and immunostained for neural markers. N=3.
Representative figures were shown. Scale bar = 200 µm.
68
Figure 18. ING5 sustains self-renewal over serial passages in the absence of growth factors.
(A) The sphere formation rates of iPB-ctr and iPB-ING5 overexpressing cells at three successive passages in the absence of EGF and FGF treatment. Expression of ING5 was induced by cumate for 3 days before cells were dissociated and 200 viable single cells were seeded in each well of 96- well plates and were grown for 12-14 days. 30 µL of fresh medium containing 30 µg/mL cumate was added to the existing medium every three days. Serial sphere assays were conducted by harvesting all of the cells from the last passage and replating them into new wells at the same dilution. N = 3, ** P < 0.01, *** P < 0.001 (unpaired t test). (B) Representative DIC images of spheres from the tertiary passage in iPB cells. Scale bar = 400 µm.
69
However, ING5 overexpressing cells displayed significant resistance to MEK1/2 and PI3K inhibitors and retained sphere forming ability relative to control cells (Figure 16A). We found that
ING5 elevated phosphorylation levels of AKT and ERK1/2, effectors of the PI3K and MEK pathways, suggesting ING5 induces a higher level of activation of these two pathways which made the cells more resistant to the inhibitors targeting the two pathways (Figure 16B).
The association of the PI3K/AKT and MEK/ERK pathways with stem cell maintenance was based primarily on the observation of enhanced stem cell survival and proliferation rate. To address whether the two pathways affected other stem cell properties, we treated iPB cell lines with PX-
866 and PD184352 and induced differentiation. We found that inhibition of both PI3K and
MEK1/2 reduced expression of Nestin and increased Tubb3 expression in both iPB-ING5 and iPB- ctr cells (Figure 17), indicating a differentiation-promoting effect when the two pathways were repressed. We then profiled the CD133-positive population in BT 189 cells treated with PX-866 and PD184352 and found that inhibition of PI3K/AKT signaling did not affect the CD133 positive pool, while blocking the MEK/ERK pathway with PD184352 significantly reduced the expression of CD133 in the BTIC population (Figure 18). This suggests that the two mitogenic pathways function through different mechanisms to affect stem cell properties, and that of the two pathways, the MEK/ERK pathway may be more directly related to ING5-mediated self-renewal of BTICs.
3.5 Transcriptome assay in ING5 knockdown cell lines
To gain molecular insights into how ING5 promotes stemness maintenance, we conducted transcriptome profiling using shRNA cell lines. The shR-ctr group and shR-ING5 group were hierarchically clustered, indicating consistency among replicates (Figure 19). The shR-ING5
70
Figure 19. ING5 overexpression induces activation of the PI3K/AKT and MEK/ERK signaling pathways. (A) Sphere formation rates of pCI-transfected cells under treatment with protein kinase inhibitors at the indicated concentrations. N=3. Values displayed as mean ± SEM.
* P < 0.05 (unpaired t test). (B) Protein and phosphorylated protein levels of effectors in the PI3K and MEK pathways. iPB cells were induced for 3 days and then dissociated, plated at 2 × 104 /mL, and treated with 1 µM PX-866 or 2 µM PD184352 in the presence of cumate for 48 hours before lysis.
71
Figure 20. Inhibiting the PI3K/AKT and MEK/ERK signaling pathways promotes neuronal differentiation.
72
Figure 21. Inhibiting the PI3K/AKT and MEK/ERK signaling pathways promotes neuronal differentiation. Immunofluorescence of Nestin and Tubb3 in differentiated iPB control (top panels) or iPB-ING5 (bottom panels) cells treated with 1 µM PX-866 or 2 µM PD184352. iPB cells were induced for 2 days. After induction cells were dissociated, seeded on poly-L-ornithine coated coverslips, and treated with 1 µM PX-866 or 2 µM PD184352 in the presence of cumate for 5 days before fixation. Representative fields were shown. Scale bar = 200 µm.
73
Figure 22. MEK/ERK pathway inhibition decreases the CD133 positive population of BTICs.
(A) Flow cytometry analysis of the CD133 positive population in PX-866 and PD184352 treated
BT 189 cells, gated by isotype control. BT 189 cells were dissociated from spheres and seeded in
24 well plates at 2 × 104 cells/mL with treatment of 1 µM PX-866 and 2 µM PD184352 in the absence of growth factors. After treated for three days, cells were dissociated and analyzed for
CD133 positive fraction. N=3. Representative plots were shown. (B) Statistics of the flow cytometry data. MFI = median fluorescence intensity.
74
Figure 23. Transcriptome analysis in shRNA knockdown cell lines. shRNA cells were induced for 7 days with 100 ng/mL dox before the two groups of cells (shR-ctr and shR-ING5) were collected for RNA extraction. The four samples were composed of two independent preparations of samples. The hierarchical clustering of microarray data of the four samples shows two clusters
(shR-ctr vs. shR-ING5) based on ING5 expression status. The color scale is in log base 2.
75 group displayed an overall silencing of gene expression, and most of the differentially expressed genes came from ones highly expressed in the control group while being suppressed by ING5 knockdown. We performed gene set enrichment analysis (GSEA) - GO term (Gene Ontology) to identify ING5-regulated pathways and biological processes, and found that calcium channel activity was the top ranked gene set enriched in the non-targeting (ctr) control group (Table 1,
Figure 20A, 20B). There were four additional GO terms related to cation channels among the top
25 terms (Table 1). GSEA-KEGG pathway analysis also revealed an enrichment of calcium signaling genes in ctr samples (Figure 20C), consistent with ING5 positively regulating the activity of calcium channels and other cation channels to influence intracellular ion homeostasis.
On the GO-enriched gene list (Figure 20A), we saw a number of calcium channel components upregulated by ING5, including the subunits of L-type (CACNA1F, CACNA1S,
CACNA1D, CACNA1C), P/Q-type (CACNA1A) voltage-gated calcium channels and transient receptor potential cation channels permeable to calcium (TRPC3, TRPC5, TRPC4, TRPM1).
Retesting the effect of ING5 using real-time qPCR in a subset of genes validated positive regulation by ING5 (Figure 21).
3.6 Determining the regulatory function of ING5 on intracellular calcium levels
Calcium signaling positively correlates with GBM cell survival (Ishiuchi et al. 2002, Kang et al.
2010) and was reported to promote stemness features in cancer stem cells (Chai et al. 2015). Using the calcium dye Fluo3-AM in live cells we found that intracellular calcium levels were increased in iPB-ING5 cells (Figure 22A) and decreased in shR-ING5 cells (Figure 22B) as compared to controls. We also tested for correlations between ING5 expression and
76
Table 1. Top 25 Gene Ontology hits enriched in shR-ctr group using GSEA analysis. (NES
= normalized enrichment score.)
Rank Gene Ontology Name NES 1 CALCIUM CHANNEL ACTIVITY 1.839764 2 GROWTH FACTOR BINDING 1.735562 3 MORPHOGENESIS OF AN EPITHELIUM 1.687653 4 DEFENSE RESPONSE TO BACTERIUM 1.675038 5 VOLTAGE GATED CATION CHANNEL ACTIVITY 1.652875 6 REGULATION OF CYTOKINE SECRETION 1.638789 7 STRUCTURAL MOLECULE ACTIVITY 1.627298 8 POSITIVE REGULATION OF IMMUNE RESPONSE 1.624614 9 CYTOKINE SECRETION 1.605037 10 REGULATION OF IMMUNE RESPONSE 1.592373 11 HEMATOPOIETIN INTERFERON CLASSD200 DOMAIN CYTOKINE RECEPTOR ACTIVITY 1.58685 12 ACTIVATION OF IMMUNE RESPONSE 1.578876 13 VOLTAGE GATED POTASSIUM CHANNEL ACTIVITY 1.578127 14 CATION CHANNEL ACTIVITY 1.578122 15 PROTEIN SECRETION 1.556662 16 VOLTAGE GATED CHANNEL ACTIVITY 1.542501 17 JAK STAT CASCADE 1.53783 18 SECRETORY GRANULE 1.514519 19 G PROTEIN SIGNALING ADENYLATE CYCLASE ACTIVATING PATHWAY 1.512888 20 FEMALE GAMETE GENERATION 1.508166 21 STRUCTURAL CONSTITUEctr OF MUSCLE 1.50177 22 NEGATIVE REGULATION OF SIGNAL TRANSDUCTION 1.497667 23 ICOSANOID METABOLIC PROCESS 1.489995 24 REGULATION OF PROTEIN SECRETION 1.486827 25 CELL CELL ADHESION 1.485738
77
Figure 24. GSEA analysis shows that Calcium channel activity is induced by ING5. Gene list
(A) and enrichment plot (B) showing enrichment of genes upregulated in the shR-ctr group within the GO term Calcium Channel Activity, with a positive correlation. Black bars indicate the position of genes in the ranked list of differentially expressed genes between shR-ctr and shR-ING5 groups.
The green line represents the enrichment score. (C) Gene enrichment plot of KEGG Calcium signaling pathway.
78
Figure 25. ING5 upregulates genes encoding for calcium channel components. Validation of genes upregulation by ING5 from the microarray data using real-time qPCR. iPB cells were induced by cumate at 30 µg/mL for 3 days before cells were harvested for RNA extraction and qPCR analysis. N=3. Values displayed as mean ± SEM. * P < 0.05 and ** P < 0.01 (unpaired t test).
79
Figure 26. ING5 increases intracellular calcium levels in live cells.
80
Figure 27. ING5 increases intracellular calcium levels in live cells. Live cell calcium imaging in iPB (A) and shRNA (B) cell lines. iPB cells were induced by cumate at 30 µg/mL for 3 days and shRNA cells were induced by dox at 100ng/mL for 7 days. Then cells were dissociated and seeded on poly-L-ornithine coated plates at a density of 6-8 × 104 /mL for 24 hours in the absence of growth factors. Attached cells were then incubated in medium containing 7 µM Fluo3-AM at
37 °C for 30 min. After washing with PBS, cells were de-esterified for 30 min in fresh medium
(with no serum and growth factors) and examined under a fluorescence microscope.
Representative fields were shown. Scale bar = 400 µm.
81 calcium level at the single cell level by co-transfecting pCI-ING5 with pCI-mCherry. 20-30% of the cell population was co-transfected and mCherry positive cells were measured for Fluo3-AM fluorescence intensity. Calcium levels were elevated in cells transfected with pCI-ING5 compared to pCI vector (Figure 23). Quantification of Fluo3-AM fluorescence by flow cytometry also showed that knockdown of ING5 in shR-ING5 cells reduced the Calcium-High cell population
(Figure 24), whereas the Calcium-High population was increased by ING5 overexpression in iPB-
ING5 cells (Figure 25).
3.7 Determining the effects of calcium levels on BTIC self-renewal
Next we asked whether ING5-induced calcium elevation correlated with stemness using two calcium inducers that function through independent ways: ionomycin, which transports calcium ions through the plasma membrane, and cyclopiazonic acid (CPA) that releases calcium from intracellular stores. Both ionomycin and CPA increased the CD133 positive stem cell pool in
BTICs as estimated by flow cytometry (Figure 26).
We also treated cells with relatively low doses of ionomycin and CPA which increased calcium levels without causing apoptotic cell death and noted that treatment with ionomycin and
CPA stimulated sphere formation in both shR-ING5 and control cells, with the rate of sphere formation in knockdown cells nearing the level seen in untreated control cells (Figure 27). In contrast, treatment with the calcium chelator BAPTA-AM, or the L-type (Nifedipine) and T-type
(SKF96365) calcium channel blockers reduced intracellular calcium levels and decreased sphere forming ability in shRNA cell lines (Figure 27). This suggests that ING5 is essential for maintaining intracellular calcium at a relatively high level, which is required for maintenance of self-renewal in BTICs.
82
Figure 28. High ING5 expression level is correlated with high intracellular calcium levels.
Fluo3-AM fluorescence intensity of mCherry positive cells in BT 189 cells co-transfected with pCI-ctr/pCI-ING5 and pCI-mCherry. BT 189 cells were attached on poly-L-ornithine plates for 24 hours and transfected with pCI plasmids (either pCI-ING5 or empty vector) and pCI-mCherry at a ratio of 20:1, and all mCherry positive cells were supposed to be co-transfected with pCI plasmids. Two days after transfection, cells were loaded with Fluo3-AM as described in Figure
22 and examined under a fluorescence microscope. More than 200 mCherry positive cells
(regardless of a strong or weak signal) were captured and the calcium level was quantified by subtracting the initial fluorescence intensity from the fluorescence after excitation, using ImageJ software. N=3. Total cell number > 200 cells for each group, * P < 0.05 (unpaired t test).
83
Figure 29. ING5 knockdown decreases the proportion of cells with high intracellular calcium levels. (A) Flow cytometry analysis of Fluo3-AM fluorescence intensity in shRNA cells, gated by the BAPTA-AM treated control. shRNA cells were induced by dox for 6-8 days before dissociated and the suspended cells were loaded with 5 µM Fluo3-AM for 45 min at 37 °C in dark, washed with PBS twice, and checked on the flow cytometer. N=3. Representative plots were shown. (B)
Statistics of the flow cytometry data. MFI = median fluorescence intensity.
84
Figure 30. ING5 overexpression increases the proportion of cells with high intracellular calcium levels. (A) Flow cytometry analysis of Fluo3-AM fluorescence intensity in iPB cells, gated by the BAPTA-AM treated control. iPB cells were induced by cumate for 3 days before dissociated and the suspended cells were loaded with 5 µM Fluo3-AM for 45 min at 37 °C in dark, washed with PBS twice, and checked on the flow cytometer. N=3. Representative plots were shown. (B) Statistics of the flow cytometry data. MFI = median fluorescence intensity.
85
Figure 31. Calcium level elevation increases CD133 positive cells in BTICs. (A) Flow cytometry analysis of CD133 positive cells in BT 189 cells treated with calcium inducers ionomycin or CPA, gated by isotype control. BT 189 cells were dissociated from spheres, plated at 2 × 104 /mL in 12 well plates and treated with ionomycin (20 nM) or CPA (100 nM) for 48 hours in the absence of growth factors. Then cells were dissociated for FACS analysis. N=3.
Representative plots were shown. (B) Statistics of the flow cytometry data. MFI = median fluorescence intensity.
86
Figure 32. Calcium level affects BTIC self-renewal ability. Sphere formation rate in shRNA cell lines treated with calcium modulators ionomycin, CPA, BAPTA-AM, Nifedipine or
SKF96365 at the indicated concentrations for 12 days. shRNA cells were induced by dox for 6-8 days and were then dissociated, seeded on 96-well plates (200 viable cells per well). Cells were kept growing into spheres for 14 days in the presence of calcium modulators, 100 ng/mL dox and no growth factors. N=3. Values displayed as mean ± SEM. ** P < 0.01 compared to DMSO treatment of the shR-ctr group, # P < 0.05 and ## P < 0.01 compared to DMSO treatment of the shR-ING5 group (unpaired t test).
87
Since calcium signaling induces various cellular functions through the well-known calcium effectors PKC, calcineurin (CaN) and calmodulin-dependent kinases (CAMK), to further look at the molecular mechanism of calcium signaling, we treated cells with the CAMKII inhibitor KN-
93 and the CaN inhibitor Cyclosporin A (CsA) and found that CsA treatment abolished the enhanced sphere formation ability in iPB-ING5 cells. However KN-93 did not affect sphere forming rates (Figure 28), indicating the calcium signaling was transduced through the CaN axis which could induce the PI3K and MAPK pathways. None of the above reagents caused apoptotic cell death at the concentrations used in this study as estimated by both cell morphology and PARP cleavage.
3.8 Induction of the FSH pathway by ING5
Analysis of microarray data using IPA yielded a high z-score for the follicle stimulating hormone
(FSH) pathway (Table 2), which was also shown in one of the top ranked GO term in GSEA,
Female Gamete Generation (Table 1). Both the FSH receptor (FSHR) and FSHB ligand were induced by ING5 in microarray experiments and this was confirmed using quantitative PCR
(Figure 29A). Both genes were decreased in differentiating cells, further suggesting that the FSH pathway was possibly involved in the stem cell features of BTICs (Figure 29B). FSHR is a G protein-coupled receptor (GPCR) that activates the MAPK and calcium signaling pathways
(Bhartiya and Singh 2015) and induces stem cell signaling through OCT4 (Zhang et al. 2013, Liu et al. 2015). Although the FSH pathway was shown to function in follicular development and estradiol production in mammalian ovaries and germ cell development in both females and males
(Siegel et al. 2013), the physiological function and level of activation of the FSH pathway in brain cells was unexplored. TCGA gene expression data indicated that mRNA levels of FSHR and FSHB
88
Figure 33. Calcineurin is a downstream effector of calcium signaling in self-renewal regulation. Sphere formation rate in iPB cells treated with the CAMKII inhibitor KN-93 or the
CaN inhibitor Cyclosporin A at indicated concentrations. iPB cells were induced for 3 days and were then dissociated, seeded on 96-well plates (200 viable cells per well). Cells were kept growing into spheres for 14 days in the presence of KN-93 or CsA, 30 µg/mL cumate and no growth factors. N=3. Values displayed as mean ± SEM. ** P < 0.01 compared to iPB-ctr/DMSO and # P < 0.05 compared to iPB-ING5/DMSO (unpaired t test).
89
Table 2. Gene list of Quantity of FSH from the IPA downstream function.
QUANTITY OF FSH Z-SCORE: 1.982
GENE ID Fold change (ctr vs. shR) FSHR 4.31 NOS1 3.12 THRB 2.82 CGA 2.02 FSHB 1.88
90
Figure 34. ING5 induces the FSH pathway in BTICs.
91
Figure 35. ING5 induces the FSH pathway in BTICs. (A) RT-qPCR analysis of genes related to hormone and steroidogenesis functions in ING5 overexpressing cells. BT 189 cells were transfected with pCI plasmids and three days after transfection cells were harvested and RNA was extracted for qPCR analysis. N = 3. Values displayed as mean ± SEM. * P < 0.05, ** P < 0.01
(unpaired t test). (B) The expression levels of FSHB and FSHR in BT 189 cells before and after differentiation for 1-5 days. BT 189 cells were dissociated and either grow into spheres in a non- adherent condition or seeded on poly-L-ornithine coated 6 well plates for differentiation induction in the presence of 1 % FBS. N=3. (C) Immunostaining for ING5 and FSHR in iPB cells. iPB cells were induced with cumate for 3 days before the spheres were dissociated and seeded on poly-L- ornithine coated 24 well plates. After attached for 24 hours, cells were fixed and immunostained for ING5 and FSHR. N=2. Representative fields were shown. Scale bar = 100 µm.
92 were relatively low in GBM tissues, but we found that ING5 strongly elevated FSH receptor levels in BTICs (Figure 29C).
3.9 Determining the role of the ING5-induced FSH pathway in BTIC self-renewal
To address whether the FSH signaling induced by ING5 affected BTIC self-renewal, we performed sphere formation assays using an FSHR neutralizing antibody to block the pathway. Cells treated with FSHR antibody showed reduced sphere-forming ability, similar to effects seen with the calcium channel blocker SKF96365 and BAPTA, which reduced intracellular calcium levels
(Figure 30A). We found that when intracellular calcium levels were elevated in control cells by ionomycin and CPA, the sphere forming rate was increased, but was lower than in iPB-ING5 cells
(Figure 30A) suggesting that calcium signals transduced some, but not all of the effects of ING5 on BTIC self-renewal. Therefore, calcium and FSH signaling pathways may co-operatively promote self-renewal maintenance downstream of ING5. We further compared the effect of combined treatment with FSHR antibody and BAPTA to the single treatments (Figure 30B), and found a partially additive effect of inhibiting both pathways, suggesting the two pathways may have independent functions in regulating BTIC self-renewal.
To further test the effect of the FSH pathway on other stem cell properties of BTICs, we treated cells with FSHR neutralizing antibody and induced differentiation with 1% serum.
Blocking FSHR increased levels of the neuronal differentiation marker Tubb3 compared to cells treated with an IgG isotype control (Figure 31A) and reduced the number of CD133 positive cells in the BTIC population (Figure 31B), indicating the FSH pathway played a role in promoting stem cell properties and preventing neuronal differentiation. Since FSH has been reported to activate calcium-signaling pathways, we asked if ING5-induced calcium elevation was also associated with
93
Figure 36. Calcium level and the FSH signaling pathway positively regulate self-renewal ability of BTICs.
94
Figure 37. Calcium level and the FSH signaling pathway positively regulate self-renewal ability of BTICs. (A) Sphere formation assays for iPB cells treated with calcium modulators and
FSHR blocking antibody at the indicated concentrations. iPB cells were induced for 3 days and were then dissociated, seeded on 96-well plates. Cells were kept growing into spheres for 14 days in the presence of 30 µg/mL cumate, calcium modulators and FSHR blocking antibody, and no growth factors. N=3. Values displayed as mean ± SEM. * P < 0.05 and ** P < 0.01 compared to
DMSO control or IgG control (for Anti-FSHR treated) of the iPB-ctr group; # P < 0.05 ## P < 0.01 compared to DMSO control or IgG control (for Anti-FSHR treated) of the iPB-ING5 (unpaired t test). (B) Sphere formation rates for iPB cells treated with FSHR antibody or BAPTA alone, and the combination of both. N=3. Values displayed as mean ± SEM. * P < 0.05 and ** P < 0.01
(unpaired t test).
95
Figure 38. The FSH pathway promotes stem cell properties of BTICs.
96
Figure 39. The FSH pathway promotes stem cell properties of BTICs. (A)
Immunofluorescence of Nestin and Tubb3 in BT 189 cells treated with FSHR neutralizing antibody and mouse IgG as isotype control. BT 189 cells were dissociated and induced to differentiate on poly-L-ornithine coated coverslips in the presence of 1 % FBS and 0.5 µg/mL
FSHR antibody or IgG control antibody for 5 days. Representative fields were shown. Scale bar =
200 µm. (B) Flow cytometry analysis of CD133 positive cells in BT 189 cells treated with FSHR neutralizing antibody or IgG control. BT 189 cells were seeded at 2 × 104 /mL in 24 well plates and treated with 0.5 µg/mL FSHR antibody or IgG control antibody for 3 days in the absence of growth factors. Cells were then dissociated for FACS analysis. N=3. Representative plots were shown. MFI = median fluorescence intensity.
97 the FSH pathway. However, we did not observe a change in calcium levels by blocking FSHR
(data not shown), suggesting that the FSH and calcium signaling pathways probably function independently to mediate the effects of ING5 on BTIC stem cell maintenance.
3.10 Determining the role of the PHD motif in ING5-induced stemness maintenance
Since some activities of ING proteins appear to be independent of their ability to target HAT and
HDAC complexes (Sarker et al. 2008), we tested whether the PHD motif of ING5, which targets the H3K4me3 histone mark, was dispensable for the effect of ING5 on self-renewal. We constructed an iPB stable cell line overexpressing ING5 deleted for the PHD (iPB-ΔPHD), and found that the sphere formation rate in iPB-ΔPHD cells was not significantly different from control cells (Figure 32), suggesting the self-renewal capability induced by ING5 is dependent on its PHD motif, consistent with the epigenetic functions of ING5 in regulating gene expression.
Based upon the data presented, we proposed a model for ING5 functions in the maintenance of self-renewal of BTICs (Figure 33). In the absence of growth factors, ING5 induces FSH and calcium signaling by promoting transcription of the FSH receptor and ligand genes, and various plasma membrane calcium channel genes. The FSH and calcium signaling pathways further activate PI3K/AKT and MEK/ERK signaling to enhance stem cell features and the expression of stemness factors (OCT4, OLIG2 and Nestin). Gene activation by ING5 is dependent on its PHD motif to target ING5-associated histone acetyltransferase complexes onto the promoter regions of target genes.
98
Figure 40. The PHD motif is required for the stem cell maintenance function of ING5. (A)
Western blot shows the protein levels of wildtype ING5, ING5 with a FLAG tag, and PHD-deleted
ING5 (black arrows) in three iPB cell lines. iPB cells were induced with 30 µg/mL cumate for 3 days and cells were lysed for western blot assay. N=3. (B) Sphere formation assay in iPB cell lines overexpressing wild-type and PHD-deleted ING5 (ΔPHD). iPB cells were induced for 3 days and were then dissociated, seeded on 96-well plates. Cells were kept growing into spheres for 14 days in the presence of 30 µg/mL cumate and no growth factors. N=4. Values displayed as mean ±
SEM. ** P < 0.01(unpaired t test).
99
Figure 41. Model for how ING5 functions in the maintenance of BTIC self-renewal. ING5 with its PHD motif binds to the promoters and recruits HAT complexes to promote the expression of genes involved in the FSH pathway and calcium plasma membrane entry. The induced FSH and calcium signaling pathways further activate PI3K/AKT and MEK/ERK signaling in the absence of growth factors EGF and FGF, to enhance stem cell features and expression of stemness factors
OCT4, OLIG2 and Nestin.
100
3.11 The correlation between ING5 expression levels and GBM prognosis
Since ING5 maintained the stem cell features of BTICs, which would be expected to make GBMs more resistant to therapy, we asked if this was reflected in the clinical outcome of GBM patients.
Kaplan-Meier analysis of TCGA clinical data with patients divided into ING5-High and ING5-
Low groups showed a clear trend of lower survival associated with high ING5 expression, although the P-value fell short of reaching statistical significance (Figure 34). Since GBM is a very heterogeneous cancer, we tested whether ING5 was more relevant to the prognosis of tumors displaying more stem cell-like features. Stem cell-like BTIC lines are associated with a Proneural signature, indicating the Proneural subtype is probably driven more by the properties of stem cells than the proliferative nature of progenitor cells (Cusulin et al. 2015). Consistent with this, we found that high ING5 expression (red curve) correlated strongly with lower survival in the Proneural subtype, but not in other subtypes (Figure 35). Crosstabulation analysis on gene expression across
174 GBM samples showed that ING5 was positively correlated with expression of the stem cell factor SOX2 (Table 3). SOX2 is a master regulator and persistent marker for stemness in pluripotent cells and neural stem cells (Wegner and Stolt 2005). The correlation of ING5 with survival was most strongly seen in the SOX2-Low group, but not in the SOX2-High group (Figure
36), indicating a potential functional interaction between these two stem cell regulators.
101
Figure 42. ING5 levels negatively correlate with survival of GBM. Kaplan-Meier survival analysis of TCGA GBM patients with high and low levels of ING5 expression (stratified by mean value, n = 114, P = 0.077). Clinical data obtained from TCGA.
102
Figure 43. The subtype-specific correlation of ING5 expression level with GBM survival.
ING5 expression levels negatively correlate with survival in the Proneural subtype of GBM patients (n = 24). High expression levels of ING5 do not show significant correlation with survival in Classical (n = 30), Mesenchymal (n = 35) or Neural (n = 23) subtypes. The G-CIMP subtype was excluded from the analysis due to low sample number. Clinical data were obtained from TCGA.
103
Table 3. Crosstabulation of ING5 and SOX2 showing positive correlation of gene expression across GBM samples (P = 3.0178E-9).
SOX2 Median Total Low High
ING5 Median Low 62 23 85
High 23 61 84
Total 85 84 169
P-value =3.0178E-9
104
Figure 44. The potential effect of ING5 expression on GBM survival is affected by SOX2 status. ING5 expression level is significantly correlated with lower survival of GBM patients in the SOX2-Low group while no correlation was seen in the SOX2-High group (ING5 and SOX2 stratified by the median values). Clinical data were obtained from TCGA.
105
CHAPTER FOUR: DISCUSSION
Cancer is a stem cell disease. Recent studies suggest cancer stem cells are at the top of an intratumoral differentiation hierarchy, and are more responsible for tumor progression and recurrence than non-stem cancer cells, particularly in forms of cancers that are resistant to treatment. Brain tumor initiating cells are believed to originate from disturbed neurogenesis programs in combination with mitogenic mutations, however, how their long-term tumor initiating capability and intratumoral hierarchy are maintained is not well understood. The stemness of
BTICs was known to be associated with their epigenetic states, and here we show that ING5, a member of the ING family of epigenetic regulators, promotes self-renewal of BTICs, inhibits their differentiation, and does this by affecting the expression of genes that directly control the calcium signaling and the follicle stimulating hormone pathways.
The first indication of ING5 being involved in stem cell regulation came from our preliminary observation that ING5 expression was correlated with the pluripotent status of embryonic stem cells. The ING5 mRNA and protein levels were significantly higher in undifferentiated ESC clones compared to cells at the second day of differentiation. We thought this pattern of expression might be related to the function of ING5 specifically in stem cells but not in differentiated cells, and that downregulation of ING5 might be necessary for cells to undergo differentiation. We extended this observation into BTICs. Similar to the pluripotent cells, multipotent BTICs also showed ING5 downregulation during differentiation in vitro, which was consistently seen in three BTIC lines with different genetic backgrounds. This prompted us to further look at the functions of ING5 in BTIC stem cell properties. The direct function of ING5 in stem cell regulation was reported in a study that screened for chromatin modifiers in stemness
106 maintenance of epidermal stem cells (Mulder et al. 2012). Consistent with our observation in
ESCs, ING5 was identified as a stemness maintenance factor which prevented cell differentiation.
ING5 showed genetic interactions with other epigenetic factors and co-localized with the histone marker H3K4me3 on the promoter of genes activated in stem cells but silenced during differentiation. These observations are in line with its role in histone mark targeting and activity in modifying histone acetylation. So, we proposed that ING5 might regulate the stem cell properties in BTICs through epigenetic mechanisms.
In this study we found that ING5 played a role in promoting self-renewal maintenance in
BTICs while repressing lineage commitment of BTICs into mature neural cells. Cancer stem cells display abnormal features compared to the genetically intact stem cells, including the hyperproliferative activity, increased symmetric division frequency and disturbed differentiation programs. To distinguish stemness properties from the proliferation-driven phenotypes, we characterized several independent stem cell features of BTICs using different methods. We analyzed the sphere formation capability from single cells, CD133 positive cell profiling in the whole population, cell division mode analysis and differentiation propensity of cells in both ING5 overexpression and knockdown systems. All of these experiments indicate that ING5 functions to promote the stemness of BTICs and contributes to sustain their long-term propagation in vitro, especially in the absence of growth factors. And we found that this function was not associated with its role in DNA synthesis, which depends on the histone acetylation of ING5-HBO1 complex on DNA replication initiation sites (Doyon et al. 2006). Cell cycle analysis showed no effect of
ING5 on the distribution of cells in different cell cycle phases, suggesting the enhanced sphere formation capability of BTICs by ING5 was not due to an increase of proliferation rate, but from the sustained self-renewal ability of stem cells. We also found that the core stem cell factors OCT4
107 and neural stem cell/progenitor markers OLIG2, CD133 and Nestin were upregulated by ectopic expression of ING5, which might synergistically function with ING5 to promote BTIC stemness.
In our CD133 FACS analyses, we observed a fluctuation of the percentage of CD133 positive cells from different experiments. The percentage of CD133+ cells in the control groups varies from
42% (Figure 9), 76% (Figure 18), 45.8% (Figure 26) to 72% (Figure 31B). This discrepancy could be generated due to the following reasons: 1) Although all of the control cells were from the
BT 189 cell line, they went through different experimental procedures in specific experiment context. For example, the cells in Figure 9 were iPB stable cell lines constructed from BT 189 cells. They were selected and induced after they were introduced with PiggyBac transposon elements. The selection and induction processes can possibly change to some extent the cell properties and the expression of CD133. The control cells in Figure 18 were BT 189 cells treated with DMSO control (0.1 % v/v) for 3 days, while the BT 189 cells were treated with DMSO (0.1
% v/v) for 2 days for the control in Figure 26. Since cells were seeded at 2 × 104 cells/mL in both cases, different time of growth could result in a difference in growth kinetics and therefore affect the CD133 positive populations. It has been reported that the levels of CD133 in glioma cells fluctuates during cell cycle which can lead to a 300% change of CD133 positive cells (Barrantes-
Freer et al. 2015). The control cells in Figure 31B were BT 189 cells treated with 0.5 µg/mL IgG for 3 days. Therefore, different treatment, growth conditions (with or without cumate induction) and growth time might result in an inconsistent percentage of CD133 in these experiments. 2)
Another factor that may affect the percentage of CD133 is the gating of negative control (isotype antibody). Although the staining and washing steps followed the same protocol among these experiments as described in 2.10, gating of cell populations depends on the intensity of a small number of the brightest control cells which may vary in different experiments. However, the
108 experimental operation and cell gating is consistent among all groups within each experiment, so the control and experimental groups could be compared in the same experiment context.
Since the growth factor-RTK signaling pathways can support the self-renewal of BTIC population in culture and help preserve their original tumor initiating and stem cell properties (Lee et al. 2006), we tested whether ING5 had a different effect on BTICs in the absence of growth factors. We found that ING5 enhanced the self-renewal ability to a greater extent when growth factors are absent from the culture medium than when the growth factors were present (Figure
15). Over serial passages under growth factor deprivation, control cells gradually lost self-renewal ability and could not grow into spheres. ING5 overexpression, continuously induced by cumate in the iPB-ING5 cell line, maintained self-renewal ability and sphere growth at the same level seen during the first passage. To address whether ING5 induced the RTK signaling pathways independent of the activation due to growth factor binding, we targeted three major RTK signaling pathways by protein kinase inhibitors and found that the PI3K/AKT and MEK/ERK pathways were induced by ING5. Consistent with our observation, ING5 was recently reported to activate the AKT signaling in gastric cancer cells, which enhanced chemoresistance in these cells (Gou et al. 2015). The PI3K and RAS/ERK pathways have been shown to block differentiation and induce dedifferentiation from mature lineages during gliomagenesis (Holland et al. 2000), however the downstream effectors and target genes in these processes were less clear. ERK signaling has been reported to antagonize neuronal differentiation by promoting the transactivation function of
ASCL1 which leads to OLIG2 expression. Our results confirmed that both ERK and PI3K pathways inhibited neuronal differentiation in BTICs. Moreover, we found that the MEK/ERK pathway had an impact on the proportion of CD133 positive cells, but the PI3K pathway did not have such an effect, suggesting the two pathways regulated the stem cell properties of BTICs
109 through distinct mechanisms and the MEK/ERK pathway might be more directly linked to stemness features induced by ING5. ING5 also increases the expression of OLIG2, indicating it may regulate a subset of stem cell-related genes through the MEK/ERK pathway.
We then further investigated the function of ING5 in BTIC regulation by examining the
BTIC transcriptome. The ING family is highly conserved and ING5 shares over 90 percent sequence identity with ING4 (He et al. 2005). However, the two proteins have divergent functions probably through incorporating into distinct histone modification complexes and showing different expression patterns. ING5 is a stoichiometric unit of three HATs, MOZ, MORF and HBO1, which are essential regulators of embryonic development and stem cells. ING4 is associated with HBO1 but not MOZ/MORF complexes, and does not show decreased expression during stem cell differentiation. We performed transcriptome assays using Affymetrix Human Gene 2.0 ST arrays and compared differentially expressed genes (DEGs) between ING5 knockdown cells (shR-ING5) and non-targeting (shR-ctr) cells. There was a larger proportion of genes downregulated than upregulated in the shR-ING5 group. We observed that ING5 could promote self-renewal as well as suppress differentiation, which can be mediated by activation of stemness factors or repression of differentiation-promoting factors. Here our transcriptome data showed that ING5 functioned primarily to activate rather than repress gene expression. This suggests a fundamentally different mechanism from several other reported epigenetic regulators in BTIC maintenance. For example, the EZH2-PRC2 complex causes promoter H3K27me3 modification and silences the expression of BMPR1B , a critical factor in the BMP pathway, leading to astrocyte differentiation in BTICs
(Lee et al. 2008). The silenced BMP pathway bestowed BTICs with a resistance to differentiation and higher tumor initiating capability. This indicates that distinct epigenetic complexes regulate the gene expression program and stemness of BTIC in diverse ways.
110
To gain functional insights from the transcriptome data, we used the Gene Set Enrichment
Analysis website-based tool to search for biological functions enriched in the DEGs. The Gene
Ontology term suggested a significant enrichment for calcium channel activity. We validated the upregulation of a set of calcium channel component genes by ING5 and monitored the intracellular calcium levels of BTICs. We revealed that calcium levels were regulated by ING5, probably through the plasma membrane entry of calcium. This was an interesting observation considering the versatile functions of calcium as a second messenger in various cell signaling pathways. In
GBM, a number of tyrosine kinase receptors and GPCRs are activated by extracellular signals, all of which can potentially mobilize the second messenger calcium. However, the role of calcium in
BTIC regulation is not well understood besides a limited number of studies suggesting that it promoted survival and invasion of non-stem GBM cells (Holland et al. 2000, Kang et al. 2010).
We investigated whether this high level of intracellular calcium maintained by ING5 was related to the self-renewal properties of BTICs, by increasing or decreasing the concentration of calcium ions with multiple calcium modulators. We found that increasing calcium levels enhanced the sphere formation capabilities of cells while reduction of intracellular calcium decreased sphere formation rates (Figure 27 and Figure 30). Calcium also increased the CD133 positive population in BTICs, suggesting that calcium signaling was indeed involved in a mechanism for stemness maintenance. We further narrowed down the downstream effectors of calcium signaling to the calcineurin (CaN) axis. CaN is a calcium and calmodulin-dependent serine/threonine protein phosphatase which activates downstream proteins by removing their phosphate groups. It has been reported that the CaN-NFATc4 (nuclear factor of activated T cell, cytoplasmic 4) signaling axis is a major regulator of neural stem cell self-renewal and proliferation in a hypoxic microenvironment within the brain. Considering the hypoxic perivascular niche of BTICs, it is likely that CaN
111 signaling is also involved in the hypoxia-induced gene expression program in these cells to promote their self-renewal (Moreno et al. 2015). CaN also links calcium signaling with the
MEK/ERK pathway through one of its substrates, Kinase Suppressor of Ras 2 (KSR2), which promotes MEK phosphorylation by Raf (Brennan et al. 2011). In addition, KSR2 is a potential target gene with a 9-fold increase in response to ING5 according to our microarray result. These results indicate that ING5 may regulate the stem cell properties of BTICs through the Calcium-
CaN axis which may also stimulate and synergize with the MEK/ERK signaling pathway.
From the sphere formation experiments with calcium inducers we noticed that the increase of calcium level alone was insufficient to elevate the sphere formation capability to the level of
ING5 overexpressing cells (Figure 30), suggesting there were other signaling pathways that transduced the effects of ING5. We performed pathway analysis using the IPA program and found a number of downstream functions related to hormone signaling pathways were activated by ING5, such as steroidogenesis and germ cell development. Among these functions, the quantity of FSH
(Table 2), quantity of steroid (Table 4) and quantity of thyroid hormone (Table 5) showed high
Z-Scores. This indicates that ING5 might activate a network of genes regulating hormone synthesis and signaling that are essential for germ cell development and growth stimulation. Among the target genes within this functional category, the Nuclear Receptor Subfamily 5 Group A Member
1 (NR5A1, or steroidogenic factor 1, SF1) is a core factor which induces steroidogenesis and the expression of a variety of hormones. Interestingly, SF1 has been shown to activate OCT4 expression by an SF1-binding element recognized in the OCT4 proximal promoter (Yang et al.
2007), and it has been shown to induce a ground state pluripotency by reprogramming the epiblast stem cells (Guo and Smith 2010). This steroidogenesis-hormone network induced by ING5 in
112
Table 4. Gene list of Quantity of Steroid from the IPA downstream function.
QUANTITY OF STEROID Z-SCORE: 1.718 GENE ID Fold change (ctr vs. shR) NR1I2 6.95 NDST3 6.56 FGL1 5.76 HGF 5.60 VCAM1 5.07 SLC6A4 5.06 GH1 4.52 FSHR 4.31 RAG2 3.69 PNLIP 3.50 APOC1 3.31 CYP8B1 3.11 PEX5L 3.07 KLB 2.96 SCNN1A 2.11 ATF3 2.06 CGA 2.02 IL1R1 1.96 FSHB 1.88 PTGIR 1.78 ATF -6.37
113
Table 5. Gene list of Quantity of Thyroid Hormone from the IPA downstream function.
QUANTITY OF THYROID HORMONE Z-SCORE: 2.117 GENE ID Fold change (ctr vs. shR) VGF 4.39 CD28 2.35 CGA 2.02 FRK 2.86 LEP 5.31 NPY -5.10 TTR 2.03 TRH 7.63 AGPAT2 3.72
114
BTICs may potentially indicate a universal mechanism for ING5 in maintaining stemness in various pluripotent and multipotent stem cell systems.
We wanted to address whether these hormone signaling pathways actually mediated the functions of ING5 in BTICs, and focused on the FSH pathway because both the ligand (FSHB) and the receptor of the pathway (FSHR) were elevated by ING5 (Figure 29A). Although there are modest expression levels of FSHB and FSHR in GBM tissues according to TCGA data, we observed a huge increase of the number of FSH receptors on the cell membrane of ING5 overexpressing cells (Figure 29C). Moreover, blocking the FSH pathway alone diminished the self-renewal ability of BTICs. In ovarian cancer cells FSH induces OCT4 expression through the
ERK pathway and promotes cell invasion (Liu et al. 2015). We speculated that FSH might act through the same mechanism of OCT4 activation in the maintenance of BTIC self-renewal.
Although there is relatively less knowledge about these hormone pathways in the brain compared to endocrine tissues, they have been reported to regulate important events during mammalian neurodevelopment and neural stem cell properties. These hormone pathways usually play a neurotrophic role involving intracellular kinase signaling, calcium signaling and the modulation of growth factors. For example, estrogen promotes proliferation of neural stem cells/progenitors in both embryos and adults of mice through the ERK signaling pathway
(Brannvall, Korhonen, and Lindholm 2002, Okada et al. 2010). The pregnancy hormones chorionic gonadotropin and progesterone promote neuroectodermal cell fate and generation of rossets in early embryos (Gallego et al. 2010). Thyroid hormone can serve as a growth factor for glioma cells through interaction with membrane receptors and the downstream ERK pathway (Davis et al.
2006). Here we suggest that the follicle stimulating hormone pathway plays a role in the self- renewal of BTICs, besides its canonical function in ovarian cells. This is not without precedent
115 since crosstalk between FSH and calcium signaling pathways has been reported. In ovarian granulosa cells, FSH evokes an increase of intracellular calcium level (Flores, Veldhuis, and Leong
1990). We also tested whether the FSH pathway affected intracellular calcium levels but did not observe changes in calcium when FSHR was blocked by antibody treatment. However, it is possible that the downstream function of FSH is mediated by the calcium signal activated by ING5, since it has been reported that the calcium-calmodulin system was involved in the steroidogenesis regulated by FSH (Carnegie and Tsang 1984, Lai et al. 2014). Alternatively, FSH may act independently of calcium signaling by inducing the ERK pathway and expression of OCT4 as shown in two previous studies (Zhang et al. 2013) (Liu et al. 2015) and supported by our results of sphere assays with combined treatment (Figure 30B). The interaction between these two pathways in BTICs is worth further investigations.
To determine whether the PHD motif of ING5 was critical for its function in BTICs, we constructed an iPB cell line overexpressing the PHD-deleted ING5 protein (ΔPHD). Ectopic expression of ING5-ΔPHD in BTICs showed indistinguishable sphere formation capability from control cells, which was significantly lower than the wildtype ING5 overexpressing cells (Figure
32B). This suggests that the PHD motif that specifically targets ING5 to the H3K4me3 mark is indispensable for the function of ING5 in BTIC self-renewal. This may provide insights into the pharmaceutical targeting of ING5 to antagonize stemness properties of BTICs.
We also found correlations between ING5 expression levels and the clinical outcomes of
GBM patients. We found that high ING5 levels correlated strongly with poor survival in patients with the Proneural subtype of GBM. Since the BTIC lines exhibiting typical stem cell properties, rather than progenitor properties resemble the Proneural subtype from transcription profiling, this correlation between ING5 and the Proneural group is consistent with its function in promoting
116
BTIC stemness. A strong positive correlation between ING5 and SOX2 expression was also observed, confirming the idea that the BTIC population is likely responsible, to a large extent, for the poor prognosis of GBM patients. It is also intriguing that the ING5 high expression is associated with poor prognosis in the SOX2-Low group of patients, but not in SOX2-High group, implicating a functional interaction between the two factors in BTIC stemness and tumor progression. Although the expression of SOX2 was not directly regulated by ING5, they share some common target genes, including OCT4, KLF4, genes encoding for integrins and cytokines
(Berezovsky et al. 2014), a subset of which were found in the ING5 microarray data and validated using qPCR. Therefore, ING5 and SOX2 may co-occupy particular promoters to cooperatively regulate a set of target genes. Interestingly, SOX2 and OLIG2 have been shown to induce reprogramming of the non-stem GBM cells into BTICs, which engages extensive epigenetic remodeling (Suva et al. 2014). Given that ING5 is associated with both factors, it would be interesting to see whether ING5 also regulates the epigenetic landscape of BTICs during differentiation and reprogramming.
117
CHAPTER FIVE: CONCLUSIONS
We found that ING5 was highly expressed in undifferentiated BTICs and immediately downregulated upon induced differentiation into mature neural lineages, which was consistent with our observation in the differentiation of mouse embryonic stem cells. We further investigated the function of ING5 in BTIC stem cell features by performing sphere formation assays, CD133 positive cell profiling and in vitro differentiation analysis. We found ING5 promoted the stemness of BTICs while prevented lineage differentiation, and this effect was more prominent under conditions of growth factor deprivation. ING5 overexpression significantly sustained self-renewal of BTICs over serial sphere passages in the absence of growth factors, due to the higher levels of activation of the PI3K and ERK pathways. Transcriptome analysis revealed the intracellular calcium level and follicle stimulating hormone pathways were induced by ING5 and by conducting sphere formation assays we found that both functions were required for the maintenance of BTIC self-renewal. Taken together, our results suggest that ING5 is a stem cell regulator in BTICs and enhances its stemness features. Since BTICs drive tumor initiation and recurrence in glioblastomas and their stem cell properties make them difficult to eradicate, our results provide novel insights into the molecular mechanisms of BTIC stemness and suggest therapeutic strategies that may prove useful in targeting these cells. Based on our observations in vitro, it would be worthwhile to investigate the effects of ING5 on tumorigenicity using mouse models to further test its potential as a cancer stem cell target in vivo. In addition, it will be interesting to test whether inhibiting the
FSH and calcium signaling pathways in combination with inhibitors of the PI3K and RAS/ERK pathways would provide better efficacy for BTIC elimination and tumor suppression.
118
REFERENCES
2008. "Comprehensive genomic characterization defines human glioblastoma genes and core pathways." Nature 455 (7216):1061-8. doi: 10.1038/nature07385. Aguirre-Ghiso, J. A. 2007. "Models, mechanisms and clinical evidence for cancer dormancy." Nat Rev Cancer 7 (11):834-46. doi: 10.1038/nrc2256. Al-Hajj, M., M. S. Wicha, A. Benito-Hernandez, S. J. Morrison, and M. F. Clarke. 2003. "Prospective identification of tumorigenic breast cancer cells." Proc Natl Acad Sci U S A 100 (7):3983-8. doi: 10.1073/pnas.0530291100. Alcantara Llaguno, S., J. Chen, C. H. Kwon, E. L. Jackson, Y. Li, D. K. Burns, A. Alvarez- Buylla, and L. F. Parada. 2009. "Malignant astrocytomas originate from neural stem/progenitor cells in a somatic tumor suppressor mouse model." Cancer Cell 15 (1):45-56. doi: 10.1016/j.ccr.2008.12.006. Alcantara Llaguno, S. R., Z. Wang, D. Sun, J. Chen, J. Xu, E. Kim, K. J. Hatanpaa, J. M. Raisanen, D. K. Burns, J. E. Johnson, and L. F. Parada. 2015. "Adult Lineage-Restricted CNS Progenitors Specify Distinct Glioblastoma Subtypes." Cancer Cell 28 (4):429-40. doi: 10.1016/j.ccell.2015.09.007. Alvarez-Buylla, A., and J. M. Garcia-Verdugo. 2002. "Neurogenesis in adult subventricular zone." J Neurosci 22 (3):629-34. Ambros, I. M., J. Hata, V. V. Joshi, B. Roald, L. P. Dehner, H. Tuchler, U. Potschger, and H. Shimada. 2002. "Morphologic features of neuroblastoma (Schwannian stroma-poor tumors) in clinically favorable and unfavorable groups." Cancer 94 (5):1574-83. Androutsellis-Theotokis, A., R. R. Leker, F. Soldner, D. J. Hoeppner, R. Ravin, S. W. Poser, M. A. Rueger, S. K. Bae, R. Kittappa, and R. D. McKay. 2006. "Notch signalling regulates stem cell numbers in vitro and in vivo." Nature 442 (7104):823-6. doi: 10.1038/nature04940. Awe, J. P., and J. A. Byrne. 2013. "Identifying candidate oocyte reprogramming factors using cross-species global transcriptional analysis." Cell Reprogram 15 (2):126-33. doi: 10.1089/cell.2012.0060. Ayuso-Sacido, A., J. A. Moliterno, S. Kratovac, G. S. Kapoor, D. M. O'Rourke, E. C. Holland, J. M. Garcia-Verdugo, N. S. Roy, and J. A. Boockvar. 2010. "Activated EGFR signaling increases proliferation, survival, and migration and blocks neuronal differentiation in post-natal neural stem cells." J Neurooncol 97 (3):323-37. doi: 10.1007/s11060-009- 0035-x. Bachoo, R. M., E. A. Maher, K. L. Ligon, N. E. Sharpless, S. S. Chan, M. J. You, Y. Tang, J. DeFrances, E. Stover, R. Weissleder, D. H. Rowitch, D. N. Louis, and R. A. DePinho. 2002. "Epidermal growth factor receptor and Ink4a/Arf: convergent mechanisms governing terminal differentiation and transformation along the neural stem cell to astrocyte axis." Cancer Cell 1 (3):269-77. Bao, S., Q. Wu, R. E. McLendon, Y. Hao, Q. Shi, A. B. Hjelmeland, M. W. Dewhirst, D. D. Bigner, and J. N. Rich. 2006. "Glioma stem cells promote radioresistance by preferential activation of the DNA damage response." Nature 444 (7120):756-60. doi: 10.1038/nature05236.
119
Barrantes-Freer, A., M. Renovanz, M. Eich, A. Braukmann, B. Sprang, P. Spirin, L. A. Pardo, A. Giese, and E. L. Kim. 2015. "CD133 Expression is Not Synonymous to Immunoreactivity for AC133 and Fluctuates Throughout the Cell Cycle in Glioma Stem- Like Cells." PloS One 10 (6): e0130519. Beier, D., P. Hau, M. Proescholdt, A. Lohmeier, J. Wischhusen, P. J. Oefner, L. Aigner, A. Brawanski, U. Bogdahn, and C. P. Beier. 2007. "CD133(+) and CD133(-) Glioblastoma- Derived Cancer Stem Cells show Differential Growth Characteristics and Molecular Profiles." Cancer Research 67 (9): 4010-4015. Berezovsky, A. D., L. M. Poisson, D. Cherba, C. P. Webb, A. D. Transou, N. W. Lemke, X. Hong, L. A. Hasselbach, S. M. Irtenkauf, T. Mikkelsen, and A. C. deCarvalho. 2014. "Sox2 promotes malignancy in glioblastoma by regulating plasticity and astrocytic differentiation." Neoplasia 16 (3):193-206, 206 e19-25. doi: 10.1016/j.neo.2014.03.006. Berger, P. L., S. B. Frank, V. V. Schulz, E. A. Nollet, M. J. Edick, B. Holly, T. T. Chang, G. Hostetter, S. Kim, and C. K. Miranti. 2014. "Transient induction of ING4 by Myc drives prostate epithelial cell differentiation and its disruption drives prostate tumorigenesis." Cancer Res 74 (12):3357-68. doi: 10.1158/0008-5472.can-13-3076. Bernstein, B. E., T. S. Mikkelsen, X. Xie, M. Kamal, D. J. Huebert, J. Cuff, B. Fry, A. Meissner, M. Wernig, K. Plath, R. Jaenisch, A. Wagschal, R. Feil, S. L. Schreiber, and E. S. Lander. 2006. "A bivalent chromatin structure marks key developmental genes in embryonic stem cells." Cell 125 (2):315-26. doi: 10.1016/j.cell.2006.02.041. Bhartiya, D., and J. Singh. 2015. "FSH-FSHR3-stem cells in ovary surface epithelium: basis for adult ovarian biology, failure, aging, and cancer." Reproduction 149 (1):R35-48. doi: 10.1530/REP-14-0220. Bidlingmaier, S., X. Zhu, and B. Liu. 2008. "The Utility and Limitations of Glycosylated Human CD133 Epitopes in Defining Cancer Stem Cells." Journal of Molecular Medicine (Berlin, Germany) 86 (9): 1025-1032. Bignami, A., L. F. Eng, D. Dahl, and C. T. Uyeda. 1972. "Localization of the glial fibrillary acidic protein in astrocytes by immunofluorescence." Brain Res 43 (2):429-35. Bondy, M. L., M. E. Scheurer, B. Malmer, J. S. Barnholtz-Sloan, F. G. Davis, D. Il'yasova, C. Kruchko, B. J. McCarthy, P. Rajaraman, J. A. Schwartzbaum, S. Sadetzki, B. Schlehofer, T. Tihan, J. L. Wiemels, M. Wrensch, and P. A. Buffler. 2008. "Brain tumor epidemiology: consensus from the Brain Tumor Epidemiology Consortium." Cancer 113 (7 Suppl):1953-68. doi: 10.1002/cncr.23741. Bonnet, D., and J. E. Dick. 1997. "Human acute myeloid leukemia is organized as a hierarchy that originates from a primitive hematopoietic cell." Nat Med 3 (7):730-7. Bose, P., S. Thakur, S. Thalappilly, B. Y. Ahn, S. Satpathy, X. Feng, K. Suzuki, S. W. Kim, and K. Riabowol. 2013. "ING1 induces apoptosis through direct effects at the mitochondria." Cell Death Dis 4:e788. doi: 10.1038/cddis.2013.321. Brannvall, K., L. Korhonen, and D. Lindholm. 2002. "Estrogen-receptor-dependent regulation of neural stem cell proliferation and differentiation." Mol Cell Neurosci 21 (3):512-20. Brat, D. J., R. G. Verhaak, K. D. Aldape, W. K. Yung, S. R. Salama, L. A. Cooper, E. Rheinbay, C. R. Miller, M. Vitucci, O. Morozova, A. G. Robertson, H. Noushmehr, P. W. Laird, A. D. Cherniack, R. Akbani, J. T. Huse, G. Ciriello, L. M. Poisson, J. S. Barnholtz-Sloan, M. S. Berger, C. Brennan, R. R. Colen, H. Colman, A. E. Flanders, C. Giannini, M. Grifford, A. Iavarone, R. Jain, I. Joseph, J. Kim, K. Kasaian, T. Mikkelsen, B. A. Murray,
120
B. P. O'Neill, L. Pachter, D. W. Parsons, C. Sougnez, E. P. Sulman, S. R. Vandenberg, E. G. Van Meir, A. von Deimling, H. Zhang, D. Crain, K. Lau, D. Mallery, S. Morris, J. Paulauskis, R. Penny, T. Shelton, M. Sherman, P. Yena, A. Black, J. Bowen, K. Dicostanzo, J. Gastier-Foster, K. M. Leraas, T. M. Lichtenberg, C. R. Pierson, N. C. Ramirez, C. Taylor, S. Weaver, L. Wise, E. Zmuda, T. Davidsen, J. A. Demchok, G. Eley, M. L. Ferguson, C. M. Hutter, K. R. Mills Shaw, B. A. Ozenberger, M. Sheth, H. J. Sofia, R. Tarnuzzer, Z. Wang, L. Yang, J. C. Zenklusen, B. Ayala, J. Baboud, S. Chudamani, M. A. Jensen, J. Liu, T. Pihl, R. Raman, Y. Wan, Y. Wu, A. Ally, J. T. Auman, M. Balasundaram, S. Balu, S. B. Baylin, R. Beroukhim, M. S. Bootwalla, R. Bowlby, C. A. Bristow, D. Brooks, Y. Butterfield, R. Carlsen, S. Carter, L. Chin, A. Chu, E. Chuah, K. Cibulskis, A. Clarke, S. G. Coetzee, N. Dhalla, T. Fennell, S. Fisher, S. Gabriel, G. Getz, R. Gibbs, R. Guin, A. Hadjipanayis, D. N. Hayes, T. Hinoue, K. Hoadley, R. A. Holt, A. P. Hoyle, S. R. Jefferys, S. Jones, C. D. Jones, R. Kucherlapati, P. H. Lai, E. Lander, S. Lee, L. Lichtenstein, Y. Ma, D. T. Maglinte, H. S. Mahadeshwar, M. A. Marra, M. Mayo, S. Meng, M. L. Meyerson, P. A. Mieczkowski, R. A. Moore, L. E. Mose, A. J. Mungall, A. Pantazi, M. Parfenov, P. J. Park, J. S. Parker, C. M. Perou, A. Protopopov, X. Ren, J. Roach, T. S. Sabedot, J. Schein, S. E. Schumacher, J. G. Seidman, S. Seth, H. Shen, J. V. Simons, P. Sipahimalani, M. G. Soloway, X. Song, H. Sun, B. Tabak, A. Tam, D. Tan, J. Tang, N. Thiessen, T. Triche, Jr., D. J. Van Den Berg, U. Veluvolu, S. Waring, D. J. Weisenberger, M. D. Wilkerson, T. Wong, J. Wu, L. Xi, A. W. Xu, L. Yang, T. I. Zack, J. Zhang, B. A. Aksoy, H. Arachchi, C. Benz, B. Bernard, D. Carlin, J. Cho, D. DiCara, S. Frazer, G. N. Fuller, J. Gao, N. Gehlenborg, D. Haussler, D. I. Heiman, L. Iype, A. Jacobsen, Z. Ju, S. Katzman, H. Kim, T. Knijnenburg, R. B. Kreisberg, M. S. Lawrence, W. Lee, K. Leinonen, P. Lin, S. Ling, W. Liu, Y. Liu, Y. Liu, Y. Lu, G. Mills, S. Ng, M. S. Noble, E. Paull, A. Rao, S. Reynolds, G. Saksena, Z. Sanborn, C. Sander, N. Schultz, Y. Senbabaoglu, R. Shen, I. Shmulevich, R. Sinha, J. Stuart, S. O. Sumer, Y. Sun, N. Tasman, B. S. Taylor, D. Voet, N. Weinhold, J. N. Weinstein, D. Yang, K. Yoshihara, S. Zheng, W. Zhang, L. Zou, T. Abel, S. Sadeghi, M. L. Cohen, J. Eschbacher, E. M. Hattab, A. Raghunathan, M. J. Schniederjan, D. Aziz, G. Barnett, W. Barrett, D. D. Bigner, L. Boice, C. Brewer, C. Calatozzolo, B. Campos, C. G. Carlotti, Jr., T. A. Chan, L. Cuppini, E. Curley, S. Cuzzubbo, K. Devine, F. DiMeco, R. Duell, J. B. Elder, A. Fehrenbach, G. Finocchiaro, W. Friedman, J. Fulop, J. Gardner, B. Hermes, C. Herold-Mende, C. Jungk, A. Kendler, N. L. Lehman, E. Lipp, O. Liu, R. Mandt, M. McGraw, R. McLendon, C. McPherson, L. Neder, P. Nguyen, A. Noss, R. Nunziata, Q. T. Ostrom, C. Palmer, A. Perin, B. Pollo, A. Potapov, O. Potapova, W. K. Rathmell, D. Rotin, L. Scarpace, C. Schilero, K. Senecal, K. Shimmel, V. Shurkhay, S. Sifri, R. Singh, A. E. Sloan, K. Smolenski, S. M. Staugaitis, R. Steele, L. Thorne, D. P. Tirapelli, A. Unterberg, M. Vallurupalli, Y. Wang, R. Warnick, F. Williams, Y. Wolinsky, S. Bell, M. Rosenberg, C. Stewart, F. Huang, J. L. Grimsby, A. J. Radenbaugh, and J. Zhang. 2015. "Comprehensive, Integrative Genomic Analysis of Diffuse Lower-Grade Gliomas." N Engl J Med 372 (26):2481-98. doi: 10.1056/NEJMoa1402121. Bredel, M., D. M. Scholtens, G. R. Harsh, C. Bredel, J. P. Chandler, J. J. Renfrow, A. K. Yadav, H. Vogel, A. C. Scheck, R. Tibshirani, and B. I. Sikic. 2009. "A network model of a
121
cooperative genetic landscape in brain tumors." JAMA 302 (3):261-75. doi: 10.1001/jama.2009.997. Brennan, C. W., R. G. Verhaak, A. McKenna, B. Campos, H. Noushmehr, S. R. Salama, S. Zheng, D. Chakravarty, J. Z. Sanborn, S. H. Berman, R. Beroukhim, B. Bernard, C. J. Wu, G. Genovese, I. Shmulevich, J. Barnholtz-Sloan, L. Zou, R. Vegesna, S. A. Shukla, G. Ciriello, W. K. Yung, W. Zhang, C. Sougnez, T. Mikkelsen, K. Aldape, D. D. Bigner, E. G. Van Meir, M. Prados, A. Sloan, K. L. Black, J. Eschbacher, G. Finocchiaro, W. Friedman, D. W. Andrews, A. Guha, M. Iacocca, B. P. O'Neill, G. Foltz, J. Myers, D. J. Weisenberger, R. Penny, R. Kucherlapati, C. M. Perou, D. N. Hayes, R. Gibbs, M. Marra, G. B. Mills, E. Lander, P. Spellman, R. Wilson, C. Sander, J. Weinstein, M. Meyerson, S. Gabriel, P. W. Laird, D. Haussler, G. Getz, and L. Chin. 2013. "The somatic genomic landscape of glioblastoma." Cell 155 (2):462-77. doi: 10.1016/j.cell.2013.09.034. Brennan, D. F., A. C. Dar, N. T. Hertz, W. C. Chao, A. L. Burlingame, K. M. Shokat, and D. Barford. 2011. "A Raf-induced allosteric transition of KSR stimulates phosphorylation of MEK." Nature 472 (7343):366-9. doi: 10.1038/nature09860. Bruggeman, S. W., D. Hulsman, E. Tanger, T. Buckle, M. Blom, J. Zevenhoven, O. van Tellingen, and M. van Lohuizen. 2007. "Bmi1 controls tumor development in an Ink4a/Arf-independent manner in a mouse model for glioma." Cancer Cell 12 (4):328- 41. doi: 10.1016/j.ccr.2007.08.032. Burney, M. J., C. Johnston, K. Y. Wong, S. W. Teng, V. Beglopoulos, L. W. Stanton, B. P. Williams, A. Bithell, and N. J. Buckley. 2013. "An epigenetic signature of developmental potential in neural stem cells and early neurons." Stem Cells 31 (9):1868-80. doi: 10.1002/stem.1431. Butler, R., J. Robertson, and J. M. Gallo. 2000. "Mutually exclusive expression of beta(III)- tubulin and vimentin in adrenal cortex carcinoma SW13 cells." FEBS Lett 470 (2):198- 202. Caceres, A., G. A. Banker, and L. Binder. 1986. "Immunocytochemical localization of tubulin and microtubule-associated protein 2 during the development of hippocampal neurons in culture." J Neurosci 6 (3):714-22. Campos, E. I., M. Martinka, D. L. Mitchell, D. L. Dai, and G. Li. 2004. "Mutations of the ING1 tumor suppressor gene detected in human melanoma abrogate nucleotide excision repair." Int J Oncol 25 (1):73-80. Caren, H., S. H. Stricker, H. Bulstrode, S. Gagrica, E. Johnstone, T. E. Bartlett, A. Feber, G. Wilson, A. E. Teschendorff, P. Bertone, S. Beck, and S. M. Pollard. 2015. "Glioblastoma Stem Cells Respond to Differentiation Cues but Fail to Undergo Commitment and Terminal Cell-Cycle Arrest." Stem Cell Reports 5 (5):829-42. doi: 10.1016/j.stemcr.2015.09.014. Carnegie, J. A., and B. K. Tsang. 1984. "The calcium-calmodulin system: participation in the regulation of steroidogenesis at different stages of granulosa cell differentiation." Biol Reprod 30 (2):515-22. Ceccarelli, M., F. P. Barthel, T. M. Malta, T. S. Sabedot, S. R. Salama, B. A. Murray, O. Morozova, Y. Newton, A. Radenbaugh, S. M. Pagnotta, S. Anjum, J. Wang, G. Manyam, P. Zoppoli, S. Ling, A. A. Rao, M. Grifford, A. D. Cherniack, H. Zhang, L. Poisson, C. G. Carlotti, Jr., D. P. Tirapelli, A. Rao, T. Mikkelsen, C. C. Lau, W. K. Yung, R.
122
Rabadan, J. Huse, D. J. Brat, N. L. Lehman, J. S. Barnholtz-Sloan, S. Zheng, K. Hess, G. Rao, M. Meyerson, R. Beroukhim, L. Cooper, R. Akbani, M. Wrensch, D. Haussler, K. D. Aldape, P. W. Laird, D. H. Gutmann, H. Noushmehr, A. Iavarone, and R. G. Verhaak. 2016. "Molecular Profiling Reveals Biologically Discrete Subsets and Pathways of Progression in Diffuse Glioma." Cell 164 (3):550-63. doi: 10.1016/j.cell.2015.12.028. Cengiz, B., E. Gunduz, M. Gunduz, L. B. Beder, R. Tamamura, C. Bagci, N. Yamanaka, K. Shimizu, and H. Nagatsuka. 2010. "Tumor-specific mutation and downregulation of ING5 detected in oral squamous cell carcinoma." Int J Cancer 127 (9):2088-94. doi: 10.1002/ijc.25224. Chaffer, C. L., I. Brueckmann, C. Scheel, A. J. Kaestli, P. A. Wiggins, L. O. Rodrigues, M. Brooks, et al. 2011. "Normal and Neoplastic Nonstem Cells can Spontaneously Convert to a Stem-Like State." Proceedings of the National Academy of Sciences of the United States of America 108 (19): 7950-7955. Chai, S., X. Xu, Y. Wang, Y. Zhou, C. Zhang, Y. Yang, Y. Yang, H. Xu, R. Xu, and K. Wang. 2015. "Ca2+/calmodulin-dependent protein kinase IIgamma enhances stem-like traits and tumorigenicity of lung cancer cells." Oncotarget 6 (18):16069-83. doi: 10.18632/oncotarget.3866. Chang, K. C., C. Wang, and H. Wang. 2012. "Balancing self-renewal and differentiation by asymmetric division: insights from brain tumor suppressors in Drosophila neural stem cells." Bioessays 34 (4):301-10. doi: 10.1002/bies.201100090. Chen, J., Y. Li, T. S. Yu, R. M. McKay, D. K. Burns, S. G. Kernie, and L. F. Parada. 2012. "A restricted cell population propagates glioblastoma growth after chemotherapy." Nature 488 (7412):522-6. doi: 10.1038/nature11287. Cheung, K. J., Jr., and G. Li. 2002. "p33(ING1) enhances UVB-induced apoptosis in melanoma cells." Exp Cell Res 279 (2):291-8. Cox, C. V., P. Diamanti, R. S. Evely, P. R. Kearns, and A. Blair. 2009. "Expression of CD133 on Leukemia-Initiating Cells in Childhood ALL." Blood 113 (14): 3287-3296. Curley, M. D., V. A. Therrien, C. L. Cummings, P. A. Sergent, C. R. Koulouris, A. M. Friel, D. J. Roberts, M. V. Seiden, D. T. Scadden, B. R. Rueda, and R. Foster. 2009. "CD133 expression defines a tumor initiating cell population in primary human ovarian cancer." Stem Cells 27 (12):2875-83. doi: 10.1002/stem.236. Cusulin, C., C. Chesnelong, P. Bose, M. Bilenky, K. Kopciuk, J. A. Chan, J. G. Cairncross, S. J. Jones, M. A. Marra, H. A. Luchman, and S. Weiss. 2015. "Precursor States of Brain Tumor Initiating Cell Lines Are Predictive of Survival in Xenografts and Associated with Glioblastoma Subtypes." Stem Cell Reports 5 (1):1-9. doi: 10.1016/j.stemcr.2015.05.010. Dahlstrand, J., V. P. Collins, and U. Lendahl. 1992. "Expression of the class VI intermediate filament nestin in human central nervous system tumors." Cancer Res 52 (19):5334-41. Dalerba, P., S. J. Dylla, I. K. Park, R. Liu, X. Wang, R. W. Cho, T. Hoey, A. Gurney, E. H. Huang, D. M. Simeone, A. A. Shelton, G. Parmiani, C. Castelli, and M. F. Clarke. 2007. "Phenotypic characterization of human colorectal cancer stem cells." Proc Natl Acad Sci U S A 104 (24):10158-63. doi: 10.1073/pnas.0703478104. Davis, F. B., H. Y. Tang, A. Shih, T. Keating, L. Lansing, A. Hercbergs, R. A. Fenstermaker, A. Mousa, S. A. Mousa, P. J. Davis, and H. Y. Lin. 2006. "Acting via a cell surface receptor, thyroid hormone is a growth factor for glioma cells." Cancer Res 66 (14):7270-5. doi: 10.1158/0008-5472.can-05-4365.
123
Dick, J. E. 2008. "Stem cell concepts renew cancer research." Blood 112 (13):4793-807. doi: 10.1182/blood-2008-08-077941. Dolecek, T. A., J. M. Propp, N. E. Stroup, and C. Kruchko. 2012. "CBTRUS statistical report: primary brain and central nervous system tumors diagnosed in the United States in 2005- 2009." Neuro Oncol 14 Suppl 5:v1-49. doi: 10.1093/neuonc/nos218. Doyon, Y., C. Cayrou, M. Ullah, A. J. Landry, V. Cote, W. Selleck, W. S. Lane, S. Tan, X. J. Yang, and J. Cote. 2006. "ING tumor suppressor proteins are critical regulators of chromatin acetylation required for genome expression and perpetuation." Mol Cell 21 (1):51-64. doi: 10.1016/j.molcel.2005.12.007. Eapen, S. A., S. J. Netherton, K. P. Sarker, L. Deng, A. Chan, K. Riabowol, and S. Bonni. 2012. "Identification of a novel function for the chromatin remodeling protein ING2 in muscle differentiation." PLoS One 7 (7):e40684. doi: 10.1371/journal.pone.0040684. Ellis, P., B. M. Fagan, S. T. Magness, S. Hutton, O. Taranova, S. Hayashi, A. McMahon, M. Rao, and L. Pevny. 2004. "SOX2, a persistent marker for multipotential neural stem cells derived from embryonic stem cells, the embryo or the adult." Dev Neurosci 26 (2- 4):148-65. doi: 10.1159/000082134. Erfani, P., J. Tome-Garcia, P. Canoll, F. Doetsch, and N. M. Tsankova. 2015. "EGFR promoter exhibits dynamic histone modifications and binding of ASH2L and P300 in human germinal matrix and gliomas." Epigenetics 10 (6):496-507. doi: 10.1080/15592294.2015.1042645. Fasano, C. A., J. T. Dimos, N. B. Ivanova, N. Lowry, I. R. Lemischka, and S. Temple. 2007. "shRNA knockdown of Bmi-1 reveals a critical role for p21-Rb pathway in NSC self- renewal during development." Cell Stem Cell 1 (1):87-99. doi: 10.1016/j.stem.2007.04.001. Favaro, R., I. Appolloni, S. Pellegatta, A. B. Sanga, P. Pagella, E. Gambini, F. Pisati, S. Ottolenghi, M. Foti, G. Finocchiaro, P. Malatesta, and S. K. Nicolis. 2014. "Sox2 is required to maintain cancer stem cells in a mouse model of high-grade oligodendroglioma." Cancer Res 74 (6):1833-44. doi: 10.1158/0008-5472.can-13-1942. Fearon, E. R., P. J. Burke, C. A. Schiffer, B. A. Zehnbauer, and B. Vogelstein. 1986. "Differentiation of leukemia cells to polymorphonuclear leukocytes in patients with acute nonlymphocytic leukemia." N Engl J Med 315 (1):15-24. doi: 10.1056/nejm198607033150103. Feng, X., S. Bonni, and K. Riabowol. 2006. "HSP70 induction by ING proteins sensitizes cells to tumor necrosis factor alpha receptor-mediated apoptosis." Mol Cell Biol 26 (24):9244- 55. doi: 10.1128/mcb.01538-06. Feng, X., Y. Hara, and K. Riabowol. 2002. "Different HATS of the ING1 gene family." Trends Cell Biol 12 (11):532-8. Feuring-Buske, M., B. Gerhard, J. Cashman, R. K. Humphries, C. J. Eaves, and D. E. Hogge. 2003. "Improved engraftment of human acute myeloid leukemia progenitor cells in beta 2-microglobulin-deficient NOD/SCID mice and in NOD/SCID mice transgenic for human growth factors." Leukemia 17 (4):760-3. doi: 10.1038/sj.leu.2402882. Fidler, I. J., and M. L. Kripke. 1977. "Metastasis results from preexisting variant cells within a malignant tumor." Science 197 (4306):893-5. Flavahan, W. A., Y. Drier, B. B. Liau, S. M. Gillespie, A. S. Venteicher, A. O. Stemmer- Rachamimov, M. L. Suva, and B. E. Bernstein. 2016. "Insulator dysfunction and
124
oncogene activation in IDH mutant gliomas." Nature 529 (7584):110-4. doi: 10.1038/nature16490. Flores, J. A., J. D. Veldhuis, and D. A. Leong. 1990. "Follicle-stimulating hormone evokes an increase in intracellular free calcium ion concentrations in single ovarian (granulosa) cells." Endocrinology 127 (6):3172-9. doi: 10.1210/endo-127-6-3172. Fraichard, A., O. Chassande, G. Bilbaut, C. Dehay, P. Savatier, and J. Samarut. 1995. "In vitro differentiation of embryonic stem cells into glial cells and functional neurons." J Cell Sci 108 ( Pt 10):3181-8. Gallego, M. J., P. Porayette, M. M. Kaltcheva, R. L. Bowen, S. Vadakkadath Meethal, and C. S. Atwood. 2010. "The pregnancy hormones human chorionic gonadotropin and progesterone induce human embryonic stem cell proliferation and differentiation into neuroectodermal rosettes." Stem Cell Res Ther 1 (4):28. doi: 10.1186/scrt28. Galli, R., E. Binda, U. Orfanelli, B. Cipelletti, A. Gritti, S. De Vitis, R. Fiocco, C. Foroni, F. Dimeco, and A. Vescovi. 2004. "Isolation and characterization of tumorigenic, stem-like neural precursors from human glioblastoma." Cancer Res 64 (19):7011-21. doi: 10.1158/0008-5472.can-04-1364. Gallo, M., F. J. Coutinho, R. J. Vanner, T. Gayden, S. C. Mack, A. Murison, M. Remke, R. Li, N. Takayama, K. Desai, L. Lee, X. Lan, N. I. Park, D. Barsyte-Lovejoy, D. Smil, D. Sturm, M. M. Kushida, R. Head, M. D. Cusimano, M. Bernstein, I. D. Clarke, J. E. Dick, S. M. Pfister, J. N. Rich, C. H. Arrowsmith, M. D. Taylor, N. Jabado, D. P. Bazett-Jones, M. Lupien, and P. B. Dirks. 2015. "MLL5 Orchestrates a Cancer Self-Renewal State by Repressing the Histone Variant H3.3 and Globally Reorganizing Chromatin." Cancer Cell 28 (6):715-29. doi: 10.1016/j.ccell.2015.10.005. Gao, X., J. T. McDonald, L. Hlatky, and H. Enderling. 2013. "Acute and fractionated irradiation differentially modulate glioma stem cell division kinetics." Cancer Res 73 (5):1481-90. doi: 10.1158/0008-5472.can-12-3429. Garkavtsev, I., I. A. Grigorian, V. S. Ossovskaya, M. V. Chernov, P. M. Chumakov, and A. V. Gudkov. 1998. "The candidate tumour suppressor p33ING1 cooperates with p53 in cell growth control." Nature 391 (6664):295-8. doi: 10.1038/34675. Garkavtsev, I., A. Kazarov, A. Gudkov, and K. Riabowol. 1996. "Suppression of the novel growth inhibitor p33ING1 promotes neoplastic transformation." Nat Genet 14 (4):415- 20. doi: 10.1038/ng1296-415. Garkavtsev, I., and K. Riabowol. 1997. "Extension of the replicative life span of human diploid fibroblasts by inhibition of the p33ING1 candidate tumor suppressor." Mol Cell Biol 17 (4):2014-9. Gil-Perotin, S., M. Marin-Husstege, J. Li, M. Soriano-Navarro, F. Zindy, M. F. Roussel, J. M. Garcia-Verdugo, and P. Casaccia-Bonnefil. 2006. "Loss of p53 induces changes in the behavior of subventricular zone cells: implication for the genesis of glial tumors." J Neurosci 26 (4):1107-16. doi: 10.1523/jneurosci.3970-05.2006. Gomes, W. A., M. F. Mehler, and J. A. Kessler. 2003. "Transgenic overexpression of BMP4 increases astroglial and decreases oligodendroglial lineage commitment." Dev Biol 255 (1):164-77. Gonzales, M. 2001. "The 2000 World Health Organization classification of tumours of the nervous system." J Clin Neurosci 8 (1):1-3. doi: 10.1054/jocn.2000.0829.
125
Gonzalez-Perez, O., R. Romero-Rodriguez, M. Soriano-Navarro, J. M. Garcia-Verdugo, and A. Alvarez-Buylla. 2009. "Epidermal growth factor induces the progeny of subventricular zone type B cells to migrate and differentiate into oligodendrocytes." Stem Cells 27 (8):2032-43. doi: 10.1002/stem.119. Gou, W. F., D. F. Shen, X. F. Yang, S. Zhao, Y. P. Liu, H. Z. Sun, R. J. Su, J. S. Luo, and H. C. Zheng. 2015. "ING5 suppresses proliferation, apoptosis, migration and invasion, and induces autophagy and differentiation of gastric cancer cells: a good marker for carcinogenesis and subsequent progression." Oncotarget 6 (23):19552-79. doi: 10.18632/oncotarget.3735. Gu, J., Y. Liu, A. P. Kyritsis, and M. L. Bondy. 2009. "Molecular epidemiology of primary brain tumors." Neurotherapeutics 6 (3):427-35. doi: 10.1016/j.nurt.2009.05.001. Gunduz, M., M. Ouchida, K. Fukushima, H. Hanafusa, T. Etani, S. Nishioka, K. Nishizaki, and K. Shimizu. 2000. "Genomic structure of the human ING1 gene and tumor-specific mutations detected in head and neck squamous cell carcinomas." Cancer Res 60 (12):3143-6. Guo, G., and A. Smith. 2010. "A genome-wide screen in EpiSCs identifies Nr5a nuclear receptors as potent inducers of ground state pluripotency." Development 137 (19):3185- 92. doi: 10.1242/dev.052753. Guo, Y., S. Liu, P. Wang, S. Zhao, F. Wang, L. Bing, Y. Zhang, E. A. Ling, J. Gao, and A. Hao. 2011. "Expression profile of embryonic stem cell-associated genes Oct4, Sox2 and Nanog in human gliomas." Histopathology 59 (4):763-75. doi: 10.1111/j.1365- 2559.2011.03993.x. Gupta, P. B., C. M. Fillmore, G. Jiang, S. D. Shapira, K. Tao, C. Kuperwasser, and E. S. Lander. 2011. "Stochastic State Transitions Give Rise to Phenotypic Equilibrium in Populations of Cancer Cells." Cell 146 (4): 633-644. Han, X., X. Feng, J. B. Rattner, H. Smith, P. Bose, K. Suzuki, M. A. Soliman, M. S. Scott, B. E. Burke, and K. Riabowol. 2008. "Tethering by lamin A stabilizes and targets the ING1 tumour suppressor." Nat Cell Biol 10 (11):1333-40. doi: 10.1038/ncb1792. He, G. H., C. C. Helbing, M. J. Wagner, C. W. Sensen, and K. Riabowol. 2005. "Phylogenetic analysis of the ING family of PHD finger proteins." Mol Biol Evol 22 (1):104-16. doi: 10.1093/molbev/msh256. Hegi, M. E., A. C. Diserens, T. Gorlia, M. F. Hamou, N. de Tribolet, M. Weller, J. M. Kros, J. A. Hainfellner, W. Mason, L. Mariani, J. E. Bromberg, P. Hau, R. O. Mirimanoff, J. G. Cairncross, R. C. Janzer, and R. Stupp. 2005. "MGMT gene silencing and benefit from temozolomide in glioblastoma." N Engl J Med 352 (10):997-1003. doi: 10.1056/NEJMoa043331. Hemmati, H. D., I. Nakano, J. A. Lazareff, M. Masterman-Smith, D. H. Geschwind, M. Bronner- Fraser, and H. I. Kornblum. 2003. "Cancerous stem cells can arise from pediatric brain tumors." Proc Natl Acad Sci U S A 100 (25):15178-83. doi: 10.1073/pnas.2036535100. Heppner, G. H. 1984. "Tumor heterogeneity." Cancer Res 44 (6):2259-65. Hermansen, S. K., K. G. Christensen, S. S. Jensen, and B. W. Kristensen. 2011. "Inconsistent Immunohistochemical Expression Patterns of Four Different CD133 Antibody Clones in Glioblastoma." The Journal of Histochemistry and Cytochemistry : Official Journal of the Histochemistry Society 59 (4): 391-407.
126
Hide, T., T. Takezaki, Y. Nakatani, H. Nakamura, J. Kuratsu, and T. Kondo. 2009. "Sox11 prevents tumorigenesis of glioma-initiating cells by inducing neuronal differentiation." Cancer Res 69 (20):7953-9. doi: 10.1158/0008-5472.can-09-2006. Hockfield, S., and R. D. McKay. 1985. "Identification of major cell classes in the developing mammalian nervous system." J Neurosci 5 (12):3310-28. Holland, E. C., J. Celestino, C. Dai, L. Schaefer, R. E. Sawaya, and G. N. Fuller. 2000. "Combined activation of Ras and Akt in neural progenitors induces glioblastoma formation in mice." Nat Genet 25 (1):55-7. doi: 10.1038/75596. Howe, L., T. Kusch, N. Muster, R. Chaterji, J. R. Yates, 3rd, and J. L. Workman. 2002. "Yng1p modulates the activity of Sas3p as a component of the yeast NuA3 Hhistone acetyltransferase complex." Mol Cell Biol 22 (14):5047-53. Hussein, S. M., M. C. Puri, P. D. Tonge, M. Benevento, A. J. Corso, J. L. Clancy, R. Mosbergen, M. Li, D. S. Lee, N. Cloonan, D. L. Wood, J. Munoz, R. Middleton, O. Korn, H. R. Patel, C. A. White, J. Y. Shin, M. E. Gauthier, K. A. Le Cao, J. I. Kim, J. C. Mar, N. Shakiba, W. Ritchie, J. E. Rasko, S. M. Grimmond, P. W. Zandstra, C. A. Wells, T. Preiss, J. S. Seo, A. J. Heck, I. M. Rogers, and A. Nagy. 2014. "Genome-wide characterization of the routes to pluripotency." Nature 516 (7530):198-206. doi: 10.1038/nature14046. Ignatova, T. N., V. G. Kukekov, E. D. Laywell, O. N. Suslov, F. D. Vrionis, and D. A. Steindler. 2002. "Human cortical glial tumors contain neural stem-like cells expressing astroglial and neuronal markers in vitro." Glia 39 (3):193-206. doi: 10.1002/glia.10094. Iizuka, M., O. F. Sarmento, T. Sekiya, H. Scrable, C. D. Allis, and M. M. Smith. 2008. "Hbo1 Links p53-dependent stress signaling to DNA replication licensing." Mol Cell Biol 28 (1):140-53. doi: 10.1128/mcb.00662-07. Ishiuchi, S., K. Tsuzuki, Y. Yoshida, N. Yamada, N. Hagimura, H. Okado, A. Miwa, H. Kurihara, Y. Nakazato, M. Tamura, T. Sasaki, and S. Ozawa. 2002. "Blockage of Ca(2+)- permeable AMPA receptors suppresses migration and induces apoptosis in human glioblastoma cells." Nat Med 8 (9):971-8. doi: 10.1038/nm746. Jackson, E. L., J. M. Garcia-Verdugo, S. Gil-Perotin, M. Roy, A. Quinones-Hinojosa, S. VandenBerg, and A. Alvarez-Buylla. 2006. "PDGFR alpha-positive B cells are neural stem cells in the adult SVZ that form glioma-like growths in response to increased PDGF signaling." Neuron 51 (2):187-99. doi: 10.1016/j.neuron.2006.06.012. Jaenisch, R., and R. Young. 2008. "Stem cells, the molecular circuitry of pluripotency and nuclear reprogramming." Cell 132 (4):567-82. doi: 10.1016/j.cell.2008.01.015. Jiang, Y., M. Boije, B. Westermark, and L. Uhrbom. 2011. "PDGF-B Can sustain self-renewal and tumorigenicity of experimental glioma-derived cancer-initiating cells by preventing oligodendrocyte differentiation." Neoplasia 13 (6):492-503. Kaadige, M. R., and D. E. Ayer. 2006. "The polybasic region that follows the plant homeodomain zinc finger 1 of Pf1 is necessary and sufficient for specific phosphoinositide binding." J Biol Chem 281 (39):28831-6. doi: 10.1074/jbc.M605624200. Kageyama, R., and T. Ohtsuka. 1999. "The Notch-Hes pathway in mammalian neural development." Cell Res 9 (3):179-88. doi: 10.1038/sj.cr.7290016. Kamitani, H., S. Taniura, K. Watanabe, M. Sakamoto, T. Watanabe, and T. Eling. 2002. "Histone acetylation may suppress human glioma cell proliferation when p21 WAF/Cip1 and gelsolin are induced." Neuro Oncol 4 (2):95-101.
127
Kang, S. S., K. S. Han, B. M. Ku, Y. K. Lee, J. Hong, H. Y. Shin, A. G. Almonte, D. H. Woo, D. J. Brat, E. M. Hwang, S. H. Yoo, C. K. Chung, S. H. Park, S. H. Paek, E. J. Roh, S. J. Lee, J. Y. Park, S. F. Traynelis, and C. J. Lee. 2010. "Caffeine-mediated inhibition of calcium release channel inositol 1,4,5-trisphosphate receptor subtype 3 blocks glioblastoma invasion and extends survival." Cancer Res 70 (3):1173-83. doi: 10.1158/0008-5472.can-09-2886. Kataoka, H., P. Bonnefin, D. Vieyra, X. Feng, Y. Hara, Y. Miura, T. Joh, H. Nakabayashi, H. Vaziri, C. C. Harris, and K. Riabowol. 2003. "ING1 represses transcription by direct DNA binding and through effects on p53." Cancer Res 63 (18):5785-92. Kelly, J. J., O. Stechishin, A. Chojnacki, X. Lun, B. Sun, D. L. Senger, P. Forsyth, R. N. Auer, J. F. Dunn, J. G. Cairncross, I. F. Parney, and S. Weiss. 2009. "Proliferation of human glioblastoma stem cells occurs independently of exogenous mitogens." Stem Cells 27 (8):1722-33. doi: 10.1002/stem.98. Kelly, P. N., A. Dakic, J. M. Adams, S. L. Nutt, and A. Strasser. 2007. "Tumor growth need not be driven by rare cancer stem cells." Science 317 (5836):337. doi: 10.1126/science.1142596. Kichina, J. V., M. Zeremski, L. Aris, K. V. Gurova, E. Walker, R. Franks, A. Y. Nikitin, H. Kiyokawa, and A. V. Gudkov. 2006. "Targeted disruption of the mouse ing1 locus results in reduced body size, hypersensitivity to radiation and elevated incidence of lymphomas." Oncogene 25 (6):857-66. doi: 10.1038/sj.onc.1209118. Kippin, T. E., D. J. Martens, and D. van der Kooy. 2005. "p21 loss compromises the relative quiescence of forebrain stem cell proliferation leading to exhaustion of their proliferation capacity." Genes Dev 19 (6):756-67. doi: 10.1101/gad.1272305. Kleinsmith, L. J., and G. B. Pierce, Jr. 1964. "MULTIPOTENTIALITY OF SINGLE EMBRYONAL CARCINOMA CELLS." Cancer Res 24:1544-51. Konno, D., G. Shioi, A. Shitamukai, A. Mori, H. Kiyonari, T. Miyata, and F. Matsuzaki. 2008. "Neuroepithelial progenitors undergo LGN-dependent planar divisions to maintain self- renewability during mammalian neurogenesis." Nat Cell Biol 10 (1):93-101. doi: 10.1038/ncb1673. Kosodo, Y., K. Roper, W. Haubensak, A. M. Marzesco, D. Corbeil, and W. B. Huttner. 2004. "Asymmetric distribution of the apical plasma membrane during neurogenic divisions of mammalian neuroepithelial cells." EMBO J 23 (11):2314-24. doi: 10.1038/sj.emboj.7600223. Kueh, A. J., M. P. Dixon, A. K. Voss, and T. Thomas. 2011. "HBO1 is required for H3K14 acetylation and normal transcriptional activity during embryonic development." Mol Cell Biol 31 (4):845-60. doi: 10.1128/mcb.00159-10. Kuzmichev, A., Y. Zhang, H. Erdjument-Bromage, P. Tempst, and D. Reinberg. 2002. "Role of the Sin3-histone deacetylase complex in growth regulation by the candidate tumor suppressor p33(ING1)." Mol Cell Biol 22 (3):835-48. Kwon, C. H., D. Zhao, J. Chen, S. Alcantara, Y. Li, D. K. Burns, R. P. Mason, E. Y. Lee, H. Wu, and L. F. Parada. 2008. "Pten haploinsufficiency accelerates formation of high-grade astrocytomas." Cancer Res 68 (9):3286-94. doi: 10.1158/0008-5472.can-07-6867. Lachyankar, M. B., N. Sultana, C. M. Schonhoff, P. Mitra, W. Poluha, S. Lambert, P. J. Quesenberry, N. S. Litofsky, L. D. Recht, R. Nabi, S. J. Miller, S. Ohta, B. G. Neel, and
128
A. H. Ross. 2000. "A role for nuclear PTEN in neuronal differentiation." J Neurosci 20 (4):1404-13. Ladewig, J., P. Koch, and O. Brustle. 2013. "Leveling Waddington: the emergence of direct programming and the loss of cell fate hierarchies." Nat Rev Mol Cell Biol 14 (4):225-36. doi: 10.1038/nrm3543. Lai, W. A., Y. T. Yeh, W. L. Fang, L. S. Wu, N. Harada, P. H. Wang, F. C. Ke, W. L. Lee, and J. J. Hwang. 2014. "Calcineurin and CRTC2 mediate FSH and TGFbeta1 upregulation of Cyp19a1 and Nr5a in ovary granulosa cells." J Mol Endocrinol 53 (2):259-70. doi: 10.1530/jme-14-0048. Lalonde, M. E., N. Avvakumov, K. C. Glass, F. H. Joncas, N. Saksouk, M. Holliday, E. Paquet, K. Yan, Q. Tong, B. J. Klein, S. Tan, X. J. Yang, T. G. Kutateladze, and J. Cote. 2013. "Exchange of associated factors directs a switch in HBO1 acetyltransferase histone tail specificity." Genes Dev 27 (18):2009-24. doi: 10.1101/gad.223396.113. Lapidot, T., C. Sirard, J. Vormoor, B. Murdoch, T. Hoang, J. Caceres-Cortes, M. Minden, B. Paterson, M. A. Caligiuri, and J. E. Dick. 1994. "A cell initiating human acute myeloid leukaemia after transplantation into SCID mice." Nature 367 (6464):645-8. doi: 10.1038/367645a0. Lasorella, A., R. Benezra, and A. Iavarone. 2014. "The ID proteins: master regulators of cancer stem cells and tumour aggressiveness." Nat Rev Cancer 14 (2):77-91. doi: 10.1038/nrc3638. Lassman, A. B., C. Dai, G. N. Fuller, A. J. Vickers, and E. C. Holland. 2004. "Overexpression of c-MYC promotes an undifferentiated phenotype in cultured astrocytes and allows elevated Ras and Akt signaling to induce gliomas from GFAP-expressing cells in mice." Neuron Glia Biol 1 (2):157-63. doi: 10.1017/s1740925x04000249. Lee, J., S. Kotliarova, Y. Kotliarov, A. Li, Q. Su, N. M. Donin, S. Pastorino, B. W. Purow, N. Christopher, W. Zhang, J. K. Park, and H. A. Fine. 2006. "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 9 (5):391- 403. doi: 10.1016/j.ccr.2006.03.030. Lee, J., M. J. Son, K. Woolard, N. M. Donin, A. Li, C. H. Cheng, S. Kotliarova, Y. Kotliarov, J. Walling, S. Ahn, M. Kim, M. Totonchy, T. Cusack, C. Ene, H. Ma, Q. Su, J. C. Zenklusen, W. Zhang, D. Maric, and H. A. Fine. 2008. "Epigenetic-mediated dysfunction of the bone morphogenetic protein pathway inhibits differentiation of glioblastoma- initiating cells." Cancer Cell 13 (1):69-80. doi: 10.1016/j.ccr.2007.12.005. Lee, J. V., A. Carrer, S. Shah, N. W. Snyder, S. Wei, S. Venneti, A. J. Worth, Z. F. Yuan, H. W. Lim, S. Liu, E. Jackson, N. M. Aiello, N. B. Haas, T. R. Rebbeck, A. Judkins, K. J. Won, L. A. Chodosh, B. A. Garcia, B. Z. Stanger, M. D. Feldman, I. A. Blair, and K. E. Wellen. 2014. "Akt-dependent metabolic reprogramming regulates tumor cell histone acetylation." Cell Metab 20 (2):306-19. doi: 10.1016/j.cmet.2014.06.004. Lendahl, U., L. B. Zimmerman, and R. D. McKay. 1990. "CNS stem cells express a new class of intermediate filament protein." Cell 60 (4):585-95. Li, C., D. G. Heidt, P. Dalerba, C. F. Burant, L. Zhang, V. Adsay, M. Wicha, M. F. Clarke, and D. M. Simeone. 2007. "Identification of pancreatic cancer stem cells." Cancer Res 67 (3):1030-7. doi: 10.1158/0008-5472.can-06-2030.
129
Li, L., J. Mignone, M. Yang, M. Matic, S. Penman, G. Enikolopov, and R. M. Hoffman. 2003. "Nestin expression in hair follicle sheath progenitor cells." Proc Natl Acad Sci U S A 100 (17):9958-61. doi: 10.1073/pnas.1733025100. Li, S., P. Mattar, R. Dixit, S. O. Lawn, G. Wilkinson, C. Kinch, D. Eisenstat, D. M. Kurrasch, J. A. Chan, and C. Schuurmans. 2014. "RAS/ERK signaling controls proneural genetic programs in cortical development and gliomagenesis." J Neurosci 34 (6):2169-90. doi: 10.1523/jneurosci.4077-13.2014. Li, X., T. Nishida, A. Noguchi, Y. Zheng, H. Takahashi, X. Yang, S. Masuda, and Y. Takano. 2010. "Decreased nuclear expression and increased cytoplasmic expression of ING5 may be linked to tumorigenesis and progression in human head and neck squamous cell carcinoma." J Cancer Res Clin Oncol 136 (10):1573-83. doi: 10.1007/s00432-010-0815- x. Li, Y., H. Deng, L. Lv, C. Zhang, L. Qian, J. Xiao, W. Zhao, Q. Liu, D. Zhang, Y. Wang, J. Yan, H. Zhang, Y. He, and J. Zhu. 2015. "The miR-193a-3p-regulated ING5 gene activates the DNA damage response pathway and inhibits multi-chemoresistance in bladder cancer." Oncotarget 6 (12):10195-206. doi: 10.18632/oncotarget.3555. Ligon, K. L., S. P. Fancy, R. J. Franklin, and D. H. Rowitch. 2006. "Olig gene function in CNS development and disease." Glia 54 (1):1-10. doi: 10.1002/glia.20273. Ligon, K. L., E. Huillard, S. Mehta, S. Kesari, H. Liu, J. A. Alberta, R. M. Bachoo, M. Kane, D. N. Louis, R. A. Depinho, D. J. Anderson, C. D. Stiles, and D. H. Rowitch. 2007. "Olig2- regulated lineage-restricted pathway controls replication competence in neural stem cells and malignant glioma." Neuron 53 (4):503-17. doi: 10.1016/j.neuron.2007.01.009. Linzen, U., R. Lilischkis, R. Pandithage, B. Schilling, A. Ullius, J. Luscher-Firzlaff, E. Kremmer, B. Luscher, and J. Vervoorts. 2015. "ING5 is phosphorylated by CDK2 and controls cell proliferation independently of p53." PLoS One 10 (4):e0123736. doi: 10.1371/journal.pone.0123736. Liu, L., J. Zhang, C. Fang, Z. Zhang, Y. Feng, and X. Xi. 2015. "OCT4 mediates FSH-induced epithelial-mesenchymal transition and invasion through the ERK1/2 signaling pathway in epithelial ovarian cancer." Biochem Biophys Res Commun 461 (3):525-32. doi: 10.1016/j.bbrc.2015.04.061. Liu, N., J. Wang, J. Wang, R. Wang, Z. Liu, Y. Yu, and H. Lu. 2013. "ING5 is a Tip60 cofactor that acetylates p53 in response to DNA damage." Cancer Res 73 (12):3749-60. doi: 10.1158/0008-5472.can-12-3684. Liu, Q., D. H. Nguyen, Q. Dong, P. Shitaku, K. Chung, O. Y. Liu, J. L. Tso, J. Y. Liu, V. Konkankit, T. F. Cloughesy, P. S. Mischel, T. F. Lane, L. M. Liau, S. F. Nelson, and C. L. Tso. 2009. "Molecular properties of CD133+ glioblastoma stem cells derived from treatment-refractory recurrent brain tumors." J Neurooncol 94 (1):1-19. doi: 10.1007/s11060-009-9919-z. Loewith, R., M. Meijer, S. P. Lees-Miller, K. Riabowol, and D. Young. 2000. "Three yeast proteins related to the human candidate tumor suppressor p33(ING1) are associated with histone acetyltransferase activities." Mol Cell Biol 20 (11):3807-16. Lopez-Bertoni, H., B. Lal, A. Li, M. Caplan, H. Guerrero-Cazares, C. G. Eberhart, A. Quinones- Hinojosa, M. Glas, B. Scheffler, J. Laterra, and Y. Li. 2015. "DNMT-dependent suppression of microRNA regulates the induction of GBM tumor-propagating phenotype by Oct4 and Sox2." Oncogene 34 (30):3994-4004. doi: 10.1038/onc.2014.334.
130
Louis, D. N., H. Ohgaki, O. D. Wiestler, W. K. Cavenee, P. C. Burger, A. Jouvet, B. W. Scheithauer, and P. Kleihues. 2007. "The 2007 WHO classification of tumours of the central nervous system." Acta Neuropathol 114 (2):97-109. doi: 10.1007/s00401-007- 0243-4. Ma, S., K. H. Tang, Y. P. Chan, T. K. Lee, P. S. Kwan, A. Castilho, I. Ng, et al. 2010. "miR- 130b Promotes CD133(+) Liver Tumor-Initiating Cell Growth and Self-Renewal Via Tumor Protein 53-Induced Nuclear Protein 1." Cell Stem Cell 7 (6): 694-707. Markowetz, F., K. W. Mulder, E. M. Airoldi, I. R. Lemischka, and O. G. Troyanskaya. 2010. "Mapping dynamic histone acetylation patterns to gene expression in nanog-depleted murine embryonic stem cells." PLoS Comput Biol 6 (12):e1001034. doi: 10.1371/journal.pcbi.1001034. Martinez-Diaz, H., B. K. Kleinschmidt-DeMasters, S. Z. Powell, and A. T. Yachnis. 2003. "Giant cell glioblastoma and pleomorphic xanthoastrocytoma show different immunohistochemical profiles for neuronal antigens and p53 but share reactivity for class III beta-tubulin." Arch Pathol Lab Med 127 (9):1187-91. doi: 10.1043/1543- 2165(2003)127<1187:gcgapx>2.0.co;2. McKenzie, J. L., O. I. Gan, M. Doedens, and J. E. Dick. 2005. "Human short-term repopulating stem cells are efficiently detected following intrafemoral transplantation into NOD/SCID recipients depleted of CD122+ cells." Blood 106 (4):1259-61. doi: 10.1182/blood-2005- 03-1081. Meerbrey, K. L., G. Hu, J. D. Kessler, K. Roarty, M. Z. Li, J. E. Fang, J. I. Herschkowitz, A. E. Burrows, A. Ciccia, T. Sun, E. M. Schmitt, R. J. Bernardi, X. Fu, C. S. Bland, T. A. Cooper, R. Schiff, J. M. Rosen, T. F. Westbrook, and S. J. Elledge. 2011. "The pINDUCER lentiviral toolkit for inducible RNA interference in vitro and in vivo." Proc Natl Acad Sci U S A 108 (9):3665-70. doi: 10.1073/pnas.1019736108. Melcer, S., H. Hezroni, E. Rand, M. Nissim-Rafinia, A. Skoultchi, C. L. Stewart, M. Bustin, and E. Meshorer. 2012. "Histone modifications and lamin A regulate chromatin protein dynamics in early embryonic stem cell differentiation." Nat Commun 3:910. doi: 10.1038/ncomms1915. Meletis, K., V. Wirta, S. M. Hede, M. Nister, J. Lundeberg, and J. Frisen. 2006. "p53 suppresses the self-renewal of adult neural stem cells." Development 133 (2):363-9. doi: 10.1242/dev.02208. Merson, T. D., M. P. Dixon, C. Collin, R. L. Rietze, P. F. Bartlett, T. Thomas, and A. K. Voss. 2006. "The transcriptional coactivator Querkopf controls adult neurogenesis." J Neurosci 26 (44):11359-70. doi: 10.1523/jneurosci.2247-06.2006. Miraglia, S., W. Godfrey, A. H. Yin, K. Atkins, R. Warnke, J. T. Holden, R. A. Bray, E. K. Waller, and D. W. Buck. 1997a. "A Novel Five-Transmembrane Hematopoietic Stem Cell Antigen: Isolation, Characterization, and Molecular Cloning." Blood 90 (12): 5013- 5021. ———. 1997b. "A Novel Five-Transmembrane Hematopoietic Stem Cell Antigen: Isolation, Characterization, and Molecular Cloning." Blood 90 (12): 5013-5021. Molofsky, A. V., S. G. Slutsky, N. M. Joseph, S. He, R. Pardal, J. Krishnamurthy, N. E. Sharpless, and S. J. Morrison. 2006. "Increasing p16INK4a expression decreases forebrain progenitors and neurogenesis during ageing." Nature 443 (7110):448-52. doi: 10.1038/nature05091.
131
Moon, J. H., S. Kwon, E. K. Jun, A. Kim, K. Y. Whang, H. Kim, S. Oh, B. S. Yoon, and S. You. 2011. "Nanog-induced dedifferentiation of p53-deficient mouse astrocytes into brain cancer stem-like cells." Biochem Biophys Res Commun 412 (1):175-81. doi: 10.1016/j.bbrc.2011.07.070. Moreno, M., V. Fernandez, J. M. Monllau, V. Borrell, C. Lerin, and N. de la Iglesia. 2015. "Transcriptional Profiling of Hypoxic Neural Stem Cells Identifies Calcineurin-NFATc4 Signaling as a Major Regulator of Neural Stem Cell Biology." Stem Cell Reports 5 (2):157-65. doi: 10.1016/j.stemcr.2015.06.008. Mulder, K. W., X. Wang, C. Escriu, Y. Ito, R. F. Schwarz, J. Gillis, G. Sirokmany, G. Donati, S. Uribe-Lewis, P. Pavlidis, A. Murrell, F. Markowetz, and F. M. Watt. 2012. "Diverse epigenetic strategies interact to control epidermal differentiation." Nat Cell Biol 14 (7):753-63. doi: 10.1038/ncb2520. Nagarajan, R. P., and J. F. Costello. 2009. "Molecular epigenetics and genetics in neuro- oncology." Neurotherapeutics 6 (3):436-46. doi: 10.1016/j.nurt.2009.04.002. Nam, H. S., and R. Benezra. 2009. "High levels of Id1 expression define B1 type adult neural stem cells." Cell Stem Cell 5 (5):515-26. doi: 10.1016/j.stem.2009.08.017. Natsume, A., M. Ito, K. Katsushima, F. Ohka, A. Hatanaka, K. Shinjo, S. Sato, S. Takahashi, Y. Ishikawa, I. Takeuchi, H. Shimogawa, M. Uesugi, H. Okano, S. U. Kim, T. Wakabayashi, J. P. Issa, Y. Sekido, and Y. Kondo. 2013. "Chromatin regulator PRC2 is a key regulator of epigenetic plasticity in glioblastoma." Cancer Res 73 (14):4559-70. doi: 10.1158/0008-5472.can-13-0109. Niola, F., X. Zhao, D. Singh, R. Sullivan, A. Castano, A. Verrico, P. Zoppoli, D. Friedmann- Morvinski, E. Sulman, L. Barrett, Y. Zhuang, I. Verma, R. Benezra, K. Aldape, A. Iavarone, and A. Lasorella. 2013. "Mesenchymal high-grade glioma is maintained by the ID-RAP1 axis." J Clin Invest 123 (1):405-17. doi: 10.1172/jci63811. Nixon, R. A., and T. B. Shea. 1992. "Dynamics of neuronal intermediate filaments: a developmental perspective." Cell Motil Cytoskeleton 22 (2):81-91. doi: 10.1002/cm.970220202. Nouman, G. S., J. J. Anderson, K. M. Wood, J. Lunec, A. G. Hall, M. M. Reid, and B. Angus. 2002. "Loss of nuclear expression of the p33(ING1b) inhibitor of growth protein in childhood acute lymphoblastic leukaemia." J Clin Pathol 55 (8):596-601. O'Brien, C. A., A. Pollett, S. Gallinger, and J. E. Dick. 2007. "A human colon cancer cell capable of initiating tumour growth in immunodeficient mice." Nature 445 (7123):106-10. doi: 10.1038/nature05372. Ogden, A. T., A. E. Waziri, R. A. Lochhead, D. Fusco, K. Lopez, J. A. Ellis, J. Kang, et al. 2008. "Identification of A2B5+CD133- Tumor-Initiating Cells in Adult Human Gliomas." Neurosurgery 62 (2): 505-14; discussion 514-5. Ohgaki, H. 2009. "Epidemiology of brain tumors." Methods Mol Biol 472:323-42. doi: 10.1007/978-1-60327-492-0_14. Okada, M., A. Makino, M. Nakajima, S. Okuyama, S. Furukawa, and Y. Furukawa. 2010. "Estrogen stimulates proliferation and differentiation of neural stem/progenitor cells through different signal transduction pathways." Int J Mol Sci 11 (10):4114-23. doi: 10.3390/ijms11104114.
132
Oki, E., Y. Maehara, E. Tokunaga, Y. Kakeji, and K. Sugimachi. 1999. "Reduced expression of p33(ING1) and the relationship with p53 expression in human gastric cancer." Cancer Lett 147 (1-2):157-62. Omuro, A., and L. M. DeAngelis. 2013. "Glioblastoma and other malignant gliomas: a clinical review." JAMA 310 (17):1842-50. doi: 10.1001/jama.2013.280319. Ostrom, Q. T., H. Gittleman, J. Fulop, M. Liu, R. Blanda, C. Kromer, Y. Wolinsky, C. Kruchko, and J. S. Barnholtz-Sloan. 2015. "CBTRUS Statistical Report: Primary Brain and Central Nervous System Tumors Diagnosed in the United States in 2008-2012." Neuro Oncol 17 Suppl 4:iv1-iv62. doi: 10.1093/neuonc/nov189. Pena, P. V., F. Davrazou, X. Shi, K. L. Walter, V. V. Verkhusha, O. Gozani, R. Zhao, and T. G. Kutateladze. 2006. "Molecular mechanism of histone H3K4me3 recognition by plant homeodomain of ING2." Nature 442 (7098):100-3. doi: 10.1038/nature04814. Peretto, P., A. Merighi, A. Fasolo, and L. Bonfanti. 1999. "The subependymal layer in rodents: a site of structural plasticity and cell migration in the adult mammalian brain." Brain Res Bull 49 (4):221-43. Perez-Campo, F. M., J. Borrow, V. Kouskoff, and G. Lacaud. 2009. "The histone acetyl transferase activity of monocytic leukemia zinc finger is critical for the proliferation of hematopoietic precursors." Blood 113 (20):4866-74. doi: 10.1182/blood-2008-04- 152017. Pietras, A., A. M. Katz, E. J. Ekstrom, B. Wee, J. J. Halliday, K. L. Pitter, J. L. Werbeck, N. M. Amankulor, J. T. Huse, and E. C. Holland. 2014. "Osteopontin-CD44 signaling in the glioma perivascular niche enhances cancer stem cell phenotypes and promotes aggressive tumor growth." Cell Stem Cell 14 (3):357-69. doi: 10.1016/j.stem.2014.01.005. Quintana, E., M. Shackleton, M. S. Sabel, D. R. Fullen, T. M. Johnson, and S. J. Morrison. 2008. "Efficient tumour formation by single human melanoma cells." Nature 456 (7222):593- 8. doi: 10.1038/nature07567. Rajarajacholan, U. K., S. Thalappilly, and K. Riabowol. 2013. "The ING1a tumor suppressor regulates endocytosis to induce cellular senescence via the Rb-E2F pathway." PLoS Biol 11 (3):e1001502. doi: 10.1371/journal.pbio.1001502. Reynolds, B. A., and S. Weiss. 1992. "Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system." Science 255 (5052):1707-10. Riederer, B., R. Cohen, and A. Matus. 1986. "MAP5: a novel brain microtubule-associated protein under strong developmental regulation." J Neurocytol 15 (6):763-75. Rodini, C. O., D. E. Suzuki, N. Saba-Silva, A. Cappellano, J. E. de Souza, S. Cavalheiro, S. R. Toledo, and O. K. Okamoto. 2012. "Expression analysis of stem cell-related genes reveal OCT4 as a predictor of poor clinical outcome in medulloblastoma." J Neurooncol 106 (1):71-9. doi: 10.1007/s11060-011-0647-9. Ross, S. E., M. E. Greenberg, and C. D. Stiles. 2003. "Basic helix-loop-helix factors in cortical development." Neuron 39 (1):13-25. Russell, M. W., M. A. Soliman, D. Schriemer, and K. Riabowol. 2008. "ING1 protein targeting to the nucleus by karyopherins is necessary for activation of p21." Biochem Biophys Res Commun 374 (3):490-5. doi: 10.1016/j.bbrc.2008.07.076. Saito, M., K. Kumamoto, A. I. Robles, I. Horikawa, B. Furusato, S. Okamura, A. Goto, T. Yamashita, M. Nagashima, T. L. Lee, V. J. Baxendale, O. M. Rennert, S. Takenoshita, J. Yokota, I. A. Sesterhenn, G. E. Trivers, S. P. Hussain, and C. C. Harris. 2010. "Targeted
133
disruption of Ing2 results in defective spermatogenesis and development of soft-tissue sarcomas." PLoS One 5 (11):e15541. doi: 10.1371/journal.pone.0015541. Sarker, K. P., H. Kataoka, A. Chan, S. J. Netherton, I. Pot, M. A. Huynh, X. Feng, A. Bonni, K. Riabowol, and S. Bonni. 2008. "ING2 as a novel mediator of transforming growth factor- beta-dependent responses in epithelial cells." J Biol Chem 283 (19):13269-79. doi: 10.1074/jbc.M708834200. Schafer, A., E. Karaulanov, U. Stapf, G. Doderlein, and C. Niehrs. 2013. "Ing1 functions in DNA demethylation by directing Gadd45a to H3K4me3." Genes Dev 27 (3):261-73. doi: 10.1101/gad.186916.112. Schoenhals, M., A. Kassambara, J. De Vos, D. Hose, J. Moreaux, and B. Klein. 2009. "Embryonic stem cell markers expression in cancers." Biochem Biophys Res Commun 383 (2):157-62. doi: 10.1016/j.bbrc.2009.02.156. Scott, M., F. M. Boisvert, D. Vieyra, R. N. Johnston, D. P. Bazett-Jones, and K. Riabowol. 2001. "UV induces nucleolar translocation of ING1 through two distinct nucleolar targeting sequences." Nucleic Acids Res 29 (10):2052-8. Scott, M., P. Bonnefin, D. Vieyra, F. M. Boisvert, D. Young, D. P. Bazett-Jones, and K. Riabowol. 2001. "UV-induced binding of ING1 to PCNA regulates the induction of apoptosis." J Cell Sci 114 (Pt 19):3455-62. Shackleton, M., E. Quintana, E. R. Fearon, and S. J. Morrison. 2009. "Heterogeneity in cancer: cancer stem cells versus clonal evolution." Cell 138 (5):822-9. doi: 10.1016/j.cell.2009.08.017. Sheikh, B. N., M. P. Dixon, T. Thomas, and A. K. Voss. 2012. "Querkopf is a key marker of self-renewal and multipotency of adult neural stem cells." J Cell Sci 125 (Pt 2):295-309. doi: 10.1242/jcs.077271. Shi, J., W. Shi, L. Ni, X. Xu, X. Su, L. Xia, F. Xu, J. Chen, and J. Zhu. 2013. "OCT4 is epigenetically regulated by DNA hypomethylation of promoter and exon in primary gliomas." Oncol Rep 30 (1):201-6. doi: 10.3892/or.2013.2456. Shi, X., T. Hong, K. L. Walter, M. Ewalt, E. Michishita, T. Hung, D. Carney, P. Pena, F. Lan, M. R. Kaadige, N. Lacoste, C. Cayrou, F. Davrazou, A. Saha, B. R. Cairns, D. E. Ayer, T. G. Kutateladze, Y. Shi, J. Cote, K. F. Chua, and O. Gozani. 2006. "ING2 PHD domain links histone H3 lysine 4 methylation to active gene repression." Nature 442 (7098):96-9. doi: 10.1038/nature04835. Shimazaki, T., T. Shingo, and S. Weiss. 2001. "The ciliary neurotrophic factor/leukemia inhibitory factor/gp130 receptor complex operates in the maintenance of mammalian forebrain neural stem cells." J Neurosci 21 (19):7642-53. Shinoura, N., Y. Muramatsu, M. Nishimura, Y. Yoshida, A. Saito, T. Yokoyama, T. Furukawa, A. Horii, M. Hashimoto, A. Asai, T. Kirino, and H. Hamada. 1999. "Adenovirus- mediated transfer of p33ING1 with p53 drastically augments apoptosis in gliomas." Cancer Res 59 (21):5521-8. Shiseki, M., M. Nagashima, R. M. Pedeux, M. Kitahama-Shiseki, K. Miura, S. Okamura, H. Onogi, Y. Higashimoto, E. Appella, J. Yokota, and C. C. Harris. 2003. "p29ING4 and p28ING5 bind to p53 and p300, and enhance p53 activity." Cancer Res 63 (10):2373-8. Siegel, E. T., H. G. Kim, H. K. Nishimoto, and L. C. Layman. 2013. "The molecular basis of impaired follicle-stimulating hormone action: evidence from human mutations and mouse models." Reprod Sci 20 (3):211-33. doi: 10.1177/1933719112461184.
134
Signaroldi, E., P. Laise, S. Cristofanon, A. Brancaccio, E. Reisoli, S. Atashpaz, M. R. Terreni, C. Doglioni, G. Pruneri, P. Malatesta, and G. Testa. 2016. "Polycomb dysregulation in gliomagenesis targets a Zfp423-dependent differentiation network." Nat Commun 7:10753. doi: 10.1038/ncomms10753. Singh, S. K., I. D. Clarke, M. Terasaki, V. E. Bonn, C. Hawkins, J. Squire, and P. B. Dirks. 2003. "Identification of a cancer stem cell in human brain tumors." Cancer Res 63 (18):5821-8. Singh, S. K., C. Hawkins, I. D. Clarke, J. A. Squire, J. Bayani, T. Hide, R. M. Henkelman, M. D. Cusimano, and P. B. Dirks. 2004. "Identification of human brain tumour initiating cells." Nature 432 (7015):396-401. doi: 10.1038/nature03128. Soliman, M. A., P. Berardi, S. Pastyryeva, P. Bonnefin, X. Feng, A. Colina, D. Young, and K. Riabowol. 2008. "ING1a expression increases during replicative senescence and induces a senescent phenotype." Aging Cell 7 (6):783-94. doi: 10.1111/j.1474- 9726.2008.00427.x. Sommer, I., and M. Schachner. 1981. "Monoclonal antibodies (O1 to O4) to oligodendrocyte cell surfaces: an immunocytological study in the central nervous system." Dev Biol 83 (2):311-27. Spyropoulou, A., C. Piperi, C. Adamopoulos, and A. G. Papavassiliou. 2013. "Deregulated chromatin remodeling in the pathobiology of brain tumors." Neuromolecular Med 15 (1):1-24. doi: 10.1007/s12017-012-8205-y. Sturm, D., H. Witt, V. Hovestadt, D. A. Khuong-Quang, D. T. Jones, C. Konermann, E. Pfaff, M. Tonjes, M. Sill, S. Bender, M. Kool, M. Zapatka, N. Becker, M. Zucknick, T. Hielscher, X. Y. Liu, A. M. Fontebasso, M. Ryzhova, S. Albrecht, K. Jacob, M. Wolter, M. Ebinger, M. U. Schuhmann, T. van Meter, M. C. Fruhwald, H. Hauch, A. Pekrun, B. Radlwimmer, T. Niehues, G. von Komorowski, M. Durken, A. E. Kulozik, J. Madden, A. Donson, N. K. Foreman, R. Drissi, M. Fouladi, W. Scheurlen, A. von Deimling, C. Monoranu, W. Roggendorf, C. Herold-Mende, A. Unterberg, C. M. Kramm, J. Felsberg, C. Hartmann, B. Wiestler, W. Wick, T. Milde, O. Witt, A. M. Lindroth, J. Schwartzentruber, D. Faury, A. Fleming, M. Zakrzewska, P. P. Liberski, K. Zakrzewski, P. Hauser, M. Garami, A. Klekner, L. Bognar, S. Morrissy, F. Cavalli, M. D. Taylor, P. van Sluis, J. Koster, R. Versteeg, R. Volckmann, T. Mikkelsen, K. Aldape, G. Reifenberger, V. P. Collins, J. Majewski, A. Korshunov, P. Lichter, C. Plass, N. Jabado, and S. M. Pfister. 2012. "Hotspot mutations in H3F3A and IDH1 define distinct epigenetic and biological subgroups of glioblastoma." Cancer Cell 22 (4):425-37. doi: 10.1016/j.ccr.2012.08.024. Su, Z., T. Zang, M. L. Liu, L. L. Wang, W. Niu, and C. L. Zhang. 2014. "Reprogramming the fate of human glioma cells to impede brain tumor development." Cell Death Dis 5:e1463. doi: 10.1038/cddis.2014.425. Suva, M. L., E. Rheinbay, S. M. Gillespie, A. P. Patel, H. Wakimoto, S. D. Rabkin, N. Riggi, A. S. Chi, D. P. Cahill, B. V. Nahed, W. T. Curry, R. L. Martuza, M. N. Rivera, N. Rossetti, S. Kasif, S. Beik, S. Kadri, I. Tirosh, I. Wortman, A. K. Shalek, O. Rozenblatt-Rosen, A. Regev, D. N. Louis, and B. E. Bernstein. 2014. "Reconstructing and reprogramming the tumor-propagating potential of glioblastoma stem-like cells." Cell 157 (3):580-94. doi: 10.1016/j.cell.2014.02.030.
135
Suzuki, S., Y. Nozawa, S. Tsukamoto, T. Kaneko, H. Imai, and N. Minami. 2013. "ING3 is essential for asymmetric cell division during mouse oocyte maturation." PLoS One 8 (9):e74749. doi: 10.1371/journal.pone.0074749. Takahashi, K., and S. Yamanaka. 2006. "Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors." Cell 126 (4):663-76. doi: 10.1016/j.cell.2006.07.024. Tallen, G., I. Kaiser, S. Krabbe, U. Lass, C. Hartmann, G. Henze, K. Riabowol, and A. von Deimling. 2004. "No ING1 mutations in human brain tumours but reduced expression in high malignancy grades of astrocytoma." Int J Cancer 109 (3):476-9. doi: 10.1002/ijc.11715. Tallen, G., and K. Riabowol. 2014. "Keep-ING balance: tumor suppression by epigenetic regulation." FEBS Lett 588 (16):2728-42. doi: 10.1016/j.febslet.2014.03.011. Tallen, U. G., M. Truss, F. Kunitz, S. Wellmann, B. Unryn, B. Sinn, U. Lass, S. Krabbe, N. Holtkamp, C. Hagemeier, R. Wurm, G. Henze, K. T. Riabowol, and A. von Deimling. 2008. "Down-regulation of the inhibitor of growth 1 (ING1) tumor suppressor sensitizes p53-deficient glioblastoma cells to cisplatin-induced cell death." J Neurooncol 86 (1):23- 30. doi: 10.1007/s11060-007-9436-x. Tamannai, M., S. Farhangi, M. Truss, B. Sinn, R. Wurm, P. Bose, G. Henze, K. Riabowol, A. von Deimling, and G. Tallen. 2010. "The inhibitor of growth 1 (ING1) is involved in trichostatin A-induced apoptosis and caspase 3 signaling in p53-deficient glioblastoma cells." Oncol Res 18 (10):469-80. TCGA. 2008. "Comprehensive genomic characterization defines human glioblastoma genes and core pathways." Nature 455 (7216):1061-8. doi: 10.1038/nature07385. Thalappilly, S., X. Feng, S. Pastyryeva, K. Suzuki, D. Muruve, D. Larocque, S. Richard, M. Truss, A. von Deimling, K. Riabowol, and G. Tallen. 2011. "The p53 tumor suppressor is stabilized by inhibitor of growth 1 (ING1) by blocking polyubiquitination." PLoS One 6 (6):e21065. doi: 10.1371/journal.pone.0021065. Thomas, T., L. M. Corcoran, R. Gugasyan, M. P. Dixon, T. Brodnicki, S. L. Nutt, D. Metcalf, and A. K. Voss. 2006. "Monocytic leukemia zinc finger protein is essential for the development of long-term reconstituting hematopoietic stem cells." Genes Dev 20 (9):1175-86. doi: 10.1101/gad.1382606. Tohyama, T., V. M. Lee, L. B. Rorke, M. Marvin, R. D. McKay, and J. Q. Trojanowski. 1992. "Nestin expression in embryonic human neuroepithelium and in human neuroepithelial tumor cells." Lab Invest 66 (3):303-13. Tokunaga, E., Y. Maehara, E. Oki, K. Kitamura, Y. Kakeji, S. Ohno, and K. Sugimachi. 2000. "Diminished expression of ING1 mRNA and the correlation with p53 expression in breast cancers." Cancer Lett 152 (1):15-22. Toyama, T., H. Iwase, P. Watson, H. Muzik, E. Saettler, A. Magliocco, L. DiFrancesco, P. Forsyth, I. Garkavtsev, S. Kobayashi, and K. Riabowol. 1999. "Suppression of ING1 expression in sporadic breast cancer." Oncogene 18 (37):5187-93. doi: 10.1038/sj.onc.1202905. Uhrbom, L., C. Dai, J. C. Celestino, M. K. Rosenblum, G. N. Fuller, and E. C. Holland. 2002. "Ink4a-Arf loss cooperates with KRas activation in astrocytes and neural progenitors to generate glioblastomas of various morphologies depending on activated Akt." Cancer Res 62 (19):5551-8.
136
Ullah, M., N. Pelletier, L. Xiao, S. P. Zhao, K. Wang, C. Degerny, S. Tahmasebi, C. Cayrou, Y. Doyon, S. L. Goh, N. Champagne, J. Cote, and X. J. Yang. 2008. "Molecular architecture of quartet MOZ/MORF histone acetyltransferase complexes." Mol Cell Biol 28 (22):6828-43. doi: 10.1128/mcb.01297-08. Vanner, R. J., M. Remke, M. Gallo, H. J. Selvadurai, F. Coutinho, L. Lee, M. Kushida, R. Head, S. Morrissy, X. Zhu, T. Aviv, V. Voisin, I. D. Clarke, Y. Li, A. J. Mungall, R. A. Moore, Y. Ma, S. J. Jones, M. A. Marra, D. Malkin, P. A. Northcott, M. Kool, S. M. Pfister, G. Bader, K. Hochedlinger, A. Korshunov, M. D. Taylor, and P. B. Dirks. 2014. "Quiescent sox2(+) cells drive hierarchical growth and relapse in sonic hedgehog subgroup medulloblastoma." Cancer Cell 26 (1):33-47. doi: 10.1016/j.ccr.2014.05.005. Verhaak, R. G., K. A. Hoadley, E. Purdom, V. Wang, Y. Qi, M. D. Wilkerson, C. R. Miller, L. Ding, T. Golub, J. P. Mesirov, G. Alexe, M. Lawrence, M. O'Kelly, P. Tamayo, B. A. Weir, S. Gabriel, W. Winckler, S. Gupta, L. Jakkula, H. S. Feiler, J. G. Hodgson, C. D. James, J. N. Sarkaria, C. Brennan, A. Kahn, P. T. Spellman, R. K. Wilson, T. P. Speed, J. W. Gray, M. Meyerson, G. Getz, C. M. Perou, and D. N. Hayes. 2010. "Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1." Cancer Cell 17 (1):98-110. doi: 10.1016/j.ccr.2009.12.020. Vieyra, D., R. Loewith, M. Scott, P. Bonnefin, F. M. Boisvert, P. Cheema, S. Pastyryeva, M. Meijer, R. N. Johnston, D. P. Bazett-Jones, S. McMahon, M. D. Cole, D. Young, and K. Riabowol. 2002. "Human ING1 proteins differentially regulate histone acetylation." J Biol Chem 277 (33):29832-9. doi: 10.1074/jbc.M200197200. Vieyra, D., D. L. Senger, T. Toyama, H. Muzik, P. M. Brasher, R. N. Johnston, K. Riabowol, and P. A. Forsyth. 2003. "Altered subcellular localization and low frequency of mutations of ING1 in human brain tumors." Clin Cancer Res 9 (16 Pt 1):5952-61. Vieyra, D., T. Toyama, Y. Hara, D. Boland, R. Johnston, and K. Riabowol. 2002. "ING1 isoforms differentially affect apoptosis in a cell age-dependent manner." Cancer Res 62 (15):4445-52. Vlismas, A., R. Bletsa, D. Mavrogianni, G. Mamali, M. Pergamali, V. Dinopoulou, G. Partsinevelos, P. Drakakis, D. Loutradis, and A. A. Kiessling. 2016. "Microarray Analyses Reveal Marked Differences in Growth Factor and Receptor Expression Between 8-Cell Human Embryos and Pluripotent Stem Cells." Stem Cells Dev 25 (2):160-77. doi: 10.1089/scd.2015.0284. Walzak, A. A., N. Veldhoen, X. Feng, K. Riabowol, and C. C. Helbing. 2008. "Expression profiles of mRNA transcript variants encoding the human inhibitor of growth tumor suppressor gene family in normal and neoplastic tissues." Exp Cell Res 314 (2):273-85. doi: 10.1016/j.yexcr.2007.07.029. Wang, J., P. O. Sakariassen, O. Tsinkalovsky, H. Immervoll, S. O. Boe, A. Svendsen, L. Prestegarden, et al. 2008. "CD133 Negative Glioma Cells Form Tumors in Nude Rats and Give Rise to CD133 Positive Cells." International Journal of Cancer 122 (4): 761- 768. Wang, Y., E. Kim, X. Wang, B. G. Novitch, K. Yoshikawa, L. S. Chang, and Y. Zhu. 2012. "ERK inhibition rescues defects in fate specification of Nf1-deficient neural progenitors and brain abnormalities." Cell 150 (4):816-30. doi: 10.1016/j.cell.2012.06.034.
137
Wegner, M., and C. C. Stolt. 2005. "From stem cells to neurons and glia: a Soxist's view of neural development." Trends Neurosci 28 (11):583-8. doi: 10.1016/j.tins.2005.08.008. Wetzel, M., D. R. Premkumar, B. Arnold, and I. F. Pollack. 2005. "Effect of trichostatin A, a histone deacetylase inhibitor, on glioma proliferation in vitro by inducing cell cycle arrest and apoptosis." J Neurosurg 103 (6 Suppl):549-56. doi: 10.3171/ped.2005.103.6.0549. Woltjen, K., I. P. Michael, P. Mohseni, R. Desai, M. Mileikovsky, R. Hamalainen, R. Cowling, W. Wang, P. Liu, M. Gertsenstein, K. Kaji, H. K. Sung, and A. Nagy. 2009. "piggyBac transposition reprograms fibroblasts to induced pluripotent stem cells." Nature 458 (7239):766-70. doi: 10.1038/nature07863. Wu, Y., B. Sun, W. Shi, L. Ni, J. Chen, G. Cai, and J. Shi. 2016. "OCT4 is up-regulated by DNA hypomethylation of promoter in recurrent gliomas." Neoplasma 63 (3):378-84. doi: 10.4149/306_150919n492. Xing, Y. N., X. Yang, X. Y. Xu, Y. Zheng, H. M. Xu, Y. Takano, and H. C. Zheng. 2011. "The altered expression of ING5 protein is involved in gastric carcinogenesis and subsequent progression." Hum Pathol 42 (1):25-35. doi: 10.1016/j.humpath.2010.05.024. Yang, H. M., H. J. Do, D. K. Kim, J. K. Park, W. K. Chang, H. M. Chung, S. Y. Choi, and J. H. Kim. 2007. "Transcriptional regulation of human Oct4 by steroidogenic factor-1." J Cell Biochem 101 (5):1198-209. doi: 10.1002/jcb.21244. Yin, A. H., S. Miraglia, E. D. Zanjani, G. Almeida-Porada, M. Ogawa, A. G. Leary, J. Olweus, J. Kearney, and D. W. Buck. 1997. "AC133, a Novel Marker for Human Hematopoietic Stem and Progenitor Cells." Blood 90 (12): 5002-5012. You, L., K. Yan, J. Zou, H. Zhao, N. R. Bertos, M. Park, E. Wang, and X. J. Yang. 2015. "The chromatin regulator Brpf1 regulates embryo development and cell proliferation." J Biol Chem 290 (18):11349-64. doi: 10.1074/jbc.M115.643189. Yu, G. Z., M. H. Zhu, Z. Zhu, C. R. Ni, J. M. Zheng, and F. M. Li. 2004. "Genetic alterations and reduced expression of tumor suppressor p33(ING1b) in human exocrine pancreatic carcinoma." World J Gastroenterol 10 (24):3597-601. Zhang, Z., Y. Zhu, Y. Lai, X. Wu, Z. Feng, Y. Yu, R. C. Bast, Jr., X. Wan, X. Xi, and Y. Feng. 2013. "Follicle-stimulating hormone inhibits apoptosis in ovarian cancer cells by regulating the OCT4 stem cell signaling pathway." Int J Oncol 43 (4):1194-204. doi: 10.3892/ijo.2013.2054. Zhao, N., Y. Guo, M. Zhang, L. Lin, and Z. Zheng. 2010. "Akt-mTOR signaling is involved in Notch-1-mediated glioma cell survival and proliferation." Oncol Rep 23 (5):1443-7. Zhao, Q. W., Y. W. Zhou, W. X. Li, B. Kang, X. Q. Zhang, Y. Yang, J. Cheng, S. Y. Yin, Y. Tong, J. Q. He, H. P. Yao, M. Zheng, and Y. J. Wang. 2015. "Aktmediated phosphorylation of Oct4 is associated with the proliferation of stemlike cancer cells." Oncol Rep 33 (4):1621-9. doi: 10.3892/or.2015.3752. Zheng, H. C., P. Xia, X. Y. Xu, H. Takahashi, and Y. Takano. 2011. "The nuclear to cytoplasmic shift of ING5 protein during colorectal carcinogenesis with their distinct links to pathologic behaviors of carcinomas." Hum Pathol 42 (3):424-33. doi: 10.1016/j.humpath.2009.12.018. Zheng, H., H. Ying, H. Yan, A. C. Kimmelman, D. J. Hiller, A. J. Chen, S. R. Perry, G. Tonon, G. C. Chu, Z. Ding, J. M. Stommel, K. L. Dunn, R. Wiedemeyer, M. J. You, C. Brennan, Y. A. Wang, K. L. Ligon, W. H. Wong, L. Chin, and R. A. DePinho. 2008. "p53 and Pten
138
control neural and glioma stem/progenitor cell renewal and differentiation." Nature 455 (7216):1129-33. doi: 10.1038/nature07443. Zhu, Y., F. Guignard, D. Zhao, L. Liu, D. K. Burns, R. P. Mason, A. Messing, and L. F. Parada. 2005. "Early inactivation of p53 tumor suppressor gene cooperating with NF1 loss induces malignant astrocytoma." Cancer Cell 8 (2):119-30. doi: 10.1016/j.ccr.2005.07.004. Ziller, M. J., R. Edri, Y. Yaffe, J. Donaghey, R. Pop, W. Mallard, R. Issner, C. A. Gifford, A. Goren, J. Xing, H. Gu, D. Cacchiarelli, A. M. Tsankov, C. Epstein, J. L. Rinn, T. S. Mikkelsen, O. Kohlbacher, A. Gnirke, B. E. Bernstein, Y. Elkabetz, and A. Meissner. 2015. "Dissecting neural differentiation regulatory networks through epigenetic footprinting." Nature 518 (7539):355-9. doi: 10.1038/nature13990.